<%BANNER%>

Analysis of Actin Filament Polymerization on Biomimetic Particles

University of Florida Institutional Repository
Permanent Link: http://ufdc.ufl.edu/UFE0021366/00001

Material Information

Title: Analysis of Actin Filament Polymerization on Biomimetic Particles
Physical Description: 1 online resource (149 p.)
Language: english
Creator: Sturm, Colin Dane
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: acta, actin, actoclampin, bead, biomolecular, conjugation, filament, fluorescence, liposome, matlab, microscopy, microsphere, nanosphere, polymerization, sem, silanize, single, tem, tirf, vesicle
Chemical Engineering -- Dissertations, Academic -- UF
Genre: Chemical Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Actin, a monomeric globular protein found in all eukaryotic cells, polymerizes into filaments generating force in several essential cellular processes, including cell adhesion, cell movement, and cell division. Invasive bacteria, such as Listeria monocytogenes, use actin for motility inside eukaryotic host cells. Listeria produces a surface protein called ActA which is able to form an actin-rich ?rocket tail? at one pole of the bacterium. While it is known that the polymerizing filaments provide the propulsive force on the bacterial surface, the mechanism by which filaments assemble and push the surface is unknown. It had been previously widely assumed that filaments were not attached to the surface and must be free in order to elongate. Our group has argued that elongating filaments are persistently attached to the bacterial surface through a processive filament end-tracking motor, termed actoclampin. To help determine whether filaments elongate attached or unattached, forces produced by actin filaments were analyzed in actin rocket tails through the use of transmission electron microscopy (TEM) and single filaments were observed to be directly tethered at their elongating ends to immobilized ActA-coated beads in vitro. Changing parameters in vitro such as particle size, ActA density, and the time of polymerization, provides control over the number of filaments found on each particle. Using total internal reflection fluorescence microscopy (TIRF), the polarity of the filaments could be determined to show barbed end elongation at the surface of the bead. However, the number of filaments elongating off a bead was difficult determine due to the resolution limits of fluorescent microscopy. For this reason, we also used TEM as a means of assessing the state of filaments surrounding the ActA-coated beads. These results show several beads with one to three filaments attached at the surface. The combined results of TIRF and TEM show strong evidence of barbed-end elongation of filaments attached at the surface of biomimetic particles, suggesting insertional polymerization mediated by an end-tracking motor.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Colin Dane Sturm.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Dickinson, Richard B.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021366:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021366/00001

Material Information

Title: Analysis of Actin Filament Polymerization on Biomimetic Particles
Physical Description: 1 online resource (149 p.)
Language: english
Creator: Sturm, Colin Dane
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: acta, actin, actoclampin, bead, biomolecular, conjugation, filament, fluorescence, liposome, matlab, microscopy, microsphere, nanosphere, polymerization, sem, silanize, single, tem, tirf, vesicle
Chemical Engineering -- Dissertations, Academic -- UF
Genre: Chemical Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Actin, a monomeric globular protein found in all eukaryotic cells, polymerizes into filaments generating force in several essential cellular processes, including cell adhesion, cell movement, and cell division. Invasive bacteria, such as Listeria monocytogenes, use actin for motility inside eukaryotic host cells. Listeria produces a surface protein called ActA which is able to form an actin-rich ?rocket tail? at one pole of the bacterium. While it is known that the polymerizing filaments provide the propulsive force on the bacterial surface, the mechanism by which filaments assemble and push the surface is unknown. It had been previously widely assumed that filaments were not attached to the surface and must be free in order to elongate. Our group has argued that elongating filaments are persistently attached to the bacterial surface through a processive filament end-tracking motor, termed actoclampin. To help determine whether filaments elongate attached or unattached, forces produced by actin filaments were analyzed in actin rocket tails through the use of transmission electron microscopy (TEM) and single filaments were observed to be directly tethered at their elongating ends to immobilized ActA-coated beads in vitro. Changing parameters in vitro such as particle size, ActA density, and the time of polymerization, provides control over the number of filaments found on each particle. Using total internal reflection fluorescence microscopy (TIRF), the polarity of the filaments could be determined to show barbed end elongation at the surface of the bead. However, the number of filaments elongating off a bead was difficult determine due to the resolution limits of fluorescent microscopy. For this reason, we also used TEM as a means of assessing the state of filaments surrounding the ActA-coated beads. These results show several beads with one to three filaments attached at the surface. The combined results of TIRF and TEM show strong evidence of barbed-end elongation of filaments attached at the surface of biomimetic particles, suggesting insertional polymerization mediated by an end-tracking motor.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Colin Dane Sturm.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Dickinson, Richard B.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021366:00001


This item has the following downloads:


Full Text





ANALYSIS OF ACTIN FILAMENT POLYMERIZATION ON BIOMIMETIC PARTICLES


By

COLIN STURM
















A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2007


































2007 Colin Sturm

































To my father and mother for their support and inspiration









ACKNOWLEDGMENTS

I acknowledge the support and guidance of my adviser, Dr. Richard Dickinson. He was

helpful throughout my tenure at the University of Florida, encouraging a self-guided experience

with the perfect amount of direction that has given me the qualities of a great researcher. I

would like to thank my committee members for their time and patience spent with me.

Specifically, I appreciate the biochemistry knowledge and practical experience Dr. Daniel Purich

has given me. Dr. Purich would visit the laboratory daily to check on the students and give

words of advice including the not so occasional joke to keep spirits high. Being the teaching

assistant for Dr. Jason Butler and taking his complex fluids course helped me to understand

chemical engineering principles and gave me a better approach at solving difficult problems.

Even at busy times, Dr. Butler always had time to spend discussing a problem. The time I spent

with Dr. Yiider Tseng, gave me insight into what being a graduate student is about and the trials

many students go through which I am thankful for.

A great deal of motivation and researching skill came from Dr. Joseph Phillips who

worked as a post-doctorate for Dr. Purich. Dr. Phillips and I spent much of our time preparing

experiments and critically thinking of the next step in the process. I greatly appreciate all of Dr.

Phillips' help and know that much of my success came from working closely with him. I thank

Dr. William Zeile who provided much of my basic understanding of laboratory experiments and

took time to discuss my experimental problems always having helpful suggestions. I thank my

fellow laboratory mates Kimberly Interliggi, Luzelena Caro, and Gaurav Misra who went

through the good and bad with me and were supportive.

There are several people that have helped me throughout my graduate years. A short list

includes Matt Monroe and Nikul Sheth who were great roommates and friends, Darren McDuff

and Jonathan Bricker who provided plenty of laughter and discussion, and Michael Matlock and









Chris Bussum who always encouraged me and were always there for me both being truly great

friends. Finally, I would like to thank the staff at the University of Florida Interdisciplinary

Center for Biotechnology Research (ICBR) Core Laboratories for all their help with my electron

microscopy experiments.









TABLE OF CONTENTS

page

A CK N O W LED G M EN T S ................................................................. ........... ............. .....

LIST OF TABLES .............. .......................................................9

LIST OF FIGURES .................................. .. .... ..... ................. 10

ABSTRAC T ................................................... ............... 15

CHAPTER

1 A C TIN B A SED M O TILITY .............................................. ......................................... 17

A ctin is E essential for C ell M utility ........................................................................... .... 17
A ctin Sequestering Proteins .................................................................................... 18
Actin Based Propulsion of Listeria monocytogenes................................. ..................18
Experimental Evidence of Actin Dynamics .................... ........... .................19
Models for Force Generation by Actin Polymerization ...............................................21
Distinguishing Between Actin Polymerization Models ............................... ................ 26

2 ENERGY DENSITY IN BENT ACTIN FILAMENTS OF AN ACTIN ROCKET TAIL ...31

In tro d u ctio n ................... ................... ...................1..........
M e th o d s ........................................................................... 3 1
R e su lts ................... ...................3...................7..........
D iscu ssio n ................... ...................3...................7..........

3 ACTIN PROPELLED VESICLES ................................. .......................... .........46

In tro du ctio n ................... ...................4...................6..........
M materials an d M eth o d s ..................................................................................................... 4 8
B ov in e B rain E x tract ................................................................ ...............................4 8
B radford A ssay ...........................................................................................................49
Actin Purification from Rabbit Muscle ......................... 49
A cetone pow der.......................................... 49
Purification of actin from acetone powder ........................................................50
Fluorescent labeling of actin ......................................... .......... 51
Preparing Listeria monocytogenes on Agar Plates .............................................................52
Purification of ActA-His6 from Listeria monocytogenes ........... ...........53
Fluorescent Labeling of A ctA -H is6-Cys ........................................................................54
V esicle Preparation ............................................. 55
M o utility A ssa y ........................................................................................................... 5 5
C re atin e k in a se ................................................................................................... 5 6
Protease inhibitors and DTT ...............................................56



6









R e su lts ................................................................................................................................ 5 6
D iscu ssio n ................... ...................5...................9..........

4 ELECTRON MICROSCOPY OF ACTIN FILAMENTS....................................68

In tro d u ctio n ................... ...................6...................8..........
M materials and M methods ................. ....... ............................. ...........70
Functionalized 500 nm B eads .................................................................................. 70
Preparation of Listeria monocytogenes Overexpressing ActA ......................................70
Functionalized 50 nm B eads ................................................... ............................71
ActA and BSA Conjugated to 50 nm Beads ....................................... ............... 71
Flow Chamber for Exchange Experiments ..................................................................72
Beads / Listeria w ith A ctin Rocket Tails ........................................ ..... ............... 72
Preparation of Sample for Viewing with EM ........................ ............................73
Critical point dryer (CPD ) ............................................. ..... ....................... 73
Sputter coating for SEM ................................................. .............................. 73
R otary shadow for TEM ................................................. .............................. 74
Post shadowing / pre TEM treatment........................................ 74
Polyvinyl Formal Coated TEM Copper Grids...................................... ...............74
Filam ent Shearing from Fluid Flow ........................................ .......................... 75
R esu lts ................... ................................................................. ................7 6
D isc u ssio n ................... .............................................................. ................7 8

5 SINGLE ACTIN FILAMENT POLARIZATION DETERMINED BY MULTIPLE
LABELED ACTIN MONOMERS INCORPORATED INTO ACTIN FILAMENTS..........89

Introduction ................... .......................................................... ................. 89
M materials and M methods ...................................... .. ......... ......... .....91
Color Change A ssay .................. ................................. ........ .. ............ 91
Im ag e A n aly sis ................................................................ 9 1
R e su lts ................... ...................9...................2..........
D iscu ssio n ................... ...................9...................4..........

6 CONCLUSIONS AND FUTURE WORK .......................... 101

D isc u ssio n .............................................................................................................................1 0 1
Suggestions for Future W ork ................................................................. 103

APPENDIX

A MATLAB ALGORITHM TO DETERMINE ENERGY STORED IN BENT
F IL A M E N T S .........................................................................................................10 6

B M YOSIN SUBFRAGM ENT-1 ......................................... ................................. 118

In tro d u ctio n ................... ...................1.............................8
M yosin Purification .............................................................................................. ........118









Purification of S using Papain ............. .............................. ...................................119
Purification of S1 using a-Chymotrypsin............................. ................................. 120

C IM A G E J M A C R O S ......... .. ............. .......................................................................... 12 1

LIST OF REFERENCES ............... ................................................. 140

B IO G R A PH IC A L SK E T C H ........................................... ......... ................... .......................... 149















































8









LIST OF TABLES


Table page

3-1 Necessary ratios for a parallel dilution to perform a Bradford assay. ...................................61

3-2 List of buffer volumes based on 1 kg of rabbit muscle. ......................................................61

3-3 Reference table for amount (g) of ammonium sulfate ((NH4)2SO4) to add at 4C ..............61

3-4 Differences in experimental procedure for vesicle motility. ...............................................62

3-5 A ctA conjugation w ith vehicles ...................... .... ..................... ................... ............... 62









LIST OF FIGURES


Figure page

1-1 Actin filaments polymerizing at the leading edge of a cell. ................................................28

1-2 Simplified cartoon of an actin monomer with a nucleotide bound inside the actin cleft.......28

1-3 The treadmilling process of an actin filament. ............................................ ............... 29

1-4 The treadmilling of an actin filament is enhanced with profilin and cofilin........................29

1-5 The elastic B row nian ratchet m odel. ........................................................... .....................30

1-6 Generalized end-tracking motor persistently bound to an actin filament ............................30

2-1 Notation used to describe a flexible rod / actin filament.................................................. 41

2-2 A 500 nm polystyrene bead functionalized with ActA and exposed to a motility assay.......41

2-3 Cartoon of an actin filament projecting an image on a two-dimensional plane .................42

2-4 Close up of an actin rocket tail that has points plotted along the filaments .........................42

2-5 Close up from Figure 2-4 showing points plotted along a single filament ...........................43

2-6 Analysis of the bending energy of filament in Figure 2-5...................... ............... 43

2-7 Output from algorithm with filaments numbered and points plotted with a blue line. ..........44

2-8 Histogram of all dE/ds values in one actin rocket tail ................. ..................................44

2-9 Histogram of all dE/ds values along 670 filaments in 12 actin rocket tails .........................45

3-1 U nilam ellar bilipid layer vesicle......................................................................... 63

3-2 V esicle conform action change. ...................................................................... ....................63

3-3 Vesicles emanating from a central vesicle mass. ...................................... ............... 64

3-4 Vesicles exposed to rhodamine actin and then Oregon-green actin .....................................64

3-5 Color change experim ent with a vesicle ........................................................ .............. 65

3-6 Color change experim ent with a vesicle ........................................................ .............. 65

3-7 H istogram of vesicle velocities...................................................................... ...................65

3-8 V esicle velocities versus vesicle radius ....................................................... ...................66









3-9 Theoretical particle velocity at different actin concentrations. ............................................66

3-10 Overlay of rhodamine labeled actin propelling an Oregon-green labeled vesicle. .............67

3-11 Large unlabeled vesicles conjugated with green fluorescent ActA ...............................67

4-1 Listeria propelled by an actin rocket tail jutting from a larger actin network......................81

4-2 Listeria with a small actin comet tail in its early stages................... ..............81

4-3 A 500 nm polystyrene bead with an actin cloud viewed using TEM ....................................82

4-4 A 500 nm polystyrene bead with an actin rocket tail viewed using TEM ..............................82

4-5 Tw o actin rocket tails converge ............................................... .......................... ...............83

4-6 Three 500 nm polystyrene beads combine to form one actin rocket tail.............................83

4-7 Single actin filaments emanating from 50 nm silica beads. .................................................84

4-8 A single actin filament associated with a single 50 nm silica bead................... ............84

4-9 Several single actin filaments associated with single 50 nm silica beads. ..........................85

4-10 Histogram of filament lengths with bin size of 100 nm. ............. ...................................... 85

4-11 Filam ent number per bead versus tim e .............. ..................................... ............... 86

4-12 Number of filaments normalized to the total number of beads versus time......................86

4-14 Time lapse of single actin filaments emanating from 50 nm beads. ...................................88

5-1 Fluorescent actin change hypothetical scenarios......................................... ............... 96

5-2 High pass filter removing frequencies larger than 50 pixels ............................................96

5-3 Low pass filter removing particles smaller than 3 pixels. ................... ................... .......... 97

5-4 Low and high pass filters on a mock filament.......... ................... ...... ..97

5-5 Compilation of 158 color change events observed ................... ................................ 98

5-6 Both fluorescent channels and overlay of a two color filament. .........................................98

5-7 Both fluorescent channels and overlay of a two color filament. .........................................99

5-8 Both fluorescent channels and overlay of a two color filament. .........................................99









5-9 Both fluorescent channels and overlay of a two color filament. .........................................99

5-10 Histogram for beads and filaments of 20 image sets.......... .....................................100












ADF

ADP

AFM

AMP-PNP

APES

Arps

ATP

BBE

BHI

BS3

BSA

CPD

DMF

DMSO

DTT

EDTA

EGTA

EM

FRAP

GME

GUV

LUV

N-WASP

NIH


LIST OF ABBREVIATIONS

Actin depolymerizing factor

Adenosine diphosphate

Atomic force microscope

Adenyl-5'-yl imidodiphosphate

3-aminopropyltriethoxysilane

Actin related proteins

Adenosine triphosphate

Bovine brain extract

Brain-heart infusion media

Bis(sulfosuccinimidyl suberate)

Bovine serum albumin

Critical point dryer

Dimethylformamide

Dimethyl sulfoxide

Dithiothreitol

Ethylenediamine tetraacetic acid

Ethylene glycol tetraacetic acid

Electron microscopy

Fluorescence recovery after photobleaching

Glycine methyl ester

Giant unilamellar vesicle

Large unilamellar vesicle

Neural Wiskott-Aldrich syndrome protein

National Institutes of Health









Nitrilotriacetic acid


PBS Phosphate buffered saline

PI Protease inhibitors

PMSF Phenylmethanesulphonyl fluoride

S1 Subfragment-1

SEM Scanning electron microscope

SUV Small unilamellar vesicle

TEM Transmission electron microscope

TIRF Total internal reflection fluorescence

TLCK NU-p-tosyl-L-lysine chloromethyl ketone

VASP Vasodilator-stimulated phosphoprotein

VCA Verprolin/cofilin homology/acidic


NTA









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

ANALYSIS OF ACTIN FILAMENT POLYMERIZATION ON BIOMIMETIC PARTICLES

By

Colin Sturm

December 2007

Chair: Richard B. Dickinson
Major: Chemical Engineering

Actin, a monomeric globular protein found in all eukaryotic cells, polymerizes into

filaments generating force in several essential cellular processes, including cell adhesion, cell

movement, and cell division. Invasive bacteria, such as Listeria monocytogenes, use actin for

motility inside eukaryotic host cells. Listeria produces a surface protein called ActA which is

able to form an actin-rich "rocket tail" at one pole of the bacterium. While it is known that the

polymerizing filaments provide the propulsive force on the bacterial surface, the mechanism by

which filaments assemble and push the surface is unknown. It had been previously widely

assumed that filaments were not attached to the surface and must be free in order to elongate.

Our group has argued that elongating filaments are persistently attached to the bacterial surface

through a processive filament end-tracking motor, termed actoclampin. To help determine

whether filaments elongate attached or unattached, forces produced by actin filaments were

analyzed in actin rocket tails through the use of transmission electron microscopy (TEM) and

single filaments were observed to be directly tethered at their elongating ends to immobilized

ActA-coated beads in vitro. Changing parameters in vitro such as particle size, ActA density,

and the time of polymerization, provides control over the number of filaments found on each

particle. Using total internal reflection fluorescence microscopy (TIRF), the polarity of the









filaments could be determined to show barbed-end elongation at the surface of the bead.

However, the number of filaments elongating off a bead was difficult determine due to the

resolution limits of fluorescent microscopy. For this reason, we also used TEM as a means of

assessing the state of filaments surrounding the ActA-coated beads. These results show several

beads with one to three filaments attached at the surface. The combined results of TIRF and

TEM show strong evidence of barbed-end elongation of filaments attached at the surface of

biomimetic particles, suggesting insertional polymerization mediated by an end-tracking motor.









CHAPTER 1
ACTIN BASED MOTILITY

Actin is Essential for Cell Motility

Cell motility is essential for physiological processes such as development, wound healing,

and defense against infection (1). Cells crawl by extending protrusions at their leading edge that

adhere to the substratum and allow the cell to pull itself forward (Figure 1-1). These protrusions

involve the polymerization of actin, which is a highly conserved (changing little throughout

evolution) globular protein (42 kDa) and the most abundant intracellular protein in most

eukaryotic cells (2). In its unpolymerized form, actin is referred to as G-actin and has two

globular regions with a hinge connecting the two domains resulting in a deep cleft (3). In this

cleft is a nucleotide binding region that can bind an Mg2+ ion completed with adenosine

diphosphate (ADP) or adenosine triphosphate (ATP).

Actin monomers polymerize into 7-nm diameter semi-flexible filaments (4) filamentouss

actin or F-actin) consisting of two proto-filaments that wrap around each other in a right-handed

helix with a 37-nm pitch and a persistence length of about 15 [tm (5-7). Actin filaments are

polar, with the filament (+)-end (also known as the barbed end) polymerizing faster than the

(-)-end (a.k.a. the pointed end) (2). This polarity results in the subunit only at the filament

(-)-end having an exposed cleft (Figure 1-2). Following binding of actin-ATP to the (+)-end, the

nucleotide undergoes hydrolysis (ATP to ADP) followed by phosphate release, and actin-ADP

depolymerizes from the (-)-end. Monomer-bound ADP is then exchanged for ATP in the

cytoplasm allowing the monomer to be recycled for (+)-end assembly. The actin-ATP critical

concentration for (+)-end assembly is 0.1 tM and that for (-)-end assembly is 0.6 pM (8). At

steady-state polymerization, the concentration of actin resides between the (+)-end and (-)-end









critical concentrations, with the (+)-end growing and the (-)-end shrinking, in a process known as

"treadmilling" (Figure 1-3).

Actin Sequestering Proteins

Actin concentration in eukaryotic cells is typically between 100 and 200 tM (9). This high

concentration of G-actin relative to the critical concentration is maintained by actin sequestering

proteins such as thymosin-p4 and profilin, which bind to G-actin thereby effectively decreasing

the concentration of monomeric actin relative to filamentous actin. Thymosin-p4 and profilin are

the main proteins responsible for actin sequestering. Thymosin-p4 binds G-actin in solution

(Kd is 0.7 [M) (9) to prevent its polymerization. Profilin similarly binds G-actin but actually

promotes polymerization by catalyzing nucleotide exchange (ADP to ATP) on actin monomers

and shuttling monomers to the filament (+)-end (10). Actin depolymerizing factor/cofilin

(ADF/cofilin) is not as significant in sequestering actin monomers as thymosin-p4 or profilin but

does promote depolymerization of actin filaments (11). Cofilin binds to F-actin-ADP and causes

the filament to twist tighter increasing the helical repeat of an actin filament from 37 nm to 27

nm (9) essentially breaking sections of the filament off resulting in depolymerization. Figure 1-4

shows the pathways and roles of profilin and cofilin interacting with a treadmilling actin

filament.

Actin Based Propulsion of Listeria monocytogenes

Listeria monocytogenes is a bacterial pathogen that infects cattle and causes severe food

poisoning in humans (12). After Listeria is phagocytosed by a host cell, the bacterium secrets

enzymes that break down the phagosome thereby releasing the bacterium into the host cells

cytoplasm. Once free in the cell, Listeria polymerizes actin filaments at one pole of the

bacterium surface to propel itself within the cytoplasm and to translocate between cells. Listeria

requires only a single bacterial protein, ActA for propulsion, and it commandeers other necessary









components from the host cell cytoplasm (13-17). ActA activates the Arp2/3 complex, which

nucleates new filaments at the bacterial surface (16, 18-20). ActA also binds to

vasodilator-stimulated phosphoprotein (VASP) (21) to promote actin (+)-end assembly. VASP

contains several oligo-proline sequences that bind profilin thereby providing a bacterial surface

bound pool of profilin actin at filament (+)-ends (12). Our group has proposed that ActA-VASP

stimulates filament growth in the actoclampin end-tracking motor model for actin based force

generation (discussed below) (22). There are also other proteins known to operate similarly to

VASP such as neural Wiskott-Aldrich syndrome protein (N-WASP) for ./nge//lt or vaccinia

motility (23-26) or the verprolin/cofilin homology/acidic (VCA) domain (C-terminal of

N-WASP) (27). Although there are other possible end-tracking motors such as N-WASP or

VCA that behave similar to ActA producing actin filaments with (+)-ends at a motile surface,

Listeria and ActA have been instrumental in determining essential factors for actin

polymerization and allow for analysis of motility in a relatively simple system (28).

Experimental Evidence of Actin Dynamics

The first observation of actin was by W.D. Halliburton in 1887 (29) who extracted a

protein from muscle he named myosin-ferment which coagulated preparations of myosin.

However, he was unable to further characterize the protein so the discovery went unnoticed for

almost 80 years (30). Brun6 Straub is credited with the discovery of actin in 1942. He

developed the first technique to isolate substantial amounts of pure actin (31) that was so

effective the technique has remained relatively unchanged to this day. More than 60 years have

passed and the mechanism of force generation by actin polymerization remains controversial.

Several methods have been employed to characterize actions role in force production. The most

insight on the mechanism of actin force generation has risen from the study of biomimetic









particle motility including the propulsion of particles such as polystyrene beads, oil droplets, and

vesicles (32-42).

Analysis of biomimetic particle systems is important because conditions can be controlled

in vitro, experiments are easily reproduced, and the possible propulsion mechanism can be

studied. Cameron et al. was the first to coat polystyrene beads with ActA and observe motility

of an artificial load by actin polymerization (32) concluding ActA is the only bacterial protein

necessary to induce actin polymerization with no dependence on the bacterium itself. Cameron

et al. took the same actin propelled ActA coated polystyrene beads and observed the system

using electron microscopy (EM) (33) revealing filaments persistently bound to the bead surface.

Filaments were found bound to 50 nm, 200 nm, and 500 nm beads although only one 50 nm bead

with a filament was observed. Schwartz et al. found that flattened particles coated with ActA are

just as motile as spherical particles ruling out any actin dependencies on geometric shape (34).

In an effort to retard motility velocities, methyl cellulose (an inert viscous solution used in

many food and cosmetic products) was used to slow motile Listeria in a cell extract (35) and

N-WASP coated beads in a reconstituted motility medium (36). Both studies found that

increasing the viscous drag force could not stop motility even at high concentrations of methyl

cellulose, and only the first study observed an effect of methyl cellulose on motility at all, which

may be explained by enhanced diffusion-limitations (43). Studies on VCA-coated particles

revealed velocities to be inversely proportional to particle radius (37) and large beads (> 1.5 [tm)

exhibited saltatory motion with recurring phases of the actin rocket tail cycling from a dense

actin network to a loose actin network (correlating to low and high velocities) (38)

(Figure 3-7A). To estimate the forces on a polystyrene particle, a flexible cantilever was

attached to an N-WASP coated particle and pulled while a micropipette held the actin rocket tail









stationary giving an estimate of 0.25 nN/im2 of stress on the particle (39). An atomic force

microscope (AFM) has also been used to measure actin network forces (44) resulting in a

pressure on the AFM cantilever of-1 nN/um2.

Deformable particles such as vesicles or oil droplets were used to estimate the magnitude

of forces generated by the actin rocket tail at the surface of the particle. Upadhyaya et al. (40)

and Giardini et al. (41) both exploited the pliable characteristic of vesicles to determine the

stresses generated by the actin rocket tail. By analyzing the vesicle tear drop shape formed from

an actin rocket tail Upadhyaya et al. (40) estimated a compressive stress on the sides of the

vesicle ranging from 3-4 nN/tm2 and a tensile stress on the rear of the vesicle of 6-8 nN/tm2.

The same tear drop shape was observed with oil droplets (42) (Figure 3-2). With all of the

experimental data available regarding actin polymerization, several models have been proposed

to describe actions interaction with a motile surface and the mechanism for force production.

Models for Force Generation by Actin Polymerization

Different models have been proposed to describe the mechanism of force generation by

actin polymerization. The main models that have garnered the most attention and credibility are

the elastic Brownian ratchet model (45), the tethered Brownian ratchet model (46), the

autocatalytic model (47), and the actoclampin model (22). The first three models mentioned are

based on the same principle of force production from monomer addition to free filament ends

(derived from the model proposed by Hill and Kirschner (48)) and differ only slightly in how

actin filaments are modeled. The actoclampin model is fundamentally different in that it

considers filament ends attached to the motile surface while polymerizing and utilizes the

additional hydrolysis energy to explain force production.

Terrell Hill initially hypothesized how monomer addition to a cytoskeletal filament could

produce a force in 1981 (49) and later expanded on the idea in 1982 (48). Hill's model









suggested a motile surface would move slightly away from polymerizing cytoskeletal filaments

by Brownian motion, with the stiff filaments preventing Brownian motion of the motile object

backward. Monomer addition to the filament end closest to the motile surface was accomplished

once the motile object fluctuated far enough away for a monomer to add essentially creating a

ratchet pushing the motile surface forward.

Many experiments showed thermal fluctuation of the motile surface to be insufficient to

produce the observed motion (45). For example, Listeria and ,.\/lgel/ both hijack actin

polymerization for motility, however, both bacteria move at the same rate even though .\/nge//t

is much larger than Listeria. According to the Hill model Listeria would have a greater velocity

than .\/nlgl//t because the Brownian fluctuations of Listeria would be greater allowing for a

faster rate of actin polymerization. To account for this observation of different sized loads

moving at the same rate, Mogilner and Oster later proposed a modification to the Brownian

ratchet model, the elastic Brownian ratchet model (45). This modification to the model suggests

filaments fluctuate away from the motile surface by Brownian motion instead of the surface

moving by Brownian motion. The elastic Brownian ratchet model predicts that if a filament is

sufficiently angled to the surface (> 10) and the filament is longer than 75 nm then an actin

filament will undergo Brownian motion away from the motile surface enough for a single

monomer addition (2.7 nm because the proto-filaments are helical) (Figure 1-5). The filament

then straightens due to its persistence length pushing the surface forward with the newly added

monomer. This cycle creates a Brownian ratchet that ensures there is a net forward movement

(45). Once a monomer has added to one filament, this lengthened filament supports some of the

load making monomer addition to other filaments easier. Mogilner and Oster (45) estimated the

stall force for an actin filament by the elastic Brownian ratchet model to be 1.8 pN based on a









free filament length of 150 nm (distance between motile surface and actin network) and a

persistence length of 1 im. The main ideas that make the Brownian ratchet model different from

Hill's model are filament ends fluctuate from Brownian motion and the motion of the object is

from the collective treadmilling of numerous actin filaments behind the motile object.

Work by Kuo and McGrath (50) gave more detail on the forces involved in actin

polymerization by optically tracking motile Listeria using a laser and photodiode to monitor

motion. Several conclusions came from this work that negated non-tethered filaments. First,

bacteria do not fluctuate enough for intercalation of G-actin monomers. Second, mean squared

displacement analysis suggests bacteria do not diffuse as predicted by Brownian ratchet

simulations. Third, bacteria fluctuate 20 times less than neighboring particles pointing to actin

tail attachment to the bacteria surface. Last, Listeria was observed to take regular steps of

5.4 nm (the diameter of G-actin monomer). The Brownian ratchet model would not predict a

persistent stepping of 5.4 nm because filament fluctuation must be less than the monomer

diameter due to the necessary angle the filament makes with the motile surface. Gerbal et al.

(51) used an optical trap as well, however, they used the trap to determine that greater than

10 pN force is required to separate a bacterium from its rocket tail.

To account for the experimentally observed attachment between the growing filament

network and the motile surface, Mogilner and Oster (46) modified their elastic Brownian ratchet

model to include two types of actin filaments, working and attached filaments. In this model,

once steady-state is achieved, filaments nucleate at the surface of the motile object while being

attached to protein complexes and under tension. Eventually these attached filaments will

dissociate and grow freely which are referred to as working filaments. These growing filaments









are under compression and are the same filaments as described in the elastic Brownian ratchet

model. Working filaments ultimately become capped and lose contact with the surface (46).

A separate model was proposed by Carlsson to explain the experimental findings of actin

polymerization in which filaments (daughter filaments) are autocatalytically branched from

existing filaments (mother filaments). Carlsson's motivation came from experimental work

showing mother daughter length correlation (52), total internal fluorescence studies showing

filament branches forming along filament sides (53, 54), and confocal microscopy showing

branches forming at the barbed end preferring newer filaments (55). The main difference of the

autocatalytic model in comparison to the Brownian ratchet model is the importance of filament

branching. Carlsson hypothesized that the formation rate of new branches is proportional to the

number of filaments or amount of polymerized actin near the motile surface. Therefore motile

objects with a larger diameter will have more autocatalytic filaments at the surface (due to the

increased surface area) distributing the load resulting in a growth velocity nearly independent of

applied force (47, 56). The autocatalytic model simply looks at how filaments interact with each

other in an actin network and is fundamentally the same as the Brownian ratchet model.

The above models, which rely on the free energy of monomer addition to free filament

ends to generate force, have a thermodynamic maximum force (stall force) (57)


Fm k In BT( A (1-1)
ax T(+)crnt

where kB is the Boltzmann constant, Tis temperature (kBTis 4.1 pN-nm), 6 is the added filament

length per subunit (2.7 nm), AT is free G-actin-ATP, and AT(+)cnt is the G-actin-ATP (+)-end

critical concentration (0.1 pM). Assuming AT must be less than the (-)-end critical concentration

(0.6 [M) for steady-state treadmilling, then Fma is less than 2.7 pN per active filament.









The actoclampin end-tracking motor model was first proposed in 2002 by Dickinson and

Purich (22) to explain how end-tracking motors both keep filaments persistently attached to a

motile surface and are capable of utilizing ATP hydrolysis for mechanical work. An

actoclampin is an end-tracking motor that facilitates the insertional polymerization of actin

filaments by affinity-modulated interactions between multivalent end-tracking proteins

(e.g. VASP, N-WASP, VCA peptide from WASP-family proteins) and the filament (+)-end.

Dickinson et al. (58) proposed that, while one end-tracking subunit binds one actin subfilament

end, another binds to a free G-actin-ATP monomer in solution and loads it onto the (+)-end of

the other actin subfilament. The energy released upon hydrolysis of ATP attenuates the binding

affinity of VASP to the actin filament, thereby releasing the filament end and binding another

G-actin-ATP monomer in solution. This process continues alternating attachment to the filament

while loading monomers resulting in constant filament attachment with insertional

polymerization (Figure 1-6). Actin has a conformation change (3) when ATP is hydrolyzed to

ADP which could explain why the clamp loses affinity for F-actin-ADP.

Importantly, by capturing energy released by ATP hydrolysis (up to 14 kBT (58)) and

allowing force-independent monomer binding from, the actoclampin model predicts a weak

dependence of filament elongation rate on compressive or tensile forces up to several pN (22).

In contrast, the Brownian ratchet models predict exponential force-velocity relationship and

much lower thermodynamic stall forces (< 2.7 pN). Consequently, according to the actoclampin

model, forward motion of a propelled particle should limit the elongation rate (or detachment) of

the most slowly elongating filaments in the actin tail. These filaments are under tension,

balanced by the compressive forces of other filaments at the motile surface. This "push-pull"

force balance explains many of the experimental observations of actin-based particle propulsion,









including monomer-sized stepping motion of Listeria (22), deformation of vesicles into tear-drop

shapes (55), and saltatory motion of larger particles (43). Any of these motile particles would be

affected by capping protein (such as gelsolin or CapG which bind to filament (+)-ends

preventing polymerization) if the filaments were not tethered, however, when capping protein

concentrations are increased in motility assays, motile particles are not hindered (59-61).

Filaments not being affected by capping protein suggests the filament (+)-ends are protected

possibly by the tethering protein complex of ActA-VASP as in the actoclampin model.

Distinguishing Between Actin Polymerization Models

The purpose of the present study is to conduct experiments to determine whether actin

filaments generate force while processively attached to the motile surface. Electron microscope

images of actin rocket tails trailing 500 nm beads were analyzed to determine the energy density

of the rocket tail to distinguish between the free filament and actoclampin models. Untethered

filaments are limited in the amount of energy that can be stored in and translated to the actin

network due to a small maximum force generation from only considering monomer addition and

filaments buckle at a much lower force than if tethered. The actoclampin model predicts much

higher energy storage in the actin rocket tail compared to free filament models.

Actin propelled vesicles were produced and analyzed in Chapter 3. The goal of producing

actin propelled vesicles was to measure the shape change of the vesicle and relate that to the

force generated by the actin rocket tail. Although force estimates were not accomplished, vesicle

velocity versus vesicle radius was determined. Calculations by Dickinson and Purich (43)

predict an inverse correlation of diameter to velocity which was confirmed in the findings of

Chapter 3 whereas free filament models would not predict the size and velocity to be directly

correlated. This difference between actoclampin and free filament models is because Dickinson









and Purich (43) determined particles >2 [tm in diameter will show saltatory motion from

diffusion-limited actin polymerization.

To determine whether filaments elongate by processive insertional polymerization,

electron microscopy of 50 nm beads with single filaments show several thousand single

filaments associated with single beads, demonstrating that individual filaments remain attached

to the motile surface during elongation. To determine the direction of elongation relative to the

bead, a fluorescent color change experiment was performed (Chapter 5) to demonstrate that the

point of monomer insertion was at the bead surface. These findings provide strong evidence

filaments are polymerizing from the bead surface and not randomly associating with the bead or

adsorbing from solution onto the bead. All of the findings in this study support the actoclampin

model.








Actin


Crawling
Cell


Edge


Figure 1-1. Actin filaments polymerizing at the leading edge of a cell producing filopodial
extensions that attach to the substratum temporarily holding the front of the cell in
place. The rear of the cell is then pulled forward allowing the cell to crawl.


ATP ADP



mm 1


ADP rich
(-)-end


Pol
Polymerization
- k".


Depolymerization


Filamentous
Actin
Hydrolysis


ATP rich
(+)-end


Figure 1-2. Simplified cartoon of an actin monomer with a nucleotide bound inside the actin
cleft. Actin monomers bind ATP in solution and polymerize into actin filaments
with the cleft oriented toward the (-)-end. The ATP is hydrolyzed and monomers
eventually depolymerize returning to solution to exchange ADP for ATP.











Pi

FPATP FADP-Pi F.ADP

-~-mom"ar i~~


r


G.ATP
A w F1ATP
AG = -1-2- k ',
jad|


AGr 14kBT
A hydrolysis


ATP


G-ATP G-ADP
ADP


22 k8T


'F-ADP-Pi F-ADP
AG =-0 ,T '
Pi-release B
AG 6 kBT
\,GADP G-ATP

AG =-Ok kT
AGexchange
The treadmilling process where an actin filament polymerizes actin-ATP, hydrolyzes
ATP, and depolymerizes actin-ADP at which point the process starts over with actin
nucleotide exchange (58). The free energy changes (AG) of each step in the
treadmilling process is calculated with one ATP hydrolyzed (-22 kBT (62)) per
monomer (Pi is phosphate, kB is Boltzmann's constant and Tis temperature).


P.
ATP i phosphate
hydrolysis release


SctinATP
addition


actin.ATF
Release


ATP
zD P
nucleotide
exchange
ATP

A p


1 tbprofilin
binding


ADF
Binding

rADF]




ADeF--"
release


ADF-
Mediated
Release


ADF


> actin-ATP

> actin-ADP-Pi

y> actin-ADP


profilin


ADF] ADF/cofilin


The treadmilling of an actin filament is enhanced with profilin and cofilin with
various pathways shown (58). The dashed box represents a polymerizing filament
sans profilin and cofilin.


Figure 1-3.


profilin (c
Release







profilin-
ushered
addition


Figure 1-4.


r>









Cross-linked
Actin Network











Figure 1-5. The elastic Brownian ratchet model hypothesizes free filaments fluctuate from a
motile surface by Brownian motion enough for a monomer to intercalate to the
(+)-end (45). The filament returns to its original configuration with an additional
monomer at the (+)-end. This additional length pushes the surface forward while the
filament is held in place by cross-links in the rocket tail.

A. End-tracking stepping motor




Monomer Release!
Addition Advancement

G acn binding region
SF -actn binding region
B. End-tracking stepping motor Profin


-4


P


I Monomer Monomer Release
Binding Transfer IAdvancement
Figure 1-6. Generalized end-tracking motor persistently bound to an actin filament (63).
A) End-tracking motor with only an F-actin binding region. B) End-tracking motor
with a G-actin binding region that loads a monomer and is then held by the F-actin
binding region.









CHAPTER 2
ENERGY DENSITY IN BENT ACTIN FILAMENTS OF AN ACTIN ROCKET TAIL

Introduction

Actin filaments are considered semi-flexible filaments with an average flexural rigidity

compared to other cytoskeletal filaments (64). This flexural rigidity is an important property

which allows actin to convert chemical energy into mechanical energy in order to drive a cell

forward or push bacteria through a cell. One way of characterizing the flexural rigidity of

cytoskeletal filaments is by the persistence length 2, which is the characteristic correlation length

in the orientation of a thermally undulating filament (6, 64). Reported values of the persistence

length for actin are in the range 10 + 5 [tm (65).

We propose a method for estimating actin filament bending energy in the actin rocket tail

behind a motile particle. Although single filament analysis of actin rocket tails is difficult to

accomplish because of the small size of filaments and dynamic change in filament growth,

electron microscopy (EM) allows visualization of static actin comet structures at the nano-level.

To produce an actin rocket tail that can be viewed with EM, beads are conjugated with ActA and

incubated in a cell extract. The beads with actin tails are then coated with a thin metal for EM

and imaged (32, 33).

Methods

To simplify the physical model for bent actin filaments, the actin filaments are considered

to be semi-flexible rods. As shown in Figure 2-1, each point along the filament is designated by

the position vector r(s) which encompasses the Cartesian coordinates [x(s), y(s), z(s)]. The arc

length is defined as s and the unit tangent vector is designated as N(s),

N(s) = dr/ds. (2-1)









The change in the unit tangent along the length is proportional to the unit normal to the

curve n (64),

dN/ds = Cn. (2-2)

If the unit normal vectors are significantly close and extended to the origin then it can be

deduced that C is the local radius of curvature which is necessary for relating the change in

tangent vector to the radius of curvature,

C 1/R. (2-3)

The energy associated with the bending of a rod has been solved (66) and has the form

E B
E-- (2-4)
L 2R2

where E is energy L is filament length, and B is the bending modulus, related to the persistence

length as

B = LkBT (2-5)

where ks is Boltzmann's constant (1.38 x 10-23 J/K), and Tis temperature (K).

The deformation energy per unit length of the bent filament is proportional to the square of

the radius of curvature (64) or equivalent to the square of Equation 2-3. Squaring Equation 2-3

and combining with Equations 2-3 and 2-4 results in the energy per unit length. Since the

curvature is not required to be constant, the energy of bending can be expressed as the integral

along the length of the filament or as the change in energy per change in length (Equation 2-8)

(67),

dE B dN (2-6)
ds 2 ds)

Electron microscopy images of actin comet tails (Figure 2-2) were analyzed to estimate the

energy stored in actin filaments. Filaments undergo tension and compression due to elongation









forcing a network of bent crosslinked filaments, which is evident from EM images (Figure 2-2).

Bent filaments are a result of monomer addition to crosslinked filaments, which are constrained

from extending, forcing the filament to bend. The bending energy can be determined by

measuring the curvature of the filaments and using Equation 2-6. Due to the necessary treatment

of the actin rocket tail for EM, some twisting and bending may occur to the filaments. The

additional energy associated with EM treatment is considered negligible because filaments that

have only one interaction with another filament are typically straight suggesting the EM

treatment has not affected the configuration. Filaments that are bent or twisted are usually

restrained by at least two points of crosslinking or filament interaction which is to be expected

from actin polymerization (33).

An algorithm was developed to determine the amount of energy stored in actin filaments as

measured from EM images. The angles of bent filaments were measured in the plane of focus.

However, one problem with determining the amount of bending in an actin filament from an EM

image is that only the two-dimension projections of the filament arcs are measurable for a

three-dimensional filament. To adjust for the missing third-dimensional data, a correlation

between the two-dimensional projected curvature of a filament and the actual three-dimensional

curvature is needed. The below derivation relates the change in energy per length of filament in

three-dimensional space to the change in energy per length of filament in two-dimensional to

estimate the amount of energy spent in bending the filament. N* is the unit tangent vector in the

EM image and s* is the projected length of the filament in the EM image. Theta (0) is the angle

with respect to the x-axis in the plane of the EM image while phi (q ) is the angle in to or out of

the EM image (Figure 2-3),

N cos6O
N* = (2-7)
Ssmin9









(- sin d>
dN*= cosn dO. (2-8)
yCosO0

The viewed length is equal to the change of the angle along the actual length of the filament (s).

Therefore, the viewed length is a function of the changing angle of the filament as it goes in to or

out of the focal plane,

L
s*= jsin (s)ds, (2-9)
0

ds* = sin q(s)ds. (2-10)

Equation 2-8 divided by Equation 2-10 is the viewed change in angle divided by the

viewed change in length,

sin 0"
dN*_ sinq dO
dN sin ]dO (2-11)
ds* cos0 ds
V sin )

The square of Equation 2-11 is needed to relate the two-dimensional values to the actual

three-dimensional orientation of the filaments,

dN (-sin ) +(cos )2 dO 2 dN* 1 dO
2 (2-12)
ds*) sin2( ds) [ds*) sin+2 ds)

The three-dimensional expression ofdN/ds must be determined to equate with the

two-dimensional change in tangent vector with respect to length,

sin cos 0
N = sin sin0 (2-13)



sin q sin 0 + cos 0 cos 0
dN= sin cos + cos sin0 d0d| (2-14)
sin








Dividing dNby ds and squaring gives (dNIds)2, which is used in the energy equation

(Equation 2-6). Through manipulation of Equation 2-14, (dN/ds)2 can be represented as a sum of

changing angles,

(dN>2 d S- + sin (2-15)
ds ds ds

The (dOlds)2 component in Equation 2-15 can be obtained by rearranging Equation 2-12.

Equation 2-15 can then be written in terms of the two-dimensional change of the filament,

(dN7 2 =do2 4 d2 dN 2 2
N- +sin4 (2-16)
ds ds) ds)

Equation 2-16 gives a relation between three-dimensions and two-dimensions; however, there is

still a (do/ds)2 term that cannot be determined directly from the two-dimensional images. A

uniform distribution is assumed for the orientation of each filament therefore the changing and

turning of the filaments should be approximately the same for both the 0 direction and the 0

direction. Thus, the assumption was made that twice the sin2 0(dO/ds)2 term will give

reasonable results. Since the filaments can grow and point in any direction, a uniform

distribution of filaments is used to account for the different orientations of filaments and to

account for the 0 direction. The distribution used is assumed to be a uniform distribution

around a sphere,

1
p(, 0)= (2-17)
47r

With the adjustment in Equation 2-16 due to the (dq/ds)2 term and combining

Equation 2-16 with Equation 2-17, the average change in angle per change in length of the

three-dimensional filament is related to the two-dimensional data through Equation 2-18,









/(adNj\ 1 i i2 4 (dN *2.d d 16 YdN *
-- 2)=- 2sin4 0 s1in6 dd=- (2-18)
ds) / 4r \ ds*) 15 ds*)

The integration of Equation 2-18 substituted into Equation 2-6 gives an approximation of the

average amount of energy per length of a bent filament,

dE\ B16 dN\*
= 16 dN(2-19)
ds/ 2 15 ds*)

Using the change in position vector instead of the change in tangent vector allows for

easier measurements of the filaments. The position vector is defined as r = + yj and the

second derivative of the position vector is equivalent to the first derivative of the tangent vector,

d 2 T dN 2
C- (2-20)
ds2 ) \ds* "

Substituting Equation 2-20 into Equation 2-19 gives the energy equation as a function of

the position vector. The addition of a correction factor and expanding the second derivative of

the position vector yields the relation of energy to filament position

AE' 8A N 8 r, 2r, + r, N -
kT 15 As + (2-21)

where a is the error in measuring the filament position (2 nm as determined by the images), and

r,,, is the measured position of the filament. Since there will be some intrinsic error in

measuring the filament position, the measured position is smoothed to a minimum to reflect as

close to the actual filament position as possible. The energy equation (Equation 2-21) is

minimized for the points to find the maximum likelihood estimation for the filament contour.









Results

Cameron et al. (33) prepared several beads with actin rocket tails for EM as described

above. Additional unpublished images of actin rocket tails, beyond those published in Cameron

et al. (2001), were generously provided by Tatiana Svitkina, Department of Biology, University

of Pennsylvania. Using the National Institutes of Health's (NIH) image software ImageJ, several

hundred filaments across 12 images were mapped. For consistency, several criteria were

followed to increase the accuracy of determining the change in energy of a filament. Only

filaments that could be clearly distinguished with lengths greater than 20 nm were measured. A

minimum of 50 filaments per image were measured to give a broad representation of filaments

found in the actin rocket tail. Points along the filament were kept at a distance of 2 nm apart and

measured points were taken as close to the axis of the filament as possible.

Figure 2-4 and Figure 2-5 show filaments plotted by the above criteria. An algorithm

(Appendix A) was created in Matlab to optimize the points along the filaments and determine the

energy density of the filaments (Figure 2-6 and Figure 2-7). This energy density in the filaments

is a result of actin polymerizing and bending filaments in the rocket tail or bending itself between

the load and rocket tail. Figure 2-6A is the optimized filament position and Figure 2-6B shows

the levels of energy density by means of a contour graph. Since the filament in Figure 2-5 is

almost straight the energy along the filament changes slightly with a maximum energy density of

0.86 pN. This process is carried out for each of the filaments shown in Figure 2-4 with the final

result for the actin rocket tail used shown in Figure 2-7 and a histogram of the results in

Figure 2-8.

Discussion

Several hundred filaments were analyzed resulting in an energy density ranging from

0.2 pN for slightly bent filaments up to 259 pN for extremely bent filaments (Figure 2-9). The









average energy density for the 670 filaments measured in 12 images was 3.1 pN which can be

compared to the theoretical models. The elastic Brownian ratchet model (45, 46) or similar

models relies on force generation solely from monomer addition to free (untethered) filament

(+)-ends. The theoretical maximum energy per added length for monomer addition is 2.7 pN,

under conditions that promote (-)-end depolymerization actinn concentration < 0.6 [M). In order

to produce a large energy density (around the maximum of 2.7 pN) in the rocket tail, actin

polymerization would have to continue at optimal conditions for the life of the rocket tail. When

only monomer addition is the energy source for filament bending, an energy density in the rocket

tail of the maximum 2.7 pN would not be observed because some energy is lost in the forward

motion of the bead and filaments typically have a large incidence angle with the bead reducing

the amount of energy transferred to the rocket tail (Figure 2-2).

In contrast, the actoclampin model utilizes the energy of hydrolysis to drive actin

polymerization of tethered filament (+)-ends, resulting in a larger energy source for motility than

would be provided by monomer addition alone. The total treadmilling cycle of actin

polymerization provides up to 22 kBT of energy with a majority coming from ATP hydrolysis

(-14 kBT) (58). The actoclampin estimates are more than enough to account for the energy

density observed in the EM images of actin rocket tails. Another advantage of the actoclampin

model over untethered models is the amount of force a filament can withstand before buckling.

Dickinson et al. (58) showed mathematically that a tethered filament can handle an order of

magnitude greater force (considering filaments perpendicular to load surface) over an untethered

filament. As untethered filaments become more glancing to the load surface, less force is

necessary to buckle the filament. Even when thermodynamic considerations are not considered,

tethered filaments have a mechanical advantage over untethered filaments.









Other groups have estimated network forces of actin rocket tails by a variety of methods.

Weisner et al. (36) estimated the polymerization force of an actin rocket tail to be greater than

50 pN using beads coated with N-WASP by studying the velocity of the beads through a viscous

solution. McGrath et al. (35) estimated the actin rocket tail polymerization force to be -10 pN

on motile Listeria similar to the approach by Weisner et al. Analysis of actin propelled vesicles

by Giardini et al. (41) and Upadhyaya et al. (40) show compressive stresses of 3 to 4 nN/um2 and

tensile stresses of 6 to 8 nN/um2. Upadhyaya et al. (40) estimated single filaments could produce

-10 pN of force. All of these estimations give a broad view of polymerization force by an actin

network but are limited in showing energy associated with individual filaments.

There are some limitations of the measurement of bent filaments to estimate the energy

change per filament length. First, the observed bending of a filament may or may not be due to

the actual growth of the filament but could be from the cumulative effect of other filament

growth. Because filaments overlap and end interactions are unknown, the filament ends are

considered straight and therefore have a value of 0. Also, stored mechanical energy could

dissipate with deformation of the rocket tail and depolymerization of filaments which would

lower the overall measured energy density. Second, not all of the filaments can be measured in

the actin tail due to hindrance from the exterior layer of filaments blocking visualization of

interior filaments. This leaves out filaments that might be straight which would lower the

average if incorporated in the analysis or bent filaments which would increase the average of

energy. Also, filaments under tension are not distinguishable from straight filaments which

would affect the energy density. To address this issue, a large population of filaments was used

to give an overall representation of filaments found in the actin tail. Third, the angle coming out

or into the image is unattainable due to the two-dimensional aspect of TEM. Since this angle is









not measured the data lost is only of filaments bent in or out of the image. Therefore, this

additional data would most likely increase the determined energy of the actin tail, although this

additional angle is made up for with the estimation made in Equation 2-18. Another concern is

the cross over of filaments with each other. This cross over prevents the accurate measure of

filament lengths and angles. To counteract this problem, only filaments that could be

distinguished for at least 20 nm were measured. Finally, the amount of perturbation from EM

treatment is unknown. However, the affect of EM treatment seems to be trivial since free

filaments or filaments with one point of interaction with another filament are usually straight.

Whereas filaments that were bent or twisted were usually restrained by at least two crosslinks

(33). Also, filaments could be affected when the rocket tail is bound to the substratum, however,

filaments closest to the substratum are not visible and periphery filaments not multiply bound to

the rocket tail were avoided.

In conclusion, the estimated energy density exceeds what should be provided by monomer

addition alone. Although not a strong proof of one model for force generation over another,

these results argue against Hill type models and favor the actoclampin model. The actoclampin

model not only explains the high energy density (compared to the untethered filament models)

but also explains the physical filament attachment to the bead surface as observed in the EM

images (Figure 2-2).





















Figure 2-1. Notation used to describe a flexible rod / actin filament where s is the arc length, r(s)
is the position vector, N(s) is the unit tangent vector, and nl and n2 are normal
vectors.


figure 2-2. A 3)U nm polystyrene Dead tunctionalized witt ActA ana exposed to a motility
assay to induce an actin rocket tail. The sample is then treated and viewed with
TEM (Unpublished image provided by Tatiana Svitkina, Department of Biology,
University of Pennsylvania and used by permission).









A
'IV
z





/ \

N*(s *) .... '
Cartoon of an actin filament projecting an image on a two-dimensional plane where
N(s) is the unit tangent vector of a three-dimensional filament of arc length s, N*(s*)
is the unit tangent vector of a three-dimensional filament projected onto a
two-dimensional plane or arc length s*, 0 is the angle of the two-dimensional
filament, and 0q is the angle out of the plane of the two-dimensional image. The eye
facing down is analogous to looking at an EM image.


Figure 2-4. Close up of an actin rocket tail that has points plotted along the filaments. The white
box designates the zoomed in section shown in Figure 2-5.


Figure 2-3.


















Figure 2-5. Close up from Figure 2-4 showing points plotted along a single filament.


860 670 880 890 900 910 920 930 940
x position in image
Maximum dE/ds = 0.86 pN
Analysis of the bending energy of filament in Figure 2-5. A) The red dotted line is
the actual position of the data found. The blue line with circles is the best-fit line.
B) A twenty level contour plot of the energy change along the filament with blue
being lowest energy and red being highest. The dashed lines point to the maximum
energy value of 0.86 pN.


Figure 2-6.














(D

E



> -810


>, -10oM


rt


900 100 1100
x position i image


1200 130C


Figure 2-7. Output from algorithm with filaments numbered and points plotted with a blue line.
The largest energy found (at filament 47) is shown by the dashed lines.


400

350

300

> 250
Cr

: 200

LL 150

100

50


0 10 20 30 40 50 60 70
dE/ds (pN)
Figure 2-8. Histogram with a 0.65 bin size of all dE/ds values (excluding filament end points)
along 53 filaments in one actin rocket tail.


