<%BANNER%>

Record for a UF thesis. Title & abstract won't display until thesis is accessible after 2010-08-31.

Permanent Link: http://ufdc.ufl.edu/UFE0022336/00001

Material Information

Title: Record for a UF thesis. Title & abstract won't display until thesis is accessible after 2010-08-31.
Physical Description: Book
Language: english
Creator: Harpley, Crystal
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: Medicine -- Dissertations, Academic -- UF
Genre: Medical Sciences thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Statement of Responsibility: by Crystal Harpley.
Thesis: Thesis (M.S.)--University of Florida, 2008.
Local: Adviser: Gulig, Paul A.
Electronic Access: INACCESSIBLE UNTIL 2010-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022336:00001

Permanent Link: http://ufdc.ufl.edu/UFE0022336/00001

Material Information

Title: Record for a UF thesis. Title & abstract won't display until thesis is accessible after 2010-08-31.
Physical Description: Book
Language: english
Creator: Harpley, Crystal
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: Medicine -- Dissertations, Academic -- UF
Genre: Medical Sciences thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Statement of Responsibility: by Crystal Harpley.
Thesis: Thesis (M.S.)--University of Florida, 2008.
Local: Adviser: Gulig, Paul A.
Electronic Access: INACCESSIBLE UNTIL 2010-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022336:00001


This item has the following downloads:


Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20101221_AAAAAU INGEST_TIME 2010-12-21T14:48:07Z PACKAGE UFE0022336_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 108127 DFID F20101221_AAATLO ORIGIN DEPOSITOR PATH harpley_c_Page_035.jpg GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
169c66510c4643931b8b061737d8a583
SHA-1
d6a92ef76ffb18472c109c3f6d88a9db7b0d579c
88431 F20101221_AAATMC harpley_c_Page_055.jpg
dc9f9bf1ad45cab0ca6f0012cc36588d
55c56c8c3c5da67bf1ea1c826f23cbf57c661d70
111680 F20101221_AAATLP harpley_c_Page_036.jpg
f193a5a809e9b7ea3e8d4eb175a55f2a
221725ef5bad3dd8f2aeefb78fd4c386eae72c6c
110304 F20101221_AAATMD harpley_c_Page_060.jpg
aef3d9206b876d35ccc1f52a2e4e4ace
54695a1cb0a6e363e036054c2c4ad364e46c2b56
115482 F20101221_AAATLQ harpley_c_Page_037.jpg
6f99b6fe744725d4f6d6f8bdba7ee140
a1aa281dd3b615b73755fc52bc77d16065d77701
110428 F20101221_AAATME harpley_c_Page_061.jpg
a9e2bf80afa064f1a116f2e756d1977f
12d8c73dee1d189b1f4f86bbaf33079b24d601b9
114020 F20101221_AAATLR harpley_c_Page_038.jpg
5ab0797715f36eadaf5fb65451c5ecd0
6ede8c52b597289a7c81d158dbebb8b25edcae40
117279 F20101221_AAATMF harpley_c_Page_062.jpg
65060f7cf777e87bd966e1c6cbd4f708
9f0cf7c313cabed91314eeee33a43b1b9aa96a54
115709 F20101221_AAATLS harpley_c_Page_039.jpg
8da5a706a152f7091212ca8395b96e34
faba324451c8233d90d151b5158b7a031ce21825
113006 F20101221_AAATMG harpley_c_Page_063.jpg
2006a50b17ed344129baa5f047105b5c
b10ab3c945046aeb47d33e2ea04b4ea5ba5891e4
116380 F20101221_AAATLT harpley_c_Page_040.jpg
f67d0845c21884d5b014c118e1ba2ff8
d9685eeee0ba2fbc175e9d37c746f52822c96027
105659 F20101221_AAATMH harpley_c_Page_064.jpg
dd18edca7383fd694583f9024cfa303e
7c243c1b9c6d8bf6242fd84e14d6e33aae7106be
112583 F20101221_AAATLU harpley_c_Page_042.jpg
fcf73ee57f0d763ad7a62ea64c44242a
815ff06beb3873c46890bb89581cb96eb01e5065
114360 F20101221_AAATMI harpley_c_Page_066.jpg
936eaa82f650f26bd7c6b0c893689f8d
e88fee2fc39101b2eedfc5222d58d7153144e9e6
116082 F20101221_AAATLV harpley_c_Page_046.jpg
9b30b90a1dd4e04197e4fee0ac1498f9
65192ddb194335c319172c980e48d85fe8e2b668
114429 F20101221_AAATMJ harpley_c_Page_067.jpg
43597e5ba868afcf0480ad936c1d379f
34cdf161baef99a7071040925505c6c043fbc6d3
110843 F20101221_AAATLW harpley_c_Page_047.jpg
e5bf66ce3d2f9d9cb975f76cffe4eb53
c67ced083118fbcd449fe546d57f7615ef765864
114385 F20101221_AAATMK harpley_c_Page_068.jpg
26f1b5d249dc8627da1b0fc48ed48b45
524722109b6f2f33af9e8eeb893f43e54c656aa3
110941 F20101221_AAATLX harpley_c_Page_048.jpg
8ddcf208892abd1b7e4741c8b6460b9d
15479453807fd873f38a4c2e5861e6da865bc959
108865 F20101221_AAATML harpley_c_Page_069.jpg
1a8535952ed46e265622837b134a21a0
fc2bd2219fc27faa32ebe4fed4853425f029cbf4
114045 F20101221_AAATLY harpley_c_Page_050.jpg
aeabbc648d23fcc9db821713ec510e6e
e553652f88a2db319d7692698d5438c46a6f3ecf
115821 F20101221_AAATNA harpley_c_Page_093.jpg
43a37e803cfa288bedb142d158ae534d
41dbbd2c4842a24ffce2970f3829da576a895754
119611 F20101221_AAATMM harpley_c_Page_071.jpg
898d7d0bf2d78a256d24e13ba7a5802f
602f68850de452039ded565796a3b3e4a086e4bf
107566 F20101221_AAATLZ harpley_c_Page_051.jpg
ebb7ed6bd956dd73d33a124918a3375d
2ae1e0ed38ccb99da1d18b68b8679214bb4e9f8e
F20101221_AAATNB harpley_c_Page_094.jpg
62c86b043d2a54bd0da94b61b9e4659c
ae707a37d51bdf7c357a7f350f2d835bd2f3c2d1
108263 F20101221_AAATMN harpley_c_Page_072.jpg
08da713890fab4a4a82cd0d6453d4ae1
d3c7449aee8fe578d7a233aeee5bc17eefe6fd9e
110156 F20101221_AAATNC harpley_c_Page_095.jpg
7db0818f78a91a56dfe479d597fad3c1
365c56f9bc7f9867fe99a2625b0363dc02f4452f
112706 F20101221_AAATMO harpley_c_Page_074.jpg
cd3572c5056eac18f73d57654d753fe6
289f06943a3372886df9cbb833e2614f3db81d63
55583 F20101221_AAATND harpley_c_Page_099.jpg
f32018ebad2669c6cc83ad44ca0e08ce
bb51c03b324dad652c070f73a635ec5a1d5796ba
114667 F20101221_AAATMP harpley_c_Page_075.jpg
fef09ad64bf8b5226ba38e9f6619882c
0791fc72b380ef462ce30d01f24b49b71f1676ed
45222 F20101221_AAATNE harpley_c_Page_100.jpg
c1272de961813e650e610f384ab03bb3
c669ef9ae28d6945578d90975f2909426f51579b
113605 F20101221_AAATMQ harpley_c_Page_079.jpg
885c72aad9640196c046c7c6773ec8ca
735a17ce242665c3939e39083325442584945c53
77213 F20101221_AAATNF harpley_c_Page_103.jpg
c632218771f22c0ee935c568619f5add
b22ffba55b510ee4cabce1c9016295c531add96c
113827 F20101221_AAATMR harpley_c_Page_081.jpg
e459d1e33348c1d0ca5c070a835692b4
7cf6010afd1e17cedd37d961948bdcb0ea3349d5
19828 F20101221_AAATNG harpley_c_Page_104.jpg
d1afc95e483747d091542ecbce20a8c9
79bd10d1384dbdda6f7e608b3fd6131f07fb2c8d
101363 F20101221_AAATMS harpley_c_Page_082.jpg
6eb8c4d7adde30563548aaf84a1cbe61
38e39a2fae12f2e8c9958d01a4f66649ab5332f3
46357 F20101221_AAATNH harpley_c_Page_105.jpg
0d29e9f247f58aeb188e3b77fc30e615
5d3a688b73fdd4a22d9f60a614db687b583217ba
111858 F20101221_AAATMT harpley_c_Page_083.jpg
3fad626d0f808502a47873a92d395485
862fde9e22c4ec18a8c1040a9d5e25b2d88fe0f5
54053 F20101221_AAATNI harpley_c_Page_106.jpg
13d4f0e21ce2f518abae77891493006b
c457a07e01f8272cb0a3abe6de32b242b8006f72
113301 F20101221_AAATMU harpley_c_Page_085.jpg
51e83386bb5a1a6e809b41920ff65779
5b03fe899939dcbfe52dae4c36386d28406fd747
67000 F20101221_AAATNJ harpley_c_Page_107.jpg
ea216bc3ed5706aac1768df59bed51d8
bea6bb02d18b95b3bb5cc7c930a0df79535c5d88
114153 F20101221_AAATMV harpley_c_Page_088.jpg
2a2abd18e18788e569cc9aa9ebe2b5e2
49217026103e49fd33508beb822b719167257da1
72509 F20101221_AAATNK harpley_c_Page_109.jpg
8220f9691778f12cc0ca72a4ce82ac3e
7b24399da8bce1afc97800c2b2281ef385cbeb33
119156 F20101221_AAATMW harpley_c_Page_089.jpg
c7ea942ea5d7bd5c1daf4b95deb9ae37
88ba27c6e4c126e6031ad1fa4e34d60285141fc3
45854 F20101221_AAATNL harpley_c_Page_110.jpg
091aa5bbeccc39d063ec9a34b927d74f
9fcface56ff4c1cec1db10bfcaf840f53407158a
112892 F20101221_AAATMX harpley_c_Page_090.jpg
580ab4f0be82f9b7eed4b94d6ed5844d
d9021186c5625566721d0d2e6a0da242652a6b23
60372 F20101221_AAATNM harpley_c_Page_111.jpg
af14bc7e3a3791595d9e3d0f87fbfbd6
89ea0c654bec397d11b51c87067f36c5f0d07fa0
105152 F20101221_AAATMY harpley_c_Page_091.jpg
1919eea93c54007f9756451e469753a3
dc24c40bb4868362717ca0e408662b4e40f333b0
113181 F20101221_AAATOA harpley_c_Page_136.jpg
41d0c2510041341c7dfbc2f33edeab9a
898db4e8c4077bdad346aef36ff142cf757651b6
82185 F20101221_AAATNN harpley_c_Page_114.jpg
1838fdeaa76eb0f49b24f65de504ad63
4a37e5643785d23bfed0b1ce1a77988c931d7b56
114098 F20101221_AAATMZ harpley_c_Page_092.jpg
85c293b3c0d91caadb7394f067a470f2
e6d8c5c62def4ad164ef5706641775175087bc21
111412 F20101221_AAATOB harpley_c_Page_137.jpg
180d82e590e040f38cc09fb7585416e4
5598875c8cd95aecafd842cf1d19e6a3fa3e5893
41653 F20101221_AAATNO harpley_c_Page_118.jpg
6e2951bafc05a4be6c96d9f6287d4f46
93bb8b9a854ef8bf9a2c8f919556f145454b973c
109954 F20101221_AAATOC harpley_c_Page_139.jpg
aff1fb7ff5e99129247a83cdb1e761aa
28328dfade9c3774317c0171a91921718b15d545
109449 F20101221_AAATNP harpley_c_Page_119.jpg
c63b5003258de42b70deb5437c601b3e
ee9ac601ff660f1a8861350c438cdd6559f9b5d0
116650 F20101221_AAATOD harpley_c_Page_142.jpg
6afbb27798e0abe9dfd809c294c2a6cd
f7cb83da3928e6a2ecd6935594ffa59da837288d
108165 F20101221_AAATNQ harpley_c_Page_120.jpg
2569a50ed24c40a3d52d61cb42956c3c
f6c9fab5f9ad212025418ded4a7d742ef6a48f72
113980 F20101221_AAATOE harpley_c_Page_147.jpg
cb40d6355104aff59dfa28a257d7ab38
170c2df561f774340035a3cdb058885fe4c81fad
116323 F20101221_AAATNR harpley_c_Page_121.jpg
bf2517cdb84f2ea5a164ec3eaae7c5d4
5b38edc88bd57bfdcacf03736e61964424e28a91
111309 F20101221_AAATOF harpley_c_Page_148.jpg
75d1e8b74c2c81e4c71a471facf7ed31
df9011907948c95a33e8da72e15179f61cf71537
114059 F20101221_AAATNS harpley_c_Page_122.jpg
fb4c01fcca4d065ce5ab3dc0edebc45d
97bf8b62f3b1805be05224cd9e8f7b3d1a079fcf
115077 F20101221_AAATOG harpley_c_Page_151.jpg
4c1c2fac7cb53c105ae38546d3382268
9b25b652e03c38aa3f001b18ffb3e6565467437b
111446 F20101221_AAATNT harpley_c_Page_125.jpg
669e623d9a61d0d5676fb96138599caf
37dc2c7ff396e544a35ffe90a5179c83b7788927
69723 F20101221_AAATOH harpley_c_Page_152.jpg
5bbaa73edd019e9ca3419cdd328fc949
b9463be4d5f8eb8a0e161e71f0b3ee788579b6ea
116425 F20101221_AAATNU harpley_c_Page_126.jpg
dd29397ec9204a35c9731ab635852e72
0363fc7ec2b13c917a5871c3a3f833bc6144f66e
35931 F20101221_AAATOI harpley_c_Page_153.jpg
91e8dce3f70d5452ff54d1436b51bbb4
e6fb987ed772c2a62a84463be49c3f66ea03bf57
107738 F20101221_AAATNV harpley_c_Page_128.jpg
eeffd1e0a6ff9e399bc900eb74720224
41f8d7dcc5cb4bcb25fcd046c3956380e475d358
129730 F20101221_AAATOJ harpley_c_Page_156.jpg
2e0a4d0e39f1fe28452d062fcfe406bf
5e5d834ef55b5d236172a8d6e9664f10dd0d4c1f
111373 F20101221_AAATNW harpley_c_Page_129.jpg
720d929e58cb7bc135445967da7e1d52
c2b7a2ca382702a63ff61884b1bd3bf28eaaa222
134714 F20101221_AAATOK harpley_c_Page_158.jpg
1ad0e5b2ff6b7afccc6ae40511ca7115
9c15d3eaaac011ae50e7bbb91a0f00ef5798777b
112626 F20101221_AAATNX harpley_c_Page_130.jpg
47584433348547364c9fb739b033954f
fbd66c14147a4027543a0bf789b32beda755948b
129943 F20101221_AAATOL harpley_c_Page_159.jpg
afcdca3530e6a6d4777c360314b7e0d1
08d22c715669f43ba1d14f0a30460204f26774bb
115302 F20101221_AAATNY harpley_c_Page_134.jpg
2ad45fba48b7931c73ca9d04aaeecff3
266292c3754aeb745fc84f6c45514ad9fe2747e6
1051982 F20101221_AAATPA harpley_c_Page_016.jp2
07e687311a5e256a786865196d3b41e1
e571414686adf0d50a1326e7d3ea63741e46ae9d
141603 F20101221_AAATOM harpley_c_Page_160.jpg
a50f2c88e08d471f306346947b125231
52af4bf6ec1dfb7f98fcd4b6d2a536f4a589eb4c
113956 F20101221_AAATNZ harpley_c_Page_135.jpg
71c318fa5e71bf3e9a6472ba5d1860af
c06b079deb6f58bcd623371ae0b4232ad40b56d0
1051947 F20101221_AAATPB harpley_c_Page_017.jp2
a89899f0c2e5f02153a01a3694d02fde
4e3e80ba31adccbff7c2d6725ae05528e3190413
131955 F20101221_AAATON harpley_c_Page_161.jpg
01a6f2dc07f35af3dbce0a13c34ef21e
94f2d4f830d86a712e844849b1e20ab1655b580e
1051936 F20101221_AAATPC harpley_c_Page_020.jp2
5119a0c82019e014134a9649f8200c6f
8379eeb5aa2d0dbe76a903bd1d7e9c38776adf49
67982 F20101221_AAATOO harpley_c_Page_162.jpg
c81a0ffa29d69fd89c24e55c614f35ec
05784c8e4124a198a20a258edfb81b3d2dba0fe1
1051979 F20101221_AAATPD harpley_c_Page_021.jp2
1b679007968f400d50468ff4ad54b213
f8fa076504200aaae16171cb425716faabfc69e6
254689 F20101221_AAATOP harpley_c_Page_001.jp2
6bfc13630ca4982cb4316e172b17d1cb
dfc986510895e8768f247e3f4e7590fe22bd146b
1051986 F20101221_AAATPE harpley_c_Page_022.jp2
c72e9c7385e02337c17824b55ea675c9
b56f811ff2928f35d5e785b58bc47449c6ecc620
26522 F20101221_AAATOQ harpley_c_Page_002.jp2
2a1819534baec122698408269d50a6af
04b3cc222f72905690cbca1c00efe0ab5d4efcbb
1051949 F20101221_AAATPF harpley_c_Page_023.jp2
0b76ca93a423ee459169bf3c64889947
37ac92832793f84a9864bea8049a4b8fc049ec3e
81480 F20101221_AAATOR harpley_c_Page_003.jp2
4354a5ed14aaf93e230eaefac34d50c6
ff328681076fdc5f6c4cc53226d19f2642a20f77
1051985 F20101221_AAATPG harpley_c_Page_024.jp2
f3fc8b48f807a537e35d4840ee6b9588
876c58380d01ece84603c0db26fa2dcfec0e66e1
F20101221_AAATOS harpley_c_Page_005.jp2
baa0aca7f8d72506580a383f6885f771
c4c73deb571f0d9d9a8fe7ab1a9244e9374dd3d3
1051954 F20101221_AAATPH harpley_c_Page_026.jp2
c4329133e45bd50ae5d4430cd7ac5f76
bc795160f2021d20f2d177ab53c8a5aae4ada82b
F20101221_AAATOT harpley_c_Page_006.jp2
61e7a25fed68c6c4f9630c00d44d6877
52570df8271f6e1f6f4bdcba43462c8b71b4beba
1051946 F20101221_AAATPI harpley_c_Page_027.jp2
bec78e2b8312d936f06a9ffc219bc0a2
9b96396a07f6ccf0b1f342fb73d9dabaa4d73eeb
388136 F20101221_AAATOU harpley_c_Page_007.jp2
45250ff5563272d37a87aba48d075e78
9dd696dcb3c030226331f37b7cdcb8b7e07613ca
1051940 F20101221_AAATPJ harpley_c_Page_028.jp2
26c28e67f727765247e2ab0e966c276b
f6dc4d9ed6175c5ba4feb85364415e04ba0cf6dd
519415 F20101221_AAATOV harpley_c_Page_011.jp2
e2f988588e83305885267be550abf4a2
4ea7da89ced62bf335b5eafbae5e4b5bbf85f50a
1051967 F20101221_AAATPK harpley_c_Page_030.jp2
cb09c0ca6ba6181e4d84d31c5f558984
421898b9840e12c7741288a3b993678a29f75115
256095 F20101221_AAATOW harpley_c_Page_012.jp2
b02b5e4b24c0c1ca0699985914c38a2b
781270da62ad542b332cfea602ebd14eae98e084
471051 F20101221_AAATQA harpley_c_Page_056.jp2
fb7f55b22a0ecb03ed105747fd11d64e
19aa52310db8b8b94b39a894956c4b0d8f13d65f
1051929 F20101221_AAATPL harpley_c_Page_031.jp2
288c342faa9e89bac389a5545b4d82f0
60e5cc5bc4af4ead3191aa0d599f8ca26efc371a
F20101221_AAATOX harpley_c_Page_013.jp2
ca8581f596ba7673f54734c1d5cbc386
63076839d4377e9900b6ae0c3f6801a5c42962bd
535820 F20101221_AAATPM harpley_c_Page_034.jp2
6ca331ec052a28f258af6861de93835f
d31f98b56db5f8b2f02fb4866d9c751e99b8ced0
F20101221_AAATOY harpley_c_Page_014.jp2
c6c1b8fcc1cb72114a408044f74eda2e
1205754ad4660162baedeba5b10c08d7f19a4522
F20101221_AAATQB harpley_c_Page_058.jp2
2ccd1789db9e74dbec93cb852814fa2d
33b283ae23b558973874b475531b5176dfa6a213
F20101221_AAATPN harpley_c_Page_037.jp2
732fda21cf469b65deeb000ee24f3505
b10fbf1c0fce557224f084574ca1adf3a2b585b9
1051960 F20101221_AAATOZ harpley_c_Page_015.jp2
c31c19bbd6ec775ff4cd58f8e1c2ba16
5d3993d9ac35d1a5002a2b2c5fc6abc0e3c80e19
1051952 F20101221_AAATQC harpley_c_Page_059.jp2
dc6030ce754a9f83d9ded0616db2dbb7
1f3bcd55447c8cc81fecd7e2bf486681c1a6c11a
F20101221_AAATPO harpley_c_Page_038.jp2
ca70a762635ecaf0b38515bbd3130a37
1c43285debd82dfc79e7cc6254ba1c9ef10aa4aa
1051951 F20101221_AAATQD harpley_c_Page_060.jp2
1c045ddc9c91c5a8b3c178107ff136d5
07445024804b425c24a17143d0c2b0de0605f543
F20101221_AAATPP harpley_c_Page_039.jp2
ecb5105a9fae83806411dad5c23a3be2
47cdcd52b356ea9c7ea40437bf283047786a9d9b
1051969 F20101221_AAATQE harpley_c_Page_065.jp2
fb70272f5163262294b30d09f40277c5
9cd856ebf2a8f2e812ab53ce8060009c1d73385c
1051970 F20101221_AAATPQ harpley_c_Page_040.jp2
608a5d4cdd0edd8012293b998af97c95
19b91024d739fb743ead5f339451b0cb4b7fda13
1051959 F20101221_AAATQF harpley_c_Page_066.jp2
b159f3bd5d7c28256a41f009e7fa2900
08cc2d078bf7c9ce3fd8c69789b51d70d1524972
1051961 F20101221_AAATPR harpley_c_Page_042.jp2
e04f9aa8333e8eff35503186a3c57832
2a47e76d4e09d63de95634e49fe19ee545f39e1e
1051938 F20101221_AAATQG harpley_c_Page_068.jp2
9970b9e5a8040f6f34f974f0a5f99918
a7a183804a3f290b8eb008a58158c57ba16e0c70
1051983 F20101221_AAATPS harpley_c_Page_043.jp2
130e7c849e91d8a58f1834d3e0accec9
344ae06869f358d3856032d76ee92dfc23e3b6e3
1051981 F20101221_AAATQH harpley_c_Page_069.jp2
259a849bce2a18fed124c614131dc49a
b983052551249e7ae1d6dd68366bda4556135f2f
1019199 F20101221_AAATPT harpley_c_Page_044.jp2
1bf905eef3403f1f1388b2c90d47d793
3b96efb5a12b37d59141e0c234be00685c0b9fea
F20101221_AAATQI harpley_c_Page_070.jp2
3efcc4e18bf75588d85f8e7230ea14b5
3f2dd2a6464e257caeae83d7832f55b428f6d49d
1051944 F20101221_AAATPU harpley_c_Page_046.jp2
fca70cf74abb1cc76ade04284a4911e3
7e42654461da26e9a33e4d877bcd4db7bb29262f
1051906 F20101221_AAATQJ harpley_c_Page_071.jp2
020ffd710b96fb9152b0dab3bbdb14d2
473b2b7f38b955cded35131e9c71fc5b7289db92
F20101221_AAATPV harpley_c_Page_048.jp2
0c96462c325d50c579bf6f424736a0a7
1714f4fb8b805f72faf66b7833398a82770ee64e
F20101221_AAATQK harpley_c_Page_073.jp2
9ca80a69f8b098296fb8f9b215d310f1
cf2b6b73535eb56e97385fcd3ce84370a44d9260
1051971 F20101221_AAATPW harpley_c_Page_051.jp2
22c660aa5d176b789c60dce954d6fd9a
b7399bec06b84c2c99ccb7f90ec0b5b96356b2f1
1051943 F20101221_AAATQL harpley_c_Page_074.jp2
d2dbb46c4f0fad4750170cc7e738a12e
676e7805a0ad043401e96ddee228eb69e2330bd8
1051937 F20101221_AAATPX harpley_c_Page_052.jp2
8d6585b8464c7741ab548bdbc5dd7b7c
7a3c1949c9abd124fcd6dc74b578892352ae330d
1051945 F20101221_AAATRA harpley_c_Page_095.jp2
e5e53d3faf397a2085817af27177f83b
feebd0926a436d689e89c93c040cea268c97beb1
F20101221_AAATQM harpley_c_Page_075.jp2
2716e431477df8033b4e43108cb93f35
f929274f402a6ed9637c4dbb9bd47aff1f228b56
1051978 F20101221_AAATPY harpley_c_Page_053.jp2
b03ef166b66a1767fec1ce1d28205a47
aeabee9771e358befcd2ba7553103d3cdc31e3a6
80854 F20101221_AAATRB harpley_c_Page_096.jp2
cf20661e0583084a6cc10424b4e7a700
fcd446e61c4bf809c15189e038d64d116fea26eb
F20101221_AAATQN harpley_c_Page_076.jp2
bc6f827321e94547455e70d290a75662
e354cd5faf6a716f90492b9c116a28431cef507c
831660 F20101221_AAATPZ harpley_c_Page_054.jp2
1469be6a92c92463dbf0a7d4a768cf71
763397dcdd65cdbb1e8b8534c7d0ab6e05922652
F20101221_AAATQO harpley_c_Page_078.jp2
ace9bb0375352e9cc3bfb38a4b168eb9
c85802077ee18410312c33b6086d01dcda570def
566617 F20101221_AAATRC harpley_c_Page_099.jp2
5fb6393aa946e0c5a4b1e328257e69e8
c25919b6eee26a9f53159218b85ac67128338c4e
1051907 F20101221_AAATQP harpley_c_Page_079.jp2
0b80a1bd7f481716ae2cf790b42f0131
4b62418caf6ccec2ef2dc5538eeca246bd9fb13d
470732 F20101221_AAATRD harpley_c_Page_100.jp2
d1ba77563fd21cec5ab0b365fd23d220
45f0d5273837353039c1b9d75d382efca1127da4
F20101221_AAATQQ harpley_c_Page_080.jp2
e40dc1a20ce4bc7cca4c06bc5a07def2
326ec5fa93ddf6fb1eb52c0463d39b66e831146c
407799 F20101221_AAATRE harpley_c_Page_101.jp2
ba821f097a54cc11fd516d5e3c511c60
cf9f8b0478ab78de600723060eb03cbf7d0809a0
1051980 F20101221_AAATQR harpley_c_Page_081.jp2
68ef698dd08fe3222763a892982f7af0
3c9841394bed6b10003f062011ead978f3dd3815
672425 F20101221_AAATRF harpley_c_Page_102.jp2
339852086fcc2da586175317f308a00e
0fdef5a85bcdc7c0492c7a427f64c78928397c7d
F20101221_AAATQS harpley_c_Page_082.jp2
1c3e31994e4c14fb56f8ab50fd3e9f62
5276b7a2ec786e201d6653b0deba544ac20ee9d9
799388 F20101221_AAATRG harpley_c_Page_103.jp2
90c2980ca8eae09d15ccba8845b84839
a0459ca2118c5195a97af27c56855f967fc9d4f1
1051972 F20101221_AAATQT harpley_c_Page_083.jp2
c2bdee6642d5edfd25a127c75c1ce598
e033519c86489b0101aab90d2468a91fef633c53
162366 F20101221_AAATRH harpley_c_Page_104.jp2
ae02a35910b8bcb151a5c1458e2f14d8
0f4e4097ea1654ad4c9b50c1bb9b986d2eacc2de
F20101221_AAATQU harpley_c_Page_084.jp2
cdfe9c696dea30d6f5e216c514887d93
a05a0fb96262a06f67835d10f2f8aba1e6da401e
483516 F20101221_AAATRI harpley_c_Page_105.jp2
f945132ea4e09b558ee798a39f11a060
2f30290d0c519fc86d9a10688d7f93308018cda4
F20101221_AAATQV harpley_c_Page_085.jp2
b1120c3ed7f1217d57e67b89dcec4de7
b53a5237af7e789c4c4c7b6ecf61c164c34cafa9
628699 F20101221_AAATRJ harpley_c_Page_107.jp2
937c032332c8ab202700f1f7c9cb40d4
8384f8eb46abe1f26e07c5d0929df44c687487d8
1051963 F20101221_AAATQW harpley_c_Page_090.jp2
4cf22ea464525a65e23790cc741aeeb8
f433dc868be5682410e8f1e076e84fe222f16706
466291 F20101221_AAATRK harpley_c_Page_108.jp2
4bb709d5f48d0ddbd95bb27f5f61be58
481a9ea0f7bd366e862caa85effe9833eb23e3b9
1051950 F20101221_AAATQX harpley_c_Page_091.jp2
9052c801cbe2f29ddb916160ce82b70a
49844229dc5a4baeef7162e89312861ef2d12ee8
1051965 F20101221_AAATSA harpley_c_Page_139.jp2
6b94237e1800be88431d0606b48ec100
e555bb7fbffdf00c6cb46970922aee6d3976ec41
729822 F20101221_AAATRL harpley_c_Page_109.jp2
d67786d4d1ed25056014b365f7f88f66
6490b693f49728a0b65eed52005f2e9ea93d13dc
F20101221_AAATSB harpley_c_Page_140.jp2
69e9c5a8276ed9babaa2aa1b346a8d03
4d503848e52a787c67041ed293a9d2cdc9ed4d01
611668 F20101221_AAATRM harpley_c_Page_111.jp2
c25bce4409c5958eeee67fc5d59024a0
3f510bbb0a547df41b1c57d4f86ad05aae4c2ad6
F20101221_AAATQY harpley_c_Page_092.jp2
4307baaf9bf083a6bcd7aee4d394a174
6a67347f9e0b4288ead616c8b9944ca3042dd435
1051976 F20101221_AAATSC harpley_c_Page_141.jp2
16f70cdfae7d50a2d1d7581be9770235
8c958c74b8b1d4cb68347459a7cd3e38ca7b77f6
686113 F20101221_AAATRN harpley_c_Page_116.jp2
5e1811d7e4c156e999a6efa836119982
c01c43078cacc1ce6427d0a8bfe4fb0f5b630e3f
F20101221_AAATQZ harpley_c_Page_094.jp2
1f457ed7dc1e8d1fb2f6df6cf6fc166b
50cdb2d6476aca3cdeefeefc9c9e3752b551e2b5
467271 F20101221_AAATRO harpley_c_Page_118.jp2
b9d0f355da4606b8086b7151e805d5e2
34496a9ef713bb588e1662906cf97c999dda29d9
F20101221_AAATSD harpley_c_Page_145.jp2
4dab637774cbed5fe59bf57fd6f61372
7305df304041b65fa616f3f6081dd7a7243833e1
F20101221_AAATRP harpley_c_Page_120.jp2
89971e4ee9f7c04bc7c7bac68f962d7c
6cc7b482952bf3b861320ca157560b4fdd687b4e
F20101221_AAATSE harpley_c_Page_147.jp2
6c70fc7f4230cc1eb92cb1ae1c42ec4b
55244ef143f533ee2f426186413eddd2b62ed010
1051911 F20101221_AAATRQ harpley_c_Page_121.jp2
e473eb7e6b271090002156af1b448ebf
ddf3075ec7458576cf749968ec679c20145b299b
F20101221_AAATSF harpley_c_Page_148.jp2
566d99704bce4186bdd100b6b3e5ffda
0d3bf01f56ca5a0de3b71f2d4d7315d62884209b
F20101221_AAATRR harpley_c_Page_123.jp2
accc33c5bf040def1a304b6777c741d2
ffd690fb414a5ad9f59edbde5f1dd86344ff8424
F20101221_AAATSG harpley_c_Page_149.jp2
38e29eb76f08dcaa1290c49c90fad49a
20b4d3fbbc5005bda345a1a995e671a6784d892b
F20101221_AAATRS harpley_c_Page_128.jp2
046090fd9c77b10dc200db0a2b9c4e92
a0d7246b0f6767c5647a6e3f3d8755bcfb44e79c
F20101221_AAATSH harpley_c_Page_150.jp2
4cfd1945e44148658a9a0182972c9c54
9e9ce75cbadaab6f9c5528a11e53234d7972a476
1051958 F20101221_AAATRT harpley_c_Page_130.jp2
e283eeac70886e55813eb8e87ad4175c
1f341b899d73f6fa75645d79e4f307a909e3a8a9
F20101221_AAATSI harpley_c_Page_151.jp2
eaa64212970fcc595d5b97b65846f2e9
c54284672b55a398b1680b50e39378adcb0b3d2d
F20101221_AAATRU harpley_c_Page_131.jp2
4ec83d6386f2d18312464bd3d0890655
5bd8c34a1998ab7f19d3a58ae2e64680b5bcbb04
372070 F20101221_AAATSJ harpley_c_Page_153.jp2
cfb6856a74763d69f1660e4a1fbdfbe9
78bf28363bb3c3e0116bac99ff640727316a2b5d
F20101221_AAATRV harpley_c_Page_133.jp2
6090384d2e6872617108974d5e4c3ae1
86af0261276a002bcbd6916a4103554b0889da7d
1051968 F20101221_AAATSK harpley_c_Page_154.jp2
e48972ec712dc6f92482125e68fdb705
9f99820411aa375ad591b1e32fddca1b351aaf15
1051977 F20101221_AAATRW harpley_c_Page_134.jp2
a4b8cf6c5d3809e1191dfeb2bb31640e
06d96453d4308509d9f4fb870124a2fb92232696
F20101221_AAATSL harpley_c_Page_158.jp2
8752377aa807d7e38d5f5b5a69b4d696
19454eef470ee11387255ef3ad63d598896db083
F20101221_AAATRX harpley_c_Page_136.jp2
a5f0c2310eb71ac6e086e22d6460550f
468d27cd440a4016d1e2fc85574859d1ba3e7e16
25271604 F20101221_AAATTA harpley_c_Page_023.tif
42a4eb0b484955115779d356392c9ca2
a140f751b5f481fdcffcb49e4a9eecc34b429831
731463 F20101221_AAATSM harpley_c_Page_162.jp2
d31de4ff2fabdc53ac4fa9ab09922066
61d59311063252b9f3d5bc674efb25aff7af70b0
F20101221_AAATRY harpley_c_Page_137.jp2
0887d5524e4793b34200991ded19d0b0
388653ca835d6299f08ff36febd7b7b28e0fcba1
F20101221_AAATTB harpley_c_Page_024.tif
dff4c7de74ed996c8883bd86f2ec76a7
6e6a002d5444dcdfce00a5a7e3a52c78f4db6f99
256534 F20101221_AAATSN harpley_c_Page_163.jp2
1459283ba1c514803d399075c408b922
4551fa24051c9b69fec9c907c165b16656e6827d
F20101221_AAATRZ harpley_c_Page_138.jp2
01fa5c4cdaadc7e31a7ebbfcd636b29c
d5b506bc48243b645ea5f913aa1588af4ed697d2
F20101221_AAATTC harpley_c_Page_026.tif
3b6f9eddb10d7bc352af0550d0c72a22
2f8add7c831d74fa4a98038546e73deb0cd97db3
F20101221_AAATSO harpley_c_Page_001.tif
4e2324a1a79953c99142e0a044faf1c6
ca06f14445d91f0fa0197aa47272a64e81f45d07
F20101221_AAATTD harpley_c_Page_027.tif
e4b99a8ee82eadfb8df5802be4839cca
c4057ca7b609b07763ad43e97ea3fecc357b4e91
F20101221_AAATSP harpley_c_Page_002.tif
ba4c786352f6b24e59d9f36cd44b646e
fbd453869af151904807a6c9e4d14fbd49d4afa7
F20101221_AAATSQ harpley_c_Page_003.tif
a3f8cd7c5a0dc19c98c2ca7ae6e820c1
8e0103a15502f9248571ed68ba765e6ec4dcbf1c
F20101221_AAATTE harpley_c_Page_028.tif
8f99d07d822d264ce144939e87201cef
d6a58f2be5ed9a466fe7936dc979c674ca8822ab
F20101221_AAATSR harpley_c_Page_005.tif
bbb9be3a23fbade2a0b35150355a1a34
9832ead354ed6ba4a2e1de538ea0ad8ad8dbd72b
F20101221_AAATTF harpley_c_Page_033.tif
f57156c8a11cf46bc2282ccf8b3d369e
4bd9f7858657de36fa23dc27acf75bc7d46f9b1d
F20101221_AAATSS harpley_c_Page_008.tif
0301d3bc0f0b0249111a926d2cdb4edb
cc287a31453d0f84fba1a2fa4285a1c072232397
F20101221_AAATTG harpley_c_Page_037.tif
72ff15c3ceb9a5179daba0c735688f5b
46a0b178ea3c46ca8165c4c688cbb0dff5c928af
F20101221_AAATST harpley_c_Page_010.tif
9a2595d473a5548bcbd87f7db6c5ab13
2ead14a422a4fd678777519a4784daf8bc9ddf2c
F20101221_AAATTH harpley_c_Page_038.tif
e43e06c1947f7e65070f4757131d06a1
ef2c1b7be2e69318b12fd21f6c1d8c45cf40fabe
F20101221_AAATSU harpley_c_Page_011.tif
74a2287b1976ff9d622f273860786af8
f9e5ccea90ef9d8bfe2b791500e8dd7717511192
F20101221_AAATTI harpley_c_Page_040.tif
ddd895f0c0b7a4fb31724016176aedec
15c8c2acd95387c541b0c8b2e48c301f51b69a18
F20101221_AAATSV harpley_c_Page_012.tif
d44edd18fad4aa349d08d150e9491f85
a31f5b64ebb63efd3f7c9961c0e6a04c78169115
F20101221_AAATTJ harpley_c_Page_042.tif
ef8260f86ddf47570aec1882446de068
aedf49520159a36ec4db33ffc4c03b431b46b34a
F20101221_AAATSW harpley_c_Page_014.tif
8eb4fe85c6ed30010b1b1ac8bb69cad4
9be304169314868168b5b1001483593dd854fdd4
F20101221_AAATTK harpley_c_Page_043.tif
3e5b68be37073412faa41434b756bc17
f14ce29a7800677aa2f7449f5510843dfa45884b
F20101221_AAATSX harpley_c_Page_015.tif
083a34bae57a2e7d4c3d7d9023b04220
17f32f8d519870f938a3bcab4b75d4229c21af9e
F20101221_AAATUA harpley_c_Page_071.tif
dfeb2a348c5873432df39aeea0d1fc19
77d5626d3e0dd570a8ac4da175310340a8617c81
F20101221_AAATTL harpley_c_Page_045.tif
34e5cbd00ed54dd3f3a7cfb09a26cb2d
da9e4834dd1fc1db78c539c816cfd58ac41c3018
F20101221_AAATSY harpley_c_Page_020.tif
51d4729d5e6976e4cc3242e4f24f9713
40a19869d6d3076e33bfb264128dc93bb76d1c19
F20101221_AAATUB harpley_c_Page_072.tif
6d945d3dd2a4bb08c8485f772580e6bf
7e5b6b1c7553181c8a4737aa6ac69feb170f42aa
F20101221_AAATTM harpley_c_Page_046.tif
da52512379fc38edde073ad6efa67c50
b6371d3461044c44ad750bd15f810195b3441d9b
F20101221_AAATSZ harpley_c_Page_022.tif
929d733ebb7999fa797a9048cf2ca32d
7abbb28c7f51563b2ec82e5d51172a26aa5c0771
F20101221_AAATUC harpley_c_Page_077.tif
56a2ca6f41724f8ce852406c5f6c30b8
aac7e124bbcdd7fe712092f6b44a96d1aeb0db0b
F20101221_AAATTN harpley_c_Page_047.tif
1010d4d4d826467f261ae55b1d677ca7
0fc34b1a626c632986cf33e29f3a8e2ef3764190
F20101221_AAATUD harpley_c_Page_079.tif
e086a4f90b1d1af0f94d02142173df68
53aa241304d32afe7faae390bff878e783db7068
F20101221_AAATTO harpley_c_Page_048.tif
131855daa0c1430dc83455f58045d3b8
83be546a5d94b535b8ea6186b3bfb4a40eb792a8
F20101221_AAATUE harpley_c_Page_080.tif
d2ffabf4d8658a4136f3360c82cd34f8
46f6c78671b4b822787de37c6352e9b221947924
F20101221_AAATTP harpley_c_Page_051.tif
e200adb075c110b552f9db72c1f6f24b
930c16046aedb56a9532cf4f71b054303376d3dc
F20101221_AAATTQ harpley_c_Page_052.tif
c9fdb695b9c6cc9c713c2b4e7b4c0cba
220665f59cab5aa3c4bd2dcda557aaca7d0f2dfd
67868 F20101221_AAAUAA harpley_c_Page_158.pro
76c41642db82ed3b3924ba1a37b0616b
92d9660a69cd8a584cf9c33f05ee083db416f854
F20101221_AAATUF harpley_c_Page_081.tif
f12952d5d9fb13b0dadfb21f0f6599ba
deb7cfd5ebdc5f59558132be33e3dbd3bd9b17da
F20101221_AAATTR harpley_c_Page_055.tif
266828a94c7ca4351aa824474b28b85d
a5c614c572b3908f43e4d8ec2be97fd8706160de
71586 F20101221_AAAUAB harpley_c_Page_160.pro
a573f530f634518c3e9062cc85bed82e
ea9be92d847e359286c742f4ae633b5ab86b8fe4
F20101221_AAATUG harpley_c_Page_082.tif
3064d443590ef4034f5fcd30d847250e
57a0b2486362e5f06bdd9b417c5a2638a1ed0737
F20101221_AAATTS harpley_c_Page_059.tif
76523f312b4f81973945a7b44f6282f9
f5eab2edfe494c5b668cabbb4c2a6f73342a50a1
66958 F20101221_AAAUAC harpley_c_Page_161.pro
43eb6edbff3bcf4b5cb47181ef579de2
50ea165caafbe955bbe7f6f70b7d4aedd178485c
F20101221_AAATUH harpley_c_Page_083.tif
9cce1c55ec16ca0e0efebba8a5c0946b
12d5d5530af1c1aaedc051911cd451c345d27354
F20101221_AAATTT harpley_c_Page_061.tif
171b71522910d432e70b98a8ad2102b6
f8d0b8605db3f11b5b6b66934f50df65d70bf5bd
10708 F20101221_AAAUAD harpley_c_Page_163.pro
0c715479f496b76a6d278aade574e26b
709080cc25f07f1e5abba9545e3bb1a38e47e4dd
F20101221_AAATUI harpley_c_Page_084.tif
9a4cd8e69301909ecad5c336400bd2c1
a2d3b237b0e7edeb227164b8c50537fffcd6888b
F20101221_AAATTU harpley_c_Page_063.tif
99ebe5978df5f80b861bd2397aab2347
6d47c9421da51d02e2a83e0d1f0f128189e19597
488 F20101221_AAAUAE harpley_c_Page_001.txt
8542de684a09c2754b6edd0e6d4d133c
bbce3029cf86202cd0931ad4d4cb15ee640f3cbd
F20101221_AAATUJ harpley_c_Page_088.tif
7f1774abeef7ee900c076d95fbe8a650
2d2e5c483c56a9794297cc5bc86c866c3cf8be64
F20101221_AAATTV harpley_c_Page_065.tif
ebed25acef762ff61c8bd00e9e748787
5f818046619f78a3feafbb0e47526ea05a1b4b3c
92 F20101221_AAAUAF harpley_c_Page_002.txt
955bf9fa20a39691eea77c7a487df930
c9aa2d0e50c469378f239185fe39c8f9b1056864
F20101221_AAATUK harpley_c_Page_090.tif
c82f3815a27e67a97bb388d73b8c57b3
3e6c74b86bc3754f006c50799eaaa1bbed7aa909
F20101221_AAATTW harpley_c_Page_066.tif
1d6275838ad79c7eb7f56c5f79b9a317
db4d41487eeb234e9f8d742573588f6517d6269e
439 F20101221_AAAUAG harpley_c_Page_004.txt
7c99bcb20dcf60ba2619d52f5cac5c65
264de1daaf5d8fdf691070e4ce932db341a3e90a
F20101221_AAATVA harpley_c_Page_111.tif
f8542b920db6b4e6f871ba6a4ff6a508
7b327a73d0f07fc177c199cc1c9e93c41d176b85
F20101221_AAATUL harpley_c_Page_091.tif
689a32e54d441f79b43b508fae41a530
e6c7a52b8d7b64e9934e015561888004b972118b
F20101221_AAATTX harpley_c_Page_068.tif
b07b5feeae451fb2468bd299e2578e44
8cda2c797db022ad7ba5cf9c797f8d5dc32c7420
3451 F20101221_AAAUAH harpley_c_Page_005.txt
77f164570e4284f1cfd730364041993b
825ccdd6809b760c1a4d7af7e871de13c809b563
F20101221_AAATVB harpley_c_Page_112.tif
749f5351cb8433014c020d064e2501b4
01dfa192790fbf570b9ccf2b8b329732674f75d2
F20101221_AAATUM harpley_c_Page_092.tif
52516fa29bb647ae2ed079ebffcd20e2
a96b634aa3b7e2147c0033217ef910fda1797b06
F20101221_AAATTY harpley_c_Page_069.tif
7be1a2ca491ffc206a69c52ea7525477
e8441a7382b8eee8e07437e89d5bb3789228c3eb
974 F20101221_AAAUAI harpley_c_Page_008.txt
459f81fc03ad0dad0b515ddbd1e4a404
922373e7011a9a0a4ab6ed3ffe1154caed2a5cb0
F20101221_AAATVC harpley_c_Page_113.tif
4d2c138132ec55775f2cd886e47e8d43
4391df572c74abe796136c16940960b83627049e
F20101221_AAATUN harpley_c_Page_093.tif
cb114bbaaf6ca7ef1438db498573f05e
f62ac77f30f584fc5054c3f416b0754763eedcd8
F20101221_AAATTZ harpley_c_Page_070.tif
e4fd3b58d320eeae0e037504182a708a
b48613c38799347ac8356b1c1275a768e3f0f8f4
2621 F20101221_AAAUAJ harpley_c_Page_009.txt
c0d2755497965c13618a1644f712a5d3
b351ef51991868f3bec12e651a5da494f93d21f4
F20101221_AAATVD harpley_c_Page_116.tif
fac87256361e2ea6cb5cce58c5c65410
750ab3961b365c80cb33d065ca1e58c6c469250d
F20101221_AAATUO harpley_c_Page_094.tif
6bb057702f4decef32c0fd4f4b584d72
f7a646c0a808b86b50c87786ef931eac73e334f9
318 F20101221_AAAUAK harpley_c_Page_010.txt
feefb0ebb0589831244bf212f0e7de1d
5ad667cc4e9a581704e91d88ab805973b398b48f
F20101221_AAATVE harpley_c_Page_117.tif
0dc48d0d9f0db6da095c463021f65819
793555af35c11a93acf819519201a951f83a8786
F20101221_AAATUP harpley_c_Page_095.tif
9db06729c51ba29b875be407526aceed
5ec9edf3151fb4e2e49e341cf76aa4c8862baa03
486 F20101221_AAAUAL harpley_c_Page_012.txt
fa0119e482ecf9e4f1fb5bbe588deee0
a6e0531c37a10b2230579618793849e6cfc68c68
F20101221_AAATVF harpley_c_Page_118.tif
f64b3f5e038d0f72f3109ac6c7f8bd9f
0e5e3787344b87d091dd1b4ea5622565f28e077a
F20101221_AAATUQ harpley_c_Page_096.tif
b98d0c8b92ab88ad11b1058b06c75bdb
8d1f105a8bc20c8c9cd70084ef142aeb8a7323d6
1781 F20101221_AAAUBA harpley_c_Page_044.txt
49018c29643ddc7a335f0a5e2df369da
bc1681975bb25d49d33a98caa8f727f697c7d619
1995 F20101221_AAAUAM harpley_c_Page_014.txt
c411263a1bdfd662cd83f4bd6f7ad3e3
c5eaf8a6dd05f07304e2833828ec7b2113dc7d6b
F20101221_AAATUR harpley_c_Page_097.tif
ff71f4b573402d398883c51aad8de336
eddb3b5fe6ea8d2c0210f0165d61db175e938e5f
2197 F20101221_AAAUBB harpley_c_Page_045.txt
56a2707b20520d399f62aa1b00c844aa
69720385bbfe3fb8570e191828b658fe14c42d9a
2103 F20101221_AAAUAN harpley_c_Page_015.txt
6e758e25b36d377134ff1cfcefc35589
ac2da7e3004597baea3058b6ba62af454053ef52
F20101221_AAATVG harpley_c_Page_119.tif
a8cea921773a3caa595279fa6f62e7bf
60e1d95cfdd134c5638680b782242f4c05054940
F20101221_AAATUS harpley_c_Page_098.tif
73213bb779bb622338a881884078b61b
0987af520bcb8b941df183b4a39ba59d454bbc63
2090 F20101221_AAAUBC harpley_c_Page_048.txt
106b2eff8f137686879d2c4cabe3959d
ff2758da10269f54e4c5f45930f27ff83e251a35
2070 F20101221_AAAUAO harpley_c_Page_016.txt
2b9a5c86f4b5e7716455c287a5838187
9557461fa51f74472fda341b48c444f43debac19
F20101221_AAATVH harpley_c_Page_121.tif
2424657effb434b8a7bb5101ca9b5907
5e396bdcd3c82db8a7988ac9f3527ac2a6b96646
F20101221_AAATUT harpley_c_Page_102.tif
99db4a4acbdc4b60e7ed33da5ba7d358
9db197a7446e96a82c2128a67119984384787f2a
2099 F20101221_AAAUBD harpley_c_Page_050.txt
9b1b8125ff9135f548c3f3d321cae06a
75c6e0fc783350b2030ff44f18d032fb4ded3713
2140 F20101221_AAAUAP harpley_c_Page_019.txt
65cf5030827115c449c862f85563a118
be656d6d9241f538d493c91a601b055a6b464b92
F20101221_AAATVI harpley_c_Page_123.tif
7a2021b3d186f4a264271742294f6860
0630f199076568860d43eed6fe3c90224794dd51
F20101221_AAATUU harpley_c_Page_103.tif
69ef4ee3a636e65926b0936c779b23e0
cf99d47e981f6aa851ba30f483a2e0306d8feee5
1817 F20101221_AAAUBE harpley_c_Page_055.txt
bcae98090918d9cdfdd8bd0677bc2357
084a1b3e042d7a30616950f0a15ba4bff10f7ae6
F20101221_AAATVJ harpley_c_Page_127.tif
6e8d854b52ade50280a6becce38b76bd
a45c276bca2ca84c10eeec245d5d6d23630b32cf
F20101221_AAATUV harpley_c_Page_104.tif
b24525c7b418f531143d5dc5d994a1f1
a9f9e472cab9d1ea3e01cff29e7a7b0c94de7a1c
2119 F20101221_AAAUBF harpley_c_Page_058.txt
59f12ce69d0e00bdf55d1d47f606f720
ed1c4f6ba6bda775813fb61dd0bfd05f08b65d16
2323 F20101221_AAAUAQ harpley_c_Page_024.txt
0e9a1a7183ee2bd0ab6d5e22084a9d4e
1193285b33fc1902dec9dbedda35887f29445be4
F20101221_AAATVK harpley_c_Page_128.tif
0a614361ef23905409c67beae25f9856
bb5206f745201b95a2f41e78f4e96c8f51050b35
F20101221_AAATUW harpley_c_Page_105.tif
e09a5f4892b2adb52b87c8224ce5c2e1
cfb65ffc4f142ea4e58681f84f3d4143dae839c0
2117 F20101221_AAAUBG harpley_c_Page_059.txt
4d3a41fc9b5d9afe0b565d1e8b4a7c28
cf2f690c27e2be71c0201bef0361fd4f62e8ae0f
2223 F20101221_AAAUAR harpley_c_Page_025.txt
2f423d63c20208286b49cd67f829bdb7
96f286db48971355e83d604c663e3a59187c489d
F20101221_AAATVL harpley_c_Page_130.tif
4bd09170e6ef8446f4e1b8fa349cbaf1
4e538c1f728ad4ecf7b15b0ade16c67ee994cace
F20101221_AAATUX harpley_c_Page_108.tif
4993408dda42964fd966bcfe9796efcb
1040882824efd6b2f6572978198fa5184eb78d0a
F20101221_AAATWA harpley_c_Page_158.tif
3e79bf3f16e638587a08b7bd8d141682
978e55f502e802a5dc3ae028a7594e387e773fae
2107 F20101221_AAAUBH harpley_c_Page_060.txt
46cad01014cc32914d8497f9f9275d4d
20167265656fbfe6ad12ad8961a82ed61c83a7c3
2150 F20101221_AAAUAS harpley_c_Page_028.txt
4cf077e67533c6cff409c435703f6aba
2331309ab9dc19b166ca1d6d067468c5faadc0d5
F20101221_AAATVM harpley_c_Page_131.tif
728af87d304d2dd383d664e1bcd30447
5526c854f14044f2ddac0628bf2b758ec0f50dac
F20101221_AAATUY harpley_c_Page_109.tif
058f55d39d70ec5e2cb2cbdff660122b
5a77c028fc92ee925929815b738fdefca12554b3
F20101221_AAATWB harpley_c_Page_159.tif
28507eb41f876200e8fe7aceb0c9ee0b
e2979e58df37ee31fd12ad9919caf868f7d6680f
F20101221_AAAUBI harpley_c_Page_061.txt
a34cc8a628d7c6e087778d714eb3cb5f
ef803d97bfe74256a8f22d8d5ebd5dd28836208a
2064 F20101221_AAAUAT harpley_c_Page_029.txt
09ae7db456c7a91246c21f232df9db60
2f3717d74daaf58b990fc171d896527d401e0e90
F20101221_AAATVN harpley_c_Page_133.tif
d4acb3fa272b0c8f7e338dbd2b58b201
310781cee8f7b5d73f0b8aaadb8934284eee8186
F20101221_AAATUZ harpley_c_Page_110.tif
9d12e820c3523bee7a6227bef8d37763
e896434afbc27d7f626528716466ad68cd621adf
F20101221_AAATWC harpley_c_Page_160.tif
acb4ab7dbb50b54099d0dfeb721c68e7
c9aeb8b31c4e347847d0259b1656a17db8c6d14b
2207 F20101221_AAAUBJ harpley_c_Page_062.txt
33b8bf9cbedbbd12ffc4160ec7095013
3ab3b665951b745c2efc03a5a4643e16f9f9a10e
1993 F20101221_AAAUAU harpley_c_Page_031.txt
1a4b90fb2555a61cf47f4b205c118afe
34a2035c5a2e15cc5f2e2f12087360f84f288cb2
F20101221_AAATVO harpley_c_Page_134.tif
18b0b58f73b96ddc360af137918d1647
689488e2ccbb5b610209d942b0205b890b00f537
F20101221_AAATWD harpley_c_Page_161.tif
2af3b077db91b6911629bf33e3137c05
027fc01457f8923b27e495e440275703b4aa6926
2151 F20101221_AAAUBK harpley_c_Page_063.txt
a3ca36f61aad64b228bac987889fa67d
6be76a26f7422076ffff1c4115e559632edcacd7
394 F20101221_AAAUAV harpley_c_Page_032.txt
67ea918dae3ab0461908cdf941f10677
7081a7b959259d922fa1573ba5c6286b519af81b
F20101221_AAATVP harpley_c_Page_135.tif
209003d1b02b7121ee22f160b6c08a0f
09091494d733e74e3668340d27e383c7b517a76f
F20101221_AAATWE harpley_c_Page_162.tif
913484a7979906b2905ed578cbcfd199
6ce25ea5ffbec39ee4630d8c5d498f7c3983b30f
2029 F20101221_AAAUBL harpley_c_Page_064.txt
b130ea58b7c27be91ca8cd8f550f745b
ad4db5e810eaa75e4bcd5392aa9bf84526f0a14d
1181 F20101221_AAAUAW harpley_c_Page_034.txt
d7186930c2f5eb54b421a017dbb77712
485b4b20f5ba29d781e75fbb889fbd84c868e367
F20101221_AAATVQ harpley_c_Page_139.tif
09d128f0cdc86512083ad6c7a7d7ac5b
9d3c417256d4c73bf5c8ddff24a4034f63909e48
F20101221_AAATWF harpley_c_Page_163.tif
321f9fc65f9ad281fc08544e22a53c40
239240900dedd48f988a902e4cacd5138f195b08
2157 F20101221_AAAUCA harpley_c_Page_088.txt
25a87562d9db6263f39c89a73826a76d
82e41ace014dea6201b1e6118d2b2a6fae071814
2168 F20101221_AAAUBM harpley_c_Page_066.txt
0331eff7a992072f37641e68ecd88125
5ef41ad0a44f82c263d70448f435ead996a287e4
2093 F20101221_AAAUAX harpley_c_Page_035.txt
7c5a55d1bef3623a5cd40a51eb6ef303
d7f55a69cb97cf748dd1829c2ad25cfc7167645c
F20101221_AAATVR harpley_c_Page_141.tif
8277cf8e4ece28c5de9abd800b8fb952
ab5371bd517d304fdfc50887282bdb4149b4fa52
935 F20101221_AAATWG harpley_c_Page_002.pro
f7283b636a158f14d5678420b6bdd53d
bf6e4d5324a26424aee389045f2617664056fcaa
F20101221_AAAUCB harpley_c_Page_090.txt
fa6058b84ea1734b9c0d53d68a02a1f4
94385c7d622a20e6f00c1dbcd70731a5d552c824
2089 F20101221_AAAUBN harpley_c_Page_069.txt
abf1d8beea4c62a81a83c7e9c91fa23e
d5c65c5f2ee8c9ba3bdd79663dcb8670fc5d644b
F20101221_AAAUAY harpley_c_Page_038.txt
7d71d2fc53a451970f36651eff92d443
08cc9620b28ba12075d43250a30fce346f7f39c7
F20101221_AAATVS harpley_c_Page_142.tif
95be1a95fc36449aaaf7e88d4f790b25
f2c482968d31f10db666f4e442f46c6792847ce9
2180 F20101221_AAAUCC harpley_c_Page_092.txt
29a3e8974307a0d8925024928f7190fb
c5bd382c02b40ff6ad0c221e8cee9edfba6f69b3
2291 F20101221_AAAUBO harpley_c_Page_071.txt
300a71776cd52b4ec003cbbdf941a3ef
115431b232eab45347ab94f02cdfa049852e1ab0
2178 F20101221_AAAUAZ harpley_c_Page_041.txt
f3b18c37b8a5bc93473f4b19c43e9f82
846ffdbdb8020f83fda606aa976da96aa532280f
F20101221_AAATVT harpley_c_Page_143.tif
4886f1c3f0ee68cab0c6efb77d1467f0
2e71c15d322a6106532a040b006ebd9505964ea6
3368 F20101221_AAATWH harpley_c_Page_003.pro
56d1535f0af7a58164ab44288949db86
1dd7dc8b61a0e3d2eb8045b4e61d80de25e8f5d7
2202 F20101221_AAAUCD harpley_c_Page_093.txt
89006b1a305f7bf2f88553e6da50dbf0
68940cd11444a433f4094b038aa6c5f57eeeeb80
2069 F20101221_AAAUBP harpley_c_Page_072.txt
31090f5a050237ce750ea7bd3297dcee
484e49de076a1ff5cd828d9019b9e21c55151392
F20101221_AAATVU harpley_c_Page_146.tif
ce932cd641f2d82b5ff1608af242513e
04dbd0c8eb096f6511e3e9155cf6cbbceb653adb
10069 F20101221_AAATWI harpley_c_Page_004.pro
b8afb0e4f4cd4738b1b91062ea80de54
b74302e1acb28f29b90d193abc17c0b3a4b62a5a
2210 F20101221_AAAUCE harpley_c_Page_094.txt
3f6e6b4c9cdcd036e23fe839ca1f3657
399aa59ff90f92f4328849a29ac91266e823dda1
2245 F20101221_AAAUBQ harpley_c_Page_073.txt
10a03ec18b1e0b1318b526d150957275
0ca8993b421cb7d770d5ba9f2dc9199ce3bc36ed
78019 F20101221_AAATWJ harpley_c_Page_005.pro
1be97309b916df59f4410d07bf16b4b0
bf59b6114aa85f957595bd12f45dbee7cbefecdb
F20101221_AAATVV harpley_c_Page_149.tif
790c8dcedee0859cd56fd607877a00be
afd04339d9db875c48197099261eb714dfabc87f
133 F20101221_AAAUCF harpley_c_Page_096.txt
fa34a3a031b2ae18251321fbb664afa9
dcf0e560484f41bb183bf751e1cfbc9168cf1d20
93980 F20101221_AAATWK harpley_c_Page_006.pro
9e915d788076c21da522616ed193d705
450c2b5e034e57c8b4d38524fc38723037e27623
F20101221_AAATVW harpley_c_Page_150.tif
01585bb1b4b01403b428d809d80dd125
214b14b55f920ac4442b4171a5103b2dc8f74d36
1098 F20101221_AAAUCG harpley_c_Page_099.txt
3b122ef5b2adbde1eca90327832e63bd
fb13136843d354a5b7ac131659bae83dad4230e5
1960 F20101221_AAAUBR harpley_c_Page_076.txt
be42161bcb9d662adbea9c0105777656
803982c4a8adc68d43dd87742568a376f8b62fa4
53835 F20101221_AAATXA harpley_c_Page_027.pro
e866a863f9320f37a474e038d3fb5b22
a9fd9f94dda2b86c677fb9cab921f09692f429ae
10490 F20101221_AAATWL harpley_c_Page_007.pro
8458cdf967ed6a6cbeb65561b2e82891
7156dea788a432ac2cab185219141d3fcc42073e
F20101221_AAATVX harpley_c_Page_151.tif
350a7e28c96c0cf59a8e3d079bdc6da8
25f2f1819b029003a1706282c6cbaaa25fe54b27
521 F20101221_AAAUCH harpley_c_Page_100.txt
d165d3037aa05f3757ea85c110323a95
33afc19ea875460e845f80dc6d30970ae6d4423d
2022 F20101221_AAAUBS harpley_c_Page_078.txt
a227d402c61489f3ae240a930ea309f7
d2c6ec3c48508d34c24ebe29f92fefdf7b174569
54676 F20101221_AAATXB harpley_c_Page_028.pro
cfdf4a80bc929e6791d7541ce13983b1
0ca87bde8385d8260f230c51361cc8ed9d6a7d64
22793 F20101221_AAATWM harpley_c_Page_008.pro
967107f8dfbff0a42aba66e0895149b6
e84a6c473ed9d0b8fa37184c7737547f038ab225
F20101221_AAATVY harpley_c_Page_155.tif
eb52522b33c253bccfd3159555c98e67
206e3960e25c7398d30dda92edee1821f484bf52
604 F20101221_AAAUCI harpley_c_Page_101.txt
9c93b62ce9e73298f513741c472705e6
e4d2a26871080adc8388720f984bcc972a4f7417
2115 F20101221_AAAUBT harpley_c_Page_080.txt
5f8d8eab8702eb699815c6b461c9bfe6
21740462054528788091864fcd03dab93f7a4d29
34229 F20101221_AAASUA harpley_c_Page_016.QC.jpg
948abd5295ab8cbc28c0b369af270f2c
f21bce7f184f163a59140cb5e59f6cc86d6ef4c9
55814 F20101221_AAATXC harpley_c_Page_030.pro
686f76c20a9b9732b9f01738171ee1cf
723de74d4fbe17ba9a5ec88c6cc86118b1f3fb61
61922 F20101221_AAATWN harpley_c_Page_009.pro
cc57f965d206552d2eccfc2282656d48
fa3e85e65f93bb4dd9bd1123bb5fbfa776eb625c
F20101221_AAATVZ harpley_c_Page_157.tif
d189b7afad4e4588fcaa010104ffda0a
0203bb5715ff362edf7f8deeb899916ea59c690d
1437 F20101221_AAAUCJ harpley_c_Page_102.txt
52c209a350ce52abe3179528a5d14484
61ea5070f4f6fb34a795054d0303afd6a2c7bda1
2138 F20101221_AAAUBU harpley_c_Page_081.txt
361a5982d6d872511c60d28195c18f72
fe6edfcb37e04dda906c90f52c8c9bc46b07dc72
1051975 F20101221_AAASUB harpley_c_Page_072.jp2
50f5cd1ce2d830cb7cccb5a7dfa6c452
1f296a577e08d3efe869442fc5c1361095881adb
47362 F20101221_AAATXD harpley_c_Page_031.pro
5e5df45b7ad3168046a9eb45df5ebc84
a2267eec49769495003b15fa62fdedaf2095bcce
20892 F20101221_AAATWO harpley_c_Page_011.pro
5e5bc38b6a4a51d997575cb625a0602f
ac8d843383349330096291e092fe9b25c305d826
462 F20101221_AAAUCK harpley_c_Page_104.txt
fab255bef8420c0276ce54ffceb80090
dd919f9b606d5e392eb445907e8940609582d35a
1899 F20101221_AAAUBV harpley_c_Page_082.txt
ec116d6b565175795f5785823b4fe0a8
83bedcd8ad470f084c1c324c1b281cc8d8014faf
133661 F20101221_AAASUC harpley_c_Page_157.jpg
fce616e52382695d292700e6d61b86f1
2125a6e319bb2a305dc1e6f21b7adaf3e4641297
9168 F20101221_AAATXE harpley_c_Page_032.pro
a775727a41c911b92e7961dbc68dc69e
e4cc865a4ea8fd8f12db7cf4f9ab35288b23051b
10736 F20101221_AAATWP harpley_c_Page_012.pro
aaff2cfd22ef5c14fb97559a44553a36
913c71c723fd6102ff42308c1ce9665466b5d7dc
828 F20101221_AAAUCL harpley_c_Page_105.txt
4f9e10bf9c37b7bdcbed38fe4a7b580c
3428d4fbc3aabaab63045b8d578f276ef442fff4
2092 F20101221_AAAUBW harpley_c_Page_083.txt
ce72e40a8712193e4589d6679073d9b2
c6eff596d53495142d871feab2e7dfb1c7f49453
51597 F20101221_AAASUD harpley_c_Page_091.pro
1ea676d653a54f04db9777dbdaa91f38
7aaea4a0a17a654f855ffc891e8dd551789f80a3
8303 F20101221_AAATXF harpley_c_Page_033.pro
324d64ea1254d410401692689a96f251
b36138186a49f04e48015508855eb0ba1d5dc5ba
50577 F20101221_AAATWQ harpley_c_Page_014.pro
ed4600095a745611661a7227f55e8bc5
a618ca0cfe75664aa8843ca31edce0c75c41eddf
2153 F20101221_AAAUDA harpley_c_Page_127.txt
7e8a20a945608c02a5f6102ff00b3e9c
1f147aaa89e664f92cddb5c46a29d467da8dfc44
772 F20101221_AAAUCM harpley_c_Page_108.txt
812c4e998589a4c0051800f3484bbd0a
2257e69acddd64e66c015d8d2b9e614ba84d88d9
F20101221_AAAUBX harpley_c_Page_085.txt
329689b6522db32bc58d4ffc783d9429
8ae71b45004dab3727612d5433bd3733f7f614ae
F20101221_AAASUE harpley_c_Page_136.tif
fdbd02e8284231cc9a3a7d55b9ca7708
3b63c4bbc398ff671aa2627bcca23777cdd1f0f9
51035 F20101221_AAATXG harpley_c_Page_035.pro
33be2fcc83f4eff60542d28e730c8a7d
8c74635e70e05aefe8a168fba04b48dfe97a9b84
36869 F20101221_AAASTP harpley_c_Page_048.QC.jpg
559f3f04601182a3cf2f5451c56ed308
ad23bd095989b30e82721d9dd4a6d94e67900e8d
50808 F20101221_AAATWR harpley_c_Page_015.pro
9eaf97f22902de6fc81f6e1548bebd30
e6f45c3fc75896e8518c26906010e6bcab55c879
2096 F20101221_AAAUDB harpley_c_Page_128.txt
e5bbb2d83ecd7946471cf7057735d6e5
e43cc3e300b533943268edb6000339ce68e8cc75
1197 F20101221_AAAUCN harpley_c_Page_109.txt
7b0889421a3e00b342c390cb2e2c5158
5f97c6c44364d8c25ad8d8de8590eb7bfcc27bae
F20101221_AAAUBY harpley_c_Page_086.txt
c663b20adab0fd3052e83551200ed0a8
b220089982a4952ca59b7b52ce55259ccf3f7eff
1051898 F20101221_AAASUF harpley_c_Page_049.jp2
9a46fde232ebf2ae618e49b5e1bfd711
eecdfcd791f3d9369cf5b58b3bffacfd6b9a8ff5
52835 F20101221_AAATXH harpley_c_Page_036.pro
4d7714ec8ecdc59754bc3936b7078eb5
3a4fcfb45d054d6731456bde6a1d79422f29d3a5
F20101221_AAASTQ harpley_c_Page_036.txt
2013e810a301e2cfea680978df393b83
c903204a0c35f949e33c748bf1ed89bb34cd2994
52671 F20101221_AAATWS harpley_c_Page_016.pro
51c002b6f445576c47eaba83c5828f9f
eff5f3867ad975d95e04d76c2e0747e7d2fcbd75
2171 F20101221_AAAUDC harpley_c_Page_130.txt
403d9d34f86f13e2e1df44cc92661fa9
b2bb639fc55d8fa9d7420ffc298c9d6971653bce
777 F20101221_AAAUCO harpley_c_Page_110.txt
4eadf7f055ac114a08b7c17bafc3c75c
d4e58bb7307c6b505e3870ca33c17fc2e9d59d31
2084 F20101221_AAAUBZ harpley_c_Page_087.txt
8cc7e624fc931230a449344d8fb9163d
686a811706c29d89ac8b019bc0d8cf3cfa394d03
2193 F20101221_AAASTR harpley_c_Page_147.txt
9c0c679e1b44aa9bafcb87c904da81a3
fad7d14bf08bdc824a2b925d883e20131df3a8e9
54226 F20101221_AAATWT harpley_c_Page_017.pro
ca9672be9c11de0972f814f371e7d99f
c51d5579961fa6070b898b8306506af0660cb388
2102 F20101221_AAATAA harpley_c_Page_077.txt
4052a565a44caa8eeb9c09bcc0499b14
f4b171042037aeec04d23eac162e6d36577fa701
2008 F20101221_AAAUDD harpley_c_Page_131.txt
6f28fca70540a45b91dcc6d4cecf14c7
e26e286f3123e011486b4e6dad32fffc6697441c
1050 F20101221_AAAUCP harpley_c_Page_112.txt
07bcab488eb4d3babc9f649c62febacf
77d7acf25586b798179276b781eda43bb2d81fd7
1051913 F20101221_AAASUG harpley_c_Page_129.jp2
987b2d6e0777dc63ad3dd326a4ab5f65
e3f259868946d5b1213fae69664e482121ac2379
56186 F20101221_AAATXI harpley_c_Page_038.pro
d7995ff79453b214ede41ed9b1f81aa8
8847f2eb1d3aa2dcfe2bf2245cb290a7e1fd0b5e
53917 F20101221_AAASTS harpley_c_Page_022.pro
8667f1784edc8cabb62e88c29eb60b05
f5765170656c09e7db2027f4dbf8b0f0e139f2a4
54697 F20101221_AAATWU harpley_c_Page_018.pro
b9c48d05a5f61aade1ce72bcf6dc9e04
8a40b4056ad30049130e5a0393a7b70cd45fc74e
18942 F20101221_AAATAB harpley_c_Page_102.QC.jpg
a789ae47a4f58b5ba051030060a088f1
f001cd7327626ebb9176b8382f1f42f8d9eb7cb8
2111 F20101221_AAAUDE harpley_c_Page_132.txt
fa051107218f11210559c741c071b41e
f80fcd62f0b827df51cfe16cde0bb971fae8663c
1034 F20101221_AAAUCQ harpley_c_Page_113.txt
381c2f4ef22c59b3ca7f608eb632a719
76da50fd2a239f1cd9bc1979dfc67ed902f8e4e3
2243 F20101221_AAASUH harpley_c_Page_089.txt
634b15b3359aae2a6f500e80653e0735
05ee12de485ed309975777640623803fae2576cb
56485 F20101221_AAATXJ harpley_c_Page_040.pro
edebbd2200e644476f4567a3b12c7707
f449820281a3bd02027866379f689558c4ba4029
3854 F20101221_AAASTT harpley_c_Page_056thm.jpg
06669ce2b648567254c3f69700a849f2
f0bca93a3bfaf7d6ffc1b7f03fb96c320f04c6b3
50870 F20101221_AAATWV harpley_c_Page_020.pro
1a9731a3f12f0298279573d8eda6cfb5
673e82e1b9f01d8627a9e9cf960725bfc4035edf
8906 F20101221_AAATAC harpley_c_Page_078thm.jpg
45c48d171c4804dd33d2955d10a6ff60
64d332f51fe0914f2ad89d57ce8625745992e263
2186 F20101221_AAAUDF harpley_c_Page_134.txt
95c4577a093e7d1992ed071ca075e1c5
316c5ead451e1dc982415e8b8f98eb698c6e04eb
1264 F20101221_AAAUCR harpley_c_Page_114.txt
f04fb3d3a76563ba6c9b212b671d295e
5189494ee3c838f00ab754560b8bab33462f5791
111677 F20101221_AAASUI harpley_c_Page_059.jpg
a9ac8fdaf5e82a227d7bb150fb164e6c
63e4c8c940c6d906e76c7d9b2f92fc42dbc6963a
55723 F20101221_AAATXK harpley_c_Page_041.pro
5a956f82dc1b76a497b61ae6741864e8
a864f36521a28e366478ae4f4e9c2d527f4b072c
34038 F20101221_AAASTU harpley_c_Page_076.QC.jpg
50071a09ddac062592fb318c3c79695c
90414d76b35e4cbcbf79e3553db6cdd884dc12e8
54306 F20101221_AAATWW harpley_c_Page_021.pro
cf01aeb897efe735a64305e1db0af0ff
8e3f33acf6c3782d4dd66db3eaef7cc3e8dd7c0a
21514 F20101221_AAATAD harpley_c_Page_116.pro
ed617082f271e081bae01b322b0c49ca
ac3c7c3794815ff858f4e974a1d1a3136732b995
2173 F20101221_AAAUDG harpley_c_Page_135.txt
6544f96c46884be15b4a4530cb3bcc6f
ea0c88112e17cf633e37da5d6ab4b78a9800e4a4
54891 F20101221_AAATYA harpley_c_Page_068.pro
4b991d6ac3774119d62308fdda956353
6718a6224493811718725c87fc09693bde39a15b
996789 F20101221_AAASUJ harpley_c_Page_097.jp2
afce2684f10c3a2a783557c927e82e72
76cde29a495004fd9aaf0009f913d7722affdaba
55135 F20101221_AAATXL harpley_c_Page_045.pro
b03c98f2f0ecd3f34314bd99b114cc2f
c9041ae79a9ca6dfd5a45ad91e6a0202b7bba925
8990 F20101221_AAASTV harpley_c_Page_067thm.jpg
ca30e4a66504024e46060049e8bc9a83
4528a29a77e3088f13e3eff74f68596609df80ec
52123 F20101221_AAATWX harpley_c_Page_023.pro
8d4e05a781b863bd6bc9875e9d0c33ae
84ad10f75b7a505dd99a0fb4c1e08bffe02dabde
797 F20101221_AAATAE harpley_c_Page_056.txt
6ad05c980c20a2f06c39a0f1bf46e6f9
5655995df6d023fe6833e93dc08943e2f8176d55
2121 F20101221_AAAUDH harpley_c_Page_138.txt
a28f389ae7e8e77a80fa9f0d49989e5d
3ad1983022af2f06757c3e902a7763bc7ab1f7be
633 F20101221_AAAUCS harpley_c_Page_115.txt
39ce08280651ec4c6ae641a73d940e5d
374c3b3a6bdc920c4e43bb836b872164e8a6f98a
53325 F20101221_AAATYB harpley_c_Page_070.pro
5af8aa45aacda025eec3ccf22e3fd3a8
758a8dde93513a35c9179f18f8bcc030e9269d1c
865636 F20101221_AAASUK harpley_c_Page_117.jp2
19ea294a0270d9cb73dce2116c83727e
b1f06f68ed8c73b86ac89702741d99b3bfe6dcab
56032 F20101221_AAATXM harpley_c_Page_046.pro
4623a68e820b70a822de341af2bbd286
481021315450b0e9db52acdf94c13f77c960aeab
8292 F20101221_AAASTW harpley_c_Page_057thm.jpg
6aa2e5ac96bd670c5a20f65d09afa741
cb0908ffb368c797df453c530dcea0258bb956b0
56802 F20101221_AAATWY harpley_c_Page_025.pro
1f77f2bf8685215e2905e0dd77c785b0
bd7430506b1645c92f90b0fd8272a45d34d06b09
53510 F20101221_AAATAF harpley_c_Page_086.pro
4b0df255bd418b72ea494a6a58725d93
42e90243d0ba69554ca234e336c0d9c600d07ad5
F20101221_AAAUDI harpley_c_Page_139.txt
d171c7471a8b204fe72b2d016cab9d03
b8a36928cc0051f43301c28eaae17606e8acc01c
1254 F20101221_AAAUCT harpley_c_Page_116.txt
33dc3d6647b3e229e48c0f656e204dfe
e8f293dee51d9070ac3717b3343fbca7e2b7e54f
58651 F20101221_AAATYC harpley_c_Page_071.pro
ea1376a147396a58c05330d7bfae1046
a6f0841d297a6d7780d205e7365fed8d7fa1bda4
54856 F20101221_AAASUL harpley_c_Page_150.pro
7aae1ca661ee571de47e3b51c80c58c3
b3b83bfdfe16e0a48fb937075199893b6932262e
52977 F20101221_AAATXN harpley_c_Page_047.pro
e7452cd4bf7ce3c71c3b3fa18aa35422
3e9e89375c1f7d3d62dc56a102d8ec63920a55c0
55686 F20101221_AAASTX harpley_c_Page_134.pro
193fa072e9ff7bb5460d17b5526a8fb5
fcc939978e0f26d6472a89551e3f14551ac3f0f0
57744 F20101221_AAATWZ harpley_c_Page_026.pro
84039a63776faa4c9dad5e85d20c7476
bb292aafb8475577e76ec3290441f1e056cced71
2158 F20101221_AAATAG harpley_c_Page_136.txt
a2b947f2b01ce77b832e327b53845a6e
591e44bff994976a23b029e7c11b4ca9914ce6cc
F20101221_AAASVA harpley_c_Page_154.tif
0308153c276f03086736600ef323970b
f67c8bb7778034f58ca5305741117947c8354cb6
2045 F20101221_AAAUDJ harpley_c_Page_140.txt
e7e2c5c1a4d60a4f8353b03e3713f643
4022f891027ba54fd6528b5b274f604d5b7e7781
940 F20101221_AAAUCU harpley_c_Page_117.txt
49ec02248a6d6664e9f4fdc926cbfc67
91f1df5d7605fbd06593c74b0a17847291c36558
52623 F20101221_AAATYD harpley_c_Page_072.pro
5e425009c7a5df9fff8483fbf2753f6c
1896826c9294f4a76ddb976541c350d7e62837ed
F20101221_AAASUM harpley_c_Page_101.tif
15230ed6b85939bcdc2e8b8a92db90b8
fcc757f5789d69d4ad04ed94b41d05aff3f06c1d
52730 F20101221_AAATXO harpley_c_Page_048.pro
a13ec30b2093b32c4051cb165ead4b62
de5b75d37c2c0afa0e053fe3985c61921c6bdd5d
F20101221_AAASTY harpley_c_Page_126.jp2
76bc5dfbc047a99c20fc1ed8ebfd2e4a
7313816062ed65466efbae00cd64a2645d8dfd96
8971 F20101221_AAATAH harpley_c_Page_090thm.jpg
3c08c8ad8336dc91ed95ca73417de890
d935dbb81b40cf4370f56fda105246df19768f58
37425 F20101221_AAASVB harpley_c_Page_028.QC.jpg
f4feea12a7d2083f32bcef3a48d753ef
79e9a50d991e779e65d7156a844306b756b80fbe
2078 F20101221_AAAUDK harpley_c_Page_141.txt
1cf29f4b0087023389afcacb810269e8
0c2ee4bc387dabcd4f59b6fe8967a33777baefcb
477 F20101221_AAAUCV harpley_c_Page_118.txt
d3ee9cbf5520e5ff4056dea91c00699f
d7125f171d1363e71feee860b2ecdd6921a55324
53625 F20101221_AAATYE harpley_c_Page_074.pro
09326c60adf221bf713a1c5da4e98797
739d44471fff2ff1665ee14f44093593681d876e
2139 F20101221_AAASUN harpley_c_Page_042.txt
352b285f0df5da17095e2a727bed2371
b3b9af0ad586234ea9706495348889195e312661
51750 F20101221_AAATXP harpley_c_Page_049.pro
905d40a6320fcc6fd794f93bbd6e7cd0
5f78dfa2489d339e606c92d09f80d4a8399760ba
F20101221_AAASTZ harpley_c_Page_029.tif
0c9a7bdfa072e305488ccc0966b74238
62c2ed924cfafccdb4d9121d0d5d9101e40cb685
2194 F20101221_AAATAI harpley_c_Page_121.txt
907601099dff98605ad018282f4b05ba
aa244d91882e1f003e56af329a9eb53ceebc5d5e
2183 F20101221_AAASVC harpley_c_Page_143.txt
4c39663a187b27c1f46c8de17d77be2c
d0e0a03e1f364db26549a32fd6b13e30f3f078cf
2208 F20101221_AAAUDL harpley_c_Page_142.txt
2185b2591ff742f0e5aa9b4f9807ff69
c1f74ea2bc6aca52b18f222fb98fab2d087c0024
F20101221_AAAUCW harpley_c_Page_119.txt
7e89b1321e3fb9998ea49b29b3e1efbe
84d21d4a910a52bbf7fa457ad9907d94599fceff
53355 F20101221_AAATYF harpley_c_Page_077.pro
c01d6896e9b4489f089c8f78ba96c415
569cf76f80d0d3fc1eeb8cdb6ebb1cae8a5d8d06
F20101221_AAASUO harpley_c_Page_053.tif
ecc20e9b1208f511a019a1b7ccc3185e
a20bbc8d9f5f3566cdff11d78414c3bc7639e567
52438 F20101221_AAATXQ harpley_c_Page_052.pro
bce393ea476e6287a9b0d0e8703165be
385ac1e918fbdbc64e8270bba03213d91772375d
36236 F20101221_AAATAJ harpley_c_Page_140.QC.jpg
c895f8feef896152e86b08dc7256eda6
eb9dd309a94aa12a5a9c80911e9b2e7aa3f9eef4
55643 F20101221_AAASVD harpley_c_Page_122.pro
4ac68a5169c4d4c9987dbadc912c5a93
07dcc7b75de4658f0124b0d989173ef98d0af482
923612 F20101221_AAAUEA harpley_c.pdf
8ed583acbbd3bd783e73aa5318190ec2
12cb147a61ed2fd25da2748f38a2ff8e0311f7c9
2039 F20101221_AAAUDM harpley_c_Page_145.txt
670369701d28e9fd950deb35d312c2d4
562730ff08f7f80bc484db65fb075405fa028f73
2185 F20101221_AAAUCX harpley_c_Page_122.txt
e0a752be48d84de8b4178605602108cf
7c28d02e388c48206e64a2ce2223011379fd6106
54091 F20101221_AAATYG harpley_c_Page_079.pro
798c4d0a926e2f6bcbc88c5e18708a89
2d19ee10a0c5c51825fc84f2d1ae94a365bebcd1
2508 F20101221_AAASUP harpley_c_Page_154.txt
46949e0102b5b5953ae6c93e9e6e6852
58db7dac86b5231d0ea894e07b05b79604f44722
51195 F20101221_AAATXR harpley_c_Page_053.pro
e7546c1d6404ff00cf3e4026f3da591c
aaf3360601da5885d2e7e9da637fae12c177a2ac
2205 F20101221_AAATAK harpley_c_Page_040.txt
94c91a2814cce1206e81e13421ee28de
990f154d64c4c131bc97d43b53464028a2a54a19
F20101221_AAASVE harpley_c_Page_056.tif
4cf3dcf23fa9a53f1f97a26eb1dcd1b7
1a6172f2953b506044229ed2248e2e4051cb2d18
8323 F20101221_AAAUEB harpley_c_Page_001.QC.jpg
e4ae54cc3ae9803a6bf1c892393106d5
02a459443ff5b1448ef607977809eabbce1b5497
2252 F20101221_AAAUDN harpley_c_Page_146.txt
82b2061567d991629adfe61c0e5937c3
3c40930b8e31d1398f769beff480f6124f587378
2118 F20101221_AAAUCY harpley_c_Page_123.txt
618007d9bde9e26fb6a181d0e138e0c5
69b09bf6f629e79ebd6147057ddde5e216dc6fd0
53675 F20101221_AAATYH harpley_c_Page_080.pro
c3e9675723faaf972a03f32cbf49431f
21b3632455808cada05b65fb129d7b2ca1509b89
F20101221_AAASUQ harpley_c_Page_035.tif
09915abed425d84e566954f0ea3854f6
3e4074e47d67f3985c29ef0252805b9df608c9e3
36596 F20101221_AAATXS harpley_c_Page_054.pro
eb13997547890bd93e7dafaa7b75e9fd
db85297d7b7f652304b2006df7e3ccc6ad2156f4
2220 F20101221_AAATAL harpley_c_Page_030.txt
dc39df5cae5609de656fa375f932642e
723379ef99e2a483e1de12aa19e4a44d53727148
411591 F20101221_AAASVF harpley_c_Page_110.jp2
c5411424597eca6597b7a4697848c99e
a54453a359a58f40a841b65d145fbcee223b6893
1188 F20101221_AAAUEC harpley_c_Page_002.QC.jpg
86ef624e6295fa9a4aa0f5ecf3874c2a
de3ae82932b8a8b75111e2400de87a431936324c
2146 F20101221_AAAUDO harpley_c_Page_148.txt
27ff6771c9092002a0104a850260d78d
6700c3c15c64c87f74e0d95694c8aa3e4a67086b
2125 F20101221_AAAUCZ harpley_c_Page_125.txt
65890a1976cf8e15298385161e2394c0
a7567484a6a9bb3afcae4ddd221e2d7134488094
54616 F20101221_AAATYI harpley_c_Page_081.pro
ae936b8b4d24bfc00101028d828093dd
f8178ea9e246637691b0cb8b98b8068da25093ab
52581 F20101221_AAASUR harpley_c_Page_029.pro
304ae6f7ca2d68f3cab300bce97fae13
1cbe844010d4f30d974c302c18f8b575fc0a8e0d
41545 F20101221_AAATXT harpley_c_Page_055.pro
c29c5dc1c81b089c8bd5509602ed073f
d52549a348b3f38e43a5e621ad668d32d1454e15
1051953 F20101221_AAATBA harpley_c_Page_125.jp2
8aa247ccdd5212c8721d71a914abccd0
e4ce8a12811595251045b2f0616fbb3d5ac9a27c
112913 F20101221_AAATAM harpley_c_Page_150.jpg
fa42cb0470b5ce646d7dcc49f9ac54be
4772a904c00a7c069ac7e60176d07f81d86409b5
998 F20101221_AAASVG harpley_c_Page_111.txt
a980f4574067a5190e50c6d6f606ca87
f4075627bbcdf74f303a6c621dc9cda1efe953b6
616 F20101221_AAAUED harpley_c_Page_002thm.jpg
ee7a9fe5c2454697cbfba5f91aaa9324
d882cea7a8baa7886a6e2a1cb62184892743ec9b
2149 F20101221_AAAUDP harpley_c_Page_150.txt
dd0291435c7a5a659d6558aa88f84b61
8cab2d97b8c7faccc69bdf3438cbf03c3fa35229
45954 F20101221_AAASUS harpley_c_Page_056.jpg
c88a1af96fc94ea797a2b4329ac4f5b0
c45462525283dc088ee7fd06a8ca8780def74830
53971 F20101221_AAATXU harpley_c_Page_058.pro
f0f402abdb48aa7ed0f59f8f1f06d629
b5c14411ed2e8c982d991fde303a8532b2ba183b
10681 F20101221_AAATBB harpley_c_Page_118.pro
243c594cb0f885519c5270559ceabecb
f8e16731d63e8a40ed2b52427d58412964cc124f
13439 F20101221_AAATAN harpley_c_Page_034.QC.jpg
58107edab1c53dd1e441a639177e4155
139bbe15b5f7956b4893b8bcafc5128e8292da55
1033 F20101221_AAAUEE harpley_c_Page_003thm.jpg
bbb1fd0ad2f2f25f905aafaab7667223
002ce3f52cb269cf1a9e3f9d4565366b761ee931
F20101221_AAAUDQ harpley_c_Page_151.txt
ec853c8a82bf102a780e2bca56457b94
f10d3a7e1085ca1c1e1c7210eb5bd3227c229fe3
47820 F20101221_AAATYJ harpley_c_Page_082.pro
9fdbcc94c2df19a86e84dc66323c3d0f
85c43271c8256c722c28c67e0f4eafed9c7b3e44
53629 F20101221_AAATXV harpley_c_Page_060.pro
fbca3707a1bb8befecfd6094b2990deb
696581efd023dbb195b5f99fb4ce0c600ddda534
9207 F20101221_AAASUT harpley_c_Page_038thm.jpg
055ca7f1768cf4705acf5a99804f2fac
696b0f8a5138b3a4403aede43b77ac7a609a3ae6
35408 F20101221_AAATBC harpley_c_Page_053.QC.jpg
27253fc98239fd012543108ade638a0e
24a0fda3f12387eb9733519c598343cf69ca08dc
38054 F20101221_AAATAO harpley_c_Page_066.QC.jpg
8ff1365829865f2aa3bcd4d0dc171881
a0121d346b40aa114d35f959a33915455fabfc45
F20101221_AAASVH harpley_c_Page_161.jp2
2015aabdefcdf7c709fe925cdf9c8fe7
d8b1071c483c74539da81684add5bfc8a9fe7f18
7968 F20101221_AAAUEF harpley_c_Page_004.QC.jpg
20ccf0819cac1383cdc78fa8c1c42e30
f16c7da0d5b82f679f4095525ccf9e48a044d195
712 F20101221_AAAUDR harpley_c_Page_153.txt
555a06fa44a013488ff882fc6dbeef92
bb730d1cf06235216b47cffe7705c0365e0afdda
53374 F20101221_AAATYK harpley_c_Page_083.pro
fe639a555c4adddc8f4f737b0abb95e5
41915021095a4a6a40752a36d6ed2025badee45f
56312 F20101221_AAATXW harpley_c_Page_062.pro
84f88a2663e87459351017c45b2791a2
0f2f5732f7474309c7947123405b7636678a561d
8789 F20101221_AAASUU harpley_c_Page_135thm.jpg
e1ad396630324bbc4b0a7b9d1794151c
2b866d58d2620634fd1b5c97aeb47fa535a8d98d
F20101221_AAATBD harpley_c_Page_140.tif
b43cc2a3a49555e988d71eb96408386a
3ffa579d042faf03fbba8f5d6e9dc6c2b3bd6313
111230 F20101221_AAATAP harpley_c_Page_141.jpg
37e474f2ba4a1e3672887b521afc2581
ee909b0cfe7ca9256b284b7057dec57dbcc9fdb0
53981 F20101221_AAASVI harpley_c_Page_119.pro
11844e71b34c9fbae1675644db8f3a4d
1614465452814c47b4fce7aef05a261499ec3ab7
1967 F20101221_AAAUEG harpley_c_Page_004thm.jpg
848df0a6c5385c2763936a02e6ed7048
439dd60c47de47a7053ae77239701636420370d2
2801 F20101221_AAAUDS harpley_c_Page_155.txt
c5501306b6f7dd87c74b128d0308f008
b2a5fa73e5b7e282042af341cdbc48105d2406f0
29152 F20101221_AAATZA harpley_c_Page_114.pro
a241f47cd1b4052ca0c7c11c9e25b80e
18b6431e24d9bbc2643fdc4eee9abf5d8b3b9b78
53050 F20101221_AAATYL harpley_c_Page_087.pro
ef81f0278dbb2012ada4ed0285a8d1d4
2130a83b0194cf60b50c5dde2c7a054adaf4070b
51384 F20101221_AAATXX harpley_c_Page_064.pro
fad3ec2632d850d1e4f099da591f0c6e
d92dd4441748896750b261e63715d96849df29d2
26585 F20101221_AAASUV harpley_c_Page_163.jpg
c9235b308bff169e89204a48ccff1128
c7360672a79e3e7918c2fef2ed3ba4f964dc4a4f
23525 F20101221_AAATBE harpley_c_Page_099.pro
dd1d8196fe4dc530a19c1805e2d950cf
82d9f6d2eb684da9987fafc0f147958277fac16e
9090 F20101221_AAATAQ harpley_c_Page_065thm.jpg
381f37aa1e79ccb3f4857e86b783a857
cc94387de8ad7b0214c5acea32f3dad047bad38f
39184 F20101221_AAASVJ harpley_c_Page_024.QC.jpg
d94049910150b0f79f392aa2ac147bc6
75f118660433c0a06e59f17e0976c24b24245b14
27466 F20101221_AAAUEH harpley_c_Page_005.QC.jpg
e34eb0296cf9a82e56cc4b1b693efc90
72a02be7d8c572f0ffe9b263fe4e4302e504f3bf
13643 F20101221_AAATZB harpley_c_Page_115.pro
3e207f242acda7b4ba11ac9445dcb77b
0ecfc8cbfb029b44d3e03758cddc8e12599034de
57249 F20101221_AAATYM harpley_c_Page_089.pro
e3acd2a64fd2c9564acf0aa6d6a38eaa
980f7d760710b7e02180ab7702180ad8b5b83850
55651 F20101221_AAATXY harpley_c_Page_065.pro
4d36b77d81dcf252c4ff07362ac70c56
b3b0df069c6b47eb6edec6ec76c8a1caf7f2d7fd
5977 F20101221_AAASUW harpley_c_Page_114thm.jpg
290036a565bece682433eefebe689d60
fb330f381860fbbb15580bbb81266552e428cac1
57046 F20101221_AAATBF harpley_c_Page_149.pro
7c351c5a150f0e9368424ec1db27f6c5
64499b9df7cc3ca598f9f85fe018fbe3565e8970
28126 F20101221_AAASVK harpley_c_Page_097.QC.jpg
e3a97ca39b5bf952406b130b21146fff
09dca7e5013837c0d22479a05edbadf4470ce3c1
9132 F20101221_AAAUEI harpley_c_Page_006thm.jpg
4ab370e1d250cf5f468aefab67da21c5
fae6daae7132d1b50316bfafc2c4794ded79fb87
2750 F20101221_AAAUDT harpley_c_Page_156.txt
39ac485f6448758103b2e83f057d516a
4272a66d3816e084fa323fb514986e0645d8cc32
21575 F20101221_AAATZC harpley_c_Page_117.pro
eb52ed761b86b3d8f78f1715808d87f8
50ec27f57fe94b210397a8df56bf725b2e27f9ba
56222 F20101221_AAATYN harpley_c_Page_093.pro
925efa511f3b54803f2c2648b15c0bc1
3978096ce3776b37ba7e038a939cb1c57aac794e
56364 F20101221_AAATXZ harpley_c_Page_067.pro
9231a5fd48458bd510db87574f99c9e3
40aa71e2ebde1bc257b611a3d6344f9bd48931fb
F20101221_AAASUX harpley_c_Page_019.tif
b330242f6ae6a1f6815ce7274abc4286
3ced1aaa3f429971759d24ddeb3225491b1f109f
2108 F20101221_AAATBG harpley_c_Page_074.txt
949d0d109120e68a6fed5fe7335f13d8
7a3d4e4543712776dd6f2034c0dc80a0e9249107
8893 F20101221_AAASWA harpley_c_Page_150thm.jpg
29c64fa20ae575a66de283ac42b2acc8
96a00f3e4cf434cf173ba953ff5d168aaa900c40
2190 F20101221_AAATAR harpley_c_Page_018.txt
4cbade8a10caa08598cacab636c0c950
5dbe6bd9802c6ccf7bf881f4c5d1cba8e9f6a970
55507 F20101221_AAASVL harpley_c_Page_092.pro
9519da7dc064dc5ad534047c25588b60
84f32644fd83123294bbff14711890c48a506c2f
1892 F20101221_AAAUEJ harpley_c_Page_007thm.jpg
6f3e32574663f850a1ba7140f484ee77
fee209142f366499f3c29a94166622a2bf23cf5b
2720 F20101221_AAAUDU harpley_c_Page_158.txt
729e3e687c8b0395e1e504163883853a
4970d17d48237d2c64fb50cb76dfac466353d2ed
56034 F20101221_AAATZD harpley_c_Page_121.pro
911fd1bcb30cd3dab0f40cc0baa36bb0
23e103b57b6d27603a53f134c7588e2373ea5a36
53928 F20101221_AAATYO harpley_c_Page_095.pro
3988e1bed94c6044c7342c93f7ee7122
9a86bea20b0082006dc101a2ee4738a2d95fc2b0
35770 F20101221_AAASUY harpley_c_Page_120.QC.jpg
52c0adde4ef19c1a50e15f9bb663a641
2a37d16be1709bd19ea7fa379e1bc316123d42d0
37361 F20101221_AAATBH harpley_c_Page_088.QC.jpg
d1415f6d5edcfec909b96fa343ba8d86
e8d7f6a9dbe1ad6ff8089e86860b43aead7e81e7
9095 F20101221_AAASWB harpley_c_Page_127thm.jpg
91653544598a572b0e62ae0415e5e47a
fd90170bad880367a6267abec33634f8ffc45232
36215 F20101221_AAATAS harpley_c_Page_059.QC.jpg
7a51e5605d7ec2afe85814d5ebd821c5
447fcbd419b6f635bb67fb91eef4d5ad863010ef
36200 F20101221_AAASVM harpley_c_Page_087.QC.jpg
0ea8be9323e066e25678650c3e80abf5
069e826e658db376af91ee98bd4adf34dd9af8db
14695 F20101221_AAAUEK harpley_c_Page_008.QC.jpg
5f956000d8b7c1800f8f136b75f084b4
00143fac1d3efe367f91f93e0c4fbdd0970636d9
2860 F20101221_AAAUDV harpley_c_Page_160.txt
7b083b16d40dc4c9fdc7adeb904d3e94
8467dea77547972666315c3bac6abbe3fc191313
53942 F20101221_AAATZE harpley_c_Page_123.pro
f932cb8149bff09f8bbe3b1087336089
ebc99c978bbcacf34e1550b17046ac6b2877920a
43791 F20101221_AAATYP harpley_c_Page_097.pro
f3f9c78129a67404efa0c84a44c0ef56
814ed9503eeaa02db1bfa888550f86d663e024ae
28214 F20101221_AAASUZ harpley_c_Page_102.pro
edcada849e0e98e0edca1c9ed185eeeb
d3842ff463a750f80defa460f3d589eae68fbba0
F20101221_AAATBI harpley_c_Page_122.jp2
b166bc052358551247a27f772e5d341f
ab07ab64b09a5237551ecb9ce8d7a0a1c057b3f8
124436 F20101221_AAASWC harpley_c_Page_154.jpg
82ca624bd3605286b578fe524ef657da
6d8593282e312cf4fe957d80ce9792a73af4c32a
9091 F20101221_AAATAT harpley_c_Page_063thm.jpg
e390f04e7cefd29020b46a158e7b6312
71f9fbb23963b37e3107b004ac23bd2f953d06fe
8757 F20101221_AAASVN harpley_c_Page_017thm.jpg
cd7aeacff9d21e00a6a5152baf83b816
c886f64d5e04c3696dabe178de97b08c4a2fa77c
3727 F20101221_AAAUEL harpley_c_Page_008thm.jpg
a37612b8d4a254cffbce71aa507541c9
c18537918b6f60aa8fa5dae91807506922fc946d
2695 F20101221_AAAUDW harpley_c_Page_161.txt
c088fbfb52cdc26f7df9ca21186c48d3
bf3a13ab2e87fab5996f8dd8f441486256d700a3
54151 F20101221_AAATZF harpley_c_Page_125.pro
29ee8b9c11d6ec473cb5c6bdcbb569c3
ea622f7321180c91391d76a4a8b62895468fd4c5
47466 F20101221_AAATYQ harpley_c_Page_098.pro
a6a0b5203f43604f3ec5167de349aece
2b22de9a679226378fe4582b86d2861b6cf3a7fb
9005 F20101221_AAATBJ harpley_c_Page_141thm.jpg
54b000f62fd3f535f9f1ac143161673c
bb4b2599c2b322a139a9f8b40a64812c2b002527
F20101221_AAASWD harpley_c_Page_125.tif
92966d759303efab16080cc94616c4cf
6164eb729bb1e66d9d2770451aac50c2a06c9d9b
111258 F20101221_AAATAU harpley_c_Page_022.jpg
84bd7b43e6bb58d8d83a9cdbe7ee0d60
79e22a9b1e68e7036ab47b73031ea9c786ff366d
35896 F20101221_AAASVO harpley_c_Page_137.QC.jpg
19381d94c7a0d62622ec4e00172df354
3867f400be301a290a698a1d2031f7e4c801773a
8790 F20101221_AAAUFA harpley_c_Page_020thm.jpg
821519c1c392c6d25f54af48a5b69e55
073077519f68a3f8b016ee10ed2706df2ef55c09
8825 F20101221_AAAUEM harpley_c_Page_009thm.jpg
e5d3cfaa5dc465cb089640760ec4acb7
9a01d8716791fd8a4c595b3a7a470914fc78f41a
1381 F20101221_AAAUDX harpley_c_Page_162.txt
cc2d440a66d6e9958d0e02f05b23240a
ceecbca93caf7cfcd997f98d9cacccc97ee6d37f
57017 F20101221_AAATZG harpley_c_Page_126.pro
85f353e20da3965a212d487e3bc24a00
d066a1bf37e0220fd090dd7e68019917e034049b
11392 F20101221_AAATYR harpley_c_Page_100.pro
cd66a6fe302e5c606e68799fe713862b
8b19d016a59dcedb2b64b35a667c5eae3324ea6f
1944 F20101221_AAATBK harpley_c_Page_097.txt
79d39c2c5fa6e5b6c4b14b3c97401742
28f45cebeccd119ce58e59750c619815713fe71f
F20101221_AAASWE harpley_c_Page_006.tif
d5043dce48c141b27be8a867b77d4e3c
9108743fa2775630796ad1c0129d4a33d676e693
40033 F20101221_AAATAV harpley_c_Page_071.QC.jpg
5ded0f8ff49f6154e2c0da4e4095f58c
2f4dd53d1c171473b1212645cb6807edbd91095a
1051924 F20101221_AAASVP harpley_c_Page_143.jp2
628c47656b81d66991b616e5b0d16c49
a1b3fedc92df3a1a9336a00682f6e923af5d9bc2
35171 F20101221_AAAUFB harpley_c_Page_021.QC.jpg
fe55f3b846a9cc4c127091309f3df9a6
f4892ec5362e99fca725d1394a6931a534568680
7294 F20101221_AAAUEN harpley_c_Page_010.QC.jpg
640ec524bbafb45467f4d6138cf89089
07b77eb1ca208bf3829327e587755780f13b039e
465 F20101221_AAAUDY harpley_c_Page_163.txt
82af456994a33db96c3adaa043f561bd
0860119ed65a8c8d540c7a375ce014973f6b8b54
53085 F20101221_AAATZH harpley_c_Page_128.pro
8849c614c40e7c31028ef384c6c89d03
7119de1ea957da6e9a110231a631a26079a0ddeb
33259 F20101221_AAATYS harpley_c_Page_103.pro
191020fd753caf0642f5eb82c6f5de40
b9a0e3b828bf78a6ff58b8b4b8da6eff69f3fa38
54846 F20101221_AAATBL harpley_c_Page_127.pro
3cc0e8bd1663e248a3857a9c552f2806
476536c2fc016ce57be509c37911bde7c1fb2ce3
2152 F20101221_AAASWF harpley_c_Page_068.txt
7b0367dfd1efc952ae339bcf8c53b200
c7731f3c2aa46e28ab4599ff4d307cc8c97f2de7
18699 F20101221_AAATAW harpley_c_Page_056.pro
db8b1409938b6cb8a5a2d46c5757e931
cb195fadf97f8510d72282a580076f9a9eb841ee
F20101221_AAASVQ harpley_c_Page_025.tif
6becad0ada06ae8925301288315a8e69
6a0e4f9da6b515b0661559b9d10514b08cc46f29
8758 F20101221_AAAUFC harpley_c_Page_021thm.jpg
3a638ee26fd3a21545c2ddacea4286ee
dbe6692d930a3256df244653fc9cdb324fb24522
17403 F20101221_AAAUEO harpley_c_Page_011.QC.jpg
65bfaf6fd89cae465db22cc650a615ae
0a6e4df77a675dea91504de238ed6bc7f35ffa00
2056 F20101221_AAAUDZ harpley_c_Page_001thm.jpg
1d60d4891f76f462d1e61a8cc0a7de98
0921fae5a00c058675d1bb545176491fb37e2978
54647 F20101221_AAATZI harpley_c_Page_129.pro
7bf251b0094b10b7119c2e4f0cc55b96
4f6c35a8b66d9dedd952079c01af1fda382c9ecb
5964 F20101221_AAATYT harpley_c_Page_104.pro
84607ad63f8116d3904c01b3dd57b8bb
27f1e71780f977921d5a121f8842feb6bff9abfe
8996 F20101221_AAATBM harpley_c_Page_028thm.jpg
c76148527fbfe9ec5355ce0bfba20aff
2661e7826f8c40d4e56b994ea923764f94e01e1f
9172 F20101221_AAASWG harpley_c_Page_130thm.jpg
e8d73fcaa75357cd45f2f19053efe82e
40e65eb5c21eaa85b76b8478bc8c0a5d263da0b7
37857 F20101221_AAATAX harpley_c_Page_158.QC.jpg
858a19cd3dffeafe670326349720f54d
51dcd47ed10ef8eb0733f15580eec5eced0f3007
987869 F20101221_AAASVR harpley_c_Page_055.jp2
d8e551293dead12b1386ac9dec9c179b
1b655e3e6173807bbe3a233c9168097a6e72d51d
1051931 F20101221_AAATCA harpley_c_Page_029.jp2
51f117cbb45f50ed5f17718ef9f55965
6aed8f8e141db1ca3284a8efc6fa62522f923ec8
36298 F20101221_AAAUFD harpley_c_Page_022.QC.jpg
74c35c5de1129d20d966874a61f6d523
28990db70967a3f20ce76844730bd7c01675d21a
4508 F20101221_AAAUEP harpley_c_Page_011thm.jpg
2f2ceb99164a53df6e0204e6dd3750ea
4253831417fcf9b09b4415310dfe38011bc2f3d7
55454 F20101221_AAATZJ harpley_c_Page_130.pro
1c3b8ba5c170643421f23211ab9f0625
17a6b545fbcd5fe3dc4171df7bae16310c2e84dc
17833 F20101221_AAATYU harpley_c_Page_105.pro
5e0436c50a30050f25b327eb06980318
6b13b241956ce88ac354e5a7c419052623db797d
6809 F20101221_AAATBN harpley_c_Page_005thm.jpg
43d5deb112787885f19c1790961be510
7f665a6227942e12ba4b724cf6a67cb0be9396c0
F20101221_AAASWH harpley_c_Page_078.tif
9d9bfdda4ce57076531e55ebe32b43c3
897a232b8740fc0cf82186a2e9ed6853831f7ba2
52851 F20101221_AAATAY harpley_c_Page_115.jpg
53443e87681f6b07737b9c155d6b8b0c
2e40e20889d7056ace0e13fa60a3a3d1b2cd6c5b
F20101221_AAASVS harpley_c_Page_137.txt
f2ffbdc89649a17008e4db594913dd90
169155a82012fb7840cc7616b29ebda2d779d8c5
2043 F20101221_AAATCB harpley_c_Page_051.txt
e1051ca3eaf3ca4be8d6d0808ff678a8
c3b6cda6025b8deb8b75f52386182303c77c44e0
9293 F20101221_AAAUFE harpley_c_Page_024thm.jpg
4652b68995f0c74b1dc5fdbdcb37d337
c12172365a28a26fe7600682367954728482cc23
9191 F20101221_AAAUEQ harpley_c_Page_012.QC.jpg
bb2ecd7b7475319eb5d5fa2c1435edb8
22f83163210a1dd20e2b1918b7b6a794ea014022
22241 F20101221_AAATYV harpley_c_Page_107.pro
8ee4e1e0248b2f139653819549520af0
e05c117bafadd7e085facee3009a740d56786d90
38931 F20101221_AAATBO harpley_c_Page_039.QC.jpg
f62b4ed01e8567d3792be0415bfbc871
b3f4e3622b365d7755edc35de69efae1ec4e59ab
5455 F20101221_AAATAZ harpley_c_Page_152thm.jpg
c5eeb5af4f3bc2a1a175f93d052dbe68
86e94958365754db6a9059096da08ad10de728fc
107462 F20101221_AAASVT harpley_c_Page_016.jpg
e3b43dddbed30c8fce20c8f66885b7f9
b354ba6d476008fb4bec4f77be28816c5bd2f943
F20101221_AAATCC harpley_c_Page_067.txt
03fe461bae1ca6ca4228e4bd053f2a3a
8200f6ad64743b291492de28de8cd2c6bf5d18be
38894 F20101221_AAAUFF harpley_c_Page_025.QC.jpg
468c4c90c3179de738cd853e404a6305
fd30d3f65a830061398273624af2c863c0847f7f
2414 F20101221_AAAUER harpley_c_Page_012thm.jpg
0320977b5f4fa00d6ff1255c82f4bdc9
86bbb51ee2b3b7e77e719ef2b0ce17c45aec2658
50760 F20101221_AAATZK harpley_c_Page_131.pro
297a70e5d384580ffa2e8ad1eab266c1
c52bbc893fea911ef12ff0d36ddf6fd2a084c964
27046 F20101221_AAATYW harpley_c_Page_109.pro
e4b5df6b1a2303eeee768836481b6e07
e41f8f46bc34f3385b9058f1dd5ccfaaee3e4ff3
16703 F20101221_AAATBP harpley_c_Page_108.pro
e614cab8870855b257d57f23469021a9
e83fd730404dca89440ab0778a1b66788e7cf2e3
110954 F20101221_AAASWI harpley_c_Page_080.jpg
d6d530ebe433ed86e27f5c0974ed3dc8
5c94950d34ed31cebfeed7d725f24dd7176e65e0
36495 F20101221_AAASVU harpley_c_Page_136.QC.jpg
27c12fda9c73b97e4aaebcba19c9094e
b988e4aaf0346ab39e89e505abbdbc8c55b34521
9007 F20101221_AAATCD harpley_c_Page_048thm.jpg
43166487ba4d0dafec8064a9c049cc29
bb4f1508f36eb34f67e984b3deaed0c3840a91ce
39309 F20101221_AAAUFG harpley_c_Page_026.QC.jpg
169014ce7c3c9ecc97b9d925ee78253f
1f9e08bbaf84b41fb58f69f96409378300c6cbd2
34473 F20101221_AAAUES harpley_c_Page_014.QC.jpg
7723d913da74011c97cad32b3bab6678
5598c3a9ecf5d0960ae9778732bf73be9dc61065
54840 F20101221_AAATZL harpley_c_Page_135.pro
d6bb44873aeb569383050750fcabb430
a866d059530de9a91e228bb3f09e6581f66a7fb0
16506 F20101221_AAATYX harpley_c_Page_110.pro
8a19f41e2de25d81b3e3c2ef6a607a97
9e552aed1225fd30238f170998ae6fc1f700305b
F20101221_AAATBQ harpley_c_Page_132.jp2
c60c1ad24ce1a3f1fdcebbae5f540a4b
c2e1a9bb1cc3d517f6489744b917b41cb2868699
55093 F20101221_AAASWJ harpley_c_Page_088.pro
caf946f9d67b56f5559f0c4ff91ee5f9
1cc03349ae9ac644509bdcc703a054a39a72e130
F20101221_AAASVV harpley_c_Page_023.txt
26e5909a7f6c9e6a8072ce016b14aeed
297feaed43c41f4d5fceebf2a8cb74cc58a2d471
36419 F20101221_AAATCE harpley_c_Page_150.QC.jpg
3b21920c047f96e2eda33cb7f9e97d14
fbb773056c4dcfe9f156e0ee5565bf944677ea60
9192 F20101221_AAAUFH harpley_c_Page_026thm.jpg
d0b828cb6c908bbe7b691f5775702f45
0912e7160987050ed8b962e761fadbecdef381c5
8222 F20101221_AAAUET harpley_c_Page_014thm.jpg
56b146f09c42378f02e3cf6814fe9012
0796cf72168077e9419636f4b9720031c93bb115
55123 F20101221_AAATZM harpley_c_Page_136.pro
6857ef87560e5c4ad6f7a2ba79df76cf
f4dd8eb750ea4b3489af9887012b4d64fbbf81c2
22497 F20101221_AAATYY harpley_c_Page_111.pro
1b15c908c1533f1a9892de261717409e
b0dd9015862a0c25100ca1081696702128195d24
36136 F20101221_AAATBR harpley_c_Page_019.QC.jpg
0819d5deaa813a51350c3afd92082d14
724974f624d74c8a253c54a55266ef63b46335dd
111105 F20101221_AAASWK harpley_c_Page_145.jpg
6b5278f2a7db1a28b441d15d91a52f95
2e426a4fa04323d8af0d23ae2204db87e8ec0270
53871 F20101221_AAASVW harpley_c_Page_059.pro
4eda03f91eaf534f6c8e4073a7cefa85
cfdf6acac64d767801e94d32a3bdb5017fbb6c4d
53199 F20101221_AAATCF harpley_c_Page_069.pro
92cfc999b95779ef02d9bd82f9cb70cb
57a398d64f6a132f6f070ad71a245a71cf3e0308
36922 F20101221_AAAUFI harpley_c_Page_027.QC.jpg
a80c1c0f4b606d44647dc190dcbf1a51
8414773de8da907b7c990742e92a3130a390cf46
53815 F20101221_AAATZN harpley_c_Page_137.pro
1a9cedf382458fe893ac6b530f4c68ca
5bece82d50c94698196df8978a8cac953da58201
21629 F20101221_AAATYZ harpley_c_Page_113.pro
48e4ca9d6d91fc2c6d38b98f09099415
6079e7cb71b5d68eb830928cd7957cb99902c396
1051948 F20101221_AAASXA harpley_c_Page_086.jp2
d58b54a2d3ac76a0f432352b56d35915
934ca5cafe4ba09b8b98a28f68ba64d050e758e5
37889 F20101221_AAASWL harpley_c_Page_079.QC.jpg
e602010cff340b39ea3a7f0ef6a09db4
da7c2468839ce5487b290d5ffd9c946aabe5354b
37699 F20101221_AAASVX harpley_c_Page_135.QC.jpg
12be40640cb02a4e9bc0b5b27206b5db
51ba2bb278d1d722c238d91bb0d43ae49c074820
79886 F20101221_AAATCG harpley_c_Page_117.jpg
023a8d458419825420ccf28ed5874ced
85b0d94e4dab33914572aed018ec9a66037fbabe
8685 F20101221_AAAUFJ harpley_c_Page_027thm.jpg
7a4e33700fba20e76c1dd70d3f92a590
410379e7d7b5eb682fac3329611d45f1e3b614ec
F20101221_AAAUEU harpley_c_Page_015.QC.jpg
7a74ccb7312dca2f4b6cdd1be779c95b
e40c5f3a2d4b5178d328ed200237da524455b37d
53808 F20101221_AAATZO harpley_c_Page_138.pro
2e04b588479ac715406c5ed30e1a1760
56f1653305689eda168cd0067fb02b1c5daec3d4
386321 F20101221_AAASXB harpley_c_Page_010.jp2
0821a463f3023c9805d7cd11f1233c7a
656342148d6e00b254de0e4a7a037542d8b929ea
F20101221_AAATBS harpley_c_Page_115.tif
4770979ce64ad7479433fa588fd7426b
dfbb63103b24286f6a8964b12ec095263f978fe3
F20101221_AAASWM harpley_c_Page_016.tif
f33277bc6a864abdb5b22da1168d6019
8c618dc94728919bc8f430d6f3a78d539a69da93
2142 F20101221_AAASVY harpley_c_Page_144.txt
6743ca8ceb4874535b34c84314ad50e0
eb6bdb1ea859f7284e555744ca44a904ccafcaa1
112270 F20101221_AAATCH harpley_c_Page_077.jpg
cd316f7340b9dc4dd675b29ce9597c16
78855cfbab0c4d679002015a7eddf87304353594
36664 F20101221_AAAUFK harpley_c_Page_029.QC.jpg
d7d6dfe2c3e9269531e66e91b4585639
d028fde6a0e63d926a39f5216e5df6b8eb36f582
8722 F20101221_AAAUEV harpley_c_Page_016thm.jpg
aa992de56d0a953102b9872be3d7b775
d0cee31f1b6312448064dc3c6e1f586e48c14764
53530 F20101221_AAATZP harpley_c_Page_139.pro
e9d326be77a3896286c31d3817a1cc90
cfe7dd9678d2fa70da6b5985834a4288a2ad1ef8
53883 F20101221_AAASXC harpley_c_Page_061.pro
1fd66a863178c2093abb51cfdf8b4a48
09c2025b560c940a1640ffdef57ddbde586a731a
F20101221_AAATBT harpley_c_Page_076.tif
f19350d520c50b91266febdb266ec343
28f39efa73c416c5be915f4258987c0d39b12017
1051919 F20101221_AAASWN harpley_c_Page_119.jp2
d97c79c917be57e7479046d9dc1c6efe
60e46c1ff98aa878bbff029db9fb7ea0df731e2a
7927 F20101221_AAASVZ harpley_c_Page_010.pro
8bcac6665e85e69c1e79957c752be13a
7b6356f1e5b73d74952bc7584af50e6a9bf96ebf
23053 F20101221_AAATCI harpley_c_Page_010.jpg
1da9f0b19ebaff20daf01481f74d17ef
e6af88950c2f378d490202b8d8db361b69897542
8814 F20101221_AAAUFL harpley_c_Page_029thm.jpg
94a614d1c4825f98c65be79f5155109f
6594b42afafc592349e1ce0c5c394ea3d35bcfb9
35986 F20101221_AAAUEW harpley_c_Page_017.QC.jpg
f4755352531fa4f393eb37b9b0b7fff2
50c4fc71b011581c03259a49ed260deda94857e3
52000 F20101221_AAATZQ harpley_c_Page_140.pro
b72b2521c7d3e1043b566083c78f0721
25e5b5c87360641452227c7823f0d581401fc866
36268 F20101221_AAASXD harpley_c_Page_047.QC.jpg
f0df390c4afce066e767891bc622127e
32756436aeae787020cbfd85bdac0c72106ab932
326378 F20101221_AAATBU harpley_c_Page_032.jp2
3c426375683c509676da5de6b8cea064
14afa7fbf9a47f766147395f49333bc6b3b0540d
2136 F20101221_AAASWO harpley_c_Page_021.txt
2b983781c9f1bf70fed5a05e8a41249c
14e32f520bc41335988d53de41d56a7341832229
525293 F20101221_AAATCJ harpley_c_Page_112.jp2
028375fc3a48f1d9cc53870f02cbb9a3
c303a990377b10cd12e43a392748232452779a14
36389 F20101221_AAAUGA harpley_c_Page_042.QC.jpg
48a5f325387cf7f7282403ab0fb66cdd
6331ecee5fbb071988dcd12121a85f5616794fc9
35913 F20101221_AAAUFM harpley_c_Page_030.QC.jpg
bc8292ec37cd8bb5cdd545cba03ff54b
f432f5e8c41bf1e3cd7b160694f852f9766ac9af
36406 F20101221_AAAUEX harpley_c_Page_018.QC.jpg
e4f2c9e7b86dd4d1934e3323ba18718b
bc377ae9785b8f1d2f12d7f128ba3b8786a32083
52708 F20101221_AAATZR harpley_c_Page_141.pro
a9a3a0df62e52089f9658862ee7ccc20
a809c60cea82b58d2b10c63e6e53f228e1e19a94
F20101221_AAASXE harpley_c_Page_146.jp2
b1c5d5a7e33ba6be130b567418461620
2b14abdbea537e8bf076ed6a3f0b30dd38495ef0
1051956 F20101221_AAATBV harpley_c_Page_018.jp2
b3ac97d370c0c65f6b9238d69eb8ef00
c777cdb5d8ae97439f54f727946d7d6e773ea4fe
21535 F20101221_AAASWP harpley_c_Page_109.QC.jpg
9c3b0a70fd5b1958bd7d1f92e626049f
cd761d4d89b29918ab3ef5b3823fd755a27caecd
58546 F20101221_AAATCK harpley_c_Page_024.pro
c14cc84695daf755f2878579a2c0112b
62c498e32180b21cf46fe55030c84d5e6c4ccc6b
8229 F20101221_AAAUGB harpley_c_Page_043thm.jpg
e61e3bb0423d5515e306ce56ba33c973
b52e72304face8f2e78f1001d02518c55ab434d9
25445 F20101221_AAAUFN harpley_c_Page_031.QC.jpg
5a052ddd175b72fc99530581890f323c
22a02a06c5932e594c85ee971df32cde660273ce
8810 F20101221_AAAUEY harpley_c_Page_019thm.jpg
215285c9ee4e96c978a447a14a49b71f
e99818d32b1ddbfe173679f67b10b9ae594f1c9e
56014 F20101221_AAATZS harpley_c_Page_147.pro
b53330aa77bd9e573a3bbafb48f3f3ee
680fadee62d542247c185f3b7f8c650a81741402
8704 F20101221_AAASXF harpley_c_Page_163.QC.jpg
93e52e77c25ac04af00c75cca207ad4e
397a20c02c74f4ebfb2499b1b660eb115ca23afb
112570 F20101221_AAATBW harpley_c_Page_041.jpg
4cb6e59a38cb448e72522be0f8a7a79c
d52435b11ee585900910971a3da66aaf4e5244d3
9344 F20101221_AAASWQ harpley_c_Page_149thm.jpg
eebcf8cb7eaf2945d3dbb3e65553f7b5
bccdf02a7c81b418b04ffda3457f3a6d1537d96c
F20101221_AAATCL harpley_c_Page_007.tif
89564812858504f2414658d87efcc0b7
8ec67a93f46a4a2fa374cfeea6e65d151bfae399
7980 F20101221_AAAUGC harpley_c_Page_044thm.jpg
a5b036d448761b056218e46de26ddd4e
3a0d66c5d704d45322f77b91fa583602be5b2f52
11170 F20101221_AAAUFO harpley_c_Page_032.QC.jpg
f845c09ee768d0b51af6e167cadc899e
42e421bdad2bf2974ba36ff10c68056a33f01bfc
35140 F20101221_AAAUEZ harpley_c_Page_020.QC.jpg
85868c4b2e647da74fca0450e2c69353
e347d5c2c9fa0553ebbe9e1cf7d91c39a6c5eb5b
54766 F20101221_AAATZT harpley_c_Page_148.pro
68f5d05b9eb095a537d0b4a8253cc43e
3dfcbd9211b7665c1e93fc2fb73c9b2a22f65b4a
66713 F20101221_AAASXG harpley_c_Page_102.jpg
c9fb237f9c7bc01cbfebbc21187ab2e6
e66a04139904a4f20933a7710f29a701f8c9a5a0
F20101221_AAATBX harpley_c_Page_034.tif
7db836a81818ba6a45e570dfabfb7daf
79143e706fa80099378c4a1df2635cdfdfa19325
F20101221_AAASWR harpley_c_Page_100.tif
a020817cd20979c01d1d587586913218
698c3e0479d74c0322b8db0110cc3df317013338
36587 F20101221_AAATDA harpley_c_Page_052.QC.jpg
a21eadde5210e396f5ad4c822bc136c9
3db3cfdf9723fbbf6189c3333d90b880d6790106
27099 F20101221_AAATCM harpley_c_Page_001.jpg
438370ae899ed5c0694e5403752254d8
8f967aecd18acbb52b30741b42fc472f25cf5dfc
8960 F20101221_AAAUGD harpley_c_Page_045thm.jpg
cb5f61cb31d5fd61f5007832b56ef699
376dcd7f04c8b2bd342b9446b650a33eaeeb8971
3401 F20101221_AAAUFP harpley_c_Page_032thm.jpg
1b0f3ec9c3f310b57f76ad1337de6335
0c4c247b1868badf398173ac156b75c291d90f8e
55832 F20101221_AAATZU harpley_c_Page_151.pro
0ce7295c7c41436df4631bfa9a1aed04
737897ad634fb387006dc6c1e6405ab5e59045e8
54398 F20101221_AAASXH harpley_c_Page_113.jpg
bdc59e7322385c031005b75d5e165fa9
60de62201db48a068bda5e990edafb7fdab9f16e
F20101221_AAATBY harpley_c_Page_089.tif
e77c03a4f0b911c230a93dbbfc83ed8e
601c815a4a96b53643e7c65bc430a849bdf1d44e
113127 F20101221_AAASWS harpley_c_Page_084.jpg
dc5c411579c93722ca0df91ffe1ed4cd
22f429f228b37bd75f687de30d5c9c3f58d7ed96
113806 F20101221_AAATDB harpley_c_Page_124.jpg
71fde4d6bacf7e65d71be2bc90c2b6b5
705ca995199a5d546fe6221784ec5b995928222f
2019 F20101221_AAATCN harpley_c_Page_053.txt
668d4f49135c3b08822dc1699332c36b
94b8b1d01b87cfd9f8af9a6f2dff121893b775a2
38371 F20101221_AAAUGE harpley_c_Page_046.QC.jpg
56dbc521dc8225a9c9cf9cbf09cf26d3
8babafcd77af89e3fe2355fded5454c90f69834e
11835 F20101221_AAAUFQ harpley_c_Page_033.QC.jpg
155a855b8e4993e7e31385e53411b1d2
941f68882cef40cd2714152197fdea629750952b
32249 F20101221_AAATZV harpley_c_Page_152.pro
88aa72ef4dc36c05f4bbf1ad4f4d8d50
2a6b2474a2f0b396453b3dac490d7e87fa9d6b58
8936 F20101221_AAATBZ harpley_c_Page_155thm.jpg
f690b59bea23d78449033331d2bbe34c
0f547f9aadc60b923d817b8376d32f66b358520e
2241 F20101221_AAASWT harpley_c_Page_149.txt
bcd7ad791667e989e9dc38161aad8379
0de50016d04d8c3119091a1f94c85ee011d13b61
2033 F20101221_AAATDC harpley_c_Page_020.txt
e53d5a6ad3d8ea5ae252fae88f88b5d5
3d36cc4133d37ede461630cc1633750133fe9938
56399 F20101221_AAATCO harpley_c_Page_142.pro
e5c6a5e2a40bb88609a3b67a2ec55ea3
d52522143121b4cb5a42f8d44537413061fc5776
110833 F20101221_AAASXI harpley_c_Page_123.jpg
0704d9e32227663cc1b78e335d366427
3e4b8ad5e8142a15cbcabbc4b84a9c14fbe525a7
9129 F20101221_AAAUGF harpley_c_Page_046thm.jpg
5c3dea0a5183a38f23ba17bcb4a28b56
b8e7620c69db01f2388f8993dced637d497b1f1c
35348 F20101221_AAAUFR harpley_c_Page_035.QC.jpg
7addf4dee007fe4690ab1029a3b957e1
10dc4ab83f7f1a9ff1dc7a06141bdc297dd25344
62472 F20101221_AAATZW harpley_c_Page_154.pro
31798dde843901fbfb1a027c11dda405
171b158e972791c8424672dd4f5c3646dfb94419
F20101221_AAASWU harpley_c_Page_156.tif
63ed293f6c1d4e090c62adb79fcb5738
3256fd54f64ea53ce95446d889809a95e859723d
F20101221_AAATDD harpley_c_Page_010thm.jpg
ad8b128a22ded88999d514c7f91f60a1
4adfae5fa21dff0c178281a650649eb0f8821aa8
114237 F20101221_AAATCP harpley_c_Page_144.jpg
50ddfab5169eb5e167b754180c5ec35c
c0e03e7a10d038c42597fe9437b479114cc671c3
35577 F20101221_AAAUGG harpley_c_Page_049.QC.jpg
814c1d9209cec5f2b778f4cca15f0586
d32b83e20f58abf832fc538fb96762da0f2ecb18
8738 F20101221_AAAUFS harpley_c_Page_035thm.jpg
4b7d298286c18c9440469c369a640740
a2759bd9791de2b3f48767e769fb150633bd72a5
69124 F20101221_AAATZX harpley_c_Page_155.pro
d15839626831ab64f0d7db22a6e94632
7b4a7e6064c57ce6530c16a4a1162fff028fb7ff
F20101221_AAASWV harpley_c_Page_060.tif
f15890c030d9a22f9222e1b7ded93d58
052b82685025a57b98c4c6aec105255dd4a7b3cf
6305 F20101221_AAATDE harpley_c_Page_031thm.jpg
93b2fdb1e9e0dc5ad4526d3ce7cc3c7a
55d9d2a7837e5bd840d2485fe07a22436219d73d
2123 F20101221_AAATCQ harpley_c_Page_017.txt
8cdf7c96fd903dfd1f61a64ec684d7c8
f738922e40ec0c5c752169cfccdb1c4120db08e8
F20101221_AAASXJ harpley_c_Page_073.tif
ddbe419e871afde057c0d6a7ea0f273f
00daced74ef3fb8d1c574abeecc04547ab2675a7
8551 F20101221_AAAUGH harpley_c_Page_049thm.jpg
782cfc63b618e99493242092bf1196b4
bd54b56bf1e66ce6df0136fad95e6b95ad979b70
9070 F20101221_AAAUFT harpley_c_Page_036thm.jpg
d85abdf2342230d07053cf12fdd57cf0
afe486281f232c5ae2c9d7496b1d4d0d77e51b13
68499 F20101221_AAATZY harpley_c_Page_156.pro
1b82cb4fbf119d9cee87874782ea564a
6d56f8df5583faddc082d65901a5fbd9e8fd612e
57483 F20101221_AAASWW harpley_c_Page_073.pro
d1778a0416d45538b04cc7a85d1ca74b
a8740864037de72a856a6992dd65137c816c5388
F20101221_AAATDF harpley_c_Page_087.tif
9e68e365fb7ba9981bf13143d5787d6b
371bae0a07b7c454222f61ba1d8fa0c20e44536a
F20101221_AAATCR harpley_c_Page_062.jp2
422fb17630e5ab6226861e9289799b62
914f48f7ccb1f258e7eb2e942015987fcde5ed7d
54310 F20101221_AAASXK harpley_c_Page_019.pro
1b77e579939ef10173e3ddd6934c1570
e7635d9c3b086d65ea8c35910c0ea426b0074849
8785 F20101221_AAAUGI harpley_c_Page_050thm.jpg
c5e642ada33492caf1ef29697aaf3172
360446a005a45180a5461eb8e3c38d5e56ec08f4
37844 F20101221_AAAUFU harpley_c_Page_037.QC.jpg
a79435ca76c2875b5d5cc837ceb791c5
924b806f293e31dd0ccc85ab69a62495a62b5e14
70956 F20101221_AAATZZ harpley_c_Page_157.pro
b78fb386abbe52ec072b93993f1318e7
b6eb4a9c7cef1d8bc9590809b6e873ba0cf273ec
53273 F20101221_AAASWX harpley_c_Page_120.pro
b05bb80e2111390a3e13e0592aec4b41
bc36e5452a728fba44e954ed4f7290b1978c696c
F20101221_AAATDG harpley_c_Page_041.tif
d4400aa97fb3c94bec3a15678ecd2bc5
70ffadaba06ab18298bcf9b2a29e7bc55dfed4d4
F20101221_AAASYA harpley_c_Page_129.txt
97eb3db115aa87a80f5837d15ad8cb98
192695caff15198c531814c6b281e1b6c614a3ca
9178 F20101221_AAATCS harpley_c_Page_025thm.jpg
138299505844559060a1472e5aad8912
c4dd43709429b38b2a03a4081fc691457c6f32b8
111107 F20101221_AAASXL harpley_c_Page_127.jpg
49f127aab60665cf34ac4ac8a3a004dd
7fdeb5af84109a490faed26f557a96d9ed02b59f
8726 F20101221_AAAUGJ harpley_c_Page_053thm.jpg
4a682518c07ba901bbfd7b8d8f42ce51
3bf065715d22150f0790677dddf00ebc20462ed4
F20101221_AAASWY harpley_c_Page_135.jp2
64576031b95648c150f9ee5f78fcfa70
01a48bfcd8a969ad19a2d43ea74fa916b92e1f7c
F20101221_AAATDH harpley_c_Page_049.tif
43eb846d847f831100ecab4dd33bcff8
f3c88c327d745ae4632cd619e7c85b89343e3d48
F20101221_AAASYB harpley_c_Page_030.tif
5b4547263e99291a1113ef1244740b54
79ab7ebc36e17473fd0e649f3a859fd7ac6cecbc
7236 F20101221_AAASXM harpley_c_Page_097thm.jpg
782122f012c8d9e506042d9aa8b80d02
ecba2f2f4fc4e18c5eca96c621a7872d0f26ca1e
25397 F20101221_AAAUGK harpley_c_Page_054.QC.jpg
06b68c02d77c39722bd7c7ec1763a093
1ae60476124c4f7b0c289cff3ede97b1483e0b0f
9193 F20101221_AAAUFV harpley_c_Page_037thm.jpg
f4ae473bcb4ee90ffeaf05144ebe3040
f21bbda1758fbb1784c8f6ce18c609ff06c83ce4
979 F20101221_AAASWZ harpley_c_Page_107.txt
d67f0984b6df07ac798c37ad76b80c98
50fa562141674e4c098d0332fb970c30cbbd8aff
197 F20101221_AAATDI harpley_c_Page_003.txt
9fba1d14486bf07e360294e2be0f8873
2fdda63aa01c2892a6c8377f867e8a3561400970
34986 F20101221_AAASYC harpley_c_Page_091.QC.jpg
1b1c9e35e8b90cac6a76f470687fa9f5
f4f30e0139cb0ae4e97fef91fc3dfa4d062d3594
2305 F20101221_AAATCT harpley_c_Page_057.txt
fcfb495cf8814a3560a259ec16674f61
cb63e334b3654ff039e53246ea97f9576619c7c9
F20101221_AAASXN harpley_c_Page_036.tif
5ccb7bc8a654aff3eedd68a6a5bfe85b
42053759e4370965cada276bc8aec09061e2bf4f
6296 F20101221_AAAUGL harpley_c_Page_054thm.jpg
e9ed2acced1ca55d8cc207450df0b8c9
f8473132e98a5f74a1a95c11528590e3c010ad45
38413 F20101221_AAAUFW harpley_c_Page_038.QC.jpg
6ad020486b19364ef6807db14132c7b7
afd71b0fd7ba4137d796d7858786befcac81a5b9
36723 F20101221_AAATDJ harpley_c_Page_045.QC.jpg
1a575922b81d325db6ca2e0110b743f8
85a671d4a6c1694f952f5e7ded1445ab0972f0aa
F20101221_AAASYD harpley_c_Page_050.jp2
eb43e9aa3829437b61799ccc1adacb82
5ab65a38739d48737415a842b6c9a40580506a82
107897 F20101221_AAATCU harpley_c_Page_078.jpg
ac758aa3fed02d7773714ba4e3a2a957
cc345fde267c48a6fd9ddbcf59b44f62193c448a
106770 F20101221_AAASXO harpley_c_Page_057.jpg
5996a193cc17c55f9b6ee476b8e3def1
73e34384317b4ab856b0cbe4f22c8b3401e58ee9
35831 F20101221_AAAUHA harpley_c_Page_070.QC.jpg
4d54edda5e900a0c0df734d7a7f06f33
dfca956f516c21af2f1e6715f90af1f961f8d154
25043 F20101221_AAAUGM harpley_c_Page_055.QC.jpg
64addcc0aba45bd32ce36a14c9173d24
ef9da473a647b66b0690305f05fd7c93f6a1b4e6
38283 F20101221_AAAUFX harpley_c_Page_040.QC.jpg
776cf4237b2929583349839ab0805891
c75c3fb3e01ffb04e554017e7f36c4c5f3bb6049
53503 F20101221_AAATDK harpley_c_Page_084.pro
29a973071356b0b95acdf28ddb0bbe0e
5c2a40115ac952919ae72be7629088bea46c27ec
2593 F20101221_AAASYE harpley_c_Page_003.QC.jpg
4568ed19581d94f485cf98918d974818
0d99d040175fb35c44b9b5fe9776feea0ed9ae7e
94462 F20101221_AAATCV harpley_c_Page_097.jpg
2a4feaf1b0ef8adef4a0f49b767a6216
00dd2c23a11db6e9c0ac0b03d843fcca2c226cae
F20101221_AAASXP harpley_c_Page_062.tif
09331bd34dcd13b31c8db9b71427f221
ea5129384fdaa4bbdc344657e730725e4d3e73b0
8905 F20101221_AAAUHB harpley_c_Page_070thm.jpg
ee8daea35838e3fe236fa6fd37a55297
b9e3eac8c84e09e45a8030857fe33661eb3705c1
33628 F20101221_AAAUGN harpley_c_Page_057.QC.jpg
04f929bc44d368f28ceee93fce57bb15
81edf4426b281a3ecf774f132554e403bf46e5f5
9402 F20101221_AAAUFY harpley_c_Page_040thm.jpg
b2e0a092aa37957af4e4563d97166c7a
f968589692222ea24dd746b3074ccb0a7af09675
2097 F20101221_AAATDL harpley_c_Page_070.txt
971ece093160b8d242c5bdf1ba32cb5f
7a2b95c79e00876c245b3cf9f6bcfaea71bee796
8652 F20101221_AAASYF harpley_c_Page_145thm.jpg
24445bc259ad7d7a376b601d3be7ba39
45db4c64fa7204d6001081c02c0c56aa7a301e72
2114 F20101221_AAATCW harpley_c_Page_047.txt
8e574c3dcf213319c681e000638f9a67
81cfba1381b7a3faf8210f0eaa7bea08187eeb96
52078 F20101221_AAASXQ harpley_c_Page_051.pro
9b5e3af31c4e9d700c742371811bb1c5
e76740ae534c41cf25b50408c20d399c7e3bd89d
35710 F20101221_AAAUHC harpley_c_Page_072.QC.jpg
12964399ca46bb1c7a8af876215e9645
f049ad205f33a75c3f421bd55cc3e73e62b713b7
9229 F20101221_AAAUGO harpley_c_Page_058thm.jpg
e2452c90130bea2bd9a2f4fd14695e35
95b4f1e9dc3d2cb473dbbd271dd711e479887a2e
8939 F20101221_AAAUFZ harpley_c_Page_041thm.jpg
917e5de4a279717f7b903be768d0f7c1
ad280b32ba22451549123a7e212a2c690b5df8a9
F20101221_AAATEA harpley_c_Page_045.jp2
ebb5a5e9cd0024c58af9bc24dcad850a
0787671d017d4ff436d9c04e039cfa704a1497b5
F20101221_AAATDM harpley_c_Page_129.tif
d60f03bd840a2b871bedcddad9328221
2d3ef596ce25325e1daa481a60353020b24feeff
F20101221_AAASYG harpley_c_Page_036.jp2
d0403f1a93abf411d6f252fdf651ac85
9ae0ef9c0cf8d4762e69a3a2d344957c5e630bf7
2126 F20101221_AAATCX harpley_c_Page_120.txt
b665367bc15b2438e57816dcc361a47a
057f17788d7aeac9b0d2efba95e40315faf99670
F20101221_AAASXR harpley_c_Page_074.tif
e59aa3faa8e1a5373b7dc67002eea026
0ef39b817b8b21952ba275292781189de9265d9b
9147 F20101221_AAAUHD harpley_c_Page_072thm.jpg
6658fb72c59f0b82d187f3ee21d39eb6
14c691cc4a3f1f5dd3b328652bcf78bb7fd736d3
8867 F20101221_AAAUGP harpley_c_Page_059thm.jpg
1232531830904125454c2ebd3d67eb3e
6a554d0f405191ea6717efdbf66dbc47e96a5ca1
9187 F20101221_AAATEB harpley_c_Page_157thm.jpg
ad0b06aa5a6b6cbd6bb98e2784bfeeaf
803a440da30e04569623f7593321dcf92bd3b0a2
102637 F20101221_AAATDN harpley_c_Page_076.jpg
98197dd3d7c4a5102b8c8d8ebf3baad5
ec19248025877b7805faf731af6079385c6e1109
33428 F20101221_AAASYH harpley_c_Page_082.QC.jpg
e1f86e032bba9c11a258c4b368cc2f77
a6af87d417c140f73da71327399acb05d760c4c6
34943 F20101221_AAATCY harpley_c_Page_078.QC.jpg
f0cf2254f18738f89bd865ed1d7a6c2e
b392ac3701e35c302af36f5802d6a5184fac0e66
8801 F20101221_AAASXS harpley_c_Page_083thm.jpg
123105bc6eb479df548819b330d00d49
1d18355127d521a50187079ff5af1120b7a4ff93
39236 F20101221_AAAUHE harpley_c_Page_073.QC.jpg
ef84e7c4f4f201f619bfc687c354b430
3507d3491fa6229e60c06b564c1823cc301a40a9
35909 F20101221_AAAUGQ harpley_c_Page_061.QC.jpg
1d2de8320bdac6b83e18f62c61e125fc
fde7047564ea77a228d2567c8b090fc96a336865
1051923 F20101221_AAATEC harpley_c_Page_087.jp2
496b4a72b0b139bf7f56b0ef2a19158c
cabfa672644813d8ba893d97fc3050b011219c86
54250 F20101221_AAATDO harpley_c_Page_112.jpg
47ee376dec91661ac8c25e0db0b07822
5aaf3923f84c87e2065f820dc52e7d78b121dd21
9118 F20101221_AAASYI harpley_c_Page_151thm.jpg
298b9a74da7b2db7dbd7ec4cd0f45d4a
30d1b1a0bfe9688ff1362a95e3f8fc1eaf61be72
55223 F20101221_AAATCZ harpley_c_Page_075.pro
e3847f44ec564bae75a3bbdeb240c1fc
ad7bacbd234a1eb51931593cbe5f49b8ffc63a29
8533 F20101221_AAASXT harpley_c_Page_001.pro
7b4719858e48a04bcc903de28c8ba38f
ababc1bd5ce781b9030bf0c5d09868febb14107e
9406 F20101221_AAAUHF harpley_c_Page_073thm.jpg
2da5ecb22aca52fa8f3f2ea864f81763
549db51e14dbbef5f06eff381df04135a06e3893
8865 F20101221_AAAUGR harpley_c_Page_061thm.jpg
d0fd5ee68d7aa8c47c816f413c8bd550
d9e8be9bdc54fb0c0ca36f7c9c79e59980d3b0d9
111798 F20101221_AAATED harpley_c_Page_045.jpg
1464567019c2bf9d7fa39798637e670b
c2cbf992cd9bedeb5d9506d1d2b7388f7d9e4009
117991 F20101221_AAATDP harpley_c_Page_146.jpg
71713da5a8e070f8602d5fd9bcda3449
19ea6cedf61cd8bd60b51c0e6946c6e12cb9d112
F20101221_AAASYJ harpley_c_Page_155.jp2
c30b75c0a669f9b3d5cf95fafd98f503
894199c39a894e950a34af03454f530f2f9cd479
4646 F20101221_AAASXU harpley_c_Page_162thm.jpg
33ba3536bdd540d0d31ff1e31eede56f
5536baaa87620f610c02f701c5b32d746cea3073
36416 F20101221_AAAUHG harpley_c_Page_074.QC.jpg
42e9064fac9c3cfe55004523495d87af
703365b135d3d7f3757698d403dcefc5e2245a07
37494 F20101221_AAAUGS harpley_c_Page_063.QC.jpg
491b250f01c6bedee42944b24b1dd1f1
709d54e7350d8f64e2a505b26411112739e7b97f
118624 F20101221_AAATEE harpley_c_Page_073.jpg
ab78a4c1a8dc7ae5b7ea4df276971803
52a73267240330e3b09c57f4d0db059b2ccb7c17
F20101221_AAATDQ harpley_c_Page_142.jp2
25390edfd49fe69605308bef353bdb64
e8ac9e28c636abefeb30a9b5423f482227ba0658
101619 F20101221_AAASXV harpley_c_Page_098.jpg
a9bcd7f729def8b7bbe379fc5ff59dcd
97f01d05e53a286023e0783f973f05e2a0783e9d
38055 F20101221_AAAUHH harpley_c_Page_075.QC.jpg
d6b30a7318c82c4f37441533bb9da022
568f2229cca804207f6e0f9d77cd64a01d9cb38f
34391 F20101221_AAAUGT harpley_c_Page_064.QC.jpg
edb3e2044f81ff7afa33d2025ed26935
8ca6ef3bd5bca5ec6c8afbae7528e5e985a727f4
2034 F20101221_AAATDR harpley_c_Page_091.txt
1234a1aea109d4e916d31511e6e15c49
e89b0e4967e488acd8a2a6d3865e37da1e8264a8
35452 F20101221_AAASYK harpley_c_Page_060.QC.jpg
c50f697e2ccf72b5f414b6f7a1132dcd
38363f1ca29821afff931bdeb72b562e807349ad
F20101221_AAASXW harpley_c_Page_027.txt
a2ac99f8804a4002f086a2eaff496aa3
1551cda304c214b8be62b9635f31b84132b73a47
9085 F20101221_AAATEF harpley_c_Page_052thm.jpg
585891f1c13e5f6fecc5686ec1849b33
9752f2f66040ef517fb433b1ba35ce9e81ae1be0
F20101221_AAAUHI harpley_c_Page_075thm.jpg
11cb28993b06f9a6f3cd53d5bb90525b
3350489d91502d53fdb3d8a5c347175d452f8675
8616 F20101221_AAAUGU harpley_c_Page_064thm.jpg
d9b6044dba67816f9d3c18ee56fa43d2
1640fb43839d448b3bdd2a7030e50df7be1a5717
F20101221_AAASZA harpley_c_Page_039.tif
83787888e5eac0a835a987f09f1fe46a
b57e1864f8c02e49488f9f68f2fe01fccc7d1408
104484 F20101221_AAATDS harpley_c_Page_131.jpg
943fe8397587a9068e33289926f4621d
aa2a534031a5cd93acd6c4b16462673992359b2e
F20101221_AAASYL harpley_c_Page_099.tif
2e12f1dfbc0181a4637ebd0e309a30cc
b8d4f4964106d70028e748b5adea7783b07b1202
F20101221_AAASXX harpley_c_Page_084.txt
01338c7cf068f211ab8204e50cf3f8a6
1806092fde3332e70b7544f6b290f09ab1b3b020
42243 F20101221_AAATEG harpley_c_Page_101.jpg
09e9deacf24d1d16587ca154aac4a27d
41075b0664776ac0c9b686fe99ef96f14c0b6fb7
36464 F20101221_AAAUHJ harpley_c_Page_077.QC.jpg
807f4d2cad95f7c37d047f297673fc3a
dcbd5b95e6d7bf30ffd44f93314a86bace895f9b
36950 F20101221_AAAUGV harpley_c_Page_065.QC.jpg
36d50b7f5dbc1089983b25657ffc484a
3666d3d20210cd5236bed0a756c000af58fb0a0d
F20101221_AAASZB harpley_c_Page_153.tif
f33e9e5ac5be084745cd32265073a10c
cc335408ea08bfb6cccc922d7b3e9e81e6876caf
8898 F20101221_AAATDT harpley_c_Page_084thm.jpg
321ff5f65733451e45c6d213c6572727
71a61e7d14a2d4177c615351a0e954e2b62cc26a
54858 F20101221_AAASYM harpley_c_Page_039.pro
9893e8a32d29e3bb0392f7e5576e8c00
e6f6ddda31df12a1ba6b0be0d9d05f374f08a390
106765 F20101221_AAASXY harpley_c_Page_053.jpg
dfeb52e048d8df1a4c647da648ea9f6c
2ce16b7073235ed5263dad32bd7fb4c8c8d98597
48583 F20101221_AAATEH harpley_c_Page_108.jpg
13acad38189fa3fde7bfa55b0e5cf88c
4f9efd1184a86c3716753067ce8509364155bb6d
9218 F20101221_AAAUHK harpley_c_Page_079thm.jpg
5b02d77c63270caa1330978e22269991
b1d1a345f566ddd563d5b8bd9d5ee56f02d30f44
F20101221_AAASZC harpley_c_Page_088thm.jpg
d328e7150e97ad0c73437ddfc2dfc23b
8c54f6ca0774894e4c80372a034ef5530529c851
F20101221_AAASYN harpley_c_Page_128thm.jpg
05c227a2504e2aa1f2e9bf9545de70eb
9261076825d755401b9dba5aef3dcaf0203cc9db
114489 F20101221_AAASXZ harpley_c_Page_143.jpg
398b08b974be00960b742f2e3aee584d
cb26f05558746434eebefe5d39da913d7a4ee316
51797 F20101221_AAATEI harpley_c_Page_145.pro
d4832ef0a018c2d8a943174585b20e90
017a87f34d775d0218b3bddf54e86f280dff9b2c
38108 F20101221_AAAUHL harpley_c_Page_080.QC.jpg
e3ed10476508d7d8ff8d84e559e220e7
54c4afdb1d6e9de126900da0ddb985cba076857a
37648 F20101221_AAAUGW harpley_c_Page_067.QC.jpg
969366d4fcb07697551c3c2aca731cb5
12dedac52b93532c6c882e7df15bda1d026ee912
114783 F20101221_AAASZD harpley_c_Page_086.jpg
431bf3c1851aa5ce34850377f8c4e984
24b60d59aed8f1aa3f2145f54793841b3d3ca25c
6681 F20101221_AAATDU harpley_c_Page_055thm.jpg
e9998d3179acaab6d92c15de903dc93a
b9a2779fe6ec19b161c5a081342a7a1ed6e81639
F20101221_AAASYO harpley_c_Page_057.tif
89b0e03b72671b36ee5d48357e712bcc
4b85acc1f6f549f8acc00c460f802c8eafc576ec
44329 F20101221_AAATEJ harpley_c_Page_044.pro
0e99b51522f3485e16e9c469de9c8fb8
4f7d39fb6af06f4c2e661362011acbb990360829
37793 F20101221_AAAUIA harpley_c_Page_094.QC.jpg
50096169a5a75e808e76f64875fd554d
4f90f98bafd46c480d90e1f95e030b74fe2e937e
8732 F20101221_AAAUHM harpley_c_Page_080thm.jpg
549fb6fce11ee0a1cdfab85c6d99f65d
a2537c12d96001fc2bc8bfa089ca22e5507c7224
37498 F20101221_AAAUGX harpley_c_Page_068.QC.jpg
860e7f35456ae7b910c623ade4c04592
a8c564904ddd20fed140e5984717b8c37794f75c
523247 F20101221_AAASZE harpley_c_Page_106.jp2
e9f22bd9c4d0d67cd3e4ff30528f17b4
3086c6adb6e0ccf857cec88c469d2d521955af15
F20101221_AAATDV harpley_c_Page_157.jp2
8d92f7662125bb8352b71d51c9e0083b
4f15a7f719182db44425b76d9557e44589442589
F20101221_AAASYP harpley_c_Page_047.jp2
9167a0396d56b0b63ea47f4ac14338eb
a4c2b1b2c8ac9bac39f962aa108f867aaec163a7
20706 F20101221_AAATEK harpley_c_Page_112.pro
c4f135a47cb3a485c37f8cf766716c95
8bf5ece5bc67a320bd48f76987ca1a1ddf1b8632
9268 F20101221_AAAUIB harpley_c_Page_094thm.jpg
e6ab636c61580a16e810704d059ec98e
ef2fdc04f2a680733fbd9f2c3bbf36ae5fbb767f
36684 F20101221_AAAUHN harpley_c_Page_081.QC.jpg
8f26800ada300ac468c88f84880d3392
aae25c14b80ba40353fc6d7c7f406e4670f2650e
9310 F20101221_AAAUGY harpley_c_Page_068thm.jpg
ad8e92a9e6f137ded7ef0d23c9be300e
0ad5bf26d58ed1d780111175f54d0360c85ffd8c
9041 F20101221_AAASZF harpley_c_Page_060thm.jpg
aea705563282a6bf262ee7b0d3f877e8
7d728f2e00cd8bb50e4c9c15298057b6e1211e06
9314 F20101221_AAATDW harpley_c_Page_062thm.jpg
1dcf62eb5c01d977797ba64e90e3a478
4c95ce3f77d3821e027b92f2d9f15ac8bb33f956
9236 F20101221_AAASYQ harpley_c_Page_143thm.jpg
f7c55cbb889a3ba432e6b54c23663ea4
b65abcd5a4c02e33dd9101348c59b863e04ccc72
F20101221_AAATEL harpley_c_Page_004.tif
32210da961f034537c75a60628c4a4c5
9a1dc51cc5a6b6ed177a09505219f916b35192ff
36187 F20101221_AAAUIC harpley_c_Page_095.QC.jpg
ee84b8fcaecd0ed994466a841226025a
493d69f41d8ca5cf84647749733707af9e9afb4b
9119 F20101221_AAAUHO harpley_c_Page_081thm.jpg
5fc77ca18353da552ffaf2ad620a3c5e
6cbadf3f3526edf5f5924b0f42f8d2bf7c85597e
8735 F20101221_AAAUGZ harpley_c_Page_069thm.jpg
e1ff1846b046db53fe9c9755227eccd8
face2fdb07feb45dcffaf77e319e97be3c1baec3
1054 F20101221_AAASZG harpley_c_Page_106.txt
a40bc340de298cf0b99640d45771ef2b
aff4f36f8f0f498a07e4426e100d104cfe3d34e7
51430 F20101221_AAATDX harpley_c_Page_078.pro
9a17bd7e8b117a6207723fb36bd95f4b
b40fc414e30b3d8da03176da86dda3dd8e0e6466
56424 F20101221_AAASYR harpley_c_Page_133.pro
e8aca63b625ae5bf3245023116e31881
b3a14d902d8e03a65c959f0af0884f47831e826b
112061 F20101221_AAATFA harpley_c_Page_138.jpg
1d2e7c31677d769272c5324fb59499a3
fe00edbb1a1f846e8bcf9f05ffe943b70ff499ca
F20101221_AAATEM harpley_c_Page_049.txt
f81497e89f5e7855387af89921f6b4df
611693256c8b8edb385ec0fdb929f085ce644c01
8737 F20101221_AAAUID harpley_c_Page_095thm.jpg
b8dfe8eebd4a62be68a8a19d2c719233
31b4cfd81fb2e7ce065198ff16823d2a9add2bab
36505 F20101221_AAAUHP harpley_c_Page_083.QC.jpg
ae8da0e3db6264e1749a2a4e91d6e5f5
334649b6561bdb6b6fcf291e7eb00e6a6a5a5c1b
49175 F20101221_AAASZH harpley_c_Page_076.pro
a326a6526c09a5f2c18ce886b3aff598
411737f0a378ecaadcb64fddec9c4d113160f2a9
9180 F20101221_AAATDY harpley_c_Page_147thm.jpg
5ff4f7a21710c6368152272d48d7f64a
a6126588213d84c6b835902e6c9f7cc7d95a6b32
31625 F20101221_AAASYS harpley_c_Page_044.QC.jpg
d08b53a35bde55b4b95d101619f55bf3
86ba087abefbe4cb17aeaacee2ce9cc2410adcde
113510 F20101221_AAATFB harpley_c_Page_028.jpg
38c477375e9399daaec75d2e197561e5
1d6f734239c42ba5cc567a34abf28969be9bfa22
54544 F20101221_AAATEN harpley_c_Page_042.pro
a7599429495862a730c90f4bc3a82da6
048a174415aabc40dcece533522524b8051925b3
3455 F20101221_AAAUIE harpley_c_Page_096.QC.jpg
4cab60abc911c741fbd3e97987f64045
0263fb9c2642e2b492333e38345629f1fc738594
36645 F20101221_AAAUHQ harpley_c_Page_084.QC.jpg
21b4504eee45dd643c582116b18ac658
f0221bc9811c2c6b978a2f2079acdfa96a671612
37173 F20101221_AAASZI harpley_c_Page_041.QC.jpg
c28a3b60f7b1d66627a1c6563ac626d0
d602216ff2cec3d2fbfc8d4552667edfba8dadbe
F20101221_AAATDZ harpley_c_Page_145.tif
dab0e4ff685a56ff955687e444d0aa7f
d55ee9f953f6932b80f15bbd5ee17a94799700dd
1051900 F20101221_AAASYT harpley_c_Page_160.jp2
b46cfa28ecd50b5abae827a4e0d42886
d09939ab889ac4e8d8a07c84210b390ea8a1083f
53746 F20101221_AAATFC harpley_c_Page_085.pro
fd8023e41b339e6da6fcb5b4edc11507
b107e2f34d1af7a6a58f467d2dad8d407c356371
111562 F20101221_AAATEO harpley_c_Page_132.jpg
c30922bb32f0abb99afd03da56ba4af7
fdec23235b831333f46f503d24191f7e6473e431
972 F20101221_AAAUIF harpley_c_Page_096thm.jpg
b4980a859ebb808b7e2740729b3eb7a4
9bea299dcebd3ca47d3a6a9a4a7265b42ec3e953
37272 F20101221_AAAUHR harpley_c_Page_085.QC.jpg
b79805a914b4dea38f9cef9490225e90
db547167daa7869833811af90f367b7a43764f98
1051974 F20101221_AAASZJ harpley_c_Page_089.jp2
3c7abb580bef47680f16c9371063b186
7c08ad3217790900f9a55a0b413b1ac9d8217704
238114 F20101221_AAASYU harpley_c_Page_004.jp2
0046be8aeacc021eaa367f75d9b4d3a0
f64a0c5bca21c49978998e9608fce196b65c4e29
F20101221_AAATFD harpley_c_Page_061.jp2
c4b723ae092574ae9242107e272476c0
4e790217b0b4afa4e110cf2340192df5fcea1aea
8167 F20101221_AAATEP harpley_c_Page_076thm.jpg
abc84d450d2efcc317490a4231e63a81
396aaf3190fa09c3599c84f2a3d5634a9bfc170f
28793 F20101221_AAAUIG harpley_c_Page_098.QC.jpg
737db0aec3110e6eaeac98285e46fe21
fbbe164b22f663aaa9962a7dcf406984fd9132f3
37375 F20101221_AAAUHS harpley_c_Page_086.QC.jpg
e2cda4c31172f050fbd207430d0fe0a1
b88543fac67fb630256357dc16b98a0df1b4448f
14766 F20101221_AAASZK harpley_c_Page_105.QC.jpg
6d74642133d61ca7fe0890a68b28b282
573c72aea16ccd5db3d9302c78f4456aed6d6732
8526 F20101221_AAASYV harpley_c_Page_091thm.jpg
5c289dd556ba8585642c60eef0add0f7
efc5a261202d41a9c3886d5791af74cdbf54c023
3238 F20101221_AAATFE harpley_c_Page_096.pro
69e1f7207d35795589a040c2cafe8870
ea9b709e55bc83f32b3c937ce11b512f235526db
55720 F20101221_AAATEQ harpley_c_Page_124.pro
884dc1ba56a601eb7b4d7f4ac1f17724
e29a1140fa0044e312bfeb43409d47d19605f7df
6920 F20101221_AAAUIH harpley_c_Page_098thm.jpg
cd72dac813c5cb87b7220bed0b223648
94ec2e6fba442ab91ab911edcb779b07f50e56f6
8986 F20101221_AAAUHT harpley_c_Page_086thm.jpg
c120a92b661f5b868cb16628dc0afbc7
e5a9106dcb7b00b4a3a1ffb3a83d62f543fc4323
2187 F20101221_AAASYW harpley_c_Page_124.txt
0aac88faf385a4f231e531edb03fc211
80fa17a635d0a7f1dc0e1c352fda575dbdd27374
2116 F20101221_AAATFF harpley_c_Page_022.txt
8164ceb69a479efbf27e9967b02de336
30a6e88a9a068acbd24cb3eb719c1b0f96e5a6be
2058 F20101221_AAATER harpley_c_Page_052.txt
86cd9b7c111c61528fd7708aeeb62241
4a458641775b4112f6f7b33911abdd9effca9236
3710 F20101221_AAAUII harpley_c_Page_099thm.jpg
c6aba4167256a4250553da60d8ef1888
a5b239727f7260e2696e0084bc11c8a2c65f78be
8828 F20101221_AAAUHU harpley_c_Page_087thm.jpg
605b0a112002a2ccc90c266b1e87bfcc
f18cd28721007056ebd4aabc896549a47d372b53
F20101221_AAASZL harpley_c_Page_041.jp2
9e53ff41749c4ca7dbebff4365d43165
ceb80a591f8ca5cdfcc5756c27d1d230f5eff8f8
35502 F20101221_AAASYX harpley_c_Page_119.QC.jpg
7c23f231ec896ca4f6cc89b1440f4ee1
ec25ebcfa4eec7978ebe4e9bb3c3a565e50e9928
49662 F20101221_AAATFG harpley_c_Page_043.pro
6a24de9917e342687c8c0443f0261935
e948403e4f969aca355bb7f85e459fd51b34b1d6
F20101221_AAATES harpley_c_Page_009.tif
bbeff12be097fd7cdfc4160a85ec332d
43d1a69e2e33cc98cfff1a360e02f3ae8112914a
15303 F20101221_AAAUIJ harpley_c_Page_100.QC.jpg
2491b39cb5a0e9b26d6de99b97c2ba61
9ded5da108bc70481c9b8320a0a0464249e2ce8f
9262 F20101221_AAAUHV harpley_c_Page_089thm.jpg
6df33922869b1061f99ae374dce2ac2b
dfde0c4cf815514ab51439f2f62b76d4af132124
F20101221_AAASZM harpley_c_Page_122.tif
8f2da0a3f5cce2e91e0db16449fc18f8
24cf3063b44227354b5d6d5d69bd49da4107fa87
34563 F20101221_AAASYY harpley_c_Page_023.QC.jpg
d37af0e521c99cb5eb737e370d6ce52f
3bbd86b171a7ab5247fb490590a6b508e1c2d151
F20101221_AAATFH harpley_c_Page_093.jp2
9c5b091af5fa74272b788210ba4cf20e
f4092644655dcad0fa6a30b92ab7e7ea74392afd
33623 F20101221_AAATET harpley_c_Page_162.pro
ed8df648cbbcdf8052ce077d2d4c42d0
cbef173fbf0d0db9a375b92febef3dfdfb44c7be
5546 F20101221_AAAUIK harpley_c_Page_100thm.jpg
242d5e0131d489c8480b8a375ddada0b
8f8adc63cbd42d79378c3c0ee5334d3780b209e9
36483 F20101221_AAAUHW harpley_c_Page_090.QC.jpg
be97cc85a5226aecc5cdd4297cddbdfb
347db7c543e767c0b12075a2006dc2bc9d6bb88f
F20101221_AAASZN harpley_c_Page_126.tif
a277499b4552ef70d672a598f224085e
192b20584414178f390c44c3af9babfb8f1a1b87
37087 F20101221_AAASYZ harpley_c_Page_156.QC.jpg
d3ba859874077f28aca0fd3c29428de7
9a7bee6793818b626a08502ea683ca50d0c02bd0
30986 F20101221_AAATFI harpley_c_Page_013.QC.jpg
d5b9fc53ec00f190df1d7e10f58447ef
6b8b1a505a39734a8c9ef2ea56c0c2a830a37664
2856 F20101221_AAATEU harpley_c_Page_157.txt
f35ee00026a3fabdc44c578dc07234f6
eeef8ec60eb49dbf3d5c5731ff944a0b7a4d6592
12952 F20101221_AAAUIL harpley_c_Page_101.QC.jpg
bce909dc30d403efcadd2f258f3ba854
e14b4bdd72ca5d7d86ec0bff9d00e127b02d4c52
476 F20101221_AAASZO harpley_c_Page_007.txt
df44485af3a0321372ab10f53a4204bd
d9d7fb313d208329c96056fd878058a165be0566
37070 F20101221_AAATFJ harpley_c_Page_124.QC.jpg
159380898ebf0674444eb20dfa8afc04
65a95ac138ee910d9131b7b842b46b8d7659467d
14377 F20101221_AAAUJA harpley_c_Page_110.QC.jpg
ff8808b70e03fe7c4aa1e7cbc33f8386
d069a5dc85e48625374cd55059377947d6916652
4043 F20101221_AAAUIM harpley_c_Page_101thm.jpg
ffc33ce0acb1e2eda1cb8486230b56fd
8afc7836cfef5d5eb2ac5c9f22def21ba70e5c00
37074 F20101221_AAAUHX harpley_c_Page_092.QC.jpg
84ae0237dafd07359b478f7335386de2
0641a5e4d7a2eefc4f2854718775dabc66675f73
385991 F20101221_AAASZP harpley_c_Page_033.jp2
6a32d01ef290e6ce26fdd4357ba63785
f823cd324c4b4dce472976b40f551c649e4d33c9
105712 F20101221_AAATFK harpley_c_Page_020.jpg
4f0320235192b90da795321e0df5d3ef
badcd97253e5689b0e1474b4543b14e8ae27dd08
4020 F20101221_AAATEV harpley_c_Page_034thm.jpg
4d9b43b500febecc1f8649070efbe864
f1ac16482d47d6f0669fe1376ca75991c50da7dd
4060 F20101221_AAAUJB harpley_c_Page_110thm.jpg
b6448e0af76159a8e9072762573e0791
3f1e8750ac4ec01d6d12303bdc081c0919e0b6ad
5193 F20101221_AAAUIN harpley_c_Page_102thm.jpg
bd592246bef06b101ab9607ee5f53eb3
330ae5b636e9ddb7644a67b0dc80f5484b997cd5
9162 F20101221_AAAUHY harpley_c_Page_092thm.jpg
4da6f306d93bdc5c0123c6bf975a6ac1
36733e79efda77d00d2b55329260b2e7678fc192
9224 F20101221_AAASZQ harpley_c_Page_096.jpg
fdef976d3081dd07343fab06e4050472
ad3a530152b9b5269e7c56c1cfef1f056f993757
55215 F20101221_AAATFL harpley_c_Page_037.pro
4fe1ca386f7fe86873ae387223d191c8
b5c422e9a5d9ae21e41305c187269a75cdfd1d16
113828 F20101221_AAATEW harpley_c_Page_058.jpg
24bc0cb2ffadf2e43a4a1bda2fc4f0da
50807567fe2bf3d3ba46c72a253106210c6e9bf1
5202 F20101221_AAAUJC harpley_c_Page_111thm.jpg
aff5234e3059eb810c38b6342f5f429a
43eca95afb94da04ac87de15690f3161b5c60d5e
21169 F20101221_AAAUIO harpley_c_Page_103.QC.jpg
8b6a8f073d5d2f9be915e4c53297048d
0b0c54e42292eaa36bcc3366bf57b2f445c4d7e2
9112 F20101221_AAAUHZ harpley_c_Page_093thm.jpg
b79b12e21ce7feaf9f6ec6b98c2c8112
45378d2a7a042e31fa9e1089dcb5c9d601d81d45
1958 F20101221_AAASZR harpley_c_Page_043.txt
5191ab263d4cd43b517527c253f27fed
b7c82f535e0ea084449f8e386fa22154640d1fc7
2159 F20101221_AAATGA harpley_c_Page_075.txt
d863f4aa8c423fdbf0d8dd91e4152f3f
e8215e6beff7c634af4a4a83cb880d67e0072c27
6904 F20101221_AAATFM harpley_c_Page_007.QC.jpg
0a699cb0e13934329b3b45e5028e420b
f1ad3785387a0bbf202cc6c5f3aabe9699d1dc4d
116645 F20101221_AAATEX harpley_c_Page_149.jpg
c8cc3d728608b507015af38f23c6ccac
21429bc0eaf503f24ca85bea04ac17f8c08244d9
16760 F20101221_AAAUJD harpley_c_Page_112.QC.jpg
c2fb6341b85587c2c987c9a9bb099939
c860689b65ef4912aff7af49a9940d713e27b26e
5540 F20101221_AAAUIP harpley_c_Page_103thm.jpg
ededba77d703e55ef704edddcb207a78
5a2096dc066bb5018838295e29eb92ead5451989
57272 F20101221_AAASZS harpley_c_Page_146.pro
0de8988b5004e157b3ee96c9916f24ba
c599a3114e2bfb3c349deeb34e55d9c36367dae2
18743 F20101221_AAATGB harpley_c_Page_111.QC.jpg
5e05754a26344085d5c645a7b293ef91
82ca389e62e2ef05e91e81da77a2549c2048cc69
95039 F20101221_AAATFN harpley_c_Page_044.jpg
38ca498cd27e1e306a15e73953072e7a
cbe0c7fe244b7727f280eed7f9b29fc6916fd6bf
36616 F20101221_AAATEY harpley_c_Page_161.QC.jpg
86ba5ff9bd92bae13059258f8bbf3ca5
37cb520f5502441563ddda14623faf999ce715c0
16221 F20101221_AAAUJE harpley_c_Page_113.QC.jpg
f42fe4dbac39fb24280668e24e167a1c
f8487d2e7c2354813fc1800f16885207518d839a
7145 F20101221_AAAUIQ harpley_c_Page_104.QC.jpg
6415b4f18d43cd18454de9fb050b256c
aa30196b5d1f2b2c27ab3def19b96b8b0cee1248
3645 F20101221_AAASZT harpley_c_Page_033thm.jpg
2170206d3eb7fb80dec563387a61fc71
ba44b3f1e5ad8e8aff0972667fa6122f17006406
37051 F20101221_AAATGC harpley_c_Page_009.QC.jpg
ec6e755a10e2377a07442e255212462b
9139ce4a53c73206c135793320704302d7c39774
F20101221_AAATFO harpley_c_Page_067.tif
b7f5c14ce7dfbdc988e56b866aebbd17
4642b28a0b9e9c60086d202c83fe301239482595
8641 F20101221_AAATEZ harpley_c_Page_022thm.jpg
3c30de988bdbec00459f9fd4a9aa006a
82639032afde3b2448cd566cb652dc679a65f22c
4623 F20101221_AAAUJF harpley_c_Page_113thm.jpg
9e79e1d89448d6eeca2926a49ef05347
dcdbba191e577106b02c32662774d2a3cdb7f223
2272 F20101221_AAAUIR harpley_c_Page_104thm.jpg
b072ee0a1312d7717a3862c6763ddd63
1477e398a37ea5c7437d2528d39e7d78f290a4ad
F20101221_AAASZU harpley_c_Page_057.jp2
1343e92c57f1916f8526befc44e0942f
050e97266a4a27c93f9d0a24ee1dc846ebda6ddb
9318 F20101221_AAATGD harpley_c_Page_071thm.jpg
f78d4506f211fba3fa73a1a53df82cb3
22c326ca52617e0866519475ae44c16ff776c114
2145 F20101221_AAATFP harpley_c_Page_039.txt
01fed7ae6fd69be55faa7e26a20705b0
d0bbfa0a8bf5435f59867a9063e2b759ae0e6a8f
15797 F20101221_AAAUJG harpley_c_Page_115.QC.jpg
ca58c834a747c422721698d2ae2009e8
79a478db2bd0fdc43c7677acef8de6772ac94b5b
4826 F20101221_AAAUIS harpley_c_Page_105thm.jpg
06c9b753bf00329c2aadebd9e8c4b455
9ad113eb5b2e4b4d11aa6f065767d766b6f5697b
8819 F20101221_AAASZV harpley_c_Page_077thm.jpg
b0f769459cd155107c9483c368072583
f9f51928b22b641b1cf20a5f823019c77c25def7
F20101221_AAATGE harpley_c_Page_009.jp2
db32b667da9b983155432f3f41c02738
00f33fa40044318048d347f88b441eb7c5c76bb4
F20101221_AAATFQ harpley_c_Page_147.tif
30a406ccf02951b3dcaacabea65d2fa6
9f14b8b50b3d96f7963afe9cb263e4f5338f05e7
4833 F20101221_AAAUJH harpley_c_Page_115thm.jpg
316b52626f4fe7124f5524b0574581e7
e7671f5adef7acf46bd761fbf9d0fe7d13eee8f6
16753 F20101221_AAAUIT harpley_c_Page_106.QC.jpg
0d7b26035f074076e4d67414f722ee99
f27f8dd2a1ec30da7484b6ecf5b5ec11c3e237b3
36372 F20101221_AAASZW harpley_c_Page_051.QC.jpg
ca5a10bf6ad029fc861081527fea5350
46cc8938d90a031fc6e1f895eb8c5c3308c657d8
F20101221_AAATGF harpley_c_Page_065.txt
0d57f89d8aa1f25ea4a92b5904b92c77
04d277bdf6985a177d45e0c4250b1d85cec17ff6
483344 F20101221_AAATFR harpley_c_Page_115.jp2
e5a14f7fbc3c2d22da9630f5cfcb3370
fd105c7b997baa82105009349c3b33466cf02424
20193 F20101221_AAAUJI harpley_c_Page_116.QC.jpg
521c43db078b8bcd1804ef55e4a72ac7
98ea4d6066541a6f78d623b9d979098037b4865f
4459 F20101221_AAAUIU harpley_c_Page_106thm.jpg
75b2df1713777fe29460feccb79d5d22
3f677b09839edd2a28b879302a7f65273e78d4e8
1051893 F20101221_AAASZX harpley_c_Page_025.jp2
1ad1bf997073c2ab2f6c73820a822187
0c2cf4befe100bb899ea99a38fed6e84d45b0d60
F20101221_AAATGG harpley_c_Page_095.txt
feaf9eec6f78815426b91cab6058db14
b5794332fd748f9c5767d487e206ca60ae3aa3ba
548647 F20101221_AAATFS harpley_c_Page_113.jp2
392841ebcf7c9e52c5d1fd6104d95d3b
fdc86bbe8cf87d01c5305740060c000054fa5261
5418 F20101221_AAAUJJ harpley_c_Page_116thm.jpg
f9b9f11a3259d5b35b302636d3754552
d5dbc09180ad30f5400555c8d47fe41c17343910
20183 F20101221_AAAUIV harpley_c_Page_107.QC.jpg
ee5ea760744f939db3030d2e3519a867
a47559629a23deacc2f36f5ef93e79d27122bf71
F20101221_AAASZY harpley_c_Page_124.tif
227ac4fdde1f5705c9d226c671b3b4ce
5a8181f47d76c3d7f6bc670df8bf0fe78ee322b4
1051896 F20101221_AAATGH harpley_c_Page_019.jp2
f97bc421ef5879b2d0c1447b1456aedb
83be685196bf34b7808a9f96ba58f6990d882aef
F20101221_AAATFT harpley_c_Page_067.jp2
113c8834b8fa1245291fd646f6e51ea5
11bb6980feb4fd15c2c1badf7b26eb9ef710d839
7185 F20101221_AAAUJK harpley_c_Page_117thm.jpg
d9101cea1ec8a4716d50dfccb41c175d
13c88b3d40e2df588c33d0b688a42e60d881ef37
5742 F20101221_AAAUIW harpley_c_Page_107thm.jpg
360f0fa0e970db1918848c2a0920fc5f
3a1afa75a6139b1512503fa2a304ef18d7c3d1d2
796304 F20101221_AAASZZ harpley_c_Page_114.jp2
277164365d613f432ed79e451bf25f06
976f0d321a2f1096d9efbfc46afc9d1672f913bd
F20101221_AAATGI harpley_c_Page_017.tif
b76acc5e8ec561351cb1128d42f5f80e
3e1144e94fa9589007ba6d492800345f7d7114ef
53446 F20101221_AAATFU harpley_c_Page_132.pro
92069df48528c53657f3cfe97c9fb5a8
5e26880a38fbddad7b695cd6039b51608da36e2f
12288 F20101221_AAAUJL harpley_c_Page_118.QC.jpg
06d43714a793e85de511af8db6ac68e0
5a59c4bbcd9662b919427fe7ada510c3ad53fa1f
15465 F20101221_AAAUIX harpley_c_Page_108.QC.jpg
ec57f54d57976ca34615d7dfc8eac3c1
599ee72416e871120685965350477c83f08bad92
F20101221_AAATGJ harpley_c_Page_156.jp2
b85d339589d7e0856a0e14955acda25b
4e0063357fc4adab6e86476abd5070378b870bde
20894 F20101221_AAATFV harpley_c_Page_106.pro
f122620f1559d89b41c99338f5aca366
f2b1baca121ca21a55077152a2bb05b75694f79f
36563 F20101221_AAAUKA harpley_c_Page_127.QC.jpg
7b67bf13b783ad5942d20c90a7263f0e
d732a9f714475df8c503cf24e0fcba47fc216d4c
3998 F20101221_AAAUJM harpley_c_Page_118thm.jpg
d92499b346cf2d6b4aa3082cf93129cf
36715384c794bfe4fed521ab3691a23110bd6dcd
F20101221_AAATGK harpley_c_Page_148.tif
f1c385be0301ccf95eb5fb38349a6ef6
b3012c99609506357e1ed0e7e7525a800355d42e
36229 F20101221_AAAUKB harpley_c_Page_129.QC.jpg
e475a8fe7379ec533ec6ada94ef72c84
b86227febf83d5fb5334201313473213e1dfd6da
8746 F20101221_AAAUJN harpley_c_Page_119thm.jpg
878f76ae292e88f9b82ade43ea1566c5
6d1ed26b6675c0870a2b9fa7a4906549127dd53f
4362 F20101221_AAAUIY harpley_c_Page_108thm.jpg
4af1756bba2f534f193c60cd439b3240
c6da377733408c5faaa5aa9572edf135cf8598ec
F20101221_AAATGL harpley_c_Page_085.tif
4fd93955f47abd02992261c1f64e3c46
7c944b6a56c1c34875d9c486fcda95397cbd62fd
1051866 F20101221_AAATFW harpley_c_Page_127.jp2
45be74968457652a4e520d5a415be111
a5e331885de014ecb71d8b8155a4eb98f40553a2
8826 F20101221_AAAUKC harpley_c_Page_129thm.jpg
af4e462c2d8e8e43bc0de4291ee14e01
27998b659a2d0ed5cca2b9811161afe412cc1bcc
8911 F20101221_AAAUJO harpley_c_Page_120thm.jpg
0eafd28f46eb3b7285e11f9fd514eb0c
26aae5a5def55379b0917a1c3ff3c7458fb2f81b
5581 F20101221_AAAUIZ harpley_c_Page_109thm.jpg
0058684b06b51efdf477cf666052ed08
4e7eb1d2831442abc8d599699fea4e831b33249c
2237 F20101221_AAATGM harpley_c_Page_126.txt
a0a0feb9d5266eadc099773b3dbe2471
ae9823478a4ca340b9c33c762a2da4b6da551540
13478 F20101221_AAATFX harpley_c_Page_101.pro
18f368bce46e3996472ab083ed899138
a364513802cf1542b7d38f2d1aa4eafc55c56cfb
F20101221_AAATHA harpley_c_Page_153.pro
c99d0e574ce86c20b368757c9a8b60a3
a3d303ceb945cb9408df20cda26377196de30e89
8347 F20101221_AAAUKD harpley_c_Page_131thm.jpg
df667e3e9d5618e73ea149c1dd30ba44
cf0c69137be545dbe5442e9a44617998872404d0
37838 F20101221_AAAUJP harpley_c_Page_121.QC.jpg
7f1ca4a3d82d1962561b970bb2f4324f
ca117df692f15a82c8a4e9fa07d6824ac9b7fad5
1278 F20101221_AAATGN harpley_c_Page_152.txt
f7fa6be61b5f31de0dab3021b2e4786d
a0b626a395a99656ad603c1a2f560af53d491702
F20101221_AAATFY harpley_c_Page_086.tif
3c39b683da499a713fedf9bbdbff89b9
73813b7da040552e47fd7c7a107a2c1895db7760
37066 F20101221_AAATHB harpley_c_Page_036.QC.jpg
46e326d65c0ea9d4256819800942c177
1eb8bb522488ba18242fba817a2d08c38db6a484
36043 F20101221_AAAUKE harpley_c_Page_132.QC.jpg
051c2a99b0c6eadd88b5a1c561f63268
73a7f7cce1c24380db6c58f6ae3b0c586a68efba
9203 F20101221_AAAUJQ harpley_c_Page_121thm.jpg
ff8b66b8305973984596a249d51088d0
4fd34ebe550afc6c710af4d8c1a129c88c000618
8706 F20101221_AAATGO harpley_c_Page_082thm.jpg
e063da06ae2a2f514293f0018bceecb2
bb1f9cbc86a1590d5d99a95ebe463216ece70c7d
54542 F20101221_AAATFZ harpley_c_Page_144.pro
59b79c05f11648d92926ac95d3109de7
2854773007df55bb4af8924d7bf11d3766373a76
8378 F20101221_AAATHC harpley_c_Page_015thm.jpg
5ebad102cf3ef76a43c038f26076fce3
deaf31017789032f01fdc69eb36e9ce93cc07ed2
8969 F20101221_AAAUKF harpley_c_Page_132thm.jpg
b0f49baf1dad453a2ba558cf469ebae1
79244d74c3f1a0138312003d16185d34c1c7a66e
37241 F20101221_AAAUJR harpley_c_Page_122.QC.jpg
848753512c30aa984f64b153eb7cecba
815e47e10038565e27f4a81620194b8a501c957a
1051912 F20101221_AAATGP harpley_c_Page_098.jp2
55ecaee714eaed21bba7aea3b4bbb4a7
60f6e81c20b5e1a1ee43d2b678aa4c4c06eec1e8
34295 F20101221_AAATHD harpley_c_Page_043.QC.jpg
806e678e1b7d39ff4d849bc4ca0702c6
348fffc2a0e537242bac542794a5b45b16df3de5
9280 F20101221_AAAUKG harpley_c_Page_133thm.jpg
86b187021ef94c0e2179dc41f34a871a
3003db784572907ca234e3acacdf102932728683
9188 F20101221_AAAUJS harpley_c_Page_122thm.jpg
e768678e5bb5c2ade29b5f5b7f44f409
4d8bff3a7c625c953881855e5a9321e43b9cf0bd
15335 F20101221_AAATGQ harpley_c_Page_099.QC.jpg
b93393f740ec88622109d3a4ad2d1c47
d1ba0c3afeca8795d4bc9e11b262482a7a91e6d9
8777 F20101221_AAATHE harpley_c_Page_047thm.jpg
5085b7ac0eecfe74f2d48212d97f80a2
5a0a80a8ecccb1c2bf108adca92e8ddbd990d117
37264 F20101221_AAAUKH harpley_c_Page_134.QC.jpg
dd294fc80718dd902f722c0586d1a7f1
a463165448a2fb4d9efd9beaacbc3e478f8ce637
36055 F20101221_AAAUJT harpley_c_Page_123.QC.jpg
dc41f9a05c5305fdda5af35de93abfb3
e7e78da984d1dce5d77c15d35965f63ced437349
133396 F20101221_AAATGR harpley_c_Page_155.jpg
df1dbb20497e76b49265889d4af1fa53
da94e02830fbc468deb5d036b33283921b620214
163870 F20101221_AAATHF harpley_c_Page_006.jpg
b7d4f0975326f48d9a27677ea0c151dc
20e770454bb7fb1a8deab522edd6725313ab87ca
9094 F20101221_AAAUKI harpley_c_Page_134thm.jpg
ca3a75d3bfcf3ce90133905d9ee40a96
318acd03bbacb33f46d2bb1d7b3f762073f32da3
8959 F20101221_AAAUJU harpley_c_Page_123thm.jpg
a6023c0cee49f74474c99118a65a25f0
4c36336a09890b780f50729deb5999183f311337
8945 F20101221_AAATGS harpley_c_Page_074thm.jpg
ab5b713a461f46a312cc4d07173554de
cdf9e9adba32cf3caf4483718acfd9f11a220159
36209 F20101221_AAATHG harpley_c_Page_148.QC.jpg
43fe5ae94187b109d86167bf07c3fdf3
f4b187f52fbe5c4d48a49cb3860bf527d6e2b27e
9100 F20101221_AAAUKJ harpley_c_Page_136thm.jpg
ef4607e18ce44201bea2454edcd02c74
791bf910353df146cb54438f651451e4512c5d24
9198 F20101221_AAAUJV harpley_c_Page_124thm.jpg
0d9446e83d0ad2ff1fc52d715c4e3972
9cc74cb304c17899833c6867fa5195cafe27173d
F20101221_AAATGT harpley_c_Page_137.tif
a0030a6182e0d82bb663bd8d2feea06d
761fd5e1ccb9574bfd7405cdb6466d108ce80ce4
26153 F20101221_AAATHH harpley_c_Page_012.jpg
38969738bdfc9f96a01bf93aabd64305
f9ebac2c2d3fa2597c9831de82646ab33749664a
8918 F20101221_AAAUKK harpley_c_Page_137thm.jpg
073f922057d3efc4bb248f4af9af79cb
493a356e320e1f6f067904d1fecdc89335e16af2
36544 F20101221_AAAUJW harpley_c_Page_125.QC.jpg
73f2ecbbd0eff20c37e8e4e86ddd61d9
b83c9a54e425ac1af309ed39c9637bd70ffe5c1a
20505 F20101221_AAATGU harpley_c_Page_034.pro
e6eace574d73d72ed054e00e3498bba7
69e86b2fbd6f284039bd8e8eab123335caee3ce5
95341 F20101221_AAATHI harpley_c_Page_031.jpg
27199da0226e133ae02936ef706ff061
fe4dd6a0f190c95095cea5b47580b72bbe976fa8
36255 F20101221_AAAUKL harpley_c_Page_138.QC.jpg
ce935bbd59107f76a08ed16c75303fea
04d626edb4fe42d9eb44e3254e015f02036bbb47
9103 F20101221_AAAUJX harpley_c_Page_125thm.jpg
36073f7e0112d7e3b20ceb8556c432a7
818af3cdee317af6b033efe72e56c7c61c0b368e
35198 F20101221_AAATGV harpley_c_Page_128.QC.jpg
02b6f9c4d448fa06658e1443d612ea81
599705c0dcf439ef887f7ca0e4a33d2767be15fe
8581 F20101221_AAATHJ harpley_c_Page_023thm.jpg
4a1068888ab77e53eeefd14a7a302b7a
64b9e1dcbdf0fe1ff4ceb99b508938cf242063c4
37602 F20101221_AAAULA harpley_c_Page_151.QC.jpg
93f2d072bb1964d66364496f5c8649d7
d3f09e627db02edda33cc88193d94a08d889d871
9020 F20101221_AAAUKM harpley_c_Page_138thm.jpg
8bd396bd5429c701450357ed58499cdc
5be19fccf1683ca079b429b237893916d157f7e8
37851 F20101221_AAAUJY harpley_c_Page_126.QC.jpg
a8be57927d01b2f50ee413ccc52d9012
44964e39a1bbd902c0de2a284240cb8dcc38083a
F20101221_AAATGW harpley_c_Page_058.tif
fe0fc8429f32158a35a79b5611e46e11
58220973afab0be81b0692c031bbf18470779254
F20101221_AAATHK harpley_c_Page_154thm.jpg
9f62cb53fc3a026c821c28cb2166d751
0d8cc60ebc1b89db311a3cb950402e05e4766a11
22386 F20101221_AAAULB harpley_c_Page_152.QC.jpg
63a718d31d0aebcd6448e8ef783d0c3b
1590e3b879bf28ef0b39554c082af4a343acfa5c
36003 F20101221_AAAUKN harpley_c_Page_139.QC.jpg
857af3040025ad4a261d8352909983ed
224f191051a3ac8b0211c3af52e7d31326de4ed6
55748 F20101221_AAATHL harpley_c_Page_143.pro
1f22fca76e906928cf83fb2119d5e5c3
8759a9b9bfce464f7bbb19541d146a6dd37d2da9
11677 F20101221_AAAULC harpley_c_Page_153.QC.jpg
436e15fe98d9d149d9958eeb22a244d9
7d9171c243e5c93734e0d7e44dff77ee386a94d0
8967 F20101221_AAAUKO harpley_c_Page_139thm.jpg
5c23ce49c638ba94afb1cbdd881b1bf7
3c0409a63128ba5bd8fd1d38e15269503c58ccf9
F20101221_AAAUJZ harpley_c_Page_126thm.jpg
d83175d5036986d103ede825b78be5f1
c2dc0a056017491dcc995ebb526fdccf48b6d45e
F20101221_AAATGX harpley_c_Page_144.tif
57034535d7b2ed9b884adbb8597b9b04
cee094e2f03bf8f122538bfd2a6b5dacd3eab3f7
111953 F20101221_AAATIA harpley_c_Page_030.jpg
6dfef3a5a5c05764d6f7be6654a26afd
70dce3d8cfe2dd8db4315da48affef289a264b22
8884 F20101221_AAATHM harpley_c_Page_018thm.jpg
8ee7513319af2b80753382c7d82a3501
a85b3acc85a25f3c66922423136117c8aa3ce99c
2963 F20101221_AAAULD harpley_c_Page_153thm.jpg
54d61a162437418d4746541e375772aa
66292849cb4acfe7e2d33bed086a20ae8e7fff35
F20101221_AAAUKP harpley_c_Page_140thm.jpg
860c81849e56d16d1e93cfecca0b63c1
1f208a4e9ee940a56ae820b0db65517ff9605622
37802 F20101221_AAATGY harpley_c_Page_143.QC.jpg
718167030441e5d568aa8bb99a95f719
f4e07e64897473b151300627118925934091d92e
4018 F20101221_AAATIB harpley_c_Page_006.txt
60bdbe678f727e051e98bc0ac3813532
ceccdec77a0a69514febc2660db479c3790ca206
F20101221_AAATHN harpley_c_Page_107.tif
81e616ebe2d992b11e6a63504bf2b5e0
0f472edd0d365db334b57482823dbdc9517f101c
35905 F20101221_AAAULE harpley_c_Page_154.QC.jpg
ed9d10a3b74c566d8136165b158639e5
55755cc3de1cf27f1c4d3b31068852fbd19a4eb9
36540 F20101221_AAAUKQ harpley_c_Page_141.QC.jpg
ade2c44aafafdc26044c70d5169b2465
e966627c5f63c03764b01837f14b5e1c365bfcc6
1515 F20101221_AAATGZ harpley_c_Page_103.txt
eb6aeba750257ff791c1eafe78be6ee5
59ae73e4900e51fae08e94bd838339a391997641
1455 F20101221_AAATIC harpley_c_Page_054.txt
a01b6e81382a77a26c9354e8277f434d
fce4756502f7f0d49726b90ddfa4600f30568e4c
F20101221_AAATHO harpley_c_Page_132.tif
878a1212a6d6164bb3618c19e95579e1
71059d74af82a86cc29dbfe35cbe7ee087c5deb2
36525 F20101221_AAAULF harpley_c_Page_155.QC.jpg
436585fccf428b05c4f9e508b397919b
1a8ef86d6390d0ebf8100258ab3e3857142fa633
38095 F20101221_AAAUKR harpley_c_Page_142.QC.jpg
e141c780d71264792f39cb0f3e048bda
f6d68623e625a85686c2a7360281c99b69dd70d1
F20101221_AAATID harpley_c_Page_054.tif
7ba8936b9c54212ca20535696dcf28b7
49740fd0761b8ff69bc720f105058fe26d5634a9
38945 F20101221_AAATHP harpley_c_Page_062.QC.jpg
a6ba89dad3f14392c377232f6a4b85aa
403d82442e488324dfb1de02d17b09f6b8a4222f
9259 F20101221_AAAULG harpley_c_Page_156thm.jpg
9b6e78f1281ada6e56bd2a02550a3389
95323bdf6b7b2c6b5168465d87245216625c2ba3
F20101221_AAAUKS harpley_c_Page_142thm.jpg
b23389a012d1512513d7062da2950d9b
ee952fa0c387940eeeda80334b4b58e2159c5793
1051917 F20101221_AAATIE harpley_c_Page_035.jp2
a000bd595ff37f88a2b5871822a5f440
e92e5b82642127a73acb484539f5443b78a7d363
1051935 F20101221_AAATHQ harpley_c_Page_088.jp2
b67497b5cb84659cf94646e9fa387d7e
9e20a9ba2c3ba30a285f9e8d70c43d29dd7426ce
37336 F20101221_AAAULH harpley_c_Page_157.QC.jpg
80b794d4ae764816f2b8afdaabeb20e2
0a1d0fdabb1018dfe6194c5bf108320f0010dc63
37455 F20101221_AAAUKT harpley_c_Page_144.QC.jpg
71443074e59f8ea0657e96300dcbc2ce
9d88c2874cdf9a763fb4bcc2c7341e79413e61e9
F20101221_AAATIF harpley_c_Page_148thm.jpg
fabe13cf88e7607c6c087c9efd1feeeb
f50d8e920c6fca66758436080ece454ff0d69efe
F20101221_AAATHR harpley_c_Page_124.jp2
495f48c3c19b521aa732aa34af809c3d
28587464743b33039d7ade483bfad527b5e85464
9186 F20101221_AAAULI harpley_c_Page_158thm.jpg
8c7081857f830992ad2147961c8a478d
1ed821fa1ba87b122de2e88f92b82e3cb4d21292
F20101221_AAAUKU harpley_c_Page_144thm.jpg
0bb6d14f21ad98f7953518cc6a11eafd
ea054fbbfa9dd3eefcc8fa2e2ff8bf982465d9e5
37749 F20101221_AAATIG harpley_c_Page_006.QC.jpg
cbbbc8bf3efeb0233a919beb9f00378a
91cd8caa9470e45ba6a06eb992613113f0e62de4
111862 F20101221_AAATHS harpley_c_Page_087.jpg
154293b0838acefdb1e15a59c831afe9
d2ece5597e5277feb8c8d25823cf151b89ab8874
36011 F20101221_AAAULJ harpley_c_Page_159.QC.jpg
da288627eec9d2245f6f628dee5afa25
8e0e691921dfa00dbef50530f45018e37d6e31e9
F20101221_AAAUKV harpley_c_Page_145.QC.jpg
95416052f0c07a555033757d4ef55d1c
fa23edb40279a95a504b1ba6f508a5644820741b
8965 F20101221_AAATIH harpley_c_Page_085thm.jpg
0f62a058fdc11ff5abd85d4bffb2b178
c55d9cd12be7692e9e5cb9304984fc8e0a5a26b7
2211 F20101221_AAATHT harpley_c_Page_046.txt
d068f6e334f922879d6c2c2c5dd50b24
162336597169e082a8ee8c532d3183f9d522c151
9031 F20101221_AAAULK harpley_c_Page_159thm.jpg
a7a280c14a4532952e164f4d0671d788
8ba8338b88733de4e176bf6469cff0f092efb14b
38376 F20101221_AAAUKW harpley_c_Page_146.QC.jpg
fd89eebea9dc939fedf991b06131b32f
c1a8d6d0490f56b8ac6a3c88698fe004a78cecdf
F20101221_AAATII harpley_c_Page_018.tif
1e12fa1bd887a19cc3a4a135e7a752ca
1df445f1c62028a2d30eae0df69ae724539e5fdd
F20101221_AAATHU harpley_c_Page_063.jp2
8fa6e91088a7913fd3ca511536166d1f
58c8ed3ee825432ab645d699a8251013b2ca3f44
38685 F20101221_AAAULL harpley_c_Page_160.QC.jpg
b556bf7f6c13286e8333b3c70fe0a8e0
5383274af6d43c9716df7d391b4907ab40d74f0a
9331 F20101221_AAAUKX harpley_c_Page_146thm.jpg
577132de51d185e4dd677b8dba640bce
86ce2e3b50458a6e051f307c9606394d242e6d78
38115 F20101221_AAATIJ harpley_c_Page_050.QC.jpg
082a288fff7d0b33780a8caf8ad78fc3
f4495a1f1eec7b73ab99acc153ab13b4c7b74379
55263 F20101221_AAATHV harpley_c_Page_066.pro
651cdb6911760e6efd412afcbafdb263
c7f6989d054d8b00cf9778290d841c24b7c02eb9
9398 F20101221_AAAULM harpley_c_Page_160thm.jpg
fd4a1a4e38bfcc9e5a2dbd0038cbab73
f6a3742a6aed075b29d05c03d00fc808e2cbbb59
37061 F20101221_AAAUKY harpley_c_Page_147.QC.jpg
2dc5ef1475d150b7bddc54fb5ab83222
e640ec4b12aa3c0f52c80e60ce81fc6aa1bbcbc9
F20101221_AAATIK harpley_c_Page_075.tif
1d5a1b609bc411a8f206c60a621b3681
37f7fb54c612be798731e9cc73479f0e26b1ab46
7596 F20101221_AAATHW harpley_c_Page_013thm.jpg
5d897ca8ea30f95396b0ce7becf4ef07
0ea716019d16f36ac34e38be6c54c7ce40a45150
9282 F20101221_AAAULN harpley_c_Page_161thm.jpg
26f9027d2593e5c7729645dc9cdbc464
d1ef7a112a9f7aabfe5e53cedd678bcfc4cd20a1
38133 F20101221_AAAUKZ harpley_c_Page_149.QC.jpg
bd25aff9051e134335b29542489b0243
b19209c699a99751546b99d684552f7dbad34e64
37068 F20101221_AAATIL harpley_c_Page_130.QC.jpg
27a2bd8f720b18edebdd84ec5b2020bb
e39399f25d4f0957b630147e993580170b395799
2256 F20101221_AAATHX harpley_c_Page_026.txt
3fe5b6128c58adf8bd47a64ca536f01f
890e730af0e115c1ed0cda32d9073fe834b94849
18495 F20101221_AAAULO harpley_c_Page_162.QC.jpg
a719ba6a579d2282a4a7169e360605f7
870637d6c90e62381a78cf43ff1a9e65bff7fbd5
F20101221_AAATJA harpley_c_Page_159.jp2
410c8ad6f48eb924df954db1860883c7
46c506cbf4f8d8d14ec3ef8e67083dfc9c6410b5
F20101221_AAATIM harpley_c_Page_120.tif
4fa70fab6cc6c6c395734a3e2e51ffd6
5fff4bfad3eec6e5138a5bc58146778bafd233e9
2101 F20101221_AAAULP harpley_c_Page_163thm.jpg
f5b10a2c32849b695603e8184ebe962d
41318b7b24eb260941f414313a6e73ae3339fd39
F20101221_AAATJB harpley_c_Page_138.tif
c0e1851f1c06ce8d4ab8aa378155905b
bc4e781f69998963da3eb0e2918d6a7139b56f0e
2209 F20101221_AAATIN harpley_c_Page_133.txt
5bda20b8b2e6e3b061c736c35d75a5e2
fb5ce282e6e53a8bf3409ececb36bbc3c3eaae18
108412 F20101221_AAATHY harpley_c_Page_140.jpg
abb2c7da2ec4c48942a3c1889dcb6418
32e7ebc00cbd464e42197758e17d1419a86ca4aa
189085 F20101221_AAAULQ UFE0022336_00001.mets FULL
114a694eb2a8d91674a7a080b7560532
3cf3857227c858b6e77a4f7ebd6c821468a735de
64150 F20101221_AAATIO harpley_c_Page_159.pro
4a42707719f8d4d09cccaeabf6fcc8fc
e639ef8b8a3f27ae85f5cbd31005160b9c18b565
912 F20101221_AAATHZ harpley_c_Page_011.txt
fa956736a65e19bb3633ab84f889fdca
735dbc82dce2c652c6f7962fb4e84b2546d70b26
F20101221_AAATJC harpley_c_Page_021.tif
e27ffb38bf329e8baf87266e185e1844
e57d962c454701a799f62f0cc19359bff65a5768
9415 F20101221_AAATIP harpley_c_Page_039thm.jpg
64ad5297772e1822cf313dc0067fb2ac
7973410b3c9f3daa98b6026bab56842378889f67
54922 F20101221_AAATJD harpley_c_Page_063.pro
8294b981c7882f4d948ab8a634719e6d
21b23afc5c77be687ef8cac6befb916ea3b94c18
35098 F20101221_AAATIQ harpley_c_Page_069.QC.jpg
47bdad214b5ab68dd5fbc4cc82a6d07d
bc3e51c750088b22584be66e2b82cb6018a28537
8779 F20101221_AAATJE harpley_c_Page_030thm.jpg
fc583c02c34d7fc46a87710550552aa2
66fcd6f29fe7a68f0cfe83627ed132cac4816067
115543 F20101221_AAATIR harpley_c_Page_133.jpg
5e70ea41c36625296df387aed8126b7e
f6b9628c3e0f2d3459c7fa5e20c2c499c0df2f98
37367 F20101221_AAATJF harpley_c_Page_093.QC.jpg
1b4b0669d3cca09e7865bdec81e5781d
0c0bae9d991842c2d80e044d6f4fb3d5aeb0696d
F20101221_AAATIS harpley_c_Page_152.tif
6a5f50461e9d067937ea6683d0689401
2fdc0c13c70902ab627a16727f36822b162916ab
966420 F20101221_AAATJG harpley_c_Page_008.jp2
0c9327abe018b4a2afa889a7c8c04106
1181c107a32bced86857d9de64d1d9e144cd0a5c
F20101221_AAATIT harpley_c_Page_144.jp2
13d80df6bad0dacdd54823007bde5a1d
92ac39f8743dc55717318130d76c75123116fafc
1051918 F20101221_AAATJH harpley_c_Page_077.jp2
b5f10deb89fd98e7467c1655f6f33a91
df5f6cfef4eb48c122851b106dd0585946997205
F20101221_AAATIU harpley_c_Page_114.tif
38d4c3f0cc8b6e3157036f111094a738
a1e248706562198e350f0259da625fa2878e4866
37982 F20101221_AAATJI harpley_c_Page_133.QC.jpg
c19b05b91e1d029d3163af12c4675585
8c9bd7f3a5a97af16b18573fee1d4a642ad1d093
39680 F20101221_AAATIV harpley_c_Page_089.QC.jpg
978a5f8fa6abff63e857fbaacf5d940c
e64accb89a115227f2fe5e70e7f043b36f877de5
8811 F20101221_AAATJJ harpley_c_Page_051thm.jpg
32479068fc438f4423db5b314d7aa7ef
f0f9c6b0ec29b59ebee830cdcebf3d9fb5114c06
67764 F20101221_AAATIW harpley_c_Page_116.jpg
6d68be12ccbf7310becead817b81b29e
77964ecc46a7fe71acb6460f1a0817c8b0481856
F20101221_AAATJK harpley_c_Page_044.tif
348810c548a74bf21bc68b497838c28b
a668818ace18faf1bae3dcf0b7c2852cb33c1ead
4708 F20101221_AAATIX harpley_c_Page_112thm.jpg
a474fd8c6ac9c0c11b51ed5ce129c92e
9900a4d37b5e18f30d7b7388f1eb01be97b9d633
F20101221_AAATJL harpley_c_Page_079.txt
5462589964c2240890d4837a76624b8d
f5e5a8b5f4349e4dc0d9702658c14a7785161bf2
112805 F20101221_AAATIY harpley_c_Page_065.jpg
a062c2ef19dbf1013f39491ce0253d75
0426842eb4c88851ad3086453843ee0b974651b2
14160 F20101221_AAATKA harpley_c_Page_056.QC.jpg
72b9fa3a6be15afe1a66f2caedaf5eb0
a4e4b6a1106d58a897929e7886cf064693cbd89b
37800 F20101221_AAATJM harpley_c_Page_058.QC.jpg
7eb8fd2a87386786338ba620eaeb474b
67406d11b4362535ce3b3a796d573f8366a10d96
F20101221_AAATKB harpley_c_Page_032.tif
2ecc31460d7686596a37fb8536c93b8b
4c96fefcba4ac2412384c6b058e14367786b636f
53234 F20101221_AAATJN harpley_c_Page_057.pro
22e6ecebc40e2384458fd40211f488b0
2a6301cf864dd5fc7d748a455bfe4428442785bb
46105 F20101221_AAATIZ harpley_c_Page_013.pro
fb281e361fefb1e259349a1a6a066582
f1a4dc22a39944e3f5b3a9e5f5bdde78f15ffb28
24686 F20101221_AAATKC harpley_c_Page_117.QC.jpg
04c9da98fceb7aed1123af9c7fd106a6
185a478022ea898087ab497c5007a5920cae5b51
2017 F20101221_AAATJO harpley_c_Page_013.txt
1d2b89e665df1b2960e8934173787812
e0f46c28d35a8e00d8e3ebb62be62b2181f06e73
F20101221_AAATKD harpley_c_Page_064.tif
5eb0fdf007bcba03a0ffc56a59cf72cc
9c5a6b68023b4d89b649d9e8fab3cdcf8a6db8e6
9114 F20101221_AAATJP harpley_c_Page_066thm.jpg
7230df09060de5fd6ccabf1d72c0acf3
1714f29e4045efbebb058018d6cb862b2da94aeb
105986 F20101221_AAATKE harpley_c_Page_049.jpg
4fbdef0c3ead869589ef689ec3f79c9e
27c5e85307c77d84e4b6c7e6be7bde3720c310a8
103554 F20101221_AAATJQ harpley_c_Page_043.jpg
18761069d06b1caecd8de1ce23037d96
3c8e2e602e71a716350824a00b78fe336acf5c7a
2585 F20101221_AAATKF harpley_c_Page_159.txt
130be79188ad5b8b3b827347aa40c185
adff5cdd17be6bef88b6d67f503c8b5aee129a2b
F20101221_AAATJR harpley_c_Page_013.tif
9b48fa6c0250eacc7a2e9fef486865a2
24af39e61747e5a56a24bd0cacf25fa287b24cf6
54069 F20101221_AAATKG harpley_c_Page_090.pro
e3cb96ffc8c73e58a67cc3daf487d8ca
08ed406e1c932493fda7bbc33bba69843df88486
F20101221_AAATJS harpley_c_Page_031.tif
13b383659c245b0b9fb421aa91bee0ab
ca36d44afa042080c78c09691c22750cad62b354
F20101221_AAATKH harpley_c_Page_037.txt
11373b9ed066d9126f8138d6bf841907
7fb0a5f3b552ea46a56779977a95821db5ccf56e
F20101221_AAATJT harpley_c_Page_050.tif
2b2a7753e9a33e2798d3e84db8105ffd
27bd8b820b5b5d13946e1133457653a347d542b6
F20101221_AAATKI harpley_c_Page_064.jp2
f81ee70964e22064be5892ab833757be
c1d9f47f90b4e9b33eb058a81a390cb4413e5506
109683 F20101221_AAATJU harpley_c_Page_070.jpg
a5f2547a46b61ce9a117a3216e6f586e
0fc4eba0225ef83ad8398e2b873f209a137ec8d3
53471 F20101221_AAATKJ harpley_c_Page_050.pro
2efbc343275e3d0a714e80f1f604e4ab
3a86899ad18fd23dc9b843faeaf21f8decfe4149
2068 F20101221_AAATJV harpley_c_Page_098.txt
ec7789fecf3d874cf7cda182c1de4351
87105fd02630d49db886eb51e0a7bb6253769949
22659 F20101221_AAATKK harpley_c_Page_114.QC.jpg
7a17e1b8a265991d56503bff2b3e2b7f
26d00411101da7fedb77c3af21beda92c16e6b01
749442 F20101221_AAATJW harpley_c_Page_152.jp2
e35d6c1a5450158f587f4f978ec2f85d
8c9f006cb870aa6e7a215ef5d10169c286dee845
F20101221_AAATKL harpley_c_Page_033.txt
641c3c5aada8c2bd2488a9ccbb304544
b3d33a24896f781ff1715de62fcb8bb62bf7744a
9253 F20101221_AAATJX harpley_c_Page_042thm.jpg
8fb755569472e72d650585f6e67c51d8
14a1015b667d3124f60fba82335d9a1731000330
103663 F20101221_AAATLA harpley_c_Page_015.jpg
786ca9746b4842b4ba8f1ad535636129
cf4996982e59169c79b534104e6e7700e9242f56
56476 F20101221_AAATKM harpley_c_Page_094.pro
6ce64e0761f27511bea705cb36d134cd
fb5062eb8eb194c6421421864544b12bcf74c04d
F20101221_AAATJY harpley_c_Page_106.tif
62fed97ae8fd6f365afb115c3a374730
c19d04465c55ac8baf55a74c117d7b312b2309e7
111261 F20101221_AAATLB harpley_c_Page_017.jpg
3c9ed410e4b2aa558aecd2e7c021f87e
85f6c8d547ac35cd0879c7d4d016430e6afde0dd
244863 F20101221_AAATKN UFE0022336_00001.xml
ff123fefbce2c730e7f66e26a83d8a16
56a910f2402fed2617fcd9da9498148bee9ad47a
34596 F20101221_AAATJZ harpley_c_Page_131.QC.jpg
9d25fa67bbb462f2fb4bde363efdf3ad
99ac22da964cfbee185c8d17f3fb91f571fbe60e
F20101221_AAATLC harpley_c_Page_018.jpg
d202920df59063a7a5aab94bb6d4f949
122d238e832cc683b628b13e73a80443e1b38189
110366 F20101221_AAATLD harpley_c_Page_019.jpg
28e247cffe727b6217a9dfc720ca3578
c23e4efd9a041286deb43655229fb6f59f3715ce
109687 F20101221_AAATLE harpley_c_Page_021.jpg
6d723dd03a8ad61c781dc441e85e31c1
88c47a5d2a7e33a32c92703c03d7239e4f3231d1
3960 F20101221_AAATKQ harpley_c_Page_002.jpg
459c5d4be10e656bddfd81ca908bb362
0ca1477d7afead97eb0a72e2991acb702b0cc926
107975 F20101221_AAATLF harpley_c_Page_023.jpg
1a4ba98f82c67551de5cdb206fdca146
1d4e69f72b2dd642c869b7a3d516049277d32662
9054 F20101221_AAATKR harpley_c_Page_003.jpg
e43418386a04bc475fd1e633bead6a4a
4866b2a5fa69f8347ae61b27a9026f08a3bc3ec8
117458 F20101221_AAATLG harpley_c_Page_024.jpg
cd948571a773a8ee901afafc2d938e7a
fb2dc78274a8c3b4b4b9e482c75dbe58d058d3af
24841 F20101221_AAATKS harpley_c_Page_004.jpg
512f56faffebf346502ef4e22e9d6e08
db3aae098aca51d2984f5671ed9393c73d8c5eff
118783 F20101221_AAATLH harpley_c_Page_025.jpg
c8dc94132d7469f80da3eb82c4623039
4ec9659d97ac91409f052cdbf9bbc2fcb33e0363
125602 F20101221_AAATKT harpley_c_Page_005.jpg
9bd24c71e269d19ef8cf5032425275c1
e976f87010f4e8be946fd89a0a816abdbe1b8a7b
118005 F20101221_AAATLI harpley_c_Page_026.jpg
c383828e9e6c5325759138fe39593555
8d3ee16bd0385c33f1fcf6741e7cfebf20060ebb
24901 F20101221_AAATKU harpley_c_Page_007.jpg
8b29f05382f113ab4af751adce691219
652b4f09f3d3fafba942fcb0a2287cff26a7d798
111092 F20101221_AAATLJ harpley_c_Page_027.jpg
89553fd905e1e22f4f69fe44e581bb17
e552084e3e62591143629536c8597eb6eba0b858
53373 F20101221_AAATKV harpley_c_Page_008.jpg
8d5101a73feb5b9b8ab894cdf3a4b68c
44e1af1a9c5e5f9547c7a4cdc4eb0f53b8015622
109136 F20101221_AAATLK harpley_c_Page_029.jpg
e3ae52eaa4d9b4883ee99dac01d07725
9a1947cc3c1b7962d1cfe652d844aa4d949d6c08
129795 F20101221_AAATKW harpley_c_Page_009.jpg
097a12efc524ff78670f501c8513d84d
9a9e8255d0c2145ce68d73ccde807c3c57b08309
F20101221_AAATLL harpley_c_Page_032.jpg
022a68cd6a0267ccfe58030b69524035
edec2966aaabee2267dc7d664f4df9166ada4b6a
50564 F20101221_AAATKX harpley_c_Page_011.jpg
2346a75bfd7631d784d0e83f300861ba
a1bde40aa9341ce8b7d8dab956c403324c1be31d
39178 F20101221_AAATLM harpley_c_Page_033.jpg
4b8af6c0ec1a6c5d8d476689396d00a7
1f2b40327eb2d67647080603b7ea99b05b86a6cb
98492 F20101221_AAATKY harpley_c_Page_013.jpg
24e1924eab84a2ec58e9e448105a1aa4
10bb4612370679ba61d8b740e022f0eb683fb2d4
109682 F20101221_AAATMA harpley_c_Page_052.jpg
851fbec205e9c19c1361a0ca3fca752b
7215f57238adc68ed2dc768be8ecc60358ba88e6
48026 F20101221_AAATLN harpley_c_Page_034.jpg
20315a7dac6fb238715ec3f239d7fce4
7a0d3b7f6603e6ef6782915f0952f0e6edfe5d98
104797 F20101221_AAATKZ harpley_c_Page_014.jpg
753eead535649a7005f946e4e24238f1
ff6708d77ce4ae4c24dcd5a92a8024086210af5d
77365 F20101221_AAATMB harpley_c_Page_054.jpg
575bb356a846f344529f65f2ee648732
d055ebc0cbd0266913da0717a6503b80818c5e3c







