Title: Cell adhesion
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00100761/00001
 Material Information
Title: Cell adhesion characterization of adhesive forces and effect of topography
Physical Description: Book
Language: English
Creator: Zhao, Lee Cheng, 1979-
Publisher: University of Florida
Place of Publication: Gainesville Fla
Gainesville, Fla
Publication Date: 2000
Copyright Date: 2000
 Subjects
Subject: Biomedical Engineering thesis, M.S   ( lcsh )
Dissertations, Academic -- Biomedical Engineering -- UF   ( lcsh )
Genre: government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )
 Notes
Summary: ABSTRACT: Two aspects of cell adhesion were studied. First, the forces of interaction in the biotin/avidin and the selectin/sialyl Lewis X (SLe superscript X) systems were measured by bonding microspheres to cantilevers of the atomic force microscope (AFM). The effect of surface topography on cell adhesion was also studied. Avidin/Biotin-- Biotin covalently linked to poly(ethylene glycol) was grafted onto silica microspheres (r = 5mm) using 3-aminopropyltriethoxysilane (APS). Avidin was grafted onto glass slides with glutaraldehyde (Glut) and APS. Slides with only Glut were used as the control. The microspheres were bonded to AFM tips, and ten measurements of the force of interaction were taken. The Derjaguin approximation between spheres and flat plates was used to normalize force values. Free biotin at 0.5, 1, and 2 wt% was added. An attractive force of 0.96 +- 0.15 nN/micrometer (mean +- SD) at a distance of 116.5 +- 26.9 nm was observed between avidin and biotin. The addition of free biotin into solution had no effect on the force. Select/SLe superscript X-- SLe superscript X covalently linked to bovine serum albumin (BSA) was grafted onto silica microspheres (r = 5mm) with Glut and APS. Microspheres with BSA grafted directly onto the surface served as the control for nonspecific interactions. Porcine vascular endothelial cells (ECs) were grown using cell culture techniques on polydimethylsiloxane (PDMS) elastomer substrates.
Summary: ABSTRACT (cont.): Ten independent force measurements were made on test samples and controls as they were brought into contact with selectin expressing ECs. There was a short-range nonspecific interaction between the microsphere and the cell surface, and a long-range specific interaction between SLe superscript X and the selectins. The short-range attractive interaction between SLe superscript X and ECs was 0.212 +- 0.040 mN/micrometer at 14.2 +- 3.1 nm, while that between BSA and ECs was 0.261 +- 0.062 mN/micrometer at 16.2 +- 2.9 nm. Between SLeX and ECs, a second interaction of 230.4 +- 40.4 pN was measured at 106.7 +- 26.0 nm. This force appeared to involve uncoiling of the selectin and rupture of the receptor-ligand bond. Topography-- ECs were grown on textured PDMS substrates with 5 micrometer deep and 5, 10, and 20 micrometer wide grooves. The ECs were investigated with optical, scanning electron, and scanning laser confocal microscopes. The ECs proliferated on PDMS substrates coated with fibronectin or treated with radiofrequency plasma, but not on plain PDMS. The ECs also deformed the PDMS substrate, and appear to bridge the grooves between ridges.
Summary: KEYWORDS: cell adhesion, selectin, sialyl Lewis X, avidin, biotin, AFM, atomic force microscope, topography
Thesis: Thesis (M.S.)--University of Florida, 2000.
Bibliography: Includes bibliographical references (p. 69-77).
System Details: System requirements: World Wide Web browser and PDF reader.
System Details: Mode of access: World Wide Web.
Statement of Responsibility: by Lee Cheng Zhao.
General Note: Title from first page of PDF file.
General Note: Document formatted into pages; contains ix, 79 p.; also contains graphics.
General Note: Vita.
 Record Information
Bibliographic ID: UF00100761
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 50744159
alephbibnum - 002640212
notis - ANA7043

Downloads

This item has the following downloads:

LCZhao_Thesis ( PDF )


Full Text











CELL ADHESION: CHARACTERIZATION OF ADHESIVE FORCES AND EFFECT
OF TOPOGRAPHY





















By

LEE CHENG ZHAO


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

UNIVERSITY OF FLORIDA


2000















ACKNOWLEDGMENTS


This thesis was created with the assistance of a great number of great people. I

will try to acknowledge them all. First, my parents, Dr. Alex Zhao and Ms. Wei Zhu

gave me love, encouragement, and criticism. They are the best parents any child could

have. My grandparents were extraordinary role models. As a child, I had to look no

further for examples of integrity, achievement, and devotion to duty than my

grandparents. I have only fond memories of my youth, my uncles, aunts, and cousins

were steady companions during my childhood in China. I look forward to seeing them

again soon.

Dr. Anthony Brennan, my advisor has been integral to this thesis, and to my

success as a person. I thank him for four years of mentoring and encouragement. Above

all, he believed in me. No goal was too high to be achievable. Without him, none of this

would have been possible. The other members of my supervisory committee have also

been of great assistance. I especially appreciate their understanding of the rushed timing

of this Master's thesis and defense. Besides helping me with my thesis, Dr. Christopher

Batich influenced my interests in polymers and biomaterials considerably through his

classes. I would like to thank Dr. Tran-Son-Tay for taking the time to review and critique

my thesis. I greatly appreciate Dr. Andrew Rapoff for agreeing to be a substitute member

of my committee at the last moment.









The students in the Brennan group, and of the Materials Science and Engineering

Department as whole have made working here enjoyable. They have also been

marvelous friends. In 1996 and 1997, Dr. Tom Miller showed amazing patience while

skillfully teaching a clumsy and naive 17-year old undergrad how to work with organic

reactions. Ms. Jennifer Russo was a wonderful collaborator and friend during my

undergraduate research in 1997 and 1998. During my internship at Exxon Chemical in

Baton Rouge, LA, Dr. Michael Zamora was a terrific mentor, who helped me gain a

sense of how engineering in industry was conducted.

Adam Feinberg, Wade Wilkerson, and Chuck Seegert helped me with the cell

adhesion portion of the thesis. Chuck created the silicon wafers and the polystyrene

templates on which silicone substrates were to be cast. Chuck also kindly allowed me

access to his comprehensive library of biomedical textbooks. Wade took the tedious job

of manufacturing the silicone substrates used in chapter 4 of this thesis. He also

performed the similarly tedious work of treating samples with radiofrequency glow

discharge plasma. Adam graciously took the scanning electron micrographs (Fig 2-4, 3-

2, 4-3, 4-5, and 4-6) shown in this thesis, and built a database of the different pictures.

Adam, Wade, and Chuck were constant sources of advice and assistance in the cell

adhesion work. I could not have accomplished this work without their help.

Many others have helped with the thesis by being great coworkers. Jeremy

Mehlem deserves commendation for his charm and his tenacity-he has been in the

Brennan group for over four years. I am eternally indebted to Jeanne Macdonald for her

help with the formatting of this thesis, and for her friendship. I can thank Jamie Rhodes

for introducing me to some of the extracurricular aspects of college. Mark Schwarz,









Jesse Arnold, John Fultz, Luxsamee Plangsangmas, Sohyun Park, Nikhil Kothurkar, Clay

Bohn, and Art Gavrin have all made my times here very memorable.

Brenden Hauser is the most reliable and dedicated undergraduate I have ever

worked with. He was of tremendous assistance in the work detailed in Chapters 2 and 3.

He helped me perform many of the grafting reactions. He also had the patience and

manual dexterity to place microspheres onto AFM tips, and to perform AFM force

measurements. I am truly indebted to him for all his help.

Members of Dr. Eugene Goldberg's group also gave me much help with my

thesis. Bob Hadba showed me many of the details regarding reacting proteins onto

different substrates. Kaustubh Rau, Laurie Jenkins, and Mike Grumski performed some

of the x-ray photoelectron spectroscopy reported in my thesis. Paul Martin and Chris

Widenhouse gave me advice on sample preparation for scanning electron microscopy.

Several employees of the Major Analytical Instrumentation Center were of great

help. Jacques Cuneo taught me how to use the atomic force microscope in 1997, and has

been a good source of advice and insight ever since. Amelia Dempere provided me with

AFM tips, and Wayne Acree was responsible for coating the SEM samples with gold and

palladium.

Much of the research would have come to naught without the dedicated efforts of

the secretaries. Jackie Hulsey, while in Materials from 1996 to 1998, and in Biomedical

Engineering until 1999, was a great help. Her bright outlook and remarkable skills made

every job easier. Shelly Burleson has been an excellent replacement for Jackie. Her

sense of humor makes the office interesting.









Shadowing Dr. Keith Ozaki helped me realize that I wanted to go into medicine

(and probably academic surgery) as a career. He has also been extremely generous with

his time in advising, and in allowing me to work in his laboratory. I could not have found

a better role model for my future career in medicine. Zaher Abouhamze and John

Rectenwald in Dr. Ozaki's research group gave me assistance countless times. I thank

them for being two of the nicest people I have ever worked with, and I only hope that I

get to work with people like them in the future. Nina Klingman was extremely

accommodating, even after my numerous errors that jeopardized the integrity of her lab.

I thank her for giving me a second (and third and fourth...) chance. Bert Herrara from

Dr. Edward Block's group provided the porcine vascular endothelial cells used in this

study. I thank him for being very accessible and friendly.

I would also like to thank my roommates (and good friends), Chris Williams and

Andy Sinclair, for putting up with my complaints, rants, and general nastiness for the past

two years. I wish that I could be as nice as they are. I also appreciate my friends Katie

Thomas, Julee Shamhart, Mandy Ewing, and Conor Hardeman for listening to me and

battling overwhelming ennui.

During my time both as an undergraduate in Materials and as a graduate student,

Martha McDonald helped me plan out my course schedule, and my life. Dr. Richard

Connell first recommended me to work for Dr. Brennan. For that, and for his advice, I

am deeply grateful. Mrs. Diana Brantley deserves thanks for putting me in touch with

Dr. Abbaschian, who helped me greatly during the writing of the thesis.















TABLE OF CONTENTS
Page


A C K N O W L E D G M E N T S ......... .. ............. ...................................................................ii

A B ST R A C T ................................v iii............................

CHAPTERS

1. IN T R O D U C T IO N ........................................................... .................. .... ................ 1

C ell A dhesion in B biology ................................................... .. ... ............. .............. 1
Quantitative M odels of Cell Adhesion .................................................. .............. 6
Techniques for Force Measurements in Polymers ...................................................... 8
R review of Polym er Interactions......................................... ............... .............. 10
Experimental Outline ........................................... .............. .. 12

2. BIOTIN-AVIDIN INTERACTION.......... ........................................... ....... ...... 13

B ack g rou n d .................................................................. 13
M materials and M methods ............ ...................................................... ........................ .. 19
R results and D discussion ......................................... .................... ..... 25
F u tu re W o rk .................................................................................................... 2 8

3. SELECTIN-SIALYL LEWIS X INTERACTIONS.............. ............ ...............29

B ack g ro u n d .................................................................................... 2 9
M materials and M methods ................................................ .. .... .. .......... .. 37
R results and D discussion .................. .......................................... .. 43
Future W ork .................................. ................... ........ .. ...................... 47

4. EFFECT OF TOPOGRAPHY ON CELL ADHESION...........................................50

B background ............. .. ...............50.........................
M materials and M methods ......... .. ................................. ...... ............ ............ 57
R results and D iscussion........ .... .... .................... ........ ..... .............. ........ .... 61
F u tu re W o rk ....................................................................................................6 6


5. CONCLUSION ............................................... ........ .. .... ........ 67










L IST O F R E FE R E N C E S ...................................... ....... ............................... ................... .. ... 69

B IO G R A PH IC A L SK E TCH ......................................................................... .............. 78















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

CELL ADHESION: CHARACTERIZATION OF ADHESIVE FORCES AND EFFECT
OF TOPOGRAPHY

By

Lee Cheng Zhao

August 2000


Chairman: Anthony Brennan
Major Department: Biomedical Engineering

Two aspects of cell adhesion were studied. First, the forces of interaction in the

biotin/avidin and the selectin/sialyl Lewis X (SLex) systems were measured by bonding

microspheres to cantilevers of the atomic force microscope (AFM). The effect of surface

topography on cell adhesion was also studied.

Avidin/Biotin

Biotin covalently linked to poly(ethylene glycol) was grafted onto silica

microspheres (r = 5mm) using 3-aminopropyltriethoxysilane (APS). Avidin was grafted

onto glass slides with glutaraldehyde (Glut) and APS. Slides with only Glut were used as

the control. The microspheres were bonded to AFM tips, and ten measurements of the

force of interaction were taken. The Derjaguin approximation between spheres and flat

plates was used to normalize force values. Free biotin at 0.5, 1, and 2 wt% was added.









An attractive force of 0.96 0.15 nN/im (mean SD) at a distance of 116.5 +

26.9 nm was observed between avidin and biotin. The addition of free biotin into

solution had no effect on the force.

Select/SLex

SLex covalently linked to bovine serum albumin (BSA) was grafted onto silica

microspheres (r = 5mm) with Glut and APS. Microspheres with BSA grafted directly

onto the surface served as the control for nonspecific interactions. Porcine vascular

endothelial cells (ECs) were grown using cell culture techniques on polydimethylsiloxane

(PDMS) elastomer substrates. Ten independent force measurements were made on test

samples and controls as they were brought into contact with selection expressing ECs.

There was a short-range nonspecific interaction between the microsphere and the

cell surface, and a long-range specific interaction between SLex and the selections. The

short-range attractive interaction between SLex and ECs was 0.212 0.040 mN/pm at

14.2 3.1 nm, while that between BSA and ECs was 0.261 0.062 mN/pm at 16.2 2.9

nm. Between SLeX and ECs, a second interaction of 230.4 40.4 pN was measured at

106.7 26.0 nm. This force appeared to involve uncoiling of the selection and rupture of

the receptor-ligand bond.

Topography

ECs were grown on textured PDMS substrates with 5 pm deep and 5, 10, and 20

pm wide grooves. The ECs were investigated with optical, scanning electron, and

scanning laser confocal microscopes.

The ECs proliferated on PDMS substrates coated with fibronectin or treated with

radiofrequency plasma, but not on plain PDMS. The ECs also deformed the PDMS

substrate, and appear to bridge the grooves between ridges.















CHAPTER 1

INTRODUCTION



Cell Adhesion in Biology

Cellular adhesion molecules are critical to the function of life. The coordination

of dynamic systems of cellular adhesion affect processes such as blood vessel formation,

immunological response, wound healing, blood clotting, embryonic development, and

tissue organization in general. Cell adhesion creates many unexpected phenomena. For

example, when embryonic cells are dissociated, the cells will reassemble into tissue

resembling the original in vitro. Altered adhesion properties are also a characteristic of

cancer cells. Thus, the applications for an improved understanding of cell adhesion are

numerous. The specificity of cell adhesion comes from combinatorial expression and

interaction among a large number of adhesion receptors. Adhesion receptors also act as

signaling molecules by connecting to the cytoskeleton. The signaling pathways allow

adhesion receptors to influence cell survival and gene expression. Most cells have

anchorage dependence, meaning that they will not proliferate or survive unless they are

adhering to a substrate. Signal transduction has typically fallen under the jurisdiction of

receptors for growth factors. Only in the past decade have cell adhesion receptors

become associated with signaling. Thus, one must consider receptors for growth factors

and adhesion receptors as parts of an integrated system.