CI










7000


2flCr


150


5000
100
O 4000
C

3000
L-
2000 10 60 110 180


1000



0 50 100 150 200 250 300
dE/ds (pN)
Figure 2-9. Histogram with a 5 pN bin size of all dE/ds values (excluding filament end points)
along 670 filaments in 12 actin rocket tails. The inset graph is the same data with
the first two bins removed and the three largest values removed (239 pN, 257 pN,
and 259 pN).









CHAPTER 3
ACTIN PROPELLED VESICLES

Introduction

Actin based motility of bacteria or nonbiological cargo such as beads (32, 37, 38, 68, 69)

are useful for studying tail velocities and the actin rocket tail network. Bacteria and beads are

rigid and do not deform under the forces exerted on the surface by actin filaments. Here, we

focus on the use of compliant vesicles as a potential method for estimating actin polymerization

forces. Vesicles are convenient particles to study compared to actin propelled beads for several

reasons. First, vesicle lipid concentrations can be controlled and lipids are free to diffuse in the

membrane. Second, vesicles mimic the plasma membrane of protruding lamellipodia and

cellular liposomes which allows for greater insight into how actin polymerization affects

motility. Finally, due to the vesicles pliable surface, actin polymerization forces associated with

the vesicle surface can be analyzed through measurement of vesicle deformation (40).

Vesicles (or liposomes) are composed of self-assembled amphiphilic lipids (70), which

have a polar hydrophilic head group and a hydrophobic aliphatic chain (the tail group). Because

of the solubility difference between the two groups the lipids form ordered structures in aqueous

solutions. Depending on the concentration of lipids, length of the aliphatic chain, and the

number of tails associated with a polar head group determines the structures formed in the

aqueous solution. The lipids aggregate to form self-closed spherical particles where one or more

lipid membranes encapsulate part of the solvent (70) (Figure 3-1). These vesicles are described

based on their size and number of membranes. Small unilamellar vesicles (SUV) are typically

50 to 100 nm in diameter and consist of a single bilayer of lipids. Giant unilamellar vesicles

(GUV) are greater than 1 atm in diameter and large unilamellar vesicles (LUV) range in size









between SUVs and GUVs. The same size designation is used for vesicles containing multiple

layers of lipid bilayers using the term multilamellar instead of unilamellar.

To nucleate new filaments at the vesicle surface, ActA must be bound to the vesicle to

promote actin polymerization. When these vesicles conjugated with ActA are exposed to a cell

cytoplasmic extract, actin polymerizes on the vesicle surface and the resulting filaments

aggregate on one vesicle hemisphere allowing for symmetry breaking and propulsion (lipids are

mobile in the membrane). As the vesicle is propelled forward, the vesicle undergoes a

conformation change from its original spherical form, creating a tear drop shape (Figure 3-2).

Using osmotic pressure, membrane tension, and comparing the shape change to the original

spherical shape, the stresses due to actin polymerization can be estimated at different locations

along the vesicle surface (40, 41).

Two groups have exploited vesicle deformation by actin polymerization to estimate actin

forces (40, 41). Both groups used similar approaches in measuring the forces, using vesicles

developed with similar methods and mechanochemical models (Table 3-4). The key

assumptions made were: force associated with membrane bending is negligible, vesicles are

unilamellar, vesicles are not stretched and have no osmotic pressure when in the rest state

(without actin polymerization), hydrodynamic drag is negligible (<100 fN for a 3-[tm-diameter

vesicle), and actin polymerization combined with membrane tension balances the osmotic

pressure inside the vesicle (40, 41).

Vesicles were exposed to a motility assay and once motile, the vesicles went through

several cycles of changing from spherical to tear drop shape suggesting filaments were attached

to the surface and some of the filaments suddenly detach from the trailing edge due to the

increase in membrane tension. Based on a reaction-diffusion model, Dickinson and Purich (43),









explained these deformations as arising from a depletion of available monomers and a reduced

rate of polymerization of tethered filaments at the center of the actin rocket tail. Since the

filaments are attached to the surface the faster periphery filaments push the vesicle forward while

the lagging interior filaments create a pulling force on the rear of the vesicle which produces the

tear drop shape (Figure 3-2).

The original goal of this study was to produce large actin propelled vesicles (> 5 [tm) and

explore the forces exerted on the surface of the vesicle from the rocket tail. Determining the

surface force was to be accomplished by anchoring the rocket tail and aspirating the front of the

vesicle with a micropipette. The actin rocket tail force on the surface would then be estimated by

analyzing the balancing opposing forces of the increased osmotic pressure of the vesicle versus

the polymerization force of the rocket tail. However, reproducing the larger motile vesicles

using the published protocols was unsuccessful, and we had to develop a new protocol to

produce motile vesicles. Although vesicles forces were not measured, the vesicle velocities were

measured and correlate well with Dickinson and Purich's (43) theoretical findings.

Materials and Methods

Bovine Brain Extract

Whole bovine brain was stored at -700C until needed. Brain was removed from storage

and weighed. The brain was crushed with a mortar and pestle in a cold room while under liquid

N2. The powder was transferred to a dounce homogenizer where it was mixed with an equal

volume of Tris MgCl2 buffer including PIs (10 mM Tris HC1 pH 7.5, 2 mM MgC12, 10 [g/mL

pepstatin A, 10 [g/mL leupeptin, 10 [g/mL chymostatin, and 1 mM PMSF). The mixture was

dounce homogenized 30 times while on ice being careful to avoid bubbles. The homogenized

brain was sonicated on ice with a tip sonicator for 30 second bursts and 1 minute rests at a level

of 25% on a power level of 1.5 again avoiding bubbles. The solution was transferred to a









centrifuge tube and centrifuged at 17,000 g (15,000 rpm in a Ti-60 rotor) for 20 minutes at 40C.

The supernatant was saved and centrifuged again at 118,000 g (41,000 rpm in a Ti-60 rotor) for

1 hour at 40C. Bradford assay was used to measure the protein concentration which should result

in approximately 10 mg/mL. The supernatant was aliquoted and stored at -700C until needed

(71).

Bradford Assay

The Bradford assay was used to get an approximation of protein concentration. The

sample was diluted and compared to a standard Bradford curve of appropriate protein (BSA)

(72). Table 3-1 lists the concentrations to use in a parallel dilution for the Bradford assay.

Actin Purification from Rabbit Muscle

Acetone powder

Rabbit muscle was frozen and stored at -700C until needed. All volumes used were based

on the measured mass of the rabbit muscle. Table 3-2 lists the volumes of buffers needed for

1 kg of rabbit muscle. To prepare acetone powder, a section of rabbit muscle was weighed and

stored at -200C overnight. The next day the rabbit muscle was placed at room temperature for

2 hours. The extraction buffer (450 mM KC1 pH 6.2 with KOH, 150 mM KH2PO4, and 0.1 mM

EDTA) was supplemented with 1 mM ATP, 0.2 mM DTT, and 0.5 mM MgC12. Fat and

connective tissue were removed from the muscle with a scalpel. The muscle was placed in a

clean beaker on ice. Two hundred mM PMSF in DMSO was prepared and used to supplement

the extraction buffer to 0.5 mM PMSF. The meat was ground two times in a cold room with a

meat grinder. Extraction buffer was added with PMSF to the ground muscle and stirred for

30 minutes. The mixture was poured into a bucket with cold water and stirred in the cold room.

The solution was then filtered through 4 layers of cheesecloth and placed over a chilled extractor

with a vacuum attachment. The muscle was removed from the cheesecloth and placed into a









clean beaker. Extraction buffer without PMSF was added and stirred for 30 minutes at 40C. The

mixture was poured into a bucket with cold water (Table 3-2) and stirred in the cold room. The

solution was filtered through 4 layers of cheesecloth and placed over a chilled extractor with a

vacuum attachment. Muscle was removed and placed in a clean beaker at room temperature with

the appropriate volume of 0.4% NaHCO3 added and stirred for 45 minutes. The solution was

filtered through 4 layers of cheesecloth and placed over a chilled extractor with a vacuum

attachment. The muscle was placed into a clean beaker in the cold room 1 mM Tris was added

and stirred for 5 minutes. The solution was filtered through 4 layers of cheesecloth and placed

over a chilled extractor with a vacuum attachment. The previous two steps were repeated.

Muscle was added to 1/6 the total volume of ice-cold acetone (-20C) and stirred for 5 minutes

on ice in the cold room. The solution was filtered through 4 layers of cheesecloth and placed

over a chilled extractor with a vacuum attachment. The previous two steps were repeated.

Muscle was added to 1/9 the total volume of ice-cold acetone (-20C) and stirred for 5 minutes

on ice in the cold room. The solution was filtered through 4 layers of cheesecloth and placed

over a chilled extractor with a vacuum attachment. The previous two steps were repeated five

more times. After the last wash, the acetone powder was placed on a dry 3 mm Whatman paper

(or 3 sheets of Whatman #1), covered, and allowed to dry overnight in a desiccator under house

vacuum to avoid humidity. The final powder was aliquoted and store at -200C (73).

Purification of actin from acetone powder

Two grams of acetone powder was weighed out to make 2 mg of actin. A 5% acetone

powder solution in G-buffer was made by adding 2 g of powder to 40 mL G-buffer (5 mM Tris

pH 8 with KOH, 0.01% NaN3, 0.1 mM CaC12, 0.2 mM ATP, 0.2 mM DTT, stored at 40C with

ATP and DTT added just before use). The slurry was stirred with a glass rod in a glass beaker

on ice for 25 minutes (every 2 minutes for 15 seconds). The solution was filtered through









6 layers of cheesecloth and centrifuged at 40C (30 minutes at 30,000 g using Ti-45 at

19,000 rpm). The supernatant volume was measured and supplemented to 2 mM MgC12. The

solution was stirred for 5 minutes at room temperature. Solid KC1 was slowly added to a final

concentration of 0.07 M and stirred for 1 hour at room temperature. The solution was

centrifuged for 3 hours at 120,000 g at 120C with a Ti-45 at 38,000 rpm. The supernatant was

removed and 1 to 2 mL G-buffer was added to each pellet to wash. Then a total of 6 mL

G-buffer was distributed among all the pellets (73). The next morning the pellets were broken

up and dissolved. The absorbances at 280 nm and 290 nm were measured to get an estimate of

the protein concentrations. The pellets were then dialyzed against 1 L of G-buffer with

2 changes for at least 3 days. After dialysis the absorbance at 280 nm and 290 nm was measured

and the concentration was calculated using the molar extinction coefficients, S280 = 1.11 L/g-cm

and 8290 = 0.63 L/g-cm which are accurate between 0.1 and 1 absorbance units.

Fluorescent labeling of actin

Filamentous actin was diluted to 60 [iM in 2 mL of labeling buffer (20 mM HEPES

pH 7.5, 0.1 M KC1, 2 mM MgC12, 3 mM NaN3, 0.3 mM ATP, with ATP added just before use).

The diluted F-actin was then dialyzed in 1 L of labeling buffer for 2 hours with one solution

exchange to remove any amine containing buffers from solution. One mg of dye was dissolved

in 100 iL DMF or DMSO immediately before mixing with actin. The dye was sonicated

1 minute to mix the dye thoroughly. While vortexing the F-actin solution, the dye was added

dropwise (slowly) to a final concentration of 300 [M (5 fold concentration) (74). The mix was

incubated at 40C overnight while mixing end-over-end. The reaction was stopped by

supplementing to 50 mM lysine. The supernatant was then centrifuged (45 minutes at 290,000 g

using a TLA 100.2 rotor at 90,000 rpm). The pellet was resuspended in 2 mL G-buffer by

slowly adding buffer and triturating with a glass rod. The resuspended protein was then









sonicated for 1 minute in a bath sonicator. The mix was swirled for 30 minutes at 40C using a

vortex setting of level 1. The solution was then dialyzed against three changes of 1 L G-buffer

for 36 hours. The absorbance of the actin at 280 nm and the absorbance of the dye were

measured. The major fractions (> 4 pM) were pooled and polymerized for 45 minutes at room

temperature by supplementing to 1 mM ATP and lx P-buffer (10x P-buffer is 0.5 M KC1, 20

mM MgC12 which is always mixed with G-buffer). The supernatant was then centrifuged

(45 minutes at 290,000 g using a TLA 100.2 rotor at 90,000 rpm). The pellet was resuspended in

G-buffer to a desired final G-actin concentration. The solution was then dialyzed against

3 changes of 1 L G-buffer for 36 hours and the supernatant was centrifuged (45 minutes at

290,000 g using a TLA 100.2 rotor at 90,000 rpm). The absorbance at 280 nm and Xmax for the

dye were measured and the G-actin concentration was calculated by using the Beer-Lambert law

adjusted for the dye (75)


[G-actin]= I [A28onm AmaCF] (3-1)
1280nm

where I is the length of the absorbance chamber (1 cm), e is the extinction coefficient, A280nm is

the actin absorbance at 280 nm, Ahmax is the actin absorbance at the maximum absorbance of the

fluorophore, and CF is the correction factor specific for each fluorophore.

Preparing Listeria monocytogenes on Agar Plates

Brain heart infusion (BHI) media with agar was prepared and autoclaved. Ten pg/mL of

chloramphenicol was added to the media to prevent bacterial growth (other than bacteria immune

to the antibiotic). The media was distributed among Petri dishes and cooled. Listeria was

streaked on the agar media and incubated at 37 C overnight. The Listeria plates were stored at

4C and individual colonies were acquired as necessary.









Purification of ActA-His6 from Listeria monocytogenes

A Listeria culture from a previously prepared agar plate was removed and grown overnight

in a culture tube containing 50 mL BHI with 10 [g/mL chloramphenicol at 370C while shaking.

The media containing incubated Listeria was added to 1 L of BHI containing 10 [g/mL

chloramphenicol and grown overnight at 370C while shaking. The broth was added to a large

centrifuge tube and cooled on ice. Listeria was pelleted with a JA-10 rotor at 5,000 rpm at 4C

for 20 minutes. The supernatant volume was measured and 50% ammonium sulfate was added

at 4C slowly over the span of an hour to salt out the protein. Ammonium sulfate increases the

overall volume when salting out a protein so the appropriate amount was calculated to include

the specific volume of ammonium sulfate. The mass of salt per initial volume (G) (Table 3-3) is

solved for

m
VfC, = VC, +- (3-2)
MW

m C,V, m +M,C G +M2C,
C =- (3-3)
MwV, V, M (V, +mV) M (1+GV,)


M,(Cf -C,) 519.08g/L (3.93M 0.5 -3.93M 0.0)
G= = 301.84g/L (3-4)
1 CfMw 1- 3.93M0.5.132.14g/mol 0.54mL
1000mL/L

where Vf is the final volume, V, is the initial volume, m is the mass of salt to add, Vs is the

specific volume of ammonium sulfate (0.54 mL/g), Cf is the final concentration of ammonium

sulfate (product of final percentage and molar saturation at 40C of 3.93 M which varies with

temperature), C, is the initial concentration of ammonium sulfate, and Mw is the molecular

weight (132.14 g/mol).

The ammonium sulfate was allowed to reach equilibrium with the protein for 3 hours while

stirring at 40C. The precipitate was then centrifuged with a JA-10 rotor at 8,000 rpm for









30 minutes at 40C. Less than 10 mL of HKB (20 mM Hepes pH 7.4 with KOH, 50 mM KC1)

with one tablet of complete protease inhibitors mini tablet was used per one liter of original

bacterial broth. The solution was then clarified by centrifuging with a JA-20 rotor at 15,000 rpm

for 15 minutes at 40C. Five mL of Talon Ni-NTA resin was equilibrated according to

manufacturer specifications and mixed with the supernatant for 30 minutes at room temperature

to bind the ActA to the resin. The resin was then washed according to manufacturer

specifications and the protein eluted using HKB with 250 mM imidazole and fractions were

collected. Remaining protein was eluted using HKB with 500 mM imidazole and was discarded.

The kept fractions were measured using a Bradford assay to get a rough estimate of the protein

concentration. Fractions with the majority of ActA were dialyzed overnight in HKB and a

Bradford assay was used to determine the final concentration (76).

Fluorescent Labeling of ActA-His6-Cys

Strain DP-L4363 has an additional cysteine group as part of the protein which a thiol group

was attached to labeling the protein with a fluorophore. First, the absorbance was measured at

280 nm and at the wavelength of the fluorophore. Dithiothreitol was added at 20 times the

concentration of ActA to break up any cystine formation and incubated at 40C for 30 minutes

while rotating. Quick dialysis was done to remove DTT by doing 4 dialysis changes over a

4 hour period using a total of 1 L of HKB. The maleimide fluorophore was added at a

concentration of 20 times of ActA and incubated overnight at 40C while rotating (wrapped in

foil). P-mercaptoethanol was added at 20 times the ActA concentration to bind any remaining

dye and incubated at 40C while rotating for 30 minutes. Quick dialysis was done as before and

the absorbance was measured at 280 nm and at the fluorophore wavelength to determine the

amount of labeling (40).









Vesicle Preparation

Vesicles are formed when a dried lipid layer is exposed to an aqueous solution. Vesicles

are made with a rotary evaporator (rotovap) which reduces the pressure of a continuously rotated

round bottom flask. Lipids are dried on a round bottom flask and rehydrated with the buffer of

choice (77). Making vesicles with 0.1 mg to 0.2 mg of lipids using a 100 mL round bottom flask

yields a large amount of vesicles per milliliter. The lipids are then rehydrated with 1 to 2 mL

buffer.

Small vesicles were made by filling the rotovap tank with de-ionized water and heating to

65C. Glass was cleaned by using a modified acid wash technique (74). The appropriate amount

of lipids was added to the round bottom flask, attached to the rotovap, and a stream of N2 was

applied for 5 minutes. Nitrogen was stopped, a vacuum applied, and the rotovap set to a rotation

of 60 rpms. Once the chloroform was evaporated leaving a layer of lipids on the surface the

flask was removed and placed under a N2 stream for 30 minutes. Lipids were rehydrated with a

buffer (Table 3-4) and allowed to rotate on the rotovap for several hours or overnight. Large

vesicles were made in the same method as above except using a beaker and no agitation. The

vesicles were also allowed to rehydrate undisturbed for 2 to 3 days (78).

Vesicles have a Ni-NTA lipid incorporated to specifically bind to the His6 tag on ActA.

ActA was bound to the Ni-NTA lipid by mixing with prepared vesicles and incubating overnight

at 4C without shaking. If the vesicles and ActA are agitated the vesicles will not be motile

when added to a motility assay. Table 3-5 lists the ratios and dilutions used to bind ActA to

vesicles.

Motility Assay

A motility assay is a mixture of components that induces actin polymerization when used

with an object having ActA bound to the object's surface. The motility assay consists of an ATP









regenerating component, a protein degradation preventing component, a thiol reducing agent, a

cell extract, and an object with ActA bound to its surface.

Creatine kinase

The creatine kinase mixture is necessary for regenerating ATP from ADP and Pi (79). The

creatine kinase mixture was made as a 10x solution containing 100 mM phosphocreatine

disodium salt hydrate enzymatic, 20 mM ATP, 20 mM EGTA, 20 mM MgCl2, and creatine

phosphokinase at 51 units/mL. One tenth of this mixture was mixed with the motility assay to

produce the correct concentration needed for ATP regeneration. The mixture was made and

stored at -700C.

Protease inhibitors and DTT

Protease inhibitors (PI) are mixed together as a 10x solution with 10 [g/mL for each

inhibitor leupeptin, chymostatin, and pepstatin A. The PIs are mixed in DMSO and stored

at -70C. Dithiothreitol (DTT) is the thiol reducing agent and was made fresh daily to a

concentration of 100 mM.

Results

Several variations of vesicle motility were attempted. The best results (the most vesicles

with actin rocket tails) were obtained by diluting BBE by half, an [ActA]/[Ni-NTA] ratio of 0.02,

and an actin concentration ranging from 3 pM to 5 pM. We found that ratio of [ActA]/[Ni-NTA]

cannot be too high or excess ActA will promote actin polymerization in solution (producing

background filaments), and it cannot be too low or there will not be enough filaments associated

with a vesicle to visualize the actin rocket tail or produce motility. An actin concentration in the

range of 3 tM to 5 tM was found to produce long, fast growing actin rocket tails. Gel-filtered

actin and non-gel-filtered actin were compared in motility assays with vesicles and Listeria. No

difference was observed in using either actin so non-gel-filtered actin was used throughout this









study. It is likely that the gel-filtered actin did not change motility outcomes because the actin

was added back to the cell extract, which already contains actin oligomers and other protein that

are normally removed in gel-filtering.

Many of the vesicle rocket tails (>50%) were found to emanate from a mass of vesicles

(Figure 3-3) most likely due to the lipids tendency to aggregate, however, vesicle aggregation did

not prevent vesicle motility. The velocity of actin propelled vesicles was determined by using

actin monomers labeled with different fluorophores. Vesicles were exposed to one color of actin

for a few minutes and then another color of actin for a few minutes. The length of the rocket tail

for each color was measured and velocities were determined from the amount of time the actin

was in the flow chamber. Most vesicles had two color actin rocket tails while a few vesicles

only incorporated one color of actin and produced short rocket tails (Figure 3-4). One color

rocket tails are probably due to lack of filament nucleation (only showing the second color of

actin added) or a poisoning of the motor protein complex preventing intercalation of new actin

(only first color of actin appearing in rocket tail). Figure 3-5 is an actin rocket tail exposed to

3 VM of actin for 5 minutes of rhodamine labeled actin and then 5 minutes of Oregon-green

labeled actin with a measured velocity of 2.5 ptm/min for the Oregon-green portion and

3.5 ptm/min for the rhodamine portion. Figure 3-6 shows vesicles exposed to 5 [tM of actin for

3 minutes of rhodamine labeled actin, 2 minutes of Oregon-green labeled actin, and 3 minutes of

rhodamine labeled actin. The velocities in the green portion are 3.5 atm/min, 3.1 Pm/min, and

3.3 jpm/min. The velocities of the first rhodamine section (farthest from the vesicle) are

3.5 jpm/min, 3.6 jpm/min, and 3.7 jpm/min. The average velocity of 41 actin rocket tails

(5 1 atM actin) propelling vesicles taken from 8 experiments was 2.8 1 pm/min with a









velocity range of 1.2 to 5.2 km/min (Figure 3-7) and the average radius of the analyzed vesicles

was 0.4 + 0.1 [m (Figure 3-8).

These findings are in agreement with the reaction-diffusion model proposed by Dickinson

and Purich (43). In this model, vesicle speed is determined by rate of monomer addition to

filament ends at the tails center. This rate is determined by the net rate of monomer addition,

which is limited by monomer diffusion (with characteristic diffusion length taken as the particle

radius, R). Accounting for reaction and diffusion as rate-limiting processes in series, the vesicle

speed V can be approximated by

kCd
V (3-5)
1+kfpR/D

where kf is the monomer on-rate constant (10 kM-s-1= 0.017 km3/s), C is the bulk monomer

concentration, d = 2.7 nm is the added filament length per monomer, p is the filament end

density (-103 mn-2) at the vesicle surface, and D is the monomer diffusivity in the tail

(-5 km2/s). Parameter values are justified in Dickinson and Purich (43). The measured speeds

are plotted against particle size in Figure 3-8 and compared to the predicted speed from

Equation 3-5. Although the scatter in the data obscures any dependence on particle size, the

measured speeds do agree well with the predictions.

No saltatory motion or tear drop shapes were observed in the analyzed vesicle velocities.

Although, vesicles were too small to observe possible shape conformations, saltatory motion

would be apparent in the rocket tail by fluorescent fluctuations in the rocket tail if deformations

to the vesicle had happened. This lack of saltatory motion for small vesicles is consistent with

the findings of Plastino et al. (71), who found saltatory motion of motile particles only for

diameters exceeding about 3 microns. This transition was explained by Dickinson & Purich (43)









as the result of more-uniform filament elongation rates on the smaller due to smaller diffusion

gradients.

Most vesicles observed were either too small to observe fluctuations in vesicle shape or too

rare to locate larger vesicles in time to capture time-lapse. Therefore larger vesicles with a shape

change from actin filaments were observed after polymerization had come to a stop, although

only a handful of these events were observed. Figure 3-10 shows vesicles with actin tails that

have variations in fluorescent intensity along the tail length. The brighter portions (in the shape

of a tear drop) in the rocket tails are from an accumulation of actin filaments. For fluorescent

fluctuations to occur, the vesicle forward motion must lag while filaments continue to

polymerize.

Larger vesicles (>3 [tm) were successfully prepared (Figure 3-11), but there was

insufficient time for enough experiments to allow detailed analysis of motility. Of the large

motile vesicles created, some vesicles were observed to be distended as if in transition to a tear

drop shape or conformation change.

Discussion

As argued in Dickinson & Purich (43), the observations of saltatory motion in vesicles

arise naturally from the actoclampin model which proposes filaments are persistently attached to

the surface. Filaments along the periphery have a consistent supply of actin monomers whereas

interior filaments must wait for monomers to diffuse through the rocket tail (without being

incorporated into other filaments) before elongation can occur. Monomer diffusion creates an

actin monomer gradient with a lower concentration of monomers available to the interior

filaments as compared to the exterior filaments. Lagging interior filaments prevent the rear of

the vesicle from moving forward while exterior filaments continue to push the rest of the vesicle









forward creating the tear drop shape which is a result of diffusion-limited filament elongation

instead of stress-dependent filament elongation (58).

Although surface forces could not be determined in our experiments, the velocity of

vesicles could be measured by determining rocket tail growth from fluorescent color change.

The average velocity of vesicles was 2.8 1 tm/min with a velocity range of 1.2 to 5.2 [tm/min

(Figure 3-7) with a noted higher velocity associated with the rhodamine portion of the actin

rocket tail (Figure 3-5 and Figure 3-6). This velocity difference could be due to the actin

monomer (labeled with rhodamine) being more competent or possibly actin labeled with

different fluorophores incorporated into actin filaments at different rates. The vesicle velocity

versus vesicle radius (Figure 3-8) agrees with theoretical predictions (43), as did the lack of

saltatory motion of these vesicles smaller than 1 to 2 [m in diameter. These smaller vesicles are

reaction-limited and not diffusion-limited, which would create the characteristic tear drop shape

observed for large vesicles. Giardini et al. (41) reported an average velocity of 3 tm/min

ranging from 0.6 to 4.2 [tm/min while Upadhyaya et al. (40) reported 0.8 0.2 [tm/min. While

the results of this study are more in line with Giardini et al., Upadhyaya et al. could have

reported a slower velocity because they observed larger vesicles or because their motility assay

was not as potent thereby reducing the actin polymerization rates. Other groups have reported

velocities of actin propelled beads (1 to 6 am/min) (33) and actin propelled Listeria

(3.0 + 0.2 am/min) (59), similar to the velocities we found for vesicle motility. This similarity of

actin propelled velocities among different types of particles of different size suggests that actin

polymerization depends primarily on the actin concentration and not on the type of cargo being

propelled.









Table 3-1. Necessary ratios for a parallel dilution to perform a Bradford assay using a stock of
0.2 mg/mL BSA.
Concentration Volume Volume
(pg/mL) BSA (pl) H20 (dl)
0 0 1000
1 5 995
5 25 975
10 50 950
15 75 925
20 100 900
25 125 875


Table 3-2. List of buffer volumes based on 1 kg of rabbit muscle.
Buffer Volume (L)
Extraction 5
NaHCO3 4
Tris 8
Acetone 12
H20 40


Table 3-3. Reference table for amount (g) of ammonium sulfate ((NH4)2S04) to add at 4C.
0% 26.3 53.4 81.2 109.9 139.5 170.0 201.4 233.8 267.3 301.8 337.5
5% 26.7 54.1 82.4 111.6 141.6 172.6 204.6 237.6 271.6 306.8
10% 27.0 54.9 83.7 113.3 143.8 175.3 207.9 241.4 276.1
15% 27.5 55.8 85.0 115.1 146.1 178.2 211.2 245.4
20% 27.9 56.6 86.3 116.9 148.5 181.1 214.7
25% 28.3 57.5 87.6 118.8 150.9 184.1
30% 28.7 58.4 89.1 120.7 153.4
35% 29.2 59.4 90.5 122.7
40% 29.7 60.3 92.0
45% 30.1 61.3
50% 30.6
55%









Table 3-4. Differences in experimental procedure for vesicle motility.
Giardini et al. (41) Upadhyaya et al. (40)


Extract used
Motile vesicle size


Rehydrating buffer


ActA labeling
Vesicle sizing
Lipid molar %, MW:
Phosphatidylcholine
Chloesterol
(Fluorescein or Oregon Green)-
phosphatidylethanolamine
Ni-NTA


Xenopus laevis egg
cytoplasm
1 to 5 im
10 mM Tris, pH 7.5
100 mM
NaCl/13%(w/w)
sucrose
Alexa 488
succinimidyl-ester
5 cycle freeze-thaw

46%, 760.09
50%, 386.66
2%, 744.05
2%, 1057.02


Bovine brain
1 to 5 rm

20 mM Hepes, pH 7.4
100 mM KC1/1 mM MgC12


Fluorescein maleimide
1 min. @ 1000xg centrifugation

90%, 760.09
0%
0.5%, 1086.25
10%, 1057.02


Table 3-5. ActA conjugation with vesicles.
[ActA] Dilution of A] AA ActA
[NTA] Vesicles (Gl)
0.02 0.5 3.25 0.065 1
0.02 0.2 1.30 0.026 1
0.02 0.1 0.65 0.013 1
0.10 0.5 3.25 0.325 1
0.10 0.2 1.30 0.130 1
0.10 0.1 0.65 0.065 1


Total
(pl)
107.7
269.2
538.5
21.5
53.8
107.7


Vesicle
(Gl)
53.8
53.8
53.8
10.8
10.8
10.8


Rehydrating
Buffer ([l)
52.8
214.4
483.6
9.8
42.1
95.9

























Figure 3-1. Unilamellar bilipid layer vesicle with a portion expanded to show the configuration
of lipids.


Figure 3-2.


Vesicle conformation change. A) Vesicle in solution with no external forces and
membrane tension balanced with osmotic pressure. B) Vesicle deformed by actin
rocket tail. C) Hypothetical forces on a vesicle surface from actin polymerization
(40, 41). Red arrows designate a compressive force, blue arrows designate an
extending force.


































Figure 3-3. Vesicles emanating from a central vesicle mass.


Figure 3-4. Vesicles exposed to rhodamine actin and then Oregon-green actin.





















Figure 3-5. Color change experiment with vesicles exposed to 5 minutes of rhodamine labeled
actin and then 5 minutes of Oregon-green labeled actin (3 aM actin).


Figure 3-6.


Color change experiment with vesicles exposed to 3 minutes of rhodamine labeled
actin, then 2 minutes of Oregon-green labeled actin, then 3 minutes of rhodamine
labeled actin (5 aM actin).


2
0
0-1


1-2


2-3 3-4


4-5


5-6


Vesicle Velocity (pm/min)

Figure 3-7. Histogram of vesicle velocities.










9
8

E

S5
6

0
-34

>1


0
0 0.2 0.4 0.6 0.8
Vesicle Radius (pm)

Figure 3-8. Vesicle velocities versus vesicle radius with data (circles) and the theoretical
velocity maximum (line (5 itm)) (43).


0.5 1 1.5 2


Particle Radius (pm)
Figure 3-9. Theoretical particle velocity at different actin concentrations (43).
























Figure 3-10. Overlay of rhodamine labeled actin propelling an Oregon-green labeled vesicle.
The red spheres are points of accumulating actin due to lagging of the trailing
vesicle edge.


Figure 3-11. Large unlabeled vesicles conjugated with green fluorescent ActA propelled by
rhodamine labeled actin.









CHAPTER 4
ELECTRON MICROSCOPY OF ACTIN FILAMENTS

Introduction

A key distinction between the actoclampin model and free filament models for actin

polymerization and force generation is whether elongating filament ends are tethered to the

motile surface. Several groups have analyzed actin rocket tails propelling beads (32, 33, 35, 36,

80) but have not reduced the number of actin filaments to show single filament interaction with a

motile surface. We show that by optimizing bead size and biochemical conditions, we can

observe single actin filament (+)-ends binding and staying persistently attached (throughout

several washes and treatment) to the surface of a bead under an electron microscope (EM).

An electron microscope functions the same as optical microscopes except EM uses a beam

of electrons instead of light. The electron beam is created by a cathode (tungsten filament) and

accelerated by a positive electrical potential through a vacuum (which reduces beam scatter) and

is then focused on the sample by electromagnets (magnetic lens) (81). In transmission electron

microscopy (TEM), the electron beam is scattered as it passes through the sample and the signal

is collected, resulting in an image. With scanning electron microscopy (SEM), the beam is

scattered by the sample as it is scanned across the sample surface, but instead of passing through,

it is reflected and the signal is collected. If the sample has a high atomic weight (such as a

metal) then details can be resolved down to the Angstrom level (82). A biomaterial sample will

not have a high atomic weight in comparison to metal and will not scatter enough of the electron

beam to produce an image. Therefore, the sample must be coated with a metal to increase the

amount of electron scatter. Because the sample is blanketed with a metal coating, some fine

details are concealed even though resolution of 5 nm is achieved. With the resolution of this









technique, single filaments can be distinguished from bundles of filaments and the orientation of

filament to bead was determined.

Cameron et al. (33) observed actin propelled polystyrene beads by EM under conditions

that reduced the number of filaments in contact with the bead surface to as few as 5 filaments

appearing bound to the particle surface by their (+)-ends, and even single-filament tails on

50-nm beads. These findings are difficult to explain by the Brownian ratchet model (45), so

Cameron et al. (33) suggested filaments might be adding monomers between transient tight

associations with the surface. Consistent with this explanation, they reported that many of the

beads were devoid of filaments and many branched filaments lacked beads suggesting filaments

nucleated at the beads surface but were subsequently detached. An alternative explanation is the

filaments elongated for some period of time before detaching spontaneously.

The observed detachment in the Cameron et al. experiments may be explained by the

method by which beads were functionalized and treated. First, ActA was ionically bound to the

bead surface which is much weaker than a covalent bond. ActA could dissociate from the bead

surface or some beads may not end up with a functional ActA on the surface at all. Second,

beads were not bound to the substrate so many of the single filaments attached to beads could

have been washed out in EM treatment.

We have performed a systematic EM study to determine whether single filament elongate

processively from 50-nm ActA-coated beads. ActA was covalently bound to the bead surface

allowing for a much stronger surface to protein bond. Beads were bound to the glass surface to

prevent wash out of sample during EM treatment. Silica beads were used for a stronger covalent

bond to ActA (silanation) and a reliable method to purify beads from excess ActA and bind

beads to a glass substrate (silica density is greater than polystyrene).









Materials and Methods


Functionalized 500 nm Beads

Five-hundred-nm amino modified polystyrene beads (Polysciences cat#07763) were

diluted with Mops buffer (MB 100 mM Mops pH 7.0 with KOH). The diluted particles were

centrifuged with a tabletop centrifuge at 14,000 rpm at 40C for 5 minutes to pellet. The

supernatant was discarded and beads were resuspend in MB and centrifuged as before.

Twenty mM bis(sulfosuccinimidyl suberate) (BS3) was prepared in MB during centrifugation

just before use. The supernatant was discarded and beads were resuspended in MB

supplemented to 2 mM BS3 and incubated at room temperature with shaking for 15 minutes.

The beads were centrifuged as before. ActA was diluted to about 200 to 250 pg/ml in MB

to an equivalent volume used when charging the particles. Beads were resuspend in prepared

ActA and incubated one hour at room temperature with shaking. The beads were centrifuged as

before. The supernatant was removed and the absorbance was measured to determine the

amount of remaining protein in solution. Beads were resuspended in glycine methyl ester

(GME) in MB (MB with 100 mM GME) then centrifuged as before and aspirated. The GME

washing step was repeated one more time and then again using only MB. Beads were then

resuspended in bovine brain extract buffer (BBEB 20 mM Hepes pH 7.5, 1 mM MgCl2, 1 mM

EGTA, 100 mM KC1, 0.2 mM CaC12, and 150 mM sucrose). Beads were stored in a sealed tube

with parafilm at 40C and are typically functional for one to two weeks.

Preparation of Listeria monocytogenes Overexpressing ActA

Listeria was prepared as described in Chapter 3. Listeria growth was stopped once in the

log phase growth and was washed twice with phosphate buffer saline (PBS). Listeria was killed

with a 20 minute incubation in 10 mM iodoacetic acid (83). Listeria was washed three times

with PBS and diluted to 1/5 the original volume in BBEB (84) and stored at -700C.









Functionalized 50 nm Beads

A 95% ethanol solution was prepared and the pH adjusted to 4.5 to 5.5 using acetic acid

(CH3COOH). Two percent 3-aminopropyltriethoxysilane (APES) in 95% ethanol solution was

prepared and mixed for 15 minutes. Beads (Polysciences cat#24040 stock is at 5.63% (v/v))

were then diluted 10x with 95% ethanol. Beads were washed two times with 95% ethanol and

pulse sonicated between washes (30 pulses at 20% level 1 power). The beads were then

aspirated and the APES solution was added and mixed for 3 minutes. The functionalized beads

were washed two times with 95% ethanol and pulse sonicated between washes (30 pulses at 20%

level 1 power). The supernatant was removed and beads allowed to cure overnight at room

temperature in humidity <60% (alternative is to heat to 1000C for 30-60 minutes to cure). Beads

were then resuspended in water, pulse sonicated (30 pulses at 20% on level 1 power), and stored

at 40C.

ActA and BSA Conjugated to 50 nm Beads

The beads surface was activated with 0.2% glutaraldehyde and incubated at room

temperature for 30 minutes while shaking. The beads were then washed in BBEB and pulse

sonicated (30 pulses at 20% level 1 power). The amount of protein to add to the beads was

determined by first calculating the surface area of beads by taking the product of the volume of

stock beads and surface area of one bead by the volume of one bead (0.338 m2 for 50 pL of

5.63% (v/v) stock beads). The mass/area of ActA (0.905 mg/m2) was then determined by taking

the molecular weight of ActA (68,000 g/mol for ActA (85) and 66,000 g/mol for BSA (86))

divided by the circumference of the Stokes radius of ActA (6.3 nm for ActA (85) and 3.5 nm for

BSA (86)). The product of bead area and protein mass/area resulted in the theoretical amount of

protein to coat the beads. Protein was added to the beads and incubated for 45 minutes at room

temperature while shaking. The beads were then washed in BBEB (the supernatant was saved to









test the amount of protein not bound) and pulse sonicated. If beads were to be fluorescently

labeled, the fluorescent BSA was calculated and conjugated to the bead at this point. Beads were

then washed 2 times in 100 mM GME in BBEB and pulse sonicated between washes. Beads

were washed once more in BBEB and stored at 40C.

Flow Chamber for Exchange Experiments

Flow chambers were made using parafilm or double sided tape arranged in two parallel

lines as support for a cover slip on top of a microscope slide. A pipette was used to introduce

material to the flow chamber while filter paper was used to extract the fluid creating the flow.

To bind beads or Listeria in the flow chamber, the cover slip was first treated with APES and

glutaraldehyde vesicless will adsorb to the glass surface and no glass treatment is necessary).

Three-aminopropyltriethoxysilane treated cover slips were made by exposing clean cover slips to

a 2% APES in 95% ethanol solution (APES solution was mixed for 15 minutes prior) for

2 minutes. Cover slips were then washed in a porcelain tray 2 times with 95% ethanol in a

beaker. Excess liquid was removed and cover slips incubated overnight at room temperature

unsealed for curing (humidity <60%).

Beads / Listeria with Actin Rocket Tails

An APES cover slip was charged with 2% glutaraldehyde for 20 minutes at room

temperature. The cover slip was washed with water, a flow chamber prepared, and diluted ActA

beads/Listeria were flowed through (an appropriate bead/Listeria cover slip surface coverage

was determined by testing different dilutions) and incubated at room temperature for 20 minutes.

Excess binding sites were quenched with BBEB containing 100 mM GME for 20 minutes. The

chamber was washed with BBEB and a motility assay was added (Chapter 3) for the desired

amount of polymerization time.









Preparation of Sample for Viewing with EM

Two percent glutaraldehyde in 0.1 M sodium cacodylate buffer was added to the flow

chamber for 20 minutes. A 0.1% tannic acid was then added for 20 minutes and rinsed with

3 exchanges of water. Uranyl acetate (UA) at 0.2% was added for 20 minutes (87). The entire

slide was submerged in water and the cover slip removed. A metal cage with lens paper at the

bottom was submerged in the water and the cover slip was placed inside the cage sample side up

(never exposed to air). Several samples were stacked in the cage separated by lens paper

(including lens paper to cover the top cover slip). The cage containing samples was transferred

to a 10% ethanol solution and incubated with mixing for 5 minutes. The cage was then

transferred in series to 20%, 40%, 60%, 80%, and 2 times in 100% for 5 minutes each, then

0.2% UA in 100% ethanol for 20 minutes, and finally 2 times in molecular sieved 100% ethanol

for 5 minutes each (87). At this point the samples were optionally placed at 40C overnight.

Critical point dryer (CPD)

The CPD was filled with molecular sieved 100% ethanol and the cage containing the

samples was placed in the CPD and sealed. Liquid CO2 was added until all ethanol was removed

from the chamber. The temperature and pressure was increased to the critical point of CO2. CO2

gas was removed until completely evacuated. The sample was removed and either coated for

SEM or TEM or placed in a desiccator until coating was done.

Sputter coating for SEM

The cover slip with sample was placed on a 13 mm SEM stub using double sided tape to

affix (sample side up). The sample was then placed in the sputter coated and filled with argon

gas. The pressure was then reduced and the sample was sputter coated with gold palladium at

45 milliamps for 30 seconds. The sample was then imaged using SEM.









Rotary shadow for TEM

Samples were placed in a rotary shadower at an angle of 450 for platinum coating. The

chamber was evacuated and the sample was coated with carbon platinum with a setting of 2.1 kV

and 60 mA. Samples were coated with a 5 nm thickness of platinum which was monitored using

a quartz thickness monitor. The sample angle was adjusted to 900 for carbon coating with a

setting of 1.9 kV and 90 mA. The sample was coated for 8 seconds resulting in approximately a

5 nm coating of carbon (to prevent sample from breaking apart).

Post shadowing / pre TEM treatment

A razor blade was used to cut 0.5 mm2 sections from edge to edge across the entire surface

of the metal coated sample. Areas that did not contain sample were colored with a black marker.

The cover slip (sample side up) was floated on 10% hydrofluoric acid (HF) until the glass cover

slip sank (about 15 minutes). A platinum loop was used to transfer the floating pieces to water.

Again the platinum loop was used to transfer pieces to a 100% solution of bleach (6.13% sodium

hypochlorite (NaOC1)) and incubated for one hour. Pieces were transferred to water and then

placed on polyvinyl formal coated 200 mesh TEM grids. Grid containing samples were dried

overnight at room temperature and later viewed with TEM with a voltage of 75 kV.

Polyvinyl Formal Coated TEM Copper Grids

Polyvinyl formal coated copper 200 mesh grids were used as support for samples in TEM

imaging. To coat grids with polyvinyl formal, formvar (0.25% to 0.4% (w/v)) was mixed with

chloroform in a glass container large enough for a microscope slide to be submersed. A glass

microscope slide was cleaned using 70% ethanol dried with a Kim wipe. The slide was coated

with a household cleaning solution and polished with lens paper to prevent the plastic from

sticking strongly to the glass surface when trying to separate in water. The glass slide was

dipped into the polyvinyl formal and allowed to incubate for a few seconds. The slide was then









removed at a constant velocity to give a smooth uniform coating. The faster the slide is removed

the thicker the film will be on the slide. Thicker layers were formed by dipping the slide several

times (allowing the chloroform to evaporate between dips). A sharp metal edge was used (such

as the edge of tweezers or a razor blade) to scrape the edges of the glass slide and to cut across

the middle of the slide so the film will separate. At approximately a 300 to 450 angle to the water

the slide was slowly submersed causing the film to separate from the slide and float away. Grids

were placed on the film (rough side up, shiny side down) avoiding any discontinuities in the film

and avoiding film edges. A note card or parafilm cut to the size of the film was placed on top

and in a sweeping motion the card was dunked and rotated up to pull the grids and film out of the

water on top of the card or parafilm which was then dried for 24 hours.

Filament Shearing from Fluid Flow

Based on EM images, filaments are obviously affected by the treatment for TEM because

filaments tend to line up in the same direction (Figure 4-7). To determine if the treatment

velocity is enough to shear a filament from the bead a simple Poiseuille flow was used to

determine the amount of shear exerted on the filament. The velocity of flow through the

chamber was solved

u =-(2 -Hz) (4-1)
Y 2,u

where /f is the pressure drop, u is the viscosity of the fluid (assumed to be as viscous as water,

0.01 g/cm-s), z is the position in the flow chamber, and H is the height of the flow chamber

(-200 [tm). To estimate the pressure drop, the average velocity through the flow chamber is

approximated to be 0.25 cm/s based on a standard flow chamber with a volume of 15 atL, flow

time of 10 s, chamber width of 3 mm, and chamber height of 200 [am. Taking the average of









Equation 4-7, the pressure drop is calculated to be 7.5 Pa/cm. With a basic understanding of the

flow chamber, the drag on the filament from the fluid motion is calculated (88)


Fd 4= M (4-2)
Sln(2A)

where L is the filament length (10 tm), A is the aspect ratio of the filament (length/diameter),

M' is the resistance matrix, and Fd is the drag force resulting in a drag force on the filament of

30 fN. This small amount of force may be enough to align filaments but is not enough to

separate an attached filament which can withstand at least 8 times this amount of extension force

(40).

Results

As a positive control for motility from substratum-bound particles, Listeria was bound to a

glass surface, exposed to a motility assay, and further treated for viewing with TEM (Figure 4-1

and Figure 4-2). Approximately 10% of Listeria observed using TEM grew actin rocket tails,

which is similar to what is observed with bound Listeria using light microscopy. Some Listeria

broke free of the glass slide and grew long twisting rocket tails (Figure 4-1). A biomimetic

system using ActA-coated 500-nm polystyrene beads was developed to emulate Listeria

actin-based motility. The 500-nm polystyrene beads were attached to a glass surface and

exposed to a motility assay. The biomimetic particles produced actin rocket tails similar to the

Listeria rocket tails (Figure 4-3 to Figure 4-6). These large particles produced too many

filaments to observe single filament interactions with the surface. The number of filaments was

reduced by reducing the bead size from 500 nm to 200 nm which produced beads with fewer

filaments but still too many to give conclusive evidence of filament attachment. The particles

were reduced again from 200 nm to 50 nm which resulted in only a few filaments associated

with each bead.









Filament density was also controlled by adjusting the ActA surface density, the extract

concentration, and the actin polymerization incubation time. The optimal ActA density for

observing single filaments corresponded to approximately 13 ActA proteins per bead. The

extract was used at full strength and actin concentrations in the range of 3 tM to 5 tM produced

single actin filaments quickly (less than one minute) (Figure 4-7 to Figure 4-9). To further

control the number of filaments, the actin polymerization incubation time was adjusted. To

determine the incubation time necessary to achieve one filament per bead, a time study was done

by incubating the beads with a motility assay for 1 minute, 2 minutes, 5 minutes, 10 minutes,

20 minutes, and 35 minutes. Due to the large number of filaments observed on the beads at time

points 20 minutes and 35 minutes, only time points 1 minute thru 10 minutes were analyzed.

To prevent human bias, 25 random images at each time point were taken. For each image,

number of particles, filament length, and number of filaments per particle were measured. The

average filament length was not statistically different between the four time points. The

histogram of filament lengths is shown in Figure 4-10. For filament lengths larger than

-100 nm, the distribution appears to be exponential which is consistent with a Poisson

distribution of elongation times. The lower counts of shorter filaments less than 100 nm suggest

that filaments grow for several subunits before stopping. At each time point the average filament

length is -500 nm, but based on velocities we have determined, a length of 500 nm would

correspond to 10 seconds of growth. Velocities determined for vesicles represent the growth rate

of an ensemble of filaments where some filaments could be lagging, slowing the overall growth

rate so single filaments could possibly be elongating at even faster rates. Therefore we can

surmise that the majority of filaments have stopped growing by the time of fixation consistent

with the reported histogram (Figure 4-10).









The average filament number per beads with filaments increased over time, due to new

filaments initiating and polymerizing from the surface (Figure 4-11). To quantify the apparent

increase of filaments, the number of filaments was normalized to the number of beads which

resulted in a steady increase of filaments per bead (Figure 4-12). Next we quantified free

filaments and found on average 7% of observed filaments were not bound to any bead

(Figure 4-12). The number of free filaments decreased over time which could be due to fewer

binding sites available for free filaments to attach. The lack of binding sites could arise from

excess protein or extract sticking to the substrate preventing filament attachment. The histogram

in Figure 4-13 shows the number of filaments per bead increases due to longer incubation times.

To confirm ActA coated beads are initiating filament growth, the same motility experiment was

tested replacing ActA coated beads with APES/glutaraldehyde treated beads or BSA treated

beads. No filaments were observed in either case on beads or on the substrate surface. Filament

polarity determined by myosin subfragment-1 (S1) was attempted but a conclusive result was not

found (the attempted protocols and outcomes are presented in Appendix B).

Discussion

Analysis of several hundred images and several thousand single filaments attached to

beads shows a persistent attachment of filaments to particles. These filaments reached a

steady-state polymerization rate within a few minutes. The observed filament lengths ranged

from 100 nm up to 4 km. In order for a filament to polymerize to a length of 4 [m at a growth

rate of about 3 km/min (Chapter 3), the filament must be associated with the bead for 1,400

monomer additions over more than a minute time period. The distribution in filament lengths,

with the exception of fewer numbers of shorter filaments, can be explained by filaments

nucleating and ceasing elongation at random times with constant rates. The fewer shorter

filaments suggests filaments grow for many cycles before stopping. Shorter filaments either









nucleated at a later time in the incubation period, or possibly the motor protein (ActA-VASP)

was phosphorylated (89) preventing continued filament elongation, or some other unknown

cause stopped polymerization. Even though filaments have stopped elongating after a relatively

short period, (on average -10 s before elongation has stopped) the filaments are still strongly

associated with the bead surface even after the several washes and treatments for EM. Single

filaments emanating from 50 nm beads before EM treatment and fixation are observed using

total internal fluorescence microscopy (TIRF) (Figure 4-14) suggesting filaments are not being

detached from the beads due to further treatments.

The observation that filaments stay persistently bound to a bead surface after

polymerization has ceased sheds light on what could be happening in an ensemble of filaments in

an actin rocket tail. In an assay of 500 nm particles (polystyrene beads or vesicles such as

Figure 3-4) with actin rocket tails, some particles have long rocket tails while others have short

dilated rocket tails. These shorter rocket tails could arise from a few quiescent filaments that

prevent the forward motion of the particle. Once a few filaments have stopped polymerizing but

are still persistently bound to the particle surface, other active filaments may continue to

polymerize until the filaments become so compressed that additional monomers are not able to

intercalate. The dilated short tails could be due to the continued polymerization of filaments

while other seized filaments prevent forward motion.

These findings are direct evidence that filaments polymerize while persistently attached to

a motile surface and cannot be accounted for by a free-ended filament model, in which case the

filament end would quickly diffuse away from the bead before this length was achieved. The

number of filament-particle associations would become almost impossible to encounter. On

average, 93% of observed filaments had ends associated with a bead. If filaments were not









polymerized at the surface of the bead but were instead polymerized in solution and later

associating with the bead surface then there would be a much higher incidence of filaments not

associated with a bead.

At longer polymerization times, several filaments were observed associated with a single

bead, although some beads had no filaments, which could be due to the functionality and

orientation of ActA on the bead surface. Unaltered ActA produced by Listeria has a

transmembrane portion that orients ActA in the correct position on the surface of the bacteria

(16, 18, 19). In this study, ActA was randomly covalently bound to the bead surface which

could prevent functionality and prevent filament polymerization if ActA is not oriented correctly

on the bead surface. Beads with no filaments could have either no ActA bound or have ActA

bound that is not in a functional orientation. The increase in filament number with time

demonstrates that filaments continue to nucleate over time (Figure 4-13). Cameron et al. (33)

were unable to observe several instances of single filaments attached to beads most likely due to

their approach of binding ActA to the bead surface and not having a method to prevent wash out

of sample from EM treatment. This study, however, has shown thousands of single filaments

associated with beads, which supports the actoclampin model and persistent filament attachment.
































Figure 4-1. Listeria (top right of image) freed from the substrate surface (possibly by the actin
rocket tail) is propelled by an actin rocket tail jutting from a larger actin network.


Figure 4-2. Listeria with a small actin comet tail in its early stages.

































Figure 4-3. A 500 nm polystyrene bead with an actin cloud viewed using TEM.


Figure 4-4. A 500 nm polystyrene bead with an actin rocket tail viewed using TEM.

































Figure 4-5. Two actin rocket tails converge.


Figure 4-6. Three 500 nm polystyrene beads combine to form one actin rocket tail.
































Figure 4-7. Single actin filaments emanating from 50 nm silica beads.


Figure 4-8. A single actin filament associated with a single 50 nm silica bead.































Figure 4-9. Several single actin filaments associated with single 50 nm silica beads. The scale
bar is 300 nm.


60 1 min. B 2 min.
50
u 40
a,
30O





0o lC

50 -- min--------

40 -
S3

-20 --

10 -



0 0000 2 'IF M M I 0 0 0 0 00
So 10 CD C CDC o o Co oDC
CM 0-3--:t-I--CD


Length (nm} Length (nm

Figure 4-10. Histogram of filament lengths with bin size of 100 nm. A) One minute actin
polymerization with 66 filaments. B) Two minute actin polymerization with
117 filaments. C) Five minute actin polymerization with 180 filaments (one outlier
of 4.7 [m not shown on graph). D) Ten minute actin polymerization with 215
filaments. Graphs A, B, and D have the same axes values as graph C.
filaments. Graphs A, B, and D have the same axes values as graph C.










U_ -I~ t - - - - 4 - - - -
m 1.4
1- I - -
S1.2 -..
E

z 0.8
S0.6 -
S0.4 -
S0.2 ----
0 -------I T -II--------------------------
0
0 2 4 6 8 10
Time (min)
Figure 4-11. Filament number per bead versus time. Blue triangles represent the average
number of filaments per beads with filaments and black squares represent total
number of filaments per total number of beads.


50.00%
45.00%
40.00%
35.00%
30.00%
25.00%
20.00%
15.00%
10.00%
5.00%
0.00%


Time (min)
Figure 4-12. Number of filaments normalized to the total number of beads versus time. Blue
diamonds represent beads with filaments per total bead count and black triangles
represent free filaments per total number of filaments.


i-------------
I -- - -- - -












* 1 minute
* 2 minute
* 5 minute
* 10 minute


120


100
C-)
80
o
w

- 60
U-
L.
40


20


0


1 2 3 4
Number of filaments attached to a bead

Figure 4-13. Histogram of the number of filaments attached to a bead versus time.


1 --















































Figure 4-14. Time lapse of single actin filaments emanating from 50 nm beads. Beads are
labeled with rhodamine and actin is labeled with Oregon-green. Motility assay
started and focused after 4 minutes. Images taken at 20 second intervals.