OPTIMIZATION OF METHODS FOR PHAGE DISPLAY USING SINGLE-CHAIN
VARIABLE FRAGMENT PHAGEMID LIBRARIES




















By

CRYSTAL J. HARPLEY


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA

2008


































O 2008 Crystal J. Harpley


























To my parents, siblings, and Matt. They, never doubted my abilities and always pushed me
harder when I doubted myself.









ACKNOWLEDGMENTS

I would like to thank my mentor, Paul A. Gulig, for teaching, guiding, and pushing me to

be a better scientist. I would also like to thank my co-workers in the Gulig lab for their

encouragement, help, and making me smile throughout the work day. Lastly, I would like to

thank my committee members, Shouguang Jin and Anita Wright, for their help and guidance.












TABLE OF CONTENTS


page

ACKNOWLEDGMENTS .............. ...............4.....


LIST OF TABLES ........._._ ...... .__ ...............8....


LIST OF FIGURES .............. ...............9.....


LI ST OF AB BREVIAT IONS ........._._ ...... .... ............... 11..


AB S TRAC T ............._. .......... ..............._ 13...


CHAPTER


1 INTRODUCTION ................. ...............15.......... ......


Methods for Bacteriological Detection and Analysis ................. ..............................15
Vibrio cholerae ................ ...............18...
Escherichia coli 0157:H7 .............. ...............20....
Phage Display ............... ...............23...
Recombinant Phage Libraries......................... ...................2
Tomlinson I + J Human Synthetic VH + VL Phagemid Libraries ................ ................ .25
New England Biolabs 12mer Peptide Phage Library ............_......__ ..............26
Tools for Production of Phages from Phagemids ................. ...............27........... ..
H elper Phage .......................... ..............2
Helper Phage with giII Mutations ....._____ .....__......._ ......__ ...........28
Helper Plasmids............... ...............29

2 MATERIALS AND METHODS .............. ...............35....


Bacterial Strains, Media, and Growth Methods............... ...............35
Biopanning of Phage Display Libraries. ............ .....__ ...............36.
Panning on Immunotubes ................. ...............36..............
Panning in Suspension............... ..............3
Panning on Microtiter Wells .............. ...............39....
Panning on Nitrocellulose Paper ................ ....__ ...............41. ....
Spot Titer of Phages ................. ...............43....... .....
Spread Titer of Phages ............ ............ ...............43...
Am plification of Phages ............................ ..__ ... ... ....... ....... .............4
High Throughput Production of Soluble Antibody Fragments (scFv antibodies) ..........44
Deoxyribonucleic Acid Manipulations ........._.. ...._._........._._. ............ 4
Plasmid Extractions ........._..... ...._... ...............44.....

Ag arose Gel El ectrophoresis .............. ...............44....
Electroporation of Plasmids .............. ...............45....
Electrocompetent Cells............... ... ........................4
Enzyme-Linked ImmunoSorbent Assays (ELISAs) ...._.. ................. .....................46











Infection Efficiency .............................. ..................4
Protein and Lipopolysaccharide (LPS) Manipulations ................. .............................48
Phenol-Water Extraction of LP S ................. ......... ...............48. ...
TRlzol Reagent Extraction of LPS ........._._. ...._. ...............49..
Extraction of Periplasmic Proteins ..........._.. ......_. ....._.._ ............4
Determination of Protein Concentration ..........._. ....... ... ........ ......_...... ............5
Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis (SDS-PAGE) ..............50
Coomassie Blue Staining. ............ ...... ._ ...............51...
Tsai-Frasch Silver Staining .............. ...............51....
W western Blot.................. ... ... .......... .... .........5
Lipopolysaccharide Saturation to Nitrocellulose Paper .................... ............... 5
Colony Blot with scFv ................. ...............54......___....

3 RE SULT S .............. ...............57....


Rationale for Study .................. ........_ ............ ............5
Specific Aim 1: Panning to V. cholerae LPS ........................... ... ......................5
The Phenol-Water Method Extracted y. cholerae LPS Most Closely Resembled the
Commercially Acquired y. cholerae LPS .............. .. .........................5
More V. cholerae LPS Can Be Bound to Nitrocellulose Paper than to a Microtiter
W ell ................... .. ....... ...... .... ... .......... .... .. ..... .. .......6
Panning to V. cholerae LPS Failed to Yield Phages that Were Specific to
V. chol erae LP S .............. ...............64...
Conclusion of Specif c Aim 1 .............. .. ...... ...............6
Specific Aim 2: Improve Panning and Screening Process of Biopanning ................... ..........67
Elution of Bound Phages by Trypsin during Panning Was Optimal between 10 and
30 M minutes .............. .. .... ........ .... ... ... ...... ... ... ........6
Screening of scFv Proteins by a High Throughput ELISA Was Acceptable ..................7 1
Screening of Phagemid Particles by a High Throughput ELISA Was Not Optimal.......73
Conclusion of Specific Aim 2 .............. .............. .............7
Specific Aim 3: Improve Phagemi d Particle Producti on ................. .......... ....... .................7
Escherichia coli TG1 Harboring the Hyperphage Genome Was Not an Optimal
Phagemid Particle Amplification Tool ............................ ... .. .................7
Helper Plasmids Were Not an Optimal Phagemid Particle Amplification Tool .............80
Homemade Hyperphage Titers Were Increased with Amplification in E. coli
M Gl655 (pGTR203) .............. ...............85....
Conclusion of Specific Aim 3 .............. ... ...... ...... .....................8
Specific Aim 4: Isolation of Specific Recombinant Phage to Stx2 Toxin of E. coli
0157: H 7 ........._..... ... ....._ .. ...............89.
Conclusion of Specific Aim 4 .............. ...............95....

4 DI SCUS SSION ........._.. ..... ._ ............... 1 19...


Specific Aim 1: Panning to K. cholerae LPS .............. ............ ..........12
Specific Aim 2: Improve Panning and Screening Process of Biopanning ................... ........ 126
Specifi c Aim 3: Improve Phagemi d Particle Producti on ................. .......... ...............135












Specific Aim 4: Isolation of Specific Recombinant Phage to Stx2 Toxin ofE. coli
0157: H7 ........._.___..... ._ __ ...............146..


5 EPILOGUE ........._.___..... ._ __ ...............153....


LIST OF REFERENCES ........._.___......___ ...............154....


BIOGRAPHICAL SKETCH ........._.___..... .__. ...............163....










LIST OF TABLES


Table Pan

2-1 Bacterial strains and plasmids used. ............. ...............55.....

3-1 Infection efficiencies ofE. coli TG1 containing various helper plasmids with
phagemid particles. ............. ...............97.....

3-2 Infection efficiencies of E. coli TG1 (M13cp-dg3-sm) with phagemid particles............_..97

3-3 Infection efficiencies of E coli TG1 (M13cp-CT-sm) and E. coli JM109 (M13cp-
CT-sm) with phagemid particles............... ...............9

3-4 Comparison of transducing units to particles per milliliter of amplified phagemid
particles. .............. ...............98....

3-5 Comparison of transducing units to particles per milliliter of amplified homemade
hyperphage produced from F- E. coli strains. ........._._.. ....__.. ...._.._._.........9










LIST OF FIGURES


Finr IPg e

1-1 Structure of M 13 phage. ............. ...............32.....

1-2 Structures of antib odi es ................. ...............33...............

1-3 Genetic map of plT2 phagemid vector from the Tomlinson scFv library. .................. .....34

3-1 Analysis of TRIzol Reagent-extracted y. cholerae N16961 LPS by SDS-PAGE. .........100

3-2 Analysis of phenol-water-extracted y. cholerae 569B LPS by SDS-PAGE. ................101

3-3 Analysis of the saturation limit of y. cholerae 569B LPS to microtiter wells by
ELI SA ............. ...............102....

3-4 Analysis of the saturation limits of primary antibodies to y. cholerae 569B LPS-
coated microtiter wells by ELISA................ ...............103

3-5 Saturation of y. cholerae 569B LPS to nitrocellulose paper. .............. ....................10

3-6 Titers of eluted phagemid particles from the first one-round panning optimization
experiment. ........... ......__ ...............106..

3-7 Titers of eluted phagemid particles from the second one-round panning optimization
experiment. ........... ......__ ...............107..

3-8 Titers of eluted phagemid particles from the third one-round panning optimization
experiment. ........... ......__ ...............108..

3-9 Analysis of eluted phagemid particles from the third one-round panning optimization
experiment by ELISA. ............. ...............109....

3-10 Analysis of anti-B SA (a-BSA) scFv proteins produced in microtiter wells by ELISA. .1 10

3-11 Analysis of a-BSA and Vc86 phagemid particles produced in microtiter wells by
ELISA ........... ..... .. ...............111...

3-12 Analysis of a-BSA phagemid particles produced in microtiter wells by ELISA. ...........1 12

3-13 Analysis of a-B SA phagemid particles produced by hyperphage by ELISA. ................. 113

3-14 Analysis of phagemid particles (1-22, 43-64) selected from panning against Stx2
toxin preparation by ELISA ........._.___..... .___ ...............114...

3-15 Analysis of phagemid particles (1, 3-5, 23-42, 65-84) selected from panning against
Stx2 toxin preparation by ELISA. ........._._._ ....__ ...............115..










3-16 Analysis of anti-Stx2 monoclonal antibodies and clones 46, 48, and 49 by ELISA.......116

3-17 Western blot analysis of phagemid clones from panning on Stx2 toxin preparation. .....1 17

3-18 Analy si s of Stx2 toxin preparati on (Toxin Technol ogi es) by SD S-PAGE ....................11 8









LIST OF ABBREVIATIONS


2xTY AG

BSA

BSG

CFU

CT

ddH20

dH20

DNA

ECL

ELISA

HC

HRP

HUS

Hye

LPS

mAb

MWCO

OD

pAb

PBS

PBST-0.01

PBST-0.05

PC

PCR


2xTY medium containing 100 Cpg/mL and 1% (w/v) glucose

Bovine serum albumin

Phosphate-buffered saline containing 0.1% (w/v)gelatin

Colony forming unit

Cholera toxin

Deionized distilled water

Deionized water

Deoxyribonucleic acid

Enhanced chemiluminescence

Enzyme-linked immunosorbent assay

Hemorrhagic colitis

Horseradish peroxidase

Hemolytic uremic syndrome

Hyperphage

Lipopolysaccharide

Monoclonal antibody

Molecular weight cut off

Optical density

Polyclonal antibody

Phosphate-buffered saline

PBS containing 0.01% (v/v) Tween-20

PBS containing 0.05% (v/v) Tween-20

Phosphate-citrate

Polymerase chain reaction










Polyethylene glycol

Plaque forming unit

Replication factor

Ribonucleic acid

Signal to noise

Single chain F variable

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis

Shiga-like toxin 2

3 ,3 ',5, 5'-tetramethylb enzi dine

Transducing unit

Multiplicity of infection

Vibrio cholerae

Volume per volume

Weight per volume


PEG

PFU

RF

RNA

S:N

scFv

SDS-PAGE

Stx-2

TMB

tu

MOI

Vc

v/v

w/v









Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

OPTIMIZATION OF METHODS FOR PHAGE DISPLAY USING SINGLE-CHAIN
VARIABLE FRAGMENT PHAGEMID LIBRARIES

By

Crystal J. Harpley

August 2008

Chair: Paul A. Gulig
Major: Medical Sciences

Detection assays for biological agents and their products are important to identify disease-

causing organisms in water, air, food, and patient samples to enable prevention or treatment of

the disease. A commonality in almost all detection assays are proteins that specifically bind to a

target molecule. While monoclonal antibodies are the most common detection reagents used in

detection assays, phage display reagents are becoming more prevalent. Phage display involves

the display of recombinant peptides on bacteriophages. These recombinant phages can be

panned against a target antigen to select recombinant peptides that bind specifically to the

antigen. Optimization of protocols and reagents used in phage display was the goal of this thesis.

Optimization of whole cell panning with the Tomlinson J human synthetic VH + VL (ScFv)

phagemid library revealed that a trypsin elution between 10 and 30 minutes eluted the highest

number of phagemid particles; this may be because the phagemid particles were degraded by

trypsin treatment after 45 minutes. A high throughput ELISA for the production and screening

of scFv proteins was developed and demonstrated that scFv proteins produced in a microtiter

well could be screened by ELISA and produce a detectable S:N even when diluted 1:10. A high

throughput ELISA for the production and screening of phagemid particles was attempted, but









phagemid particles could not be produced in a microtiter well in titers high enough to make a

high throughput ELISA acceptable for screening of phagemid particles.

Various phagemid particle amplification tools were analyzed. The hyperphage genome

was incorporated into E. coli and maintained as a plasmid to enable phagemid particle

amplification. However, these strains had low infection efficiencies with phagemid particles.

Helper plasmids were incorporated into E. coli to enable phagemid particle amplification.

However, these strains produced low yields of phagemid particles. The titer of laboratory-

produced hyperphage was improved by use of E coli MGl655 (pGTR203), expressing the M13

giIl gene. The improved high titer hyperphage was used to successfully amplify the Tomlinson J

scFv phagemid library to ensure that the redundancy of phagemid clones was maintained. Using

the newly amplified library to pan against an E. coli 0157:H7 Stx2 toxin preparation resulted in

89% of phagemid particles from the third round of panning recognizing the Stx2 toxin

preparation in an ELISA.

Previous panning procedures used in this laboratory resulted in no better than 1% of

selected phagemids being specific to the target antigen, and often no usable specific clones were

isolated. The optimization of panning procedures and reagents developed in this thesis greatly

increased the efficiency of selection of phagemid particles specific to the target antigen. Using

the improved reagents and procedures developed in this this yielded over 80% positive

phagemid clones selected against E. coli 0157:H7 flagella by another lab member. The

phagemids recognized the maj or flagellin protein by Western blot (data not shown). Therefore,

the improvements on phage display techniques and tools described in this thesis are, in fact,

useful and offer promise of success for continuing studies in the laboratory.









CHAPTER 1
INTTRODUCTION

Methods for Bacteriological Detection and Analysis

There has always been a need for detection of biological agents. One basic reason for this

need is for prevention of disease by identifying biological agents in food, water, or air. Of key

interest are organisms that have potential as bioterrorism agents. With resent bioterrorist events,

the need for rapid detection of bioterrorism agents has increased (1). Another reason for

detection of biological agents is for the purpose of diagnosing disease by detecting agents in

patient samples so that suitable treatment can be given. Detection is also of interest to

epidemiology. By detecting agents that cause disease and analyzing them through epidemiology,

measures can be implemented to control or prevent outbreaks. Various methods have been used

to detect organisms or their products.

Since the discovery of microscopic organisms, methods for detecting the organisms have

evolved. One of the first methods of detection was culturing an organism in enrichment

medium, followed by selective medium. Afterwards, the organism would go through various

biochemical and metabolic tests for identification. However, culture-based identification of

organisms can take days or even weeks. As seen in the anthrax attacks in the fall of 2001 and the

SARS virus outbreaks of 2002-2003, symptoms can occur within days of exposure and infection

can spread rapidly. These reasons enforce the need for rapid detection of biological agents (2).

The rapid detection of biological agents is necessary to treat individuals at risk, limit

transmission of the disease, improve public health surveillance and epidemiology, and monitor

environmental impact (3-5).

Methods have progressed to optimize detection of agents and their products in clinical and

environmental settings. Over the last 25 years, assays for detection and identification of agents









have improved immensely. Particularly, the reagents and detection equipment have improved to

allow for the detection and identification of agents in as little as a few minutes (2). An ideal

detection system would be rapid, sensitive, selective, inexpensive, and would not require

extensive training of personnel to operate the system. While there is no single optimal-detection

method, there are numerous methods for detecting a variety of agents in a variety of

environments.

Immunological tools are one of the most widely used and successful methods for detection

of biological agents. Since the first radioimmunoassay was developed by Yalow and Berson in

1959 for the detection of human insulin (6), immunoassays have expanded and been used for the

detection of a variety of agents. The main components of an immunoassay are summarized by

Andreotti et al. as follows: "Immunoassays rely upon four basic components regardless of the

application and underlying technology: (i) the antigen to be detected; (ii) the antibody or

antiserum used for detection; (iii) the method to separate bound antigen and antibody complexes

from unbound reactants; and (iv) the detection method. The efficacy of any given immunoassay

is dependent on two maj or factors: the efficiency of antigen-antibody complex formation and the

ability to detect these complexes."

The most important component of an immunoassay is the antibody. The discovery of

different types of antibodies has altered the range and scope of immunoassays. Polyclonal

antibodies have largely been supplanted by monoclonal antibodies. In the current age,

recombinantly engineered antibodies are supplanting monoclonal antibodies (2). The demand

for a variety of immunological assays reflects the growing number of assays developed to

optimize their use by increasing their sensitivity, speed, handling, and cost. The specific binding

of antibody to antigen has been coupled with a variety of detection applications including









fluorescence, enzymatic activity, chemiluminescence, electrochemiluminescence, metallic beads,

and many more. The detection of these complexes can be assayed on a variety of platforms such

as biosensors, flow cytometry, microarray, and lateral flow diffusion devices (1). Perhaps the

most common immunological assays are Enzyme-Linked ImmunoSorbent Assays (ELISAs).

Such assays either have antibody or antigen bound to a solid support that enables specific,

sensitive, and quantitative detection of antigen or antibody by optical-density detection of a

colorimetric signal.

Nucleic acid-based tools for detection have advanced in the past few decades. Of main

interest is Polymerase Chain Reaction (PCR), which was invented in 1983 by Kary Mullis and

coworkers (7). Polymerase chain reaction involves the amplification ofDNA by use of

oligonucleotide primers, heat stable DNA polymerase, and nucleotides in an exponential

capacity. The original PCR has since been altered to give quantitative real time-PCR (q-PCR or

kinetic PCR) and reverse transcription (RT)-PCR. Quantitative real time-PCR involves the

amplification along with quantification of the DNA, while RT-PCR involves reverse transcribing

a piece of RNA into DNA followed by PCR amplification of the DNA. Advances in PCR

chemistry and thermocyclers have shortened the length of DNA amplification from a few hours

to minutes. With small sample volumes in the amount of a few microliters and containing as

little as one bacterial cell, PCR is one of the most sensitive assays available. However, this

sensitivity increases the risk of contamination generating false positives. Field-based PCR

amplification and identification is not common due to the complexity of the system and highly

trained personnel required to operate and interpret the system. Perhaps the greatest constraint of

nucleic acid-based detection assays is the availability of genomic sequence data of biological

agents (2). Detection assays are continually being invented and improved to detect biological










agents. Because there are numerous detection environments, conditions, and agents, the

development and improvement of detection assays will continue to progress.

Vibrio cholerae

Vibrio cholerae is a gram-negative, curved rod-shaped bacterium with a single polar

flagellum found primarily in estuarine and marine environments. It is a facultative human

pathogen causing the pandemic diarrheal disease cholera (8). Cholera is characterized by

profuse watery diarrhea, vomiting, and leg cramps leading to dehydration and shock (9).

Cholera infection occurs through the ingestion of food or water contaminated with the bacterium.

Because cholera can be prevented by proper sanitation and hygiene, it is uncommon in

industrialized countries and is most prevalent in the Indian subcontinent and sub-Saharan Africa

(10).

Vibrio cholerae uses its toxin co-regulated pilus (TCP) to colonize the small intestine;

once attached to the small intestine, the bacterium secretes cholera toxin (CT) (11). Cholera

toxin binds to the epithelial cell receptor, Ghln, and is transported into the cell. Cholera toxin is

an AB toxin composed of a catalytically active A-subunit surrounded by a homopentameric B-

subunit. Once the CT is internalized, it is transported in a retrograde pathway through the Golgi

to the endoplasmic reticulum. In the endoplasmic reticulum it is retrotranslocated to the cytosol.

In the cytosol CT catalyzes ADP-ribosylation of the GTP-binding protein Gs (8) causing

adenylate cyclase to become constitutively activated. The increase in cAMP levels leads to

secretion of Cl~, HCO3-, and water from epithelial cells into the intestinal lumen causing diarrhea.

The loss of water can amount to 30 liters per day, and without proper rehydration treatment

can lead to a 30% mortality rate (12). Under-reporting of cholera infections is a great problem.

It is estimated that 3 to 5 million cases and 120,000-200,000 deaths occur worldwide annually

(13). However, this number could easily be multiplied by a factor of ten due to unreported cases.









There are many different serogroups of y. cholerae, most of which do not cause acute

diarrhea. Vibrio cholerae is classified into serovars or serogroups on the basis of its

lipopolysaccharide (LPS) O-antigen (14). There are at least 200 known serogroups, of which

serogroups 01 and 0139 are the only ones that cause epidemic or endemic cholera. The 01

serogroup can be further distinguished into three serotypes. Ogawa and Inaba are the most

common serotypes, with Hikojima being rarely reported. These serotypes can be further

classified into two biotypes, El Tor and classical, that differ in their biochemical properties and

phage susceptibilities (8).

With the enormous numbers of cases of cholera every year, there is a need for an effective

diagnosis tool for patient and environmental samples. Detecting endemic serogroups of

yK cholerae early in an outbreak is extremely important for control of an epidemic. A problem

with detection of y. cholerae is that cholera is a disease of developing countries. Outbreaks

normally occur around water-ravaged areas where laboratories are not prevalent. Therefore,

Hield-based assays are the most effective tools for early detection of y. cholerae. Because there

are many serogroups of y. cholerae, most of which present mild symptoms, it is important to

distinguish the epidemic strains from the non-epidemic strains. Vibrio cholerae 01 and 0139

are the epidemic strains. Determining features of these serogroups are the O-antigen of their

LPS and their ability to produce CT. Therefore, Hield tests usually detect these specific antigens.

There are many rapid diagnostic tests for cholera. Some of these detect CT by passive-

latex agglutination (15,16), while others detect the O-antigens of LPS from 01 (17-21) and 0139

(22-24) strains of y. cholerae. A commonly used rapid y. cholerae diagnostic tool is the

multistep colloidal gold-based colorimetric immunoassay known as SMART. This monoclonal

antibody-based test was developed for the detection of y. cholerae 01 (25,26) and 0139 (24)









strains in stool specimens. SMART is 95% sensitive and 100% specific to K. cholerae 01

strains (26) and 100% sensitive and 97% specific to 0139 strains (24). A one-step

immunochromatographic dipstick test for the detection of y. cholerae 01 and 0139 LPS in stool

samples was invented by the Institute Pasteur in Paris. This assay requires minimal technical

skill and rapidly detects thresholds of purified LPS at 10 ng/mL for y. cholerae 01 and at 50

ng/mL for y. cholerae 0139 strains in approximately 10 minutes (27). With continuing

improvement and use of methods for detection of y. cholerae, outbreaks could be lessened or

prevented with proper treatment and containment.

Escherichia coli 0157:H7

Escherichia coli 0157:H7 is an enterohemorrhagic serotype of E coli that causes

hemorrhagic colitis (HC) and hemolytic uremic syndrome (HUS). The bacterium is primarily

found on cattle farms and colonizes cattle, swine, and deer intestines with subclinical effects to

the animals (28). Disease in humans comes from the ingestion of beef, milk, vegetables, and

other products that are contaminated with E. coli 0157:H7. Symptoms from an E. coli 0157:H7

infection include mild diarrhea, abdominal pain, vomiting, bloody diarrhea, HC, strokes, and

HUS (28).

Over the past 23 years, 146 Shiga-like toxin producing E. coli (STEC) outbreaks and

sporadic cases of human illnesses have been traced to consumption of beef contaminated with

various E. coli 0157 strains (29). Most of these illnesses were caused by infection with E. coli

0157:H7. Of the 146 outbreaks and sporadic cases 89% occurred in the United States (29). The

large number of cases in the Unites States compared with the rest of the world can be explained

by high levels of beef consumption and availability of E. coli 1 57 diagnostic methods in the

United States. The Center for Disease Control and Prevention estimated that E. coli 0157:H7 is









responsible for approximately 73,000 illnesses and more than 60 deaths per year in the United

States (30).

Escherichia coli 0157:H7 produces Shiga toxins that are responsible for human disease.

These Shiga toxins produce severe cytopathic effects and have a high degree of homology with

Shiga-toxin (Stx) of \1hige//a~ dysenteriae type 1 (28). The Stx toxin is a member of the AB toxin

family. It is composed of a catalytically active A-subunit surrounded by a pentameric B-subunit.

The B subunits specifically bind to glycosphingolipid globotriosylceramide (Gb3) receptors (31)

of the renal glomerular endothelial, mesangial (32) and tubular epithelial cells (33). Upon entry

into the cell, the catalytically active Al subunit cleaves ribosomal RNA leading to the cessation

of protein synthesis and cell death (34). Not only does Stx toxin damage host cells, but it may

also increase the adherence of E coli 0157:H7 to epithelial cells leading to increased risk of

colonization. Tissue culture experiments showed that Stx toxin evoked an increase in a

eukaryotic receptor, nucleolin, that binds to the E. coli 0157 attachment factor, intimin, leading

to increased cell adherence (32).

There are two distinct toxin-converting bacteriophages (phages), 933J and 933W, in E. coli

0157:H7 that generate two genetically related toxins that are antigenically distinct but create

similar biologic effects (35). These two toxins are called Shiga toxins I and II (Stxl and 2).

Experimental studies suggest that E. coli that produces the Stx2 toxin is more virulent than

E. coli that produces Stxl or both Stxl and Stx2 toxins (36). Escherichia coli strains producing

Stx2 toxin are more frequently linked with the development of HUS than Stxl toxin-producing

strains (37,38). In mouse studies, the lethal dose of purified Stx2 toxin is 400 times lower than

that of Stxl toxin (39). In piglet studies, Stx2 toxin-producing E. coli strains caused more severe

neurologic symptoms than strains producing both Stxl and 2 toxins or only Stxl toxin (40).









Detection of E. coli 0157:H7 has become of increasing importance to the food industry in

the United States as outbreaks continue to occur. Traditional methods of detection ofE coli

0157:H7 involve plating and culturing, enumeration methods, biochemical testing, microscopy,

and flow cytometry. Other methods have been developed, including immunoassays (41),

immunomagnetic separations (42), nucleic acid probe-based methods based on hybridization and

polymerase chain reaction (PCR) (43), and DNA microarrays (44). However, many of these

assays are time-consuming and not suitable for rapid detection of E coli 0157:H7. Therefore,

biosensors have been developed for the rapid detection of E coli 0157:H7 cells and Stx toxins.

"An electrochemical biosensor is a self-contained integrated device, which is capable of

providing specific quantitative or semi-quantitative analytical information using a biological

recognition element (biochemical receptor) which is retained in direct spatial contact with an

electrochemical transduction element" (45). Some advantages of biosensors include their

continuous data acquisition ability, target specificity, fast response, mass produce feasibility, and

the simplicity of sample preparation. A quartz-crystal microbalance (QCM) has been developed

for detection of E coli 0157:H7 cells. Upon specif ic binding of E coli 0157:H7, the QCM uses

its ultra sensitive mass-measuring sensor to detect decreases in the crystal-resonance frequency

to enable detection of 2.0 x 102 CFU/mL of E coli 0157:H7 (42). An amperometric biosensor

for E. coli 0157:H7 cells made use of a dissolved-oxygen probe to enable detection of 50

CFU/mL in as little as 20 minutes of preparation and processing time. Upon binding of the

bacterial cells, a decrease in enzyme activity results in a change in oxygen concentration that was

detected with a Clark-type oxygen electrode probe (46). Polymerase chain reaction-based assays

are also very common for detection ofE. coli 0157:H7 cells (47-51) and Stx toxins (50).









Phage Display

Phage display was initially described in 1985 by George P. Smith (52) as a means to

display foreign proteins on filamentous bacteriophage. Filamentous phages are non-lytic phages

with circular ssDNA genomes. Of the filamentous phages, the Ff family (fl, fd, and M13)

phages infect F+ E. coli through binding with their F pilus. These phages are useful tools to link

genotype and phenotype of select recombinant proteins. This link was created by encoding a

foreign polypeptide in-frame with a coat protein gene of the M13 phage. Phage display could

theoretically be implemented with any phage, but filamentous phages have been the most widely

used. Of the filamentous phages, M13 is the most commonly used.

The M13 phage (Fig. 1-1) replicates in E. coli, turning the bacterium into a phage-

production factory (53). The bacteria harboring these phages do not lyse, but undergo reduced

cell growth due to the stress of phage production. The M13 phage contains a 6.4-kb, circular,

single-stranded DNA genome that encodes phage proteins I to XI. Five of these proteins are coat

proteins. The maj or coat protein (pXIII) is present in approximately 2,700 copies and protects

the genome in a cylindrical manner. The minor coat proteins pVII and plX are necessary for

efficient particle assembly, while the minor coat proteins pIII and pVI are necessary for particle

stability and infectivity (54-56). The pIII protein mediates the binding of the phage to the F pilus

and is necessary for viral uncoating and phage DNA transfer to the cytoplasm of the bacterium

(55). Host enzymes then convert the ssDNA into supercoiled dsDNA, known as the replicative

form (RF) (55). The RF is essential to the phage display system because it can be purified and

manipulated just like a plasmid. Through the manipulation of the RF of M13 came some of the

earliest cloning vectors (57). Through some of these vectors came the development of

recombinant phage libraries.









During assembly of M13, the foreign protein is fused to a coat protein and displayed on the

surface of the phage. The minor coat protein III (pIll) is the most common protein for fusions,

but the maj or coat protein XIII (pXIII) and the other minor coat proteins of M13 have also been

used for recombinant fusions (58,59). Phage displays using the pIII and pXIII proteins have

different advantages. Using the pXIII protein as the fusion protein enables high copy display of

the recombinant protein because there are over 2,700 copies of the pXIII protein on the surface

of the phage. However, a drawback in using the pXIII protein as the fusion protein is its

limitations in the size of the displayed protein (60). The pXIII protein can only display peptides

less than six amino acids in length before the function of the coat protein becomes compromised

and the number of infectious particles plummets. If the size of the display peptide increases to

eight amino acids, only 40% of the phages are infective; if the display peptide increases to 16

amino acids, less than 1% of the phages are infective. Recombinant fusions using the pIII

protein are not as restricted in the size of the display peptide (53). The pIll-fusion protein can

display peptides of 100 amino acids or greater before the ability of the pIII protein to bind to the

F pilus of E coli becomes compromised (52). Also, since there are only five copies of the pIII

protein on the surface of phage particles, the ability to select high affinity binding phage particles

is greater than that of the ability of the pXIII-fusion protein.

Recombinant Phage Libraries

Recombinant phage libraries are composed of phages that display a fused protein to a coat

protein of the phage. There are two main phage display libraries, phage and phagemid libraries.

Phage libraries are M13 with the addition of a recombinant fusion to the giII gene, while

phagemid libraries are just plasmids that contain a recombinant fusion to the giII gene. There

are advantages and disadvantages to both systems. Phagemid libraries are advantageous due to

their greater library diversity because of their higher ligation-transformation efficiency, and their









simplicity enables easier genetic manipulation than with phage vectors. Also, phagemid particles

isolated from phagemid libraries are able to be produced to generate more phagemid particles or

secrete the recombinant protein, which is usually the ultimate end product. A disadvantage of

phagemid particles is their dependence on the aid of a helper phage to provide the rest of the

M13 proteins in trans to enable phagemid particle production and assembly. Phage libraries are

advantageous due to their lack of dependence on helper phages. Because phage libraries are

M13 phage with some alterations, they are capable of propagating themselves by simple

infection ofE. coli. A disadvantage to phage libraries is that they are less stable than phagemid

libraries (61). Also, phage particles are only able to produce phages and have to be further

manipulated to be able to produce the recombinant protein alone.

Tomlinson I + J Human Synthetic VH + VL Phagemid Libraries

The Tomlinson libraries are naivev" libraries comprised of approximately 1 x 10s random

phagemids derived from non-immunized human donors (62). Naive libraries enable greater

diversification of antibody genes, increasing the probability of isolating phagemids specific to a

wide variety of targets. The Tomlinson libraries also encode greater diversity through random

side-chain diversification. These phagemid libraries encode a single chain F variable (scFv)

gene fusion to the giII gene of a library vector plasmid contained in a M13 phage. The scFv

(Fig. 1-2) is composed of a single polypeptide with VH and VL domains that are joined by a

flexible glycine-serine linker. The Tomlinson libraries were constructed by use of reverse

transcription and PCR to amplify the VH and VL antibody genes from B-lymphocytes of human

donors. Universal degenerate primers were then used to anneal to the 5' end of the exons

encoding the antibody V-gene, which is conserved in humans. The mRNA from the B

lymphocytes was converted to cDNA, which represents the VH and VL antibody genes. The

cDNA was then PCR assembled using an overlap extension technique and contained the









restriction enzyme sites for subcloning into the plT2 library vector. The library vector, plT2,

encodes an M13 origin of replication, an ampicillin resistance gene (bla), and both His6 and myc

tags (Fig. 1-3). The plT2 phagemids were electroporated into E. coli, where they were

superinfected with helper phage to generate phagemid particles displaying a scFv-pIII fusion

protein. The phagemid particles display the scFv on the surface of the phage particles and

encode the scFv gene in the phagemid genome, linking phenotype with genotype.

New England Biolabs 12mer Peptide Phage Library

The New England Biolabs 12mer Peptide Phage Library (Ph.D. System, New England

Biolabs) is a combinatorial phage library that encodes a random sequence of twelve amino acids

fused to the giII gene of the M13 genome. The 12 random amino acids are fused to the N-

terminus of the pIII protein, which is displayed on the surface of the M13 phage. The first

residue of the mature protein is the first randomized position. The peptide is followed by a (Gly-

Gly-Gly-Ser) spacer linked to the wild type pIII protein. The library is constructed in M13

phage with an insertion of the lacZa gene fragment into the genome. The insertion of the lacZa

gene fragment enables distinction of E coli that harbor phages from the library opposed to

environmental phages that do not contain the lacZa gene. This is done by blue/white screening

of phages on agar that contains X-gal and IPTG. When an E. coli strain has a functional lacZ

gene it will produce P-galactosidase, which is a heterodimer composed of an a and an 0Z peptide.

Neither of these peptides have enzymatic activity on their own; therefore, if one component is

missing then P-galactosidase enzymatic activity is lost and if they are complemented they

spontaneously combine to generate an active structure. Beta-galactosidase cleaves X-gal into a

product that becomes oxidized to generate an insoluble blue product. IPTG is an inducer of the

lac promoter which drives transcription of the lacZ gene. Therefore, if phages are produced in

E. coli cells with a nonfunctional lacZa gene and a functional lacZG1 gene and grown under









selection for the E. coli cells then bacteria containing a phage from the library, which encodes

lacZa, will have a complemented lacZ gene to enable distinction from bacteria containing a

phage from the environment. The NEB 12mer library consists of approximately 2.7 x 109

electroporated sequences that were amplified once to yield approximately 55 copies of each

sequence. Sequencing of 104 clones from the library yielded six clones (5.8%) that did not

contain a displayed peptide insert. Sequencing from the 98 other clones revealed a wide

diversity of sequences with no obvious positional biases.

Tools for Production of Phages from Phagemids

To propagate phagemid particles, an amplification tool must be supplied that encodes the

rest of the M13 genes necessary for phagemid particle production. These tools may be phages or

plasmids and are referred to as "helpers" because they help the phagemid particles propagate by

supplying the necessary phage genes in trans. Some often used phage amplification tools include

helper phage, hyperphage, phaberge phage, ex-phage, and helper plasmids.

Helper Phage

Helper phage is the most common helper tool used to produce phagemid particles. There

are many variations of helper phages, R408, VCSM13, and M13KO7 (63) that differ slightly. Of

the various helper phages M13KO7 is the most commonly used. It is a derivative of M13 that

has a couple differences including a kanamycin resistance gene and the Pl5A origin of

replication, which allows the genome to be replicated as a plasmid in E. coli.

Because helper phages are basically M13 phages, they supply all of the genes necessary for

production of phagemid particles. Just like M13, helper phages infect F E. coli through the

binding of the pIII protein to the F pilus. Helper phage amplifies phagemid particles to yield

average titers of 2 x 1010-12 phagemids/mL (64). However, because helper phage encodes a wild

type gHI gene, the phage particles produced will contain a mixture of wild-type pIII proteins and










pIll-fusion proteins. This heterogeneity in display of fusion proteins can result in progeny

phagemids bearing all wild type pIII proteins or only a monovalent display of the recombinant

pIII protein (65). The low level of display of recombinant pIII proteins results in low efficiency

of selection of recombinant phagemid particles. Because the helper phages will also be produced

and packaged, amplification of phagemids with helper phage generates a heterogeneous mix of

phage particles encoding phagemids and helper phage genomes. The number of helper phages

produced can sometimes be greater than or equal the number of phagemid particles generated

(64). To get around the problems of helper phage, many alterations have been implemented to

improve helper phage.

Helper Phage with gHIlMutations

There are many variations to helper phage. One of the key variations to helper phage is the

deletion or mutation of the giII gene in the helper phage genome, yielding hyperphage, Ex-

phage, and Phaberge phage. All three gIll-mutated helper phages are derivatives of M13KO7

that lack full giII gene functionality but still possess pIII proteins. These pIII proteins enable the

helper phages to bind to the F pilus ofE. coli and transfer their helper phage genomes into the

E. coli. Hyperphage contains a partial deletion of the giII gene, while Ex-phage and Phaberge

phage have amber stop codons within the giII gene. Ex-phage contains two amber stop codons,

and Phaberge phage contains only one. The mutations in the giII gene of the helper phages

promote multivalent display of fusion proteins, which enhances the avidity of binding of

phagemid particles to the target molecule. This increased avidity is desirable because it

increases the chances of selecting positive clones.

In comparing the antigen-binding activity of phagemid particles produced using M13KO7,

hyperphage-produced phagemid particles have over 400-fold increased activity (66), Ex-phage-

produced phagemid particles have over 100-fold increased activity (65), and Phaberge-phage-









produced phagemid particles have over 5 to 20-fold increased activity (67). Of the three giII

gene mutated helper phages, hyperphage has the highest display of recombinant fusions. This is

most likely due to partial read through of the amber stop codons in non-suppressor strains with

Ex-phage or Phaberge phage. This partial readthrough generates display of wild-type and

fusion-pIII proteins. Hyperphage does not have any readthrough of the wild-type giII gene

because of the partial deletion of the giII gene, ensuring that all pIII proteins are recombinant

fusions. Hyperphage not only has a higher display level of fusion-pIII proteins, but it also

packages phagemid particles over 100 times more efficiently than Ex-phage or Phaberge phage

(68).

Of the giIl gene-mutated helper phages, hyperphage is the most advantageous for use in

phage display systems. Hyperphage enables multivalent display of pIII fusions, which makes it

ideal in phagemid particle production. However, an issue with hyperphage is that the phagemid

particle stocks generated using hyperphage have titers of 109-10 phagemids/mL (64). This is a log

or two lower than phagemid particles produced using helper phage. However, many of the

phagemid particles generated by helper phage are useless because they bear no recombinant pIII

proteins. Therefore, even though the quantity of phagemid particles produced with hyperphage

is lower than those produced with helper phage, the quality of phagemid particles produced is

much higher.

Helper Plasmids

Helper plasmids (64) are M13-based plasmids used for phagemid particle amplification

that are engineered in three forms to overcome many disadvantages of helper phage and

hyperphage. The plasmids are M13mpl9 with a chloramphenicol resistance gene cloned in from

pBSL121 to allow for selection ofE. coli containing the helper plasmids. The M13 origin of

replication was deleted and replaced with the pl5a origin of replication from pMPM-K3. This









eliminates the ability of helper plasmids to be packaged in progeny phage particles, resulting in

progeny phages that contain the phagemids but not the helper plasmids. The giII gene of the

helper plasmid was partially deleted or fully deleted to yield three helper plasmids with varying

lengths of their giII gene.

These three helper plasmids are maintained in E. coli and provide phagemid particles with

all of the necessary structural proteins for phagemid particle amplification. One of the

advantages of the helper plasmids is that they have a full (M13cp), deleted (M13cp-dg3), or

truncated (M13cp-CT) giIl gene, which enables monovalent to multivalent display of

recombinant proteins. Multivalent display phagemid particles possess high avidity binding

ability, while monovalent display phagemid particles possess high affinity binding ability. Thus,

if high affinity monovalent display phagemid particles are desired, M13cp would be used. If

high avidity multivalent display phagemid particles are desired, M13cp-dg3 or M13cp-CT might

be used. An additional advantage is that the helper plasmids are maintained in E. coli. This

negates the need for superinfection, removing the limiting factor of the number of helper phages

or hyperphages needed to amplify phagemid particles.

This thesis describes the efforts to optimize biopanning processes and associated reagents

to improve the likelihood of isolating recombinant phages that specifically bind to biological

agents or their products. The recombinant phages that specifically bind to biological agents or

their products will eventually be used to enable detection of the agents.

The specific aims for this study are:

1. To optimize extraction methods and determine saturation limits of y. cholerae 01
LPS and use these methods to pan phage display libraries to y. cholerae LPS.

*A phenol-water method was a more effective extraction method than TRIzol Reagent
method for the extraction of y. cholerae 01 LPS and 10-100 times more LPS could be
bound to nitrocellulose paper as opposed to a microtiter well. Biopannings against









K~ cholerae 01 LPS that was immobilized onto nitrocellulose paper or a microtiter well
failed to yield phages that recognized y. cholerae 01 LPS.


2. To optimize panning and screening procedures to allow for a more efficient
biopanning process that will be more likely to isolate and detect specific recombinant
phagemid particles.

* In biopanning with the Tomlinson scFv phagemid library, a trypsin elution time of 10-30
minutes eluted the most phagemid particles. Screening scFv proteins by a high throughput
ELISA was an acceptable screening process and enabled the possible screening of
hundreds of clones per day. Screening phagemid particles by a high throughput ELISA
was not an acceptable screening process because phagemid particles could not be produced
in high enough titers in a microtiter well to make a high throughput ELISA screen
effective.


3. To optimize the production of phagemid particles to ensure high quality and high
quantity of phagemid particles.

* Phagemid particle amplification by E. coli harboring hyperphage or helper plasmids was
not an effective method to produce phagemid particles. Homemade hyperphage that was
produced in E. coli MGl655 (pGTR203) enabled high quality and high quantity
production of hyperphage. With homemade hyperphage, the Tomlinson J library was
effectively amplified to ensure that the redundancy of the clone population in the library
was maintained.


4. To isolate specific recombinant phagemid particles to E. coli 0157:H7 Stx2 toxin
using the optimization techniques discovered in previous aims.

* Using the improved homemade hyperphage and the homemade hyperphage-amplified
Tomlinson J scFv library to pan against a commercial Stx2 toxin preparation resulted in
isolation of phagemid particles that recognized the Stx2 toxin preparation.










plX


Figure 1-1. Structure of M13 phage. The M13 phage particle contains a ssDNA, circular, 6.4-kb
genome that encodes genes I-XI and the M13 ori. The M13 phage has five coat
proteins: the maj or coat protein (pXIII) that is present in 2,700 copies and the minor
coat proteins (pIll, pVI, pVII, and plX) that are present in 5 copies each.









mAb Fab scFv


VVH



CH2~ Linker
cH3,

Figre1-. trctresofanibdis.m~ (mnoloalanibdy,Fa(fgmnatie






Figrligh Srutu chain).Thies heav andightchainsomAb antbdy) Fab fantiodes arehedtogethe

by disulfide bonds, and the heavy and light chains of the scFv antibody are fused
together by a flexible glycine-serine linker.











lac promoter
colE1 ori RBS




V r a l evpe/B le ad e r

ScVariable Heavy
ba plT2 Linker

rnyc tag
amber stop codon

M113 ori
glnt



Figure 1-3. Genetic map of plT2 phagemid vector from the Tomlinson scFv library. RBS-
ribosome binding site. pelB leader peptide sequence promotes export of the scFv
protein. Variable Heavy and Variable Light peptide sequences are fused together by
a glycine-serine linker. An amber stop codon is at the junction of the c-myc tag and
the gIII gene to enable conditional expression of the scFv-pIII fusion in an amber
suppressor strain. The M13 origin of replication enables packaging into M13 phage
particles, the bla gene encodes ampicillin resistance, and the colEl origin of
replication enables maintenance as a plasmid in E. coli.









CHAPTER 2
MATERIALS AND METHODS

Bacterial Strains, Media, and Growth Methods

The bacterial strains used and their genotypes are listed in Table 2-1. All E. coli strains

were grown in LB broth (10 g tryptone, 5 g yeast extract, 5 g NaC1, and 3 mL of 1 M NaOH in 1

L water) or on LB agar plates containing 1.5% (w/v) agar. Exceptions are E. coli TGl, E. coli

DH~a (pNR100), and E. coli ER2738. Escherichia coli TG1 was grown in 2xTY broth (16 g

tryptone, 10 g yeast extract, 5 g NaC1, and 3 mL of 1 M NaOH in 1 L water) or on 2xTY agar

plates containing 1.5% (w/v) agar. Escherichia coli DH~a (pNR100) was grown in LB broth

containing 100 Cpg/mL ampicillin or on LB agar plates containing 100 Cpg/mL ampicillin and

1.5% (w/v) agar. Escherichia coli ER273 8 was grown in LB broth containing 20 Cpg/mL

tetracycline or on LB agar plates containing 20 Cpg/mL tetracycline and 1.5% (w/v) agar. All

strains of y. cholerae were grown in Luria Bertani broth containing with physiological saline

(LB-N) (10 g tryptone, 5 g yeast extract, 8.5 g NaC1, and 3 mL of 1 M NaOH in 1 L water) or on

LB-N agar plates containing 1.5% (w/v) agar. The most common medium used was 2xTY

containing 100 Cpg/mL ampicillin and 1% (w/v) glucose; therefore, it was abbreviated as 2xTY

AG. Escherichia coli strains harboring hyperphage were grown in broth or on plates as

specified, with the addition of 40 Cpg/mL kanamycin. Escherichia coli strains harboring helper

plasmids were grown in broth or on plates as specified above, with the addition of 30 Cpg/mL

chl orampheni col .

All bacterial cultures were initially grown as a standing-overnight culture. A standing-

overnight culture was made from 10 mL of specified medium that was inoculated with bacteria

from an agar plate. The standing-overnight culture was grown overnight (~16 h) in a 370C

incubator. A log-phase culture was obtained by diluting a standing-overnight culture 1:40 in









medium and incubating the culture in a 370C incubator with shaking until the optical density at

600 nm (OD600) was between 0.4 to 0.6.

The Tomlinson J Human Synthetic VH + VL phagemid library was constructed by

Medical Research Council, Cambridge, U.K. (Human Single Fold scFv Libraries I + J

(Tomlinson I + J). 2002. Cambridge, UK, MRC Laboratory of Molecular Biology, MRC Centre

for Protein Engineering.) and was obtained from the Interdisciplinary Center for Biotechnology

Research Hybridoma Core, University of Florida. The Ph. D. 12mer peptide library was

obtained from New England Biolabs (NEB, Ipswich, MA). The hyperphage (66) used to amplify

phagemid particles was obtained from Progen Biotechnik (Heidelberg, Germany).

Biopanning of Phage Display Libraries

Panning on Immunotubes

A polystyrene immunotube (Nunc, Rochester, NY) was coated with 2 mL of 50 Cpg/mL

Shiga-like toxin 2 (Stx2 (Toxin Technology, Sarasota, FL)) in phosphate-buffered saline (PBS)

(Cellgro, Manassas, VA) (137 mM NaC1, 2.7 mM KC1, 4.3 mM Na2HPO4, 1.4 mM KH2PO4, pH

7.3). The immunotube was rotated overnight on a labquake at 40C. The next day the tube was

washed three times with 4 mL of PBS to remove non-immobilized toxin. The tube was filled

with casein blocker (Pierce, Rockford, IL) containing 0.05% (v/v) Tween-20 and incubated for

two hours at room temperature (~250C). Excess blocker was removed by rinsing the tube three

times with 4 mL of PBS. The Tomlinson J library (for round one of panning) or amplified eluted

phages (for subsequent rounds of panning) was added to the immunotube at a concentration of 1

x 1012 phages in 4 mL of casein blocker (Pierce) containing 0.05% (v/v) Tween-20. The library

was incubated for one hour at room temperature on a labquake, followed by a one hour standing

incubation. The unbound library was removed by aspiration, and the weakly bound phages were

removed by washing the tube 10 times for round one and 20 times for rounds two and three with









4 mL of PBS containing 0.05% (v/v) Tween-20 (PBS-0.05T). The excess PBS-0.05T was

aspirated, and bound phages were eluted with 0.5 mL of trypsin-PB S (10% (v/v) trypsin stock

(10 mg/mL trypsin (Type XIII from Bovine Pancreas) (Sigma-Aldrich, St. Louis, MO), 50 mM

Tris-HCI (pH 7.4), 1 mM CaCl2 in water) in PBS) for 10 minutes at room temperature on a

labquake. Ten microliters of the eluted phages were immediately titered, while the rest of the

eluted phages were infected into 5 mL of E coli TG1 at an OD600 Of 0.4. The infected E. coli

was incubated for 30 minutes at 370C in a standing water bath. Following incubation, the

infected cells were isolated by centrifugation at 13,776 x g for 10 minutes at room temperature.

Cells were suspended in 0.6 mL of 2xTY medium and plated on three 2xTY AG plates to

amplify the phagemid-containing E. coli. The plates were incubated overnight at 370C. The

next day 1.5 mL of 2xTY was added to the lawns of bacteria, and the bacterial lawns were

scraped from the plate in the medium and pooled together. Of the pooled bacteria, 50 CLL was

inoculated into 100 mL of 2xTY AG medium and grown in a 370C shaking incubator until the

OD600 WAS 0.4. Once the OD was reached, 10 mL of the culture was superinfected with

homemade hyperphage at a MOI of 20 and incubated for 30 minutes in a standing 370C water

bath. The superinfection was centrifuged at 13,776 x g for 10 minutes at room temperature. The

resulting pellet was suspended in 100 mL of 2xTY containing 100 Cpg/mL ampicillin, 40 Cpg/mL

kanamycin, and 0.1% (w/v) glucose and incubated overnight in a shaking 300C water bath. The

next day the culture was centrifuged at 13,776 x g for 10 minutes at 40C. Phage-containing

supernatant was precipitated with final concentrations of 13 mM Polyethylene Glycol (PEG) and

0.55 M NaC1. PEG and NaCl were dissolved in the supernatant at room temperature and then

the PEG-containing supernatant was incubated overnight at 40C. The next day the PEG-

precipitated supernatant was centrifuged at 13,776 x g for 20 minutes at 40C. The pellet was









suspended in 2 mL of PBS and centrifuged at 13,776 x g for 5 minutes at room temperature to

remove bacterial debris. The supernatant was transferred to a new tube to be titered and stored at

40C until used in further rounds of panning or screening.

Panning in Suspension

Vibrio cholerae N16961 was grown to log phase, and 10' to 109 CellS were centrifuged at

10,621 x g for 5 minutes at room temperature. The bacterial pellet was suspended in 1 mL of

PBS and transferred to a microcentrifuge tube. The Tomlinson J library (for round one of

panning) or amplified eluted phages (for subsequent rounds of panning) was added to the

bacterial suspension at a concentration of 1 x 10l phages in 0.5 mL ofPBS. The library and

bacterial cells were incubated for 3 hours at 40C while rotating on a labquake. The cell

suspension was centrifuged at 10,621 x g for 5 minutes at room temperature. The supernatant,

containing the unbound phage library, was stored at 40C. The bacterial pellet containing the

bound phages was suspended in 0.5 mL of PBS to wash off phages that were weakly bound to

the bacteria. The suspension was centrifuged at 10,621 x g for 5 minutes at room temperature.

The supernatant was discarded, and the pellet was suspended in 0.5 mL of PB S for a second

wash. The suspended cells were transferred to a new microcentrifuge tube and centrifuged at

10,621 x g for 5 minutes at room temperature. The supernatant was aspirated, and phages bound

to the bacteria were eluted with 0.5 mL of trypsin-PBS for 10 to 60 minutes at room temperature

on a labquake. Ten microliters of the elution was immediately titered. The rest of the eluted

phages were infected into 5 mL of E coli TG1 at an OD600 Of 0.4. The infected E. coli were

incubated for 30 minutes in a 370C standing water bath. Following incubation, the infected cells

were isolated by centrifugation at 13,776 x g for 10 minutes at room temperature. The cells were

suspended in 0.6 mL of 2xTY medium and plated on three 2xTY AG plates to amplify the

phagemid-containing bacteria. The plates were incubated overnight at 370C. The next day 1.5









mL of 2xTY was added to the lawns of bacteria, and the bacterial lawns were scraped from the

plate in the medium and pooled together. Of the pooled bacteria, 50 CLL were inoculated into 50

mL of 2xTY AG medium and grown in a 370C shaking incubator till the OD600 WAS 0.4. Once

the OD was reached, 10 mL of the culture was superinfected with homemade hyperphage at a

MOI of 20 and incubated for 30 minutes in a 370C standing water bath. The superinfection was

centrifuged at 13,776 x g for 10 minutes at room temperature. The resulting pellet was

suspended in 50 mL of 2xTY containing 100 Cpg/mL ampicillin, 40 Cpg/mL kanamycin, and 0. 1%

(w/v) glucose and incubated overnight in a 300C shaking water bath. The next day the culture

was centrifuged at 13,776 x g for 10 minutes at 40C. The phage-containing supernatant was

precipitated by slowly adding 12.5 mL of PEG solution (20% (w/v) PEG and 2.5 M NaCl in

water) to the supernatant with continuous swirling. The PEG-containing supernatant was

incubated overnight at 40C. The next day the PEG-containing supernatant was centrifuged at

13,776 x g for 20 minutes at 40C. The pellet was suspended in 1 mL of PBS and centrifuged at

10,261 x g for 5 minutes at room temperature to remove bacterial debris. The supernatant was

transferred to a new tube to be titered and stored at 40C until used in further rounds of panning or

screening.

Panning on Microtiter Wells

A polysorp microtiter well (Nunc) was coated with 150 pIL of 100 Cpg/mL K. cholerae 569

lipopolysaccharide (LPS) (Sigma-Aldrich) in carbonate coating buffer (0. 1 M NaHCO3, 0.02%

(w/v) NaN3 (pH 8.6)). Prior to coating, the LPS was sonicated by a Vibra Cell (Sonics &

Materials Inc., Danbury, CT) for one minute. The microtiter well was coated overnight at 40C in

a humid chamber. The next day the microtiter well was equilibrated to room temperature for 15

minutes. The LPS was aspirated and blocked with 200 CLL of carbonate blocking buffer for one

hour at 40C. The well was washed six times with 200 CLL of 0. 1 M Tris-buffered saline (TBS)









(81 mM Tris-HC1, 20 mM Tris-Base, 154 mM NaC1, pH 7.5) containing 0.1% (v/v) Tween-20

(TB S-0. 1T) for the first round of panning and with 200 CLL of 0. 1 M TBS containing 0.5% (v/v)

Tween-20 (TB S-0.05T) for subsequent rounds of panning. The washed wells were incubated

with 2 x 1011 PFU of the Ph.D.-12mer peptide library (NEB) (for round one of panning) or

amplified phages (for subsequent rounds of panning) diluted in 100 CLL of TBS-0. 1T for one hour

at room temperature with gentle rocking. The unbound phages were aspirated and stored at 40C.

The wells were washed 10 times with TBS-0. 1T for the first round of panning and with 200 CLL

of TBS-0.05T for subsequent rounds of panning. The phages were eluted with 100 CLL of 0. 1 M

glycine (pH 2.2) for 10 minutes at room temperature with gentle agitation. The glycine

containing the eluted phages was transferred to a microcentrifuge tube and neutralized with 15

CLL of Tris-HCI (pH 9.0). Five microliters of the eluted phages were immediately titered. One

hundred and ten microliters of the remaining elution was added to 20 mL of LB diluted 1:100

with a standing-ovemnight culture of E coli ER2738. The culture was grown for 4.5 hours in a

370C shaking incubator. The turbid culture was centrifuged at 13,776 x g for 10 minutes at 40C.

The phages were precipitated by slowly adding 3.3 mL of PEG solution (20% (w/v) PEG and 2.5

M NaCl in water) to the phage-containing supernatant with continuous swirling. The PEG-

containing supernatant was incubated overnight at 40C. The next day the PEG-containing

supernatant was centrifuged at 13,776 x g for 20 minutes at 40C. The pellet was suspended in 1

mL of 0. 1 M TBS and centrifuged at 10,621 x g for 5 minutes at room temperature to remove

bacterial debris. The phages were precipitated a second time by slowly adding 167 CLL of PEG

solution to the supernatant with continuous swirling. The PEG-containing supernatant was

incubated one hour at 40C and centrifuged at 13,776 x g for 15 minutes at 40C. The pellet was

suspended in 200 CLL of 0. 1 M TBS to be titered and stored at 40C until used in further rounds of









panning or screening. Five rounds of panning were done with screening of clones after the fifth

round of panning.

Panning on Nitrocellulose Paper

Six pieces of nitrocellulose paper (Bio-Rad) were cut to the surface-area dimensions of a

microtiter well (5 mm x 20 mm). One and a half milliliters of approximately 340 Cpg/mL

K~ cholerae 5 69 LP S (phenol -water-extracte d) i n PB S wa s added to si x mi croc entri fuge tub es,

each containing a strip of nitrocellulose paper. Prior to coating, the LPS was sonicated by a

Vibra Cell (Sonics & Materials Inc.) for one minute. The nitrocellulose papers were coated

overnight at 40C on a rotating labquake. The next day two pieces of LPS-coated nitrocellulose

paper were each transferred to a new microcentrifuge tube and washed five times with 1 mL of

PBS on a labquake with five minutes per wash. Washed nitrocellulose strips were transferred to

new microcentrifuge tubes and blocked for one hour in PBS casein blocking blocker (Pierce) on

a labquake at room temperature. Blocked nitrocellulose strips were transferred to new

microcentrifuge tubes and washed five times with 1 mL of PBS on a labquake with five minutes

per wash. Each of the washed pieces of nitrocellulose papers was transferred to a new

microcentrifuge tube containing 1.5 x 10"1PFU of the Ph.D.-12mer peptide library (NEB)

diluted in 1 mL of PBS. Nitrocellulose strips were incubated with the library for one hour at

room temperature on a labquake. After one hour, the panned nitrocellulose strips were

transferred to new microcentrifuge tubes and washed five times with 1 mL of PB S on a labquake

with five minutes per wash to remove the unbound library. Washed nitrocellulose strips were

transferred to new microcentrifuge tubes to be eluted. One piece of nitrocellulose was acid

eluted by adding 250 CLL of 0. 1 M glycine (pH 2.2) to the strip and incubating it for 10 minutes at

room temperature on a labquake. The acid eluted nitrocellulose paper was removed to a new

microcentrifuge tube, and the glycine solution containing the eluted phages was neutralized with










37 pIL of 1 M Tris-HCI (pH 9.0). The acid eluted phages were stored at 40C until used for

amplification. The other piece of panned nitrocellulose paper was antigen eluted by adding 250

CLL of approximately 340 Cpg/mL of y. cholerae 569B LPS diluted in PBS to the piece of

nitrocellulose paper and incubated one hour at room temperature on a labquake. The piece of

nitrocellulose paper was removed, and the remaining LPS solution containing the eluted phages

was acid eluted as described above. Five microliters of each eluted phage solution was titered,

and the rest of the two elutions were each added to 20 mL of LB diluted 1:100 with a standing-

overnight culture ofE. coli ER2738. The cultures were grown for 4.5 hours in a 370C shaking

incubator. The turbid cultures were centrifuged at 13,776 x g for 10 minutes at 40C. Five

microliters of the amplified phages were titered. The rest of the phages were PEG precipitated

by slowly adding 3.3 mL of PEG solution to the phage-containing supernatant with continuous

swirling. The PEG-containing supernatant was incubated overnight at 40C. The next day the

PEG-containing supernatant was centrifuged at 13,776 x g for 20 minutes at 40C. The pellet was

suspended in 1 mL of PBS and centrifuged at 13,776 x g for 5 minutes at room temperature to

remove bacterial debris. The supernatant were transferred to new microcentrifuge tubes, titered,

and stored at 40C.

After the first round of panning, the amplified eluted phages were negatively panned on.

Two pieces of nitrocellulose paper (5 mm x 20 mm) were blocked for one hour in 1 mL of PB S

casein blocking buffer (Pierce) on a labquake at room temperature. The blocked nitrocellulose

strips were washed three times with 1 mL of PBS on a labquake at room temperature with five

minutes per wash. One piece of blocked nitrocellulose paper was added to the acid-eluted-

amplified phages from the round one panning and incubated one hour on a labquake at room

temperature. The other strip of blocked nitrocellulose paper was added to the LPS and acid-









eluted-amplified phages from the round one panning and incubated one hour on a labquake at

room temperature. The negatively panned pieces of nitrocellulose paper were removed, and the

remaining amplified-eluted phages were used for two more rounds of panning. After the third

round of panning the eluted phages were screen by ELISA.

Spot Titer of Phages

Escherichia coli TG1 was grown to log phase and centrifuged at 13,776 x g for 10

minutes at 40C. The resulting pellet was suspended in PBS to yield an E. coli TG1 concentration

of 1 x 1010 CFU/mL. One hundred microliters of the concentrated E. coli TG1 was spread on a

2xTY AG plate and set to dry for 30 seconds. Serially diluted phages in phosphate-buffered

saline containing 0.1% (w/v) gelatin (BSG) was dropped onto the plate in 10 CLL drops. Once the

drops dried, the plate was incubated overnight at 370C. The next day the colonies were counted,

and the approximate titer was calculated.

Spread Titer of Phages

Phages were serially diluted in BSG. One hundred microliters of the diluted phages were

added to 0.9 mL of log phase E. coli TGl. The infected E. coli TG1 was incubated 20 minutes in

a 370C standing water bath. One hundred microliters of the infections were plated on 2xTY AG

plates. The plates were incubated overnight at 370C. The next day the colonies were counted,

and the approximate titer was calculated.

Amplification of Phages

A colony of E coli TG1 containing a phagemid was picked from a plate with a sterile

toothpick and swirled in 3 mL of 2xTY AG medium. The culture was grown overnight at 370C.

The overnight culture was diluted 1:20 in 3 mL of 2xTY AG medium and grown to log phase in

a 370C shaking incubator. Three hundred microliters of the log-phase culture was infected with

hyperphage at a MOI of 10 and incubated in a 370C standing water bath for 30 minutes. The









culture was added to 30 mL of 2xTY medium containing 100 Cpg/mL ampicillin, 40 Cpg/mL

kanamycin, and 0.1% (w/v) glucose. The culture was grown overnight in a 300C shaking

incubator. The turbid culture was centrifuged at 13,776 x g for 10 minutes at 40C. The phage-

containing supematant was titered and stored at 40C.

High Throughput Production of Soluble Antibody Fragments (scFv antibodies)

A colony of E. coli HB2 1 51 containing a phagemid was picked from a plate with a sterile

toothpick and swirled in 200 CIL of 2xTY containing 100 Cpg/mL ampicillin, 0. 1% (w/v) glucose,

and 1 mM Isopropyl P-D-1 -thiogalactopyranoside (IPTG) in a 96-well polystyrene microtiter

plate (Coming, Coming, NY). The plate was incubated overnight at 370C. The next day the

plate was centrifuged at 4,667 x g for 10 minutes at 200C. Supernatants were analyzed by

ELISA.

Deoxyribonucleic Acid Manipulations

Plasmid Extractions

Plasmid extractions for cultures of 3 mL were performed with the QIAprep Spin Miniprep

kit (Qiagen, Germantown, MD), while cultures of 100 mL or greater were performed with the

Plasmid Midi kit (Qiagen). Extraction procedures were performed as directed in the instruction

manual .

Agarose Gel Electrophoresis

Deoxyribonucleic acid samples from 1-5 CLL were added to 2 CLL of 10 x Gel Loading

Buffer (Invitrogen, Carlsbad, CA). The samples were loaded onto a 0.7% (w/v) agarose gel using

Tris-borate-EDTA buffer (89 mM Tris, 89 mM boric acid, 2 mM EDTA) containing 10 Clg/mL

ethidium bromide. The gel was electrophoresed at 100 V until the loading dye was two-thirds

down the gel. The DNA bands were analyzed on a Gel Doc XR (Bio-Rad, Hercules, CA).









Electroporation of Plasmids

Electroporation was performed to transform plasmids into E. coli. DNA (3 0-1000 ng/CLL)

was added to 50 CLL of electrocompetent cells (see below) in a microcentrifuge tube. The tube

was flicked to mix the DNA with the cells. Once the DNA and cells were mixed, they were

transferred to a chilled 0.1 cm electroporation cuvette and electroporated at 1.25 kV/cm with a

MicroPulser (Bio-Rad). Nine hundred and fifty microliters of 2xTY medium was immediately

added to the transformation and transferred to a small culture tube where it was incubated in a

370C static water bath for 1 hour. The electroporation was serially diluted in BSG and plated on

appropriate plates. The plates were incubated overnight at 370C. The next day the colonies were

counted, and the number of transformations was calculated.

Electrocompetent Cells

Electrocompetent cells were made by diluting a standing overnight culture 1:100 in 1 L of

medium containing 1% (w/v) glucose and appropriate antibiotics. The culture was incubated in a

370C shaking incubator until the OD600 WAS between 0.4 and 0.6. All centrifuge rotors, wash

buffers, and centrifuge tubes that were used were pre-chilled to 40C. Once the OD600 WAS

between 0.4 and 0.6, the culture was chilled on ice in a 40C cold room for 45 minutes or more

until the culture was ~40C. The culture was centrifuged at 13,776 x g for 10 minutes at 40C, and

the supernatant was decanted. The pellets were suspended in 1.2 L chilled ddH20 by vortexing

or pipeting until the cell suspension was homogeneous. The suspended pellets were centrifuged

to pellet the bacterial cells. The supernatant was decanted, and the pellets were suspended in 400

mL of chilled ddH20. The bacteria were pelleted by centrifugation, and the supernatant was

decanted. The pellets were suspended in 200 mL of chilled ddH20 into one centrifuge tube. The

bacteria were pelleted by centrifugation, and the supernatant was decanted. The pellet was

suspended in 25 mL of chilled ddH20 and transferred to a 35 mL centrifuge tube. The bacteria









were pelleted by centrifugation, and the supernatant was decanted. The pellet was suspended in

25 mL of chilled 10% (w/v) glucose in water and centrifuged to pellet the bacteria. The

supernatant was decanted, and the bacteria were suspended in 0.5-1 mL of chilled 10% (w/v)

glucose in water. The suspended bacteria were either used directly for transformation or

aliquoted and frozen on dry ice in 95% (v/v) EtOH and stored at -800C.

Enzyme-Linked ImmunoSorbent Assays (ELISAs)

Enzyme-linked immunosorbent assays were performed for the screening and

characterization of phage particles, scFvs, LPS, bacterial protein preparations, monoclonal

antibodies, and polyclonal antibodies. The coating antigen or antibody was diluted in PBS

(Cellgro) or carbonate/bicarbonate buffer (15 mM Na2CO3 and 35 mM NaHCO3 in water), and

100 CIL was loaded onto a 96-well polystyrene flat-bottomed microtiter plate (Becton-Dickinson,

Franklin Lakes, NJ). Concentrations used of whole cells (10s tu/mL), LPS (1 Cpg/mL), toxin (10

Cpg/mL), and proteins (10 Cpg/mL) were kept relatively consistent. The coated microtiter plate was

incubated overnight in a humid chamber at 40C. The next day the microtiter plate was

equilibrated to room temperature for 30 minutes. Wells were aspirated with an ELx 800 Strip

Washer (Bio-Tek, Winooski, VT) and washed once with 300 CLL of PBS. If the coating antigen

was whole cells, LPS, or toxin, then the coating antigen was removed by vacuum or pipet and

sterilized. Two hundred microliters of PBS casein blocking buffer (Pierce) containing 0.05%

(v/v) Tween-20 was added to each well and incubated two hours at room temperature. Blocker

was aspirated by the ELx 800 Strip Washer (Bio-Tek). One hundred microliters of primary

antibody diluted in PBS casein blocking buffer (Pierce) containing 0.05% (v/v) Tween-20 was

added to each well and incubated one hour at room temperature. Concentrations used of

phagemid particles were 107 to 109 tu/mL and antibodies were 1 to 10 Cpg/mL. The primary

antibody was aspirated and washed three times with 300 CLL of PBS-0.05T by the ELx 800 Strip









Washer (Bio-Tek). One hundred microliters horseradish peroxidase (HRP) conjugated

secondary antibodies were diluted in PBS casein blocking buffer (Pierce) containing 0.05% (v/v)

Tween-20 and added to each well and incubated 30 minutes at room temperature. The secondary

antibody was aspirated and washed three times with 300 CLL of PBS-0.05T by the ELx 800 Strip

Washer (Bio-Tek). Substrate was prepared by dissolving one capsule of phosphate-citrate (PC)

buffer (0.05 M phosphate-citrate buffer (pH 5.0), 0.03% (w/v) sodium perborate) (Sigma-

Al dri ch) in 100 mL of water. A ten milligram tablet of 3, 3', 5, 5'-tetramethylb enzi dine

substrate (Sigma-Aldrich) was added to 10 mL of the PC buffer to give a final concentration of 1

mg/mL. Two hundred microliters of substrate was added to each well and incubated for 30

minutes at room temperature. The plate was read in an ELx 800 UV plate reader (Bio-Tek) at

630 nm. The data were analyzed with KcJunior (Bio-Tek) software.