The coming fruition of the Human Genome Project and privately funded efforts at

sequencing the human genome, such as the much publicized Celera Genomics effort

(Rockville, Md) will reveal many trends in the cell adhesion system. Recently, the

genome for the fruitfly Drosophila melanogaster. and in 1998, the genome for the

nematode Caenorhabditis elegans was sequenced.2 From the information contained in

these genomes, one can make several conclusions. First, in both D. melanogaster and C.

elegans there are a large number of extracellular matrix proteins (ECM), much more than

in the eukaryote Saccharomyces cerevisiae (yeast). ECM proteins appear crucial for

multicellular life. Integrins, the primary receptors for ECM proteins are fewer, with

seven types in the fly and three in the worm. There are at least 18 integrins in vertebrates.

Cadherins, another adhesion receptor has a similar increase in complexity. There are over

70 cadherin genes in humans, while the fly has only 3, and the worm 15. The

combinatorial action of the large number of adhesion molecules allows spatiotemporal

regulation, and thus, the specificity observed in cell adhesion.

The complexity and ubiquity of the cell adhesion system make a quantitative

study extremely important. A fundamental, theoretical understanding of cellular behavior

has not been completely formulated. Since the applications of cell adhesion are so varied,

I concentrated on one aspect, leukocyte adhesion to the blood vessel endothelium. In

particular, I focused on the forces between specific receptors and ligands. Before any

discussion of receptor-ligand interactions, a review of how cell adhesion affects the blood

vessel is required.

Cell adhesion is important for the immune response in the vasculature, and by

association, the creation of vascular grafts in biomaterials. First, a discussion of blood









vessels: the vascular wall is composed of three layers: the intima, the media, and the

adventitia.3 The intima is the innermost layer of the wall, and thus, is in direct contact

with blood. This layer is composed of a monolayer of endothelial cells. The media is

made of smooth muscle cells bounded by elastic laminae. The smooth muscle cells are

embedded in a matrix of collagen, proteoglycan, and elastin. This layer is prominent in

large arteries such as the aorta, and becomes thinner and less distinct as the vessel

diameter decreases. The outer most layer is the adventitia, which is formed of loose

connective tissue, fibroblasts, small nerve fibers, and capillaries. Obviously, cell

adhesion holds all three layers together, but how cells in the blood interact with the vessel

is the focus of my study.

In both the vasculature and most other tissue, the entities that mediate cell

adhesion fall into three classes: adhesion receptors, extracellular matrix (ECM)

molecules, and the intracellular adhesion plaque proteins. Cell adhesion receptors, such

as integrins, selections, cadherins, and members of the immunoglobulin-cell adhesion

molecule (Ig-CAMs) superfamily, determine the specificity of cell-cell or cell-ECM

interaction. They are transmembrane molecules that bind to ECM molecules or counter-

receptors on other cells. The ECM molecules, like fibronectin and collagen, are typically

fibrous proteins that create a complex network that interacts simultaneously with several

cell receptors. Intracellular adhesion plaque proteins form linkages between adhesion

receptors and the cytoskeleton. These linkages allow the cell to respond to changes in

adhesion properties. Some examples are a and 3 catenins that bind with cadherins. The

study of all of these adhesion molecules may be enhanced by quantitative measurements

of the force of interaction between them.









The entities that dictate cell adhesion in the endothelium are mainly cell adhesion

receptors. Selectins and integrins are used in combination in the inflammatory response

to deliver leukocytes to the site of injury. Upon activation by the injury, endothelial cells

express P- and E-selectins that weakly bind leukocytes in a process known as "rolling."

Then, integrins are used to strongly attach leukocytes to the blood vessel wall. Selectins

are adhesion receptors that recognize carbohydrate ligands. Three types of selections have

been identified: L, P, and E.4 L-selectins, the first to be discovered, is expressed by

leukocytes. This form of selection participates in the adherence of leukocytes to peripheral

lymph nodes, and the recruitment of neutrophils and monocytes to inflammatory sites. P-

selectins was discovered by investigators interested in platelet activation,5 and is rapidly

distributed to the surface of platelets upon activation by thrombin and other mediators.

E-selectins are found on the endothelium, and support the adhesion of eosinophils,

neutrophils, and monocytes.

There are many applications to an improved understanding of the leukocyte

adhesion process, since the adhesion molecules are involved in inflammation, wound

healing, and cancer metastasis. For instance, the integrin ligand, arginine-glycine-aspartic

acid (RGD) has become the basis of efforts to develop treatments.6 Telios

Pharmaceuticals of San Diego created a wound healing gel-bearing RGD to improve cell

adhesion. Anti-clotting drugs are another application-Centocor Inc. of Malvem, PA

markets ReoPro (abciximab), a drug that targets the glycoprotein IIb/IIIa receptor on the

surface of platelets. Another avenue of investigation is to treat inflammation by blocking

selections. Cancer cells display adhesion molecules that are much different from the

complement found on normal cells. For example, it has been shown that CD44, an









adhesion molecule found on leukocytes also occurs on the surface of pancreatic tumor

cells.7

Besides all the pharmaceutical implications, one additional clinical application for

the improved understanding of cell adhesion in the vasculature is the design of synthetic

vascular grafts. Prosthetic vascular grafts are used for the replacement of aneurysms, as a

site for dialysis, and to revascularize tissue when veins are unavailable. Such grafts,

especially smaller models (diameter < 4mm), are subject to failure from two modes:

thrombogenicity and anastomotic pseudointimal hyperplasia. In the former, an

inflammatory response to the foreign biomaterial (typically polyethylene terephthalate or

polytetrafluoroethylene) causes thrombosis and leads to occlusion of the vessel.

According to Greisler, occlusions due to thrombogenicity typically occur within six

months.8 After six months, failure of small diameter grafts is often associated with

anastomotic pseudointimal hyperplasia. Both of these failure modalities are due to a

mismatch in the chemical and mechanical properties between the graft and the natural

blood vessel. To prevent these graft failures, biomedical engineers attempted to seed

endothelial cells onto the surface of the grafts.9' 10 The goal of endothelial cell seeding is

mimicry of the vascular wall. To improve the efficacy of cell seeding, investigators have

coated the graft with pyrolytic carbon.11' 12 Other methods to avert failure include the use

of heparin13 to stop thrombosis, the binding of other antithrombic agents,14 and the

covalent linkage of peptides onto the graft surface15, 16 to reduce platelet adhesion. The

complexity of the vascular environment has hindered the development of a completely

successful replacement for the human blood vessel despite some 40 years of research. A









more complete theoretical understanding of cell adhesion phenomena would help the

search for a solution.



Quantitative Models of Cell Adhesion

Cell adhesion does not just depend on the receptors-ligand interactions. The

behavior of the cell in flow also affects the adhesion characteristics.'7' 18 Cellular biology

has provided a wealth of qualitative information regarding cell adhesion, but a rigorous

quantitative model is lacking. George Bell formulated the seminal theory19 in 1978, but

his model has not been comprehensively tested. Bell's theory may be understood by

considering an antibody-bearing cell and an antigen-bearing cell. There is specific

density of antibodies or antigens on the cell surface. If the forward and reverse rate

constants for the bond formation in solution are known, then the forces between the cells

may be calculated. According to Bell, two nonspecific electrical forces create an

equilibrium separation between cells. Generally, cells are negatively charged. Thus, there

is a repulsive electrostatic force. There is also an attractive force between cells due to van

der Waals interactions. The two forces operate at different distances, thereby creating an

energy minimum at the equilibrium separation distance. Bell's calculations show the

separation between lipid bilayers to be approximately 10 nm. This separation distance is

what allows specific bond formation to occur. Once the cells are in proximity, receptors

and ligands may diffuse to adjacent sites on the cell membrane and bind. Specific

adhesion between cells depends on the reaction rates of the exact receptor-ligand system.

Although the force between one receptor-ligand pair is weak, the collective whole of such

interactions on the cell surface creates a strong adhesive force.









Since Bell's work, many other theories about cell adhesion have been formulated.

Ward, Dembo, and Hammer studied the effects of ligand density on cell detachment.20

During cell attachment, it is known that receptors cluster onto the cell-substrate interface,

thus increasing both the ligand density and the adhesive strength. Ward et al assumed a

one-dimensional tape peeling model20 with two regions, a microscopic binding region and

a macroscopic nonbinding region. At the edge of the nonbinding region, there is a

tension Tmac applied at angle Omac with the substrate. Within the binding region, one

assumes immobile receptors. A critical tension for cell detachment Tcrit may be solved

based on several parameters.


Nlkb ( RtKeq
T,,t = In 1+
1 cos In,,, Ac




The critical tension is proportional to the ligand density N1, and the thermal

energy kbO, and exhibits a weak dependence on receptor number, Rt, and receptor-ligand

affinity Keq. The cell area is Aceli. Hammer et al also investigated the effect of flow on

cell adhesion.21 A simulation of leukocyte rolling on endothelium was performed by

idealizing the cell as a hard sphere with adhesive springs and microvilli on the surface.

Based on the flow velocity, a spectrum of adhesive states may be modeled, from transient

adhesion, such as the leukocyte rolling in this study, to irreversible attachment.

The models of cell adhesion attempt to deal with many terms-ligand density,

flow conditions, etc., I limited the scope of my work to measurements of the forces

between individual receptor-ligand bonds, essentially, the receptor-ligand affinity.

Analysis of the receptor-ligand interaction from a polymer physics perspective may be a









useful method toward understanding. The terms and methodologies from polymer

physics may shed new light on the topic typically investigated by biologists and

biochemists. Thus, a review of recent work in measuring the forces of adhesion between

polymers, particularly hydrophilic polymers, is necessary.



Techniques for Force Measurements in Polymers

Before any discussion of the forces in polymer interaction, one must first discuss

the techniques used to measure these forces, as the limitations of these methods have a

significant impact on what can be measured. The literature contains two types of studies

regarding the interaction forces between surfaces: theoretical and experimental. The

experimental studies have been performed primarily through two methods: the surface

force apparatus (SFA), and force measurements with the atomic force microscope (AFM).

Israelachvili pioneered characterization of forces using the surface force apparatus (SFA)

in 1978.22 The surface force apparatus consists of two crossed cylinders of an atomically

smooth substrate, held closely together with fluid in between. The substrate is

maneuvered with a micrometer and a piezoelectric. The separation between the

substrates is measured with a method known as multiple beam interference fringes.

White light is introduced perpendicular to the surfaces, and from the wavelengths of the

standing waves produced, the separation can be measured to within 0.2 nm. The force

between the substrates is measured using a spring. Studies with the SFA have focused on

silica, mica, and alumina, because of the limitation that the substrate must be atomically

smooth.









Atomic Force Microscopy (AFM) is a powerful tool for imaging the morphology

of materials, and for measuring the interaction forces between two surfaces. The

predecessor of the AFM, the scanning tunneling microscope was invented in 1981 by

Gerd Binnig and Heinrich Rohrer,23 the instrument was quickly adapted for use as an

atomic force microscope (AFM).24 For the invention of the STM, Binnig and Rohrer won

the Nobel prize in physics in 1986. The AFM has been used to measure the topography

of a surface,25' 26 the viscoelastic phase change, and the intersurface forces.

The method of using the AFM to measure intersurface forces requires the

researcher to physically attach a colloid onto the cantilever. The deflection of the

cantilever would be measured as the colloid is brought in contact with a flat surface. One

of the researchers who pioneered investigating forces using the atomic force microscope

was Pashley.22' 27-30 He attached particles onto AFM tips, and measured the interaction

forces between the two surfaces. Pashley has investigated the forces between silicon and

polypropylene in aqueous solution.30 In this study, glass spheres of 6 to 7 jpm diameter

were interacted with flat polypropylene surfaces in varying concentrations of NaCl in

water. Since then, this colloid probe technique has been extended to measuring the forces

31, 32 33 34
in polymers,31' titania,33 and gold.34

Recently, the AFM has been adapted for use in biology. Sagvolden, et al.

measured the adhesion forces of cervical carcinoma,35 and Holland, Siedlecki, and

Marchant mapped the topography of platelets.36 Within the past decade, AFM has been

used to measure individual ligand-receptor debonding forces for the avidin-biotin

system.37-39 The AFM has also been adapted to recognize specific antibodies, and thus,

discriminate between different molecules.40









Review of Polymer Interactions

The high selectivity of biological interactions may be attributed to the

combination of hydrogen bonding, van der Waals, steric, hydrophobic, and electrostatic

forces. In the force measurements, on must account for all these interactions. Thus, a

brief review of how previous researchers have separated these forces using using both the

SFA and the AFM is necessary.

There are several researchers investigating forces using the SFA, including

Israelachvili,22, 41-48 Klein,42, 45, 46, 49-56 Luckham,49, 51, 54 55 57 Fetters,45-47, 52, 53, 56 and

Claesson.58-62 One application of the surface force apparatus is to study adsorbed

polymers in conjunction with ellipsometry and adsorption studies. Since cell adhesion

molecules are all polymers, a polymer physics method of investigating these molecules

may improve the understanding of binding and dissociation. The following is a review of

some of the advances in measuring the interaction between polymers.

Klein and Luckham have investigated the forces acting between poly(ethylene

oxide) adsorbed onto mica surfaces in cyclopentane solution.49 Using the SFA, they

studied the interaction force while varying molecular weight, compression and

decompression intervals, and approach and separation rates. The onset of the interactions

between adsorbed polymers was measured at 3Rg, three times the radius of gyration.

Klein and Luckham have also measured the interaction forces between mica surfaces

immersed in toluene.54 In this study, PEO and PS of molecular weight from 40 Kg/mol to

3 10Kg/mol were introduced into toluene. Following adsorption of the polymers, force-

distances curves were measured. They found that in pure toluene, a good solvent for both

PEO and PS, showed a short-range van der Waals attraction. Since toluene is a good









solvent, both PEO and PS are well solvated in solution, simulating proteins at the surface

of cell membranes. In experiments involving nonspecific protein-protein binding, one

may expect van der Waals attraction.

Claesson has worked with adsorption of copolymers of PEO and

ethyl(hydroxyethyl)cellulose (EHEC). Hydrophobic surfaces (Langmuir-Blodgett films

deposited on mica) were coated with EHEC (mw 250 Kg/mol), and the force was

measured as a function of temperature in aqueous solution.61 It was discovered that the

force changed from repulsive to attractive as the temperature increased. The researchers

explained this phenomenon by dividing the interaction force into two components: the

elastic and the osmotic. The elastic interaction results from the loss of conformational

entropy and volume for the polymer chains as the two surfaces approach. This elastic

force is generally repulsive. The osmotic interactions come from the increase in polymer

concentration as the two surfaces are brought together. For the osmotic interaction, if the

Flory-Huggins interaction parameter, X, is larger than 0.5, then the force is attractive, but

if X < 0.5, then the force is repulsive. Attractive forces for X > 0.5 (when the polymer is

essentially insoluble) can also be attributed to entanglement between adsorbed chains on

the two surfaces. Cell adhesion receptors may be modeled as two polymer chains in

solution. While the value of the X parameter depends on the individual receptor, the

osmotic interactions are likely to be negative since proteins are soluble.

All of the work described in the previous paragraph used a neutral solvent. When

electrolytes are added to the experiment, another force becomes important. As the

surfaces approach, the electric double-layer of ions adsorbed onto the surfaces creates a

repulsive force. In one study, the researchers investigated the interactions between mica









surfaces in a pH 9.9 solution. The SFA showed repulsive forces between the mica when

no polymer was present. When poly(ethylene imine) was added, there was a weak

adhesive force upon separation of the surfaces. Claesson et al., attribute this attraction to

bridging (the polymer chain is adsorbed upon both surfaces). Bridging effects may also

be seen in the biological interactions investigated in this thesis.