88









CHAPTER 5
SINGLE ACTIN FILAMENT POLARIZATION DETERMINED BY MULTIPLE LABELED
ACTIN MONOMERS INCORPORATED INTO ACTIN FILAMENTS

Introduction

As shown with EM images, single filaments emanating from 50 nm ActA-coated particles

could be observed and characterized. Although EM provides nanometer details, the system is

not dynamic and polarity of the filaments can not be easily determined. Determining filament

polarity is important to establish whether filament (+)-ends are attached and elongating by

insertional polymerization at the bead surface, as would be expected in an end tracking motor

such as ActA-VASP, or instead (-)-ends are attached and filaments are growing with free ends

away from the beads. For example, Brown and Spudich (90) showed polylysine coated

polystyrene beads with the majority of (+)-ends of filaments pointing away from the bead

surface. In their experiments, polylysine nucleated actin filaments and caused an increase in

polymerization but does not orient the filament at the bead surface.

Our strategy for determining at which end actin is adding onto tethered filaments was to

expose the bound ActA-functionalized beads to a cell extract containing fluorescent actin of one

color for a few minutes, then switch to an extract containing actin of another color for a few

minutes. The resulting filaments were then observed under total internal reflection fluorescence

(TIRF) microscopy, which has been successfully applied previously to single filament dynamics

(54, 74) as well as insertional polymerization by substratum-bound formins (91, 92). Similarly,

motor proteins actinn and myosin) were visualized using a two-color assay with TIRF (93) and

the recruitment and dynamics of various components of actin polymerization (including

N-WASP and Arp2/3) was studied using a two-color assay visualized with TIRF (94).

Figure 5-1 illustrates some possible outcomes of a two-color single actin filament

experiment, here assuming green actin is added first followed by red actin. If the elongating









(+)-end is located at the bead surface, a span of red F-actin would appear next to the bead

adjacent to green span farther from the bead (Figure 5-1A). On the other hand, if the (-)-end is

bound to the bead surface, the colors would be reversed with the green span bound to the bead

(Figure 5-1B). A similar result would appear if the filament were initially side-bound to the bead

(Figure 5-1C), in which case green filament spans would appear on both sides of the bead and

red adjacent to one span. It is also possible that the filament both nucleates and stops elongating

while exposed to one color or the other, thereby producing a filament of uniform color

(Figure 5-1D). Finally, two separate filaments could polymerize from the surface of the bead but

overlap and appear indistinguishable as one filament. If the two filaments were of different

colors, the conjoined filaments would appear as a yellow filament adjacent to the bead and the

single-filament span would be either red or green away from the bead, depending on whether the

longer filament grew before or after the color change (Figure 5-1E).

These possible outcomes are assuming negligible (-)-end growth or shrinkage from

ATP-actin or ADP-actin. Pollard et al. (95) determined the polymerization rates of ATP-actin at

the (-)-end to be k = 1.3 iM-ls1 and k- = 0.8 s-1 and polymerization rates of ADP-actin at the

(-)-end to be k+ = 0.16 iM-ls1 and k- = 0.27 s-1. For ADP-actin, the on rate is slightly less than

the off rate causing a slow depolymerization of ADP-actin from the (-)-end, however, so slow

that depolymerization would not be noticeable. The on rate of ATP-actin is slightly more than

the off rate of ATP-actin causing a slow polymerization of ATP-actin at the (-)-end but

polymerization at the (+)-end is 10 times greater than at the (-)-end so considerably more

polymerization occurs at the (+)-end. If (-)-end addition was noticeable, then the insertional

polymerization case (Figure 5-1A) would merely have a short red portion on the end of the tail.

Filament orientation could still be determined by the much longer red portion near the bead









designating (+)-end growth. A three section filament consisting of one fluorescent actin split by

another was not seen in any of the experiments.

All two color filaments observed in this study resulted in the correct orientation for actin to

be insertionally polymerized at the bead surface (Figure 5-1A and Figure 5-5). These results

combined with EM images (Chapter 4) indicate single actin filaments are persistently attached

while actin is insertionally polymerized, consistent with the actoclampin model.

Materials and Methods

Color Change Assay

ActA functionalized 50 nm beads were bound to a flow chamber as described in Chapter 4.

Actin was labeled with either Oregon-green or rhodamine (Chapter 3). A motility assay with

5 [M non-labeled actin (black actin) was flowed through the chamber for 2 minutes. Then, a

motility assay with 5 VM Oregon-green actin (green actin) was flowed through the chamber for

1 to 2 minutes. Next, a motility assay with 5 VM rhodamine actin (red actin) was flowed through

the chamber for 1 to 2 minutes. The order of actin addition to the flow chamber did not prevent

intercalation of actin monomers to produce two color single filaments. Last, 1% glutaraldehyde

was flowed through the chamber to fix the filaments. Samples were viewed within 1 minute of

preparation using TIRF.

Image Analysis

Several macros were created to analyze images using NIH ImageJ software (Appendix C).

To better visualize filaments a macro was created to apply a high pass and low pass filter to

remove low and high frequencies from images. The laser used for TIRF produced a higher

signal in the center of the image than at the edges creating a low frequency background that was

removed by a high pass filter (Figure 5-2 had a high pass filter applied to the image). The high

pass filter removed frequencies larger than 50 pixels. When the setting for the high pass filter









was changed to remove frequencies greater than 50 pixels, the outcome of the image did not

change. Background fluorescence and small particles produced specs of noise in all of the

images. These were removed using a low pass filter which filtered objects smaller than 3 pixels

(Figure 5-3 had a low pass filter applied to the image). The combination of the high and low

pass filters greatly increased the number of filaments that could be observed per field of view

(Figure 5-4 is the combination of low and high pass filters on an image). The brightness and

contrast for each image was optimized to show the filament.

A line scan of each filament was done to show fluorescence intensity. Fluorescence

intensity varied in different regions of the image and varied between the two fluorophores used

in experiments. Therefore fluorescence was normalized for each line scan by taking each

fluorescent value and dividing by the maximum value along the line scan (for each color). This

normalized data was then plotted (Figure 5-6 to Figure 5-9) to show relative intensity along the

filament length and to show the fluorescent actin farthest from the bead was not connected to the

bead other than through the other fluorescent actin. Bead count was done using a macro to

threshold for yellow beads and a circle counter was used to tally yellow bead events

(Figure 5-10) as well as beads were counted by hand for verification. Single filaments were

counted by hand for each image.

Results

There are several advantages to fixing filaments with glutaraldehyde as opposed to viewing

filaments not bound to the substratum. First, several images can be taken while scanning the

sample without worry of depolymerization. Background fluorescence is also removed with the

flow through of glutaraldehyde for fixation (excess actin is flushed out). Second, filaments are

bound to the surface and not fluctuating from Brownian motion providing a better image for

analysis. Photobleaching of fluorescent filaments is prevalent in these experiments so static









filaments allow for the image to be focused and taken once as opposed to the several images

typically necessary with fluctuating filaments. Time lapse of single filaments is difficult because

filaments fade due to photobleaching within 10 to 15 seconds of continuous fluorescence.

Finally, since two color filaments have a low occurrence, several field of views must be taken.

Filaments reach a maximum length quickly (<30 seconds) so live images are not practical for

finding two color filaments.

There were several criteria for counting individual filaments. Filaments had to be

sufficient length (>1 pm), filaments could not overlap with other filaments at critical points such

as color change points or bead-filament attachment, filaments had to have a high signal to noise

ratio and the filament had to be continuous, the position of labeled filamentous actin did not

overlap with other labeled filamentous actin except minimally at the transition of the color

change. These same criteria were used for counting single color filaments (when applicable).

Some of the filaments appear to have variations in the fluorescence intensity. These variations

could be from background autofluorescence of protein from the extract or from the amount of

fluorescence monomer incorporated at any point in the filament. Filaments could also fluctuate

from the surface reducing the fluorescent signal output generated from the evanescent wave of

TIRF.

Figure 5-5 is a compilation of all 158 two color single filaments observed in more than

150 image sets of 12 experiments. Approximately 20 of the 158 events were considered to be of

excellent quality meeting all criteria. Figure 5-6 and Figure 5-7 show a two color filament bound

to a bead where Oregon-green was added first and rhodamine actin followed. Both fluorescent

channels are shown separately and a graph of the line scan along the filament shows the

normalized fluorescence intensity. Figure 5-8 and Figure 5-9 are the same as Figure 5-6 and









Figure 5-7 except with rhodamine actin added first and then Oregon-green actin added second.

The graph of the line scan (Figure 5-6 to Figure 5-9) clearly shows a fluorescence intensity

switch from the bead to the end of the tail. The green portion of the filament is clearly not

connected to the bead by green actin in Figure 5-6 and 5-7 as well as the red portion of the

filament is not connected to the bead by red actin in Figure 5-8 and 5-9. Figure 5-10 is a

histogram of bead and filament count for 20 image sets. Twenty-five percent of beads had single

actin filaments and 11% to 12% of beads counted had single filaments with 8% of the single

filaments having two fluorescent actin monomers incorporated. To test that filaments were being

generated by ActA beads, the same two-color experiments were performed as before except

ActA coated beads were replaced with BSA coated beads. Beads were also completely removed

and a motility assay containing actin was added to the flow chamber as before. No filaments

were observed in either of the controls.

Discussion

Over 158 events show a bead with a two color single actin filament attached. Every event

that exhibited a two color actin filament resulted in the correct polarity of (+)-end closest to the

bead independent of order of fluorescent actin added. Approximately 25% of beads counted had

an actin filament attached compared to 35% observed in EM images at a comparable time of

polymerization (5 minutes). The difference in observed filaments is due to the resolution limit of

TIRF compared to EM. A criteria of filaments being around 1 tm or greater was set for counting

filaments in TIRF. The length criterion of a single filament in EM was much lower because

filaments are easier to distinguish due to the high resolution associated with EM.

Single filaments with only one color actin were -11% of beads counted for either color and

independent of the order of fluorescent addition to samples. Out of the single filaments counted

8% were two color filaments. At first glance these numbers may seem low but can be explained









from the results of EM. From the elongation rate determined, (-3 [tm/min) the average filament

(-500 nm) measured in EM would only elongate for 10 seconds. Few filaments in EM were

observed to elongate for much longer times and distances (max -4 tm) for unknown reasons

phosphorylationn of VASP may not have occurred until a later time or some other poisoning

device had not happened). Assuming the same mechanism is occurring in these TIRF

experiments, filaments grow for 10 to 20 seconds and stop elongating but still remain

persistently attached (two chamber flow through before fixation). The addition of another

fluorescent actin would then not intercalate into the halted filament. Halted filaments would also

explain the low occurrence of two color single filaments. For a two color filament to occur, the

filament must have started elongating in the few seconds remaining before the next fluorescent

actin is added and the filament must continue to intercalate the new actin.

The observation of new actin monomers intercalating at the surface of a bead shows single

actin filaments are persistently attached while actin is insertionally polymerized. Free filament

models do not have an explanation to support this observation. A free filament would easily

diffuse away from a bead surface during the 10 to 20 seconds of actin polymerization observed

in EM and TIRF. Insertional polymerization from ActA-coated beads strongly suggests

elongation by end-tracking motors and supports the actoclampin model for ActA-VASP. Further

experiments are required to establish that VASP (rather than, e.g., adsorbed formin) is the

end-tracking protein responsible for actin assembly in these experiments.







+-


O0


0


+0


-+0


+O


Figure 5-1.


Fluorescent actin change hypothetical scenarios with green added to the assay first,
then red, attached to a yellow 50 nm bead. The + and designate the (+)-end and
(-)-end of the filament. A) Result if insertional actin polymerization is occurring.
B) Result if the actin filament is nucleated in solution and bound to the bead with the
incorrect polarity. C) Filament nucleated in solution and bound to the side of the
bead. D) Single filament growth without incorporation of two colors. E) Overlap
of filaments giving the appearance of A, except the red/green overlap would create a
yellow filament.


Figure 5-2. High pass filter removing frequencies larger than 50 pixels applied to a mock
filament. A) Mock filament with nothing applied. B) Low frequency added to the
image. C) High pass filter applied to image.


0000


00090
















A :. B
Figure 5-3. Low pass filter removing particles smaller than 3 pixels applied to a mock filament.
A) Speckles added to a mock filament image. B) Low pass filter applied to image.


D) Combination of low and high pass filters to A.
D) Combination of low and high pass filters to A.


0000






























gure 5-5. Compilation of 158
than 150 image sets


color change events observed


om 12 experiments and more


-F




Tail tBo 8ad Pa mn n LumJ
Figure 5-6. Both fluorescent channels and overlay of a two color filament where Oregon-green
actin was added to the experiment first. The graph is of a line scan along the length
of the filament. Scale bar = 1 m.























Figure 5-7.


Both fluorescent channels and overlay of a two color filament where Oregon-green
actin was added to the experiment first. The graph is of a line scan along the length
of the filament. Scale bar = l m.


00 0.5 k.0 Is
To 4 Beid Pcrirn omr
Figure 5-8. Both fluorescent channels and overlay of a two color filament where rhodamine
actin was added to the experiment first. The graph is of a line scan along the length
of the filament. Scale bar = l m.









it.t




r= t RIM PO&n* Wcn
Figure 5-9. Both fluorescent channels and overlay of a two color filament where rhodamine
actin was added to the experiment first. The graph is of a line scan along the length
of the filament. Scale bar = l m.











2500


2000


1500


1000


500


0


Figure 5-10. Histogram for beads and filaments of 20 image sets.









CHAPTER 6
CONCLUSIONS AND FUTURE WORK

Discussion

Improved understanding of actin-based motility has several advantages. The most

immediate are a better understanding of how cells work and how the body functions. With this

understanding of actin polymerization, a variety of biological issues can be approached such as

how cancer spreads through the body, or spinal muscular atrophy caused by a deficiency of actin

(96), or better medical techniques and drug delivery in fighting diseases. Less apparent and still

in the distant future are other applications such as biomimetic devices for separating and

detecting a desired particle or the control of nano-devices such as nano-switches or nano-valves.

Although actin has been studied for over 100 years the mechanism of actin polymerization is still

disputed. In this work, experiments and analysis of actin filaments were done to elucidate the

mechanism governing actin polymerization.

The mechanical energy density of filaments in an actin rocket tail was found to be -3 pN

on average (Chapter 2) which is greater than the energy that could be stored by the free energy of

monomer addition. Because free filament models rely on monomer addition to the free filament

end as a means for energy, the maximum energy density is about 2.7 pN. The energy density

measured in Chapter 2 suggests that polymerization yields more energy than could be provided

by monomer addition alone, consistent with the actoclampin mechanism, which also harnesses

the energy of ATP hydrolysis. Energy is also lost through heat dissipation (friction), filaments

depolymerizing, or periphery filaments with no means of translating energy to the rocket tail.

Although the method in Chapter 2 has some limitations, the approach provides a means to

calculate an energy density of filaments that could be applied in other venues involving static

filament images.









Vesicles were found to be propelled by actin rocket tails at -3 itm/min velocities. The

vesicle velocities reported in Chapter 3 are similar to other measured vesicle velocities (41) and

other biomimetic velocities (32, 37, 38, 59, 69) reported in the literature. Vesicle velocities

compared to vesicle diameter agree well with theoretical calculations (43) indicating the small

vesicles observed (<1.5 um) are more reaction-limited than diffusion-limited. Some vesicles

were slower than predicted which could arise from vesicle-to-vesicle variation (e.g. in filament

density) or a poisoning of the ActA-VASP mechanism slowing motility. The velocities of this

study give a basis for how well the components prepared in our laboratory function and how well

the components operate compared to published data. Vesicle shape change could not clearly be

determined from fluorescent images, but evidence of saltatory motion should have appeared in

the actin rocket tail if it were present. Saltatory motion was not observed for the analyzed small

vesicles.

Single filaments were observed in EM to be persistently bound to a bead surface.

Filaments quickly elongated then stopped for unknown reasons, possibly due to phosphorylation

of VASP or some other poisoning of the end-tracking motor. Single filaments continued to stay

attached to the bead surface well after elongation had ceased and after the sample was treated for

viewing with EM. A free filament would either diffuse or be washed away in this scenario. The

actoclampin model is the only actin polymerization model that can explain filament elongation

while staying persistently attached to a surface. The polarization of single filaments was further

explored in Chapter 5. Here filaments incorporated new actin closest to the bead surface

indicating insertional polymerization. The low number of two color filaments observed can be

explained by the fast elongation rates of actin filaments. For a two color event to occur, a

filament must have polymerized in the remaining seconds of the first actin environment and









continue to polymerize with the new batch of actin. Again, the actoclampin model can explain

insertional polymerization of filaments tethered to the particle at their (+)-ends.

The work presented in this study significantly contributes to the knowledge of actin based

motility specifically on how single actin filaments interact with a surface. Further work with the

techniques described in this study will be useful in clarifying how actin polymerization is

capable of producing forces. The results presented here support the actoclampin model and

argue against models requiring free filament ends.

Suggestions for Future Work

Further work could be done in mapping filaments from images obtained with EM. A less

dense actin tail would increase the reliability of the measurements. Because the third-dimension

is inaccessible in EM images, a simulation of an actin rocket tail propelling a bead could be

created. The resulting image could be projected on a two-dimensional surface and the same

method outlined in Chapter 2 performed on the simulated filaments to determine how accurate

the algorithm is at determining an energy density in an actin rocket tail.

Large (>3 [im) actin propelled vesicles were prepared but a lack of time prevented the

study and analysis of the system. Several experiments could be done using large motile vesicles.

The velocity and saltatory motion of large actin propelled vesicles could be further studied in the

same manner performed in Chapter 3. Actin rocket tail forces could be probed by anchoring the

tail of a motile vesicle while the front surface of the vesicle (opposite side of the actin tail) is

aspirated with a micropipette. The subsequent shape change of the vesicle could be measured to

determine the amount of force the actin rocket tail is exerting on the vesicle surface.

Fluorescence recovery after photobleaching (FRAP) could be used to measure residence times of

ActA at the surface of a vesicle. ActA would be fluorescently labeled, attached to a large vesicle

and motility induced. Then FRAP would be used to photobleach ActA and to determine if ActA









concentrations change at the vesicle surface. Phosphoinositides are known to increase actin

polymerization (85, 97-99). The affects of a phosphoinositidyl lipid could be studied and used to

accelerate nucleation time and possibly vesicle velocities. Various cargos could be inserted into

a motile vesicle to determine viability for various applications (e.g. drug delivery or lab on a chip

designs).

Further work could be done with single filaments imaged in EM. Actin filaments were

allowed to polymerize on single beads for up to 35 minutes as described in Chapter 4. Longer

incubation times could reveal if filaments might start polymerizing again or if new filaments will

continue to form. To determine polarity of filaments in EM, myosin subfragment-1 (S1)

(Appendix B) (100) could be used to show directionality of single filaments. Subfragment-1

binds to actin in a specific direction creating an arrow head on the filament with the barbed end

of the arrow pointing toward the (+)-end and the pointed end of the arrow pointed toward the

(-)-end of the filament. Directionality could also be determined by doing the same two-color

method as in Chapter 5, except using regular actin and then biotin actin (or vice-versa). Next,

streptavidin gold particles would be added which will bind to the biotin actin (101, 102) showing

where new actin is intercalating.

Further experiments with the two-color actin method could provide more insight into what

is happening with the filament. If incubation times with each color actin were reduced to

seconds (possibly 10 seconds for each color actin) there might be an increase in the number of

two-color filaments observed because of the velocities determined in Chapter 3. There could

also be a decrease in the number of two-color filaments observed because nucleation time could

play a role in how many filaments polymerize in the given incubation period. Most likely, there

is an optimal time that produces several two-color filaments in one experiment.









Non-hydrolyzable ATP (AMP-PNP) (103-105) could be bound to actin for use in the two-color

experiment. If polymerization was dependent on hydrolysis then no two-color filaments would

appear, however, if hydrolysis did not matter, two-color filaments would be observed. More

time experiments could be done with the two-color experiment. Filaments could be incubated

with one color actin for a long period of time (1 hour) and then the second color actin would be

added to see if filaments are able to polymerize new actin on existing filaments or if new

filaments can form at all.











APPENDIX A
MATLAB ALGORITHM TO DETERMINE ENERGY STORED IN BENT FILAMENTS

% v20 has the lengths^3 instead of just lengths for the X and Y matrices
% 9-17-03
% v22 has the x and y combined into one matrix for solving, also has the
% option of removing the bending portion for the end points
% corrected all endpts (ie A(1,1)=S^-2+5*L/lengths(1)^3 to
% A(1,1)=S^-2+6*L/lengths(1)^3)
% 9-18-03
% v23 put values back in for end points, adjusted values at points next to
% end points 9-22-03
% v24 solved for x and y positions in separate matrices
% S is set to 1 nm
% option to choose filament to evaluate reinstated
% 9-24-03
% v25 mass filament evaluation with option of how many to evaluate at a
% time and the average energy per filament with plots given
% 9-24-03
% v255 same as last one but now plots are with found positions and contours
% 9-25-03
% v26 all of the filaments evaluated with a contour plot of all filaments
% and their dEds on one plot
% 9-25-03
% v261 took derivative at each point instead of fitting a curve to the
% values. Used those results to get average energy.
% 10-1-03
% v271 made the contour plot relative to all filaments energy magnitude
% 10-1-03
% v30 Now I am using ImageJ (downloaded from NIH) to obtain pixel locations
% from the graphics. The data is automatically recorded by ImageJ in
% a text file. To determine the end of a filament, the last point of
% the filament is recorded twice. All data in a text file is
manipulated
% and plotted at the same time.
% 10-8-03
% v31 Gives option to have the filaments numbered or to point out the
% largest energy value or to have a legend
% 12-9-03
% v32 Changed dE/dS=(B/3)*(dT*/dS*)^2 to dE/dS=(B*8/15)*(dT*/dS*)^2
% in program it's dEds=B/3*dTds2 to dEds=B*8/15*dTds2
% also changed the lamda component. There was an error in changing
% over to position vector. The bending modulus (B) is divided by 2
% in the original energy equation, however this did not happen for
% the position vector energy equation. Therefore instead of going
% through and dividing lamda by 2 everywhere, the division is taken
% place at the beginning of the program.
% 2-2-04
% v321 Increased data set to a total of 97 filaments
% 4-14-04
% v33 When I was calculating the lengths of the filaments by finding the
% length of the hypotenuse between points I would then multiply
this by the
% nmperpix to convert pixels to nm but I have forgotten why I did
this so
% in this version I have taken that multiplication out.











% I have also fixed the y axis so that it increases going up and
% there are no longer negative values for the y axis.
% For single filaments the x and y axes are in nm and so is the
% length of the filament. The options for legend,max point, and
% filament number are set to always be on
% 4-20-04
% v34 Added a section after the last pause that is to analyze the
% filaments. The plot is now in nm. Put in safety in calculating
% lengths so that if a length is too long (>25pix) then the program
% is paused because there is likely a filament that wasn't ended
% correctly
% 4-21-04
% v35 Determines the distance of each measured point from a center line.
% Also plots the AlldEds vs DFC and a histogram of AlldEds. Took
% out the first set of plots that were made in previous programs.
% 4-22-04
% v36 Found correlation coefficient (cc) for the DFC vs AlldEds and plotted
the
% least squares fit for the graph on the graph. Found the distance
% from the surface to the point along with its correlation
% coefficient (cc) and plotted the cc for that graph
% 4-25-04
% Plots circles over data points and the radius of the circle
% corresponds to the amount of energy at that point.
% 4-29-04
% Fitting a curve to filament data using position vector
%E(ri)/kT = lambda*sum((ri+l-2ri+ri-l)/delta(s)^2)^2 *delta(s) +
% sum((ri-rmi)^2/sigma^2
%where ri=xii+yij
%Take derivative and set to zero and find the closest fit
clear
% firstbatch
% secondbatch
% smBead2445900
% lgBead2516174
% tiff2425801
% matt2445901
% tiff2445901
% colin2486116
colin2435842
%determine number of filaments in data set
mk=l;
for i=l:length(data)-1
if data(i+1,2)==data(i,2) & data(i+1,3)==data(i,3)
filmark(mk,1)=i;
mk=mk+l;
end
end
filmark=[-l;filmark]; %for the main for loop length
%end of determine number of filaments in data set


% find any points that have been triple clicked or there is only one point
% for the filament
deq=0;
for i=l:length(filmark)-1
triple=filmark(i+1)-filmark(i);











if triple==4 I triple==3 I triple==2 I triple==l, deq=filmark(i);,
disp([triple deq]),end
end
if deq>0, return, end
% go look for the numbers listed and check for problems such as triples
% end of finding points in triplet
clc
disp(' ')
disp(' ')
disp('There are')
disp(mk-1)
disp('filaments loaded')
filstart = input ('Start at which filament: ');
filend = input ('End at which filament: ');
if filstart~=filend
ex = input('Exclude any filaments (enter numbers to exclude separated by
spaces or a range (3:13), 0 for none) :','s');
EDisp = MENU('How should size of energy be displayed?','Circles','Contour
Graph','None');
else


EDisp = 3;
ex='0';
end
exc=0;
excl=str2num(ex);
% MaxE = 1;
option)
% Fnum = 1;
option)
% Lege = 1;
option)


MaxE = 2;
Fnum = 2;
Lege = 2;
% MaxE =
%Fnum =
% Lege =
B=41000;
kT=4.1;
% L=B/kT;


%marker for removing filaments

option set to always off (next three lines give

option set to always off (next three lines give

%option set to always off (next three lines give


%option set to always on (next three lines give option)
%option set to always on (next three lines give option)
%option set to always on (next three lines give option)
MENU('Point to maximum energy value?','Yes','No');
MENU('Display filament numbers on graph?','Yes','No');
MENU('Display legend with range of energy levels?','Yes','No');
%B=(lambda)*k*T [=] pN*(nm^2)
%k*T [=] pN*nm
%L=lambda [=] nm


L=10000;
L=L/2;
reason in v32.
nmperpix=actdiam/pixeldiam;
gp=l;
graph
AllXf=[];
AllYf=[];
AllF=[];
AlldEds=[];
AllavgdEds=[];
DFC=[];
avgDFC=[];
DTS=[];
avgDTS=[];
TL=[];
% hold;


change for position v32 and up, see

5nm/pixel
counter for finding min and max for


keep
keep
keep
keep
keep
keep
keep
keep
keep
keep


x values
y values
the lengths
energy values
averaged energy values
distances to center line
DFC averages
distances to surface
DTS averages
filament lengths











%used to calculate distance from this center line to each point
slope = (center(1,2)-center(2,2))/(center(1,1)-center(2,1));
center = center(1,2)-slope*center(1,1);
%end of used to calculate distance from this center line to each point
for filnumber=filstart:filend;

%excludes certain filaments
for i=l:length(excl)
if filnumber==excl(1,i), exc=l;, end
end
if exc==l, exc=0;, continue, end
%end of excludes certain filaments
clear dr2 dx2 dy2 z Xf Yf lengths tlengths g d X p pd2 dTds dTds2 dEds
avgEds f filament fb dist DistFromCenter
F=filmark(filnumber+1)-filmark(filnumber)-1; %length of each
filament
AllF=[AllF F]; %store all the lengths
separatee each filament data into it's own set
mk=l;
for i=filmark(filnumber)+2:filmark(filnumber+1)
filament(mk,1)=data(i,2);
filament(mk,2)=data(i,3);
filament(mk,3)=data(i,1);
mk=mk+l;


end
%end of separate each filament data into it's own set
%calculates the distance of each point from a designated center line
for i = 1:length(filament)
dtbead=0;
fb(i) = filament(i,2)+l/slope*filament(i,1); %b of y=
for each point
dist(i,1) = (fb(i)-
centerb)/(slope+1/slope); %perpendicular x point on center lit
dist(i,2) =
slope*dist(i,1)+centerb; %perpendicular y point on cer
line


nx+b


ne

iter


DistFromCenter(i,l) = ((filament(i,1)-dist(i,1))^2 + (filament(i,2)-
dist(i,2))^2)^0.5; %actual distance to center line
%next indent calculates the distance from the surface of the bead to
each point (assuming a perfect spherical bead)
if (pixeldiam/2)^2 DistFromCenter(i,1)^2 > 0
dtbead = pixeldiam/2 ((pixeldiam/2)^2 -
DistFromCenter(i,l)^2)^0.5;
end
DistToSurface(i,l) = ((dist(i,1)-center(1,1))^2 + (dist(i,2)-
center(1,2))^2)^0.5 + dtbead;
%end of next indent calculates the distance from the surface of the
bead to each point (assuming a perfect spherical bead)
end
DFC = [DFC;DistFromCenter]; %saves all distances
avgDFC=[avgDFC;mean(DistFromCenter(2:length(DistFromCenter)-
1))]; %doesn't include end points because they are set to 0
DTS = [DTS;DistToSurface]; %saves all distances
avgDTS=[avgDTS;mean(DistToSurface(2:length(DistToSurface)-
1))]; %doesn't include end points because they are set to 0
%end of calculates the distance of each point from a designated center
line











%determine the axes of the contour graph
xmaxi(gp,l) = max(filament(:,1));
ymaxi(gp,1) = max(filament(:,2));
xmini(gp,l) = min(filament(:,1));
ymini(gp,1) = min(filament(:,2));
gp=gp+l; %counter for finding min
and max for graph
%end of determine the axes of the contour graph
% determine lengths between points
for i=l:F-1;
a=filament(i+1,1)-filament(i,1); %a side of triangle (x)
b=filament(i+1,2)-filament(i,2); %b side of triangle (y)
c=(a^2+b^2)^0.5; %hypotenuse of triangle
%safety check to make sure filaments were ended properly
% if c>75, disp(filmark(filnumber)+i), end
if c>75, disp(filament(i,3)), end
%end of safety check to make sure filaments were ended properly
% use following line to find location of points in data
% disp(filament(i,3))
lengths(i,1)=c;
tlengths(i,1)=sum(lengths); %progressive length


end
TL=[TL;tlengths(length(tlengths)) ];
lengths
% find x portion of ri of X*ri(Xf Yf)=d
X=[];
X(1,1)= S^-2 +L/lengths(1)^3;
X(1,2)= -2*L/lengths(1)^3;
X(1,3)= L/lengths(1)^3;
d(1,1)=filament(1,1)/S^2;
X(2,1)= -2*L/lengths(2)^3;
X(2,2)=S^-2 +5*L/lengths(2)^3;
X(2,3)= -4*L/lengths(2)^3;
X(2,4)= L/lengths(2)^3;
d(2,1)=filament(2,1)/S^2;
X(F-1,F-3)= L/lengths(F-1)^3;
X(F-1,F-2)= -4*L/lengths(F-1)^3;
X(F-1,F-1)=S^-2 +5*L/lengths(F-1)^3;
X(F-1,F) = -2*L/lengths(F-1)^3;
d(F-1,1) =filament(F-1,1)/S^2;
X(F,F-2)= L/lengths(F-1)^3;
X(F,F-1)= -2*L/lengths(F-1)^3;
X(F,F) =S^-2 +L/lengths(F-1)^3;
d(F,1) =filament(F,1)/S^2;
for i=3:F-2
X(i,i-2)= L/lengths(i)^3;
X(i,i-l)= -4*L/lengths(i)^3;
X(i,i) =S^-2 +6*L/lengths(i)^3;
X(i,i+l)= -4*L/lengths(i)^3;
X(i,i+2)= L/lengths(i)^3;
d(i,l) =filament(i,l)/S^2;
end


;save all


Xf=X\d;
%end of finding x portion of ri of X*ri(Xf Yf)=d
% find y portion of ri of Y*ri(Xf Yf)=g
% X and Y matrices are identical so the following is the different end











% points for the y axis
g(l,l) =filament(1,2)/S^2;
g(2,1) =filament(2,2)/S^2;
g(F-1,1)=filament(F-1,2)/S^2;
g(F,1) =filament(F,2)/S^2;
for i=3:F-2
g(i,1)=filament(i,2)/S^2;
end
Yf=X\g;
%end of finding y portion of ri of Y*ri(Xf Yf)=g
% (dTheta/dS)^2 = (d^2ri/dS^2)^2 = (d^2x/dS^2)^2 + (d^2y/dS^2)^2
% (d^2x/dS^2) = (x(i+l)-2xi+x(i-1))/S^2
% x=l or x=N are set to 0
for i=2:length(Xf)-l
dx2(i,1) = (Xf(i+l)-2*Xf(i)+Xf(i-1))/lengths(i)^2;
dy2(i,l) = (Yf(i+l)-2*Yf(i)+Yf(i-1))/lengths(i)^2;
end
dx2=[dx2;0];
dy2=[dy2;0];
dr2 = dx2.^2+dy2.^2;
% end of derivative at each point along filament
dTds2=dr2; %(dTheta/dS)^2 =
(d^2ri/dS^2)^2
dEds=B*8/15*dTds2; %energy at each point
avgdEds=mean(dEds(2:length(dEds)-1,1)); %average energy of filament
without ends which are 0
if filstart==filend
subplot(2,1,1)
hold on
%I want the filaments plotted versus nm lengths which is done in
%the following for just one filament
Xf=(Xf xmini)*nmperpix;
Yf=(Yf ymini)*nmperpix;
filament(:,1)=(filament(:,1) xmini)*nmperpix;
filament(:,2)=(filament(:,2) ymini)*nmperpix;
xmaxi=(xmaxi-xmini)*nmperpix;
ymaxi=(ymaxi-ymini)*nmperpix;
xmini=0;
ymini=0;
%If pixels is good enough then take out from next comment up to
%this and the labels for nm
plot(Xf,Yf,'b-')
plot(filament(:,l),filament(:,2),'mx')
xlabel('nm');
ylabel('nm');
hold off
subplot(2,1,2)
hold on
plot(Xf,Yf,'b-')
ylabel('nm');
hold off
% else
% plot(Xf,Yf)
% plot(filament(:,1),filament(:,2))


end
aw=num2str(filnumber);


;distinguish filaments on


graph











if Fnum~=2
text(Xf(1),Yf(1),aw)
graph
end
AllXf=[AllXf;Xf];
AllYf=[AllYf;Yf];
AlldEds=[AlldEds;dEds];
AllavgdEds=[AllavgdEds;avgdEds];
%finds maximum dEds of all filaments
if max(dEds)>=max(AlldEds)
bigdEds=filnumber;
[M,loc]=max(dEds);
Xloc=Xf(loc);
Yloc=Yf(loc);
maxpt=[filament(loc,1) filament(loc,2)];
end
%end of finds maximum dEds of all filaments


distinguish filaments on


end
% v is the number of contour lines, z is the energy at each point
% could just set v=20 and it would be the same thing as line below (as long
% as the number of contour lines is 20 in line below)
% vd=20;
% v=(min(AlldEds):(max(AlldEds)-min(AlldEds))/vd:max(AlldEds));
v=20;
z=diag(AlldEds);
if filstart==filend
hold on
subplot(2,1,2)
contour(AllXf,AllYf,z,v);
hold off
% else
% contour(AllXf,AllYf,z,v);


% sets limits of the viewing area for the graph
if max(Xf)>max(xmaxi), xmaxi=max(Xf);, end
if min(Xf) if max(Yf)>max(ymaxi), ymaxi=max(Yf);, end
if min(Yf) xmax = max(xmaxi);
ymin = min(ymini);
xmin = min(xmini);
ymax = max(ymaxi);
% if abs(xmax-xmin) > abs(ymax-ymin)
square


makes viewing window a


ymax=ymin+.5*abs(ymax-ymin)+.5*abs(xmax-xmin);
ymin=ymin+.5*abs(ymax-ymin)-.5*abs(xmax-xmin);
else
xmax=xmin+.5*abs(ymax-ymin)+.5*abs(xmax-xmin);
xmin=xmin-.5*abs(ymax-ymin)+.5*abs(xmax-xmin);
end
end of sets limits of the viewing area for the graph
.=[xmin Xloc]; %for plot of line to maximum dEds
.=[Yloc Yloc];
=[Xloc Xloc];
=[ymin Yloc]; %for plot of line to maximum dEds
filstart==filend %if just plotting one filament
subplot(2,1,1)











hold on
if MaxE~=2
plot(xl, yl,'k--')
dEds
plot(x2, y2,'k--')
dEds
end
axis([xmin xmax ymin ymax])
subplot(2,1,2)
hold on
if MaxE~=2
plot(xl, yl,'k--')
dEds
plot(x2, y2,'k--')
dEds
end
axis([xmin xmax ymin ymax])
maxdEdsString = num2str(round
xlabel(['Maximum dE/dS found
hold off
% else
% hold on
% if MaxE~=2
% plot(xl, yl,'k--')
% plot(x2, y2,'k--')
% end
% axis([xmin xmax ymin ymax])
% mg=num2str(v');
% if Lege~=2
% legend(mg,-l);
contour lines
% end
% hold off


%draw a line from y axes to maximum

%draw a line from x axes to maximum


%sets axis on graph




%draw a line from y axes to maximum

%draw a line from x axes to maximum


%sets axis on graph
(max(dEds)*100)*.01);
= ',maxdEdsString,' pN'])


plotting more than one filament


,draw a line from y axes to maximum dEds
,draw a line from x axes to maximum dEds

%sets axis on graph
%for legend

%legend displays the range of


end
disp(' ')
disp(' ')
disp(['Filament with the largest dEds is filament number num2str(bigdEds)])
disp('at position )
disp([Xloc Yloc])
disp(' ')
disp(['Average filament length = num2str(round(mean(TL)*nmperpix)) nm'])
% pause
everythingg after this pause was added to analyze the filament data, the
%plot is now versus nm with nm length filaments
%used for plotting (gives number of data points used for each line)
FL=[1 A11F(1)];
for i = 2:length(AllF)
FL(i,:) = [FL(i-1,2)+1 FL(i-1,2)+AllF(i)];
end
%end of used for plotting (gives number of data points used for each line)


xmini=min(AllXf);
ymini=min(AllYf);
AllXf=(AllXf-xmini)*nmperpix;
AllYf=(AllYf-ymini)*nmperpix;
xmin=0;
xmax=max(AllXf);











ymin=0;
ymax=max(AllYf);
% plots filaments
for i = l:length(AllF)
hold on
plot(AllXf(FL(i,l):FL(i,2)), AllYf(FL(i,1):FL(i,2)))
end
axis([xmin xmax ymin ymax]) %sets axis on graph
% end of plots filaments
if EDisp == 1
%plot circles on data points corresponding to energy size
angle=0:0.01:2*pi;
for i = 1:length(AlldEds)
hold on
xl=AlldEds(i)*cos(angle)+AllXf(i);
yl=AlldEds(i)*sin(angle)+AllYf(i);
% plot(xl,yl,'b-')
if AlldEds(i)<=14.999, plot(xl,yl,'k-'), end
if AlldEds(i)>=15 & AlldEds(i)<=24.999, plot(xl,yl,'b-'), end
if AlldEds(i)>=25 & AlldEds(i)<=34.999, plot(xl,yl,'g-'), end
if AlldEds(i)>=35 & AlldEds(i)<=44.999, plot(xl,yl,'c-'), end
if AlldEds(i)>=45 & AlldEds(i)<=54.999, plot(xl,yl,'r-'), end
if AlldEds(i)>=55, plot(xl,yl,'m-'), end
end
%end of plot circles on data points corresponding to energy size
end
plot([100 200],[100 100],'k-')
text(100,125,'100nm')
%comments below skews contour only along x direction
% for i=l:length(z)
% if i+l<=length(z), z(i,i+1)=mean([z(i,i) z(i,i+l)]);, end
% if i+2<=length(z), z(i,i+2)=mean([z(i,i+1) z(i,i+2)]);, end
% if i+3<=length(z), z(i,i+3)=mean([z(i,i+2) z(i,i+3)]);, end
% if i-l>=l, z(i,i-l)=mean([z(i,i) z(i,i-l)]);, end
% if i-2>=2, z(i,i-2)=mean([z(i,i-1) z(i,i-2)]);, end
% if i-3>=3, z(i,i-3)=mean([z(i,i-2) z(i,i-3)]);, end
% end
%end of comments below skews contour only along x direction
% skews contour in x and y direction
% for i=l:length(AlldEds)-1
% oneZ(i)=.333*AlldEds(i)+.333*AlldEds(i+1);
% if i+2<=length(AlldEds), twoZ(i)=.25*AlldEds(i)+.25*AlldEds(i+2);, end
% end
% z=diag(AlldEds) + diag(oneZ,1) + diag(twoZ,2) + diag(oneZ,-l) + diag(twoZ,
2);
% end of skews contour in x and y direction
if EDisp == 2
v=15; %number of contour limes
contour(AllXf,AllYf,z,v);
end
% %plot a circle to represent the bead on the filament graph (next indent
section)
% % CenterBeadX = (2*((center(1,2)-
centerb)*slope+center(1,1))+(4*((center(1,2)-centerb)*slope+center(1,1))^2-
4*(slope^2+1)*(center(1,1)^2+center(1,2)^2-2*center(1,2)*centerb+centerb^2-
pixeldiam^2/4))^0.5)/2/(slope^2+1);
% % CenterBeadY = slope*CenterBeadX+centerb;











% circle = [beadcenter(1,1)-pixeldiam/2:1:beadcenter(1,1)+pixeldiam/2];
% circle = ((pixeldiam^2/4-(circlex-
beadcenter(l,l)).^2).^.5+beadcenter(1,2)-ymini).*nmperpix;
% negcircley = (-(pixeldiam^2/4-(circlex-
beadcenter(l,l)).^2).^.5+beadcenter(1,2)-ymini).*nmperpix;
% circle = (circlex-xmini).*nmperpix;
% plot(circlex,circley,'r'),plot(circlex,negcircley,'r')
% %end of plot a circle to represent the bead on the filament graph (next
indent section)
%better plot of the bead circle
angle=0:0.01:2*pi;
xl=((pixeldiam/2*cos(angle)+beadcenter(1,1))-xmini).*nmperpix;
yl=((pixeldiam/2*sin(angle)+beadcenter(1,2))-ymini).*nmperpix;
plot(xl,yl,'r')
text((beadcenter(1,1)-xmini)*nmperpix,(beadcenter(1,2)-
ymini)*nmperpix,'Bead')
%end of better plot of the bead circle
hold off
pause
%plot of dEds vs DistFromCenter
% remove 0 value dEds and corresponding distances
nonOdEds=find(AlldEds>0);
for i=l:length(non0dEds)
nonOAlldEds(i,1)=AlldEds(non0dEds(i));
nonODFC(i,1)=DFC(nonOdEds(i))*nmperpix;
nonODTS(i,1)=DTS(nonOdEds(i))*nmperpix;
end
% end of remove 0 value dEds and corresponding distances
subplot(2,2,1)
plot(nonODFC,nonOAlldEds,'.g')
ylabel('dE/ds [pN]')
xlabel('nm from center line')
[r2DFC,linecc]=corcoeff(nonODFC,nonOAlldEds);
hold on
plot(linecc(:,1),linecc(:,2))
title(['dEds vs Distance to Center of Actin Tail: R^2='
num2str(round(r2DFC*10000)*.0001)])
hold off
subplot(2,2,2)
avgDFC=avgDFC.*nmperpix;
%plot of Average dEds vs Average DFC
plot(avgDFC,AllavgdEds,'x')
ylabel('dE/ds [pN]')
xlabel('nm from center line')
[r2DFC,linecc]=corcoeff(avgDFC,AllavgdEds);
hold on
plot(linecc(:,1),linecc(:,2))
title(['Avg dEds vs Avg Distance to Center of Actin Tail: R^2='
num2str(round(r2DFC*10000)*.0001)])
hold off
end of plot of Average dEds vs Average DFC
subplot(2,2,3)
plot(nonODTS,nonOAlldEds,'.g')
ylabel('dE/ds [pN]')
xlabel('nm from surface of bead')
[r2DTS,linecc]=corcoeff(nonODTS,nonOAlldEds);
hold on











plot(linecc(:,1),linecc(:,2))
title(['dEds vs Distance to Bead Surface: R^2='
num2str(round(r2DTS*10000)*.0001)])
hold off
subplot(2,2,4)
%plot of Average dEds vs Average DTS
avgDTS=avgDTS.*nmperpix;
plot(avgDTS,AllavgdEds,'x')
ylabel('dE/ds [pN]')
xlabel('nm from surface of bead')
[r2DTS,linecc]=corcoeff(avgDTS,AllavgdEds);
hold on
plot(linecc(:,l),linecc(:,2))
title(['Avg dEds vs Avg Distance to Bead Surface: R^2='
num2str(round(r2DTS*10000)*.0001)])
hold off
%end of plot of Average dEds vs Average DTS
pause
subplot(1,1,1)
hist(nonOAlldEds,100)
xlabel('dE/ds [pN]')
ylabel('Count')
a = used to calculate length between points
actdiam = actual diameter in nm of bead
AllavgdEds = keeps all the averaged change in Energy per length
AlldEds = keeps all change in Energy per length
AllF = keep all the number of points used for each filament
AllXf = keeps all the determined X positions from the equation
AllYf = keeps all the determined Y positions from the equation
avgdEds = average energy of each filament
avgDFC = average distance from center, keeps average distance from the center
of the tail for each filament
avgDTS = average distance to surface, keeps average distance from the surface
of the bead for each filament
aw = used to label the filaments on the graph
b = used to calculate length between points
B = bending modulous (lamda *kT)
bigdEds = stores filament with the largest energy value
c = calculated length between points (pixel length)
center = b value of y=mx+b of line running through center of actin tail
d = x positions from the images
data = matrix that holds all the data points from plotting the filaments and
is locating in whichever m file is used
dEds = change in energy along length
DFC = distance from center, keeps distance from the center of the tail
dist = x and y perpendicular point from individual points on center line
DistFromCenter = actual distance from center line for each point
DistToSurface = distance to surface of the bead for each point
dr2 = combination of dx2 and dy2 to yield second derivative of position
vector
dtbead = calculates distance to bead from each individual point
dTds2 = same as dr2
DTS = distance to surface, keeps distance from the surface of the bead
dx2 = second derivative of found x values
dy2 = second derivative of found y values
ex = separate filaments to exclude in the array of filaments selected
exc = marker (0 or 1) to know if current filament is excluded











excl = lists excluded filaments in number format
F = stores number of points used for individual filaments
fb = b of y=mx+b for each point of individual filaments
filament = data for each individual filament in pixel length
filend = which filament to stop analyzing
filmark = gives the location of the last point for each filament
Fnum = option to list numbers next to the filaments (l=off, 2=on)
filnumber = for loop counter to know which filament to work with
filstart = which filament in the group to start analyzing
g = y positions from the images
gp = counter for min and max values of filament location (counts from 1 to
number of filaments to analyze)
kT = boltzmanns temperature
L = lamda, persistence length
Lege = option to display a legend listing the levels of energy on the contour
map (l=off, 2=on)
lengths = lengths between each point
loc = index point of dEds with the highest energy
M = value of the largest energy
maxdEdsString = to display a rounded maximum energy value on graph
MaxE = option to have dashed lines point out highest energy point (l=off,
2=on)
maxpt = actual position on image of largest energy level
mk = counter for number of individual filaments
nmperpix = convert pixels into nm
pixeldiam = diameter of bead in pixels from image
slope = slope of line running through center of actin tail
TL = keeps all filament total lengths in pixel length
lengths = progessive length of filament
v = number of contour lines
X = x portion of X*Xf=d, X are the amounts multiplied by the found x
positions
xl = plot horizontal line from y axes to highest energy
x2 = plot vertical line from x axes to highest energy
Xf = found x positions for each filament
Xloc = found x position of the largest energy value
xmax = x position with largest energy value
xmaxi = stores largest x value of all filaments
xmin = x position with smallest energy value
xmini = stores smallest x value of all filaments
Y = y portion of Y*Yf=g, Y are the amounts multiplied by the found y
positions
yl = plot horizontal line from y axes to highest energy
y2 = plot vertical line from y axes to highest energy
Yf = found y positions for each filament
Yloc = found y position of the largest energy value
ymax = y position with largest energy value
ymaxi = stores largest y value of all filaments
ymin = y position with smallest energy value
ymini = stores smallest y value of all filaments
z = matrix with the diagonal being all the energy levels for each point









APPENDIX B
MYOSIN SUBFRAGMENT-1

Introduction

Myosin subfragment-1 (Sl) binds to actin filaments in a specific orientation showing the

polarity of the filament. When bound to actin filaments, S1 appears on the filament as an

arrowhead with the barbed end of the arrow pointing toward the (+)-end and the pointed end of

the arrow pointing toward the (-)-end. To determine filament polarity in EM experiments, S1

was added to actin filaments to determine their orientation. Several methods were attempted to

bind S1 to actin filaments including: flowing S1 across filaments bound to a substratum, mixing

S1 with F-actin and then binding to a substratum, varying concentrations of both S1 and F-actin,

trying different buffers, cleaving S1 with papain or a-chymotrypsin (Sigma-Aldrich Co.), using

myosin prepared in the lab or from Cytoskeleton, Inc., using deactivated n-ethylmaleimide

(NEM) myosin, and trying different EM techniques such as paraformaldehyde fixation, negative

staining, or replicas. None of these variations were successful in labeling actin filaments with S1

so a color change assay (using TIRF) was performed to determine filament polarity.

Myosin Purification

The following protocol was derived from several sources (106-108). Frozen rabbit skeletal

muscle (300 g) was thawed from -70C to 4 oC overnight. Muscle was minced twice in a meat

grinder and stirred for exactly 15 minutes with 1 L of buffer A (0.3 M KC1 pH 6.5, 2 mM sodium

pyrophosphate, 0.15 M potassium phosphate buffer). Four liters of water was then added and the

solution was then poured through 4 layers of cheesecloth. Next, 7 L of water was added to the

filtered solution causing the myosin to precipitate out. The solution incubated for 3 h. at 4C

allowing myosin to precipitate. Supernatant was removed as best possible and the precipitate

was centrifuged (15,000 g for 20 minutes at 40C). The precipitate was triturated in 300 mL









buffer B (0.4 M KC1 pH 6.7 and 0.03 M potassium phosphate buffer) and centrifuged (16,000 g

for 30 minutes at 40C) discarding the precipitate. The supernatant was passed through glass wool

to remove lipids and 250 mL of water was added to the filtered solution. The solution was

allowed to incubate for 30 minutes at 40C then centrifuged (16,000 g for 30 minutes at 40C) and

the precipitate was discarded. The supernatant was diluted with 4 L of water and centrifuged

(15,000 g for 20 minutes at 0C). The precipitate was the dissolved in a minimal amount of

buffer C (0.5 M KC1 pH 6.8 and 5 mM potassium phosphate buffer). The absorbance was

measured at 280 nm using an extinction coefficient of 0.59 mL/mg-cm and myosin was stored at

-70C. If myosin was to be inactivated using NEM, an aliquot of 10 tM myosin was dialyzed

against an imidazole buffer (10 mM imidazole pH 7, 0.5 M KC1, and 10 mM EDTA) for 2 hours.

The myosin was then incubated with 1 mM NEM for 1 hour on ice (54).

Purification of S1 using Papain

Myosin (30 mg) was dialyzed into sample buffer (0.03 M KC1 pH 6.8 and 6 mM potassium

phosphate buffer) overnight. The myosin was pelleted (1,000 g) and resuspend in 1 mL of

papain digestion buffer (sample buffer with 20 mM cysteine-HCl at pH 7). The purchased

papain slurry (500 tL) was equilibrated by washing with 4 mL of papain digestion buffer. The

equilibrated papain was incubated while mixing at room temperature for 1 hour. The myosin

was then added to the prepared papain and mixed for 15 minutes. The cleaving of myosin was

stopped by adding 2 tL of 1 M iodoacetic acid. Insoluble myosin and papain was pelleted at

14,000 g for 1 minute and the supernatant containing S1 was centrifuged (100,000 g for 1 hour).

The absorbance of S1 was measured at 280 nm using an extinction coefficient of 0.8 mL/mg-cm

and S1 was stored at -700C.









Purification of S1 using a-Chymotrypsin

Myosin was dialyzed overnight against 0.05 M KC1 pH 7.0 (100, 109-112). The myosin

was then centrifuged at 1,000 g for 2 minutes and the pellet concentration was diluted to

10 mg/ml in 2x chymotrypsin digestion buffer (0.24 M NaCl pH 7, 20 mM sodium phosphate,

and 4 mM EDTA). The diluted myosin was then homogenized in a teflon-glass dounce

homogenizer on ice by hand to produce an opalescent solution. a-Chymotrypsin was dissolved

in ImM HC1 and the absorbance measured at 280 nm using an extinction coefficient of

2.04 mL/mg-cm. Myosin was digested with 0.03 mg/ml a-chymotrypsin (N-p-tosyl-L-lysine

chloromethyl ketone (TLCK) treated) at 250C for 20 minutes while shaking. Digesting was

stopped by supplementing to 1 mM PMSF. The cleaved myosin and undigested myosin was

precipitated by adding one volume of 6 mM MgCl2 and centrifuged (15,000 g for 5 minutes at

room temperature). The supernatant was then mixed with 2 volumes of saturated ammonium

sulfate at 250C (4.1 M or 767 g of (NH4)2SO4 per liter of water) and centrifuged (15,000 g for

5 minutes) to precipitate S1. The precipitate was then triturated in a small volume of 100 mM

KC1 and 20 mM Tris-HCl pH 8.0 until the solution became clear. Salt was removed from the S1

solution by passing the mixture over a Sephadex G25 column and the absorbance measured at

280 nm using an extinction coefficient of 0.80 mL/mg-cm.









