Phylogenetic analysis of symbionts in feather-feeding lice of the genus Columbicola: evidence for repeated symbiont repl...

MISSING IMAGE

Material Information

Title:
Phylogenetic analysis of symbionts in feather-feeding lice of the genus Columbicola: evidence for repeated symbiont replacements
Physical Description:
Mixed Material
Language:
English
Creator:
Smith, Wendy A.
Oaksen, Kelly F.
Johnson, Kevin P.
Reed, David L.
Carter, Tamar
Smith, Kari L.
Koga, Ryuichi
Publisher:
BioMed Central (BMC Evolutionary Biology)
Publication Date:

Notes

Abstract:
Background: Many groups of insects have obligate bacterial symbionts that are vertically transmitted. Such associations are typically characterized by the presence of a monophyletic group of bacteria living in a well-defined host clade. In addition the phylogeny of the symbiotic bacteria is typically congruent with that of the host, signifying co-speciation. Here we show that bacteria living in a single genus of feather lice, Columbicola (Insecta: Phthiraptera), present an exception to this typical pattern. Results: The phylogeny of Columbicola spp. symbionts revealed the presence of three candidate clades, with the most species-rich clade having a comb-like topology with very short internodes and long terminal branches. Evolutionary simulations indicate that this topology is characteristic of a process of repeated symbiont replacement over a brief time period. The two remaining candidate clades in our study exhibit high levels of nucleotide substitution, suggesting accelerated molecular evolution due to relaxed purifying selection or smaller effective population size, which is typical of many vertically transmitted insect symbionts. Representatives of the fast-evolving and slow-evolving symbiont lineages exhibit the same localization, migration, and transmission patterns in their hosts, implying direct replacement. Conclusions: Our findings suggest that repeated, independent symbiont replacements have taken place over the course of the relatively recent radiation of Columbicola spp. These results are compatible with the notion that lice and other insects have the capability to acquire novel symbionts through the domestication of progenitor strains residing in their local environment. Keywords: Symbiosis, Insect, Lice, Co-speciation, Symbiont replacement
General Note:
Publication of this article was funded in part by the University of Florida Open Access publishing Fund. In addition, requestors receiving funding through the UFOAP project are expected to submit a post-review, final draft of the article to UF's institutional repository, IR@UF, (www.uflib.ufl.edu/UFir) at the time of funding. The institutional Repository at the University of Florida community, with research, news, outreach, and educational materials.
General Note:
Smith et al. BMC Evolutionary Biology 2013, 13:109 http://www.biomedcentral.com/1471-2148/13/109; Pages 1-15
General Note:
doi:10.1186/1471-2148-13-109 Cite this article as: Smith et al.: Phylogenetic analysis of symbionts in feather-feeding lice of the genus Columbicola: evidence for repeated symbiont replacements. BMC Evolutionary Biology 2013 13:109.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
All rights reserved by the source institution.
System ID:
AA00016104:00001

Full Text





U A GG.
GAUem G 00
SAAUGU. AUCUGGAGGAA UA CCAUX
C U A. GAUAUs.
GI I I- I-I 0I 1 I U G.
UGAOGA GU GGAOOU UAAGd *A UGGu
*C d
G C
G 0
G-C A G A
G-0

G AuG AGcA
-C G-O
G -A A A-
J AA A GG
C C A
U A A A
G 0 0 GA
GA U A G u
A G GG-C
S AO AA U A G-C
AA CC UGGGA UGCAUUUGAG CAJG GGCA GGO C U -A
I I1-.i l I 1 G d
G O GU// U U.
UC GSGUO CU AJUAUA GA U UGU C UG GUAU0 G U -*U
UAA *AGG G U AA UA
*q. //U A A


ATAG
A C
UUA *-Au
GA^
CG. L
A J


*. GuA AA-
U AG U U AL
0 *l I
U CG G U(
A GA AG


AU CCUGAU
.GSGAGU A
* AGG


"AA (1)
GA.


A
G
G- C0
mA- UA
*A
G-UA
*C* A
A CA
A A

A *AG AG *
GUU GGGG GAG
UGA UGAACG GCC UU
G C**"
G
A
*A A G
GJU GOGG COGAG
UGA UG AOG GOO ii
***. G G


UA AG
U G
J J
G-C
GO-
GO-
U G GA C AAC
AG G C
U ACGAGC CC U UA-Udd UuuGU UGCCCSA U
U I I I I I I l I I I *
1(2) G UCUUCG AGCAG A AAGGACU U A C .C AU
A A GJ
A U C G G 1(1)
A U U-U A
ASG G C G
U- GA- G-C
U *U -A -U
G 0U0 A 0 G
U A G 0:
uUC U G U .
J.J A J

U~~~ GU C
oC-e .A-oU" .u-
U GG A CU_ A
C G

AU A A
GA UGAGU G AA U G UA A
0 U f G. *U-A-O

A I C l II /CuU. C-G
ACGGAUCAG A U C
A JASA I-C
^^r~u
AA
G AGAC AG' G A G--0


U A
-3 mA G/I G-- U
UG~ ~ UU. d
0~ X A'


GU AG A J
U-U
u- UU d -- A
U~\G Ad d (10 GG

* G 0 "
J UA


....... \ U-^ ^ U-A
r II1 1 oA .uGO-uAAU G-A A uU AA
CG A GU A A G U O *


G 0 1 JUA UG dG0 J,
G--* u' UGAGG-AU AAGA\ G
U [Ae UA 0 G: AA AGA.\ OrAti
G"Il )O I G I *A QU^O AG
GGG UCWOd t. A U -A A A*


ouoN uue, I /eu,,(L^ e ,uc
I- I C -il d GOA 11 *iU A
A AAAJJ UGGAA. GA UA *5U G GU AA
AG I) GG.\ C G A
N AAUJUGAOGGGGfUlUdGAOAAG\JA^ U




|N N N I I I I I I -III N Ab A/ L *
G r U A *
ACGC CN AOJJOG JO J AG GG (

N G O U NG U C OG d A U UAC'GA





Cil A. NNN N NN G ^
N I U G A A
N NNC A A A G J
N G N N UGA 0
N U N AAG

G N GNNN A G AA
5' J N N *U A U G.
O NN NNNNNNNNNNNr UG { -5iO
AoGGN N A *A J
INN 0 IG I
O--G N *UAA

Ui*G N
GO N
A AU N
A G N
N A NA
GUI 3'
G C
JN A
UG
eu *G1 Columbicola veigasimoni
A J j Insertion with number of bases in parenthesc
S A A-d Deletion

G 0 Disrupting non-canonical pairing (n 85)
anonical pairing (n 48)
GU
G--C I






GU G) Q Conservative non-canonical G:A pairing (n=(
A A I







U- Conservative non-canonical G:U pairing (n 4

A AS 0 Site with variability below average for all sites
SGd
A Site with variability above average for all siteI

C-
JG
-_ A








A U Probable tertiary interactions are connected with solid hnfl
d u.G
du UC
JO neto ihnmero ae nprnhs


es





6)

*1)
(v<) (n 148)
s (v>i)(n 137)


es


AG









uAG
i A G
G G CAG( iUG
AAuGC A UCUGGAGGAAUA AGCG
1 1 I 1 1 1 -3 I I I -1
GUGGCGA AUGGACCUUAAGAU iGG CGGA
G C
*A d
G d
G-C
GC

A. A
A --
GA AG
U A
*G C.
*C-G.
i AAG
C G
U- A
G UG
G U
A U
A G ( 3
S AC GA U A
AACCUGGGA UGCAU UUGA C] GGUA GGC C
c i -" i i i A GA I I I I Ui i Gi G C
U UGUGCCCC G U GUA GACU G U UGiC UGG // "UeAu3
AA G G,


NNNNNN NN
N
N
IN
IN
IN
IN
IN

5'


G UA
G- C"
U- A
G-CA
CG A-U
A G C G
ic/-GA G
U U/G C
CAC GC A
:u~A
AUG A GA AG *..
Gr OiIG GUO GOGGO CGGGAAAAGC%
AG U 'I I I 1 I U I I
A uCAAGuc'G CGd CCA AG GUC UUUCUACG
A GGAuGA A G G U U A
Gil"//CA *AO G
GiUUA / / CA, G C A A
Gi / AUGGG

A
l(6)'dg G A3



A P A
G G -G
AC G iU
G\G c CC -G
A "UoC C U:
U \ GA
C \iGA A UA

GC^ ~ GC
G *CG" A C
UF \ U G U U
(3 .. GG A (3
u c"' ,"ri -c.- o
A A U C -G
U A
A U
C- G A U
( d A *
'0GGGCCiCAC A GGGC ACUAUUGGA
.G GG O CU CAO C GG
e U Io I Ii lU*** I III I A
UCA GGGGUG CG U C CGA UGGCA
*AG A A 'A i
A UiAAU
A A
Cc
.C -- G
A- U
G-C
A G
AC


UA AG
U G
i i
G-C
G-C
G--C
U GO A
U A G~ AC AGC
U A CGA AC CC UUA-UC CUU uGU U C C d U GA U
G I III* ..1 I~ o Il l I I I U "
U.G UGCUCG AGGGGU AGAAAACu AGGGCUGGS
A-U i (, A U G UG A
A-U 'c G *C
A G U A A A
A U G- C A
C G ,C C
U U (G %. U
G U -A A 0 G
U A *C AG
e-uc, i1o!



GO A A U-A A G
GA UGC CUU GG U G U
111 1* 1 uC C -G
AG ACGAAUCAAG /uA uG C


C GA /G G ,A
Go2 eCGA GC i-A

G G G A
A C U -



G G A A i
U G C. U A -
G 'U C oO U U CA.
i A G A U G
A A A
5U<^ \ A\u ,uUA


A C C\i iUG- AAG G
A^ c i^ ^ AAAGA\ 5J^4C
A CAG C-G C \

C. *A *C 0-OiC
UGCCG A-iU G A A, A uCG
IU1 C I G 0 A AA A
GG -U AG dC



UUGGAA ^G--OAA 'C -O\ G U
A GGUJ UGU\cU AAA

GA U UAACGAGudCC AC AAG \\ iUA U.C(

II I1" I.. I 110 I1I CG A
AU UGU UCCGGA UiG U G A
CA A GAJ A GUC
G AA U UA AGAA3





SA A G CU CA
U G A*
A A A G
C N GA A A
O A N GC/G GA i

SNNNNNNNNNNN^ A''' /G u "A
G N N eC O^ ^U
C N NNNNNNNNNNN G C A Ci
"C r" A "AA 0 G
C GG
A G N
G G N
AG N U N

G--C
G A N
A G
G i
e I II C A








i A NC
G C
: Columbicola paradoxus
C G
A i
A--U









A- u i Insertion with number of bases in parentheses
A c d Deletion
G A
A
A U--
G C Disrupting non-canonical pairing (n=30)
IN A G U:










A U I- Canonical pairing (n=40)
G- i
G O
U-G O Conservative non-canonical G:A pairing (n=8)
A
*A- u Conservative non-canonical G:U pairing (n=20)
C- G
G G Site ,ith variability below average for all sites (v<1) (n=70)
Ai G I











A -G Site with variability above average for all sites (v>1) (n=84)
CU G
U C

Probable tertiary interactions are connected with solid lines


i. I-I



























AC AA

AA
AACCUGGGA UGCAUUUGA C G
gC I Io 1 1 I .
UOGG GCC CC GUGUAGAC U G U
0 UA A A


NNNNNNNNNN
NNNNN NN
N
N
N
N
N
N
N

5,


G UA

U A
G-CA
COG A-U
A G C G
u GA CAG
CC (3GG A

CA O A GA
GA AU GUO GO(
A(3 U I I1 I I
A GuCAAGGuC CGG CGA
A GGAGAC A G
GUo CA3 CA \, A A
UG iO// A 0 G C
iG G///G 0UG G G0A

CG A
CC
OC U \ \ G U G u3
A
G U C A
A *,,C O-G
A A~uiA
U \ A 0G
'C \ U (5 A -UA

G U 0
G C*
C A*
U A
A U
O-G
S3-G. GUA A
C GGGCCUCAC A CGGGG ACUAUUGGA
U. .I .I IIIIII I11 .1111.1 A
UuCC GGAGUG CG CC CGA UAGCA
AG A SG A U
A A A iA
A A
*C A.
A- U
G-C
A G
AC


AG ** *
GGO OGG GSA A GC
1 i I II I *i : i 8)
&OG GCC UUUUCUUC-
U G .. ..