Experimental Outline

In my thesis, I measured some of the adhesive forces between receptor-ligand

pairs using the atomic force microscope. First, I measured the forces between biotin

grafted onto silica microspheres and avidin grafted onto glass surfaces. To determine

whether free biotin had any effect on the force of adhesion, I varied the concentration of

free biotin in the measurements. Next, I measured the forces between sialyl Lewis X and

selection expressing endothelial cells. Selectins are adhesion molecules involved the

tethering and rolling of leukocytes in flowing blood. I also initiated an investigation of

how topography affects endothelial cell growth and adhesion to polydimethylsiloxane

substrates. Porcine vascular endothelial cells were obtained from Dr. Edward Block's

group in the Malcom Randall Veterans Adminstration medical center. Characterization

of the cells was performed with the assistance of Dr. C. Keith Ozaki.














CHAPTER 2

BIOTIN-AVIDIN INTERACTION


Background

Biotin, commonly known as the water-soluble Vitamin H, is a cofactor in several

enzymatic carboxylation reactions. Consequently, it is found in both tissue and blood. It

has a molar mass of 244 g/mol and has a high affinity for both the bacterial protein

streptavidin and the related egg-white glycoprotein avidin. The interactions between

biotin and avidin/streptavidin form two of the strongest protein-ligand complexes known.

The affinity of biotin for avidin is Ka = 1013-1015 /M,63 and the free energy of association

is 21 kcal/mol. Usually, antibody-antigen interaction has an affinity of 107 -1011 /M.

Though the strength of the bond is pH dependent, the bond is stable over a broad pH

range. Thus, the biotin/streptavidin complex has been used as a model to study the

molecular basis of ligand-macromolecule interactions. The high affinity of biotin for

streptavidin has also been used in many bioanalytical techniques. Since the biotin/avidin

system is common and well characterized by x-ray crystallographic and genetic

techniques, I decided to investigate this system first. AFM force measurements of

biological receptors in the inexpensive biotin/avidin system were stepping stone toward

more complex systems. Also, although the forces between biotin and avidin have been

measured previously,64' 65 there are some issues with the data such that a modification in

the sample preparation may generate new insights into how biotin binds to avidin. Before









discussion of the forces involved in binding, a discussion of the current knowledge about

the structure and binding of streptavidin, avidin, and biotin is in order.



Structural Properties of Streptavidin and Avidin

Streptavidin is a tetrameric protein isolated from Streptomyces avidinii with a

molecular weight of approximately 60 Kg/mol. Each subunit is folded into an eight-

stranded antiparallel 3 barrel.66-68 At one end of this barrel is the binding site for biotin.

Thus, there are four binding sites for biotin per streptavidin. This binding site contains

both amino acid residues from the interior of the barrel and that from a loop region

attached to an adjacent subunit. Unlike many proteins composed of different subunits

(i.e. hemoglobin), streptavidin does not exhibit cooperative binding.69 The loop region of

the adjacent streptavidin subunit encloses the binding pocket, so that once biotin is bound

to the pocket of the barrel, it is not solvent accessible (Fig 2.1). The valeryl side chain

carboxyl oxygens of biotin links to the side chain of Ser88 and the primary chain nitrogen

of Asn49 through hydrogen bonds. The ureido ring forms hydrogen bonds to the side

chain of Asp 128 and, the side chain of Ser45 by the two primary amines. The side chains

of Tyr43, Asn23, and Ser27 can also hydrogen bond to the carboxy on the ring. In

addition to hydrogen bonding, the biotin molecule is bound through hydrophobic

interactions. Four tryptophan residues, Trp79, Trp92, Trpl08, Trpl20, from an adjacent

streptavidin subunit interacts with the tetrahydrothiophene ring of the biotin.70 Due of all

the hydrogen bonds and the hydrophobic interactions, the release rate of biotin from

streptavidin is exceptionally slow.71









Biotin and avidin interact in a manner similar to biotin and streptavidin.70'72

Avidin, which is also a tetramer, is a glycoprotein and has a molar mass of approximately

67Kg/mol. The binding pocket in avidin also contains the aromatic side chain of Phe72,

which interacts hydrophobically with the valeryl side chain of the biotin. The valeryl

carboxyl group forms three hydrogen bonds in avidin, instead of two in streptavidin.

Avidin also has a longer loop that closes over the binding site, making it less accessible to

solvent. These differences most likely account for the higher affinity of avidin for biotin

compared to streptavidin and biotin.


Figure 2-1. The top figure is a schematic of the avidin-biotin binding
process, obtained from the web site of Dr. Helmut Grubmuller,
http://www.mpibpc.gwdg.de/abteilungen/071/strept.html. The figure
on the bottom left is a three-dimensional representation of avidin, and
the figure on the bottom right, is a representation of biotin.









Binding Characteristics of Streptavidin

My work uses the AFM to measure the binding forces in the avidin/biotin system.

Since the binding characteristics have been measured previously, a review of the

literature is necessary. Weber and coworkers66' 68, 73 have proposed that the exceedingly

strong binding of streptavidin to biotin (Kd = 10-15 M) is the polarizable carbonyl group of

the ureido ring, which forms three hydrogen bonds in the binding site. From molecular

dynamics and free energy calculations, Miyamoto and Kollman74 state that van der Waals

interactions are more important than electrostatic and hydrogen bonding interactions. By

modifying the tryptophan residues from the enclosing loop (Trp79 and Trpl08) through

site-directed mutagenesis researchers have confirmed that van der Waals interactions are

important. The mutated streptavidin had binding constants 106 lower than normal

streptavidin.75

The atomic force microscope has been used to make direct measurements of the

force required to rupture a single streptavidin-biotin bond.64'65 Florin, Moy, and Gaub

determined the rupture forces between biotin and avidin by physisorbing biotinylated

bovine serum albumin (BSA) to the tips of AFM cantilevers.64 Then, the AFM tip was

incubated with avidin, and force measurements were taken against biotinylated agarose

beads. Thus, for these experiments, the total measured force involves the physical

adsorption of BSA, the binding of the BSA-biotin to avidin, and another binding of

avidin to biotin on an agarose bead. The agarose bead yields when pressed by the

cantilever, increasing the total sum of interactions that occur between the tip bound avidin

and the surface bound biotin. Therefore each force curve is composed of the interactions

of many biotin-avidin pairs. Hence, the quantized force of interaction for a single pair









must be determined through an autocorrelation analysis. Using this analysis, the

calculated force required to rupture a single biotin-avidin pair is 160 20 pN. To

confirm this measurement, Florin et al measured the binding force after the addition of

free avidin to the fluid cell. The free avidin reduced the overall adhesive force, but did

not change the calculated force of a single biotin-avidin pair. The free biotin also reduced

the number of binding contacts.

Florin et al.64 also measured the force between avidin and desthiobiotin and

iminobiotin, two analogs that have lower binding constants than biotin. They calculated a

binding force of 125 20 pN with desthiobiotin, and 85 15 pN with iminobiotin. Later,

the same group65 compared the unbinding or rupture force of five different avidin-biotin

pairs with corresponding thermodynamic energies. The dissociation force required to

separate the bound compounds varied from 257 25 pN for streptavidin-iminobiotin to

85 10 pN for avidin-iminobiotin. There was no correlation between the free energy,

AG, and the dissociation force, but there is a direct proportionality between the enthalpy

and the dissociation force. The proportionality factor was christened the effective width

of the binding potential, reff = AH/Fdissociation, which was 95 10 nm. Moy, Florin, and

Gaub65 then suggested a model for ligand-receptor dissociation based on the free energy

and enthalpy of the process that explains the correlation between the enthalpy and the

dissociation force.

The structural analysis of avidin/biotin binding from X-ray crystallography gives a

general view of how biotin fits into avidin the classic lock-and-key model, but a more

rigorous kinetic (and thermodynamic) model is necessary to fully explain ligand binding.

Thus, atomic force microscope was used to measure the rupture force between biotin and









the aforementioned tryptophan mutants (at Trp79, Trpl08, and Trp120).75 The rupture

force was correlated with changes in equilibrium thermodynamic parameters (free energy,

enthalpy), and the activation thermodynamic barriers for the dissociation of the biotin-

streptavidin complex (free energy of dissociation, and enthalpy of dissocation). A

correlation existed only with the activation enthalpy of dissociation.75 To determine the

mechanism of unbinding, Grubmuller, Heymann, and Tavan used molecular mechanics

calculations.76 The computer simulation suggested a multiple-pathway rupture

mechanism, which involves five major unbinding steps, each corresponding to the break

of hydrogen bonds.

As stated earlier the binding forces between biotin and avidin have been measured

by Florin, Moy, and Gaub.64'65 There are several issues with their method of using biotin

that was physically adsorbed to the AFM tip. First, the measurement of forces by AFM

quantifies the weakest force available. Thus, if the force to rupture the bond between

avidin and biotin is stronger than the force to take the adsorbed biotin off the AFM tip,

then the force reported would be the force of adsorption. Since the avidin-biotin bond is

the strongest non-covalent bond in biology, this possibility cannot be ignored. I

circumvented this issue by covalently linking the biotin to microspheres bonded to AFM

tips. I also added a spacer molecule, poly(ethylene glycol) of MW = 3,400 Da, between

the biotin and the microsphere. This spacer allowed the biotin to be mobile, and thus

bind effectively with avidin. The calculation to convert the AFM data to a force-distance

graph is also different in my method. Florin et al.64 did not account for the radius of the

tip by normalizing the value for the force to force per area. Thus, any variation in the tip

radius would affect the result. This study uses the Derjaguin approximation between a









rigid sphere and a flat surface to normalize the force. To perform the normalization, one

must measure the radius of the tip, which was performed with scanning electron

micrographs of the glass microsphere bonded onto the AFM tip.



Materials and Methods

Previously, Florin, Moy, and Gaub attached biotin to AFM tips by non-

specifically adsorbing biotinylated BSA.64 The biotinylated BSA may desorb in solution,

which would render the force measurements useless. So in the following experiments,

the biotin is covalently bonded to a microsphere through 3-aminopropyltriethoxysilane

and polyethylene glycol (PEG) of 3.4 Kg/mol. Thus, there is a spacer, the PEG, between

the biotin and the microsphere, which allows the biotin the mobility to bind with avidin.

The microsphere is then bonded to the AFM tip using epoxy. A micromanipulator is used

to physically place the epoxy and the microsphere onto the AFM tip. A

micromanipulator is an instrument that allows minute movements in the x, y, and z axes.

Attached to one end of the micromanipulator are thin tungsten filaments. Using one

filament, the operator can deposit the epoxy glue onto the AFM tip, and then, using

another filament, pick up a microsphere and place it onto the tip.



Grafting of 3-Aminopropyltriethoxysilane on Glass

A silane coupling agent, 3-aminopropyl triethoxysilane (APS), was grafted to

glass. First, the glass was cleaned ultrasonically for five minutes in a chloroform and

methanol solution and then dried in air. A solution of dry toluene and APS were

combined in a round bottom flask and allowed to mix before addition of both the glass









microspheres and the glass slides. After the addition of the microspheres and the slides,

the solution was stirred before heating to the desired temperature. The reaction

proceeded at temperature for 12 hours after which time the wafers were removed, and

rinsed with toluene. To fully condense the APS, the glass was dried and placed in an

oven at 110 OC for 1 hour. The efficacy of this procedure was confirmed using x-ray

photoelectron spectroscopy (XPS) (Table 2.1).

Table 2.1. XPS data of plain glass and APS grafted glass. The increase in nitrogen and carbon
indicates grafting of the silane coupling agent.
Element Plain glass APS grafted glass
(Mass Conc%) (Mass Conc%)
Cls 15.3 31.1
Ols 46.0 34.1
Si2p 38.9 30.1
Nls 00.0 04.7


Limitations of the XPS only allowed characterization of glass slides, and not the

glass microspheres used in the experiment. However, given the same composition of

both slides and microspheres, and the same reaction used in both, one may state that the

successful grafting of the slides mean success for the microspheres.



Biotin Grafting to Glass

After the glass was treated with APS, I added biotin to the surface. I obtained co-

N-hydroxysuccinimidyl ester of poly(ethylene glycol)-carbonate, MW = 3400 g/mol,

from Shearwater Polymers (Huntsville, AL). Hereafter, this material will be referred to

as NHS-PEG-biotin.









Characterization of Biotin Grafted Surfaces

Avidin is the conjugate for biotin in biology. We used the natural affinity

between biotin and avidin to measure the concentration of biotin on the surface.

ExtrAvidin@-FITC conjugate, which absorbs at 280 nm and 495 nm, was purchased from

Sigma-Aldrich (St. Louis, MO). Fluorescein isothiocyanate (FITC), is a common

fluorescent dye attached to proteins that is used to identify the presence of those proteins.

Four samples were investigated. First, to detect autofluoresence, both the control, a glass

slide that had been reacted with APS only and the test sample, which was a glass slide

grafted with biotin, were photographed using a Zeiss fluorescent optical microscope at

10X magnification. All pictures were exposed for 1.5 seconds, and the relative intensities

were measured (Figure 2-2a). Then the ExtrAvidin-FITC conjugate was added to both

the test sample and the control, and then washed with deionized (DI) water(Figure 2-2b).

The test sample increased in fluorescence, thus showing that the biotin had been

successfully grafted onto the surface of glass slides.





































Figure 2-2. The picture on the left shows a plain glass surface that had been dipped in
ExtrAvidin -FITC, and then washed with PBS. The picture on the right shows a biotinylated
glass surface, dipped in ExtrAvidin -FITC, and then washed with PBS. The increase in
color indicates specific binding of the avidin to the biotin on the surface.



Attachment of Avidin to Glass Slides

To attach avidin to glass, APS and glutaraldehyde were first grafted onto the glass

to provide the correct reactive functionalities. The treatment with APS is the same as

outlined above. Then, an 8 % glutaraldehyde solution in phosphate buffered saline (PBS)

at pH of 7.4 was added to the APS grafted glass slides. The mixture was left overnight

with end-to-end mixing. The glass slides were removed from the glutaraldehyde and

washed (4X) with PBS. Some 0.05 % avidin-FITC was added to the glutaraldehyde









treated glass slides. As before, a fluorescent optical microscope was used to identify that

avidin was attached to the glass (Figure 2.3).




























Figure 2-3. The picture on the left is a glass surface reacted with glutaraldehyde, but no
ExtrAvidin -FITC. There was no autofluorescence. The picture on the right was reacted with
glutaraldehyde and ExtrAvidin -FITC, and washed 3 times in PBS. The increase in intensity
is indicative of the successful reaction of avidin onto the glass surface.



Atomic Force Microscopy

Force measurements were done on a Digital Instruments Nanoscope III atomic

force microscope. The experiment was conducted inside a fluid cell provided by the

manufacturer. The fluid cell allows the solution in the cell to be changed without

disturbing the experimental setup. AFM tips modified with biotin were brought into

contact with three different samples. First, plain glass slides without any modifications









were used. Next, glass slides modified with glutaraldehyde were used. Then, force

measurements were taken using glass slides with avidin grafted to the surface. Ten

measurements were taken for each substrate. To calculate the adhesion force and the

interaction energy, the radius of the particle was determined by optical microscopy and

scanning electron microscopy (Figure 2-4). The spring constant, 0.12 N/m for the tip was

obtained from the manufacturer (Digital Instruments, Santa Barbara, CA).