APPENDIX C
IMAGEJ MACROS


// Region of Interest (ROI) extractor

macro "ROI Extractor Action Tool "{
// This section asks for the directory to manipulate images and then checks what ROI number is
next
x=0;
fitdirectory = getDirectory("Select a Directory") // prompts what directory to take images
from for processing
list = getFileList(fitdirectory);
for (i = 0; i < list.length; i++) {
for (j=l; j if (startsWith(list[i], j)) {
x=j;
} } }
x=d2s(x+l,0);
Dialog.create("Image Choice"); // create dialog box named Image choice
Dialog.addMessage("While thresholding, press Y for an overlay. Close the blank image \n when
done with thresholding and all images will be saved.")
Dialog.addString("X Y Width Height", 0 0 0 0 ") //Form to enter text
Dialog.addString("ROI working with",x); // Form to enter number
Dialog.show(; // show dialog box created
xywhstring = Dialog.getStringo; // stores entered string
x = parseInt(Dialog.getString(); // stores entered number
xywhsplit = split(xywhstring) // split string entered with any delimiter
xywh= newArray(4) // create new array
for(i=0; i<=xywhsplit.length-1; i++){
xywh[i] = parselnt(xywhsplit[i]); // convert string to number
}
open(fitdirectory+"bandpass tritc.tif'); // opens tritc image
trit = getImagelDo; // image ID of original tritc image
makeRectangle(xywh[0], xywh[1], xywh[2], xywh[3]) //make new rectangle selection
run("Duplicate...", "title=bandpass tritc-1.tif');
run("Tiff..", "save="+fitdirectory+x+"bandpass tritc.tif'); // saves new bandpass image in
appropriate directory
smalltritc = getImagelD);
smalltritctitle = getTitle);
selectImage(trit);
close);
open(fitdirectory+"bandpass_fitc.tif'); // opens the fitc image
fit = getImageID(; // image ID of original fitc image
makeRectangle(xywh[0], xywh[1], xywh[2], xywh[3]) //make new rectangle selection
run("Duplicate...", "title=bandpassfitc-1.tif');









run("Tiff..", "save="+fitdirectory+x+"bandpassfitc.tif'); // saves new bandpass image in
appropriate directory
smallfitc = getImagelD);
smallfitc_title = getTitle);
selectImage(fit);
close);
run("RGB Merge...", "red="+smalltritc title+" green="+smallfitc title+" blue=*None* keep");
newImage("Close Me When Done Thresholding", "8-bit White", 16, 16, 1);
dummylmage=getImagelD();
run("Tile");
selectImage(smallfitc);
selectImage(smalltritc);
run("Brightness/Contrast...");
while (isOpen(dummylmage)) { //this pauses the macro until the small tritc image is closed
wait(50);
}
selectImage(smalltritc);
run("Tiff...", "save="+fitdirectory+x+"bandpass tritc.tif ); // saves new bandpass image in
appropriate directory
close);
selectImage(smallfitc);
run("Tiff..", "save="+fitdirectory+x+"bandpassfitc.tif'); // saves new bandpass image in
appropriate directory
close);
run("Tiff...", "save="+fitdirectory+x+"RGB.tif); // saves new bandpass image in appropriate
directory
close);
}

I End of Region of Interest (ROI) extractor


// Duplicate rectangular selection onto an image
// Write down x, y, width, height from original rectangular selection to enter here in order of x y
w h separated by spaces

macro "Reproduce Rectangle Selection (Cut Paste) Tool -" {
xywhstring=getString("X Y Width Height", 0) // enter x y width height value with spaces
from previous rect selection
xywhsplit = split(xywhstring) // split string entered with any delimiter
xywh= newArray(4) // create new array
for(i=0; i<=xywhsplit.length-1; i++){
xywh[i] = parselnt(xywhsplit[i]); // convert string to number


}
makeRectangle(xywh[0], xywh[1], xywh[2], xywh[3])
}


//make new rectangle selection










I End of Reproduce Rectangle Selection (Cut Paste) Tool

//Get coordinates of a rectangular selection or a line selection and lists in log box
//Use this in conjunction with the "Reproduce Rectangle Selection (Cut Paste) Tool"

macro "Get Selection Coord Action Tool -
COOOD22D23D24D2aD2bD2cD32D33D34D35D36D37D38D39D3aD3bD3cD42D43D44D4aD
4bD4cD53D5bD63D6bD73D7bD83D8bD93D9bDa3DabDb2Db3Db4DbaDbbDbcDc2Dc3Dc4
Dc5Dc6Dc7Dc8Dc9DcaDcbDccDd2Dd3Dd4DdaDdbDdcCOOOC 111 C222C333C444C555C666
C777C888C999CaaaCbbbCcccCdddCeeeCfffDOOD01D02D03D04D05D06D07D08DO9DOaDO
bDOcDOdDOeDOfD10D 11D12D13D14D15D16D17D18D19DlaDlbDlcDldDleDlfD20D21D2
5D26D27D28D29D2dD2eD2fD30D31D3dD3eD3fD40D41D45D46D47D48D49D4dD4eD4fD5
OD51D52D54D55D56D57D58D59D5aD5cD5dD5eD5fD60D61D62D64D65D66D67D68D69D
6aD6cD6dD6eD6fD70D71D72D74D75D76D77D78D79D7aD7cD7dD7eD7fD80D81D82D84D
85D86D87D88D89D8aD8cD8dD8eD8fD90D91D92D94D95D96D97D98D99D9aD9cD9dD9e
D9fDaODalDa2Da4Da5Da6Da7Da8Da9DaaDacDadDaeDafDbODbDb5Db6Db7Db8Db9Dbd
DbeDbfDcODclDcdDceDcfDdODd Dd5Dd6Dd7Dd8Dd9DddDdeDdfDeODelDe2De3De4De5
De6De7De8De9DeaDebDecDedDeeDefDfODflDf2Df3Df4Df5Df6Df7Df8Df9DfaDfbDfcDfdD
feDff" {
type = selectionType);
if (type==-l)
print("No selection");
else {
if (type==5){
getLine(xl, yl, x2, y2, lineWidth);
length=sqrt((xl-x2)*(xl-x2)+(yl-y2)*(yl-y2));
if (y2>yl)
angle=-acos((x2-xl)/length)*180/PI;
else
angle=acos((x2-xl)/length)* 180/PI;
print("Line,x l,yl,x2,y2,length,angle,",x l,yl,x2,y2,length,angle);
}
if (type==0){
getBoundingRect(x, y, w, h);
print("Rectangle, x, y, width, height,",x,y,w,h);
}}
restorePreviousTool;
//setTool(O); // was used to switch to rectangle selection tool but line above seems more
appropriate
}

I End of Get Selection Coord Tool

I This reproduces either a line or rectangular selection displayed on one image to all other open
images









// Given the other open images have appropriate dimensions to reproduce selection

macro "Reproduce Selection Tool -
COOOD 11D12D13D14D15D16D17D18D19DlaDlbDlcDldD1eD21D2eD31D3eD41D44D4eD
51D54D5eD61D65D6eD71D75D7eD81D86D8eD91D96D9eDalDa6DaeDblDb7DbeDclDc7D
ceDd 1 DdeDelDe2De3De4De5De6De7De8De9DeaDebDecDedDeeCOOOC 111 C222C333C444C
555C666C777C888C999CaaaCbbbCcccCdddCeeeCfffDOOD01D02D03D04D05D06D07D08DO
9DOaDObDOcDOdDOeDOfD10D 1fD20D22D23D24D25D26D27D28D29D2aD2bD2cD2dD2fD3
OD32D33D34D35D36D37D38D39D3aD3bD3cD3dD3fD40D42D43D45D46D47D48D49D4aD
4bD4cD4dD4fD50D52D53D55D56D57D58D59D5aD5bD5cD5dD5fD60D62D63D64D66D67
D68D69D6aD6bD6cD6dD6fD70D72D73D74D76D77D78D79D7aD7bD7cD7dD7fD80D82D8
3D84D85D87D88D89D8aD8bD8cD8dD8fD90D92D93D94D95D97D98D99D9aD9bD9cD9dD
9fDaODa2Da3Da4Da5Da7Da8Da9DaaDabDacDadDafDbODb2Db3Db4Db5Db6Db8Db9DbaDb
bDbcDbdDbfDcODc2Dc3Dc4Dc5Dc6Dc8Dc9DcaDcbDccDcdDcfDdODd2Dd3Dd4Dd5Dd6Dd7
Dd8Dd9DdaDdbDdcDddDdfDeODefDfODflDf2Df3Df4Df5Df6Df7Df8Df9DfaDfbDfcDfdDfeD
ff'{
type = selectionType);
if (type==-l)
print("No selection");
else {
if (type==5){
getLine(xl, yl, x2, y2, lineWidth);
print("X1-value,",xl);
print("Y1-value,",yl);
print("X2-value,",x2);
print("Y2-value,",y2);
length=sqrt((xl-x2)*(xl-x2)+(yl-y2)*(yl-y2));
print("Length,",length);
if(y2>yl)
angle=-acos((x2-xl)/length)* 180/PI;
else
angle=acos((x2-xl)/length)* 180/PI;
print("Angle,",angle);
}
if (type==0){
getBoundingRect(x, y, w, h);
print("X-value,",x);
print("Y-value,",y);
print("Width-value,",w);
print("Height-value,",h);
}}
if (nlmages==0)
print("No images are open");
else
imagesopen = newArray(nImages);
for(i=l; i<=nlmages(; i++){









selectlmage(i);
if (type==5)
makeLine(xl, yl, x2, y2);
if (type==0)
makeRectangle(x, y, w, h);
}}

I End of Reproduce Selection Tool

//Implements (TRITC-FITC)/(TRITC+FITC+0.01) for
// our single filament data, the 0.01 is to prevent infinity cases
// Note: image ids are given as negative values. To select a specific image use it's negative id.
// To select the ith image that has been open use a positive value starting at 1 to number open.

macro "Ratio Tool -
C000D17D27D37D47D57D67D77D87D97Da7Db7Db9DbaDbbDbcDbdDc7Dc9DcaDcbDccDc
dDd7Dd9DdaDdbDdcDddDe7De9DeaDebDecDedDf9DfaDfbDfcDfdC000D7aD89D8aD8bD9a
C000C 11C222C333C444C555C666C777C888C999D19DlaDlbDlcDldD29D2aD2bD2cD2d
D39D3aD3bD3cD3dD49D4aD4bD4cD4dD51D52D53D54D55D59D5aD5bD5cD5dD61D62D6
3D64D65D71D72D73D74D75D81D82D83D84D85D91D92D93D94D95C999CaaaCbbbCcccC
dddCeeeCfffD69D6aD6bDa9DaaDabCfffDOOD01D02D03D04D05D06D07D08D09D0aDObDO
cDOdDOeDOfD10D11D12D13D14D15D16D18D1eD1fD20D21D22D23D24D25D26D28D2eD
2fD30D31D32D33D34D35D36D38D3eD3fD40D41D42D43D44D45D46D48D4eD4fD50D56D
58D5eD5fD60D66D68D6cD6dD6eD6fD70D76D78D79D7bD7cD7dD7eD7fD80D86D88D8cD
8dD8eD8fD90D96D98D99D9bD9cD9dD9eD9fDa0DalDa2Da3Da4Da5Da6Da8DacDadDaeDa
fDb0DblDb2Db3Db4Db5Db6Db8DbeDbfDcODclDc2Dc3Dc4Dc5Dc6Dc8DceDcfDd0Dd Dd2
Dd3Dd4Dd5Dd6Dd8DdeDdfDe0DelDe2De3De4De5De6De8DeeDefDf0DflDf2Df3Df4Df5Df
6Df7Df8DfeDff' {
//Dialog box to get exact file names image.tif
requires("1.34m"); // make sure correct imagej version is running
if (nImages==0) // returns number of images open
print("No images are open");
else
imagesopen = newArray(nImages); //makes array size of images open
for(i=l; i<=nImages(; i++){ //for(initialize, limit, increment)
selectImage(i); //selects image as they are listed, -i=actual images, i=count of
images
imagesopen[i-1] = getTitle(; //returns title of images, arrays start at 0 which is why i-1
}
Dialog.create("Image Choice"); // create dialog box named Image choice
Dialog.addChoice("TRITC", imagesopen); // drop down box of images open
Dialog.addChoice("FITC", imagesopen, imagesopen[l]); // drop down box of images open,
with default of second image available
Dialog.show(; // show dialog box created
TRITC = Dialog.getChoiceo; // saves choice entered in first dialog box









FITC = Dialog.getChoiceo; // saves choice entered in second dialog box, order matt
here
//Adds TRITC to FITC
imageCalculator("add create 32-bit", FITC, TRITC); //adds images and makes the new
image 32-bit (necessary for these images)
X = nImage(; // save image id of image just created
selectImage(X); // make image active
run("Add...", "value=0.01"); // add 0.01 to all values to prevent infinity values latter
rename("Add"); // name image add
//Subtracts TRITC to FITC and then invert LUT
imageCalculator(" subtract create 32-bit", TRITC, FITC); //subtracts images and makes the
new image 32-bit (necessary for these images) however when subtracting the image ends up
inverted
selectImage(X + 1); // make image active
rename(" Subtract"); // name image subtract
requires(" 1.30j"); // check appropriate imagej version is running
getLut(reds, greens, blues); // get values for every pixel
for (i=0; i values after subtracting
reds[i] = 255-reds[i];
greens[i] = 255-greens[i];
blues[i] = 255-blues[i];


ers


setLut(reds, greens, blues); // sets new pixel values on image
//Divides Subtract/Add then Invert LUT
imageCalculator(" divide create 32-bit", "Subtract", "Add"); //divides
selectImage(X + 2); // make image active
rename("Ratio"); // name image Ratio
requires(" 1.30j"); // check version
getLut(reds, greens, blues); // get pixel values
for (i=0; i reds[i] = 255-reds[i];
greens[i] = 255-greens[i];
blues[i] = 255-blues[i];


setLut(reds, greens, blues);
//Closes Add and Subtract images
selectImage(X);
close);
selectImage(X);
close);


// set pixel values
// closes intermediate images


}

I End of Ratio macro

I This macro takes a fitc image and a tritc image runs bandpass on both and asks to save









// both before merging into an rgb. Make sure to open fitc first to operate correctly.

macro "Bandpass RGB Merge Action Tool -
COOOD21D22D23D24D25D26D27D28D29D2aD2bD2cD2dD2eCO00C 111C222D31D32D33D3
4D35D36D37D38D39DD3aD3D3cD3dD3eC222C333C444C555C666D41D42D43D44D45D46
D47D48D49D4aD4bD4cD4dD4eC666C777C888D51D52D53D54D55D56D57D58D59D5aD5b
D5cD5dD5eC888C999CaaaD61D62D63D64D65D66D67D68D69D6aD6bD6cD6dD6eCaaaCbb
bCcccD71D72D73D74D75D76D77D78D79D7aD7bD7cD7dD7eCcccCdddD81D82D83D84D8
5D86D87D88D89D8aD8bD8cD8dD8eCdddD91D92D93D94D95D96D97D98D99D9aD9bD9c
D9dD9eCdddCeeeCfffDalDa2Da3Da4Da5Da6Da7Da8Da9DaaDabDacDadDaeCfffDOOD01DO
2D03D04D05D06D07D08D09DOaDObDOcDOdDOeDOfD1OD 11D12D13D14D 15D16D17D18D
19D1aDlbD 1cDdDeDfD20D2fD30D3fD40D4fD50D5fD60D6fD70D7fD80D8fD90D9fDaO
DafDbODb lDb2Db3Db4Db5Db6Db7Db8Db9DbaDbbDbcDbdDbeDbfDcODclDc2Dc3Dc4Dc5
Dc6Dc7Dc8Dc9DcaDcbDccDcdDceDcfDdODd Dd2Dd3Dd4Dd5Dd6Dd7Dd8Dd9DdaDdbDdc
DddDdeDdfDeODelDe2De3De4De5De6De7De8De9DeaDebDecDedDeeDefDfODflDf2Df3Df
4Df5Df6Df7Df8Df9DfaDfbDfcDfdDfeDff' {
fitdirectory = getDirectory("Select a Directory") // prompts what directory to take images
from for processing
open(fitdirectory+"fitc.tif'); // opens the fitc image
fit = getImagelDO; // image ID of original fitc image
// fitdirectory = getDirectory("image") // this was used before "select a directory" (3 lines
up) was used
open(fitdirectory+"tritc.tif'); // opens tritc image
trit = getImagelDO; // image ID of original tritc image
run("Bandpass Filter...", "filterlarge=50 filtersmall=3 suppress=None tolerance=5 autoscale
saturate");
selectlmage(fit); // selectimage for processing
run("Bandpass Filter...", "filterlarge=50 filtersmall=3 suppress=None tolerance=5 autoscale
saturate");
run("RGB Merge...", "red=tritc.tif green=fitc.tif blue=*None* keep");
selectlmage(trit);
run("Tiff..", "save="+fitdirectory+"bandpass tritc.tif') // saves new bandpass image in
appropriate directory
close); // closes bandpass image
selectlmage(fit);
run("Tiff..", "save="+fitdirectory+"bandpass_fitc.tif') // saves new bandpass image in
appropriate directory
close); // closes bandpass image
makeRectangle(906, 691, 451, 343); // makes a rectangle selection in the bottom right
corner
run("To Selection"); // zooms to the selection just made
makeRectangle(0, 0, 0, 0); // removes selection
}

I End of Bandpass RGB Merge Tool
/////////////////////////////////////////////////////////////////////////////////////////////////////









// This macro shifts two images for overlay purpose either up/down or left/right
// Then merges images into RGB.

macro "Image Shift Action Tool -
COOOD12D13D14D15D16D17D18D22D28D32D38D42D48D52D55D56D57D58D59D5aD5bD
62D65D68D6bD72D75D78D7bD82D83D84D85D86D87D88D8bD95D9bDa5DabDb5DbbDc5
Dc6Dc7Dc8Dc9DcaDcbCOOC111C222C333C444C555C666C777C888C999CaaaCbbbCcccCd
ddCeeeCfffDOODO 1D02D03DO4DO5D06D07DO8D09DOaDObDOcDOdDOeDOfD 1 OD 1 1D 19D 1 a
D bD 1cD 1dD eD 1 fD20D21D23D24D25D26D27D29D2aD2bD2cD2dD2eD2fD30D31D33D34
D35D36D37D39D3aD3bD3cD3dD3eD3fD40D4 1D43D44D45D46D47D49D4aD4bD4cD4dD4
eD4fD50D51D53D54D5cD5dD5eD5fD60D61D63D64D66D67D69D6aD6cD6dD6eD6fD70D7
1D73D74D76D77D79D7aD7cD7dD7eD7fD80D81D89D8aD8cD8dD8eD8fD90D91D92D93D9
4D96D97D98D99D9aD9cD9dD9eD9fDaODalDa2Da3Da4Da6Da7Da8Da9DaaDacDadDaeDaf
DbODb 1Db2Db3Db4Db6Db7Db8Db9DDbaDbDbDbeDbfDcODc 1Dc2Dc3Dc4DccDcdDceDcf
DdODd Dd2Dd3Dd4Dd5Dd6Dd7Dd8Dd9DdaDdbDdcDddDdeDdfDeODelDe2De3De4De5De6
De7De8De9DeaDebDecDedDeeDefDfODflDf2Df3Df4Df5Df6Df7Df8Df9DfaDfbDfcDfdDfeDf
f'{
//Dialog box to get exact file names image.tif
requires("1.34m"); // make sure correct imagej version is running
if (nImages==0) // returns number of images open
print("No images are open");
else
imagesopen = newArray(nImages+l); //makes array size of images open
for(i=l; i<=nImages(; i++){ //for(initialize, limit, increment)
selectImage(i); //selects image as they are listed, -i=actual images, i=count of
images
imagesopen[i-1] = getTitle(; //returns title of images, arrays start at 0 which is why i-1
}
imagesopen[nImages]="*None*; // Makes a *None* selection so user can select this when no
change is needed
Dialog.create("Image Choice"); // create dialog box named Image choice
Dialog.addChoice(" TRITC image: ",imagesopen);
Dialog.addChoice("FITC image:", imagesopen, imagesopen[l]);
Dialog.addChoice("Image to move down", imagesopen); // drop down box of images open
Dialog.addNumber("Number of pixels to move down", 1); // drop down box of images open
Dialog.addChoice("Image to move left", imagesopen, imagesopen[nImages]); // drop down
box of images open, with default of second image available
Dialog.addNumber("Number of pixels to move left",1); // drop down box of images open
Dialog.show(; // show dialog box created
tritc = Dialog.getChoice(; // saves choice entered in first dialog box
fitc = Dialog.getChoiceo; // saves choice entered in second dialog box
movedown = Dialog.getChoiceo; // saves choice entered in third dialog box
moveleft = Dialog.getChoiceo; // saves choice entered in fouth dialog box, order
matters here
movedownnumber = Dialog.getNumbero; // saves number choice entered in first
number box









moveleftnumber = Dialog.getNumber();
//Check an image is selected for overlay
if (tritc== "*None*")
exit("Please select a red channel image");
if (fitc=="*None*")
exit("Please select a green channel image");
// Shift image down
if (movedown=="*None*")
q=1; //I don't know how to exit the for loop so put this useless
command here so if can do something
else{
selectImage(movedown); // selects the correct image for shifting
w=getWidth(; // gets image width
h=getHeight(+movedownnumber; // gets image height plus amount to move
run("Canvas Size...", "width="+w+" height="+h+" position=Bottom-Center zero");
if (movedown!=tritc){ // this if statement moves the other image up so an overlay
will work (images have to be same dimensions)
selectImage(tritc);
run("Canvas Size...", "width="+w+" height="+h+" position=Top-Center zero");}
else{
selectImage(fitc);
run("Canvas Size...", "width="+w+" height="+h+" position=Top-Center zero");
} }
// Shift image left (for description of what is going on below just look at descriptions for
"movedown" directly above, everything is about the same)
if (moveleft== "*None*")
q=1;
else{
selectImage(moveleft);
w=getWidth(+moveleftnumber;
h=getHeight);
run("Canvas Size...", "width="+w+" height="+h+" position=Center-Left zero");
if (moveleft!=tritc){
selectImage(tritc);
run("Canvas Size...", "width="+w+" height="+h+" position=Center-Right
zero");}
else{
selectImage(fitc);
run("Canvas Size...", "width="+w+" height="+h+" position=Center-Right zero");
} }
run("RGB Merge...", "red="+tritc+" green="+fitc+" blue=*None* keep");
}

//End of Image Shift Tool
/////////////////////////////////////////////////////////////////////////////////////////////////////









//BELOW THIS LINE ARE ADDITIONAL MACROS

macro "Purich 40X"{
run("Set Scale...", "distance=4.317 known=1 pixel=1 unit=[m global");
run(" Scale Bar...");
}

macro "Purich 60X"{
run("Set Scale...", "distance=6.4756 known=1 pixel=1 unit=[m global");
run(" Scale Bar...");
}

macro "Purich 100X"{
run("Set Scale...", "distance=10.8357 known=1 pixel=1 unit=[m global");
run(" Scale Bar...");
}

macro "Dickinson 100X"{
run("Set Scale...", "distance= 1.0762 known=1 pixel=l unit=[m global");
run(" Scale Bar...");
}

macro "Dickinson 150X"{
run(" Set Scale...", "distance= 16.6143 known= 1 pixel= 1 unit= m global");
run(" Scale Bar...");
}

I Duplicate line selection from one image to another
// Write down x, y, angle, length from original line selection to enter here

macro "Reproduce Line Selection" // Tool -
COOOC 111C222DOcDlbD2bD3aD49D59D68D77D87D96Da5Db5Dc4Dd3De3Df2C222C3333C
444C555C666C777C888C999CaaaCbbbCcccCdddCeeeCfffDOOD01D02D03D04D05D06D07D
08D09DOaDObDOdDOeDOfD10D11D12D13D14D15D16D17D18D19DlaDlcDldDleDlfD20D
21D22D23D24D25D26D27D28D29D2aD2cD2dD2eD2fD30D31D32D33D34D35D36D37D38
D39D3bD3cD3dD3eD3fD40D41D42D43D44D45D46D47D48D4aD4bD4cD4dD4eD4fD50D51
D52D53D54D55D56D57D58D5aD5bD5cD5dD5eD5fD60D61D62D63D64D65D66D67D69D6
aD6bD6cD6dD6eD6fD70D71D72D73D74D75D76D78D79D7aD7bD7cD7dD7eD7fD80D81D8
2D83D84D85D86D88D89D8aD8bD8cD8dD8eD8fD90D91D92D93D94D95D97D98D99D9aD
9bD9cD9dD9eD9fDaODalDa2Da3Da4Da6Da7Da8Da9DaaDabDacDadDaeDafDbDb lDb2Db
3Db4Db6Db7Db8Db9DbaDbbDbcDbdDbeDbfDcODc Dc2Dc3Dc5Dc6Dc7Dc8Dc9DcaDcbDcc
DcdDceDcfDd0DdlDd2Dd4Dd5Dd6Dd7Dd8Dd9DdaDdbDdcDddDdeDdfDeODelDe2De4De5
De6De7De8De9DeaDebDecDedDeeDefDfODflDf3Df4Df5Df6Df7Df8Df9DfaDfbDfcDfdDfeDf
f' {
{
x=getNumber("X value", 0) // enter x value from line made









y=getNumber("Y value", 0) // enter y value from line made
angle=getNumber("Angle value", 0) // enter angle value from line made
length=getNumber("Length value", 0) // enter length value from line made
xl = x-length*cos(angle*3.1415926/180) // calculate start x value from angle and length
yl = y+length*sin(angle*3.1415926/180) // calculate start y value from angle and length
makeLine(xl,yl,x,y) // create new line selection
}

/ End of Reproduce Line Selection Tool

I Duplicate rectangular selection onto an image
// Write down x, y, width, height from original rectangular selection to enter here

macro "Reproduce Rectangle Selection" // Tool -
COOOC 111C222D11D12D13D14D15D16D17D18D19DlaDlbDlcDldD21D2dD31D3dD41D4
dD51D5dD61D6dD71D7dD81D8dD91D98D99D9aD9bD9dDalDa8DabDadDb lDb8DbbDbdD
clDc8Dc9DcaDcbDcdDdlDddDelDe2De3De4De5De6De7De8De9DeaDebDecDedC222C333
C444C555C666C777C888C999CaaaCbbbCcccCdddCeeeCfffDOOD01D02D03D04D05D06D07
D08D09DOaDObDcDOdDOeDOfD 1OD 1 eD 1 fD2D22D23D24D25D26D27D28D29D2aD2bD2c
D2eD2fD30D32D33D34D35D36D37D38D39D3aD3bD3cD3eD3fD40D42D43D44D45D46D4
7D48D49D4aD4bD4cD4eD4fD50D52D53D54D55D56D57D58D59D5aD5bD5cD5eD5fD60D6
2D63D64D65D66D67D68D69D6aD6bD6cD6eD6fD70D72D73D74D75D76D77D78D79D7aD
7bD7cD7eD7fD80D82D83D84D85D86D87D88D89D8aD8bD8cD8eD8fD90D92D93D94D95D
96D97D9cD9eD9fDaODa2Da3Da4Da5Da6Da7Da9DaaDacDaeDafDbODb2Db3Db4Db5Db6Db
7Db9DbaDbcDbeDbfDcODc2Dc3Dc4Dc5Dc6Dc7DccDceDcfDdODd2Dd3Dd4Dd5Dd6Dd7Dd8
Dd9DdaDdbDdcDdeDdfDeODeeDefDfODflDf2Df3Df4Df5Df6Df7Df8Df9DfaDfbDfcDfdDfeDf
f'{
x=getNumber("X value", 0) {/enter x value from previous rect selection
y=getNumber("Y value", 0) // enter y value from previous rect selection
width=getNumber("Width value", 680) / enter width value from previous rect selection
height=getNumber("Height value", 518) // enter height value from previous rect selection
makeRectangle(x, y, width, height) //make new rectangle selection
}

/ End of Reproduce Rectangle Selection Tool

This macro asks what image to use. You then draw a line selection (straight, sectioned, or
freehand line) across the area of interest. Once done close the blank image causing the macro to
continue. The values along the selection are extracted and the maximum value for each color is
found. Each value is divided by the maximum normalizing each color to it's maximum. The
data is saved in three files labeled with the ROI# then either fitcdata.txt, tritcdata.txt, or
linecoordsdata.txt (so the line could be reproduced later). A graph of the data is also displayed
and all images are closed.

I Close "Log" window so errant information isn't saved with the data









if (isOpen("Log")) {
selectWindow("Log");
run("Close");
}
// Ask for directory and ROI for that directory
numberofROI=0;
linescansdone=0;
directory = getDirectory(" Select a Directory"); // asks for directory
// So I don't forget what directory I just worked on
noEndSlash = substring(directory, 0,lengthOf(directory)-1);
lastind = lastIndexOf(noEndSlash, "/");
justDirectoryNumber = substring(noEndSlash, lastind+l,lengthOf(noEndSlash));
requires("1.38m");
title = "Text Window";
title = titleitlel+"];
if (isOpen(titlel)==false)
run("New... ", "name="+title2+" type=[Text File] width=15 height=30");
print(title2,"\n"+justDirectoryNumber);
// End of: So I don't forget what directory I just worked on
filedirect = directory+"/"; // saves a forward slash with directory for ease of
use
list = getFileList(directory); // gets list of files in directory selected
for (i = 0; i < list.length; i++) {
for (j=1; j if (startsWith(list[i], j))
numberofROI = j; // saves number of file with largest number
if (startsWith(list[i], j+"linecoordsdata.txt"))
linescansdone = j; // saves number of linecoordsdata.txt with
largest number and adds one
} }
if (numberofROI==0) // checks if there is something to analyze in this directory
exit("No regions to analyze in this directory"); // if not then exits macros
if (numberofROI==linescansdone) // checks if there is something to analyze in
this directory
exit("All regions have been analyzed in this directory"); // if not then exits macros
linescansdone=linescansdone+l; // advances ROI to scan by one so next will be
selected
Dialog.create("Image Choice"); // create dialog box named Image choice
Dialog.addMessage("Create a plot of intensities."); // general message with instructions
Dialog.addMessage("There appears to be "+numberofROI+"\nregions to analyze for this set");
// general message with instructions
Dialog.addNumber("Event to plot ",linescansdone); // Form to enter a number
Dialog.addCheckbox(" Cycle through all ROI for this directory (starting at ROI number
entered)?", true) // check box
Dialog.show(; // show dialog box created
event = Dialog.getNumber(; // stores entered string









allROI = Dialog.getCheckbox);
if (allROI==false)
numberofROI=event;
for (j=event; j<=numberofROI; j++){
open(directory+j+"RGB.tif');
rgbid = getImageID();


window


// open overlay image
// save image ID


// Zooms into image so easier to trace filament
makeRectangle(0, 0, 15, 15);
run("To Selection"); // zooms 1
v)
run("Out"); //backs o
run("Out");
run("Out");
run(" Select None"); // remove


o selection made (also enlarges image


ff zoom


s selection


newImage("Close Me When Done Selecting Line", "8-bit White", 16, 16, 1); //
dummy image, closed when want to continue
dummylmage=getImageID();
run("Tile"); // tiles images
setTool(5); // selects the sectioned line tool
selectlmage(rgbid); // make sure rgb image is focused
while (isOpen(dummylmage)) { //this pauses the macro until the
small tritc image is closed
wait(10);
}
selectlmage(rgbid); // focus rgb image
// Reproduces line drawn onto all open images
type = selectionType);
if (type==-l)
print("No selection");
else


getSelectionCoordinates(x, y)
run("RGB Split");
close);
makeSelection(type,x,y);
// Getting the profile of the FITC sigi
// print("FITC DATA");
grace doesn't have a problem
profile = getProfile);
profileMax = 0;
normalizedprofile = newArray(profil
for (i=0; i if (profile[i]>profileMax)
profileMax = profile[i
}


; // get line coordinates
// split image into red, green, blue
// close blue image
// make line on green image
lal
// took this out so plotting program

// gets data of each value along selection
// initializes variable
e.length); // initializes array


1: //finds max









for (i=0; i normalizedprofile[i] = profile[i]/profileMax; // normalizes each value
print(i+" "+normalizedprofile[i]); // prints to log for saving
}
// print("%%%%%");
// print("Green Max "+profileMax);
//print("%%%%%%%%%%");
close); // closes green image
// Plot profile
Plot.create("Profile", "Tail to Bead Position", "Normalized Intensity", normalizedprofile);
// plots green data
selectWindow("Log"); // selects log window
saveAs("Text", directory+j+"fitcdata.txt"); // saves data in log window
selectWindow("Log");
run("Close");
// print("TRITC DATA");
// Getting the profile of the TRITC signal
makeSelection(type,x,y);
profile = getProfile();
profileMax = 0;
normalizedprofile = newArray(profile.length);
for (i=0; i if (profile[i]>profileMax)
profileMax = profile[i];


}
for (i


=0; i normalizedprofile[i] = profile[i]/profileMax;
print(i+" "+normalizedprofile[i]);


}
// print("%%%%%");
// print("Red Max "+profileMax);
//print("%%%%%%%%%%");
// Plot profile
Plot.setColor("red");
Plot.add("line", normalizedprofile);
Plot.setColor("green");
Plot.update();
seen on screen (needed so save will work lat
close);
selectWindow("Log");
saveAs("Text", directory+j+"tritcdat,
selectWindow("Log");
run("Close");
// print(" SELECTION COORDINAI
//print(" x y");
for (i=0; i

// sets next call to plot as red
// adds red data to plot
// sets default color of plot
// updates graph so the plot is drawn and
er)
// closes red image
// selects log window
.txt"); // saves data in log window









print(x[i]+" "+y[i]);
}
selectWindow("Log");
saveAs("Text", directory+j+"linecoordsdata.txt");
selectWindow("Log");
run("Close");

saveAs("Tiff', directory+j+"imagejgraph.tif');
wait(2000);


close);


// saves data in log window


// save graph


// close graph


This macro asks for a directory to work with. Finds all the 16-bit images of interest (the color
change images) Plots the predetermined line on the 16-bit images and gets the pixel data and
plots the values saving everything.

basedir = getDirectory("Select a Directory") // prompts what directory to take images from for
processing
//setB atchMode(true);
for (i = 1; i <=40; i++) { // changes directory
if (i<=9)


dir = basedir+"0"+i+"/";
if (i>9)


// sets directory


dir = basedir+i+"/"; // sets directory
list = getFileList(dir); // gets list of files for working directory
if (list.length==0){ // checks if no files found
print(" Stopped at directory "+dir);
exit // stops code
}


for (k = 0; k < list.length; k++) {
for (j=l; j if (startsWith(list[k], j+"linecoord")) {
open(dir+j+"bandpassfitc.tif');
linemaker);
normalizedprofile = plotter();


//16-bit fitc
// calls function to reproduce line


close);
Plot.create("Profile", "Tail to Bead Position", "Normalized Intensity");
// plots green data
Plot.setLimits(0, normalizedprofile.length-1, 0, 1); // set plot dimensions


(xmin,xmax,ymin,ymax)
Plot.setLineWidth(2);
Plot. setColor("green");
Plot.add("line", normalizedprofile);
selectWindow("Log");


// sets default color of plot
// adds red data to plot
// selects log window









saveAs("Text", dir+j+"fitcdata-16bit.txt");
window
selectWindow("Log");
run("Close");
open(dir+j+"bandpass tritc.tif'); //
linemaker(; // calls fu
normalizedprofile = plotter();
close);
Plot.setLineWidth(2);
Plot.setColor("red"); // sets nex

Plot.add("line", normalizedprofile); //
Plot.update(; // updates
and seen on screen (needed so save will work later)
selectWindow("Log"); //
saveAs("Text", dir+j+"tritcdata-16bit.txt");
window
selectWindow("Log");
run("Close");
saveAs("Tiff", dir+j +"imagej graph- 16bit.tif');
close); // close graph

}
//setB atchMode(false);
function linemaker(){
coords=File.openAsString(dir+j+"linecoordsdata.txt"); //
coordinates
coordarray = split(coords); // splits values in
xcoord = newArray(coordarray.length/2); // initiated
ycoord = newArray(coordarray.length/2); // initiated
for(i=0;i coordarray[i] = parselnt(coordarray[i]); // convert
for(i=0;i xcoord[i] = coordarray[i*2]; // saves x and y c
ycoord[i] = coordarray[(i+l)*2-1];
}
makeSelection(6, xcoord, ycoord); // reproduces lin
} // end of function
function plotter(){
profile = getProfileo; // gets data of ea
profileMax = 0; // initializes varia
profileMin = profile[0];
normalizedprofile = newArray(profile.length); //
for (i=0; i if (profile[i]>profileMax)
profileMax = profile[i]; // finds m


// saves data in log


16-bit trite
nation to reproduce line



t call to plot as red

adds red data to plot
graph so the plot is drawn

selects log window
// saves data in log



// save graph





opens saved line

to string array
s array
s array

:s from string to integer

:oordinates separately


e


;h value along selection
ible

initializes array


ax









if (profile[i] profileMin = profile[i]; // finds min
}
for (i=0; i profile[i] = profile[i] profileMin; // subtracts background
normalizedprofile[i] = profile[i]/(profileMax-profileMin); // normalizes each
value
print(i+" "+normalizedprofile[i]); // prints to log for saving
}
return normalizedprofile;
}

I This macro opens the bandpass images of each directory, sets a threshold so mostly
// beads are seen and then counts the beads and saves the data.
// Note: For some reason batch mode doesn't work here because of the result window that
// is displayed. It doesn't close properly.

fitdirectory = getDirectory("Select a Directory") // prompts what directory to take images
from for processing
setBatchMode(true);
for (i = 1; i <=40; i++) { // changes directory
if (i<=9)
listdirect = fitdirectory+"0"+i+"/"; // sets directory
if (i>9)
listdirect = fitdirectory+i+"/"; // sets directory
list = getFileList(listdirect); // gets list of files for working directory
if (list.length!=0){ // checks if no files found
ROIhere=0;
for (j=l; j if (startsWith(listj ], "1"))
ROIhere = 1; //checks that an ROI is in this directory, if so
continue, else go to next iteration
}
if (ROIhere==l){
open(listdirect+"bandpassfitc.tif'); // open fitc image
open(listdirect+"bandpass tritc.tif'); // open fitc image
run("RGB Merge...", "red=bandpass tritc.tif green=bandpassfitc.tif
blue=*None*");
x=getWidth();
y=getHeight();
// Note: the following only works on 24-bit RGB images
for (m=0; m for (n=0; n v = getPixel(m,n);
red = (v>>16)&0xff; // extract red byte (bits 23-17)
green = (v>>8)&0xff; // extract green byte (bits 15-8)









blue = v&0xff; // extract blue byte (bits 7-0)
if ((red < 190) || (green < 190)) // I is or making only
yellow show, && is and making bright red and green show as well as yellow
setPixel(m,n,0); // set non yellow pixels as
black
}
}
run("Tiff..", "save="+listdirect+"RGB_bead_count.tif");
run(" 8-bit");
setThreshold(1, 255);
run("Convert to Mask");
rename("Directory_"+i);
// Since the image is thresholded to only display yellow, those spots
remaining are
// considered to be beads so there is no size limitation or circularity limit.
run("Analyze Particles...", "size=4-Infinity circularity=0-1.00
show=Nothing clear summarize");
close);
} } }
while(nImages>0){
close);
}
selectWindow(" Summary");
saveAs("Text", fitdirectory+"bead_count.txt");
//run("Text...", "save="+fitdirectory+"bead_count.txt"); // saves results
run("Close"); // closes results
setBatchMode(false);

//BELOW THIS LINE ARE MACROS SPECIFICALLY MADE FOR SHORTCUT KEYS

//Get coordinates of a rectangular selection or a line selection and lists in log box
//Use this in conjunction with the "Reproduce Rectangle Selection (Cut Paste) Tool"

macro "Get Selection Coord Action [y]"{
type = selectionType();
if (type==-l)
print("No selection");
else {
if (type==5){
getLine(xl, yl, x2, y2, lineWidth);
length=sqrt((xl-x2)*(xl-x2)+(y -y2)*(y 1-y2));
if (y2>yl)
angle=-acos((x2-xl)/length)*180/PI;
else
angle=acos((x2-xl)/length)* 180/PI;
print("Line,x 1,yl,x2,y2,length,angle,",x 1,yl,x2,y2,length,angle);










if (type==0){
getBoundingRect(x, y, w, h);
print("Rectangle, x, y, width, height,",x,y,w,h);
}}
restorePreviousTool;
//setTool(O); // was used to switch to rectangle selection tool but line above seems more
appropriate
}

I End of Get Selection Coord Tool

I This macro uses RGB Merge... to merge two images and create a RGB image
// This macro assumes the first image opened (first image in the window list) is the red image

macro "Auto RGB Merge Action [q]"{
selectlmage(1);
tritc = getTitle);
selectlmage(2);
fitc = getTitle);
run("RGB Merge...", "red="+tritc+" green="+fitc+" blue=*None* keep");
}

I End of "Auto RGB Merge Action"
//////////////////////////////////////////////////////////////////////////////////////////////////////









LIST OF REFERENCES


1. Ridley, A., M. Peckham, and P. Clark. 2004. Cell motility: From molecules to organisms.
Wiley, Hoboken, NJ.

2. Lodish, H., A. Berk, P. Matsudaira, C. A. Kaiser, M. Krieger, M. P. Scott, L. Zipursky,
and J. Damell. 2004. Molecular cell biology. W.H. Freeman and Company, New York.

3. Graceffa, P., and R. Dominguez. 2003. Crystal structure of monomeric actin in the ATP
state: Structural basis of nucleotide-dependent actin dynamics. J. Biol. Chem.
278:34172-34180.

4. Janmey, P. A., S. Hvidt, G. F. Oster, J. Lamb, T. P. Stossel, and J. H. Hartwig. 1990.
Effect of ATP on actin filament stiffness. Nature 347:95-99.

5. Gittes, F., B. Mickey, J. Nettleton, and J. Howard. 1993. Flexural rigidity of microtubules
and actin filaments measured from thermal fluctuations in shape. J. Cell Biol.
120:923-934.

6. Isambert, H., P. Venier, A. C. Maggs, A. Fattoum, R. Kassab, D. Pantaloni, and M. F.
Carlier. 1995. Flexibility of actin filaments derived from thermal fluctuations. Effect of
bound nucleotide phalloidin and muscle regulatory proteins. J. Biol. Chem.
270:11437-11444.

7. Ott, A., M. Magnasco, A. Simon, and A. Libchaber. 1993. Measurement of the
persistence length of polymerized actin using fluorescence microscopy. Phys. R. E. Stat.
Phy. Plasmas Fluids and Rel. Inter. Topics 48:R1642-1645.

8. Alberts, B., A. Johnson, J. Lewis, M. Raff, K. Roberts, and P. Walter. 2002. Molecular
biology of the cell. Garland Science, New York.

9. Bray, D. 2001. Cell movements: From molecules to motility. Garland Publishing, New
York.

10. Dickinson, R. B., F. S. Southwick, and D. L. Purich. 2002. A direct-transfer
polymerization model explains how the multiple profilin-binding sites in the actoclampin
motor promote rapid actin-based motility. Arch. Biochem. Biophys. 406:296-301.

11. Carlier, M. F., V. Laurent, J. Santolini, R. Melki, D. Didry, G. X. Xia, Y. Hong, N. H.
Chua, and D. Pantaloni. 1997. Actin depolymerizing factor (ADF/cofilin) enhances the
rate of filament turnover: Implication in actin-based motility. J CellBiol 136:1307-1322.

12. Southwick, F. S., and D. L. Purich. 1998. Listeria and .\lNgel//al actin-based motility in
host cells. Trans. Am. Clin. Climatol. Assoc. 109:160-172.









13. Marchand, J. B., P. Moreau, A. Paoletti, P. Cossart, M. F. Carlier, and D. Pantoloni.
1995. Actin-based movement of Listeria monocytogenes: Actin assembly results from the
local maintenance of uncapped filament barbed ends at the bacterium surface. J. Cell
Biol. 130:331-343.

14. Sanger, J. M., J. W. Sanger, and F. S. Southwick. 1992. Host cell actin assembly is
necessary and likely to provide the propulsive force for intracellular movement of
Listeria monocytogenes. Infect. Immun. 60:3609-3619.

15. Smith, G. A., D. A. Portnoy, and J. A. Theriot. 1995. Asymmetric distribution of the
Listeria monocytogenes ActA protein is required and sufficient to direct actin-based
motility. Mol. Microbiol. 17:945-951.

16. Southwick, F. S., and D. L. Purich. 1994. Arrest of Listeria movement in host cells by a
bacterial ActA analogue: Implications for actin-based motility. Proc. Natl. Acad. Sci.
U.S.A. 91:5168-5172.

17. Southwick, F. S., and D. L. Purich. 1995. Inhibition of Listeria locomotion by mosquito
oostatic factor, a natural oligoproline peptide uncoupler of profilin action. Infect. Immun.
63:182-190.

18. Brundage, R. A., G. A. Smith, C. A., J. A. Theriot, and D. A. Portnoy. 1993. Expression
and phosphorylation of the Listeria monocytogenes ActA protein in mammalian cells.
Proc. Natl. Acad. Sci. U.S.A. 90:11890-11894.

19. Kocks, C., R. Hellio, P. Gounon, H. Ohayon, and P. Cossart. 1993. Polarized distribution
ofListeria monocytogenes surface protein ActA at the site of directional actin assembly.
J. Cell Sci. 105:699-710.

20. Domann, E., J. Wehland, M. Rohde, S. Pistor, M. Hartl, W. Goebel, M.
Leimeister-Wachter, M. Wuenscher, and T. Chakraborty. 1992. A novel bacterial
virulence gene in Listeria monocytogenes required for host cell microfilament interaction
with homology to the proline-rich region of vinculin. Embo. J. 11:1981-1990.

21. Gertler, F. B., K. Niebuhr, M. Reinhard, J. Wehland, and P. Soriano. 1996. Mena, a
relative of VASP and Drosophila Enabled, is implicated in the control of microfilament
dynamics. Cell 87:227-239.

22. Dickinson, R. B., and D. L. Purich. 2002. Clamped-filament elongation model for
actin-based motors. Biophys. J 82:605-617.

23. Mimuro, H., T. Suzuki, S. Suetsugu, H. Miki, T. Takenawa, and C. Sasakawa. 2000.
Profilin is required for sustaining efficient intra- and intercellular spreading of.\/nge//t
flexneri. J. Biol. Chem. 275:28893-28901.









24. Frischknecht, F., V. Moreau, S. Rottger, S. Gonfloni, I. Reckmann, G. Superti-Furga, and
M. Way. 1999. Actin-based motility of vaccinia virus mimics receptor tyrosine kinase
signalling. Nature 401:926-929.

25. Chereau, D., F. Kerff, P. Graceffa, Z. Grabarek, K. Langsetmo, and R. Dominguez. 2005.
Actin-bound structures of Wiskott-Aldrich syndrome protein (WASP)-homology domain
2 and the implications for filament assembly. Proc. Natl. Acad. Sci. U. S. A.
102:16644-16649.

26. Chereau, D., and R. Dominguez. 2006. Understanding the role of the G-actin-binding
domain of Ena/VASP in actin assembly. J. Struct. Biol.

27. Kelly, A. E., H. Kranitz, V. Dotsch, and R. D. Mullins. 2005. Actin binding to the central
domain of WASP/Scar proteins plays a critical role in the activation of the Arp2/3
complex. J. Biol. Chem. 281:10589-10597.

28. Beckerle, M. C. 1998. Spatial control of actin filament assembly: Lessons from Listeria.
Cell 95:741-748.

29. Halliburton, W. D. 1887. On muscle plasma. J. Physiol. 8:133-202.

30. Perry, S. V. 2003. When was actin first extracted from muscle? J. Mus. R. Cell
Mot. :597-599.

31. Straub, F. B. 1942. Actin. Stud. Inst. Med. Chem. Univ. Szeged 2:1-15.

32. Cameron, L. A., M. J. Footer, A. van Oudenaarden, and J. A. Theriot. 1999. Motility of
ActA protein-coated microspheres driven by actin polymerization. Proc. Natl. Acad. Sci.
U.S.A. 96:4908-4913.

33. Cameron, L. A., T. M. Svitkina, D. Vignjevic, J. A. Theriot, and G. G. Borisy. 2001.
Dendritic organization of actin comet tails. Curr. Biol. 11:130-135.

34. Schwartz, I. M., M. Ehrenberg, M. Bindschadler, and J. L. McGrath. 2004. The role of
substrate curvature in actin-based pushing forces. Curr. Biol. 14:1094-1098.

35. McGrath, J. L., N. J. Eungdamrong, C. I. Fisher, F. Peng, L. Mahadevan, T. J. Mitchison,
and S. C. Kuo. 2003. The force-velocity relationship for the actin-based motility of
Listeria monocytogenes. Curr. Biol. 13:329-332.

36. Wiesner, S., E. Helfer, D. Didry, G. Ducouret, F. Lafuma, M. F. Carlier, and D.
Pantaloni. 2003. A biomimetic motility assay provides insight into the mechanism of
actin-based motility. J. CellBiol. 160:387-398.









37. Bemheim-Groswasser, A., S. Wiesner, R. M. Golsteyn, M. F. Carlier, and C. Sykes.
2002. The dynamics of actin-based motility depend on surface parameters. Nature
417:308-311.

38. Bemheim-Groswasser, A., J. Prost, and C. Sykes. 2005. Mechanism of actin-based
motility: A dynamic state diagram. Biophys. J. 89:1411-1419.

39. Marcy, Y., J. Prost, M. F. Carlier, and C. Sykes. 2004. Forces generated during
actin-based propulsion: A direct measurement by micromanipulation. Proc. Natl. Acad.
Sci. U.S.A. 101:5992-5997.

40. Upadhyaya, A., J. R. Chabot, A. Andreeva, A. Samadani, and A. van Oudenaarden. 2003.
Probing polymerization forces by using actin-propelled lipid vesicles. Proc. Natl. Acad.
Sci. U.S.A. 100:4521-4526.

41. Giardini, P. A., D. A. Fletcher, and J. A. Theriot. 2003. Compression forces generated by
actin comet tails on lipid vesicles. Proc. Natl. Acad. Sci. U.S.A. 100:6493-6498.

42. Boukellal, H., O. Campas, J.-F. Joanny, J. Prost, and C. Sykes. 2004. Soft Listeria:
Actin-based propulsion of liquid drops. Phys. Rev. E. Stat. Nonlin. Soft Matter Phys.
69:061906.

43. Dickinson, R. B., and D. L. Purich. 2006. Diffusion rate limitations in actin-based
propulsion of hard and deformable particles. Biophys. J. 91:1548-1563.

44. Parekh, S. H., O. Chaudhuri, J. A. Theriot, and D. A. Fletcher. 2005. Loading history
determines the velocity of actin-network growth. Nat. CellBiol. 7:1119-1123.

45. Mogilner, A., and G. Oster. 1996. Cell motility driven by actin polymerization. Biophys.
J. 71:3030-3045.

46. Mogilner, A., and G. Oster. 2003. Force generation by actin polymerization II: The
elastic ratchet and tethered filaments. Biophys. J 84:1591-1605.

47. Carlsson, A. E. 2003. Growth velocities of branched actin networks. Biophys. J.
84:2907-2918.

48. Hill, T. L., and M. W. Kirschner. 1982. Subunit treadmilling of microtubules or actin in
the presence of cellular barriers: Possible conversion of chemical free energy into
mechanical work. Proc. Natl. Acad. Sci. U.S.A. 79:490-494.

49. Hill, T. L. 1981. Microfilament or microtubule assembly or disassembly against a force.
Proc. Natl. Acad. Sci. U.S.A. 78:5613-5617.

50. Kuo, S. C., and J. L. McGrath. 2000. Steps and fluctuations of Listeria monocytogenes
during actin-based motility. Nature 407:1026-1029.









51. Gerbal, F., V. Laurent, A. Ott, M. F. Carlier, P. Chaikin, and J. Prost. 2000. Measurement
of the elasticity of the actin tail ofListeria monocytogenes. Eur. Biophys. J. 29:134-140.

52. Pantaloni, D., R. Boujemaa, D. Didry, P. Gounon, and M. F. Carlier. 2000. The Arp2/3
complex branches filament barbed ends: Functional antagonism with capping proteins.
Nat. Cell Biol. 2:385-391.

53. Amann, K. J., and T. D. Pollard. 2001. The Arp2/3 complex nucleates actin filament
branches from the sides of pre-existing filaments. Nat. Cell Biol. 3:306-310.

54. Amann, K. J., and T. D. Pollard. 2001. Direct real-time observation of actin filament
branching mediated by Arp2/3 complex using total internal reflection fluorescence
microscopy. Proc. Natl. Acad. Sci. U.S.A. 98:15009-15013.

55. Ichetovkin, I., W. Grant, and J. Condeelis. 2002. Cofilin produces newly polymerized
actin filaments that are preferred for dendritic nucleation by the Arp2/3 complex. Curr.
Biol. 12:79-84.

56. Carlsson, A. E. 2001. Growth of branched actin networks against obstacles. Biophys. J.
81:1907-1923.

57. Mogilner, A., and G. Oster. 2003. Polymer motors: Pushing out the front and pulling up
the back. Curr. Biol. 13:R721-733.

58. Dickinson, R. B., L. Caro, and D. L. Purich. 2004. Force generation by cytoskeletal
filament end-tracking proteins. Biophys. J 87:2838-2854.

59. Laine, R. O., K. L. Phaneuf, C. C. Cunningham, D. Kwiatkowski, T. Azuma, and F. S.
Southwick. 1998. Gelsolin, a protein that caps the barbed ends and severs actin filaments,
enhances the actin-based motility ofListeria monocytogenes in host cells. Infect. Immun.
66:3775-3782.

60. Bear, J. E., T. M. Svitkina, M. Krause, D. A. Schafer, J. J. Loureiro, G. A. Strasser, I. V.
Maly, O. Y. Chaga, J. A. Cooper, G. G. Borisy, and F. B. Gertler. 2002. Antagonism
between Ena/VASP proteins and actin filament capping regulates fibroblast motility. Cell
109:509-521.

61. Barzik, M., T. I. Kotova, H. N. Higgs, L. Hazelwood, D. Hanein, F. B. Gertler, and D. A.
Schafer. 2005. Ena/VASP proteins enhance actin polymerization in the presence of
barbed end capping proteins. J. Biol. Chem. 280:28653-28662.

62. Howard, J. 2001. Mechanics of motor proteins and the cytoskeleton. Sinauer, Sunderland,
MA.

63. Dickinson, R. B., and D. L. Purich. 2007. Models for actin polymerization and force
generation by cytoskeletal filament end-tracking motors. J. Math. Biol. Under Review.









64. Boal, D. 2002. Mechanics of the cell. Cambridge University Press, Cambridge.

65. Kas, J., H. Strey, J. X. Tang, D. Finger, R. Ezzell, E. Sackmann, and P. A. Janmey. 1996.
F-actin, a model polymer for semiflexible chains in dilute, semidilute, and liquid
crystalline solutions. Biophys. J. 70:609-625.

66. Landau, L. D., and E. M. Lifshitz. 1986. Theory of elasticity. Reed Educational and
Professional Publishing Ltd., New York.

67. Doi, M., and S. F. Edwards. 1986. The theory of polymer dynamics. Clarendon Press;
Oxford University Press, Oxford Oxfordshire; New York.

68. Noireaux, V., R. M. Golsteyn, E. Friederich, J. Prost, C. Antony, D. Louvard, and C.
Sykes. 2000. Growing an actin gel on spherical surfaces. Biophys. J 78:1643-1654.

69. Loisel, T. P., R. Boujemaa, D. Pantaloni, and M. F. Carlier. 1999. Reconstitution of
actin-based motility of Listeria and .\l/gel//a using pure proteins. Nature 401:613-616.

70. Lasic, D. D. 1997. Liposomes in gene delivery. CRC Press LLC, Boca Raton, FL.

71. Plastino, J., I. Lelidis, J. Prost, and C. Sykes. 2004. The effect of diffusion,
depolymerization and nucleation promoting factors on actin gel growth. Eur. Biophys. J.
33:310-320.

72. Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of microgram
quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem.
72:248-254.

73. Pardee, J. D., and J. A. Spudich. 1982. Purification of muscle actin. Methods Enzymol.
85 Pt B:164-181.

74. Kuhn, J. R., and T. D. Pollard. 2005. Real-time measurements of actin filament
polymerization by total internal reflection fluorescence microscopy. Biophys. J.
88:1387-1402.

75. Haugland, R. P. 1995. Coupling of monoclonal antibodies with fluorophores. Methods
Mol. Biol. 45:205-221.

76. Welch, M. D., J. Rosenblatt, J. Skoble, D. A. Portnoy, and T. J. Mitchison. 1998.
Interaction of human Arp2/3 complex and the Listeria monocytogenes ActA protein in
actin filament nucleation. Science 281:105-108.

77. Lasic, D. D. 1988. The mechanism of vesicle formation. Biochem. J. 256:1-11.

78. Hub, H. H., U. Zimmermann, and H. Ringsdorf. 1982. Preparation of large unilamellar
vesicles. FEBSLett. 140:254-256.









79. Khan, L. A., M. Raj, and M. Amin. 1989. Coupling of the enzymic activities of myosin
ATPase and creatine kinase and its role in muscular contraction. Indian J Biochem.
Biophys. 26:148-152.