Infection Efficiency

Infection efficiency experiments were done to compare the infection efficiency of

phagemid particles to bacterial strains harboring helper plasmids to the same strains not

harboring helper plasmids. Escherichia coli standing-overnight cultures were diluted 1:40 in 3

mL of medium containing 1% (w/v) glucose for strains containing no helper plasmids and in

medium containing 1% (w/v) glucose and 30 Cpg/mL chloramphenicol for strains containing

helper plasmids. The diluted cultures were grown in a 370C shaking incubator till the OD600 WAS

between 0.4 and 0.6. Once the desired OD600 WAS obtained, 3 x 107 bacteria were infected with

phage at a MOI of 0. 1 and incubated for 20 minutes in a 370C standing water bath. The

transductions were titered using the spread-titer method. Transduced E. coli strains containing

helper plasmids were titered on 2xTY AG plates containing 30 Cpg/mL chloramphenicol.

Transduced E. coli strains containing no helper plasmids were titered on 2xTY AG plates.









Protein and Lipopolysaccharide (LPS) Manipulations

Phenol-Water Extraction of LPS

A phenol-water extraction (76) was performed to extract y. cholerae 569B LPS from

whole cells. One and a half liters of y. cholerae 569B was grown to log phase in a 370C shaking

incubator. Bacteria were pelleted at 13,776 x g for 10 minutes at 40C and suspended in 10 mL of

1% (w/v) NaCl in water. The bacteria were centrifuged and suspended in another 10 mL of 1%

(w/v) NaCl in water. Next, the bacteria were centrifuged and suspended in 10 mL of ddH20.

This suspension was warmed to 700C, along with 24 mL of 90% (w/v) phenol (made directly

before extraction). Ten milliliters of warmed 90% (w/v) phenol was added to the 10 mL cell

suspension and vortexed. The twenty milliliter suspension was warmed in a 700C water bath for

20 minutes while vortexing frequently. The phenol suspension was transferred to an ice bath

where it was swirled in ice water for 5 minutes. The phenol suspension was centrifuged at 2,284

x g for 25 minutes at 40C. The aqueous phase of the phenol extraction was transferred to a new

centrifuge tube and stored on ice. The same volume of water as that extracted was added back

into the phenol suspension tube and heated to 700C for 20 minutes with frequent vortexing. The

extraction step was repeated, and the two aqueous phase extractions were combined into the

same tube. Aqueous extractions were placed in a 700C water bath for 15 minutes. Twelve

milliliters of warmed 90% (w/v) phenol was added to the warmed aqueous extraction and

incubated at 700C for 10-15 minutes with frequent vortexing. Two back extractions were

performed on the aqueous extractions. The final extraction was dialyzed (MWCO: 3,500) for

24-48 hours at 40C. The dialyzed extraction was digested with DNase (Qiagen) (20 Clg

DNase/mL dialysis volume) and RNase (Qiagen) (40 Clg RNase/mL dialysis volume) in MgCl (1

CIL 20% (w/v) MgCl/mL dialysis volume) overnight at 370C. The digested extraction was










dialyzed (MWCO: 3,500) for 24-48 hours at 40C. The final extraction was used in a Tsai-Frasch

silver stain to analyze the LPS.

TR~zol Reagent Extraction of LPS

TRlzol Reagent (Invitrogen), a mono-phasic solution of phenol and guanidine

isothiocyanate (77), was used to extract y. cholerae N16961 LPS from whole cells. Ten

milliliters of log phase (OD600 ~ 0.4-0.6) y. cholerae N16961 were centrifuged at 13,776 x g for

10 minutes at 40C. The supernatant was decanted, and the cells were suspended in 200 CIL of

TRlzol Reagent and incubated at room temperature for 10 tol5 minutes. Twenty microliters of

chloroform per mg of cells was added to the mixture, vortexed, and incubated at room

temperature for 10 minutes. The mixture was centrifuged at 12,000 x g for 10 minutes at room

temperature. The resulting aqueous phase was transferred to a new microcentrifuge tube and

stored on ice. One hundred microliters of ddH20 was added to the TRlzol-Reagent tube, and a

back extraction was performed. Two more back extractions were performed following the first

back extraction to give a total of four extractions. The extractions were dried with a Savant

SpeedVac Concentrator (Global Medical Instrumentation, Ramsey, MN) for 2.5 hours. The

pellet was suspended in 0.5 mL of 0.375 M MgCl in 95% (v/v) EtOH that was chilled to 40C.

The suspension was centrifuged at 12,000 g for 15 minutes at room temperature. The resulting

pellet was suspended in 200 CIL of ddH20. The final LPS extraction was analyzed in a Tsai-

Frasch silver stain (see below).

Extraction of Periplasmic Proteins

Periplasmic proteins from E. coli were extracted using a Tris-EDTA-Sucrose (TES)

extraction. A standing-overnight culture of an E. coli culture was diluted 1:400 in 100 mL of LB

medium and incubated overnight in a 370C shaking incubator. The next day the cells were

centrifuged at 13,776 x g for 10 minutes at 40C. The supernatant was stored at 40C. The pellet









was suspended in 25 mL of PBS by vortexing. The suspension was centrifuged at 13,776 x g for

10 minutes at 40C. The supernatant was discarded, and the pellet was suspended in 10 mL of

TES (20% (w/v) D-sucrose, 30 mM Tris-HC1, 1 mM EDTA) in water. The suspended pellet was

incubated for 15 minutes at room temperature with occasional swirling. The suspension was

centrifuged at 13,776 x g for 10 minutes at 40C. The supernatant was discarded, and the pellet

was gently washed with 10 mL of 0.5 mM MgCl2 without dislodging the pellet. The 10 mL of

the 0.5 M MgCl2 WASh was discarded, and the pellet was suspended by vortexing in 10 mL of 0.5

mM MgCl2. The suspension was incubated in an ice bath for 15 minutes with occasional

swirling. The suspension was centrifuged at 21,525 x g for 30 minutes at 40C. The supernatant

was collected and concentrated with 10,000 MWCO Amicon (Millipore, Billerica, MA) filters

and stored at 40C. The pellet was suspended in 10 mL of PBS and stored at 40C.

Determination of Protein Concentration

Protein concentration was determined by DC-Protein Assay (Bio-Rad). Protein samples

were diluted in PBS. Five microliters of each sample was added to a microtiter well followed by

25 pIL of Reagent A and 200 pIL of Reagent B. All wells were performed in triplicate. Protein

standards used were Bovine Serum Albumin (B SA) in concentrations from 0-1.4 mg/mL. The

reaction took place for 15 minutes at room temperature and was read at 750 nm with an ELx 800

UV plate reader (Bio-Tek). Protein concentrations were analyzed using KC Junior (Bio-Tek)

software.

Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis (SDS-PAGE)

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis was performed to resolve

yK cholerae LPS and protein preparations. ReadyGel Tris-glycine 12% (w/v), 15% (w/v), and 4-

20% (w/v) PAGE gels (Bio-Rad) were used with the MiniProtean-Electrophoresis system (Bio-

Rad) for SDS-PAGE analysis of samples. The LPS samples were sonicated by a Vibra Cell









(Sonics & Materials Inc.) for 2 minutes and diluted in Laemmli sample buffer (Bio-Rad).

Protein samples were vortexed and diluted in Laemmli sample buffer. After dilution in Laemmli

sample buffer, the samples were boiled for 10 minutes and loaded in the gels. The samples were

electrophoresed for 1 hour at 100 Volts in electrode buffer (25 mM Tris, 0.17 M glycine, and

0.1% (w/v) SDS).

Coomassie Blue Staining

Proteins resolved by SDS-PAGE were stained with 0.05% (w/v) Coomassie Blue R, 50%

(v/v) methanol, and 10% (v/v) glacial acetic acid in water and agitated gently for 10 minutes.

The gel was rinsed with ddH20 to remove excess stain. The gel was destined with 5% (v/v)

methanol and 7% (v/v) glacial acetic acid in water overnight while gently shaking. The next day

a picture was taken of the gel, and the gel was dried.

Tsai-Frasch Silver Staining

Tsai-Frasch silver staining (78) was performed on SDS-PAGE resolved LPS that was

extracted from Y. cholerae. Carbohydrates resolved by SDS-PAGE were fixed in 40% (v/v)

ethanol and 5% (v/v) glacial acetic acid in water overnight. Carbohydrates were oxidized by

incubating the fixed gel in 0.7% (w/v) periodic acid, 40% (v/v) ethanol, and 5% (v/v) glacial

acetic acid in water for five minutes. The gel was washed three times in 500 mL of ddH20 with

15 minutes per wash. Reagent A (1% (v/v) 10 M NaOH and 6.7% (v/v) NH40H in water) and

Reagent B (20% (w/v) AgNO3 in Water) were made immediately before staining. The stain was

made by drop wise titration of Reagent B into Reagent A until a hazy brown color to the solution

appeared and remained. The titrated reagent was diluted in 1 15 mL of ddH20 to make the final

stain solution. The gel was placed in the staining solution and stained for 10 minutes with

agitation. The gel was washed three times in 500 mL of ddH20 with 15 minutes per wash. The

gel was developed with 1 x Developer (5 x Tsai-Frasch developer: 1.3 mM citric acid and 0.25%









(v/v) of 37% (v/v) formaldehyde in ddH20) for 3 to 10 minutes with constant gentle shaking.

The developed gel was soaked in 1% (v/v) acetic acid in water for 10 minutes and transferred

into 50 mL of ddH20. A picture was taken of the gel, and then the gel was dried.

Western Blot

Proteins resolved by SDS-PAGE were transferred onto a nitrocellulose membrane,

followed by reaction with antibodies and development with enhanced chemiluminescent (ECL)

(Pierce). A Mini-TransBlotting cell (Bio-Rad) was used for the transfer of proteins to the

nitrocellulose membrane. The transfer buffer (25 mM Tris, 192 mM glycine, and 20% (v/v)

methanol) was pre-chilled to 40C. The transfer occurred at 100 V for one hour at 40C.

Transferred proteins were analyzed by Ponceau S staining. Fifteen milliliters of 0.3% (w/v)

Ponceau S dissolved in 3% (v/v) trichloroacetic acid (TCA) was added to the transferred

membrane and incubated 10 minutes at room temperature with gentle agitation. The stained

membrane was gently washed in 25 mL of dH20 to remove excess stain. The membrane was

photographed to view the stained proteins. The membrane was blocked in 25 mL of PB S casein

blocking buffer (Pierce) containing 1% (v/v) Tween-20 for two hours at room temperature or

overnight at 40C with gentle shaking. Primary antibodies were diluted in PBS casein blocking

buffer (Pierce) containing 0.05% (v/v) Tween-20 to give a final volume of 10 mL and were

added to the blocked membrane for one hour at room temperature with gentle shaking. The

membrane was washed three times with 20 mL of PB S-0.05T while shaking for 15 minutes for

each wash. Secondary antibodies conjugated with HRP were diluted in PBS casein blocking

buffer (Pierce) containing 1% (v/v) Tween-20 to give a final volume of 10 mL and were added to

the membrane for 30 minutes at room temperature with gentle shaking. The membrane was

washed three times with 20 mL of PBS-0.05T while shaking for 10 minutes for each wash. The









blot was incubated in 8 mL of ECL substrate (Pierce) for three minutes at room temperature and

developed on CL-XPosure Film (Pierce) by an X-omat 2000 processor (Kodak, Rochester, NY).

Lipopolysaccharide Saturation to Nitrocellulose Paper

The LPS saturation experiment to nitrocellulose paper was performed to determine the

concentration of y. cholerae 569B LPS needed to saturate a piece of nitrocellulose paper.

Phenol-water-extracted y. cholerae 569B LPS was serially diluted from 100 to 1 Cpg/mL in PBS.

Twelve pieces of nitrocellulose paper were cut to dimensions of 5 mm x 20 mm. Nitrocellulose

strips were put in a microcentrifuge tube containing 256 CLL of the LPS dilutions or PB S for a

control. LPS and PBS coated strips were done in duplicates in separate tubes. The strips were

coated overnight at 40C on a labquake. Coated nitrocellulose strips were transferred to new

microcentrifuge tubes. Strips were washed 3 times with 1 mL of PBS on a labquake with 5

minutes per wash. Strips were blocked with 1 mL of casein blocking buffer (Pierce) for one hour

at room temperature on a labquake. Strips were incubated with 0.5 mL of rabbit anti-Y. cholerae

01 LPS polyclonal antibody (Accurate Chemical, Westbury, NY) at a dilution of 1:400 in casein

blocking buffer (Pierce) for one hour at room temperature on a labquake. Strips were washed 3

times with 1 mL of PBS on a labquake with 5 minutes per wash. Strips were incubated with 0.5

mL of goat anti-rabbit peroxidase conjugated monoclonal antibody (Jackson ImmunoResearch

Laboratories, West Grove, PA) at a concentration of 1:1000 in casein blocking buffer (Pierce) for

30 minutes at room temperature on a labquake. Strips were washed 3 times with 1 mL of PBS

on a labquake with 5 minutes per wash. Strips were incubated in 5 mL of ECL substrate (Pierce)

for three minutes at room temperature and developed on CL-XPosure Film (Pierce) by an X-

omat 2000 processor (Kodak).









Colony Blot with scFv

To test whether bacteria produce scFv that are specific to a target molecule, a colony blot

was developed to screen scFv in a high throughput manor. A Petri dish size piece of

nitrocellulose paper (Bio-Rad) was coated overnight at 40C with 10 mL of target antigen at a

concentration that the target antigen was panned. The next day the coated nitrocellulose paper

was washed three times with 15 mL of0. 1 M TBS with five minutes per wash on a labquake.

The nitrocellulose paper was then blocked for one hour at room temperature with 10 mL of

casein blocker (Sigma-Aldrich) on a labquake. After being blocked, the nitrocellulose paper was

overlaid onto a plate containing bacterial colonies and stamped with a metal-plate stamp. The

bacterial-colony-containing nitrocellulose was washed three times with 15 mL of 0. 1 M TBS

with five minutes per wash on a labquake. The nitrocellulose paper was incubated with Protein

L-peroxidase (Sigma-Aldrich) that was diluted 1:2,000 in 10 mL of casein blocker (Sigma-

Aldrich) for one hour at room temperature on a labquake. The nitrocellulose paper was washed

three times with 15 mL of0. 1 M TBS with five minutes per wash on a labquake. The

nitrocellulose paper was incubated in 5 mL of ECL substrate (Pierce) for three minutes at room

temperature and developed on CL-XPosure Film (Pierce) by an X-omat 2000 processor (Kodak).









Table 2-1. Bacterial strains al
Strain
E. coli DH~a




E. coli DH~a (pNR100)




E. coli EC100D





E. coli EC100D (pGTR203)
(Hye)





E. coli ER2738




E. coli HB 101 (pMJ100)



E. coli HB2151

E. coli JM109




E. coli MGl655
E. coli MGl655 (pGTR203)
(Hye)
E. coli 0157:H7 EDL933
E. coli 0157:H7 87-23


nd plasmids used.
Genotype / Description
F-9p80dlacZAM15
A(lacZYA-argF)Ul69 deoR,
recA1, endA1, hsdR17(rk
mk/), phoA, supE44, h-, thi-
1, gyrA96, relA1
F-9p80dlacZAM15
A(lacZYA-ar gF)U l69 deoR,
recA1, endA1, hsdR1 7(rk
mi), phoA, supE44, h- thi-
1, gyrA96, relA1 (pNR100)
F- mcrA A(mrr-hsdRMS-
mcrBC) (p80dlacZAM15
AlacX74 recAl1 endAl1
araD139 A(ara, leu)7697
galU galK h- rpsL nupG
pir (DHFR)
F- mcrA A(mrr-hsdRMS-
mcrBC) (p80dlacZAM15
AlacX74 recAl1 endAl1
araD139 A(ara, leu)7697
galU galK h- rpsL nupG
pir (DHFR) (pGTR203)
(Hy Q)
F' lac14, A(lacZ)M 15
proA B zzf: :Tnl0(TetR) /
jhuA2, supE, thi, a(lac-
proAB), A(hsdldSmcrB)5,
(r m McrBC )
F-, hsdS20(rB- mB), recAB3,
ara-14, proA2 lacYI, galK2,
rpsL20(SmR), xyl-5 mtl-1,
supE44, h-, (pMJ100)
ara, A(lac-pro), thi/
F'proA B lacP4ZAM15
endA1, recA1, gyrA96, thi,
hsdR17 (rk mk ), relA1,
supE44, A( lac-proAB) /F'
traD36, proAB,
laqI9ZAM15
F ilvG rfb-50, rph-1
F ilv(T, rfb-50, rph-1
(pGTR203) (Hye)
Stx1 Stx2 gyrA
Stxf~, Stx2-, gyrA


Source / Reference
Bethesda Research
Laboratories, Rockville,
MD


(69)




Epicentre, Madison, WI





This paper






New England Biolabs,
Ipswich, MA



(70)



Amersham Pharmacia,
Piscataway, NJ
Invitrogen, Carlsbad, CA




(71)
This paper

(72)
(73)









Table 2-1. Continued
Strain
E. coli TG1


E. coli TG1 (pGTR203)



Hye


M13cp

M13cp-CT

M13cp-dg3

pGTR203

pMJ100

pNR100
K. cholerae 569B
K. cholerae N16961


Genotype / Description
A(lac-proAB), supE, thi,
hsdD5/F' traD36, proA B,
lac4, lacZAM15
A(lac-proAB), supE, thi,
hsdD5/F' traD36, proA B,
lacP4, lacZAM 15
(pGTR203)
Hyperphage genome
(contains all M13 genes
except gIll)
M13mpl9; CamR; full
length gIII
M13mpl9; CamR; truncated
8III
M13mpl9; CamR; deleted
8III
pACYC184 with M13 gIll;
CamR
pBluescript with slt-II;
AmpR
Stx2 toxoid; AmpR
Classical, Inaba
El Tor, Inaba


Source / Reference
MRC, Cambridge, UK


Gopal Sapparapu



(66)


(64)

(64)

(64)

Gopal Sapparapu

(70)

(69)
(74)
(75)









CHAPTER 3
RESULTS

Rationale for Study

There exists a strong need for rapid, sensitive, and selective methods of detection for

biological agents. Detecting biological organisms and their products has many uses. One reason

is for diagnosis of patient samples. Detection and identification of a disease-causing agent in

patient samples allows for the proper treatment to be given. Detecting biological agents in food,

water, and air allows for preventative measures to be enforced to prevent or inhibit the spread of

disease. A common component in almost all detection assays is a protein that specifically binds

to a target molecule. This thesis describes the optimization of protocols used to isolate specific

proteins that bind to target molecules by using phage display.

The specific aims of this study are:

1. To optimize extraction methods and determine saturation limits of y. cholerae LPS
and use these methods to pan phage display libraries to y. cholerae LPS.
2. To optimize panning and screening procedures to allow for a more efficient
biopanning process that will be more likely to isolate and detect specific recombinant
phagemid particles.
3. To optimize the production of phagemid particles to ensure high quality and high
quantity of phagemid particles.
4. To isolate specific recombinant phagemid particles to E. coli 0157:H7 Stx2 toxin
using the optimization techniques discovered in previous aims.

Specific Aim 1: Panning to V. cholerae LPS

The first goal of this work was to isolate phage display reagents that specifically

recognized V. cholerae 01 LPS. There are many serogroups of y. cholerae but only Ol and

0139 cause epidemic cholera. Therefore, 01 LPS serves as a useful target because it is a

distinguishing feature of y. cholerae 01 strains. Previous attempts in the laboratory to obtain

phage display reagents that specifically recognized y. cholerae 01 LPS may have failed because

the K cholerae 01 LPS was panned with the Tomlinson scFv phagemid library. Phagemid









particles, which only encode the giII gene, require phage amplification tools to provide phage

genes I-XI to enable phagemid particle production. We chose the Ph.D. 12mer phage library

(NEB) because the Tomlinson scFv phagemid library was disadvantageous due to the poor

quality of current methods to amplify phagemid particles. Because the Ph.D. 12mer phage

library is M13 phage, it does not require the aid of tools to amplify phages like a phagemid

library requires. Before the Ph.D. 12mer phage display library was to be panned against

yK cholerae 01 LPS, an extraction method for y. cholerae 01 LPS and the best solid support for

LPS binding had to be determined. Two LPS extraction methods were compared to determine

which method generated the best quality of LPS to pan against. Also, two binding supports

(microtiter well and nitrocellulose) paper were analyzed to determine which binding support

could bind the most LPS.

The Phenol-Water Method Extracted E cholerae LPS Most Closely Resembled the
Commercially Acquired E cholerae LPS

The phenol-water method for the extraction of LPS (76) was the previous lab procedure

used to extract LPS. However, a new method for LPS extraction involving TRlzol Reagent

(Invitrogen) was described by Yi and Hackett (77) that was stated to be a more efficient method

for LPS extraction than the phenol-water method. The TRlzol Reagent LPS extraction can be

executed in one day, while the phenol-water LPS extraction requires nearly a week. The TRlzol

Reagent method for LPS extraction is also stated to be a cleaner method than the conventional

phenol-water method for LPS extraction by extracting LPS with less degradation and

contamination. TRlzol Reagent is composed of phenol and guanidinium thiocyanate in aqueous

phase, and upon addition of chloroform the TRIzol Reagent can be used to extract LPS.

Vibrio cholerae N16961 LPS was extracted by the TRlzol Reagent extraction of LPS (see

Materials and Methods). The TRlzol Reagent-extracted y. cholerae LPS was resolved by SDS-









PAGE and stained with Coomassie blue for visualization of proteins (Fig. 3-1A) and with Tsai-

Frasch silver stain for visualization of carbohydrates (see Materials and Methods) (Fig. 3-1B).

The Coomassie blue stain of the TRlzol Reagent-extracted LPS, which was diluted 1:2 in

Laemmli sample buffer, showed no detectable proteins. The limit of detection for proteins with

Coomassie blue is 0.3-1 Cpg of protein/band (79); therefore, there was less than 0.3 Cpg of any

single protein loaded onto the gel that was stained by Coomassie blue. The Tsai-Frasch silver

stain for the TRlzol Reagent LPS extraction showed non-staining bands between 20 and 30 kDa.

The Tsai-Frasch silver stain primarily stains carbohydrates and poorly stains proteins and lipids

(78); therefore, the non-staining bands may be lipids or proteins. However, because the

Coomassie blue stain for the TRlzol-extracted LPS showed no detectable proteins and due to the

thickness of the non-staining bands, which correlates to concentration, then the non-staining

bands were most likely due to lipid contamination. The LPS banding pattern of the TRlzol

reagent LPS extraction in the silver stain was very similar to the commercial K. cholerae 569B

LPS (Sigma-Aldrich). On the silver stain, both LPS preparations had bands at approximately 14

and 20 kDa and no bands larger than 50 kDa. Previously published silver stains of y. cholerae

01 LPS had bands at ~10 and 14 kDa (which was stated to be the lipid A-core of LPS), and ~20-

50 kDa (which was stated to be the lipid A-core plus repeating O-antigens) (80,81).

To determine the concentration of the TRlzol-Reagent-extracted LPS, the intensities of

the bands were compared to those of the commercial LPS in the silver stained gel (Fig. 3-1B).

The intensity of the band of the commercial LPS (1 mg/mL) that was diluted 1:5 was

approximately 1.5 times stronger than the intensity of the band of the TRIzol-Reagent-extracted

LPS that was diluted 1:2. Therefore, we estimated the concentration of the TRlzol-Reagent-









extracted LPS was approximately 260 Cpg/mL. The purity of this LPS extraction was less than

0.2 Cpg of protein/yg of LPS.

The phenol-water LPS extraction (76) was performed to compare results with the TRIzol

Reagent LPS extraction. The phenol-water LPS extraction method differs little from the original

phenol-water LPS extraction described by Westphal, Luderitz, and Bister in 1952 (82). The

phenol-water LPS extraction method utilizes the property that most proteins but not LPS are

soluble in phenol, and that LPS is soluble in water. At temperatures above 680C, phenol and

water are miscible. Upon cooling to 5-100C and centrifugation, three phases result: an aqueous

phase containing LPS, a phenol phase containing proteins, and a solid phase containing water-

and phenol-insoluble compounds. Removal and purification of the LPS-containing aqueous

phase follows the separation of the phases.

One and a half liters of log phase (OD600 ~0.4-0.6) K. cholerae 569B was used for

extraction of LPS by the phenol-water extraction protocol (see Materials and Methods). The

four aqueous extractions were dialyzed and digested by DNase (Qiagen) and RNase (Qiagen)

because nucleic acids are also extracted into the aqueous phase. Prior to digestion the dialyzed

extraction was diluted 1:100 in ddH20, and the absorbance was measured at 260 nm and 280 nm

(A260 and A280). The A260, which correlates to DNA concentration, predigestion was 0.277 and

post-digestion was 0.148. The A280, which correlates to protein concentration, predigestion was

0.127 and post-digestion was 0.059. The final phenol-water extraction was examined for

proteins by Coomassie blue stain (Fig. 3-2A) and for carbohydrates by Tsai-Frasch silver stain

(see Materi als and Method s) (F ig. 3 -2B). The C oomas si e b lue stai n for phenol -water- extracted

yK cholerae LPS showed a protein band at approximately 15 kDa. This could possibly have been

due to the 0.04 mg/mL RNase, 13.7 kDa, present in the extraction. The band intensity correlated









with the concentration of RNase present in the sample. Therefore, the protein contamination was

most likely not from bacterial proteins that were co-extracted. The Tsai-Frasch silver stain for

phenol-water-extracted y. cholerae LPS showed similar banding patterns to that of the

yK cholerae 569B LPS (Sigma-Aldrich). On the silver stain, both LPS preparations had bands at

approximately 13, 20, and 23 kDa. The phenol-water extraction showed distinct banding around

20-50 kDa that correlates with published data of where the banding of the O-antigen of LPS

occurs. The commercially obtained LPS had smearing around 20-50 kDa. Vibrio cholerae 01

LPS has 12-18 O-antigen groups (83); therefore, the additional banding in the phenol-water-

extracted LPS may be due to the ability of our phenol-water method to extract LPS with higher

numbers of O-antigen side chains than that of the method used for the extraction of the

commercial LPS that was phenol extracted and purified by gel-filtration chromatography.

To determine the concentration of the phenol-water-extracted LPS, the intensities of the

bands were compared to those of the commercial LPS in the silver stained gel (Fig. 3-2B). The

intensities of the bands of the commercial LPS (1 mg/mL) that was diluted 1:8 were between the

intensities of the bands of the phenol-water-extracted LPS that was diluted 1:2 and 1:4.

Therefore, we estimated the concentration of the commercial LPS at a dilution of 1:8 was

comparable to the concentration of the phenol-water-extracted LPS at a dilution of 1:3. Because

the commercial LPS was at a concentration of 1 mg/mL, the concentration of the phenol-water-

extracted LPS was approximately 375 Cpg/mL. The purity of this LPS extraction was

approximately 1 Cpg of protein/yg of LPS. However, if the protein band from the Coomassie blue

stain was from RNase, the phenol-water-extracted LPS would have less than 0.3 Cpg of bacterial

protein contamination/ Cpg of LPS, which is the same protein concentration in the TRIzol-

extracted LPS.









Based on the results described above, we decided to use the phenol-water method for the

extract on of y. cholerae 0 1 LP S over the TRlzol Reagent method. The phenol-water- extracted

LPS most closely resembled y. cholerae 569B LPS (Sigma-Aldrich) when examined by silver

stain. The phenol-water extraction method even appears to be a more sensitive and delicate

extraction method than that used to extract the commercial LPS, because more precise banding

from O-antigens was detected in the phenol-water-extracted LPS than in the commercial LPS by

silver stain. The phenol-water LPS extraction extracted more protein than the TRlzol Reagent

LPS extraction; however, for using the LPS extraction for panning it is more important to have

intact O-antigens than minor protein contamination.

More V. cholerae LPS Can Be Bound to Nitrocellulose Paper than to a Microtiter Well

In most panning procedures the target molecule to be panned against is immobilized onto a

solid support. We wanted to compare the saturation limits for LPS binding to a microtiter well

and to a piece of nitrocellulose paper of equal dimensions. Enzyme-linked immunosorbent

assays were performed to determine the saturation limits of the amount of phenol-water-

extracted LPS that could be bound to a microtiter well. A 96-well Maxisorp plate (Nunc) was

coated with 0. 1-200 Cpg/mL of y. cholerae 569B LPS that was extracted by the phenol-water

method. Two different primary antibodies were used to detect the LPS: a mouse anti-

K~ cholerae 01 LPS monoclonal antibody, a-Vc 01 LPS mAb, (Austral Biologicals, San Ramon,

CA) and a rabbit anti-Y. cholerae 01 polyclonal antibody, a-Vc 01 LPS pAb (Accurate

Chemical, Westbury, NY). The primary antibodies were used at their manufacturer

recommended concentration or dilution: a concentration of 1 Cpg/mL for a-Vc 01 LPS mAb and

a dilution of 1:100 for the a-Vc 01 LPS pAb. Standard ELISA procedures were used for the

assay (see Materials and Methods) (Fig. 3-3). When using the a-Vc 01 LPS mAb a significant

difference (p<0.05) between the signal to noise (S:N) values (signal on antigen-coated









well/signal on PB S-treated well) of LPS bound to a-Vc 01 LPS mAb reacting to HRP-

conjugated anti-mouse antibody was not reached until the concentration of the LPS used to coat

the microtiter plate decreased from 1 Cpg/mL to 0.1 Cpg/mL (0.1 to 0.01 Cpg LPS) (p = 0.01). When

using the a-Vc 01 LPS pAb, a significant difference in the S:N was not reached until the

concentration of the LPS decreased from 10 Cpg/mL to 1 Cpg/mL (p = 0.01). Both the a-Vc 01

LPS mAb and the a-Vc 01 LPS pAb when used at the manufacturer recommended

concentrations saturated LPS-coated microtiter wells when the microtiter wells were coated with

LPS at a concentration of 10 Cpg/mL. For the sake of conserving the LPS, the concentration of

LPS used to coat microtiter wells was decided to be 1 Cpg/mL. While there was a significant

difference in the S:N when using 10 Cpg/mL of LPS as opposed to 1 Cpg/mL of LPS with the a-Vc

01 LPS pAb, the S:N difference was only 0.5. This decrease in the S:N by 0.5 did not merit the

need to use 10-fold more LPS; therefore, for further experiments y. cholerae LPS was used at a

concentration of 1 Cpg/mL to coat microtiter wells.

The saturation limits for the primary antibodies were analyzed using 1 Cpg/mL K. cholerae

LPS to coat a microtiter plate in an ELISA experiment. The a-Vc 01 LPS mAb was serially

diluted to concentrations from 1.3 to 0.04 Cpg/mL by 2-fold dilutions. The a-Vc 01 LPS pAb

was serially diluted from 1:100 to 1:3,200 by 2-fold dilutions. When using 1 Cpg/mL LPS to coat

a microtiter well, the saturation limit of the a-Vc 01 LPS mAb was not reached (Fig. 3-4A).

There were significant differences (p=0.01-0.03) between the S:N of each dilution step when

using the a-Vc 01 LPS mAb at concentrations from 1.3-0.04 Cpg/mL. Therefore, when using 1

Cpg/mL LPS the saturation limit for the a-Vc 01 LPS mAb was greater than 1.3 Cpg/mL. When

using 1 Cpg/mL LPS to coat a microtiter well, a significant difference in the measured S:N

between dilution steps when using the serially diluted a-Vc 01 LPS pAb was not reached until a









dilution of 1:400 (p=0.01) (Fig. 3-4B). Therefore, the saturation limit of the a-Vc 01 LPS pAb

was reached at a dilution of 1:400.

Because a microtiter well coated with 1 Cpg/mL LPS was never saturated by the a-Vc 01

LPS mAb, the a-Vc 01 LPS pAb was used to characterize the saturation limits of nitrocellulose

paper. Pieces of nitrocellulose paper were cut to dimensions (5 mm x 20 mm) that were

approximately equivalent to the surface area of a microtiter well. The same phenol-water-

extracted Y. cholerae LPS used for the microtiter well saturation experiment was also used to

characterize the saturation of nitrocellulose with LPS. Serially diluted LPS from 100-1 Cpg/mL

was used to saturate strips of nitrocellulose paper. These strips of LPS-coated nitrocellulose

paper were examined in an altered Western blot protocol (see Materials and Methods for LPS

Saturation to Nitrocellulose Paper) using the a-Vc 01 LPS pAb at a dilution of 1:400 (Fig. 3-5).

The nitrocellulose paper was not saturated with LPS until 100 Cpg/mL LPS (250 Cpg) was used to

coat the nitrocellulose paper. Therefore, nitrocellulose paper can bind approximately 250 to

2,500 times more LPS than a microtiter well.

Panning to V. cholerae LPS Failed to Yield Phages that Were Specific to V. cholerae LPS

Vibrio cholerae 569B LPS (Sigma-Aldrich) was immobilized onto a polystyrene

Maxisorp microtiter well and panned against the NEB Ph.D. 12mer phage display library (see

Materials and Methods). Five rounds of panning were performed with amplification of phages

after the first four rounds of panning. Each round of panning promotes selection of phages that

specifically bind to the target molecule. Amplifying the selected phages should increase the

percentage of specific phages present in the library to pan with. Panning and amplifying for five

rounds should greatly improve the chances of isolating phages that specifically bind to the target

molecule.









One hundred phage clones from the fifth round of panning were amplified and screened by

ELISA against y. cholerae 569B LPS, carbonate coating buffer, and PBS-treated microtiter

wells. The a-Vc 01 LPS mAb was also tested against the three coating antigens and generated a

S:N of 15.5, proving that the LPS successfully bound to the microtiter plate. Anti-BSA

phagemid particles were used in the ELISA against BSA (10Cpg/mL), carbonate coating buffer,

and PBS-treated microtiter wells and generated a S:N of 22, proving that the ELISA was a

successful assay for the detection of phages. All 100 clones gave a S:N <1.4. A positive signal

was classified as a S:N >2. Therefore, all of the clones screened by ELISA were negative,

meaning that they failed to specifically bind to K. cholerae 569B LPS.

Panning the Ph.D. 12mer phage library against y. cholerae 01 LPS immobilized onto a

microtiter well failed to yield phage clones specific to K. cholerae 01 LPS. It was therefore

attempted to improve the panning procedure. Instead of using a microtiter well as a binding

support for LPS, nitrocellulose paper was used as a binding support. Nitrocellulose paper can

bind approximately 250 to 2,500 times more LPS than a microtiter well; therefore, if more

antigen is present to pan against then the chances of isolating clones that specifically bind to LPS

may improve. Also, the elution method was tested. The previous elution method was with

glycine (pH 2.2). To determine if a glycine elution or an elution with a high concentration of the

target antigen is more successful in eluting phages, both elution methods were tested in parallel.

The theory behind an antigen elution is that a high concentration of antigen in solution will

rapidly bind phages bound to the antigen on the nitrocellulose paper when the phages are

transiently released. The end result is that these phages will elutee" off of the antigen-coated

nitrocellulose paper and bind to antigen in solution. Antigen elution is supposed to be a less

harsh treatment than acidic or basic elutions, which may denature the peptides.









Vibrio cholerae 569B LPS that was phenol-water-extracted was immobilized onto a strip

of nitrocellulose paper and panned with the Ph.D. 12mer phage display library (NEB) (see

Materials and Methods). Two parallel panning experiments were performed; one panning used a

glycine (pH 2.2) elution, while the other panning used a y cholerae 569B LPS elution. Three

rounds of panning were performed with a negative panning performed between rounds one and

two. The negative panning involved panning the library against a strip of nitrocellulose paper

that was blocked in casein blocking buffer (Sigma-Aldrich) to remove phages that were specific

to nitrocellulose paper or casein blocking buffer. Amplification of phages was performed after

rounds one and two to increase the number of selected phages to improve the chances of

isolating phages specific to the target molecule. Two hundred clones for each panning were

selected from the round 3 elution and amplified for screening of phages by ELISA. The 400

phage clones were screened in an ELISA against y. cholerae 569B LPS and PBS. The a-Vc 01

LPS mAb was also tested against LPS and PBS-treated wells and generated a S:N of 10, proving

that the LPS successfully bound to the microtiter plate. Anti-B SA phagemid particles were used

in the ELISA against BSA (10Cpg/mL) and PBS-treated microtiter wells and generated a S:N of

15, proving that the ELISA was a successful assay for the detection of phages. All 400 clones

gave a S:N less than 1.7. Therefore, all of the clones screened were negative and failed to

specifically bind to K. cholerae 01 LPS.

Conclusion of Specific Aim 1

In comparing the phenol-water and the TRIzol Reagent methods for extraction of

yK cholerae 01 LPS, the phenol-water method was better. The phenol-water extraction for LPS

yielded LPS that most closely resembled commercially obtained y. cholerae 01 LPS by silver

stain of SDS-PAGE-resolved LPS. The phenol-water method even extracted LPS that appeared

to have more O-antigens than the commercially obtained y. cholerae 01 LPS. In comparing the









saturation limits of y. cholerae 01 LPS bound to microtiter wells and to nitrocellulose paper, the

nitrocellulose paper bound approximately 250 to 2,500 times more LPS than did microtiter wells.

Panning the Ph.D. 12mer phage display library against y. cholerae 01 LPS immobilized onto

nitrocellulose and onto microtiter wells failed to yield phages that were specific to y. cholerae

01 LPS in an ELISA.

Specific Aim 2: Improve Panning and Screening Process of Biopanning

There are many steps in the biopanning and screening process. By analyzing certain key

steps and optimizing them, the chances of isolating phagemid particles that are specific to the

target molecule panned against should increase. Optimization of the panning process was

performed with K. cholerae 01 whole cells as a target. Whole cells were chosen as the target

because they are the target antigen for the biosensors under development by our collaborators.

Optimization of the concentration of whole cells were analyzed to determine if higher or lower

concentrations of the antigen were more likely to select for specific phagemid particles.

Increasing the concentration of antigen should increase the amount of phagemid particles

selected. This could mean an increase in phagemid particles that are specific and/or non-specific

to the target molecule. Decreasing the concentration of antigen should decrease the amount of

phagemid particles selected. This could mean a decrease in phagemid particles that are specific

and/or non-specific to the target molecule. An ideal concentration would select a high number of

phagemid particles that are specific to the target molecule and a low number of phagemid

particles that are not specific to the target molecule.

The Tomlinson scFv libraries have a trypsin cleavable site between the scFv peptide and

the pIII peptide of the phagemid particle. Cleaving this site separates the phagemid particle from

the antigen, leaving the scFv bound to the antigen. Optimizing the time of trypsin elution may

improve the chances of isolating specific phagemid particles. Eluting with trypsin for too short









of a time may fail to elute specific phagemid particles, but eluting for too long of a time may

increase the amount of nonspecific phagemid particles eluted or may even cause degradation of

the phagemid particles.

Elution of Bound Phages by Trypsin during Panning Was Optimal between 10 and 30
Minutes

A one-round panning procedure was done to determine the optimal concentration of whole

cells in suspension and the optimal trypsin elution time for biopanning. The purpose of the

experiment was to compare the differences between using 1 x 10" or 1 x 109 whole cells of

yK cholerae N16961 and trypsin elution times of 15, 30, 45, and 60 minutes in the biopanning

process. Four parallel pannings were performed. Two were panned with the Tomlinson J library

(2.5 x 1010 tu), one against 1 x 10s and the other against 1 x 109 whole cells of y. cholerae in

suspension. The two other pannings used the Tomlinson J library (2.5 x 1010 tu) containing 1%

(~2.5 x 10" tu) of Vc86 phagemid particles (an scFv-producing phagemid particle previously

isolated in this lab by Dr. Rebecca Moose-Clemente from the Tomlinson scFv library that

recognizes y. cholerae 1019 and E. coli 0157:H7 whole cells). Panning against a library that is

already enriched with phagemid particles that recognize the target molecule should help

determine the optimal panning parameters. If more phagemid particles are eluted with the

enriched library than the non-enriched library at a certain panning parameter then it would lead

to the conclusion that the phagemid particles that recognized the target molecule were eluted at

that parameter. One panning used 1 x 10s, and the other used 1 x 109 whole cells of y. cholerae

in suspension. Standard suspension panning procedures (see Materials and Methods) were

performed. When the trypsin was added to the bound-phagemid particles, an aliquot of the

eluted phagemid particles was removed at 15, 30, 45, and 60 minutes after the start of the elution

and was titered immediately (Fig. 3-6). Elution of phagemid particles with trypsin was optimal









between 15 and 30 minutes. The phagemid particle titer decreased by approximately 80-93%

from 30 to 45 minutes for the four pannings. More phagemid particles were eluted when panned

against 1 x 109 whole cells. However, there were not enough data to do statistical analysis to

determine if using 1 x 109 whole cells eluted significantly more phagemid particles than using 1

x 10s whole cells in the biopanning process.

Because the best whole cell concentration and trypsin elution time was not definitively

obtained in the first experiment, a second one-round panning was done to analyze these factors.

The purpose of the experiment was to compare the differences between using 1 x 10 1 x 10s,

and 1 x 109 whole cells of y. cholerae N16961 and trypsin elution times of 10, 20, and 30

minutes in the biopanning process. The same protocol was used as the first one-round panning

protocol above, except an additional concentration of whole cells was added, so there were six

parallel pannings instead of four. When the trypsin was added to the bound-phagemid particles,

an aliquot of the eluted phagemid particles was removed at 10, 20, and 30 minutes after the start

of the elution and was titered immediately (Fig. 3-7). There were no significant differences in

the titers of eluted phagemid particles when comparing the concentrations of whole cells and

trypsin elution times used in the second one-round panning experiment. The panning was

expected to yield distinct differences between eluted titers of phagemid particles from the

libraries with and without 1% Vc86 phagemid particles. A certain parameter was supposed to be

optimal and yield higher elution titers from the library with 1% Vc86 phagemid particles.

Because there were not any significant differences in the titers of the eluted phagemid particles at

the different parameters an optimal panning condition could not be determined. To determine

which panning parameter elutes the most phagemid particles that recognize y. cholerae whole









cells, another one-round panning was performed with analysis of phagemid particle specifieity

via ELISA.

A third one-round panning was performed to analyze whole cell concentration and trypsin

elution time further. The purpose of the experiment was to compare the differences between

using 1 x 10s and 1 x 109 whole cells of y. cholerae N16961 and trypsin elution times of 10 and

30 minutes in the biopanning process. The same protocol was used as for the first one-round

panning protocol above, except that only two elution time points were taken. When the trypsin

was added to the bound phagemid particles, an aliquot of the eluted phagemid particles was

removed at 10 and 30 minutes after the start of the elution and titered immediately (Fig. 3 -8).

When panning with the Tomlinson J scFv phagemid library that contains 1% (~2.5 x 10s tu)

Vc86 phagemid particles, the amount of eluted phagemid particles did not significantly change

(p=0.50) from 10 to 30 minutes when panning with 1 x 109 whole cells. When panning with the

library that contains 1% (~2.5 x 10s tu) Vc86 phagemid particles, the amount of eluted phagemid

particles significantly decreased (p=0.04) from 10 to 30 minutes when panning with 1 x 10s

whole cells. Because the amount of eluted phagemid particles either stayed the same or

decreased from 10 to 30 minutes in the pannings with 1% Vc86 phagemid particles, it suggested

that eluting for 30 minutes had no advantage over eluting for 10 minutes.

To test if the eluted phagemid particles at the various time points were specific to

K~ cholerae, an ELISA was performed. The eluted phagemid particles from the third one-round

panning experiment were amplified in E. coli TG1 (Hye-lg), which is E. coli TG1 that was

already infected with hyperphage, and used at a concentration of 1 x 10s tu/mL in an ELISA.

The phagemid particles were analyzed by ELISA against y. cholerae whole cells (3 x 10s

CFU/mL) and PBS-treated microtiter wells. A standard ELISA protocol (see Materials and









Methods) was performed (Fig. 3-9). Of the different panning variations, the amplified phagemid

particles from the panning with the library that contained 1% Vc86 phagemid particles, 1 x 10s

whole cells y. cholerae, and an elution time of 30 minutes generated a significantly higher S:N

(p=0.01-0.04) in an ELISA compared to the amplified phagemid particles from the other panning

variations. This suggested that panning with 10s whole cells with a trypsin elution time of 30

minutes was optimal for selecting phagemid particles that were specific to the panned whole

cells compared to using 109 whole cells or a trypsin elution of 10 minutes.

Screening of scFv Proteins by a High Throughput ELISA Was Acceptable

Once the panning procedure is complete, screening of phages or scFvs is performed. The

previous method of screening scFvs involved inducing phagemid-containing bacteria to produce

scFvs in culture tubes or flasks and screening the scFvs by ELISA. However, this method was

very time consuming. Optimizing the screening process to enable high throughput screening of

clones should increase the chances of finding clones that specifically bind to the target molecule.

A colony blot was attempted to try to enable high throughput screening of scFv proteins. The

colony blot (see Materials and Methods) involved the transfer of bacterial colonies that were

grown to secrete scFvs onto a piece of nitrocellulose paper that was coated with a target antigen

and blocked with casein blocking buffer. Following the transfer, the piece of nitrocellulose

paper was washed with TBS to remove unbound scFv proteins, incubated with Protein-L

peroxidase (Sigma-Aldrich) to bind to the kappa light chains of the scFv, washed with TBS to

remove excess Protein-L peroxidase, and developed with ECL (Pierce). Upon development with

ECL, scFv-producing clones specific to BSA failed to produce a positive signal against a piece

of nitrocellulose paper coated in 10 Cpg/mL BSA but did produce a positive signal against pieces

of nitrocellulose paper coated in PBS or y. cholerae whole cells. Also, clones that were not

specific to y. cholerae whole cells in an ELISA produced positive signals against pieces of









nitrocellulose paper coated with y. cholerae whole cells, PBS, or 10 Cpg/mL BSA. The colony

blot was determined to be neither selective nor specific. Because the colony blot was not

functioning, a way to improve the culture tube method for the production of scFvs was

performed.

A high throughput ELISA was contrived to enable screening of hundreds of clones per

day. This high throughput procedure involved picking bacterial colonies containing a phagemid

and growing them overnight in a 96-well plate to produce scFvs (see Materials and Methods).

The next day the 96-well plate containing the turbid cultures was centrifuged to remove the

bacterial cells. These scFvs in the supernatant were examined in an ELISA for screening of their

specificity. To determine how well scFvs could be produced in a microtiter well, as opposed to a

culture tube or flask, a bacterial colony containing a phagemid that encodes an scFv specific to

BSA was grown in a microtiter well. When analyzed by ELISA, the anti-BSA (a-B SA) clone

was diluted 1:2 in casein blocking buffer and produced a S:N of 11.5 in an ELISA against BSA.

Anti-BSA scFv produced in a flask produced an average S:N value between 10 andl5; therefore,

producing scFv proteins in a microtiter well may not be optimal, but it was acceptable for the

purpose of a high throughput screen.

We were successful in producing scFvs in a microtiter well. The next test was to

determine how far the scFv can be diluted and still produce a positive signal. The a-BSA clone

was grown in a microtiter well to produce scFvs. The produced scFvs were diluted 1:2, 1:4, 1:6,

and 1:10 in casein blocking buffer and analyzed by ELISA (Fig. 3-10). Diluting the scFv

supernatant 1:10 in casein blocking buffer yielded a S:N of 5.6. This S:N was high enough for

detection of a positive clone and was not significantly different (p=0. 13) than the S:N produced

when using a-BSA phagemid particles that were diluted 1:6. Therefore, diluting scFv









supernatants as far as 1:10 with other clones can enable detection of positive clones. However,

the a-BSA clone was the strongest clone that we had; therefore, weaker clones might fail to be

detected if analyzed at a dilution of 1:10. For the purpose of a high throughput ELISA screen for

scFvs I would recommend diluting the scFv proteins only 1:4 to 1:6 to lessen the chances of not

detecting weaker binding scFvs.

Screening of Phagemid Particles by a High Throughput ELISA Was Not Optimal

Once the panning procedure is complete, screening of phagemid particles or scFvs is

performed. Previous methods of screening phagemid particles involved the amplification of

eluted phagemid particles in culture tubes or flasks in 2xTY containing 100 Cpg/mL ampicillin, 40

Cpg/mL kanamycin, and 0.1% (w/v) glucose at 300C overnight, followed by screening by ELISA.

This process was time-consuming; therefore, it was tested if phagemid particle production could

be performed in a high throughput manor as with scFvs. Vc86 and a-BSA phagemids in E. coli

TG1 (Hye-sm) (E. coli TG1 that was previously infected with hyperphage) were produced in a

96-well plate in 2xTYcontaining 100 Cpg/mL ampicillin, 40 Cpg/mL kanamycin, and with or

without 0. 1% (w/v) glucose at 300C and 370C overnight. Protocols recommend producing

phagemid particles in medium that has 0.1% (w/v) glucose, but due to catabolite repression of

the lac promoter, which drives transcription of the giII gene, phagemid particle production might

increase without the presence of glucose in the growth medium. The next day, the bacterial cells

were removed by centrifugation, and the phage-containing supernatants were examined in an

ELISA at a dilution of 1:2 in casein blocking buffer (Fig. 3-11). There was not a significant

difference in the S:N (p=0.07-0.48) of phagemid particles produced in microtiter wells with and

without 0. 1% (w/v) glucose. Vc86 phagemid particles produced in microtiter wells with or

without glucose at 300C and 370C did not produce a S:N higher than one. It was also noted that

the cultures in the microtiter plates were only slightly turbid. The Vc86 phagemid particles









produced in a flask (S:N~-4) yielded a significantly higher S:N (p=0.05) compared to Vc86

phagemid particles produced in microtiter wells. When a-BSA phagemid particles were

produced at 300C, the S:N was significantly higher (p<0.04) than when a-BSA phagemid

particles were produced at 370C at either glucose concentration. The a-BSA phagemid particles

produced in a flask (S:N~-12) yielded a significantly higher (p=0.02) S:N compared to a-BSA

phagemid particles produced in microtiter wells (S:N~7).

The experiment was repeated a second time with only the a-B SA phagemid particles

(Fig. 3-12). Anti-BSA phagemid particles produced in microtiter wells yielded significantly

higher signals (p<0.03) when produced at 300C than at 370C at either glucose concentration.

There was not a significant difference (p=0.34) in the signals produced by phagemid particles

that were produced with or without glucose at 300C. When phagemid particles were produced in

microtiter wells at 370C, there was a significantly higher (p=0.02) signal when phagemids were

produced with no glucose compared to 0.1% (w/v) glucose. There was not a significant

difference in the signal (p=0.09) produced by phagemid particles produced in microtiter wells

compared to phagemid particles produced in a flask.

When phagemid particles were diluted 1:2, phagemid particles yielded significantly higher

signals in an ELISA when produced at 300C compared to 370C. Whether the growth medium

included 0.1% (w/v) glucose or not when phagemid particles were incubated at 300C did not

significantly affect the signal of the phagemid particles; therefore, 0. 1% (w/v) glucose was not

enough to cause catabolite repression of the lac promoter. Also, phages produced in flasks

yielded higher signals than phages produced in microtiter wells. To determine if low titers of

phagemid particles produced in a microtiter well were a factor in the lowered ELISA signals for

phagemid particles, the phagemid particles were titered. When a-BSA and Vc86 phagemid









particles were produced in microtiter wells at 300C, the titers were approximately 10' tu/mL.

When a-BSA and Vc86 phagemid particles were produced in microtiter wells at 370C, the titers

were approximately106 tu/mL. Growing phagemid particles at 300C in microtiter wells yielded

approximately 10-fold higher tu/mL titers than phagemid particles that were produced in

microtiter wells at 370C. Whether the growth medium contained 0.1% (w/v) glucose or not did

not affect the titers of phagemid particles. All phages produced in microtiter wells were

amplified by E. coli TG1 (Hye-sm). Escherichia coli TG1 (Hye-sm) is E. coli TG1 that was

already infected with hyperphage. Having hyperphage already present in E. coli removes an

additional step in phagemid particle production to enable a more high throughput screen When

a-BSA and Vc86 phagemid particles were produced in flasks at 300C with 0.1% (w/v) glucose in

the medium, the titers were approximately 10s to109 tu/mL. The phagemid particles produced in

the flasks were amplified by superinfection with hyperphage, instead of by E. coli TG1 harboring

hyperphage. Growing phagemid particles by superinfection with hyperphage may have resulted

in higher titers of phagemid particles compared to phagemid particles produced from E. coli that

already contained hyperphage. The high throughput phagemid particle ELISA was not optimal

for screening phagemid particles because phagemid particles could not be produced in high

enough titers in microtiter wells to make screening of phagemid particles effective in a high

throughput manor.

Conclusion of Specific Aim 2

When panning with the Tomlinson J human synthetic VH + VL phagemid library, trypsin

elution was best between 10 and 30 minutes. Phagemid particles could possibly be degraded by

trypsin treatment of 45 minutes or greater. When screening scFvs, a high throughput ELISA was

acceptable. The scFv proteins can be diluted 1:10 in an ELISA with other scFv proteins and still

produce high enough signals to be detected. When screening phages, screening in a high









throughput ELISA was not optimal. Phagemid particles cannot be produced in high enough

titers when produced in microtiter wells to be useful in a high throughput manor.

Specific Aim 3: Improve Phagemid Particle Production

Phagemid particle production is one of the most important steps in biopanning. Because

the only M13 gene in a phagemid is the giII gene, use of phagemids requires an amplification

tool to be provided in trans to supply the additional M13 genes to enable phagemid particle

production. There needs to be high quality and high quantity phagemid particle production

during amplification of phagemid particles to ensure that redundancy of clones in the library is

maintained and clones are produced in high enough titers to be useful in assays that use

phagemid particles. Hyperphage (Progen) was the previous phagemid particle amplification

tool. Hyperphage is an M13-based phage that contains a partial deletion of the giII gene to

enable production of phagemid particles that display recombinant pIII proteins but no wild-type

pIII proteins. However, the commercially obtained hyperphage stock was of poor quality. The

hyperphage stock was stated to contain 1012 particles/mL; however, it only had 10s tu/mL, as

determined in an infection assay. Because there was not enough Tomlinson scFv phagemid

library stock for panning experiments, the library stock needed to be amplified. The Tomlinson I

and J scFv libraries contain 1 x 10s phagemid clones. To ensure that the Tomlinson scFv

libraries are amplified to ensure that the redundancy of clones is maintained, an amplification

tool needs to enable high quality and high quantity phagemid particle production. Because

commercially obtained hyperphage had too low of titers to be an effective phagemid particle

amplification tool, other ways of phagemid particle production were analyzed.









Escherichia coli TG1 Harboring the Hyperphage Genome Was Not an Optimal Phagemid
Particle Amplification Tool

To overcome the low titers of hyperphage, the hyperphage genome was incorporated into

E. coli and maintained in the strain as a plasmid to ensure that every phagemid-containing E. coli

also contained the hyperphage genome to enable phagemid particle production. Log phase

(OD600 ~0.4) E. coli TG1 was transduced with hyperphage at a MOI of 5, incubated in a 370p

standing water bath for 30 minutes, serially diluted in BSG, and then plated on 2xTY plates

containing 40 Cpg/mL kanamycin. The plates were incubated overnight at 370C. The next day

the plates contained small (0.5 mm diameter) and large (1-1.5 mm diameter) colonies. Two

small and large colonies were single colony passage twice. The small colonies maintained a

small phenotype (0.5 mm diameter), and the large colonies maintained a large phenotype (1-1.5

mm diameter). The two strains were named E. coli TG1 (Hye-sm) and E. coli (Hye-lg).

The infection efficiencies of phagemid particles into E. coli TGl, E. coli TG1 (Hye-sm),

and E. coli (Hye-lg) were analyzed. The three E. coli strains were grown to log phase and

transduced with serially diluted a-B SA and Vc86 phagemid particles in B SG at a MOI of 0.001.

Transductions were incubated in a 370C standing water bath for 30 minutes and plated on 2xTY

AG plates containing 40 Cpg/mL kanamycin for transduced strains ofE. coli TG1 containing

hyperphage and on 2xTY AG plates for transduced E. coli TGl. The numbers of E coli cells

transduced with a-B SA phagemid particles were as follows: 1,200 transductions into E. coli

TGl, 340 transductions into E. coli TG1 (Hye-sm), and 100 transductions into E. coli TG1

(Hye-lg). When the number of a-BSA phagemid particles transduced into E. coli TG1 was set

at an infection efficiency of 100%, the number of a-B SA phagemid particles transduced into

E. coli TG1 (Hye-sm) was 28% and the number of a-B SA phagemid particles transduced into









E. coli TG1 (Hye-lg) was 0. 1% of the number of transduced phagemid particles into E. coli

TGl.

The numbers of E coli transduced with Vc86 phagemid particles were as follows: 4,300

transductions into E. coli TGl, 950 transductions into E. coli TG1 (Hye-sm), and 80

transductions into E. coli TG1 (Hye-lg). When the number of Vc86 phagemid particles

transduced into E. coli TG1 was set at an infection efficiency of 100%, the number of Vc86

phagemid particles transduced into E. coli TG1 (Hye-sm) was 22% and the number of Vc86

phagemid particles transduced into E. coli TG1 (Hye-lg) was 2% of the number of transduced

phagemid particles into E. coli TGl. Escherichia coli TG1 (Hye-sm) had an infection efficiency

that was approximately 25% of the infection efficiency ofE coli TGl, and E. coli TG1 (Hye-lg)

had an infection efficiency that was approximately 1% of the infection efficiency of E. coli TGl.

The infection efficiencies of E coli TG1 (Hye-sm/1g) were too low to for the E. coli strains to be

useful to amplify the Tomlinson scFv libraries. However, they could still be useful as an

amplification tool to amplify single clones when the transduction efficiency is not critical.

Therefore, the quality and quantity of phagemid particles produced by E. coli TG1 (Hye-sm/1g)

were analyzed.

Escherichia coli TG1 (Hye-sm) and E. coli TG1 (Hye-lg) were analyzed for their ability

to produce phagemid particles. Escherichia coli TG1 (Hye-sm) and E. coli TG1 (Hye-lg) were

transduced with a-B SA phagemid particles (as described above) and grown under conditions to

produce phagemid particles (see Materials and Methods). Escherichia coli TG1 containing a-

BSA phagemids was superinfected and grown under conditions to produce phagemid particles

(see Materials and Methods). The superinfected E. coli TG1 was used as a comparison for

strains of E coli TG1 containing hyperphage as a phagemid particle production tool. Anti-BSA









phagemid particle yields were as follows: E. coli TG1 that was superinfected with hyperphage

yielded 3 x 10 tu/mL, E. coli TG1 (Hye-sm) yielded 1 x 109 tu/mL, and E. coli TG1 (Hye-lg)

yielded 3 x 10s tu/mL. These results showed that E. coli TG1 that contained hyperphage

produced high titers of phagemid particles.

To determine the quality of a-B SA phagemid particles produced by the three E. coli

strains, the produced a-B SA phagemid particles were analyzed in an ELISA. Microtiter plates

were coated with 10 Cpg/mL BSA or only PBS. Standard ELISA procedures were followed (see

Materials and Methods) to analyze 1 x 10s tu/mL of a-B SA phagemid particles as the primary

antibody (Fig. 3-13). Anti-BSA phagemid particles produced a significantly higher S:N (p=0.01)

when produced from E. coli TG1 (Hye-sm) compared to E. coli TG1 (Hye-lg). There was not a

significant difference in the S:N (p=0.50) from a-BSA phagemid particles produced from

superinfected E. coli TG1 and E. coli TG1 (Hye-lg). Because superinfected E. coli TG1 and

E. coli TG1 (Hye-lg) yielded titers (~3 x 10s tu/mL) of a-B SA phagemid particles that were

approximately a third of that of E coli TG1 (Hye-sm) titers (~1 x 109 tu/mL), E. coli TG1

(Hye-sm)-produced a-BSA phagemid particles were analyzed in the ELISA with a third less

phage-containing supernatant volume. The amplified a-BSA phagemid particles were from the

same clone; therefore, the differences in S:N from the produced a-BSA phagemid particles was

not expected. A possible reason for the S:N differences was that a different amount of

particles/mL was used in the ELISA. The ELISA compared equal tu/mL titers of the a-B SA

phagemid particles produced. However, tu/mL is a measure of infectious particles and not total

particles. Because an ELISA measures particles with binding activity, it is possible that

superinfected E. coli TG1 and E. coli TG1 (Hye-lg) yielded similar particles/mL yields as

E. coli TG1 (Hye-sm), which would account for the differences in the S:N. The results showed









that E. coli TG1 (Hye-sm) and E. coli TG1 (Hye-lg) had infection efficiencies that were too

low for them to be effective amplification tools to amplify the Tomlinson J scFv library.

However, these strains could be used as tools to amplify single phagemid particles with

acceptable phagemid particle production and quality.

Helper Plasmids Were Not an Optimal Phagemid Particle Amplification Tool

Helper plasmids (Los Alamos National Laboratory, Los Alamos, NM) (64) are M13-based

plasmids that are maintained in E. coli to provide phagemid-containing E. coli with all of the

genes necessary to produce phagemid particles. The helper plasmids come in three forms that

either have a full-length (M13cp), a deleted (M13cp-dg3), or a truncated (M13cp-CT) giIl gene.

The helper plasmid-containing E. coli strains were analyzed for their infection efficiencies with

phagemid particles, production yield of phagemid particles, and the quality of phagemid particles

produced.

The three helper plasmids were electroporated (see Materials and Methods) into

electrocompetent E. coli TGl. All three transformations yielded colonies between 0.5-2 mm in

diameter. A small (0.5 mm) and a large (2 mm) colony from E. coli TG1 (M13-cp), E. coli TG1

(M13cp-CT), and E. coli TG1 (M13cp-dg3) were single colony passage twice. Escherichia coli

TG1 (M13cp-sm) was the only strain that maintained a consistent phenotype of 0.5 mm colonies.

The other 5 strains yielded colonies with diameters of 0.5-2 mm after each passage. To

determine if the different colony size phenotypes had an effect on the functionality of the helper

plasmids, a small and large isolate from each strain were analyzed for their infection efficiencies

with phagemid particles.

To analyze the infection efficiencies of phagemid particles into helper plasmid-containing

E. coli TGl, an infection efficiency (see Materials and Methods) experiment was performed that

compared the infection efficiencies of helper plasmid-containing E. coli with E. coli that did not









contain helper plasmids. The bacterial strains were grown to log phase and transduced with a-

BSA or Vc86 phagemid particles at a MOI of 0. 1 (Table 3-1). The number ofE coli TGlcells

that were transduced with phagemid particles was set at an infection effciency of 100%, and the

number of E coli TG1 cells harboring helpers that were transduced with phagemid particles was

compared to the infection effciency of E coli TGl. Of the six E. coli TG1 strains harboring

helper plasmids, E. coli TG1 (M13cp-dg3-sm) and E. coli TG1 (M13cp-CT-sm) were the only

strains that yielded infection effciencies grater than 0.3% of the infection effciency of E. coli

TGl. Escherichia coli TG1 (M13cp-dg3-sm) had infection effciencies of 175% and 97%, while

E. coli TG1 (M13cp-CT-sm) had infection effciencies of 7% and 13% compared to the infection

effciency of E. coli TGl. Because the small and large isolates of helper plasmid-containing

E. coli had different infection effciencies, the colony phenotype had an effect of the

functionality of the helper plasmid-containing E. coli. Because initial tests with E. coli TG1

(M13cp-dg3-sm) had an infection effciency that was as high or higher than the infection

effciency of E coli TGl, its infection effciency was further analyzed with the transduction of

additional phagemid particles.

Escherichia coli TG1 (M13cp-dg3-sm) was transduced with clone 18 (a phagemid particle

isolated from the Tomlinson I scFv library that recognizes the A27L protein of vaccinia virus),

a-AV20N3 (a phagemid particle isolated from the Tomlinson I scFv library that recognizes

rAuto (a recombinant L. monocytogenes murein hydrolase)), and Vc86 phagemid particles. The

various strains were analyzed further for their infection effciencies compared to E. coli TG1

(Table 3-2). Escherichia coli TG1 (M13cp-dg3-sm) had infection effciencies of 92%, 1 13%,

and 120% compared to the infection effciency ofE. coli TGl. Escherichia coli TG1 (M13cp-

dg3-sm) had infection effciencies that were approximately the same as E. coli TGl; therefore,









this strain was used for further analysis to determine the quantity and the quality of phagemid

particles produced. However, before E. coli TG1 (M13cp-dg3-sm) was analyzed further,

M13cp-CT-sm was examined.

To test if the helper plasmid (M13-CT-sm) could yield a higher infection efficiency in

another E. coli strain, it was extracted (see Materials and Methods) from E. coli TG1 (M13cp-

CT-sm) and electroporated (see Materials and Methods) into E. coli JM109. Escherichia coli

JM109 was chosen because it was genotypically more similar than E. coli TG1 to the bacterial

strain that Los Alamos National Laboratory used with the helper plasmids (E. coli DH~a F').

Escherichia coli JM109 and E. coli DH~a F' contain relA l and recA1i, while E. coli TG1 does

not.

Escherichia coli TG1 (M13cp-CT-sm) and E. coli JM109 (M13cp-CT-sm) were

transduced with clone 18, a-AV20N3, and Vc86 phagemid particles and were analyzed for their

infection efficiencies compared to E. coli TG1 and E. coli JM109 (Table 3-3). The average

infection efficiency of E coli TG1 (M13cp-CT-sm) was 47%, while the average infection

efficiency of E coli JM109 (M13cp-CT-sm) was 28%. Therefore, using E. coli JM109 to

maintain M13cp-CT-sm did not increase the infection efficiency.

In analyzing the infection efficiencies, E coli TG1 (M13cp-dg3-sm) had an average

infection efficiency of 120% with a standard deviation of 33% and E. coli TG1 (M13cp-CT-sm)

had an average infection efficiency of 30% with a standard deviation of 16%. Escherichia coli

TG1 (M13cp-dg3-sm) had a significantly higher infection efficiency (p=0.01) than E. coli TG1

(M13cp-CT-sm).

Escherichia coli TG1 (M13cp-dg3-sm), E. coli TG1 (M13cp-CT-sm), and E. coli TG1 that

was superinfected with hyperphage were analyzed for their ability to produce phagemid









particles. To determine if the addition of 1 mM IPTG to the growth medium would increase

phagemid particle production, phagemid particle amplification was performed in parallel with

and without the addition of 1 mM IPTG to the growth medium. Isopropyl-beta-D-

thiogalactopyranoside is a molecular mimic of a lactose metabolite that induces transcription

from the lac promoter. Because the lac promoter drives transcription of the glll-scFv fusion, it

was hypothesized that increased expression of pIll-scFv fusions would result in higher yields of

phagemid particles. Phagemid particles were amplified (see Materials and Methods) and titered

by the spot-titer method (see Materials and Methods) (Table 3-4). Phagemid particles were

amplified to titers of 106-108 tu/mL by E. coli TG1 (M13cp-dg3-sm), 105 -106 tu/mL by E. coli

TG1 (M13cp-CT-sm), and 10s -109 tu/mL by E. coli TG1 superinfected with commercial

hyperphage. Titers of phagemid particles produced with 1 mM IPTG were either the same or

lower than titers produced without IPTG. Therefore, it was determined that the addition of 1

mM IPTG did not improve phagemid particle production. Phagemid particles produced by

E. coli TG1 (M13cp-dg3-sm) and E. coli TG1 (M13cp-CT-sm) yielded low titers of phagemid

particles. An overnight shaking culture for the amplification of phagemid particles contains

approximately 109 CFU/mL. Therefore, for the helper plasmids to yield titers as low as 105

tu/mL means that only 1 in 10,000 bacteria produced an infectious phagemid particle. These

titers were too low for either helper plasmid to be used as an effective phagemid particle

production tool.

Even though helper plasmids produced phagemid particle titers too low to be used as an

amplification tool, the phagemid particles produced by E. coli TG1 (M13cp-dg3-sm), E. coli

TG1 (M13cp-CT-sm), and E. coli TG1 superinfected by hyperphage were further analyzed by

ELISA. Phagemid particles were serially diluted and analyzed in a standard ELISA procedure









(see Materials and Methods). Analysis of ELISA data showed that S:N values as high as 7 when

less than 5 x 104 tu/mL of phagemid particles were used as a primary antibody. Previous data in

the laboratory showed that phagemid particles were not detectable in an ELISA unless they were

used at concentrations of at least 1 x 106 particles/mL. Spot titering is a measure of transducing

units (number of infectious units), while an ELISA is a measure of the number of particles that

possess antibody activity. To determine if the measured tu/mL were underestimating the

phagemid particles/mL, an anti-M13 sandwich ELISA was performed to quantify the number of

phagemid particles produced by E. coli TG1 (M13cp-dg3-sm), E. coli TG1 (M13cp-CT-sm), and

E. coli TG1 superinfected by hyperphage. A microtiter plate was coated with anti-M13 pVIII

monoclonal antibody at a concentration of 10 Cpg/mL. Phagemid particles were incubated with

the coated wells at dilutions of 1:2, 1:20, or 1:200. The phagemid particles were then detected

using HRP-conjugated anti-M13 monoclonal antibody at a dilution of 1:2,500, followed by

standard ELISA procedures (see Materials and Methods). A concentrated phagemid preparation,

which had been amplified by hyperphage, was serially diluted and used as a standard for

particles/mL. The values for particles/mL were linearly extrapolated from the standard curve of

the standard phagemid particle (Table 3-4). Phagemid particles amplified by E. coli TG1

(M13cp-CT-sm) had particles/mL to tu/mL ratios between 740 and 23,000. Therefore, phagemid

particles were being produced in high quantities; however, less than 0. 1% of the phagemid

particles produced were infectious. Phagemid particles amplified by E. coli TG1 (M13cp-dg3-

sm) had particles/mL to tu/mL ratios between 1 and 190. Phagemid particles amplified by

E. coli TG1 that was superinfected with hyperphage had particles/mL to tu/mL ratios between 6

and 24. Of the three amplification tools (M13cp-dg3-sm, M13cp-CT-sm, and commercial









hyperphage), commercial hyperphage was the only tool that could amplify phagemid particles to

titers that were high enough to be used in phage display applications.

Phagemid particles amplified by helper plasmids in E. coli TG1 yielded phagemid particle

titers between 105 and 10s tu/mL. Helper plasmids amplified phagemid particles to titers too low

to be useful in biopanning. To determine if another E. coli strain could improve the titers of

helper plasmid-amplified phagemid particles, E. coli MGl655 containing helper plasmid

(M13cp-dg3-sm) was used to analyze phagemid particle production. Escherichia coli MGl655

was chosen because it has few known mutations and lacks the F plasmid. Therefore, it was

hypothesized that the stress upon the E. coli for maintaining the helper plasmids would be

lessened if it did not already have to maintain the F plasmid. Four different phagemid particles

were amplified with E. coli MGl655 (M13 cp-dg3-sm), and the resulting titers were: 2.5 x 10 ,

6.3 x 106, 2.5 x 10 and 3.9 x 107 tu/mL for an average of 2.4 x 107 tu/mL, which was still too

low to be useful in biopanning.

Homemade Hyperphage Titers Were Increased with Amplification in E. coli MG1655
(pGTR203)

Commercial hyperphage was the only phagemid particle amplification tool tested that

produced phagemid particle progeny in yields high enough to be used in biopanning. However,

the titer of the commercial hyperphage stock was not high enough to be used to amplify the

Tomlinson scFv libraries. Therefore, a way to improve homemade hyperphage production was

analyzed. Because hyperphage has a partial deletion of its giIl gene, it needs the giIl gene to be

provided in trans in the E. coli strain used to produce hyperphage particles. Homemade

hyperphage was previously produced by using pGTR203, a plasmid that encodes giII.

Escherichia coli TG1 (pGTR203) or E. coli HB2151 (pGTR203) were infected with hyperphage,

and the transduced bacteria were grown under conditions that promoted hyperphage production.









In the past, hyperphage amplified from E. coli TG1 (pGTR203) yielded titers between 107 to 10s

tu/mL and E. coli HB215 1 (pGTR203) yielded titers between 106 to 107 tu/mL. Both of these

strains are F+E coli. Because hyperphage infects bacteria by the F pilus, it was hypothesized that

some of the produced hyperphages could have been taken up by the bacteria. Therefore, it was

tested if hyperphage could be produced in higher titers when produced in an F E. coli.

Two F- E. coli strains, E. coli MGl655 and E. coli EC100D, were used to amplify

hyperphage. The strains were made electrocompetent (see Materials and Methods) and

transformed by electroporation (see Materials and Methods) with pGTR203. Escherichia coli

MGl655 (pGTR203) and E. coli EC 100D (pGTR203) were made electrocompetent and

transformed by electroporation with the hyperphage genome that had been extracted from E. coli

TG1 (hyperphage) with a QIAprep Spin Miniprep kit (Qiagen) (see Materials and Methods).

Five colonies from each transformation with hyperphage, E. coli MGl655 (pGTR203) (Hye) 1-

5 and E. coli EC 100D (pGTR203) (Hye) 1-5, were single colony passage twice and analyzed

for their ability to produce hyperphage.

Standing overnight cultures of all ten isolates were diluted 1:400 in broth and grown under

conditions that promote phage production (see Materials and Methods). The hyperphage

produced were analyzed for tu/mL by the whole plate titer method (see Materials and Methods)

and for particles/mL by an anti-M13 sandwich ELISA (described above) (Table 3-5). The titers

of hyperphage produced from the five E. coli EC 100D (pGTR203) (Hye) isolates 2. 1 x 10 & 2.5

x 10s tu/mL were significantly lower than the hyperphage produced from the five E. coli

MGl655 (pGTR203) (Hye), 3.0 x 10912.0 x 109 tu/mL (p=0.02). The ratios of particles/mL to

tu/mL of hyperphages produced from E. coli EC 100D (pGTR203) (Hye) were between 0.4 and

7.8, and the ratios of particles/mL to tu/mL of hyperphages produced from E. coli MGl655









(pGTR203) (Hye) were between 0.3 and 1.0. Therefore the amount of particles and transducing

units of hyperphage produced from E. coli EC 100D (pGTR203) (Hye) and E. coli MGl655

(pGTR203) (Hye) were approximately the same, meaning that the hyperphages produced were

infectious. Both F- E. coli strains generated higher titers of hyperphage than did the two F

E. coli strains. Concentration of hyperphage produced from E. coli MGl655 (pGTR203) (Hye)

should generate a high enough titer of hyperphage to enable high quality and high quantity

production of the Tomlinson scFv libraries.

Escherichia coli MGl655 (pGTR203) (Hye) isolate 5 yielded the highest titer of

hyperphage (5.4 x 109 tu/mL); therefore, it was used to produce a large quantity of hyperphage to

be used to amplify the Tomlinson J scFv phagemid library. One liter of hyperphage was

produced from E. coli MGl655 (pGTR203) (Hye) 5 with a titer of 2.0 x 109 tu/mL, and the

phages were concentrated by PEG precipitation (see Materials and Methods) to a volume of 2.8

mL with a final titer of 3 x 1011 tu/mL. This concentrated hyperphage was used to amplify the

Tomlinson J scFv phagemid library.

The original Tomlinson J scFv library was amplified to ensure high quality and high

quantity of phagemid particles produced. The Tomlinson J scFv phagemid library was originally

received in E. coli TGl. This stock was grown in batch culture and frozen. Therefore, clones

were not lost due to poor phagemid particle amplification with the low titer hyperphage because

we grew the original bacterial stock of the library and not the previously made phagemid particle

stock of the library that most likely lost redundancy of clones because it was amplified with low

titer commercial hyperphage. A frozen stock of E coli TG1 (Tomlinson J scFv phagemid

library) was thawed on ice and inoculated into one liter of 2xTY AG broth. This culture was

grown in a 370C shaking incubator until the culture reached log phase. Thirty minutes into the









incubation an aliquot of the growing culture was removed and titered to ensure that the library

was initially in a concentration high enough that redundancy of clones was present. The titer of

the E. coli TG1 (Tomlinson J scFv phagemid library) 30 minutes into incubation was 1.3 x 107

CFU/mL, and because the culture volume was one liter there were 1.3 x 1010 CFU. The

Tomlinson J phagemid library contains ~1 x 10s phagemid clones; therefore, there was initially a

100-fold redundancy of clones in the library. Approximately 2.3 x 1010 CFU of log phase E. coli

TG1 (Tomlinson J scFv phagemid library) was superinfected with homemade hyperphage at a

MOI of 10 and incubated for 30 minutes in a 370C standing incubation. The superinfected

bacteria were then titered on 2xTY AG plates containing 40 Cpg/mL kanamycin to ensure that

enough bacteria were superinfected so that the redundancy of the phagemid clones was

maintained. Approximately 2.9 x 1010 bacteria containing a phagemid were superinfected, which

represents about 300-fold redundancy of clones. The superinfected bacteria were then grown in

1 liter of broth under conditions that promoted phagemid particle production (see Materials and

Methods). The final yield Tomlinson J scFv phagemid particles in the amplified culture was

6.6 x 109 tl/mL. The liter of phagemid particles was then PEG precipitated to 15 mL of

4.4 x 1011 tu/mL, which yields 6.6 x 1012 tu. Titering throughout the Tomlinson J scFv library

amplification ensured that the redundancy of the library clones was maintained at each step. The

Tomlinson J scFv library that was amplified with homemade hyperphage was frozen and also

used for biopanning.

Conclusion of Specific Aim 3

Escherichia coli TG1 harboring hyperphage was not an optimal phagemid particle

amplification tool because of the low infection efficiency with phagemid particles. The infection

efficiencies of phagemid particles into E. coli TG1 (Hye-sm) and E. coli TG1 (Hye-lg) were

28% or less than the infection efficiency of phagemid particles into E. coli TGl. Phagemid









particle amplification by helper plasmids was not an optimal amplification tool because it

produced phagemid particle titers too low to be used in biopanning. Producing homemade

hyperphage in E. coli MGl655 (pGTR203) enabled high yield of hyperphage particles. The

improved high titer homemade hyperphage was used to successfully amplify the Tomlinson J

scFv phagemid library to ensure that the redundancy of phagemid clones was maintained.

Specific Aim 4: Isolation of Specific Recombinant Phage to Stx2 Toxin of E. coli 0157:H17

The Tomlinson J scFv phagemid library that was amplified with homemade hyperphage

was panned against Stx2 toxin of E coli 0157:H7. Shiga toxins (Stx) kill mammalian cells and

cause severe disease from infection with E. coli 0157:H7 cells. Escherichia coli 0157:H7 can

produce Shiga toxins 1 and/or 2. Escherichia coli harboring the Stx2 toxin appears to be more

virulent than E. coli harboring Stxl or both Stxl and 2 toxins (36-39). Therefore, Stx2 toxin was

the target molecule chosen for panning. The Stx2 toxin preparation that was used to pan against

was obtained from Toxin Technology (Sarasota, FL) and was stated to contain 50% Stx2 toxin.

The other 50% of proteins in the Stx2 toxin preparation was most likely composed of

contaminating E. coli HB 101 proteins because the Stx2 toxin preparation was prepared from

E. coli HB 101 (pMJ100) lysates. pMJ100 encodes Stx2 toxoid.

A three round biopanning process was performed to pan the Tomlinson J scFv phagemid

library against a Stx2 toxin preparation-coated immunotube (see Materials and Methods) with

amplification between panning rounds to amplify the selected phagemid particles. Phagemid

particles were eluted using trypsin treatment for 10 minutes. Phagemid particles contain a

trypsin cleavage site between the scFv and the pIll. Therefore, phagemid particles that were

bound to Stx2 toxin preparation by their scFv could be eluted from the Stx2 toxin preparation to

be amplified for further rounds of panning. Approximately 1 x 1012 phagemid particles were

added to an immunotube that was coated with 100 Cpg of Stx2 toxin preparation. Approximately









2.2 x 106 phagemid particles were eluted after round one of panning. The round-one-eluted

phages were amplified with homemade hyperphage to 2 x 1010 tu/mL in 100 mL ofbroth. The

amplified phagemid particles were then PEG precipitated to 9 x 10l tu/mL in 2.5 mL PBS.

Approximately 1 x 1012 phagemid particles from the PEG-precipitated phagemid particles from

round one were panned against Stx2 for a second round of panning. Approximately 2.0 x 107

phagemid particles were eluted after round two of panning. The round-two-eluted phages were

amplified with homemade hyperphage to 2.5 x 1010 tu/mL in 100 mL broth. The amplified

phagemid particles were then PEG precipitated to 7.8 x 10l tu/mL in 2.5 mL PBS.

Approximately 1 x 1012 phagemid particles from the PEG-precipitated phagemid particles from

round two were panned against Stx2 for a third round of panning. Approximately 1.5 x 1010

phagemid particles were eluted after the third round of panning. The increase in the number of

phagemid particles eluted (2 x 106 for round one, 2 x 10' for round two, 2 x 10'0 for round three)

after every round suggests that phagemid particles that specifically bound to Stx2 toxin

preparation were being selected and amplified.

Eighty-four phagemid particles from the third round of panning were individually

amplified with homemade hyperphage (see Materials and Methods) to be analyzed for specificity

by ELISA and Western blot experiments. Forty-four clones were initially analyzed by ELISA.

Microtiter wells coated with 10 Cpg/mL Stx2 toxin preparation were incubated with phagemid

particle-containing supernatants that were diluted 1:2 in blocking buffer, 1 Cpg/mL anti-

Stx211E10 (a-Stx2 subunit A) monoclonal antibody (Toxin Technology), 1 Cpg/mL anti-Stx2-

BBl12 (a-Stx2 subunit B) monoclonal antibody (Toxin Technology), or 1 x 10s tu/mL a-BSA

phagemid particles (negative control for determining the noise for S:N calculations) (Fig. 3-14).

Of the 44 clones screened, 40 had a S:N between 9 and 18. The two anti-Stx toxin monoclonal









antibodies had a S:N less than 2 against Stx2 toxin preparation. The Stx2 toxin preparation may

not have bound to the microtiter wells very efficiently, or the anti-Stx2 monoclonal antibodies

may not have worked very well. Conversely, the phagemid clones that were positive may have

bound to the microtiter well instead of the Stx2 toxin preparation.

To determine if the clones were binding to Stx2 or to the microtiter well, 4 of the

previously screened clones along with 44 new clones were analyzed via ELISA using the same

procedure as above with the addition of PBS-treated microtiter wells as a negative control (Fig.

3-15). Of the 48 clones, 44 had signals of 0.5 or higher against Stx2 toxin preparation-coated

microtiter wells. Four clones (clones 1, 23, 35, and 39) had signals as high or higher against

PBS-treated microtiter wells compared to Stx2 toxin preparation-coated microtiter wells. These

four clones could possibly recognize polystyrene. Because the library was panned against Stx2

toxin preparation-coated immunotubes, it was possible that some clones were selected that bound

to the polystyrene immunotube. Fifty-three of 59 clones (89%) had a S:N from 2 to 23, and 48

of 59 clones (81%) had a S:N from 10 to 23 against Stx2 toxin preparation. Therefore, the

majority of the clones screened bound to the Stx2 toxin preparation. However, 4/59 (7%) of the

clones bound to PBS-treated wells as well as Stx2-coated wells. The four clones had a signal in

an ELISA between 0.31-0.87 against PBS-treated wells. These clones most likely bound to

polystyrene. Isolation of plastic binding phages is fairly common (84). No matter how well the

panning vessel is blocked with blocking buffer, phages can still be selected that bind to the

panning vessel. Of the 59 clones screened, 2 (3%) did not generate a positive signal against

Stx2-coated microtiter wells or non-coated microtiter wells. These clones had unknown binding

specificity.









The anti-Stx monoclonal antibodies (a-Stx2 subunit A and a-Stx2 subunit B) were further

analyzed to determine the reason why they did not react in the ELISA. Microtiter wells coated

with 10 Cpg/mL Stx2 toxin preparation or carbonate bicarbonate buffer were incubated with a-

Stx2 subunit A and a-Stx2 subunit B monoclonal antibodies at concentrations of 1, 10, or 100

Cpg/mL. Bound antibodies were detected with HRP-conjugated anti-mouse antibody (Cappel) or

HRP-conjugated anti-mouse antibody (Sigma-Aldrich). Standard ELISA procedures followed

(see Materials and Methods). All parameters generated S:N values less than two. The Stx2 toxin

preparation appeared to be binding to the well because the amplified panning clones were

binding to the Stx2 with high S:N values in microtiter wells. The concentrations of anti-Stx2

monoclonal antibodies were not an issue because using the monoclonal antibodies at a

concentration as high as 100 Cpg/mL did not improve the signal of the monoclonal antibodies in

an ELISA. Two different HRP-conjugated anti-mouse antibodies were used to determine if the

secondary antibody was not recognizing the primary antibodies. Neither HRP-conjugated anti-

mouse antibody improved the signal of the monoclonal antibodies in an ELISA. Positive

controls for the anti-mouse antibodies were not performed in the ELISA to determine whether or

not the anti-mouse antibodies were functional. However, the two HRP-conjugated anti-mouse

antibodies were used within three days of these experiments and generated positive signals in an

ELISA against murine antibodies. The reason why the anti-Stx monoclonal antibodies gave low

signals with Stx2 toxin preparation in an ELISA was not determined.

Analysis by ELISA shows that the maj ority of clones isolated from the Stx2 panning

bound to the Stx2 toxin preparation in an ELISA. Because the Stx2 toxin preparation was only

50% pure for Stx2 toxin, the panning clones could have been binding to Stx2 toxin or the 50% of

impurities in the preparation. To analyze if the Stx2 toxin preparation panning clones recognized









the Stx2 toxin or the impurities, three clones were selected and examined for reactivity with the

Stx2 toxin preparation and E. coli supernatants or periplasmic break fractions. The strains

examined were E. coli 0157:H7 933 (expresses Stxl and 2 toxins), E. coli DH~a (pNR100)

(expresses Stx2 toxoid), and E. coli DH~a. Microtiter wells were coated with 0.01 mg/mL Stx2

toxin preparation, 0.07 mg/mL E. coli 0157:H7 933 periplasmic break fraction, 2 mg/mL E. coli

0157:H7 933 supernatant, 1 mg/mL E. coli DH~a (pNR100) periplasmic break fraction, 1

mg/mL E. coli DH~a periplasmic break fraction, or carbonate coating buffer. If the clones

bound to Stx2 toxin, they were expected to generate positive signals to all the antigens except for

E. coli DH~a and carbonate coating buffer. If the clones bound to impurities in the Stx2 toxin

preparation, the clones were expected to generate positive signals for all the antigens except

carbonate coating buffer. Coated wells were incubated with 10 Cpg/mL of a-Stx2 subunit A

monoclonal antibody (11E10), 10 Cpg/mL of a-Stx2 subunit B monoclonal antibody (BBl2), 1 x

109 tu/mL of phagemid particles of Stx2 clones (46, 48, and 49), or a-BSA phagemids. Standard

ELISA procedures were followed (see Materials and Methods) (Fig. 3-16). All parameters

generated S:N values of less than 1.5 except for clones 46, 48, and 49 that were analyzed with

Stx2 toxin preparation-coated microtiter wells. Because the monoclonal antibodies to Stx2 toxin

preparation did not generate positive signals in an ELISA against any of the parameters, it was

not determined if the clones from the Stx2 toxin preparation panning were reacting to Stx2 toxin

or impurities in the commercially obtained Stx2 toxin preparation.

Clones that were selected from the Stx2 toxin preparation panning and had a positive

signal by ELISA to Stx2 toxin preparation were further analyzed by Western blot to determine if

the clones were binding to Stx2 toxin or to impurities in the Stx2 toxin preparation. Six clones

were screened in a Western blot against Stx2 toxin preparation, E. coli 0157:H7 933 (expresses









Stxl and 2 toxins) supernatant, and E. coli 0157:H7 87-23 (Stx phage cured). Because the Stx2

toxin is composed of A and B subunits that are approximately 32 and 10-kDa proteins, a clone

that bound to Stx2 toxin was expected to recognize a 32 or 10-kDa protein in the Stx2 toxin

preparation and E. coli 0157:H7 933 supernatant but not in E. coli 0157:H7 87-23 supernatant.

A clone that bound to impurities was expected to recognize proteins in all three antigens.

A 4-20% polyacrylamide gel was loaded with 1.5 Cpg/well of Stx2 toxin preparation, 120

mg/well of E coli 0157:H7 933 supernatant, and 120 mg/well of E coli 0157:H7 87-23

supernatant. The resolved proteins were transferred to nitrocellulose and analyzed by Westemn

blot (see Materials and Methods). Phagemid particles (a-BSA and Stx2 toxin preparation

panning clones 46, 48, 80, 81, and 83) were used as a primary antibody at approximately 109

tu/mL, and a-Stx2 subunit A monoclonal antibody was used at 4 Cpg/mL. Anti-Stx2 subunit B

monoclonal antibody failed to generate a positive signal by Western blot. Standard Westemn blot

procedures were performed (see Materials and Methods) (Fig. 3-17). The a-Stx2 subunit A

monoclonal antibody recognized a 32-kDa (the size of the A subunit) protein in the Stx2 toxin

preparation and in the E. coli 0157:H7 933 supernatant, thereby proving that Stx2 toxin subunit

A was present in both preparations but not in the E. coli 0157:H7 87-23 supernatant as was

expected. Anti-BSA failed to recognize any of the antigens, as was expected. Clone 46 failed to

recognize any of the three antigens. Clones 48 recognized 35, 45, and 55-kDa proteins in the

Stx2 toxin preparation and a 35-kDa protein in both E. coli 0157:H7 supernatant. This result

means that clone 48 probably bound to bacterial protein impurities in the Stx2 toxin preparation.

Clones 49 and 80 bound to a 35-kDa protein in all three antigens and a 20-kDa protein in the

Stx2 toxin preparation and in E. coli 0157:H7 87-23 supernatant. Therefore, clones 49 and 80

probably bound to bacterial protein impurities in the Stx2 toxin preparation. Clones 81 and 83










appeared to recognize the same bacterial protein impurities as clones 49 and 80, except that their

band intensities were much lighter.

Western blot and ELISA analyses showed that clones obtained from panning against the

Stx2 toxin preparation recognized the Stx2 toxin preparation. The clones that have been

screened so far appeared to bind to impurities in the Stx2 toxin preparation. To determine if the

Stx2 toxin preparation contained 50% Stx2 toxin as stated by the producer, the toxin preparation

was resolved and analyzed by SDS-PAGE (Fig. 3-18). If the Stx2 toxin preparation was

comprised of 50% Stx2 toxin, 10 and 32-kDa protein bands should have been the predominant

protein bands. However, a protein band at approximately 32 kDa was a minor protein band. A

protein band at 10 kDa was a maj or protein band; however, whether these protein bands were

actually the Stx2 toxin subunits was not determined. The 10 and 32-kDa protein bands

comprised less than 10% of the Stx2 toxin preparation. Because the concentration of Stx2 toxin

was so low in the commercially obtained Stx2 toxin preparation, it was probable that the

maj ority of the phagemid particles isolated from the Stx2 panning would recognize impurities

instead of Stx2 toxin.

Conclusion of Specific Aim 4

Panning against commercially obtained Stx2 toxin preparation was successful in isolating

phagemid particles that bound to the Stx2 toxin preparation. Initial screening of clones by

Western blot showed that the clones bound to impurities in the commercially obtained Stx2 toxin

preparation and not Stx2 toxin. The commercially obtained Stx2 toxin preparation was supposed

to contain 50% Stx2 toxin; however, SDS-PAGE analysis showed that less no more than 10% of

the proteins present in the preparation may have been Stx2 toxin. Therefore, it was likely that

the maj ority of clones selected from the panning would recognize impurities in the Stx2 toxin

preparation and not Stx2 toxin. However, this panning exercise showed that the Tomlinson J









library amplified with the homemade hyperphage was a more effective tool than the previously

amplified libraries.