Figure 2-4. Occasionally, the microspheres are different than anticipated. Here, the
experimenter accidentally bonded a doublet to the AFM cantilever. The microsphere was
reacted with APS and PEG-biotin.









Effect of Concentration

Force measurements were also performed with three different concentration of

biotin in PBS (0.5 %, 1 %, and 2 %). Ten measurements were taken for each substrate,

with the appropriate solution of biotin in the AFM fluid cell. Free biotin may act as an

inhibitor by binding to the avidin. Thus the force of adhesion that is measured between

the biotin and avidin may decrease with the addition of free biotin.


Results and Discussion

Using the Derjaguin approximation, the forces between the biotin grafted AFM tip

and the glass surface were normalized to the radii of the microsphere. The forces

between biotin and the plain glass resulted in -0.24 0.08 nN/im. The forces between

biotin and glutaraldehyde grafted glass were significantly larger (p = 0.001). The increase

in the attractive force between biotin and glutaraldehyde may be attributed to hydrogen

bonding with the glutaraldehyde. The plain glass is a flat surface whereas the addition of

APS and glutaraldehyde creates a carpet of hydrophilic groups than can attract the biotin

molecule. The interaction between biotin and avidin was both significantly larger (p =

4.63 E -05) and occurred at a significantly greater distance (p = 6.36 E -06) than that

between biotin and glutaraldehyde. The increase in both the distance of interaction and

the force is due to the specific binding with the biotin by the avidin grafted to the surface

(Table 2-2).









Table 2-2. Force measurements between biotin grafted microspheres and three substrates
Microsphere Biotin-PEG Biotin-PEG Biotin-PEG
Substrate Glass Glutaraldehyde-glass Avidin-
gluteraldehyde-glass
Distance(nm) 40.5 14.0 46.0 14.5 116.5 26.9
Force (nN/nm) -0.24 0.08 -0.53 0.16 -0.96 0.15


The force measurement was also performed with varying concentrations of biotin

in solution (Table 2-3). I had expected the biotin to block the binding sites of the avidin,

and thus, reduce the force. There appears to be no variation in the distance or the strength

of the force with changing concentration.

Table 2-3. Force measurements with different concentrations of free biotin in solution.
Biotin
Botn 0 wt% 0.5 wt% 1.0 wt% 2.0 wt%
Concentration
Distance (nm) 116.5 26.9 83.5 24.2 84.0 23.3 91.0 22.2
Force (nN/Wm) -0.96 0.15 -0.99 0.27 -0.95 0.20 -0.94 0.15


A t-test was performed between the three samples. There was no significant

difference in distance or force between the 0.5 %, 1.0 %, and 2.0 % concentrations.

Differences from force measurements performed without free biotin in solution were also

tested statistically (Table 2-4). One explanation for this result is that the concentration of

biotin was not sufficient to cause a significant blocking of the binding sites on avidin.

Another possibility is that the biotin adsorbed onto the surface of the substrate, and thus,

did not bind to avidin. To fully explain why the addition of biotin did not affect the force

measurements, more experiments must be performed. First, higher concentrations of free

biotin should be used. Perhaps a critical concentration must be reached to block the

binding. Biotin may adsorb to the glass surface. Thus, a different, preferably









hydrophobic, surface grafted with avidin may be used to determine if biotin adsorption is

affecting the force measurements.

In summary, the force measurements conducted with biotin-PEG grafted

microspheres with the plain glass showed a difference in binding between that of the

plain glass, the glutaraldehyde modified surface, and the avidin grafted surface. The

increase in attraction between the avidin grafted surface and the biotin grafted

microsphere may be attributed to the specific receptor ligand binding. When free biotin

was added to solution, no significant effect on the interaction is found.


Table 2-4. List of p-values for t-tests. The first value in each box, PD is for comparisons of the
distances, and the second value, PF is for comparisons of the force measurements

0.5 % 1.0 % 2.0 %

0.5 % N/A PD = 0.96 PD = 0.48

PF = 0.68 PF = 0.59

1.0% PD = 0.96 N/A PD = 0.50

PF = 0.68 PF = 0.92

2.0 % PD = 0.48 PD = 0.50 N/A

PF = 0.59 PF = 0.92

0 % biotin, interaction PD = 0.0023 PD = 0.0018 PD = 3.08E-04
of biotin and avidin
PF = 2.69E-07 PF = 1.61E-08 PF = 5E-10

Biotin and PD = 3.23E-08 PD = 3.62E-04 PD = 4.24E-05
glutaraldhyde
PF = 9.88E-08 PF = 2.69E-07 PF = 9.94E-06

Biotin and glass PD = 5.27E-04 PD = 1.24E-04 PD = 9.60E-06

PF = 1.59E-04 PF = 4.5E-09 PF = 1E-10









Future Work

The technique of measuring forces between receptor and ligand using AFM may

be applied to other systems. As reported in the next chapter, this has already been

performed with the selectin-sialyl Lewis X system. Modification to the biotin-avidin

system may also be performed and investigated using this technique. For example,

iminobiotin and desthiobiotin are two biotin analogs that bind less effectively to avidin.

The forces between these two ligands and avidin may be investigated using the same

techniques presented here. The changes in the force may be correlated to changes in

binding affinity.















CHAPTER 3

SELECTIN-SIALYL LEWIS X INTERACTIONS



Background

In the previous chapter, I investigated the forces of adhesion between avidin and

biotin, a system that had been extensively investigated. Avidin-biotin binding is not

particularly relevant clinically, however. Thus, another system should be investigated. In

this chapter, I will provide some background on cell adhesion molecules, and then focus

on selection mediated leukocyte rolling. Cell adhesion molecules are important for many

processes, such as the growth of tumors, the function of the immune system, and reaction

to biomaterials such as neointimal hyperplasia. One particularly important process is the

attachment, adhesion, and diapedesis of leukocytes at sites of tissue injury. In the

inflammatory response, leukocytes identify and bind to the site of injury via the

combination of selections and integrins. Upon activation by the injury, endothelial cells

express E- and P-selectins, which are adhesion receptors that recognize carbohydrate

ligands. These selections weakly bind leukocytes in a process known as "rolling."6

Following the initial binding by the selections, integrins strongly bind leukocytes to the

blood vessel wall. This process of leukocyte rolling and adhesion is the focus of my

research. In the literature there is a dearth of work detailing direct force measurements of

selection binding.









Discussion of force measurements should be prefaced by a review of the

biological behavior of selections. In the past 20 years, three types of selections have been

identified: L-, P-, and E-.4 L-selectins, the first to be discovered, are expressed by

leukocytes. This form of selection participates in the adherence of leukocytes to peripheral

lymph nodes, and the recruitment of neutrophils and monocytes to inflammatory sites. P-

selectins were discovered by investigators interested in platelet activation,5 and is rapidly

distributed to the surface of platelets upon activation by thrombin and other mediators.

E-selectins are found on the endothelium, and support the adhesion of eosinophils,

neutrophils, and monocytes.

Since selections are important for the mediation of leukocyte rolling, they are

important in the overall immune system. For example, patients with AIDS or leukemia

often have elevated levels of serum L-selectin.77 Another disease, leukocyte adhesion

deficiency II, involves recurrent bacterial infections, and the dysfunction of neutrophil

motility due to the lack of sialyl Lewis X.78 Selectins have also been associated with

injury due to ischemia and reperfusion,79 and wound repair.80 Sialyl Lewis X is even

important for normal brain development,1 as reported by Santos-Benito et al.

A complex cascade of interactions controls the binding of leukocytes.

Chemokines82 and platelet-activating factor83 expressed on the endothelial cell surface are

recognized by leukocytes after the initial contact. In contrast to most cell adhesion

phenomenon, the recruitment of leukocytes from the flowing bloodstream is a very rapid

process, which requires a special mechanism for the establishment of cell contacts. The

selections, which are distributed exclusively in leukocytes and the vasculature, are

specialized for this purpose.









Selectin Composition

To understand why selections exhibit this behavior, a review of the literature on the

structural composition of the selections in needed. The extracellular part of all selections is

composed of three different kinds of protein domains. The amine terminus of each

selection is a 120 amino acid domain that has features similar to C-type animal lectins, so

called because the binding of carbohydrates is dependent on the presence of extracellular

calcium.84 This domain is followed by a sequence of -35-40 amino acids similar to a

repeat structure originally discovered in epidermal growth factor (EGF). The single EGF

domain is followed by several "complement binding" (CB) domains, containing -60

amino acids. Patel, Nollert, and McEver85 showed that reducing the number of CB

domains in P-selectin decreases the efficiency of rolling, which suggests that these CB

domains are involved in the extension of P-selectin a sufficient length from the plasma

membrane. Size differences between the types of selections and between animal species

involve different number of CB domains. For example in rats, the E-selectins have five

CB domains and the P-selectins have eight, while in humans, E- and P-selectins have six

and nine respectively. L-selectins in both species have only two CB domains.86 All of

the selections are anchored in the membrane by a transmembrane region followed by a

short cytoplasmic tail (Figure 3-1).









Leukocyte


P O


Lectin c

I EGF dc

* Comple


E i bindir
doma
L
Transm
*





Endothelium

Figure 3-1. Schematic of three kinds of selections in humans


lomain

,main

rmentary
ig (CB)
ins

membrane region


Since contact between most leukocytes and endothelium is initiated by selections,

their regulation is important for control of diapedesis. E- and P-selectins are only present

on the surface of endothelium activated by proinflammatory stimuli. Thus, leukocyte

rolling will only occur in inflamed tissues. L-selectins are constitutively expressed on

leukocytes. E-selectin is induced by cytokines such as interleukin-l-3 and tumor necrosis

factor-a and lipopolysaccharide in human umbilical vein endothelial cells.87' 8 Some 3-4

hours after activation, maximal levels of E-selectins are expressed on the surface. P-

selectins are stored in a-granules inside platelets, and in Weibel-Palade bodies in

endothelial cells. They are rapidly mobilized after stimulation with histamine or

thrombin.89 Expression is maximal at -5-10 minutes, and is cleared from the surface

with 60 minutes. This regulation may be an important avenue for future research.

Conceivably, force measurements may be performed on cells in varying environments.

Thus, a study of how selection expression affects the forces measured may be conducted.









Biophysical Parameter of Selectin-Ligand Interaction

Since this study involves measurements of the forces of specific receptor-ligand

interaction, a discussion of previous attempts at measuring the selectin-SLeX bond is

necessary. Based on the leukocyte rolling assumption, selections have been proposed to

have rapid rate constants for bond association and dissociation, and special mechanical

properties for tensile forces and bond dissociation.90-92 Also, the affinity of selections for

their ligands does not need to be especially high. Several groups93-98 had measured

binding between selections and the tetrasaccharide sialyl Lewis X or its stereoisomer sialyl

Lewis A to be fairly weak (dissociation constant Kd = 0.1-5 mM). Binding kinetics

between the interaction of L-selectin and its ligand, glycosylation-dependent cell adhesion

molecule-1 (GlyCAM-1) has also been determined.99 The authors showed that L-selectin

binds to GlyCAM-1 with a Kd of 108 [IM, and dissociates very fast, with a dissociation

rate constant kff of -10 s-1. Other studies with P-selectin and P-selectin glycoprotein

ligand-1 (PSGL-1) revealed a Kd of 200 nM, and a kff of -1.5 s-1.100

Other than fast association rates, high tensile strength of the selectin-ligand bond

was suggested to support the rolling function.101 Using a laminar flow chamber, the

experimenters visualized tethering and release of neutrophils on lipid bilayers containing

incorporated P-selectin. Flow subjects leukocytes to a shear force that increases the koff,

in the absence of shear, an "intrinsic kff" was determined by extrapolation to zero flow

rate. By analyzing the kinetics of the transient binding events (tethers), Alon, Hammer,

and Springerlo1 showed that the intrinsic koff for P-selectin as 1 s- The bond interaction

distance was determined to be 5 nm. An intrinsic koff of 7 s-1 was determined with similar

measurements for L-selectin, using L-selectin expressing neutrophils flowing over









substrates coated with L-selectin ligand.102 L-selectin mediated rolling appears to be

faster than rolling mediated by E- or P- selection, which agrees with the observation that

the koff of L-selectin is -10 s-1 higher than that for E- and P-selectins.103-105 The faster

rolling of leukocytes on L-selectins may also be explained by the fewer number of CB

domains. As stated earlier, the CB domains are involved in the extension of the selection

away from the membrane. Thus, the shorter L-selectin does not extend as much, which

leads to decrease time of interaction, and faster rolling.

Selectins are not the only molecules that support leukocyte rolling. Using in vitro

adhesion assays under flow, tensacin,106 vascular cell adhesion molecule-1, a4b7/mucosal

addressin cell adhesion molecule-1,107 and CD44/hyaluronanl08' 109 also allowed rolling.

Selectins are much more efficient at this task, however. For example, when the

differences between leukocyte rolling on immobilized antibodies versus LeX and SLeX

was examined,11 antibodies supported rolling only within a small range of site densities

and wall shear stresses, while the selection mediated rolling occurred throughout a wide

range. For example, the antibody PM-81 showed leukocyte rolling at a site density of

100-150 sites/pim2, a wall shear stress of 0.5 to 8.0 dyn cm-2, and a velocity of 0.2 to 0.6

lpm/sec. For leukocytes rolling on E-selectin, the range was 35-900 sites/pm2, 0.5 to 32.0

dyn cm-2, and a velocity of 1.3 to 10 lpm/sec. Shear stress above a critical threshold is

required to maintain rolling interactions in L-, E-, and P-selectins.111,112 It has been

suggested that the fluid shear may induce additional bond formation by deforming the cell

slightly after the first bond forms, thus increasing the time and area of the cell/substrate

contact. Taylor and coworkers developed a technique to explain the transition from

tethering to adhesion.113 Using a cone-plate viscometry, they showed that the binding









kinetics of selection and integrin are optimized to function at discrete shear rate and stress.

Thus, the leukocyte recruitment process requires specific flow rates and receptor

densities. Although this study measured forces in a static fluid chamber, future work may

involve measuring the adhesion at varying flow rates and receptor densities.



Selectin Ligands

The ligands for the selections are also important for a basic understanding of the

rolling process. Instead of the ligand binding through protein-protein interactions found

on most cell adhesion molecules, the ligands of the selections bind with carbohydrates

attached to either a scaffold protein or a lipid carrier molecule. By itself, the carrier

molecule is insufficient to define a selection ligand-it must be expressed in the correct

cellular environment to allow glycosylation (attachment of the carbohydrate moiety).

Oligosaccharides can bind to selections on several different molecules, even ones

not used as carriers physiologically. For example, BSA has been used as to bind

selectins.114'115 Also, oligosaccharides are not the only modifications required to make a

carrier molecule capable of binding selections. For example, tyrosine sulfonation is also

required to make PSGL-1 able to bind to selectins.116-118

Both glycolipids and glycoproteins can act as carriers for sialyl Lewis X.

Glycolipids carrying oligosaccharides such as sialyl Lewis X have been shown to support

rolling of E-selectin transfected cells, and of L-selectin expressing leukocytes.119 PSGL-1

was identified by expression cloning,120 and by affinity isolation as a 250-kDa disulfide-

linked dimer.121 It is the only ligand so far that has been demonstrated, in vivo, to

mediate leukocyte rolling on endothelium,122 and leukocyte diapedesis into inflamed









tissue. 12,124 This ligand is widely expressed in cells of lymphoid, dendritic, and myeloid

linkage.125 Thus, every leukocyte has PSGL-1. While the model system used in this study

does not involve PSGL-1, the BSA used in the experiments effectively simulates the

function of the PSGL-1.