80. Cameron, L. A., J. R. Robbins, M. J. Footer, and J. A. Theriot. 2004. Biophysical
parameters influence actin-based movement, trajectory, and initiation in a cell-free
system. Mol. Biol. Cell 15:2312-2323.

81. Bozzola, J. J. 1999. Electron microscopy: Principles and techniques for biologists. Jones
and Bartlett, Sudbury, Mass.

82. Goodhew, P. J. 2001. Electron microscopy and analysis. Taylor & Francis, New York.

83. Rosenblatt, J., B. Agnew, H. Abe, J. Bamburg, and T. Mitchison. 1997. Xenopus actin
depolymerizing factor/cofilin (XAC) is responsible for the turnover of actin filaments in
Listeria monocytogenes tails. J CellBiol. 136:1323-1332.

84. Zeile, W. L., F. Zhang, R. B. Dickinson, and D. L. Purich. 2005. Listeria's right-handed
helical rocket-tail trajectories: Mechanistic implications for force generation in
actin-based motility. CellMotil. Cytoskeleton 60:121-128.

85. Cicchetti, G., P. Maurer, P. Wagener, and C. Kocks. 1999. Actin and phosphoinositide
binding by the ActA protein of the bacterial pathogen Listeria monocytogenes. J Biol.
Chem. 274:33616-33626.

86. Hartman, J. J., and R. D. Vale. 1999. Microtubule disassembly by ATP-dependent
oligomerization of the AAA enzyme katanin. Science 286:782-785.

87. Svitkina, T. M., and G. G. Borisy. 1998. Correlative light and electron microscopy of the
cytoskeleton of cultured cells. Methods Enzymol. 298:570-592.

88. Butler, J. E., and E. S. G. Shaqfeh. 2002. Dynamic simulations of the inhomogeneous
sedimentation of rigid fibres. J. FluidMech. 468:205-237.

89. Harbeck, B., S. Hittelmaier, K. Schluter, B. M. Jockusch, and S. Illenberger. 2000.
Phosphorylation of the vasodilator-stimulated phosphoprotein regulates its interaction
with actin. J. Biol. Chem. 275:30817-30825.

90. Brown, S. S., and J. A. Spudich. 1979. Nucleation of polar actin filament assembly by a
positively charged surface. J. CellBiol. 80:499-504.

91. Kovar, D. R., and T. D. Pollard. 2004. Insertional assembly of actin filament barbed ends
in association with formins produces piconewton forces. Proc. Natl. Acad. Sci. U.S.A.
101:14725-14730.









92. Romero, S., C. Le Clainche, D. Didry, C. Egile, D. Pantaloni, and M. F. Carlier. 2004.
Formin is a processive motor that requires profilin to accelerate actin assembly and
associated ATP hydrolysis. Cell 119:419-429.

93. Saito, K., M. Tokunaga, A. H. Iwane, and T. Yanagida. 1997. Dual-colour microscopy of
single fluorophores bound to myosin interacting with fluorescently labelled actin using
anti-Stokes fluorescence. J. Microsc. 188:255-263.

94. Benesch, S., S. Polo, F. P. Lai, K. I. Anderson, T. E. Stradal, J. Wehland, and K. Rottner.
2005. N-WASP deficiency impairs EGF internalization and actin assembly at
clathrin-coated pits. J. Cell Sci. 118:3103-3115.

95. Pollard, T. D., L. Blanchoin, and R. D. Mullins. 2000. Molecular mechanisms controlling
actin filament dynamics in nonmuscle cells. Annu. Rev. Biophys. Biomol. Struct.
29:545-576.

96. Rossoll, W., S. Jablonka, C. Andreassi, A. K. Kroning, K. Karle, U. R. Monani, and M.
Sendtner. 2003. Smn, the spinal muscular atrophy-determining gene product, modulates
axon growth and localization of beta-actin mRNA in growth cones of motoneurons. J
CellBiol 163:801-812.

97. Gov, N. S., and A. Gopinathan. 2006. Dynamics of membranes driven by actin
polymerization. Biophys. J. 90:454-469.

98. Chodniewicz, D., and D. V. Zhelev. 2003. Novel pathways of F-actin polymerization in
the human neutrophil. Blood 102:2251-2258.

99. Oude Weernink, P. A., P. Schulte, Y. Guo, J. Wetzel, M. Amano, K. Kaibuchi, S.
Haverland, M. Voss, M. Schmidt, G. W. Mayr, and K. H. Jakobs. 2000. Stimulation of
phosphatidylinositol-4-phosphate 5-kinase by Rho-kinase. JBiol Chem
275:10168-10174.

100. Okamoto, Y., and T. Sekine. 1985. A streamlined method of subfragment one preparation
from myosin. J. Biochem. Tokyo 98:1143-1145.

101. Chan, B. P., A. Chilkoti, W. M. Reichert, and G. A. Truskey. 2003. Effect of streptavidin
affinity mutants on the integrin-independent adhesion of biotinylated endothelial cells.
Biomaterials 24:559-570.

102. Faulstich, H., S. Zobeley, U. Bentrup, and B. M. Jockusch. 1989. Biotinylphallotoxins:
Preparation and use as actin probes. J. Histochem. Cytochem. 37:1035-1045.

103. Jahn, W. 2007. The Association of Actin and Myosin in the Presence of
gamma-Amido-ATP Proceeds Mainly via a Complex with Myosin in the Closed
Conformation. Biochemistry 46:9654-9664.









104. Sakamaki, J., H. Honda, E. Imai, K. Hatori, K. Shimada, and K. Matsuno. 2003.
Enhancement of the sliding velocity of actin filaments in the presence of ATP analogue:
AMP-PNP. Biophysical chemistry 105:59-66.

105. Van Dijk, J., C. Fernandez, and P. Chaussepied. 1998. Effect of ATP analogues on the
actin-myosin interface. Biochemistry 37:8385-8394.

106. Hayat, M. A. 1981. Principles and techniques of electron microscopy: Biological
applications. Edward Arnold, London.

107. Szent-Gyorgyi, A. 1951. Chemistry of muscular contraction. Academic Press, New York.

108. Mommaerts, W. F. 1958. Chemical investigation of muscular tissues. Methods in medical
research 7:1-59.

109. Chaussepied, P., and A. A. Kasprzak. 1989. Isolation and characterization of the
G-actin-myosin head complex. Nature 342:950-953.

110. Toyoshima, Y. Y., S. J. Kron, E. M. McNally, K. R. Niebling, C. Toyoshima, and J. A.
Spudich. 1987. Myosin subfragment-1 is sufficient to move actin filaments in vitro.
Nature 328:536-539.

111. Weeds, A. G., and R. S. Taylor. 1975. Separation of subfragment-1 isoenzymes from
rabbit skeletal muscle myosin. Nature 257:54-56.

112. Weeds, A. G., and B. Pope. 1977. Studies on the chymotryptic digestion of myosin.
Effects of divalent cations on proteolytic susceptibility. JMolBiol 111:129-157.









BIOGRAPHICAL SKETCH

Colin Dane Sturm was born in Oklahoma City, OK, on February 14, 1978 to parents

George J. Sturm and Barbara J. Sturm. He graduated from Moore High School in Moore, OK, in

1997. In 2002 he received a Bachelor of Science in chemical engineering with a focus on

biotechnology and a minor in chemistry from the University of Oklahoma. He started his

graduate studies at the University of Florida in chemical engineering in 2002 and joined Dr.

Richard Dickinson's research group in 2003. He obtained his Doctor of Philosophy in chemical

engineering in 2007.





PAGE 1

ANALYSIS OF ACTIN FILAMENT POLYME RIZATION ON BIOMIMETIC PARTICLES By COLIN STURM A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007 1

PAGE 2

2007 Colin Sturm 2

PAGE 3

To my father and mother fo r their support a nd inspiration 3

PAGE 4

ACKNOWLEDGMENTS I acknowledge the support and guidance of my adviser, Dr. Richard Dickinson. He was helpful throughout my tenure at the University of Florida, encouraging a self-guided experience with the perfect amount of direction that has given me the quali ties of a great researcher. I would like to thank my committee members for their time and patience spent with me. Specifically, I appreciate the biochemistry knowledge and practical experience Dr. Daniel Purich has given me. Dr. Purich would visit the labor atory daily to check on the students and give words of advice including the not so occasional joke to keep spirits high. Being the teaching assistant for Dr. Jason Butler and taking his co mplex fluids course helped me to understand chemical engineering principles and gave me a better approach at solving difficult problems. Even at busy times, Dr. Butler always had time to spend discussing a problem. The time I spent with Dr. Yiider Tseng, gave me insight into what being a graduate student is about and the trials many students go through which I am thankful for. A great deal of motivation and researchi ng skill came from Dr Joseph Phillips who worked as a post-doctorate for Dr. Purich. Dr. Phillips and I spent much of our time preparing experiments and critically thinking of the next step in the process. I greatly appreciate all of Dr. Phillips help and know that much of my success came from working closely with him. I thank Dr. William Zeile who provided much of my basi c understanding of laboratory experiments and took time to discuss my experimental problems al ways having helpful suggestions. I thank my fellow laboratory mates Kimberly Interliggi, Luzelena Caro, and Ga urav Misra who went through the good and bad with me and were supportive. There are several people that have helped me throughout my graduate years. A short list includes Matt Monroe and Nikul Sheth who were great roommates and friends, Darren McDuff and Jonathan Bricker who provided plenty of la ughter and discussion, an d Michael Matlock and 4

PAGE 5

5 Chris Bussum who always encouraged me and were always there for me both being truly great friends. Finally, I would like to thank the staff at the University of Florida Interdisciplinary Center for Biotechnology Research (ICBR) Core Laborator ies for all their help with my electron microscopy experiments.

PAGE 6

TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........9 LIST OF FIGURES.......................................................................................................................10 ABSTRACT...................................................................................................................................15 CHAPTER 1 ACTIN BASED MOTILITY..................................................................................................17 Actin is Essential for Cell Motility.........................................................................................17 Actin Sequestering Proteins............................................................................................18 Actin Based Propulsion of Listeria monocytogenes ........................................................18 Experimental Evidence of Actin Dynamics....................................................................19 Models for Force Generation by Actin Polymerization..................................................21 Distinguishing Between Actin Polymerization Models.........................................................26 2 ENERGY DENSITY IN BENT ACTIN FI LAMENTS OF AN ACTIN ROCKET TAIL...31 Introduction................................................................................................................... ..........31 Methods..................................................................................................................................31 Results.....................................................................................................................................37 Discussion...............................................................................................................................37 3 ACTIN PROPELLED VESICLES.........................................................................................46 Introduction................................................................................................................... ..........46 Materials and Methods...........................................................................................................48 Bovine Brain Extract.......................................................................................................48 Bradford Assay................................................................................................................49 Actin Purification from Rabbit Muscle...........................................................................49 Acetone powder........................................................................................................49 Purification of actin from acetone powder...............................................................50 Fluorescent labeling of actin....................................................................................51 Preparing Listeria monocytogenes on Agar Plates..........................................................52 Purification of ActA-His6 from Listeria monocytogenes ................................................53 Fluorescent Labeling of ActA-His6-Cys.........................................................................54 Vesicle Preparation..........................................................................................................55 Motility Assay.................................................................................................................55 Creatine kinase.........................................................................................................56 Protease inhibitors and DTT....................................................................................56 6

PAGE 7

Results.....................................................................................................................................56 Discussion...............................................................................................................................59 4 ELECTRON MICROSCOPY OF ACTIN FILAMENTS......................................................68 Introduction................................................................................................................... ..........68 Materials and Methods...........................................................................................................70 Functionalized 500 nm Beads.........................................................................................70 Preparation of Listeria monocytogenes Overexpressing ActA.......................................70 Functionalized 50 nm Beads...........................................................................................71 ActA and BSA Conjugated to 50 nm Beads...................................................................71 Flow Chamber for Exchange Experiments......................................................................72 Beads / Listeria with Actin Rocket Tails........................................................................72 Preparation of Sample for Viewing with EM..................................................................73 Critical point dryer (CPD)........................................................................................73 Sputter coating for SEM...........................................................................................73 Rotary shadow for TEM...........................................................................................74 Post shadowing / pre TEM treatment.......................................................................74 Polyvinyl Formal Coated TEM Copper Grids.................................................................74 Filament Shearing from Fluid Flow................................................................................75 Results.....................................................................................................................................76 Discussion...............................................................................................................................78 5 SINGLE ACTIN FILAMENT POLARIZA TION DETERMINED BY MULTIPLE LABELED ACTIN MONOMERS INCORPOR ATED INTO ACTIN FILAMENTS..........89 Introduction................................................................................................................... ..........89 Materials and Methods...........................................................................................................91 Color Change Assay........................................................................................................91 Image Analysis................................................................................................................9 1 Results.....................................................................................................................................92 Discussion...............................................................................................................................94 6 CONCLUSIONS AND FUTURE WORK...........................................................................101 Discussion.............................................................................................................................101 Suggestions for Future Work................................................................................................103 APPENDIX A MATLAB ALGORITHM TO DETERMIN E ENERGY STORED IN BENT FILAMENTS........................................................................................................................106 B MYOSIN SUBFRAGMENT-1............................................................................................118 Introduction................................................................................................................... ........118 Myosin Purification............................................................................................................ ..118 7

PAGE 8

8 Purification of S1 using Papain............................................................................................119 Purification of S1 using -Chymotrypsin.............................................................................120 C IMAGEJ MACROS..............................................................................................................12 1 LIST OF REFERENCES.............................................................................................................140 BIOGRAPHICAL SKETCH.......................................................................................................149

PAGE 9

LIST OF TABLES Table page 3-1 Necessary ratios fo r a parallel dilution to perform a Bradford assay.....................................61 3-2 List of buffer volumes based on 1 kg of rabbit muscle..........................................................61 3-3 Reference table for amount (g) of ammonium sulfate ((NH4)2SO4) to add at 4C.................61 3-4 Differences in experimental procedure for vesicle motility...................................................62 3-5 ActA conjugation with vesicles............................................................................................ ..62 9

PAGE 10

LIST OF FIGURES Figure page 1-1 Actin filaments polymerizing at the leading edge of a cell....................................................28 1-2 Simplified cartoon of an actin monomer w ith a nucleotide bound inside the actin cleft.......28 1-3 The treadmilling process of an actin filament........................................................................29 1-4 The treadmilling of an actin filament is enhanced with profilin and cofilin..........................29 1-5 The elastic Brownian ratchet model.......................................................................................3 0 1-6 Generalized end-tracking motor pers istently bound to an actin filament...............................30 2-1 Notation used to describe a flexible rod / actin filament........................................................41 2-2 A 500 nm polystyrene bead functionalized with ActA and exposed to a motility assay.......41 2-3 Cartoon of an actin filament projec ting an image on a two-dimensional plane.....................42 2-4 Close up of an actin rocket tail th at has points plotted along the filaments...........................42 2-5 Close up from Figure 2-4 showing points plotted along a single filament.............................43 2-6 Analysis of the bending ener gy of filament in Figure 2-5......................................................43 2-7 Output from algorithm with filaments numb ered and points plotte d with a blue line...........44 2-8 Histogram of all dE/ds valu es in one actin rocket tail............................................................44 2-9 Histogram of all dE/ds values along 670 filaments in 12 actin rocket tails...........................45 3-1 Unilamellar bilipid layer vesicle......................................................................................... ....63 3-2 Vesicle conformation change............................................................................................... ..63 3-3 Vesicles emanating fr om a central vesicle mass....................................................................64 3-4 Vesicles exposed to rhodamine actin and then Oregon-green actin.......................................64 3-5 Color change experiment with a vesicle.................................................................................65 3-6 Color change experiment with a vesicle.................................................................................65 3-7 Histogram of vesicle velocities........................................................................................... ....65 3-8 Vesicle velocities ve rsus vesicle radius.................................................................................. 66 10

PAGE 11

3-9 Theoretical particle velocity at different actin concentrations...............................................66 3-10 Overlay of rhodamine labeled actin pr opelling an Oregon-green labeled vesicle...............67 3-11 Large unlabeled vesicles conjugated with green fluorescent ActA......................................67 4-1 Listeria propelled by an actin rocket tail ju tting from a larger actin network........................81 4-2 Listeria with a small actin comet tail in its early stages.........................................................81 4-3 A 500 nm polystyrene bead with an actin cloud viewed using TEM.....................................82 4-4 A 500 nm polystyrene bead with an actin rocket tail viewed using TEM..............................82 4-5 Two actin rocket tails converge........................................................................................... ...83 4-6 Three 500 nm polystyrene beads combine to form one actin rocket tail................................83 4-7 Single actin filaments eman ating from 50 nm silica beads....................................................84 4-8 A single actin filament associated with a single 50 nm silica bead........................................84 4-9 Several single actin filaments asso ciated with single 50 nm silica beads..............................85 4-10 Histogram of filament leng ths with bin size of 100 nm.......................................................85 4-11 Filament number per bead versus time.................................................................................86 4-12 Number of filaments normalized to the total number of beads versus time.........................86 4-14 Time lapse of single actin fila ments emanating from 50 nm beads.....................................88 5-1 Fluorescent actin cha nge hypothetical scenarios....................................................................96 5-2 High pass filter removing freque ncies larger than 50 pixels..................................................96 5-3 Low pass filter removing particles smaller than 3 pixels.......................................................97 5-4 Low and high pass filters on a mock filament........................................................................97 5-5 Compilation of 158 color change events observed.................................................................98 5-6 Both fluorescent channels and overlay of a two color filament.............................................98 5-7 Both fluorescent channels and overlay of a two color filament.............................................99 5-8 Both fluorescent channels and overlay of a two color filament.............................................99 11

PAGE 12

12 5-9 Both fluorescent channels and overlay of a two color filament.............................................99 5-10 Histogram for beads and filaments of 20 image sets..........................................................100

PAGE 13

LIST OF ABBREVIATIONS ADF Actin depolymerizing factor ADP Adenosine diphosphate AFM Atomic force microscope AMP-PNP Adenyl-5'-yl imidodiphosphate APES 3-aminopropyltriethoxysilane Arps Actin related proteins ATP Adenosine triphosphate BBE Bovine brain extract BHI Brain-heart infusion media BS3 Bis(sulfosuccinimidyl suberate) BSA Bovine serum albumin CPD Critical point dryer DMF Dimethylformamide DMSO Dimethyl sulfoxide DTT Dithiothreitol EDTA Ethylenediamine tetraacetic acid EGTA Ethylene glycol tetraacetic acid EM Electron microscopy FRAP Fluorescence recovery after photobleaching GME Glycine methyl ester GUV Giant unilamellar vesicle LUV Large unilamellar vesicle N-WASP Neural Wiskott-Aldrich syndrome protein NIH National Institutes of Health 13

PAGE 14

14 NTA Nitrilotriacetic acid PBS Phosphate buffered saline PI Protease inhibitors PMSF Phenylmethanesulphonyl fluoride S1 Subfragment-1 SEM Scanning electron microscope SUV Small unilamellar vesicle TEM Transmission electron microscope TIRF Total internal reflection fluorescence TLCK Np-tosyl-L-lysine chloromethyl ketone VASP Vasodilator-stimulated phosphoprotein VCA Verprolin/cofilin homology/acidic

PAGE 15

Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ANALYSIS OF ACTIN FILAMENT POLYME RIZATION ON BIOMIMETIC PARTICLES By Colin Sturm December 2007 Chair: Richard B. Dickinson Major: Chemical Engineering Actin, a monomeric globular protein found in all eukaryotic cells polymerizes into filaments generating force in several essential cellular processes, incl uding cell adhesion, cell movement, and cell division. Invasive bacteria, such as Listeria monocytogenes, use actin for motility inside eukaryotic host cells. Listeria produces a surface protein called ActA which is able to form an actin-rich rocket tail at one pole of the bacter ium. While it is known that the polymerizing filaments provide the propulsive force on the bacterial surface, the mechanism by which filaments assemble and push the surface is unknown. It had been previously widely assumed that filaments were not attached to the surface and must be free in order to elongate. Our group has argued that el ongating filaments are persistently att ached to the ba cterial surface through a processive filament end-tracking motor, termed actoclampin. To help determine whether filaments elongate attach ed or unattached, forces pro duced by actin filaments were analyzed in actin rocket tails through the us e of transmission electron microscopy (TEM) and single filaments were observed to be directly tethered at thei r elongating ends to immobilized ActA-coated beads in vitro. Changing parameters in vitro such as particle size, ActA density, and the time of polymerization, provides contro l over the number of filaments found on each particle. Using total internal reflection fluor escence microscopy (TIRF), the polarity of the 15

PAGE 16

16 filaments could be determined to show barb ed-end elongation at the surface of the bead. However, the number of filaments elongating off a bead was difficult determine due to the resolution limits of fluorescent microscopy. For this reason, we also used TEM as a means of assessing the state of filament s surrounding the ActA-coated beads. These results show several beads with one to three filaments attached at the surface. The combined results of TIRF and TEM show strong evidence of barbed-end elongati on of filaments attach ed at the surface of biomimetic particles, suggesting insertional polymerization mediat ed by an end-tracking motor.

PAGE 17

CHAPTER 1 ACTIN BASED MOTILITY Actin is Essential for Cell Motility Cell motility is essential for physiological processes such as development, wound healing, and defense against infection (1). Cells crawl by extending protrusi ons at their leading edge that adhere to the substratum and allow the cell to pu ll itself forward (Figure 1-1). These protrusions involve the polymerization of actin, which is a highly conser ved (changing little throughout evolution) globular protein (42 kDa) and the most abundant intracellu lar protein in most eukaryotic cells (2). In its unpolymerized form, actin is referr ed to as G-actin and has two globular regions with a hinge conne cting the two domains resulting in a deep cleft (3). In this cleft is a nucleotide bindi ng region that can bind an Mg2+ ion complexed with adenosine diphosphate (ADP) or adenosine triphosphate (ATP). Actin monomers polymerize into 7-nm diameter semi-flexible filaments (4) (filamentous actin or F-actin) consisting of two proto-filaments that wrap around each other in a right-handed helix with a 37-nm pitch and a persistence length of about 15 m (5-7). Actin filaments are polar, with the filament (+)-end (also known as the barbed end) polymerizing faster than the (-)-end (a.k.a. the pointed end) (2). This polar ity results in the subunit only at the filament (-)-end having an exposed cleft (Figure 1-2). Fo llowing binding of actin-A TP to the (+)-end, the nucleotide undergoes hydrolysis (ATP to ADP ) followed by phosphate release, and actin-ADP depolymerizes from the (-)-end. Monomer-bound ADP is then exchanged for ATP in the cytoplasm allowing the monomer to be recycled for (+)-end assembly. The actin-ATP critical concentration for (+)-end assembly is 0.1 M and that for (-)-end assembly is 0.6 M (8). At steady-state polymerization, the c oncentration of actin resides between the (+)-end and (-)-end 17

PAGE 18

critical concentrations, with the (+)-end growing and the (-)-end shrinking, in a process known as treadmilling (Figure 1-3). Actin Sequestering Proteins Actin concentration in eukaryotic ce lls is typically between 100 and 200 M (9). This high concentration of G-actin relative to the critical concentration is maintained by actin sequestering proteins such as thymosin4 and profilin, which bind to G-actin thereby effectively decreasing the concentration of monomeric actin re lative to filamentous actin. Thymosin4 and profilin are the main proteins responsible for actin sequestering. Thymosin4 binds G-actin in solution (Kd is 0.7 M) (9) to prevent its polymerization. Profilin similarly binds G-actin but actually promotes polymerization by catalyzing nucleotide exchange (ADP to ATP) on actin monomers and shuttling monomers to the filament (+)-e nd (10). Actin depolymerizing factor/cofilin (ADF/cofilin) is not as significant in sequestering actin monomers as thymosin4 or profilin but does promote depolymerization of actin filaments (11). Cofilin binds to F-actin-ADP and causes the filament to twist tighter increasing the helica l repeat of an actin filament from 37 nm to 27 nm (9) essentially breaki ng sections of the filament off resu lting in depolymerization. Figure 1-4 shows the pathways and roles of profilin and cofilin interacting with a treadmilling actin filament. Actin Based Propulsion of Listeria monocytogenes Listeria monocytogenes is a bacterial pathogen that in fects cattle and causes severe food poisoning in humans (12). After Listeria is phagocytosed by a host cell, the bacterium secrets enzymes that break down the phagosome thereby releasing the bacterium into the host cells cytoplasm. Once free in the cell, Listeria polymerizes actin filame nts at one pole of the bacterium surface to propel itself within the cytoplasm and to translocate between cells. Listeria requires only a single bacterial protein, ActA for propulsion, and it commandeers other necessary 18

PAGE 19

components from the host cell cytoplasm (13-17). ActA activates the Arp2/3 complex, which nucleates new filaments at the bacterial surface (16, 18-20). ActA also binds to vasodilator-stimulated phosphoprotein (VASP) (21) to promote actin (+)-end assembly. VASP contains several oligo-proline sequences that bind profilin thereby providing a bacterial surface bound pool of profilin actin at filament (+)-ends (12). Our group has proposed that ActAVASP stimulates filament growth in the actoclampin end-tracking motor model for actin based force generation (discussed below) (22). There are also other proteins known to operate similarly to VASP such as neural Wiskott-Aldr ich syndrome protein (N-WASP) for Shigella or vaccinia motility (23-26) or the verprolin/cofilin homology/acidic (VCA) domain (C-terminal of N-WASP) (27). Although there are other possi ble end-tracking motors such as N-WASP or VCA that behave similar to ActA producing actin filaments with (+)-ends at a motile surface, Listeria and ActA have been instrumental in determining essential factors for actin polymerization and allow for analysis of motil ity in a relatively simple system (28). Experimental Evidence of Actin Dynamics The first observation of actin was by W.D. Halliburton in 1887 (29) who extracted a protein from muscle he named myosin-ferment which coagulated preparations of myosin. However, he was unable to further characterize the protein so the discovery went unnoticed for almost 80 years (30). Brn Straub is credited with the discovery of actin in 1942. He developed the first technique to isolate substantial amounts of pure actin (31) that was so effective the technique has remained relatively un changed to this day. More than 60 years have passed and the mechanism of force generation by actin polymerization remains controversial. Several methods have been employed to character ize actins role in force production. The most insight on the mechanism of actin force genera tion has risen from the study of biomimetic 19

PAGE 20

particle motility including the propulsion of particles such as polystyrene beads, oil droplets, and vesicles (32-42). Analysis of biomimetic particle systems is important because conditions can be controlled in vitro experiments are easily reproduced, and the possible propulsion mechanism can be studied. Cameron et al. was the first to coat polystyrene beads with ActA and observe motility of an artificial load by actin polymerization (32) concluding ActA is the only bacterial protein necessary to induce actin polymer ization with no dependence on th e bacterium itself. Cameron et al. took the same actin propelled ActA coated polystyrene beads a nd observed the system using electron microscopy (EM) (33) revealing filaments persistently bound to the bead surface. Filaments were found bound to 50 nm, 200 nm, and 500 nm beads although only one 50 nm bead with a filament was observed. Schwartz et al. f ound that flattened particles coated with ActA are just as motile as spherical particles ruling out a ny actin dependencies on geometric shape (34). In an effort to retard motility velocities, met hyl cellulose (an inert viscous solution used in many food and cosmetic products) was used to slow motile Listeria in a cell extract (35) and N-WASP coated beads in a reconstituted mo tility medium (36). Both studies found that increasing the viscous drag force could not stop mo tility even at high concentrations of methyl cellulose, and only the first study observed an effect of methyl cellulose on motility at all, which may be explained by enhanced diffusion-limitatio ns (43). Studies on VCA-coated particles revealed velocities to be inversely proportional to particle radi us (37) and large beads (> 1.5 m) exhibited saltatory motion with recurring phases of the actin rocket tail cycling from a dense actin network to a loose actin network (correlating to lo w and high velocities) (38) (Figure 3-7A). To estimate the forces on a polystyrene particle, a flexible cantilever was attached to an N-WASP coated particle and pulle d while a micropipette held the actin rocket tail 20

PAGE 21

stationary giving an estimate of 0.25 nN/ m2 of stress on the particle (39). An atomic force microscope (AFM) has also been used to meas ure actin network forces (44) resulting in a pressure on the AFM cantilever of ~1 nN/ m2. Deformable particles such as vesicles or oil droplets were used to estimate the magnitude of forces generated by the actin rocket tail at the surface of the particle. Upadhyaya et al. (40) and Giardini et al. (41) both e xploited the pliable characteristic of vesicles to determine the stresses generated by the actin rocket tail. By analyzing the vesicle tear drop shape formed from an actin rocket tail Upadhyaya et al. (40) estim ated a compressive stress on the sides of the vesicle ranging from 3-4 nN/ m2 and a tensile stress on the rear of the vesicle of 6-8 nN/ m2. The same tear drop shape was observed with oil droplets (42) (Figure 3-2). With all of the experimental data available regarding actin po lymerization, several models have been proposed to describe actins interacti on with a motile surface and the mechanism for force production. Models for Force Generation by Actin Polymerization Different models have been proposed to describe the mechanism of force generation by actin polymerization. The main models that have garnered the most attention and credibility are the elastic Brownian ratchet model (45), the tethered Brownian ratchet model (46), the autocatalytic model (47), and the actoclampin mode l (22). The first thr ee models mentioned are based on the same principle of force production fr om monomer addition to free filament ends (derived from the model proposed by Hill and Kirschner (48)) and differ only slightly in how actin filaments are modeled. The actoclampin m odel is fundamentally different in that it considers filament ends attached to the mo tile surface while polymerizing and utilizes the additional hydrolysis energy to explain force production. Terrell Hill initially hypothesi zed how monomer addition to a cytoskeletal filament could produce a force in 1981 (49) and later expande d on the idea in 1982 (48). Hills model 21

PAGE 22

suggested a motile surface would move slightly aw ay from polymerizing cytoskeletal filaments by Brownian motion, with the stiff filaments prev enting Brownian motion of the motile object backward. Monomer addition to the filament e nd closest to the motile surface was accomplished once the motile object fluctuated far enough away for a monomer to add essentially creating a ratchet pushing the motile surface forward. Many experiments showed thermal fluctuation of the motile surface to be insufficient to produce the observed motion (45). For example, Listeria and Shigella both hijack actin polymerization for motility, however, both bacter ia move at the same rate even though Shigella is much larger than Listeria According to the Hill model Listeria would have a greater velocity than Shigella because the Brownian fluctuations of Listeria would be greater allowing for a faster rate of actin polymerization. To account for this observation of different sized loads moving at the same rate, Mogilner and Oster later proposed a modification to the Brownian ratchet model, the elastic Brownian ratchet mode l (45). This modification to the model suggests filaments fluctuate away from the motile surface by Brownian motion instead of the surface moving by Brownian motion. The elastic Brownian ra tchet model predicts th at if a filament is sufficiently angled to the surface (> 10) and the filament is longer than 75 nm then an actin filament will undergo Brownian motion away fr om the motile surface enough for a single monomer addition (2.7 nm because the proto-filame nts are helical) (Figure 1-5). The filament then straightens due to its persistence length pushing the surface forward with the newly added monomer. This cycle creates a Brownian ratche t that ensures there is a net forward movement (45). Once a monomer has added to one filament, this lengthened filament supports some of the load making monomer addition to other filaments eas ier. Mogilner and Os ter (45) estimated the stall force for an actin filament by the elastic Brownian ratchet model to be 1.8 pN based on a 22

PAGE 23

free filament length of 150 nm (distance be tween motile surface and actin network) and a persistence length of 1 m. The main ideas that make the Brownian ratchet model different from Hills model are filament ends fluctuate from Brownian motion and the motion of the object is from the collective treadmilling of numerous actin filaments behind the motile object. Work by Kuo and McGrath (50) gave more detail on the forces involved in actin polymerization by optically tracking motile Listeria using a laser and photodiode to monitor motion. Several conclusions came from this work that negated non-tethered filaments. First, bacteria do not fluctuate enough for intercalatio n of G-actin monomers. Second, mean squared displacement analysis suggests bacteria do no t diffuse as predicted by Brownian ratchet simulations. Third, bacteria fluctuate 20 times le ss than neighboring part icles pointing to actin tail attachment to the bacteria surface. Last, Listeria was observed to take regular steps of 5.4 nm (the diameter of G-actin monomer). Th e Brownian ratchet mode l would not predict a persistent stepping of 5.4 nm because filament fluctuation must be less than the monomer diameter due to the necessary angle the filament makes with the motile surface. Gerbal et al. (51) used an optical trap as well, however, they used the trap to determine that greater than 10 pN force is required to separate a bacterium from its rocket tail. To account for the experimentally observed attachment between the growing filament network and the motile surface, Mogilner and Oster (46) modified their elastic Brownian ratchet model to include two types of actin filaments, wo rking and attached filaments. In this model, once steady-state is achieved, filaments nucleate at the surface of the motile object while being attached to protein complexes and under tension. Eventually these attached filaments will dissociate and grow freely which ar e referred to as working filaments. These growing filaments 23

PAGE 24

are under compression and are the sa me filaments as described in the elastic Brownian ratchet model. Working filaments ultimately become ca pped and lose contact with the surface (46). A separate model was proposed by Carlsson to explain the expe rimental findings of actin polymerization in which filaments (daughter fila ments) are autocatalytic ally branched from existing filaments (mother filaments). Carlss ons motivation came from experimental work showing mother daughter length correlation (52) total internal fluorescence studies showing filament branches forming along filament sides (53, 54), and confocal microscopy showing branches forming at the barbed end preferring ne wer filaments (55). The main difference of the autocatalytic model in comparison to the Brownian ratchet model is the importance of filament branching. Carlsson hypothesized that the formati on rate of new branches is proportional to the number of filaments or amount of polymerized actin near the motile surface. Therefore motile objects with a larger diameter will have more au tocatalytic filaments at the surface (due to the increased surface area) distributing the load resulti ng in a growth velocity nearly independent of applied force (47, 56). The autocatalytic model simp ly looks at how filaments interact with each other in an actin network and is fundamentally the same as the Brownian ratchet model. The above models, which rely on the free energy of monomer addition to free filament ends to generate force, have a therm odynamic maximum force (stall force) (57) critT T BA ATk F)( maxln (1-1) where kB is the Boltzmann constant, T is temperature ( kBT is 4.1 pN-nm), is the added filament length per subunit (2.7 nm), AT is free G-actin-ATP, and AT(+)crit is the G-actin-ATP (+)-end critical concentration (0.1 M). Assuming AT must be less than the (-)-end critical concentration (0.6 M) for steady-state treadmilling, then Fmax is less than 2.7 pN per active filament. 24

PAGE 25

The actoclampin end-tracking motor model wa s first proposed in 2002 by Dickinson and Purich (22) to explain how end-tracking motors both keep filaments persistently attached to a motile surface and are capable of utilizing ATP hydrolysis for mechanical work. An actoclampin is an end-tracking motor that faci litates the insertional polymerization of actin filaments by affinity-modulated interactions between multivalent end-tracking proteins (e.g. VASP, N-WASP, VCA peptid e from WASP-family proteins) and the filament (+)-end. Dickinson et al. (58) proposed that, while one end-tracking subunit bind s one actin subfilament end, another binds to a free G-actin-ATP mono mer in solution and loads it onto the (+)-end of the other actin subfilament. Th e energy released upon hydrolysis of ATP attenuates the binding affinity of VASP to the actin filament, there by releasing the filament end and binding another G-actin-ATP monomer in solution. This process c ontinues alternating attachment to the filament while loading monomers resulting in consta nt filament attachment with insertional polymerization (Figure 1-6). Actin has a confor mation change (3) when ATP is hydrolyzed to ADP which could explain why the clam p loses affinity for F-actin-ADP. Importantly, by capturing energy released by ATP hydrolysis (up to 14 kBT (58)) and allowing force-independent monomer binding from the actoclampin model predicts a weak dependence of filament elongation rate on compressive or tensile forces up to several pN (22). In contrast, the Brownian ratchet models pred ict exponential force-velo city relationship and much lower thermodynamic stall forces (< 2.7 pN). Consequently, accord ing to the actoclampin model, forward motion of a propelled particle sh ould limit the elongation rate (or detachment) of the most slowly elongating filaments in the ac tin tail. These filame nts are under tension, balanced by the compressive forces of other filaments at the motile surface. This push-pull force balance explains many of the experimental observations of actin-based particle propulsion, 25

PAGE 26

including monomer-sized st epping motion of Listeria (22), deform ation of vesicles into tear-drop shapes (55), and saltatory motion of larger particles (43). Any of these motile particles would be affected by capping protein (such as gelsolin or CapG which bind to filament (+)-ends preventing polymerization) if the filaments we re not tethered, however, when capping protein concentrations are increased in motility assays, motile particles are not hindered (59-61). Filaments not being affected by capping protein suggests the filament (+)-ends are protected possibly by the tethering protei n complex of ActAVASP as in the actoclampin model. Distinguishing Between Actin Polymerization Models The purpose of the present study is to conduct experiments to dete rmine whether actin filaments generate force while processively att ached to the motile surface. Electron microscope images of actin rocket tails trailing 500 nm bead s were analyzed to determine the energy density of the rocket tail to distinguish between the fr ee filament and actoclampin models. Untethered filaments are limited in the amount of energy that can be stored in and translated to the actin network due to a small maximum force generati on from only considering monomer addition and filaments buckle at a much lower force than if tethered. The actoclampin model predicts much higher energy storage in the actin rocket tail compared to free filament models. Actin propelled vesicles were produced and analyzed in Chapte r 3. The goal of producing actin propelled vesicles was to measure the shape change of the vesicle and relate that to the force generated by the actin rock et tail. Although force estimate s were not accomplished, vesicle velocity versus vesicle radius was determined. Calculations by Dickinson and Purich (43) predict an inverse correla tion of diameter to velocity which was confirmed in the findings of Chapter 3 whereas free filament models would no t predict the size and velocity to be directly correlated. This difference betw een actoclampin and free filament models is because Dickinson 26

PAGE 27

27 and Purich (43) determined particles >2 m in diameter will show saltatory motion from diffusion-limited actin polymerization. To determine whether filaments elongate by processive insertional polymerization, electron microscopy of 50 nm beads with si ngle filaments show se veral thousand single filaments associated with single beads, demonstr ating that individual filaments remain attached to the motile surface during elongation. To determin e the direction of elo ngation relative to the bead, a fluorescent color change experiment was pe rformed (Chapter 5) to demonstrate that the point of monomer insertion was at the bead surface. These findings provide strong evidence filaments are polymerizing from the bead surface and not randomly associating with the bead or adsorbing from solution onto the bead. All of th e findings in this study support the actoclampin model.

PAGE 28

Figure 1-1. Actin filaments pol ymerizing at the leading edge of a cell producing filopodial extensions that attach to the substratum temporarily holding the front of the cell in place. The rear of the ce ll is then pulled forward allowing the cell to crawl. Figure 1-2. Simplified cartoon of an actin m onomer with a nucleotide bound inside the actin cleft. Actin monomers bind ATP in solu tion and polymerize into actin filaments with the cleft oriented towa rd the (-)-end. The ATP is hydrolyzed and monomers eventually depolymerize returning to solution to exchange ADP for ATP. 28

PAGE 29

Figure 1-3. The treadmilling process where an ac tin filament polymerizes actin-ATP, hydrolyzes ATP, and depolymerizes actin-ADP at which point the process star ts over with actin nucleotide exchange (58). The free energy changes ( G ) of each step in the treadmilling process is calculated with one ATP hydrolyzed (~22 kBT (62)) per monomer (Pi is phosphate, kB is Boltzmanns constant and T is temperature). PiPiPiPi (-) (+) ATP ADP actinATP actinADP Pi PiactinADPPi p p p p p p ADF ADF p p ADFprofilin ADF/cofilin Pi ATP ADP p p p p phosphate release actin ATP Addition profilin Release profilinushered addition nucleotide exchange profilin binding ADF release ADF actin ATP Release ADFMediated Release ATP hydrolysis ADF Binding ADF Figure 1-4. The treadmilling of an actin filament is enhanced with profilin and cofilin with various pathways shown (58). The dashed box represents a polymerizing filament sans profilin and cofilin. 29

PAGE 30

30 Figure 1-5. The elastic Browni an ratchet model hypothesizes free filaments fluctuate from a motile surface by Brownian motion enough fo r a monomer to intercalate to the (+)-end (45). The filament returns to its original configuration with an additional monomer at the (+)-end. This additional le ngth pushes the surface forward while the filament is held in place by cr oss-links in the rocket tail. Figure 1-6. Generalized end-tr acking motor persistently bound to an actin filament (63). A) End-tracking motor with only an F-actin binding region. B) End-tracking motor with a G-actin binding region that loads a monomer and is then held by the F-actin binding region.

PAGE 31

CHAPTER 2 ENERGY DENSITY IN BENT ACTIN FILA MENTS OF AN ACTIN ROCKET TAIL Introduction Actin filaments are considered semi-flexible f ilaments with an average flexural rigidity compared to other cytoskeletal filaments (64). This flexural rigidity is an important property which allows actin to convert chemical energy into mechanical energy in order to drive a cell forward or push bacteria through a cell. One wa y of characterizing the flexural rigidity of cytoskeletal filaments is by the persistence length which is the characteristic correlation length in the orientation of a thermally undulating filament (6, 64). Reported values of the persistence length for actin are in the range 10 5 m (65). We propose a method for estimating actin filame nt bending energy in the actin rocket tail behind a motile particle. Although si ngle filament analysis of actin rocket tails is difficult to accomplish because of the small size of filame nts and dynamic change in filament growth, electron microscopy (EM) allows visualization of st atic actin comet structur es at the nano-level. To produce an actin rocket tail th at can be viewed with EM, bead s are conjugated with ActA and incubated in a cell extract. The beads with actin tails are then coated wi th a thin metal for EM and imaged (32, 33). Methods To simplify the physical model for bent actin filaments, the actin filaments are considered to be semi-flexible rods. As shown in Figure 2-1, each point along the filament is designated by the position vector r ( s ) which encompasses the Cartesian coordinates [x( s ), y( s ), z( s )]. The arc length is defined as s and the unit tangent vector is designated as N ( s ), N ( s ) = dr/ds (2-1) 31

PAGE 32

The change in the unit tangent along the leng th is proportional to the unit normal to the curve n (64), dN / ds = Cn (2-2) If the unit normal vectors are significantly clos e and extended to the origin then it can be deduced that C is the local radius of curvature which is necessary for relating the change in tangent vector to the radius of curvature, C = 1/ R (2-3) The energy associated with the bending of a r od has been solved (66) and has the form 22 R B L E (2-4) where E is energy L is filament length, and B is the bending modulus, re lated to the persistence length as B = kBT (2-5) where kB is Boltzmanns constant (1.38 x 10-23 J/K), and T is temperature (K). The deformation energy per unit le ngth of the bent filament is proportional to the square of the radius of curvature (64) or equivalent to the square of Equation 2-3. Squaring Equation 2-3 and combining with Equations 2-3 and 2-4 resu lts in the energy per uni t length. Since the curvature is not required to be constant, the en ergy of bending can be e xpressed as the integral along the length of the filament or as the change in energy per change in length (Equation 2-8) (67), 22 ds dNB ds dE (2-6) Electron microscopy images of actin comet tails (Figure 2-2) were analyzed to estimate the energy stored in actin filaments. Filament s undergo tension and comp ression due to elongation 32

PAGE 33

forcing a network of bent crosslinked filaments, which is evident from EM images (Figure 2-2). Bent filaments are a result of monomer addition to crosslinked filaments, which are constrained from extending, forcing the filament to be nd. The bending energy can be determined by measuring the curvature of the filaments and usi ng Equation 2-6. Due to the necessary treatment of the actin rocket tail for EM, some twisting and bending may occur to the filaments. The additional energy associated with EM treatment is considered negligible because filaments that have only one interaction with another filament are typically straight suggesting the EM treatment has not affected the configuration. Filaments that are bent or twisted are usually restrained by at least two points of crosslinking or filament intera ction which is to be expected from actin polymerization (33). An algorithm was developed to determine the amount of energy stored in actin filaments as measured from EM images. The angles of bent f ilaments were measured in the plane of focus. However, one problem with determining the amount of bending in an actin filament from an EM image is that only the two-dimension projections of the filament arcs are measurable for a three-dimensional filament. To adjust for the missing third-dimensional data, a correlation between the two-dimensional projected curvature of a filament and the actual three-dimensional curvature is needed. The below de rivation relates the change in energy per length of filament in three-dimensional space to the change in energy pe r length of filament in two-dimensional to estimate the amount of energy sp ent in bending the filament. N is the unit tangent vector in the EM image and s is the projected length of the filament in the EM image. Theta ( ) is the angle with respect to the x-axis in the plane of the EM image while phi ( ) is the angle in to or out of the EM image (Figure 2-3), sin cos N (2-7) 33

PAGE 34

d dN cos sin (2-8) The viewed length is equal to th e change of the angle along the act ual length of the filament ( s ). Therefore, the viewed length is a function of the changing angle of the filament as it goes in to or out of the focal plane, Ldss s0)(sin*, (2-9) dssds )(sin* (2-10) Equation 2-8 divided by Equation 2-10 is th e viewed change in angle divided by the viewed change in length, ds d ds dN sin cos sin sin (2-11) The square of Equation 2-11 is needed to rela te the two-dimensional values to the actual three-dimensional orientation of the filaments, 2 2 2 2 2 2 2 2sin 1 sin cos sin ds d ds dN ds d ds dN (2-12) The three-dimensional expression of dN / ds must be determined to equate with the two-dimensional change in tangent vector with respect to length, cos sinsin cossin N, (2-13) dd dN sin sincoscossin coscossinsin. (2-14) 34

PAGE 35

Dividing dN by ds and squaring gives ( dN / ds )2, which is used in the energy equation (Equation 2-6). Through mani pulation of Equation 2-14, ( dN / ds )2 can be represented as a sum of changing angles, 2 2 2 2sin ds d ds d ds dN (2-15) The ( d / ds )2 component in Equation 2-15 can be obtained by rearranging Equation 2-12. Equation 2-15 can then be writte n in terms of the two-dimensi onal change of the filament, 2 4 2 2* sin ds dN ds d ds dN (2-16) Equation 2-16 gives a relation between three-dimensions and two-dimensions; however, there is still a 2dsd term that cannot be determined directly from the two-dimensional images. A uniform distribution is assumed for the orientat ion of each filament therefore the changing and turning of the filaments should be approximately the same for both the direction and the direction. Thus, the assumption was made that twice the 2 2dsdsin term will give reasonable results. Since the filaments can gr ow and point in any direction, a uniform distribution of filaments is used to account for the different or ientations of filaments and to account for the direction. The distribution used is assumed to be a uniform distribution around a sphere, 4 1 ),(p (2-17) With the adjustment in Equation 2-16 due to the 2dsd term and combining Equation 2-16 with Equation 2-17, the average cha nge in angle per change in length of the three-dimensional filament is related to th e two-dimensional data through Equation 2-18, 35

PAGE 36

2 0 2 0 2 4 2* 15 16 sin sin2 4 1 ds dN dd ds dN ds dN (2-18) The integration of Equation 2-18 substituted into Equation 2-6 gives an approximation of the average amount of energy per length of a bent filament, 2* 15 16 2 ds dNB ds dE (2-19) Using the change in position vector instead of the change in tangent vector allows for easier measurements of the filaments. The position vector is defined as jyixr and the second derivative of the position vector is equivalent to the first derivative of the tangent vector, 2 2 2 2* ds dN ds rd. (2-20) Substituting Equation 2-20 into Equation 2-19 gives the energy equation as a function of the position vector. The additi on of a correction factor and ex panding the second derivative of the position vector yields the rela tion of energy to filament position N i imi N i ii i B irr s s rrr Tk rE1 2 2 1 2 2 1 12 15 8 (2-21) where is the error in measuring the filament positi on (2 nm as determined by the images), and rm,i is the measured position of the filament. Since there will be some intrinsic error in measuring the filament position, the measured po sition is smoothed to a minimum to reflect as close to the actual filament position as possible. The en ergy equation (Equation 2-21) is minimized for the points to find the maximum likelihood estimation for the filament contour. 36

PAGE 37

Results Cameron et al. (33) prepared several beads with actin rocket tails for EM as described above. Additional unpublished images of actin rocket tails, beyond thos e published in Cameron et al. (2001), were generously provided by Tatiana Svitkina, Depa rtment of Biology, University of Pennsylvania. Using the Natio nal Institutes of Healths (NIH ) image software ImageJ, several hundred filaments across 12 images were mapped. For consistency, several criteria were followed to increase the accuracy of determining the change in energy of a filament. Only filaments that could be clearly distinguished with lengths greater than 20 nm were measured. A minimum of 50 filaments per image were measured to give a broad repr esentation of filaments found in the actin rocket tail. Po ints along the filament were kept at a distance of 2 nm apart and measured points were taken as close to the axis of the filament as possible. Figure 2-4 and Figure 2-5 show filaments plo tted by the above criteria. An algorithm (Appendix A) was created in Matlab to optimize the points along the filaments and determine the energy density of the filaments (Figure 2-6 and Fi gure 2-7). This energy de nsity in the filaments is a result of actin polymerizi ng and bending filaments in the rock et tail or bending itself between the load and rocket tail. Figure 2-6A is the optimized filament positi on and Figure 2-6B shows the levels of energy density by means of a contour graph. Since the filament in Figure 2-5 is almost straight the energy along the filament cha nges slightly with a maximum energy density of 0.86 pN. This process is carried out for each of the filaments shown in Figure 2-4 with the final result for the actin rocket tail used shown in Figure 2-7 and a histogr am of the results in Figure 2-8. Discussion Several hundred filaments were analyzed re sulting in an energy density ranging from 0.2 pN for slightly bent filaments up to 259 pN for extremely bent filaments (Figure 2-9). The 37

PAGE 38

average energy density for the 670 filaments meas ured in 12 images was 3.1 pN which can be compared to the theoretical models. The elastic Brownian ratchet model (45, 46) or similar models relies on force generation solely from m onomer addition to free (untethered) filament (+)-ends. The theoretical maximum energy per added length for monomer addition is 2.7 pN, under conditions that promote (-)-end depol ymerization (actin concentration < 0.6 M). In order to produce a large energy density (around the ma ximum of 2.7 pN) in the rocket tail, actin polymerization would have to continue at optimal conditions for the life of the rocket tail. When only monomer addition is the energy source for fila ment bending, an energy density in the rocket tail of the maximum 2.7 pN would not be observe d because some energy is lost in the forward motion of the bead and filaments typically have a large incidence angle with the bead reducing the amount of energy transferred to the rocket tail (Figure 2-2). In contrast, the actoclampin model utilizes the energy of hydrolysis to drive actin polymerization of tethered filament (+)-ends, resu lting in a larger energy source for motility than would be provided by monomer addition alone. The total treadmilling cycle of actin polymerization provides up to 22 kBT of energy with a majority coming from ATP hydrolysis (~14 kBT ) (58). The actoclampin estimates are more than enough to account for the energy density observed in the EM images of actin rock et tails. Another advantage of the actoclampin model over untethered models is th e amount of force a filament can withstand before buckling. Dickinson et al. (58) showed mathematically th at a tethered filament can handle an order of magnitude greater force (considering filaments pe rpendicular to load surface) over an untethered filament. As untethered filaments become more glancing to the load surface, less force is necessary to buckle the filament. Even when thermodynamic considerations are not considered, tethered filaments have a mechanical advantage over untethered filaments. 38

PAGE 39

Other groups have estimated network forces of actin rocket tails by a variety of methods. Weisner et al. (36) estimated the polymerization force of an actin rocket tail to be greater than 50 pN using beads coated with N-WASP by studying the velocity of the b eads through a viscous solution. McGrath et al. (35) estimated the actin rocket tail polymerization force to be ~10 pN on motile Listeria similar to the approach by Weisner et al. Analysis of actin propelled vesicles by Giardini et al. (41) and Upadhyaya et al. (40) show compressive st resses of 3 to 4 nN/um2 and tensile stresses of 6 to 8 nN/um2. Upadhyaya et al. (40) estimated single filaments could produce ~10 pN of force. All of these estimations give a broad view of polymerization force by an actin network but are limited in showing energy associated with individual filaments. There are some limitations of the measuremen t of bent filaments to estimate the energy change per filament length. First, the observed bending of a filament may or may not be due to the actual growth of the filament but could be from the cumulative effect of other filament growth. Because filaments overlap and end inte ractions are unknown, the filament ends are considered straight and therefore have a value of 0. Also, stored mechanical energy could dissipate with deformation of the rocket tail and depolymerization of filaments which would lower the overall measured energy density. Second, not all of the filaments can be measured in the actin tail due to hindrance from the exterior layer of filaments bloc king visualization of interior filaments. This leaves out filaments that might be straight which would lower the average if incorporated in the analysis or bent filaments which would in crease the average of energy. Also, filaments under tension are not di stinguishable from stra ight filaments which would affect the energy density. To address this issue, a large population of filaments was used to give an overall representation of filaments f ound in the actin tail. Third, the angle coming out or into the image is unattainable due to the twodimensional aspect of TEM. Since this angle is 39

PAGE 40

not measured the data lost is only of filaments bent in or out of the im age. Therefore, this additional data would most likel y increase the determined energy of the actin tail, although this additional angle is made up for with the estimati on made in Equation 2-18. Another concern is the cross over of filaments with each other. This cross over prevents the accurate measure of filament lengths and angles. To counteract this problem, only filaments that could be distinguished for at least 20 nm were measured Finally, the amount of perturbation from EM treatment is unknown. However, the affect of EM treatment seems to be trivial since free filaments or filaments with one point of interac tion with another filament are usually straight. Whereas filaments that were bent or twisted were usually restrained by at least two crosslinks (33). Also, filaments could be affected when th e rocket tail is bound to the substratum, however, filaments closest to the substratum are not vi sible and periphery filaments not multiply bound to the rocket tail were avoided. In conclusion, the estimated energy density exceeds what should be provided by monomer addition alone. Although not a strong proof of one model for force generation over another, these results argue against Hill type models and favor the actoclampin model. The actoclampin model not only explains the hi gh energy density (compared to the untethered filament models) but also explains the physical f ilament attachment to the bead surface as observed in the EM images (Figure 2-2). 40

PAGE 41

Figure 2-1. Notation used to describe a flexible rod / actin filament where s is the arc length, r(s) is the position vector, N(s) is the unit tangent vector, and n1 and n2 are normal vectors. Figure 2-2. A 500 nm polystyrene bead functionalized with ActA and exposed to a motility assay to induce an actin rock et tail. The sample is th en treated and viewed with TEM (Unpublished image provided by Tatiana Svitkina, Department of Biology, University of Pennsylvania and used by permission). 41

PAGE 42

Figure 2-3. Cartoon of an actin filament projecting an image on a two-dimensional plane where N(s) is the unit tangent vector of a three-dimensional filament of arc length s N*(s*) is the unit tangent vector of a three-dimensional filament projected onto a two-dimensional plane or arc length s* is the angle of the two-dimensional filament, and is the angle out of the plane of the two-dimensional image. The eye facing down is analogous to looking at an EM image. Figure 2-4. Close up of an actin rocket tail that has points plotted along the filaments. The white box designates the zoomed in section shown in Figure 2-5. 42

PAGE 43

Figure 2-5. Close up from Figure 2-4 show ing points plotted along a single filament. Figure 2-6. Analysis of the bending energy of fila ment in Figure 2-5. A) The red dotted line is the actual position of the data found. The blue line w ith circles is the best-fit line. B) A twenty level contour plot of the en ergy change along the filament with blue being lowest energy and red being highest. The dashed lines point to the maximum energy value of 0.86 pN. 43

PAGE 44

Figure 2-7. Output from algorith m with filaments numbered and poi nts plotted with a blue line. The largest energy found (at filament 47) is shown by the dashed lines. Figure 2-8. Histogram with a 0. 65 bin size of all dE/ds values (excluding filament end points) along 53 filaments in one actin rocket tail. 44

PAGE 45

45 Figure 2-9. Histogram with a 5 pN bin size of all dE/ds values (excluding filament end points) along 670 filaments in 12 actin rocket tails. The inset graph is the same data with the first two bins removed and the thr ee largest values removed (239 pN, 257 pN, and 259 pN).

PAGE 46

CHAPTER 3 ACTIN PROPELLED VESICLES Introduction Actin based motility of bacteria or nonbiologi cal cargo such as beads (32, 37, 38, 68, 69) are useful for studying tail velocities and the ac tin rocket tail network. Bacteria and beads are rigid and do not deform under the forces exerted on the surface by actin filaments. Here, we focus on the use of compliant vesicles as a pot ential method for estimati ng actin polymerization forces. Vesicles are convenient particles to study compared to actin propelled beads for several reasons. First, vesicle lipid con centrations can be controlled and lipids are free to diffuse in the membrane. Second, vesicles mimic the plas ma membrane of protruding lamellipodia and cellular liposomes which allows for greater in sight into how actin polymerization affects motility. Finally, due to the vesicles pliable surf ace, actin polymerization forces associated with the vesicle surface can be analyzed through me asurement of vesicle deformation (40). Vesicles (or liposomes) are composed of se lf-assembled amphiphili c lipids (70), which have a polar hydrophilic head group and a hydrophobic aliphatic chain (the tail group). Because of the solubility difference between the two groups the lipids form ordered structures in aqueous solutions. Depending on the concentration of lip ids, length of the aliphatic chain, and the number of tails associated with a polar head group determines the structures formed in the aqueous solution. The lipids aggregate to form self-closed spherical particles where one or more lipid membranes encapsulate part of the solvent (70) (Figure 3-1). These vesicles are described based on their size and number of membranes. Small unilamellar vesicles (SUV) are typically 50 to 100 nm in diameter and consist of a single bilayer of lipids. Giant unilamellar vesicles (GUV) are greater than 1 m in diameter and large unilamella r vesicles (LUV) range in size 46

PAGE 47

between SUVs and GUVs. The same size designation is used for vesicles containing multiple layers of lipid bilayers using the term multilamellar instead of unilamellar. To nucleate new filaments at the vesicle surf ace, ActA must be bound to the vesicle to promote actin polymerization. When these vesicl es conjugated with ActA are exposed to a cell cytoplasmic extract, actin polymerizes on th e vesicle surface and the resulting filaments aggregate on one vesicle hemisphere allowing fo r symmetry breaking and propulsion (lipids are mobile in the membrane). As the vesicle is propelled forward, the vesicle undergoes a conformation change from its orig inal spherical form, creating a tear drop shape (Figure 3-2). Using osmotic pressure, membrane tension, and comparing the shape change to the original spherical shape, the stresses due to actin polymerization can be estimated at different locations along the vesicle surface (40, 41). Two groups have exploited vesicle deformati on by actin polymerization to estimate actin forces (40, 41). Both groups used similar a pproaches in measuring the forces, using vesicles developed with similar methods and mecha nochemical models (Table 3-4). The key assumptions made were: force associated with membrane bending is negligible, vesicles are unilamellar, vesicles are not stre tched and have no osmotic pre ssure when in the rest state (without actin polymerization), hydrodynamic drag is negligible (<100 fN for a 3m-diameter vesicle), and actin polymerization combined with membrane tension balances the osmotic pressure inside the vesicle (40, 41). Vesicles were exposed to a motility assay and once motile, the vesicles went through several cycles of changing from spherical to tear drop shape suggesting filaments were attached to the surface and some of the filaments suddenly detach from the trailing edge due to the increase in membrane tension. Based on a reac tion-diffusion model, Dick inson and Purich (43), 47

PAGE 48

explained these deformations as arising from a depletion of available monomers and a reduced rate of polymerization of tethered filaments at the center of the actin rocket tail. Since the filaments are attached to the surface the faster periphery filaments push the vesicle forward while the lagging interior filaments create a pulling forc e on the rear of the vesicle which produces the tear drop shape (Figure 3-2). The original goal of this study was to produce large actin propelled vesicles (> 5 m) and explore the forces exerted on the surface of the ve sicle from the rocket tail. Determining the surface force was to be accomplished by anchoring th e rocket tail and aspirating the front of the vesicle with a micropipette. Th e actin rocket tail force on the surface would then be estimated by analyzing the balancing opposing for ces of the increased osmotic pr essure of the vesicle versus the polymerization force of the rocket tail. However, reproducing the larger motile vesicles using the published protocols was unsuccessful, and we had to develop a new protocol to produce motile vesicles. Although vesi cles forces were not measured the vesicle velocities were measured and correlate well with Dickins on and Purichs (43) theoretical findings. Materials and Methods Bovine Brain Extract Whole bovine brain was stored at -70C until needed. Brai n was removed from storage and weighed. The brain was crushed with a mortar and pestle in a cold room while under liquid N2. The powder was transferred to a dounce hom ogenizer where it was mixed with an equal volume of Tris MgCl2 buffer including PIs (10 mM Tris HCl pH 7.5, 2 mM MgCl2, 10 g/mL pepstatin A, 10 g/mL leupeptin, 10 g/mL chym ostatin, and 1 mM PMSF). The mixture was dounce homogenized 30 times while on ice being careful to avoid bubbles. The homogenized brain was sonicated on ice with a tip sonicator fo r 30 second bursts and 1 minute rests at a level of 25% on a power level of 1.5 again avoiding bubbles. The solution was transferred to a 48

PAGE 49

centrifuge tube and centrifuged at 17,000 g (15,000 rpm in a Ti-60 rotor) for 20 minutes at 4C. The supernatant was saved and centrifuged again at 118,000 g (41,000 rpm in a Ti-60 rotor) for 1 hour at 4C. Bradford assay was used to m easure the protein concentr ation which should result in approximately 10 mg/mL. The supernatant wa s aliquoted and stored at -70C until needed (71). Bradford Assay The Bradford assay was used to get an a pproximation of protein concentration. The sample was diluted and compared to a standard Bradford curve of appropriate protein (BSA) (72). Table 3-1 lists the concen trations to use in a parallel dilution for the Bradford assay. Actin Purification from Rabbit Muscle Acetone powder Rabbit muscle was frozen and stored at -70 C until needed. All volumes used were based on the measured mass of the rabbit muscle. Tabl e 3-2 lists the volumes of buffers needed for 1 kg of rabbit muscle. To prepare acetone powder, a section of rabbit muscle was weighed and stored at -20C overnight. The next day the ra bbit muscle was placed at room temperature for 2 hours. The extraction buffer (450 mM KCl pH 6.2 with KOH, 150 mM KH2PO4, and 0.1 mM EDTA) was supplemented with 1 mM AT P, 0.2 mM DTT, and 0.5 mM MgCl2. Fat and connective tissue were removed from the muscle w ith a scalpel. The muscle was placed in a clean beaker on ice. Two hundred mM PMSF in DMSO was prepared an d used to supplement the extraction buffer to 0.5 mM PMSF. The meat was ground two times in a cold room with a meat grinder. Extraction buffer was added with PMSF to the ground muscle and stirred for 30 minutes. The mixture was poured into a bucket w ith cold water and stirred in the cold room. The solution was then filtered through 4 layers of cheesecloth and placed over a chilled extractor with a vacuum attachment. The muscle was removed from the cheesecloth and placed into a 49

PAGE 50

clean beaker. Extraction buffer without PMSF was added and stirred for 30 minutes at 4C. The mixture was poured into a bucket with cold water (T able 3-2) and stirred in the cold room. The solution was filtered through 4 layers of cheesecl oth and placed over a chilled extractor with a vacuum attachment. Muscle was removed and placed in a clean beaker at room temperature with the appropriate volume of 0.4% NaHCO3 added and stirred for 45 minutes. The solution was filtered through 4 layers of cheesecloth and pl aced over a chilled extractor with a vacuum attachment. The muscle was placed into a clean beaker in the cold room 1 mM Tris was added and stirred for 5 minutes. The solution was filte red through 4 layers of cheesecloth and placed over a chilled extractor with a vacuum attachme nt. The previous two steps were repeated. Muscle was added to 1/6 the total volume of ice-cold acetone (-20C) a nd stirred for 5 minutes on ice in the cold room. The solution was filtered through 4 layers of cheesecloth and placed over a chilled extractor with a vacuum attachme nt. The previous two steps were repeated. Muscle was added to 1/9 the total volume of ice-cold acetone (-20C) a nd stirred for 5 minutes on ice in the cold room. The solution was filtered through 4 layers of cheesecloth and placed over a chilled extractor with a vacuum attachment The previous two steps were repeated five more times. After the last wash, the acetone powder was placed on a dry 3 mm Whatman paper (or 3 sheets of Whatman #1), covered, and allowe d to dry overnight in a desiccator under house vacuum to avoid humidity. The final powde r was aliquoted and store at -20C (73). Purification of actin from acetone powder Two grams of acetone powder was weighed out to make 2 mg of actin. A 5% acetone powder solution in G-buffer was made by adding 2 g of powder to 40 mL G-buffer (5 mM Tris pH 8 with KOH, 0.01% NaN3, 0.1 mM CaCl2, 0.2 mM ATP, 0.2 mM DT T, stored at 4C with ATP and DTT added just before use). The slurry was stirred with a glass rod in a glass beaker on ice for 25 minutes (every 2 minutes for 15 seconds). The solution was filtered through 50

PAGE 51

6 layers of cheesecloth and centrifuged at 4C (30 minutes at 30,000 g using Ti-45 at 19,000 rpm). The supernatant volume was measured and supplemented to 2 mM MgCl2. The solution was stirred for 5 minutes at room temper ature. Solid KCl was slowly added to a final concentration of 0.07 M and stirred for 1 hour at room temperature. The solution was centrifuged for 3 hours at 120,000 g at 12C with a Ti-45 at 38,000 rpm. The supernatant was removed and 1 to 2 mL G-buffer was added to each pellet to wash. Then a total of 6 mL G-buffer was distributed among all the pellets (73). The next mo rning the pellets were broken up and dissolved. The absorbances at 280 nm and 290 nm were measured to get an estimate of the protein concentrations. The pellets were then dialyzed against 1 L of G-buffer with 2 changes for at least 3 days. After dialysis the absorbance at 280 nm and 290 nm was measured and the concentration was calculated using the molar extinction coefficients, 280 = 1.11 L/gcm and 290 = 0.63 L/gcm which are accurate be tween 0.1 and 1 absorbance units. Fluorescent labeling of actin Filamentous actin was diluted to 60 M in 2 mL of labeling buffer (20 mM HEPES pH 7.5, 0.1 M KCl, 2 mM MgCl2, 3 mM NaN3, 0.3 mM ATP, with ATP added just before use). The diluted F-actin was then dialyzed in 1 L of labeling buffer for 2 hours with one solution exchange to remove any amine containing buffers from solution. One mg of dye was dissolved in 100 L DMF or DMSO immediately before mixing with actin. The dye was sonicated 1 minute to mix the dye thoroughly. While vorte xing the F-actin solution, the dye was added dropwise (slowly) to a final con centration of 300 M (5 fold con centration) (74). The mix was incubated at 4C overnight while mixing end-over-end. The reaction was stopped by supplementing to 50 mM lysine. The supernatan t was then centrifuged (45 minutes at 290,000 g using a TLA 100.2 rotor at 90,000 rpm). The pellet was resuspended in 2 mL G-buffer by slowly adding buffer and tritura ting with a glass rod. The resuspended protein was then 51

PAGE 52

sonicated for 1 minute in a bath sonicator. Th e mix was swirled for 30 minutes at 4C using a vortex setting of level 1. The solution was then dialyzed against three changes of 1 L G-buffer for 36 hours. The absorbance of the actin at 280 nm and the absorbance of the dye were measured. The major fractions (> 4 M) were pooled and polymerized for 45 minutes at room temperature by supplementing to 1 mM ATP and 1x P-buffer (10x P-buffer is 0.5 M KCl, 20 mM MgCl2 which is always mixed with G-buffer). The supernatant was then centrifuged (45 minutes at 290,000 g using a TLA 100.2 rotor at 90,000 rpm). The pellet was resuspended in G-buffer to a desired final G-actin concentratio n. The solution was then dialyzed against 3 changes of 1 L G-buffer for 36 hours and the supernatant was centrifuged (45 minutes at 290,000 g using a TLA 100.2 rotor at 90,000 rp m). The absorbance at 280 nm and max for the dye were measured and the G-actin concentratio n was calculated by using the Beer-Lambert law adjusted for the dye (75) CFAA l actinGnm actinG nm max 280 2801 (3-1) where l is the length of the absorbance chamber (1 cm), is the extinction coefficient, A280nm is the actin absorbance at 280 nm, A max is the actin absorbance at the maximum absorbance of the fluorophore, and CF is the correction factor specific for each fluorophore. Preparing Listeria monocytogenes on Agar Plates Brain heart infusion (BHI) media with agar wa s prepared and autoclaved. Ten g/mL of chloramphenicol was added to the media to prevent bacterial growth (other than bacteria immune to the antibiotic). The me dia was distributed among Petri dishes and cooled. Listeria was streaked on the agar media and incubated at 37 C overnight. The Listeria plates were stored at 4C and individual colonies were acquired as necessary. 52

PAGE 53

Purification of ActA-His6 from Listeria monocytogenes A Listeria culture from a previously prepared ag ar plate was removed and grown overnight in a culture tube containing 50 mL BHI with 10 g /mL chloramphenicol at 37C while shaking. The media containing incubated Listeria was added to 1 L of BHI containing 10 g/mL chloramphenicol and grown overnight at 37C wh ile shaking. The broth was added to a large centrifuge tube and cooled on ice. Listeria was pelleted with a JA-10 rotor at 5,000 rpm at 4C for 20 minutes. The supernatant volume was me asured and 50% ammonium sulfate was added at 4C slowly over the span of an hour to salt ou t the protein. Ammonium sulfate increases the overall volume when salting out a protein so the appropriate am ount was calculated to include the specific volume of amm onium sulfate. The mass of salt per initial volume ( G ) (Table 3-3) is solved for W iiffM m CVCV (3-2) s W iW siW iiW f ii fW fGVM CMG mVVM VCMm V VC VM m C 1 (3-3) Lg LmL gmL molg M M MLg VMC CCM GsWf ifW/84.301 /1000 /54.0 /14.1325.093.31 0.093.35.093.3/08.519 1 (3-4) where Vf is the final volume, Vi is the initial volume, m is the mass of salt to add, Vs is the specific volume of ammonium sulfate (0.54 mL/g), Cf is the final concen tration of ammonium sulfate (product of final percen tage and molar saturation at 4C of 3.93 M which varies with temperature), Ci is the initial concentrati on of ammonium sulfate, and MW is the molecular weight (132.14 g/mol). The ammonium sulfate was allowed to reach equilibrium with the protein for 3 hours while stirring at 4C. The precipitate was then centrifuged with a JA-10 rotor at 8,000 rpm for 53

PAGE 54

30 minutes at 4C. Less than 10 mL of HKB (20 mM Hepes pH 7.4 with KOH, 50 mM KCl) with one tablet of complete protease inhibitors mini tablet was used per one liter of original bacterial broth. The solution wa s then clarified by centrifuging with a JA-20 rotor at 15,000 rpm for 15 minutes at 4C. Five mL of Talon Ni-NTA resin was equilibrated according to manufacturer specifications and mixed with the supernatant for 30 minutes at room temperature to bind the ActA to the resin. The resin was then washed according to manufacturer specifications and the protein eluted using HKB with 250 mM imidazole and fractions were collected. Remaining protein was eluted using HKB with 500 mM imidazole and was discarded. The kept fractions were measured using a Bradford assay to ge t a rough estimate of the protein concentration. Fractions with the majority of ActA were dialyzed overnight in HKB and a Bradford assay was used to determ ine the final concentration (76). Fluorescent Labeling of ActA-His6-Cys Strain DP-L4363 has an additional cysteine grou p as part of the protein which a thiol group was attached to labeling the protein with a fluor ophore. First, the absorbance was measured at 280 nm and at the wavelength of the fluorophore. Dithiothreitol was added at 20 times the concentration of ActA to break up any cystine formation and incubated at 4C for 30 minutes while rotating. Quick dialysis was done to remove DTT by doing 4 dialysis changes over a 4 hour period using a total of 1 L of HKB The maleimide fluorophore was added at a concentration of 20 times of ActA and incubate d overnight at 4C while rotating (wrapped in foil). -mercaptoethanol was added at 20 times the ActA concentration to bind any remaining dye and incubated at 4C while rotating for 30 minutes. Quick di alysis was done as before and the absorbance was measured at 280 nm and at the fluorophore wavelength to determine the amount of labeling (40). 54

PAGE 55

Vesicle Preparation Vesicles are formed when a dried lipid layer is exposed to an aqueous solution. Vesicles are made with a rotary evaporator (rotovap) whic h reduces the pressure of a continuously rotated round bottom flask. Lipids are dried on a round bottom flask and rehydrated with the buffer of choice (77). Making vesicles with 0.1 mg to 0.2 mg of lipids using a 100 mL round bottom flask yields a large amount of vesicles per milliliter. The lipids are th en rehydrated with 1 to 2 mL buffer. Small vesicles were made by filling the rotovap tank with de-ionized water and heating to 65C. Glass was cleaned by usi ng a modified acid wash technique (74). The appropriate amount of lipids was added to the round bottom flask, at tached to the rotovap, and a stream of N2 was applied for 5 minutes. Nitrogen was stopped, a v acuum applied, and the ro tovap set to a rotation of 60 rpms. Once the chloroform was evaporated leaving a layer of lipids on the surface the flask was removed and placed under a N2 stream for 30 minutes. Lipids were rehydrated with a buffer (Table 3-4) and allowed to rotate on the rotovap for seve ral hours or overnight. Large vesicles were made in the same method as a bove except using a beaker and no agitation. The vesicles were also allowed to rehydrate undisturbed for 2 to 3 days (78). Vesicles have a Ni-NTA lipid incorporat ed to specifically bind to the His6 tag on ActA. ActA was bound to the Ni-NTA lipid by mixing with prepared vesicles and incubating overnight at 4C without shaking. If the vesicles and ActA are agitated the vesicles will not be motile when added to a motility assay. Table 3-5 lists the ratios and dilutions used to bind ActA to vesicles. Motility Assay A motility assay is a mixture of components th at induces actin polymerization when used with an object having ActA bound to the objects su rface. The motility assay consists of an ATP 55

PAGE 56

regenerating component, a protein degradation preventing component, a thiol reducing agent, a cell extract, and an object w ith ActA bound to its surface. Creatine kinase The creatine kinase mixture is necessary for regenerating ATP from ADP and Pi (79). The creatine kinase mixture was made as a 10x solution containing 100 mM phosphocreatine disodium salt hydrate enzymatic, 20 mM ATP, 20 mM EGTA, 20 mM MgCl2, and creatine phosphokinase at 51 units/mL. One tenth of this mixture was mixed with the motility assay to produce the correct concentrati on needed for ATP regeneration. The mixture was made and stored at -70C. Protease inhibitors and DTT Protease inhibitors (PI) are mixed together as a 10x solution with 10 g/mL for each inhibitor leupeptin, chymostatin, and pepstatin A. The PIs are mixed in DMSO and stored at -70C. Dithiothreitol (DTT) is the thio l reducing agent and was made fresh daily to a concentration of 100 mM. Results Several variations of vesicle motility were atte mpted. The best results (the most vesicles with actin rocket tails) were obtained by diluti ng BBE by half, an [ActA]/[Ni-NTA] ratio of 0.02, and an actin concentration ranging from 3 M to 5 M. We found that rati o of [ActA]/[Ni-NTA] cannot be too high or excess ActA will promot e actin polymerization in solution (producing background filaments), and it cannot be too low or there will not be enough filaments associated with a vesicle to visualize the actin rocket tail or produce motility. An actin concentration in the range of 3 M to 5 M was found to produce long, fast growing actin rocket tails. Gel-filtered actin and non-gel-filtered actin were compared in motility assays with vesicles and Listeria No difference was observed in using either actin so non-gel-filtered actin was used throughout this 56

PAGE 57

study. It is likely that the ge l-filtered actin did not change mo tility outcomes because the actin was added back to the cell extract, which already contains actin oligomers and other protein that are normally removed in gel-filtering. Many of the vesicle rocket tails (>50%) were found to emanate from a mass of vesicles (Figure 3-3) most likely due to th e lipids tendency to aggregate, however, vesicle aggregation did not prevent vesicle motility. The velocity of ac tin propelled vesicles was determined by using actin monomers labeled with different fluorophores. Vesicles were exposed to one color of actin for a few minutes and then another color of actin for a few minutes. The length of the rocket tail for each color was measured and velocities were determined from the amount of time the actin was in the flow chamber. Most vesicles had tw o color actin rocket tails while a few vesicles only incorporated one color of actin and produced short rocket tails (Figure 3-4). One color rocket tails are probably due to lack of filament nuc leation (only showing the second color of actin added) or a poisoning of the motor protei n complex preventing inte rcalation of new actin (only first color of actin appeari ng in rocket tail). Figure 3-5 is an actin rocket tail exposed to 3 M of actin for 5 minutes of rhodamine labele d actin and then 5 minutes of Oregon-green labeled actin with a measured velocity of 2.5 m/min for the Oregon-green portion and 3.5 m/min for the rhodamine portion. Figur e 3-6 shows vesicles exposed to 5 M of actin for 3 minutes of rhodamine labeled actin, 2 minutes of Oregon-green labeled actin, and 3 minutes of rhodamine labeled actin. The veloc ities in the green portion are 3.5 m/min, 3.1 m/min, and 3.3 m/min. The velocities of the first rhodami ne section (farthest fr om the vesicle) are 3.5 m/min, 3.6 m/min, and 3.7 m/min. The average velocity of 41 actin rocket tails (5 1 M actin) propelling vesicles taken from 8 experiments was 2.8 1 m/min with a 57

PAGE 58

velocity range of 1.2 to 5.2 m/min (Figure 3-7) and the average radius of the analyzed vesicles was 0.4 0.1 m (Figure 3-8). These findings are in agreement with the reaction-diffusion model proposed by Dickinson and Purich (43). In this model, vesicle speed is determined by rate of monomer addition to filament ends at the tails center. This rate is determined by the net rate of monomer addition, which is limited by monomer diffusion (with charact eristic diffusion length ta ken as the particle radius, R ). Accounting for reaction and diffusion as ra te-limiting processes in series, the vesicle speed V can be approximated by DRk Cdk Vf f/1 (3-5) where kf is the monomer on-rate constant (10 M-1s-1= 0.017 m3/s), C is the bulk monomer concentration, d = 2.7 nm is the added filament length per monomer, is the filament end density (~103 m-2) at the vesicle surface, and D is the monomer diffusivity in the tail (~5 m2/s). Parameter values are justified in Dick inson and Purich (43). The measured speeds are plotted against particle size in Figure 38 and compared to the predicted speed from Equation 3-5. Although the scatter in the data obscures any dependence on particle size, the measured speeds do agree well with the predictions. No saltatory motion or tear drop shapes were observed in the analyzed vesicle velocities. Although, vesicles were too sm all to observe possible shape conformations, saltatory motion would be apparent in the rocket tail by fluorescent fluctuations in the rocket tail if deformations to the vesicle had happened. This lack of saltatory motion for small vesicles is consistent with the findings of Plastino et al (71), who found saltatory moti on of motile particles only for diameters exceeding about 3 microns. This transition was explained by Dickinson & Purich (43) 58

PAGE 59

as the result of more-uniform filament elonga tion rates on the smaller due to smaller diffusion gradients. Most vesicles observed were either too small to observe fluctuations in vesicle shape or too rare to locate larger vesicles in time to capture time-lapse. Therefore larger vesicles with a shape change from actin filaments were observed after polymerization had come to a stop, although only a handful of these events were observed. Figure 3-10 shows vesicles with actin tails that have variations in fluorescent in tensity along the tail length. Th e brighter portions (in the shape of a tear drop) in the rocket tails are from an accumulation of actin filaments. For fluorescent fluctuations to occur, the ve sicle forward motion must lag while filaments continue to polymerize. Larger vesicles (>3 m) were successfully prepared (Figure 3-11), but there was insufficient time for enough experiments to allo w detailed analysis of motility. Of the large motile vesicles created, some vesicles were observed to be distended as if in transition to a tear drop shape or conformation change. Discussion As argued in Dickinson & Purich (43), the obs ervations of saltatory motion in vesicles arise naturally from the actoclampin model which proposes filaments are pers istently attached to the surface. Filaments along the periphery have a consistent supply of actin monomers whereas interior filaments must wait fo r monomers to diffuse through th e rocket tail (without being incorporated into other filaments) before elong ation can occur. Monomer diffusion creates an actin monomer gradient with a lower concentrat ion of monomers availa ble to the interior filaments as compared to the exterior filaments. Lagging interior filaments prevent the rear of the vesicle from moving forward while exterior fila ments continue to push the rest of the vesicle 59

PAGE 60

60 forward creating the tear drop shape which is a result of diffusion-limited filament elongation instead of stress-dependent filament elongation (58). Although surface forces could not be determined in our experiments, the velocity of vesicles could be measured by determining rock et tail growth from fluorescent color change. The average velocity of vesicles was 2.8 1 m/min with a velocity range of 1.2 to 5.2 m/min (Figure 3-7) with a noted higher velocity associated with the rhodamine portion of the actin rocket tail (Figure 3-5 and Figur e 3-6). This velocity difference could be due to the actin monomer (labeled with rhodamine) being more competent or possibly actin labeled with different fluorophores incor porated into actin filaments at different rates. The vesicle velocity versus vesicle radius (Figure 3-8) agrees with theoretical predic tions (43), as did the lack of saltatory motion of these vesicles smaller than 1 to 2 m in diameter. These smaller vesicles are reaction-limited and not diffusion-li mited, which would create the characteristic tear drop shape observed for large vesicles. Giardini et al (41) reported an average velocity of 3 m/min ranging from 0.6 to 4.2 m/min while Upadhyaya et al (40) reported 0.8 0.2 m/min. While the results of this study are more in line with Giardini et al ., Upadhyaya et al. could have reported a slower velocity because they observed larger vesicles or because their motility assay was not as potent thereby reduc ing the actin polymerization rate s. Other groups have reported velocities of actin propelled beads (1 to 6 m/min) (33) and actin propelled Listeria (3.0 0.2 m/min) (59), similar to the velocities we f ound for vesicle motility. This similarity of actin propelled velocities among diff erent types of particles of diffe rent size suggests that actin polymerization depends primarily on the actin co ncentration and not on the type of cargo being propelled.

PAGE 61

Table 3-1. Necessary ratios for a parallel diluti on to perform a Bradford assay using a stock of 0.2 mg/mL BSA. Concentration (g/mL) Volume BSA (l) Volume H2O (l) 0 0 1000 1 5 995 5 25 975 10 50 950 15 75 925 20 100 900 25 125 875 Table 3-2. List of buffer volumes based on 1 kg of rabbit muscle. Buffer Volume (L) Extraction 5 NaHCO3 4 Tris 8 Acetone 12 H2O 40 Table 3-3. Reference table for am ount (g) of ammonium sulfate ((NH4)2SO4) to add at 4C. 0% 26.3 53.4 81.2 109.9 139.5 170.0 201.4 233.8 267.3 301.8 337.5 5% 26.7 54.1 82.4 111.6 141.6 172.6 204.6 237.6 271.6 306.8 10% 27.0 54.9 83.7 113.3 143.8 175.3 207.9 241.4 276.1 15% 27.5 55.8 85.0 115.1 146.1 178.2 211.2 245.4 20% 27.9 56.6 86.3 116.9 148.5 181.1 214.7 25% 28.3 57.5 87.6 118.8 150.9 184.1 30% 28.7 58.4 89.1 120.7 153.4 35% 29.2 59.4 90.5 122.7 40% 29.7 60.3 92.0 45% 30.1 61.3 50% 30.6 55% 61

PAGE 62

Table 3-4. Differences in experimental procedure for vesicle motility. Giardini et al. (41) Upadhyaya et al. (40) Extract used Xenopus laevis egg cytoplasm Bovine brain Motile vesicle size 1 to 5 m 1 to 5 m Rehydrating buffer 10 mM Tris, pH 7.5 100 mM NaCl/13%(w/w) sucrose 20 mM Hepes, pH 7.4 100 mM KCl/1 mM MgCl2 ActA labeling Alexa 488 succinimidyl-ester Fluorescein maleimide Vesicle sizing 5 cycle freeze-t haw 1 min. @ 1000g centrifugation Lipid molar %, MW: Phosphatidylcholine 46%, 760.09 90%, 760.09 Chloesterol 50%, 386.66 0% (Fluorescein or Oregon Green)phosphatidylethanolamine 2%, 744.05 0.5%, 1086.25 Ni-NTA 2%, 1057.02 10%, 1057.02 Table 3-5. ActA conj ugation with vesicles. [ActA] [NTA] Dilution of Vesicles [NTA] [ActA] ActA (l) Total (l) Vesicle (l) Rehydrating Buffer (l) 0.02 0.5 3.25 0.065 1 107.7 53.8 52.8 0.02 0.2 1.30 0.026 1 269.2 53.8 214.4 0.02 0.1 0.65 0.013 1 538.5 53.8 483.6 0.10 0.5 3.25 0.325 1 21.5 10.8 9.8 0.10 0.2 1.30 0.130 1 53.8 10.8 42.1 0.10 0.1 0.65 0.065 1 107.7 10.8 95.9 62

PAGE 63

Figure 3-1. Unilamellar bilipid layer vesicle w ith a portion expanded to show the configuration of lipids. Figure 3-2. Vesicle conformation change. A) Vesicle in solution with no external forces and membrane tension balanced with osmotic pressure. B) Vesicle deformed by actin rocket tail. C) Hypothetical forces on a vesicle surface from actin polymerization (40, 41). Red arrows designate a compre ssive force, blue arrows designate an extending force. 63

PAGE 64

Figure 3-3. Vesicles emanati ng from a central vesicle mass. Figure 3-4. Vesicles expo sed to rhodamine actin and then Oregon-green actin. 64

PAGE 65

Figure 3-5. Color change experiment with vesi cles exposed to 5 minutes of rhodamine labeled actin and then 5 minutes of Oregon-green labeled actin (3 M actin). Figure 3-6. Color change experiment with vesi cles exposed to 3 minutes of rhodamine labeled actin, then 2 minutes of Oregon-green labe led actin, then 3 minutes of rhodamine labeled actin (5 M actin). 0 2 4 6 8 10 12 14 16 0 11 22 33 44 55 6 Vesicle Velocity ( m/min)Frequency Figure 3-7. Histogram of vesicle velocities. 65

PAGE 66

0 1 2 3 4 5 6 7 8 9 0 0.2 0.4 0.6 0.8 1 Vesicle Radius ( m)Vesicle Velocity ( m/min) Figure 3-8. Vesicle velocities versus vesicle radius with data (circles) and the theoretical velocity maximum (line (5 m)) (43). Figure 3-9. Theoretical particle velocity at different actin concentrations (43). 66

PAGE 67

67 Figure 3-10. Overlay of rhodamine labeled actin propelling an Oregon-green labeled vesicle. The red spheres are points of accumulating actin due to lagging of the trailing vesicle edge. Figure 3-11. Large unlabeled vesicles conjuga ted with green fluores cent ActA propelled by rhodamine labeled actin.

PAGE 68

CHAPTER 4 ELECTRON MICROSCOPY OF ACTIN FILAMENTS Introduction A key distinction between th e actoclampin model and free filament models for actin polymerization and force generation is whether elongating filament ends are tethered to the motile surface. Several groups have analyzed actin rocket tails pr opelling beads (32, 33, 35, 36, 80) but have not reduced the number of actin filaments to show single filament interaction with a motile surface. We show that by optimizing bead size and biochemical conditions, we can observe single actin filament (+)-ends binding and staying persistently attached (throughout several washes and treatment) to the surface of a bead under an electron microscope (EM). An electron microscope functions the same as optical microscopes except EM uses a beam of electrons instead of light. The electron be am is created by a cathode (tungsten filament) and accelerated by a positive electrical potential through a vacuum (whi ch reduces beam scatter) and is then focused on the sample by electromagnets (magnetic lens) (81). In transmission electron microscopy (TEM), the electron beam is scattere d as it passes through the sample and the signal is collected, resulting in an image. With scanning electron microscopy (SEM), the beam is scattered by the sample as it is scanned across th e sample surface, but instead of passing through, it is reflected and the signal is collected. If the sample has a high atomic weight (such as a metal) then details can be resolved down to th e Angstrom level (82). A biomaterial sample will not have a high atomic weight in comparison to metal and will not sca tter enough of the electron beam to produce an image. Therefore, the sample must be coated with a metal to increase the amount of electron scatter. Becau se the sample is blanketed with a metal coating, some fine details are concealed even though resolution of 5 nm is achieved. With the resolution of this 68

PAGE 69

technique, single filaments can be distinguished from bundles of f ilaments and the orientation of filament to bead was determined. Cameron et al. (33) observed actin prope lled polystyren e beads by EM under conditions that reduced the number of filaments in contact with the bead surface to as few as 5 filaments appearing bound to the particle surface by their (+)-ends, and ev en single-filament tails on 50-nm beads. These findings are difficult to e xplain by the Brownian ratchet model (45), so Cameron et al. (33) suggeste d filaments might be adding mo nomers between transient tight associations with the surface. C onsistent with this explanation, they reported that many of the beads were devoid of filaments and many branched filaments lacked beads suggesting filaments nucleated at the beads surface but were subsequently detached. An alternative explanation is the filaments elongated for some period of time before detaching spontaneously. The observed detachment in the Cameron et al. experiments may be explained by the method by which beads were functionalized and treated. First, ActA was ionically bound to the bead surface which is much weaker than a covale nt bond. ActA could dissociate from the bead surface or some beads may not end up with a f unctional ActA on the surface at all. Second, beads were not bound to the substrate so many of the single filaments attached to beads could have been washed out in EM treatment. We have performed a systematic EM study to determine whether sing le filament elongate processively from 50-nm ActA-coated beads. ActA was covalently bound to the bead surface allowing for a much stronger surface to protein bond. Beads were bound to the glass surface to prevent wash out of sample during EM treatment. Silica beads were used for a stronger covalent bond to ActA (silanation) and a reliable met hod to purify beads from excess ActA and bind beads to a glass substrate (silica de nsity is greater than polystyrene). 69

PAGE 70

Materials and Methods Functionalized 500 nm Beads Five-hundred-nm amino modified polystyrene beads (Polysciences cat#07763) were diluted with Mops buffer (MB 100 mM Mops pH 7.0 with KOH). The diluted particles were centrifuged with a tabletop centrifuge at 14,000 rpm at 4C for 5 minutes to pellet. The supernatant was discarded and beads were re suspend in MB and centrifuged as before. Twenty mM bis(sulfosuccinimidyl suberate) (BS3) was prepared in MB during centrifugation just before use. The supernatant was discarded and beads were resuspended in MB supplemented to 2 mM BS3 and incubated at room temperature with shaking for 15 minutes. The beads were centrifuged as before. ActA was diluted to about 200 to 250 g/ml in MB to an equivalent volume used when charging the particles. Beads were resuspend in prepared ActA and incubated one hour at room temperature with shaking. The beads were centrifuged as before. The supernatant was removed and th e absorbance was measured to determine the amount of remaining protein in solution. Beads were resuspended in glycine methyl ester (GME) in MB (MB with 100 mM GME) then centr ifuged as before and aspirated. The GME washing step was repeated one more time and then again using only MB. Beads were then resuspended in bovine brain extract buffer (BBEB 20 mM Hepes pH 7.5, 1 mM MgCl2, 1 mM EGTA, 100 mM KCl, 0.2 mM CaCl2, and 150 mM sucrose). Beads we re stored in a sealed tube with parafilm at 4C and are typica lly functional for one to two weeks. Preparation of Listeria monocytogenes Overexpressing ActA Listeria was prepared as described in Chapter 3. Listeria growth was stopped once in the log phase growth and was washed twice with phosphate buffer saline (PBS). Listeria was killed with a 20 minute incubation in 10 mM iodoacetic acid (83). Listeria was washed three times with PBS and diluted to 1/5 the original volume in BBEB (84) a nd stored at -70C. 70

PAGE 71

Functionalized 50 nm Beads A 95% ethanol solution was prepared and the pH adjusted to 4.5 to 5.5 using acetic acid (CH3COOH). Two percent 3-aminopropyltriethoxysila ne (APES) in 95% ethanol solution was prepared and mixed for 15 minut es. Beads (Polysciences cat #24040 stock is at 5.63% (v/v)) were then diluted 10x with 95% ethanol. Beads were washed two times with 95% ethanol and pulse sonicated between washes (30 pulses at 20% level 1 power). The beads were then aspirated and the APES solution was added and mixed for 3 minutes. The functionalized beads were washed two times with 95% ethanol and pulse sonicated between washes (30 pulses at 20% level 1 power). The supernatan t was removed and beads allowed to cure overnight at room temperature in humidity <60% (alternative is to h eat to 100C for 30-60 minutes to cure). Beads were then resuspended in water, pulse sonicate d (30 pulses at 20% on leve l 1 power), and stored at 4C. ActA and BSA Conjugated to 50 nm Beads The beads surface was activated with 0.2% glutaraldehyde and incubated at room temperature for 30 minutes while shaking. Th e beads were then washed in BBEB and pulse sonicated (30 pulses at 20% level 1 power). Th e amount of protein to add to the beads was determined by first calculating the surface area of beads by taking the pr oduct of the volume of stock beads and surface area of one bead by the volume of one bead (0.338 m2 for 50 L of 5.63% (v/v) stock beads). Th e mass/area of ActA (0.905 mg/m2) was then determined by taking the molecular weight of ActA (68,000 g/mol for ActA (85) and 66,000 g/mol for BSA (86)) divided by the circumference of th e Stokes radius of ActA (6.3 nm for ActA (85) and 3.5 nm for BSA (86)). The product of bead area and protei n mass/area resulted in th e theoretical amount of protein to coat the beads. Pr otein was added to the beads and incubated for 45 minutes at room temperature while shaking. The beads were then washed in BBEB (the supernatant was saved to 71

PAGE 72

test the amount of protein not bound) and pulse sonicated. If beads were to be fluorescently labeled, the fluorescent BSA was ca lculated and conjugated to the b ead at this point. Beads were then washed 2 times in 100 mM GME in BBEB and pulse sonicated between washes. Beads were washed once more in BBEB and stored at 4C. Flow Chamber for Exchange Experiments Flow chambers were made using parafilm or double sided tape arranged in two parallel lines as support for a cover slip on top of a microscope slide. A pipette was used to introduce material to the flow chamber while filter paper was used to extract the fluid creating the flow. To bind beads or Listeria in the flow chamber, the cover slip was first treated with APES and glutaraldehyde (vesicles will adsorb to the glass surface and no glass treatment is necessary). Three-aminopropyltriethoxysilane treated cover slip s were made by exposing clean cover slips to a 2% APES in 95% ethanol solution (APES solution was mixed for 15 minutes prior) for 2 minutes. Cover slips were then washed in a porcelain tray 2 times with 95% ethanol in a beaker. Excess liquid was removed and cover slips incubated overnight at room temperature unsealed for curing (humidity <60%). Beads / Listeria with Actin Rocket Tails An APES cover slip was charged with 2% glutaraldehyde for 20 minutes at room temperature. The cover slip was washed with wa ter, a flow chamber prepared, and diluted ActA beads/ Listeria were flowed through (an appropriate bead/Listeria cover slip surface coverage was determined by testing different dilutions) and incubated at room temperature for 20 minutes. Excess binding sites were quenched with BBEB containing 100 mM GM E for 20 minutes. The chamber was washed with BBEB and a motility assay was added (Chapter 3) for the desired amount of polymerization time. 72

PAGE 73

Preparation of Sample for Viewing with EM Two percent glutaraldehyde in 0.1 M sodi um cacodylate buffer was added to the flow chamber for 20 minutes. A 0.1% tannic acid wa s then added for 20 minutes and rinsed with 3 exchanges of water. Uranyl acetate (UA) at 0.2% was added for 20 minutes (87). The entire slide was submerged in water and the cover slip removed. A metal cage with lens paper at the bottom was submerged in the water and the cover slip was placed inside the cage sample side up (never exposed to air). Several samples were stacked in the cage separated by lens paper (including lens paper to cover th e top cover slip). The cage co ntaining samples was transferred to a 10% ethanol solution and incubated with mixing for 5 minutes. The cage was then transferred in series to 20%, 40%, 60%, 80%, and 2 times in 100% for 5 minutes each, then 0.2% UA in 100% ethanol for 20 minutes, and finally 2 times in molecular sieved 100% ethanol for 5 minutes each (87). At this point the samples were optionally placed at 4C overnight. Critical point dryer (CPD) The CPD was filled with molecular sieved 100% ethanol and the cage containing the samples was placed in the CPD and sealed. Liquid CO2 was added until all ethanol was removed from the chamber. The temperature and pressu re was increased to the critical point of CO2. CO2 gas was removed until completely evacuated. Th e sample was removed and either coated for SEM or TEM or placed in a de siccator until coating was done. Sputter coating for SEM The cover slip with sample was placed on a 13 mm SEM stub using double sided tape to affix (sample side up). The sample was then pl aced in the sputter coat ed and filled with argon gas. The pressure was then reduced and the sample was sputter coated with gold palladium at 45 milliamps for 30 seconds. The sample was then imaged using SEM. 73

PAGE 74

Rotary shadow for TEM Samples were placed in a rotary shadower at an angle of 45 for platinum coating. The chamber was evacuated and the sample was coated with carbon platinum with a setting of 2.1 kV and 60 mA. Samples were coated with a 5 nm thickness of platinum which was monitored using a quartz thickness monitor. The sample angle wa s adjusted to 90 for carbon coating with a setting of 1.9 kV and 90 mA. The sample was co ated for 8 seconds resulting in approximately a 5 nm coating of carbon (to preven t sample from breaking apart). Post shadowing / pre TEM treatment A razor blade was used to cut 0.5 mm2 sections from edge to edge across the entire surface of the metal coated sample. Areas that did not co ntain sample were colored with a black marker. The cover slip (sample side up) was floated on 10 % hydrofluoric acid (HF) until the glass cover slip sank (about 15 minutes). A platinum loop was used to transfer the floating pieces to water. Again the platinum loop was used to transfer pi eces to a 100% solution of bleach (6.13% sodium hypochlorite (NaOCl)) and incubated for one hour. Pieces were transferred to water and then placed on polyvinyl formal coated 200 mesh TEM grids. Grid containing samples were dried overnight at room temperature and later vi ewed with TEM with a voltage of 75 kV. Polyvinyl Formal Coated TEM Copper Grids Polyvinyl formal coated copper 200 mesh grid s were used as support for samples in TEM imaging. To coat grids with polyvinyl formal, formvar (0.25% to 0.4% (w/v)) was mixed with chloroform in a glass container large enough for a microscope slide to be submersed. A glass microscope slide was cleaned using 70% ethanol dried with a Kim wipe. The slide was coated with a household cleaning solution and polished w ith lens paper to prevent the plastic from sticking strongly to the glass surface when trying to separate in water. The glass slide was dipped into the polyvinyl formal and allowed to in cubate for a few seconds. The slide was then 74

PAGE 75

removed at a constant velocity to give a smooth uni form coating. The faster the slide is removed the thicker the film will be on the slide. Thicke r layers were formed by dipping the slide several times (allowing the chloroform to evaporate between dips). A sharp metal edge was used (such as the edge of tweezers or a razor blade) to scrape the edges of the glass slide and to cut across the middle of the slide so the film will separate. At approximately a 30 to 45 angle to the water the slide was slowly submersed causing the film to separate from the slide and float away. Grids were placed on the film (rough side up, shiny side down) avoiding any discon tinuities in the film and avoiding film edges. A note card or parafilm cut to the size of the film was placed on top and in a sweeping motion the card was dunked and rota ted up to pull the grids and film out of the water on top of the card or parafilm which was then dried for 24 hours. Filament Shearing from Fluid Flow Based on EM images, filaments are obviously affected by the treatment for TEM because filaments tend to line up in the same direction (Figure 4-7). To determine if the treatment velocity is enough to shear a filament from the bead a simple Poiseuille flow was used to determine the amount of shear exerted on the filament. The velocity of flow through the chamber was solved Hzzuy 22 (4-1) where is the pressure drop, is the viscosity of the fluid (assu med to be as viscous as water, 0.01 g/cms), z is the position in the flow chamber, and H is the height of the flow chamber (~200 m). To estimate the pressure drop, the av erage velocity through the flow chamber is approximated to be 0.25 cm/s based on a standard flow chamber with a volume of 15 L, flow time of 10 s, chamber width of 3 mm, and chamber height of 200 m. Taking the average of 75

PAGE 76

Equation 4-7, the pressure drop is calculated to be 7.5 Pa/cm. With a basic understanding of the flow chamber, the drag on the filament fr om the fluid motion is calculated (88) y duM A L F 12ln 4 (4-2) where L is the filament length (10 m), A is the aspect ratio of the filament (length/diameter), M-1 is the resistance matrix, and Fd is the drag force resulting in a drag force on the filament of 30 fN. This small amount of force may be e nough to align filaments but is not enough to separate an attached filament wh ich can withstand at le ast 8 times this amount of extension force (40). Results As a positive control for motility from substratum-bound particles, Listeria was bound to a glass surface, exposed to a motility assay, and further treated for viewing with TEM (Figure 4-1 and Figure 4-2). Approximately 10% of Listeria observed using TEM grew actin rocket tails, which is similar to what is observed with bound Listeria using light microscopy. Some Listeria broke free of the glass slide and grew long twis ting rocket tails (Figure 4-1). A biomimetic system using ActA-coated 500-nm polyst yrene beads was developed to emulate Listeria actin-based motility. The 500-nm polystyrene beads were attached to a glass surface and exposed to a motility assay. The biomimetic part icles produced actin rocket tails similar to the Listeria rocket tails (Figure 4-3 to Figure 4-6). These large particles produced too many filaments to observe single filament interactions with the surface. The number of filaments was reduced by reducing the bead size from 500 nm to 200 nm which produced beads with fewer filaments but still too many to give conclusive evidence of filament attachment. The particles were reduced again from 200 nm to 50 nm which resulted in only a few filaments associated with each bead. 76

PAGE 77

Filament density was also controlled by adju sting the ActA surface density, the extract concentration, and the actin polymerization in cubation time. The optimal ActA density for observing single filaments corresponded to appr oximately 13 ActA proteins per bead. The extract was used at full strength and ac tin concentrations in the range of 3 M to 5 M produced single actin filaments quickly (less than one mi nute) (Figure 4-7 to Figure 4-9). To further control the number of filaments, the actin polymerization incubation time was adjusted. To determine the incubation time necessary to achieve one filament per bead, a time study was done by incubating the beads with a motility assay for 1 minute, 2 minutes, 5 minutes, 10 minutes, 20 minutes, and 35 minutes. Due to the large nu mber of filaments observed on the beads at time points 20 minutes and 35 minutes, only time points 1 minute thru 10 minutes were analyzed. To prevent human bias, 25 random images at each time point were taken. For each image, number of particles, filament length, and number of filaments per particle were measured. The average filament length was not statistically di fferent between the four time points. The histogram of filament lengths is shown in Fi gure 4-10. For filament lengths larger than ~100 nm, the distribution appear s to be exponential which is consistent with a Poisson distribution of elongation times. The lower counts of shorter filaments less than 100 nm suggest that filaments grow for several subunits before st opping. At each time point the average filament length is ~500 nm, but based on velocities we have determined, a le ngth of 500 nm would correspond to 10 seconds of growth. Velocities dete rmined for vesicles represent the growth rate of an ensemble of filaments where some filame nts could be lagging, sl owing the overall growth rate so single filaments could possibly be elongating at even faster rates. Therefore we can surmise that the majority of filaments have st opped growing by the time of fixation consistent with the reported histogram (Figure 4-10). 77

PAGE 78

The average filament number per beads with filaments increased over time, due to new filaments initiating and polymerizing from the su rface (Figure 4-11). To quantify the apparent increase of filaments, the number of filame nts was normalized to the number of beads which resulted in a steady increase of filaments per bead (Figure 4-12). Next we quantified free filaments and found on average 7% of obser ved filaments were not bound to any bead (Figure 4-12). The number of free filaments d ecreased over time which could be due to fewer binding sites available for free fila ments to attach. The lack of binding sites could arise from excess protein or extract sticking to the substrate preventing filame nt attachment. The histogram in Figure 4-13 shows the number of filaments per bead increases due to longer incubation times. To confirm ActA coated beads ar e initiating filament growth, the same motility experiment was tested replacing ActA coated beads with APES /glutaraldehyde treated beads or BSA treated beads. No filaments were observed in either cas e on beads or on the substrate surface. Filament polarity determined by myosin s ubfragment-1 (S1) was attempted but a conclusive result was not found (the attempted protocols and outc omes are presented in Appendix B). Discussion Analysis of several hundred images and several thousand single fila ments attached to beads shows a persistent attachment of filame nts to particles. These filaments reached a steady-state polymerization rate within a few minutes. The observed filament lengths ranged from 100 nm up to 4 m. In order for a filament to polymerize to a length of 4 m at a growth rate of about 3 m/min (Chapter 3), the filament must be associated with the bead for 1,400 monomer additions over more than a minute time period. The distributi on in filament lengths, with the exception of fewer numbers of shor ter filaments, can be explained by filaments nucleating and ceasing elongation at random times with constant rates. The fewer shorter filaments suggests filaments grow for many cycles before stopping. Shorter filaments either 78

PAGE 79

nucleated at a later time in the incubation pe riod, or possibly the mo tor protein (ActAVASP) was phosphorylated (89) preventing continued filament elongation, or some other unknown cause stopped polymerization. Even though fila ments have stopped elonga ting after a relatively short period, (on average ~10 s before elongatio n has stopped) the filaments are still strongly associated with the bead surface even after the several washes and treatments for EM. Single filaments emanating from 50 nm beads before EM treatment and fixation are observed using total internal fluorescence microscopy (TIRF) (F igure 4-14) suggesting f ilaments are not being detached from the beads due to further treatments. The observation that filaments stay pers istently bound to a bead surface after polymerization has ceased sheds light on what could be happening in an ensemble of filaments in an actin rocket tail. In an assay of 500 nm particles (polystyrene beads or vesicles such as Figure 3-4) with actin rocket tail s, some particles have long rocket tails while others have short dilated rocket tails. These shorter rocket tail s could arise from a few quiescent filaments that prevent the forward motion of the particle. Once a few filaments have stopped polymerizing but are still persistently bound to the particle surface, other active filaments may continue to polymerize until the filaments become so compresse d that additional monomers are not able to intercalate. The dilated short tails could be due to the contin ued polymerization of filaments while other seized filament s prevent forward motion. These findings are direct evidence that filament s polymerize while persistently attached to a motile surface and cannot be accounted for by a free-ended filament model, in which case the filament end would quickly diffuse away from th e bead before this length was achieved. The number of filament-particle asso ciations would become almost impossible to encounter. On average, 93% of observed filaments had ends asso ciated with a bead. If filaments were not 79

PAGE 80

polymerized at the surface of the bead but we re instead polymerized in solution and later associating with the bead surface then there woul d be a much higher incidence of filaments not associated with a bead. At longer polymerization times, several filament s were observed associated with a single bead, although some beads had no filaments, wh ich could be due to the functionality and orientation of ActA on the bead su rface. Unaltered ActA produced by Listeria has a transmembrane portion that orients ActA in the correct position on the surface of the bacteria (16, 18, 19). In this study, ActA was random ly covalently bound to the bead surface which could prevent functionality and pr event filament polymerization if ActA is not oriented correctly on the bead surface. Beads with no filaments could have eith er no ActA bound or have ActA bound that is not in a functional orientation. The increase in filament number with time demonstrates that filaments cont inue to nucleate over time (Fi gure 4-13). Camer on et al. (33) were unable to observe several instances of single filaments attached to beads most likely due to their approach of binding ActA to the bead su rface and not having a method to prevent wash out of sample from EM treatment. This study, how ever, has shown thousands of single filaments associated with beads, which supports the actocl ampin model and persistent filament attachment. 80

PAGE 81

Figure 4-1. Listeria (top right of image) freed from the substrate surface ( possibly by the actin rocket tail) is propelled by an actin rocket tail jutting from a larger actin network. Figure 4-2. Listeria with a small actin comet tail in its early stages. 81

PAGE 82

Figure 4-3. A 500 nm polystyrene bead with an actin cloud viewed using TEM. Figure 4-4. A 500 nm polystyrene bead with an actin rocket ta il viewed using TEM. 82

PAGE 83

Figure 4-5. Two actin rocket tails converge. Figure 4-6. Three 500 nm polystyrene beads combine to form one actin rocket tail. 83

PAGE 84

Figure 4-7. Single actin filaments emanating from 50 nm silica beads. Figure 4-8. A single actin filament associ ated with a single 50 nm silica bead. 84

PAGE 85

Figure 4-9. Several single actin filaments associated with single 50 nm silica beads. The scale bar is 300 nm. Figure 4-10. Histogram of filament lengths w ith bin size of 100 nm. A) One minute actin polymerization with 66 filaments. B) Two minute actin polymerization with 117 filaments. C) Five minute actin polymer ization with 180 filaments (one outlier of 4.7 m not shown on graph). D) Ten mi nute actin polymerization with 215 filaments. Graphs A, B, and D have the same axes values as graph C. 85

PAGE 86

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 0246810 Time (min)Filament Number / Bead Figure 4-11. Filament number per bead versus time. Blue triangles represent the average number of filaments per beads with fila ments and black squa res represent total number of filaments per total number of beads. 0.00% 5.00% 10.00% 15.00% 20.00% 25.00% 30.00% 35.00% 40.00% 45.00% 50.00% 0246810 Time (min) Figure 4-12. Number of filament s normalized to the total number of beads versus time. Blue diamonds represent beads with filaments pe r total bead count and black triangles represent free filaments per total number of filaments. 86

PAGE 87

0 20 40 60 80 100 120 12345 Number of filaments attached to a beadFrequency 1 minute 2 minute 5 minute 10 minute Figure 4-13. Histogram of the number of f ilaments attached to a bead versus time. 87

PAGE 88

88 Figure 4-14. Time lapse of single actin fila ments emanating from 50 nm beads. Beads are labeled with rhodamine and actin is labe led with Oregon-green. Motility assay started and focused after 4 minutes. Images taken at 20 second intervals.