UA AG
U G
i i
G-C
G-C
GOGOC
U *G CA GJA AC
oA GC JJ G oAGuu
SA CGA GC CC UUA-iOCiiiGO GAUu
G I I I I I I I. I I I I I I I / I I II I C.
U G UGCUCG AGGGGU AGGAAA A GG GOUSSG
uA U A U G OACUCA o
A U C G G G
A G U A A A
G U G C A
C G G C
U *U G *U
G C A G
U-AA G U
UA C- G G -C
U G G i C C G
C -G A C G
O G G CU A U
GOA AU A A G
*** U*GI CG U A A
GA UGCCUUGGS U G A
.I I I I I* U C-G
AG ACGGAUCAAG/A J UG0
G/GA G --C


A
c U U UC -- 3U


G AG0 A
uG\ CGA AC CU UAC
ce.. c cCUU U--
G G A G3
*AA 'CA
iiG \G CA

GG GO J -U AAU G CA I CG
i AA U AUCi\ G
AC UA G-C A C\ GA
U U A G

0A 0C AA AG\ 0(A0
G O A U "G \

GGCCGC A-U A CCC AG
1 G 0 A *A. A
UUGGA G-CAA G U
UGG ,U -C UG GC_G U (A AAU
/ GG.\ \ L
SU UG A oGCtGGO OGOAOAAG \\CUA C
G A
AOCUGUUOC GG0 UG U G A
C A uio A G
A G GUe C u 0U
u u G G
C A A G
C N N A A A A

G N N G e ( .
O N A c 0uT-
SN NNNNN G O A O
i N N G i
U N G C AU C -A
C N N AG C GC
A N N iA 0
O N N CAA
C N N
A NN
U G N
G U N
GiC N
G A N
A G N
G iU I
U- A N
G-C
GiU 3'
G- U
i A
U--A Columbicola columbae
C- G
A u j Insertion with number of bases in parentheses
A--i
AA- UC d Deletion
G A
A c Disrupting non-canonical pairing (n=9)
i A 0
A i Canonical pairing (n=50)
G- U
G3 0 Conservative non-canonical G:A pairing (n=3)
i G0
A
G -C Conservative non-canonical G:U pairing (n=7)
C G
u : G Site with variability below average for all sites (v
0C A
A G Site with variability above average for all sites (v>) (n
OC -G
U G
UUr
Probable tertiary interactions are connected with solid lines


0CA
G o^'
G \GA
U G
AAUG


22)

-86)




Full Text

PAGE 1

Vibrio cholerae (53.7%) Yersinia pestis (54.3%) Dickeya dadantii (55.2%) Pantoea agglomerans (55.4%) Enterobacter hormaechi (55.1%) PE ( C. arnoldi ex Macropygia nigrirostris ) (53.3%) 2 Escherichia coli (54.8%) Shigella exneri (54.8%) Salmonella enterica (54.8%) E ( Metaseiulus occidentalis mite) (55.5%) PE ( Physconelloides zenaidurae avian bodylouse) (54.1%) 47! PE ( C. wombeyi ex Geophaps plumifera ) (52.7%) 46 PE ( C. baculoides ex Zenaida macroura ) (49.2%) 4! PE ( C. macrourae 1 ex Leptotila verreauxi ) (51.3%) 24! PE ( C. macrourae 1 ex Geotrygon montana ) (51.5%) 22! PE ( C. macrourae 1 ex Leptotila plumbeiceps ) (51.4%) 23! 55/ 0.56 60/0.72 71/ 0.70 53/0.62 B Buchnera aphidicola PE ( Acyrthosiphon pisum aphid) (50.1%) Photorhabdus luminescens (55.2%) Arsenophonus arthropodicus SE ( Pseudolynchia canariensis louse y) (49.5%)! PE ( Pediculus humanus human body louse) (48.8%)! PE ( Pediculus capitis human head louse) (53.2%)! SE ( Bactericella cockerelli psyllid) (54.4%)! Sodalis glossinidius SE ( Glossina morsitans tsetse y) (54.8% ) PE ( C. extinctus ex Patagioenas fasciata ) (53.8%) #15! PE ( C. macrourae 3 ex Zenaida macroura ) (52.1%) #26! PE ( C. macrourae 4 ex Zenaida galapagoensis ) (51.7%) #27! SE ( Paracoccus northofagicola mealy bug) (51.0%)! PE ( Pseudolynchia canariensis louse y) (49.5%)! Wigglesworthia glossinidia PE ( Glossina brevipalpis tsetse y) (51.0%)! PE ( Haematomyzus elephantis elephant louse) (45.0%)! PE ( Haematopinus eurysternus cattle louse) (47.8%)! PE ( Sitophilus rugicollis grain weevil) (52.1%)! 98/ 1.0 81/ 0.81 Blochmannia oridanus PE ( Camponotus oridanus carpenter ant) (47.7%)! Blochmannia pensylvanicus PE ( Camponotus pensylvanicus carpenter ant) (48.5%) PE ( C. rodmani ex Geopelia humeralis ) (54.5%) #38! PE ( C. fortis ex Otidiphaps nobilis ) (54.3%) #16! SE ( Cantao occelatus stinkbug) (55.3%)! PE ( C. clayae ex Treron waalia ) (54.8%) #6! PE ( C. paradoxus ex Lopholaimus antarcticus ) (49.7%) #37! PE ( C. claytoni ex Ducula rugaster ) (54.8%) #7! PE ( C. claytoni ex Ducula rugaster ) (54.8%) #8! PE ( C. claviformis ex Columba palumbus ) (54.5%) #5! PE ( C. bacillus ex Streptopelia decaocto ) (53.4%) #3! PE ( C. malenkeae ex Ducula pacica ) (54.1%) #28 PE ( C. elbeli ex Treron sieboldii ) (53.2%) #11! PE ( C. exilicornis 1 ex Macropygia amboinensis ) (54.8%) #13! PE ( C. exilicornis 1 ex Macropygia amboinensis ) (54.8%) #12! PE ( C. mjoebergi ex Geopelia placida ) (53.5%) #32! PE ( C. mjoebergi ex Geopelia striata ) (54.5%) #33! PE ( C. mjoebergi ex Geopelia striata ) (54.5%) #34! PE ( C. timmermanni ex Leptotila rufaxilla ) (53.6%) #39! PE ( C. timmermanni ex Leptotila rufaxilla ) (53.6%) #40! PE ( C. harbisoni ex Phaps histrionica ) (52.3%) #20! PE ( C. waggermani ex Patagioenas leucocephala ) (53.3%) #43! PE ( C. adamsi ex Patagioenas plumbae ) (53.3%) #1! PE ( C. guimaraesi 2 ex Chalcophaps indica ) (54.5%) #19! PE ( C. guimaraesi 1 ex Chalcophaps indica ) (54.5%) #18! PE ( C. gracilicapitis ex Leptotila jamaicensis ) (53.7%) #17! PE ( C. passerinae 2 ex Claravis pretiosa ) (54.7%) #35! PE ( C. passerinae 2 ex Claravis pretiosa ) (54.7%) #36! PE ( C. waltheri ex Geotrygon frenata ) (54.1%) #45 PE ( C. exilicornis 2 ex Phapitreron amethystinus ) (52.5%) #14! PE ( C. mckeani ex Ocyphaps lophotes ) (52.9%) #31! PE ( C. macrourae 2 ex Zenaida asiatica ) (54.3%) #25! PE ( C. masoni ex Petrophassa albipennis ) (53.8%) #29! 75 Strain HS (53.1%)! SE ( Craterina melbae louse y) (55.6%)! SE ( Curculio sikkimensis chestnut weevil) (55.4%) PE ( Sitophilus granarius grain weevil) (54.4%)! PE ( Sitophilus oryzae grain weevil) (54.9%) PE ( Sitophilus zeamais grain weevil) (55.1%)! PE ( C. tschulyschman ex Columba livia ) (54.2%) #41! PE ( C. columbae ex Columba livia ) (54.4%) #9! PE ( C. columbae ex Columba livia ) (54.4%) #10! PE ( C. veigasimoni ex Phapitreron leucotis ) (47.0%) #42! PE ( C. koopae ex Geophaps scripta ) (55.0%) #21 PE ( C. masoni ex Petrophassa rupennis ) (53.8%) #30! 0.09 substitutions/site! 53/ 0.57 55/ 0.82 74/1.0 89/ 0.84 99/ 1.0 60/ 0.59 88/ 1.0 C A 50/ 0.56



PAGE 1

# Louse Host Location Louse Voucher Code Gene GenBank Accession 1 C. adamsi Patagioenas plumbea Guyana 04.24.99.03 16S rDNA JQ063407 2 C. arnoldi Macropygia nigrirostris Papua New Guinea 05.14.03.05 16S rDNA JQ963434 GroEL JQ063386 3 C. bacillus Streptopelia decaocto Netherlands 11.15.99.01 16S rDNA JQ063440 4 C. baculoides Zenaida macroura USA 10.19.98.01 16S rDNA JQ063412 5 C. claviformis Columba palumbus UK 01.20.03.16 16S rDNA JQ063427 FusA JQ063396 6 C. clayae Treron waalia Ghana 03 .21.00.09 16S rDNA JQ063451 FusA JQ063405 7 C. claytoni Ducula rufigaster Papua New Guinea 07.26.04.06 16S rDNA JQ063433 8 C. claytoni Ducula rufigaster Papua New Guinea 08.19.03.14 16S rDNA JQ063430 FusA JQ063395 9 C. columbae Columba livia USA 06.29.98.3 16S rDNA JQ063426 FusA JQ063397 GroEL JQ063388 10 C. columbae Columba livia Australia 07.15.02 16S rDNA JQ063439 11 C. elbeli Treron sieboldii China 06.06.05.04 16S rDNA JQ063437 GroEL JQ063392

PAGE 2

12 C. exi licornis 1 Macropygia amboinensis Papua New Guinea 08.19.03.07 16S rDNA JQ063414 FusA JQ063398 GroEL JQ063387 13 C. exilicornis 1 Macropygia amboinensis Papua New Guinea 08.19.0308 16S rDNA JQ063449 14 C. exilicornis 2 Phapitreron amethystinu s Phillipines 05.26.99.06 16S rDNA JQ063408 15 C. extinctus Patagioenas fasciata USA 01.20.03.01 16S rDNA JQ063416 16 C. fortis Otidiphaps nobilis Papua New Guinea 05.14.03.07 16S rDN A JQ063452 FusA JQ063402 GroEL JQ063390 17 C. gracilicapiti s Leptotila jamaicensis Mexico 09.29.98.04 16S rDNA JQ063444 18 C. guimaraesi 1 Chalcophaps indica Vanuatu 07.26.04.04 16S rDNA JQ063431 19 C. guimaraesi 2 Chalcophaps indica Australia 07.20.04.12 16S rDNA JQ063435 20 C. harbisoni Phaps histrionica Aust ralia 05.14.03.09 16S rDNA JQ063411 21 C. koopae Geophaps scripta Australia 01.08.03.10 16S rDNA JQ063432 22 C. macrourae 1 Geotrygon montana Mexico 09.29.98.01 16S rDNA JQ063436 FusA JQ063406 23 C. macrourae 1 Leptotila plumbeiceps Mexico 10.19.9 8.04 16S rDNA JQ063448 24 C. macrourae 1 Leptotila verreauxi Mexico 10.19.98.02 16S rDNA JQ063446 25 C. macrourae 2 Zenaida asiatica USA 09.29.98.5 16S rDNA JQ063447 26 C. macrourae 3 Zenaida macroura USA 02.01.99.09 16S rDNA JQ063450 GroEL JQ0633 93 27 C. macrourae 4 Zenaida galapagoensis Galapagos 07.01.99.02 16S rDNA JQ063425 28 C. malenkeae Ducula pacifica Vanuatu 01.27.04.02 16S rDNA JQ063417

PAGE 3

29 C. masoni Petrophassa albipennis Australia 05.14.03.13 16S rDNA JQ063441 30 C. masoni Petrophas sa rufipennis Australia 01.27.04.12 16S rDNA JQ063442 FusA JQ063403 31 C. mckeani Ocyphaps lophotes Australia 01.20.03.10 16S rDNA JQ063420 FusA JQ063400 GroEL JQ063391 32 C. mjoebergi Geopelia placida Australia 05.14.03.17 16S rDNA JQ0 63445 FusA JQ063404 33 C. mjoebergi Geopelia striata Hawaii 01.20.03.13 16S rDNA JQ063419 34 C. mjoebergi Geopelia striata Hawaii 03.21.00.05 16S rDNA 35 C. passerinae 2 Claravis pretiosa Mexico 09.29.98.03 16S rDNA JQ063428 36 C. passerinae 2 C laravis pretiosa Mexico 02.01.99.06 16S rDNA JQ063429 37 C. paradoxus Lopholaimus antarcticus Australia 01.27.04.05 16S rDNA JQ063423 38 C. rodmani Geopelia humeralis Australia 05.14.03.12 16S rDNA JQ063443 39 C. timmermanni Leptotila rufaxilla Guyana 0 1.08.03.07 16S rDNA JQ063422 40 C. timmermanni Leptotila rufaxilla Guyana 04.24.99.02 16S rDNA JQ063421 FusA JQ063401 41 C. tschulyschman Columba livia USA 05.07.09.01 16S rDNA JQ063415 42 C. veigasimoni Phapitreron leucotis Phillipines 05.26.99.0 3 16S rDNA JQ063438 43 C. waggermani Patagioenas leucocephala USA 11.15.99.08 16S rDNA JQ063409 44 C. waiteae Columba leucomela Australia 01.27.04.08 FusA JQ063399 GroEL JQ063389 45 C. waltheri Geotrygon frenata Peru 01.20.03.04 16S rDNA JQ063418

PAGE 4

46 C. wombeyi Geophaps plumifera Australia 01.08.03.16 16S rDNA JQ063410 47 Physconelloides zenaidurae Zenaida macroura USA N/A 16S rDNA JQ063413