Hypothesis

The characteristics of selections that allow leukocyte rolling are distinctive.

Selectins are cellular brakes: they must attach quickly, bind weakly, and then release the

ligand. Under flow, fast interactions between selection and ligand are essential because

leukocytes are rushing past the endothelium. They must also bind weakly, otherwise,

either the selection or the ligand will be torn out of the cell membrane by the momentum of

the leukocyte. To investigate the nature of the forces of adhesion between selections and

ligands, we have chosen to use Atomic Force Microscopy (AFM). The movement of the

AFM cantilever effectively simulates the rolling process. The initial contact between the

ligands on the leukocyte and the endothelium is reproduced when the tip is extended onto

the cell membrane. At this time, the sialyl Lewis X binds to the selection on the

endothelial cell surface. The retraction of the AFM tip simulates the rolling of the

leukocyte along the endothelium. As the leukocyte rolls, the ligand-receptor binding of

SLex to selection will be broken. The same occurs when the AFM tip retracts. During

retraction, the AFM will measure the force required to pull a selection to full extension,

and the break the binding between SLex and the selection. The understanding of this how

leukocytes roll may be applied toward improved pharmaceuticals-drug delivery devices

that target inflamed areas with blood vessels that express selections.









Materials and Methods

SLeX covalently linked to bovine serum albumin (BSA) was grafted onto silica

microspheres (r = 5pm) using glutaraldehyde and 3-aminopropyltriethoxysilane (APS).

Microspheres with BSA grafted directly onto the surface served as control for nonspecific

interactions. The microspheres were bonded to AFM cantilevers. Porcine vascular

endothelial cells (ECs) were grown using cell culture techniques on radiofrequency

plasma treated polydimethylsiloxane elastomer substrates. Ten independent observations

were made on test samples and controls as they were brought into contact with selection

expressing ECs. The Derjaguin approximation between spheres and flat plates was used

to normalize force values.



Grafting of 3-Aminopropyltriethoxysilane on Glass Slides and Glass Microspheres

A silane coupling agent, 3-aminopropyl triethoxysilane (APS), was grafted to

glass. First, the glass was cleaned with five-minute sonicating wash using chloroform

and methanol. After the wash, the glass was dried in air. A solution of dry toluene and

silane coupling agent were combined in a round bottom flask and allowed to mix before

addition of the silicon wafers. After the addition of the wafers, the solution was allowed

to stir before heating to the desired temperature. The reaction proceeded at temperature

for 12 hours after which time the wafers were removed, and rinsed with toluene. To cure

the silane coupling agent, the glass was dried and placed in an oven at 110 OC for 1 hour.

The same procedure was performed in Chapter 2. The efficacy of this procedure was

confirmed using x-ray photoelectron spectroscopy (XPS) (Table 2-1).









Limitations of the XPS only allowed characterization of glass slides, and not the

glass microspheres used in the experiment. However, given the same composition of

both slides and microspheres, and the same reaction used in both, one may state that the

successful grafting of the slides mean success for the microspheres.



Addition of Glutaraldehyde

An 8 % glutaraldehyde solution in phosphate buffered saline (PBS) at pH of 7.4

was added to the APS grafted microspheres. The mixture was left overnight with end-to-

end mixing, which allowed the aldehyde groups to react with the primary amine on the

APS.



Addition of BSA-SLex or BSA

Some 0.01 % of SLex-BSA, purchased from Oxford GlycoSciences, Ltd (Oxford,

England), in PBS was added to the glutaraldehyde/microsphere mixture. For the control

samples, 0.01 % BSA was added to the glutaraldehyde modified microspheres.

Carboxylic acid functional groups on the BSA react with the glutaraldehyde to covalently

link the SLex to the microsphere. After the modification, the microsphere was bonded to

the large, thick AFM cantilevers on contact mode tips. The microsphere is then bonded

to the AFM tip using epoxy. A micromanipulator is used to physically place the epoxy

and the microsphere onto the AFM tip. A micromanipulator is an instrument that allows

minute movements in the x, y, and z axes. Attached to one end of the micromanipulator

are thin tungsten filaments. Using one filament, the operator can deposit the epoxy glue

onto the AFM tip, and then, using another filament, pick up a microsphere and place it










onto the tip. See Figure 3-2 for scanning electron micrographs of the AFM tip with the

SLex modified tips. XPS was also performed on microspheres modified with

glutaraldehyde, and BSA to demonstrate the efficacy of the reactions (Table 3-1).


Table 3-1. Comparison of XPS data from plain glass, glass modified with glutaraldehyde, and
glass reacted with bovine serum albumen (BSA).

Plain Glass Glutaraldehyde-APS- BSA-Glutaraldehyde-
Plain Glass
Element Glass APS-Glass
(Mass Conc%) Ga
(Mass Conc%) (Mass Conc%)
Cls 15.3 31.1 57.6
Ols 46.0 40.2 28.6
Si2p 38.9 21.3 00.0
Nls 00.0 07.5 13.8



















I Eu. I *















































Figure 3-2. Scanning Electron Micrographs of SLeX grafted microspheres on
AFM cantilevers.









Cells Grown on PDMS Substrates

T-2 Silastic silicone elastomer resin and base were obtained from Dow Corning.

The exact composition of the T-2 Silastic is proprietary, but the basic components of

vinyl functionalized PDMS, and a platinum crosslinking agent are present. The base and

the resin are mixed together in a 10:1 ratio, degassed in a vacuum, and cast on silicon

substrate for 24 hours.



Plasma Treatment of PDMS Substrates

PDMS substrates were treated by radio frequency glow discharge plasma. Using a

RF Plasma Inc. HFS 401S instrument, set at 50 watts, the PDMS samples were plasma

treated under 40 mTorr of vacuum in Argon for 5 minutes. This treatment created

hydrophilic groups on the hydrophobic PDMS.



Porcine Vascular Endothelial Cell (PVEC) Culture

PVECs were obtained from Dr. Edward Block's group. The PVECs were treated

with trypsin, and then immersed in 12 mL of cell culture media of the following

composition. Some 4 % fetal bovine serum, 0.5 % penicillin/streptocillin, were added to

10 % RPMI medium 1640. Then, 25 jtL of gentamin, and 20 jtL of amphoterin B were

added to 50 mL of the RPMI medium 1640 mixed with fetal bovine serum and

penicillin/streptocillin.









PVECs Grown on PDMS

Plasma treated PDMS substrates were cut into 2.5 cm squares, and placed in 6-

well culture plates. The 6-well plates and PDMS were placed in a sterilization bag, and

treated with ethylene oxide (ETO). After the ETO treatment, the plates were degassed in

a sterile hood for 24 hours. Then, some 2 mL the PVECs in cell culture media were

added to the 6-well plates containing PDMS substrates. The experimenters adhered to

strict sterile procedures at all times.



Force Measurements

Force measurements were done on a Digital Instruments Nanoscope III atomic

force microscope. The experiment was conducted inside the fluid cell provided by the

manufacturer. The fluid cell allows the solution to be exchanged without disturbing the

experimental setup. A PDMS substrate grown with live PVECs was removed from the

cell culture media and placed on an AFM substrate. The tip modified with SLex was

placed in the tip holder, and measurements were taken directly on PVECs in phosphate

buffered saline, pH 7.4. To calculate the adhesion force and the interaction energy, the

radius of the particle was determined by optical microscopy and scanning electron

microscopy. The spring constant, 0.12 N/m for the tip was obtained from the

manufacturer (Digital Instruments, Santa Barbara, CA). To determine the nonspecific

force of adhesion, force measurements were also performed in 7 M urea. Urea is a

chaotropic agent that denatures proteins. Thus, the interactions measured would not

involve specific binding between the sialyl Lewis X and the selections.









Results and Discussion

The force curves showed two kinds of interactions. At small distances, there was

a large interaction between the microsphere and the cell surface in both the test samples

and the controls. At larger distances, there was a weaker interaction only between SLex

and the selections (Table 3.1). See Figure 3-3 for the force curves. Since there was no

interaction between the controls (BSA on microspheres, no SLeX), we hypothesized that

the long-range interaction was the specific binding of SLex to P- and E- selections on the

surface of the cell.

Table 3-2. Data for interaction between glutaraldehyde reacted and sialyl Lewis X reacted
microspheres and porcine vascular endothelial cells. The interaction between sialyl Lewis X and
the endothelial cells showed two interactions.

Micro e Ge SLex in 7M SLex-BSA, SLex-BSA
Microsphere Glutaraldehyde '
__________________Urea 1st interaction 2nd Interaction
Substrate PVECs PVECs PVECs PVECs
Distance (nm) 16.2 + 2.9 21.3 + 6.1 14.2 + 3.1 106.7 + 26.0
Force (nN/pm) -0.27 + 0.06 -0.28 + 0.07 -0.21 + 0.04 -0.46 + 0.08


Using the Derjaguin approximation between a sphere of 5 jpm radius and a flat

surface, the short-range attractive interaction between SLex and ECs was calculated to be

0.21 0.04 [N/pm (meanstandard deviation) at a distance of 14.2 3.1 nm. The

interaction between BSA and ECs was calculated to be 0.270.06 jN/pm at 16.2 2.9

nm. After performing a t-test on the values of the attractive force, a p-value of 0.03355

was found. Thus, there is likely to be a significant difference between the strength of the

interaction. A t-test was also performed on the distance of the force. In this case, a p-

value 0.14999 was calculated. The difference in distances of the force may not be

significant.






44


Between SLex and endothelial cells, a second interaction of 230.440.4 pN was

measured at 106.7 26.0 nm. There was no corresponding interaction on the control

samples. Using the Derjaguin approximation, this value came to an interaction of

0.460.08 [N/pm. This force appeared to involve uncoiling of the selection and rupture of

the receptor-ligand bond. The mechanism for leukocyte rolling appears to involve the

binding of the SLex to endothelial cells at short distances. Then the selection acts as an

uncoiling tether as the leukocyte rolls away from the endothelium. When the selection is

fully extended, the selectin-ligand bond ruptures. Figure 3-4 is a schematic of the

leukocyte rolling process.






Sialyl Lewis X with PVEC

2.0 Retract
-n : 1st Interaction
o : Extend
S2nd Interaction
CD
z I
-F 0I
...................


0 0 100.0 200.0

Separation (nm)



Figure 3-3. Sample force-separation curve showing both the initial nonspecific
interaction and the second snecific interaction.


























Figure 3-4. Schematic of how the leukocyte rolling process is
simulated by the AFM.

When the SLex grafted microspheres were brought into contact with endothelial cells in

7M urea, there was only one interaction. The long-range interaction did not occur. At a

distance of 21.3 6.1 nm, there was an attractive interaction of 0.28 0.07 nN/pm

calculated using the Derjaguin approximation. A t-test was performed between the

distance of the urea experiment and the first peak of the sialyl Lewis X experiment. A p-

value of 8.89E-06 indicated that there was a significant difference in the distance of

interaction. Between the forces of interaction in the urea experiment and the sialyl Lewis

X experiment, p = 0.067. When the urea data and the force between the glutaraldehyde

grafted microspheres and the ECs were compared, there were no significant differences in

distance (p = 0.556) or strength of the force. Based on this data, one can conclude that

the urea changed the binding between selections and sialyl Lewis X because the forces and

the distance of interaction between SLex grafted microspheres and ECs in PBS and in 7M

urea are different. The forces and distance of interaction were not demonstrably different

between SLex and ECs in urea, and glutaraldehyde grafted microspheres and ECs in PBS.









The denaturing effect of 7M urea may have altered the selections such that the binding

between SLex and the selections was quantitatively similar to the nonspecific binding

between glutaraldehyde grafted microspheres and ECs.

The interaction at distances greater than the distances for the second interaction

between sialyl Lewis X grafted microspheres and ECs were also investigated. The slope

of both the retraction and the extension curves between 200 nm and 700 nm were

recorded (Table 3-3). The value of the force at 700 nm (in nN/[tm) was subtracted from

the value at 200 nm, and divided by 500 nm, resulting in a slope (in [iN). All of the

slopes were extremely close to zero, which indicates that at distances above 200 nm, the

tip and the cell are not in contact. The lack of contact supports the explanation of selection

tethering given above. T-tests between the treatments indicate no significant differences

(P < 10E-4 in all cases).

Table 3-3. Slope measured between 200 nm and 700 nm separation.
Glutaraldehyde SLex in 7M Urea SLex-BSA
Slope of Retraction (iN) -3.2E-5 1.2E-3 7.0E-6 6.3E-4 4.7E-6 8.5E-4
Slope of Extension (iN) -6.0E-5 7.9E-4 -1.3E-5 1.4E-3 -3.2E-5 2.9E-4


These experiments demonstrate, for the first time by AFM, the fundamental forces

involved in selectin-ligand interactions on cell surfaces. The long-range attraction

between SLex and endothelial cells is indicative of SLex tethering by the selection. This

weak tethering slows down leukocytes as they roll, thus making selections vital for

leukocyte recruitment and diapedesis in the high shear stress vasculature. The short-

range attraction is due to adhesion between the cell membrane and the microsphere.

There are some differences between the strength of this short-range attraction for the test

samples and the controls. The differences may be due to variables in the graft density of









BSA on the microsphere surface. The graft density may vary because the BSA-SLex

would be less reactive, since the end modified with SLex would not react with the

glutaraldehyde.



Future Work

This initial measurement of selectin-SLex interactions on the surface of cells leads

to several intriguing possibilities in further basic research, and pharmaceutical and

biomaterials development. First, the selections may be further characterized and described

in the terms and relationships of polymer physics. This will hasten the integration of

physics, chemistry, and biology. This technique may be extended to other cell surface

adhesion molecules. Integrins are the involved in the leukocyte adhesion process after

selection mediated rolling. The effects of hormones, cytokines, and other cellular

signaling molecules on ligand-receptor binding can also be studied.

A better fundamental understanding of the selections may lead to developments in

pharmaceuticals and biomaterials. One possibility is the mapping of cell surfaces. The

distribution of selections, integrins, and antibodies on a cell surface may be mapped using

their binding properties. Then the cell surface composition may be related to the effect of

pharmaceuticals. The interactions between biomaterials and cells can also be studied

using such a technique.



Structure/Property Relations of Individual Selectin Domains

An investigation into how each portion of the selection molecule affects binding

and rolling can be performed. As stated in the background, the exact composition of the









selection differs among species and type (L-, E-, P-). A modification in the selection

composition, such as the reduction in the number of CB domains may change the distance

of the interaction, and reduce the effectiveness of leukocyte rolling.



Investigation into Integrin Mediated Adhesion

After being slowed by rolling, integrins act to bind the leukocyte tightly to the

endothelium surface prior to diapedesis. The forces of this interaction may be

investigated in a similar fashion to the SLex-selectin experiment. The ligand for integrins

is the repeat unit of the amino acids arginine, glycine, and aspartic acid, denoted by the

abbreviation RGD.126 The RGD tripeptide appears in fibronectin, vitronectin,

osteopontin, collagen, thrombospondin, fibrinogen, and von Willebrand factor. This

family of adhesion proteins and their integrin receptors act to provide cells with signals

for polarity, differentiation, position, traction for migration, anchorage. An RGD

containing molecule may be grafted to an AFM tip, and brought into contact with cells

expressing integrins. The forces of interaction may be measured and compared with the

selection experiment outlined previously.