PAGE 89

CHAPTER 5 SINGLE ACTIN FILAMENT POLARIZATION DETERMINED BY MULTIPLE LABELED ACTIN MONOMERS INCORPORATED INTO ACTIN FILAMENTS Introduction As shown with EM images, single filaments emanating from 50 nm ActA-coated particles could be observed and characteri zed. Although EM provides nanometer details, the system is not dynamic and polarity of the f ilaments can not be easily determined. Determining filament polarity is important to estab lish whether filament (+)-ends are attached and elongating by insertional polymerization at the bead surface, as would be expected in an end tracking motor such as ActAVASP, or instead (-)-ends are attached and filaments are growing with free ends away from the beads. For example, Brown and Spudich (90) showed polylysine coated polystyrene beads with the majority of (+)-ends of filaments pointing away from the bead surface. In their experiments, polylysine nucleat ed actin filaments and caused an increase in polymerization but does not orient th e filament at the bead surface. Our strategy for determining at which end act in is adding onto tethered filaments was to expose the bound ActA-functionalized beads to a cell extract cont aining fluorescent actin of one color for a few minutes, then switch to an extr act containing actin of another color for a few minutes. The resulting filaments were then obs erved under total internal reflection fluorescence (TIRF) microscopy, which has been successfully applied previously to single filament dynamics (54, 74) as well as insertiona l polymerization by substratum-bound formins (91, 92). Similarly, motor proteins (actin and myosin) were visuali zed using a two-color assa y with TIRF (93) and the recruitment and dynamics of various components of actin polymerization (including N-WASP and Arp2/3) was studied using a twocolor assay visualized with TIRF (94). Figure 5-1 illustrates some possible outcome s of a two-color si ngle actin filament experiment, here assuming green actin is added first followed by red ac tin. If the elongating 89

PAGE 90

(+)-end is located at the bead surface, a span of red F-actin w ould appear next to the bead adjacent to green span farther from the bead (Fig ure 5-1A). On the other hand, if the (-)-end is bound to the bead surface, the colors would be reversed with the gree n span bound to the bead (Figure 5-1B). A similar result would appear if th e filament were initially side-bound to the bead (Figure 5-1C), in which case green filament sp ans would appear on both sides of the bead and red adjacent to one span. It is also possible that the filament both nucleates and stops elongating while exposed to one color or the other, th ereby producing a filament of uniform color (Figure 5-1D). Finally, two separate filaments could polymerize from the surface of the bead but overlap and appear indistinguishab le as one filament. If the two filaments were of different colors, the conjoined filaments would appear as a yellow filament adjacent to the bead and the single-filament span would be either red or gr een away from the bead, depending on whether the longer filament grew before or afte r the color change (Figure 5-1E). These possible outcomes are assuming neglig ible (-)-end growth or shrinkage from ATP-actin or ADP-actin. Pollard et al. (95) dete rmined the polymerization rates of ATP-actin at the (-)-end to be k+ = 1.3 M-1s-1 and k= 0.8 s-1 and polymerization rate s of ADP-actin at the (-)-end to be k+ = 0.16 M-1s-1 and k= 0.27 s-1. For ADP-actin, the on rate is slightly less than the off rate causing a slow depolymerization of ADP-actin from the (-)-end, however, so slow that depolymerization would not be noticeable. The on rate of ATP-actin is slightly more than the off rate of ATP-actin causing a slow pol ymerization of ATP-ac tin at the (-)-end but polymerization at the (+)-end is 10 times greater than at the (-)-end so considerably more polymerization occurs at the (+)-end. If (-)-e nd addition was noticeable, then the insertional polymerization case (Figure 5-1A) would merely ha ve a short red portion on the end of the tail. Filament orientation could still be determined by the much longer red portion near the bead 90

PAGE 91

designating (+)-end growth. A th ree section filament consisting of one fluorescent actin split by another was not seen in any of the experiments. All two color filaments observed in this study resulted in the co rrect orientation for actin to be insertionally polymerized at the bead surf ace (Figure 5-1A and Figure 5-5). These results combined with EM images (Chapter 4) indicate single actin filaments are persistently attached while actin is insertionally polymerized, consistent with the actoclampin model. Materials and Methods Color Change Assay ActA functionalized 50 nm beads were bound to a flow chamber as described in Chapter 4. Actin was labeled with either Oregon-green or rhodamine (Chapter 3). A motility assay with 5 M non-labeled actin (black actin ) was flowed through the chamber for 2 minutes. Then, a motility assay with 5 M Oregon-green actin (green actin) was flowed through the chamber for 1 to 2 minutes. Next, a motility assay with 5 M rhodamine actin (red actin) was flowed through the chamber for 1 to 2 minutes. The order of ac tin addition to the flow chamber did not prevent intercalation of actin monomers to produce two co lor single filaments. Last, 1% glutaraldehyde was flowed through the chamber to fix the filament s. Samples were viewed within 1 minute of preparation using TIRF. Image Analysis Several macros were created to analyze imag es using NIH ImageJ software (Appendix C). To better visualize filaments a macro was create d to apply a high pass an d low pass filter to remove low and high frequencies from images. The laser used for TIRF produced a higher signal in the center of the image than at the e dges creating a low freque ncy background that was removed by a high pass filter (Figure 5-2 had a hi gh pass filter applied to the image). The high pass filter removed frequencies larger than 50 pixels. When the setting for the high pass filter 91

PAGE 92

was changed to remove frequencies greater than 50 pixels, the outcome of the image did not change. Background fluorescence and small partic les produced specs of noise in all of the images. These were removed using a low pass filter which filtered objects smaller than 3 pixels (Figure 5-3 had a low pass filter applied to the image). The combination of the high and low pass filters greatly increased the number of filaments that could be obser ved per field of view (Figure 5-4 is the combination of low and high pass filters on an image). The brightness and contrast for each image was optimized to show the filament. A line scan of each filament was done to show fluorescence intensity. Fluorescence intensity varied in different regions of the im age and varied between the two fluorophores used in experiments. Therefore fluorescence wa s normalized for each line scan by taking each fluorescent value and dividing by the maximum value along the line scan (for each color). This normalized data was then plotted (Figure 5-6 to Figure 5-9) to show relative intensity along the filament length and to show the fluorescent actin farthest from th e bead was not connected to the bead other than through the othe r fluorescent actin. Bead c ount was done using a macro to threshold for yellow beads and a circle count er was used to tally yellow bead events (Figure 5-10) as well as beads were counted by hand for verificat ion. Single filaments were counted by hand for each image. Results There are several advantages to fixing filame nts with glutaraldehyde as opposed to viewing filaments not bound to the substratum. First, se veral images can be taken while scanning the sample without worry of depolymerization. Background fluorescence is also removed with the flow through of glutaraldehyde for fixation (excess actin is flushed out). Second, filaments are bound to the surface and not fluctuating from Brownian motion providing a better image for analysis. Photobleaching of fluorescent filaments is prevalent in these experiments so static 92

PAGE 93

filaments allow for the image to be focused a nd taken once as opposed to the several images typically necessary with fluctuating filaments. Time lapse of single filaments is difficult because filaments fade due to photobleaching within 10 to 15 seconds of continuous fluorescence. Finally, since two color filaments have a low occurre nce, several field of vi ews must be taken. Filaments reach a maximum length quickly (<30 seconds) so live images are not practical for finding two color filaments. There were several criteria fo r counting individual filaments. Filaments had to be sufficient length (>1 m), filaments could not overlap with ot her filaments at critical points such as color change points or bead-f ilament attachment, filaments had to have a high signal to noise ratio and the filament had to be continuous, th e position of labeled filamentous actin did not overlap with other labeled fila mentous actin except minimally at the transition of the color change. These same criteria were used for c ounting single color filament s (when applicable). Some of the filaments appear to have variations in the fluorescence intensity. These variations could be from background autofluorescence of prot ein from the extract or from the amount of fluorescence monomer incorporated at any point in the filament. Filaments could also fluctuate from the surface reducing the fluorescent signal ou tput generated from the evanescent wave of TIRF. Figure 5-5 is a compilation of all 158 two color single filame nts observed in more than 150 image sets of 12 experiments. Approximately 20 of the 158 events were considered to be of excellent quality meeting all criteria. Figure 5-6 and Figure 5-7 show a two color filament bound to a bead where Oregon-green was added first and rhodamine actin followed. Both fluorescent channels are shown separately and a graph of the line scan along the filament shows the normalized fluorescence intensity. Figure 5-8 an d Figure 5-9 are the same as Figure 5-6 and 93

PAGE 94

Figure 5-7 except with rhodamine actin added firs t and then Oregon-green actin added second. The graph of the line scan (Figure 5-6 to Fi gure 5-9) clearly shows a fluorescence intensity switch from the bead to the end of the tail. The green portion of the f ilament is clearly not connected to the bead by green actin in Figure 5-6 and 5-7 as well as the red portion of the filament is not connected to the bead by red actin in Figure 5-8 and 5-9. Figure 5-10 is a histogram of bead and filament count for 20 imag e sets. Twenty-five per cent of beads had single actin filaments and 11% to 12% of beads counted had single filaments with 8% of the single filaments having two fluorescent acti n monomers incorporated. To test that filaments were being generated by ActA beads, the same two-color experiments were performed as before except ActA coated beads were replaced with BSA coated beads. Beads were also completely removed and a motility assay containing actin was added to the flow chamber as before. No filaments were observed in either of the controls. Discussion Over 158 events show a bead with a two color single actin filament attached. Every event that exhibited a two color actin filament resulted in the correct polarity of (+)-end closest to the bead independent of order of fluorescent actin added. Approxima tely 25% of beads counted had an actin filament attached compared to 35% obs erved in EM images at a comparable time of polymerization (5 minutes). The difference in obse rved filaments is due to the resolution limit of TIRF compared to EM. A crite ria of filaments being around 1 m or greater was set for counting filaments in TIRF. The length criterion of a si ngle filament in EM was much lower because filaments are easier to distinguish due to the high resolution associated with EM. Single filaments with only one color actin were ~11% of bead s counted for either color and independent of the order of fluor escent addition to samples. Out of the single filaments counted 8% were two color filaments. At first glance these numbers may seem low but can be explained 94

PAGE 95

from the results of EM. From the elongation rate determined, (~3 m/min) the average filament (~500 nm) measured in EM would only elongate for 10 seconds. Few filaments in EM were observed to elongate for much longer times and distances (max ~4 m) for unknown reasons (phosphorylation of VASP may not have occurred until a later time or some other poisoning device had not happened). Assuming the same mechanism is occurring in these TIRF experiments, filaments grow for 10 to 20 seconds and stop elongating but still remain persistently attached (two chamber flow thr ough before fixation). The addition of another fluorescent actin would then not in tercalate into the halted filame nt. Halted filaments would also explain the low occurrence of two color single fila ments. For a two color filament to occur, the filament must have started elongating in the fe w seconds remaining before the next fluorescent actin is added and the filament must continue to intercalate the new actin. The observation of new actin monomers intercal ating at the surface of a bead shows single actin filaments are persistently attached while ac tin is insertionally polymerized. Free filament models do not have an explanation to support th is observation. A free filament would easily diffuse away from a bead surface during the 10 to 20 seconds of actin polymerization observed in EM and TIRF. Insertional polymerizati on from ActA-coated beads strongly suggests elongation by end-tracking motors and supports th e actoclampin model for ActAVASP. Further experiments are required to esta blish that VASP (rather than, e. g., adsorbed formin) is the end-tracking protein responsible for actin assembly in these experiments. 95

PAGE 96

Figure 5-1. Fluorescent actin cha nge hypothetical scenarios with gr een added to the assay first, then red, attached to a yellow 50 nm bead The + and designate the (+)-end and (-)-end of the filament. A) Result if insertional actin polymerization is occurring. B) Result if the actin filament is nucleated in solution and bound to the bead with the incorrect polarity. C) Filament nucleated in solution and bound to the side of the bead. D) Single filament growth without incorporation of two colors. E) Overlap of filaments giving the appearance of A, except the red/green overlap would create a yellow filament. Figure 5-2. High pass filter re moving frequencies larger than 50 pixels applied to a mock filament. A) Mock filament with nothing applied. B) Low frequency added to the image. C) High pass filter applied to image. 96

PAGE 97

Figure 5-3. Low pass filter removing particles smalle r than 3 pixels applied to a mock filament. A) Speckles added to a mock filament image. B) Low pass filter applied to image. Figure 5-4. Low and high pass filters on a mock filament. A) Speckles and low frequency added to a mock filament. B) Low pass filter of A. C) High pass filter of A. D) Combination of low and high pass filters to A. 97

PAGE 98

Figure 5-5. Compilation of 158 color change events observed from 12 experiments and more than 150 image sets. Figure 5-6. Both fluorescent channels and overl ay of a two color filament where Oregon-green actin was added to the experiment first. Th e graph is of a line s can along the length of the filament. Scale bar = 1 m. 98

PAGE 99

Figure 5-7. Both fluorescent channels and overl ay of a two color filament where Oregon-green actin was added to the experiment first. Th e graph is of a line s can along the length of the filament. Scale bar = 1 m. Figure 5-8. Both fluorescent channels and overl ay of a two color filament where rhodamine actin was added to the experiment first. Th e graph is of a line s can along the length of the filament. Scale bar = 1 m. Figure 5-9. Both fluorescent channels and overl ay of a two color filament where rhodamine actin was added to the experiment first. Th e graph is of a line s can along the length of the filament. Scale bar = 1 m. 99

PAGE 100

100 Figure 5-10. Histogram for beads and filaments of 20 image sets.

PAGE 101

CHAPTER 6 CONCLUSIONS AND FUTURE WORK Discussion Improved understanding of actin-based motility has several advantages. The most immediate are a better understand ing of how cells work and how the body functions. With this understanding of actin polymerizatio n, a variety of biological issues can be approached such as how cancer spreads through the body, or spinal muscular atrophy cau sed by a deficiency of actin (96), or better medical technique s and drug delivery in fighting dise ases. Less apparent and still in the distant future are other applications such as biomimetic devices for separating and detecting a desired particle or the control of nanodevices such as nano-switches or nano-valves. Although actin has been studied for over 100 years the mechanism of actin polymerization is still disputed. In this work, experi ments and analysis of actin fila ments were done to elucidate the mechanism governing actin polymerization. The mechanical energy density of filaments in an actin rocket tail was found to be ~3 pN on average (Chapter 2) which is greater than the energy that could be stored by the free energy of monomer addition. Because free filament models rely on monomer addition to the free filament end as a means for energy, the maximum energy de nsity is about 2.7 pN. The energy density measured in Chapter 2 suggests that polymeriza tion yields more energy than could be provided by monomer addition alone, consis tent with the actoclampin mech anism, which also harnesses the energy of ATP hydrolysis. Energy is also lo st through heat dissipati on (friction), filaments depolymerizing, or periphery filaments with no mean s of translating energy to the rocket tail. Although the method in Chapter 2 has some lim itations, the approach provides a means to calculate an energy density of f ilaments that could be applied in other venues involving static filament images. 101

PAGE 102

Vesicles were found to be propelled by actin ro cket tails at ~3 m/min velocities. The vesicle velocities reported in Chapter 3 are simila r to other measured vesicle velocities (41) and other biomimetic velocities (32, 37, 38, 59, 69) reported in the literature. Vesicle velocities compared to vesicle diameter agree well with th eoretical calculations (4 3) indicating the small vesicles observed (<1.5 um) are more reaction-li mited than diffusion-limited. Some vesicles were slower than predicted which could arise from vesicle-to-vesicle variation (e.g. in filament density) or a poisoning of the ActAVASP mechanis m slowing motility. The velocities of this study give a basis for how well th e components prepared in our laboratory function and how well the components operate compared to published data Vesicle shape change could not clearly be determined from fluorescent images, but evidence of saltatory motion shou ld have appeared in the actin rocket tail if it were present. Saltato ry motion was not observed for the analyzed small vesicles. Single filaments were observed in EM to be persistently bound to a bead surface. Filaments quickly elongated then stopped for unknown reasons, possibly due to phosphorylation of VASP or some other poisoning of the end-tracki ng motor. Single filaments continued to stay attached to the bead surface we ll after elongation had ceas ed and after the samp le was treated for viewing with EM. A free filament would either diffuse or be washed away in this scenario. The actoclampin model is the only ac tin polymerization model that can explain filament elongation while staying persistently attached to a surface. The polarization of singl e filaments was further explored in Chapter 5. Here filaments inco rporated new actin closest to the bead surface indicating insertional po lymerization. The low number of two color filaments observed can be explained by the fast elongation ra tes of actin filaments. For a two color event to occur, a filament must have polymerized in the remaining seconds of the first actin environment and 102

PAGE 103

continue to polymerize with the new batch of ac tin. Again, the actoclampin model can explain insertional polymerization of filaments tethered to the particle at their (+)-ends. The work presented in this study significantl y contributes to the know ledge of actin based motility specifically on how single actin filaments interact with a surface. Further work with the techniques described in this study will be useful in clarif ying how actin polymerization is capable of producing forces. The results pres ented here support the actoclampin model and argue against models requiring free filament ends. Suggestions for Future Work Further work could be done in mapping fila ments from images obtained with EM. A less dense actin tail would increase the reliability of the measurements. Because the third-dimension is inaccessible in EM images, a simulation of an actin rocket tail propelling a bead could be created. The resulting image could be proj ected on a two-dimensional surface and the same method outlined in Chapter 2 performed on the si mulated filaments to determine how accurate the algorithm is at determining an energy density in an actin rocket tail. Large (>3 m) actin propelled vesicles were prepared but a lack of time prevented the study and analysis of the system. Several experi ments could be done using large motile vesicles. The velocity and saltatory motion of large actin propelled vesicles could be further studied in the same manner performed in Chapter 3. Actin rock et tail forces could be probed by anchoring the tail of a motile vesicle while the front surface of the vesicle (opposite side of the actin tail) is aspirated with a micropipette. The subsequent shap e change of the vesicle could be measured to determine the amount of force the actin rocket tail is exerting on the vesicle surface. Fluorescence recovery after photobleaching (FRAP) c ould be used to measure residence times of ActA at the surface of a vesicle. ActA would be fluorescently labeled, attached to a large vesicle and motility induced. Then FRAP would be used to photobleach ActA and to determine if ActA 103

PAGE 104

concentrations change at the vesicle surface. Phosphoinositides are known to increase actin polymerization (85, 97-99). The affects of a phos phoinositidyl lipid could be studied and used to accelerate nucleation time and possibly vesicle velocitie s. Various cargos could be inserted into a motile vesicle to determine viability for various applications (e.g. drug delivery or lab on a chip designs). Further work could be done with single filaments imaged in EM. Actin filaments were allowed to polymerize on single beads for up to 35 minutes as described in Chapter 4. Longer incubation times could reveal if filaments might start polymerizing again or if new filaments will continue to form. To determine polarity of filaments in EM, myosin subfragment-1 (S1) (Appendix B) (100) could be used to show directionality of si ngle filaments. Subfragment-1 binds to actin in a specific dir ection creating an arrow head on th e filament with the barbed end of the arrow pointing toward the (+)-end and th e pointed end of the arrow pointed toward the (-)-end of the filament. Directionality could also be determined by doing the same two-color method as in Chapter 5, except using regular actin and then biotin actin (o r vice-versa). Next, streptavidin gold particles would be added which will bind to the bi otin actin (101, 102) showing where new actin is intercalating. Further experiments with the tw o-color actin method could provi de more insight into what is happening with the filament. If incubati on times with each color actin were reduced to seconds (possibly 10 seconds for each color actin) th ere might be an increa se in the number of two-color filaments observed because of the veloci ties determined in Chapter 3. There could also be a decrease in the number of two-colo r filaments observed because nucleation time could play a role in how many filame nts polymerize in the given incuba tion period. Most likely, there is an optimal time that produces several two-color filaments in one experiment. 104

PAGE 105

105 Non-hydrolyzable ATP (AMP-PNP) (103-105) could be bound to actin for use in the two-color experiment. If polymerization was dependent on hydrolysis then no two-color filaments would appear, however, if hydrolysis did not matter, tw o-color filaments would be observed. More time experiments could be done with the two-colo r experiment. Filaments could be incubated with one color actin for a long period of time (1 hour) and then the second color actin would be added to see if filaments are able to polymerize new actin on existing filaments or if new filaments can form at all.

PAGE 106

APPENDIX A MATLAB ALGORITHM TO DETERMINE ENER GY STORED IN BENT FILAMENTS % v20 has the lengths^3 instead of just lengths for the X and Y matrices % 9-17-03 % v22 has the x and y combined into one matrix for solving, also has the % option of removing the bending portion for the end points % corrected all endpts (ie A(1,1)=S^-2+5*L/lengths(1)^3 to % A(1,1)=S^-2+6*L/lengths(1)^3) % 9-18-03 % v23 put values back in for end points, adjusted values at points next to % end points 9-22-03 % v24 solved for x and y positions in separate matrices % S is set to 1 nm % option to choose filament to evaluate reinstated % 9-24-03 % v25 mass filament evaluation with option of how many to evaluate at a % time and the average energy per filament with plots given % 9-24-03 % v255 same as last one but now plots are with found positions and contours % 9-25-03 % v26 all of the filaments evaluated with a contour plot of all filaments % and their dEds on one plot % 9-25-03 % v261 took derivative at each point instead of fitting a curve to the % values. Used those results to get average energy. % 10-1-03 % v271 made the contour plot relative to all filaments energy magnitude % 10-1-03 % v30 Now I am using ImageJ (downloaded from NIH) to obtain pixel locations % from the graphics. The data is automatically recorded by ImageJ in % a text file. To determine the end of a filament, the last point of % the filament is recorded twice. All data in a text file is manipulated % and plotted at the same time. % 10-8-03 % v31 Gives option to have the filaments numbered or to point out the % largest energy value or to have a legend % 12-9-03 % v32 Changed dE/dS=(B/3)*(dT*/dS*)^2 to dE/dS=(B*8/15)*(dT*/dS*)^2 % in program it's dEds=B/3*dTds2 to dEds=B*8/15*dTds2 % also changed the lamda component. There was an error in changing % over to position vector. The bending modulus (B) is divided by 2 % in the original energy equation, however this did not happen for % the position vector energy equation. Therefore instead of going % through and dividing lamda by 2 everywhere, the division is taken % place at the beginning of the program. % 2-2-04 % v321 Increased data set to a total of 97 filaments % 4-14-04 % v33 When I was calculating the lengths of the filaments by finding the % length of the hypotenuse between points I would then multiply this by the % nmperpix to convert pixels to nm but I have forgotten why I did this so % in this version I have taken that multiplication out. 106

PAGE 107

% I have also fixed the y axis so that it increases going up and % there are no longer negative values for the y axis. % For single filaments the x and y axes are in nm and so is the % length of the filament. The options for legend,max point, and % filament number are set to always be on % 4-20-04 % v34 Added a section after the last pause that is to analyze the % filaments. The plot is now in nm. Put in safety in calculating % lengths so that if a length is too long (>25pix) then the program % is paused because there is likely a filament that wasn't ended % correctly % 4-21-04 % v35 Determines the distance of each measured point from a center line. % Also plots the AlldEds vs DFC and a histogram of AlldEds. Took % out the first set of plots that were made in previous programs. % 4-22-04 % v36 Found correlation coefficient (cc) for the DFC vs AlldEds and plotted the % least squares fit for the graph on the graph. Found the distance % from the surface to the point along with its correlation % coefficient (cc) and plotted the cc for that graph % 4-25-04 % Plots circles over data points and the radius of the circle % corresponds to the amount of energy at that point. % 4-29-04 % Fitting a curve to filament data using position vector %E(ri)/kT = lambda*sum((ri+1-2ri+ri-1)/delta(s)^2)^2 *delta(s) + % sum((ri-rmi)^2/sigma^2 %where ri=xii+yij %Take derivative and set to zero and find the closest fit clear % firstbatch % secondbatch % smBead2445900 % lgBead2516174 % tiff2425801 % matt2445901 % tiff2445901 % colin2486116 colin2435842 %determine number of filaments in data set mk=1; for i=1:length(data)-1 if data(i+1,2)==data(i,2) & data(i+1,3)==data(i,3) filmark(mk,1)=i; mk=mk+1; end end filmark=[-1;filmark]; %for the main for loop length %end of determine number of filaments in data set % find any points that have been triple clicked or there is only one point % for the filament deq=0; for i=1:length(filmark)-1 triple=filmark(i+1)-filmark(i); 107

PAGE 108

if triple==4 | triple==3 | triple==2 | triple==1, deq=filmark(i);, disp([triple deq]),end end if deq>0, return, end % go look for the numbers listed and check for problems such as triples % end of finding points in triplet clc disp(' ') disp(' ') disp('There are') disp(mk-1) disp('filaments loaded') filstart = input ('Start at which filament: '); filend = input ('End at which filament: '); if filstart~=filend ex = input('Exclude any filaments (enter numbers to exclude separated by spaces or a range (3:13), 0 for none) :','s'); EDisp = MENU('How should size of energy be displayed?','Circles','Contour Graph','None'); else EDisp = 3; ex='0'; end exc=0; %marker for removing filaments excl=str2num(ex); % MaxE = 1; %option set to always off (next three lines give option) % Fnum = 1; %option set to always off (next three lines give option) % Lege = 1; %option set to always off (next three lines give option) MaxE = 2; %option set to always on (next three lines give option) Fnum = 2; %option set to always on (next three lines give option) Lege = 2; %option set to always on (next three lines give option) % MaxE = MENU('Point to maximum energy value?','Yes','No'); % Fnum = MENU('Display filament numbers on graph?','Yes','No'); % Lege = MENU('Display legend with range of energy levels?','Yes','No'); B=41000; %B=(lambda)*k*T [=] pN*(nm^2) kT=4.1; %k*T [=] pN*nm % L=B/kT; %L=lambda [=] nm L=10000; L=L/2; %change for position_v32 and up, see reason in v32. nmperpix=actdiam/pixeldiam; %nm/pixel gp=1; %counter for finding min and max for graph AllXf=[]; %keep all x values AllYf=[]; %keep all y values AllF=[]; %keep all the lengths AlldEds=[]; %keep all energy values AllavgdEds=[]; %keep all averaged energy values DFC=[]; %keep all distances to center line avgDFC=[]; %keep all DFC averages DTS=[]; %keep all distances to surface avgDTS=[]; %keep all DTS averages TL=[]; %keep all filament lengths % hold; 108

PAGE 109

%used to calculate distance from this center line to each point slope = (center(1,2)-center(2,2))/(center(1,1)-center(2,1)); centerb = center(1,2)-slope*center(1,1); %end of used to calculate distance from this center line to each point for filnumber=filstart:filend; %excludes certain filaments for i=1:length(excl) if filnumber==excl(1,i), exc=1;, end end if exc==1, exc=0;, continue, end %end of excludes certain filaments clear dr2 dx2 dy2 z Xf Yf lengths tlengths g d X p pd2 dTds dTds2 dEds avgEds f filament fb dist DistFromCenter F=filmark(filnumber+1)-filmark(filnumber)-1; %length of each filament AllF=[AllF F]; %store all the lengths %seperate each filament data into it's own set mk=1; for i=filmark(filnumber)+2:filmark(filnumber+1) filament(mk,1)=data(i,2); filament(mk,2)=data(i,3); filament(mk,3)=data(i,1); mk=mk+1; end %end of seperate each filament data into it's own set %calculates the distance of each point from a designated center line for i = 1:length(filament) dtbead=0; fb(i) = filament(i,2)+1/slope*filament(i,1); %b of y=mx+b for each point dist(i,1) = (fb(i)centerb)/(slope+1/slope); %perpendicular x point on center line dist(i,2) = slope*dist(i,1)+centerb; %perpendicular y point on center line DistFromCenter(i,1) = ((filament(i,1)-dist(i,1))^2 + (filament(i,2)dist(i,2))^2)^0.5; %actual distance to center line %next indent calculates the distance from the surface of the bead to each point (assuming a perfect spherical bead) if (pixeldiam/2)^2 DistFromCenter(i,1)^2 > 0 dtbead = pixeldiam/2 ((pixeldiam/2)^2 DistFromCenter(i,1)^2)^0.5; end DistToSurface(i,1) = ((dist(i,1)-center(1,1))^2 + (dist(i,2)center(1,2))^2)^0.5 + dtbead; %end of next indent calculates the distance from the surface of the bead to each point (assuming a perfect spherical bead) end DFC = [DFC;DistFromCenter]; %saves all distances avgDFC=[avgDFC;mean(DistFromCenter(2:length(DistFromCenter)1))]; %doesn't include end points because they are set to 0 DTS = [DTS;DistToSurface]; %saves all distances avgDTS=[avgDTS;mean(DistToSurface(2:length(DistToSurface)1))]; %doesn't include end points because they are set to 0 %end of calculates the distance of each point from a designated center line 109

PAGE 110

%determine the axes of the contour graph xmaxi(gp,1) = max(filament(:,1)); ymaxi(gp,1) = max(filament(:,2)); xmini(gp,1) = min(filament(:,1)); ymini(gp,1) = min(filament(:,2)); gp=gp+1; %counter for finding min and max for graph %end of determine the axes of the contour graph % determine lengths between points for i=1:F-1; a=filament(i+1,1)-filament(i,1); %a side of triangle (x) b=filament(i+1,2)-filament(i,2); %b side of triangle (y) c=(a^2+b^2)^0.5; %hypotenuse of triangle %safety check to make sure filaments were ended properly % if c>75, disp(filmark(filnumber)+i), end if c>75, disp(filament(i,3)), end %end of safety check to make sure filaments were ended properly % use following line to find location of points in data % disp(filament(i,3)) lengths(i,1)=c; tlengths(i,1)=sum(lengths); %progressive length end TL=[TL;tlengths(length(tlengths))]; %save all lengths % find x portion of ri of X*ri(Xf Yf)=d X=[]; X(1,1)= S^-2 +L/lengths(1)^3; X(1,2)= -2*L/lengths(1)^3; X(1,3)= L/lengths(1)^3; d(1,1)=filament(1,1)/S^2; X(2,1)= -2*L/lengths(2)^3; X(2,2)=S^-2 +5*L/lengths(2)^3; X(2,3)= -4*L/lengths(2)^3; X(2,4)= L/lengths(2)^3; d(2,1)=filament(2,1)/S^2; X(F-1,F-3)= L/lengths(F-1)^3; X(F-1,F-2)= -4*L/lengths(F-1)^3; X(F-1,F-1)=S^-2 +5*L/lengths(F-1)^3; X(F-1,F) = -2*L/lengths(F-1)^3; d(F-1,1) =filament(F-1,1)/S^2; X(F,F-2)= L/lengths(F-1)^3; X(F,F-1)= -2*L/lengths(F-1)^3; X(F,F) =S^-2 +L/lengths(F-1)^3; d(F,1) =filament(F,1)/S^2; for i=3:F-2 X(i,i-2)= L/lengths(i)^3; X(i,i-1)= -4*L/lengths(i)^3; X(i,i) =S^-2 +6*L/lengths(i)^3; X(i,i+1)= -4*L/lengths(i)^3; X(i,i+2)= L/lengths(i)^3; d(i,1) =filament(i,1)/S^2; end Xf=X\d; %end of finding x portion of ri of X*ri(Xf Yf)=d % find y portion of ri of Y*ri(Xf Yf)=g % X and Y matrices are identical so the following is the different end 110

PAGE 111

% points for the y axis g(1,1) =filament(1,2)/S^2; g(2,1) =filament(2,2)/S^2; g(F-1,1)=filament(F-1,2)/S^2; g(F,1) =filament(F,2)/S^2; for i=3:F-2 g(i,1)=filament(i,2)/S^2; end Yf=X\g; %end of finding y portion of ri of Y*ri(Xf Yf)=g % (dTheta/dS)^2 = (d^2ri/dS^2)^2 = (d^2x/dS^2)^2 + (d^2y/dS^2)^2 % (d^2x/dS^2) = (x(i+1)-2xi+x(i-1))/S^2 % x=1 or x=N are set to 0 for i=2:length(Xf)-1 dx2(i,1) = (Xf(i+1)-2*Xf(i)+Xf(i-1))/lengths(i)^2; dy2(i,1) = (Yf(i+1)-2*Yf(i)+Yf(i-1))/lengths(i)^2; end dx2=[dx2;0]; dy2=[dy2;0]; dr2 = dx2.^2+dy2.^2; % end of derivative at each point along filament dTds2=dr2; %(dTheta/dS)^2 = (d^2ri/dS^2)^2 dEds=B*8/15*dTds2; %energy at each point avgdEds=mean(dEds(2:length(dEds)-1,1)); %average energy of filament without ends which are 0 if filstart==filend subplot(2,1,1) hold on %I want the filaments plotted versus nm lengths which is done in %the following for just one filament Xf=(Xf xmini)*nmperpix; Yf=(Yf ymini)*nmperpix; filament(:,1)=(filament(:,1) xmini)*nmperpix; filament(:,2)=(filament(:,2) ymini)*nmperpix; xmaxi=(xmaxi-xmini)*nmperpix; ymaxi=(ymaxi-ymini)*nmperpix; xmini=0; ymini=0; %If pixels is good enough then take out from next comment up to %this and the labels for nm plot(Xf,Yf,'b-') plot(filament(:,1),filament(:,2),'mx') xlabel('nm'); ylabel('nm'); hold off subplot(2,1,2) hold on plot(Xf,Yf,'b-') ylabel('nm'); hold off % else % plot(Xf,Yf) % plot(filament(:,1),filament(:,2)) end aw=num2str(filnumber); %distinguish filaments on graph 111

PAGE 112

if Fnum~=2 text(Xf(1),Yf(1),aw) %distinguish filaments on graph end AllXf=[AllXf;Xf]; AllYf=[AllYf;Yf]; AlldEds=[AlldEds;dEds]; AllavgdEds=[AllavgdEds;avgdEds]; %finds maximum dEds of all filaments if max(dEds)>=max(AlldEds) bigdEds=filnumber; [M,loc]=max(dEds); Xloc=Xf(loc); Yloc=Yf(loc); maxpt=[filament(loc,1) filament(loc,2)]; end %end of finds maximum dEds of all filaments end % v is the number of contour lines, z is the energy at each point % could just set v=20 and it would be the same thing as line below (as long % as the number of contour lines is 20 in line below) % vd=20; % v=(min(AlldEds):(max(AlldEds)-min(AlldEds))/vd:max(AlldEds)); v=20; z=diag(AlldEds); if filstart==filend hold on subplot(2,1,2) contour(AllXf,AllYf,z,v); hold off % else % contour(AllXf,AllYf,z,v); end % sets limits of the viewing area for the graph if max(Xf)>max(xmaxi), xmaxi=max(Xf);, end if min(Xf)max(ymaxi), ymaxi=max(Yf);, end if min(Yf) abs(ymax-ymin) %makes viewing window a square % ymax=ymin+.5*abs(ymax-ymin)+.5*abs(xmax-xmin); % ymin=ymin+.5*abs(ymax-ymin)-.5*abs(xmax-xmin); % else % xmax=xmin+.5*abs(ymax-ymin)+.5*abs(xmax-xmin); % xmin=xmin-.5*abs(ymax-ymin)+.5*abs(xmax-xmin); % end % end of sets limits of the viewing area for the graph x1=[xmin Xloc]; %for plot of line to maximum dEds y1=[Yloc Yloc]; x2=[Xloc Xloc]; y2=[ymin Yloc]; %for plot of line to maximum dEds if filstart==filend %if just plotting one filament subplot(2,1,1) 112

PAGE 113

hold on if MaxE~=2 plot(x1, y1,'k--') %draw a line from y axes to maximum dEds plot(x2, y2,'k--') %draw a line from x axes to maximum dEds end axis([xmin xmax ymin ymax]) %sets axis on graph subplot(2,1,2) hold on if MaxE~=2 plot(x1, y1,'k--') %draw a line from y axes to maximum dEds plot(x2, y2,'k--') %draw a line from x axes to maximum dEds end axis([xmin xmax ymin ymax]) %sets axis on graph maxdEdsString = num2str(round(max(dEds)*100)*.01); xlabel(['Maximum dE/dS found = ',maxdEdsString,' pN']) hold off % else %plotting more than one filament % hold on % if MaxE~=2 % plot(x1, y1,'k--') %draw a line from y axes to maximum dEds % plot(x2, y2,'k--') %draw a line from x axes to maximum dEds % end % axis([xmin xmax ymin ymax]) %sets axis on graph % mg=num2str(v'); %for legend % if Lege~=2 % legend(mg,-1); %legend displays the range of contour lines % end % hold off end disp(' ') disp(' ') disp(['Filament with the largest dEds is filament number num2str(bigdEds)]) disp('at positon') disp([Xloc Yloc]) disp(' ') disp(['Average filament length = num2str(round(mean(TL)*nmperpix)) nm']) % pause %everthing after this pause was added to analyze the filament data, the %plot is now versus nm with nm length filaments %used for plotting (gives number of data points used for each line) FL=[1 AllF(1)]; for i = 2:length(AllF) FL(i,:) = [FL(i-1,2)+1 FL(i-1,2)+AllF(i)]; end %end of used for plotting (gives number of data points used for each line) xmini=min(AllXf); ymini=min(AllYf); AllXf=(AllXf-xmini)*nmperpix; AllYf=(AllYf-ymini)*nmperpix; xmin=0; xmax=max(AllXf); 113

PAGE 114

ymin=0; ymax=max(AllYf); % plots filaments for i = 1:length(AllF) hold on plot(AllXf(FL(i,1):FL(i,2)), AllYf(FL(i,1):FL(i,2))) end axis([xmin xmax ymin ymax]) %sets axis on graph % end of plots filaments if EDisp == 1 %plot circles on data points corresponding to energy size angle=0:0.01:2*pi; for i = 1:length(AlldEds) hold on x1=AlldEds(i)*cos(angle)+AllXf(i); y1=AlldEds(i)*sin(angle)+AllYf(i); % plot(x1,y1,'b-') if AlldEds(i)<=14.999, plot(x1,y1,'k-'), end if AlldEds(i)>=15 & AlldEds(i)<=24.999, plot(x1,y1,'b-'), end if AlldEds(i)>=25 & AlldEds(i)<=34.999, plot(x1,y1,'g-'), end if AlldEds(i)>=35 & AlldEds(i)<=44.999, plot(x1,y1,'c-'), end if AlldEds(i)>=45 & AlldEds(i)<=54.999, plot(x1,y1,'r-'), end if AlldEds(i)>=55, plot(x1,y1,'m-'), end end %end of plot circles on data points corresponding to energy size end plot([100 200],[100 100],'k-') text(100,125,'100nm') %comments below skews contour only along x direction % for i=1:length(z) % if i+1<=length(z), z(i,i+1)=mean([z(i,i) z(i,i+1)]);, end % if i+2<=length(z), z(i,i+2)=mean([z(i,i+1) z(i,i+2)]);, end % if i+3<=length(z), z(i,i+3)=mean([z(i,i+2) z(i,i+3)]);, end % if i-1>=1, z(i,i-1)=mean([z(i,i) z(i,i-1)]);, end % if i-2>=2, z(i,i-2)=mean([z(i,i-1) z(i,i-2)]);, end % if i-3>=3, z(i,i-3)=mean([z(i,i-2) z(i,i-3)]);, end % end %end of comments below skews contour only along x direction % skews contour in x and y direction % for i=1:length(AlldEds)-1 % oneZ(i)=.333*AlldEds(i)+.333*AlldEds(i+1); % if i+2<=length(AlldEds), twoZ(i)=.25*AlldEds(i)+.25*AlldEds(i+2);, end % end % z=diag(AlldEds) + diag(oneZ,1) + diag(twoZ,2) + diag(oneZ,-1) + diag(twoZ,2); % end of skews contour in x and y direction if EDisp == 2 v=15; %number of contour limes contour(AllXf,AllYf,z,v); end % %plot a circle to represent the bead on the filament graph (next indent section) % % CenterBeadX = (2*((center(1,2)centerb)*slope+center(1,1))+(4*((center(1,2)-centerb)*slope+center(1,1))^24*(slope^2+1)*(center(1,1)^2+center(1,2)^2-2*center(1,2)*centerb+centerb^2pixeldiam^2/4))^0.5)/2/(slope^2+1); % % CenterBeadY = slope*CenterBeadX+centerb; 114

PAGE 115

% circlex = [beadcenter(1,1)-pixeldiam/2:1:beadcenter(1,1)+pixeldiam/2]; % circley = ((pixeldiam^2/4-(circlexbeadcenter(1,1)).^2).^.5+beadcenter(1,2)-ymini).*nmperpix; % negcircley = (-(pixeldiam^2/4-(circlexbeadcenter(1,1)).^2).^.5+beadcenter(1,2)-ymini).*nmperpix; % circlex = (circlex-xmini).*nmperpix; % plot(circlex,circley,'r'),plot(circlex,negcircley,'r') % %end of plot a circle to represent the bead on the filament graph (next indent section) %better plot of the bead circle angle=0:0.01:2*pi; x1=((pixeldiam/2*cos(angle)+beadcenter(1,1))-xmini).*nmperpix; y1=((pixeldiam/2*sin(angle)+beadcenter(1,2))-ymini).*nmperpix; plot(x1,y1,'r') text((beadcenter(1,1)-xmini)*nmperpix,(beadcenter(1,2)ymini)*nmperpix,'Bead') %end of better plot of the bead circle hold off pause %plot of dEds vs DistFromCenter % remove 0 value dEds and corresponding distances non0dEds=find(AlldEds>0); for i=1:length(non0dEds) non0AlldEds(i,1)=AlldEds(non0dEds(i)); non0DFC(i,1)=DFC(non0dEds(i))*nmperpix; non0DTS(i,1)=DTS(non0dEds(i))*nmperpix; end % end of remove 0 value dEds and corresponding distances subplot(2,2,1) plot(non0DFC,non0AlldEds,'.g') ylabel('dE/ds [pN]') xlabel('nm from center line') [r2DFC,linecc]=corcoeff(non0DFC,non0AlldEds); hold on plot(linecc(:,1),linecc(:,2)) title(['dEds vs Distance to Center of Actin Tail: R^2=' num2str(round(r2DFC*10000)*.0001)]) hold off subplot(2,2,2) avgDFC=avgDFC.*nmperpix; %plot of Average dEds vs Average DFC plot(avgDFC,AllavgdEds,'x') ylabel('dE/ds [pN]') xlabel('nm from center line') [r2DFC,linecc]=corcoeff(avgDFC,AllavgdEds); hold on plot(linecc(:,1),linecc(:,2)) title(['Avg dEds vs Avg Distance to Center of Actin Tail: R^2=' num2str(round(r2DFC*10000)*.0001)]) hold off end of plot of Average dEds vs Average DFC subplot(2,2,3) plot(non0DTS,non0AlldEds,'.g') ylabel('dE/ds [pN]') xlabel('nm from surface of bead') [r2DTS,linecc]=corcoeff(non0DTS,non0AlldEds); hold on 115

PAGE 116

plot(linecc(:,1),linecc(:,2)) title(['dEds vs Distance to Bead Surface: R^2=' num2str(round(r2DTS*10000)*.0001)]) hold off subplot(2,2,4) %plot of Average dEds vs Average DTS avgDTS=avgDTS.*nmperpix; plot(avgDTS,AllavgdEds,'x') ylabel('dE/ds [pN]') xlabel('nm from surface of bead') [r2DTS,linecc]=corcoeff(avgDTS,AllavgdEds); hold on plot(linecc(:,1),linecc(:,2)) title(['Avg dEds vs Avg Distance to Bead Surface: R^2=' num2str(round(r2DTS*10000)*.0001)]) hold off %end of plot of Average dEds vs Average DTS pause subplot(1,1,1) hist(non0AlldEds,100) xlabel('dE/ds [pN]') ylabel('Count') a = used to calculate length between points actdiam = actual diameter in nm of bead AllavgdEds = keeps all the averaged change in Energy per length AlldEds = keeps all change in Energy per length AllF = keep all the number of points used for each filament AllXf = keeps all the determined X positions from the equation AllYf = keeps all the determined Y positions from the equation avgdEds = average energy of each filament avgDFC = average distance from center, keeps average distance from the center of the tail for each filament avgDTS = average distance to surface, keeps average distance from the surface of the bead for each filament aw = used to label the filaments on the graph b = used to calculate length between points B = bending modulous (lamda *kT) bigdEds = stores filament with the largest energy value c = calculated length between points (pixel length) centerb = b value of y=mx+b of line running through center of actin tail d = x positions from the images data = matrix that holds all the data points from plotting the filaments and is locating in whichever m file is used dEds = change in energy along length DFC = distance from center, keeps distance from the center of the tail dist = x and y perpendicular point from individual points on center line DistFromCenter = actual distance from center line for each point DistToSurface = distance to surface of the bead for each point dr2 = combination of dx2 and dy2 to yield second derivative of position vector dtbead = calculates distance to bead from each individual point dTds2 = same as dr2 DTS = distance to surface, keeps distance from the surface of the bead dx2 = second derivative of found x values dy2 = second derivative of found y values ex = separate filaments to exclude in the array of filaments selected exc = marker (0 or 1) to know if current filament is excluded 116

PAGE 117

117 excl = lists excluded filaments in number format F = stores number of points used for individual filaments fb = b of y=mx+b for each point of individual filaments filament = data for each individual filament in pixel length filend = which filament to stop analyzing filmark = gives the location of the last point for each filament Fnum = option to list numbers next to the filaments (1=off, 2=on) filnumber = for loop counter to know which filament to work with filstart = which filament in the group to start analyzing g = y positions from the images gp = counter for min and max values of filament location (counts from 1 to number of filaments to analyze) kT = boltzmanns temperature L = lamda, persistence length Lege = option to display a legend listing the levels of energy on the contour map (1=off, 2=on) lengths = lengths between each point loc = index point of dEds with the highest energy M = value of the largest energy maxdEdsString = to display a rounded maximum energy value on graph MaxE = option to have dashed lines point out highest energy point (1=off, 2=on) maxpt = actual position on image of largest energy level mk = counter for number of individual filaments nmperpix = convert pixels into nm pixeldiam = diameter of bead in pixels from image slope = slope of line running through center of actin tail TL = keeps all filament total lengths in pixel length tlengths = progessive length of filament v = number of contour lines X = x portion of X*Xf=d, X are the amounts multiplied by the found x positions x1 = plot horizontal line from y axes to highest energy x2 = plot vertical line from x axes to highest energy Xf = found x positions for each filament Xloc = found x position of the largest energy value xmax = x position with largest energy value xmaxi = stores largest x value of all filaments xmin = x position with smallest energy value xmini = stores smallest x value of all filaments Y = y portion of Y*Yf=g, Y are the amounts multiplied by the found y positions y1 = plot horizontal line from y axes to highest energy y2 = plot vertical line from y axes to highest energy Yf = found y positions for each filament Yloc = found y position of the largest energy value ymax = y position with largest energy value ymaxi = stores largest y value of all filaments ymin = y position with smallest energy value ymini = stores smallest y value of all filaments z = matrix with the diagonal being all the energy levels for each point

PAGE 118

APPENDIX B MYOSIN SUBFRAGMENT-1 Introduction Myosin subfragment-1 (S1) binds to actin fila ments in a specific orientation showing the polarity of the filament. When bound to actin filaments, S1 appears on the filament as an arrowhead with the barbed end of the arrow poi nting toward the (+)-end and the pointed end of the arrow pointing toward the (-)-end. To determine filament polarity in EM experiments, S1 was added to actin filaments to determine their or ientation. Several methods were attempted to bind S1 to actin filaments including: flowing S1 across filame nts bound to a substratum, mixing S1 with F-actin and then binding to a substratum, varying concentr ations of both S1 and F-actin, trying different buffers, cleav ing S1 with papain or -chymotrypsin (Sigma-Aldrich Co.), using myosin prepared in the lab or from Cytoskel eton, Inc., using deactivated n-ethylmaleimide (NEM) myosin, and trying different EM techniques such as paraformaldehyde fixation, negative staining, or replicas. None of th ese variations were successful in labeling actin filaments with S1 so a color change assay (using TIRF) was performed to determine filament polarity. Myosin Purification The following protocol was derived from severa l sources (106-108). Fr ozen rabbit skeletal muscle (300 g) was thawed from -70C to 4 C overnight. Muscle was minced twice in a meat grinder and stirred for exactly 15 minutes with 1 L of buffer A (0.3 M KCl pH 6.5, 2 mM sodium pyrophosphate, 0.15 M potassium phosphate buffer). Four liters of water was then added and the solution was then poured through 4 layers of chees ecloth. Next, 7 L of water was added to the filtered solution causing the myosin to precipitate out. The solution incubated for 3 h. at 4C allowing myosin to precipitate. Supernatant wa s removed as best possible and the precipitate was centrifuged (15,000 g for 20 minut es at 4C). The precipitate was triturated in 300 mL 118

PAGE 119

buffer B (0.4 M KCl pH 6.7 and 0.03 M potassium phosphate buffer) and centrifuged (16,000 g for 30 minutes at 4C) discarding the precipitate. The supernatant was passed through glass wool to remove lipids and 250 mL of water was adde d to the filtered solution. The solution was allowed to incubate for 30 minutes at 4C then centrifuged (16,000 g for 30 minutes at 4C) and the precipitate was discarded. The supernatan t was diluted with 4 L of water and centrifuged (15,000 g for 20 minutes at 0C). The precipitat e was the dissolved in a minimal amount of buffer C (0.5 M KCl pH 6.8 and 5 mM potassium phosphate buffer). The absorbance was measured at 280 nm using an extinction coefficient of 0.59 mL/mgcm and myosin was stored at -70C. If myosin was to be inac tivated using NEM, an aliquot of 10 M myosin was dialyzed against an imidazole buffer (10 mM imidazole pH 7, 0.5 M KCl, and 10 mM EDTA) for 2 hours. The myosin was then incubated with 1 mM NEM for 1 hour on ice (54). Purification of S1 using Papain Myosin (30 mg) was dialyzed into sample bu ffer (0.03 M KCl pH 6.8 and 6 mM potassium phosphate buffer) overnight. The myosin was pell eted (1,000 g) and resuspend in 1 mL of papain digestion buffer (sample buffer with 20 mM cysteineHCl at pH 7). The purchased papain slurry (500 L) was equilibrated by washing with 4 mL of papain digestion buffer. The equilibrated papain was incuba ted while mixing at room temper ature for 1 hour. The myosin was then added to the prepared papain and mi xed for 15 minutes. The cleaving of myosin was stopped by adding 2 L of 1 M iodoacetic acid. Insoluble myosin and papain was pelleted at 14,000 g for 1 minute and the supernatant contai ning S1 was centrifuged (100,000 g for 1 hour). The absorbance of S1 was measured at 280 nm us ing an extinction coeffi cient of 0.8 mL/mgcm and S1 was stored at -70C. 119

PAGE 120

120 Purification of S1 using -Chymotrypsin Myosin was dialyzed overnight against 0.05 M KCl pH 7.0 (100, 109-112). The myosin was then centrifuged at 1,000 g for 2 minutes and the pellet concentr ation was diluted to 10 mg/ml in 2x chymotrypsin digestion buffer (0 .24 M NaCl pH 7, 20 mM sodium phosphate, and 4 mM EDTA). The diluted myosin wa s then homogenized in a teflon-glass dounce homogenizer on ice by hand to pr oduce an opalescent solution. -Chymotrypsin was dissolved in 1mM HCl and the absorbance measured at 280 nm using an extinction coefficient of 2.04 mL/mgcm. Myosin was digested with 0.03 mg/ml -chymotrypsin (Np-tosyl-L-lysine chloromethyl ketone (TLCK) tr eated) at 25C for 20 minutes wh ile shaking. Digesting was stopped by supplementing to 1 mM PMSF. The cleaved myosin and undigested myosin was precipitated by adding one volume of 6 mM MgCl2 and centrifuged (15,000 g for 5 minutes at room temperature). The supernatant was then mixed with 2 volumes of saturated ammonium sulfate at 25C (4.1 M or 767 g of (NH4)2SO4 per liter of water) and centrifuged (15,000 g for 5 minutes) to precipitate S1. Th e precipitate was then triturat ed in a small volume of 100 mM KCl and 20 mM Tris-HCl pH 8.0 until the solution became clear. Salt was removed from the S1 solution by passing the mixture over a Sephadex G 25 column and the absorbance measured at 280 nm using an extinction coefficient of 0.80 mL/mgcm.

PAGE 121

APPENDIX C IMAGEJ MACROS ////////////////////////////////////////////////////// //////////////// //////////////// /////////////// // Region of Interest (ROI) extractor ////////////////////////////////////////////////////// //////////////// //////////////// /////////////// macro "ROI Extractor Action Tool "{ // This section asks for the directory to manipul ate images and then checks what ROI number is next x=0; fitdirectory = getDirectory("Select a Directory" ) // prompts what directory to take images from for processing list = getFileList(fitdirectory); for (i = 0; i < list.length; i++) { for (j=1; j
PAGE 122

run("Tiff...", "save="+fitdirectory+x+"bandpa ss_fitc.tif"); // saves new bandpass image in appropriate directory smallfitc = getImageID(); smallfitc_title = getTitle(); selectImage(fit); close(); run("RGB Merge...", "red="+sma lltritc_title+" green="+smallfitc _title+" blue=*None* keep"); newImage("Close Me When Done Thres holding", "8-bit White", 16, 16, 1); dummyImage=getImageID(); run("Tile"); selectImage(smallfitc); selectImage(smalltritc); run("Brightness/Contrast..."); while (isOpen(dummyImage)) { // this pauses th e macro until the small tritc image is closed wait(50); } selectImage(smalltritc); run("Tiff...", "save="+fitdirectory+x+"bandpass _tritc.tif"); // saves new bandpass image in appropriate directory close(); selectImage(smallfitc); run("Tiff...", "save="+fitdirectory+x+"bandpa ss_fitc.tif"); // saves new bandpass image in appropriate directory close(); run("Tiff...", "save="+fitdirectory+x+"RGB.tif"); // saves new bandpass image in appropriate directory close(); } ////////////////////////////////////////////////////// //////////////// //////////////// /////////////// // End of Region of Inte rest (ROI) extractor ////////////////////////////////////////////////////// //////////////// //////////////// /////////////// ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// // Duplicate rectangular se lection onto an image // Write down x, y, width, height from original rectangular selection to ente r here in order of x y w h separated by spaces ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// macro "Reproduce Rectangle Select ion (Cut Paste) Tool "{ xywhstring=getString("X Y Widt h Height", 0) // enter x y width height value with spaces from previous rect selection xywhsplit = split(xywhstring) // spl it string entered with any delimiter xywh= newArray(4) // create new array for(i=0; i<=xywhsplit.length-1; i++){ xywh[i] = parseInt(xywhsplit[i ]); // convert string to number } makeRectangle(xywh[0], xywh[1], xywh[2], xywh[3]) //make new rectangle selection } 122

PAGE 123

////////////////////////////////////////////////////// //////////////// //////////////// //////////////// // End of Reproduce Rectangle Se lection (Cut Paste) Tool ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// //Get coordinates of a recta ngular selection or a line se lection and lists in log box //Use this in conjunction with the "Reprodu ce Rectangle Selection (Cut Paste) Tool" ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// macro "Get Selection Coord Action Tool C000D22D23D24D2aD2bD2cD32D33D34D35D36D37D38D39D3aD3bD3cD42D43D44D4aD 4bD4cD53D5bD63D6bD73D7bD83D8bD93D9bDa3DabDb2Db3Db4DbaDbbDbcDc2Dc3Dc4 Dc5Dc6Dc7Dc8Dc9DcaDcbDccDd2Dd3D d4DdaDdbDdcC000C111C222C333C444C555C666 C777C888C999CaaaCbbbCcccCdddCeeeCfffD00D01D 02D03D04D05D06D07D08D09D0aD0 bD0cD0dD0eD0fD10D11D12D 13D14D15D16D17D18D19D1aD 1bD1cD1dD1eD1fD20D21D2 5D26D27D28D29D2dD2eD2fD30D31D3dD3eD3 fD40D41D45D46D47D48D49D4dD4eD4fD5 0D51D52D54D55D56D57D58D59D5aD5cD5dD 5eD5fD60D61D62D64D65D66D67D68D69D 6aD6cD6dD6eD6fD70D71D72D74D75D76D77 D78D79D7aD7cD7dD7e D7fD80D81D82D84D 85D86D87D88D89D8aD8cD8dD8eD8fD90D9 1D92D94D95D96D97D98D99D9aD9cD9dD9e D9fDa0Da1Da2Da4Da5Da6Da7Da8Da9D aaDacDadDaeDafDb0D b1Db5Db6Db7Db8Db9Dbd DbeDbfDc0Dc1DcdDceDcfDd 0Dd1Dd5Dd6Dd7Dd8Dd9DddDde DdfDe0De1De2De3De4De5 De6De7De8De9DeaDebDecDedD eeDefDf0Df1Df2Df3Df4Df5Df6Df7Df8Df9DfaDfbDfcDfdD feDff"{ type = selectionType(); if (type==-1) print("No selection"); else { if (type==5){ getLine(x1, y1, x2, y2, lineWidth); length=sqrt((x1x2)*(x1-x2)+(y1-y2)*(y1-y2)); if (y2>y1) angle=-acos((x2-x1)/length)*180/PI; else angle=acos((x2-x1)/length)*180/PI; print("Line,x1,y1,x2,y2,length,angle,",x1,y1,x2,y2,length,angle); } if (type==0){ getBoundingRect(x, y, w, h); print("Rectangle, x, y, width, height,",x,y,w,h); }} restorePreviousTool; //setTool(0); // was used to switch to rectan gle selection tool but line above seems more appropriate } ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// // End of Get Selection Coord Tool ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// // This reproduces either a line or rectangular selection displayed on one image to all other open images 123

PAGE 124

// Given the other open images have appropr iate dimensions to reproduce selection ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// macro "Reproduce Selection Tool C000D11D12D13D14D15D16D17D 18D19D1aD1bD1cD1dD1eD21D2eD31D3eD41D44D4eD 51D54D5eD61D65D6eD71D75D7eD81D86D8eD91D96D9eDa1Da6DaeDb1Db7DbeDc1Dc7D ceDd1DdeDe1De2De3De4De5De6De7De8 De9DeaDebDecDedDeeC000C111C222C333C444C 555C666C777C888C999CaaaCbbbCcccCdddCeeeCfffD00D01D02D03D04D05D06D07D08D0 9D0aD0bD0cD0dD0eD0fD10D 1fD20D22D23D24D25D26D27D 28D29D2aD2bD2cD2dD2fD3 0D32D33D34D35D36D37D38D39D3aD3bD3cD3dD3fD40D42D43D45D46D47D48D49D4aD 4bD4cD4dD4fD50D52D53D55D56D57D58D 59D5aD5bD5cD5dD5fD60D62D63D64D66D67 D68D69D6aD6bD6cD6dD6fD70D72D73D74D 76D77D78D79D7aD7bD7cD7dD7fD80D82D8 3D84D85D87D88D89D8aD8bD8cD8dD8fD90D 92D93D94D95D97D98D99D9aD9bD9cD9dD 9fDa0Da2Da3Da4Da5Da7Da8Da9DaaDabDacDadDafDb0Db2Db3Db4Db5Db6Db8Db9DbaDb bDbcDbdDbfDc0Dc2Dc3Dc4Dc5Dc6Dc8Dc9 DcaDcbDccDcdDcfDd0Dd2Dd3Dd4Dd5Dd6Dd7 Dd8Dd9DdaDdbDdcDddDdfDe0D efDf0Df1Df2Df3Df4Df5Df6Df7Df8Df9DfaDfbDfcDfdDfeD ff"{ type = selectionType(); if (type==-1) print("No selection"); else { if (type==5){ getLine(x1, y1, x2, y2, lineWidth); print("X1-value,",x1); print("Y1-value,",y1); print("X2-value,",x2); print("Y2-value,",y2); length=sqrt((x1x2)*(x1-x2)+(y1-y2)*(y1-y2)); print("Length,",length); if (y2>y1) angle=-acos((x2-x1)/length)*180/PI; else angle=acos((x2-x1)/length)*180/PI; print("Angle,",angle); } if (type==0){ getBoundingRect(x, y, w, h); print("X-value,",x); print("Y-value,",y); print("Width-value,",w); print("Height-value,",h); }} if (nImages==0) print("No images are open"); else imagesopen = newArray(nImages); for(i=1; i<=nImages(); i++){ 124

PAGE 125

selectImage(i); if (type==5) makeLine(x1, y1, x2, y2); if (type==0) makeRectangle(x, y, w, h); }} ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// // End of Reproduce Selection Tool ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// // Implements (TRITC-FITC)/(TRITC+FITC+0.01) for // our single filament data, the 0. 01 is to prevent infinity cases // Note: image ids are given as nega tive values. To select a specific image use it's negative id. // To select the ith image that has been open use a positive value starting at 1 to number open. ////////////////////////////////////////////////////// //////////////// //////////////// /////////////// macro "Ratio Tool C000D17D27D37D47D57D67D77D 87D97Da7Db7Db9DbaDbbDbc DbdDc7Dc9DcaDcbDccDc dDd7Dd9DdaDdbDdcDddDe7De9DeaDebDecDedDf9DfaDfbDfcDfdC000D7aD89D8aD8bD9a C000C111C222C333C444C555C666C777C888C999D19D1aD1bD1cD1dD29D2aD2bD2cD2d D39D3aD3bD3cD3dD49D4aD4bD4cD4dD51D 52D53D54D55D59D5aD5bD5cD5dD61D62D6 3D64D65D71D72D73D74D75D81D82D83D 84D85D91D92D93D94D95C999CaaaCbbbCcccC dddCeeeCfffD69D6aD6bDa9DaaDabCfffD00D 01D02D03D04D05D06D07D08D09D0aD0bD0 cD0dD0eD0fD10D11D12D13D 14D15D16D18D1eD1fD20D21D 22D23D24D25D26D28D2eD 2fD30D31D32D33D34D35D36D38D3eD3fD40D41D42D43D44D45D46D48D4eD4fD50D56D 58D5eD5fD60D66D68D6cD6dD6eD6fD70D76D 78D79D7bD7cD7dD7eD7fD80D86D88D8cD 8dD8eD8fD90D96D98D99D9bD9cD9dD9eD9fDa0 Da1Da2Da3Da4Da5Da6Da8DacDadDaeDa fDb0Db1Db2Db3Db4Db5Db6Db8Db eDbfDc0Dc1Dc2Dc3Dc4Dc5Dc6Dc8DceDcfDd0Dd1Dd2 Dd3Dd4Dd5Dd6Dd8DdeDdfDe0De1De2De3De4 De5De6De8DeeDefDf0Df1Df2Df3Df4Df5Df 6Df7Df8DfeDff"{ //Dialog box to get exact file names image.tif requires("1.34m"); // make sure correct imagej version is running if (nImages==0) // returns number of images open print("No images are open"); else imagesopen = newArray(nImages); //makes array size of images open for(i=1; i<=nImages(); i++) { //for(initialize, limit, increment) selectImage(i); //selects image as they are listed, -i=actual images, i=count of images imagesopen[i-1] = getTitle(); //returns title of images, arrays start at 0 which is why i-1 } Dialog.create("Image Choice"); // create dialog box named Image choice Dialog.addChoice("TRITC", imagesopen); // drop down box of images open Dialog.addChoice("FITC", imagesopen, imagesopen[1]); // drop down box of images open, with default of second image available Dialog.show(); // show dialog box created TRITC = Dialog.getChoice(); // save s choice entered in first dialog box 125

PAGE 126

FITC = Dialog.getChoice(); // saves choice entered in second dialog box, order matters here //Adds TRITC to FITC imageCalculator("add create 32-bit", FITC, TRITC); //adds images and makes the new image 32-bit (necessary for these images) X = nImage(); // save image id of image just created selectImage(X); // make image active run("Add...", "value=0.01"); // add 0.01 to all values to prevent infinity values latter rename("Add"); // name image add //Subtracts TRITC to FITC and then invert LUT imageCalculator("subtract creat e 32-bit", TRITC, FITC); //subt racts images and makes the new image 32-bit (necessary for these images) however when subtracting the image ends up inverted selectImage(X + 1); // make image active rename("Subtract"); // name image subtract requires("1.30j"); // check appropr iate imagej version is running getLut(reds, greens, blues); // get values for every pixel for (i=0; i
PAGE 127

// both before merging into an rgb. Make su re to open fitc first to operate correctly. ////////////////////////////////////////////////////// //////////////// //////////////// /////////////// macro "Bandpass RGB Merge Action Tool C000D21D22D23D24D25D26D27D28D29D2aD2bD2cD2dD2eC000C111C222D31D32D33D3 4D35D36D37D38D39D3aD3bD3cD3dD3eC2 22C333C444C555C666D 41D42D43D44D45D46 D47D48D49D4aD4bD4cD4dD4eC666C777C888D51D52D53D54D55D56D57D58D59D5aD5b D5cD5dD5eC888C999CaaaD61D62D63D64D65D66D67D68D69D6aD6bD6cD6dD6eCaaaCbb bCcccD71D72D73D74D75D76D77D78D79D7aD 7bD7cD7dD7eCcccCdddD81D82D83D84D8 5D86D87D88D89D8aD8bD8cD8dD8eCdddD91 D92D93D94D95D96D97D98D99D9aD9bD9c D9dD9eCdddCeeeCfffDa1Da2Da3Da4Da5Da6D a7Da8Da9DaaDabDacDadDaeCfffD00D01D0 2D03D04D05D06D07D08D09D0aD0bD0cD0dD 0eD0fD10D11D12D13D14D15D16D17D18D 19D1aD1bD1cD1dD1eD1fD20D2f D30D3fD40D4fD50D5fD60D6fD70D7fD80D8fD90D9fDa0 DafDb0Db1Db2Db3Db4Db5Db6Db7Db8Db9DbaD bbDbcDbdDbeDbfDc0Dc1Dc2Dc3Dc4Dc5 Dc6Dc7Dc8Dc9DcaDcbDccDcdDceDcfDd 0Dd1Dd2Dd3Dd4Dd5D d6Dd7Dd8Dd9DdaDdbDdc DddDdeDdfDe0De1De2De3De4De5De6De7De8 De9DeaDebDecDedDeeD efDf0Df1Df2Df3Df 4Df5Df6Df7Df8Df9DfaDfbDfcDfdDfeDff"{ fitdirectory = getDirectory("Select a Directory" ) // prompts what directory to take images from for processing open(fitdirectory+"fitc.tif"); // opens the fitc image fit = getImageID(); // image ID of original fitc image // fitdirectory = getDirectory("image") // this was used before "select a directory" (3 lines up) was used open(fitdirectory+"tritc.tif"); // opens tritc image trit = getImageID(); // imag e ID of original tritc image run("Bandpass Filter...", "filte r_large=50 filter_small=3 suppress=None tolerance=5 autoscale saturate"); selectImage(fit); // selectimage for processing run("Bandpass Filter...", "filte r_large=50 filter_small=3 suppress=None tolerance=5 autoscale saturate"); run("RGB Merge...", "red =tritc.tif green=fitc.tif blue=*None* keep"); selectImage(trit); run("Tiff...", "save="+fitdirectory+"bandpass_ tritc.tif") // saves new bandpass image in appropriate directory close(); // closes bandpass image selectImage(fit); run("Tiff...", "save="+fitdirectory+"bandpass_ fitc.tif") // saves new bandpass image in appropriate directory close(); // closes bandpass image makeRectangle(906, 691, 451, 343); // makes a rectangle selection in the bottom right corner run("To Selection"); // zooms to the selection just made makeRectangle(0, 0, 0, 0); // removes selection } ////////////////////////////////////////////////////// //////////////// //////////////// /////////////// // End of Bandpass RGB Merge Tool ////////////////////////////////////////////////////// //////////////// //////////////// /////////////// 127

PAGE 128

// This macro shifts two images for overla y purpose either up/down or left/right // Then merges images into RGB. ////////////////////////////////////////////////////// //////////////// //////////////// /////////////// macro "Image Shift Action Tool C000D12D13D14D15D16D17D18D22D28D32D 38D42D48D52D55D56D57D58D59D5aD5bD 62D65D68D6bD72D75D78D7bD82D83D84D 85D86D87D88D8bD95D9bDa5DabDb5DbbDc5 Dc6Dc7Dc8Dc9DcaDcbC000C111C222C333C 444C555C666C777C888C999CaaaCbbbCcccCd ddCeeeCfffD00D01D02D03D04D05D06D07D08D09D0aD0bD0cD0dD0eD0fD10D11D19D1a D1bD1cD1dD1eD1fD20D21D23D24D25D26D27D29D2aD2bD2cD2dD2eD2fD30D31D33D34 D35D36D37D39D3aD3bD3cD3dD3eD3fD40 D41D43D44D45D46D47D49D4aD4bD4cD4dD4 eD4fD50D51D53D54D5cD5dD5eD5fD60D61 D63D64D66D67D69D6aD6cD6dD6eD6fD70D7 1D73D74D76D77D79D7aD7cD7dD7eD7fD80D81D89D8aD8cD8dD8eD8fD90D91D92D93D9 4D96D97D98D99D9aD9cD9dD9eD9fDa0Da1Da2 Da3Da4Da6Da7Da8Da9DaaDacDadDaeDaf Db0Db1Db2Db3Db4Db6Db7Db8Db9DbaDbcDbdD beDbfDc0Dc1Dc2Dc3Dc4DccDcdDceDcf Dd0Dd1Dd2Dd3Dd4Dd5Dd6Dd7Dd8Dd9DdaDdbDdcDddDdeDdfDe0De1De2De3De4De5De6 De7De8De9DeaDebDecDedDeeDef Df0Df1Df2Df3Df4Df5Df6Df7Df8Df9DfaDfbDfcDfdDfeDf f"{ //Dialog box to get exact file names image.tif requires("1.34m"); // make sure correct imagej version is running if (nImages==0) // returns number of images open print("No images are open"); else imagesopen = newArray(nImages+1); //makes array size of images open for(i=1; i<=nImages(); i++) { //for(initialize, limit, increment) selectImage(i); //selects image as they are listed, -i=actual images, i=count of images imagesopen[i-1] = getTitle(); //returns title of images, arrays start at 0 which is why i-1 } imagesopen[nImages]="*None*"; // Makes a *None* se lection so user can select this when no change is needed Dialog.create("Image Choice"); // create dialog box named Image choice Dialog.addChoice("TRITC image:",imagesopen); Dialog.addChoice("FITC image:", imagesopen, imagesopen[1]); Dialog.addChoice("Image to move down", imagesopen); // drop down box of images open Dialog.addNumber("Number of pixels to m ove down",1); // drop down box of images open Dialog.addChoice("Image to move left", imagesopen, imagesopen[nImages]); // drop down box of images open, with default of second image available Dialog.addNumber("Number of pixels to m ove left",1); // drop down box of images open Dialog.show(); // show dialog box created tritc = Dialog.getChoice(); // sa ves choice entered in first dialog box fitc = Dialog.getChoice(); // sa ves choice entered in second dialog box movedown = Dialog.getChoice(); // save s choice entered in third dialog box moveleft = Dialog.getChoice(); // saves c hoice entered in fouth dialog box, order matters here movedownnumber = Dialog.getNumber(); // saves number choice entered in first number box 128

PAGE 129

moveleftnumber = Dialog.getNumber(); //Check an image is selected for overlay if (tritc=="*None*") exit("Please select a red channel image"); if (fitc=="*None*") exit("Please select a green channel image"); // Shift image down if (movedown=="*None*") q=1; // I don't know how to exit the for loop so put this useless command here so if can do something else{ selectImage(movedown); // selects the correct image for shifting w=getWidth(); // gets image width h=getHeight()+movedownnumber; // gets image height plus amount to move run("Canvas Size...", "width="+w+" hei ght="+h+" position=Bottom-Center zero"); if (movedown!=tritc){ // this if statem ent moves the other image up so an overlay will work (images have to be same dimensions) selectImage(tritc); run("Canvas Size...", "width="+w+" hei ght="+h+" position=T op-Center zero");} else{ selectImage(fitc); run("Canvas Size...", "width="+w+" he ight="+h+" position= Top-Center zero"); } } // Shift image left (for description of what is going on below just l ook at descriptions for "movedown" directly above, ev erything is about the same) if (moveleft=="*None*") q=1; else{ selectImage(moveleft); w=getWidth()+moveleftnumber; h=getHeight(); run("Canvas Size...", "width="+w+" hei ght="+h+" position=Center-Left zero"); if (moveleft!=tritc){ selectImage(tritc); run("Canvas Size...", "width="+w+" height="+h+" position=Center-Right zero");} else{ selectImage(fitc); run("Canvas Size...", "width="+w+" hei ght="+h+" position=Ce nter-Right zero"); } } run("RGB Merge...", "red="+tritc+" green="+fitc+" blue=*None* keep"); } ////////////////////////////////////////////////////// //////////////// //////////////// /////////////// // End of Image Shift Tool ////////////////////////////////////////////////////// //////////////// //////////////// /////////////// 129

PAGE 130

//BELOW THIS LINE ARE ADDITIONAL MACROS ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// macro "Purich 40X"{ run("Set Scale...", "distance=4.317 known=1 pixel=1 unit=m global"); run("Scale Bar..."); } ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// macro "Purich 60X"{ run("Set Scale...", "distance=6.4756 known=1 pixel=1 unit=m global"); run("Scale Bar..."); } ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// macro "Purich 100X"{ run("Set Scale...", "distance=10.835 7 known=1 pixel=1 unit=m global"); run("Scale Bar..."); } ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// macro "Dickinson 100X"{ run("Set Scale...", "distance=11.076 2 known=1 pixel=1 unit=m global"); run("Scale Bar..."); } ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// macro "Dickinson 150X"{ run("Set Scale...", "distance=16.614 3 known=1 pixel=1 unit=m global"); run("Scale Bar..."); } ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// // Duplicate line selection from one image to another // Write down x, y, angle, length from or iginal line selection to enter here ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// macro "Reproduce Line Se lection" // Tool C000C111C222D0cD1bD2bD3aD49D59D68D77D87D96Da5Db5Dc4Dd3De3Df2C222C333C 444C555C666C777C888C999CaaaC bbbCcccCdddCeeeCfffD00D01 D02D03D04D05D06D07D 08D09D0aD0bD0dD0eD0fD10D11D12D13D14D 15D16D17D18D19D1aD1cD1dD1eD1fD20D 21D22D23D24D25D26D27D28D29D2aD2cD2dD 2eD2fD30D31D32D33D34D35D36D37D38 D39D3bD3cD3dD3eD3fD40D41D42D43D44D45D46D47D48D4aD4bD4cD4dD4eD4fD50D51 D52D53D54D55D56D57D58D5aD5bD5cD5dD5eD5fD60D61D62D63D64D65D66D67D69D6 aD6bD6cD6dD6eD6fD70D71D72D73D74D75D76D78D79D7aD7bD7cD7dD7eD7fD80D81D8 2D83D84D85D86D88D89D8aD8bD8cD8dD8eD8fD90D91D92D93D94D95D97D98D99D9aD 9bD9cD9dD9eD9fDa0Da1Da2Da3Da4Da6Da7 Da8Da9DaaDabDacDad DaeDafDb0Db1Db2Db 3Db4Db6Db7Db8Db9DbaDbbDbcDbdDbeDbfDc0Dc 1Dc2Dc3Dc5Dc6Dc7Dc8Dc9DcaDcbDcc DcdDceDcfDd0Dd1Dd2Dd4Dd5Dd6Dd7Dd8Dd9DdaDdbDdcDddDdeDdfDe0De1De2De4De5 De6De7De8De9DeaDebDecDedD eeDefDf0Df1Df3Df4Df5Df6Df7Df8Df9DfaDfbDfcDfdDfeDf f"{ { x=getNumber("X value", 0) // en ter x value from line made 130

PAGE 131

y=getNumber("Y value", 0) // en ter y value from line made angle=getNumber("Angle va lue", 0) // enter angle value from line made length=getNumber("Length value", 0) // enter length value from line made x1 = x-length*cos(angle*3.1415926/ 180) // calculate start x value from angle and length y1 = y+length*sin(angle*3.1415926/180) // calculat e start y value from angle and length makeLine(x1,y1,x,y) // create new line selection } ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// // End of Reproduce Line Selection Tool ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// // Duplicate rectangular se lection onto an image // Write down x, y, width, height from orig inal rectangular selection to enter here ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// macro "Reproduce Rectangle Selection" // Tool C000C111C222D11D12D13D14D15D16D17D18D19D1aD1bD1cD1dD21D2dD31D3dD41D4 dD51D5dD61D6dD71D7dD81D8d D91D98D99D9aD9bD9dDa1Da8DabDadDb1Db8DbbDbdD c1Dc8Dc9DcaDcbDcdDd1DddDe1De2De3De4D e5De6De7De8De9DeaDebDecDedC222C333 C444C555C666C777C888C999CaaaC bbbCcccCdddCeeeCfffD00D0 1D02D03D04D05D06D07 D08D09D0aD0bD0cD0dD0eD0fD10D1eD1f D20D22D23D24D25D26D27D28D29D2aD2bD2c D2eD2fD30D32D33D34D35D36D37D38D39D3aD3bD3cD3eD3fD40D42D43D44D45D46D4 7D48D49D4aD4bD4cD4eD4fD50D52D53D54D55D56D57D58D59D5aD5bD5cD5eD5fD60D6 2D63D64D65D66D67D68D69D6aD6bD6cD6eD 6fD70D72D73D74D75D76D77D78D79D7aD 7bD7cD7eD7fD80D82D83D84D85D86D87D88D89D8aD8bD8c D8eD8fD90D92D93D94D95D 96D97D9cD9eD9fDa0Da2Da3Da4Da5Da6Da7Da9DaaDacDaeDafDb0Db2Db3Db4Db5Db6Db 7Db9DbaDbcDbeDbfDc0Dc2Dc3Dc4Dc5Dc 6Dc7DccDceDcfDd0D d2Dd3Dd4Dd5Dd6Dd7Dd8 Dd9DdaDdbDdcDdeDdfDe0DeeDefDf0Df1Df2Df3 Df4Df5Df6Df7Df8Df9DfaDfbDfcDfdDfeDf f"{ { x=getNumber("X value", 0) // enter x va lue from previous rect selection y=getNumber("Y value", 0) // enter y va lue from previous rect selection width=getNumber("Width value", 680) // enter width value from prev ious rect selection height=getNumber("Height value" 518) // enter height value from previous rect selection makeRectangle(x, y, width, height ) //make new rectangle selection } ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// // End of Reproduce Rectangle Selection Tool //////////////// //////////////////// ///////////////////////////////////////// This macro asks what image to use. You then draw a line selection (s traight, sectioned, or freehand line) across the area of interest. Once done close the blank image causing the macro to continue. The values along the selection are ex tracted and the maximum value for each color is found. Each value is divided by the maximum no rmalizing each color to it's maximum. The data is saved in three files labeled with the ROI# then either fitcdata.txt, tritcdata.txt, or linecoordsdata.txt (so the line could be reproduced la ter). A graph of the data is also displayed and all images are closed. ////////////////////////////////////////////////////////////////////////////////////////// //////////////// //////////////// // Close "Log" window so errant info rmation isn't saved with the data 131

PAGE 132

if (isOpen("Log")) { selectWindow("Log"); run("Close"); } // Ask for directory and ROI for that directory numberofROI=0; linescansdone=0; directory = getDirectory("Select a Directory"); // asks for directory // So I don't forget what di rectory I just worked on noEndSlash = substring(directory, 0,lengthOf(di rectory)-1); lastind = lastIndexOf(noEndSlash, "/"); justDirectoryNumber = substring(noEndS lash, lastind+1,lengthOf(noEndSlash)); requires("1.38m"); title1 = "Text Window"; title2 = "["+title1+"]"; if (isOpen(title1)==false) run("New... ", "name="+title2+" type =[Text File] width=15 height=30"); print(title2,"\n"+ju stDirectoryNumber); // End of: So I don't forget what directory I just worked on filedirect = directory+"/"; // saves a forward slash with directory for ease of use list = getFileList(directory); // gets list of files in directory selected for (i = 0; i < list.length; i++) { for (j=1; j
PAGE 133

allROI = Dialog.getCheckbox(); if (allROI==false) numberofROI=event; for (j=event; j<=numberofROI; j++){ open(directory+j+"RGB.tif "); // open overlay image rgbid = getImageID(); // save image ID // Zooms into image so easier to trace filament makeRectangle(0, 0, 15, 15); run("To Selection"); // zooms to selection made (also enlarges image window) run("Out"); // backs off zoom run("Out"); run("Out"); run("Select None"); // removes selection newImage("Close Me When Done Select ing Line", "8-bit White", 16, 16, 1); // dummy image, closed when want to continue dummyImage=getImageID(); run("Tile"); // tiles images setTool(5); // selects the sectioned line tool selectImage(rgbid); // make sure rgb image is focused while (isOpen(dummyImage)) { // this pauses the macro until the small tritc image is closed wait(10); } selectImage(rgbid); // focus rgb image // Reproduces line drawn onto all open images type = selectionType(); if (type==-1) print("No selection"); else getSelectionCoordinates(x, y); // get line coordinates run("RGB Split"); // spl it image into red, green, blue close(); // close blue image makeSelection(type,x,y); // make line on green image // Getting the profile of the FITC signal // print("FITC DATA"); // t ook this out so plotting program grace doesn't have a problem profile = getProfile(); // gets data of each value along selection profileMax = 0; // initializes variable normalizedprofile = newArray(profile .length); // initializes array for (i=0; iprofileMax) profileMax = profile[i]; // finds max } 133

PAGE 134

for (i=0; iprofileMax) profileMax = profile[i]; } for (i=0; i
PAGE 135

print(x[i]+" "+y[i]); } selectWindow("Log"); saveAs("Text", directory+j+"linecoordsdat a.txt"); // saves data in log window selectWindow("Log"); run("Close"); saveAs("Tiff", directory+j+"imagejgraph.tif"); // save graph wait(2000); close(); // close graph } //////////////////////////////////// //////////////// //////////////// /////////////////////////////// This macro asks for a directory to work with. Fi nds all the 16-bit images of interest (the color change images) Plots the predetermined line on th e 16-bit images and gets the pixel data and plots the values saving everything. //////////////////////////////////// //////////////// //////////////// /////////////////////////////// basedir = getDirectory("Select a Directory") // prompts what direct ory to take images from for processing //setBatchMode(true); for (i = 1; i <=40; i++) { // changes directory if (i<=9) dir = basedir+"0"+i+"/"; // sets directory if (i>9) dir = basedir+i+"/"; // sets directory list = getFileList(dir); // gets list of files for working directory if (list.length==0){ // checks if no files found print("Stopped at directory "+dir); exit // stops code } for (k = 0; k < list.length; k++) { for (j=1; j
PAGE 136

saveAs("Text", dir+j+"fitcdata-16bit.txt"); // saves data in log window selectWindow("Log"); run("Close"); open(dir+j+"bandpass_tr itc.tif"); // 16-bit tritc linemaker(); // calls function to reproduce line normalizedprofile = plotter(); close(); Plot.setLineWidth(2); Plot.setColor("red"); // sets next call to plot as red Plot.add("line", normalizedprofile ); // adds red data to plot Plot.update(); // updates graph so the plot is drawn and seen on screen (needed so save will work later) selectWindow("Log"); // selects log window saveAs("Text", dir+j+"tritcdata -16bit.txt"); // saves data in log window selectWindow("Log"); run("Close"); saveAs("Tiff", dir+j+"imagejgraph-16bit.tif"); // save graph close(); // close graph } } } } //setBatchMode(false); function linemaker(){ coords=File.openAsStrin g(dir+j+"linecoordsdata.txt"); // opens saved line coordinates coordarray = split(coords); // splits values into string array xcoord = newArray(coordarray.l ength/2); // initiates array ycoord = newArray(coordarray.l ength/2); // initiates array for(i=0;iprofileMax) profileMax = profile[i]; // finds max 136

PAGE 137

if (profile[i]9) listdirect = fitdirectory+i+"/"; // sets directory list = getFileList(listdirect); // gets list of files for working directory if (list.length!=0){ // checks if no files found ROIhere=0; for (j=1; j>16)&0xff; // extract red byte (bits 23-17) green = (v>>8)&0xff; // extract green byte (bits 15-8) 137

PAGE 138

blue = v&0xff; // extract blue byte (bits 7-0) if ((red < 190) || (green < 190)) // || is or making only yellow show, && is and making bright re d and green show as well as yellow setPixel(m,n,0); // set non yellow pixels as black } } run("Tiff...", "save="+li stdirect+"RGB_bead_count.tif"); run("8-bit"); setThreshold(1, 255); run("Convert to Mask"); rename("Directory_"+i); // Since the image is thresholded to only display yellow, those spots remaining are // considered to be beads so there is no size limitation or circularity limit. run("Analyze Particles...", "s ize=4-Infinity ci rcularity=0-1.00 show=Nothing clear summarize"); close(); } } } while(nImages>0){ close(); } selectWindow("Summary"); saveAs("Text", fitdirec tory+"bead_count.txt"); //run("Text...", "save="+fitdirectory+"bead_count.txt"); // saves results run("Close"); // closes results setBatchMode(false); ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// //BELOW THIS LINE ARE MACROS SPECIFICALLY MADE FOR SHORTCUT KEYS ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// //Get coordinates of a recta ngular selection or a line se lection and lists in log box //Use this in conjunction with the "Reprodu ce Rectangle Selection (Cut Paste) Tool" ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// macro "Get Selection Coord Action [y]"{ type = selectionType(); if (type==-1) print("No selection"); else { if (type==5){ getLine(x1, y1, x2, y2, lineWidth); length=sqrt((x1x2)*(x1-x2)+(y1-y2)*(y1-y2)); if (y2>y1) angle=-acos((x2-x1)/length)*180/PI; else angle=acos((x2-x1)/length)*180/PI; print("Line,x1,y1,x2,y2,length,angle,",x1,y1,x2,y2,length,angle); 138

PAGE 139

139 } if (type==0){ getBoundingRect(x, y, w, h); print("Rectangle, x, y, width, height,",x,y,w,h); }} restorePreviousTool; //setTool(0); // was used to switch to rectan gle selection tool but line above seems more appropriate } ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// // End of Get Selection Coord Tool ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// // This macro uses RGB Merge... to me rge two images and create a RGB image // This macro assumes the first image opened (fir st image in the window list) is the red image ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// macro "Auto RGB Merge Action [q]"{ selectImage(1); tritc = getTitle(); selectImage(2); fitc = getTitle(); run("RGB Merge...", "red="+tritc+" green="+fitc+" blue=*None* keep"); } ////////////////////////////////////////////////////// //////////////// //////////////// //////////////// // End of "Auto RGB Merge Action" ////////////////////////////////////////////////////// //////////////// //////////////// ////////////////

PAGE 140

LIST OF REFERENCES 1. Ridley, A., M. Peckham, and P. Clark. 2004. Cell motility: From molecules to organisms. Wiley, Hoboken, NJ. 2. Lodish, H., A. Berk, P. Matsudaira, C. A. Ka iser, M. Krieger, M. P. Scott, L. Zipursky, and J. Darnell. 2004. Molecular cell biology. W.H. Freeman and Company, New York. 3. Graceffa, P., and R. Dominguez. 2003. Crystal structure of monome ric actin in the ATP state: Structural basis of nucleotide-dependent actin dynamics. J. Biol. Chem. 278:34172-34180. 4. Janmey, P. A., S. Hvidt, G. F. Oster, J. Lamb, T. P. Stossel, and J. H. Hartwig. 1990. Effect of ATP on actin filament stiffness. Nature 347:95-99. 5. Gittes, F., B. Mickey, J. Nettleton, and J. Howard. 1993. Flexural ri gidity of microtubules and actin filaments measured from thermal fluctuations in shape. J. Cell Biol. 120:923-934. 6. Isambert, H., P. Venier, A. C. Maggs, A. Fattoum, R. Kassab, D. Pantaloni, and M. F. Carlier. 1995. Flexibility of actin filaments derived from thermal fluctuations. Effect of bound nucleotide phalloidin and muscle regulatory proteins. J. Biol. Chem. 270:11437-11444. 7. Ott, A., M. Magnasco, A. Simon, and A. Libchaber. 1993. Measurement of the persistence length of polymerized ac tin using fluorescence microscopy. Phys. R. E. Stat. Phy. Plasmas Fluids and Rel. Inter. Topics 48:R1642-1645. 8. Alberts, B., A. Johnson, J. Lewis, M. Raff, K. Roberts, and P. Walter. 2002. Molecular biology of the cell. Garland Science, New York. 9. Bray, D. 2001. Cell movements: From molecules to motility. Garland Publishing, New York. 10. Dickinson, R. B., F. S. Southwick, and D. L. Purich. 2002. A direct-transfer polymerization model explains how the multiple profilin-binding sites in the actoclampin motor promote rapid actin-based motility. Arch. Biochem. Biophys. 406:296-301. 11. Carlier, M. F., V. Laurent, J. Santolini, R. Melki, D. Didry, G. X. Xia, Y. Hong, N. H. Chua, and D. Pantaloni. 1997. Actin depolym erizing factor (ADF/cofilin) enhances the rate of filament turnover: Implication in actin-based motility. J Cell Biol 136:1307-1322. 12. Southwick, F. S., and D. L. Purich. 1998. Listeria and Shigella actin-based motility in host cells. Trans. Am. Clin. Climatol. Assoc. 109:160-172. 140

PAGE 141

13. Marchand, J. B., P. Moreau, A. Paoletti, P. Cossart, M. F. Carlier, and D. Pantoloni. 1995. Actin-based movement of Listeria monocytogenes : Actin assembly results from the local maintenance of uncapped filament barbed ends at the bacterium surface. J. Cell Biol. 130:331-343. 14. Sanger, J. M., J. W. Sanger, and F. S. Southwick. 1992. Host cell actin assembly is necessary and likely to provide the propulsi ve force for intracellular movement of Listeria monocytogenes Infect. Immun. 60:3609-3619. 15. Smith, G. A., D. A. Portnoy, and J. A. Theriot. 1995. Asymmetric distribution of the Listeria monocytogenes ActA protein is required and sufficient to direct actin-based motility. Mol. Microbiol. 17:945-951. 16. Southwick, F. S., and D. L. Purich. 1994. Arrest of Listeria movement in host cells by a bacterial ActA analogue: Implica tions for actin-based motility. Proc. Natl. Acad. Sci. U.S.A. 91:5168-5172. 17. Southwick, F. S., and D. L. Purich. 1995. Inhibition of Listeria locomotion by mosquito oostatic factor, a natural oligoproline peptide uncoupler of profilin action. Infect. Immun. 63:182-190. 18. Brundage, R. A., G. A. Smith, C. A., J. A. Theriot, and D. A. Portnoy. 1993. Expression and phosphorylation of the Listeria monocytogenes ActA protein in mammalian cells. Proc. Natl. Acad. Sci. U.S.A. 90:11890-11894. 19. Kocks, C., R. Hellio, P. Gounon, H. Ohayon, and P. Cossart. 1993. Polarized distribution of Listeria monocytogenes surface protein ActA at the site of directional actin assembly. J. Cell Sci. 105:699-710. 20. Domann, E., J. Wehland, M. Rohde, S. Pistor, M. Hartl, W. Goebel, M. Leimeister-Wachter, M. Wuenscher, and T. Chakraborty. 1992. A novel bacterial virulence gene in Listeria monocytogenes required for host cell microfilament interaction with homology to the proline-rich region of vinculin. Embo. J. 11:1981-1990. 21. Gertler, F. B., K. Niebuhr, M. Reinhard J. Wehland, and P. Soriano. 1996. Mena, a relative of VASP and Drosophila Enabled, is implicated in the control of microfilament dynamics. Cell 87:227-239. 22. Dickinson, R. B., and D. L. Purich 2002. Clamped-filament elongation model for actin-based motors. Biophys. J. 82:605-617. 23. Mimuro, H., T. Suzuki, S. Suetsugu, H. Miki, T. Takenawa, and C. Sasakawa. 2000. Profilin is required for sustaining efficient intraand intercellular spreading of Shigella flexneri. J. Biol. Chem. 275:28893-28901. 141

PAGE 142

24. Frischknecht, F., V. Moreau, S. Rttger, S. Gonfloni, I. Reckmann, G. Superti-Furga, and M. Way. 1999. Actin-based motility of vaccini a virus mimics receptor tyrosine kinase signalling. Nature 401:926-929. 25. Chereau, D., F. Kerff, P. Graceffa, Z. Gr abarek, K. Langsetmo, and R. Dominguez. 2005. Actin-bound structures of Wiskott-Aldrich syndrome protein (W ASP)-homology domain 2 and the implications for filament assembly. Proc. Natl. Acad. Sci. U. S. A. 102:16644-16649. 26. Chereau, D., and R. Dominguez. 2006. Unde rstanding the role of the G-actin-binding domain of Ena/VASP in actin assembly. J. Struct. Biol. 27. Kelly, A. E., H. Kranitz, V. Dtsch, and R. D. Mullins. 2005. Actin binding to the central domain of WASP/Scar proteins plays a crit ical role in the activation of the Arp2/3 complex. J. Biol. Chem. 281:10589-10597. 28. Beckerle, M. C. 1998. Spa tial control of actin filame nt assembly: Lessons from Listeria Cell 95:741-748. 29. Halliburton, W. D. 1887. On muscle plasma. J. Physiol. 8:133-202. 30. Perry, S. V. 2003. When was actin first extracted from muscle? J. Mus. R. Cell Mot. :597-599. 31. Straub, F. B. 1942. Actin. Stud. Inst. Med. Chem. Univ. Szeged 2:1-15. 32. Cameron, L. A., M. J. Footer, A. van Oude naarden, and J. A. Theriot. 1999. Motility of ActA protein-coated microspheres driven by actin polymerization. Proc. Natl. Acad. Sci. U.S.A. 96:4908-4913. 33. Cameron, L. A., T. M. Svitkina, D. Vignj evic, J. A. Theriot, and G. G. Borisy. 2001. Dendritic organization of actin comet tails. Curr. Biol. 11:130-135. 34. Schwartz, I. M., M. Ehrenberg, M. Bindschad ler, and J. L. McGrath. 2004. The role of substrate curvature in ac tin-based pushing forces. Curr. Biol. 14:1094-1098. 35. McGrath, J. L., N. J. Eungdamrong, C. I. Fi sher, F. Peng, L. Mahadevan, T. J. Mitchison, and S. C. Kuo. 2003. The force-velocity re lationship for the actin -based motility of Listeria monocytogenes Curr. Biol. 13:329-332. 36. Wiesner, S., E. Helfer, D. Didry, G. Du couret, F. Lafuma, M. F. Carlier, and D. Pantaloni. 2003. A biomimetic motility assay provides insight into the mechanism of actin-based motility. J. Cell Biol. 160:387-398. 142

PAGE 143

37. Bernheim-Groswasser, A., S. Wiesner, R. M. Golsteyn, M. F. Carlier, and C. Sykes. 2002. The dynamics of actin-based mo tility depend on surface parameters. Nature 417:308-311. 38. Bernheim-Groswasser, A., J. Prost, a nd C. Sykes. 2005. Mechanism of actin-based motility: A dynamic state diagram. Biophys. J. 89:1411-1419. 39. Marcy, Y., J. Prost, M. F. Carlier, and C. Sykes. 2004. Forces generated during actin-based propulsion: A direct m easurement by micromanipulation. Proc. Natl. Acad. Sci. U.S.A. 101:5992-5997. 40. Upadhyaya, A., J. R. Chabot, A. Andreeva, A. Samadani, and A. van Oudenaarden. 2003. Probing polymerization forces by usi ng actin-propelled lipid vesicles. Proc. Natl. Acad. Sci. U.S.A. 100:4521-4526. 41. Giardini, P. A., D. A. Fletcher, and J. A. Theriot. 2003. Compression forces generated by actin comet tails on lipid vesicles. Proc. Natl. Acad. Sci. U.S.A. 100:6493-6498. 42. Boukellal, H., O. Campas, J.-F. Joanny, J. Prost, and C. Sykes. 2004. Soft Listeria : Actin-based propulsion of liquid drops. Phys. Rev. E. Stat. Nonlin. Soft Matter Phys. 69:061906. 43. Dickinson, R. B., and D. L. Purich. 2006. Diffusion rate limitations in actin-based propulsion of hard and deformable particles. Biophys. J. 91:1548-1563. 44. Parekh, S. H., O. Chaudhuri, J. A. Ther iot, and D. A. Fletcher. 2005. Loading history determines the velocity of actin-network growth. Nat. Cell Biol. 7:1119-1123. 45. Mogilner, A., and G. Oster. 1996. Cell motility driven by actin polymerization. Biophys. J. 71:3030-3045. 46. Mogilner, A., and G. Oster. 2003. For ce generation by actin polymerization II: The elastic ratchet and tethered filaments. Biophys. J. 84:1591-1605. 47. Carlsson, A. E. 2003. Growth velocities of branched actin networks. Biophys. J. 84:2907-2918. 48. Hill, T. L., and M. W. Kirschner. 1982. Subunit treadmilling of micr otubules or actin in the presence of cellular barriers: Possible conversion of chemical free energy into mechanical work. Proc. Natl. Acad. Sci. U.S.A. 79:490-494. 49. Hill, T. L. 1981. Microfilament or microtubu le assembly or disassembly against a force. Proc. Natl. Acad. Sci. U.S.A. 78:5613-5617. 50. Kuo, S. C., and J. L. McGrat h. 2000. Steps and fluctuations of Listeria monocytogenes during actin-based motility. Nature 407:1026-1029. 143

PAGE 144

51. Gerbal, F., V. Laurent, A. Ott, M. F. Carlier, P. Chaikin, and J. Prost. 2000. Measurement of the elasticity of the actin tail of Listeria monocytogenes Eur. Biophys. J. 29:134-140. 52. Pantaloni, D., R. Boujemaa, D. Didry, P. Gounon, and M. F. Carlier. 2000. The Arp2/3 complex branches filament barbed ends: F unctional antagonism with capping proteins. Nat. Cell Biol. 2:385-391. 53. Amann, K. J., and T. D. Pollard. 2001. The Arp2/3 complex nucleates actin filament branches from the sides of pre-existing filaments. Nat. Cell Biol. 3:306-310. 54. Amann, K. J., and T. D. Pollard. 2001. Di rect real-time observation of actin filament branching mediated by Arp2/3 complex using total internal reflection fluorescence microscopy. Proc. Natl. Acad. Sci. U.S.A. 98:15009-15013. 55. Ichetovkin, I., W. Grant, and J. Condeel is. 2002. Cofilin produces newly polymerized actin filaments that are preferred for dendritic nucleation by the Arp2/3 complex. Curr. Biol. 12:79-84. 56. Carlsson, A. E. 2001. Growth of branched actin networks against obstacles. Biophys. J. 81:1907-1923. 57. Mogilner, A., and G. Oster. 2003. Polymer motors: Pushing out the front and pulling up the back. Curr. Biol. 13:R721-733. 58. Dickinson, R. B., L. Caro, and D. L. Purich. 2004. Force generation by cytoskeletal filament end-tracking proteins. Biophys. J. 87:2838-2854. 59. Laine, R. O., K. L. Phaneuf, C. C. Cunni ngham, D. Kwiatkowski, T. Azuma, and F. S. Southwick. 1998. Gelsolin, a protein that caps the barbed ends and severs actin filaments, enhances the actin-based motility of Listeria monocytogenes in host cells. Infect. Immun. 66:3775-3782. 60. Bear, J. E., T. M. Svitkina, M. Krause, D. A. Schafer, J. J. Loureiro, G. A. Strasser, I. V. Maly, O. Y. Chaga, J. A. Cooper, G. G. Borisy, and F. B. Gertler. 2002. Antagonism between Ena/VASP proteins and actin fila ment capping regulates fibroblast motility. Cell 109:509-521. 61. Barzik, M., T. I. Kotova, H. N. Higgs, L. Hazelwood, D. Hanein, F. B. Gertler, and D. A. Schafer. 2005. Ena/VASP protei ns enhance actin polymeri zation in the presence of barbed end capping proteins. J. Biol. Chem. 280:28653-28662. 62. Howard, J. 2001. Mechanics of motor proteins and the cytoskeleton. Sinauer, Sunderland, MA. 63. Dickinson, R. B., and D. L. Purich. 2007. Models for actin polymerization and force generation by cytoskeletal f ilament end-tracking motors. J. Math. Biol. Under Review. 144

PAGE 145

64. Boal, D. 2002. Mechanics of the cell. Cambridge University Press, Cambridge. 65. Kas, J., H. Strey, J. X. Tang, D. Finger, R. Ezzell, E. Sackmann, and P. A. Janmey. 1996. F-actin, a model polymer for semiflexible chains in dilute, semidilute, and liquid crystalline solutions. Biophys. J. 70:609-625. 66. Landau, L. D., and E. M. Lifshitz. 1986. Theory of elasticity. Reed Educational and Professional Publishing Ltd., New York. 67. Doi, M., and S. F. Edwards. 1986. The theory of polymer dynamics. Clarendon Press; Oxford University Press, Oxford Oxfordshire; New York. 68. Noireaux, V., R. M. Golsteyn, E. Friederich, J. Prost, C. Antony, D. Louvard, and C. Sykes. 2000. Growing an actin gel on spherical surfaces. Biophys. J. 78:1643-1654. 69. Loisel, T. P., R. Boujemaa, D. Pantaloni and M. F. Carlier. 1999. Reconstitution of actin-based motility of Listeria and Shigella using pure proteins. Nature 401:613-616. 70. Lasic, D. D. 1997. Liposomes in gene delivery. CRC Press LLC, Boca Raton, FL. 71. Plastino, J., I. Lelidis, J. Prost, a nd C. Sykes. 2004. The effect of diffusion, depolymerization and nucleation promoting factors on actin gel growth. Eur. Biophys. J. 33:310-320. 72. Bradford, M. M. 1976. A rapid and sensitiv e method for the quantitation of microgram quantities of protein utilizing the principle of prot ein-dye binding. Anal. Biochem. 72:248-254. 73. Pardee, J. D., and J. A. Spudich 1982. Purification of muscle actin. Methods Enzymol. 85 Pt B:164-181. 74. Kuhn, J. R., and T. D. Pollard. 2005. R eal-time measurements of actin filament polymerization by total internal reflection fluorescence microscopy. Biophys. J. 88:1387-1402. 75. Haugland, R. P. 1995. Coupling of monoclonal antibodies with fluorophores. Methods Mol. Biol. 45:205-221. 76. Welch, M. D., J. Rosenblatt, J. Skoble, D. A. Portnoy, and T. J. Mitchison. 1998. Interaction of human Arp2/3 complex and the Listeria monocytogenes ActA protein in actin filament nucleation. Science 281:105-108. 77. Lasic, D. D. 1988. The mechanism of vesicle formation. Biochem. J. 256:1-11. 78. Hub, H. H., U. Zimmermann, and H. Ri ngsdorf. 1982. Preparation of large unilamellar vesicles. FEBS Lett. 140:254-256. 145

PAGE 146

79. Khan, L. A., M. Raj, and M. Amin. 1989. Coupling of the enzymic activities of myosin ATPase and creatine kinase and its role in muscular contraction. Indian J. Biochem. Biophys. 26:148-152. 80. Cameron, L. A., J. R. Robbins, M. J. Footer, and J. A. Theriot. 2004. Biophysical parameters influence actin-based movement trajectory, and initi ation in a cell-free system. Mol. Biol. Cell 15:2312-2323. 81. Bozzola, J. J. 1999. Electron microscopy: Pr inciples and techniques for biologists. Jones and Bartlett, Sudbury, Mass. 82. Goodhew, P. J. 2001. Electron microscopy a nd analysis. Taylor & Francis, New York. 83. Rosenblatt, J., B. Agnew, H. Abe, J. Bamburg, and T. Mitchison. 1997. Xenopus actin depolymerizing factor/cofilin (XAC) is responsible for the turnover of actin filaments in Listeria monocytogenes tails. J. Cell Biol. 136:1323. 84. Zeile, W. L., F. Zhang, R. B. Dickinson, and D. L. Purich. 2005. Listeria's right-handed helical rocket-tail trajectories: Mechanis tic implications for force generation in actin-based motility. Cell Motil. Cytoskeleton 60:121-128. 85. Cicchetti, G., P. Maurer, P. Wagener, and C. Kocks. 1999. Actin and phosphoinositide binding by the ActA protein of the bacterial pathogen Listeria monocytogenes J. Biol. Chem. 274:33616-33626. 86. Hartman, J. J., and R. D. Vale. 1999. Microtubule disassembly by ATP-dependent oligomerization of the AAA enzyme katanin. Science 286:782-785. 87. Svitkina, T. M., and G. G. Borisy. 1998. Correlative light and elec tron microscopy of the cytoskeleton of cultured cells. Methods Enzymol. 298:570-592. 88. Butler, J. E., and E. S. G. Shaqfeh. 2002. Dynamic simulations of the inhomogeneous sedimentation of rigid fibres. J. Fluid Mech. 468:205-237. 89. Harbeck, B., S. Httelmaier, K. Schluter, B. M. Jockusch, and S. Illenberger. 2000. Phosphorylation of the vasodilator-stimulat ed phosphoprotein regulates its interaction with actin. J. Biol. Chem. 275:30817-30825. 90. Brown, S. S., and J. A. Spudich. 1979. Nu cleation of polar actin filament assembly by a positively charged surface. J. Cell Biol. 80:499-504. 91. Kovar, D. R., and T. D. Pollard. 2004. Insertional assembly of actin filament barbed ends in association with formin s produces piconewton forces. Proc. Natl. Acad. Sci. U.S.A. 101:14725-14730. 146

PAGE 147

92. Romero, S., C. Le Clainche, D. Didry, C. Egile, D. Pantaloni, and M. F. Carlier. 2004. Formin is a processive motor that requires profilin to accelerate actin assembly and associated ATP hydrolysis. Cell 119:419-429. 93. Saito, K., M. Tokunaga, A. H. Iwane, and T. Yanagida. 1997. Dual-c olour microscopy of single fluorophores bound to myosin interacting with fluores cently labelled actin using anti-Stokes fluorescence. J. Microsc. 188:255-263. 94. Benesch, S., S. Polo, F. P. Lai, K. I. Ande rson, T. E. Stradal, J. Wehland, and K. Rottner. 2005. N-WASP deficiency impairs EGF inte rnalization and actin assembly at clathrin-coated pits. J. Cell Sci. 118:3103-3115. 95. Pollard, T. D., L. Blanchoin, and R. D. Mullins. 2000. Molecular mechanisms controlling actin filament dynamics in nonmuscle cells. Annu. Rev. Biophys. Biomol. Struct. 29:545-576. 96. Rossoll, W., S. Jablonka, C. Andreassi, A. K. Kroning, K. Karle, U. R. Monani, and M. Sendtner. 2003. Smn, the spinal muscular atrophy-determining gene product, modulates axon growth and localization of beta-actin mRNA in gr owth cones of motoneurons. J Cell Biol 163:801-812. 97. Gov, N. S., and A. Gopinathan. 2006. Dynamics of membranes driven by actin polymerization. Biophys. J. 90:454-469. 98. Chodniewicz, D., and D. V. Zhelev. 2003. N ovel pathways of F-actin polymerization in the human neutrophil. Blood 102:2251-2258. 99. Oude Weernink, P. A., P. Schulte, Y. G uo, J. Wetzel, M. Amano, K. Kaibuchi, S. Haverland, M. Voss, M. Schmidt, G. W. Mayr, and K. H. Jakobs. 2000. Stimulation of phosphatidylinositol-4-phosphate 5-kinase by Rho-kinase. J Biol Chem 275:10168-10174. 100. Okamoto, Y., and T. Sekine. 1985. A stream lined method of subfra gment one preparation from myosin. J. Biochem. Tokyo 98:1143-1145. 101. Chan, B. P., A. Chilkoti, W. M. Reichert, a nd G. A. Truskey. 2003. Effect of streptavidin affinity mutants on the integrin-independent adhesion of biotinylated endothelial cells. Biomaterials 24:559-570. 102. Faulstich, H., S. Zobeley, U. Bentrup, and B. M. Jockusch. 1989. Biotinylphallotoxins: Preparation and use as actin probes. J. Histochem. Cytochem. 37:1035-1045. 103. Jahn, W. 2007. The Association of Actin and Myosin in the Presence of gamma-Amido-ATP Proceeds Mainly via a Complex with Myosin in the Closed Conformation. Biochemistry 46:9654-9664. 147

PAGE 148

148 104. Sakamaki, J., H. Honda, E. Imai, K. Hatori, K. Shimada, and K. Matsuno. 2003. Enhancement of the sliding velo city of actin filaments in the presence of ATP analogue: AMP-PNP. Biophysical chemistry 105:59-66. 105. Van Dijk, J., C. Fernandez, and P. Ch aussepied. 1998. Effect of ATP analogues on the actin-myosin interface. Biochemistry 37:8385-8394. 106. Hayat, M. A. 1981. Principles and te chniques of electron microscopy: Biological applications. Edward Arnold, London. 107. Szent-Gyrgyi, A. 1951. Chemistry of muscul ar contraction. Academic Press, New York. 108. Mommaerts, W. F. 1958. Chemical investigation of muscular tissues. Methods in medical research 7:1-59. 109. Chaussepied, P., and A. A. Kasprzak. 1989. Isolation and characterization of the G-actin-myosin head complex. Nature 342:950-953. 110. Toyoshima, Y. Y., S. J. Kron, E. M. McNa lly, K. R. Niebling, C. Toyoshima, and J. A. Spudich. 1987. Myosin subfragment-1 is suffici ent to move actin filaments in vitro. Nature 328:536-539. 111. Weeds, A. G., and R. S. Taylor. 1975. Separation of subfragment-1 isoenzymes from rabbit skeletal muscle myosin. Nature 257:54-56. 112. Weeds, A. G., and B. Pope. 1977. Studies on the chymotryptic digestion of myosin. Effects of divalent cations on proteolytic susceptibility. J Mol Biol 111:129-157.

PAGE 149

BIOGRAPHICAL SKETCH Colin Dane Sturm was born in Oklahoma City, OK, on February 14, 1978 to parents George J. Sturm and Barbara J. Sturm. He gra duated from Moore High School in Moore, OK, in 1997. In 2002 he received a Bachelor of Scie nce in chemical engineering with a focus on biotechnology and a minor in chemistry from th e University of Oklahoma. He started his graduate studies at the University of Florida in chemical engineering in 2002 and joined Dr. Richard Dickinson's research group in 2003. He obtained his Doctor of Philosophy in chemical engineering in 2007. 149