PAGE 1

RESEARCHARTICLEOpenAccessPhylogeneticanalysisofsymbiontsin feather-feedingliceofthegenus Columbicola : evidenceforrepeatedsymbiontreplacementsWendyASmith1,KellyFOakeson1,KevinPJohnson2,DavidLReed3,TamarCarter3,KariLSmith1,RyuichiKoga4, TakemaFukatsu4,DaleHClayton1andColinDale1*AbstractBackground: Manygroupsofinsectshaveobligatebacterialsymbiontsthatareverticallytransmitted.Such associationsaretypicallycharacterizedbythepresenceofamonophyleticgroupofbacterialivinginawell-defined hostclade.Inadditionthephylogenyofthesymbioticbacteriaistypicallycongruentwiththatofthehost, signifyingco-speciation.Hereweshowthatbacterialivinginasinglegenusoffeatherlice, Columbicola (Insecta:Phthiraptera),presentanexceptiontothistypicalpattern. Results: Thephylogenyof Columbicola spp.symbiontsrevealedthepresenceofthreecandidateclades,withthe mostspecies-richcladehavingacomb-liketopologywithveryshortinternodesandlongterminalbranches. Evolutionarysimulationsindicatethatthistopologyischaracteristicofaprocessofrepeatedsymbiontreplacement overabrieftimeperiod.Thetworemainingcandidatecladesinourstudyexhibithighlevelsofnucleotide substitution,suggestingacceleratedmolecularevolutionduetorelaxedpurifyingselectionorsmallereffective populationsize,whichistypicalofmanyverticallytransmittedinsectsymbionts.Representativesofthefast-evolving andslow-evolvingsymbiontlineagesexhibitthesamelocalization,migration,andtransmissionpatternsintheir hosts,implyingdirectreplacement. Conclusions: Ourfindingssuggestthatrepeated,independentsymbiontreplacementshavetakenplaceoverthe courseoftherelativelyrecentradiationof Columbicola spp.Theseresultsarecompatiblewiththenotionthatlice andotherinsectshavethecapabilitytoacquirenovelsymbiontsthroughthedomesticationofprogenitorstrains residingintheirlocalenvironment. Keywords: Symbiosis,Insect,Lice,Co-speciation,SymbiontreplacementBackgroundManyinsectsmaintainobligate,primaryendosymbiotic bacteriathatprovidenutrientsthatarelackingintheir naturaldiet.Associationsbetweenprimarysymbionts andtheirinsecthostsareoftenancientinorigin,and havefacilitatedtheexploitationofnewecologicalniches byinsects[1].Theverticaltransmissionofprimary symbiontsoftenresultsinhost-symbiontco-speciation, asevidencedbytopologicalcongruencebetweenthe insectandbacterialphylogenies[2,3]. Featherlice(Insecta:Phthiraptera)areobligate,permanentectoparasitesofbirdsandmammalsthatspendtheir entirelifecycleonthehost[4].Thegenus Columbicola contains88describedmorpho-species,allofwhichparasitizecolumbiformbirds(pigeonsanddoves)[5].Someof thesemorpho-speciesarefurtherdividedintomolecularly distinctcrypticspecies[6].Speciesof Columbicola are relativelyhost-specific,withmostknownfromonlya singlespeciesofbirdhost.Transmissionoflicebetween birdsoccursmainlyduringperiodsofdirectcontact,as occursbetweenparentbirdsandtheiroffspringinthenest [7].However, Columbicola arealsoknowntodisperse phoreticallyonhippoboscidlouseflies,whicharewinged parasitesofbirds[7,8].Truetotheirname,featherlice *Correspondence: colin.dale@utah.edu1DepartmentofBiology,UniversityofUtah,257South1400East,SaltLake City,UT84112,USA Fulllistofauthorinformationisavailableattheendofthearticle 2013Smithetal.;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsoftheCreative CommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,distribution,and reproductioninanymedium,providedtheoriginalworkisproperlycited.Smith etal.BMCEvolutionaryBiology 2013, 13 :109 http://www.biomedcentral.com/1471-2148/13/109

PAGE 2

feedprimarilyonfeathers,secretions,deadskinandother dermal “ debris ” [9].Featherspresentanutritionallychallengingdietbecausetheyconsistmostlyofkeratins,which aredifficulttodigestandhaveaminoacidcompositions thataremarkedlybiased[10].Inaddition,theavailability ofvitaminsandco-factorsisexpectedtobelimitedina dietcomprisingmostlyfeathers[11]. Whileabacterialendocellularsymbiontwasobserved microscopicallyintheabdomenof Columbicolacolumbae in1931[12],sequencingandphylogeneticanalysisonly recentlyrevealedthatthisbacteriumisacloserelativeof thetsetseflysymbiont, Sodalisglossinidius ,whichisin turnisamemberofawell-establishedsymbiontcladethat isfoundinadiverserangeofinsecthosts[13]. Insitu hybridizationexperimentsdemonstratedthatthesymbiont of C.columbae ishousedwithinspecializedbacteriocytes andpassedtooffspringviamaternal(ovarial)transmission [13].Thefunctionofthe C.columbae symbiosisiscurrentlyundefined,butitseemslikelythatthesymbiosishas anutritionalbasisbecauseofthefactthatkeratin-rich feathersrepresentanutritionallyincompletediet[14]. Thepurposeofthecurrentstudywastoperformabroad characterizationofbacterialsymbiontdiversityinmore than40membersofthegenus Columbicola ,obtainedfrom pigeonsanddovescollectedinaworldwidesurvey Since Columbicola symbiontsareendocellularinbacteriocytes, wetestedwhetherthesesymbiontsexhibitpatternsofcospeciationtypicaloflongestablished,obligateassociations foundinotherinsects.However,incontrast,ourmolecular phylogeneticanalysesrevealedstrikingdiversityandevolutionarydynamicsinthehost-symbiontassociationsofthis singleinsectgenus.Weproposeandtestseveralhypothesestoaccountfortheseunexpectedfindings.ResultsIdentificationof Columbicola spp.symbiontsWeinitiallysequenced4816SrRNAclonesfromindividualsof C.columbae and C.baculoides .Onlyasingle16S rRNAsequencewasidentifiedfromeachhostspecies.For eachoftheother Columbicola spp.inthestudy,wesequencedaminimumoffour16SrRNAgeneclones.No withinspeciessequenceheterogeneitywasobserved,indicatingthateachofthe Columbicola spp.screenedinthis studyharborsonlyasinglebacterialsymbiont.Structuralanalysisof16SrRNAsequencesof Columbicola spp.symbiontsIntheinitial16SrRNAgenephylogenycontainingallof thesequencesderivedfromthesymbiontsof Columbicola spp.,thesequencesderivedfrom C.veigasimoni and C. paradoxus exhibitedunusuallylongbranches,indicating substantiallyhigherevolutionaryratesthanthesequences oftheother Columbicola spp.symbiontsofthesameclade (Additionalfile1).Inaddition,the C.veigasimoni and C. paradoxus symbiont16SrRNAsequencesexhibitedunusuallylowG+Ccontentsrelativetotheothermembers ofthesameclade(Additionalfile1).Thesepatternssuggestthepossibilitythatthesehighlydivergentsequences mightrepresentnon-functionalcopiesofthe16SrRNA geneinthesesymbiontgenomes. Secondarystructureanalysesofthe16SrRNAsequencesusingahomologymodel[15]revealedthatthe C.veigasimoni and C.paradoxus symbiont16SrRNA sequencesexhibitunusuallyhighratiosofdisruptive: conservativenucleotidesubstitutions.Forexample,in the C.veigasimoni 16SrRNAsequence,85outofatotal of180substitutions(47.2%)arepredictedtoencode disruptivechanges(causingp utativestem-looptransitions;Additionalfile2).Similarly,inthe C.paradoxus 16SrRNAsequence,disruptivesubstitutionscomprise 30outofatotalof98substitutions(30.6%;Additional file3).Incontrast,the C.columbae 16SrRNAsequence thatresidesonarelativelyshortbranchhasonly9outof atotal69substitutions(13%)thatarecharacterizedasdisruptive(Additionalfile4).Indeed,statisticalanalysesshow thattheratiosofdisruptive/conservativesubstitutionsin boththe C.veigasimoni and C.paradoxus symbiont16S rRNAsequencesaresignificantlyhigherthaninthe C. columbae symbiont16SrRNAsequence(Fisher ’ sexact test; P <0.001).Furthermore,usingthe16SrRNAvariabilitymapderivedbyWuyts etal .[16],wedeterminedthat the C.veigasimoni and C.paradoxus symbiont16SrRNA sequencesalsohavesignificantlymoresubstitutionsinsites thatnormallydisplaylowvariability(Fisher ’ sexacttest; P <0.0001).Allthesedatasuggestthatthe C.veigasimoni and C.paradoxus symbiont16SrRNAsequencesarenot evolvinginaccordancewiththefunctionalconstraintsthat affectother16SrRNAsequences,includingthatofthe C.columbae symbiont.However,itshouldalsobenoted thatthesesequencesareremarkablyfreeofindels,which haverecentlybeenshowntoaccumulaterapidlyinthe pseudogenesof Sodalis -alliedsymbionts[17].Thus,based ontheavailabledata,wecannotdetermineifthehighly disrupted16SrRNAobtainedfromthe C.veigasimoni and C.paradoxus symbiontsarefunctional.Becauseofthepossibilitythatthe C.veigasimoni and C.paradoxus symbionts maintainadditional(functional)paralogouscopiesof16S rRNAthatwerenotamplifiedbytheuniversalprimers usedinthisstudy,weelectedtoexcludethe C.veigasimoni and C.paradoxussymbiont16SrRNAgenesequences fromsubsequentmolecularphylogeneticanalyses.Phylogeneticanalysisof16SrRNAgenesequencesof Columbicola spp.symbiontsInthe16SrRNAphylog eny,thesymbiontsof Columbicola spp.wereassignedtothreecladesintheGammaproteobacteria(Figure1).Atthisstage,becauseoftherelative paucityofrepresentationincladesBandC,itshouldbeSmith etal.BMCEvolutionaryBiology 2013, 13 :109 Page2of15 http://www.biomedcentral.com/1471-2148/13/109

PAGE 3

A B CVibrio cholerae (53.7%) Yersinia pestis (54.3%) Pantoea agglomerans (55.4%) Enterobacter hormaechi (55.1%) (54.8%) Salmonella enterica (54.8%) E ( Metaseiulus occidentalis -mite) (55.5%) PE ( C. wombeyi ex Geophaps plumifera ) (52.7%) #46 PE ( Physconelloides zenaidurae PE ( C. baculoides ex Zenaida macroura ) (49.2%) #4 PE ( C. macrourae 1ex Geotrygon montana ) (51.5%) #22 PE ( C. macrourae 1ex Leptotila plumbeiceps ) (51.4%) #23 PE ( C. macrourae 1ex Leptotila verreauxi ) (51.3%) #24 Escherichia coli (54.8%) Buchnera aphidicola PE( Acyrthosiphon pisum aphid) (50.1%) 99/ 1.0 53/ 0.57 51 PE( C. arnoldi ex Macropygia nigrirostris ) (53.3%) #2 Arsenophonus arthropodicus SE ( Pseudolynchia canariensis PE ( Pediculus capitis human head louse) (53.2%) PE ( Pediculus humanus human body louse) (48.8%)*Photorhabdus luminescens (55.2%) SE ( Paracoccus northofagicola mealy bug) (51.0%) PE ( C. macrourae 3ex Zenaida macroura ) (52.1%) #26 PE ( C. macrourae 4 ex Zenaida galapagoensis ) (51.7%) #27 PE ( Sitophilus rugicollis grain weevil) (52.1%) PE ( Pseudolynchia canariensis Wigglesworthia glossinidia PE ( Glossina brevipalpis PE ( Haematomyzus elephantis elephant louse) (45.0%) PE ( Haematopinus eurysternus cattle louse) (47.8%) Blochmannia pensylvanicus PE ( Camponotus pensylvanicus carpenter ant) (48.5%) PE ( carpenter ant) (47.7%) 62/ 0.88 66/0.84 98/1.0 Dickeya dadantii (55.2%) SE ( Bactericella cockerelli psyllid) (54.4%) Sodalis glossinidius SE ( Glossina morsitans PE ( C. extinctus ex Patagioenas fasciata ) (53.8%) #15 SE ( Cantao occelatus stinkbug) (55.3%) SE ( Curculio sikkimensis chestnut weevil) (55.4%) PE ( Sitophilus granarius grain weevil) (54.4%) Strain HS (55.1%) PE ( Sitophilus zeamais grain weevil) (55.1%) PE ( Sitophilus oryzae grain weevil) (54.9%) SE ( Craterina melbae PE ( C. tschulyschman ex Columba livia ) (54.2%) #41 PE ( C. columbae ex Columba livia ) (54.4%) #9 PE ( C. columbae ex Columba livia ) (54.4%) #10 PE ( C. exilicornis 1ex Macropygia amboinensis ) (54.8%) #13 PE ( C. exilicornis 1ex Macropygia amboinensis ) (54.8%) #12 PE ( C. harbisoni ex Phaps histrionica ) (52.3%) #20 PE ( C. waggermani ex Patagioenas leucocephala ) (53.3%) #43 PE ( C. adamsi ex Patagioenas plumbae ) (53.3%) #1 PE ( C. timmermanni ex Leptotila rufaxilla ) (53.6%) #40 PE ( C. timmermanni ex Leptotila rufaxilla ) (53.6%) #39 PE ( C. mjoebergi ex Geopelia placida ) (53.5%) #32 PE ( C. mjoebergi ex Geopelia striata ) (54.5%) #34 PE ( C. mjoebergi ex Geopelia striata ) (54.5%) #33 PE ( C. guimaraesi 2ex Chalcophaps indica ) (54.5%) #19 PE ( C. guimaraesi 1ex Chalcophaps indica ) (54.5%) #18 PE ( C. gracilicapitis ex Leptotila jamaicensis ) (53.7%) #17 PE ( C. passerinae 2ex Claravis pretiosa ) (54.7%) #36 PE ( C. passerinae 2ex Claravis pretiosa ) (54.7%) #35 PE ( C. clayae ex Treron wallia ) (54.8%) #6 73/0.96 52/ 0.58 *PE ( C. rodmani ex Geopelia humeralis ) (54.5%) #38 PE ( C. fortis ex Otidiphaps nobilis ) (54.3%) #16 PE ( C. claytoni ex ) (54.8%) #7 PE ( C. claytoni ex ) (54.8%) #8 PE ( C. malenkeae ex ) (54.1%) #28 PE ( C. koopae ex Geophaps scripta ) (55.0%) #21 PE ( C. elbeli ex Treron sieboldii ) (53.2%) #11 PE ( C. bacillus ex Streptopelia decaocto ) (53.4%) #3 PE ( C. claviformis ex Columba palumbus ) (54.5%) #5 PE ( C. waltheri ex Geotrygon frenata ) (54.1%) #45 PE ( C. exilicornis 2ex Phapitreron amethystinus ) (52.5%) #14 PE ( C. mckeani ex Ocyphaps lophotes ) (52.9%) #31 PE ( C. macrourae 2ex Zenaida asiatica ) (54.3%) #25 PE ( C. masoni ex ) (53.8%) #30 PE ( C. masoni ex Petrophassa albipennis ) (53.8%) #29 69/ 0.76 55/ 0.88 89/ 0.88 86/ 1.0 99/ 1.0 67/ 0.78 56/ 0.56 57/ 0.71* *0.09 substitutions/site Figure1 (Seelegendonnextpage.) Smith etal.BMCEvolutionaryBiology 2013, 13 :109 Page3of15 http://www.biomedcentral.com/1471-2148/13/109