Cellular Environment and Pharmaceutical Effects

The interplay between leukocytes and endothelial cells involve two modes of

communication. The first type, physical interaction mediated by receptor and counter-

receptor molecules was investigated in this thesis, through the AFM investigation of

selectin/sialyl Lewis X interaction. The second type occurs through the production of

soluble mediators. These soluble polypeptide mediators are known as cytokines. These









molecules stimulate the expression of P- and E-selectins on endothelial cells. The

addition of such signaling molecules may change the forces and distances of interaction,

which can be characterized by the AFM. Similarly pharmaceuticals that affect the

immune system also may be characterized by AFM.



Cell Surface Mapping

The increased interaction due to selectin-SLex binding can be used to create maps

of cell surfaces. As the modified AFM tip is scanned across a cell, areas with selections

will be contrasted against areas lacking in selections. Thus, the density of selections on a

cell surface may be measured. Recently, Hinterdorfer's group used molecular recognition

of an antigen by an antibody to map a mica surface adsorbed with lysozymes.40'127 The

same type of recognition imaging may be used with selections, and applied to studies of

the surfaces of ex vivo vasculature. The density of selection expression may be

characterized as a function of blood flow and distance from injury.














CHAPTER 4

EFFECT OF TOPOGRAPHY ON CELL ADHESION



Background

The previous chapters of this thesis have focused on specific receptor-ligand

interactions. To related the sub-micron scales of interaction investigated by the AFM to

cellular level effects in the hierarchical system of cell adhesion, I also looked at the how

topography affects cell adhesion. Specifically, I investigated how porcine vascular

endothelial cells will adhere to micropattemed polydimethylsiloxane surfaces. It is

known that cells display different behavior based on the topography of the surface. As

mentioned earlier, many cells show anchorance dependence, meaning that cell growth and

proliferation requires an adhesive surface. Thus, a better understanding of topographical

effects is an important step in the creation of improved tissue scaffolding. Also, the

physical effects of topography may be probed by the AFM. Although no study was

performed in this thesis, which elements within the cell are affected by topography as

evaluated by the AFM is a logical progression of this research. A review of the literature

about topographical effects on cell adhesion is necessary to show what work has been

done before.

Cell contacts often occur through a modality known as "focal adhesions." Focal

adhesions involve integrins, which were introduced earlier as important mediators of the

leukocyte rolling and adhesion process.128 These integrins bind to adhesive proteins such

50









as fibronectin through an RGD sequence. Several intracellular proteins are also involved

in binding, including talin, vinculin, and a-actinin. The a-actinin binds to actin

microfilaments, causing alignment of the cell.129 Since adhesive proteins such as

fibronectin and vitronectin are present in most cell culture media, their adsorption onto

biomaterial surfaces can significantly affect cell adhesion studies. Once the proteins have

adsorbed to the biomaterial, cells adhere through integrin binding, and then begin to

produce endogenous fibronectin to be deposited on the substrate. The deposition of

fibronectin depends on the displacement of other proteins already adsorbed onto the

substrate. The displacement is occurs more readily in hydrophilic substrates than

hydrophobic substrates.130,131 The relative ease of protein displacement does not

necessarily explain whether hydrophilic or hydrophobic substrates will promote cell

adhesion, however. The initial deposition of fibronectin may cause hydrophobic

substrates to be conducive to cell adhesion.

To study the effect of topography on cell adhesion, a precise pattern of topography

must be first generated. The most common method for producing patterns in PDMS is to

first lithographically produce the pattern on a silicon wafer, and then replicate that pattern

by casting onto the substrate. The type of lithography used is optimized for the size of

the pattern. The smallest patterns are produced with direct write laser lithography and

AFM lithography while those on the 2-10 jpm scale are produced with UV

photolithography.132

Patterns created on a substrate affects how cells bind and align, but the

mechanism of this effect, referred with the encompassing term "contact guidance" have

not been explained sufficiently. The width, the depth, the spacing between patterns, and









the preparation method all seem to have an effect.132 Cell adhesion is also affected by the

surface chemistry, the surface free energy and the kinetics of the binding.132-134 Some

general phenomena have been observed, however. First, groove widths in the sub-

cellular range influences cell alignment in the direction of the grooves.128,135 The depth

of the groove has some influence, as deeper grooves allow bridging by the cell.136

Surface features in the range of 1-5 [im seem to promote cellular conformation.

Schmidt and von Recum performed experiments with cell adhesion to patterned

substrates.133 Using microtextured silicone made by casting on a photoresist coated glass

mask or silicon wafer. They varied the depth of the feature, and the spacing between

different texture events. How murine peritoneal macrophages adhere to wells that ranged

from smooth to 10 jam in width was studied. The arrays of these wells were arranged at

variable distances from each other, causing an anisotropic pattern. The wells were also

arranged a constant distance of 20.4 jam as an isotropic control. They found that some

portion of the pattern must be anisotropic to invoke cell guidance. They also postulated

that texture affects how cells adhere through both the surface area of adhesion and the

perimeter of the area.

Contact guidance is also dependent on the type of cell or protein adhesion studied.

In a later study by von Recum and others, patterns with groove widths of 2, 5, and 10 Pam

were coated with rat dermal fibroblasts.134 The grooves were 0.5 jam deep, with ridges of

the same width acting as separation. By measuring the alignment of the ECM protein

deposited, the 2 jam pattern caused the most alignment, whereas the 5 and 10 Pam patterns

behaved more like a smooth surface. Variation of the chemical groups on a surface has

also been studied. Britland et al. used photolithography to create a pattern of









aminosilane (AS) and a pattern of hydrophobic silane (HS) on glass slides.137 The baby

hamster kidney cells used in the study adhered preferentially on the aminosilane pattern.

Cells also showed elongation along 10 jam AS tracks. One may view the AS pattern as a

ligand that binds cells with greater affinity than the HS. Xiao and Truskey examined the

ligand affinity issue.138 This group also sees greater adhesion with increased ligand

density, which they attribute to greater surface area. It is not a linear relationship. Using

cyclic and linear compounds, they showed that similar compounds could have

dramatically different affinities for the same receptor. This indicates that the combination

of chemistry and topographical cues has a significant effect on cell adhesion. Besides the

ligand affinity, ligand density also affects cell adhesion. As discussed in Chapter 1,

Hammer's group was able to create a restricted mathematical model that solved how the

ligand density affects the kinetics of cell detachment.20

Many of the studies done on patterned substrates have been conducted with

PDMS. It provides an easily processed, low energy substrate that is well understood.

Due to the low modulus, a precise reproduction of the pattern may be difficult to achieve,

however. Thus, polystyrene, which allows improved precision of patterning, has been

studied. Using polystyrene as the substrate, Walboomers et al. observed that the

sharpness of the ridges was more important to contact guidance than the reduction of

artifacts left behind from lithography.139 The investigators also hypothesized that the

contact guidance was caused by mechanical stress exerted by the topography.

Several groups focused on a patterned surface chemistry as a means of producing

controlled cell growth on a surface. Potember, Matsuzawa, and Liesi grew dissociated

embryonic rat hippocampal neurons on photolithigraphically patterned diethylenetriamine









(DETA) substrates.14 A synthetic peptide, P1543 was used to make the surface bioactive,

with plain DETA substrate as a control. Self-assembled monolayers prepared from n-

octadecyltrichlorosilane (OTS) were used as a nonadhesive component. As a proof of

nonadhesion, cells did not attach to the OTS substrates after 1.5 hours of incubation. The

patterned DETA substrates were the poorest surfaces for neuron adhesion, producing

clumps of short cells that did not mature, while the P1543 coated substrates promoted

directional growth of 150-m neurites.

Healy et al. used a completely chemical pattern on their substrate to examine the

kinetics of organization and mineralization of rat calvaria bone cells.141 The patterns

were made by photolithographically producing a smooth surface of alternating surface

chemistry. The pattern was alternating 50 ism strips of dimethyldichlorosilane (DMS)

and N-(2-aminoethyl)-3-aminopropyl-trimethoxysilane (EDS). The experimenters

concluded that the charge of the molecules on the surface was not the dominant

mechanism for determining cell adhesion with bone cells. Rather, differences in cell

adhesion were due to the adsorption of vitronectin onto the EDS regions, and the lack of

adsorption in the DMS regions. The Healy group also evaluated the strength of cell

adhesion through a probabilistic analysis.142 Using a Weibull distribution, the

investigators found a wide variation in the strength of adhesion in a cell population. This

variation may be due to differences in the surface density of ligands, the density of

receptor-ligand bonds, and the areas of focal contacts.

While most of the previous work described in vitro cell adhesion, a significant

amount of in vivo work has also been performed. Since biomaterials often respond

differently in vivo, a review of the literature is necessary to evaluate the differences. In an









attempt to promote the formation of a stable pseudo-neointima, patterns were made of

Dacron polyester fibrils (250 jpm long and 25 jpm in diameter, variable spacing between

fibrils) on polyurethane vascular patches. These patches were then implanted in sheep,

along with a non-patterned control. One week after implantation, the entirety of the

texture surface was covered by thrombus, while the non-textured surface was only

partially covered. The authors hypothesized that the cause of the increase in thrombosis

on the texture substrate was due to changes in flow conditions-the textured surface

reduces local flow velocity.128

In another study, patterned silicone disks were implanted under the skin of New

Zealand White rabbits.143 The authors hypothesized that the microtexture would

interlock with the fibrous capsule that typically forms around implants. This interlocking

would limit movement of capsule, thus reducing mechanical irritation, and therefore, the

capsule thickness. The thickness of the fibrous capsule did not vary significantly between

the textured and smooth surfaces, however. The authors attributed this result to

insufficiencies in the dimensions of the grooves and ridges. SEM showed that fibroblasts

did not orient parallel to the grooves in vivo, which contradicts previous in vitro

studies.144 Since the work related in my thesis is an in vitro study, some of the trends

observed here may be expected to differ in vivo.

Walboomer and colleagues have extensively studied contact guidance both in

vitro and in vivo, and may give some insights into how the two environments affects cell

adhesion. In one study, rat dermal fibroblasts were cultured on polystyrene (PS), titanium

coated polystyrene (PSTi), silicone (SIL), and poly-L-lactic acid (PLL) of depths of 0.5

[m, and groove widths between 1 and 10 rm.145 Treatment with radiofrequency glow









discharge plasma (RF plasma) to increase the hydrophilicity was another variable. The

common topography between the surfaces indicates that differences in proliferation and

orientation may be attributed to surface chemistry effects. The RF plasma treatment

increased cell proliferation. For the untreated samples, PLL and SIL showed the most

proliferation, with less for PS and PSTi. Cell alignment seems dependent on both the

groove width and the surface chemistry. So for PS and SIL, cells did not align on 10 ipm

wide grooves, but did on 1 [pm grooves. No such correlation with groove width was

observed with PSTi, however. Cells aligned regardless of the width. The fibroblasts

were most strongly aligned on the PLL, and increase in alignment correlated with

decreasing groove width. In another study, Walboomers et al. examined the effect of

depth (0.5, 1.0, 1.8, and 5.4 tm) and width (1, 2, 5, 10, and 20 tm) of fibroblast

alignment on polystyrene.146 The cell orientation seems to increase with groove depth:

1.8 and 5.4 pm showed the most alignment. Fibroblasts exhibited bridging of the deeper

grooves, which the authors attributed to high membrane stiffness that do not allow the

cell to bend into the deep grooves. Contrary to previous observations on PS,145 the

groove width did not correlate to significant differences in cell alignment. For in vivo

studies, Walboomer et al. implanted microtextured PS in goats147 and microtextured

silicone in guinea pigs.148 The increased complexity of the animal model leads to

different observations in vivo and in vitro. In vivo, the interplay of several cell types leads

to the formation of a fibrous capsule around the textured implant. In both studies,

showed no significant differences in morphology of the capsules around textured and

smooth implants.









In summary, the previous work mainly involved phenomenological observations

of how and whether cells proliferated and aligned to various substrates. The exact

cellular mechanisms of contact guidance have not been elucidated. The staining methods

outlined in this chapter are a first step toward determining which portion of the cell is

affected by topography. Furthermore, the techniques involving AFM force measurements

related in Chapters 2 and 3 may be applied to this study. The changes in the expression

of cell adhesion molecules due to topography may be monitored by AFM.



Materials and Methods

We grew porcine vascular endothelial cells on polydimethyl siloxane substrates

(PDMS). The PDMS substrates were cast on polystyrene templates made through

replication of microtextured silicon wafers. Other members of the Brennan research

group generously provided the polystyrene templates and the microtextured silicon

wafers, so the details of their manufacture will not be divulged here. After treatment with

RF plasma, substrates were sterilized. Cells were obtained from Dr. Edward Block's

group at the Malcom Randall Veterans Adminstration medical center, and grown using

standard cell culture techniques. The cell morphology on PDMS substrates was observed

through optical microscopy, scanning electron microscopy, and scanning laser confocal

microscopy.



Synthesis of PDMS

T-2 Silastic silicone elastomer resin and base were obtained from Dow Corning.

The exact composition of the T-2 Silastic is proprietary, but the basic components of









vinyl functionalized PDMS, and a platinum crosslinking agent are present. The base and

the resin are mixed together in a 10:1 ratio, degassed in a vacuum, and cast on patterned

silicon substrates for 24 hours. In this study the depth of the topography was a constant 5

lm. Groove width was varied between 5, 10, and 20 im. Nobs were also made with

separations of 5, 10, and 20 im.



Porcine Vascular Endothelial Cell (PVEC) Culture

PVECs were obtained from Dr. Edward Block's group. The PVECs were treated

with trypsin, and then immersed in 12 mL of cell culture media of the following

composition. Some 4 % fetal bovine serum, 0.5 % penicillin/streptocillin, was added to

10 % RPMI medium 1640. Then, 25 jtL of gentamin, and 20 jtL of amphoterin B were

added to 50 mL of the RPMI medium 1640 mixed with fetal bovine serum and

penicillin/streptocillin.



Plasma Treatment of PDMS Substrates

Some of the PDMS substrates were treated by radio frequency glow discharge

plasma to improve cell adhesion by creating hydrophilic groups on the hydrophobic

PDMS. Using a RF Plasma Inc. HFS 401S instrument, set at 50 watts, the PDMS

samples were plasma treated at 40 mTorr of vacuum in Argon for 5 minutes.



Fibronectin Coating of PDMS Substrates

Another method to improve cell adhesion was by coating the PDMS substrate

with fibronectin. After sterilization, the substrates were soaked in a 4 [ig/mL solution of









fibronectin (Sigma-Aldrich, St. Louis, MO) for two hours. Then, the solution was

siphoned off, and endothelial cells were added immediately.



PVECs Grown on PDMS

Plasma treated PDMS substrates were cut into 2.5 cm squares, and placed in 6-

well culture plates. The 6-well plates and PDMS were placed in a sterilization bag, and

treated with ethylene oxide (ETO). After the ETO treatment, the plates were degassed in

a sterile hood for 24 hours. Then, some 2mL the PVECs in cell culture media were added

to the 6-well plates containing PDMS substrates. The experimenter adhered to strict

sterile procedures at all times. Observations were made after three days of growth. The

cell culture media was changed every other day.



Optical Microscopy

The cells grown on PDMS were placed under an optical microscope, and viewed

at 10X. The cells were not stained. Pictures were taken with a digital camera.



Scanning Electron Microscopy

The experimenters used a JEOL 6400 scanning electron microscope to view the

microtextured substrates. After fixing the cells with Trump's solution (glutaraldehyde

and formaldehyde), the samples were washed with phosphate buffered saline, pH = 7.4.