Table 3-1. Infection efficiencies ofE. cobi TG1 containing various helper plasmids with
phagemid particles. Bacterial strains were transduced with anti-BSA (a-BSA) and
Vc86 phagemid particles at a MOI of 0. 1, and transduced bacteria were enumerated.
Infection efficiency is the number of transductions into a bacterial strain divided by
the number of transductions into E. cobi TG1 and is represented as a percentage.
Infection Infection Average
efficiency efficiency infection
Transductions with Transductions with Vc86 efficiency
Strain with a-BSA a-BSA (%) with Vc86 (%) (%)
E. cobi TG1 1.2 x 106 100 3.6 x 106 100 100

E. cobi TG1 9.3 x 102 0.1 6.9 x 103 0.3 0.2
(M13cp-sm)
E. cobi TG1 8.3 x 102 0.1 9.6 x 103 0.3 0.2
(M13cp-lg)
E. cobi TG1 7.9 x 104 7 4.8 x 105 13 10
(M13cp-CT-sm)
E. cobi TG1 <1.3 x 102 <0.01 <1.3 x 102 <0.01 <0.01
(M13cp-CT-lg)
E. cobi TG1 2.1 x 106 175 3.6 x 106 97 136
(M13cp-dg3-sm)
E. cobi TG1 <1.3 x 102 <0.01 <1.3 x 102 <0.01 <0.01
(M13cp-dg3-lg)

Table 3-2. Infection efficiencies of E cobi TG1 (M13cp-dg3-sm) with phagemid particles.
Bacterial strains were transduced with phagemid particles at a MOI of 0. 1, and
transduced bacteria were enumerated. Infection efficiency is the number of
transductions with a phagemid into a bacterial strain divided by the number of
transductions with the same phagemid into E. cobi TGl.
Infection
Strain Phagemid Transductions efficiency (%)
E. cobi TG1 Vc86 5.3 x 106 100
E. cobi TG1 (M13cp-dg3-sm) Vc86 4.9 x 106 92
E. cobi TG1 Clone 18 2.4 x 106 100
E. cobi TG1 (M13cp-dg3-sm) Clone 18 2.7 x 106 113
E. cobi TG1 a-AV20N3 1.5 x 106 100
E. cobi TG1 (M13cp-dg3-sm) a-AV20N3 1.8 x 106 120


































Table 3-4. Comparison of transducing units to particles per milliliter of amplified phagemid
particles. The spot-titer method was used to determine tu/mL. Particles/mL was
determined via a sandwich ELISA with anti-M13 antibodies. (n = 2 wells)
Phagemid
particle Amplification (Particles/mL)/
amplified tool used tu/mL Particles/mL (tu/mL)
Vc86 M13cp-CT-sm 8.7 x 105 2.0 x 1010 23000
Vc86 M13cp-CT-sm 4.6 x 106 3.4 x 109 740
clone 18 M13 cp-dg3 -sm 2.4 x 10s 2.2 x 10s 1
a-AV20N3 M13cp-dg3-sm 4.4 x 107 7.0 x 107 2
a-BSA M13cp-dg3-sm 1.3 x 10s 1.8 x 10s 1
Vc86 M13cp-dg3-sm 8.7 x 106 1.7 x 109 190
clone 18 Hyperphage 7.1 x10s 1.1 x1010 15
a-AV20N3 Hyperphage 2.0 x 109 1.1 x 1010 6
a-BSA Hyperphage 5.8 x 10s 1.4 x 1010 24


Table 3-3. Infection efficiencies of E coli TG1 (M13cp-CT-sm) and E. coli JM109 (M13cp-CT-
sm) with phagemid particles. Bacterial strains were transduced with phagemid
particles at a MOI of 0. 1, and transduced bacteria were enumerated. Infection
efficiency is the number of transductions with a phagemid particle into a helper
plasmid-containing bacterial strain divided by the number of transductions with the
same phagemid particle into the bacterial strain without the helper plasmid.
Infection
Strain Phagemid Transductions efficiency %


E. coli TG1
E. coli TG1 (M13cp-CT-sm)
E. coli TG1
E. coli TG1 (M13cp-CT-sm)
E. coli TG1
E. coli TG1 (M13cp-CT-sm)
E. coli JM109
E. coli JM109 (M13cp-CT-sm)
E. coli JM109
E. coli JM109 (M13cp-CT-sm)
E. coli JM109
E. coli JM109 (M13cp-CT-sm)


Vc86
Vc86
Clone 18
Clone 18
a-AV20N3
a-AV20N3
Vc86
Vc86
Clone 18
Clone 18
a-AV20N3
a-AV20N3


2.4 x 10
1.3 x 10
1.4 x 10
5.5 x 104
9.3 x 104
4.4 x 104
6.8 x 10
1.5 x 10
1.7 x10
6.1 x104
1.4 x 10
3.7 x 104


100
54
100
39
100
47
100
22
100
36
100
26




























\ j \ _I j


Table 3-5. Comparison of transducing units to particles per milliliter of amplified homemade
hyperphage produced from F- E. coli strains. The spot-titer method was used to
determine tu/mL. Particles/mL was determined via a sandwich ELISA with anti-M13
antibodies. (n = 2 wells).


(Particles/mL)/
(tu/mL)


Hyperphage production strain
E.coli EC100D (pGTR203) (Hye) 1
E.coli EC 100D (pGTR203) (Hye) 2
E.coli EC 100D (pGTR203) (Hye) 3
E.coli EC 100D (pGTR203) (Hye) 4
E.coli EC100D (pGTR203) (Hye) 5
E.coli MGl655 (pGTR203) (Hye) 1
E.coli MGl655 (pGTR203) (Hye) 2
E.coli MGl655 (pGTR203) (Hye) 3
E.coli MGl655 (pGTR203) (Hye) 4
E.coli MGl655 (pGTR203) (Hye) 5


tu/mL
5.3 x 10
9.4 x 10
3.2 x 10
2.9 x 106
6.0 x 10
1.0 x109
8.0 x 10
3.0 x109
4.3 x 109
5.4 x 109


Particles/mL
4.2 x 10
3.5 x10s

< 1.0 x107
2.1 x10s
1.0 x109
7.3 x 10s
1.3 x 109
1.4 x109
1.5 x109








A


B


250
150
100
ye


-~r


20


Figure 3-1. Analysis of TRlzol Reagent-extracted y. cholerae N16961 LPS by SDS-PAGE.
Samples were resolved on a 12% (w/v) polyacrylamide gel and stained with A)
Coomassie blue stain or B) Tsai-Frasch silver stain. (1) TRIzol-extracted
yl cholerae N16961 LPS. (2) y. cholerae 569B LPS (Sigma-Aldrich).


MWV


MWV "1


250
150
100
75

50
37
25;


so -'
37


25
20 & 1



15 -
10










A


MW 1 2 3 4 5 6


MWV 1 2 3 4 5

250(
150
100
75 .
50 --,,
37
25 -
20
15
10


F figure 3 -2. Analy si s of phenol -water-extracte d y. cholerae 5 69B LP S by SD S-PAGE. Samples
were resolved on a 4-20% (w/v) polyacrylamide gel and stained with A) Coomassie
blue stain or B) Tsai-Frasch silver stain. (1-3) y. cholerae 569B LPS (Sigma-
Aldrich) at dilutions of 1:4 (1), 1:8 (2), and 1:16 (3). (4-6) Phenol -water-extracted
yK cholerae 569B LPS at dilutions of 1:2 (4), 1:4 (5), and 1:8 (6).









10



*,-e ac ~c-Vc 01
6
/ LPS mAb
z (1 p-g/mL)
-m-- a-Vc 01
2 LPS pAb
(1:100)

0.1 1 10 100 1000

V. cholerae LPS (pg/mL)

Figure 3-3. Analysis of the saturation limit of y. cholerae 569B LPS to microtiter wells by
ELISA. Signal to noise (S:N) ratios were calculated for the reactivity of mouse anti-
K~ cholerae 01 LPS monoclonal antibody (a-Vc 01 LPS mAb) (1 Cpg/mL) and rabbit
anti-K cholerae 01 LPS polyclonal antibody (a-Vc 01 LPS pAb) (1:100) to
K~ cholerae 569B LPS (0.1-200 Cpg/mL)-coated microtiter wells. Bound primary
antibodies were detected with either HRP-conjugated goat anti-rabbit antibody or
HRP-conjugated goat anti-mouse antibody. Wells were developed with TMB
substrate, and their absorbances read at 630 nm. (n=3 wells). There was a significant
decrease (*p=0.01) in S:N when the concentration of LPS decreased from 1 to 0.1
Cpg/mL when using the a-Vc 01 LPS mAb. There was a significant decrease
(*p=0.01) in the S:N when the concentration of LPS decreased from 10 to 1 Cpg/mL
when using the a-Vc 01 LPS pAb.









A 12


10




2 6








0.0 0.5 1.0 1.5

a-Vc 01 LPS mAb (1 pg/mL)


Figure 3-4. Analysis of the saturation limits of primary antibodies to K. cholerae 569B LPS-
coated microtiter wells by ELISA. Primary antibodies: A) mouse anti-Y. cholerae
01 LPS monoclonal antibody (a-Vc 01 LPS mAb) (0.04-1.3 Cpg/mL) and B) rabbit
anti-K cholerae 01 LPS polyclonal antibody (a-Vc 01 LPS pAb) (1:100-1:1,600)
were reacted with K. cholerae 569B LPS (1 Cpg/mL)-coated microtiter wells. Bound
primary antibodies were detected with either HRP-conjugated goat anti-rabbit
antibody or HRP-conjugated goat anti-mouse antibody. Wells were developed with
TIVB substrate and their absorbances read at 630 nm. The S:N were calculated for
every primary antibody concentration or dilution (n=3 wells). In (A) comparisons of
S:N between every monoclonal antibody concentration used yielded significant
differences (p<0.03) in S:N; therefore, the saturation limit of the monoclonal antibody
could not be determined. In (B) a significant decrease (*p=0.01) in the S:N was not
reached until the polyclonal antibody dilution decreased from 1:400 (ratio dilution of
0.003) to 1:800(ratio dilution of 0.001); therefore, the saturation limit of the
polyclonal antibody was reached at a dilution of 1:400.










B 12


10


Ratio of Dilution, Dilution
z I/I 10.01, 1:100
v, /I 10.005, 1:200
0.003, 1:400
4 ~0.001, l:soo
2


0 0.002 0.004 0.006 0.008 0.01

Ratio of Dilution


Figure 3-4. Continued.








I
E
II,
CR

Cf)
PL
I
Q)

O
~
O
3j


32


10


Figure 3-5. Saturation of y. cholerae 569B LPS to nitrocellulose paper. Vibrio cholerae 569B
LPS was incubated with nitrocellulose paper at concentrations from 1-100 Cpg/mL.
The primary antibody used to detect the LPS was rabbit anti-Y. cholerae 01 LPS
polyclonal antibody (1:400). Bound primary antibody was detected with HRP-
conjugated goat anti-rabbit antiserum and detected with ECL substrate. The
nitrocellulose paper became saturated with y. cholerae 569B LPS at a LPS
concentration of 100 Cpg/mL. The strips incubated with 0 Cpg/mL LPS failed to stain
significantly .


II



II




lil


100


CII




I!111~


3 1









100
90
E 80
3 70
S60 -- a lo cells ve
O ~109 cells Vc
50 aC18e16cl~s
4-5 109 cells Vc, 1% Vc86
c, 40- ocelVc1%V8

m, 20
10


15 30 45 60

Trypsin Elution (min)

Figure 3-6. Titers of eluted phagemid particles from the first one-round panning optimization
experiment. The Tomlinson J library +/- 1% Vc86 phagemid particles was panned
against 10s and 109 whole cells of y. cholerae N16961 in suspension. Phagemid
particles that were bound to the y. cholerae whole cells were eluted with 10% (v/v)
trypsin stock (10 mg/mL trypsin, 50 mM Tris-HCI (pH 7.4), 1 mM CaCl2 in water) in
PBS for 15, 30, 45, and 60 minutes. Eluted phagemid particles were enumerated via
the spot-titer method.














-e- 107 cells Vc
S108 cells Vc
-A 109 CGIIS Vc


15 20 25 30


Trypsin Elution (min)


120

100

80

60

40

20

0


107 cells Vc
108 cells Vc
109 CGIIS Vc


g_

Ip


Trypsin Elution (min)


Figure 3-7. Titers of eluted phagemid particles from the second one-round panning optimization
experiment. The Tomlinson J library +/- 1% Vc86 phagemid particles was panned
against 10 10s, and 109 whole cells of y. cholerae N16961 in suspension. Phagemid
particles that were bound to the y. cholerae whole cells were eluted in 10% (v/v)
trypsin stock (10 mg/mL trypsin, 50 mM Tris-HCI (pH 7.4), 1 mM CaCl2 in water) in
PBS for 10, 20, and 30 minutes. Eluted phagemid particles were enumerated via the
spot-titer method. A) Panning with the Tomlinson J library B) Panning with the
Tomlinson J library (+) 1% Vc86 phagemid particles.

























)10 20 30 41


O


108 cells Vc
109 cells Vc
10" cells Vc, 1% Vc86
10" cells Vc, 1% Vc86


Trypsin Elution (min)

Figure 3-8. Titers of eluted phagemid particles from the third one-round panning optimization
experiment. The Tomlinson J library +/- 1% Vc86 phagemid particles was panned
against 10s and 109 whole cells of y. cholerae N16961 in suspension. Phagemid
particles that were bound to the y. cholerae whole cells were eluted in 10% (v/v)
trypsin stock (10 mg/mL trypsin, 50 mM Tris-HCI (pH 7.4), 1 mM CaCl2 in water) in
PBS for 10 and 30 minutes. Eluted phagemid particles were enumerated via the spot-
titer method.




























108 108 108 108 109 109 109 109 a-Vc Vc86
10 10 30 30 10 10 30 30 01
min min min min min min min min LPS
+Vc86 +Vc86 +Vc86 +Vc86 mAb


Figure 3-9. Analysis of eluted phagemid particles from the third one-round panning optimization
experiment by ELISA. Microtiter wells coated with 3 x 10s CFU/mL K. cholerae
whole cells or PB S and were incubated with 1 x 10s tu/mL of amplified eluted
phagemid particles or 10 Cpg/mL of mouse anti-Y. cholerae 01 LPS monoclonal
antibody. Bound phagemid particles were detected with HRP-conjugated anti-M13
secondary antibody, and the monoclonal antibody was detected with HRP-conjugated
goat anti-mouse antibody. Wells were developed with TMB substrate, and their
absorbances were read at 630 nm (n=3 wells). There was a significantly higher S:N
(*p=0.01-0.04) from the amplified eluted phagemid particles from the panning with
10s whole cells with a 30 minute elution using the library containing 1% Vc86 when
compared to the amplified eluted phagemid particles from the other pannings.










14

12

10


to 6





1:2 1:4 1:6 1:10

Dilution of a(-BSA scFv Supernatant in Blocking Buffer


Figure 3-10. Analysis of anti-BSA (a-BSA) scFv proteins produced in microtiter wells by
ELISA. Microtiter wells were coated with 10Cpg/mL BSA or PBS. Anti-BSA scFv
particles were produced in a microtiter well and isolated by centrifugation and the
scFv-containing supernatants were diluted 1:2, 1:4, 1:6, or 1:8 in blocking buffer and
incubated with the coated wells. Bound scFv particles were detected with HRP-
conjugated Protein L. Wells were developed with TMB substrate, and their
absorbances were read at 630 nm. (n=2 wells).

























V V V V V B B B B B

30 30 37 37 30 30 30 37 37 30
0 0.1 0 0.1 0.1 0 0.1 0 0.1 0.1
W W W W F W W W W F


Phagemid
Particle

Temp. (oC)
% Glucose
Grown in


Figure 3-11. Analysis of a-BSA and Vc86 phagemid particles produced in microtiter wells by
ELISA. Phagemid particles Vc86 (V) and a-BSA (B) were produced at either 300C
or 370C in medium containing either 0% or 0. 1% glucose in a microtiter well (W) or
in a flask (F). Microtiter wells were coated with 10Cpg/mL BSA, 3 x 10s CFU/mL
V. cholerae whole cells, or PBS. Phagemid particles were diluted 1:2 in blocking
buffer and incubated with the coated wells. Bound phagemid particles were detected
with HRP-conjugated anti-M13 secondary antibody. Wells were developed with
TMB substrate, and their absorbances were read at 630 nm. (*) represents p<0.05 and
(**) represents p<0.01. (n=2 wells)










0.8


0.7

0.6


0.3


0.2

0.1

0.0
30oC, 0% 30oC, 0.1% 37oC, 0% 37oC, 0.1% 30oC, 0.1%
glucose glucose glucose glucose glucose, in
flask

ar-BSA Phagemid Particles


Figure 3-12. Analysis of a-B SA phagemid particles produced in microtiter wells by ELISA.
Microtiter wells were coated with 10Cpg/mL BSA. Phagemid particles were diluted
1:2 in blocking buffer and incubated with the coated wells. Bound phagemid
particles were detected with HRP-conjugated anti-M13 secondary antibody. Wells
were developed with TMB substrate, and their absorbances were read at 630 nm (n=2
wells). Phagemid particles produced in a microtiter well at 300C produced
significantly higher (*p=0.03) S:N values than phagemid particles produced in a
microtiter well at 370C at the same glucose concentration.














15


S10







Superinfected with (Hye-s m) (Hye-lg)
Hye
a-BSA Phagemid Particles


Figure 3-13. Analysis of anti-BSA (a-BSA) phagemid particles produced by hyperphage by
ELISA. Microtiter wells were coated with 10 Cpg/mL B SA or PB S and incubated with
1 x 10s tu/mL a-BSA phagemid particles produced from E. coli TG1 strains already
infected with hyperphage (Hye-sm, Hye-lg) or from E. coli TG1 that was
superinfected with hyperphage. Bound phagemid particles were detected with HRP-
conjugated anti-M13 secondary antibody. Wells were developed with TMB
substrate, and their absorbances were read at 630 nm (n=3 wells). The S:N from a-
BSA phagemid particles produced from (Hye-sm) was significantly lower (*p<0.01)
than the S:N from a-BSA phagemid particles produced from (Hye-lg) or from
superinfection with hyperphage.





I I I I I I 1 111111 1 I I I 1 111111 1 I I I I 11111 1 I I I I I II,,


20

15

10


Figure 3-14. Analysis of phagemid particles (1-22, 43-64) selected from panning against Stx2
toxin preparation by ELISA. Microtiter wells were coated with 10 Cpg/mL of Stx2
toxin preparation (Toxin Technologies) and incubated with clones 1 through 22 and
43 through 64 diluted 1:2 in blocking buffer, 1 Cpg/mL anti-Stx2 subunit A
monoclonal antibody (11E10), 1 Cpg/mL anti-Stx2 subunit B monoclonal antibody
(BBl12), or anti-BSA phagemid particles diluted 1:2 in blocking buffer. Microtiter
wells coated with 10 Cpg/mL BSA were incubated with anti-BSA phagemid particles
that were diluted 1:2 in blocking buffer. Bound phagemid particles were detected
with HRP-conjugated anti-M13 monoclonal antibody. Bound monoclonal antibodies
were detected with HRP-conjugated goat affinity purified anti-mouse antibody.
Wells were developed with TMB substrate, and their absorbances were read at 630
nm (n=2 wells). The noise used to calculate S:N was the signal of anti-BSA
phagemid particles that were incubated with Stx2-coated wells.


















10









Figure 3-15. Analysis of phagemid particles (1, 3-5, 23-42, 65-84) selected from panning against
Stx2 toxin preparation by ELISA. Microtiter wells were coated with 10 Cpg/mL of
Stx2 toxin or carbonate coating buffer and incubated with clones 1, 3 through 5, 23
through 42, and 65 through 84 diluted 1:2 in blocking buffer. Bound phagemid
particles were detected with HRP-conjugated anti-M13 monoclonal antibody. Wells
were developed with TMB substrate, and their absorbances were read at 630 nm (n=2
wells).





















S:N




Stx2
0157 PP


mAb -
mAb DH5a PP
11E10 clone clone(NR00 P
BB12 clone
4648 a-BSA
49


Figure 3-16. Analysis of anti-Stx2 monoclonal antibodies and clones 46, 48, and 49 by ELISA.
Microtiter wells were coated with 10 Cpg/mL of Stx2 toxin, 70 Cpg/mL ofE. coli
0157:H7 933 periplasmic break fraction, 2 mg/mL of E. coli 0157:H7 933
supernatant (0157 Sup.), Img/mL of E coli DH~a (pNR100) periplasmic break
fraction, and Img/mL of E coli DH~a periplasmic break fraction. (PP) represents
periplasmic break fractions. Bound phagemid particles were detected with HRP-
conjugated anti-M13 monoclonal antibody. Bound monoclonal antibodies were
detected with HRP-conjugated goat affinity purified anti-mouse antibody. Wells
were developed with TMB substrate, and their absorbances were read at 630 nm (n=2
wells).










46 48

T +-T+ -MVW


15
10


49 80 81 83
MW T+- T +- T+- T+-




250
250
15
10 ),


Figure 3-17. Western blot analysis of phagemid clones from panning on Stx2 toxin preparation.
Samples were resolved by SDS-PAGE on a 4-20% (w/v) polyacrylamide gel and
analyzed by Western blot. Samples: (T) -1.5Cpg/well of Stx2 toxin, (+) 120 Cpg/well
E. coli 0157:H7 933 (contains Stxl and 2 toxins) supernatant, and (-) -120 Cpg/well
E. coli 0157:H7 87-23 (does not produce Stx toxin) supernatant. Primary antibodies:
Anti-Stx2 toxin subunit A monoclonal antibody (Mab); Stx2 phagemid clones 46, 48,
80, 81, 83. A) Western blot was developed with TMB and exposed for 30 seconds.
B) Western blot was developed with TMB and exposed for 5 minutes.


Mab a-BSA

MnWT+- T+-





r *



1







MWV Stx2


-32: A subunit



-10: B subun it


Figure 3-18. Analysis of Stx2 toxin preparation (Toxin Technologies) by SDS-PAGE. Five
micrograms of Stx2 toxin preparation was resolved on a 4-20% (w/v) polyacrylamide
gel and stained with Coomassie blue stain. Stx2 toxin is comprised of a 32-kDa A
subunit and five 10-kDa B subunits.


250
150
1100
75

50
37


25 "-e-
20 "''"
15 ~
10 .-em(









CHAPTER 4
DISCUSSION

Detection of biological agents and their products has become a prime importance in the

United States (85) since the anthrax attacks of 2001 (86). Early detection of biological agents is

important to prevent or inhibit the spread of disease. Detecting biological agents in water, air,

and food enables the proper course of action to be implemented to lessen or prevent the effects of

the organism. Detection of biological agents or their products is also important for diagnosis of a

disease. Detection of the disease-causing agent in patient samples enables the proper treatment

to given.

There are vast numbers of detection assays, and a common component to most of them is a

protein that binds specifically to a target antigen. Monoclonal antibodies are the default tool that

scientists use today to detect target molecules; however, monoclonal antibodies have some

drawbacks that are causing the use of phage display reagents to become more prevalent and take

the place of monoclonal antibodies (2). First, the generation of monoclonal antibodies requires

killing animals and takes months. Also, monoclonal antibodies are rarely isolated to pathogenic

or self antigens, causing limitations in the scope of the target antigen for the generation of

monoclonal antibodies. Phage display is advantageous because it lacks many of the drawbacks

of monoclonal antibodies. In phage display, if using a non-immunized library, the killing of

animals is not required and the isolation of specific recombinant phages can be done in a couple

weeks instead of months. Also, the scope of antigen used for phage display is far greater than

that used for monoclonal antibodies because phage display enables the selection of recombinant

proteins to pathogenic and self antigens. Phage display was developed in 1985 by George P.

Smith as a means to display foreign proteins on a filamentous phage (52). The sequence for the

foreign peptide is encoded in frame with a coat protein gene of the phage genome. This enables









the foreign peptide to be fused to the coat protein, which effectively links the phenotype and

genotype of the phage. Therefore, every phage that displays a foreign protein also encodes its

peptide sequence in the genome.

This thesis describes efforts to optimize phage display methods to increase the chances of

isolating recombinant antibodies or peptides for the potential use in detection assays.

Specific Aim 1: Panning to K cholerae LPS

Vibrio cholerae is a human pathogen that is the causative agent of cholera. Cholera is a

diarrheal disease that affects millions of people worldwide annually (9). Detecting K. cholerae

in the environment and patient samples is the first step to preventing and treating the disease. To

detect Y. cholerae we need to isolate a recombinant phage that binds specifically and sensitively

to Y. cholerae. There are over 200 serogroups of y. cholerae that are classified on the basis of

the O-antigen of their LPS; however, only serogroups 01 and 0139 cause endemic acute

diarrhea. Therefore, the LPS of these two serogroups would be excellent targets for detection.

Vibrio cholerae 01 LPS was chosen as a target for biopanning. Because y. cholerae 01

LPS was no longer sold commercially, it had to be extracted in the laboratory before any panning

experiments could be performed. The previous method to extract LPS in the laboratory was a

phenol-water extraction (76). However, a new LPS extraction was developed using a

commercial RNA isolation reagent, TRlzol Reagent, that could extract LPS in a fraction of the

time of the phenol-water extraction (77). The TRlzol method was even claimed to be able to

extract LPS with less degradation and contamination than the phenol-water method. To

determine which method was better, both methods were used to extract y. cholerae 01 LPS and

the extracted LPS was analyzed and compared to commercially acquired y. cholerae 01 LPS

(Sigma-Aldrich) that had been phenol extracted and purified by gel-filtration chromatography.









The criteria for the optimal LPS extraction were to have little or no protein and lipid

contamination and to have similar banding patterns to the commercial K. cholerae 01 LPS.

Vibrio cholerae 01 LPS was extracted using the phenol-water and TRIzol methods. The

LPS was resolved by SDS-PAGE and analyzed by Coomassie blue staining and Tsai-Frasch

silver staining (Figs. 3-1 and 3-2). Vibrio cholerae 01 LPS that was extracted by the TRIzol

reagent had less protein contamination than did the LPS that was extracted by the phenol-water

method. The relative purity of the TRIzol-extracted LPS was less than 0.3 Cpg of protein/yg of

LPS, while the relative purity of the phenol-water-extracted LPS was approximately 1 Cpg of

protein/yg of LPS. However, the protein contamination in the phenol-water LPS extraction was

most likely due to the RNase that was used in the extraction. The Coomassie blue stain for the

phenol-water-extracted LPS showed a protein band at approximately 15 kDa. The molecular

weight of RNase is 13.7 kDa. The phenol-water LPS extraction contained 0.04 mg/mL of

RNase, which correlates to the band intensity of the 15-kDa protein band. Because the phenol-

water LPS extraction only contained approximately 0.02 mg/mL of DNase, MW of 31IkDa, the

DNase might not have been concentrated enough to be seen by Coomassie blue stain. Therefore,

both LPS extractions probably had less than 0.3 Cpg of bacterial protein contamination for every

Cpg of LPS extracted. The TRlzol extracted LPS showed non-staining bands between 20 to 30

kDa on the silver stain (Fig. 3-1B). The Tsai-Frasch silver stain primarily stains carbohydrates

and poorly stains proteins and lipids (78). Because the Coomassie blue stain for the TRlzol-

extracted LPS showed no detectable proteins and due to the thickness of the non-staining bands,

which correlates to concentration, the non-staining bands are most likely due to lipid

contamination. Previously published silver stains of y. cholerae 01 LPS had bands at ~10 and

14 kDa which were stated to be the lipid A-core of LPS, and ~20-50 kDa which were stated to be









the lipid A-core plus repeating O-antigens (80,81). The non-staining bands were not the lipid A

because the lipid A contains two glucosamines that are stainable by silver stain. Therefore, the

TRlzol-extraction appears to have also extracted cellular lipids. Therefore, the TRIzol-extracted

LPS appeared to have more lipid contamination than the phenol-water-extracted LPS.

When the resolved LPS extractions were compared by silver stain to the commercial LPS,

the phenol-water-extracted LPS most closely resembled the commercial LPS. In the silver stain

for TRIzol-extracted LPS and commercial LPS, the TRlzol-extracted LPS had bands at 14 and

20 kDa with non-staining bands between 20 and 30 kDa, the commercial LPS had bands at 14,

20, and 23 kDa. In the silver stain for phenol-water-extracted LPS and commercial LPS, both

LP S prep arati on s had b and s at 1 3, 2 0, and 2 3 kD a. The phenol -water-extracte d LP S had

additional banding from 23-50 kDa, while the commercial LPS had more of a smear from 23-50

kDa. Previously published silver stains of y. cholerae 01 LPS have O-antigen banding/smears

from 20-50 kDa (80,81). Vibrio cholerae 01 LPS has 12 to 18 O-antigen groups (83); therefore,

the additional banding in the phenol-water-extracted LPS may be due to the ability of the phenol-

water method to extract LPS with more O-antigens than that of the method used for the

extraction of the commercial LPS which was phenol extracted and purified by gel-filtration

chromatography. Because the phenol-water-extracted LPS most closely resembled the

commercial LPS and appeared to have less lipid contamination than the TRlzol-extracted LPS,

the phenol -water-extracte d LP S was u se d for later exp erim ents.

Before y. cholerae 01 LPS was panned, a solid support capable of binding a high quantity

of yK cholerae 01 LPS was analyzed. In the past the Ph.D. 12mer phage library was panned

against antigens that were immobilized onto a microtiter well. Previous experiments from this

laboratory suggested that nitrocellulose paper could bind more LPS than a microtiter well. It was









hypothesized that having more LPS present to bind phages would enable more phages that

specifically bound to the LPS to be selected. However, increasing the amount of LPS present

may also increase the amount of nonspecific phages selected as well as the amount of specific

phages selected. Therefore, more antigen present to pan against may not necessarily improve the

chances of selecting phages specific to the antigen.

To determine which solid support could bind more LPS, the saturation limit of LPS was

analyzed on a microtiter well and a piece of nitrocellulose paper with the same surface area as a

microtiter well. The microtiter well became saturated with LPS when coated with 0.1-1 Cpg LPS,

and the nitrocellulose paper became saturated with LPS when coated with ~250 Cpg LPS.

Therefore, nitrocellulose paper could bind approximately 250-2,500 times more LPS than a

microtiter well of equal area. Nitrocellulose paper has previously been used as a solid support

for proteins in panning procedures involving glutathione S-transferase and hen egg white

lysozyme (87). Nitrocellulose paper was chosen as a binding support because it had a high

capacity to bind proteins (~100 Cpg/cm2)

To determine if panning against more antigen increases the selection of phages specific for

the antigen, both microtiter wells and nitrocellulose paper were used as solid supports for

panning against y. cholerae 01 LPS. Vibrio cholerae 01 LPS (Sigma-Aldrich) was

immobilized onto a polystyrene Maxisorp microtiter well and panned with the NEB Ph.D. 12mer

phage display library. Five rounds of panning were performed with amplification of phages after

the first four rounds of panning. Panning promotes the selection of phages that specifically bind

to the target antigen over phages that are not specific to the antigen. Amplifying pools of eluted

phages between pannings enriches the amount of specific phages. Therefore, each additional

round of panning should increase the amount of specific phages in the pool of eluted phages.









One hundred phages from the fifth round of panning were individually amplified and analyzed in

an ELISA against y. cholerae 01 LPS, carbonate coating buffer, and PBS.

Anti-Vc 01 LPS mAb generated a S:N of 15.5, proving that the LPS successfully bound to

the microtiter plate. Anti-B SA phagemid particles were used in the ELISA against BSA

(10Cpg/mL) and generated a S:N of 22, proving that the ELISA successfully detected phages. All

100 clones gave S:N <1.4, whereas a positive signal was classified as a S:N >2. Therefore, all of

the clones screened by ELISA were negative for binding to y. cholerae 569B LPS.

Because panning the Ph.D. 12mer library against y. cholerae 01 LPS that was

immobilized onto a microtiter well failed to yield phages that were positive to y. cholerae 01

LPS, the panning procedure was altered in an attempt to improve the chances of selecting phages

that specifically bind to the antigen. The first alteration was to change the antigen binding

support from a microtiter well to nitrocellulose. It was determined that nitrocellulose paper

could bind more LPS than a microtiter well; therefore, with more LPS present to pan against

more specific phages may be selected. Also, the mechanism of phage elution was analyzed.

Previous panning procedures with the Ph.D. library used a glycine (pH 2.2) elution. There are

numerous elution methods, pH, salt, pressure, temperature, antigen, none of which are optimal

for every antigen-antibody complex. The key is to find a method that effectively dissociates the

antigen-antibody complex without causing harm to either the antibody or the antigen (88). An

acidic elution interferes with the electrostatic and hydrophobic interactions of the antigen-

antibody complex and causes them to dissociate. Acidic elutions do not work for every antigen-

antibody complex; sometimes the low pH of the elutant can denature the antibodies or antigens.

Elution with antigen elutes the antibody from the immobilized antigen-antibody complex by

competing with the immobilized antigen for binding with the antibody. While pH elutions are









the most common elutions used in phage display, elutions by antigen are not uncommon and

have even been noted to be more successful than pH elutions because they decrease the elution

of phages that are not specific to the antigen (89-91).

Panning the Ph.D. 12mer library against phenol-water-extracted y. cholerae 01 LPS that

was immobilized onto nitrocellulose paper and eluted with either glycine (pH 2.2) or y. cholerae

01 LPS was performed. Three rounds of panning were performed with amplification of phages

after rounds one and two. A subtractive panning was performed after round one. This

"negative" panning was performed to decrease the amount of nonspecific phages in the library.

The negative panning involved panning the library against blocked nitrocellulose paper;

therefore, clones specific to the blocker, nitrocellulose paper, or polystyrene should be removed

from the library. Two hundred clones from each panning were individually amplified and

analyzed by ELISA. The 400 phage clones and a-Vc 01 LPS mAb were screened in an ELISA

against Y. cholerae 569B LPS. Anti-Vc 01 LPS mAb generated a S:N of 10, proving that the

LPS successfully bound to the microtiter plate. Anti-BSA phagemid particles were used in the

ELISA against BSA (10Cpg/mL) and generated a S:N of 15, proving that the ELISA was a

successful assay for the detection of phages. All 400 clones gave a S:N less than 1.7; therefore,

all of the clones screened were negative for binding to y. cholerae 01 LPS.

Previous attempts to select recombinant phages to LPS have failed to generate highly

specific recombinant phages that generally lack a consensus sequence (92-94). Also, the

isolation of recombinant phages that bind to LPS is less common than for the isolation of

recombinant phages that bind to protein. This is most likely because protein-protein interactions

are usually stronger than protein-carbohydrate interactions. The antibody-antigen complex is

held together by noncovalent interactions such as hydrogen bonds, van der Waals forces,









coulombic interactions, and hydrophobic bonds (95). The strength of the antibody-antigen

complex depends on the strength of the bonds used to hold it together. Protein-protein

interactions make use of all noncovalent bonds, but carbohydrate-protein interactions primarily

make use of hydrogen bonds and rarely make use of hydrophobic or coulombic interactions.

Also, protein-protein interactions can be held together by numerous bonds due to the

conformational structure of the protein-protein interaction. The carbohydrate-protein interaction

is usually only bound by a single domain due to the linearity of the carbohydrate structure.

Therefore, the low density and the limited variety of bonds that occur in carbohydrate-protein

interactions make LPS a difficult target to select recombinant proteins that bind to it with high

affinity.

Specific Aim 2: Improve Panning and Screening Process of Biopanning

Biopanning is performed to select phagemid particles that specifically bind to a target. To

increase the chances of selecting phagemid particles that are specific to a target, the optimization

of the biopanning and screening process was performed. Optimization of the panning process

was performed with y. cholerae 01 whole cells as a target, and the concentration of whole cells

and the trypsin elution time were analyzed. The concentration of whole cells used in biopanning

usually ranges from 10' to 109 whole cells (96). The goal was to use enough whole cells to

enable capture of specific recombinant phagemid particles and keep the capture of phagemid

particles not specific to the whole cells to a minimum. The trypsin elution time was also

analyzed. The Tomlinson libraries have a trypsin cleavage site between the pIII protein and the

scFv particle. Therefore, trypsin is used to cleave the phagemid particle from the scFv peptide

that is bound to the target. The Tomlinson protocol recommends a trypsin elution of 10 minutes.

The trypsin elution time was analyzed to maximize the number of eluted phagemid particles that

were specific to the target. Eluting for too short of a time may fail to elute specific phagemid









particles, but eluting for too long of a time may only increase the elution of nonspecific

phagemid particles or even degrade the phagemid particles.

Comparative pannings were performed to determine the optimal concentration of

yK cholerae whole cells and optimal trypsin elution time. Parallel pannings were performed with

the Tomlinson J scFv library with and without 1% Vc86 phagemid particles spiked into the

library. Vc86 is a clone isolated from the Tomlinson I scFv library that recognizes an unknown

antigen on Y. cholerae whole cells. Because ~1011 phagemids are being used in the pannings

there only exists a 100-fold redundancy of clones in the non-spiked library. Therefore,

enumeration of eluted phagemid particles after one round of panning may not yield significant

differences in eluted titers under various test parameters. Spiking the library with 1% phagemid

particles that are specific to y. cholerae whole cells should enable distinction of an optimal

panning parameter that selects and elutes the highest number of specific recombinant phagemid

particles. We were looking for a combination of the concentration of whole cells and trypsin

elution time to elute more phagemid particles from the Vc86-spiked library than the non-spiked

library. Because the only difference between the pannings was that the Vc86-spiked library had

a higher initial concentration of specific phagemid particles to K. cholerae whole cells, an

increase in eluted phagemid particles with a certain parameter should correlate to an increase in

specific phagemid particles being eluted.

The first one-round panning used 1 x 10s or 1 x 109 whole cells of y. cholerae with trypsin

elutions of 15, 30, 45, and 60 minutes (Fig. 3-6). Elution of phagemid particles with trypsin was

optimal between 15 and 30 minutes. The phagemid particle titer decreased by 80 to 93% from

30 to 45 minutes for the four pannings. The decrease in the titer of phagemid particles was most

likely due to degradation of the phagemid particles by trypsin. Therefore, trypsin elution should









not exceed 30 minutes. More phagemid particles were eluted when panned against 1 x 109 whole

cells for both the spiked and nonspiked libraries. However, there were not enough data to do

statistical analysis to determine if using 1 x 109 whole cells eluted significantly more phagemid

particles than using 1 x 10s whole cells in the biopanning process.

The best whole cell concentration and trypsin elution time were not definitively obtained

from the first one-round panning; therefore, a second one-round panning was performed to

further analyze the trypsin elution time and whole cell concentration. The second one-round

panning used 107, 108, or 109 whole cells of y. cholerae with a trypsin elution of 10, 20, and 30

minutes (Fig. 3-7). There were no significant differences in the titers of eluted phagemid

particles when comparing the concentrations of whole cells and trypsin elution times used in the

second one-round panning experiment for both library pannings. This could mean that there may

not be a significant advantage in using the different elution times or concentrations of whole

cells for selecting specific recombinant phages.

Because using titering to analyze the differences in the amount of eluted phagemid

particles under various parameters failed to definitively determine a panning condition that

eluted the highest number of specific recombinant phagemid particles, another one-round

panning was performed with analysis of phagemid particle specificity by ELISA. Following

elution of the phagemid particles, the phagemid particles were amplified in batch and analyzed

by ELISA to determine which panning parameter eluted the most phagemid particles that were

specific to y. cholerae whole cells.

The third one-round panning used 10s or 109 whole cells of y. cholerae with trypsin

elutions of 10 and 30 minutes (Fig. 3-8). The eluted phagemid particles were amplified in batch

and analyzed by ELISA. Of the different panning variations, the amplified phagemid particles









from the panning with the library that contained 1% Vc86 phagemid particles, 10s whole cells

yK cholerae, and an elution time of 30 minutes generated a significantly higher S:N (p=0.01-

0.04) in an ELISA compared to the amplified phagemid particles from the other panning

variations. This suggests that panning with 10"s whole cells with a trypsin elution time of 30

minutes is optimal for selecting phagemid particles that are specific to the panned whole cells

compared to using 109 whole cells or a trypsin elution of 10 minutes. There might have been an

excess of whole cells when panning against 109 whole cells. This excess may have promoted the

binding of nonspecific phages; therefore, fewer whole cells present may have been more efficient

for panning because with fewer whole cells present the specific phagemid particles should

preferentially bind the whole cells and the number of nonspecific phagemid particles would be

lessened. Eluting for 10 minutes may not have been long enough for the trypsin to cleave all the

specific phagemid particles from the whole cells; therefore, eluting for 30 minutes was optimal

because it may have given the trypsin enough time to elute the phagemid particles from the

whole cells.

After the biopanning process is complete, the selected scFvs or phages are screened for

their specifieity to the target molecule. The previous method to screen scFv proteins was to

grow amber suppressor-free phagemid-containing bacteria under conditions to secrete scFv

proteins in culture tubes or flasks. However, this process was very time consuming; therefore, a

way to screen scFv proteins in a high throughput manor was investigated. High throughput

screening of scFv proteins is very common. A Biacore A 100 array system is capable of

screening hundreds of scFv proteins per day (97). This system screens scFv proteins from crude

bacterial extracts on a sensor chip surface. The antibodies that are captured onto the chip are

analyzed and ranked by the percentage of bound scFv proteins remaining on the chip after









dissociation of nonspecific scFv proteins in buffer. His-tagged scFvs can be screened in an

automated high throughput approach by measure of surface plasmon resonance using a Qiagen

BioRobot 3000 LS (98). A chromatography press is capable of purifying and screening

hundreds of scFv proteins per day (99). This system involves His-tagged scFvs being purified by

a cation exchange column and immobilized metal ion affinity chromatography. The proteins

were then quantified by a bicinchoninic acid (BCA) assay and analyzed by SDS-PAGE. While

there are many high throughput screens for scFv proteins, most of the methods involve expensive

equipment that is not readily available to our laboratory. Therefore, our goal was to develop a

high throughput screen for scFv proteins.

A colony blot was attempted to screen scFv proteins. The colony blot involved the transfer

of bacterial colonies that were grown to secrete scFv proteins onto an antigen-coated piece of

nitrocellulose paper. The scFv proteins that bound to the nitrocellulose via the antigen were then

detected with Protein-L peroxidase and identified with ECL. However, when this system was

attempted, the colony blot was determined to be neither specific nor selective. When the a-B SA

clone was screened by this method, it did not generate a positive signal on BSA-coated

nitrocellulose but did generate a positive signal on PBS and y. cholerae-coated nitrocellulose

paper. Also, when clones selected in a panning against y. cholerae whole cells were screened by

this method, they generated positive signals to BSA, PBS, and y. cholerae-coated nitrocellulose

papers. It appeared as if random clones bound to the nitrocellulose paper despite it being

blocked with casein buffer and that all clones that produced scFvs were detected. For this

problem to occur many steps or just one step could have gone wrong in the experiment. It is

possible that not enough antigen was used to coat the nitrocellulose, the nitrocellulose did not

capture the antigen, the nitrocellulose was poorly blocked, or the washes were not stringent










enough. The effort required to troubleshoot the colony blot was determined to not be worth the

time. Therefore, a way to improve the old method for producing scFvs was developed.

To enable high throughput production of scFv proteins, a clone was grown in a 96-well

plate instead of a culture tube or flask, and the scFv-containing supernatant was examined in an

ELISA. When a-BSA clone was grown in a 96-well plate, its scFv-containing supernatant

diluted 1:2 yielded a S:N of 11.5 against BSA in an ELISA. Anti-BSA scFv produced in a flask

yielded a S:N between 10 and 15; therefore, producing scFv proteins in a microtiter well may not

be optimal, but it is acceptable for the purpose of a high throughput screen. Next was to

determine how far the microtiter well-produced scFvs could be diluted and still generate a

sufficient signal to be detected in an ELISA. This test was to determine how many clones could

be mixed together (i.e., diluted) and screened in a single microtiter well to enable even greater

high throughput screening of scFvs. The a-BSA clone was grown in a microtiter well to produce

scFvs. The supernatants were diluted 1:2, 1:4, 1:6, and 1:10 in casein blocking buffer and

analyzed by ELISA (Fig. 3-10).

When a-BSA scFv-containing supernatant was diluted 1:10 it still yielded a S:N of 5.6 in

an ELISA. This S:N is high enough for detection of positive clones. However, the a-BSA clone

was a very strong clone; therefore, the maj ority of clones selected will have a weaker signal than

the a-BSA clone. To lessen the chance of not detecting scFvs with weak affinity or low titers, it

was decided to only dilute scFv clones 1:4 to 1:6 for high throughput screening by ELISA. This

would still enable screening of hundreds of clones per day, a vast improvement to the previous

method of screening clones in culture tubes or flasks which only enabled screening of a couple

dozen clones per day.









Because the high throughput screen for scFv proteins was successful, it was examined if

a similar high throughout ELISA could be applied to the screening of phagemid particles.

Previous methods of screening phages involved producing phagemid particles in a culture tube

or flask and screening the supernatants by ELISA. This method was time consuming; therefore,

a high throughput screen was investigated. Previous high throughput ELISAs have been

developed for the screening of phages; however, while their screening is rapid, the production of

the phages is time consuming. These high throughput ELISAs require overnight production of

phages in a culture tube or flask followed by PEG precipitation of the phages for the

concentrated phages to be screened by ELISA (100,101). We hoped to develop an even more

high throughput ELISA by not only screening phagemid particles in a microtiter well but also

growing the phagemid particles in a microtiter well.

Phagemid particles from the a-BSA and Vc86 clones were transduced into E. coli TG1

(Hye-sm), which harbors the hyperphage genome as a plasmid. Producing phagemid particles

in E. coli TG1 (Hye-sm) removes the additional step of superinfection with hyperphage;

therefore, reducing the time required to produce phagemid particles. Colonies from the

transduction were picked and grown in a microtiter well to produce phagemid particles. Growth

temperature and glucose concentration of the medium were analyzed. The Tomlinson protocol

recommends producing phagemid particles in medium that has 0.1% (w/v) glucose. However,

this concentration of glucose could promote catabolite repression of the lac promoter, which

promotes transcription of the gHI gene. Therefore, it was examined if phagemid particles would

be produced better in medium with or without 0. 1% (w/v) glucose.

The a-B SA and Vc86 phagemid particles were produced in a microtiter well at 300C or

370C with and without 0. 1% (w/v) glucose. The phage-containing supernatants were diluted 1:2









in blocking buffer and analyzed by ELISA (Fig. 3-11). There was not a significant difference in

the S:N (p=0.07-0.48) of phagemid particles produced in microtiter wells with and without 0. 1%

(w/v) glucose. Therefore, the 0.1% (w/v) glucose in the medium was probably not enough

glucose to cause catabolite repression of the lac promoter. Vc86 phagemid particles produced in

microtiter wells with or without glucose at 300C and 370C did not produce a S:N greater than 1,

whereas Vc86 phagemid particles produced in a flask (S:N ~4) yielded a significantly higher S:N

(p=0.05). Observation of the turbidities of the bacterial cultures that were grown in a microtiter

well revealed only slightly turbid cultures compared to bacterial cultures that were grown in a

flask. The growth conditions may not have been optimal for the bacteria to grow when they had

the added stress of producing phagemid particles. It was possible that the bacterial cultures that

were grown to produce phagemid particles in a microtiter well experienced less aeration than

bacterial cultures grown in a flask or culture tube. The decreased aeration could decrease

bacterial growth and reduce phagemid particle production. Escherichia coli is a facultative

anaerobe which means it makes ATP, which drives biosynthesis, by aerobic respiration or

fermentation. Aerobic respiration generates 36 ATP molecules from one molecule of glucose,

while fermentation only generates 2 ATP molecules from one molecule of glucose. Because

biosynthesis is fueled by ATP, aerobic respiration is more effective for bacterial growth.

When a-BSA phagemid particles were produced at 300C, the S:N was significantly higher

(p<0.04) than when a-BSA phagemid particles were produced at 370C at either glucose

concentration in a microtiter well. It appeared that phagemid particle production was more

permissive at 300C opposed to 370C. The a-B SA phagemid particles produced in a flask (S:N

~12) yielded a significantly higher (p=0.02) S:N compared to a-BSA phagemid particles

produced in microtiter wells (S:N~7) when analyzed at a 1:2 dilution in an ELISA. The reduced









S:N of phagemid particles used in an ELISA could have been due to less phagemid particles

present in the phagemid particle-containing supernatant that was produced in a microtiter well as

opposed to the phagemid particle-containing supernatant that was produced in a flask. This

could have been due to reduced aeration in the microtiter well causing decreased bacterial

growth and thus reduced quantities of phagemid particles produced in a microtiter well.

Phagemid particle production in microtiter wells was repeated a second time with a-BSA

phagemid particles (Fig. 3-12). Anti-BSA phagemid particles yielded significantly higher

signals (p<0.03) when produced at 300C as opposed to 370C at either glucose concentration.

These data support the previous data that 300C is a more permissive temperature than 370C for

phagemid particle production. There was not a significant difference (p=0.34) in the signals

produced by phagemid particles that were produced with or without glucose at 300C. These data

also support the previous data that 0. 1% (w/v) glucose in the medium does not provide an

advantage or hindrance to phagemid particle production. While the second experiment did not

yield a significant difference (p=0.09) in signals produced by a-B SA phagemid particles

produced in a flask opposed to in a microtiter well, the microtiter well-produced phagemid

particles yielded lower signals than did phagemid particles produced in a flask.

Because phagemid particles that were produced from the same clone resulted in various

ELISA signals when phage-containing supernatant were analyzed at a dilution of 1:2 it was

possible that the reason for the differences in ELISA signals from phage-containing supernatant

were due to differences in phagemid particle titers produced under the various conditions.

Therefore, the titers of a-B SA and Vc86 phagemid particles produced in a microtiter well and in

a flask were analyzed. Phagemid particles produced at 300C in a microtiter well yielded

approximately 107 tu/mL while phagemid particles produced at 370C in a microtiter well yielded









approximately 106 tu/mL. Producing phagemid particles at 300C in microtiter wells yielded

approximately 10-fold higher tu/mL titers than phagemid particles that were produced in

microtiter wells at 370C. Whether the growth medium contained 0.1% (w/v) glucose or not did

not affect the titers of phagemid particles. When a-BSA and Vc86 phagemid particles were

produced in flasks at 300C with 0.1% (w/v) glucose in the medium, the titers were

approximately 10s tol109 tu/mL. Producing phagemid particles in a microtiter well yielded higher

titers and signals in an ELISA when produced at 300C. Whether glucose was in the media or not

did not affect phagemid particle titers or ELISA signals. Phagemid particles could not be

produced in a microtiter well in high enough titers to make a high throughput screen acceptable.

When clone Vc86, a clone of average strength, was produced in a microtiter well, it was not able

to generate a positive signal. Therefore, the high throughput ELISA screen for phagemid

particles was not acceptable for screening of phagemid particles.

Specific Aim 3: Improve Phagemid Particle Production

Amplifieation of phagemid particles is an essential step in biopanning. A phagemid

contains the giII gene of M13 but lacks the rest of the M13 genes needed for phagemid particle

production. Therefore, a phagemid requires an amplification tool to provide the rest of the M13

genes in trans to enable phagemid particle production. It is essential that there is high quality

and high quantity phagemid particle production during amplification to ensure that the

redundancy of clones in the library is maintained and that the library is in high enough titers to

be useful in assays.

The most commonly used amplification tool for phagemids is helper phage. There are

many variations of helper phages, R408, VCSM13, and M13KO7 (63) that differ slightly. Of the

helper phages, M13KO7 is the most commonly used. Because helper phages contain a wild type

giII gene, they are not optimal for phagemid particle amplification because the produced









phagemid particles will express non-recombinant pIII proteins along with recombinant pIII

proteins. To overcome the drawback of helper phages encoding a wild type giII gene, many giII

gene mutated variations of helper phages were engineered. Many of these strains either have a

deletion of the giII gene or encode amber stop codons within the giII gene to lessen the

expression of wild type pIII proteins (65-67). Of the giII gene mutated helper phages,

hyperphage is the most effective for producing high titers of phagemid particles that have

multivalent display of pIll-fusion proteins (66). Hyperphage was the previous amplification tool

used by our laboratory to amplify phagemid particles; however, the titer of the hyperphage stock

was too low to effectively amplify every phagemid particle in the library. Therefore, a way to

improve phagemid particle production was investigated.

To overcome the problem of the low titer of the stock of hyperphage from effectively

infecting every phagemid-containing bacterium, the hyperphage genome was transduced into

E. coli to be maintained as a plasmid to ensure that every phagemid-containing bacterium also

contained hyperphage to enable amplification of phagemid particles. When hyperphage was

transduced into E. coli TGl, small (0.5 mm diameter) and large (1-1.5 mm diameter) colonies

resulted. When the small and large colonies were passage, the small colonies maintained a

small phenotype (0.5 mm diameter) and the large colonies maintained a large phenotype (1-1.5

mm diameter). The two strains were named E. coli TG1 (Hye-sm) and E. coli (Hye-lg). The

reason for the variation in the colony size was not known. Because colonies were grown under

kanamycin resistance for hyperphage, both strains contained hyperphage, or at least they

contained the kanamycin resistance gene that hyperphage carries. The differences in the colony

sizes suggested that the bacteria that formed the colonies were not replicating at the same rate.

The smaller colonies might be small because they are under stress from replicating the









hyperphage genome along with the E. coli genome, and the large colonies might be larger

because they are capable of growing faster because they are under less stress because they may

not be replicating everything. To determine if the two strains were functionally different, their

infection efficiencies were compared to E. coli TG1 and their ability to produce phagemid

particles was compared to E. coli TG1 containing a phagemid that was superinfected with

hyperphage.

The infection efficiencies of a-B SA and Vc86 into E. coli TGl, E. coli TG1 (Hye-sm),

and E. coli (Hye-lg) were analyzed. When the number of a-B SA phagemid particles transduced

into E. coli TG1 was set at an infection efficiency of 100%, the number of a-BSA phagemid

particles transduced into E. coli TG1 (Hye-sm) was 28% and the number of a-BSA phagemid

particles transduced into E. coli TG1 (Hye-lg) was 0. 1% of the number of transduced phagemid

particles into E. coli TGl. When the number of Vc86 phagemid particles transduced into E. coli

TG1 was set at an infection efficiency of 100%, the number of Vc86 phagemid particles

transduced into E. coli TG1 (Hye-sm) was 22% and the number of Vc86 phagemid particles

transduced into E. coli TG1 (Hye-lg) was 2% of the number of transduced phagemid particles

into E. coli TGl. Escherichia coli TG1 (Hye-sm) had an infection efficiency that was

approximately 25% of the infection efficiency of E. coli TGl, and E. coli TG1 (Hye-lg) had an

infection efficiency that was approximately 1% of the infection efficiency of E. coli TGl. It

appears as ifE. coli TG1 has a reduced ability to be transduced with phagemid particles when it

is harboring hyperphage. The differences in the infection efficiencies between E. coli TG1

(Hye-sm) and E. coli TG1 (Hye-lg) show that their functionalities are different. The reason for

E. coli TG1 (Hye-lg) generating larger colonies may have been because it shed a function

required in transduction and therefore was able to grow faster. For example, the production of









the F pilus could have been compromised by the stress from E. coli TG1 replicating the

hyperphage genome.

Both hyperphage-containing strains ofE. coli TG1 had infection efficiencies that were too

low for the strains to be used to effectively amplify clones in a library. While the hyperphage-

containing E. coli strains have infection efficiencies too low to be used to amplify the library

their infection efficiencies are still acceptable for amplifying single clones because the

transduction efficiency is not critical for the amplification of a single clone. Therefore, the

quality and quantity of phagemid particles produced by E. coli TG1 (Hye-sm/1g) were analyzed.

Anti-BSA phagemid particles were produced by E. coli TG1 (Hye-sm), E. coli TG1

(Hye-lg), and E. coli TG1 that was superinfected with hyperphage. Anti-BSA phagemid particle

yields were as follows: E. coli TG1 that was superinfected with hyperphage yielded 3 x 10

tu/mL, E. coli TG1 (Hye-sm) yielded 1 x 109 tu/mL, and E. coli TG1 (Hye-lg) yielded 3 x 10s

tu/mL. While the ability of hyperphage-containing E. coli TG1 to be transduced with phagemid

particles was reduced compared to the ability of E coli TGl, its ability to produce phagemid

particles was not reduced. Phagemid particles produced by hyperphage-containing bacteria

yielded titers high enough to be used as a phagemid particle amplification tool for single clones.

The qualities of a-B SA phagemid particles produced by the three E. coli strains were

analyzed by ELISA (Fig. 3-13). Anti-BSA phagemid particles produced a significantly higher

S:N (p=0.01) when produced from E. coli TG1 (Hye-sm) compared to E. coli TG1 (Hye-lg).

There was not a significant difference in the S:N (p=0.50) from a-B SA phagemid particles

produced from superinfected E. coli TG1 and E. coli TG1 (Hye-lg). Because the titer of

phagemid particles produced from E. coli TG1 (Hye-sm), ~1 x 109 tu/mL, was three times

higher than the titer of phagemid particles produced from E. coli TG1 (Hye-lg) or E. coli TG1









that was superinfected with hyperphage, ~3 x 10s tu/mL, a third less phage-containing

supernatant was used in the ELISA for phagemid particles that were produced from E. coli TG1

(Hye-sm). The decrease in the S:N from a-BSA phagemid particles produced from E. coli TG1

(Hye-sm) could have resulted if the particles/mL of the phagemid particles produced from all

three strains were the same. Because hyperphage-containing E. coli TG1 was not going to be

used to amplify the library, further analysis of it was stopped, so that the time used to analyze it

could be put to better use finding a more effective phagemid particle amplification tool.

Hyperphage-containing bacteria was not acceptable as an amplification tool for the library

because of their low infection efficiencies, but E. coli TG1 (Hye-sm) was acceptable for use as

an amplification tool for the amplification of single clones.

Helper plasmids are another amplification tool for the production of phagemid particles

(64). They are M13-based plasmids that are maintained in E. coli to provide phagemids with all

the genes necessary to produce phagemid particles. Helper plasmids were engineered in three

forms that differ in the length of their giII genes. The helper plasmids contain a full-length

(M13cp), a deleted (M13cp-dg3), or a truncated (M13cp-CT) giIl gene. These three helper

plasmids were transformed into E. coli and analyzed for their infection efficiency with phagemid

particles, their ability to produce phagemid particles, and the quality of the phagemid particles

produced.

When the helper plasmids were transformed into E. coli TGl, the resulting colony sizes

varied from a diameter of 0.5 to 2 mm. The reasons for various colony sizes are hypothesized

above. A small (0.5 mm) colony and a large (2 mm) colony from each transformation were

passage to determine if the colony size phenotype would be maintained. Escherichia coli TG1

(M13cp-sm) was the only strain that maintained a consistent phenotype of 0.5 mm colonies, the









rest resulted in colonies that had a diameter between 0.5 and 2 mm. A small and large of each

strain were further analyzed to determine if the different colony phenotypes also had different

functionality.

The six helper plasmid-containing E. coli TG1 strains were analyzed for their infection

effciencies with two phagemid particles in relation to the infection effciencies of E coli TG1

that did not contain any helper plasmids. Escherichia coli TG1 (M13cp-dg3-sm) and E. coli

TG1 (M13cp-CT-sm) were the only strains that yielded infection effciencies greater than 0.3%

of the infection effciency of E coli TG1 (Table 3-1). Escherichia coli TG1 (M13cp-dg3-sm)

had infection effciencies of 175% and 97%, while E. coli TG1 (M13cp-CT-sm) had infection

effciencies of 7% and 13% compared to the infection effciency of E coli TGl. The differences

in the infection effciencies between the small and large strains of helper plasmid-containing

E. coli revealed that they had different functionalities. Except for E. coli TG1 (M13cp-dg3-sm),

all of the other helper plasmid strains had reduced infection effciencies compared to E. coli

TGl. Therefore, the cost for the E. coli cells maintaining the helper plasmids resulted in a

decrease in their infection effciencies. Because E. coli TG1 (M13cp-dg3-sm) and E. coli TG1

(M13cp-CT-sm) had higher infection effciencies compared to their large phenotypes, it

suggested that the large strains had a defect in some component of the cell involved in

transduction.

Escherichia coli TG1 (M13cp-dg3-sm) had the highest infection effciencies of the helper

plasmid strains; therefore, it was further tested to determine if the high infection effciencies

could be maintained when tested with additional phagemid particles. Escherichia coli TG1

(M13cp-dg3-sm) had infection effciencies of 92%, 113%, and 120% compared to the infection

effciencies ofE. coli TG1 with Vc86, clonel8, and a-AV20N3 (Table 3-2). The infection









efficiency ofE coli TG1 (M13cp-dg3-sm) was approximately the same as E. coli TGl;

therefore, E. coli TG1 did not suffer any reduction in infection efficiency due to maintaining

M13 cp-dg3 -sm.

Escherichia coli TG1 (M13cp-CT-sm) was the only other helper plasmid strain besides

E. coli TG1 (M13cp-dg3-sm) that had an infection efficiency close to that of E coli TGl.

However, its infection efficiency was still too low for it to be used as an amplification tool for

the library. Therefore, M13cp-CT-sm was transformed into E. coli JM109 to try and increase its

infection efficiency. Escherichia coli DH~a F' was the bacterial strain that the engineers of the

helper plasmids used to work with the helper plasmids; therefore, we tested if an E. coli strain

more similar to E. coli DH~a F' would generate higher infection efficiencies if it contained

M13cp-CT-sm. Escherichia coli JM109 and E. coli DH~a F' contain relA1 and recA1 while

E. coli TG1 does not. However, infection efficiency experiments with E. coli JM109 (M13cp-

CT-sm) with three different phagemids yielded lower infection efficiencies than did E. coli TG1

(M13cp-CT-sm). The average infection efficiency of E coli TG1 (M13cp-CT-sm) was 47%,

while the average infection efficiency ofE coli JM109 (M13cp-CT-sm) was 28%. Therefore,

using E. coli JM109 to maintain M13cp-CT-sm did not increase its infection efficiency.

When the infection efficiencies of E coli TG1 (M13cp-dg3-sm) were analyzed, this strain

yielded an average infection efficiency of 120133% and E. coli TG1 (M13cp-CT-sm) had an

average infection efficiency of 30116%. Escherichia coli TG1 (M13cp-dg3-sm) had a

significantly higher infection efficiency (p=0.01) than E. coli TG1 (M13cp-CT-sm).

Escherichia coli TG1 (M13cp-dg3-sm), E. coli TG1 (M13cp-CT-sm), and E. coli TG1 that

was superinfected with hyperphage were analyzed for their abilities to produce phagemid

particles. Phagemid particles were produced with and without 1 mM IPTG in the growth









medium to analyze if IPTG had an effect on phagemid particle production yield. Because the lac

promoter drives transcription of the glll-scFv fusion it was hypothesized that the expressing

more pIll-scFv fusions would result in more phagemid particles produced. However, when

phagemid particles were titered, the phagemid particles produced in growth medium that

contained IPTG had the same or lower titers of phagemid particles than when phagemid particles

were produced without IPTG in the growth medium. Because there was not an increase in the

phagemid particle yields when they were produced with IPTG, it suggests that the amount of pIII

protein was not a limiting factor in phagemid particle production. For the titers that had a

decrease in phagemid particle yields when produced with IPTG the increase in pIII protein may

have been harmful to phage assembly because the excess of pIII protein may have been toxic.

Phagemid particles were amplified to titers of 106-10s tu/mL by E. coli TG1 (M13cp-dg3-sm),

105 -106 tu/mL by E. coli TG1 (M13cp-CT-sm), and 10s -109 tu/mL by E. coli TG1 superinfected

with commercial hyperphage. Titers of phagemid particles produced from E. coli TG1 that

contained helper plasmids were too low for the helper plasmids to be an effective amplification

tool. Phagemid particle production cultures usually contain ~109 CFU/mL; therefore, for the

helper plasmids to only generate 105 tu/mL means that only 1 in every 10,000 bacteria produced

an infectious phagemid particle. The use of helper plasmids as an amplification tool did not

generate high enough titers of phagemid particles for them to effectively amplify a library.

Even though the helper plasmids were determined to be an ineffective amplification tool,

they were further analyzed for the quality of phagemid particles produced by ELISA. Analysis

by ELISA revealed that phagemid particles produced by helper plasmids resulted in a S:N of 7

when less than 5 x 104 tu/mL of phagemid particles were used as a primary antibody. Data

obtained from our laboratory showed that phagemid particles are not detectable in an ELISA









unless they are used at concentrations of at least 1 x 106 particles/mL (data not shown). Because

the spot titer method measures the number of infectious phagemid particles, it is possible that the

total number of phagemid particles far exceeded the number of infectious phagemid particles.

To determine if the total number of phagemid particles is greater than the number of infectious

particles an anti-M13 sandwich ELISA was used to enumerate the particles/mL of phagemid

particle-containing supernatants (Table 3-4). Phagemid particles amplified by E. coli TG1

(M13cp-CT-sm) had particle/mL to tu/mL ratios between 740 and 23,000, E. coli TG1 (M13cp-

dg3-sm) had particle/mL to tu/mL ratios between 1 and 190, while phagemid particles amplified

by E. coli TG1 that was superinfected with hyperphage had particle/mL to tu/mL ratios between

6 and 24. Phagemid particles produced by helper plasmid-containing E. coli had particle/mL to

tu/mL ratios that were much higher than those for phagemid particles produced by E. coli that

was superinfected hyperphage. In biopanning the number of phagemid particles is not important,

the number of infectious phagemid particles is what matters because if a phagemid particle is not

infectious then it can not enter a bacterial cell and be produced to eventually be selected and

screened. Therefore, even though the number of particles produced by helper plasmids was high,

the number of infectious particles produced was low, making the helper plasmids an inadequate

amplification tool for the library.

The helper plasmids yielded phagemid particle progeny at titers between 105 tol0s tu/mL.

In a last attempt to get the helper plasmids to generate high titers of phagemid particles, the

M13 cp-dg3 -sm helper plasmid was incorporated into E. coli MGl65 5 to determine if a less

complex E. coli strain could generate higher titers of phagemid particles. Escherichia coli

MGl655 was chosen because it has few known mutations and does not contain the F-plasmid.

Four different phagemid particles were amplified with E. coli MGl655 (M13cp-dg3-sm), and the









resulting titers were: 2.5 x 10 6.3 x 106, 2.5 x 10 and 3.9 x 107 tu/mL. Producing phagemid

particles from E. coli MGl655 that contained a helper plasmid did not increase the titer of

phagemid particles produced. The helper plasmids were extensively analyzed and determined to

be poor amplification tools. Therefore, hyperphage was the best amplification tool, but it was

available in titers too low to be used effectively.

To overcome the low titers of commercial hyperphage, we made homemade hyperphage.

Homemade hyperphage was previously made by transducing F E. coli cells that harbored a

plasmid that encoded the giIl gene, pGTR203, with hyperphage and growing these transduced

cells under conditions that promoted phage production. In the past, hyperphage amplified from

E. coli TG1 (pGTR203) yielded titers between 107 to 10s tu/mL and E. coli HB2151 (pGTR203)

yielded titers between 106 to 107 tu/mL. Many liters of hyperphage had to be produced and

concentrated to generate hyperphage in a titer high enough to be used for a single panning

experiment. Therefore, a way to increase the titer of homemade hyperphage was investigated to

generate enough hyperphage to be used in multiple experiments. Escherichia coli TG1 and

E. coli HB2151 are both F+ strains. Because hyperphage infects bacteria by the F pilus, it was

hypothesized that some of the hyperphage produced were being taken back into the bacteria

which would decrease the titers of the hyperphage produced. Therefore, it was investigated if

hyperphage could be produced in higher titers when produced from a F- E. coli strain.

Escherichia coli MGl655 and E. coli EC 100D were the two F- strains of E. coli chosen to

amplify hyperphage. These strains were transformed with pGTR203 encoding the giII gene and

the hyperphage genome, and Hyve isolates from each strain were grown under conditions to

promote phage production. The particles/mL and the tu/mL were analyzed (Table 3-5). The

average titer of hyperphage produced from the five E. coli EC 100D (pGTR203) (Hy@) isolates









was 2. 1 x 10 12.5 x 10s tu/mL. The average titer of hyperphage produced from the Hyve E. coli

MGl655 (pGTR203) (Hye) isolates was 3.0 x 109 12.0 x 109 tu/mL. The titer of hyperphage

produced from the E. coli MGl655 (pGTR203) (Hye) isolates was significantly higher (p=0.02)

than the titer of hyperphage produced from the E. coli EC 100D (pGTR203) (Hye) isolates. The

higher titers produced from E. coli MGl655 might be because it has less known mutations than

E. coli EC 100D. The ratios of particles/mL to tu/mL of hyperphages produced from E. coli

EC 100D (pGTR203) (Hye) were between 0.4 and 7.8. The ratios of particles/mL to tu/mL of

hyperphages produced from E. coli MGl655 (pGTR203) (Hye) were between 0.3 and 1.0. The

particle/mL to tu/mL ratios of hyperphage produced from the F- strains were approximately the

same, which means that approximately all of the phagemid particles produced were infectious.

Escherichia coli MGl655 (pGTR203) (Hye) produced higher titers of hyperphage than did both

F' strains. Whether or not the higher titers of hyperphage were due to being produced in an F-

strain was not determined because E. coli EC 100D (pGTR203) (Hye) produced approximately

the same titers of hyperphage as E. coli TG1 (pGTR203) (Hye). What was determined was that

the bacterial strain used to produce hyperphage had an effect on hyperphage production.

Because E. coli MGl655 (pGTR203) (Hye) 5 produced the highest titers of hyperphage, it

was used to produce a large batch culture of hyperphage. The hyperphage was then concentrated

to approximately 1 x 1012 tu in 3 mL, which was concentrated enough to amplify the Tomlinson J

scFv phagemid library. The Tomlinson J scFv phagemid library was amplified with homemade

hyperphage with quality control checks by titering to ensure that the amplification of the library

generated high quality and high quantity of phagemid particles in a way that the redundancy of

the library was maintained.









Specific Aim 4: Isolation of Specific Recombinant Phage to Stx2 Toxin of E. coli 0157:H7

Once the Tomlinson J scFv library was amplified and homemade hyperphage was

improved to enable high quality and high quantity amplification of the library, a target molecule

had to be selected for panning. Escherichia coli 0157:H7 is an enterohemorrhagic serotype of

E. coli that is responsible for approximately 73,000 illnesses and more than 60 deaths per year in

the United States (30). Stx toxins kill mammalian cells and are the maj or causes of damage in

E. coli 0157:H7-infected individuals. Escherichia coli 0157:H7 can produce Stx toxins 1

and/or 2. Escherichia coli that produces Stx2 toxin appears to be more virulent than E. coli that

produces Stxl or both Stxl and 2 toxins (36-40). Therefore, Stx2 toxin would be a good target

for detection in animal and patient samples.

A Stx2 toxin preparation was obtained that was stated to be 50% pure for Stx2 toxin. The

other 50% of the preparation most likely contained bacterial impurities from E. coli HB 101

because the Stx2 toxin preparation was made from E. coli HB 101 (pMJ100-encodes Stx2 toxin).

A 100% pure Stx2 preparation would have been ideal; however, if the preparation contains 50%

Stx2 toxin then the most of the clones selected should theoretically be specific to Stx2 toxin.

The Tomlinson J scFv library that was amplified with homemade hyperphage was panned

against the Stx2 toxin preparation. Three rounds of panning were performed with amplification

of eluted phagemid particles by homemade hyperphage between each panning. Every panning

used 1 x 1012 phagemid particles. The number of eluted phagemid particles for the three

pannings were as follows: 2 x 106 for round one, 2 x 10' for round two, and 2 x 10'0 for round

three. The increase in the numbers of eluted phagemid particles suggested that phagemid

particles specific to Stx2 toxin preparation were being selected and amplified at each round.

Clones from the third round of panning were individually amplified and screened by ELISA.

Analysis of 59 clones resulted in 53/59 clones (89%) having a S:N from 2 to 23 with 48/59









clones (81%) having a S:N from 10 to 23 against the Stx2 toxin preparation. Therefore, the

panning was successful in isolating clones that bound to Stx2 toxin preparation. Not all of the

clones that were selected bound to Stx2 toxin preparation; 4/59 (7%) of the clones bound to PBS-

treated wells as well as Stx2-coated wells. The four clones had a signal in an ELISA between

0.31 and 0.87 against PBS-treated wells. These clones most likely bound to polystyrene. Even

though the immunotube was blocked with blocking buffer, the unintentional isolation of clones

that bound to the panning vessel is common (84). Of the clones screened, 2/59 (3%) did not bind

to Stx2 toxin preparation or a microtiter well. These clones have unknown binding specificity.

It is common to select clones that have unknown binding specificity. The fact that only

approximately 3% of my clones had unknown binding specificity proves that the panning was

successful in isolating specific recombinant phagemid particles.

Most of the clones selected from panning recognized the Stx2 toxin preparation; however,

the Stx2 toxin preparation was only 50% pure for Stx2 toxin. Therefore, it was unknown if the

clones bound to Stx2 toxin or bacterial impurities in the Stx2 toxin preparation. To determine if

the selected clones bound to Stx2 toxin, two anti-Stx2 toxin monoclonal antibodies were

examined. One antibody recognized the A-subunit of Stx2 toxin, while the other antibody

recognized the B-subunit of Stx2 toxin. However, these monoclonal antibodies never generated

a positive signal in an ELISA against the Stx2 toxin preparation. Using the monoclonal

antibodies at the manufacturer recommended concentration of 1 Cpg/mL failed to generate a

positive signal in an ELISA. It was initially reasoned that the monoclonal antibodies may need

to be used at a higher concentration to increase the S:N values. Increasing the concentration of

monoclonal antibodies in an ELISA to 100 Cpg/mL still failed to produce a positive signal in an

ELISA. Therefore, it was hypothesized that the HRP-conjugated anti-mouse secondary antibody









that recognized the anti-Stx2 toxin monoclonal antibodies was not recognizing the anti-Stx2

monoclonal antibodies. However, when another HRP-conjugated anti-mouse secondary

antibody was used to recognize the anti-Stx2 toxin monoclonal antibodies it also failed to

produce a positive signal in an ELISA. Both anti-mouse antibodies had been tested within three

days of these experiments against murine monoclonal antibodies and produced a positive signal.

However, positive controls for the anti-mouse antibodies were not used in the ELISA that

analyzed the two anti-mouse antibodies against Stx2 toxin preparation to determine whether or

not they were functional. It was not determined why the monoclonal antibodies did not

recognize the Stx2 toxin preparation in an ELISA. One reason might be that there was not

enough Stx2 toxin with a correctly folded binding site in the Stx2 toxin preparation for the

monoclonal antibodies to recognize. Also, it is possible that the batches of monoclonal

antibodies that we received were somehow defective.

Because the anti-Stx2 toxin monoclonal antibodies could not be used to determine the

binding specificity of the selected panning clones in an ELISA, other means of determining the

binding specificities of the clones were analyzed. Three clones that bound to Stx2 toxin

preparation were screened against Stx2 toxin, E. coli 0157:H7 933 (expresses Stx 1 and 2

toxins) periplasmic break fraction, E. coli 0157:H7 933 supernatant, E. coli DH~a (pNR100)

(expresses Stx2 toxoid) periplasmic break fraction, and E. coli DH~a periplasmic break fraction.

A clone that bound to the E. coli bacterial impurities in the Stx2 toxin preparation would be

expected to bind to all of the antigens, but a clone that bound to Stx2 toxin would be expected to

bind to all of the antigens except the E. coli DH~a periplasmic break fraction. However, the

three selected clones only recognized the Stx2 toxin preparation and none of the other antigens;

therefore, it was not determined if the clones were actually binding to Stx2 toxin.










Analysis of clones by ELISA proved that the maj ority of the selected clones from panning

bound to the Stx2 toxin preparation but failed to determine if the clones bound to Stx2 toxin

because the anti-Stx2 monoclonal antibodies and comparative analysis of +/- Stx2 bacterial

preparations failed to generate positive signals in an ELISA. However, when both of the anti-

Stx2 toxin monoclonal antibodies were examined in a Western blot against SDS-PAGE resolved

Stx-2 toxin preparation, the anti-Stx2 A subunit toxin monoclonal antibody successfully

recognized the Stx2 toxin A subunit in the Stx2 toxin preparation. Therefore, analysis by

Western blot could at least determine if any of the clones recognized the A subunit of Stx2 toxin.

Six clones were screened in a Westemn blot against SDS-PAGE resolved Stx2 toxin

preparation, E. coli 0157:H7 933 (expresses Stxl and 2 toxins) supernatant, and E. coli 0157:H7

87-23 (Stx phage cured) supernatant (Fig. 3-17). Because the Stx2 toxin is composed of A and B

subunits that are approximately 32 and 10-kDa proteins, a clone that recognized Stx2 toxin was

expected to recognize a protein band at 32 kDa or 10 kDa on the Stx2 toxin preparation and the

E. coli 0157:H7 933 supernatant but not on the E. coli 0157:H7 87-23 supernatant. If a clone

recognized bacterial impurities then it would be expected to recognize proteins in all three

antigens. The anti-Stx2 A subunit monoclonal antibody recognized a 32-kDa protein in the

commercial Stx2 toxin preparation and in the E. coli 0157:H7 933 supernatant proving that Stx2

A subunit was present in both preparations and not in E. coli 0157:H7 87-23 supernatant, as was

expected. Anti-BSA was also analyzed against the three antigens as a negative control to ensure

that phagemid particles did not generate artifact bands. Anti-BSA failed to recognize any of the

antigens, as was expected. Clone 46 failed to recognize any of the three antigens. It is possible

that the binding site of the target that it recognized in the Stx2 toxin preparation was denatured

by SDS-PAGE. Clone 48 recognized 35, 45, and 55-kDa proteins in the Stx2 toxin preparation









and a 35-kDa protein in both E. coli 0157:H7 supernatants, which means that it probably

recognizes bacterial protein impurities in the commercial Stx2 toxin. Clones 49 and 80 bound to

a 35-kDa protein in all three antigens and a 20-kDa protein in the Stx2 toxin preparation and in

E. coli 0157:H7 87-23. Therefore, clones 49 and 80 probably bound to bacterial protein

impurities in the commercial Stx2 toxin. Clones 81 and 83 appeared to recognize the same

bacterial protein impurities as clones 49 and 80, except that their band intensities were much

lighter.

Screening by ELISA showed that 89% of the clones bound to the Stx2 toxin preparation.

Western blot analysis of six clones revealed that none recognized Stx2 toxin and that five of the

clones most likely recognized bacterial impurities in the Stx2 toxin preparation. Because both of

the anti-Stx2 monoclonal antibodies did not recognize Stx2 toxin in the Stx2 toxin preparation in

an ELISA and the anti-Stx2 B subunit monoclonal antibody did not recognize Stx2 B subunit in

a Western blot, we assumed that there was not a high concentration of Stx2 toxin in the Stx2

toxin preparation. The Stx2 toxin preparation was stated to be 50% pure for Stx2 toxin by SDS-

PAGE analysis. To determine if this was true, the Stx2 toxin preparation was resolved by SDS-

PAGE and stained by Coomassie blue (Fig. 3-18). If the Stx2 toxin preparation was 50% Stx2

toxin, 32-kDa and 10-kDa protein bands should have been the most prominent bands. However,

SDS-PAGE analysis of the Stx2 toxin preparation showed that a 32-kDa protein band was a

minor protein band that represented less than 5% of the protein in the preparation. There was a

band at 10 kDa that had a strong band intensity. These bands might not have even represented

Stx2 toxin subunits but may have just been proteins with a similar molecular weight. What is

certain is the amount of Stx2 toxin in the Stx2 toxin preparation did not comprise 50% or even

10% of the proteins in the Stx2 toxin preparation. Therefore, it is probable that the majority of









the clones screened from the panning recognized bacterial impurities from E. coli and not Stx2

toxin. Extensive screening of additional clones may result in isolation of a clone specific to Stx2

toxin; however, it is recommended to try panning again with a better Stx2 toxin preparation. The

panning was successful in selecting phagemid particles that bound specifically to the Stx2 toxin

preparation, the phagemid particles selected just did not have the desired binding specificity. It

is recommended to use the newly amplified Tomlinson scFv library with the aid of the improved

homemade hyperphage to pan against another target with a high degree of purity.

This thesis details the optimization of phage display protocols and reagents. Analysis of

extraction methods for LPS were compared and it was determined that a phenol-water method,

opposed to a TRlzol method, extracted y. cholerae 01 LPS that most closely resembled

commercially acquired y. cholerae 01 LPS. Also, binding platforms were analyzed to

determine which platform could bind the most LPS. It was determined that nitrocellulose paper

could bind approximately 250 to 2,500 times more LPS than that a microtiter well. Optimization

of whole cell panning with the Tomlinson J scFv phagemid library revealed that a trypsin elution

between 10 and 30 minutes eluted the highest number of phagemid particles; this may be due to

phagemid particle degradation occurring with trypsin treatment after 45 minutes. Screening of

phage display reagents was analyzed and a high throughput ELISA was developed for the

production and screening of scFv proteins that was able to generate S:N values of produced scFv

proteins from an ELISA when diluted 1:10. A similar high throughput ELISA was developed for

the production and screening of phagemid particles; however, this method was not acceptable

due to low titers produced of phagemid particles in a microtiter well.

The optimization of phagemid particle production was performed. Multiple phagemid

particle amplification tools were analyzed for their infection efficiencies, yields of phagemid









particles produced, and the quality of phagemid particles produced. Hyperphage-containing

E. coli resulted in unacceptable infection efficiencies and helper plasmid-containing E. coli

resulted in unacceptable phagemid particle production titers. However, improvement of

hyperphage that was produced in our laboratory was performed by use of E coli MGl655

(pGTR203) which generated high titer and high quality hyperphage. This high quality

hyperphage was used to successfully amplify the Tomlinson J scFv phagemid library to ensure

that the redundancy of phagemid clones was maintained. This library was then panned against

an Stx2 toxin preparation and resulted in 89% of phagemid particles from the third round of

panning recognizing the Stx2 toxin preparation in an ELISA. The efficiency of this panning was

greatly improved from previous pannings from this laboratory, which usually resulted in no

better than 1% of selected phagemids being specific to the target antigen, and often no isolation

of usable specific clones. Therefore, the optimization of methods and reagents in this thesis will

greatly improve our laboratory and other laboratories chances of selecting recombinant phages

specific to a target molecule.









CHAPTER 5
EPILOGUE

The homemade hyperphage-amplified Tomlinson J library produced in these studies was

used to pan against E. coli 0157:H7 flagella and yielded over 80% positive phagemid clones.

The phagemids recognized the maj or flagellin protein by Western blot. The high throughput

scFv screen was used in the E. coli 0157:H7 flagellum proj ect and proved useful in identifying

rare scFv-secreting clones. Therefore, the improvements on phage display techniques and tools

described in this thesis are, in fact, useful and offer promise of success for continuing studies in

the laboratory.










LIST OF REFERENCES


1. Lim, D.V., Simpson, J.M., Kearns, E.A., and Kramer, M.F. 2005. Current and developing
technologies for monitoring agents of bioterrorism and biowarfare. Clin. M~icrobiol. Rev.
18:583-607.

2. Peruski, L.F., Jr. and Peruski, A.H. 2003. Rapid diagnostic assays in the genomic biology
era: detection and identification of infectious disease and biological weapon agents.
Biotechniques 35:840-846.

3. Cieslak, T.J., Rowe, J.R., Kortepeter, M.G., Madsen, J.M., Newmark, J., Christopher,
G.W., Culpepper, R.C., and Eitzen, E.M., Jr. 2000. A field-expedient algorithmic
approach to the clinical management of chemical and biological casualties. M~il. M~ed.
165:659-662.

4. Franz, D.R., Jahrling, P.B., Friedlander, A.M., McClain, D.J., Hoover, D.L., Bryne,
W.R., Pavlin, J.A., Christopher, G.W., and Eitzen, E.M., Jr. 1997. Clinical recognition
and management of patients exposed to biological warfare agents. JAM~A 278:3 99-411i.

5. Franz, D.R., Jahrling, P.B., McClain, D.J., Hoover, D.L., Byrne, W.R., Pavlin, J.A.,
Christopher, G.W., Cieslak, T.J., Friedlander, A.M., and Eitzen, E.M., Jr. 2001. Clinical
recognition and management of patients exposed to biological warfare agents. Clin. Lab
M~ed. 21:435-473.

6. YALOW, R. S. and BERSON, S.A. 1959. Assay of plasma insulin in human subj ects by
immunological methods. Nature 184 (Suppl 21):1648-1649.

7. Saiki, R.K., Scharf, S., Faloona, F., Mullis, K.B., Horn, G.T., Erlich, H.A., and Arnheim,
N. 1992. Enzymatic amplification ofbeta-globin genomic sequences and restriction site
analysis for diagnosis of sickle cell anemia. 1985. Biotechnology 24:476-480.

8. Kaper, J.B., Morris, J.G., Jr., and Levine, M.M. 1995. Cholera. Clin. M~icrobiol. Rev.
8:48-86.

9. Sack, D.A., Sack, R.B., Nair, G.B., and Siddique, A.K. 2004. Cholera. Lancet 363:223-
233.

10. Gaffga, N.H., Tauxe, R.V., and Mintz, E.D. 2007. Cholera: a new homeland in Africa?
Am. J. Trop. M~ed. Hyg. 77:705-713.

11. Broeck, D.V., Horvath, C., and De Wolf, M. 2007. Pathogens in focus Vibrio Cholerae:
Cholera toxin. Thze International Journal ofBiochemistry & Cell Biology 39: 1771-1775.

12. Sanchez, J. and Holmgren, J. 2005. Virulence factors, pathogenesis and vaccine
protection in cholera and ETEC diarrhea. Curr. Opin. Immunol. 17:388-398.

13. 2001. Cholera vaccines. Wkly. Epideemiol. Rec. 76: 117-124.










14. Ansaruzzaman, M., Bhuiyan, N.A., Safa, A., Sultana, M., McUamule, A., Mondlane, C.,
Wang, X.Y., Deen, J.L., von Seidlein, L., Clemens, J.D., Lucas, M., Sack, D.A., and
Balakrish, N.G. 2007. Genetic diversity of El Tor strains of Vibrio cholerae 01 with
hybrid traits isolated from Bangladesh and Mozambique. Int. J. M~ed. M~icrobiol. 297:443-
449.

15. Almeida, R.J., Hickman-Brenner, F.W., Sowers, E.G., Puhr, N.D., Farmer, J.J., III, and
Wachsmuth, I.K. 1990. Comparison of a latex agglutination assay and an enzyme-linked
immunosorbent assay for detecting cholera toxin. J. Clin. M~icrobiol. 28:128-130.

16. Yam, W.C., Lung, M.L., and Ng, M.H. 1992. Evaluation and optimization of a latex
agglutination assay for detection of cholera toxin and Escherichia coli heat-labile toxin.
J. Clin. M~icrobiol. 30:2518-2520.

17. Andrade, J.R., da Silva, J.L., Barbosa, M.F., and Caldas, C.A. 1992. Preliminary report
on the application of the coagglutination test for rapid diagnosis of cholera in the upper
Solimoes River area in the Brazilian Amazon region. Braz. J. M~ed. Biol. Res. 25:375-
378.

18. Carillo, L., Gilman, R.H., Mantle, R.E., Nunez, N., Watanabe, J., Moron, J., Quispe, V.,
and Ramirez-Ramos, A. 1994. Rapid detection of Vibrio cholerae 01 in stools of
Peruvian cholera patients by using monoclonal immunodiagnostic kits. Loyaza Cholera
Working Group in Peru. J. Clin. M~icrobiol. 32:856-857.

19. Colwell, R.R., Hasan, J.A., Huq, A., Loomis, L., Siebeling, R.J., Torres, M., Galvez, S.,
Islam, S., Tamplin, M.T., and Bernstein, D. 1992. Development and evaluation of a rapid,
simple, sensitive, monoclonal antibody-based co-agglutination test for direct detection of
Vibrio cholerae 01. FEM~S2icrobiol. Lett. 76:215-219.

20. Jesudason, M.V., Thangavelu, C.P., and Lalitha, M.K. 1984. Rapid screening of fecal
samples for Vibrio cholerae by a coagglutination technique. J. Clin. M~icrobiol. 19:712-
713.

21. Rahman, M., Sack, D.A., Wadood, A., Yasmin, M., and Latif, A. 1989. Rapid
identification of Vibrio cholerae serotype 01 from primary isolation plates by a
coagglutination test. J. M~ed. M~icrobiol. 28:39-41.

22. Agarwal, V., Biswas, M., Pathak, A.A., and Saoji, A.M. 1995. Rapid detection of Vibrio
cholerae 0139 in faecal specimens by coagglutination. Indian1 M.2ed. Res. 101:55-56.

23. Hasan, J.A., Huq, A., Nair, G.B., Garg, S., Mukhopadhyay, A.K., Loomis, L., Bernstein,
D., and Colwell, R.R. 1995. Development and testing of monoclonal antibody-based
rapid immunodiagnostic test kits for direct detection of Vibrio cholerae 0139 synonym
Bengal. J. Clin. M~icrobiol. 33:2935-2939.










24. Qadri, F., Hasan, J.A., Hossain, J., Chowdhury, A., Begum, Y.A., Azim, T., Loomis, L.,
Sack, R.B., and Albert, M.J. 1995. Evaluation of the monoclonal antibody-based kit
Bengal SMART for rapid detection of Vibrio cholerae 0139 synonym Bengal in stool
samples. J. Clin. M~icrobiol. 33:732-734.

25. Hasan, J.A., Bernstein, D., Huq, A., Loomis, L., Tamplin, M.L., and Colwell, R.R. 1994.
Cholera DFA: an improved direct fluorescent monoclonal antibody staining kit for rapid
detection and enumeration of Vibrio cholerae Ol. FEM~S2icrobiol. Lett. 120:143-148.

26. Hasan, J.A., Huq, A., Tamplin, M.L., Siebeling, R.J., and Colwell, R.R. 1994. A novel kit
for rapid detection of Vibrio cholerae Ol. J. Clin. M~icrobiol. 32:249-252.

27. Nato, F., Boutonnier, A., Rajerison, A., Grosjean, P., Dartevelle, S., Guenole, A.,
Bhuiyan, N.A., Sack, D.A., Nair, G.B., Fournier, J.A., and Chanteau, S. 2003. One-step
immunochromatographic dipstick tests for rapid detection of Vibrio cholerae 01 and
0139 in stool samples. Clinical and Diagnostic Laboratory Immunology 10:476-478.

28. Nataro, J.P. and Kaper, J.B. 1998. Diarrheagenic Escherichia coli. Clin. M~icrobiol. Rev.
11:142-201.

29. Hussein, H.S. 2007. Prevalence and pathogenicity of Shiga toxin-producing Escherichia
coli in beef cattle and their products. J. Anim Sci. 85:E63-E72.

30. Mead, P.S., Slutsker, L., Dietz, V., McCaig, L.F., Bresee, J.S., Shapiro, C., Griffin, P.M.,
and Tauxe, R.V. 1999. Food-related illness and death in the United States. Emerg. Infect.
Dis. 5:607-625.

31. Boyd, B. and Lingwood, C. 1989. Verotoxin receptor glycolipid in human renal tissue.
Nephron 51:207-210.

32. Robinson, L.A., Hurley, R.M., Lingwood, C., and Matsell, D.G. 1995. Escherichia coli
verotoxin binding to human paediatric glomerular mesangial cells. Ped'iatr. Nephrol.
9:700-704.

33. Takeda, T., Dohi, S., Igarashi, T., Yamanaka, T., Yoshiya, K., and Kobayashi, N. 1993.
Impairment by verotoxin of tubular function contributes to the renal damage seen in
haemolytic uraemic syndrome. 1. Infect. 27:339-341.

34. O'Brien, A.D., Tesh, V.L., Donohue-Rolfe, A., Jackson, M.P., Olsnes, S., Sandvig, K.,
Lindberg, A.A., and Keusch, G.T. 1992. Shiga toxin: biochemistry, genetics, mode of
action, and role in pathogenesis. Curr. Top. M~icrobiol. Immunol. 180:65-94.

35. Strockbine, N.A., Marques, L.R., Newland, J.W., Smith, H.W., Holmes, R.K., and
O'Brien, A.D. 1986. Two toxin-converting phages from Escherichia coli 0157:H7 strain
933 encode antigenically distinct toxins with similar biologic activities. Infect. Immun.
53:135-140.










36. Yamagami, S., Motoki, M., Kimura, T., Izumi, H., Takeda, T., Katsuura, Y., and
Matsumoto, Y. 2001. Efficacy of postinfection treatment with anti-Shiga toxin (Stx) 2
humanized monoclonal antibody TMA-15 in mice lethally challenged with Stx-producing
Escherichia coli. J. Infect. Dis. 184:738-742.

37. Ostroff, S.M., Tarr, P.I., Neill, M.A., Lewis, J.H., Hargrett-Bean, N., and Kobayashi,
J.M. 1989. Toxin genotypes and plasmid profiles as determinants of systemic sequelae in
Escherichia coli 0157:H7 infections. J. Infect. Dis. 160:994-998.

38. Boerlin, P., McEwen, S.A., Boerlin-Petzold, F., Wilson, J.B., Johnson, R.P., and Gyles,
C.L. 1999. Associations between virulence factors of Shiga toxin-producing Escherichia
coli and disease in humans. J. Clin. M~icrobiol. 37:497-503.

39. Tesh, V.L., Burris, J.A., Owens, J.W., Gordon, V.M., Wadolkowski, E.A., O'Brien, A.D.,
and Samuel, J.E. 1993. Comparison of the relative toxicities of Shiga-like toxins type I
and type II for mice. Infect. Immun. 61:3392-3402.

40. Donohue-Rolfe, A., Kondova, I., Oswald, S., Hutto, D., and Tzipori, S. 2000.
Escherichia coli 0157:H7 strains that express Shiga toxin (Stx) 2 alone are more
neurotropic for gnotobiotic piglets than are isotypes producing only Stxl or both Stxl
and Stx2. 1 Infect. Dis. 181:1825-1829.

41. Magliulo, M., Simoni, P., Guardigli, M., Michelini, E., Luciani, M., Lelli, R., and Roda,
A. 2007. A rapid multiplexed chemiluminescent immunoassay for the detection of
Escherichia coli 0157:H7, Yersinia enterocolitica, Salmonella tyiphimurium, and Listeria
monocytogenes pathogen bacteria. J. Agric. Food Chem. 5 5:493 3-493 9.

42. Wang, L.J., Wu, C.S., Hu, Z.Y., Zhang, Y.F., Li, R., and Wang, P. 2008. Sensing
Escherichia coli 0157:H7 via frequency shift through a self-assembled monolayer based
QCM immunosensor. J. Zhejiang. Univ Sci. B 9:121-131.

43. Johnston, L.M., Elhanafi, D., Drake, M., and Jaykus, L.A. 2005. A simple method for the
direct detection of Salmonella and Escherichia coli 0157:H7 from raw alfalfa sprouts
and spent irrigation water using PCR. J. FoodProt. 68:2256-2263.

44. Jin, H.Y., Tao, K.H., Li, Y.X., Li, F.Q., and Li, S.Q. 2005. Microarray analysis of
Escherichia coli 0157:H7. World J. Gastroenterol. 11:5811-5815.

45. Thevenot, D.R., Toth, K., Durst, R.A., and Wilson, G.S. 2001. Electrochemical
biosensors: recommended definitions and classification. Biosens. Bioelectron. 16:121-
131.

46. Theegala, C.S., Small, D.D., and Monroe, W.T. 2008. Oxygen electrode-based single
antibody amperometric biosensor for qualitative detection of E coli and bacteria in water.
J. Environ. Sci. Health A Tox. Hazard. Subst. Environ. Eng 43:478-487.










47. Liu, Y., Gilchrist, A., Zhang, J., and Li, X.F. 2008. Detection of viable but nonculturable
Escherichia coli 0157:H7 bacteria in drinking water and river water. Appl. Environ.
M~icrobiol. 74:1502-1507.

48. Kalnauwakul, S., Phengmak, M., Kongmuang, U., Nakaguchi, Y., and Nishibuchi, M.
2007. Examination of diarrheal stools in Hat Yai City, South Thailand, for Escherichia
coli 0157 and other diarrheagenic Escherichia coli using immunomagnetic separation
and PCR method. .Sambeastl~ Asian J. Trop. M~ed. Public Health 38:871-880.