PAGE 4

notedthatthesedesignationsaretentative,andnot supportedbyextremelyhighbootstrapvalues.CladeA containsthelargestnumberof Columbicola spp.symbionts andisrepresentedbythesymbiontof C.columbae ,the tsetseflysymbiont( Sodalisglossinidius ),andsymbiontsof grainweevilsofthegenus Sitophilus .CladeAalsocontains severallongestablishedprimaryendosymbiontsincluding Wigglesworthiaglossinidia and Blochmannia spp.,further supportingthenotionofamonophyleticoriginofthese bacteriaandthe Sodalis -alliedsymbionts[18].CladeBis representedbyseveral Columbicola spp.symbionts,includingthesymbiontof C.baculoides ,andsymbiontsidentified fromtheavianbodylouse Physconelloideszenaidurae and themite Metaseiulusoccidentalis [19].ThesolerepresentativeofcladeC,thesymbiontof C.arnoldi ,isdistantlyallied totheaphidprimaryendosymbiont Buchneraaphidicola .Phylogeneticanalysisofmultiplegenesequencesof Columbicola spp.symbiontsOnthebasisofcombinedsequencedataof16SrRNA, fusA and groEL genes,thesymbiontsof Columbicola spp. werealsodividedintothreedistinctcladesA,BandCin theGammaproteobacteria(Figure2).Herethenumberof analyzedtaxawassmallerbecauseallthreegeneswerenot alwayssuccessfullyamplifiedbyPCRfromthelousesamples.However,thephylogeneticrelationshipswereentirely concordantwiththeanalysisofthe16SrRNAsequence dataalone(Figure1).Notably,bothMLandBayesiansupportvalueswerehigherforthethreecladesof Columbicola spp.symbiontsinthecombinedtreeincomparisontothe treederivedfrom16SrDNAalone.Thisreflectsthefact thattheprotein-codingsequencesevolveinamorestable manner,allowingustogenerateanalignmentwithless ambiguitywithrespecttooutgroups.Star-likephylogenyinthecladeAof Columbicola spp. symbiontsInthe16SrRNAgenephylogeny(Figure1),thesymbiont sequencesfromdifferentindividualsofthesamespecies/ crypticspecies/haplogroups formedmonophyleticgroups withhighstatisticalsupport.Inmanycasesthesesequences wereidentical.However,deeper relationshipsbetweenthe symbiontsof Columbicola spp.werenotwellresolvedregardlessofthereconstructionmethodemployed.In particular,theinternodesconnectingtherepresentativesof cladeAwereextremelyshortwithlittleornostatistical support,althoughsubstantialsequencedivergencewasobservedamongtherepresentativesofcladeA,asevidenced bytherelativelylongbranchesleadingtoterminalnodes.In cladeA,consequently,thephylogenyexhibitedacomb-or star-likeappearance,exceptfor thefollowingstatisticallysupportedterminalclust ersofrecentorigin:(i) C.fortis #16 and C.rodmani# 38;(ii) C.adamsi #1and C.waggermani #43;and(iii) C.columbae #9,10and C.tschulyshman #41 (Figure1).Inthephylogenybasedonthecombined16S rRNA, fusA and groEL genesequences(Figure2),no clusterswithsignificantstatisticalsupportwereidentified incladeA,whichconsistentlycontainedlongterminal branchesandcorroboratedthestar-likephylogeneticrelationshipinthecladeAsymbiontsof Columbicola spp.Lackofhost-symbiontphylogeneticcongruenceUsingtheShimodairaandHasegawa(S-H)test[20],the phylogenetictreeof Columbicola spp.[21]wasfoundto besignificantlydifferenttothesymbiont16SrRNAtree (differenceinlnL=575.24, P <0.001)andthesymbiont treederivedfromthecombineddataset(differenceinln L=46.43, P <0.001). Co-phylogeneticanalysis[22]ofthe16SrRNAgene datasetreconstructed17potentialco-speciationevents betweenthehostandsymbiontlineages(Figure3).However,thisnumberofco-speciationeventswasnotsignificantlyhigherthanthatexpectedbychance( P >0.05). Usingthecombineddatasetof16SrRNA, fusA and groEL genes,co-phylogeneticanalysisreconstructedonly6potentialco-speciationevents(Figure3),againnomorethan thatexpectedbychance( P >0.05).MolecularevolutionaryrateestimationsRelativeratetestsrevealedthatthesubstitutionratesin 16SrRNAgenesofthecladeA( C.columbae ),cladeB ( C.baculoides ),andcladeC( C.arnoldi )symbiontswere significantlyhigherthanthoseoftheirfree-livingrelatives.CladeBandcladeCsymbiontsexhibited1.42fold and1.35foldhigherrates,respectively,whichweresupportedby P -values<0.001,whilethecladeAsymbiont showeda1.17foldhigherrate,whichwassupportedby a P -value<0.01(Table1). (Seefigureonpreviouspage.) Figure1 16SrRNAphylogenyof Columbicola spp.symbionts. Phylogenyof Columbicola spp.symbionts(bold)andrelatedbacteriabasedon maximumlikelihoodandBayesiananalysesofa1.46-kbpfragmentof16SrRNAgenesequences.Insectsymbiontsaredesignedbytheprefix “ PE ” (primaryendosymbiont), “ SE ” (secondaryendosymbiont)or “ E ” (ifunknown),followedbyinsecthostnameandcommonname(orlatinnameof birdhostfor Columbicola spp.)Thenumbersadjacenttonodesindicatemaximumlikelihoodbootstrapvalues(toleftofdiagonalline)and Bayesianposteriorprobabilities,whereapplicable(torightofline),fornodeswithbootstrapsupport>50%andBayesianposteriorprobabilities >0.5. Asterisksindicatenodeswith100%bootstrapsupportandBayesianposteriorprobability=1.Theboldarrowhighlightsthelocationofthesequence derivedfromstrainHS,therecentlycharacterizedprogenitorofthe Sodalis -alliedsymbionts.NumbersinparenthesesrepresenttheG+Ccontentof the16SrRNAsequences.FinalnumberscorrespondtosamplenumbersinAdditionalfile5. Smith etal.BMCEvolutionaryBiology 2013, 13 :109 Page4of15 http://www.biomedcentral.com/1471-2148/13/109

PAGE 5

Vibrio cholerae Yersinia pestis Buchnera aphidicola PE( Acyrthosiphon pisum aphid) PE( C. arnoldi ex Macropygia nigrirostris ) #2 PE ( C. macrourae 1ex Geotrygon montana ) #22 Salmonella enterica Escherichia coli Sodalis glossinidius SE ( Glossina morsitans Strain HS PE ( Sitophilus oryzae grain weevil) PE ( C. masoni ex ) #30 PE ( C. claytoni ex ) #8 PE ( C. claviformis ex Columba palumbus ) #5 PE ( C. clayae ex Treron wallia ) #6 PE ( C. exilicornis 1ex Macropygia amboinensis ) #12 PE ( C. waiteae ex Columba leucomela ) #44 PE ( C. timmermanni ex Leptotila rufaxilla ) #40 PE ( C. fortis ex Otidiphaps nobilis ) #16 PE ( C. macrourae 3ex Zenaida macroura ) #26 PE ( C. elbeli ex Treron sieboldii ) #11 Wigglesworthia glossinidia PE ( Glossina brevipalpis tsetse) PE ( carpenter ant) Blochmannia pensylvanicus PE ( Camponotus pensylvanicus carpenter ant) PE ( C. columbae ex Columba livia ) #9 PE ( C. mjoebergi ex Geopelia placida ) #32 PE ( C. mckeani ex Ocyphaps lophotes ) #31 0.09 substitutions/site 98/1.0 85/1.0 51 73/ 1.0 77/ 0.88 71/ 1.0 80/ 1.0 87/ 1.0 C B A Figure2 Multigenephylogenyof Columbicola spp.symbionts. Phylogenyof Columbicola spp.symbiontsderivedfrommaximumlikelihood andBayesiananalysesofacombineddatasetconsistingof16SrRNA, fusA and groEL genesequences.ConventionsasinFigure1. Smith etal.BMCEvolutionaryBiology 2013, 13 :109 Page5of15 http://www.biomedcentral.com/1471-2148/13/109

PAGE 6

Invivolocalizationof C.baculoides symbiontInmalesof C.baculoides ,fluorescent insitu hybridization detectedthesymbiontcellswithinbacteriocytesthat clusteredonbothsidesoftheabdominalcavity(Figure4A, BandC).Inyoungfemalesof C.baculoides ,thesymbiont cellsexhibitedthesamelocalizationasinmales(not shown).Inmaturefemales,bycontrast,thesymbiontcells werefoundinovarialtissues,localizedinapairofspecializedtransmissionorganscalle dovarialampullae(Figure4D andE),andverticallytransmittedfromtheovarialampullaetotheposteriorpoleofdevelopingoocytes(Figure4F). Thelocalization,migrationandtransmissionpatternsof thecladeBsymbiontof C.baculoides arealmostidentical tothoseofthecladeAsymbiontof C.columbae [13].SimulatingsymbiontreplacementsIncommonwiththecomplete tree(Figure1),the16S rRNAtreecomprisingonlystrainHSandthecladeA Columbicola spp.symbiontsalsohadacomb-liketopology withlowstatisticalsupport( Figure5A).WhenthediscretetimeMonteCarlosimulationwasoptimizedtoyieldprogenitoranddescendantdivergencelevelsmatchingthose obtainedfromtherealdata(Figure5A),theresultingtree (Figure5B)wasfoundtobeverysimilartothetree depictedinFigure5Abothintermsofitscomb-liketopologyandlowstatisticalsupport.Inaddition,bothtrees hadarelativelywidevarianceinterminalbranchlengths. Sinceoursimulationsonlyintroducepointmutationsand thereforeyieldunambiguouss equencealignments,thisindicatesthatthelackofresolutionintherealtreesisindeed afunctionoftheunderlyingevolutionaryprocess,rather thananartifactofthetree-buildingprocess. SincethetreepresentedinFigure5Bwasgenerated fromasimulationinwhichdescendantsevolvedatarate thatwaseightfoldhigherthantheirprogenitor,wewere curioustotesttheeffectofincreasingtheprogenitorevolutionaryratesuchthatitmatchedthatofthedescendants.Interestingly,whenthesimulationwasperformed witharateofprogenitorsequenceevolutionthatequaled thatofthedescendants,theresultingtreewasfoundto haveasignificantlyincreasedmeanbootstrapsupport (Figure5C).Whilethetopologyofthetreeretaineda comb-likeappearance,allnodeswereresolvedwithhigh bootstrapsupportinaccordancewiththetemporalorder Figure3 Comparisonofthephylogeniesofrepresentativespeciesof Columbicola spp.andtheirsymbioticbacteria. Columbicola trees arefrommaximumlikelihoodanalysisofsequencesofthemitochondrialcytochromeoxidaseIgene,mitochondrial12SrRNAgene,andnuclear elongationfactor1alphagene[21].SymbionttreesarefromFigure1(left)andFigure2(right)inthecurrentpaper.Connectinglinesillustrate host-symbiontassociations.Bulletednodesareco-speciationeventsinferredfromreconciliationanalysis[22]. Smith etal.BMCEvolutionaryBiology 2013, 13 :109 Page6of15 http://www.biomedcentral.com/1471-2148/13/109

PAGE 7

oftheirdescentfromtheprogenitor.Theoverallincrease inthelevelofbootstrapsupportforthetreeinFigure5C couldbeduetoeither(i)asimpleincreaseintheamount ofsignalinthephylogeneticanalysisor(ii)anincreasein theratioofsignaltonoisearisingfromanincreased progenitor:descendantmutationrate.Toresolvethis questionweperformedathirdsimulationruninwhich therateofdescendantsequenceevolutionwaselevateda furthereightfoldincomparisontotheratesusedinthe previousrun,yieldingthetreepresentedinFigure5D. Althoughtheratesofprogenitorevolutionareidenticalin therunsyieldingFigures5Cand5D,themeanbootstrap valuesforthetreepresentedinFigure5Darereducedtoa levelcomparabletothatobservedinFigure5B.Thisindicatesthat,underthescenarioofrepeatedsymbiontreplacementsmodeledinoursimulation,anincreaseinthe relativerateofevolutionofthedescendants,asobserved intherealdata,hasastrongnegativeimpactontheability toresolverelationshipsbetweentheprogenitoranddescendants.Finally,weperformedanadditionalsimulation run(withparametersidenticaltothoseusedtogenerate Figure5B)inwhichonedescendantspeciatedmidway throughthesimulationrun,tomimicahost-symbiont co-speciationevent.Notably,therelationshipbetweenthe resultingdescendants(labeledcospec1andcospec2in Figure5E)wasresolvedwith100%bootstrapsupport, indicatingthatitistheunderlyingevolutionaryprocess, ratherthanthedata,alignmentpropertiesorphylogenetic methodthatresultsinlowbootstrapsupportforour comb-liketrees.Thus,atafundamentallevel,ourresults showthatarelativelyunsupportedcomb-liketreetopologyisexpectedunderascenarioinwhichmultiple independentsymbiontreplacementstakeplaceinalimitedtimeframe,becausedescendantshaveveryfew synapomorphiesthathaveaccruedintheslow-evolving progenitorlineage.However,whenthesymbioticdescendantsco-speciatewiththeirinsecthost,synapomorphies areseededintheancestralsymbioticlineageatanelevated rate,yieldingasignalthatcanbeusedtoinfertheirphylogeneticrelationships.Evidenceforrecentsymbiontreplacementeventsin Columbicola spp.Thepairwiseestimatesofsynonymoussequencedivergenceforthe fusA and groEL sequencesallowedusto estimatethetimesincedivergenceofstrainHSandthe various Columbicola spp.cladeAsymbiontsatbetween 54,000and367,000years(Table2).Incomparisonwith ratesofdivergenceobtainedforsomeancientprimary insectsymbiontsfoundinotherinsecthosts,theacquisitionofthe Columbicola spp.cladeAsymbiontsseems tobearelativelyrecentevent,thatisfaryoungerthan theradiationofthemajorityof Columbicola spp.DiscussionWeundertookanextensivemolecularsurveyofbacterial symbiontsassociatedwithfeather-feedingliceofthegenus Columbicola ,whicharefoundonmostspeciesofpigeons anddovesacrosstheglobe.Themolecularphylogenetic analysesgroupedthesymbiontsof Columbicola spp.into threeputativeclades,designatedA,BandC,inthe Gammaproteobacteria,indicatingpolyphyleticevolutionaryoriginsofthesesymbionts.Whilestatisticalsupport forthemonophylyofthethreecladeswasonlymarginal Table1Relative-ratetestscomparingmolecularevolutionaryratesof16SrRNAgenesequencesbetweendifferent lineagesofthesymbiontsof Columbicola spp.andfree-livingrelativesLineage1Lineage2OutgroupK11K22K1-K2K1/K2 Pvalue3Symbiontof C.columbae (cladeA)[JQ063426] Dickeyadadantii [AY360397] Vibriocholerae [CP001486.1] 0.1240.1060.0181.1700.012 Symbiontof C.baculoides (cladeB)[JQ063412] Salmonellaenterica [AP011957] Vibriocholerae [CP001486.1] 0.1430.1010.0421.4162.1910-5Symbiontof C.arnoldi (cladeC)[JQ963434] Escherichiacoli [U00096] Vibriocholerae [CP001486.1] 0.1420.1050.0371.3529.1810-6Symbiontof C.columbae (cladeA)[JQ063426] Symbiontof C.baculoides (cladeB)[JQ063412] Vibriocholerae [CP001486.1] 0.1240.143 0.0190.8670.105 Symbiontof C.baculoides (cladeB)[JQ063412] Symbiontof C.arnoldi (cladeC)[JQ963434] Vibriocholerae [CP001486.1] 0.1430.1420.0011.0070.981 Symbiontof C.arnoldi (cladeC)[JQ963434] Symbiontof C.columbae (cladeA)[JQ063426] Vibriocholerae [CP001486.1] 0.1420.1240.0181.1450.065 Symbiontof C.columbae (cladeA)[JQ063426] Sodalisglossinidius (cladeA)[AF548135] Vibriocholerae [CP001486.1] 0.1240.1060.0180.1690.001 Sodalisglossinidius (cladeA)[AF548135] Dickeyadadantii [AY360397] Vibriocholerae [CP001486.1] 0.1060.106000.9981Estimatedmeandistancebetweenlineage1andthelastcommonancestoroflineages1and2.2Estimatedmeandistancebetweenlineage2andthelastcommonancestoroflineages1and2.3P -valuewasgeneratedusingtheprogramRRTree[ 23 ].Smith etal.BMCEvolutionaryBiology 2013, 13 :109 Page7of15 http://www.biomedcentral.com/1471-2148/13/109

PAGE 8

inthecaseofthe16SrRNAtree,strongersupportwas observedinthecaseofthecombinedtree,whichincorporateddatafromprotein-codinggenes.Eachrepresentative lousesamplescreenedinourstudywasassociatedwitha singlebacterialsymbiontderivedfromoneofthethree clades.Thesamepatternwasconsistentlyobservedin multipleindividualsofthesamelousespeciesandincrypticspecies,indicatingthatthereissomelevelofstability inthelouse-symbiontassociations.Basedontheseresults, weconcludethateachofthecolumbiformfeatherliceof thegenus Columbicola isassociatedwithasingleprimary bacterialsymbiont,whilethesymbiontsassociatedwithdifferentspeciesof Columbicola arephylogeneticallydiverse. Inmanyobligatehost-symbiontassociations,suchas thoseobservedbetweenaphidsand Buchneraaphidicola andtsetsefliesand Wigglesworthiaglossinidia ,aspecific hosttaxonisknowntohaveadoptedasinglesymbiont thatismaintainedovermacroevolutionarytimethrough repeatedboutsofco-speciation,yieldingcongruenthost andsymbiontphylogenies[3,24].Inthisstudywedemonstratedthatevolutionarypatternsofhost-symbiontassociationaredistinctinthe Columbicola featherlice:the primarysymbiontsareofpolyphyleticevolutionaryorigins anddonotexhibitanysignificantdegreeofhost-symbiont co-speciation.Similarpolyphyleticprimarysymbionts havebeenreportedfromweevilsofthesubfamily Dryophthorinae[25,26],andsuckingliceofthesuborder Anoplura[27].However,thecaseof Columbicola spp.is remarkableinthatthesepolyphyleticprimarysymbionts arefoundwithinasinglegenusofhostinsects. Relativeratetestsrevealedthatmolecularevolutionary ratesareelevatedinrepresentativesofallthreesymbiont cladesassociatedwith Columbicola spp.,butthelevels ofaccelerationaremorepronouncedincladesBandC relativetocladeA(Figure1,Table1).Thispattern suggestthatcladeBandcladeCmayrepresentmore ancientsymbiontlineagesthathaveexperiencedalonger historyofhost-symbiontco-evolution,withanassociatedacceleratedevolutionaryrate.Thisinturnimplies thatrepresentativesofcladeAhaveamorerecentorigin ofsymbiosis.Thedifferentsymbiontlineagesmayhave beenacquiredbythedifferent Columbicola lineages independentlyor,alternatively,representativesofclade AmayhavereplacedtheputativelymoreancientcladeB andClineages.Thereplacementscenarioseemsmore likely,giventhatall Columbicola speciesexaminedin thisstudywerefoundtoharboronlyasinglesymbiont, andthatrepresentativesofbothcladeAandBexhibit thesamelocalization,migrationandtransmissionpatternsaccordingtoFISH-basedmicroscopicanalyses (Figure4)[13]. Previousstudiessuggestedthat Columbicola feather licediversifiedmainlyinthePaleogene[28,29].Thus,if replacementshaveoccurred,theymusthaveoccurred Figure4 Fluorescent insitu hybridizationof C.baculoidessymbiontcells. ( A )Abdominalimageofanadultmale.Signalsofthesymbiont cellsaredetectedinbacteriocyteclusterslocatedonbothsidesoftheabdominalbodycavity.Whitesquaresindicatetheareasofpanels B and C .( B C )Enlargedimagesofthebacteriocyteclustersinpanel A .Signalsofthesymbiontcellsarelocalizedinthecytoplasmofthebacteriocytes. ( D )Abdominalimageofanadultfemale.Signalsofthesymbiontcellsarelocalizedinovarialampullaelocatedatthebaseoftheovaries.( E )An enlargedimageofanovarialampulla.( F )Asnapshotofsymbionttransmissionfromanovarialampullatoadevelopingoocyte.Redandgreen signalsindicatesymbiont16SrRNAandhostnuclearDNA,respectively.Abbreviations:oa:ovarialampulla,ov:ovariol,vg:vagina,pp:posterior poleofoocyte. Smith etal.BMCEvolutionaryBiology 2013, 13 :109 Page8of15 http://www.biomedcentral.com/1471-2148/13/109

PAGE 9

sincethistime.Similarsymbiontreplacementshavebeen reportedfromweevilsofthefamilyDryophthoridae, wheretheancientsymbiontlineage Nardonella wasreplacedbyseveralsymbiontlineagesthatarepredictedto beofmorerecentorigin[25,26].Indeed,replacementsinvolving Sitophilus spp.andtheircladeAsymbionts,SOPE andSZPE,arenowpredictedtohavetakenplaceveryrecently(<28,000years),basedonanalysisofgenome-wide substitutionratesbetweenSOPEandstrainHS[17].Also, inaphidsofthetribeCerataphidini,theancientsymbiont Buchneraaphidicola isthoughttohavebeenreplaced byfungalsymbiontlineages[30-32].Inthecaseofthe Columbicola spp.cladeAsymbiontsitisimportantto notethatourestimatesoftimesincedivergencefrom strainHSindicateanextremelyrecentoriginforthese symbioticassociations(<0.4MY;Table2).Sincethe radiationofthe Columbicola spp.complexisestimatedto haveencompassedapproximately57millionyears[28,29], wepredictthattherehasbeenlittleopportunityfor co-speciationbetween Columbicola spp.andtheirnewly acquiredcladeAsymbionts.Thus,itisnotsurprisingthat wefoundverylittleevidenceofcongruencebetweenthe phylogeniesof Columbicola spp.andtheircladeAsymbionts.Theonlyexceptionswerethecasesofveryrecent divergencebetween C.columbae and C.tschulyshman bothofwhichoccuronRockPigeons,andthesymbionts of Columbicolamacrourae 3(onMourningDove)and C. macrourae 4(onGalapagosDove).Thesetwodovesare predictedtohavedivergedfromoneotherlessthan2millionyearsago[29],andbasedongeneticdivergences,their licearepredictedtohavedivergedevenmorerecently (around0.2MYA,[21]). Figure5 BootstrappedMLtreesderivedfromareducedcladeAdatasetandsequencesderivedfromevolutionarysimulations. The treedepictingthereducedcladeAdataset(panel A )comprisesonlystrainHSand Columbicola spp.symbionts(withterminalnodeslabeled accordingtothenumberslistedinAdditionalFile5).Treesinpanels B – E arederivedfromtheoutputgeneratedbysimulationwithevolutionary rates(substitutions/site/cycle)of0.002(progenitor)and0.016(descendants)forpanel B and E ,0.016forboththeprogenitoranddescendants forpanel C and0.016(progenitor)and0.128(descendants)forpanel D .Terminalnodesrepresentingtheprogenitorsequencearelabeled “ prog ” anddescendantsarelabeledwiththeprefix “ d ” ,followedbythecyclenumberoftheirbirthinthe5000cyclesimulation.Thetreeinpanel E was obtainedfromadatasetinwhichonedescendantwaspermittedtospeciateatcycle2500,givingriseto “ cospec1 ” and “ cospec2 ” .Mean bootstrapvalues(MBV),depictedatthefootofeachtreewerecomputedfrombootstrapvaluesobtainedforallinternalnodesineachtree.Only individualbootstrapvalues>50%aredepictedinthefigure. Smith etal.BMCEvolutionaryBiology 2013, 13 :109 Page9of15 http://www.biomedcentral.com/1471-2148/13/109

PAGE 10

Molecularphylogeneticanalysesoffeatherlicerevealseveralwell-supportedcladeswithinthegenus Columbicola [21].Thus,itisstrikingthatalmostalloftheinternal branchesincladeAareshort,generatingacomb-liketree topologywithverylittleoverallbootstrapsupport(Figure1). WhiletheunresolvednatureofcladeAcouldbeascribed toartifactualissuesrelatedtothealignmentofdiversesequencesinthetreespresentedinFigures1and2,reconstructionofthe16SrRNAtreewithonlythe Columbicola spp.cladeAsymbionts(resultinginarelativelyunambiguousalignment)didnotimprovetheresolutionofthe tree(Figure5A). InweevilsoftheDryophthoridae,ithasbeensuggested thatsymbiontreplacementsmighthavebeendrivenby majorchangesintheinsectdiet[25].Incontrast,all Columbicola speciesareobligateparasitesofcolumbiform birdsthatliveonadietoffeathers,secretionsanddead skin.Experimentaltransfersof Columbicola spp.between differentspeciesofpigeonsanddovesshowthattheselice arecapableoffeeding,survivingandreproducingon heterospecifichosts[33].Hence,symbiontreplacements in Columbicola spp.areunlikelytobeattributableto dietarychanges.Previousstudieshavesuggestedthat biologicalvectorssuchasparasiticwaspsandmitesmight facilitatesymbionttransfersandreplacementsacross differenthostspecies[34,35].However,neitherparasitoid waspsnorectoparasiticmiteshavebeenreportedfrom Columbicola spp.[4].Anotherpossibilityisthathorizontal symbionttransfersaremediatedbetweendifferentlouse speciesbyinterspecificmating,asreportedforfacultative symbiontsinthepeaaphid[36].Whilethereisnoevidenceofinterspecificmatingin Columbicola spp.,itis notablethattheseliceundergohostswitchingthrough phoreticdispersalonhippoboscidlouseflies[8].This couldatleastfacilitatecontactbetweenmalesandfemales ofdifferentlousespecies.Itisalsonoteworthythatmale liceoftenremainincopulawiththeirfemalepartnersfor severalhours[4],whichcouldprovideawindowforhorizontal(maletofemale)symbionttransfers. However,phylogeneticlinesofevidencedonotfavorthe above-mentionedhypothesisofhorizontalsymbionttransfersbetween Columbicola spp.Ifdifferent Columbicola specieshadbeenundergoingoccasionalsymbionttransfers,theresultingsymbiontphylogenetictreewouldbe expectedtobeofcompactshape,withrelativelyshort terminalbranches,asobservedforfacultativeinsect symbionts,suchas Wolbachia Rickettsia Hamiltonella Regiella and Serratia [37-39].Contrarytothisexpectation, thephylogeniesofthesymbiontsof Columbicola spp.are characterizedbylongterminalbranchesandveryshort internodes,givingthetreestheircomb-liketopologies (Figure1,Figure2,Figure5A). Toaccountforthetreetopologiesobservedinour studyweproposeanalternativehypothesisthatinvolves repeatedsymbiontacquisitionsfromacommonbacterial “ progenitor ” thatisubiquitousintheenvironment.Such acandidatewasrecentlyisolatedfromahumanwound obtainedfollowingimpalementwithadeadtreebranch [17].Theresultingisolate,named “ strainHS ” ,wasfound tobeamemberofthe Sodalis -alliedcladeofinsectsymbionts,designatedascladeAinthecurrentstudy(Figure1, Figure2).Furthermore,itwasshownthatthegeneinventoriesoftwodistinct Sodalis -alliedsymbiontsrepresent reducedsubsetsofstrainHS,supportingthenotionthat anancestralrelativeofstrainHSservedasaprogenitor fortheseinsectassociates.Thisprogenitorhypothesisis compatiblewiththefollowingobservationsandevolutionarypatternsderivedfromourstudy:First,thesymbionts of Columbicola spp.,inparticularthoseofthecladeA,are closelyrelatedtosymbiontsofphylogeneticallydistant insecthoststhatencompassdiversegeographicaland ecologicalhabitats,suchastsetseflies,louseflies,grain weevils,chestnutweevils,longicornbeetlesandstinkbugs [40-47].Second,thesymbiontlineagesof Columbicola spp.,tsetsefliesandgrainweevils,whichshow~100% infectionfrequencies[13,40,41],suggestiveofrelative stabilityandcontinuityoftheassociations,aregenerally characterizedbylongterminalbranchesinthephylogeny (Figure1).Third,incontrast,strainHSandthesymbiont lineagesofchestnutweevilsandstinkbugs,whichshowlow infectionfrequencies[45-47],whichmaybeindicativeof instabilityand/oratemporaryn atureoftheseassociations, Table2Pairwiseestimatesofsynonymousdivergence( d S) andestimatesoftimesincedivergence(TSD)between strainHSandsymbiontsof Columbicola spp., Sodalis glossinidius andthe Sitophilusoryzae symbiont,SOPEgroELfusA Species d STSD(years) d STSD(years) C.clayae (#6)n/a1n/a10.428260,522 C.mjoebergi (#33)n/a1n/a10.421256,261 C.masoni (#30)n/a1n/a10.341207,565 C.timmermanni (#40)n/a1n/a10.310188,696 C.fortis (#16)0.345214,3880.273166,174 C.mckeani (#31)0.394245,4390.224136,348 C.exilicornis (#12)0.11772,8520.236143,652 C.waiteae 0.238148,3200.201122,347 C.columbae (#10)0.200124,7250.223135,739 C.claviformis (#5)n/a1n/a10.13682,782 C.claytoni (#7)n/a1n/a10.08954,173 C.macrourae (#26)0.590367,213n/a1n/a1C.elbeli (#11)0.542337,209n/a1n/a1S.glossinidius 0.245152,5280.16399,217 SOPE0.04528,0000.04628,0001Symbiontsequencenotavailableforpairwisecomparison.SeeAdditionalfile 5 .Smith etal.BMCEvolutionaryBiology 2013, 13 :109 Page10of15 http://www.biomedcentral.com/1471-2148/13/109

PAGE 11

arecharacterizedbyverysh ortterminalbranchesinthe phylogeny(Figure1).Thus,wepostulatethatthecladeA symbiontsof Columbicola spp.wereacquiredfromaprogenitorlineagethatpersistsintheenvironmentwithalow rateofmolecularevolution,andthatestablishmentofthe vertically-transmittedendosymbioticlifestyleresultsin acceleratedmolecularevolutionofthesymbiontgenes [48,49],givingrisetothelongterminalbranchesobserved inourphylogenies.Basedontheseassumptions,older Columbicola symbiontlineagesareanticipatedtohave experiencedacceleratedmolecularevolutionforlonger periodsoftimeandarethusexpectedtoexhibitlonger terminalbranches,whereasrecentsymbiontlineagesare expectedtohaveexperiencedreducedevolutionaryaccelerationandthusexhibitshorter terminalbranches(Figure5). Inordertodeterminehowaprocessofrecentsymbiontacquisitionmightinfluencethereconstructionof phylogenetictreesencompassingsuchevents,wedevelopedan insilico simulationofthisevolutionaryprocess, parameterizedusingmetaheuristicapproachesthatallow thesimulationtocloselyapproximatesequencedivergencesobservedbetweenstrainHSandthe Columbicola spp.cladeAsymbionts.Dataobtainedfromthesimulationyieldedatree(Figure5B)thatisstrikinglysimilarto thetreeobtainedfromtherealcladeAdataset(Figure5A). Thisindicatesthataprocessofsymbiontacquisition(or replacement)inwhichaslowly-evolvingprogenitorgives risetomorerapidlyevolvingsymbioticdescendantsis expectedtoyieldatreewithacomb-liketopologythathas littlestatisticalsupport.Furthersimulationrunswithelevatedevolutionaryratesdemonstratedthatlowlevelsof bootstrapsupportareobtainedwhentherateofevolution ofthesymbiontsequencesisincreasedrelativetothe progenitor,However,whenasymbiontwaspermittedto speciateinthesimulation,mimickinghost-symbiontcospeciationandfacilitatingsynapomorphiesintheresulting lineages,theresultingnodewasresolvedwithahighlevel ofbootstrapsupport.Thus,weconcludethatthecombliketopologyandlowstatisticalsupportobservedinclade Aisananticipatedoutcomeofanunderlyingevolutionary processinwhichrepeatedsymbiontreplacementoracquisitioneventsoccurinashorttimeperiod.Wealsoinfer thatthefewhighlysupportednodesinthe Columbicola spp.cladeAsymbiontsreflectveryrecenthost-symbiont co-speciationevents.ConclusionsInconclusion,wehavedemonstratedunexpecteddiversity andevolutionarydynamicso fthebacterialsymbiontsin featherliceofthegenus Columbicola .Toaccountforthe peculiarevolutionarypatternsobservedin Columbicolasymbiontassociations,weproposeahypothesisofrepeated, recentsymbiontacquisitio n/replacementeventsfroma commonenvironmental “ progenitor ” lineagesuchasthe recentlydiscovered “ strainHS ” [17].Usingasimulationbasedapproach,weshowthataprocessofsymbiontreplacementleadstoacharacteristiccomb-liketopologyin phylogenetictreesderivedfromthesymbionts.Thepolyphyleticbacterialsymbiontsof Columbicola spp.highlight thediversityandcomplexityofinsect-microbesymbiotic systems,andprovideinsightsintohowsuchsymbioticassociationshaveestablishedanddiversifiedinnature.MethodsSamplecollectionandDNAsequencingSamplesoflicewerecollectedfromwildbirdsusingthe postmortemrufflingprocedure[50].DNAwasextracted fromindividuallousesamplesbyfirstremovingthehead andplacingboththeheadandabdomeninextractionbufferATL(Qiagen).DNAwasthenisolatedusingaQiagen DNAeasyTissueExtractionKit.PriortoDNAextraction, bodypartswereremovedandmountedinbalsamon microscopeslidesasmorphologicalvouchers. DNAextractswereusedastemplateforPCRamplificationofasegmentofthebacterial16SrRNAgene (1.46-kb).Forasubsetofoursampleswealsoamplified a0.76-kbfragmentoftheelongationfactorEF-G( fusA ) geneanda1.49kbfragmentoftheheatshockchaperone ( groEL )gene.Weusedtheuniversalprimers27F(5 ’ AGAGTTTGATCCTGGCTCAG-3 ’ )and1492R(5 ’ -TAC GGTTACCTTGTTACGACTT-3 ’ )toamplify16SrRNA gene,thedegenerateprimersGroELF1(5 ’– ATGGGC WGCWAAAGAYGTRAAAT – 3 ’ )andGroELR1(5 ’– TCGGTRGTGATMATCAGRCCRGC-3 ’ )(designedfrom analignmentofinsectsymbiontsandotherfreeliving membersoftheGammaproteobacteria)toamplifythe groEL gene,andFusAF(5 ’ -CATCGGCATCATGGC NCAYATHGA-3 ’ )andFusAR(5 ’ -CAGCATCGGCTG CAYNCCYTTRTT-3 ’ )[51]toamplifythe fusA gene. ThePCRproductswerepurifiedusingtheQiagengel extractionkitandconcentratedinMicroconcolumns (Millipore).PurifiedproductswereclonedintoaTOPO 2.1vector(Invitrogen).Sangersequencingreactionswere performedon48clonesderivedfrom C.columbae and C.baculoides DNA,andaminimumoffourclonesderivedfromeachoftheotherDNAsamples.Sequences wereresolvedandcheckedinthesoftwarepackage Lasergene(DNAStar,Inc.Madison,Wisconsin).AllsymbiontsequencesweredepositedintheDDBJ/EMBL/ GenBanknucleotidesequencedatabasesundertheaccessionnumberslistedinAdditionalfile5.MolecularphylogeneticanalysesInordertoreconstructthephylogenyofthe Columbicola spp.symbionts,MUSCLE[52]wasusedtoalignsequencesofthe16SrRNAgenealone,andacombined datasetcomprising16SrRNA, fusA and groEL genes.The alignmentswereinspectedandadjustedmanuallyandareSmith etal.BMCEvolutionaryBiology 2013, 13 :109 Page11of15 http://www.biomedcentral.com/1471-2148/13/109

PAGE 12

availableuponrequestfromK.F.O.Sequencesfromother symbiontsandfree-livingbacteriawereselectedforinclusiononthebasisofsequencesimilarity,usingtheBLAST searchtool(NCBI).Weusedthisapproachtoensurethat, foreach Columbicola spp.symbiontsequence,thethree mostcloselyrelatedsymbiontsequencesandthemost closelyrelatedfree-livingbacterialsequencefromGenBank wererepresentedinourdataset.Theremainingfree-living taxawereselectedtoprovideappropriateresolutionwithin thefamilyEnterobacteriaceae(Additionalfile6). Vibrio cholerae wasselectedasanoutgroupbecauseitrepresents adistantlyrelatedmemberofthefamilyEnterobacteriaceae. WethenusedModeltest[53]andJModelTest[54]for BayesianInformationCriteriatoinferthemostappropriate modelofsequenceevolution(GTR+I+G)forsubsequent analyses.Analysisof16SrRNAgenesequenceswasperformedusingthemaximumlikelihood(ML)approach implementedinPhyML[55],with25randomstartingtrees and100bootstrapreplicates.BayesianposteriorprobabilitieswereestimatedusingMrBayes3.1.2[56].Runswere carriedoutfor1.5milliongenerationsusingthedefault parametersof4chains(3heatedandonecold)andstopped whenthestandarddeviationofsplitfrequenciesconverged tolessthan0.00001.Thefirst4000generationswere discardedasburn-inbasedonthestabilizationofloglikelihoodvaluesatthispoint.Consensustreeswerebuiltbased onthe50%majorityruleconsensus.Forthecombined analysis,wefirstusedapartitionhomogeneitytest[57-59] totestforconflictbetweenthethreesequencesinour combineddataset(16SrRNA, fusA and groEL genes).Since therewereseveraltaxafromwhichwecouldonlyobtain sequencedatafortwoofthethreeloci,absentsequences weretreatedasmissingdatai nthetree-buildingsoftware. Forthecombinedanalysis,datawerepartitionedandparameterswereestimatedseparatelyforeachgene.Analysisof16SrRNAsecondarystructureInordertodeterminetheimpactofsubstitutionsinthe Columbicola spp.symbiontsequencesonthestructure oftheir16SrRNAmolecules,wemappedsubstitutions fromthesesequencesontothesecondarystructureof the16SrRNAsequencederivedfromthemostclosely relatedoutgroup( Yersiniapestis )forwhichasecondary structurehasbeendeduced[60].Theresultinghomologymodelswerethenusedtocomputetherelativeratiosofsubstitutionsresultingin(i)changesthatpreserve thesecondarystructureofthemoleculeand(ii)changes thatinduceperturbationsinstructure(i.e.stem-loop transitions)[15].Wethenfurtheranalyzedthepositions ofsubstitutionsinthe Columbicola spp.symbiont16S rRNAsequencesinaccordancewithaposition-specific variabilitymapcomputedpreviouslyusing3,407bacterial16SrRNAsequences[16].Inouranalysis,substitutionswerescoredaccordingtoabinaryscheme[15],in whichsitesaredesignatedashavingsubstitutionrates thatareeitherhigherorlowerthantheaverageforall sitesanalyzedinthevariabilitymapstudy[16].FluorescentinsituhybridizationOligonucleotideprobesspecifictothe16SrRNAsequence from C.baculoides (5 ’ -GTTTTCTGTTACCGTTCGATT3 ’ and5 ’ -TTGCTTTTTCCTTCTTACTG-3 ’ )wereused forwhole-mountfluorescent insitu hybridizationasdescribedpreviously[13].Insectsobtainedfromacolonyof C.baculoides maintainedoncaptivemourningdoves ( Zenaidamacroura )werefixedinCarnoy ’ ssolutionfor twodays.Thelicewerethenwashedthreetimesinethanolandimmersedin6%(v/v)H2O2inethanolfor7d toquenchtheautofluorescenceofinsecttissues.After quenching,theinsectswerewashedthreetimesinethanol andthendecapitatedandpuncturedrepeatedlywithafine tungstenneedlethroughouttheabdomen.Theywerethen washedtwiceinethanol,threetimesinphosphatebufferedsalinecontaining0.3%TritonX-100(PBSTx), andequilibratedinhybridizationbuffer(20mMTris – HCl [pH8.0],0.9MNaCl,0.01%sodiumdodecylsulfate,30% formamide).TheprobesandSYTOXgreenwereaddedat finalconcentrationsof100nMand0.5 M,respectively, andthespecimenswereincubatedovernightatroom temperature.Thespecimenswerethenwashedseveral timesinPBSTx,mountedinSlowfade(Invitrogen)and observedunderbothanepifluorescencemicroscope (Axiophoto;Zeiss)andalaserconfocalmicroscope (PASCAL5;Zeiss).Toconfirmspecificdetectionofthe symbionts,aseriesofcontrolexperimentswereconducted. Theseconsistedofano-probecontrol,RNasedigestion control,andacompetitive-su ppressioncontrolwithexcess unlabeledprobeasdescribedpreviously[61].Co-phylogeneticanalysesWeusedtwoalternativeappro achestotestforcongruence betweenlouseandsymbiontphylogenies.First,weconductedShimodaira-Hasegawa(S-H)tests[20]onthehost tree[21]andtheMLtreesfromthe16SrRNAgenealone (Figure1)andcombinedsequences(Figure2)fromthe symbionts.Thesetreeswereprunedtoincludeonlyasingle representativesequencefromeachlouse/symbionttaxon. Thismethodwasusedtoassesswhetherthesymbiontdata canbeusedtorejectthelousephylogenetictree. Asasecondmethod,wereconstructedthenumberof co-speciationeventsbetweenthesymbiontandlouse treesusingreconciliationanalysis[62]asimplemented inTreeMap1[22].Thisanalysiswasagainperformed usingboththe16SrRNAgenetree(Figure1)andthe combinedtree(Figure2).AsintheS-Htestanalysis,both thelouseandsymbionttreeswereprunedtoincludeonly asinglerepresentativeofeachlouse/symbionttaxonto avoidartificiallybiasingtheresultsinfavorofcongruence.Smith etal.BMCEvolutionaryBiology 2013, 13 :109 Page12of15 http://www.biomedcentral.com/1471-2148/13/109

PAGE 13

Thesymbiontphylogenywasrandomized10,000timesto determineifthenumberofinferredco-speciationevents wasgreaterthanthatexpectedbychance[22,63].EvolutionarysimulationWeusedadiscrete-timeMonteCarlosimulationtomodel theevolutionaryscenarioofrepeatedsymbiontreplacementbyanenvironmentalprogenitor.Inthesimulation, the16SrRNAsequenceofstrainHSwasusedasacandidateprogenitorsequenceandpermittedtoaccumulate randommutationatauser-definedrateoverthecourseof 5000simulationcycles.Overthistime,theevolvingstrain HSsequencegaverisetoauser-definednumberofdescendantsequencesthatsubsequentlyevolveindependently atauser-definedrateuntilthe5000simulationcyclesare complete.Thetimingofdescendantbirthisgovernedbya 2-parameterBetadistribution,withvaluesof = =2, whichwerestochasticallyoptimizedalongwiththemutationratesoftheprogenitoranddescendantstoyieldan averagepairwisedistanceandvariancebetweentheprogenitoranddescendantsequencesthatcloselymatches themeanpairwisedistanceandvarianceestimatedfrom pairwisecomparisonsbetweenthe16SrRNAsequences ofstrainHSandthe Columbicola spp.symbiontsofclade A.Sequencesobtainedattheendofeachsimulationrun werealignedunambiguouslyandusedfortheconstruction ofphylogenetictrees,usingthesamemethodsasforthe constructionofthetreeinFigure1.Forcomparison,an additionaltreewasalsoconstructedforthe16SrRNA sequencesofstrainHSandthecladeAsymbiontsof Columbicola spp.alone.EstimatingdivergencetimesbetweenstrainHSandother membersofcladeABasedongenome-wideestimatesofsynonymoussubstitutionrates,thedivergencebet weenstrainHSandtheweevil symbiont,SOPE,waspredictedtohavetakenplacearound 28,000yearsago(17).ToestimatethedatesofsequencedivergencebetweenstrainHSandothercladeAsymbionts wecomputedpairwisesynonymoussitedivergencesforthe fusA and groEL sequenceslistedinAdditionalfiles5and6 usingtheKumarmethodimp lementedinMEGA[64].The resultingestimatesofsequen cedivergencewerethenused toobtaindatesofdivergencebasedonextrapolationfrom thestrainHS-SOPEcomparisons.AdditionalfilesAdditionalfile1: Phylogenyof Columbicola spp.symbiontsand relatedbacteriabasedona1.46-kbfragmentof16SrRNA. Insect symbiontsaredesignedbytheprefix “ PE ” (primaryendosymbiont), “ SE ” (secondaryendosymbiont)orE(ifunknown),followedbyhostnameand commonname(forthosenotderivedfrom Columbicola spp.)The numbersadjacenttonodesindicatemaximumlikelihoodbootstrap values(abovetheline)andBayesianposteriorprobabilities,where applicable(belowtheline),fornodeswithbootstrapsupport>50%and Bayesianposteriorprobabilities>0.5.Asterisksindicatenodeswith100% bootstrapsupportandBayesianposteriorprobability=1.Theboldarrow highlightsthelocationofthesequencederivedfromstrainHS,therecently characterizedprogenitorofthe Sodalis -alliedsymbionts.Numbersin parenthesesrepresenttheG+Ccontentofthe16SrRNAsequences.Final numberscorrespondtothelistprovidedinSupplementaryTable1. Additionalfile2: Homologymodeldepictingthe C.veigasimoni symbiont16SrRNAsequencemappedontothepredicted Y.pestis 16SrRNAstructure. Homologywasdeducedfromanalignment generatedinMuscle,andadjustedmanuallytoaccountforindels. Substitutionsinthesymbiont16SrRNAarehighlightedinbold. Substitutionswithahigher-than-averagerateofvariability(v>1)are highlightedwithredspots,whereasthosewithalower-than-averagerate ofvariability(v<1)arehighlightedwithbluespots.Thecountsof differentsubstitutiontypesaredisplayedinparenthesesinthekey. Additionalfile3: Homologymodeldepictingthe C.paradoxus symbiont16SrRNAsequencemappedontothepredicted Y.pestis 16SrRNAstructure. Homologywasdeducedfromanalignment generatedinMuscle,andadjustedmanuallytoaccountforindels. Substitutionsinthesymbiont16SrRNAarehighlightedinbold. Substitutionswithahigher-than-averagerateofvariability(v>1)are highlightedwithredspots,whereasthosewithalower-than-averagerate ofvariability(v<1)arehighlightedwithbluespots.Thecountsof differentsubstitutiontypesaredisplayedinparenthesesinthekey. Additionalfile4: Homologymodeldepictingthe C.columbae symbiont16SrRNAsequencemappedontothepredicted Y.pestis 16SrRNAstructure. Homologywasdeducedfromanalignment generatedinMuscle,andadjustedmanuallytoaccountforindels. Substitutionsinthesymbiont16SrRNAarehighlightedinbold. Substitutionswithahigher-than-averagerateofvariability(v>1)are highlightedwithredspots,whereasthosewithalower-than-averagerate ofvariability(v<1)arehighlightedwithbluespots.Thecountsof differentsubstitutiontypesaredisplayedinparenthesesinthekey. Additionalfile5: Lousespecimensusedinthecurrentstudy. Informationrelatingtothecollection,maintenanceandstorageoflouse specimensusedinthecurrentstudy,alongaccessionnumbersof sequencesdepositedintheGenbankdatabase. Additionalfile6: AdditionalDNAsequencesusedinthecurrent study. Genbankaccessionnumbersofadditionalsequencesusedin phylogeneticanalysesinthecurrentstudy. Abbreviation ML: Maximumlikelihood. Competinginterests Theauthorsdeclarethattheyhavenocompetinginterests. Authors ’ contributions DHCcollectedmanyoftheliceusedinthestudy.WAS,KFO,KPJ,DLR,TF, DHCandCDconceivedanddesignedthestudy.WAS,DLR,TCandKLS performedmoleculargeneticexperimentsleadingtosequencecollection. WAS,KPJandKFOperformedphylogeneticanalyses,andCDperformed evolutionarysimulations.RKandTFperformed invivo localizationofthe C. baculoides symbiont.Allauthorswereinvolvedindraftingthemanuscript andhavereadandapprovedthefinalversion. Acknowledgements TheauthorsthankM.Reed,S.Bush,J.AllenandA.Claytonforvariousformsof assistance.WealsothankT.Davis,J.Weckstein,S.Deko,N.Hillgarth,M.Robbins, P.LoiandJ.Kirchmanforcollectingliceusedinthisstudy.Allprocedures followedguidelinesoftheInstitutionalAnimalCareandUseCommitteeofthe UniversityofUtah.ThisworkwassupportedbyNationalScienceFoundation grantDEB0614565toDHCandCD,DBI0102112andDEB0555024toDLR, NationalScienceFoundationDoctoralDissertationImprovementGrantDEB 0608329toWASandDHCandNationalInstitutesofHealthgrant 1R01AI095736.TFwassupportedbytheProgramforPromotionofBasicand AppliedResearchesforInnovationsinBio-orientedIndustry(BRAIN).Smith etal.BMCEvolutionaryBiology 2013, 13 :109 Page13of15 http://www.biomedcentral.com/1471-2148/13/109

PAGE 14

Authordetails1DepartmentofBiology,UniversityofUtah,257South1400East,SaltLake City,UT84112,USA.2IllinoisNaturalHistorySurvey,UniversityofIllinois,1816 S.OakStreet,Champaign,IL61820,USA.3FloridaMuseumofNaturalHistory, UniversityofFlorida,Gainesville,FL32611,USA.4BioproductionResearch Institute,NationalInstituteofAdvancedIndustrialScienceandTechnology (AIST),Tsukuba305-8566,Japan. Received:5September2012Accepted:24May2013 Published:31May2013 References1.DouglasAE: Mycetocytesymbiosisininsects. BiolRevCambPhilosSoc 1989, 64: 409 – 434. 2.MoranNA,MunsonMA,BaumannP,IshikawaH: Amolecularclockin endosymbioticbacteriaiscalibratedusingtheinsecthosts. ProcRoySoc LondonSerB 1993, 253: 167 – 171. 3.BaumannP: Biologyofbacteriocyte-associatedendosymbiontsofplant sap-suckinginsects. AnnuRevMicrobiol 2005, 59: 155 – 189. 4.JohnsonKP,ClaytonDH: Thebiology,ecology,andevolutionofchewing lice. In Thechewinglice:worldchecklistandbiologicaloverview. 24thedition. EditedbyPriceRD,HellenthalRA,PalmaRL,JohnsonKP,ClaytonDH. Illinois:NaturalHistorySurveySpecialPublication;2003:449 – 476. 5.BushSE,PriceRD,ClaytonDH: Descriptionsofeightnewspeciesof featherliceinthegenus Columbicola (Phthiraptera:Philopteridae),with acomprehensiveworldchecklist. JParasitol 2009, 95: 286 – 294. 6.MalenkeJR,JohnsonKP,ClaytonDH: Hostspecializationdifferentiates crypticspeciesoffeather-feedinglice. Evolution 2009, 63: 1427 – 1438. 7.HarbisonCW,BushSE,MalenkeJR,ClaytonDH: Comparativetransmission dynamicsofcompetingparasitespecies. Ecology 2008, 89: 3186 – 3194. 8.HarbisonCW,ClaytonDH: Communityinteractionsgovernhostswitchingwith implicationsforhost-parasitecoevolutionaryhistory. PNAS 2011, 108: 9525 – 9529. 9.ClaytonDH,AdamsRJ,BushSE: Phthiraptera,theChewingLice .In Parasitic diseasesofwildbirds. EditedbyAtkinsonCT,ThomasNJ,HunterDB.Ames, Iowa:Wiley-Blackwell;2008:515 – 526. 10.GillespieJM,FrenkelMJ: Thediversityofkeratins. CompBiochemPhysiol 1974, 47B: 339 – 346. 11.WaterhouseDF: Digestionininsects. AnnRevEntomol 1957, 2: 1 – 18. 12.RiesE:DieSymbiosederLauseundFederlinge. ZMorpholOekolTiere 1931, 20: 233 – 367. 13.FukatsuT,KogaR,SmithWA,TanakaK,NikohN,Sasaki-FukatsuK, YoshizawaK,DaleC,ClaytonDH: Bacterialendosymbiontoftheslender pigeonlouse Columbicolacolumbae ,alliedtoendosymbiontsofgrain weevilsandtsetseflies. AppEnvironMicrobiol 2007, 73: 6660 – 6669. 14.MarshallRC,OrwinDFG,GillespieJM: Structureandbiochemistryof mammalianhardkeratin. ElectronMicroscRev 1991, 4: 47 – 83. 15.PeiAY,OberdorfWE,NossaCW,AgarwalA,ChokshiP,GerzEA,JinZ,LeeP, YangL,PolesM,BrownSM,SoteroS,DesantisT,BrodieE,NelsonK,PeiZ: Diversityof16SrRNAgeneswithinindividualprokaryoticgenomes. ApplEnvironMicrobiol 2010, 76: 3886 – 3897. 16.WuytsJ,VandePeerY,DeWachterR: Distributionofsubstitutionrates andlocationofinsertionsitesinthetertiarystructureofribosomalRNA. NuclAcidsRes 2001, 29: 5017 – 5028. 17.ClaytonAL,OakesonKF,GutinM,PontesA,DunnDM,vonNiederhausern AC,WeissRB,FisherM,DaleC: Anovelhuman-infectionderived bacteriumprovidesinsightsintotheevolutionaryoriginsofmutualistic insect-bacterialsymbioses. PLoSGenet 2012, 8 (11):e1002990.doi:10.1371/ journal.pgen.1002990. 18.HerbeckJT,DegnanPH,WernegreenJJ: Non-homogeneousmodelof sequenceevolutionindicatesindependentoriginsofprimary endosymbiontswithintheenterobacteriales(gamma-Proteobacteria). MolBiolEvol 2004, 22: 20 – 532. 19.HoyMA,JeyaprakashA: Microbialdiversityinthepredatorymite Metaseiulusoccidentalis (Acari:Phytoseiidae)anditsprey, Tetranychus urticae (Acari:Tetranychidae). BiolControl 2005, 32: 427 – 441. 20.ShimodiraH,HasegawaM: Multiplecomparisonsoflog-likelihoodswith applicationstophylogeneticinference. MolBiolEvol 1999, 16: 1114 – 1116. 21.JohnsonKP,ReedDL,HammondParkerSL,KimD,ClaytonDH: Phylogeneticanalysisofnuclearandmitochondrialgenessupports speciesgroupsfor Columbicola (Insecta:Phthiraptera). MolPhylEvol 2007, 45: 506 – 518. 22.PageRDM:TreeMapforMacintosh,ver.1.0b ;1995.http://taxonomy.zoology. gla.ac.uk/rod/treemap.html. 23.Robinson-RechaviM,HuchonD: RRTree:Relative-RateTestsbetweengroups ofsequencesonaphylogenetictree. Bioinformatics 2000, 16: 296 – 297. 24.MoranNA,McCutcheonJP,NakabachiA: Genomicsandevolutionof heritablebacterialsymbionts. AnnuRevGenet 2008, 42: 165 – 190. 25.LefevreCH,CharlesA,VallierB,DelobelB,FarrellB,HeddiA: Endosymbiont phylogenesisinthedryophthoridaeweevils:evidenceforbacterial replacement. MolBiolEvol 2004, 21: 965 – 973. 26.ConordCL,DespresA,VallierA,BalmandA,MiquelC,ZundelS,Lemperiere G,HeddiA: Long-termevolutionarystabilityofbacterialendosymbiosis incurculionoidea:additionalevidenceofsymbiontreplacementinthe dryophthoridaefamily. MolBiolEvol 2008, 25: 859 – 868. 27.HypsaV,KrizekJ: Molecularevidenceforpolyphyleticoriginofthe primarysymbiontsofsuckinglice(Phthiraptera,Anoplura). MicrobEcol 2007, 54: 242 – 251. 28.PereiraSL,JohnsonKP,ClaytonDH,BakerAJ: Mitochondrialandnuclear DNAsequencessupportaCretaceousoriginofColumbiformesand dispersal-drivenradiationinthePaleogene. SystBiol 2007, 56: 656 – 672. 29.JohnsonKP,WecksteinJD: TheCentralAmericanlandbridgeasan engineofdiversificationinNewWorlddoves. JBiogeography 2011, 38: 1069 – 1076. 30.FukatsuT,IshikawaH: Anoveleukaryoticextracellularsymbiontinan aphid, Astegopteryxstyraci (Homoptera,Aphididae,Hormaphidinae). JInsectPhysiol 1992, 38: 765 – 773. 31.FukatsuT,IshikawaH: Phylogeneticpositionofyeast-likesymbiontof Hamiltonaphisstyraci (Homoptera,Aphididae)basedon18SrDNA sequence. InsectBiochemMolBiol 1996, 26: 383 – 388. 32.FukatsuT,AokiS,KurosuU,IshikawaH: PhylogenyofCerataphidiniaphids revealedbytheirsymbioticmicroorganismsandbasicstructureoftheir galls:Implicationsforhost-symbiontcoevolutionandevolutionofsterile soldiercastes. ZoologSci 1994, 11: 613 – 623.33.BushSE,ClaytonDH: Theroleofbodysizeinhostspecificity:Reciprocal transferexperimentswithfeatherlice. Evolution 2006, 60: 2158 – 2167. 34.HuigensME,deAlmeidaRP,BoonsPA,LuckRF,StouthamerR: Natural interspecificandintraspecifichorizontaltransferofparthenogenesisinducing Wolbachia in Trichogramma wasps. ProcRSocLondon, SerB 2004, 271: 509 – 515. 35.JaenikeJ,PolakM,FiskinA,HelouM,MinhasM: Interspecifictransmission ofendosymbiotic Spiroplasma bymites. BiolLett 2007, 3: 23 – 25. 36.MoranNA,DunbarH: Sexualacquisitionofbeneficialsymbiontsin aphids. PNAS 2006, 103: 12803 – 12806. 37.WerrenJH,ZhangW,GuoLR: Evolutionandphylogenyof Wolbachia : reproductiveparasitesofarthropods. ProcRSocLondon,SerB 1995, 261: 55 – 63. 38.RussellJA,LatorreA,Sabater-MuozB,MoyaA,MoranNA: Side-stepping secondarysymbionts:widespreadhorizontaltransferacrossandbeyond theAphidoidea. MolEcol 2003, 12: 1061 – 1075. 39.WeinertLA,WerrenJH,AebiA,StoneGN,JigginsFM: Evolutionand diversityof Rickettsia bacteria. BMCBiol 2009, 7: 6. 40.DaleC,MaudlinI: Sodalis gen.nov.and Sodalisglossinidius sp.nov.,a microaerophilicsecondaryendosymbiontofthetsetsefly Glossina morsitansmorsitans IntJSystBacteriol 1999, 49: 267 – 275. 41.HeddiA,NardonP: Sitophilusoryzae L:amodelforintracellularsymbiosis intheDryophthoridaeweevils(Coleoptera). Symbiosis 2005, 39: 1 – 11. 42.NovakovaE,HypsaV: Anew Sodalislineagefrombloodsuckingfly Craterinamelbae (Diptera,Hippoboscoidea)originatedindependently oftsetsefliessymbiont Sodalisglossinidius FEMSMicrobiolLett 2007, 269: 131 – 135. 43.TojuH,HosokawaT,KogaR,NikohN,MengXY,KimuraN,FukatsuT: “ Candidatus Curculioniphilusbuchneri, ” anovelcladeofbacterial endocellularsymbiontsfromweevilsofthegenus Curculio ApplEnviron Microbiol 2010, 76: 275 – 282. 44.GrunwaldS,PilhoferM,HllW: Microbialassociationsingutsystemsof wood-andbark-inhabitinglonghornedbeetles[Coleptera: Cerambycidae]. SystApplMicrobiol 2010, 33: 25 – 34. 45.KaiwaN,HosokawaT,KikuchiY,NikohN,MengXY,KimuraN,ItoM, FukatsuT: Primarygutsymbiontandsecondary,Sodalis-alliedsymbiont oftheScutelleridstinkbug. ApplEnvironMicrobiol 2010, 76: 3486 – 3494. 46.KaiwaN,HosokawaT,KikuchiY,NikohN,MengXY,KimuraN,ItoM, FukatsuT: Bacterialsymbiontsofthegiantjewelstinkbug Eucoryssus grandis (Hemiptera:Scutelleridae). ZoolSci 2011, 28: 169 – 174.Smith etal.BMCEvolutionaryBiology 2013, 13 :109 Page14of15 http://www.biomedcentral.com/1471-2148/13/109

PAGE 15

47.TojuH,FukatsuT: Diversityandinfectionprevalenceofendosymbiontsin naturalpopulationsofthechestnutweevil:relevanceoflocalclimate andhostplants. MolEcol 2011, 20: 853 – 868. 48.MoranNA: AcceleratedevolutionandMuller ’ srachetinendosymbiotic bacteria. PNAS 1996, 93: 2873 – 2878. 49.WernegreenJJ: Genomeevolutioninbacterialendosymbiontsofinsects. NatRevGenet 2002, 3: 850 – 861. 50.ClaytonDH,DrownDM: Criticalevaluationoffivemethodsforquantifying chewinglice(Insecta:Phthiraptera). JParasitol 2001, 87: 1291 – 1300. 51.DaleC,PlagueGR,WangB,OchmanH,MoranNA: TypeIIIsecretion systemsandtheevolutionofmutualisticendosymbiosis. PNAS 2002, 99: 12397 – 12402. 52.EdgarRC: MUSCLE:amultiplesequencealignmentmethodwithreduced timeandspacecomplexity. BMCBioinforma 2004, 5: 113. 53.PosadaD,CrandallKA: MODELTEST:testingthemodelofDNA substitution. Bioinformatics 1998, 14: 817 – 818. 54.PosadaD: jModelTest:phylogeneticmodelaveraging. MolBiolEvol 2008, 25: 1253 – 1256. 55.GuindonS,GascuelO: Asimple,fast,andaccuratealgorithmtoestimate largephylogeniesbymaximumlikelihood. SystBiol 2003, 52: 696 – 704. 56.HuelsenbeckJP,RonquistF: MRBAYES:Bayesianinferenceofphylogenetic trees. Bioinformatics 2001, 17: 754 – 755. 57.FarrisJS,KllersjM,KlugeAG,BultC: Testingsignificanceof incongruence. Cladistics 1994, 10: 315 – 319. 58.FarrisJS,KllersjM,KlugeAG,BultC: Constructingasignificancetestfor incongruence.SystBiol 1995, 44: 570 – 572. 59.SwoffordDL: PAUP*.Phylogeneticanalysisusingparsimony(*andother methods) ,Version4. Sunderland,Massachusetts:SinauerAssociates;2003. 60.CannoneJJ,SubramanianS,SchnareMN,CollettJR,D ’ SouzaLM,DuY, FengB,LinN,MadabusiLV,MllerKM,PandeN,ShangZ,YuN,GutellRR: TheComparativeRNAweb(CRW)Site:anonlinedatabaseof comparativesequenceandstructureinformationforribosomal,intron, andotherRNAs. BMCBioinforma 2002, 3: 2. 61.SakuraiM,KogaR,TsuchidaT,MengXY,FukatsuT: Rickettsia symbiontin thepeaaphid Acyrthosiphonpisum :novelcellulartropism,effectonhost fitness,andinteractionwiththeessentialsymbiont Buchnera ApplEnvironMicrobiol 2005, 71: 4069 – 4075. 62.PageRDM: Componentanalysis:avaliantfailure? Cladistics 1990, 6: 119 – 136. 63.PageRDM: Temporalcongruenceandcladisticanalysisofbiogeography andcospeciation. SystZool 1990, 39: 205 – 226. 64.TamuraK,PetersonD,PetersonN,StecherG,NeiM,KumarS: MEGA5: MolecularEvolutionaryGeneticsAnalysisusingMaximumLikelihood, EvolutionaryDistance,andMaximumParsimonyMethods. MolBiolEvol 2011, 28: 2731 – 2739.doi:10.1186/1471-2148-13-109 Citethisarticleas: Smith etal. : Phylogeneticanalysisofsymbiontsin feather-feedingliceofthegenus Columbicola :evidenceforrepeated symbiontreplacements. BMCEvolutionaryBiology 2013 13 :109. Submit your next manuscript to BioMed Central and take full advantage of: € Convenient online submission € Thorough peer review € No space constraints or color “gure charges € Immediate publication on acceptance € Inclusion in PubMed, CAS, Scopus and Google Scholar € Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit Smith etal.BMCEvolutionaryBiology 2013, 13 :109 Page15of15 http://www.biomedcentral.com/1471-2148/13/109


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 EB9MKG8Y3_FVEYAG INGEST_TIME 2014-03-24T22:10:48Z PACKAGE AA00016104_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

Isolate 16S rDNA accession # FusA accession # GroEL accession# Buchnera aphidicola str ain APS BA000003 BA000003 BA000003 Candidatus Arsenophonus arthropodicus DQ115536 Candidatus Blochmannia floridanus BX248583.1 BX248583.1 BX248583.1 Candidat us Blochmannia pennsylvanicus strain BPEN CP00016 CP00016 CP00016 Endosymbiont of Haematopinus eurysternus DQ076661 Endosymbiont of Pediculus capitis DQ076659 Endosymbiont of Pediculus humanus DQ076660 Endosymbiont of Polyplax serrata DQ076667 Enterobacter endosymbiont of Metaseiulus occidentalis clone pAJ240 AY753173 Enterobacter hormaechei subsp. steigerwaltii strain EN562T AJ853890 Dickeya dadantii strain S3 1 AY360397 Escherichia coli K12 U00096 U00096 U00096 Pantoea agglomerans strain NCTC9381T AJ251466 Photorhabdus luminescens subsp. laumondii TTO1. BX571859 Primary endosymbiont of Sitophilus granarius AY126638 Primary endosymbiont of Sitophilus oryzae AF548142 JX524200 AF005236.1 Primary endosymbiont of Sitophilus rug icollis AY126639 Primary endosymbiont of Sitophilus zeamais AF548137 Primary symbiont of Pseudolynchia canariensis DQ115535 Salmonella enterica subsp. enterica serovar typhimurium AP011957 AP011957 AP011957.1 Secondary endosymbiont of Bactericera cockerelli AF263557 Secondary endosymbiont of Cantao ocellatus AB541010

PAGE 2

Secondary endosymbiont of Craterina melbae CMS06 EF174495 Secondary endosymbiont of Curculio sikkimensis isolate S86_4 AB541010 Secondary endosymbiont of Paracoccus notho fagicola AF476109 Shigella flexneri 2002017 CP001383 CP001383 CP001383 Sodalis glossinidius endosymbiont of tsetse Glossina morsitans AF548135 AP008232 AF326971.1 Strain HS JX444565 JX524201 JX444566 Vibrio cholerae MJ 1236 CP001486.1 CP001486.1 CP 001486.1 Wigglesworthia glossinidia endosymbiont of Glossina brevipalpis BA00021 BA00021 BA00021 Yersinia pestis strain KIM AE009952 AE009952 AE009952.1