Then, a sequential wash was performed using DI water and 100 % ethanol in 20 vol%

increments. For example, initially, the sample was washed with 100 % DI water. Then,

the sample was washed with 20 vol% ethanol in DI water. The ethanol was gradually









increased to 100 % ethanol. The samples were allowed to dry in air, and mounted on

SEM sample holders with carbon glue. The samples were then coated with gold and

palladium.



Scanning Laser Confocal Microscopy

Two methods of staining were used. First, von Willebrand factor, a blood

glycoprotein required for platelet adhesion that is made by endothelial cells, was stained.

The endothelial cells grown on PDMS were washed twice with phosphate buffered saline

(PBS), and fixed with acetone at -20 OC for 15 minutes. After another two washes in

PBS, the substrates were placed in a 0.46 [g/mL solution of goat anti-human von

Willebrand factor biotinylated IgG for 1 hour. The sample was then washed in PBS

twice, and placed in 5 [g/mL solution of streptavidin-fluorescein for 1 hour. After

another wash with PBS, the sample was inverted onto a glass coverslip. Silicone vacuum

grease was placed on the outer edges of the sample to create some separation between the

sample and the glass, so as to not crush the cells and deform the patterns.

The second method used fluorescein labeled GSL-1 isolectin B4 (Burlingame,

CA) to stain carbohydrates on the cell membrane. After washing with PBS, fixing wit

acetone, and a second wash with PBS (same details as above), the fixed endothelial cells

on PDMS were placed in a 10 pLg/mL solution of the GSL-1 isolectin B4. Since the

isolectin was already labeled with fluorescein, no further staining was necessary. The

sample was washed, and placed on a glass coverslip with the same methods reported in

the previous paragraph. Images were taken with a Zeiss LSM 430 scanning laser

confocal microscope.









Results and Discussion

Porcine vascular endothelial cells adhered minimally to plain PDMS substrates

(Fig 4.1). When the substrate was treated with fibronectin, cell adhesion was improved

(Fig 4.2). Scanning electron microscopy showed that addition of fibronectin obscured the

topography, however (Fig 4.3). Thus, RF plasma was used to modify the surface. The

RF plasma treated surface improved cell adhesion (Fig 4.4). The morphology of the cells

on the PDMS substrate was investigated using the SEM. Some cells appear to pull the

ridges toward themselves (Fig 4.5). When the PDMS substrate is viewed along the

topography, some cells appear to be bridging the groove (Fig 4.6). Cells observed under

the scanning laser confocal microscope show deformation of the ridges (Fig 4.7), but the

resolution was not high enough to observe bridging. Walboomers et al. previously

observed similar deformation of silicone substrates with rat dermal fibroblasts.145
































Figure 4-1. Plain PDMS. Efforts at growing endothelial cells the
surface did not lead to a confluent layer. There are very few cells on
the surface. The ridges are separated by 10 gim.


bv ~~i*

a


Figure 4-2. Endothelial cells grown on textured substrate coated with
fibronectin. The separation between the on the upper right is 5 gin, and
between the ridges on the lower left is 10 Gnm.


































Figure 4-3. SEM of PDMS substrate coated with fibronectin. The
fibronectin seem to obscure the features, and deform the ridges.


Figure 4-4. Optical microscopy picture of PDMS substrate treated with
RF plasma. Notice the confluent cell layer in the middle region with no
topography. The separation between ridges on the upper right is 5 G[m,
and on the lower left is 10 Gm.

















. .. .


Si..ii


Figure 4-5. SEM of a cell deflecting the ridges of the PDMS substrate.


.. .... ....

"
"m::


.... ...


Figure 4-6. SEM of 5gm high ridges. A cell appears to be bridging the
10im groove between the ridges.


....... ...





















































Figure 4-7. Scanning laser confocal microscopy image of endothelial cells stained with
goat anti-human von Willebrand factor biotinylated IgG on PDMS substrates. The
sample is at a small tilt, which means that the focal plane only captures a few cells
(indicated by the green coloring). Thus, cells are not present in greater quantities on the
bottom of the sample. Notice that the cells appear to distort the PDMS ridges.









This thrust of my thesis is inchoate. The methodologies of observing cell adhesion (i.e.

cell staining for laser confocal microscopy, sample preparation for SEM) were developed,

but deficiencies in time precluded a systematic study. Several issues remain unresolved.

The preparation of samples for SEM requires desiccation of the cells, which may create

artifacts. The staining intensity and resolution of the laser confocal microscope must be

improved to allow the creation of meaning three-dimensional projections from pictures

taken at a various focal planes.



Future Work

Endothelial cell adhesion on the microtextured PDMS substrates, as characterized

through laser confocal microscopy, will be correlated to the width of ridges, the width of

the valleys, and the height of the ridges. The topography for maximal and minimal cell

adhesion may be determined. Other types of cells may be introduced to the system.

Smooth muscle cells may be used in conjunction with endothelial cells to determine if the

interplay of the two cell types creates variations in contact guidance. Since the ultimate

goal of this study is improved biomaterials for clinical applications, a study of the

behavior microtextured devices in vivo is necessary. The patterned surfaces will be

converted to a device, and implanted into an animal model.















CHAPTER 5

CONCLUSION



In my thesis I investigated two aspects of cell adhesion. At the molecular level,

specific receptor-ligand interactions were probed. In the avidin-biotin system, I observed

a significant increase in the force of adhesion, which indicates that avidin grafted onto

glass substrates bound specifically to biotin. I also showed that the addition of free biotin

to the environment in which the force measurements were taken did not alter the force of

adhesion specifically. In the selectin-sialyl Lewis X system, I observed interactions at

two distances. First, a nonspecific interaction attributed to the adhesion of the

microsphere to the cell membrane occurred at close range. At longer distances, a second

interaction occurred. Based on the model of leukocyte rolling, the second interaction

indicates elastic tethering of the sialyl Lewis X by the selection. This is the first

measurement of the adhesive forces of selections on endothelial cells by atomic force

microscopy.

At the cellular level, I investigated the effect of topography on endothelial cell

proliferation and orientation. Porcine vascular endothelial cells were grown on PDMS

substrates. The proliferation of the cells was improved when the substrate was coated

with fibronectin or made more hydrophilic with RF plasma. The cells appear to deform

the ridges on the PDMS, and possibly bridge the groove between ridges.









More work is required in both of these topics. By measuring the adhesive forces

with the AFM, an investigator may determine both the density and the affinity of

expressed of cell adhesion molecules. This may lead to a better understanding of how

topography affects cell adhesion. The surface densities of selections or other cell adhesion

molecules may correlate with differences in the topography of the substrate. While the

two aspects of cell adhesion covered in my thesis may seem dissimilar-they occur at

different levels of complexity, with further work in the molecular level, the mechanisms

behind contact guidance may be elucidated.















LIST OF REFERENCES



1. Rubin, G.M., Yandell, M.D., Wortman, J.R., Gabor Miklos, G.L., Nelson, C.R.,
Hariharan, I.K., Fortini, M.E., Li, P.W., Apweiler, R., Fleischmann, W., Cherry,
J.M., Henikoff, S., Skupski, M.P., Misra, S., Ashburner, M., Bimey, E., Boguski
M.S., Brody, T., Brokstein, P., Celniker, S.E., Chervitz, S.A., Coates, D., Cravchik,
A., Gabrielian, A., Galle, R.F., Gelbart, W.M., George, R.A., Goldstein, L.S., Gong,
F., Guan, P., Harris, N.L., Hay, B.A., Hoskins, R.A., Li, J., Li, Z., Hynes, R.O.,
Jones, S.J., Kuehl, P.M., Lemaitre, B., Littleton, J.T., Morrison, D.K., Mungall, C.,
O'Farrell, P.H., Pickeral, O.K., Shue, C., Vosshall, L.B., Zhang, J., Zhao, Q., Zheng,
X.H., Zhong, F., Zhong, W., Gibbs, R., Venter, J.C., Adams, M.D. & Lewis, S.
(2000) Science 287, 2204-15.

2. Chervitz, S., Aravind, L., Sherlock, G., Ball, C.A., Koonin, E.V., Dwight, S.S.,
Harris, M.A., Dolinski, K., Mohr, S., Smith, T., Weng, S., Cherry, J.M. & Botstein,
D. (1998) Science 282, 2022-8.

3. Moore, W. S. (1993) (W. B. Saunders Company, Philadelphia), pp. 870.

4. Bevilacqua, M. P., Nelson, R. M., Mannori, G. & Cecconic, O. (1994) Annu Rev
Med 45, 361-378.

5. Bevilacqua, M. P. & Nelson, R. M. (1993) J Clin Invest 91, 379-387.

6. Travis, J. (1993) Science 40, 906-908.

7. Tanabe, K. K. & Saya, H. (1994) Crit Rev Oncog 5, 201-12.

8. Greisler, H. P. (1991) Xanthus Intelligence Unit Report No. 3 (R. G. Landes
Company, Austin, TX).

9. Schneider, A., Melmed, R. N., Schwalb, H., Karck, M., Vlodavsky, I. & Uretzky, G.
(1992) J Vasc Surg 15, 649-656.

10. Magometschnigg, H., Kadletz, M., Vodrazka, M., Dock, W., Grimm, M.,
Grabenwoger, M., Minar, E., Staudacher, M., Fenzl, G. & Wolner, E. (1992) J Vasc
Surg 15, 527-35.









11. Cenni, E., Arciola, C. R., Ciapetti, G., Granchi, D., Savarino, L., Stea, S.,
Cavedagna, D., Curti, T., Falsone, G. & Pizzoferrato, A. (1995) Biomaterials 16,
973-976.

12. Sbarbati, R., Giannessi, D., Cenni, M. C., Lazzerini, G., Verni, F. & Caterina, R. D.
(1991) Biomaterials 14, 491-498.

13. Paulsson, M., Gouda, I., Larm, O. & Ljungh, A. (1994) J Biomed Mater Res 28,
311-317.

14. Phaneuf, M. D., Berceli, S. A., Bide, M. J., Quist, W. C. & LoGerfo, F. W. (1997)
Biomaterials 18, 755-765.

15. Nicol, A., Gowda, D. C. & Urry, D. W. (1992) J Biomed Mater Res 26, 393-413.

16. Ito, Y., Kajihara, M. & Imanishi, Y. (1991) J Biomed Mater Res 25, 1325-1337.

17. Cozens-Roberts, C., Lauffenburger, D. A. & Quinn, J. A. (1990) Biophys J 58, 841-
856.

18. Cozens-Roberts, C., Quinn, J. A. & Lauffenburger, D. A. (1990) Biophys J. 58,
107-125.

19. Bell, G. I. (1978) Science 200, 618-627.

20. Ward, M. D., Dembo, M. & Hammer, D. A. (1995) Ann Biomed Eng 23, 322-331.

21. Hammer, D. A. & Apte, S. M. (1992) Biophys. J. 63, 35-57.

22. Israelachvili, J. N. & Adams, G. E. (1978) J Chem Soc Faraday Trans. 1 74, 975.

23. Binnig, G., Rohrer, H., Gerber, C. & Weibel, E. (1982) Phys Rev Lett 49, 57-61.

24. Binnig, G., Quate, C. F. & Gerber, C. (1986) Phys Rev Lett 56, 930-933.

25. Suzuki, A., Yamazaki, M., Kobiki, Y. & Suzuki, H. (1997) Macromolecules 30,
2350-2354.

26. Suzuki, A., Yamazaki, M. & Kobiki, Y. (1995) J Chem Phys 104, 1751-1757.

27. Craig, V. S. J., Hyde, A. M. & Pashley, R. M. (1996) Langmuir 12, 3557-3562.

28. Ducker, W. A., Senden, T. J. & Pashley, R. M. (1991) Nature 353, 239-241.

29. Ducker, W. A., Senden, T. J. & Pashley, R. M. (1992) Langmuir 8, 1831-1836.

30. Meagher, L. & Pashley, R. M. (1995) Langmuir 11, 4019-4024.









31. Karaman, M. E., Meagher, L. & Pashley, R. M. (1993) Langmuir 9, 1220-1227.

32. Meagher, L. & Craig, V. S. J. (1994) Langmuir 10, 2736.

33. Larson, I., Drummond, C. J. & Chan, D. Y. C. (1993) J Am Chem Soc 115, 25.

34. Biggs, S., Chow, M. K., Zukoski, C. F. & Grieser, F. (1993) J Colloid Interface Sci
160, 511.

35. Sagvolden, G., Giaever, I., Pettersen, E. O. & Feder, J. (1999) Proc Natl Acad Sci
USA 96, 471-476.

36. Holland, N. B., Siedlecki, C. A. & Marchant, R. E. (1999) J Biomed Mater Res 45,
167-174.

37. Florin, E., Moy, V. & Gaub, H. (1994) Science 264, 415.

38. Moy, V., Florin, E. & Gaub, H. (1994) Coll Surf A 93, 343.

39. Ludwig, M., Moy, V., Rief, M., Florin, E. & Gaub, H. (1994) Microsc Microanal
Microstruct 5, 321.

40. Raab, A., Han, W., Badt, D., Smith-Gill, S. J., Lindsay, S. M., Schindler, H. &
Hinterdorfer, P. (1999) Nat Biotechnol 17, 902-905.

41. Ducker, W. A., Xu, Z. & Israelachvili, J. N. (1994) Langmuir 10, 3279-3289.

42. Israelachvili, J. N., Tirrell, M., Klein, J. & Almog, Y. a. (1984) Macromolecules 17,
204-209.

43. Israelachvili, J. N. & Wennerstrom, H. (1996) Nature 379, 219-225.

44. Kamiyama, Y. & Israelachvili, J. (1992) Macromolecules 25, 5081-5088.

45. Klein, J., Kamiyama, Y., Yoshizawa, H., Israelachvili, J. N., Fetters, L. J. & Pincus,
P. (1992) Macromolecules 25, 2062-2064.

46. Klein, J., Kamiyama, Y., Yoshizawa, H., Israelachvili, J. N., Fredrickson, G. H.,
Pincus, P. & Fetters, L. J. (1993) Macromolecules 26, 5552-5560.

47. Ruths, M., Yoshizawa, H., Fetters, L. J. & Israelachvili, J. N. (1996)
Macromolecules 29, 7193-7203.

48. Wong, J. Y., Kuhl, T. L., Israelachvili, J. N., Mullah, N. & Zalipsky, S. (1997)
Science 275, 820-822.

49. Klein, J. & Luckham, P. (1982) Nature 300, 429-431.









50. Klein, J. (1983) J Chem Soc Faraday Trans 179, 99-118.

51. Klein, J. & Luckham, P. F. (1984) Macromolecules 17, 1041-1048.

52. Klein, J., Kumacheva, E., Mahalu, D., Perahia, D. & Fetters, L. J. (1994) Nature
370, 634-636.

53. Kumacheva, E., Klein, J., Pincus, P. & Fetters, L. J. (1993) Macromolecules 26,
6477-6482.

54. Luckham, P. F. & Klein, J. (1985) Macromolecules 18, 721-728.

55. Luckham, P. & Klein, J. (1990) J Chem Soc Faraday Trans I 86, 1363-1368.

56. Taunton, H. J., Toprakcioglu, C., Fetters, L. J. & Klein, J. (1990) Macromolecules
23, 571-580.

57. Briscoe, B., Luckham, P. & Zhu, S. (1996) Macromolecules 29, 6208-6211.

58. Claesson, P. (1993) Colloids Surf A: Phys Eng Aspects 77, 109-118.

59. Claesson, P. M., Ederth, T., Bergeron, V. & Rutland, M. W. (1996) Adv Colloid
Interface Sci 67, 119-183.

60. Claesson, P. M., Blomberg, E., Paulson, O. & Malmsten, M. (1996) Colloids Surf
A: Phys Eng Aspects 112, 131-139.

61. Malmsten, M., Claesson, P. M., Pezron, E. & Pezron, I. (1990) Langmuir 6, 1572-
1578.

62. Malmsten, M. & Claesson, P. M. (1991) Langmuir 7, 988-994.

63. Green, N. M. (1975) Adv Protein Chem 29, 85-133.

64. Florin, E.-L., Moy, V. T. & Gaub, H. E. (1994) Science 264, 415-417.

65. Moy, V. T., Florin, E.-L. & Gaub, H. E. (1994) Science 266, 257-259.

66. Weber, P. C., Ohlendorf, D. H., Wendoloski, J. J. & Salemme, F. R. (1989) Science
243, 85-88.

67. Hendrickson, W. A., Pahler, A., Smith, J. L., Satow, Y., Merrit, E. A. &
Phizackerley, R. P. (1989) Proc Natl Acad Sci USA 86, 2190-2194.