49. Stefan, A., Scaramagli, S., Bergami, R., Mazzini, C., Barbanera, M., Perelle, S., and
Fach, P. 2007. Real-time PCR and enzyme-linked fluorescent assay methods for
detecting Shiga-toxin-producing Escherichia coli in mincemeat samples. Can. J
M~icrobiol. 53:337-342.

50. Zhang, W., Bielaszewska, M., Pulz, M., Becker, K., Friedrich, A.W., Karch, H., and
Kuczius, T. 2008. New immuno-PCR assay for detection of low concentrations of shiga
toxin 2 and its variants. J. Clin. M~icrobiol. 46:1292-1297.

51. Auvray, F., Lecureuil, C., Tache, J., Leclerc, V., Deperrois, V., and Lombard, B. 2007.
Detection, isolation and characterization of Shiga toxin-producing Escherichia coli in
retail-minced beef using PCR-based techniques, immunoassays and colony hybridization.
Lett. Appl. M~icrobiol. 45:646-651.

52. Smith, G.P. 1985. Filamentous fusion phage: novel expression vectors that display cloned
antigens on the virion surface. Science 228:1315-1317.

53. Carmen, S. and Jermutus, L. 2002. Concepts in antibody phage display. Brief: Funct.
G~enomic. Proteomic. 1: 189-203.

54. Endemann, H. and Model, P. 1995. Location of filamentous phage minor coat proteins in
phage and in infected cells. Journal of2~olecular Biology 250:496-506.

55. Henry, T.J. and Pratt, D. 1969. The proteins of bacteriophage M13. Proc. Natl. Acad'. Sci.
U. S. A 62:800-807.

56. Russel, M. and Model, P. 1989. Genetic analysis of the filamentous bacteriophage
packaging signal and of the proteins that interact with it. J. Virol. 63:3284-3295.

57. Messing, J. 1991. Cloning in M13 phage or how to use biology at its best. Gene 100:3-
12.

58. Kwasnikowski, P., Kristensen, P., and Markiewicz, W.T. 2005. Multivalent display
system on filamentous bacteriophage pVII minor coat protein. J. Immunol. Methods
307:135-143.

59. Gao, C., Mao, S., Lo, C.H., Wirsching, P., Lerner, R.A., and Janda, K.D. 1999. Making
artificial antibodies: a format for phage display of combinatorial heterodimeric arrays.
Proc. Natl. Acad'. Sci. U. S. A 96:6025-6030.










60. Iannolo, G., Minenkova, O., Petruzzelli, R., and Cesareni, G. 1995. Modifying
fi1amentous phage capsid: limits in the size of the maj or capsid protein. J. Mol1. Biol.
248:835-844.

61. Scott,J.K. and Barbas,C.F.I. 2001. Phage display vectors. In: Phage Display Laboratory
Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 2.1-2.19 pp.

62. Tomlinson, I.M., Walter, G., Marks, J.D., Llewelyn, M.B., and Winter, G. 1992. The
repertoire of human germline VH Sequences reveals about fifty groups of VH Segments
with different hypervariable loops. J. Mol1. Biol. 227:776-798.

63. Rakonjac, J., Jovanovic, G., and Model, P. 1997. Filamentous phage infection-mediated
gene expression: construction and propagation of the gIII deletion mutant helper phage
R408d3. Gene 198:99-103.

64. Chasteen, L., Ayriss, J., Pavlik, P., and Bradbury, A.R. 2006. Eliminating helper phage
from phage display. Nucleic Acids Res. 34:el45.

65. Baek, H., Suk, K.H., Kim, Y.H., and Cha, S. 2002. An improved helper phage system for
efficient isolation of specific antibody molecules in phage display. Nucleic Acids Res.
30:el8.

66. Rondot, S., Koch, J., Breitling, F., and Dubel, S. 2001. A helper phage to improve single-
chain antibody presentation in phage display. Nat. Biotechnol. 19:75-78.

67. Soltes, G., Barker, H., Marmai, K., Pun, E., Yuen, A., and Wiersma, E.J. 2003. A new
helper phage and phagemid vector system improves viral display of antibody Fab
fragments and avoids propagation of insert-less virions. J. Immunol. Methods 274:233-
244.

68. Soltes, G., Hust, M., Ng, K.K., Bansal, A., Field, J., Stewart, D.I., Dubel, S., Cha, S., and
Wiersma, E.J. 2007. On the influence of vector design on antibody phage display. J.
Biotechnol. 127:626-637.

69. Wen, S.X., Teel, L.D., Judge, N.A., and O'Brien, A.D. 2006. Genetic toxoids of Shiga
toxin types 1 and 2 protect mice against homologous but not heterologous toxin
challenge. Vaccine 24: 1142-1148.

70. Weinstein, D.L., Jackson, M.P., Perera, L.P., Holmes, R.K., and O'Brien, A.D. 1989. In
vivo formation of hybrid toxins comprising Shiga toxin and the Shiga-like toxins and role
of the B subunit in localization and cytotoxic activity. Infect. Immun. 57:3743-3750.

71. Guyer, M. S., Reed, R.R., Steitz, J.A., and Low, K.B. 1980. Identification of a sex-factor-
affinity site in Escherichia coli as gw. Cold. Spring. Harb. Symp. Quant. Biol. 45: 135-
140.










72. Riley, L.W., Remis, R.S., Helgerson, S.D., McGee, H.B., Wells, J.G., Davis, B.R.,
Hebert, R.J., Olcott, E.S., Johnson, L.M., Hargrett, N.T., Blake, P.A., and Cohen, M.L.
1983. Hemorrhagic colitis associated with a rare Escherichia coli serotype. N. Engl. J
M~ed. 308:681-685.

73. Karpman, D., Connell, H., Svensson, M., Scheutz, F., Alm, P., and Svanborg, C. 1997.
The role of lipopolysaccharide and Shiga-like toxin in a mouse model ofEscherichia coli
0157:H7 infection. J. Infect. Dis. 175:611-620.

74. Bik, E.M., Gouw, R.D., and Mooi, F.R. 1996. DNA fingerprinting of Vibrio cholerae
strains with a novel insertion sequence element: a tool to identify epidemic strains. J
Clin. M~icrobiol. 34:1453-1461.

75. Heidelberg, J.F., Eisen, J.A., Nelson, W.C., Clayton, R.A., Gwinn, M.L., Dodson, R.J.,
Haft, D.H., Hickey, E.K., Peterson, J.D., Umayam, L., Gill, S.R., Nelson, K.E., Read,
T.D., Tettelin, H., Richardson, D., Ermolaeva, M.D., Vamathevan, J., Bass, S., Qin, H.,
Dragoi, I., Sellers, P., McDonald, L., Utterback, T., Fleishmann, R.D., Nierman, W.C.,
and White, O. 2000. DNA sequence of both chromosomes of the cholera pathogen Vibrio
cholerae. Nature 406:477-483.

76. Evans,T.J. 2000. Septic Shock Methods and Protocols. In Methods in Molecular
Medicine. Humana Press, 19-23.

77. Yi, E.C. and Hackett, M. 2000. Rapid isolation method for lipopolysaccharide and lipid
A from gram-negative bacteria. Analyst 125:651-656.

78. Tsai, C.M. and Frasch, C.E. 1982. A sensitive silver stain for detecting
lipopolysaccharides in polyacrylamide gels. Anal. Biochem. 119:115-119.

79. Sasse, J. and Gallagher, S.R. 2003. Detection of proteins on blot transfer membranes.
Curr. Protoc. Mol1. Biol. Chapter 10:Unitl0.

80. Waldor, M.K., Colwell, R., and Mekalanos, J.J. 1994. The Vibrio cholerae 0139
serogroup antigen includes an O-antigen capsule and lipopolysaccharide virulence
determinants. Proc. Natl. Acad'. Sci. U. S. A 91:11388-11392.

81. Villeneuve, S., Boutonnier, A., Mulard, L.A., and Fournier, J.M. 1999. Immunochemical
characterization of an Ogawa-Inaba common antigenic determinant of Vibrio cholerae
01. Microbiology 145 ( Pt 9):2477-2484.

82. Westphal, O., Luderitz, O., and Bister, F. 1952. Uber Die Extraktion Von Bakterien Mit
Phenol Wasser. Zeitschrift fur Naturforschung Section B-A Journal of Chemical Sciences
7:148-155.

83. Villeneuve, S., Souchon, H., Riottot, M.M., Mazie, J.C., Lei, P., Glaudemans, C.P.,
Kovac, P., Fournier, J.M., and Alzari, P.M. 2000. Crystal structure of an anti-
carbohydrate antibody directed against Vibrio cholerae 01 in complex with antigen:
molecular basis for serotype specificity. Proc. Natl. Acad'. Sci. U. S. A 97:8433-8438.










84. Adey, N.B., Mataragnon, A.H., Rider, J.E., Carter, J.M., and Kay, B.K. 1995.
Characterization of phage that bind plastic from phage-displayed random peptide
libraries. Gene 156:27-31.

85. Persell, D.J. and Robinson, C.H. 2008. Detection and early identification in bioterrorism
events. Fant. Conanunity Health 31:4-16.

86. 2001. Update: Investigation of bioterrorism-related anthrax and interim guidelines for
clinical evaluation of persons with possible anthrax. MWdWR M~orb. Mortal. Wkly. Rep.
50:941-948.

87. Nakamura, M., Watanabe, H., Nishimiya, Y., Tsumoto, K., Ishimura, K., and Kumagai, I.
2001. Panning of a phage VH library using nitrocellulose membranes: application to
selection of a human VH library. J. Biochent. 129:209-212.

88. Yarmush, M.L., Antonsen, K.P., Sundaram, S., and Yarmush, D.M. 1992.
Immunoadsorption: strategies for antigen elution and production of reusable adsorbents.
Biotechnol. Prog. 8: 168-178.

89. Nagumo, Y., Oguri, H., Tsumoto, K., Shindo, Y., Hirama, M., Tsumuraya, T., Fujii, I.,
Tomioka, Y., Mizugaki, M., and Kumagai, I. 2004. Phage-display selection of antibodies
to the left end of CTX3C using synthetic fragments. J. Inanunol. Methods 289: 137-146.

90. Goletz, S., Christensen, P.A., Kristensen, P., Blohm, D., Tomlinson, I., Winter, G., and
Karsten, U. 2002. Selection of large diversities of antiidiotypic antibody fragments by
phage display. J. Mol1. Biol. 315:1087-1097.

91. Meulemans, E.V., Slobbe, R., Wasterval, P., Ramaekers, F.C., and van Eys, G.J. 1994.
Selection of phage-displayed antibodies specific for a cytoskeletal antigen by competitive
elution with a monoclonal antibody. J. Mol1. Biol. 244:353-360.

92. Zhu, Y., Ho, B., and Ding, J.L. 2003. Sequence and structural diversity in endotoxin-
binding dodecapeptides. Biochina. Biophys. Acta 1611:234-242.

93. Thomas, C.J., Sharma, S., Kumar, G., Visweswariah, S.S., and Surolia, A. 2003.
Biopanning of endotoxin-specific phage displayed peptides. Biochens. Biophys. Res.
Conanun. 307:133-138.

94. Noda, K., Yamasaki, R., Hironaka, Y., and Kitagawa, A. 2001. Selection of peptides that
bind to the core oligosaccharide of R-form LPS from a phage-displayed heptapeptide
library. FEM~S2icrobiol. Lett. 205:349-354.

95. Harlow,E. and Lane,D. 1988. Antibodies a Laboratory Manual. Cold Spring Harbor
Laboratory, Cold Spring Harbor, NY. 23-34 pp.

96. Paoli, G.C., Chen, C.Y., and Brewster, J.D. 2004. Single-chain Fv antibody with
specificity for Listeria nzonocytogenes. J. Inanunol. Methods 289: 147-155.










97. Leonard, P., Safsten, P., Hearty, S., McDonnell, B., Finlay, W., and O'Kennedy, R. 2007.
High throughput ranking of recombinant avian scFv antibody fragments from crude
lysates using the Biacore A100. J. Inanunol. Methods 323:172-179.

98. Su, B., Hrin, R., Harvey, B.R., Wang, Y.J., Ernst, R.E., Hampton, R.A., Miller, M.D.,
Strohl, W.R., An, Z., and Montgomery, D.L. 2007. Automated high-throughput
purification of antibody fragments to facilitate evaluation in functional and kinetic based
assays. J. Inanunol. Methods 322:94-103.

99. Bannister, D., Wilson, A., Prowse, L., Walsh, M., Holgate, R., Jermutus, L., and
Wilkinson, T. 2006. Parallel, high-throughput purification of recombinant antibodies for
in vivo cell assays. Biotechnol. Bioeng. 94:931-937.

100. Lobel, L.I., Morseman, J.P., Zeng, X., Lustbader, J.W., Chen, H., and Allnutt, F.C. 2001.
Development of a fluorescence based high throughput assay for antagonists of the human
chorionic gonadotropin receptor extracellular domain: analysis of peptide inhibitors. J
Bionzol. Screen. 6:151-158.

101. Vanhercke, T., Ampe, C., Tirry, L., and Denolf, P. 2005. Rescue and in situ selection and
evaluation (RISE): a method for high-throughput panning of phage display libraries. J.
Bionzol. Screen. 10: 108-117.









BIOGRAPHICAL SKETCH

Crystal Harpley (Mazur) was born in Columbus, OH; and shortly after moved to

Wellington, FL, where she completed grade school. In 2006 she graduated with high honors

from the University of Florida, with a major in microbiology and cell sciences and a minor in

chemistry. During the process of finishing her master' s thesis, Crystal Harpley got married and

is now Crystal Mazur.





PAGE 1

1 OPTIMIZATION OF METHODS FOR PH AGE DISPLAY USI NG SINGLE-CHAIN VARIABLE FRAGMENT PHAGEMID LIBRARIES By CRYSTAL J. HARPLEY A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Crystal J. Harpley

PAGE 3

3 To my parents, siblings, and Matt. They, never doubted my abilities and always pushed me harder when I doubted myself.

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank m y mentor, Paul A. Gulig, for teaching, guiding, and pushing me to be a better scientist. I would also like to thank my co-workers in the Gulig lab for their encouragement, help, and making me smile thro ughout the work day. Lastly, I would like to thank my committee members, Shouguang Jin and Anita Wright, for their help and guidance.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES.........................................................................................................................9 LIST OF ABBREVIATIONS........................................................................................................ 11 ABSTRACT...................................................................................................................................13 CHAP TER 1 INTRODUCTION..................................................................................................................15 Methods for Bacteriological Detection and Analysis.............................................................15 Vibrio cholerae .......................................................................................................................18 Escherichia coli O157:H 7......................................................................................................20 Phage Display.........................................................................................................................23 Recombinant Phage Libraries................................................................................................. 24 Tomlinson I + J Human Synthetic VH + VL Phagemid Libraries.................................... 25 New England Biolabs 12mer Peptide Phage Library...................................................... 26 Tools for Production of Phages from Phagemids................................................................... 27 Helper Phage...................................................................................................................27 Helper Phage with gIII Mutations ................................................................................... 28 Helper Plasmids...............................................................................................................29 2 MATERIALS AND METHODS........................................................................................... 35 Bacterial Strains, Media, and Growth Methods...................................................................... 35 Biopanning of Phage Display Libraries.................................................................................. 36 Panning on Immunotubes................................................................................................36 Panning in Suspension..................................................................................................... 38 Panning on Microtiter Wells........................................................................................... 39 Panning on Nitrocellulose Paper.....................................................................................41 Spot Titer of Phages........................................................................................................43 Spread Titer of Phages.....................................................................................................43 Amplification of Phages.................................................................................................. 43 High Throughput Production of Soluble An tibody Fragm ents (scFv antibodies).......... 44 Deoxyribonucleic Acid Manipulations................................................................................... 44 Plasmid Extractions......................................................................................................... 44 Agarose Gel Electrophoresis...........................................................................................44 Electroporation of Plasmids............................................................................................45 Electrocompetent Cells......................................................................................................... ..45 Enzyme-Linked ImmunoSorbent Assays (ELISAs)............................................................... 46

PAGE 6

6 Infection Efficiency........................................................................................................... .....47 Protein and Lipopolysacchar ide (LP S) Manipulations........................................................... 48 Phenol-Water Extraction of LPS.....................................................................................48 TRIzol Reagent Extraction of LPS..................................................................................49 Extraction of Periplasmic Proteins.................................................................................. 49 Determination of Protein Concentration......................................................................... 50 Sodium Dodecyl Sulfate-Polyacrylam ide Gel Electrophoresis (SDS-PAGE) ................ 50 Coomassie Blue Staining................................................................................................. 51 Tsai-Frasch Silver Staining............................................................................................. 51 Western Blot....................................................................................................................52 Lipopolysaccharide Saturation to Nitrocellulose Paper .................................................. 53 Colony Blot with scFv.....................................................................................................54 3 RESULTS...............................................................................................................................57 Rationale for Study.................................................................................................................57 Specific Aim 1: Panning to V. cholerae LPS.........................................................................57 The Phenol-Water Method Extracted V. cholerae LPS Most Closely Resembled the Comm ercially Acquired V. cholerae LPS...................................................................58 More V. cholerae LPS Can Be Bound to N itrocellulose Paper th an to a Microtiter Well ..............................................................................................................................62 Panning to V. cholerae LPS Failed to Yi eld Phages that W ere Specific to V. cholerae LPS........................................................................................................... 64 Conclusion of Specific Aim 1......................................................................................... 66 Specific Aim 2: Improve Panning and Screening Process of Biopanning .............................67 Elution of Bound Phages by Trypsin duri ng Panning W as Optimal between 10 and 30 Minutes................................................................................................................... 68 Screening of scFv Proteins by a High Throughput ELISA W as Acceptable.................. 71 Screening of Phagemid Particles by a High Throughput ELISA W as Not Optimal....... 73 Conclusion of Specific Aim 2......................................................................................... 75 Specific Aim 3: Improve Phag em id Particle Production........................................................ 76 Escherichia coli TG1 Harboring the Hyperphage Geno me Was Not an Optimal Phagemid Particle Amplification Tool........................................................................ 77 Helper Plasmids Were Not an Optimal Phage mid Particle Amplification Tool.............80 Homemade Hyperphage Titers Were Incre ased with Amplification in E. coli MG1655 (pGTR203)...................................................................................................85 Conclusion of Specific Aim 3......................................................................................... 88 Specific Aim 4: Isolation of Specific Reco m binant Phage to Stx2 Toxin of E. coli O157:H7..............................................................................................................................89 Conclusion of Specific Aim 4......................................................................................... 95 4 DISCUSSION.......................................................................................................................119 Specific Aim 1: Panning to V. cholerae LPS .......................................................................120 Specific Aim 2: Improve Panning and Screening Process of Biopanning ...........................126 Specific Aim 3: Improve Phag em id Particle Production...................................................... 135

PAGE 7

7 Specific Aim 4: Isolation of Specific Recom binant Phage to Stx2 Toxin of E. coli O157:H7............................................................................................................................146 5 EPILOGUE...........................................................................................................................153 LIST OF REFERENCES.............................................................................................................154 BIOGRAPHICAL SKETCH.......................................................................................................163

PAGE 8

8 LIST OF TABLES Table Page 2-1 Bacterial strains and plasmids used................................................................................... 55 3-1 Infection efficiencies of E. coli TG1 containing various helper plasm ids with phagemid particles............................................................................................................. 97 3-2 Infection efficiencies of E. coli TG1 (M13cp-dg3-sm ) with phagemid particles.............. 97 3-3 Infection efficiencies of E. coli TG1 (M13cp-CT-sm ) and E. coli JM109 (M13cpCT-sm) with phagemid particles........................................................................................ 98 3-4 Comparison of transducing units to pa rticle s per milliliter of amplified phagemid particles..............................................................................................................................98 3-5 Comparison of transducing units to pa rticle s per milliliter of amplified homemade hyperphage produced from FE. coli strains...................................................................... 99

PAGE 9

9 LIST OF FIGURES Figure Page 1-1 Structure of M13 phage..................................................................................................... 32 1-2 Structures of antibodies......................................................................................................33 1-3 Genetic map of pIT2 phagemid vector from the Tomlinson scFv library i .......................34 3-1 Analysis of TRIzol Reagent-extracted V. cholerae N16961 LPS by SDS-PAGE. .........100 3-2 Analysis of phenol-water-extracted V. cholerae 569B LPS by SDS-PAGE. .................. 101 3-3 Analysis of the saturation limit of V. cholerae 56 9B LPS to microtiter wells by ELISA..............................................................................................................................102 3-4 Analysis of the saturation limits of primary antibodies to V. cholerae 569B LPScoated m icrotiter wells by ELISA.................................................................................... 103 3-5 Saturation of V. cholerae 569B LPS to nitrocellulose paper. .......................................... 105 3-6 Titers of eluted phagemid particles from the first one-round panning optimization experiment..................................................................................................................... ...106 3-7 Titers of eluted phagemid particles from the second one-round panning optimization experiment..................................................................................................................... ...107 3-8 Titers of eluted phagemid particles from the third one-round panning optimization experiment..................................................................................................................... ...108 3-9 Analysis of eluted phagemid particle s from the third one-round panning optimization experiment by ELISA......................................................................................................109 3-10 Analysis of anti-BSA ( -BSA) scFv proteins produced in m icrotiter wells by ELISA.. 110 3-11 Analysis of -BSA and Vc86 phagem id particles produced in microtiter wells by ELISA..............................................................................................................................111 3-12 Analysis of -BSA phagem id particles produced in microtiter wells by ELISA............ 112 3-13 Analysis of -BSA phagem id particles produced by hyperphage by ELISA..................113 3-14 Analysis of phagemid particles (1-22, 43-64) selected from panning against Stx2 toxin preparation by ELISA.............................................................................................114 3-15 Analysis of phagemid particles (1, 3-5, 23-42, 65-84) selected from panning against Stx2 toxin preparation by ELISA....................................................................................115

PAGE 10

10 3-16 Analysis of anti-Stx2 monoclonal an tibodies and clones 46, 48, and 49 by ELISA.. ..... 116 3-17 Western blot analysis of phagem id clones from panni ng on Stx2 toxin preparation...... 117 3-18 Analysis of Stx2 toxin preparation (Toxin Technologies ) by SDS-PAGE......................118

PAGE 11

11 LIST OF ABBREVIATIONS 2xTY AG 2xTY m edium containing 100 g/mL and 1% (w/v) glucose BSA Bovine serum albumin BSG Phosphate-buffered saline c ontaining 0.1% (w/v) gelatin CFU Colony forming unit CT Cholera toxin ddH2O Deionized distilled water dH2O Deionized water DNA Deoxyribonucleic acid ECL Enhanced chemiluminescence ELISA Enzyme-linked immunosorbent assay HC Hemorrhagic colitis HRP Horseradish peroxidase HUS Hemolytic uremic syndrome Hy Hyperphage LPS Lipopolysaccharide mAb Monoclonal antibody MWCO Molecular weight cut off OD Optical density pAb Polyclonal antibody PBS Phosphate-buffered saline PBST-0.01 PBS containing 0.01% (v/v) Tween-20 PBST-0.05 PBS containing 0.05% (v/v) Tween-20 PC Phosphate-citrate PCR Polymerase chain reaction

PAGE 12

12 PEG Polyethylene glycol PFU Plaque forming unit RF Replication factor RNA Ribonucleic acid S:N Signal to noise scFv Single chain F variable SDS-PAGE Sodium dodecyl sulfatepolyacrylamide gel electrophoresis Stx-2 Shiga-like toxin 2 TMB 3,3 ,5,5 -tetramethylbenzidine tu Transducing unit MOI Multiplicity of infection Vc Vibrio cholerae v/v Volume per volume w/v Weight per volume

PAGE 13

13 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science OPTIMIZATION OF METHODS FOR PH AGE DISPLAY USI NG SINGLE-CHAIN VARIABLE FRAGMENT PHAGEMID LIBRARIES By Crystal J. Harpley August 2008 Chair: Paul A. Gulig Major: Medical Sciences Detection assays for biological agents and their products are im portant to identify diseasecausing organisms in water, air, food, and patient samples to enable prevention or treatment of the disease. A commonality in almo st all detection assays are proteins that specifically bind to a target molecule. While monoclonal antibodies are the most common detection reagents used in detection assays, phage display reagents are becoming more prev alent. Phage display involves the display of recombinant peptides on bact eriophages. These recombinant phages can be panned against a target antigen to select reco mbinant peptides that bind specifically to the antigen. Optimization of protocols and reagents used in phage displa y was the goal of this thesis. Optimization of whole cell panning with the Tomlinson J human synthetic VH + VL (scFv) phagemid library revealed that a trypsin eluti on between 10 and 30 minutes eluted the highest number of phagemid particles; this may be because the phagemid particles were degraded by trypsin treatment after 45 minutes. A high th roughput ELISA for the production and screening of scFv proteins was developed and demonstrated that scFv proteins produced in a microtiter well could be screened by ELISA and produce a de tectable S:N even when diluted 1:10. A high throughput ELISA for the production and screeni ng of phagemid particles was attempted, but

PAGE 14

14 phagemid particles could not be produced in a microtiter well in titers high enough to make a high throughput ELISA acceptable for screening of phagemid particles. Various phagemid particle amplification t ools were analyzed. The hyperphage genome was incorporated into E. coli and maintained as a plasmid to enable phagemid particle amplification. However, these strains had low in fection efficiencies with phagemid particles. Helper plasmids were incorporated into E. coli to enable phagemid particle amplification. However, these strains produced low yields of phagemid particles. The titer of laboratoryproduced hyperphage was improved by use of E. coli MG1655 (pGTR203), expressing the M13 gIII gene. The improved high titer hyperphage was used to successfully amplify the Tomlinson J scFv phagemid library to ensure that the redund ancy of phagemid clones was maintained. Using the newly amplified libra ry to pan against an E. coli O157:H7 Stx2 toxin preparation resulted in 89% of phagemid particles from the thir d round of panning recognizing the Stx2 toxin preparation in an ELISA. Previous panning procedures used in this la boratory resulted in no better than 1% of selected phagemids being specific to the target antigen, and often no usable specific clones were isolated. The optimization of panning procedures and reagents developed in this thesis greatly increased the efficiency of sel ection of phagemid partic les specific to the ta rget antigen. Using the improved reagents and procedures develope d in this theis yielded over 80% positive phagemid clones selected against E. coli O157:H7 flagella by another lab member. The phagemids recognized the major flagellin protein by Western blot (data no t shown). Therefore, the improvements on phage display techniques and t ools described in this thesis are, in fact, useful and offer promise of success for continuing studies in the laboratory.

PAGE 15

15 CHAPTER 1 INTRODUCTION Methods for Bacteriological Detection and Analysis There has always been a need for detection of biological agents. One basic reason for this need is for prevention of disease by identifying biol ogical agents in food, water, or air. Of key interest are organism s that have po tential as bioterrorism agents. With resent bioterrorist events, the need for rapid detection of bioterrorism agents has increased (1). Another reason for detection of biological agents is for the purpos e of diagnosing disease by detecting agents in patient samples so that suitable treatment can be given. Detection is also of interest to epidemiology. By detecting agents that cause disease and analyzing them through epidemiology, measures can be implemented to control or prev ent outbreaks. Various methods have been used to detect organisms or their products. Since the discovery of microscopic organisms, methods for detecting the organisms have evolved. One of the first methods of detecti on was culturing an organism in enrichment medium, followed by selective medium. Afterwards, the organism would go through various biochemical and metabolic tests for identificatio n. However, culture-based identification of organisms can take days or even weeks. As seen in the anthrax attacks in the fall of 2001 and the SARS virus outbreaks of 2002-2003, symptoms can occur within days of exposure and infection can spread rapidly. These reasons enforce the need for rapid detection of biological agents (2). The rapid detection of biological agents is necessary to treat individuals at risk, limit transmission of the disease, improve public he alth surveillance and epidemiology, and monitor environmental impact (3-5). Methods have progressed to optimize detection of agents and their produc ts in clinical and environmental settings. Over the last 25 years, as says for detection and identification of agents

PAGE 16

16 have improved immensely. Particularly, the reag ents and detection equipment have improved to allow for the detection and identification of agents in as little as a few minutes (2). An ideal detection system would be rapid, sensitive, selective, inexpensive, and would not require extensive training of personnel to operate the system. While th ere is no single optimal-detection method, there are numerous methods for detectin g a variety of agents in a variety of environments. Immunological tools are one of th e most widely used and successful methods for detection of biological agents. Since the first radioimmunoassay was developed by Yalow and Berson in 1959 for the detection of human insulin (6), imm unoassays have expanded and been used for the detection of a variety of agents. The main components of an immunoassay are summarized by Andreotti et al. as follows: Immunoassays rely upon four basic components regardless of the application and underlying technolo gy: (i) the antigen to be de tected; (ii) the antibody or antiserum used for detection; (iii) the method to separate bound antigen and antibody complexes from unbound reactants; and (iv) the detection method. The e fficacy of any given immunoassay is dependent on two major factors: the efficien cy of antigen-antibody complex formation and the ability to detect these complexes. The most important component of an imm unoassay is the antibody. The discovery of different types of antibodies has altered the range and scope of immunoassays. Polyclonal antibodies have largely been supplanted by mo noclonal antibodies. In the current age, recombinantly engineered antibodies are suppl anting monoclonal antibodies (2). The demand for a variety of immunological a ssays reflects the growing number of assays developed to optimize their use by increasing their sensitivity, speed, handling, and cost. The specific binding of antibody to antigen has been coupled with a variety of detection ap plications including

PAGE 17

17 fluorescence, enzymatic activity, chemiluminescence, electrochemiluminescence, metallic beads, and many more. The detection of these complexes can be assayed on a variety of platforms such as biosensors, flow cytometry, microarray, and lateral flow diffusion devices (1). Perhaps the most common immunological assa ys are Enzyme-Linked ImmunoSorbent Assays (ELISAs). Such assays either have antibody or antigen bound to a solid support that enables specific, sensitive, and quantitative dete ction of antigen or antibody by op tical-density detection of a colorimetric signal. Nucleic acid-based tools for detection have a dvanced in the past few decades. Of main interest is Polymerase Chain Reaction (PCR), which was invented in 1983 by Kary Mullis and coworkers (7). Polymerase chain reaction involves the amplification of DNA by use of oligonucleotide primers, heat stable DNA polym erase, and nucleotides in an exponential capacity. The original PCR has since been altered to give quantitative real time-PCR (q-PCR or kinetic PCR) and reverse transcription (RT )-PCR. Quantitative real time-PCR involves the amplification along with quantif ication of the DNA, while RT-P CR involves reverse transcribing a piece of RNA into DNA followed by PCR amplification of the DNA. Advances in PCR chemistry and thermocyclers have shortened th e length of DNA amplification from a few hours to minutes. With small sample volumes in th e amount of a few microliters and containing as little as one bacterial cell, PCR is one of the most sensitive assays available. However, this sensitivity increases the risk of contaminati on generating false positiv es. Field-based PCR amplification and identification is not common du e to the complexity of the system and highly trained personnel required to operate and interpret the system. Perhaps the greatest constraint of nucleic acid-based detection assays is the availability of genomic sequence data of biological agents (2). Detection assays are continually being invented and improve d to detect biological

PAGE 18

18 agents. Because there are numerous detecti on environments, conditions, and agents, the development and improvement of detection assays will continue to progress. Vibrio cholerae Vibrio cholerae is a g ram-negative, curved rod-sh aped bacterium with a single polar flagellum found primarily in estuarine and marine environments. It is a facultative human pathogen causing the pandemic diarrheal disease cholera (8). Cholera is characterized by profuse watery diarrhea, vomiting, and leg cram ps leading to dehydration and shock (9). Cholera infection occurs through th e ingestion of food or water contaminated with the bacterium. Because cholera can be prevented by proper sa nitation and hygiene, it is uncommon in industrialized countries and is most prevalent in the Indian subcontinent and sub-Saharan Africa (10). Vibrio cholerae uses its toxin co-regulated pilus (TCP ) to colonize the small intestine; once attached to the small intestine, the bacter ium secretes cholera toxi n (CT) (11). Cholera toxin binds to the epit helial cell receptor, GM1, and is transported into the cell. Cholera toxin is an AB toxin composed of a catalytically active A-subunit surrounded by a homopentameric Bsubunit. Once the CT is internal ized, it is transported in a retr ograde pathway through the Golgi to the endoplasmic reticulum. In the endoplasmic re ticulum it is retrotranslocated to the cytosol. In the cytosol CT catalyzes ADP-ribosy lation of the GTP-binding protein Gs (8) causing adenylate cyclase to become constitutively activ ated. The increase in cAMP levels leads to secretion of Cl-, HCO3 -, and water from epithelial cells into th e intestinal lumen causing diarrhea. The loss of water can amount to 30 liters per day, and without proper rehydration treatment can lead to a 30% mortality rate (12). Under-repo rting of cholera infections is a great problem. It is estimated that 3 to 5 million cases and 120,000-200,000 deaths occur worldwide annually (13). However, this number could easily be multip lied by a factor of ten due to unreported cases.

PAGE 19

19 There are many different serogroups of V. cholerae, most of which do not cause acute diarrhea. Vibrio cholerae is classified into serovars or serogroups on the basis of its lipopolysaccharide (LPS) O-antigen (14). There are at least 200 known serogroups, of which serogroups O1 and O139 are the only ones that ca use epidemic or endemic cholera. The O1 serogroup can be further distinguished into thr ee serotypes. Ogawa and Inaba are the most common serotypes, with Hikojima being rarely reported. These sero types can be further classified into two biotypes, El Tor and classica l, that differ in their biochemical properties and phage susceptibilities (8). With the enormous numbers of cases of cholera every year, there is a need for an effective diagnosis tool for patient a nd environmental samples. De tecting endemic serogroups of V. cholerae early in an outbreak is extremely importan t for control of an epidemic. A problem with detection of V. cholerae is that cholera is a disease of developing countries. Outbreaks normally occur around water-ravaged areas where la boratories are not prevalent. Therefore, field-based assays are the most eff ective tools for early detection of V. cholerae. Because there are many serogroups of V. cholerae most of which present mild symptoms, it is important to distinguish the epidemic strains from the non-epidem ic strains. Vibrio cholerae O1 and O139 are the epidemic strains. Determining features of these serogroups are the O-antigen of their LPS and their ability to produce CT. Therefore, field tests usually detect these specific antigens. There are many rapid diagnostic tests for choler a. Some of these detect CT by passivelatex agglutination (15,16), while others detect the O-antigens of LPS from O1 (17-21) and O139 (22-24) strains of V. cholerae A commonly used rapid V. cholerae diagnostic tool is the multistep colloidal gold-based colorimetric im munoassay known as SMART. This monoclonal antibody-based test was developed for the detection of V. cholerae O1 (25,26) and O139 (24)

PAGE 20

20 strains in stool specimens. SMART is 95% sensitive and 100% specific to V. cholerae O1 strains (26) and 100% sensitive and 97% sp ecific to O139 strains (24). A one-step immunochromatographic dipstick test for the detection of V. cholerae O1 and O139 LPS in stool samples was invented by the Institute Pasteur in Paris. This assay requires minimal technical skill and rapidly detects thresholds of purified LPS at 10 ng/mL for V. cholerae O1 and at 50 ng/mL for V. cholerae O139 strains in approximately 10 minutes (27). With continuing improvement and use of methods for detection of V. cholerae, outbreaks could be lessened or prevented with proper treatment and containment. Escherichia coli O157:H7 Escherich ia coli O157:H7 is an enterohemorrhagic serotype of E. coli that causes hemorrhagic colitis (HC) and hemolytic uremic syndrome (HUS). The bacterium is primarily found on cattle farms and colonizes cattle, swine, a nd deer intestines with subclinical effects to the animals (28). Disease in humans comes from the ingestion of beef milk, vegetables, and other products that are contaminated with E. coli O157:H7. Symptoms from an E. coli O157:H7 infection include mild diarrhea abdominal pain, vomiting, bloody diarrhea, HC, strokes, and HUS (28). Over the past 23 years, 146 Shiga-like toxin producing E. coli (STEC) outbreaks and sporadic cases of human illnesses have been traced to consumption of beef contaminated with various E. coli O157 strains (29). Most of these illn esses were caused by infection with E. coli O157:H7. Of the 146 outbreaks and sporadic cases 89% occurred in the United States (29). The large number of cases in the Unites States compared with the rest of the world can be explained by high levels of beef consumption and availability of E. coli O157 diagnostic methods in the United States. The Center for Disease Control and Prevention estimated that E. coli O157:H7 is

PAGE 21

21 responsible for approximately 73,000 illnesses and more than 60 deaths per year in the United States (30). Escherichia coli O157:H7 produces Shiga toxins that ar e responsible for human disease. These Shiga toxins produce severe cytopathic effects and have a high degree of homology with Shiga-toxin (Stx) of Shigella dysenteriae type 1 (28). The Stx toxin is a member of the AB toxin family. It is composed of a catalytically active A-subunit surrounded by a pentameric B-subunit. The B subunits specifically bind to glycosphingolip id globotriosylceramide (Gb3) receptors (31) of the renal glomerular endotheli al, mesangial (32) and tubular ep ithelial cells (33). Upon entry into the cell, the catalytically active A1 subunit cleaves ri bosomal RNA leading to the cessation of protein synthesis and cell d eath (34). Not only does Stx toxin damage host cells, but it may also increase the adherence of E. coli O157:H7 to epithelial cells l eading to increased risk of colonization. Tissue culture experiments show ed that Stx toxin evoked an increase in a eukaryotic receptor, nucleolin, that binds to the E. coli O157 attachment factor, intimin, leading to increased cell adherence (32). There are two distinct toxinconverting bacteriophages ( phages), 933J and 933W, in E. coli O157:H7 that generate two genetic ally related toxins that are an tigenically distinct but create similar biologic effects (35). These two toxins are called Shiga toxins I and II (Stx1 and 2). Experimental studies suggest that E. coli that produces the Stx2 toxin is more virulent than E. coli that produces Stx1 or both Stx1 and Stx2 toxins (36). Escherichia coli strains producing Stx2 toxin are more frequently linked with th e development of HUS than Stx1 toxin-producing strains (37,38). In mouse studies, the lethal dose of purified St x2 toxin is 400 times lower than that of Stx1 toxin (39). In pi glet studies, Stx2 toxin-producing E. coli strains caused more severe neurologic symptoms than strains producing bot h Stx1 and 2 toxins or only Stx1 toxin (40).

PAGE 22

22 Detection of E. coli O157:H7 has become of increasing im portance to the food industry in the United States as outbreaks continue to o ccur. Traditional met hods of detection of E. coli O157:H7 involve plating and cu lturing, enumeration methods, biochemical testing, microscopy, and flow cytometry. Other methods have been developed, including immunoassays (41), immunomagnetic separations ( 42), nucleic acid probe-based me thods based on hybridization and polymerase chain reaction (PCR) (43), and DNA mi croarrays (44). However, many of these assays are time-consuming and not suitable for rapid detection of E. coli O157:H7. Therefore, biosensors have been developed for the rapid detection of E. coli O157:H7 cells and Stx toxins. An electrochemical biosensor is a self-contai ned integrated device, which is capable of providing specific quantitative or semi-quantitative analytical information using a biological recognition element (biochemical receptor) which is retained in direct spatial contact with an electrochemical transduction elem ent (45). Some advantages of biosensors include their continuous data acquisition ability, target specifi city, fast response, mass produce feasibility, and the simplicity of sample preparation. A quartzcrystal microbalance (QCM ) has been developed for detection of E. coli O157:H7 cells. Upon specific binding of E. coli O157:H7, the QCM uses its ultra sensitive mass-measuring sensor to detect decreases in the crystal-resonance frequency to enable detection of 2.0 x 102 CFU/mL of E. coli O157:H7 (42). An amperometric biosensor for E. coli O157:H7 cells made use of a dissolvedoxygen probe to enable detection of 50 CFU/mL in as little as 20 minutes of prep aration and processing time. Upon binding of the bacterial cells, a decrease in en zyme activity results in a change in oxygen concentration that was detected with a Clark-type oxygen electrode probe (46). Polymerase chain reaction-based assays are also very common for detection of E. coli O157:H7 cells (47-51) and Stx toxins (50).

PAGE 23

23 Phage Display Phage display was initially described in 1985 by George P. Sm ith (52) as a means to display foreign proteins on fila mentous bacteriophage. Filamentous phages are non-lytic phages with circular ssDNA genomes. Of the filament ous phages, the Ff family (f1, fd, and M13) phages infect F+ E. coli through binding with their F pilus. These phages are useful tools to link genotype and phenotype of select recombinant proteins. This link was created by encoding a foreign polypeptide in-frame with a coat protei n gene of the M13 phage. Phage display could theoretically be implemented with any phage, but filamentous phages have been the most widely used. Of the filamentous phages, M13 is the most commonly used. The M13 phage (Fig. 1-1) replicates in E. coli, turning the bacterium into a phageproduction factory (53). The b acteria harboring these phages do not lyse, but undergo reduced cell growth due to the stress of phage producti on. The M13 phage contains a 6.4-kb, circular, single-stranded DNA genome that enc odes phage proteins I to XI. Fi ve of these proteins are coat proteins. The major coat protein (pXIII) is pr esent in approximately 2,700 copies and protects the genome in a cylindrical manner. The minor coat proteins pVII and pIX are necessary for efficient particle assembly, while the minor coat proteins pIII and pVI are necessary for particle stability and infectivity (54-56). The pIII protein mediates the binding of the phage to the F pilus and is necessary for viral uncoa ting and phage DNA transfer to the cytoplasm of the bacterium (55). Host enzymes then convert the ssDNA into supercoiled dsDNA, known as the replicative form (RF) (55). The RF is essential to the pha ge display system because it can be purified and manipulated just like a plasmid. Through the ma nipulation of the RF of M13 came some of the earliest cloning vectors (57). Through some of these vectors came the development of recombinant phage libraries.

PAGE 24

24 During assembly of M13, the foreign protein is fused to a coat protein and displayed on the surface of the phage. The minor coat protein II I (pIII) is the most common protein for fusions, but the major coat protein XIII (pXIII) and the other minor coat proteins of M13 have also been used for recombinant fusions (58,59). Phage di splays using the pIII and pXIII proteins have different advantages. Using the pXIII protein as the fusion protein enab les high copy display of the recombinant protein because there are over 2,700 copies of the pXIII protein on the surface of the phage. However, a drawback in using the pXIII protein as the fusion protein is its limitations in the size of the displayed protein (6 0). The pXIII protein can only display peptides less than six amino acids in length before the function of the coat protein becomes compromised and the number of infectious partic les plummets. If the size of the display peptide increases to eight amino acids, only 40% of the phages are inf ective; if the display peptide increases to 16 amino acids, less than 1% of the phages are in fective. Recombinant fusions using the pIII protein are not as restricted in the size of the display peptide (53). Th e pIII-fusion protein can display peptides of 100 amino acids or greater before the ability of the pIII protein to bind to the F pilus of E. coli becomes compromised (52). Also, since there are only five copies of the pIII protein on the surface of phage particles, the abilit y to select high affinity binding phage particles is greater than that of the abil ity of the pXIII-fusion protein. Recombinant Phage Libraries Recom binant phage libraries are composed of pha ges that display a fused protein to a coat protein of the phage. There are two main phage display libraries, phage and phagemid libraries. Phage libraries are M13 with the addition of a recombinant fusion to the gIII gene, while phagemid libraries are just plasmids that contain a recombinant fusion to the gIII gene. There are advantages and disadvantages to both systems. Phagemid libra ries are advantageous due to their greater library diversity be cause of their higher ligation-transformation efficiency, and their

PAGE 25

25 simplicity enables easier genetic manipulation than with phage vectors. Also, phagemid particles isolated from phagemid libraries are able to be produced to generate more phagemid particles or secrete the recombinant protein, which is usually the ultimate end product. A disadvantage of phagemid particles is their depende nce on the aid of a helper phage to provide the rest of the M13 proteins in trans to enable phagemid partic le production and assembly. Phage libraries are advantageous due to their lack of dependence on helper phages. Because phage libraries are M13 phage with some alterati ons, they are capable of propa gating themselves by simple infection of E. coli. A disadvantage to phage lib raries is that they are less stable than phagemid libraries (61). Also, phage particles are only able to produce phages and have to be further manipulated to be able to produce the recombinant protein alone. Tomlinson I + J Human Synthetic VH + VL Phagemid Libraries The Tomlinson libraries are nave libra ries comprised of approximately 1 x 108 random phagemids derived from non-immunized human d onors (62). Nave libraries enable greater diversification of antibody genes, increasing the probability of isolating phagemids specific to a wide variety of targets. The Tomlinson libraries also encode greater diversity through random side-chain diversification. These phagemid libra ries encode a single chain F variable (scFv) gene fusion to the gIII gene of a library vector plasmid contained in a M13 phage. The scFv (Fig. 1-2) is composed of a single polypeptide with VH and VL domains that are joined by a flexible glycine-serine linker. The Tomlinson libraries were construc ted by use of reverse transcription and PCR to amplify the VH and VL antibody genes from B-lymphocytes of human donors. Universal degenerate primers were then used to anneal to the 5 end of the exons encoding the antibody V-gene, which is conserved in humans. The mRNA from the B lymphocytes was converted to cDNA, which represents the VH and VL antibody genes. The cDNA was then PCR assembled using an overl ap extension technique and contained the

PAGE 26

26 restriction enzyme sites for subcloning into the pIT2 library vector. The library vector, pIT2, encodes an M13 origin of replication, an ampicillin resistance gene ( bla ), and both His6 and myc tags (Fig. 1-3). The pIT2 phagemids were electroporated into E. coli, where they were superinfected with helper phage to generate phagemid particles displaying a scFv-pIII fusion protein. The phagemid particles display the sc Fv on the surface of the phage particles and encode the scFv gene in the phagemid genome, linking phenotype with genotype. New England Biolabs 12mer Peptide Phage Library The New England Biolabs 12m er Peptide Ph age Library (Ph.D. System, New England Biolabs) is a combinatorial phage library that encodes a random sequence of twelve amino acids fused to the gIII gene of the M13 genome. The 12 random amino acids are fused to the Nterminus of the pIII protein, which is displaye d on the surface of the M13 phage. The first residue of the mature protein is the first random ized position. The peptide is followed by a (GlyGly-Gly-Ser) spacer linked to the wild type pI II protein. The library is constructed in M13 phage with an insertion of the lacZ gene fragment into the genome. The insertion of the lacZ gene fragment enables distinction of E. coli that harbor phages from the library opposed to environmental phages that do not contain the lacZ gene. This is done by blue/white screening of phages on agar that contains X-gal and IPTG. When an E. coli strain has a functional lacZ gene it will produce -galactosidase, which is a he terodimer composed of an and an peptide. Neither of these peptides have enzymatic activity on their own; therefore, if one component is missing then -galactosidase enzymatic activity is lost and if they are complemented they spontaneously combine to generate an active structure. Beta-galactosidase cleaves X-gal into a product that becomes oxidized to generate an insol uble blue product. IPTG is an inducer of the lac promoter which drives transcription of the lacZ gene. Therefore, if phages are produced in E. coli cells with a nonfunctional lacZ gene and a functional lacZ gene and grown under

PAGE 27

27 selection for the E. coli cells then bacteria containing a pha ge from the library, which encodes lacZ will have a complemented lacZ gene to enable distinction from bacteria containing a phage from the environment. The NEB 12me r library consists of approximately 2.7 x 109 electroporated sequences that were amplified onc e to yield approximately 55 copies of each sequence. Sequencing of 104 clones from the lib rary yielded six clones (5.8%) that did not contain a displayed peptide insert. Sequenci ng from the 98 other clones revealed a wide diversity of sequences with no obvious positional biases. Tools for Production of Phages from Phagemids To propagate phagem id particles, an amplification tool must be supplied that encodes the rest of the M13 genes necessary for phagemid part icle production. These tools may be phages or plasmids and are referred to as helpers because they help the phagemid particles propagate by supplying the necessary phage genes in trans. Some often used phage amplification tools include helper phage, hyperphage, phaberge pha ge, ex-phage, and helper plasmids. Helper Phage Helper phage is the m ost common helper tool used to produce phagemid particles. There are many variations of helper phages, R408, VCSM13, and M13KO7 ( 63) that differ slightly. Of the various helper phages M13KO7 is the most commonly used. It is a derivative of M13 that has a couple differences including a kanamyci n resistance gene and the P15A origin of replication, which allows the genome to be replicated as a plasmid in E. coli. Because helper phages are basically M13 phages, they supply all of the genes necessary for production of phagemid particles. Just like M13, helper phages infect F+ E. coli through the binding of the pIII protein to the F pilus. Help er phage amplifies phagemid particles to yield average titers of 2 x 1010-12 phagemids/mL (64). However, because helper phage encodes a wild type gIII gene, the phage particles produced will contai n a mixture of wild-type pIII proteins and

PAGE 28

28 pIII-fusion proteins. This heterogeneity in di splay of fusion proteins can result in progeny phagemids bearing all wild type pIII proteins or only a monovalent display of the recombinant pIII protein (65). The low level of display of re combinant pIII proteins results in low efficiency of selection of recombinant phage mid particles. Because the help er phages will also be produced and packaged, amplification of phagemids with helper phage generates a heterogeneous mix of phage particles encoding phagemids and helper phage genomes. The number of helper phages produced can sometimes be greater than or e qual the number of phagemid particles generated (64). To get around the problems of helper phag e, many alterations have been implemented to improve helper phage. Helper Phage with gIII Mutations There are many variations to help er phage. One of the key varia tions to helper phage is the deletion or mutation of the gIII gene in the helper phage genome, yielding hyperphage, Exphage, and Phaberge phage. All three gIII -mutated helper phages ar e derivatives of M13KO7 that lack full gIII gene functionality but sti ll possess pIII proteins. Thes e pIII proteins enable the helper phages to bi nd to the F pilus of E. coli and transfer their helper phage genomes into the E. coli. Hyperphage contains a partial deletion of the gIII gene, while Ex-phage and Phaberge phage have amber stop codons within the gIII gene. Ex-phage contains two amber stop codons, and Phaberge phage contains only one. The mutations in the gIII gene of the helper phages promote multivalent display of fusion proteins, which enhances the avidity of binding of phagemid particles to the target molecule. This increased avidity is desirable because it increases the chances of selecting positive clones. In comparing the antigen-binding activity of phagemid particles produced using M13KO7, hyperphage-produced phagemid particles have ov er 400-fold increased activity (66), Ex-phageproduced phagemid particles have over 100-fold increased activity (65), and Phaberge-phage-

PAGE 29

29 produced phagemid particles have over 5 to 20fold increased activity (67). Of the three gIII gene mutated helper phages, hyperphage has the highe st display of recombinant fusions. This is most likely due to partial read through of the amber stop codons in non-suppressor strains with Ex-phage or Phaberge phage. This partial readthrough generates display of wild-type and fusion-pIII proteins. Hyperphage does not have any readthrough of the wild-type gIII gene because of the partial deletion of the gIII gene, ensuring that all pIII proteins are recombinant fusions. Hyperphage not only ha s a higher display level of fu sion-pIII proteins, but it also packages phagemid particles over 100 times more efficiently than Ex-phage or Phaberge phage (68). Of the gIII gene-mutated helper phages, hyperphage is the most advantageous for use in phage display systems. Hyperphage enables multivalent display of pIII fusions, which makes it ideal in phagemid particle production. However, an issue with hyperphage is that the phagemid particle stocks generated usi ng hyperphage have titers of 109-10 phagemids/mL (64). This is a log or two lower than phagemid particles produced using helper phage. However, many of the phagemid particles generated by helper phage are useless because they bear no recombinant pIII proteins. Therefore, even though the quantity of phagemid particles prod uced with hyperphage is lower than those produced with helper phage the quality of phagemid particles produced is much higher. Helper Plasmids Helper plasm ids (64) are M13-based plasmids used for phagemid particle amplification that are engineered in three forms to overc ome many disadvantages of helper phage and hyperphage. The plasmids are M13mp19 with a chloramphenicol resistance gene cloned in from pBSL121 to allow for selection of E. coli containing the helper plasmids. The M13 origin of replication was deleted and replaced with the p15a origin of replication from pMPM-K3. This

PAGE 30

30 eliminates the ability of helper plasmids to be packaged in progeny phage particles, resulting in progeny phages that contain the phagemid s but not the helper plasmids. The gIII gene of the helper plasmid was partially delete d or fully deleted to yield thr ee helper plasmids with varying lengths of their gIII gene These three helper plasmids are maintained in E. coli and provide phagemid particles with all of the necessary structural proteins for phagemid particle amplification. One of the advantages of the helper plasmids is that they have a full (M13cp), deleted (M13cp-dg3), or truncated (M13cp-CT) gIII gene, which enables monovalent to multivalent display of recombinant proteins. Multivalent display phagemid particles possess high avidity binding ability, while monovalent display phagemid partic les possess high affinity binding ability. Thus, if high affinity monovalent display phagemid pa rticles are desired, M13cp would be used. If high avidity multivalent display phagemid partic les are desired, M13cp-dg3 or M13cp-CT might be used. An additional advantage is that the helper plasmids are maintained in E. coli. This negates the need for superinfection, removing the limiting factor of the number of helper phages or hyperphages needed to amplify phagemid particles. This thesis describes the efforts to optimi ze biopanning processes and associated reagents to improve the likelihood of is olating recombinant phages that specifically bind to biological agents or their products. The recombinant phages that specifically bind to biological agents or their products will eventually be used to enable detection of the agents. The specific aims for this study are: 1. To optimize extraction methods a nd determine saturation limits of V. cholerae O1 LPS and use these methods to pa n phage display libraries to V. cholerae LPS. A phenol-water method was a more effectiv e extraction method than TRIzol Reagent method for the extraction of V. cholerae O1 LPS and 10-100 times more LPS could be bound to nitrocellulose paper as opposed to a microtiter well. Biopannings against

PAGE 31

31 V. cholerae O1 LPS that was immobilized onto nitrocellulose paper or a microtiter well failed to yield phages that recognized V. cholerae O1 LPS. 2. To optimize panning and screening proce dures to allow for a more efficient biopanning process that will be more likely to isolate and detect specific recombinant phagemid particles. In biopanning with the Tomlinson scFv phage mid library, a trypsin elution time of 10-30 minutes eluted the most phagemid particles. Screening scFv protei ns by a high throughput ELISA was an acceptable screening process and enabled the possible screening of hundreds of clones per day. Screening phagemid particles by a high throughput ELISA was not an acceptable screening process because phagemid particles could not be produced in high enough titers in a mi crotiter well to make a high throughput ELISA screen effective. 3. To optimize the production of phagemid par ticles to ensure high quality and high quantity of phagemid particles. Phagemid particle amplification by E. coli harboring hyperphage or helper plasmids was not an effective method to produce phagemid particles. Homemade hyperphage that was produced in E. coli MG1655 (pGTR203) enabled high quality and high quantity production of hyperphage. With homemade hyperphage, the Tomlinson J library was effectively amplified to ensure that the re dundancy of the clone population in the library was maintained. 4. To isolate specific recombinant phagemid particles to E. coli O157:H7 Stx2 toxin using the optimization techniques discovered in previous aims. Using the improved homemade hyperphage and the homemade hyperphage-amplified Tomlinson J scFv library to pan against a commercial Stx2 toxin preparation resulted in isolation of phagemid particles that recognized the Stx2 toxin preparation.

PAGE 32

32 Figure 1-1. Structure of M13 phage. The M13 phage particle c ontains a ssDNA, circular, 6.4-kb genome that encodes genes I-XI and the M13 ori The M13 phage has five coat proteins: the major coat protein (pXIII) that is present in 2,700 copies and the minor coat proteins (pIII, pVI, pVII, and pI X) that are present in 5 copies each. pIII pVI pVIII pVII pIX M13 oriIV XI I VI X V VII IX XIII III II ~6.4-kb ssDNApIII pVI pVIII pVII pIX M13 oriIV XI I VI X V VII IX XIII III II ~6.4-kb ssDNA

PAGE 33

33 Figure 1-2. Structures of an tibodies. mAb (monoclonal an tibody), Fab (fragment antigen binding), scFv (single-chai n variable fragment), VH (variable heavy chain), VL (variable light chain), CH1-3 (constant heavy chain domains 1-3), and CL (constant light chain). The heavy and light chains of mAb and Fab antibodies are held together by disulfide bonds, and the heavy and light chains of the scFv antibody are fused together by a flexible glycine-serine linker. mAb FabscFvVHVHVHVLVLVLLinker CLCLCH1CH1CH3CH2

PAGE 34

34 Figure 1-3. Genetic map of pIT2 phagemid vect or from the Tomlinson scFv library. RBSribosome binding site. pelB leader peptide sequence promotes export of the scFv protein. Variable Heavy and Variable Light peptide seque nces are fused together by a glycine-serine linker. An amber stop codon is at the junction of the c-myc tag and the gIII gene to enable conditional expression of the scFv-pIII fusion in an amber suppressor strain. The M13 or igin of replication enable s packaging into M13 phage particles, the bla gene encodes ampicill in resistance, and th e colE1 origin of replication enables maintenance as a plasmid in E. coli. M13 ori pelB leader lac promoter colE1 ori

PAGE 35

35 CHAPTER 2 MATERIALS AND METHODS Bacterial Strains, Media, and Growth Methods The bacte rial strains used and their genotypes are listed in Table 2-1. All E. coli strains were grown in LB broth (10 g tryptone, 5 g yeast extract, 5 g NaCl, and 3 mL of 1 M NaOH in 1 L water) or on LB agar plates contai ning 1.5% (w/v) agar. Exceptions are E. coli TG1, E. coli DH5 (pNR100), and E. coli ER2738. Escherichia coli TG1 was grown in 2xTY broth (16 g tryptone, 10 g yeast extract, 5 g NaCl, and 3 mL of 1 M NaOH in 1 L water) or on 2xTY agar plates containing 1.5% (w/v) agar. Escherichia coli DH5 (pNR100) was grown in LB broth containing 100 g/mL ampicillin or on LB agar plates containing 100 g/mL ampicillin and 1.5% (w/v) agar. Escherichia coli ER2738 was grown in LB broth containing 20 g/mL tetracycline or on LB ag ar plates containing 20 g/mL tetracycline and 1 .5% (w/v) agar. All strains of V. cholerae were grown in Luria Bertani broth containing with physiological saline (LB-N) (10 g tryptone, 5 g yeast extract, 8.5 g Na Cl, and 3 mL of 1 M NaOH in 1 L water) or on LB-N agar plates containing 1.5% (w/v) agar The most common medium used was 2xTY containing 100 g/mL ampicillin and 1% (w/v) glucose; therefore, it was abbreviated as 2xTY AG. Escherichia coli strains harboring hyperphage were grown in broth or on plates as specified, with the addition of 40 g/mL kanamycin. Escherichia coli strains harboring helper plasmids were grown in brot h or on plates as specified above, with the addition of 30 g/mL chloramphenicol. All bacterial cultures were initially grown as a standing-overnight culture. A standingovernight culture was made from 10 mL of specifi ed medium that was inoculated with bacteria from an agar plate. The standing-overnight culture was grown overnight (~16 h) in a 37C incubator. A log-phase culture was obtained by diluting a sta nding-overnight culture 1:40 in

PAGE 36

36 medium and incubating the culture in a 37C incu bator with shaking until the optical density at 600 nm (OD600) was between 0.4 to 0.6. The Tomlinson J Human Synthetic VH + VL phagemid library was constructed by Medical Research Council, Cambridge, U.K. (Human Single Fold scFv Libraries I + J (Tomlinson I + J). 2002. Cambridge, UK, MRC La boratory of Molecula r Biology, MRC Centre for Protein Engineering.) and was obtained from the Interdisciplinary Center for Biotechnology Research Hybridoma Core, University of Flor ida. The Ph. D. 12mer peptide library was obtained from New England Biolabs (NEB, Ipswich, MA). The hyperphage (66) used to amplify phagemid particles was obtained from Progen Biotechnik (Heidelberg, Germany). Biopanning of Phage Display Libraries Panning on Immunotubes A polystyrene imm unotube (Nunc, Rochester, NY) was coated with 2 mL of 50 g/mL Shiga-like toxin 2 (Stx2 (Toxin Technology, Sara sota, FL)) in phosphate-buffered saline (PBS) (Cellgro, Manassas, VA) (137 mM NaCl, 2.7 mM KCl, 4.3 mM Na2HPO4, 1.4 mM KH2PO4, pH 7.3). The immunotube was rotated overnight on a la bquake at 4C. The next day the tube was washed three times with 4 mL of PBS to remove non-immobilized toxin. The tube was filled with casein blocker (Pierce, Rockford, IL) containing 0.05% (v /v) Tween-20 and incubated for two hours at room temperature (~25C). Excess blocker was removed by rinsing the tube three times with 4 mL of PBS. The Tomlinson J library (for round one of panning) or amplified eluted phages (for subsequent rounds of panning) was added to the immunot ube at a concentration of 1 x 1012 phages in 4 mL of casein blocker (Pierce) containing 0.05% (v/v) Tween-20. The library was incubated for one hour at room temperat ure on a labquake, followed by a one hour standing incubation. The unbound library was removed by aspiration, and the weakly bound phages were removed by washing the tube 10 times for round one and 20 times for rounds two and three with

PAGE 37

37 4 mL of PBS containing 0.05% (v/v) Tween-20 (PBS-0.05T). The excess PBS-0.05T was aspirated, and bound phages were eluted with 0.5 mL of trypsin-PBS (10% (v/v) trypsin stock (10 mg/mL trypsin (Type XIII from Bovine Panc reas) (Sigma-Aldrich, St. Louis, MO), 50 mM Tris-HCl (pH 7.4), 1 mM CaCl2 in water) in PBS) for 10 minut es at room temperature on a labquake. Ten microliters of the eluted phages were immediately titered, while the rest of the eluted phages were infected into 5 mL of E. coli TG1 at an OD600 of 0.4. The infected E. coli was incubated for 30 minutes at 37C in a stan ding water bath. Following incubation, the infected cells were isolated by centrifugation at 13,776 x g for 10 minutes at room temperature. Cells were suspended in 0.6 mL of 2xTY medi um and plated on three 2xTY AG plates to amplify the phagemid-containing E. coli. The plates were incubate d overnight at 37C. The next day 1.5 mL of 2xTY was added to the lawn s of bacteria, and the bacterial lawns were scraped from the plate in the medium and pool ed together. Of the pooled bacteria, 50 L was inoculated into 100 mL of 2xT Y AG medium and grown in a 37C shaking incubator until the OD600 was 0.4. Once the OD was reached, 10 mL of the culture was superinfected with homemade hyperphage at a MOI of 20 and incuba ted for 30 minutes in a standing 37C water bath. The superinfection was cen trifuged at 13,776 x g for 10 minutes at room temperature. The resulting pellet was suspended in 100 mL of 2xTY containing 100 g/mL ampicillin, 40 g/mL kanamycin, and 0.1% (w/v) glucose and incubated overnight in a shaking 30C water bath. The next day the culture was centr ifuged at 13,776 x g for 10 minutes at 4C. Phage-containing supernatant was precipitated with final concentra tions of 13 mM Polyethylene Glycol (PEG) and 0.55 M NaCl. PEG and NaCl were dissolved in th e supernatant at room temperature and then the PEG-containing supernatant was incubated overnight at 4C. The next day the PEGprecipitated supernatant was cen trifuged at 13,776 x g for 20 minutes at 4C. The pellet was

PAGE 38

38 suspended in 2 mL of PBS and centrifuged at 13 ,776 x g for 5 minutes at room temperature to remove bacterial debris. The supe rnatant was transferred to a new tube to be titered and stored at 4C until used in further rounds of panning or screening. Panning in Suspension Vibrio cholerae N16961 was grown to log phase, and 107 to 109 cells were centrifuged at 10,621 x g for 5 minutes at room temperature. Th e bacterial pellet was su spended in 1 mL of PBS and transferred to a microcentrifuge tube The Tomlinson J library (for round one of panning) or amplified eluted phages (for subs equent rounds of panning) was added to the bacterial suspension at a concentration of 1 x 1011 phages in 0.5 mL of PB S. The library and bacterial cells were incubated for 3 hours at 4C while rota ting on a labquake. The cell suspension was centrifuged at 10,621 x g for 5 minutes at room temperature. The supernatant, containing the unbound phage library, was stored at 4C. The bacteria l pellet containing the bound phages was suspended in 0.5 mL of PBS to wa sh off phages that were weakly bound to the bacteria. The suspension was centrifuged at 10,621 x g for 5 minutes at room temperature. The supernatant was discarded, and the pellet was suspended in 0.5 mL of PBS for a second wash. The suspended cells were transferred to a new microcentrifuge tu be and centrifuged at 10,621 x g for 5 minutes at room temperature. The supernatant was aspirated, and phages bound to the bacteria were eluted with 0.5 mL of tryps in-PBS for 10 to 60 minutes at room temperature on a labquake. Ten microliters of the elution was immediately titered. The rest of the eluted phages were infected into 5 mL of E. coli TG1 at an OD600 of 0.4. The infected E. coli were incubated for 30 minutes in a 37C standing water bath. Following incubation, the infected cells were isolated by centrifugation at 13,776 x g for 10 mi nutes at room temperature. The cells were suspended in 0.6 mL of 2xTY medium and plat ed on three 2xTY AG plates to amplify the phagemid-containing bacteria. Th e plates were incubated overni ght at 37C. The next day 1.5

PAGE 39

39 mL of 2xTY was added to the lawns of bacteria, and the bacterial lawns were scraped from the plate in the medium and pooled toge ther. Of the pooled bacteria, 50 L were inoculated into 50 mL of 2xTY AG medium and grown in a 37C shaking incubator till the OD600 was 0.4. Once the OD was reached, 10 mL of the culture was superinfected with homemade hyperphage at a MOI of 20 and incubated for 30 minutes in a 37 C standing water bath. The superinfection was centrifuged at 13,776 x g for 10 minutes at room temperature. The resulting pellet was suspended in 50 mL of 2xTY containing 100 g/mL ampicillin, 40 g/mL kanamycin, and 0.1% (w/v) glucose and incubated overnight in a 30C shaking water bath. The next day the culture was centrifuged at 13,776 x g for 10 minutes at 4C. The phage-containing supernatant was precipitated by slowly adding 12.5 mL of PEG solution (20% (w/v) PEG and 2.5 M NaCl in water) to the supernatant with continuous swirling. The PEG-containing supernatant was incubated overnight at 4C. The next day th e PEG-containing supernatant was centrifuged at 13,776 x g for 20 minutes at 4C. The pellet was suspended in 1 mL of PBS and centrifuged at 10,261 x g for 5 minutes at room temperature to re move bacterial debris. The supernatant was transferred to a new tube to be titered and stored at 4C until used in further rounds of panning or screening. Panning on Microtiter Wells A polysorp m icrotiter well (Nunc) was coated with 150 L of 100 g/mL V. cholerae 569 lipopolysaccharide (LPS) (Sigma-Aldri ch) in carbonate coating buffer ( 0.1 M NaHCO3, 0.02% (w/v) NaN3 (pH 8.6) ). Prior to coating, the LPS was sonicated by a Vibra Cell (Sonics & Materials Inc., Danbury, CT) for one minute. The microtiter well was coated overnight at 4C in a humid chamber. The next day the microtiter we ll was equilibrated to room temperature for 15 minutes. The LPS was aspirated and blocked with 200 L of carbonate blocking buffer for one hour at 4C. The well was washed six times with 200 L of 0.1 M Tris-buffered saline (TBS)

PAGE 40

40 (81 mM Tris-HCl, 20 mM Tris-Base, 154 mM Na Cl, pH 7.5) containi ng 0.1% (v/v) Tween-20 (TBS-0.1T) for the first round of panning and with 200 L of 0.1 M TBS cont aining 0.5% (v/v) Tween-20 (TBS-0.05T) for subsequent rounds of panning. The washed wells were incubated with 2 x 1011 PFU of the Ph.D.-12mer peptide librar y (NEB) (for round one of panning) or amplified phages (for subsequent rounds of panning) diluted in 100 L of TBS-0.1T for one hour at room temperature with gentle rocking. The unbound phages were aspirated and stored at 4C. The wells were washed 10 times with TBS-0.1T for the first round of panning and with 200 L of TBS-0.05T for subsequent rounds of pa nning. The phages were eluted with 100 L of 0.1 M glycine (pH 2.2) for 10 minutes at room temper ature with gentle agitation. The glycine containing the eluted phages was transferred to a microcentrifuge tube and neutralized with 15 L of Tris-HCl (pH 9.0). Five microliters of the eluted phages were immediately titered. One hundred and ten microliters of the remaining el ution was added to 20 mL of LB diluted 1:100 with a standing-overnight culture of E. coli ER2738. The culture was grown for 4.5 hours in a 37C shaking incubator. The turbid culture wa s centrifuged at 13,776 x g for 10 minutes at 4C. The phages were precipitated by slowly adding 3.3 mL of PEG so lution (20% (w/v) PEG and 2.5 M NaCl in water) to the phage-containing su pernatant with continuous swirling. The PEGcontaining supernatant was incubated overnight at 4C. The next day the PEG-containing supernatant was centrifuged at 13,776 x g for 20 minut es at 4C. The pellet was suspended in 1 mL of 0.1 M TBS and centrifuged at 10,621 x g for 5 minutes at room temperature to remove bacterial debris. The phages were preci pitated a second time by slowly adding 167 L of PEG solution to the supernatant with continuous swirling. The PEG-containing supernatant was incubated one hour at 4C and centrifuged at 13, 776 x g for 15 minutes at 4C. The pellet was suspended in 200 L of 0.1 M TBS to be titered and stored at 4C until used in further rounds of

PAGE 41

41 panning or screening. Five rounds of panning were done with screening of clones after the fifth round of panning. Panning on Nitrocellulose Paper Six pieces o f nitrocellulose paper (Bio-Rad) were cut to the surface-area dimensions of a microtiter well (5 mm x 20 mm). One a nd a half milliliters of approximately 340 g/mL V. cholerae 569 LPS (phenol-water-extracted ) in PBS was added to si x microcentrifuge tubes, each containing a strip of nitrocellulose paper. Prior to coating, the LPS was sonicated by a Vibra Cell (Sonics & Materials Inc.) for one minut e. The nitrocellulose papers were coated overnight at 4C on a rotating labquake. The next day two pieces of LPScoated nitrocellulose paper were each transferred to a new microcentrifuge tube and washed five times with 1 mL of PBS on a labquake with five minutes per wash. Wash ed nitrocellulose strips were transferred to new microcentrifuge tubes and bl ocked for one hour in PBS casein blocking blocker (Pierce) on a labquake at room temperature. Blocked nitrocellulose stri ps were transferred to new microcentrifuge tubes and washed five times with 1 mL of PBS on a labquake with five minutes per wash. Each of the washed pieces of nitrocellulose papers was transferred to a new microcentrifuge tube containing 1.5 x 1011 PFU of the Ph.D.-12mer peptide library (NEB) diluted in 1 mL of PBS. Nitroc ellulose strips were incubated with the library for one hour at room temperature on a labquake. After one hour, the panned nitrocellulose strips were transferred to new microcentrifuge tubes and washed five times with 1 mL of PBS on a labquake with five minutes per wash to remove the unbound library. Washed nitrocellulose strips were transferred to new microcentrifuge tubes to be eluted. One pi ece of nitrocellulose was acid eluted by adding 250 L of 0.1 M glycine (pH 2.2) to the stri p and incubating it for 10 minutes at room temperature on a labquake. The acid eluted nitrocellulose paper was removed to a new microcentrifuge tube, and the glyc ine solution containing the eluted phages was neutralized with

PAGE 42

42 37 L of 1 M Tris-HCl (pH 9.0). The acid eluted phages were stored at 4C until used for amplification. The other piece of panned nitroc ellulose paper was antigen eluted by adding 250 L of approximately 340 g/mL of V. cholerae 569B LPS diluted in PBS to the piece of nitrocellulose paper and incubated one hour at room temperature on a labquake. The piece of nitrocellulose paper was removed, and the rema ining LPS solution containing the eluted phages was acid eluted as described above. Five microl iters of each eluted phage solution was titered, and the rest of the two elutions were each adde d to 20 mL of LB diluted 1:100 with a standingovernight culture of E. coli ER2738. The cultures were grown for 4.5 hours in a 37C shaking incubator. The turbid cultures were centrifuged at 13,776 x g for 10 minutes at 4C. Five microliters of the amplified phages were titered. The rest of the phages were PEG precipitated by slowly adding 3.3 mL of PEG solution to the phage-containing supern atant with continuous swirling. The PEG-containing supern atant was incubated overnight at 4C. The next day the PEG-containing supernatant was ce ntrifuged at 13,776 x g for 20 minutes at 4C. The pellet was suspended in 1 mL of PBS and centrifuged at 13 ,776 x g for 5 minutes at room temperature to remove bacterial debris. The supernatants were transferred to new microc entrifuge tubes, titered, and stored at 4C. After the first round of panning, the amplifie d eluted phages were negatively panned on. Two pieces of nitrocellulose paper (5 mm x 20 mm) were blocked for one hour in 1 mL of PBS casein blocking buffer (Pierce) on a labquake at room temperature. The blocked nitrocellulose strips were washed three times with 1 mL of PBS on a labquake at room temperature with five minutes per wash. One piece of blocked nitroc ellulose paper was added to the acid-elutedamplified phages from the round one panning an d incubated one hour on a labquake at room temperature. The other strip of blocked nitrocellulose paper was added to the LPS and acid-

PAGE 43

43 eluted-amplified phages from the round one panni ng and incubated one hour on a labquake at room temperature. The negatively panned pieces of nitrocellulose paper were removed, and the remaining amplified-eluted phages were used for two more rounds of panning. After the third round of panning the eluted phages were screen by ELISA. Spot Titer of Phages Escherich ia coli TG1 was grown to log phase an d centrifuged at 13,776 x g for 10 minutes at 4C. The resulting pellet was suspended in PBS to yield an E. coli TG1 concentration of 1 x 1010 CFU/mL. One hundred microlit ers of the concentrated E. coli TG1 was spread on a 2xTY AG plate and set to dry for 30 seconds. Serially diluted phages in phosphate-buffered saline containing 0.1% (w/v) gelatin (BSG) was dropped onto the plate in 10 L drops. Once the drops dried, the plate was incubate d overnight at 37C. The next day the colonies were counted, and the approximate titer was calculated. Spread Titer of Phages Phages were serially diluted in BSG. One hundred m icroliters of the diluted phages were added to 0.9 mL of log phase E. coli TG1. The infected E. coli TG1 was incubated 20 minutes in a 37C standing water bath. One hundred microlit ers of the infections were plated on 2xTY AG plates. The plates were incubated overnight at 37C. The next day the colonies were counted, and the approximate titer was calculated. Amplification of Phages A colony of E. coli TG1 containing a phagem id was picked from a plate with a sterile toothpick and swirled in 3 mL of 2xTY AG medium. The culture was grown overnight at 37C. The overnight culture was diluted 1:20 in 3 mL of 2xTY AG medium and grown to log phase in a 37C shaking incubator. Three hundred microlit ers of the log-phase cult ure was infected with hyperphage at a MOI of 10 and incubated in a 37 C standing water bath for 30 minutes. The

PAGE 44

44 culture was added to 30 mL of 2xTY medium containing 100 g/mL ampicillin, 40 g/mL kanamycin, and 0.1% (w/v) glucose. The cultu re was grown overnight in a 30C shaking incubator. The turbid culture was centrifuge d at 13,776 x g for 10 minutes at 4C. The phagecontaining supernatant was tit ered and stored at 4C. High Throughput Production of Soluble An tibody Fragments (scFv antibodies) A colony of E. coli HB2151 containing a phagemid was picked from a plate with a sterile toothpick and swirled in 200 L of 2xTY containing 100 g/mL ampicillin, 0.1% (w/v) glucose, and 1 mM Isopropyl -D-1-thiogalactopyranoside (IPTG) in a 96-well polystyr ene microtiter plate (Corning, Corning, NY). The plate was incubated overnight at 37C. The next day the plate was centrifuged at 4,667 x g for 10 minutes at 20C. Supernatants were analyzed by ELISA. Deoxyribonucleic Acid Manipulations Plasmid Extractions Plasm id extractions for cultures of 3 mL were performed with the QIAprep Spin Miniprep kit (Qiagen, Germantown, MD), while cultures of 100 mL or greater were performed with the Plasmid Midi kit (Qiagen). Extrac tion procedures were performed as directed in the instruction manual. Agarose Gel Electrophoresis Deoxyribonucleic acid sam ples from 1-5 L were added to 2 L of 10 x Gel Loading Buffer (Invitrogen, Carlsbad, CA). The samples we re loaded onto a 0.7% (w /v) agarose gel using Tris-borate-EDTA buffer (89 mM Tris, 89 mM boric acid, 2 mM EDTA) containing 10 g/mL ethidium bromide. The gel was electrophoresed at 100 V until the loadin g dye was two-thirds down the gel. The DNA bands were analyzed on a Gel Doc XR (Bio-Rad, Hercules, CA).

PAGE 45

45 Electroporation of Plasmids Electroporation was perform ed to transform plasmids into E. coli. DNA (30-1000 ng/ L) was added to 50 L of electrocompetent cells (see below) in a microcentrifuge tube. The tube was flicked to mix the DNA with the cells. Once the DNA and cells were mixed, they were transferred to a chilled 0.1 cm electroporation cuvette and elec troporated at 1.25 kV/cm with a MicroPulser (Bio-Rad). Nine hundred and fift y microliters of 2xTY medium was immediately added to the transformation and tr ansferred to a small culture tube where it was incubated in a 37C static water bath for 1 hour. The electropora tion was serially diluted in BSG and plated on appropriate plates. The pl ates were incubated overnight at 37C The next day the colonies were counted, and the number of transformations was calculated. Electrocompetent Cells Electrocom petent cells were made by diluting a standing overnight culture 1:100 in 1 L of medium containing 1% (w/v) glucose and appropriat e antibiotics. The culture was incubated in a 37C shaking incubator until the OD600 was between 0.4 and 0.6. All centrifuge rotors, wash buffers, and centrifuge tubes that were used were pre-chilled to 4C. Once the OD600 was between 0.4 and 0.6, the culture was chilled on ice in a 4C cold room for 45 minutes or more until the culture was ~4C. The culture was centr ifuged at 13,776 x g for 10 minutes at 4C, and the supernatant was decanted. The pelle ts were suspended in 1.2 L chilled ddH2O by vortexing or pipeting until the cell suspension was homogeneous. The suspended pellets were centrifuged to pellet the bacterial cells. The supernatant was decanted, and the pellets were suspended in 400 mL of chilled ddH2O. The bacteria were pelleted by centrifugation, and the supernatant was decanted. The pellets were suspended in 200 mL of chilled ddH2O into one centrifuge tube. The bacteria were pelleted by centrifugation, and the supernatant was d ecanted. The pellet was suspended in 25 mL of chilled ddH2O and transferred to a 35 mL ce ntrifuge tube. The bacteria

PAGE 46

46 were pelleted by centrifugation, and the supernatan t was decanted. The pellet was suspended in 25 mL of chilled 10% (w/v) gluc ose in water and centrifuged to pellet the bacteria. The supernatant was decanted, and the bacteria were suspended in 0.5-1 mL of chilled 10% (w/v) glucose in water. The suspended bacteria were either used directly for transformation or aliquoted and frozen on dry ice in 95 % (v/v) EtOH and stored at -80C. Enzyme-Linked ImmunoSorbent Assays (ELISAs) Enzym e-linked immunosorbent assays we re performed for the screening and characterization of phage partic les, scFvs, LPS, bacterial pr otein preparations, monoclonal antibodies, and polyclonal antibodies. The co ating antigen or antibody was diluted in PBS (Cellgro) or carbonate/bicarbonate buffer (15 mM Na2CO3 and 35 mM NaHCO3 in water), and 100 L was loaded onto a 96-well polystyrene flatbottomed microtiter plate (Becton-Dickinson, Franklin Lakes, NJ). Concentr ations used of whole cells (108 tu/mL), LPS (1 g/mL), toxin (10 g/mL), and proteins (10 g/mL) were kept relatively consiste nt. The coated microtiter plate was incubated overnight in a humid chamber at 4C. The next day the microtiter plate was equilibrated to room temp erature for 30 minutes. Wells were aspirated with an ELX 800 Strip Washer (Bio-Tek, Winooski, VT) and washed once with 300 L of PBS. If the coating antigen was whole cells, LPS, or toxin, then the coati ng antigen was removed by vacuum or pipet and sterilized. Two hundred microliters of PBS casein blocking buffer (Pierce) containing 0.05% (v/v) Tween-20 was added to each well and incubate d two hours at room temperature. Blocker was aspirated by the ELX 800 Strip Washer (Bio-Tek). On e hundred microliters of primary antibody diluted in PBS casein blocking buffer (Pierce) containing 0.05% (v/v) Tween-20 was added to each well and incubated one hour at ro om temperature. Concentrations used of phagemid particles were 107 to 109 tu/mL and antibodies were 1 to 10 g/mL. The primary antibody was aspirated and washed three times with 300 L of PBS-0.05T by the ELX 800 Strip

PAGE 47

47 Washer (Bio-Tek). One hundred microliter s horseradish peroxidase (HRP) conjugated secondary antibodies were diluted in PBS casein blocking buffer (Pierce) containing 0.05% (v/v) Tween-20 and added to each well an d incubated 30 minutes at room temperature. The secondary antibody was aspirated and washed three times with 300 L of PBS-0.05T by the ELX 800 Strip Washer (Bio-Tek). Substrate was prepared by dissolving one capsule of phosphate-citrate (PC) buffer (0.05 M phosphate-citrate buffer (pH 5.0) 0.03% (w/v) sodium perborate) (SigmaAldrich) in 100 mL of water. A ten milligram tablet of 3, 3, 5, 5-tetramethylbenzidine substrate (Sigma-Aldrich) was added to 10 mL of th e PC buffer to give a final concentration of 1 mg/mL. Two hundred microliters of substrate was added to each well and incubated for 30 minutes at room temperature. The plate was re ad in an ELx 800 UV plate reader (Bio-Tek) at 630 nm. The data were analyzed with KcJunior (Bio-Tek) software. Infection Efficiency Infection efficiency experim ents were done to compare the infection efficiency of phagemid particles to bacterial strains harboring helper plasmids to the same strains not harboring helper plasmids. Escherichia coli standing-overnight cultures were diluted 1:40 in 3 mL of medium containing 1% (w/v) glucose for strains containing no helper plasmids and in medium containing 1% (w/v) glucose and 30 g/mL chloramphenicol for strains containing helper plasmids. The diluted cultures were grown in a 37C shaking incubator till the OD600 was between 0.4 and 0.6. Once the desired OD600 was obtained, 3 x 107 bacteria were infected with phage at a MOI of 0.1 and incubated for 20 mi nutes in a 37C standi ng water bath. The transductions were titered using th e spread-titer method. Transduced E. coli strains containing helper plasmids were titered on 2xTY AG plates containing 30 g/mL chloramphenicol. Transduced E. coli strains containing no helper plasmi ds were titered on 2xTY AG plates.

PAGE 48

48 Protein and Lipopolysaccharide (LPS) Manipulations Phenol-Water Extraction of LPS A phenol-water extraction (76) was perform ed to extract V. cholerae 569B LPS from whole cells. One and a half liters of V. cholerae 569B was grown to log phase in a 37C shaking incubator. Bacteria were pell eted at 13,776 x g for 10 minutes at 4C and suspended in 10 mL of 1% (w/v) NaCl in water. The bacteria were ce ntrifuged and suspended in another 10 mL of 1% (w/v) NaCl in water. Next, the bacteria were centrifuged and susp ended in 10 mL of ddH2O. This suspension was warmed to 70C, along with 24 mL of 90% (w/v) ph enol (made directly before extraction). Ten milliliters of warmed 90% (w/v) phenol was added to the 10 mL cell suspension and vortexed. The twenty milliliter suspension was warmed in a 70C water bath for 20 minutes while vortexing frequently. The phenol suspension was transferred to an ice bath where it was swirled in ice water for 5 minutes The phenol suspension was centrifuged at 2,284 x g for 25 minutes at 4C. The aqueous phase of the phenol extraction was transferred to a new centrifuge tube and stored on ice. The same vol ume of water as that extracted was added back into the phenol suspension tube and heated to 70C for 20 minutes with frequent vortexing. The extraction step was repeated, and the two aqueous phase extractions were combined into the same tube. Aqueous extractions were placed in a 70C water bath for 15 minutes. Twelve milliliters of warmed 90% (w/v) phenol was added to the warmed aqueous extraction and incubated at 70C for 10-15 minut es with frequent vortexing. Two back extractions were performed on the aqueous extractions. The final extraction was dialyzed (MWCO: 3,500) for 24-48 hours at 4C. The dialyzed extraction was digested with DNase (Qiagen) (20 g DNase/mL dialysis volume) and RN ase (Qiagen) (40 g RNase/mL dialysis volume) in MgCl (1 L 20% (w/v) MgCl/mL dialysis volume) overni ght at 37C. The di gested extraction was

PAGE 49

49 dialyzed (MWCO: 3,500) for 24-48 h ours at 4C. The final extrac tion was used in a Tsai-Frasch silver stain to analyze the LPS. TRIzol Reagent Extraction of LPS TRIzol Rea gent (Invitrogen), a monophasic solution of phenol and guanidine isothiocyanate (77), was used to extract V. cholerae N16961 LPS from whole cells. Ten milliliters of log phase (OD600 ~ 0.4-0.6) V. cholerae N16961 were centrifuged at 13,776 x g for 10 minutes at 4C. The supernatant was decante d, and the cells were suspended in 200 L of TRIzol Reagent and incubated at room temperat ure for 10 to15 minutes. Twenty microliters of chloroform per mg of cells was added to th e mixture, vortexed, and incubated at room temperature for 10 minutes. The mixture was ce ntrifuged at 12,000 x g for 10 minutes at room temperature. The resulting aqueous phase was transferred to a new mi crocentrifuge tube and stored on ice. One hundred microliters of ddH2O was added to the TRIzol-Reagent tube, and a back extraction was performed. Two more back extractions were performed following the first back extraction to give a total of four extractio ns. The extractions were dried with a Savant SpeedVac Concentrator (Globa l Medical Instrumentation, Ramsey, MN) for 2.5 hours. The pellet was suspended in 0.5 mL of 0.375 M MgCl in 95% (v/v) EtOH that was chilled to 4C. The suspension was centrifuged at 12,000 g for 15 mi nutes at room temperature. The resulting pellet was suspended in 200 L of ddH2O. The final LPS extracti on was analyzed in a TsaiFrasch silver stain (see below). Extraction of Periplasmic Proteins Periplasmic proteins from E. coli were extracted using a Tris-EDTA-Sucrose (TES) extraction. A standing-overnight culture of an E. coli culture was diluted 1: 400 in 100 mL of LB medium and incubated overnight in a 37C shaki ng incubator. The next day the cells were centrifuged at 13,776 x g for 10 minutes at 4C. Th e supernatant was stored at 4C. The pellet

PAGE 50

50 was suspended in 25 mL of PBS by vortexing. The suspension was centrifuged at 13,776 x g for 10 minutes at 4C. The supernatant was discar ded, and the pellet was suspended in 10 mL of TES (20% (w/v) D-sucrose, 30 mM Tris-HCl, 1 mM EDTA) in water. Th e suspended pellet was incubated for 15 minutes at room temperature with occasional swirling. The suspension was centrifuged at 13,776 x g for 10 minutes at 4C. The supernatant was discarded, and the pellet was gently washed with 10 mL of 0.5 mM MgCl2 without dislodging the pe llet. The 10 mL of the 0.5 M MgCl2 wash was discarded, and the pellet wa s suspended by vortexing in 10 mL of 0.5 mM MgCl2. The suspension was incubated in an ice bath for 15 minutes with occasional swirling. The suspension was centrifuged at 21,525 x g for 30 minutes at 4C. The supernatant was collected and concentrated with 10,000 MWCO Amicon (Millipore, B illerica, MA) filters and stored at 4C. The pellet was suspende d in 10 mL of PBS and stored at 4C. Determination of Protein Concentration Protein concentration was determ ined by DC-Protein Assay (Bio-Rad). Protein samples were diluted in PBS. Five microliters of each sample was added to a microtiter well followed by 25 L of Reagent A and 200 L of Reagent B. All wells were performed in triplicate. Protein standards used were Bovine Serum Albumin (BSA ) in concentrations from 0-1.4 mg/mL. The reaction took place for 15 minutes at room temper ature and was read at 750 nm with an ELx 800 UV plate reader (Bio-Tek). Protein concentratio ns were analyzed using KC Junior (Bio-Tek) software. Sodium Dodecyl Sulfate-Polyacrylami de Gel Electropho resis (SDS-PAGE) Sodium dodecyl sulfate-polyacrylamide gel electrophoresis was performed to resolve V. cholerae LPS and protein preparations ReadyGel Tris-glycine 12% (w/v), 15% (w/v), and 420% (w/v) PAGE gels (Bio-Rad) were used with the MiniProtean-Electr ophoresis system (BioRad) for SDS-PAGE analysis of samples. The LPS samples were sonicated by a Vibra Cell

PAGE 51

51 (Sonics & Materials Inc.) for 2 minutes and d iluted in Laemmli sample buffer (Bio-Rad). Protein samples were vortexed and diluted in La emmli sample buffer. After dilution in Laemmli sample buffer, the samples were boiled for 10 minut es and loaded in the gels. The samples were electrophoresed for 1 hour at 100 Volts in elec trode buffer (25 mM Tris, 0.17 M glycine, and 0.1% (w/v) SDS). Coomassie Blue Staining Proteins resolved by SDS-PAGE we re stained with 0.05% (w/v) Coomassie Blue R, 50% (v/v) methanol, and 10% (v/v) glaci al acetic acid in water and ag itated gently for 10 minutes. The gel was rinsed with ddH2O to remove excess stain. The ge l was destained with 5% (v/v) methanol and 7% (v/v) glacial ac etic acid in water overnight while gently shaking. The next day a picture was taken of the gel, and the gel was dried. Tsai-Frasch Silver Staining Tsai-Frasch silver staining ( 78) was performed on SDS-PAGE resolved LPS that was extracted from V. cholerae. Carbohydrates resolved by SDS-P AGE were fixed in 40% (v/v) ethanol and 5% (v/v) glacial acetic acid in water overnight. Carbohydrates were oxidized by incubating the fixed gel in 0.7% (w/v) periodic acid, 40% (v/v) ethanol, and 5% (v/v) glacial acetic acid in water for five minutes. The ge l was washed three times in 500 mL of ddH2O with 15 minutes per wash. Reagent A (1 % (v/v) 10 M NaOH and 6.7% (v/v) NH4OH in water) and Reagent B (20% (w/v) AgNO3 in water) were made immediatel y before staining. The stain was made by drop wise titration of Reagent B into Re agent A until a hazy brown color to the solution appeared and remained. The titrated r eagent was diluted in 115 mL of ddH2O to make the final stain solution. The gel was placed in the stai ning solution and stained for 10 minutes with agitation. The gel was washed three times in 500 mL of ddH2O with 15 minutes per wash. The gel was developed with 1 x Developer (5 x Tsai -Frasch developer: 1.3 mM citric acid and 0.25%

PAGE 52

52 (v/v) of 37% (v/v) formaldehyde in ddH2O) for 3 to 10 minutes with constant gentle shaking. The developed gel was soaked in 1% (v/v) acet ic acid in water for 10 minutes and transferred into 50 mL of ddH2O. A picture was taken of the gel, and then the gel was dried. Western Blot Proteins resolved by SDS-PAGE we re tran sferred onto a nitrocellulose membrane, followed by reaction with antibodies and develo pment with enhanced chemiluminescent (ECL) (Pierce). A Mini-TransBlotting cell (Bio-Rad) wa s used for the transfer of proteins to the nitrocellulose membrane. The transfer buffer (25 mM Tris, 192 mM glycine, and 20% (v/v) methanol) was pre-chilled to 4C. The transf er occurred at 100 V for one hour at 4C. Transferred proteins were analyzed by Ponceau S staining. Fifteen milliliters of 0.3% (w/v) Ponceau S dissolved in 3% (v/v) trichloroace tic acid (TCA) was adde d to the transferred membrane and incubated 10 minutes at room te mperature with gentle agitation. The stained membrane was gently washed in 25 mL of dH2O to remove excess stain. The membrane was photographed to view the stained proteins. The membrane was blocked in 25 mL of PBS casein blocking buffer (Pierce) contai ning 1% (v/v) Tween-20 for two hours at room temperature or overnight at 4C with gentle shaking. Primary antibodies were diluted in PBS casein blocking buffer (Pierce) containing 0.05% (v/v) Tween-20 to give a final volume of 10 mL and were added to the blocked membrane for one hour at room temperature with gentle shaking. The membrane was washed three times with 20 mL of PBS-0.05T while shaking for 15 minutes for each wash. Secondary antibodies conjugated w ith HRP were diluted in PBS casein blocking buffer (Pierce) containing 1% (v /v) Tween-20 to give a final volume of 10 mL and were added to the membrane for 30 minutes at room temperat ure with gentle shaking. The membrane was washed three times with 20 mL of PBS-0.05T while shaking for 10 minutes for each wash. The

PAGE 53

53 blot was incubated in 8 mL of ECL substrate (Pierce) for three minutes at room temperature and developed on CL-XPosure Film (Pierce) by an X-omat 2000 processor (Kodak, Rochester, NY). Lipopolysaccharide Saturation to Nitroce llulose Paper The LPS saturation experiment to nitrocellu lose paper was performed to determine the concentration of V. cholerae 569B LPS needed to saturate a piece of nitrocellulose paper. Phenol-water-extracted V. cholerae 569B LPS was serially diluted from 100 to 1 g/mL in PBS. Twelve pieces of nitrocellulose paper were cut to dimensions of 5 mm x 20 mm. Nitrocellulose strips were put in a microcentrifuge tube containing 256 L of the LPS dilutions or PBS for a control. LPS and PBS coated stri ps were done in duplicates in sepa rate tubes. The strips were coated overnight at 4C on a labquake. Coated ni trocellulose strips we re transferred to new microcentrifuge tubes. Strips were washed 3 times with 1 mL of PBS on a labquake with 5 minutes per wash. Strips were blocked with 1 mL of casein blocking buffer (Pierce) for one hour at room temperature on a labqua ke. Strips were incubated with 0.5 mL of rabbit anti-V. cholerae O1 LPS polyclonal antibody (Accurate Chemical, We stbury, NY) at a dilution of 1:400 in casein blocking buffer (Pierce) for one hour at room te mperature on a labquake. Strips were washed 3 times with 1 mL of PBS on a labqua ke with 5 minutes per wash. Strips were incubated with 0.5 mL of goat anti-rabbit peroxi dase conjugated monoclonal an tibody (Jackson ImmunoResearch Laboratories, West Grove, PA) at a concentratio n of 1:1000 in casein blocking buffer (Pierce) for 30 minutes at room temperature on a labquake. St rips were washed 3 times with 1 mL of PBS on a labquake with 5 minutes per wash. Strips were incubated in 5 mL of ECL substrate (Pierce) for three minutes at room temperature and de veloped on CL-XPosure Film (Pierce) by an Xomat 2000 processor (Kodak).

PAGE 54

54 Colony Blot with scFv To test whether bacteria produce scF v that ar e specific to a target molecule, a colony blot was developed to screen scFv in a high thr oughput manor. A Petri dish size piece of nitrocellulose paper (Bio-Rad) wa s coated overnight at 4C with 10 mL of target antigen at a concentration that the target an tigen was panned. The next day th e coated nitrocellulose paper was washed three times with 15 mL of 0.1 M TB S with five minutes per wash on a labquake. The nitrocellulose paper was then blocked for one hour at room temperature with 10 mL of casein blocker (Sigma-Aldrich) on a labquake. Afte r being blocked, the nitrocellulose paper was overlaid onto a plate containing bacterial colonies and stamped with a metal-plate stamp. The bacterial-colony-containing nitr ocellulose was washed three times with 15 mL of 0.1 M TBS with five minutes per wash on a labquake. The nitrocellulose paper was incubated with Protein L-peroxidase (Sigma-Aldrich) that was dilute d 1:2,000 in 10 mL of casein blocker (SigmaAldrich) for one hour at room temperature on a labquake. The nitr ocellulose paper was washed three times with 15 mL of 0.1 M TBS with five minutes per wash on a labquake. The nitrocellulose paper was incubate d in 5 mL of ECL substrate (Pie rce) for three minutes at room temperature and developed on CL-XPosure Film (Pierce) by an X-omat 2000 processor (Kodak).

PAGE 55

55 Table 2-1. Bacterial strains and plasmids used. Strain Genotype / Description Source / Reference E. coli DH5 F80dlacZ M15 ( lacZYA-argF )U169 deoR, recA1, endA1 hsdR 17(rk mk +), phoA, supE44 -, thi1, gyrA96, relA1 Bethesda Research Laboratories, Rockville, MD E. coli DH5 (pNR100) F80dlacZ M15 ( lacZYA-argF )U169 deoR, recA1, endA1, hsdR17 (rk mk +), phoA supE44, thi1, gyrA96, relA1 (pNR100) (69) E. coli EC100D Fmcr A (mrr-hsd RMS mcr BC) 80dlacZ M15 lacX74 recA1 end A1 araD139 ( ara, leu )7697 gal U gal K rps L nupG pir+(DHFR) Epicentre, Madison, WI E. coli EC100D (pGTR203) (Hy ) Fmcr A (mrr-hsd RMS mcr BC) 80dlacZ M15 lacX74 recA1 end A1 araD139 ( ara, leu )7697 gal U gal K rps L nupG pir+(DHFR) (pGTR203) (Hy ) This paper E. coli ER2738 F lacIq, ( lacZ )M15 proA+B+, zzf::Tn10( TetR ) / fhuA2, supE, thi, ( lacproAB ), ( hsdMS-mcrB )5, (r m McrBC) New England Biolabs, Ipswich, MA E. coli HB101 (pMJ100) F-, hsdS20(rBmB), recAB 3, ara-14, proA 2 lacYl galK 2, rpsL 20(SmR), xyl-5 mtl -l, supE44, -, (pMJ100) (70) E. coli HB2151 ara, ( lac-pro), thi / F' proA+B+, lacIqZ M15 Amersham Pharmacia, Piscataway, NJ E. coli JM109 endA1, recA1, gyrA96, thi, hsdR17 (rk mk +), relA 1, supE44, ( lac-proAB ) /F traD36, proAB laq IqZ M15 Invitrogen, Carlsbad, CA E. coli MG1655 F, -, ilvG, rfb -50, rph -1 (71) E. coli MG1655 (pGTR203) (Hy ) F, -, ilvG, rfb -50, rph -1 (pGTR203) (Hy ) This paper E. coli O157:H7 EDL933 Stx1+, Stx2+, gyrA (72) E. coli O157:H7 87-23 Stx1-, Stx2-, gyrA (73)

PAGE 56

56 Table 2-1. Continued Strain Genotype / Description Source / Reference E. coli TG1 (lac proAB ), supE, thi, hsdD 5/F traD36, proA+B, lacIq, lacZ M15 MRC, Cambridge, UK E. coli TG1 (pGTR203) (lac proAB ), supE, thi, hsdD 5/F traD36, proA+B, lacIq, lacZ M15 (pGTR203) Gopal Sapparapu Hy Hyperphage genome (contains all M13 genes except gIII ) (66) M13cp M13mp19; CamR; full length gIII (64) M13cp-CT M13mp19; CamR; truncated gIII (64) M13cp-dg3 M13mp19; CamR; deleted gIII (64) pGTR203 pACYC184 with M13 gIII; CamR Gopal Sapparapu pMJ100 pBluescript with slt-II; AmpR (70) pNR100 Stx2 toxoid; AmpR (69) V. cholerae 569B Classical, Inaba (74) V. cholerae N16961 El Tor, Inaba (75)

PAGE 57

57 CHAPTER 3 RESULTS Rationale for Study There exists a strong need for rapid, sensitiv e, and selective m ethods of detection for biological agents. Detecting bi ological organisms and their products has many uses. One reason is for diagnosis of patient samples. Detecti on and identification of a disease-causing agent in patient samples allows for the pr oper treatment to be given. De tecting biological agents in food, water, and air allows for preventative measures to be enforced to prevent or inhibit the spread of disease. A common component in almost all detection assays is a protein that specifically binds to a target molecule. This thesis describes the optimization of protocols us ed to isolate specific proteins that bind to target mo lecules by using phage display. The specific aims of this study are: 1. To optimize extraction methods and determine saturation limits of V. cholerae LPS and use these methods to pan phage display libraries to V. cholerae LPS. 2. To optimize panning and screening proce dures to allow for a more efficient biopanning process that will be more likely to isolate and detect specific recombinant phagemid particles. 3. To optimize the production of phagemid par ticles to ensure high quality and high quantity of phagemid particles. 4. To isolate specific recombinant phagemid particles to E. coli O157:H7 Stx2 toxin using the optimization techniques discovered in previous aims. Specific Aim 1: Panning to V. cholerae LPS The f irst goal of this work was to isolate phage display reagents that specifically recognized V. cholerae O1 LPS. There are many serogroups of V. cholerae but only O1 and O139 cause epidemic cholera. Therefore, O1 LP S serves as a useful target because it is a distinguishing feature of V. cholerae O1 strains. Previous attemp ts in the laboratory to obtain phage display reagents that specifically recognized V. cholerae O1 LPS may have failed because the V. cholerae O1 LPS was panned with the Tomlinson scFv phagemid library. Phagemid

PAGE 58

58 particles, which only encode the gIII gene, require phage amplifica tion tools to provide phage genes I-XI to enable phagemid particle production. We chose the Ph.D. 12mer phage library (NEB) because the Tomlinson scFv phagemid lib rary was disadvantageous due to the poor quality of current methods to amplify phagemid particles. Because the Ph.D. 12mer phage library is M13 phage, it does not require the ai d of tools to amplif y phages like a phagemid library requires. Before the Ph.D. 12mer pha ge display library was to be panned against V. cholerae O1 LPS, an extraction method for V. cholerae O1 LPS and the best solid support for LPS binding had to be determined. Two LPS extraction methods were compared to determine which method generated the best quality of L PS to pan against. Also, two binding supports (microtiter well and nitrocellulo se) paper were analyzed to determine which binding support could bind the most LPS. The Phenol-Water Method Extracted V. cholerae LPS Most Closely Resembled the Commerc ially Acquired V. cholerae LPS The phenol-water method for the extraction of LPS (76) was the previous lab procedure used to extract LPS. However, a new met hod for LPS extraction involving TRIzol Reagent (Invitrogen) was described by Yi a nd Hackett (77) that was stated to be a more efficient method for LPS extraction than the phenol-water method. The TRIzol Reagent LPS extraction can be executed in one day, while the phenol-water LPS ex traction requires nearly a week. The TRIzol Reagent method for LPS extraction is also stated to be a cleaner method than the conventional phenol-water method for LPS extraction by ex tracting LPS with less degradation and contamination. TRIzol Reagent is composed of phenol and guanidinium thiocyanate in aqueous phase, and upon addition of chloroform the TRIz ol Reagent can be used to extract LPS. Vibrio cholerae N16961 LPS was extracted by the TRIzol Reagent extraction of LPS (see Materials and Methods). Th e TRIzol Reagent-extracted V. cholerae LPS was resolved by SDS-

PAGE 59

59 PAGE and stained with Coomassie blue for visua lization of proteins (Fi g. 3-1A) and with TsaiFrasch silver stain for visualization of carbohydr ates (see Materials and Methods) (Fig. 3-1B). The Coomassie blue stain of the TRIzol Reagent-extracted LPS, which was diluted 1:2 in Laemmli sample buffer, showed no detectable proteins. The limit of detection for proteins with Coomassie blue is 0.3-1 g of protein/band (79); theref ore, there was less than 0.3 g of any single protein loaded onto the gel that was stai ned by Coomassie blue. The Tsai-Frasch silver stain for the TRIzol Reagent LPS extraction sh owed non-staining bands be tween 20 and 30 kDa. The Tsai-Frasch silver stain prim arily stains carbohydrates and poor ly stains proteins and lipids (78); therefore, the non-staini ng bands may be lipids or proteins. However, because the Coomassie blue stain for the TRIzol-extracted L PS showed no detectable proteins and due to the thickness of the non-staining bands, which correl ates to concentrati on, then the non-staining bands were most likely due to lipid contamination. The LPS banding pattern of the TRIzol reagent LPS extraction in the silver stai n was very similar to the commercial V. cholerae 569B LPS (Sigma-Aldrich). On the s ilver stain, both LPS preparations had bands at approximately 14 and 20 kDa and no bands larger than 50 kDa. Previously published silver stains of V. cholerae O1 LPS had bands at ~10 and 14 kDa (which was stated to be the lipid A-core of LPS), and ~2050 kDa (which was stated to be the lipid Acore plus repeating O-antigens) (80,81). To determine the concentration of the TRIzol-Reagent-extract ed LPS, the intensities of the bands were compared to those of the commerci al LPS in the silver stained gel (Fig. 3-1B). The intensity of the band of the commercial LPS (1 mg/mL) that was diluted 1:5 was approximately 1.5 times stronger than the intensit y of the band of the TRIzol-Reagent-extracted LPS that was diluted 1:2. Therefore, we estim ated the concentration of the TRIzol-Reagent-

PAGE 60

60 extracted LPS was approximately 260 g/mL. The purity of this LPS extraction was less than 0.2 g of protein/ g of LPS. The phenol-water LPS extraction (76) was perf ormed to compare results with the TRIzol Reagent LPS extraction. The phe nol-water LPS extraction method di ffers little from the original phenol-water LPS extraction desc ribed by Westphal, Luderitz, a nd Bister in 1952 (82). The phenol-water LPS extraction method utilizes the property that most prot eins but not LPS are soluble in phenol, and that LPS is soluble in water. At temp eratures above 68C, phenol and water are miscible. Upon cooling to 5-10C and centrifugation, thre e phases result : an aqueous phase containing LPS, a phenol phase containing proteins, and a solid phase containing waterand phenol-insoluble compounds. Removal and pur ification of the LP S-containing aqueous phase follows the separation of the phases. One and a half liters of log phase (OD600 ~0.4-0.6) V. cholerae 569B was used for extraction of LPS by the phenol-water extraction protocol (see Material s and Methods). The four aqueous extractions were dialyzed and di gested by DNase (Qiagen) and RNase (Qiagen) because nucleic acids are also extracted into the aqueous phase. Prior to digestion the dialyzed extraction was diluted 1:100 in ddH2O, and the absorbance was measured at 260 nm and 280 nm (A260 and A280). The A260, which correlates to DNA concentration, predigestion was 0.277 and post-digestion was 0.148. The A280, which correlates to protein concentration, predigestion was 0.127 and post-digestion was 0.059. The final phenol-water extraction was examined for proteins by Coomassie blue stain (Fig. 3-2A) an d for carbohydrates by Tsai-Frasch silver stain (see Materials and Methods) (Fig. 3-2B). The Coomassie blue st ain for phenol-water-extracted V. cholerae LPS showed a protein band at approximately 15 kDa. This could possibly have been due to the 0.04 mg/mL RNase, 13.7 kDa, present in the extraction. The ba nd intensity correlated

PAGE 61

61 with the concentration of RNase present in the sa mple. Therefore, the protein contamination was most likely not from bacterial prot eins that were co-extracted. The Tsai-Frasch silver stain for phenol-water-extracted V. cholerae LPS showed similar banding patterns to that of the V. cholerae 569B LPS (Sigma-Aldrich). On the silver stain, both LPS preparations had bands at approximately 13, 20, and 23 kDa. The phenol-wat er extraction showed distinct banding around 20-50 kDa that correlates with published data of where the banding of the O-antigen of LPS occurs. The commercially obtained LPS had smearing around 20-50 kDa. Vibrio cholerae O1 LPS has 12-18 O-antigen groups (83); therefore, the additional banding in the phenol-waterextracted LPS may be due to the ability of our phenol-water method to extract LPS with higher numbers of O-antigen side chains than that of the method used for the extraction of the commercial LPS that was phenol extracted and purified by gel-filtration chromatography. To determine the concentration of the phenol-water-extracted LPS, the intensities of the bands were compared to those of the commercial L PS in the silver stained gel (Fig. 3-2B). The intensities of the bands of the commercial LPS (1 mg/mL) that was diluted 1:8 were between the intensities of the bands of th e phenol-water-extracted LPS th at was diluted 1:2 and 1:4. Therefore, we estimated the concentration of the commercial LPS at a dilution of 1:8 was comparable to the concentration of the phenol-wat er-extracted LPS at a dilution of 1:3. Because the commercial LPS was at a concentration of 1 mg/mL, the concentration of the phenol-waterextracted LPS was approximately 375 g/mL. The purity of this LPS extraction was approximately 1 g of protein/ g of LPS. However, if the protein band from the Coomassie blue stain was from RNase, the phenol-water-e xtracted LPS would have less than 0.3 g of bacterial protein contamination/ g of LPS, which is the same prot ein concentration in the TRIzolextracted LPS.

PAGE 62

62 Based on the results described above, we d ecided to use the phenol-water method for the extraction of V. cholerae O1 LPS over the TRIzol Reagent method. The phenol-water-extracted LPS most closely resembled V. cholerae 569B LPS (Sigma-Aldrich) when examined by silver stain. The phenol-water extraction method even appears to be a more sensitive and delicate extraction method than that used to extract th e commercial LPS, because more precise banding from O-antigens was detected in the phenol-wate r-extracted LPS than in the commercial LPS by silver stain. The phenol-water LPS extraction ex tracted more protein than the TRIzol Reagent LPS extraction; however, for using the LPS extrac tion for panning it is more important to have intact O-antigens than minor protein contamination. More V. cholerae LPS Can Be Bound to Nitrocellulose Paper than to a Microtiter Well In m ost panning procedures the target molecule to be panne d against is immobilized onto a solid support. We wanted to compare the satu ration limits for LPS binding to a microtiter well and to a piece of nitrocellulo se paper of equal dimensions. Enzyme-linked immunosorbent assays were performed to determine the sa turation limits of the amount of phenol-waterextracted LPS that could be bound to a microtite r well. A 96-well Maxisorp plate (Nunc) was coated with 0.1-200 g/mL of V. cholerae 569B LPS that was extracted by the phenol-water method. Two different primary antibodies were used to detect the LPS: a mouse antiV. cholerae O1 LPS monoclonal antibody, Vc O1 LPS mAb (Austral Biologi cals, San Ramon, CA) and a rabbit antiV. cholerae O1 polyclonal antibody, Vc O1 LPS pAb (Accurate Chemical, Westbury, NY). The primary antibodies were used at their manufacturer recommended concentration or d ilution: a concen tration of 1 g/mL for Vc O1 LPS mAb and a dilution of 1:100 for the Vc O1 LPS pAb. Standard ELISA procedures were used for the assay (see Materials and Methods) (Fig. 3-3). When using the Vc O1 LPS mAb a significant difference (p 0.05) between the signal to noise (S :N) values (signal on antigen-coated

PAGE 63

63 well/signal on PBS-treated well) of LPS bound to -Vc O1 LPS mAb reacting to HRPconjugated anti-mouse antibody was not reached until the concentration of the LPS used to coat the microtiter plate decreased from 1 g/mL to 0.1 g/mL (0.1 to 0.01 g LPS) (p = 0.01). When using the Vc O1 LPS pAb, a significant difference in th e S:N was not reached until the concentration of the LPS decreased from 10 g/mL to 1 g/mL (p = 0.01). Both the -Vc O1 LPS mAb and the Vc O1 LPS pAb when used at the manufacturer recommended concentrations saturated LPS-coated microtiter we lls when the microtiter wells were coated with LPS at a concentration of 10 g/mL. For the sake of conservi ng the LPS, the concentration of LPS used to coat microtiter wells was decided to be 1 g/mL. While there was a significant difference in the S:N when using 10 g/mL of LPS as opposed to 1 g/mL of LPS with the Vc O1 LPS pAb, the S:N difference was only 0.5. This d ecrease in the S:N by 0.5 did not merit the need to use 10-fold more LPS; therefore, for further experiments V. cholerae LPS was used at a concentration of 1 g/mL to coat microtiter wells. The saturation limits for the primary antibodies were analyzed using 1 g/mL V. cholerae LPS to coat a microtiter plate in an ELISA experiment. The -Vc O1 LPS mAb was serially diluted to concentrations from 1.3 to 0.04 g/mL by 2-fold dilutions. The Vc O1 LPS pAb was serially diluted from 1:100 to 1:3 ,200 by 2-fold dilutions. When using 1 g/mL LPS to coat a microtiter well, the saturation limit of the Vc O1 LPS mAb was not reached (Fig. 3-4A). There were significant differences (p=0.01-0.03) between the S:N of each dilution step when using the Vc O1 LPS mAb at concentrations from 1.3-0.04 g/mL. Therefore, when using 1 g/mL LPS the sa turation limit for the Vc O1 LPS mAb was greater than 1.3 g/mL. When using 1 g/mL LPS to coat a microtiter well, a significant difference in the measured S:N between dilution steps when using the serially diluted Vc O1 LPS pAb was not reached until a

PAGE 64

64 dilution of 1:400 (p=0.01) (Fig. 3-4B). Therefore, the saturation limit of the Vc O1 LPS pAb was reached at a dilution of 1:400. Because a microtiter well coated with 1 g/mL LPS was never saturated by the Vc O1 LPS mAb, the Vc O1 LPS pAb was used to characterize the saturation limits of nitrocellulose paper. Pieces of nitrocellulose paper were cut to dimensions (5 mm x 20 mm) that were approximately equivalent to the surface area of a microtiter well. The same phenol-waterextracted V. cholerae LPS used for the microtiter well saturation experiment was also used to characterize the saturatio n of nitrocellulose with LPS. Serially diluted LPS from 100-1 g/mL was used to saturate strips of nitrocellulose pa per. These strips of L PS-coated nitrocellulose paper were examined in an altered Western blot protocol (see Materi als and Methods for LPS Saturation to Nitrocellulose Paper) using the Vc O1 LPS pAb at a dilution of 1:400 (Fig. 3-5). The nitrocellulose paper was not saturated with LPS until 100 g/mL LPS (250 g) was used to coat the nitrocellulose paper. Therefore, n itrocellulose paper can bind approximately 250 to 2,500 times more LPS than a microtiter well. Panning to V. cholerae LPS Failed to Yield Phages that W ere Specific to V. cholerae LPS Vibrio cholerae 569B LPS (Sigma-Aldrich) was i mmobilized onto a polystyrene Maxisorp microtiter well and panned against the NEB Ph.D. 12mer phage display library (see Materials and Methods). Five r ounds of panning were performed with amplification of phages after the first four round s of panning. Each round of panning promotes selection of phages that specifically bind to the target molecule. Amp lifying the selected phages should increase the percentage of specific phages present in the libra ry to pan with. Panning and amplifying for five rounds should greatly improve the chances of isolating phages that specifically bind to the target molecule.

PAGE 65

65 One hundred phage clones from the fifth round of panning were amplified and screened by ELISA against V. cholerae 569B LPS, carbonate coating buffer, and PBS-treated microtiter wells. The -Vc O1 LPS mAb was also tested against the three coating antig ens and generated a S:N of 15.5, proving that the LPS successfully bound to the microtiter plate. Anti-BSA phagemid particles were used in the ELISA against BSA (10 g/mL), carbonate coating buffer, and PBS-treated microtiter wells and generated a S:N of 22, proving that the ELISA was a successful assay for the detecti on of phages. All 100 clones gave a S:N <1.4. A positive signal was classified as a S:N 2. Therefore, all of the clones screened by ELISA were negative, meaning that they failed to specifically bind to V. cholerae 569B LPS. Panning the Ph.D. 12mer phage library against V. cholerae O1 LPS immobilized onto a microtiter well failed to yield phage clones specific to V. cholerae O1 LPS. It was therefore attempted to improve the panning procedure. Instead of using a microtiter well as a binding support for LPS, nitrocellulose paper was used as a binding support. Nitrocellulose paper can bind approximately 250 to 2,500 times more LPS than a microtiter well; therefore, if more antigen is present to pan against then the chances of isolating clone s that specifically bind to LPS may improve. Also, the elution method was te sted. The previous elution method was with glycine (pH 2.2). To determine if a glycine elutio n or an elution with a high concentration of the target antigen is more successful in eluting phages both elution methods were tested in parallel. The theory behind an antigen elution is that a high concentration of antigen in solution will rapidly bind phages bound to the antigen on the nitrocellulose paper when the phages are transiently released. The end result is that th ese phages will elute o ff of the antigen-coated nitrocellulose paper and bind to antigen in solution. Antigen elution is supposed to be a less harsh treatment than acidic or basic elu tions, which may denature the peptides.

PAGE 66

66 Vibrio cholerae 569B LPS that was phenol-water-extract ed was immobilized onto a strip of nitrocellulose paper and panned with the Ph.D. 12mer phage display library (NEB) (see Materials and Methods). Two pa rallel panning experiments were pe rformed; one panning used a glycine (pH 2.2) elution, while the other panning used a V. cholerae 569B LPS elution. Three rounds of panning were performed with a negati ve panning performed between rounds one and two. The negative panning involved panning the libra ry against a strip of nitrocellulose paper that was blocked in casein blocking buffer (Sigma -Aldrich) to remove phages that were specific to nitrocellulose paper or casein blocking buffer. Amplification of phages was performed after rounds one and two to increase the number of selected phages to improve the chances of isolating phages specific to the target mol ecule. Two hundred clones for each panning were selected from the round 3 elution and amplif ied for screening of phages by ELISA. The 400 phage clones were screened in an ELISA against V. cholerae 569B LPS and PBS. The -Vc O1 LPS mAb was also tested against LPS and PBS-tr eated wells and generate d a S:N of 10, proving that the LPS successfully bound to the microtiter pl ate. Anti-BSA phagemid particles were used in the ELISA against BSA (10 g/mL) and PBS-treated microtiter wells and generated a S:N of 15, proving that the ELISA was a successful assay for the detection of phages. All 400 clones gave a S:N less than 1.7. Therefore, all of the clones screened were negative and failed to specifically bind to V. cholerae O1 LPS. Conclusion of Specific Aim 1 In com paring the phenol-water and the TR Izol Reagent methods for extraction of V. cholerae O1 LPS, the phenol-water method was bett er. The phenol-water extraction for LPS yielded LPS that most closely resembled commercially obtained V. cholerae O1 LPS by silver stain of SDS-PAGE-resolved LPS. The phenol-wat er method even extracted LPS that appeared to have more O-antigens than the commercially obtained V. cholerae O1 LPS. In comparing the

PAGE 67

67 saturation limits of V. cholerae O1 LPS bound to microtiter wells a nd to nitrocellulose paper, the nitrocellulose paper bound approxi mately 250 to 2,500 times more LPS than did microtiter wells. Panning the Ph.D. 12mer phage display library against V. cholerae O1 LPS immobilized onto nitrocellulose and onto microtiter wells failed to yield phages that were specific to V. cholerae O1 LPS in an ELISA. Specific Aim 2: Improve Panning a nd Screening Process of Biopanning There are many steps in the bi opanning and screening process. By analyzing certain key steps and optim izing them, the chances of isola ting phagemid particles th at are specific to the target molecule panned against should increa se. Optimization of the panning process was performed with V. cholerae O1 whole cells as a target. Whole cells were chosen as the target because they are the target antigen for the bios ensors under development by our collaborators. Optimization of the concentration of whole cells were analyzed to determine if higher or lower concentrations of the antigen were more likely to select for specific phagemid particles. Increasing the concentration of antigen shoul d increase the amount of phagemid particles selected. This could mean an increase in phagemid particles that are specif ic and/or non-specific to the target molecule. Decreasing the concentr ation of antigen should decrease the amount of phagemid particles selected. This could mean a decrease in phagemi d particles that are specific and/or non-specific to the target molecule. An id eal concentration would se lect a high number of phagemid particles that are specific to the target molecule and a low number of phagemid particles that are not specific to the target molecule. The Tomlinson scFv libraries have a trypsin cleavable site between the scFv peptide and the pIII peptide of the phagemid particle. Cleaving this site separates the phagemid particle from the antigen, leaving the scFv bound to the antigen. Optimizing the time of trypsin elution may improve the chances of isolating specific phagemid particles. Eluting with trypsin for too short

PAGE 68

68 of a time may fail to elute specific phagemid pa rticles, but eluting for too long of a time may increase the amount of nonspecific phagemid par ticles eluted or may even cause degradation of the phagemid particles. Elution of Bound Phages by Trypsin durin g Panning Was Optimal betw een 10 and 30 Minutes A one-round panning procedure was done to dete rmine the optimal concentration of whole cells in suspension and the optimal trypsin elution time for biopanning. The purpose of the experiment was to compare the differences between using 1 x 108 or 1 x 109 whole cells of V. cholerae N16961 and trypsin elution times of 15, 30, 45, and 60 minutes in the biopanning process. Four parallel pannings were performed. Two were panned with the Tomlinson J library (2.5 x 1010 tu), one against 1 x 108 and the other against 1 x 109 whole cells of V. cholerae in suspension. The two other pannings us ed the Tomlinson J library (2.5 x 1010 tu) containing 1% (~2.5 x 108 tu) of Vc86 phagemid particles (an scFv-p roducing phagemid particle previously isolated in this lab by Dr. Rebecca Moose-Clem ente from the Tomlinson scFv library that recognizes V. cholerae 1019 and E. coli O157:H7 whole cells). Panning against a library that is already enriched with phagemid particles th at recognize the target molecule should help determine the optimal panning parameters. If more phagemid particles are eluted with the enriched library than the non-enri ched library at a certain panning parameter then it would lead to the conclusion that the phagemid particles that recognized the target molecule were eluted at that parameter. One panning used 1 x 108, and the other used 1 x 109 whole cells of V. cholerae in suspension. Standard suspension panning pr ocedures (see Materi als and Methods) were performed. When the trypsin was added to th e bound-phagemid particles, an aliquot of the eluted phagemid particles was removed at 15, 30, 45, and 60 minutes after the start of the elution and was titered immediately (Fig. 3-6). Elution of phagemid particles w ith trypsin was optimal

PAGE 69

69 between 15 and 30 minutes. The phagemid par ticle titer decreased by approximately 80-93% from 30 to 45 minutes for the four pannings. Mo re phagemid particles were eluted when panned against 1 x 109 whole cells. However, there were not enough data to do statis tical analysis to determine if using 1 x 109 whole cells eluted significantly more phagemid particles than using 1 x 108 whole cells in the biopanning process. Because the best whole cell concentration a nd trypsin elution time was not definitively obtained in the first experiment, a second one-rou nd panning was done to analyze these factors. The purpose of the experiment was to compare the differences between using 1 x 107, 1 x 108, and 1 x 109 whole cells of V. cholerae N16961 and trypsin elution times of 10, 20, and 30 minutes in the biopanning process. The same pr otocol was used as the first one-round panning protocol above, except an additional concentratio n of whole cells was added, so there were six parallel pannings instead of four. When the trypsin was added to the bound-phagemid particles, an aliquot of the eluted phagemid particles was removed at 10, 20, and 30 minutes after the start of the elution and was titered im mediately (Fig. 3-7). There were no significant differences in the titers of eluted phagemid particles when comparing the con centrations of whole cells and trypsin elution times used in the second one -round panning experiment. The panning was expected to yield distinct differences between eluted titers of phagemid particles from the libraries with and without 1% Vc86 phagemid partic les. A certain parameter was supposed to be optimal and yield higher elution titers from th e library with 1% Vc86 phagemid particles. Because there were not any significa nt differences in the titers of the eluted phagemid particles at the different parameters an optimal panning cond ition could not be determined. To determine which panning parameter elutes the most phagemid particles that recognize V. cholerae whole

PAGE 70

70 cells, another one-round panning was performed with analysis of phagemid particle specificity via ELISA. A third one-round panning was performed to anal yze whole cell concentration and trypsin elution time further. The purpose of the experi ment was to compare the differences between using 1 x 108 and 1 x 109 whole cells of V. cholerae N16961 and trypsin elution times of 10 and 30 minutes in the biopanning process. The same protocol was used as for the first one-round panning protocol above, except that only two elution time points were taken. When the trypsin was added to the bound phagemid particles, an a liquot of the eluted phagemid particles was removed at 10 and 30 minutes after the start of th e elution and titered im mediately (Fig. 3-8). When panning with the Tomlinson J scFv pha gemid library that contains 1% (~2.5 x 108 tu) Vc86 phagemid particles, the amount of eluted phagemid particles did not significantly change (p=0.50) from 10 to 30 minutes when panning with 1 x 109 whole cells. When panning with the library that contains 1% (~2.5 x 108 tu) Vc86 phagemid particles, the amount of eluted phagemid particles significantly decreas ed (p=0.04) from 10 to 30 minutes when panning with 1 x 108 whole cells. Because the amount of eluted phagemid particles either stayed the same or decreased from 10 to 30 minutes in the pannings w ith 1% Vc86 phagemid particles, it suggested that eluting for 30 minutes had no adva ntage over eluting for 10 minutes. To test if the eluted phagemid particles at the various time points were specific to V. cholerae an ELISA was performed. The eluted phagemid particles from the third one-round panning experiment were amplified in E. coli TG1 (Hy -lg), which is E. coli TG1 that was already infected with hyperphage, an d used at a concentration of 1 x 108 tu/mL in an ELISA. The phagemid particles were analyzed by ELISA against V. cholerae whole cells (3 x 108 CFU/mL) and PBS-treated microtiter wells. A standard ELISA protocol (see Materials and

PAGE 71

71 Methods) was performed (Fig. 3-9). Of the diffe rent panning variations, the amplified phagemid particles from the panning with the library that contained 1% Vc86 phagemid particles, 1 x 108 whole cells V. cholerae, and an elution time of 30 minutes generated a significantly higher S:N (p=0.01-0.04) in an ELISA compared to the amp lified phagemid particles from the other panning variations. This suggested that panning with 108 whole cells with a trypsin elution time of 30 minutes was optimal for selecting phagemid part icles that were specifi c to the panned whole cells compared to using 109 whole cells or a trypsin elution of 10 minutes. Screening of scFv Proteins by a Hig h Throughput ELISA Was Acceptable Once the panning procedure is com plete, screen ing of phages or scFvs is performed. The previous method of screening scFvs involved in ducing phagemid-containi ng bacteria to produce scFvs in culture tubes or flasks and screeni ng the scFvs by ELISA. However, this method was very time consuming. Optimizing the screening process to enable high throughput screening of clones should increase the chances of finding clones that specifically bind to the target molecule. A colony blot was attempted to try to enable hi gh throughput screening of scFv proteins. The colony blot (see Materials and Meth ods) involved the transfer of bacterial colonies that were grown to secrete scFvs onto a piece of nitrocellulo se paper that was coated with a target antigen and blocked with casein blocking buffer. Following the transfer, the piece of nitrocellulose paper was washed with TBS to remove unbound scFv proteins, incubated with Protein-L peroxidase (Sigma-Aldrich) to bind to the kappa light chains of the scF v, washed with TBS to remove excess Protein-L peroxidase, and developed with ECL (Pierce). Upon development with ECL, scFv-producing clones specific to BSA fa iled to produce a positive signal against a piece of nitrocellulose paper coated in 10 g/mL BSA but did produce a positive signal against pieces of nitrocellulose paper coated in PBS or V. cholerae whole cells. Also, clones that were not specific to V. cholerae whole cells in an ELISA produced positive signals against pieces of

PAGE 72

72 nitrocellulose paper coated with V. cholerae whole cells, PBS, or 10 g/mL BSA. The colony blot was determined to be ne ither selective nor specific. Because the colo ny blot was not functioning, a way to improve the culture tube method for the production of scFvs was performed. A high throughput ELISA was contrived to enable screening of hundreds of clones per day. This high throughput procedure involved pi cking bacterial colonies containing a phagemid and growing them overnight in a 96-well plate to produce scFvs (see Materials and Methods). The next day the 96-well plate containing the tu rbid cultures was centrifuged to remove the bacterial cells. These scFvs in the supernatant we re examined in an ELISA for screening of their specificity. To determine how well scFvs could be produced in a microtiter well, as opposed to a culture tube or flask, a bacteria l colony containing a phagemid that encodes an scFv specific to BSA was grown in a microtiter well. Wh en analyzed by ELISA, the anti-BSA ( -BSA) clone was diluted 1:2 in casein blocking buffer and pr oduced a S:N of 11.5 in an ELISA against BSA. Anti-BSA scFv produced in a flask produced an average S:N value between 10 and15; therefore, producing scFv proteins in a mi crotiter well may not be optimal, but it was acceptable for the purpose of a high throughput screen. We were successful in producing scFvs in a microtiter well. The next test was to determine how far the scFv can be diluted and still produce a positive signal. The -BSA clone was grown in a microtiter well to produce scFvs. The produced scFvs were diluted 1:2, 1:4, 1:6, and 1:10 in casein blocking buffer and analyzed by ELISA (Fig. 3-10). Diluting the scFv supernatant 1:10 in casein blocking buffer yielde d a S:N of 5.6. This S:N was high enough for detection of a positive clone and was not significantly different (p=0.13) than the S:N produced when using -BSA phagemid particles that were d iluted 1:6. Therefore, diluting scFv

PAGE 73

73 supernatants as far as 1:10 with other clones can enable detection of positive clones. However, the -BSA clone was the strongest cl one that we had; therefore, weaker clones might fail to be detected if analyzed at a dilution of 1:10. Fo r the purpose of a high throughput ELISA screen for scFvs I would recommend diluting the scFv proteins only 1:4 to 1:6 to less en the chances of not detecting weaker binding scFvs. Screening of Phagemid Particles by a Hig h Throughput ELISA Was Not Optimal Once the panning procedure is com plete, scre ening of phagemid particles or scFvs is performed. Previous methods of screening pha gemid particles involved the amplification of eluted phagemid particles in culture tubes or flasks in 2xTY containing 100 g/mL ampicillin, 40 g/mL kanamycin, and 0.1% (w/v) glucose at 30 C overnight, followed by screening by ELISA. This process was time-consuming; therefore, it was tested if phagemid particle production could be performed in a high throughput manor as with scFvs. Vc86 and -BSA phagemids in E. coli TG1 (Hy -sm) ( E. coli TG1 that was previously infected wi th hyperphage) were produced in a 96-well plate in 2xTYcontaining 100 g/mL ampicillin, 40 g/mL kanamycin, and with or without 0.1% (w/v) glucose at 30C and 37 C overnight. Protoc ols recommend producing phagemid particles in medium that has 0.1% (w/v ) glucose, but due to catabolite repression of the lac promoter, which drives transcription of the gIII gene, phagemid particle production might increase without the presence of glucose in the gr owth medium. The next day, the bacterial cells were removed by centrifugation, and the phage-cont aining supernatants were examined in an ELISA at a dilution of 1:2 in casein blocking buf fer (Fig. 3-11). There was not a significant difference in the S:N (p=0.07-0.48) of phagemid part icles produced in micr otiter wells with and without 0.1% (w/v) glucose. Vc86 phagemid particles produced in micro titer wells with or without glucose at 30C and 37C di d not produce a S:N higher than on e. It was also noted that the cultures in the microtiter plates were onl y slightly turbid. Th e Vc86 phagemid particles

PAGE 74

74 produced in a flask (S:N ~4) yielded a significantly higher S:N (p=0.05) compared to Vc86 phagemid particles produced in microtiter wells. When -BSA phagemid particles were produced at 30C, the S:N was signifi cantly higher (p<0.04) than when -BSA phagemid particles were produced at 37C at either glucose concentration. The -BSA phagemid particles produced in a flask (S:N ~12) yielded a significantly high er (p=0.02) S:N compared to -BSA phagemid particles produced in microtiter wells (S:N~7). The experiment was repeated a second time with only the -BSA phagemid particles (Fig. 3-12). Anti-BSA phagemid particles produced in microtiter wells yielded significantly higher signals (p<0.03) when produced at 30C than at 37C at either glucose concentration. There was not a significant difference (p=0.34) in the signals produced by phagemid particles that were produced with or wit hout glucose at 30C. When phage mid particles were produced in microtiter wells at 37C, there was a significan tly higher (p=0.02) signal when phagemids were produced with no glucose compared to 0.1% (w /v) glucose. There was not a significant difference in the signal (p=0.09) produced by pha gemid particles produced in microtiter wells compared to phagemid particles produced in a flask. When phagemid particles were diluted 1:2, pha gemid particles yielded significantly higher signals in an ELISA when produ ced at 30C compared to 37C. Whether the growth medium included 0.1% (w/v) glucose or not when phagemi d particles were incubated at 30C did not significantly affect the signal of the phagemid particles; theref ore, 0.1% (w/v) glucose was not enough to cause catabolite repression of the lac promoter. Also, phages produced in flasks yielded higher signals than phage s produced in microtiter wells. To determine if low titers of phagemid particles produced in a microtiter well were a factor in the lowered ELISA signals for phagemid particles, the phagemid particles were titered. When -BSA and Vc86 phagemid

PAGE 75

75 particles were produced in mi crotiter wells at 30C, the titers were approximately107 tu/mL. When -BSA and Vc86 phagemid particles were produced in microtiter wells at 37C, the titers were approximately106 tu/mL. Growing phagemid particles at 30C in microtiter wells yielded approximately 10-fold higher tu/mL titers than phagemid particles that were produced in microtiter wells at 37C. Whether the growth medium contained 0.1% (w/v) glucose or not did not affect the titers of phagem id particles. All phages produced in microtiter wells were amplified by E. coli TG1 (Hy -sm). Escherichia coli TG1 (Hy -sm) is E. coli TG1 that was already infected with hyperphage. Having hyperphage already present in E. coli removes an additional step in phagemid par ticle production to enable a more high throughput screen When -BSA and Vc86 phagemid particles were produced in flasks at 30C with 0.1% (w/v) glucose in the medium, the titers were approximately108 to109 tu/mL. The phagemid particles produced in the flasks were amplified by superinf ection with hyperphage, instead of by E. coli TG1 harboring hyperphage. Growing phagemid particles by superi nfection with hyperphage may have resulted in higher titers of phagemid particles comp ared to phagemid particles produced from E. coli that already contained hyperphage. The high throug hput phagemid particle ELISA was not optimal for screening phagemid particle s because phagemid particles could not be produced in high enough titers in microtiter wells to make scr eening of phagemid particles effective in a high throughput manor. Conclusion of Specific Aim 2 When panning with the Tom linson J human synthetic VH + VL phagemid library, trypsin elution was best between 10 and 30 minutes. Phag emid particles could possibly be degraded by trypsin treatment of 45 minutes or greater. Wh en screening scFvs, a high throughput ELISA was acceptable. The scFv proteins can be diluted 1:10 in an ELISA with other scFv proteins and still produce high enough signals to be detected. When screening phages, screening in a high

PAGE 76

76 throughput ELISA was not optimal. Phagemid particles cannot be produced in high enough titers when produced in microtiter wells to be useful in a high throughput manor. Specific Aim 3: Improve Phagemid Particle Production Phagem id particle production is one of the most important steps in biopanning. Because the only M13 gene in a phagemid is the gIII gene, use of phagemids requires an amplification tool to be provided in trans to supply the additional M13 genes to enable phagemid particle production. There needs to be high quality and high quantity phagemid particle production during amplification of phagemid pa rticles to ensure that redundanc y of clones in the library is maintained and clones are produced in high enou gh titers to be useful in assays that use phagemid particles. Hyperphage (Progen) was the previous phagemid particle amplification tool. Hyperphage is an M13-based phage that contains a par tial deletion of the gIII gene to enable production of phagemid particles that display recombinant pIII proteins but no wild-type pIII proteins. However, the commercially obt ained hyperphage stock was of poor quality. The hyperphage stock was st ated to contain 1012 particles/mL; however, it only had 108 tu/mL, as determined in an infection assay. Because there was not enough Tomlinson scFv phagemid library stock for panning experiment s, the library stock needed to be amplified. The Tomlinson I and J scFv libraries contain 1 x 108 phagemid clones. To ensure that the Tomlinson scFv libraries are amplified to ensure that the redundancy of clones is maintained, an amplification tool needs to enable high qua lity and high quantity phagemid particle production. Because commercially obtained hyperphage had too low of titers to be an effective phagemid particle amplification tool, other ways of phage mid particle production were analyzed.

PAGE 77

77 Escherichia coli TG1 Harboring th e Hyperphage Geno me Was Not an Optimal Phagemid Particle Amplification Tool To overcome the low titers of hyperphage, the hyperphage genome was incorporated into E. coli and maintained in the strain as a plasmi d to ensure that every phagemid-containing E. coli also contained the hyperphage genome to enab le phagemid particle production. Log phase (OD600 ~0.4) E. coli TG1 was transduced with hyperphage at a MOI of 5, incubated in a 37C standing water bath for 30 minutes, serially d iluted in BSG, and then plated on 2xTY plates containing 40 g/mL kanamycin. The plates were incubated overnight at 37C. The next day the plates contained small (0.5 mm diameter) and large (1-1.5 mm diameter) colonies. Two small and large colonies were si ngle colony passaged twice. Th e small colonies maintained a small phenotype (0.5 mm diameter), and the large colonies maintained a large phenotype (1-1.5 mm diameter). The two strains were named E. coli TG1 (Hy -sm) and E. coli (Hy -lg). The infection efficiencies of phagemid particles into E. coli TG1, E. coli TG1 (Hy -sm), and E. coli (Hy -lg) were analyzed. The three E. coli strains were grown to log phase and transduced with serially diluted -BSA and Vc86 phagemid particle s in BSG at a MOI of 0.001. Transductions were incubated in a 37C standi ng water bath for 30 minutes and plated on 2xTY AG plates containing 40 g/mL kanamycin for transduced strains of E. coli TG1 containing hyperphage and on 2xTY AG plates for transduced E. coli TG1. The numbers of E. coli cells transduced with -BSA phagemid particles were as follows: 1,200 transductions into E. coli TG1, 340 transductions into E. coli TG1 (Hy -sm), and 100 transductions into E. coli TG1 (Hy -lg). When the number of -BSA phagemid particles transduced into E. coli TG1 was set at an infection efficiency of 100%, the number of -BSA phagemid particles transduced into E. coli TG1 (Hy -sm) was 28% and the number of -BSA phagemid particles transduced into

PAGE 78

78 E. coli TG1 (Hy -lg) was 0.1% of the number of tr ansduced phagemid particles into E. coli TG1. The numbers of E. coli transduced with Vc86 phagemid particles were as follows: 4,300 transductions into E. coli TG1, 950 transductions into E. coli TG1 (Hy -sm), and 80 transductions into E. coli TG1 (Hy -lg). When the number of Vc86 phagemid particles transduced into E. coli TG1 was set at an infection efficiency of 100%, the number of Vc86 phagemid particles transduced into E. coli TG1 (Hy -sm) was 22% and the number of Vc86 phagemid particles transduced into E. coli TG1 (Hy -lg) was 2% of the number of transduced phagemid particles into E. coli TG1. Escherichia coli TG1 (Hy -sm) had an infection efficiency that was approximately 25% of the infection efficiency of E. coli TG1, and E. coli TG1 (Hy -lg) had an infection efficiency that was approxi mately 1% of the infection efficiency of E. coli TG1. The infection efficiencies of E. coli TG1 (Hy -sm/lg) were too low to for the E. coli strains to be useful to amplify the Tomlinson scFv libraries. However, they could still be useful as an amplification tool to amplify single clones when the transducti on efficiency is not critical. Therefore, the quality and quantit y of phagemid particles produced by E. coli TG1 (Hy -sm/lg) were analyzed. Escherichia coli TG1 (Hy -sm) and E. coli TG1 (Hy -lg) were analyzed for their ability to produce phagemid particles. Escherichia coli TG1 (Hy -sm) and E. coli TG1 (Hy -lg) were transduced with -BSA phagemid particles (as described above) and grown under conditions to produce phagemid particles (see Materials and Methods). Escherichia coli TG1 containing BSA phagemids was superinfected and grown under conditions to produce phagemid particles (see Materials and Methods ). The superinfected E. coli TG1 was used as a comparison for strains of E. coli TG1 containing hyperphage as a phagemid particle production tool. Anti-BSA

PAGE 79

79 phagemid particle yields were as follows: E. coli TG1 that was superinfected with hyperphage yielded 3 x 10 8 tu/mL, E. coli TG1 (Hy -sm) yielded 1 x 109 tu/mL, and E. coli TG1 (Hy -lg) yielded 3 x 108 tu/mL. These results showed that E. coli TG1 that contained hyperphage produced high titers of phagemid particles. To determine the quality of -BSA phagemid particle s produced by the three E. coli strains, the produced -BSA phagemid particles were analyzed in an ELISA. Microtiter plates were coated with 10 g/mL BSA or only PBS. Standard ELI SA procedures were followed (see Materials and Methods) to analyze 1 x 108 tu/mL of -BSA phagemid particles as the primary antibody (Fig. 3-13). Anti-BSA phagemid particles produced a significantly higher S:N (p=0.01) when produced from E. coli TG1 (Hy -sm) compared to E. coli TG1 (Hy -lg). There was not a significant difference in the S:N (p=0.50) from -BSA phagemid particles produced from superinfected E. coli TG1 and E. coli TG1 (Hy -lg). Because superinfected E. coli TG1 and E. coli TG1 (Hy -lg) yielded titers (~3 x 108 tu/mL) of -BSA phagemid particles that were approximately a third of that of E. coli TG1 (Hy -sm) titers (~1 x 109 tu/mL), E. coli TG1 (Hy -sm)-produced -BSA phagemid particles were analy zed in the ELISA with a third less phage-containing supernatan t volume. The amplified -BSA phagemid particles were from the same clone; therefore, the differe nces in S:N from the produced -BSA phagemid particles was not expected. A possible reason for the S:N differences was that a different amount of particles/mL was used in the ELISA. The ELISA compared equal tu/mL titers of the -BSA phagemid particles produced. However, tu/mL is a measure of infectious particles and not total particles. Because an ELISA measures partic les with binding activity, it is possible that superinfected E. coli TG1 and E. coli TG1 (Hy -lg) yielded similar particles/mL yields as E. coli TG1 (Hy -sm), which would account for the differences in the S:N. The results showed

PAGE 80

80 that E. coli TG1 (Hy -sm) and E. coli TG1 (Hy -lg) had infection efficiencies that were too low for them to be effective amplification t ools to amplify the Tomlinson J scFv library. However, these strains could be used as t ools to amplify single pha gemid particles with acceptable phagemid particle production and quality. Helper Plasmids Were Not an Optimal Ph agemid Particle Amplifica tion Tool Helper plasmids (Los Alamos National Labor atory, Los Alamos, NM) (64) are M13-based plasmids that are maintained in E. coli to provide phagemid-containing E. coli with all of the genes necessary to produce phagemid particles. The helper plasmids come in three forms that either have a full-length (M13cp), a dele ted (M13cp-dg3), or a truncated (M13cp-CT) gIII gene. The helper plasmid-containing E. coli strains were analyzed for their infection efficiencies with phagemid particles, production yield of phagemid pa rticles, and the quality of phagemid particles produced. The three helper plasmids were electropor ated (see Materials and Methods) into electrocompetent E. coli TG1. All three transformations yiel ded colonies between 0.5-2 mm in diameter. A small (0.5 mm) and a large (2 mm) colony from E. coli TG1 (M13-cp), E. coli TG1 (M13cp-CT), and E. coli TG1 (M13cp-dg3) were single colony passaged twice. Escherichia coli TG1 (M13cp-sm) was the only strain that maintain ed a consistent phenotype of 0.5 mm colonies. The other 5 strains yielded co lonies with diameters of 0.52 mm after each passage. To determine if the different colony size phenotypes ha d an effect on the functionality of the helper plasmids, a small and large isolate from each strain were analyzed for their infection efficiencies with phagemid particles. To analyze the infection efficiencies of phage mid particles into helper plasmid-containing E. coli TG1, an infection efficiency (see Materials and Methods) experiment was performed that compared the infection efficienci es of helper plasmid-containing E. coli with E. coli that did not

PAGE 81

81 contain helper plasmids. The bacterial strains were grown to log phase and transduced with BSA or Vc86 phagemid particles at a MOI of 0.1 (Table 3-1). The number of E. coli TG1cells that were transduced with phagemid particles was set at an infection efficiency of 100%, and the number of E. coli TG1 cells harboring helpers that were transduced with phagemid particles was compared to the infection efficiency of E. coli TG1. Of the six E. coli TG1 strains harboring helper plasmids, E. coli TG1 (M13cp-dg3-sm) and E. coli TG1 (M13cp-CT-sm) were the only strains that yielded inf ection efficiencies grater than 0.3% of the infection efficiency of E. coli TG1. Escherichia coli TG1 (M13cp-dg3-sm) had infection efficiencies of 175% and 97%, while E. coli TG1 (M13cp-CT-sm) had infection efficiencies of 7% and 13% compared to the infection efficiency of E. coli TG1. Because the small and large isolates of helper plasmid-containing E. coli had different infection effi ciencies, the colony phenotype had an effect of the functionality of the he lper plasmid-containing E. coli. Because initial tests with E. coli TG1 (M13cp-dg3-sm) had an infection efficiency th at was as high or highe r than the infection efficiency of E. coli TG1, its infection efficiency was further analyzed with the transduction of additional phagemid particles. Escherichia coli TG1 (M13cp-dg3-sm) was transduced w ith clone 18 (a phagemid particle isolated from the Tomlinson I scFv library that recognizes the A27L protein of vaccinia virus), -AV20N3 (a phagemid particle isolated from the Tomlinson I scFv library that recognizes rAuto (a recombinant L. monocytogenes murein hydrolase)), and Vc86 phagemid particles. The various strains were analyzed further for their infection efficiencies compared to E. coli TG1 (Table 3-2). Escherichia coli TG1 (M13cp-dg3-sm) had infection efficiencies of 92%, 113%, and 120% compared to the infection efficiency of E. coli TG1. Escherichia coli TG1 (M13cpdg3-sm) had infection efficiencies that were approximately the same as E. coli TG1; therefore,

PAGE 82

82 this strain was used for further analysis to determine the quantity and the quality of phagemid particles produced. However, before E. coli TG1 (M13cp-dg3-sm) was analyzed further, M13cp-CT-sm was examined. To test if the helper plasmid (M13-CT-sm) c ould yield a higher inf ection efficiency in another E. coli strain, it was extracted (see Materials and Methods) from E. coli TG1 (M13cpCT-sm) and electroporated (see Materials and Methods) into E. coli JM109. Escherichia coli JM109 was chosen because it was genotypically more similar than E. coli TG1 to the bacterial strain that Los Alamos National Laborat ory used with the helper plasmids ( E. coli DH5 F). Escherichia coli JM109 and E. coli DH5 F contain relA 1 and recA 1, while E. coli TG1 does not. Escherichia coli TG1 (M13cp-CT-sm) and E. coli JM109 (M13cp-CT-sm) were transduced with clone 18, -AV20N3, and Vc86 phagemid particle s and were analyzed for their infection efficiencies compared to E. coli TG1 and E. coli JM109 (Table 3-3). The average infection efficiency of E. coli TG1 (M13cp-CT-sm) was 47%, while the average infection efficiency of E. coli JM109 (M13cp-CT-sm) was 28%. Therefore, using E. coli JM109 to maintain M13cp-CT-sm did not increase the infection efficiency. In analyzing the infection efficiencies, E coli TG1 (M13cp-dg3-sm) had an average infection efficiency of 120% with a standard deviation of 33% and E. coli TG1 (M13cp-CT-sm) had an average infection efficiency of 30% with a standard deviation of 16%. Escherichia coli TG1 (M13cp-dg3-sm) had a significantly highe r infection efficiency (p=0.01) than E. coli TG1 (M13cp-CT-sm). Escherichia coli TG1 (M13cp-dg3-sm), E. coli TG1 (M13cp-CT-sm), and E. coli TG1 that was superinfected with hyperphage were an alyzed for their abil ity to produce phagemid

PAGE 83

83 particles. To determine if th e addition of 1 mM IPTG to th e growth medium would increase phagemid particle production, phagemid particle amplification was performed in parallel with and without the addition of 1 mM IPTG to the growth medium. Isopropyl-beta-Dthiogalactopyranoside is a molecular mimic of a lactose metabolite that induces transcription from the lac promoter. Because the lac promoter drives transcription of the gIII -scFv fusion, it was hypothesized that increased e xpression of pIII-scFv fusions would result in higher yields of phagemid particles. Phagemid particles were am plified (see Materials a nd Methods) and titered by the spot-titer method (see Mate rials and Methods) (Table 3-4) Phagemid particles were amplified to titers of 106-108 tu/mL by E. coli TG1 (M13cp-dg3-sm), 105 -106 tu/mL by E. coli TG1 (M13cp-CT-sm), and 108 -109 tu/mL by E. coli TG1 superinfected with commercial hyperphage. Titers of phagemid particles produced with 1 mM IPTG were either the same or lower than titers produced without IPTG. Theref ore, it was determined that the addition of 1 mM IPTG did not improve phagemid particle production. Phagemid particles produced by E. coli TG1 (M13cp-dg3-sm) and E. coli TG1 (M13cp-CT-sm) yielded low titers of phagemid particles. An overnight shaki ng culture for the amplification of phagemid particles contains approximately 109 CFU/mL. Therefore, for the helper pl asmids to yield titers as low as 105 tu/mL means that only 1 in 10,000 bacteria produ ced an infectious phagemid particle. These titers were too low for either helper plasmid to be used as an effective phagemid particle production tool. Even though helper plasmids produced phagemid particle titers too low to be used as an amplification tool, the phag emid particles produced by E. coli TG1 (M13cp-dg3-sm), E. coli TG1 (M13cp-CT-sm), and E. coli TG1 superinfected by hyperphage were further analyzed by ELISA. Phagemid particles were serially diluted and analyzed in a standard ELISA procedure

PAGE 84

84 (see Materials and Methods). Anal ysis of ELISA data showed that S:N values as high as 7 when less than 5 x 104 tu/mL of phagemid particles were used as a primary antibody. Previous data in the laboratory showed that phagemid particles were not detectable in an ELISA unless they were used at concentrations of at least 1 x 106 particles/mL. Spot titering is a measure of transducing units (number of infectious units), while an ELISA is a measure of the number of particles that possess antibody activity. To determine if the measured tu/mL were underestimating the phagemid particles/mL, an anti-M 13 sandwich ELISA was performed to quantify the number of phagemid particles produced by E. coli TG1 (M13cp-dg3-sm), E. coli TG1 (M13cp-CT-sm), and E. coli TG1 superinfected by hyperphage. A microtit er plate was coated with anti-M13 pVIII monoclonal antibody at a concentration of 10 g/mL. Phagemid particles were incubated with the coated wells at dilutions of 1:2, 1:20, or 1: 200. The phagemid particles were then detected using HRP-conjugated anti-M13 monoclonal an tibody at a dilution of 1:2,500, followed by standard ELISA procedures (see Materials and Me thods). A concentrated phagemid preparation, which had been amplified by hyperphage, was seri ally diluted and used as a standard for particles/mL. The values for particles/mL were li nearly extrapolated from the standard curve of the standard phagemid particle (Table 3-4). Phagemid particles amplified by E. coli TG1 (M13cp-CT-sm) had particles/mL to tu/mL ra tios between 740 and 23,000. Therefore, phagemid particles were being produced in high quantities; however, less than 0.1% of the phagemid particles produced were infectious Phagemid particles amplified by E. coli TG1 (M13cp-dg3sm) had particles/mL to tu/mL ratios between 1 and 190. Phagemid particles amplified by E. coli TG1 that was superinfected with hyperphage had particles/mL to tu/mL ratios between 6 and 24. Of the three amplification tools (M13cp-dg3-sm, M13cp-CT-sm, and commercial

PAGE 85

85 hyperphage), commercial hyperphage was the only t ool that could amplify phagemid particles to titers that were high enough to be us ed in phage display applications. Phagemid particles amplifie d by helper plasmids in E. coli TG1 yielded phagemid particle titers between 105 and 108 tu/mL. Helper plasmids amplified phagemid particles to titers too low to be useful in biopanning. To determine if another E. coli strain could improve the titers of helper plasmid-amplified phagemid particles, E. coli MG1655 containing helper plasmid (M13cp-dg3-sm) was used to analyze phagemid particle production. Escherichia coli MG1655 was chosen because it has few known mutations and lacks the F plasmid. Therefore, it was hypothesized that the stress upon the E. coli for maintaining the helper plasmids would be lessened if it did not already have to maintain the F plasmid. Four different phagemid particles were amplified with E. coli MG1655 (M13cp-dg3-sm), and the resulting titers were: 2.5 x 107, 6.3 x 106, 2.5 x 107, and 3.9 x 107 tu/mL for an average of 2.4 x 107 tu/mL, which was still too low to be useful in biopanning. Homemade Hyperphage Titers Were Increas ed with Amplification in E. coli MG1655 (pGTR203) Commercial hyperphage was the only phagemid particle amplif ication tool tested that produced phagemid particle progeny in yields high enough to be used in biopanning. However, the titer of the commercial hyperphage stock was not high enough to be used to amplify the Tomlinson scFv libraries. Therefore, a way to improve homemade hyperphage production was analyzed. Because hyperphage has a partial deletion of its gIII gene, it needs the gIII gene to be provided in trans in the E. coli strain used to produce hyperp hage particles. Homemade hyperphage was previously produced by using pGTR203, a plasmid that encodes gIII. Escherichia coli TG1 (pGTR203) or E. coli HB2151 (pGTR203) were infected with hyperphage, and the transduced bacteria were grown under conditions that promoted hyperphage production.

PAGE 86

86 In the past, hyperphage amplified from E. coli TG1 (pGTR203) yielded titers between 107 to 108 tu/mL and E. coli HB2151 (pGTR203) yielded titers between 106 to 107 tu/mL. Both of these strains are F+ E. coli. Because hyperphage infects bacteria by the F pilus, it was hypothesized that some of the produced hyperphages could have been taken up by the bacteria. Therefore, it was tested if hyperphage could be produced in higher titers when produced in an FE. coli. Two FE. coli strains, E. coli MG1655 and E. coli EC100D, were used to amplify hyperphage. The strains were made electr ocompetent (see Materials and Methods) and transformed by electroporation (see Ma terials and Methods) with pGTR203. Escherichia coli MG1655 (pGTR203) and E. coli EC100D (pGTR203) were made electrocompetent and transformed by electroporation with the hyperphage genome that had been extracted from E. coli TG1 (hyperphage) with a QIAprep Spin Miniprep kit (Qiagen) (see Materials and Methods). Five colonies from each transformation with hyperphage, E. coli MG1655 (pGTR203) (Hy ) 15 and E. coli EC100D (pGTR203) (Hy ) 1-5, were single colony passaged twice and analyzed for their ability to produce hyperphage. Standing overnight cultures of all ten isolates were diluted 1:400 in broth and grown under conditions that promote phage production (s ee Materials and Methods). The hyperphage produced were analyzed for tu/mL by the whole plate titer method (see Materials and Methods) and for particles/mL by an anti-M13 sandwich ELI SA (described above) (T able 3-5). The titers of hyperphage produced from the five E. coli EC100D (pGTR203) (Hy ) isolates 2.1 x 108 2.5 x 108 tu/mL were significantly lower than the hyperphage produced from the five E. coli MG1655 (pGTR203) (Hy ), 3.0 x 109.0 x 109 tu/mL (p=0.02). The ratios of particles/mL to tu/mL of hyperphages produced from E. coli EC100D (pGTR203) (Hy ) were between 0.4 and 7.8, and the ratios of particles/mL to tu/mL of hyperphages produced from E. coli MG1655

PAGE 87

87 (pGTR203) (Hy ) were between 0.3 and 1.0. Therefore th e amount of particles and transducing units of hyperphage produced from E. coli EC100D (pGTR203) (Hy ) and E. coli MG1655 (pGTR203) (Hy ) were approximately the same, meani ng that the hyperphages produced were infectious. Both FE. coli strains generated higher titers of hyperphage than did the two F+ E. coli strains. Concentration of hyperphage produced from E. coli MG1655 (pGTR203) (Hy ) should generate a high enough titer of hyperphage to enable high quality and high quantity production of the Tomlinson scFv libraries. Escherichia coli MG1655 (pGTR203) (Hy ) isolate 5 yielded the highest titer of hyperphage (5.4 x 109 tu/mL); therefore, it was used to pr oduce a large quantity of hyperphage to be used to amplify the Tomlinson J scFv phagemid library. One liter of hyperphage was produced from E. coli MG1655 (pGTR203) (Hy ) 5 with a titer of 2.0 x 109 tu/mL, and the phages were concentrated by PEG precipitation (see Materials and Methods) to a volume of 2.8 mL with a final titer of 3 x 1011 tu/mL. This concentrated hyperphage was used to amplify the Tomlinson J scFv phagemid library. The original Tomlinson J scFv library was amplified to ensure high quality and high quantity of phagemid particles produced. The Tom linson J scFv phagemid library was originally received in E. coli TG1. This stock was grown in batch cu lture and frozen. Therefore, clones were not lost due to poor phagemid particle amp lification with the low titer hyperphage because we grew the original bacterial stock of the library and not the previously made phagemid particle stock of the library that most likely lost redundancy of clones because it was amplified with low titer commercial hyperphage. A frozen stock of E. coli TG1 (Tomlinson J scFv phagemid library) was thawed on ice and inoculated into one liter of 2xTY AG broth. This culture was grown in a 37C shaking incubator until the culture reached log phase. Thirty minutes into the

PAGE 88

88 incubation an aliquot of the grow ing culture was removed and titered to ensure that the library was initially in a concentration high enough that redundancy of clones was present. The titer of the E. coli TG1 (Tomlinson J scFv phagemid library ) 30 minutes into incubation was 1.3 x 107 CFU/mL, and because the culture volum e was one liter there were 1.3 x 1010 CFU. The Tomlinson J phagemid library contains ~1 x 108 phagemid clones; therefore, there was initially a 100-fold redundancy of clones in th e library. Approximately 2.3 x 1010 CFU of log phase E. coli TG1 (Tomlinson J scFv phagemid library) was s uperinfected with homemade hyperphage at a MOI of 10 and incubated for 30 minutes in a 37C standing incubation. The superinfected bacteria were then titered on 2xTY AG plates containing 40 g/mL kanamycin to ensure that enough bacteria were superinf ected so that the redundancy of the phagemid clones was maintained. Approximately 2.9 x 1010 bacteria containing a phagemid were superinfected, which represents about 300-fold redundancy of clones. Th e superinfected bacteria were then grown in 1 liter of broth under conditions that promoted phagemid particle production (see Materials and Methods). The final yield Tomlinson J scFv pha gemid particles in the amplified culture was 6.6 x 109 tu/mL. The liter of phagemid particles was then PEG precipitated to 15 mL of 4.4 x 1011 tu/mL, which yields 6.6 x 1012 tu. Titering throughout the Tomlinson J scFv library amplification ensured that the redun dancy of the library clones was maintained at each step. The Tomlinson J scFv library that was amplified w ith homemade hyperphage was frozen and also used for biopanning. Conclusion of Specific Aim 3 Escherich ia coli TG1 harboring hyperphage was not an optimal phagemid particle amplification tool because of the low infection e fficiency with phagemid particles. The infection efficiencies of phagemid particles into E. coli TG1 (Hy -sm) and E. coli TG1 (Hy -lg) were 28% or less than the infection effi ciency of phagemid particles into E. coli TG1. Phagemid

PAGE 89

89 particle amplification by helper plasmids was not an optimal amplification tool because it produced phagemid particle titers too low to be used in biopanning. Producing homemade hyperphage in E. coli MG1655 (pGTR203) enabled high yield of hyperphage particles. The improved high titer homemade hyperphage was used to successfully amplify the Tomlinson J scFv phagemid library to ensure that the redundancy of phagemid clones was maintained. Specific Aim 4: Isolation of Sp ecific Recombinant Phage to Stx2 Toxin of E. coli O157:H7 The Tom linson J scFv phagemid library th at was amplified with homemade hyperphage was panned against Stx2 toxin of E. coli O157:H7. Shiga toxins (Stx) kill mammalian cells and cause severe disease from infection with E. coli O157:H7 cells. Escherichia coli O157:H7 can produce Shiga toxins 1 and/or 2. Escherichia coli harboring the Stx2 toxin appears to be more virulent than E. coli harboring Stx1 or both Stx1 and 2 toxins (36-39). Therefore, Stx2 toxin was the target molecule chosen for panning. The Stx2 toxin preparation that wa s used to pan against was obtained from Toxin Technolog y (Sarasota, FL) and was stated to contain 50% Stx2 toxin. The other 50% of proteins in the Stx2 toxin preparation was most likely composed of contaminating E. coli HB101 proteins because the Stx2 toxi n preparation was prepared from E. coli HB101 (pMJ100) lysates. p MJ100 encodes Stx2 toxoid. A three round biopanning process was perfor med to pan the Tomlinson J scFv phagemid library against a Stx2 toxin preparation-coated immunotube (see Materials and Methods) with amplification between panning roun ds to amplify the selected pha gemid particles. Phagemid particles were eluted using tr ypsin treatment for 10 minutes. Phagemid particles contain a trypsin cleavage site between the scFv and the p III. Therefore, phagemid particles that were bound to Stx2 toxin preparation by their scFv could be eluted from the Stx2 toxin preparation to be amplified for further rounds of panning. Approximately 1 x 1012 phagemid particles were added to an immunotube th at was coated with 100 g of Stx2 toxin preparation. Approximately

PAGE 90

90 2.2 x 106 phagemid particles were eluted after round one of panning. The round-one-eluted phages were amplified with homemade hyperphage to 2 x 1010 tu/mL in 100 mL of broth. The amplified phagemid particles were then PEG precipitated to 9 x 1011 tu/mL in 2.5 mL PBS. Approximately 1 x 1012 phagemid particles from the PEG-preci pitated phagemid particles from round one were panned against Stx2 for a second round of panning. Approximately 2.0 x 107 phagemid particles were eluted after round two of panning. The round-two-eluted phages were amplified with homemade hyperphage to 2.5 x 1010 tu/mL in 100 mL broth. The amplified phagemid particles were then PEG precipitated to 7.8 x 1011 tu/mL in 2.5 mL PBS. Approximately 1 x 1012 phagemid particles from the PEG-preci pitated phagemid particles from round two were panned against Stx2 for a th ird round of panning. Approximately 1.5 x 1010 phagemid particles were eluted af ter the third round of panning. Th e increase in the number of phagemid particles eluted (2 x 106 for round one, 2 x 107 for round two, 2 x 1010 for round three) after every round suggests that phagemid par ticles that specifica lly bound to Stx2 toxin preparation were being se lected and amplified. Eighty-four phagemid particles from the third round of panning were individually amplified with homemade hyperphage (see Materials and Methods) to be analyzed for specificity by ELISA and Western blot experime nts. Forty-four clones were initially analyzed by ELISA. Microtiter wells coated with 10 g/mL Stx2 toxin preparation were incubated with phagemid particle-containing supernatants that we re diluted 1:2 in blocking buffer, 1 g/mL antiStx211E10 ( -Stx2 subunit A) monoclona l antibody (Toxin Technology), 1 g/mL anti-Stx2BB12 ( -Stx2 subunit B) monoclonal anti body (Toxin Technology), or 1 x 108 tu/mL -BSA phagemid particles (negative cont rol for determining the noise for S:N calculations) (Fig. 3-14). Of the 44 clones screened, 40 had a S:N between 9 and 18. The two anti-Stx toxin monoclonal

PAGE 91

91 antibodies had a S:N less than 2 against Stx2 to xin preparation. The Stx2 toxin preparation may not have bound to the microtiter wells very e fficiently, or the anti-S tx2 monoclonal antibodies may not have worked very well. Conversely, th e phagemid clones that were positive may have bound to the microtiter well instead of the Stx2 toxin preparation. To determine if the clones were binding to Stx2 or to the microt iter well, 4 of the previously screened clones along with 44 new cl ones were analyzed via ELISA using the same procedure as above with the addi tion of PBS-treated microtiter wells as a negative control (Fig. 3-15). Of the 48 clones, 44 had signals of 0.5 or higher against Stx2 toxi n preparation-coated microtiter wells. Four clones (clones 1, 23, 35, a nd 39) had signals as high or higher against PBS-treated microtiter wells compared to Stx2 toxi n preparation-coated microtiter wells. These four clones could possibly recognize polystyrene. Because the library was panned against Stx2 toxin preparation-coated immunot ubes, it was possible that some clones were selected that bound to the polystyrene immunotube. Fifty-three of 59 clones (89%) had a S:N from 2 to 23, and 48 of 59 clones (81%) had a S:N from 10 to 23 agai nst Stx2 toxin preparation. Therefore, the majority of the clones screened bound to the Stx2 toxin preparation. Ho wever, 4/59 (7%) of the clones bound to PBS-treated wells as well as Stx2 -coated wells. The four clones had a signal in an ELISA between 0.31-0.87 agains t PBS-treated wells. These clones most likely bound to polystyrene. Isolation of plastic binding phages is fairly common (84). No matter how well the panning vessel is blocked with blocking buffer, phages can still be selected that bind to the panning vessel. Of the 59 clones screened, 2 (3 %) did not generate a positive signal against Stx2-coated microtiter wells or non-coated microtiter wells. These clones had unknown binding specificity.

PAGE 92

92 The anti-Stx monoclonal antibodies ( -Stx2 subunit A and -Stx2 subunit B) were further analyzed to determine the reason why they did no t react in the ELISA. Microtiter wells coated with 10 g/mL Stx2 toxin preparation or carbonate bicarbonate buffer were incubated with Stx2 subunit A and -Stx2 subunit B monoclonal antibodies at concentrations of 1, 10, or 100 g/mL. Bound antibodies were detected with HR P-conjugated anti-mouse antibody (Cappel) or HRP-conjugated anti-mouse antibody (Sigma-Aldri ch). Standard ELISA procedures followed (see Materials and Methods). All parameters generated S:N values less than two. The Stx2 toxin preparation appeared to be binding to the we ll because the amplified panning clones were binding to the Stx2 with high S:N va lues in microtiter wells. The concentrations of anti-Stx2 monoclonal antibodies were not an issue b ecause using the monoclonal antibodies at a concentration as high as 100 g/mL did not improve the signal of the monoclonal antibodies in an ELISA. Two different HRP-conjugated anti-mous e antibodies were used to determine if the secondary antibody was not recogn izing the primary antibodies. Neither HRP-conjugated antimouse antibody improved the signal of the m onoclonal antibodies in an ELISA. Positive controls for the anti-mouse antibodies were not pe rformed in the ELISA to determine whether or not the anti-mouse antibodies we re functional. However, the two HRP-conjugated anti-mouse antibodies were used within three days of these experiments and generated positive signals in an ELISA against murine antibodies. The reason why the anti-Stx monoclonal antibodies gave low signals with Stx2 toxin preparation in an ELISA was not determined. Analysis by ELISA shows that the majority of clones isolated from the Stx2 panning bound to the Stx2 toxin preparati on in an ELISA. Because the Stx2 toxin preparation was only 50% pure for Stx2 toxin, the panning clones could have been binding to St x2 toxin or the 50% of impurities in the preparation. To analyze if th e Stx2 toxin preparation panning clones recognized

PAGE 93

93 the Stx2 toxin or the impurities, three clones were selected and examined for reactivity with the Stx2 toxin preparation and E. coli supernatants or periplasmic break fractions. The strains examined were E. coli O157:H7 933 (expresses Stx1 and 2 toxins), E. coli DH5 (pNR100) (expresses Stx2 toxoid), and E. coli DH5 Microtiter wells were coated with 0.01 mg/mL Stx2 toxin preparation, 0.07 mg/mL E. coli O157:H7 933 periplasmic break fraction, 2 mg/mL E. coli O157:H7 933 supernatant, 1 mg/mL E. coli DH5 (pNR100) periplasmic break fraction, 1 mg/mL E. coli DH5 periplasmic break fraction, or carbon ate coating buffer. If the clones bound to Stx2 toxin, they were expected to genera te positive signals to a ll the antigens except for E. coli DH5 and carbonate coating buffer. If the clones bound to impurities in the Stx2 toxin preparation, the clones were expected to gene rate positive signals for all the antigens except carbonate coating buffer. Coated wells were incubated with 10 g/mL of -Stx2 subunit A monoclonal antibody (11E10), 10 g/mL of -Stx2 subunit B monoclonal antibody (BB12), 1 x 109 tu/mL of phagemid particles of Stx2 clones (46, 48, and 49), or -BSA phagemids. Standard ELISA procedures were followed (see Materials and Methods) (F ig. 3-16). All parameters generated S:N values of less than 1.5 except fo r clones 46, 48, and 49 that were analyzed with Stx2 toxin preparation-coated microtiter wells. Because the monoclonal antibodies to Stx2 toxin preparation did not generate posit ive signals in an ELISA against any of the parameters, it was not determined if the clones from the Stx2 toxi n preparation panning were reacting to Stx2 toxin or impurities in the commercially obtained Stx2 toxin preparation. Clones that were selected from the Stx2 toxin preparation panning and had a positive signal by ELISA to Stx2 toxin preparation were furt her analyzed by Western blot to determine if the clones were binding to Stx2 toxin or to im purities in the Stx2 toxin preparation. Six clones were screened in a Western blot against Stx2 toxin preparation, E. coli O157:H7 933 (expresses

PAGE 94

94 Stx1 and 2 toxins) supernatant, and E. coli O157:H7 87-23 (Stx phage cu red). Because the Stx2 toxin is composed of A and B subunits that are approximately 32 and 10-kDa proteins, a clone that bound to Stx2 toxin was expect ed to recognize a 32 or 10-kDa protein in the Stx2 toxin preparation and E. coli O157:H7 933 supernatant but not in E. coli O157:H7 87-23 supernatant. A clone that bound to impurities was expected to recognize proteins in all three antigens. A 4-20% polyacrylamide gel was loaded with 1.5 g/well of Stx2 toxin preparation, 120 mg/well of E. coli O157:H7 933 supernatant, and 120 mg/well of E. coli O157:H7 87-23 supernatant. The resolved proteins were transf erred to nitrocellulose and analyzed by Western blot (see Materials and Methods ). Phagemid particles ( -BSA and Stx2 toxin preparation panning clones 46, 48, 80, 81, and 83) were used as a primary antibody at approximately 109 tu/mL, and -Stx2 subunit A monoclonal antibody was used at 4 g/mL. Anti-Stx2 subunit B monoclonal antibody failed to generate a positive si gnal by Western blot. Standard Western blot procedures were performed (see Mate rials and Methods) (Fig. 3-17). The -Stx2 subunit A monoclonal antibody recognized a 32 -kDa (the size of the A subunit) protein in the Stx2 toxin preparation and in the E. coli O157:H7 933 supernatant, there by proving that Stx2 toxin subunit A was present in both preparations but not in the E. coli O157:H7 87-23 supernatant as was expected. Anti-BSA failed to recognize any of th e antigens, as was expected. Clone 46 failed to recognize any of the three antig ens. Clones 48 recognized 35, 45, and 55-kDa proteins in the Stx2 toxin preparation and a 35-kDa protein in both E. coli O157:H7 supernatants. This result means that clone 48 probably bound to bacterial protein impurities in the Stx2 toxin preparation. Clones 49 and 80 bound to a 35-kDa protein in all th ree antigens and a 20-kDa protein in the Stx2 toxin preparation and in E. coli O157:H7 87-23 supernatant. Therefore, clones 49 and 80 probably bound to bacterial protein impurities in the Stx2 toxin preparation. Clones 81 and 83

PAGE 95

95 appeared to recognize the same bacterial protein impurities as cl ones 49 and 80, except that their band intensities were much lighter. Western blot and ELISA analyses showed that clones obtained from panning against the Stx2 toxin preparation recognized the Stx2 toxin preparation. The clones that have been screened so far appeared to bind to impurities in the Stx2 toxin preparation. To determine if the Stx2 toxin preparation contained 50% Stx2 toxin as stated by th e producer, the toxin preparation was resolved and analyzed by SDS-PAGE (Fi g. 3-18). If the Stx2 toxin preparation was comprised of 50% Stx2 toxin, 10 and 32-kDa prot ein bands should have been the predominant protein bands. However, a protein band at a pproximately 32 kDa was a minor protein band. A protein band at 10 kDa was a major protein ba nd; however, whether thes e protein bands were actually the Stx2 toxin subunits was not determined. The 10 and 32-kDa protein bands comprised less than 10% of the Stx2 toxin prepar ation. Because the concentration of Stx2 toxin was so low in the commercially obtained St x2 toxin preparation, it was probable that the majority of the phagemid particles isolated from the Stx2 panning would recognize impurities instead of Stx2 toxin. Conclusion of Specific Aim 4 Panning against comm ercially obtained Stx2 toxi n preparation was successful in isolating phagemid particles that bound to the Stx2 toxin pr eparation. Initial sc reening of clones by Western blot showed that the clones bound to impur ities in the commercially obtained Stx2 toxin preparation and not Stx2 toxin. The commercially obtained Stx2 toxin preparation was supposed to contain 50% Stx2 toxin; however SDS-PAGE analysis showed th at less no more than 10% of the proteins present in the prepar ation may have been Stx2 toxin. Therefore, it was likely that the majority of clones selected from the panni ng would recognize impurities in the Stx2 toxin preparation and not Stx2 toxin. However, this panning exercise showed that the Tomlinson J

PAGE 96

96 library amplified with the homemade hyperphage was a more effective tool than the previously amplified libraries.

PAGE 97

97 Table 3-1. Infection efficiencies of E. coli TG1 containing various helper plasmids with phagemid particles. Bacterial strains were transduced with anti-BSA ( -BSA) and Vc86 phagemid particles at a MOI of 0.1, and transduced bacteria were enumerated. Infection efficiency is the number of trans ductions into a bacterial strain divided by the number of transductions into E. coli TG1 and is represented as a percentage. Strain Transductions with -BSA Infection efficiency with -BSA (%) Transductions with Vc86 Infection efficiency with Vc86 (%) Average infection efficiency (%) E. coli TG1 1.2 x 106 1003.6 x 106 100 100 E. coli TG1 (M13cp-sm) 9.3 x 102 0.16.9 x 103 0.3 0.2 E. coli TG1 (M13cp-lg) 8.3 x 102 0.19.6 x 103 0.3 0.2 E. coli TG1 (M13cp-CT-sm) 7.9 x 104 74.8 x 105 13 10 E. coli TG1 (M13cp-CT-lg) <1.3 x 102 <0.01<1.3 x 102<0.01 <0.01 E. coli TG1 (M13cp-dg3-sm) 2.1 x 106 1753.6 x 106 97 136 E. coli TG1 (M13cp-dg3-lg) <1.3 x 102 <0.01<1.3 x 102<0.01 <0.01 Table 3-2. Infecti on efficiencies of E. coli TG1 (M13cp-dg3-sm) with phagemid particles. Bacterial strains were transduced with phagemid particles at a MOI of 0.1, and transduced bacteria were enumerated. Infection efficiency is the number of transductions with a phagemid into a b acterial strain divided by the number of transductions with the same phagemid into E. coli TG1. Strain Phagemid Transductions Infection efficiency (%) E. coli TG1 Vc86 5.3 x 106 100 E. coli TG1 (M13cp-dg3-sm) Vc86 4.9 x 106 92 E. coli TG1 Clone 18 2.4 x 106 100 E. coli TG1 (M13cp-dg3-sm) Clone 18 2.7 x 106 113 E. coli TG1 -AV20N3 1.5 x 106 100 E. coli TG1 (M13cp-dg3-sm) -AV20N3 1.8 x 106 120

PAGE 98

98 Table 3-3. Infecti on efficiencies of E. coli TG1 (M13cp-CT-sm) and E. coli JM109 (M13cp-CTsm) with phagemid particles. Bacteria l strains were transduced with phagemid particles at a MOI of 0.1, and transduced bacteria were enumerated. Infection efficiency is the number of transductions with a phagemid particle into a helper plasmid-containing bacterial strain divide d by the number of transductions with the same phagemid particle into the bacter ial strain without the helper plasmid. Strain Phagemid Transductions Infection efficiency (%) E. coli TG1 Vc86 2.4 x 105 100 E. coli TG1 (M13cp-CT-sm) Vc86 1.3 x 105 54 E. coli TG1 Clone 18 1.4 x 105 100 E. coli TG1 (M13cp-CT-sm) Clone 18 5.5 x 104 39 E. coli TG1 -AV20N3 9.3 x 104 100 E. coli TG1 (M13cp-CT-sm) -AV20N3 4.4 x 104 47 E. coli JM109 Vc86 6.8 x 105 100 E. coli JM109 (M13cp-CT-sm) Vc86 1.5 x 105 22 E. coli JM109 Clone 18 1.7 x 105 100 E. coli JM109 (M13cp-CT-sm) Clone 18 6.1 x 104 36 E. coli JM109 -AV20N3 1.4 x 105 100 E. coli JM109 (M13cp-CT-sm) -AV20N3 3.7 x 104 26 Table 3-4. Comparison of transducing units to particles per milliliter of amplified phagemid particles. The spot-titer method was used to determine tu/mL. Particles/mL was determined via a sandwich ELISA with anti-M13 antibodies. (n = 2 wells) Phagemid particle amplified Amplification tool used tu/mL Particles/mL (Particles/mL)/ (tu/mL) Vc86 M13cp-CT-sm 8.7 x 105 2.0 x 1010 23000 Vc86 M13cp-CT-sm 4.6 x 106 3.4 x 109 740 clone 18 M13cp-dg3-sm 2.4 x 108 2.2 x 108 1 -AV20N3 M13cp-dg3-sm 4.4 x 107 7.0 x 107 2 -BSA M13cp-dg3-sm 1.3 x 108 1.8 x 108 1 Vc86 M13cp-dg3-sm 8.7 x 106 1.7 x 109 190 clone 18 Hyperphage 7.1 x 108 1.1 x 1010 15 -AV20N3 Hyperphage 2.0 x 109 1.1 x 1010 6 -BSA Hyperphage 5.8 x 108 1.4 x 1010 24

PAGE 99

99 Table 3-5. Comparison of transducing units to particles per milliliter of amplified homemade hyperphage produced from FE. coli strains. The spot-titer method was used to determine tu/mL. Particles/mL was dete rmined via a sandwich ELISA with anti-M13 antibodies. (n = 2 wells). Hyperphage production strain tu/mL Particles/mL (Particles/mL)/ (tu/mL) E.coli EC100D (pGTR203) (Hy ) 1 5.3 x 107 4.2 x 108 7.8 E.coli EC100D (pGTR203) (Hy ) 2 9.4 x 107 3.5 x 108 3.7 E.coli EC100D (pGTR203) (Hy ) 3 3.2 x 108 1.1 x 108 0.4 E.coli EC100D (pGTR203) (Hy ) 4 2.9 x 106 < 1.0 x 107 <3 E.coli EC100D (pGTR203) (Hy ) 5 6.0 x 108 2.1 x 108 0.4 E.coli MG1655 (pGTR203) (Hy ) 1 1.0 x 109 1.0 x 109 1.0 E.coli MG1655 (pGTR203) (Hy ) 2 8.0 x 108 7.3 x 108 0.9 E.coli MG1655 (pGTR203) (Hy ) 3 3.0 x 109 1.3 x 109 0.4 E.coli MG1655 (pGTR203) (Hy ) 4 4.3 x 109 1.4 x 109 0.3 E.coli MG1655 (pGTR203) (Hy ) 5 5.4 x 109 1.5 x 109 0.3

PAGE 100

100 Figure 3-1. Analysis of TRIzol Reagent-extracted V. cholerae N16961 LPS by SDS-PAGE. Samples were resolved on a 12% (w/v) pol yacrylamide gel and stained with A) Coomassie blue stain or B) Tsai-Frasc h silver stain. (1) TRIzol-extracted V. cholerae N16961 LPS. (2) V. cholerae 569B LPS (Sigma-Aldrich).

PAGE 101

101 Figure 3-2. Analysis of phenol-water-extracted V. cholerae 569B LPS by SDS-PAGE. Samples were resolved on a 4-20% (w/v) polyacrylamide gel and stained with A) Coomassie blue stain or B) Tsai-F rasch silver stain. (1-3) V. cholerae 569B LPS (SigmaAldrich) at dilutions of 1:4 (1), 1:8 (2), and 1:16 (3). (4-6) Phenol-water-extracted V. cholerae 569B LPS at dilutions of 1:2 (4), 1:4 (5), and 1:8 (6).

PAGE 102

102 Figure 3-3. Analysis of the saturation limit of V. cholerae 569B LPS to microtiter wells by ELISA. Signal to noise (S:N) ratios were calculated for the react ivity of mouse antiV. cholerae O1 LPS monoclonal antibody ( -Vc O1 LPS mAb) (1 g/mL) and rabbit antiV. cholerae O1 LPS polyclonal antibody ( -Vc O1 LPS pAb) (1:100) to V. cholerae 569B LPS (0.1-200 g/mL)-coated microtiter wells. Bound primary antibodies were detected with either HRP-conjugate d goat anti-rabbit antibody or HRP-conjugated goat anti-mouse antibody. Wells were developed with TMB substrate, and their absorbances read at 630 nm. (n=3 wells). There was a significant decrease (*p=0.01) in S:N when the concentration of LPS decreased from 1 to 0.1 g/mL when using the -Vc O1 LPS mAb. There was a significant decrease (*p=0.01) in the S:N when the concentration of LPS decreased from 10 to 1 g/mL when using the -Vc O1 LPS pAb. 0 2 4 6 8 10 0.11101001000 V. cholerae LPS ( g/mL)S:N -Vc O1 LPS mAb (1 g/mL) -Vc O1 LPS pAb (1:100)**

PAGE 103

103 A Figure 3-4. Analysis of the saturati on limits of primary antibodies to V. cholerae 569B LPScoated microtiter wells by ELISA. Primary antibodies: A) mouse antiV. cholerae O1 LPS monoclonal antibody ( -Vc O1 LPS mAb) (0.04-1.3 g/mL) and B) rabbit antiV. cholerae O1 LPS polyclonal antibody ( -Vc O1 LPS pAb) (1:100-1:1,600) were reacted with V. cholerae 569B LPS (1 g/mL)-coated microtiter wells. Bound primary antibodies were detected with either HRP-conjugated goat anti-rabbit antibody or HRP-conjugated goat anti-mouse antibody. Wells were developed with TMB substrate and their absorbances read at 630 nm. The S:N were calculated for every primary antibody concentrat ion or dilution (n=3 wells). In (A) comparisons of S:N between every monoclonal antibody con centration used yi elded significant differences (p<0.03) in S:N; therefore, th e saturation limit of the monoclonal antibody could not be determined. In (B) a significant decrease (*p=0.01) in the S:N was not reached until the polyclonal antibody dilu tion decreased from 1:400 (ratio dilution of 0.003) to 1:800(ratio dilution of 0.001); therefore, the satu ration limit of the polyclonal antibody was reached at a dilution of 1:400. 0 2 4 6 8 10 12 0.00.51.01.5 -Vc O1 LPS mAb (1 g/mL)S:N

PAGE 104

104 Ratio of Dilution, Dilution 0.01, 1:100 0.005, 1:200 0.003, 1:400 0.001, 1:800 <0.001, 1:1,600 B Figure 3-4. Continued. 0 2 4 6 8 10 12 00.0020.0040.0060.0080.01 Ratio of DilutionS:N *

PAGE 105

105 Figure 3-5. Saturation of V. cholerae 569B LPS to nitrocellulose paper. Vibrio cholerae 569B LPS was incubated with nitrocellulose paper at concentrations from 1-100 g/mL. The primary antibody used to detect the LPS was rabbit antiV. cholerae O1 LPS polyclonal antibody (1:400). Bound prim ary antibody was detected with HRPconjugated goat anti-rabbit antiserum and detected with ECL substrate. The nitrocellulose paper became saturated with V. cholerae 569B LPS at a LPS concentration of 100 g/mL. The strips incubated with 0 g/mL LPS failed to stain significantly.

PAGE 106

106 0 10 20 30 40 50 60 70 80 90 100 15304560 Trypsin Elution (min)Phages Eluted /107 (tu/mL) Figure 3-6. Titers of eluted phagemid partic les from the first one-round panning optimization experiment. The Tomlinson J library +/ 1% Vc86 phagemid particles was panned against 108 and 109 whole cells of V. cholerae N16961 in suspension. Phagemid particles that were bound to the V. cholerae whole cells were eluted with 10% (v/v) trypsin stock (10 mg/mL trypsin, 50 mM Tris-HCl (pH 7.4), 1 mM CaCl2 in water) in PBS for 15, 30, 45, and 60 minutes. Eluted pha gemid particles were enumerated via the spot-titer method. 108 cells Vc 109 cells Vc 108 cells Vc, 1% Vc86 109 cells Vc, 1% Vc86

PAGE 107

107 A B Figure 3-7. Titers of eluted phagemid partic les from the second one-round panning optimization experiment. The Tomlinson J library +/ 1% Vc86 phagemid particles was panned against 107, 108, and 109 whole cells of V. cholerae N16961 in suspension. Phagemid particles that were bound to the V. cholerae whole cells were eluted in 10% (v/v) trypsin stock (10 mg/mL trypsin, 50 mM Tris-HCl (pH 7.4), 1 mM CaCl2 in water) in PBS for 10, 20, and 30 minutes. Eluted phage mid particles were enumerated via the spot-titer method. A) Panning with the Tomlinson J library B) Panning with the Tomlinson J library (+) 1% Vc86 phagemid particles. 0 20 40 60 80 100 120 1015202530 Trypsin Elution (min)Phages Eluted/107 (tu/mL) 0 20 40 60 80 100 120 102030 Trypsin Elution (min)Phages Eluted/106 (tu/mL) 107 cells Vc 108 cells Vc 109 cells Vc 107 cells Vc 108 cells Vc 109 cells Vc

PAGE 108

108 0 2 4 6 8 10 12 010203040 Trypsin Elution (min)Phages Eluted/107 (tu/mL) Figure 3-8. Titers of eluted phagemid partic les from the third one-round panning optimization experiment. The Tomlinson J library +/ 1% Vc86 phagemid particles was panned against 108 and 109 whole cells of V. cholerae N16961 in suspension. Phagemid particles that were bound to the V. cholerae whole cells were eluted in 10% (v/v) trypsin stock (10 mg/mL trypsin, 50 mM Tris-HCl (pH 7.4), 1 mM CaCl2 in water) in PBS for 10 and 30 minutes. Eluted phagemid particles were enumer ated via the spottiter method. 10 8 cells Vc 109 cells Vc 108 cells Vc, 1% Vc86 109 cells Vc, 1% Vc86

PAGE 109

109 Figure 3-9. Analysis of eluted phagemid par ticles from the third one-round panning optimization experiment by ELISA. Microtit er wells coated with 3 x 108 CFU/mL V. cholerae whole cells or PBS and we re incubated with 1 x 108 tu/mL of amplified eluted phagemid particles or 10 g/mL of mouse antiV. cholerae O1 LPS monoclonal antibody. Bound phagemid particles were de tected with HRP-conjugated anti-M13 secondary antibody, and the monoclonal anti body was detected with HRP-conjugated goat anti-mouse antibody. Wells were deve loped with TMB substrate, and their absorbances were read at 630 nm (n=3 we lls). There was a si gnificantly higher S:N (*p=0.01-0.04) from the amplified eluted phagemid particles from the panning with 108 whole cells with a 30 minute elution us ing the library containing 1% Vc86 when compared to the amplified eluted phagem id particles from the other pannings. 0 2 4 6 8 10 10 min 10 min +Vc86 30 min 30 min +Vc86 10 min 10 min +Vc86 30 min 30 min +Vc86 -Vc O1 LPS mAb Vc86 S:N 10 8 10 8 10 8 10 8 10 9 10 9 10 9 10 9

PAGE 110

110 Figure 3-10. Analysis of anti-BSA ( -BSA) scFv proteins produced in microtiter wells by ELISA. Microtiter wells were coated with 10 g/mL BSA or PBS. Anti-BSA scFv particles were produced in a microtiter well and isolated by centrifugation and the scFv-containing supernatants were diluted 1:2, 1:4, 1:6, or 1:8 in blocking buffer and incubated with the coated wells. Bound sc Fv particles were detected with HRPconjugated Protein L. We lls were developed with TMB substrate, and their absorbances were read at 630 nm. (n=2 wells). 0 2 4 6 8 10 12 14 1:21:41:61:10 Dilution of -BSA scFv Supernatant in Blocking BufferS:N

PAGE 111

111 Figure 3-11. Analysis of -BSA and Vc86 phagemid particles produced in microtiter wells by ELISA. Phagemid particles Vc86 (V) and -BSA (B) were produced at either 30C or 37C in medium containing either 0% or 0.1% glucose in a microtiter well (W) or in a flask (F). Microtiter wells were coated with 10 g/mL BSA, 3 x 108 CFU/mL V. cholerae whole cells, or PBS. Phagemid particles were dilu ted 1:2 in blocking buffer and incubated with the coated wells Bound phagemid particles were detected with HRP-conjugated anti-M13 secondary antibody. Wells were developed with TMB substrate, and their ab sorbances were read at 630 nm. (*) represents p<0.05 and (**) represents p<0.01. (n=2 wells) Phagemid Particle V V V V V B B B B B Temp. (C) 30 30 37 37 30 30 30 37 37 30 % Glucose 0 0.1 0 0.1 0.1 0 0.1 0 0.1 0.1 Grown in W W W W F W W W W F 0 2 4 6 8 10 12 14S:N ** **

PAGE 112

112 Figure 3-12. Analysis of -BSA phagemid particles produced in microtiter wells by ELISA. Microtiter wells were coated with 10 g/mL BSA. Phagemid particles were diluted 1:2 in blocking buffer and incubated w ith the coated wells. Bound phagemid particles were detected with HRP-conj ugated anti-M13 secondary antibody. Wells were developed with TMB substrate, and th eir absorbances were read at 630 nm (n=2 wells). Phagemid particles produced in a microtiter well at 30C produced significantly higher (*p=0.03) S:N values th an phagemid particles produced in a microtiter well at 37C at the same glucose concentration. 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.830C, 0% glucose 30C, 0.1% glucose 37C, 0% glucose 37C, 0.1% glucose 30C, 0.1% glucose, in flask -BSA Phagemid ParticlesSignal

PAGE 113

113 Figure 3-13. Analysis of anti-BSA ( -BSA) phagemid particles produced by hyperphage by ELISA. Microtiter wells were coated with 10 g/mL BSA or PBS and incubated with 1 x 108 tu/mL -BSA phagemid particles produced from E. coli TG1 strains already infected with hyperphage (Hy -sm, Hy -lg) or from E. coli TG1 that was superinfected with hyperphage. Bound phagemid particles were detected with HRPconjugated anti-M13 secondary antibody. Wells were developed with TMB substrate, and their absorbances were read at 630 nm (n=3 wells). The S:N from BSA phagemid particles produced from (Hy -sm) was significantly lower (*p<0.01) than the S:N from -BSA phagemid particles produced from (Hy -lg) or from superinfection with hyperphage. 0 5 10 15 20 Superinfected with Hy (Hy -sm)(Hy -lg) a-BSA Phagemid ParticlesS:N

PAGE 114

114 Figure 3-14. Analysis of phagemid particles (1-22, 43-64) selected from panning against Stx2 toxin preparation by ELISA. Microt iter wells were coated with 10 g/mL of Stx2 toxin preparation (Toxin Technologies) a nd incubated with cl ones 1 through 22 and 43 through 64 diluted 1:2 in blocking buffer, 1 g/mL anti-Stx2 subunit A monoclonal antibody (11E10), 1 g/mL anti-Stx2 subunit B monoclonal antibody (BB12), or anti-BSA phagemid particles dilu ted 1:2 in blocking buffer. Microtiter wells coated with 10 g/mL BSA were incubated with anti-BSA phagemid particles that were diluted 1:2 in blocking buffer. Bound phagemid particles were detected with HRP-conjugated anti-M13 monoclona l antibody. Bound monoclonal antibodies were detected with HRP-conjugated goa t affinity purified anti-mouse antibody. Wells were developed with TMB substrate, and their absorbances were read at 630 nm (n=2 wells). The noise used to calculate S:N was the signal of anti-BSA phagemid particles that were incubated with Stx2-coated wells. 0 5 10 15 20 251 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 0 2 1 2 2 4 3 4 4 4 5 4 6 4 7 4 8 4 9 5 0 5 1 5 2 5 3 5 4 5 5 5 6 5 7 5 8 5 9 6 0 6 1 6 2 6 3 6 4 B B 1 2 1 1 E 1 0 a B SAS:N

PAGE 115

115 Figure 3-15. Analysis of phagemid particles (1, 3-5, 23-42, 65-84) selected from panning against Stx2 toxin preparation by ELISA. Micr otiter wells were coated with 10 g/mL of Stx2 toxin or carbonate coating buffer a nd incubated with clones 1, 3 through 5, 23 through 42, and 65 through 84 diluted 1:2 in blocking buffer. Bound phagemid particles were detected with HRP-conj ugated anti-M13 monoclonal antibody. Wells were developed with TMB substrate, and th eir absorbances were read at 630 nm (n=2 wells). 0 5 10 15 20 251 3 4 5 2 3 2 4 2 5 2 6 2 7 2 8 2 9 3 0 3 1 3 2 3 3 3 4 3 5 3 6 3 7 3 8 3 9 4 0 4 1 4 2 6 5 6 6 6 7 6 8 6 9 7 0 7 1 7 2 7 3 7 4 7 5 7 6 7 7 7 8 7 9 8 0 8 1 8 2 8 3 8 4S:N

PAGE 116

116 Figure 3-16. Analysis of anti-Stx2 monoclona l antibodies and clones 46, 48, and 49 by ELISA. Microtiter wells were coated with 10 g/mL of Stx2 toxin, 70 g/mL of E. coli O157:H7 933 periplasmic break fraction, 2 mg/mL of E. coli O157:H7 933 supernatant (O157 Sup.), 1mg/mL of E. coli DH5 (pNR100) periplasmic break fraction, and 1mg/mL of E. coli DH5 periplasmic break fraction. (PP) represents periplasmic break fractions. Bound phagemi d particles were detected with HRPconjugated anti-M13 monoclonal antibody. Bound monoclonal antibodies were detected with HRP-conjugate d goat affinity purified an ti-mouse antibody. Wells were developed with TMB substrate, and th eir absorbances were read at 630 nm (n=2 wells). mAb 11E10 mAb BB12 clone 46 clone 48 clone 49 -BSA DH5 PP DH5 (pNR100) PP O157 Sup. O157 PP Stx2 0 2 4 6 8 10 S:N

PAGE 117

117 Figure 3-17. Western blot analys is of phagemid clones from pa nning on Stx2 toxin preparation. Samples were resolved by SDS-PAGE on a 4-20% (w/v) polyacrylamide gel and analyzed by Western blot. Samples: (T) -1.5 g/well of Stx2 toxin, (+) 120 g/well E. coli O157:H7 933 (contains Stx1 and 2 t oxins) supernatant, and (-) -120 g/well E. coli O157:H7 87-23 (does not produce Stx toxi n) supernatant. Primary antibodies: Anti-Stx2 toxin subunit A monoclonal an tibody (Mab); Stx2 phagemid clones 46, 48, 80, 81, 83. A) Western blot was developed with TMB and exposed for 30 seconds. B) Western blot was developed with TMB and exposed for 5 minutes.

PAGE 118

118 Figure 3-18. Analysis of Stx2 toxin prepara tion (Toxin Technologies) by SDS-PAGE. Five micrograms of Stx2 toxin preparation was resolved on a 4-20% (w/v) polyacrylamide gel and stained with Coomassie blue stain. Stx2 toxin is comprised of a 32-kDa A subunit and five 10-kDa B subunits.

PAGE 119

119 CHAPTER 4 DISCUSSION Detection of biological agents and their produ cts has becom e a prime importance in the United States (85) since the anth rax attacks of 2001 (86). Early de tection of biological agents is important to prevent or inhibit th e spread of disease. Detecting biological agents in water, air, and food enables the proper course of action to be implemented to lessen or prevent the effects of the organism. Detection of biologi cal agents or their products is al so important for diagnosis of a disease. Detection of the disease-causing agent in patient samples enab les the proper treatment to given. There are vast numbers of detection assays, a nd a common component to most of them is a protein that binds specifically to a target antigen. Monoclonal anti bodies are the default tool that scientists use today to detect target molecules; however, monoclonal antibodies have some drawbacks that are causing the use of phage displa y reagents to become more prevalent and take the place of monoclonal antibodies (2). First, the genera tion of monoclonal antibodies requires killing animals and takes months. Also, monoclona l antibodies are rarely isolated to pathogenic or self antigens, causing limitations in the scope of the target antigen for the generation of monoclonal antibodies. Phage display is advant ageous because it lacks many of the drawbacks of monoclonal antibodies. In phage display, if using a non-immunized library, the killing of animals is not required and the isolation of sp ecific recombinant phages can be done in a couple weeks instead of months. Also, the scope of anti gen used for phage display is far greater than that used for monoclonal antibodies because phage display enables the se lection of recombinant proteins to pathogenic and self antigens. Ph age display was developed in 1985 by George P. Smith as a means to display foreign proteins on a filamentous phage (52). The sequence for the foreign peptide is encoded in frame with a coat protein gene of the phage genome. This enables

PAGE 120

120 the foreign peptide to be fused to the coat pr otein, which effectively links the phenotype and genotype of the phage. Therefore, every phage that displays a fo reign protein also encodes its peptide sequence in the genome. This thesis describes efforts to optimize phage display methods to increase the chances of isolating recombinant antibodies or peptides for the potential use in detection assays. Specific Aim 1: Panning to V. cholerae LPS Vibrio cholerae is a hum an pathogen that is the causa tive agent of cholera. Cholera is a diarrheal disease that affects millions of people worldwide annually (9). Detecting V. cholerae in the environment and patient samples is the first step to preventing and treating the disease. To detect V. cholerae we need to isolate a recombinant phage that binds specifically and sensitively to V. cholerae. There are over 200 serogroups of V. cholerae that are clas sified on the basis of the O-antigen of their LPS; however, only se rogroups O1 and O139 cause endemic acute diarrhea. Therefore, the LPS of these two serogr oups would be excellent targets for detection. Vibrio cholerae O1 LPS was chosen as a target for biopanning. Because V. cholerae O1 LPS was no longer sold commercially, it had to be extracted in the laboratory before any panning experiments could be performed. The previous method to extract LPS in the laboratory was a phenol-water extraction (76). However, a new LPS extraction was developed using a commercial RNA isolation reagent, TRIzol Reagent, that could extract LPS in a fraction of the time of the phenol-water extraction (77). The TRIzol method was even claimed to be able to extract LPS with less degradation and contamination than the phenol-water method. To determine which method was better, both methods were used to extract V. cholerae O1 LPS and the extracted LPS was analyzed and compared to commercially acquired V. cholerae O1 LPS (Sigma-Aldrich) that had been phenol extracted and purified by gel-filtration chromatography.

PAGE 121

121 The criteria for the optimal LPS extraction we re to have little or no protein and lipid contamination and to have similar banding patterns to the commercial V. cholerae O1 LPS. Vibrio cholerae O1 LPS was extracted using the phenol -water and TRIzol methods. The LPS was resolved by SDS-PAGE and analyzed by Coomassie blue staining and Tsai-Frasch silver staining (Figs. 3-1 and 3-2). Vibrio cholerae O1 LPS that was extracted by the TRIzol reagent had less protein contamination than did the LPS that was extracted by the phenol-water method. The relative purity of the TR Izol-extracted LPS was less than 0.3 g of protein/ g of LPS, while the relative purity of the phenol -water-extracted LPS was approximately 1 g of protein/ g of LPS. However, the protein contamin ation in the phenol-water LPS extraction was most likely due to the RNase that was used in th e extraction. The Coomassie blue stain for the phenol-water-extracted LPS show ed a protein band at approxim ately 15 kDa. The molecular weight of RNase is 13.7 kDa. The phenol-w ater LPS extraction contained 0.04 mg/mL of RNase, which correlates to the band intensity of the 15-kDa protein band. Because the phenolwater LPS extraction only cont ained approximately 0.02 mg/mL of DNase, MW of 31kDa, the DNase might not have been concen trated enough to be seen by Coom assie blue stain. Therefore, both LPS extractions probably had less than 0.3 g of bacterial protein contamination for every g of LPS extracted. The TRIzol extracted LPS showed non-staining bands between 20 to 30 kDa on the silver stain (Fig. 3-1B). The Tsai-Fra sch silver stain primarily stains carbohydrates and poorly stains proteins and lipids (78). Because the Coomassie blue stain for the TRIzolextracted LPS showed no detectable proteins a nd due to the thickness of the non-staining bands, which correlates to concentr ation, the non-staining bands ar e most likely due to lipid contamination. Previously published silver stains of V. cholerae O1 LPS had bands at ~10 and 14 kDa which were stated to be the lipid A-core of LPS, and ~20-50 kDa which were stated to be

PAGE 122

122 the lipid A-core plus repeating O-antigens (80,81). The non-stai ning bands were not the lipid A because the lipid A contains two glucosamines that are stainable by silver stain. Therefore, the TRIzol-extraction appears to have also extracted cellular lipids. Therefore, the TRIzol-extracted LPS appeared to have more lipid contamin ation than the phenol-water-extracted LPS. When the resolved LPS extractions were comp ared by silver stain to the commercial LPS, the phenol-water-extracted LPS most closely resembled the commercial LPS. In the silver stain for TRIzol-extracted LPS and commercial LPS, the TRIzol-extracted LPS had bands at 14 and 20 kDa with non-staining bands between 20 and 30 kDa, the commercial LPS had bands at 14, 20, and 23 kDa. In the silver stain for phenol -water-extracted LPS and commercial LPS, both LPS preparations had bands at 13, 20, and 23 kDa. The phenol-water-extracted LPS had additional banding from 23-50 kDa, while the co mmercial LPS had more of a smear from 23-50 kDa. Previously published silver stains of V. cholerae O1 LPS have O-antigen banding/smears from 20-50 kDa (80,81). Vibrio cholerae O1 LPS has 12 to 18 O-antigen groups (83); therefore, the additional banding in the phenol -water-extracted LPS may be due to the ability of the phenolwater method to extract LPS with more O-antige ns than that of the method used for the extraction of the commercial LPS which was phenol extracted and purified by gel-filtration chromatography. Because the phenol-water-ext racted LPS most closely resembled the commercial LPS and appeared to have less lipid contamination than the TRIzol-extracted LPS, the phenol-water-extracted LPS was used for later experiments. Before V. cholerae O1 LPS was panned, a solid support capable of binding a high quantity of V. cholerae O1 LPS was analyzed. In the past th e Ph.D. 12mer phage library was panned against antigens that were immobilized onto a microtiter well. Previous experiments from this laboratory suggested that nitroce llulose paper could bind more LPS than a microtiter well. It was

PAGE 123

123 hypothesized that having more LPS present to bind phages would enable more phages that specifically bound to the LPS to be selected. However, increasing the amount of LPS present may also increase the amount of nonspecific phage s selected as well as the amount of specific phages selected. Therefore, more antigen presen t to pan against may not necessarily improve the chances of selecting phages specific to the antigen. To determine which solid support could bind more LPS, the saturation limit of LPS was analyzed on a microtiter well and a piece of nitroce llulose paper with the same surface area as a microtiter well. The microtiter well became saturated with LPS when coated with 0.1-1 g LPS, and the nitrocellulose paper became satu rated with LPS when coated with ~250 g LPS. Therefore, nitrocellulose paper could bind approximately 250-2,500 times more LPS than a microtiter well of equal area. Nitrocellulose pa per has previously been used as a solid support for proteins in panning proce dures involving glutathione S-tr ansferase and hen egg white lysozyme (87). Nitrocellulose paper was chosen as a binding support because it had a high capacity to bind proteins (~100 g/cm2). To determine if panning against more antigen increases the sel ection of phages specific for the antigen, both microtiter well s and nitrocellulose paper were used as solid supports for panning against V. cholerae O1 LPS. Vibrio cholerae O1 LPS (Sigma-Aldrich) was immobilized onto a polystyrene Maxisorp microtiter well and panned with the NEB Ph.D. 12mer phage display library. Five rounds of panning were performed with amplification of phages after the first four rounds of panning. Panning promotes the selection of phages that specifically bind to the target antigen over phage s that are not specific to the an tigen. Amplifying pools of eluted phages between pannings enriches the amount of specific phages. Therefore, each additional round of panning should increase the amount of specific phages in the pool of eluted phages.

PAGE 124

124 One hundred phages from the fifth round of panning were individually amplified and analyzed in an ELISA against V. cholerae O1 LPS, carbonate coating buffer, and PBS. Anti-Vc O1 LPS mAb generated a S:N of 15.5, proving that the LPS successfully bound to the microtiter plate. Anti-BSA phagemid pa rticles were used in the ELISA against BSA (10 g/mL) and generated a S:N of 22, proving that the ELISA successfully detected phages. All 100 clones gave S:N <1.4, whereas a positive signal was classified as a S:N 2. Therefore, all of the clones screened by ELISA were negative for binding to V. cholerae 569B LPS. Because panning the Ph.D. 12mer library against V. cholerae O1 LPS that was immobilized onto a microtiter well failed to yield phages that were positive to V. cholerae O1 LPS, the panning procedure was altered in an attempt to improve the chances of selecting phages that specifically bind to the antigen. The first alteration was to change the antigen binding support from a microtiter well to nitrocellulose. It was determined that nitrocellulose paper could bind more LPS than a microtiter well; ther efore, with more LPS present to pan against more specific phages may be selected. Also, the mechanism of phage elution was analyzed. Previous panning procedures with the Ph.D. library used a glycine (pH 2. 2) elution. There are numerous elution methods, pH, salt, pressure, te mperature, antigen, none of which are optimal for every antigen-antibody complex. The key is to find a method that effectively dissociates the antigen-antibody complex without causing harm to either the antibody or the antigen (88). An acidic elution inte rferes with the electrostatic and hydrophobic interactions of the antigenantibody complex and causes them to dissociate. Acidic elutions do not work for every antigenantibody complex; sometimes the low pH of the elutant can denature the antibodies or antigens. Elution with antigen elutes the antibody from the immobili zed antigen-antibody complex by competing with the immobilized antigen for bindi ng with the antibody. While pH elutions are

PAGE 125

125 the most common elutions used in phage displa y, elutions by antigen are not uncommon and have even been noted to be more successful than pH elutions because th ey decrease the elution of phages that are not speci fic to the antigen (89-91). Panning the Ph.D. 12mer library against phenol-water-extracted V. cholerae O1 LPS that was immobilized onto nitroce llulose paper and eluted with either glycine (pH 2.2) or V. cholerae O1 LPS was performed. Three rounds of panning were performed with amplification of phages after rounds one and two. A subtractive pa nning was performed after round one. This negative panning was performed to decrease th e amount of nonspecific phages in the library. The negative panning involved pann ing the library against bloc ked nitrocellulose paper; therefore, clones specific to the blocker, nitrocellulose paper, or polystyrene should be removed from the library. Two hundred clones from each panning were individually amplified and analyzed by ELISA. Th e 400 phage clones and -Vc O1 LPS mAb were screened in an ELISA against V. cholerae 569B LPS. Anti-Vc O1 LPS mAb gene rated a S:N of 10, proving that the LPS successfully bound to the microtiter plate. Anti-BSA phagemid particles were used in the ELISA against BSA (10 g/mL) and generated a S:N of 15, proving that the ELISA was a successful assay for the detection of phages. All 400 clones gave a S:N less than 1.7; therefore, all of the clones screened were negative for binding to V. cholerae O1 LPS. Previous attempts to select recombinant pha ges to LPS have failed to generate highly specific recombinant phages that generally la ck a consensus sequence (92-94). Also, the isolation of recombinant phages that bind to LPS is less common than for the isolation of recombinant phages that bind to protein. This is most likely because protei n-protein interactions are usually stronger than protei n-carbohydrate interactions. Th e antibody-antigen complex is held together by noncovalent in teractions such as hydrogen bonds, van der Waals forces,

PAGE 126

126 coulombic interactions, and hydrophobic bonds ( 95). The strength of the antibody-antigen complex depends on the strength of the bonds used to hold it togeth er. Protein-protein interactions make use of all noncovalent bonds, but carbohydrate-pro tein interactions primarily make use of hydrogen bonds and rarely make use of hydrophobic or coulombic interactions. Also, protein-protein interac tions can be held together by numerous bonds due to the conformational structure of the protein-protein interaction. Th e carbohydrate-prote in interaction is usually only bound by a single domain due to th e linearity of the car bohydrate structure. Therefore, the low density and the limited va riety of bonds that occu r in carbohydrate-protein interactions make LPS a difficult target to select recombinant proteins that bind to it with high affinity. Specific Aim 2: Improve Panning a nd Screening Process of Biopanning Biopanning is perform ed to select phagemid partic les that specifically bind to a target. To increase the chances of selecti ng phagemid particles that are speci fic to a target, the optimization of the biopanning and screening process was pe rformed. Optimization of the panning process was performed with V. cholerae O1 whole cells as a ta rget, and the concentr ation of whole cells and the trypsin elution time were analyzed. Th e concentration of whole cells used in biopanning usually ranges from 107 to 109 whole cells (96). The goal was to use enough whole cells to enable capture of specific recombinant phagemid particles and keep the capture of phagemid particles not specific to the w hole cells to a minimum. The trypsin elution time was also analyzed. The Tomlinson libraries have a trypsin cleavage site between the pIII protein and the scFv particle. Therefore, trypsin is used to cl eave the phagemid particle from the scFv peptide that is bound to the target. Th e Tomlinson protocol recommends a trypsin elution of 10 minutes. The trypsin elution time was analyzed to maximi ze the number of eluted phagemid particles that were specific to the target. Eluting for too short of a time may fail to elute specific phagemid

PAGE 127

127 particles, but eluting for too long of a time may only in crease the elution of nonspecific phagemid particles or even de grade the phagemid particles. Comparative pannings were performed to determine the optimal concentration of V. cholerae whole cells and optimal trypsin elution time. Parallel pannings were performed with the Tomlinson J scFv library w ith and without 1% Vc86 phagemi d particles spiked into the library. Vc86 is a clone isolated from the To mlinson I scFv library th at recognizes an unknown antigen on V. cholerae whole cells. Because ~1011 phagemids are being used in the pannings there only exists a 100-fold redundancy of clones in the non-spiked library. Therefore, enumeration of eluted phagemid particles afte r one round of panning may not yield significant differences in eluted titers unde r various test parameters. Spiking the library with 1% phagemid particles that are specific to V. cholerae whole cells should enable di stinction of an optimal panning parameter that selects and elutes the highest number of specific recombinant phagemid particles. We were looking for a combination of the concentration of whole cells and trypsin elution time to elute more phagemid particles from the Vc86-spiked library than the non-spiked library. Because the only difference between th e pannings was that the Vc86-spiked library had a higher initial concentration of specific phagemid particles to V. cholerae whole cells, an increase in eluted phagemid particles with a cert ain parameter should correlate to an increase in specific phagemid particles being eluted. The first one-round panning used 1 x 108 or 1 x 109 whole cells of V. cholerae with trypsin elutions of 15, 30, 45, and 60 minutes (Fig. 3-6). Elution of phagem id particles with trypsin was optimal between 15 and 30 minutes. The phagemid particle titer decreased by 80 to 93% from 30 to 45 minutes for the four pannings. The decrea se in the titer of phagemid particles was most likely due to degradation of the phagemid particle s by trypsin. Therefore, trypsin elution should

PAGE 128

128 not exceed 30 minutes. More phagemid partic les were eluted when panned against 1 x 109 whole cells for both the spiked and nonspiked libraries However, there were not enough data to do statistical analysis to determine if using 1 x 109 whole cells eluted sign ificantly more phagemid particles than using 1 x 108 whole cells in the biopanning process. The best whole cell concentra tion and trypsin elution time were not definitively obtained from the first one-round panning; therefore, a second one-round panning was performed to further analyze the trypsin elution time and whole cell concentration. The second one-round panning used 107, 108, or 109 whole cells of V. cholerae with a trypsin elution of 10, 20, and 30 minutes (Fig. 3-7). There were no significant di fferences in the titers of eluted phagemid particles when comparing the con centrations of whole ce lls and trypsin elution times used in the second one-round panning experiment for both library pannings. This could mean that there may not be a significant advantage in using the diffe rent elution times or concentrations of whole cells for selecting specific recombinant phages. Because using titering to analyze the differe nces in the amount of eluted phagemid particles under various parameters failed to definitively determine a panning condition that eluted the highest number of specific recombinant phagemid particles, another one-round panning was performed with analysis of phagemi d particle specificity by ELISA. Following elution of the phagemid particles, the phagemid particles were amplified in batch and analyzed by ELISA to determine which panning parameter el uted the most phagemid particles that were specific to V. cholerae whole cells. The third one-round panning used 108 or 109 whole cells of V. cholerae with trypsin elutions of 10 and 30 minutes (Fig. 3-8). The el uted phagemid particles were amplified in batch and analyzed by ELISA. Of the different panning variations, the amplified phagemid particles

PAGE 129

129 from the panning with the library that contained 1% Vc86 phagemid particles, 108 whole cells V. cholerae, and an elution time of 30 minutes gene rated a significantly higher S:N (p=0.010.04) in an ELISA compared to the amplifie d phagemid particles from the other panning variations. This suggests that panning with 108 whole cells with a tryps in elution time of 30 minutes is optimal for selecting phagemid partic les that are specific to the panned whole cells compared to using 109 whole cells or a trypsin elution of 10 minutes. There might have been an excess of whole cells wh en panning against 109 whole cells. This excess may have promoted the binding of nonspecific phages; therefore, fewer whol e cells present may have been more efficient for panning because with fewer whole cells pr esent the specific phage mid particles should preferentially bind the whole cells and the number of nonspecific phagemid particles would be lessened. Eluting for 10 minutes may not have been long enough for the trypsin to cleave all the specific phagemid particles from the whole cells; therefore, eluting for 30 minutes was optimal because it may have given the trypsin enough tim e to elute the phagemid particles from the whole cells. After the biopanning process is complete, the selected scFvs or phages are screened for their specificity to the target molecule. The previous met hod to screen scFv proteins was to grow amber suppressor-free phagemid-containing bacteria under conditions to secrete scFv proteins in culture tubes or flasks. However, th is process was very time consuming; therefore, a way to screen scFv proteins in a high th roughput manor was investig ated. High throughput screening of scFv proteins is very common. A Biacore A 100 array system is capable of screening hundreds of scFv proteins per day (97). This system sc reens scFv proteins from crude bacterial extracts on a sensor chip surface. Th e antibodies that are captured onto the chip are analyzed and ranked by the per centage of bound scFv proteins remaining on the chip after

PAGE 130

130 dissociation of nonspecific scFv proteins in buffer. His-tagged scFvs can be screened in an automated high throughput approach by measure of surface plasmon resonance using a Qiagen BioRobot 3000 LS (98). A chromatography press is capable of purifying and screening hundreds of scFv proteins per day (99). This sy stem involves His-tagged scFvs being purified by a cation exchange column and immobilized meta l ion affinity chromatography. The proteins were then quantified by a bicinchoninic acid ( BCA) assay and analyzed by SDS-PAGE. While there are many high throughput screens for scFv prot eins, most of the methods involve expensive equipment that is not readily av ailable to our laboratory. Therefore, our goal was to develop a high throughput screen for scFv proteins. A colony blot was attempted to screen scFv proteins. The colony blot involved the transfer of bacterial colonies that were grown to secr ete scFv proteins onto an antigen-coated piece of nitrocellulose paper. The scFv proteins that bou nd to the nitrocellulose via the antigen were then detected with Protein-L peroxidase and identified with ECL. However, when this system was attempted, the colony blot was determined to be neither specific nor selective. When the -BSA clone was screened by this method, it did not generate a positive signal on BSA-coated nitrocellulose but did generate a positive signal on PBS and V. cholerae -coated nitrocellulose paper. Also, when clones se lected in a panning against V. cholerae whole cells were screened by this method, they generated posi tive signals to BSA, PBS, and V. cholerae -coated nitrocellulose papers. It appeared as if random clones bound to the nitrocellulose paper despite it being blocked with casein buffer and th at all clones that produced scFv s were detected. For this problem to occur many steps or just one step co uld have gone wrong in the experiment. It is possible that not enough antigen was used to coat the nitrocellulose, the nitrocellulose did not capture the antigen, the nitrocel lulose was poorly blocked, or th e washes were not stringent

PAGE 131

131 enough. The effort required to tr oubleshoot the colony blot was de termined to not be worth the time. Therefore, a way to improve the old method for producing scFvs was developed. To enable high throughput production of scFv proteins, a clone was grown in a 96-well plate instead of a culture tube or flask, and the scFv-containing supernatant was examined in an ELISA. When -BSA clone was grown in a 96-well pl ate, its scFv-containing supernatant diluted 1:2 yielded a S:N of 11.5 against BSA in an ELISA. Anti-BSA scFv produced in a flask yielded a S:N between 10 and 15; th erefore, producing scFv proteins in a microtiter well may not be optimal, but it is acceptable for the purpose of a high thro ughput screen. Next was to determine how far the microtiter well-produced scFvs could be diluted and still generate a sufficient signal to be detected in an ELISA. This test was to determine how many clones could be mixed together (i.e., diluted) and screened in a single microtit er well to enable even greater high throughput screening of scFvs. The -BSA clone was grown in a microtiter well to produce scFvs. The supernatants were diluted 1:2, 1:4, 1:6, and 1:10 in casein blocking buffer and analyzed by ELISA (Fig. 3-10). When -BSA scFv-containing supernatant was dilute d 1:10 it still yielded a S:N of 5.6 in an ELISA. This S:N is high enough for de tection of positive clones. However, the -BSA clone was a very strong clone; therefore, the majority of clones selected will have a weaker signal than the -BSA clone. To lessen the chance of not detecting scFvs with weak affinity or low titers, it was decided to only dilute scFv clones 1:4 to 1:6 for high throughput screening by ELISA. This would still enable screening of hundreds of clones per day, a vast improvement to the previous method of screening clones in culture tubes or flasks which only enable d screening of a couple dozen clones per day.

PAGE 132

132 Because the high throughput screen for scFv proteins was successful, it was examined if a similar high throughout ELISA could be applied to the screening of phagemid particles. Previous methods of screening phages involved pr oducing phagemid particles in a culture tube or flask and screening the supe rnatants by ELISA. This method was time consuming; therefore, a high throughput screen was investigated. Previous high throughput ELISAs have been developed for the screening of pha ges; however, while their screen ing is rapid, the production of the phages is time consuming. These high thro ughput ELISAs require ov ernight production of phages in a culture tube or flask followed by PEG precipitation of the phages for the concentrated phages to be screened by ELISA (100,101). We hoped to develop an even more high throughput ELISA by not only screening phagemid particles in a microtiter well but also growing the phagemid particles in a microtiter well. Phagemid particles from the -BSA and Vc86 clones were transduced into E. coli TG1 (Hy -sm), which harbors the hyperphage genome as a plasmid. Producing phagemid particles in E. coli TG1 (Hy -sm) removes the additional step of superinfection with hyperphage; therefore, reducing the time required to produ ce phagemid particles. Colonies from the transduction were picked and grown in a microtiter well to produce phagemid particles. Growth temperature and glucose concentr ation of the medium were analyzed. The Tomlinson protocol recommends producing phagemid particles in medi um that has 0.1% (w/v) glucose. However, this concentration of glucose could promote catabolite repression of the lac promoter, which promotes transcription of the gIII gene. Therefore, it was examin ed if phagemid particles would be produced better in medium with or without 0.1% (w/v) glucose. The -BSA and Vc86 phagemid particles were produced in a microtiter well at 30C or 37C with and without 0.1% (w/v) glucose. The phage-containing supernat ants were diluted 1:2

PAGE 133

133 in blocking buffer and analyzed by ELISA (Fig. 311). There was not a significant difference in the S:N (p=0.07-0.48) of phagemid particles produced in microtiter wells with and without 0.1% (w/v) glucose. Therefore, the 0.1% (w/v) glucose in the medium was probably not enough glucose to cause catabol ite repression of the lac promoter. Vc86 phagemid particles produced in microtiter wells with or without glucose at 30 C and 37C did not produce a S:N greater than 1, whereas Vc86 phagemid particles produced in a flas k (S:N ~4) yielded a significantly higher S:N (p=0.05). Observation of the turbidities of the ba cterial cultures that were grown in a microtiter well revealed only slightly turbid cultures compar ed to bacterial cultures that were grown in a flask. The growth conditions may not have been op timal for the bacteria to grow when they had the added stress of producing phagemid particles. It was possible that the bacterial cultures that were grown to produce phagemid particles in a microtiter well experienced less aeration than bacterial cultures grown in a flask or culture tube. The decreased aeration could decrease bacterial growth and reduce phagemid particle production. Escherichia coli is a facultative anaerobe which means it makes ATP, which driv es biosynthesis, by aerobic respiration or fermentation. Aerobic respiration generates 36 ATP molecules from one molecule of glucose, while fermentation only generates 2 ATP molecule s from one molecule of glucose. Because biosynthesis is fueled by ATP, aerobic respira tion is more effective for bacterial growth. When -BSA phagemid particles were produced at 30C, the S:N was significantly higher (p<0.04) than when -BSA phagemid particles were produ ced at 37C at either glucose concentration in a microtiter we ll. It appeared that phagemi d particle production was more permissive at 30C opposed to 37C. The -BSA phagemid particles pr oduced in a flask (S:N ~12) yielded a significantly hi gher (p=0.02) S:N compared to -BSA phagemid particles produced in microtiter wells (S:N~7) when analyzed at a 1:2 dilution in an ELISA. The reduced

PAGE 134

134 S:N of phagemid particles used in an ELISA c ould have been due to less phagemid particles present in the phagemid particle -containing supernatant that was produced in a microtiter well as opposed to the phagemid particle-containing supern atant that was produced in a flask. This could have been due to reduced aeration in the microtiter well causi ng decreased bacterial growth and thus reduced quantities of phagemi d particles produced in a microtiter well. Phagemid particle production in microtit er wells was repeated a second time with -BSA phagemid particles (Fig. 3-12). Anti-BSA pha gemid particles yielded significantly higher signals (p<0.03) when produced at 30C as opposed to 37C at either glucose concentration. These data support the previous da ta that 30C is a more permissi ve temperature than 37C for phagemid particle production. There was not a significant difference (p=0.34) in the signals produced by phagemid particles that were produced with or without glucose at 30C. These data also support the previous data that 0.1% (w/v) glucose in th e medium does not provide an advantage or hindrance to phagemid particle prod uction. While the second experiment did not yield a significant difference (p= 0.09) in signals produced by -BSA phagemid particles produced in a flask opposed to in a microtiter well, the microtiter well-produced phagemid particles yielded lower signals than di d phagemid particles produced in a flask. Because phagemid particles that were produced from the same clone resulted in various ELISA signals when phage-containing supernatants were analyzed at a dilution of 1:2 it was possible that the reason for the differences in ELISA signals from phage-containing supernatants were due to differences in phagemid particle titers produced under the various conditions. Therefore, the titers of -BSA and Vc86 phagemid particles produced in a microtiter well and in a flask were analyzed. Phagemid particles produced at 30C in a microtiter well yielded approximately 107 tu/mL while phagemid particles produced at 37C in a microtiter well yielded

PAGE 135

135 approximately 106 tu/mL. Producing phagemid particles at 30C in microtiter wells yielded approximately 10-fold higher tu/mL titers than phagemid particles that were produced in microtiter wells at 37C. Whether the growth medium contained 0.1% (w/v) glucose or not did not affect the titers of ph agemid particles. When -BSA and Vc86 phagemid particles were produced in flasks at 30C with 0.1% (w/v ) glucose in the medium, the titers were approximately108 to109 tu/mL. Producing phagemid particles in a microtiter well yielded higher titers and signals in an ELISA when produced at 30C. Whether glucose was in the media or not did not affect phagemid particle titers or ELI SA signals. Phagemid particles could not be produced in a microtiter well in high enough titers to make a high throughput screen acceptable. When clone Vc86, a clone of average strength, was produced in a microtiter well, it was not able to generate a positive signal. Therefore, the high throughput ELISA screen for phagemid particles was not acceptable for screening of phagemid particles. Specific Aim 3: Improve Phagemid Particle Production Am plification of phagemid particles is an essential step in biopanning. A phagemid contains the gIII gene of M13 but lacks the rest of the M13 genes needed for phagemid particle production. Therefore, a phagemid re quires an amplification tool to provide the rest of the M13 genes in trans to enable phagemid particle production. It is esse ntial that there is high quality and high quantity phagemid particle production during amplification to ensure that the redundancy of clones in the library is maintained and that the library is in high enough titers to be useful in assays. The most commonly used amplification tool fo r phagemids is helper phage. There are many variations of helper phages, R408, VCSM13, a nd M13KO7 (63) that differ slightly. Of the helper phages, M13KO7 is the most commonly used Because helper phages contain a wild type gIII gene, they are not optimal for phagemid particle amplification because the produced

PAGE 136

136 phagemid particles will express non-recombinant pIII proteins along with recombinant pIII proteins. To overcome the drawback of helper phages encoding a wild type gIII gene, many gIII gene mutated variations of helper phages were engineered. Many of these strains either have a deletion of the gIII gene or encode amber stop codons within the gIII gene to lessen the expression of wild type pIII proteins (65-67). Of the gIII gene mutated helper phages, hyperphage is the most effective for producing high titers of phagemid particles that have multivalent display of pIII-fusion proteins (66). Hyperphage was the previous amplification tool used by our laboratory to amplify phagemid particles; however, th e titer of the hyperphage stock was too low to effectively amplify every phagemid particle in the library. Therefore, a way to improve phagemid particle production was investigated. To overcome the problem of the low titer of the stock of hyperphage from effectively infecting every phagemid-containing bacterium, the hyperphage genome was transduced into E. coli to be maintained as a plasmid to ensure that every phagemid-cont aining bacterium also contained hyperphage to enable amplification of phagemid particles. When hyperphage was transduced into E. coli TG1, small (0.5 mm diameter) and large (1-1.5 mm diameter) colonies resulted. When the small and large colonies were passaged, the small colonies maintained a small phenotype (0.5 mm diameter) and the large colonies maintained a large phenotype (1-1.5 mm diameter). The two strains were named E. coli TG1 (Hy -sm) and E. coli (Hy -lg). The reason for the variation in the colony size was not known. Because colonies were grown under kanamycin resistance for hyperphage, both strain s contained hyperphage, or at least they contained the kanamycin resistance gene that hyperphage carries. The differences in the colony sizes suggested that the bacteria that formed the colonies were not replicating at the same rate. The smaller colonies might be small because they are under stress from replicating the

PAGE 137

137 hyperphage genome along with the E. coli genome, and the large colonies might be larger because they are capable of growing faster beca use they are under less stress because they may not be replicating everything. To determine if the two strains we re functionally different, their infection efficiencies were compared to E. coli TG1 and their ability to produce phagemid particles was compared to E. coli TG1 containing a phagemid th at was superinfected with hyperphage. The infection efficiencies of -BSA and Vc86 into E. coli TG1, E. coli TG1 (Hy -sm), and E. coli (Hy -lg) were analyzed. When the number of -BSA phagemid particles transduced into E. coli TG1 was set at an infection efficiency of 100%, the number of -BSA phagemid particles transduced into E. coli TG1 (Hy -sm) was 28% and the number of -BSA phagemid particles transduced into E. coli TG1 (Hy -lg) was 0.1% of the number of transduced phagemid particles into E. coli TG1. When the number of Vc86 pha gemid particles transduced into E. coli TG1 was set at an infection efficiency of 100%, the number of Vc86 phagemid particles transduced into E. coli TG1 (Hy -sm) was 22% and the number of Vc86 phagemid particles transduced into E. coli TG1 (Hy -lg) was 2% of the number of transduced phagemid particles into E. coli TG1. Escherichia coli TG1 (Hy -sm) had an infection efficiency that was approximately 25% of the infection efficiency of E. coli TG1, and E. coli TG1 (Hy -lg) had an infection efficiency that was approximate ly 1% of the infection efficiency of E. coli TG1. It appears as if E. coli TG1 has a reduced ability to be transd uced with phagemid particles when it is harboring hyperphage. The differences in the infection efficiencies between E. coli TG1 (Hy -sm) and E. coli TG1 (Hy -lg) show that their functionali ties are d ifferent. The reason for E. coli TG1 (Hy -lg) generating larger colonies may ha ve been because it shed a function required in transduction and theref ore was able to grow faster. For example, the production of

PAGE 138

138 the F pilus could have been compromised by the stress from E. coli TG1 replicating the hyperphage genome. Both hyperphage-containing strains of E. coli TG1 had infection efficiencies that were too low for the strains to be used to effectively amplify clones in a librar y. While the hyperphagecontaining E. coli strains have infection efficiencies too low to be used to amplify the library their infection efficiencies are still acceptable for amplif ying single clones because the transduction efficiency is not critical for the am plification of a single cl one. Therefore, the quality and quantity of phagemid particles produced by E. coli TG1 (Hy -sm/lg) were analyzed. Anti-BSA phagemid particles were produced by E. coli TG1 (Hy -sm), E. coli TG1 (Hy -lg), and E. coli TG1 that was superinfected with hyper phage. Anti-BSA phagemid particle yields were as follows: E. coli TG1 that was superinfected with hyperphage yielded 3 x 10 8 tu/mL, E. coli TG1 (Hy -sm) yielded 1 x 109 tu/mL, and E. coli TG1 (Hy -lg) yielded 3 x 108 tu/mL. While the ability of hyperphage-containing E. coli TG1 to be transduced with phagemid particles was reduced comp ared to the ability of E. coli TG1, its ability to produce phagemid particles was not reduced. Phagemid partic les produced by hyperphage -containing bacteria yielded titers high enough to be us ed as a phagemid particle amplification tool for single clones. The qualities of -BSA phagemid particles produced by the three E. coli strains were analyzed by ELISA (Fig. 3-13). Anti-BSA pha gemid particles produced a significantly higher S:N (p=0.01) when produced from E. coli TG1 (Hy -sm) compared to E. coli TG1 (Hy -lg). There was not a significant difference in the S:N (p=0.50) from -BSA phagemid particles produced from superinfected E. coli TG1 and E. coli TG1 (Hy -lg). Because the titer of phagemid particles produced from E. coli TG1 (Hy -sm), ~1 x 109 tu/mL, was three times higher than the titer of phag emid particles produced from E. coli TG1 (Hy -lg) or E. coli TG1

PAGE 139

139 that was superinfected with hyperphage, ~3 x 108 tu/mL, a third less phage-containing supernatant was used in the ELISA for phagemid particles that were produced from E. coli TG1 (Hy -sm). The decrease in the S:N from -BSA phagemid particles produced from E. coli TG1 (Hy -sm) could have resulted if th e particles/mL of the phagemi d particles produced from all three strains were the same. Because hyperphage-containing E. coli TG1 was not going to be used to amplify the library, furthe r analysis of it was stopped, so that the time used to analyze it could be put to better use finding a more eff ective phagemid particle amplification tool. Hyperphage-containing bacteria was not acceptable as an amplif ication tool for the library because of their low infection efficiencies, but E. coli TG1 (Hy -sm) was acceptable for use as an amplification tool for the am plification of single clones. Helper plasmids are another amplification t ool for the production of phagemid particles (64). They are M13-based plas mids that are maintained in E. coli to provide phagemids with all the genes necessary to produce phage mid particles. Helper plasmids were engineered in three forms that differ in the length of their gIII genes. The helper plasmi ds contain a full-length (M13cp), a deleted (M13cp-dg3) or a truncated (M13cp-CT) gIII gene. These three helper plasmids were transformed into E. coli and analyzed for their infection efficiency with phagemid particles, their ability to produ ce phagemid particles, and the qua lity of the phagemid particles produced. When the helper plasmids were transformed into E. coli TG1, the resulting colony sizes varied from a diameter of 0.5 to 2 mm. The reasons for various colony sizes are hypothesized above. A small (0.5 mm) colony and a large (2 mm) colony from each transformation were passaged to determine if the colony size phenotype would be maintained. Escherichia coli TG1 (M13cp-sm) was the only strain that maintained a consistent phenotype of 0.5 mm colonies, the

PAGE 140

140 rest resulted in colonies that had a diameter between 0.5 and 2 mm. A small and large of each strain were further analyzed to determine if the different colony phenotyp es also had different functionality. The six helper plasmid-containing E. coli TG1 strains were analyzed for their infection efficiencies with two phagemid particles in relation to the infec tion efficiencies of E. coli TG1 that did not contain a ny helper plasmids. Escherichia coli TG1 (M13cp-dg3-sm) and E. coli TG1 (M13cp-CT-sm) were the only strains that yielded infection efficiencies greater than 0.3% of the infection efficiency of E. coli TG1 (Table 3-1). Escherichia coli TG1 (M13cp-dg3-sm) had infection efficiencies of 175% and 97%, while E. coli TG1 (M13cp-CT-sm) had infection efficiencies of 7% and 13% compared to the infection efficiency of E. coli TG1. The differences in the infection efficiencies be tween the small and large strain s of helper plasmid-containing E. coli revealed that they had different functionalities. Except for E. coli TG1 (M13cp-dg3-sm), all of the other helper plasmid strains had reduced infection efficiencies compared to E. coli TG1. Therefore, the cost for the E. coli cells maintaining the helper plasmids resulted in a decrease in their infectio n efficiencies. Because E. coli TG1 (M13cp-dg3-sm) and E. coli TG1 (M13cp-CT-sm) had higher infection efficiencies compared to their large phenotypes, it suggested that the larg e strains had a defect in some component of the cell involved in transduction. Escherichia coli TG1 (M13cp-dg3-sm) had the highest inf ection efficiencies of the helper plasmid strains; therefore, it was further tested to determine if the high infection efficiencies could be maintained when tested with additional phagemid particles. Escherichia coli TG1 (M13cp-dg3-sm) had infection efficiencies of 92%, 113%, and 120% compared to the infection efficiencies of E. coli TG1 with Vc86, clone18, and -AV20N3 (Table 3-2). The infection

PAGE 141

141 efficiency of E. coli TG1 (M13cp-dg3-sm) was approximately the same as E. coli TG1; therefore, E. coli TG1 did not suffer any reduction in inf ection efficiency due to maintaining M13cp-dg3-sm. Escherichia coli TG1 (M13cp-CT-sm) was the only othe r helper plasmid strain besides E. coli TG1 (M13cp-dg3-sm) that had an infec tion efficiency close to that of E. coli TG1. However, its infection efficiency was still too low for it to be used as an amplification tool for the library. Therefore, M13c p-CT-sm was transformed into E. coli JM109 to try and increase its infection efficiency. Escherichia coli DH5 F was the bacterial strain that the engineers of the helper plasmids used to work with the help er plasmids; therefore, we tested if an E. coli strain more similar to E. coli DH5 F would generate higher infecti on efficiencies if it contained M13cp-CT-sm. Escherichia coli JM109 and E. coli DH5 F contain relA1 and recA1 while E. coli TG1 does not. However, infection efficiency experiments with E. coli JM109 (M13cpCT-sm) with three different phagemids yiel ded lower infection efficiencies than did E. coli TG1 (M13cp-CT-sm). The average infection efficiency of E. coli TG1 (M13cp-CT-sm) was 47%, while the average infection efficiency of E. coli JM109 (M13cp-CT-sm) was 28%. Therefore, using E. coli JM109 to maintain M13cp-CT-sm did not increase its infection efficiency. When the infection efficiencies of E coli TG1 (M13cp-dg3-sm) were analyzed, this strain yielded an average infecti on efficiency of 120% and E. coli TG1 (M13cp-CT-sm) had an average infection efficiency of 30%. Escherichia coli TG1 (M13cp-dg3-sm) had a significantly higher infection efficiency (p=0.01) than E. coli TG1 (M13cp-CT-sm). Escherichia coli TG1 (M13cp-dg3-sm), E. coli TG1 (M13cp-CT-sm), and E. coli TG1 that was superinfected with hyperphage were anal yzed for their abilities to produce phagemid particles. Phagemid particles were produced with and without 1 mM IPTG in the growth

PAGE 142

142 medium to analyze if IPTG had an effect on phagemid particle production yield. Because the lac promoter drives transcription of the gIII -scFv fusion it was hypothesi zed that the expressing more pIII-scFv fusions would result in more phagemid particles produced. However, when phagemid particles were titere d, the phagemid particles produ ced in growth medium that contained IPTG had the same or lower titers of phagemid particles than when phagemid particles were produced without IPTG in th e growth medium. Because ther e was not an increase in the phagemid particle yields when they were produced with IPTG, it suggests that the amount of pIII protein was not a limiting factor in phagemid particle production. For the titers that had a decrease in phagemid particle yields when produ ced with IPTG the increase in pIII protein may have been harmful to phage assembly because th e excess of pIII protein may have been toxic. Phagemid particles were amplified to titers of 106-108 tu/mL by E. coli TG1 (M13cp-dg3-sm), 105 -106 tu/mL by E. coli TG1 (M13cp-CT-sm), and 108 -109 tu/mL by E. coli TG1 superinfected with commercial hyperphage. Titers of phagemid particles produced from E. coli TG1 that contained helper plasmids were too low for the he lper plasmids to be an effective amplification tool. Phagemid particle produc tion cultures usually contain ~109 CFU/mL; therefore, for the helper plasmids to only generate 105 tu/mL means that only 1 in every 10,000 bacteria produced an infectious phagemid particle. The use of help er plasmids as an amplification tool did not generate high enough titers of phagemid particle s for them to effectively amplify a library. Even though the helper plasmids were determin ed to be an ineffec tive amplification tool, they were further analyzed for the quality of phagemid particles produced by ELISA. Analysis by ELISA revealed that phagemid particles produced by helper plasmids resulted in a S:N of 7 when less than 5 x 104 tu/mL of phagemid particles were used as a primary antibody. Data obtained from our laboratory show ed that phagemid particles are not detectable in an ELISA

PAGE 143

143 unless they are used at concen trations of at least 1 x 106 particles/mL (data not shown). Because the spot titer method measures the number of infectious phagemid pa rticles, it is possible that the total number of phagemid particles far exceeded the number of infectious phagemid particles. To determine if the total number of phagemid par ticles is greater than the number of infectious particles an anti-M13 sandwich ELI SA was used to enumerate the particles/mL of phagemid particle-containing supernatants (Table 3-4). Phagemid particles amplified by E. coli TG1 (M13cp-CT-sm) had particle/mL to tu/mL ratios between 740 and 23,000, E. coli TG1 (M13cpdg3-sm) had particle/mL to tu/mL ratios between 1 and 190, while phagemid particles amplified by E. coli TG1 that was superinfected w ith hyperphage had particle/mL to tu/mL ratios between 6 and 24. Phagemid particles produced by helper plasmid-containing E. coli had particle/mL to tu/mL ratios that were much higher than those for phagemid particles produced by E. coli that was superinfected hyperphage. In biopanning the number of phagemid partic les is not important, the number of infectious phagemid particles is what matters because if a pha gemid particle is not infectious then it can not enter a bacterial cell and be produced to eventua lly be selected and screened. Therefore, even t hough the number of partic les produced by helper plasmids was high, the number of infectious particle s produced was low, making the helper plasmids an inadequate amplification tool for the library. The helper plasmids yielded phagemid particle progeny at titers between 105 to108 tu/mL. In a last attempt to get the helper plasmids to generate high titers of phagemid particles, the M13cp-dg3-sm helper plasmid was incorporated into E. coli MG1655 to determine if a less complex E. coli strain could generate higher ti ters of phagemid particles. Escherichia coli MG1655 was chosen because it has few known muta tions and does not cont ain the F-plasmid. Four different phagemid particles were amplified with E. coli MG1655 (M13cp-dg3-sm), and the

PAGE 144

144 resulting titers were: 2.5 x 107, 6.3 x 106, 2.5 x 107, and 3.9 x 107 tu/mL. Producing phagemid particles from E. coli MG1655 that contained a he lper plasmid did not in crease the titer of phagemid particles produced. The helper plasmids were extensively analyzed and determined to be poor amplification tools. Therefore, hyperpha ge was the best amplification tool, but it was available in titers too low to be used effectively. To overcome the low titers of commercial hyperphage, we made homemade hyperphage. Homemade hyperphage was previo usly made by transducing F+ E. coli cells that harbored a plasmid that encoded the gIII gene, pGTR203, with hyperphage and growing these transduced cells under conditions that promoted phage produc tion. In the past, hyperphage amplified from E. coli TG1 (pGTR203) yielded titers between 107 to 108 tu/mL and E. coli HB2151 (pGTR203) yielded titers between 106 to 107 tu/mL. Many liters of hyperphage had to be produced and concentrated to generate hyperphage in a tite r high enough to be used for a single panning experiment. Therefore, a way to increase the t iter of homemade hyperphage was investigated to generate enough hyperphage to be used in multiple experiments. Escherichia coli TG1 and E. coli HB2151 are both F+ strains. Because hyperphage infect s bacteria by the F pilus, it was hypothesized that some of the hyperphage produced were being taken back into the bacteria which would decrease the titers of the hyperphage produced. Therefore, it was investigated if hyperphage could be produced in high er titers when produced from a FE. coli strain. Escherichia coli MG1655 and E. coli EC100D were the two Fstrains of E. coli chosen to amplify hyperphage. These strains were transformed with pGTR203 encoding the gIII gene and the hyperphage genome, and five isolates from each strain were grown under conditions to promote phage production. The particles/mL and the tu/mL were analyzed (Table 3-5). The average titer of hyperphage produced from the five E. coli EC100D (pGTR203) (Hy ) isolates

PAGE 145

145 was 2.1 x 108.5 x 108 tu/mL. The average titer of hyperphage produced from the five E. coli MG1655 (pGTR203) (Hy ) isolates was 3.0 x 109 .0 x 109 tu/mL. The titer of hyperphage produced from the E. coli MG1655 (pGTR203) (Hy ) isolates was significantly higher (p=0.02) than the titer of hyperphage produced from the E. coli EC100D (pGTR203) (Hy ) isolates. The higher titers produced from E. coli MG1655 might be because it has less known mutations than E. coli EC100D. The ratios of particles/mL to tu/mL of hyperphages produced from E. coli EC100D (pGTR203) (Hy ) were between 0.4 and 7.8. The ratios of particles/mL to tu/mL of hyperphages produced from E. coli MG1655 (pGTR203) (Hy ) were between 0.3 and 1.0. The particle/mL to tu/mL ratios of hyperphage produced from the Fstrains were approximately the same, which means that approximately all of the phagemid particles produced were infectious. Escherichia coli MG1655 (pGTR203) (Hy ) produced higher titers of hyperphage than did both F+ strains. Whether or not the hi gher titers of hyperphage were due to being produced in an Fstrain was not determined because E. coli EC100D (pGTR203) (Hy ) produced approximately the same titers of hyperphage as E. coli TG1 (pGTR203) (Hy ). What was determined was that the bacterial strain used to produce hyper phage had an effect on hyperphage production. Because E. coli MG1655 (pGTR203) (Hy ) 5 produced the highest titers of hyperphage, it was used to produce a large batc h culture of hyperphage. The hyper phage was then concentrated to approximately 1 x 1012 tu in 3 mL, which was concentrated enough to amplify the Tomlinson J scFv phagemid library. The Tomlinson J scFv ph agemid library was amplified with homemade hyperphage with quality control ch ecks by titering to ensure that the amplification of the library generated high quality and high quantity of phagemid particles in a way that the redundancy of the library was maintained.

PAGE 146

146 Specific Aim 4: Isolation of Specific Recombinant Phage to Stx2 Toxin of E. coli O157:H7 Once the Tom linson J scFv library was amplified and homemade hyperphage was improved to enable high quality and high quantity am plification of the library, a target molecule had to be selected for panning. Escherichia coli O157:H7 is an enterohemorrhagic serotype of E. coli that is responsible for approximately 73,000 illnesses and more than 60 deaths per year in the United States (30). Stx toxins kill mammalia n cells and are the major causes of damage in E. coli O157:H7-infected individuals. Escherichia coli O157:H7 can produce Stx toxins 1 and/or 2. Escherichia coli that produces Stx2 toxin appears to be more virulent than E. coli that produces Stx1 or both Stx1 and 2 toxins (36-40). Therefore, Stx2 toxin would be a good target for detection in animal and patient samples. A Stx2 toxin preparation was obt ained that was stated to be 50% pure for Stx2 toxin. The other 50% of the preparation most likely contained bacterial impurities from E. coli HB101 because the Stx2 toxin preparation was made from E. coli HB101 (pMJ100-encodes Stx2 toxin) A 100% pure Stx2 preparation would have been id eal; however, if the preparation contains 50% Stx2 toxin then the most of the clones selected sh ould theoretically be specific to Stx2 toxin. The Tomlinson J scFv library that was amplified with homemade hyperphage was panned against the Stx2 toxin preparation. Three rounds of panning were performed with amplification of eluted phagemid particles by homemade hyperphage between each panning. Every panning used 1 x 1012 phagemid particles. The number of el uted phagemid particles for the three pannings were as follows: 2 x 106 for round one, 2 x 107 for round two, and 2 x 1010 for round three. The increase in the num bers of eluted phagemid particles suggested that phagemid particles specific to Stx2 toxin preparation we re being selected and amplified at each round. Clones from the third round of panning were i ndividually amplified and screened by ELISA. Analysis of 59 clones resulted in 53/59 clones (89%) having a S:N from 2 to 23 with 48/59

PAGE 147

147 clones (81%) having a S:N from 10 to 23 against the Stx2 toxin preparation. Therefore, the panning was successful in isolating clones that bo und to Stx2 toxin preparation. Not all of the clones that were selected bound to Stx2 toxin preparation; 4/59 (7%) of the clones bound to PBStreated wells as well as Stx2-coated wells. The four clones had a signal in an ELISA between 0.31 and 0.87 against PBS-treated wells. These cl ones most likely bound to polystyrene. Even though the immunotube was blocked w ith blocking buffer, the unintentional isolation of clones that bound to the panning vessel is common (84). Of the clones sc reened, 2/59 (3%) did not bind to Stx2 toxin preparation or a microtiter well These clones have unknown binding specificity. It is common to select clones that have unknown binding specificity. The fact that only approximately 3% of my clones had unknown bindi ng specificity proves that the panning was successful in isolating specific r ecombinant phagemid particles. Most of the clones selected from panning rec ognized the Stx2 toxin preparation; however, the Stx2 toxin preparation was only 50% pure fo r Stx2 toxin. Therefore, it was unknown if the clones bound to Stx2 toxin or bacterial impurities in the Stx2 toxin preparation. To determine if the selected clones bound to Stx2 toxin, two anti-Stx2 toxin monoclonal antibodies were examined. One antibody recognized the A-su bunit of Stx2 toxin, wh ile the other antibody recognized the B-subunit of Stx2 toxin. However, these monoclonal antibodies never generated a positive signal in an ELISA against the St x2 toxin preparation. Using the monoclonal antibodies at the manufacturer re commended concentration of 1 g/mL failed to generate a positive signal in an ELISA. It was initially reasoned that the monoclonal antibodies may need to be used at a higher concentra tion to increase the S:N values. Increasing the concentration of monoclonal antibodies in an ELISA to 100 g/mL still failed to produce a positive signal in an ELISA. Therefore, it was hypothesized that the HRP-conjugated anti-mouse secondary antibody

PAGE 148

148 that recognized the anti-Stx2 toxin monoclonal antibodies was not recognizing the anti-Stx2 monoclonal antibodies. However, when another HRP-conjugated anti-mouse secondary antibody was used to recognize the anti-Stx2 t oxin monoclonal antibodies it also failed to produce a positive signal in an ELISA. Both anti-mouse antibodies had been tested within three days of these experiments against murine monocl onal antibodies and produced a positive signal. However, positive controls for the anti-mouse an tibodies were not used in the ELISA that analyzed the two anti-mouse antibodies against Stx2 toxin preparation to determine whether or not they were functional. It was not dete rmined why the monoclonal antibodies did not recognize the Stx2 toxin preparati on in an ELISA. One reason might be that there was not enough Stx2 toxin with a correctly folded binding site in the Stx2 toxin preparation for the monoclonal antibodies to recogni ze. Also, it is possible that the batches of monoclonal antibodies that we received were somehow defective. Because the anti-Stx2 toxin monoclonal antibodi es could not be used to determine the binding specificity of the select ed panning clones in an ELISA, other means of determining the binding specificities of the clones were analyz ed. Three clones that bound to Stx2 toxin preparation were screen ed against Stx2 toxin, E. coli O157:H7 933 (expresses Stx 1 and 2 toxins) periplasmic break fraction, E. coli O157:H7 933 supernatant, E. coli DH5 (pNR100) (expresses Stx2 toxoid) peri plasmic break fraction, and E. coli DH5 periplasmic break fraction. A clone that bound to the E. coli bacterial impurities in the Stx2 toxin preparation would be expected to bind to all of the an tigens, but a clone that bound to Stx2 toxin would be expected to bind to all of the antigens except the E. coli DH5 periplasmic break fraction. However, the three selected clones only recognized the Stx2 to xin preparation and none of the other antigens; therefore, it was not determined if the cl ones were actually binding to Stx2 toxin.

PAGE 149

149 Analysis of clones by ELISA proved that the ma jority of the selected clones from panning bound to the Stx2 toxin preparation but failed to determine if th e clones bound to Stx2 toxin because the anti-Stx2 monoclona l antibodies and comparative an alysis of +/Stx2 bacterial preparations failed to generate positive signals in an ELISA. However, when both of the antiStx2 toxin monoclonal antibodies were examined in a Western blot against SDS-PAGE resolved Stx-2 toxin preparation, the anti-Stx2 A s ubunit toxin monoclonal antibody successfully recognized the Stx2 toxin A subunit in the Stx2 toxin preparation. Therefore, analysis by Western blot could at least determine if any of the clones recognized the A subunit of Stx2 toxin. Six clones were screened in a Western bl ot against SDS-PAGE resolved Stx2 toxin preparation, E. coli O157:H7 933 (expresses Stx1 and 2 toxins) supernatant, and E. coli O157:H7 87-23 (Stx phage cured) supernatant (Fig. 3-17). Because the Stx2 toxin is composed of A and B subunits that are approximately 32 and 10-kDa prot eins, a clone that rec ognized Stx2 toxin was expected to recognize a protein band at 32 kDa or 10 kDa on the Stx2 toxin preparation and the E. coli O157:H7 933 supernatant but not on the E. coli O157:H7 87-23 supernatant. If a clone recognized bacterial impurities th en it would be expected to re cognize proteins in all three antigens. The anti-Stx2 A subunit monoclonal antibody recognized a 32-kDa protein in the commercial Stx2 toxin preparation and in the E. coli O157:H7 933 supernatant proving that Stx2 A subunit was present in both preparations and not in E. coli O157:H7 87-23 supernatant, as was expected. Anti-BSA was also analyzed against the three antigens as a negative control to ensure that phagemid particles did not ge nerate artifact bands. Anti-BSA failed to recognize any of the antigens, as was expected. Clone 46 failed to rec ognize any of the three antigens. It is possible that the binding site of the targ et that it recognized in the Stx2 toxin preparation was denatured by SDS-PAGE. Clone 48 recognized 35, 45, and 55-kD a proteins in the St x2 toxin preparation

PAGE 150

150 and a 35-kDa protein in both E. coli O157:H7 supernatants, which means that it probably recognizes bacterial protein impuriti es in the commercial Stx2 toxin Clones 49 and 80 bound to a 35-kDa protein in all three antigens and a 20-kD a protein in the Stx2 toxin preparation and in E. coli O157:H7 87-23. Therefore, clones 49 an d 80 probably bound to bacterial protein impurities in the commercial Stx2 toxin. Clones 81 and 83 appeared to recognize the same bacterial protein impurities as clones 49 and 80, except that thei r band intensities were much lighter. Screening by ELISA showed that 89% of the clones bound to the Stx2 toxin preparation. Western blot analysis of six clones revealed that none recognized Stx2 toxin and that five of the clones most likely recognized bacterial impurities in the Stx2 toxin preparation. Because both of the anti-Stx2 monoclonal antibodies did not recognize Stx2 toxin in the Stx2 toxin preparation in an ELISA and the anti-Stx2 B subunit monoclona l antibody did not recognize Stx2 B subunit in a Western blot, we assumed that there was not a high concentration of Stx2 toxin in the Stx2 toxin preparation. The Stx2 toxin preparation was stated to be 50% pur e for Stx2 toxin by SDSPAGE analysis. To determine if this was true the Stx2 toxin preparation was resolved by SDSPAGE and stained by Coomassie blue (Fig. 3-18). If the Stx2 toxin preparation was 50% Stx2 toxin, 32-kDa and 10-kDa protein bands should have been the most prominent bands. However, SDS-PAGE analysis of the Stx2 toxin preparation showed that a 32-kDa protein band was a minor protein band that represente d less than 5% of the protein in the preparation. There was a band at 10 kDa that had a strong band intensity. These bands might not have even represented Stx2 toxin subunits but may have just been protei ns with a similar molecular weight. What is certain is the amount of Stx2 toxin in the Stx2 toxin preparation did not comprise 50% or even 10% of the proteins in the Stx2 t oxin preparation. Theref ore, it is probable th at the majority of

PAGE 151

151 the clones screened from the panning recognized bacterial impurities from E. coli and not Stx2 toxin. Extensive screening of add itional clones may result in isolati on of a clone specific to Stx2 toxin; however, it is re commended to try panning again with a better Stx2 toxin preparation. The panning was successful in selecting phagemid par ticles that bound specifica lly to the Stx2 toxin preparation, the phagemid particles selected just did not have the desired binding specificity. It is recommended to use the newly amplified Tomlin son scFv library with the aid of the improved homemade hyperphage to pan against another target with a high degree of purity. This thesis details the optimization of phage display protocols and r eagents. Analysis of extraction methods for LPS were compared and it was determined that a phenol-water method, opposed to a TRIzol method, extracted V. cholerae O1 LPS that most closely resembled commercially acquired V. cholerae O1 LPS. Also, binding platforms were analyzed to determine which platform could bind the most LPS. It was determined th at nitrocellulose paper could bind approximately 250 to 2,500 times more L PS than that a microtiter well. Optimization of whole cell panning with the Tomlinson J scFv phagemid library revealed that a trypsin elution between 10 and 30 minutes eluted th e highest number of phagemid pa rticles; this may be due to phagemid particle degradation occu rring with trypsin treatment af ter 45 minutes. Screening of phage display reagents was analyzed and a high throughput ELISA was developed for the production and screening of scFv prot eins that was able to generate S:N values of produced scFv proteins from an ELISA when diluted 1:10. A similar high throughput ELISA was developed for the production and screening of phagemid partic les; however, this method was not acceptable due to low titers produced of phagemid particles in a microtiter well. The optimization of phagemid particle production was performed. Multiple phagemid particle amplification tools were analyzed for th eir infection efficiencies yields of phagemid

PAGE 152

152 particles produced, and the quality of phagemi d particles produced. Hyperphage-containing E. coli resulted in unacceptable infection effici encies and helper plasmid-containing E. coli resulted in unacceptable phagemid particle pr oduction titers. However, improvement of hyperphage that was produced in our laboratory was performed by use of E. coli MG1655 (pGTR203) which generated high titer and hi gh quality hyperphage. This high quality hyperphage was used to successfully amplify the Tomlinson J scFv phagemid library to ensure that the redundancy of phagemid clones was mainta ined. This library was then panned against an Stx2 toxin preparation and resulted in 89% of phagemid particles fr om the third round of panning recognizing the Stx2 toxin preparation in an ELISA. Th e efficiency of this panning was greatly improved from previous pannings from this laboratory, which usually resulted in no better than 1% of selected phagemids being specif ic to the target antige n, and often no isolation of usable specific clones. Theref ore, the optimization of methods a nd reagents in this thesis will greatly improve our laboratory and other laboratories chances of selecting recombinant phages specific to a target molecule.

PAGE 153

153 CHAPTER 5 EPILOGUE The hom emade hyperphage-amplified Tomlinson J library produced in these studies was used to pan against E. coli O157:H7 flagella and yielded ov er 80% positive phagemid clones. The phagemids recognized the major flagelli n protein by Western blot. The high throughput scFv screen was used in the E. coli O157:H7 flagellum project and proved useful in identifying rare scFv-secreting clones. Therefore, the im provements on phage display techniques and tools described in this thesis are, in fact, useful and offer promise of success for continuing studies in the laboratory.

PAGE 154

154 LIST OF REFERENCES 1. Lim, D.V., Simpson, J.M ., Kearns, E.A., and Kramer, M. F. 2005. Current and developing technologies for monitoring agents of bioterrorism and biowarfare. Clin. Microbiol. Rev. 18:583-607. 2. Peruski, L.F., Jr. and Peruski, A.H. 2003. Rapid diagnostic assays in the genomic biology era: detection and identificat ion of infectious disease and biological weapon agents. Biotechniques 35:840-846. 3. Cieslak, T.J., Rowe, J.R., Kortepeter, M.G., Madsen, J.M., Newmark, J., Christopher, G.W., Culpepper, R.C., and Eitzen, E.M., Jr. 2000. A field-expedient algorithmic approach to the clinical management of chemical and biol ogical casualties. Mil. Med. 165:659-662. 4. Franz, D.R., Jahrling, P.B., Friedlander, A.M., McClain, D.J., Hoover, D.L., Bryne, W.R., Pavlin, J.A., Christopher, G.W., and Eitzen, E.M., Jr. 1997. Clinical recognition and management of patients exposed to biological warfare agents. JAMA 278:399-411. 5. Franz, D.R., Jahrling, P.B., McClain, D. J., Hoover, D.L., Byrne, W.R., Pavlin, J.A., Christopher, G.W., Cieslak, T.J ., Friedlander, A.M., and Eitzen, E.M., Jr. 2001. Clinical recognition and management of patients exposed to biological warfare agents. Clin. Lab Med. 21:435-473. 6. YALOW, R.S. and BERSON, S.A. 1959. Assay of plasma insulin in human subjects by immunological methods. Nature 184 (Suppl 21):1648-1649. 7. Saiki, R.K., Scharf, S., Faloona, F., Mulli s, K.B., Horn, G.T., Erlich, H.A., and Arnheim, N. 1992. Enzymatic amplification of beta-globi n genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. 1985. Biotechnology 24:476-480. 8. Kaper, J.B., Morris, J.G., Jr., and Levine, M.M. 1995. Cholera. Clin. Microbiol. Rev. 8:48-86. 9. Sack, D.A., Sack, R.B., Nair, G.B., and Siddique, A.K. 2004. Cholera. Lancet 363:223233. 10. Gaffga, N.H., Tauxe, R.V., and Mintz, E.D. 2007. Cholera: a ne w homeland in Africa? Am. J. Trop. Med. Hyg. 77:705-713. 11. Broeck, D.V., Horvath, C., and De Wolf, M. 2007. Pathogens in focus Vibrio Cholerae: Cholera toxin. The International Journal of Biochemistry & Cell Biology 39:1771-1775. 12. Sanchez, J. and Holmgren, J. 2005. Viru lence factors, pathogenesis and vaccine protection in cholera and ETEC diarrhea. Curr. Opin. Immunol. 17:388-398. 13. 2001. Cholera vaccines. Wkly. Epidemiol. Rec. 76:117-124.

PAGE 155

155 14. Ansaruzzaman, M., Bhuiyan, N.A., Safa, A., Sultana, M., McUamule, A., Mondlane, C., Wang, X.Y., Deen, J.L., von Seidlein, L., Cl emens, J.D., Lucas, M., Sack, D.A., and Balakrish, N.G. 2007. Genetic dive rsity of El Tor strains of Vibrio cholerae O1 with hybrid traits isolated from Bangladesh and Mozambique. Int. J. Med. Microbiol. 297:443449. 15. Almeida, R.J., Hickman-Brenner, F.W., So wers, E.G., Puhr, N.D., Farmer, J.J., III, and Wachsmuth, I.K. 1990. Comparison of a latex a gglutination assay and an enzyme-linked immunosorbent assay for detecting cholera toxin. J. Clin. Microbiol. 28:128-130. 16. Yam, W.C., Lung, M.L., and Ng, M.H. 1992. Evaluation and optimization of a latex agglutination assay for dete ction of cholera toxin and Escherichia coli heat-labile toxin. J. Clin. Microbiol. 30:2518-2520. 17. Andrade, J.R., da Silva, J.L., Barbosa, M.F., and Caldas, C.A. 1992. Preliminary report on the application of the coagglutination test for rapid diagnosis of cholera in the upper Solimoes River area in the Brazilian Amazon region. Braz. J. Med. Biol. Res. 25:375378. 18. Carillo, L., Gilman, R.H., Mantle, R.E., N unez, N., Watanabe, J., Moron, J., Quispe, V., and Ramirez-Ramos, A. 1994. Rapid detection of Vibrio cholerae O1 in stools of Peruvian cholera patients by using monoc lonal immunodiagnostic kits. Loyaza Cholera Working Group in Peru. J. Clin. Microbiol. 32:856-857. 19. Colwell, R.R., Hasan, J.A., Huq, A., Loomis L., Siebeling, R.J., Torres, M., Galvez, S., Islam, S., Tamplin, M.T., and Bernstein, D. 1992. Development and evaluation of a rapid, simple, sensitive, monoclonal antibody-based co-a gglutination test for direct detection of Vibrio cholerae 01. FEMS Microbiol. Lett. 76:215-219. 20. Jesudason, M.V., Thangavelu, C.P., and Lalitha, M.K. 1984. Rapid screening of fecal samples for Vibrio cholerae by a coagglutination technique. J. Clin. Microbiol. 19:712713. 21. Rahman, M., Sack, D.A., Wadood, A., Yasmin, M., and Latif, A. 1989. Rapid identification of Vibrio cholerae serotype O1 from primary isolation plates by a coagglutination test. J. Med. Microbiol. 28:39-41. 22. Agarwal, V., Biswas, M., Pathak, A.A., and Saoji, A.M. 1995. Rapid detection of Vibrio cholerae 0139 in faecal specimens by coagglutination. Indian J. Med. Res. 101:55-56. 23. Hasan, J.A., Huq, A., Nair, G.B., Garg, S., Mukhopadhyay, A.K., Loomis, L., Bernstein, D., and Colwell, R.R. 1995. Development and testing of monoclonal antibody-based rapid immunodiagnostic test kits for direct detection of Vibrio cholerae O139 synonym Bengal. J. Clin. Microbiol. 33:2935-2939.

PAGE 156

156 24. Qadri, F., Hasan, J.A., Hossain, J., Chowdhury, A., Begum, Y.A., Azim, T., Loomis, L., Sack, R.B., and Albert, M.J. 1995. Evaluati on of the monoclonal antibody-based kit Bengal SMART for rapid detection of Vibrio cholerae O139 synonym Bengal in stool samples. J. Clin. Microbiol. 33:732-734. 25. Hasan, J.A., Bernstein, D., Huq, A., Loomis, L., Tamplin, M.L., and Colwell, R.R. 1994. Cholera DFA: an improved direct fluorescen t monoclonal antibody staining kit for rapid detection and enumeration of Vibrio cholerae O1. FEMS Microbiol. Lett. 120:143-148. 26. Hasan, J.A., Huq, A., Tamplin, M.L., Siebeling, R.J., and Colwell, R.R. 1994. A novel kit for rapid detection of Vibrio cholerae O1. J. Clin. Microbiol. 32:249-252. 27. Nato, F., Boutonnier, A., Rajerison, A., Grosjean, P., Dartevelle, S., Guenole, A., Bhuiyan, N.A., Sack, D.A., Nair, G.B., Four nier, J.A., and Chanteau, S. 2003. One-step immunochromatographic dipstick tests for rapid detection of Vibrio cholerae O1 and O139 in stool samples. Clinical and Diagnostic Laboratory Immunology 10:476-478. 28. Nataro, J.P. and Kaper, J.B. 1998. Diarrheagenic Escherichia coli Clin. Microbiol. Rev. 11:142-201. 29. Hussein, H.S. 2007. Prevalence and pa thogenicity of Shiga toxin-producing Escherichia coli in beef cattle and their products. J. Anim Sci. 85:E63-E72. 30. Mead, P.S., Slutsker, L., Dietz, V., McCai g, L.F., Bresee, J.S., Shapiro, C., Griffin, P.M., and Tauxe, R.V. 1999. Food-related illne ss and death in the United States. Emerg. Infect. Dis. 5:607-625. 31. Boyd, B. and Lingwood, C. 1989. Verotoxin rece ptor glycolipid in human renal tissue. Nephron 51:207-210. 32. Robinson, L.A., Hurley, R.M., Lingwood, C., and Matsell, D.G. 1995. Escherichia coli verotoxin binding to human paediatric glomerular mesangial cells. Pediatr. Nephrol. 9:700-704. 33. Takeda, T., Dohi, S., Igarashi, T., Yama naka, T., Yoshiya, K., and Kobayashi, N. 1993. Impairment by verotoxin of tubular function contributes to the renal damage seen in haemolytic uraemic syndrome. J. Infect. 27:339-341. 34. O'Brien, A.D., Tesh, V.L., Donohue-Rolfe A., Jackson, M.P., Olsnes, S., Sandvig, K., Lindberg, A.A., and Keusch, G.T. 1992. Shiga to xin: biochemistry, genetics, mode of action, and role in pathogenesis. Curr. Top. Microbiol. Immunol. 180:65-94. 35. Strockbine, N.A., Marques, L.R., Ne wland, J.W., Smith, H.W., Holmes, R.K., and O'Brien, A.D. 1986. Two toxin-converting phages from Escherichia coli O157:H7 strain 933 encode antigenically distinct toxins with similar biologic activities. Infect. Immun. 53:135-140.

PAGE 157

157 36. Yamagami, S., Motoki, M., Kimura, T., Izumi, H., Takeda, T., Katsuura, Y., and Matsumoto, Y. 2001. Efficacy of postinfection tr eatment with anti-Shiga toxin (Stx) 2 humanized monoclonal antibody TM A-15 in mice lethally challenged with Stx-producing Escherichia coli J. Infect. Dis. 184:738-742. 37. Ostroff, S.M., Tarr, P.I., Neill, M.A., Le wis, J.H., Hargrett-Bean, N., and Kobayashi, J.M. 1989. Toxin genotypes and plasmid profiles as determinants of systemic sequelae in Escherichia coli O157:H7 infections. J. Infect. Dis. 160:994-998. 38. Boerlin, P., McEwen, S.A., Boerlin-Petzo ld, F., Wilson, J.B., Johnson, R.P., and Gyles, C.L. 1999. Associations between virule nce factors of Shiga toxin-producing Escherichia coli and disease in humans. J. Clin. Microbiol. 37:497-503. 39. Tesh, V.L., Burris, J.A., Owens, J.W., Gordon, V.M., Wadolkowski, E.A., O'Brien, A.D., and Samuel, J.E. 1993. Comparison of the relative toxicities of Shig a-like toxins type I and type II for mice. Infect. Immun. 61:3392-3402. 40. Donohue-Rolfe, A., Kondova, I., Oswal d, S., Hutto, D., and Tzipori, S. 2000. Escherichia coli O157:H7 strains that express Sh iga toxin (Stx) 2 alone are more neurotropic for gnotobiotic pi glets than are isotypes producing only Stx1 or both Stx1 and Stx2. J. Infect. Dis. 181:1825-1829. 41. Magliulo, M., Simoni, P., Gu ardigli, M., Michelini, E., Luciani, M., Lelli, R., and Roda, A. 2007. A rapid multiplexed chemiluminescent immunoassay for the detection of Escherichia coli O157:H7, Yersinia enterocolitica Salmonella typhimurium and Listeria monocytogenes pathogen bacteria. J. Agric. Food Chem. 55:4933-4939. 42. Wang, L.J., Wu, C.S., Hu, Z.Y., Zha ng, Y.F., Li, R., and Wang, P. 2008. Sensing Escherichia coli O157:H7 via frequency shift through a self-assembled monolayer based QCM immunosensor. J. Zhejiang. Univ Sci. B 9:121-131. 43. Johnston, L.M., Elhanafi, D., Drake, M., and Jaykus, L.A. 2005. A simple method for the direct detection of Salmonella and Escherichia coli O157:H7 from raw alfalfa sprouts and spent irrigation water using PCR. J. Food Prot. 68:2256-2263. 44. Jin, H.Y., Tao, K.H., Li, Y.X., Li, F.Q ., and Li, S.Q. 2005. Microarray analysis of Escherichia coli O157:H7. World J. Gastroenterol. 11:5811-5815. 45. Thevenot, D.R., Toth, K., Durst, R.A., and Wilson, G.S. 2001. Electrochemical biosensors: recommended definitions and classification. Biosens. Bioelectron. 16:121131. 46. Theegala, C.S., Small, D.D., and Monroe, W.T. 2008. Oxygen electrode-based single antibody amperometric biosensor for qualitative detection of E. coli and bacteria in water. J. Environ. Sci. Health A Tox. Hazard. Subst. Environ. Eng 43:478-487.

PAGE 158

158 47. Liu, Y., Gilchrist, A., Zhang, J., and Li, X.F. 2008. Detection of viable but nonculturable Escherichia coli O157:H7 bacteria in drinking water and river water. Appl. Environ. Microbiol. 74:1502-1507. 48. Kalnauwakul, S., Phengmak, M., Kongmu ang, U., Nakaguchi, Y., and Nishibuchi, M. 2007. Examination of diarrheal stools in Hat Yai City, South Thailand, for Escherichia coli O157 and other diarrheagenic Escherichia coli using immunomagnetic separation and PCR method. Southeast Asian J. Trop. Med. Public Health 38:871-880. 49. Stefan, A., Scaramagli, S., Bergami, R., Mazzini, C., Barbanera, M., Perelle, S., and Fach, P. 2007. Real-time PCR and enzyme-linked fluorescent assay methods for detecting Shiga-toxin-producing Escherichia coli in mincemeat samples. Can. J. Microbiol. 53:337-342. 50. Zhang, W., Bielaszewska, M., Pulz, M., Becker, K., Friedrich, A.W., Karch, H., and Kuczius, T. 2008. New immuno-PCR assay for detection of low concentrations of shiga toxin 2 and its variants. J. Clin. Microbiol. 46:1292-1297. 51. Auvray, F., Lecureuil, C., Tache, J., Leclerc, V., Deperrois, V., and Lombard, B. 2007. Detection, isolation and characte rization of Shiga toxin-producing Escherichia coli in retail-minced beef using PCR-based techni ques, immunoassays and colony hybridization. Lett. Appl. Microbiol. 45:646-651. 52. Smith, G.P. 1985. Filamentous fusion phage: novel expression vectors that display cloned antigens on the virion surface. Science 228:1315-1317. 53. Carmen, S. and Jermutus, L. 2002. Concepts in antibody phage display. Brief. Funct. Genomic. Proteomic. 1:189-203. 54. Endemann, H. and Model, P. 1995. Location of filamentous phage minor coat proteins in phage and in infected cells. Journal of Molecular Biology 250:496-506. 55. Henry, T.J. and Pratt, D. 1969. The proteins of bacteriophage M13. Proc. Natl. Acad. Sci. U. S. A 62:800-807. 56. Russel, M. and Model, P. 1989. Genetic analysis of the fila mentous bacteriophage packaging signal and of the prot eins that interact with it. J. Virol. 63:3284-3295. 57. Messing, J. 1991. Cloning in M13 phage or how to use biology at its best. Gene 100:312. 58. Kwasnikowski, P., Kristensen, P., an d Markiewicz, W.T. 2005. Multivalent display system on filamentous bacteriopha ge pVII minor coat protein. J. Immunol. Methods 307:135-143. 59. Gao, C., Mao, S., Lo, C.H., Wirsching, P., Lerner, R.A., and Janda, K.D. 1999. Making artificial antibodies: a format for phage disp lay of combinatorial heterodimeric arrays. Proc. Natl. Acad. Sci. U. S. A 96:6025-6030.

PAGE 159

159 60. Iannolo, G., Minenkova, O., Petruzzelli, R., and Cesareni, G. 1995. Modifying filamentous phage capsid: limits in the size of the major capsid protein. J. Mol. Biol. 248:835-844. 61. Scott,J.K. and Barbas,C.F.I. 2001. Phage display vectors. In: Ph age Display Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 2.1-2.19 pp. 62. Tomlinson, I.M., Walter, G., Marks, J.D., Llewelyn, M.B., and Winter, G. 1992. The repertoire of human germline VH sequences reveals about fifty groups of VH segments with different hypervariable loops. J. Mol. Biol. 227:776-798. 63. Rakonjac, J., Jovanovic, G., and Model, P. 1997. Filamentous phage infection-mediated gene expression: construction and propagation of the gIII deletion mutant helper phage R408d3. Gene 198:99-103. 64. Chasteen, L., Ayriss, J., Pavlik, P., and Bradbury, A.R. 2006. Eliminating helper phage from phage display. Nucleic Acids Res. 34:e145. 65. Baek, H., Suk, K.H., Kim, Y.H., and Cha, S. 2002. An improved helper phage system for efficient isolation of specific an tibody molecules in phage display. Nucleic Acids Res. 30:e18. 66. Rondot, S., Koch, J., Breitling, F., and Dubel, S. 2001. A helper phage to improve singlechain antibody presentation in phage display. Nat. Biotechnol. 19:75-78. 67. Soltes, G., Barker, H., Marmai, K., P un, E., Yuen, A., and Wiersma, E.J. 2003. A new helper phage and phagemid vector system improves viral display of antibody Fab fragments and avoids propagation of insert-less virions. J. Immunol. Methods 274:233244. 68. Soltes, G., Hust, M., Ng, K.K., Bansal, A., Fi eld, J., Stewart, D.I., Dubel, S., Cha, S., and Wiersma, E.J. 2007. On the influence of vector design on antibody phage display. J. Biotechnol. 127:626-637. 69. Wen, S.X., Teel, L.D., Judge, N.A., and O'Brien, A.D. 2006. Genetic toxoids of Shiga toxin types 1 and 2 protect mice agains t homologous but not heterologous toxin challenge. Vaccine 24:1142-1148. 70. Weinstein, D.L., Jackson, M.P., Perera, L.P., Holmes, R.K., and O'Brien, A.D. 1989. In vivo formation of hybrid toxins comprising Shig a toxin and the Shiga-like toxins and role of the B subunit in localiza tion and cytotoxic activity. Infect. Immun. 57:3743-3750. 71. Guyer, M.S., Reed, R.R., Steitz, J.A., and Low, K.B. 1980. Identification of a sex-factoraffinity site in Escherichia coli as gw. Cold. Spring. Harb. Symp. Quant. Biol. 45:135140.

PAGE 160

160 72. Riley, L.W., Remis, R.S., Helgerson, S.D., McGee, H.B., Wells, J.G., Davis, B.R., Hebert, R.J., Olcott, E.S., Johnson, L.M., Harg rett, N.T., Blake, P.A., and Cohen, M.L. 1983. Hemorrhagic colitis associated with a rare Escherichia coli serotype. N. Engl. J. Med. 308:681-685. 73. Karpman, D., Connell, H., Svensson, M., Scheutz, F., Alm, P., and Svanborg, C. 1997. The role of lipopolysaccharide and Sh iga-like toxin in a mouse model of Escherichia coli O157:H7 infection. J. Infect. Dis. 175:611-620. 74. Bik, E.M., Gouw, R.D., and M ooi, F.R. 1996. DNA fingerprinting of Vibrio cholerae strains with a novel insertion sequence element: a tool to identify epidemic strains. J. Clin. Microbiol. 34:1453-1461. 75. Heidelberg, J.F., Eisen, J.A., Nelson, W. C., Clayton, R.A., Gwinn, M.L., Dodson, R.J., Haft, D.H., Hickey, E.K., Peterson, J.D., Um ayam, L., Gill, S.R., Nelson, K.E., Read, T.D., Tettelin, H., Richardson, D., Ermolaeva, M.D., Vamathevan, J., Bass, S., Qin, H., Dragoi, I., Sellers, P., McDonald, L., Utterback, T., Fleishmann, R.D., Nierman, W.C., and White, O. 2000. DNA sequence of both chromosomes of the cholera pathogen Vibrio cholerae. Nature 406:477-483. 76. Evans,T.J. 2000. Septic Shock Methods and Protocols. In Methods in Molecular Medicine. Humana Press, 19-23. 77. Yi, E.C. and Hackett, M. 2000. Rapid is olation method for lipopolysaccharide and lipid A from gram-negative bacteria. Analyst 125:651-656. 78. Tsai, C.M. and Frasch, C.E. 1982. A sensitive silver stain for detecting lipopolysaccharides in polyacrylamide gels. Anal. Biochem. 119:115-119. 79. Sasse, J. and Gallagher, S.R. 2003. Detecti on of proteins on blot transfer membranes. Curr. Protoc. Mol. Biol. Chapter 10:Unit10. 80. Waldor, M.K., Colwell, R., and Mekalanos, J.J. 1994. The Vibrio cholerae O139 serogroup antigen includes an O-antigen capsule and lipopolysaccharide virulence determinants. Proc. Natl. Acad. Sci. U. S. A 91:11388-11392. 81. Villeneuve, S., Boutonnier, A., Mulard, L.A., and Fournier, J.M. 1999. Immunochemical characterization of an Ogawa-Inab a common antigenic determinant of Vibrio cholerae O1. Microbiology 145 ( Pt 9):2477-2484. 82. Westphal, O., Luderitz, O., and Bister, F. 1952. Uber Die Extraktion Von Bakterien Mit Phenol Wasser. Zeitschrift fur Naturforschung Section B-A Journal of Chemical Sciences 7:148-155. 83. Villeneuve, S., Souchon, H., Riottot, M.M ., Mazie, J.C., Lei, P., Glaudemans, C.P., Kovac, P., Fournier, J.M., and Alzari, P.M. 2000. Crystal structure of an anticarbohydrate antibody directed against Vibrio c holerae O1 in complex with antigen: molecular basis for serotype specificity. Proc. Natl. Acad. Sci. U. S. A 97:8433-8438.

PAGE 161

161 84. Adey, N.B., Mataragnon, A.H., Rider, J.E., Carter, J.M., and Kay, B.K. 1995. Characterization of phage that bind plas tic from phage-displayed random peptide libraries. Gene 156:27-31. 85. Persell, D.J. and Robinson, C.H. 2008. Det ection and early identific ation in bioterrorism events. Fam. Community Health 31:4-16. 86. 2001. Update: Investigation of bioterrorism-related anthrax and interim guidelines for clinical evaluation of pers ons with possible anthrax. MMWR Morb. Mortal. Wkly. Rep. 50:941-948. 87. Nakamura, M., Watanabe, H., Nishimiya, Y., Tsumoto, K., Ishimura, K., and Kumagai, I. 2001. Panning of a phage VH library using nitrocellulose membranes: application to selection of a human VH library. J. Biochem. 129:209-212. 88. Yarmush, M.L., Antonsen, K.P., Sundaram, S., and Yarmush, D.M. 1992. Immunoadsorption: strategies for antigen el ution and production of reusable adsorbents. Biotechnol. Prog. 8:168-178. 89. Nagumo, Y., Oguri, H., Tsumoto, K., Shindo, Y., Hirama, M., Tsumuraya, T., Fujii, I., Tomioka, Y., Mizugaki, M., and Kumagai, I. 2004. Phage-display selection of antibodies to the left end of CTX3C using synthetic fragments. J. Immunol. Methods 289:137-146. 90. Goletz, S., Christensen, P.A., Kristense n, P., Blohm, D., Tomlin son, I., Winter, G., and Karsten, U. 2002. Selection of large divers ities of antiidiotypic antibody fragments by phage display. J. Mol. Biol. 315:1087-1097. 91. Meulemans, E.V., Slobbe, R., Wasterval, P., Ramaekers, F.C., and van Eys, G.J. 1994. Selection of phage-displayed antibodies specifi c for a cytoskeletal antigen by competitive elution with a monoclonal antibody. J. Mol. Biol. 244:353-360. 92. Zhu, Y., Ho, B., and Ding, J.L. 2003. Seque nce and structural diversity in endotoxinbinding dodecapeptides. Biochim. Biophys. Acta 1611:234-242. 93. Thomas, C.J., Sharma, S., Kumar, G. Visweswariah, S.S., and Surolia, A. 2003. Biopanning of endotoxin-specific phage displayed peptides. Biochem. Biophys. Res. Commun. 307:133-138. 94. Noda, K., Yamasaki, R., Hironaka, Y., and Kitagawa, A. 2001. Selection of peptides that bind to the core oligosaccharide of R-fo rm LPS from a phage-displayed heptapeptide library. FEMS Microbiol. Lett. 205:349-354. 95. Harlow,E. and Lane,D. 1988. Antibodies a Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spri ng Harbor, NY. 23-34 pp. 96. Paoli, G.C., Chen, C.Y., and Brewst er, J.D. 2004. Single-chain Fv antibody with specificity for Listeria monocytogenes J. Immunol. Methods 289:147-155.

PAGE 162

162 97. Leonard, P., Safsten, P., Hearty, S., McDonnell, B., Finlay, W., and O'Kennedy, R. 2007. High throughput ranking of recombinant av ian scFv antibody fragments from crude lysates using the Biacore A100. J. Immunol. Methods 323:172-179. 98. Su, B., Hrin, R., Harvey, B.R., Wang, Y. J., Ernst, R.E., Hampton, R.A., Miller, M.D., Strohl, W.R., An, Z., and Montgomery, D.L. 2007. Automated high-throughput purification of antibody fragment s to facilitate evaluation in functional and kinetic based assays. J. Immunol. Methods 322:94-103. 99. Bannister, D., Wilson, A., Prowse, L., Walsh, M., Holgate, R., Jermutus, L., and Wilkinson, T. 2006. Parallel, high-throughput pu rification of recombinant antibodies for in vivo cell assays. Biotechnol. Bioeng. 94:931-937. 100. Lobel, L.I., Morseman, J.P., Zeng, X., Lust bader, J.W., Chen, H., and Allnutt, F.C. 2001. Development of a fluorescence based high thr oughput assay for antagon ists of the human chorionic gonadotropin receptor extracellular domain: analysis of peptide inhibitors. J. Biomol. Screen. 6:151-158. 101. Vanhercke, T., Ampe, C., Tirry, L., and Denol f, P. 2005. Rescue and in situ selection and evaluation (RISE): a method for high-thr oughput panning of phage display libraries. J. Biomol. Screen. 10:108-117.

PAGE 163

163 BIOGRAPHICAL SKETCH Crystal Harpley (Mazur) was born in Colu m bus, OH; and shortly after moved to Wellington, FL, where she completed grade scho ol. In 2006 she graduated with high honors from the University of Florida, with a major in microbiology a nd cell sciences and a minor in chemistry. During the process of finishing her masters thesis, Crystal Harpley got married and is now Crystal Mazur.