68. Weber, P. C., Pantoliano, M. W. & Salemme, F. R. (1995) Acta Crystalllogr D 51,
590-596.

69. Jones, M. L. & Kurzban, G. P. (1995) Biochemistry 34, 11750-11756.









70. Livnah, O., Bayer, E. A., Wilchek, M. & Sussman, J. L. (1993) Proc Natl Acad Sci
USA 90, 5076-5080.

71. Chilkoti, A. & Stayton, P. S. (1995) J Am Chem Soc 117, 10622-10628.

72. Pugliese, L., Coda, A., Malcovati, M. & Bolognesi, M. (1993) J Mol Biol 231, 698-
710.

73. Weber, P. C., Wendoloski, J. J., Pantoliano, M. W. & Salemme, F. R. (1992) J Am
Chem Soc 114, 3197-3200.

74. Miyamoto, S. & Kollman, P. A. (1993) Proteins 16, 226-245.

75. Chilkoti, A., Boland, T., Ratner, B. D. & Stayton, P. S. (1995) Biophys J 69, 2125-
2130.

76. Grubmuller, H., Heymann, B. & Tavan, P. (1996) Science 271, 997-999.

77. Tedder, T. F., Steeber, D. A., Chen, A. & Engel, P. (1995) FASEB J 9, 866-873.

78. Etzioni, A., Phillips, L. M., Paulson, J. C. & Harlan, J. M. (1995) Ciba Found Symp
189, 51.

79. Seekamp, A., Till, G. O., Mulligan, M. S., Paulson, J. C., Anderson, D. C.,
Miyasaka, M. & Ward, P. A. (1994) Am J Pathol 144, 592-598.

80. Subramaniam, M., Saffaripour, S., Van De Water, L. V. D., Frenette, P. S.,
Mayadas, T. N., Hynes, R. O. & Wagner, D. D. (1997) Am J Pathol 150, 1701-
1709.

81. Santos-Benito, F. F., Fernandez-Meyorales, A. & Martin-Lornas, M. (1992) J Exp
Med 176, 915-918.

82. Baggiolini, M., Dewald, B. & Moser, B. (1997) Annu Rev Immunol 15, 675-705.

83. Zimmerman, G. A., McIntyre, T. M. & Prescott, S. M. (1996) J Clin Invest 98,
1699-1701.

84. Drickamer, K. (1988) J Biol Chem 268, 9557-9560.

85. Patel, K. D., Nollert, M. U. & McEver, R. P. (1995) J Cell Biol 131, 1893-1902.

86. Vestweber, D. & Blanks, J. E. (1999) Physiol Rev 79, 181-213.

87. Bevilacqua, M. P., Pober, J. S., Mendrick, D. L., Cotran, R. S. & Jr, M. A. G.
(1987) Proc Natl Acad Sci USA 84, 9238-9242.









88. Pober, J. S., Lapierre, L. A., Stolpen, A. H., Brock, T. A., Springer, A. T., Fiers, W.,
Bevilacqua, M. P., Mendrick, D. L. & Jr, M. A. G. (1987) J Immunol 138, 3319-
3324.

89. Geng, J.-G., Bevilacqua, M. P., Moore, K. L., McIntyre, T. M., Prescott, S. M.,
Kim, J. M., Bliss, G. A., Zimmerman, G. A. & McEver, R. P. (1990) Nature 343,
757-760.

90. Dembo, M., Torney, D. C., Saxman, K. & Hammer, D. (1988) Proc R Soc Lond B
Biol Sci 234, 55-83.

91. Hammer, D. A. & Apte, S. M. (1992) Biophys J 63, 35-57.

92. Lawrence, M. B. & Springer, T. A. (1991) Cell 65.

93. Brandley, B. K., Kiso, M., Abbas, S., Nikrad, P., Srivasatava, O., Foxall, C., Oda,
Y. & Hasegawa, A. (1993) Glycobiology 3, 633-641.

94. Cooke, R. M., Hale, R. S., Lister, S. G., Shah, G. & Weir, M. P. (1994)
Biochemistry 33, 10591-10596.

95. Foxalll, C., Watson, S. R., Dowbenko, D., Fennie, C., Lasky, L. A., Kiso, M.,
Hasegawa, A., Asa, D. & Brandley, B. K. (1992) J Cell Biol 117, 895-902.

96. Jacob, G. S., Kirmaier, C., Abbas, S. Z., Howard, S. C., Steininger, C. N., Welply,
J. K. & Scudder, P. (1995) Biochemistry 34, 1210-1217.

97. Nelson, R. M., Dolich, S., Aruffo, A., Cecconi, O. & Bevilacqua, M. P. (1993) J
Clin Invest 91, 1157-1166.

98. Sanders, W. J., Katsumoto, T. R., Bertozzi, C. R., Rosen, S. D. & Kiessling, L. L.
(1996) Biochemistry 35, 14862-14867.

99. Nicholson, M. W., Barclay, A. N., Singer, M. S., Rosen, S. D. & Merwe, A. V. D.
(1998) J Biol Chem 273, 749-755.

100. Mehta, P., Cummings, R. D. & McEver, R. P. (1998) J Cell Biochem S24A, X5-
210.

101. Alon, R., Hammer, D. A. & Springer, T. A. (1995) Nature 374, 539-542.

102. Alon, R., Chen, S., Puri, K. D., Finger, E. B. & Springer, T. A. (1997) J Cell Biol
138, 1169-1180.

103. Jung, U., Bullard, D. C., Tedder, T. F. & Ley, K. (1996) Am J Physiol 271, H2740-
H2747.


104. Kunkel, E. J. & Ley, K. (1996) Circ Res 79, 1196-1204.









105. Puri, K. D., Finger, E. B. & Springer, T. A. (1997) J Immunol 158, 405-413.

106. Clark, R. A., erickson, H. P. & Springer, T. A. (1997) J Cell Biol 137, 755-765.

107. Berlin, C., Bargatze, R. F., Campbell, J. J., Andrian, U. H. V., Szabo, M. C.,
Hasslen, S. R., Nelson, R. D., Berg, E. L., Erlandsen, S. L. & Butcher, E. C. (1995)
Cell 80, 413-422.

108. Clark, R. A., Alon, R. & Springer, T. A. (1996) J Cell Biol 134, 1075-1087.

109. Dengrendele, H. C., Estess, P., Picker, L. J. & Siegelman, M. H. (1996) J Exp Med
183, 1119-1130.

110. Chen, S., Alon, R., Fuhlbrigge, R. C. & Springer, T. A. (1997) Pro Natl Acad Sci
USA 94, 3172-3177.

111. Finger, E. B., Puri, K. D., Alon, R., Lawrence, M. B., Andrian, U. H. V. & Springer,
T. A. (1996) Nature 379, 266-269.

112. Lawrence, M. B., Kansas, G. S., Kunkel, E. J. & Ley, K. (1997) J Cell Biol 136,
717-727.

113. Taylor, A. D., Neelamegham, S., Hellums, J. D., Smith, C. W. & Simon, S. I.
(1996) Biophys J 71, 3488-3500.

114. Berg, E. L., Magnani, J., Warnock, R. A., Robinson, M. K. & Butcher, E. C. (1992)
Biochem Biophy Res Commun 184, 1048-1055.

115. Berg, E. L., Robinson, M. K., Mansson, O., Butcher, E. C. & Magnani, J. L. (1991)
J Biol Chem 266, 14869-14872.

116. Pouyani, T. & Seed, B. (1995) Cell 83, 333-343.

117. Sako, D., Comess, K. M., Barone, K. M., Camphausen, R. T., Cumming, D. A. &
Shaw, G. D. (1995) Cell 83, 323-331.

118. Wilkins, P. P., McEver, R. P. & Cummings, R. D. (1996) J Biol Chem 270, 22677-
22680.

119. Alon, R., Feizi, T., Yuen, C. T., Fuhlbrigge, R. C. & Springer, T. A. (1995) J
Immunol 154, 5356-5366.

120. Labow, M. A., Norton, C. R., Rumberger, J. M., Lombard-Gillooly, K. M., Shuster,
D. J., Hubbard, J., Bertko, R., Knaack, P. A., Terry, R. W., Harbison, M. L.,
Kontgen, F., Stewart, C. L., McIntyre, K. W., Will, P. C., Burns, D. K. & Wolitzky,
B. A. (1994) Immunity 1, 709-720.









121. Moore, K. L., Stults, N. L., Diaz, S., Smith, D. F., Cummings, R. D., Varki, A. &
McEver, R. P. (1992) J Cell Biol 118, 445-456.

122. Norman, K. E., Moore, K. L., McEver, R. P. & Ley, K. (1995) Blood 86, 4417-
4421.

123. Borges, E., Pendl, G., Eytner, R., Moll, T., Steegmaier, M., Zollner, O. &
Vestweber, D. (1997) Blood 90, 1934-1942.

124. Borges, E., Tietz, W., Steegmaier, M., Moll, T., Hallmann, R., Hamann, A. &
Vestweber, D. (1997) J Exp Med 185, 573-578.

125. Laszik, Z., Jansen, P. J., Cummings, R. D., Tedder, T. F., McEver, R. P. & Moore,
K. L. (1996) Blood 88, 3010-3021.

126. Ruoslahti, E. & Pierschbacher, M. D. (1987) Science 238, 491-497.

127. Willemsen, O. H., Snel, M. M. E., Werf, K. O. v. d., Grooth, B. G. d., Greve, J.,
Hinterdorfer, P., Gruber, H. J., Schindler, H., Kooyk, Y. v. & Figdor, C. G. (1998)
Biophys J 75, 2220-2228.

128. Fujisawa, N., Poole-Warren, L. A., Woodard, J. C., Bertram, C. D. & Schindhelm,
K. (1999) Biomaterials 20, 955-962.

129. Dubreiul, R. R. (1991) BioEssays 13, 219-226.

130. Schakenraad, J. M., Busscher, H. J., Wildevuur, C. R. & Arends, J. (1986) J
Biomed Mater Res 20, 773-84.

131. Bentley, K. L. & Klebe, R. J. (1985) J Biomed Mater Res 19, 757-769.

132. Curtis, A. & WIlkinson, C. (1997) Biomaterials 18, 1579-1583.

133. Schmidt, J. A. & Recum, A. F. v. (1992) Biomaterials 13, 1059-1069.

134. den Braber, E. T. d., Ruijter, J. E. d., Ginsel, L. A., Recum, A. F. v. & Jansen, J. A.
(1998) J Biomed Mater Res 40, 291.

135. Britland, S., Morgan, H., Wojiak-Stondart, B., Riehle, M., Curtis, A. & Wilkinson,
C. (1996) Exp Cell Res 228, 313-325.

136. den Braber, E. T., de Ruijter, J. E., Smits, H. T. J., Ginsel, L. A., von Recum, A. F.
& Jansen, J. A. (1995) J Biomed Mater Res 29, 511-518.

137. Britland, S., Clark, P., Connolly, P. & Moores, G. (1992) Exp Cell Res 198, 124-
129.

138. Xiao, Y. & Truskey, G. A. (1996) Biophys. J. 71, 2869-2884.









139. Walboomers, X. F., Croes, H. J. E., Ginsel, L. A. & Jansen, J. A. (1998)
Biomaterials 19, 1861-1868.

140. Potember, R. S., Matsuzawa, M. & Liesi, P. (1995) Synthetic Metals 71, 1997-
1999.

141. Healy, K. E., Thomas, C. H., Rezania, A., Kim, J. E., McKeown, P. J., Lom, B. &
Hockberger, P. E. (1996) Biomaterials 17, 195-208.

142. Rezania, A., Thomas, C. H. & Healy, K. E. (1997) Ann Biomed Eng 25, 190-203.

143. den Braber, E. T. d., Ruijter, J. E. d. & Jansen, J. A. (1997) J Biomed Mater Res 37,
539-547.

144. Recum, A. F. v. & Kooten, T. G. v. (1995) J Biomater Sci Polymer Edn 7, 181-198.

145. Walboomers, X. F., Croes, H. J. E., Ginsel, L. A. & Jansen, J. A. (1999) J Biomed
Mater Res 47, 204-212.

146. Walboomers, X. F., Monaghan, W., Curtis, A. S. G. & Jansen, J. A. (1999) J
Biomed Mater Res 46, 212-220.

147. Walboomers, X. F., Croes, H. J. E., Ginsel, L. A. & Jansen, J. A. (1998) J Biomed
Mater Res 42, 634-641.

148. Walboomers, X. F. & Jansen, J. A. (2000) Biomaterials 21, 629-636.















BIOGRAPHICAL SKETCH


In 1979, the author was born in Tianjin, a small industrial city of 10 million

people, in the People's Republic of China. His father and grandfather, both engineers,

named him "engineering achiever." They hoped, that as the first male child of both his

father's and mother's families, the author would be successful. For seven years, he lived

an insouciant life as the spoiled scion of two well-to-do families. In 1987, the author and

his mother immigrated to Gainesville, Florida, to join his father, who was a Ph.D.

candidate in Engineering Mechanics at the University of Florida.

Like a fly in amber, he was trapped in Gainesville by fate and circumstance for the

next 13 years. After graduating from the international baccalaureate program at Eastside

high school, the author enrolled at the University of Florida in the summer of 1996, and

majored in Materials Science and Engineering. In May of 1996, he began an illustrious

relationship with Dr. Anthony Brennan. Initially, the author worked on sol-gel organic-

inorganic hybrid composites with Tom Miller, then a Ph.D. candidate. Then, in 1998, the

author switched to working with the atomic force microscope, which is an integral part of

his thesis. In the summer of 1998, the author worked for Exxon Chemical, under the

supervision of Dr. Michael Zamora, a former student of Dr. Brennan.

After his experience at Exxon, the author decided to switch to biomedical

engineering, and started a combined BS/MS program. To finish the program in 4 years,









the author took a prodigious number of classes, including 10 in the fall of 1998. As part

of the biomedical engineering program, he engaged in clinical shadowing in the spring of

1999. While shadowing Dr. Keith Ozaki, a vascular surgeon, who graciously contributed

to the author's thesis, the author realized that he wanted to be a surgeon. So the author

took the MCAT, and applied to medical school. During the summer of 1999, with the

reluctant permission of Dr. Brennan, the author traveled and studied in Europe. Upon his

return, the author worked on his thesis while also interviewing for medical school. On

June 20th, 2000, the author will begin a two-year stint as a secondary school science

teacher in Ghana for the Peace Corps. Then in 2002, the author will matriculate at

Northwestern University Medical School.




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs