Biomaterials and the Foreign Body Reaction: Surface Chemistry Dependent Macrophage Adhesion, Fusion, Apoptosis, and Cytokine Production

Home , Foreign body reaction, Giant cell

BIOMATERIALS AND THE FOREIGN BODY REACTION: SURFACE CHEMISTRY DEPENDENT MACROPHAGE ADHESION, FUSION, APOPTOSIS, AND CYTOKINE PRODUCTION

JACQUELINE ANN JONES

Submitted in partial fulfillment of the requirements

For the degree of Doctor of Philosophy

Dissertation Advisor: James Morley Anderson, M.D., Ph.D.

Department of Biomedical Engineering

CASE WESTERN RESERVE UNIVERSITY

May, 2007 CASE WESTERN RESERVE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

We hereby approve the dissertation of

______

candidate for the Ph.D. degree *.

(signed)______(chair of the committee)

______

(date) ______

*We also certify that written approval has been obtained for any proprietary material contained therein.

iii Dedication

This work is dedicated to…

My ever loving and ever supportive parents:

My mother, Iris Quiñones Jones, who gave me the freedom to be and to dream, inspires me with her passion and courage, and taught me the true meaning of friendship.

My father, Glen Michael Jones, a natural-born teacher who taught me to be an inquisitive student of life, inspires me with his strength and perseverance, and gave me his kind, earnest heart that loves deeply and always strives to improve the lives of others.

And to my younger siblings:

Angie, the yin to my yang whose resoluteness and grace continually inspires me & Jeff, my comrade whose wit and charisma brings me joy …both of whom I am ever proud.

Thank you. I am truly blessed to know and love you.

iv Table of Contents

Chapter I: Introduction

Chapter II: Proteomic Analysis and Quantification of Cytokines and Chemokines from Biomaterial Surface-Adherent Macrophages and Foreign Body Giant Cells

Chapter III: Matrix Metalloproteinases and Their Inhibitors in the Foreign Body Reaction on Biomaterials

Chapter IV: The Effects of Hydrophobic, Hydrophilic, & Ionic Surface Chemistries on Cellular Behaviors

Chapter V: In vitro and in vivo Macrophage Behaviors on Surface Modified Polyurethanes

Chapter VI: Dynamic Systems Model of Cellular Interactions with Biomaterial Surfaces

Chapter VII: Conclusions

Appendix: Self-Assembled Monolayers as a Model Material System for Material-Dependant Analysis of Macrophage/FBGC Cellular Behaviors

v List of Tables

Table 1.1. Adhesive Nature of Common Biomaterials.

Table 1.2. Water Contact Angles of Clinically and Experimentally Utilized Polymers

Table 1.3. Important Cytokines Involved in Inflammation.

Table 1.4. Summary of Previous Findings of Macrophage-Derived Cytokine mRNA Expression.

Table 2.1. Surface Chemistries of the Photografted Polymers.

Table 2.2. Potential Differences in Cytokines/Chemokines Due to Material, Time, or IL-4 Addition as Analyzed by Cytokine Arrays

Table 3.1. Matrix Metalloproteinases and their Functions

Table 3.2. MMPs/TIMPs Detected in Cell Cultures at Day 7 using Antibody Arrays

Table 3.3. MMP/TIMP Concentrations in Cell Cultures with RGD Substrate Over Time (ng/mL)

Table 4.1. Comparison of Alternatively Activated and Classically Activated Macrophages

Table 5.1. Chemical Description of Polyurethane Materials

Table 5.2. Water Contact Angle Measurements on Polyurethane Materials

Table 5.3. Differential Cell Analysis of Cell Layers at Day 4

Table 6.1. Estimated Rate Constants Utilized in Model Simulations

Table 6.2. Effects of Material-Dependent Parameters on Cellular Behavior

Table 6.3. Sensitivity Ranges for Each Parameter

Table A.1. AFM Results for in-house SAM Surfaces Prior to and After Treatment

Table A.2. Mixed Monocyte and Lymphocyte Non-Adherent Cell Population

vi Table A.3.A. Commonly Used and Chosen Self-Assembled Monolayer Synthesis Techniques (Part I)

Table A.3.B. Commonly Used and Chosen Self-Assembled Monolayer Synthesis Techniques (Part II)

vii List of Figures

Figure 1.1. Biological Responses to an Implanted Biomaterial Device.

Figure 1.2. Relationship of Aqueous Contact Angles and Cell Adhesion. Adapted from reference [11]. Conditionally (non)adhesive surfaces will typically depend on the degree of hydration of the hydrophilic surface. Hence, hydrated PEG/PEO is conditionally (non)adhesive, while glass is adhesive. Aqueous adhesion o tension is defined as τ = γlvcosθ(dyne/cm).

Figure 1.3. The Progression of Monocytes, Macrophages, and Foreign Body Giant Cells.

Figure 1.4. Cascade of Events Involved in Apoptosis.

Figure 1.5. The tissue/implant interface with protein adsorption, macrophage adhesion and activation, cytokine and growth factor production, and cellular synthesis and proliferation.

Figure 2.1. Cellular/Biomaterial/Cytokine Interactions of Inflammation and Wound Healing.

Figure 2.2. Adherent Cell Densities (A) and FBGC Formation (B). Data represents the mean ± the standard error of the mean (n=3). “*” indicates statistical significance in comparison to PET (p<0.05).

Figure 2.3. IL-1β (A), IL-6 (B), and IL-10 (C) Cytokine Production from Macrophages and FBGCs Adherent to Photograft Polymers. Mean ± SEM, n=4.

Figure 2.4. Effect of IL-4 on IL-10 (A&B), IL-8 (C&D), and MIP-1β (E&F) Production at days 7 and 10. Closed: without IL-4. Open: with added IL-4. Mean ± SEM, n=4.

Figure 2.5. Chemokine IL-8 (A) and MIP-1β (B) Expression over Time. “*” indicates a statistical difference in comparison to all other materials, while “^” indicates a statistical difference in comparison to PET. Mean ± SEM, n=4.

Figure 2.6. Cellular activation as a Function of IL-1β (A), IL-6 (B) and IL-10 (C) Production per Cell. Mean ± SEM, n=3.

Figure 2.7. IL-8 (A) and MIP-1β (B) Production per Cell. Mean ± SEM, n=3.

viii Figure 3.1. Macrophage/FBGC Secretion of Matrix Metalloproteinase-9 (A) and Tissue Inhibitors of Matrix Metalloproteinases, TIMP-1 (B) and TIMP-2 (C). Mean ± SEM, n=4.

Figure 3.2. Effect of IL-4 on MMP-9 (A&B), TIMP-1 (C&D), TIMP-2 (E&F) Production at Days 7 and 10. Open: with added IL-4. Closed: without IL-4. “*” indicates statistical difference between with and without IL-4. Mean ± SEM, n=4. Figure 3.3. MMP-9 (A), TIMP-1 (B), and TIMP-2 (C) Production per Cell. Mean ± SEM, n=3.

Figure 3.4. The Effects of MMP Pharmalogical Inhibitors ECG (A), NNGH (B), CL-82198 (C), and Actinonin (D) on Macrophage Adhesion (■) and Fusion (●). Mean ± SEM, n=3.

Figure 4.1. Disparate Effects of Hydrophobic and Hydrophilic Surfaces on Macrophage Adhesion (A) and Fusion (B). Mean ± SEM, n=3. “*” indicates that these values for the hydrophilic surface are statistically less than the hydrophobic surface values (p<0.05).

Figure 4.2. Disparate Effects of Hydrophilic Non-Ionic and Ionic Surfaces on Macrophage Adhesion (A) and Fusion (B). Mean ± SEM, n=3. “*” indicates that these values for the hydrophilic/ionic surfaces are statistically greater than the hydrophilic/neutral surface values (p<0.05).

Figure 4.3. Direct Comparison of Cellular Adhesion (A), Cytokine Concentration (B), and Cellular Activation (C) on Hydrophobic and Hydrophilic/Neutral Surfaces. Cellular activation equals cytokine concentration measured in 1mL of media normalized to the total number of adherent cells. Mean ± SEM, n=3,4. “*” indicates that these values for the hydrophilic surface are statistically less than the hydrophobic surface values (p<0.05).

Figure 4.4. Cytokine/Chemokine Concentration (A-D) and Cellular Activation (E-H) on Hydrophobic and Hydrophilic/Neutral Surfaces. Cellular activation equals cytokine/chemokine concentration measured in 1mL of media normalized to the total number of adherent cells. Notated numbers indicate the x fold increase between the values for BDEDTC and PAAm. Mean ± SEM, n=3. “*” indicates that values statistically greater on the hydrophilic surface compared to the hydrophobic surface values (p<0.05).

Figure 4.5. MMP/TIMP Concentration (A-C) and Cellular Activation (D-F) on Hydrophobic and Hydrophilic/Neutral Surfaces. Cellular activation equals cytokine concentration measured in 1mL of media

ix normalized to the total number of adherent cells. Notated numbers indicate the x fold change between the values for BDEDTC and PAAm. Mean ± SEM, n=3. “*” indicates that values statistically greater on the hydrophilic surface compared to the hydrophobic surface values (p<0.05).

Figure 4.6. Cytokine/Chemokine Concentration (A-E) and Cellular Activation (E-J) on Hydrophilic/Neutral and Hydrophilic/Ionic Surfaces. Cellular activation equals cytokine/chemokine concentration measured in 1mL of media normalized to the total number of adherent cells. Notated numbers indicate the x fold increase between the values for PAAm and PAANa or DMAPAAmMeI. Mean ± SEM, n=3. “*” indicates that values statistically greater on the hydrophilic/neutral surface compared to the hydrophilic/ionic surface values (p<0.05).

Figure 4.7. MMP/TIMP Concentration (A-C) and Cellular Activation (D-F) on Hydrophilic/Neutral and Hydrophilic/Ionic Surfaces. Cellular activation equals cytokine/chemokine concentration measured in 1mL of media normalized to the total number of adherent cells. Notated numbers indicate the x fold increase between the values for PAAm and PAANa or DMAPAAmMeI. Mean ± SEM, n=3. “*” indicates that values statistically greater on the hydrophilic/neutral surface compared to the hydrophilic/ionic surface values (p<0.05).

Figure 5.1. Monocyte/Macrophage Adhesion on Modified and Unmodified Polyurethane Materials, In Vitro. Mean ± SEM, n=3.

Figure 5.2. Monocyte/Macrophage Adhesion on Silicone Modified and Unmodified Polyurethane Materials, In Vivo. Mean ± SEM, n=4.

Figure 5.3. Foreign Body Giant Cell Formation on Modified and Unmodified Polyurethane Materials at Days 7 (A) and 10 (B), In Vitro. Surfaces with missing datapoints (i.e. E80A, E80A-F, E80A-P at day 7) had less than 1 percentage of fused cells on the surface. Mean ± SEM, n=3.

Figure 5.4. Optical Micrographs of Adherent Macrophages and IL-4 Induced FBGCs on PDMS and the Silicone Modified and Unmodified Polyurethanes at Day 10, In Vitro. Adherent cells are shown on the surfaces of Elasthane 80A (a,b); PurSil 20 80A (c,d); Bionate 80A (e,f); CarboSil 20 90A (g,h); and PDMS (i,j). Scale bars represent 100μm.

Figure 5.5. FBGC Formation on Silicone Modified and Unmodified Polyurethane Materials, In Vivo. Surfaces with missing data

x points (i.e. E80A and B80A at day 4) had less than 1 percentage of fused cells on the surface. Mean ± SEM, n=3.

Figure 5.6. Annexin V Apoptosis Analysis of Cells Adherent to Unmodified and Modified Polyurethane Materials, In Vitro. Mean ± SEM, n=3.

Figure 5.7. Viable and Apoptotic Adherent Cell Densities on the Unmodified and Modified Polyurethane Materials, In Vitro. Mean ± SEM, n=3.

Figure 5.8. Annexin V Apoptosis Analysis of Cells Adherent to Silicone Modified and Unmodified Polyurethane Materials, In Vivo. Surfaces with missing datapoints (PDMS at days 7 and 21) are not available. Mean ± SEM, n=3.

Figure 6.1. Schematic of In Vitro Cell Culture System.

Figure 6.2. System Diagram.

Figure 6.3. Comparison of Experimental and Computed Percentages.

Figure 6.4. Additional Analysis of Adherent Cell Populations over Time.

Figure A.1. Self-Assembled Monolayer Structure. X= Terminal functional o group: CH3, COOH, or OH. α= molecular tilt, typically 30 .

Figure A.2. Advancing and Receding Contact Angles on SAM (in-house) Surfaces. Untreated, dry samples were examined at days 0 (A) and 40 (D) after synthesis. Treated samples were incubated in serum-free media at 37oC for 5 (B) and 18 (C) days post synthesis. Mean ± standard deviation, n=6,3. “#” indicates statistical difference between advancing and receding value sets and advancing and receding values for gold (p<0.05).

Figure A.3. Changes in Contact Angles on Methyl SAM (in-house) Surfaces within the 1st day in an Aqueous Environment. Samples were incubated in serum-free media at 37oC. Mean ± Standard deviation, n=3. “*” indicates contact angles for the –CH3 terminated surface that are statistically different than angles for the Au surface (p<0,05).

Figure A.4. Advancing and Receding Contact Angles on SAM (out-of-house) Surfaces. Mean ± Standard deviation, n=3. “*” Each of the material sets is statistically different than the other materials (p<0.05).

xi Figure A.5. AFM Images of the Gold Control (A,E) and the Methyl (B,F), Carboxyl (C,G), and Hydroxyl (D,H) SAM Surfaces Immediately following Synthesis (A-D) and After 2 hours of Treatment (E-H). Samples treated were incubated in serum-free media at 37oC for 2 hours.

Figure A.6. Cellular Adhesion on SAM Surfaces in-house (A) and out-of- house (B) over Time. Mean ± SEM, n=3. “^” indicates a significant decrease in adhesion occurs after this timepoint (p=0.002). “*” indicates that the values for these materials are statistically different from the methyl SAM surface (p=0.03). .

Figure A.7. Macrophage Fusion into FBGCs on SAM Surfaces in-house (A) and out-of-house (B) over Time. Mean ± SEM, n=3. “^” indicates that the values for this material are statistically greater than gold (p=0.04). “*” indicates that the values at this timepoint are significantly greater than at day 3 (p<0.05).

Figure A.8. Cellular Apoptosis on SAM Surfaces (in-house) over Time. Percent apoptosis is the percentage of total number of adherent nuclei that are apoptotic. Mean ± SEM, n=3.

Figure A.9. Cellular Detachment on SAM Surfaces (in-house) at each Timepoint. Mean ± SEM, n=3,4. “*” indicates that the values at 2 hours are significant greater than at subsequent timepoints (p<0.05).

Figure A.10. Cytokine (A-C) and Chemokine (D-E) Production from Cell Cultures on in-house SAM Surfaces. Mean ± SEM, n=3. “*” indicates that the values at this timepoint are statistically greater than at subsequent timepoints (p<0.05).

Figure A.11. MMP (A) and TIMP (B-C) Production from Cell Cultures on in- house SAM Surfaces. Mean ± SEM, n=3. “*” indicates values statistically greater than at day 3 (p<0.05).

xii Acknowledgements

This work would not be possible without…

My mentor, James M. Anderson, the godfather of my career Thank you for inspiring and cultivating the scientist in me! Words cannot express my gratitude and admiration.

My academic advisor, Roger E. Marchant, whose advice set the ball in motion,

Bonnie Lou Berry, the heart of our lab, a remarkable woman of diverse interests and talents that reminds me of the importance of balance,

Erica Colton, who not only made each experiment possible but also has my deepest admiration for her open heart and kindness to all,

Amy K. McNally, a sweet and patient soul who always challenges me to think outside of the box,

My committee members: James Anderson, Roger Marchant, Anne Hiltner, Horst von Recum, and Gerald Saidel, each of whom expanded my intellectual limits,

My partners in crime: Jasmine Patel, David Chang, Danielle Soranno, and Gabriela Voskerician who always know how to have fun procrastinating and equally enjoy discussing the big and little questions in life,

My cohorts through it all: Mahrokh Dadsetan, Abby Qin, Bill Brodbeck, Terry Collier, Analiz Rodrigues, Elizabeth Christenson, Mike Wiggins, Sara MacEwan, and Matt MacEwan, each of whom it has been a privileged to work with and know,

Finally, last but never least, my cheering section, my support system, my dear friends: Deanna, Samantha, Tony, Tonya, Amanda, Jeff, Jackie, Josh, Ray, and the ladies night crew without whom I would have lost my sanity years ago and laughed till I cried significantly less.

Thank you. Thank you all.

xiii List of Abbreviations

In alphabetical order AFM = atomic force microscopy B80A, B90A = Bionate 80/90A BD = 1, 4-butanediol BDEDTC = poly(styrene-co-benzyl N,N-diethyldithiocarbamate) CMC = carboxylmethyl cellulose CSil90A = Carbosil 90A CSLM = confocal scanning laser microscopy DMAPAAmMEI = methyl iodide of poly[3-(dimethylamino)propyl[acrylamide E80A = Elasthane 80A E80A-F = Elasthane 80A with fluorocarbon surface modifying endgroups E80A-P = Elasthane 80A with polyethylene oxide surface modifying endgroups ECG = epigallocatechin gallate ECM = extracellular matrix EGF = epidermal growth factor ENA = epithelial-derived neutrophil activating protein FBGC = foreign body giant cell FC=fluorocarbon FEP = fluorinated ethylene propylene GDNF = glial cell-derived neurotrophic factor GRO = growth related oncogene HGF = hepatocyte growth factor IGFBP = insulin-like growth factor binding protein IL = interleukin MCP = macrophage chemotactic protein MDC = macrophage-derived chemokine MDI = 4, 4’-methylene bisphenyl diisocyanate MIP = macrophage inflammatory protein MMP = matrix metalloproteinase OM = optical microscopy NAP = neutrophil activating protein NK = natural killer NNGH = N-Isobutyl-N-(4-methoxyphenylsulfonyl)-glycylhydroxamic acid PAAm = polyacrylamide PAANa = sodium salt of poly(acrylic acid) PARC = pulmonary and activation-regulated chemokine PBS = phosphate buffer saline PC = polycarbonate PDGF = platelet-derived growth factor PDMS = polydimethylsiloxane PE = polyethylene PEG = polyethylene glycol PEO = polyethylene oxide PET = polyethylene terephthalate

xiv PHECD = poly (1,6 hexyl 1, 2-ethyl carbonate) diol p-HEMA = poly(hydroxylethyl metacrylate) PMMA = poly(methyl methacrylate) PMN = polymorphonuclear cell PP = polypropylene PS = polystyrene PSil80A = PurSil 80A PTFE = polytetrafluoroethylene PTMO = polytetramethylene oxide PU = polyurethane PVC = polyvinylchloride RANTES = regulated upon activation, normal T-cell expressed, and presumably secreted RGD = lysine, glycine, aspartic acid amino acid sequence SAM = self-assembled monolayer SEM = standard error of the mean SFM = serum-free media SME = surface modifying endgroup TARC = thymus and activation regulated chemokine TCPS = tissue culture polystyrene TGF = transforming growth factor TIMP = Tissue inhibitor of matrix metalloproteinases TNF = Tumor necrosis factor XPS = X-ray photoscpectroscopy

xv Biomaterials and the Foreign Body Reaction: Surface Chemistry Dependent Macrophage Adhesion, Fusion, Apoptosis, and Cytokine Production

Abstract

Jacqueline Ann Jones

The foreign body reaction has proven to be a hindrance to the functionality of implanted biomedical devices. The ability to direct this reaction, inflammation, and wound healing via the material-dependent control of key cellular components, macrophages and foreign body giant cells (FBGCs), is vital to the development of future biomedical devices. This dissertation addresses the hypothesis that surface chemistry directs adherent macrophage/FBGC behavior.

Specifically, this research endeavored to elucidate the relationship of surface hydrophobicity, hydrophilicity, and ionic chemistry with macrophage adhesion, activation, fusion into FBGCs, apoptosis, and production of cytokines, chemokines, matrix metalloproteinases (MMPs), and tissue inhibitors of MMPs

(TIMPs) using model material systems. In addition, a dynamic mathematical model was developed in order to further understand and predict these relationships.

Utilizing an in vitro human monocyte culture system and surface-modified biomaterials displaying hydrophobic, hydrophilic, and/or ionic chemistries, it has been demonstrated that material surface chemistry influences macrophage adhesion and fusion ultimately directing the cytokines/chemokines/MMPs/TIMPs

xvi released from biomaterial-adherent macrophages/FBGCs. Hydrophilic/neutral surfaces significantly inhibited adhesion and fusion in comparison to the hydrophobic and hydrophilic/ionic surfaces. Adherent cells on these cell-limiting hydrophilic/neutral surfaces produced greater quantities of each protein analyzed in comparison to the adhesion supporting surfaces indicating an increased activation state in these cells. This finding directly contradicts previous dogma that cell activation correlates with cellular adhesion prompting additional analysis into this phenomenon. In addition, the cytokine/chemokine profiles produced by adherent cells at earlier timepoints shifted from a more classically activated state to an alternatively activated state at later timepoints suggesting that a phenotypic switch occurs in these biomaterial-adherent cells.

Subsequent analysis using surface-modified polyurethanes confirmed that

FBGC formation was promoted by hydrophobic chemistry modifications, in vitro and in vivo. Macrophage apoptosis was promoted at earlier timepoints in vivo, while fusion was promoted at later timepoints supporting the theory that macrophages fuse as a mechanism to escape apoptosis.

This pivotal study clearly presents evidence that material surface chemistry can differentially affect macrophage/FBGC adhesion, activation, fusion, and apoptosis and the cytokine/chemokine/MMP/TIMP profiles derived from activated macrophages/FBGCs adherent to biomaterial surfaces.

xvii Chapter I

Chapter I: Introduction

The following Ph.D. research dissertation addresses the global hypothesis

that the monocyte/macrophage is the major cellular component controlling the tissue/material inflammatory, foreign body, and wound healing responses, and that the behaviors of monocytes, macrophages and foreign body giant cells

(FBGC) are differentially affected by material surface properties. In particular, this dissertation focuses on understanding the modulation of monocyte/macrophage adhesion, macrophage fusion, FBGC formation, and macrophage apoptosis as a function of surface chemistry; specifically hydrophobicity, hydrophilicity, and charge. Research has shown that these

adherent cells release cytokines, chemokines, matrix metalloproteinases

(MMPs), and inhibitors of matrix metalloproteinases (TIMPs) in order to

orchestrate cellular infiltration, activation, and matrix formation thus directing the

inflammatory and wound healing responses. For this reason, cytokine/chemokine/MMP/TIMP production profiles were analyzed for a material

surface chemistry dependency. In addition, these cellular behaviors (adhesion,

fusion, apoptosis, and cytokine production) were mathematically modeled and

analyzed for predictive capabilities. A collective understanding of these material

chemistries and cellular interactions including the resulting cytokine/chemokine profiles is imperative for determining how material surface chemistry can direct the inflammatory, foreign body and the wound healing responses and ultimately

- 1 - Chapter I

be utilized to predict the optimum application of surfaces for biomedical material

applications.

Biological Responses to Implanted Biomaterials

Medical implant devices have been utilized for more than 40 years.

Currently, more than 80,000 brands and models of medical devices are being produced by over 20,000 companies worldwide. In 2001, 3,507 new medical device products received marketing from the FDA.1 It has been estimated that 8 to 10 percent of Americans (20-25 million people) currently have an implanted medical device.2 With all implanted medical devices a certain sequence of

biological events occurs following implantation: injury, inflammatory cell

infiltration, acute inflammation, chronic inflammation, granulation tissue

formation, the foreign body reaction, and fibrosis as depicted in Figure 1.1.3,4

Variations in the intensity and time duration of these processes can be dependent on the size, shape, chemical properties, and physical properties of the biomaterial, the physical dimensions and properties of the device, and the specific tissue, organ, or species where the device is implanted.

The biocompatibility of a device is characterized by the extent of the inflammatory reaction, which attempts to neutralize or wall off the implanted device.3,5,6 Acute inflammation is of short duration (minutes to days) and is

characterized by exudate formation and the infiltration of leukocytes

(predominantly neutrophils followed by monocytes and lymphocytes).4 These leukocytes infiltrate the area in response to injury from the local vasculature via

- 2 - Chapter I

Figure 1.1: Biological Responses to an Implanted Biomaterial Device.4,6

margination, adhesion, emigration, and chemotaxis. Once at the site of injury, the leukocytes begin a series of processes including phagocytosis and the release of extracellular products. Within a few days, chronic inflammation begins and can last for weeks to months to years. Chronic inflammation is characterized in the tissue by the presence of macrophages, monocytes, and lymphocytes with neovascularization and the production of connective tissue.3 By this point, the neutrophil population, the hallmark of acute inflammation, has decreased significantly and the monocyte population diminished by differentiating into macrophages or undergoing cell death. Leukocytes of the chronic inflammation

- 3 - Chapter I

phase continue to participate in the phagocytosis of the foreign object and

release extracellular products, including cytokines and chemokines, which can

cause cellular activation, infiltration, fusion, and/or apoptosis in the other cell

populations.

The healing response and it’s hallmark, granulation tissue formation, is

initiated by monocyte and macrophage production of extracellular matrix as well

as the proliferation of fibroblast and infiltration of vascular endothelial cells to the

implant site.3,4 It is can occur as early as 3-5 days after implantation and is

characterized by the proliferation of fibroblasts and new blood vessels as well as its pink, soft granular appearance. Fibroblasts proliferate and synthesize proteoglycans and collagen. Initially, proteoglycans predominate the extracellular matrix with collagen Type I; however later, fibroblasts synthesize collagen Type III that will predominate the extracellular matrix and form the fibrous capsule. During this response, macrophages can release MMPs, which breakdown extracellular matrix components, and their inhibitors, TIMPs, further directing the extracellular matrix remodeling process. Wound healing is dependent on the extent of injury or the defect created during implantation and can occur by primary union, the healing of clean, surgical incisions that have been joined together by surgical sutures, or by secondary union, which requires a large tissue defect to be filled often resulting in increased granulation tissue formation and fibrosis or scarring.

Independent of the granulation tissue, the foreign body reaction occurs on the biomaterial and in the surrounding tissue. This reaction consists of foreign

- 4 - Chapter I

body giant cells and components of granulation tissue (i.e. macrophages,

fibroblasts, capiliaries).3,4,7 The roughness of the surface can affect the

composition of the foreign body reaction. Smoother surfaces have a thin layer of

macrophages while rough surfaces (i.e. fabrics) contain and are surrounded by

both macrophages and FBGCs. The foreign body reaction can remain at the

tissue-implant interfaces. FBGCs have been found on implants retrieved 30

years following implantation and may persist for the lifetime of the device.

Generally, a fibrous capsule composed of dense, fibroconnective scar

tissue forms surrounding the biomaterial or implant isolating the foreign object and the foreign body reaction from the surrounding tissue. This is beneficial for instances in which the foreign object needs to be isolated from the surrounding tissue; however, this process can be detrimental to the function of a biomedical device. For example, the fibrous capsule increases the impedance of the host tissue requiring a greater electrical output from a biomedical device (i.e. cardiac pacemaker or defibrillator) in order to produce the desired effect. This in turn reduces the lifetime of the device battery requiring more frequent replacement procedures. In addition, the delivery of a drug from a biomedical device can also be inhibited or blocked by a formed fibrous capsule. Regardless of application, fibrous capsule formation or fibrosis is considered to be the end-stage of the healing response. Alternatively, regeneration of the injured tissue can occur with the replacement of the damaged tissue with parenchymal cells of the same type.

This occurrence of the process depends on the persistence of the natural

- 5 - Chapter I

framework of the implant site and the proliferative capacity of the cells resident to

the implant site (i.e. labile, stable, or permanent cells).

Biomaterial properties have been shown to affect stages of inflammation,

the foreign body reaction, and fibrosis.8,9 Numerous studies have investigated

material-dependent tissue/biomaterial interactions shown in Figure 1.1 including

leukocyte infiltration and adhesion, fibroblast adhesion and proliferation, and

fibrous encapsulation.4,9,10 In addition, research has shown that biomaterial

properties and behaviors of surrounding cells (i.e. lymphocytes) do affect the

behavior of biomaterial-adherent monocytes, macrophages, and FBGCs, which

are highly involved in four of the five above discussed responses.3,4 However,

specific material surface chemistry dependent mechanisms involved in

monocyte/macrophage adhesion, differentiation, activation, fusion, apoptosis, and cytokine/chemokine/MMP/TIMP production, which modulate these responses, has yet to be determined.

Hydrophobicity and Hydrophilicity of Biomaterials

This research aims to address the effects of hydrophobic, hydrophilic, and

ionic surface chemistries on cellular behaviors. It does so by utilizing material

surfaces displaying distinct surface chemistries to draw correlations between

these surface properties and the resulting cellular behaviors. Granted the

relative nature of these classifications, it is common practice to classify polymeric

materials as hydrophobic or hydrophilic based upon the observable water contact

angles related to the wettability of the material surface. Clinically-utilized

- 6 - Chapter I

polymers present surface chemistries ranging from very hydrophobic, for

example polytetrafluoroethylene (PTFE) and poly(dimethyl siloxane) (PDMS), to

very hydrophilic such as in the case of poly(hydroxyethyl methacrylate) (p-

HEMA) and polyethylene oxide (PEO).

A significant amount of research been conducted analyzing the effects of

hydrophobicity and hydrophilicity on biological responses (i.e. protein adsorption

and cellular adhesion). Previous research has shown that very hydrophobic

materials such PTFE with a contact angles of 105o-116o inhibit cellular adhesion.

In contrast, hydrophilic materials such as tissue culture polystyrene (TCPS) are known to support cellular adhesion. Other hydrophilic materials that present highly hydrated surfaces (i.e. PEO) inhibit cellular adhesion. Continued research has attempted to explain these discrepancies and has hypothesized that cellular inhibition seen on hydrophobic surfaces results from adsorbed proteins interacting with the hydrophobic surface in a manner that changes the protein conformation making ligand moieties (i.e. RGD) needed for integrin binding and cellular adhesion inaccessible. Additional research has put forth evidence that in the later hydrophilic case the relatively longer polymeric chains (i.e. PEO) attract a sufficient amount of water molecules to create a barrier to protein adsorption, which in turn inhibits cellular adhesion.

Naturally, controversy arises with the classification of a material surface as hydrophobic or hydrophilic due to the fact that many of the material surfaces have wide and oftentimes overlapping ranges of contact angles. For example,

Vogler defined biomaterials with contact angles greater than 65o (τ=30 dyne/cm)

- 7 - Chapter I

to be hydrophobic and more wettable surfaces with contact angles less than 65o

(τ=30 dyne cm) to be hydrophilic. A more appropriate definition would be to

assign materials with contact angles in the range of 65-80 as intermediate

wetting (see Table 1.1). Saltzman utilized Vogler’s definition and further

separated hydrophobic and hydrophilic surface chemistries based upon

empirically derived correlations between cell adhesion and water contact angles

as conditionally (non)adhesive (i.e. pHEMA and PEO), adhesive (i.e. TCPS, PU,

PMMA, and PET), and nonadhesive (i.e.PTFE, PE, PC, and PP) (Figure 1.2).11

Increased Nonadhesive protein Low protein binding adsorption (i.e. albumin); Denaturing of adhesive proteins

High energy Low energy

Hydrophilic surfaces Hydrophobic surfaces

Conditionally (non)adhesive Adhesive Nonadhesive

Material Rigidity: Water Content Cell Attachment Cell Attachment Relative Mammalian Relative Mammalian

0 10 20 30 40 50 60 70 80 90 100 110 Aqueous Contact Angle (degrees)

80 60 40 20 0 -20 -40 Aqueous Adhesion Tension

Figure 1.2: Relationship of Aqueous Contact Angles and Cell Adhesion. Adapted from reference [11]. Conditionally (non)adhesive surfaces will typically depend on the degree of hydration of the hydrophilic surface. Hence, hydrated PEG/PEO is conditionally (non)adhesive, while glass is adhesive. Aqueous o adhesion tension is defined as τ = γlvcosθ(dyne/cm).

- 8 - Chapter I

The polymer surfaces utilized in this study are classified as hydrophobic, intermediate wetting, or hydrophilic based upon their relative contact angles and are analyzed accordingly. As shown in Table 1.1, the biomaterial contact angles and classifications of the materials utilized in this dissertation correlate with the respective polymeric material types. Comparative studies of cellular behaviors with respect to these relative hydrophobic and hydrophilic surface chemistries can potentially further elucidate these relationships. Additional insight into these interactions can direct future research and ultimately be utilized to develop criteria for polymer design in future biomaterial applications.

Table 1.1 Adhesive Nature of Common Biomaterials Conditionally (non)adhesive Adhesive Nonadhesive

- Polyethylene glycol / - Poly(methyl - Polytetrafluoroethylene polyethylene oxide (PEG/PEO) methacrylate) (PMMA) (PTFE) - Poly(hydroxyethyl-methacrylate) - Polyethylene - Fluorinated ethylene (pHEMA) terephthalate (PET) propylene (FEP) - Carboxymethyl cellulose (CMC) - Tissue culture - Poly(propylene - Polyvinyl alcohol polystyrene (TCPS) fumarate) - Polyacrylamide - Polyurethanes - Polyethylene (PE) - Polyvinyl pyrrolidone - Clean glass (SiO2) - Polypropylene (PP) - Dextran - Polyvinylchloride - Teflons - Alginate - Polycarbonate (PC) - Hyaluronic acid - Polylactide - Cellulose - Nylon-66 - Ambient exposed metals - Different plasma treated materials - Polylactide - Collagen

Adapted from reference [11].

- 9 -

Table 1.2 Water Contact Angles of Clinically and Experimentally Utilized Polymers11,12 Material Relative Contact Relative Hydrophobicity Angles (o) Cell Attachment Clinically Used Polymers* PFTE - Polytetrafluoroethylene Very Hydrophobic 105-116 Nonadhesive PDMS – Poly(dimethyl siloxane) Very Hydrophobic 100-104 Nonadhesive PE – Polyethylene Hydrophobic 92-97 Nonadhesive PP – Polypropylene Hydrophobic 92-97 Nonadhesive PS – Polystyrene Hydrophobic 75-91 Nonadhesive PC – Polycarbonates Hydrophobic 95 Adhesive PMMA – Poly(methyl methacrylate) Intermediate 65-80 Adhesive PU – Polyurethane Intermediate 60 Adhesive PET – Poly(ethylene terephthalate) Intermediate 65-81 Adhesive TCPS – Tissue culture polystyrene Hydrophilic 35-68 Adhesive

- 10 p-HEMA – Poly(hydroxyethyl methacrylate) Hydrophilic 60-80 Conditionally (non)adhesive PEO – Poly(ethylene oxide Hydrophilic 20-45 Conditionally (non)adhesive Photografted Polymers^ PET – Poly(ethylene terephthalate) Intermediate 71 Adhesive BDEDTC - Modified poly(ethylene terephthalate) Intermediate 68 Adhesive PAAm- Modified poly(ethylene terephthalate) Hydrophilic 46 Conditionally (non)adhesive PAANa - Modified poly(ethylene terephthalate) Hydrophilic 24 Conditionally (non)adhesive DMAPAAmMeI - Modified poly(ethylene terephthalate) Hydrophilic 31 Conditionally (non)adhesive Polyurethanes^ Elasthane 80A – Polyether urethane Intermediate 61 Adhesive Bionate 80A – Polycarbonate urethane Intermediate 62 Adhesive PurSil 20 80A – PDMS modified polyether urethane Hydrophobic 90 Adhesive CarboSil 20 90A – PDMS modified polycarbonate urethane Hydrophobic 92 Adhesive Elasthane 80A-F – Fluorocarbon modified polyether urethane Very Hydrophobic 109 Nonadhesive Elasthane 80A-P – Polyethylene oxide modified polyether urethane Intermediate 62 Adhesive * Adapted from references [11,12]. ^ See Chapters II and V for detailed description of the materials and contact angle analysis.

Chapter I

Material-Dependent Macrophage Adhesion

During acute and chronic inflammation, monocytes migrate to the injury/implant site as a result of chemotaxis and undergo a series of events depicted in Figure 1.3.6 These monocytes may adhere to the biomaterial.

Adherent monocytes may then become activated and differentiate into adherent macrophages, detach from the surface, or undergo apoptosis (programmed cell death). In vivo, a monocyte is considered to be a tissue macrophage upon leaving the vasculature. In vitro, extravasation does not occur; therefore the blood-isolated cells are referred to as monocytes until differentiation into macrophages occurs. Morphologically, adherent macrophages differ from small,

Monocyte Macrophage Foreign Body Giant Cell Blood Tissue Tissue/Biomaterial Biomaterial

Activity Chemotaxis Chemotaxis Adhesion Phenotypic Migration Migration Differentiation Expression Adhesion Signal Transduction Differentiation Activation

Figure 1.3: The Progression of Monocytes, Macrophages, and Foreign Body Giant Cells.4

rounded adherent monocytes by exhibiting a greater cytoplasmic to nuclear area ratio and a more spread morphology. Adherent macrophages release degradative enzymes, free radicals, and reactive oxygen species in an attempt to degrade and phagocytose the biomaterial. Due to the inherent size disparity, the macrophages are ineffective at removing or destroying the biomaterial implant

- 11 - Chapter I

and begin to fuse to form FBGCs in response to this coined “frustrated

phagocytosis”. Foreign body giant cells, which are defined as having three or

more nuclei, further release these degradative agents in an attempt to degrade or

phagocytose the biomaterial and continue to remain at the implant site years

later. With many polymers, such as polyurethanes, these degradative agents

can induce chain scission and degrade the polymer over time.13-16 Previous

studies have shown significant surface pitting and cracking of a polyurethane film

upon removal of adherent macrophages and FBGCs.13 Surface pitting and

cracking may propagate to full thickness cracking resulting in the failure of the device. An example of this is seen in cardiac pacemaker leads, where full thickness cracking of the polyurethane insulation caused the pacemaker lead circuitry to short and the lead to fail resulting in an inactive pacemaker and requiring a new lead replacement.17,18 In addition, both macrophages and

FBGCs can secrete a variety of cytokines, chemokines, MMPs, and TIMPs that

can modulate the surrounding environment. This will be discussed later.

Upon implantation of a biomaterial, protein adsorption occurs immediately

in accordance with the Vroman effect.19 Monocytes and macrophages recognize

these adsorbed proteins on the biomaterial surface via their integrin and protein

receptors. As a result, macrophage adhesion can occur via several adhesive

receptor-ligand interactions. The adsorption of numerous serum proteins (i.e.

complement 3b, von Willebrand factor, fibronectin, IgG, thrombospondin,

vitronectin, albumin, and a2 macroglobulin) has been shown to be material

dependent.20-28 The biomaterial chemistry and resulting type, amount, clustering,

- 12 - Chapter I and conformation of adsorbed proteins all play a role in mediating macrophage adhesion.20,24,29 Numerous studies are attempting to mediate cellular adhesion via modifications in materials that modulate various protein adsorption profiles and conformations. Examples of material modifications include lysine-derivatized polyurethane surfaces intended to increase plasminogen adsorption, polyurethane surfaces modified with amphiphilic polymers or polyethylene glycol

(PEG) to decrease protein adsorption, and the fluorination of polyurethanes to decrease protein adsorption.23-25,30

Integrin receptors on macrophages are known to bind to adsorbed blood proteins and extracellular matrix proteins (i.e. fibronectin, vitronectin, von

Willebrand factor, albumin, plasminogen) mediating macrophage adhesion and migration. Specifically, β2 integrins have been shown to mediate initial monocyte adhesion, while β1 integrins are more significant at later timepoints when macrophage fusion is occurring.31 Integrin involved adhesion, specifically, β2 integrins has also been shown to be a key mechanism for migration of leukocytes on a biomaterial.32 Another ligand-receptor interaction that macrophages utilize to adhere involves complement receptors on the macrophage extracellular membrane interacting with adsorbed complement proteins and/or IgG or IgM antibodies resulting from opsonization.28 Studies have shown that macrophage adhesion on various biomaterials (i.e. polyether urethane and fluorinated, siliconized, nitrogenized and oxygenated polystyrene surfaces) decreased from

50-100% in media that has been depleted of complement component C3, indicating that C3 is essential to macrophage adhesion. Also, it has been

- 13 - Chapter I

suggested that the β2 integrin may be utilized as the C3 receptor for adhesion.33

Research is currently being conducted in our laboratory to further understand material effects on integrin-mediated macrophage and FBGC adhesion.

IL-4-induced FBGC Formation and Cytoskeletal Arrangement

Multinucleated cells, known as foreign body giant cells (FBGCs), are a hallmark of chronic inflammation and have been found at the tissue/biomaterial interface year following implantation. As shown in Figure 1.3, these multinucleated cells are a result of macrophage fusion and can contain at minimum three nuclei up to a hundred nuclei.6 Previous studies have shown that macrophage fusion into FBGCs is induced by the lymphocyte-derived cytokine

interleukin -4 (IL-4) and that FBGC formation can be inhibited by IL-4 antibody

inhibition.34,35 IL-13, which is also secreted by TH2 lymphocytes, has been

shown to induce FBGC formation in a non-synergistic fashion with IL-4.36 IL-4 and IL-13 induced fusion greatly up-regulates mannose receptor activity that is believed to mediate the endocytosis of glycoproteins and phagocytosis of microorganisms.34,36-38 Conversely, inhibitors of mannose receptor have been

shown to prevent fusion in vitro, suggesting that the mannose receptors has an

essential role in the mechanism of IL-4 and IL-13 induced macrophage

fusion.38,39 In addition, both IL-4 and IL-13 are down-regulators of several pro- inflammatory cytokines (i.e. IL-1, tumor necrosis factor-α, and IL-6) and chemokines (i.e. IL-8 & macrophage inflammatory protein – 1α).40-44

- 14 - Chapter I

The cellular cytoskeleton regulates many critical functions of

monocytes/macrophages/FBGCs during the inflammatory response, including

adhesion, spreading, migration, phagocytosis and secretion. For this reason,

cytoskeletal arrangement of adherent monocytes, macrophages, and FBGCs has

been investigated. Immunohistochemistry and confocal scanning laser

microscopy revealed that monocytes posses a diffuse cytoplasmic staining of

adhesive structural proteins, macrophages contain distinct punctate filamentous

F-actin structures along the ventral cell membrane identified as the core of the

podosome adhesive structure, and FBGCs contain the punctate, filamentous F-

actin of the podosome at the extreme periphery of the ventral cell surface.45

Disrupting the filamentous actin has also been shown to inhibit macrophage fusion.46 It is important to understand the adhesive interactions between a cell

and biomaterial because these interactions may also rearrange the cytoskeletal

microfilaments and may cause a redistribution of surface receptors associated

with the cytoskeleton.

Apoptosis and Inflammation

The presence or activity of these adherent cells can be minimized by

designing biomaterials that promote apoptosis or programmed cell death.

Apoptosis refers to cell death that is programmed versus accidental as in

necrosis. These processes differ in their means of induction and ability to induce

an inflammatory response. Necrosis is the result of a chemical, physical, or

biological damage, while apoptosis requires active gene expression and is

- 15 - Chapter I

medicated by specific receptors and signal transduction.47 In the case of

biomaterial interactions, apoptosis can be initiated by specific receptor/ligand

interactions (i.e. TNF-α binding to TNF-RI), the presence (transforming growth

factor, TGF-β) or absence (colony stimulating factor, CSF) of specific growth hormones, and/or the disruption of cell-cell or cell-matrix interactions (i.e. disrupted integrin or cadherin ligand interactions). Apoptotic cells undergo a cascade of events that ultimately lead to the formation of sealed cell fragments or apoptotic bodies as depicted in Figure 1.4. This method of cell death is beneficial in that these sealed apoptotic bodies do not release the cell’s intracellular contents the can induce inflammation.

- 16 - Chapter I

Figure 1.4: Cascade of Events Involved in Apoptosis.47

The apoptosis cascade is complex sequence of highly regulated events

that is still not fully understood. In brief, initiation occurs via the binding of TNF-α to TNF-RI or Fas-ligand to its Fas receptor or one of the other above mentioned mechanisms.47 A signaling complex is formed and a sequence of activation of

initiator caspases occurs. These initiator caspases cause the first proteolytic

events (cytoskeletal protein cleavage). The mechanisms that maintain the

- 17 - Chapter I

location of phosphotidyl serine in the inner leaflet of the cell membrane are

destroyed via cleavage of the translocase or activation of the scramblase.

Numerous studies utilize the presence of phosphotidyl serine in the outer cell

membrane as a marker for early stage apoptosis.47-49 In addition, the initiator

caspases irreversibly activate the executor caspases in the apoptosis cascade.

Once this occurs, apoptosis is inevitable for these executor caspases cleave a

variety of proteins essential for cell survival and activate factors that cleave DNA.

Late stage apoptosis can be identified via the detection of specific DNA

fragmentation. Morphological features of apoptotic cells include cell shrinkage, condensed chromatin that is packed in close association with the nuclear

membrane, blebbing, and the presence of apoptotic bodies.47,49

The numbers of biomaterial-adherent macrophages decrease significant

by day 7 in vitro. In addition to cell detachment, apoptosis has been shown to be a key mechanism involved with minimized macrophage adhesion. A study of neutrophil apoptosis suggests that engagement of β2 integrins with its ligands in

the presence of pro-apoptotic stimuli (i.e. TNF-α) may modulate neutrophil

apoptosis.50 McNally et. al. showed that β2 integrins are prevalent in adherent

monocytes/macrophages at earlier in vitro timepoints, while β1 integrins are more prevailing at later timepoints when macrophage fusion is occurring31. It has been

demonstrated that macrophage apoptosis and fusion are inversely related and

therefore has been suggested that macrophage fusion into foreign body giant

cells may be an escape mechanism for cell apoptosis.49

- 18 - Chapter I

Macrophage-Derived Cytokines, Chemokines, MMPs, and TIMPs Involved in Inflammation, the Foreign Body Reaction, and Wound Healing

The macrophage has been suggested to be a major conductor cell of

inflammation and wound healing due to its involvement in acute and chronic

inflammation and fibrosis and its ability to produce multiple cytokines/chemokines

that direct the cellular activities of other inflammatory and wound healing. Figure

1.5 depicts some of the potential cytokine and growth factor interactions between macrophages and other inflammatory and wound healing as hypothesized by Dr.

James Anderson in 1984.3,5 These cytokines can have functions in chemotaxis,

cell activation, surface proteins/receptor up-regulation, fusion, protein production, and surface proteins/receptor up-regulation, fusion, protein production, and apoptosis amongst other processes. The particular functions of these cytokines

Figure 1.5: The tissue/implant interface with protein adsorption, macrophage adhesion and activation, cytokine and growth factor production, and cellular synthesis and proliferation.3,5

- 19 - Chapter I

of interest involved in inflammation will be discussed in Chapter II and are listed

in Table 1.3. Significant research has been conducted to investigate cytokine

production from adherent monocytes, macrophages, and FBGCs. Some studies

have been concerned with whether or not a particular surface activates adherent cells. Many of these investigations have utilized the production of IL-6, IL-8, and

TNF-α as indicators of monocyte/macrophage activation.51-53 Other studies

investigated whether or not a particular cytokine is involved in the functions of the

adherent cell.35-44,54 Recent studies in our laboratory, by Brodbeck et.al., have begun to study the surface chemistry dependent nature of macrophage cytokine expression via an analysis of macrophage-derived mRNA expression of various cytokines both in vitro and in vivo.55,56 Theses findings are summarized in Table

1.4. A genomic investigation of cytokine production indicates the potentially

produced cytokine profiles. The next step in understanding the surface chemistry

effects on cytokine/chemokine production is a proteomic investigation that

provides insight into the specific cytokine/chemokine concentrations produced in

cell culture.

Matrix metalloproteinases (MMP) are a family of over 25 zinc-dependent

proteolytic enzymes capable of degrading the structural protein components

within extracellular matrix (ECM) and at the cellular surface.57 MMPs are actively

secreted by adherent macrophages and FBGCs amongst other cells and can

modulate numerous biological processes including wound healing, the foreign

body reaction, inflammation, angiogenesis, and fibrous capsule formation.58-60

- 20 - Chapter I

Table 1.3 Important Cytokines Involved in Inflammation61 Cell Cell Types Cytokine Source Receptor Affected Function of Interest IL-1β Monocytes IL-1R1 Numerous -Modulates production of cytokines, Macrophages extracellular matrix proteins, Dendritic cells receptors B-lymphocytes IL-6 Monocytes IL-6R with B-lymphocytes -Stimulates the proliferation of B- Macrophages chains Neutrophils lymphocytes. Fibroblasts α and β T-lymphocytes -Induces immunoglobulin production Endothelial cells Monocytes -Increases neutrophil production. T-lymphocytes Macrophages -Involved in T-cell activation, growth, B-lymphocytes and differentiation. Eosinophils -Induces the proliferation of murine Mast Cells pluripotent hematopoietic progenitors. -Regulates the survival, proliferation, and differentiation of mononuclear phagocytes. MIP-1β Monocytes CCR5 Monocytes -Monokine with inflammatory and Macrophages CCR8 Lymphocytes chemokinetic properties. T-lymphocytes Thymocytes -Chemoattractant activity for B-lymphocytes Dendritic Cells monocytes, CD4+ lymphocytes, NK cells thymocytes, and dendritic cells. Neutrophils Eosinophils Basophils Mast cells Endothelial cells Fibroblasts IL-8 Monocytes CXCR1 Neutrophils -Chemotactic factor that attracts Lymphocytes CXCR2 Basophils neutrophils, basophils, Dendritic Fibroblasts T-lymphocytes cells, and T-cells, but not Endothelial cells Eosinophils monocytes. -Activates neutrophil. -Induces angiogenesis. TIMP-1 Monocytes - - - Complexes with Macrophages Acts on MMPs metalloproteinases Fibroblasts (such as collagenases) and irreversibly inactivates them. -Known to act on MMP-1, 2, 3, 7, 8, 9, 10, 11, 12, 13, & 16. Does not act on MMP-14. TIMP-2 Monocytes - - -Complexes with metalloproteinases Macrophages Acts on MMPs (such as collagenases) and Fibroblasts irreversibly inactivates them. -Known to act on MMP-1, 2, 3, 7, 8, 9, 10, 13, 14, 15, 16 & 19. MMP-9 Monocytes - - -Cleaves most if not all of the Macrophages ECM proteins constituents of the extracellular Fibroblasts matrix.

- 21 - Chapter I

Cell Cell Types Cytokine Source Receptor Affected Function of Interest IL-10 Monocytes IL-10Rα Monocytes -Modulates cytokine production. T-lymphocytes IL-10Rβ Macrophages -Inhibits production of IL-1α, IL-1β, B-lymphocytes Neutrophils IL-6, IL-8, IL-10, IL-12, GM-CSF, Eosinophils G-CSF, M-CSF, TNF-α, MIP-1α, Basophils MIP-1β, MIP-1, RANTES and LIF B-lymphocytes by activated T-lymphocytes monocytes/macrophages. -Enhances HCC4, CCR5, and IL- 1Ra. -Inhibits MHC class II antigens, ICAM-1, CD80, and CD86 on monocytes. -Enhances expression of HLA-G on trophoblasts and monocytes. -Inhibits NO production by macrophages. -Inhibits monocyte/macrophage ability to modulate turnover of ECM by inhibiting gelatinase and collegenase production and enhances TIMP production. -Enhances survival of resting B-cells and proliferation of B-cells. -Inhibits T-cell proliferation, apoptosis, and cytokine production. IL-4 T-lymphocytes IL-4R type I Macrophages -Induces macrophage fusion. Mast cells with IL-4Rγ B-lymphocytes -Participates in at least several B- Eosinophils IL-4R type T-lymphocytes cell activation processes. II with IL- -Induces the expression of class II 13Rα MHC molecules on resting B-cells. - Enhances both secretion and cell surface expression of IgE and IgG1. -Regulates the expression of the low affinity Fc receptor for IgE (CD23) on both lymphocytes and monocytes. -Regulates help T-cell differentiation into TH2 type. IL-13 T-lymphocytes IL-13Rα1 Macrophages -Promotes macrophage fusion. Mast cells IL-13Rα2 B-lymphocytes -Synergizes with IL2 in regulating NK cells IL-4Rα T-lymphocytes interferon-γ synthesis. -Regulates B-cell IgE secretion. -Modulates Th2 cell development IL-2 T-lymphocytes IL-2R T-lymphocytes -Required for T-cell proliferation. B-lymphocytes -Can stimulate B-cells, monocytes, Monocytes lymphokine-activated killer cells, NK cell and natural killer cells. PARC unknown unknown T-lymphocytes -Chemoattractant for T-cells not monocytes and granulocytes.

- 22 - Chapter I

Cell Cell Types Cytokine Source Receptor Affected Function of Interest TGF-β Numerous TGF-βR Endothelial cells -Multifunctional peptide that controls Fibroblast proliferation, differentiation, and Macrophages other functions in many cell types. Monocytes -Inhibits endothelial cell proliferation. T-lymphocytes -Increases synthesis of matrix B-lymphocytes proteins. Neutrophils -Decreases synthesis of proteases. NK cells -Increases synthesis of protease inhibitors. - Increases synthesis of cell adhesion receptors. -Blocks antibody production by B- cells. -Inhibits generation of lymphokine activated killer cells and cytotoxic T-cells. - Inhibits respiratory burst of macrophages and apoptosis of T- Cells. -Chemotactic for monocytes, lymphocytes, neutrophils, and fibroblasts. -Induces expression of IL-2, IL-2R, IL-10, and IFN-γ in T-cells. -Induces mRNA expression of IL-1α, IL-1β, TNF-α, PDGF-BB, and bFGF in resting human blood monocytes. -Autoinduction of expression of cytokines in fibroblasts and monocytes. TNF-α Macrophages p60 Endothelial cells -Stimulate cell proliferation and Monocytes p80 Macrophages induce cell differentiation. T-lymphocytes Fibroblasts -Induce cell death of some tumor B-lymphocytes Neutrophils cell lines. Fibroblasts Lymphocytes -Induces NO synthase, IL1, G-CSF, NK cells and IL-3R production in Mast cells endothelial cells. Basophils -Induces endothelial cell permeability to albumin. -Induces P-selectin, ICAM-1, and VCAM-1 in endothelial cells. -Suppresses endothelial cell proliferation. -Induces macrophage expression of MHC antigens, IL-1 and GN-CSF -Induces fibroblast proliferation and production of IL-6, IL-1, and MMPs -Suppresses fibroblast production of TIMPs and collagen synthesis -Activates neutrophils and superoxide anion generation.

- 23 - Chapter I

Cell Cell Types Cytokine Source Receptor Affected Function of Interest MDC Macrophages CCR4 T-lymphocytes -Non-catalytic metalloprotease-like Monocytes NK cells protein. B-lymphocytes Dendritic cells -Chemoattractant for T-cells, NK T-lymphocytes Monocytes cells, Dendritic cells, and monocytes. MCP-1 Fibroblasts CCR2 Monocytes -Chemotactic factor that attracts Endothelial cells Basophils monocytes, basophils, T-cells, NK Monocytes T-lymphocytes cells but not neutrophils. Macrophages NK cells Neutrophils NAP-2 Megakaryocytes CXCR1 Neutrophils -Chemoattractants and activators for Platelets CXCR2 Megakaryocytes neutrophils. Neutrophils Fibroblasts -Inhibits megakaryocytopoiesis. Monocytes Mast cells -Chemoattractant for fibroblasts and T-lymphocytes Lymphocytes increase synthesis of matrix components. -Chemoattractant for mast cells. -Inhibits lymphocytes cytolytic activity RANTES Macrophages CCR1 Monocytes -Chemoattractant for blood Endothelial cells CCR3 T-lymphocytes monocytes, memory T- helper Platelets CCR4 Eosinophils cells and Eosinophils, basophils, Eosinophils CCR5 Basophils dendritic cells. Dendritic cells -Causes the release of histamine from basophils. -Activates T-cells. PDGF-BB Platelets PDGF-Rαα Neutrophils -Stimulate growth in other wound Fibroblasts PDGF-Rαβ Monocytes healing cells (i.e. neutrophils and Endothelial cells PDGF-Rββ Fibroblasts monocytes). Monocytes -Stimulates the migration of Macrophages fibroblasts. EGF - EGF-R Epidermal cells -Stimulates the growth of various Epithelial cells epidermal and epithelial tissues in Fibroblasts vivo and in vitro and of some fibroblasts in cell culture. Eotaxin-2 unknown CCR3 Eosinophils -Chemotactic for resting T- CCL24 Basophils lymphocytes, basophils, and T-lymphocytes eosinophils. Dendritic cells -Has lower chemotactic activity for neutrophils but none for monocytes. -Activates lymphocytes. GRO Monocytes CXCR1 Neutrophils -Involved in activation of neutrophils, Macrophages CXCR2 Lymphocytes lymphocytes, and monocytes. Endothelial cells Monocytes -Chemoattractant for neutrophils, Neutrophils Basophils basophils, eosinophils, Lymphocytes Eosinophils monocytes, and lymphocytes. Fibroblasts ENA-78 Fibroblasts CXCR1 Neutrophils -Involved in neutrophil activation. Macrophages -Chemoattractant for neutrophils. Monocytes -Promotes neutrophil adhesion Neutrophils

- 24 - Chapter I

These proteolytic enzymes have multiple functions including breaking physical

ECM barriers allowing for cellular migration, proliferation, and matrix remodeling,

regulating the activities of cytokines and growth factors, and modulating cell

surface bound proteins, cell-cell adhesions and cell-surface adhesions involved

in potentially involved in migration, fusion, and apoptosis.57,62-65,63,66 MMP

production is regulated transcriptionally in cells by cytokines and other factors, while secreted MMPs are inactivated by tissue inhibitors of matrix metalloproteinases (TIMPs), which are also secreted by numerous cell types including macrophages. 64,67-72,62,63,73 A material-dependent, proteomic analysis

of MMP and TIMP production by macrophages and FBGCs in culture would

indicate additional mechanisms by which material surface chemistry affect these

adherent cells direction the subsequent biological responses.

Kinetic Modeling

Few mathematical models have been developed regarding cellular adhesion. Studies have developed models to investigate cell-cell adhesion, the spatial interactions between tumor-associated macrophages, tumor cells, and normal tissue cells, and the role of macrophages in angiogenesis.74-78 The cell-

cell adhesion model by Coombs et.al. analyzes cell-cell adhesion based on the

interactions of multiple receptor-ligand interactions with respect to the physical

properties of the cellular membranes and concentration and physical dimensions

of the ligand-receptor pairs This model’s focus is not applicable to the cell-

biomaterial interactions discussed in this research.77 Another model by

- 25 - Chapter I

Table 1.4 Summary of Previous Findings of Macrophage-Derived Cytokine mRNA Expression55,56 Differences in mRNA Expression (RT-PCR) Cytokines in vitro in vivo Day 3 Day 7 Day 10 Day 7 Day 14 Day 21 IL-1β x x x x x Modification Effect IL-1RA x x Material Modification x x Effect Effect IL-2 x x x

IL-4 x x x

IL-6 x x x Modification x Modification Effect Effect IL-8 x x Material Modification Modification Modification Effect Effect Effect Effect IL-10 x Material Material Modification x Modification Effect Effect Effect Effect IL-13 x x x

β-actin x x x x x x

TNF-α x x x x x x

TGF-β Modification Modification x Effect Effect “x” indicates cytokine production, but not necessarily a material-dependency.

Bigerelle and Anselme address the kinetics of cell adhesion with biomaterials with respect to material nature, surface topography, and surface chemistry; however the cells of interest are osteoblasts, which undergo proliferation, unlike monocytes/macrophages, and do no fuse. In addition, this model focuses on adhesion in terms of strength not cell number and does not address cellular adhesion with respect to apoptosis or cell detachment. Only one model addressed macrophage adhesion and fusion. Previous studies in our laboratory by Zhao, et.al. mathematically modeled the probability of the FBGCs size distribution (i.e. number of nuclei) at a given time in vivo on biomaterial

- 26 - Chapter I

surfaces.79,80 The parameters of the system were the initial density of FBGCs

measured, the initial density of cells or macrophages measured, time, and a

value that accounts for the cell area covered by 2-5 macrophages. The model

was capable of predicting FBGC size distribution and showed it to be material-

dependent. However, the model determines the probability of the size

distribution, without accounting for cell differentiation, detachment, and

apoptosis.

Significance

Systematical approaches have been taken to understand the individual

stages (i.e. acute inflammation, chronic inflammation, foreign body reaction, and

fibrosis) of wound healing. Monocytes and macrophages have been shown to be

key cells involved throughout the process of wound healing. Studies have

identified a variety of commonly used biomaterials to promote or inhibit

monocyte/macrophage adhesion and fusion and have suggested that monocyte,

macrophage, and FBGCs behavior is surface dependent. Components of

cellular adhesion and fusion, such as integrin and mannose receptor expression

and cytoskeletal arrangement, have be shown to differ between early timepoints

in vitro, where monocyte/macrophage adhesion predominates and at later

timepoints, when fusion is prevalent. Yet, the relationship between surface chemistry (i.e. hydrophobic, hydrophilic, and ionic chemistries) and monocyte/macrophage adhesion, fusion, and apoptosis is still left to be ascertained. In addition, previous studies have indicated that cytokine

- 27 - Chapter I

expression is surface chemistry material dependant. An in vitro large scale proteomic approach to this hypothesis would isolate and quantify the final products (cytokine/chemokine/MMP/TIMP proteins) produced by adherent monocyte, macrophage, and FBGCs and be useful in establishing surface chemistry criteria for future biomaterial applications. To conclude, a mathematical model of our in vitro cell culture system would further elucidate these material-dependent cellular behaviors and provide the first step at quantitatively predicting these interactions.

Hypothesis & Specific Aims

The Hypothesis and Specific Aims of this work are as follows:

Overall Hypothesis: Material surface chemistry directs adherent monocyte,

macrophage, and foreign body giant cell behaviors, specifically

monocyte/macrophage adhesion, macrophage activation, macrophage fusion

into FBGCs, macrophage and FBGC cytokine/chemokine/MMP/TIMP production,

and monocyte/macrophage apoptosis.

Specific Aim #1: To elucidate the relationship between hydrophobic, hydrophilic,

and ionic surface chemistries and adherent cellular behaviors using distinct

surface-modified polyethylene terephthalate (PET) materials. This research

is discussed in Chapters II, III, IV, and VI.

Specific Aim #2: To determine the effects of the addition of hydrophobic or

hydrophilic surface modifying endgroups (SMEs) to clinically used

polyurethanes on monocyte/macrophage adhesion, macrophage fusion into

- 28 - Chapter I

FBGCs, and monocyte/macrophage apoptosis (presented in Chapter V).

Specific Aim #3: To ascertain the effects of commonly used functional groups

(CH3, COOH, & OH) on adherent cellular behaviors using homogeneous self-

assembling monolayer (SAMs) surfaces. Research for specific aim #3 are

thoroughly discussed and analyzed in Appendix A.

Specific Aim #4: To develop a mechanistic model based on experimental-

derived quantitative metrics that will define the relationships of inflammatory

cell behavior on biomaterial surfaces (addressed in Chapter VI).

References

1. Administration FDA. Just the facts - Better health care with quality medical

devices: FDA on the cutting edge of device technology: Food and Drug

Administration; 2002. Publication No. FS 01-5 p.

2. Brandt EN, Weertman JR, Andersen FA, Burke AP, DeHaven KE, Frye R,

Greenberger P, Kantrowitz AR, Noah L, Refojo MF and others. Improving

Medical Implant Performance Through Retrieval Information: Challenges

and Opportunities. 2000; Bethesda, MD.

3. Anderson JM. Inflammatory reaction: The nemesis of implants. In: Zilla P,

Greisler HP, Landes RG, editors. Tissue Engineering of Vascular

Prosthetic Grafts; 1999. p 197-206.

4. Anderson JM. Biological responses to materials. Annual Reviews Material

Research 2001;31:81-110.

5. Anderson JM. Inflamatory response to implants. ASAIO 1988;11:101-107.

- 29 - Chapter I

6. Anderson JM. Multinucleated giant cells. Current Opinion in Hematology

2000;7:40-47.

7. Luttikhuizen DT, Harmsen MC, Van Luyn MJA. Cellular and molecular

dynamics in the foreign body reaction. Tissue Engineering

2006;12(7):1955-1970.

8. Ward WK, Slobodzian EP, Tiekotter KL, Wood MD. The effect of

microgeometry, implant thickness and polyurethane chemistry on the

foreign body response to subcutateneous implants. Biomaterials

2002;23:4185-4192.

9. Sanders JE, Bale SD, Neumann T. Tissue response to microfibers of

different polymers: polyester, polyethylene, polylactic acid, and

polyurethane. Journal Biomedical Materials Research 2002;62:222-227.

10. Ward WK, Slobodzian EP, Tiekotter KL, Wood MD. The effect of

microgeometry, implant thickness and polyurethane chemistry on teh

foreign body response to subcutatenous implants. Biomaterials

2002;23:4185-4192.

11. Saltzman WM. Cell interactions with polymers. Tissue Engineering:

Principles for the design of replacement organs and tissues. New York,

NY: Oxford University Press; 2004. p 348-385.

12. Vogler EA. Structure and reacitivty of water at biomaterial surfaces.

Advances in Colloid and Interface Science 1998;74:69-117.

13. Christenson EM, Anderson JM, Hiltner A. Oxidative mechanisms of

poly(carbonate urethane) and poly(ether urethane) biodegradation: In vivo

- 30 - Chapter I

and in vitro correlations. Journal Biomedical Materials Research

2004;70A(245-255).

14. Christenson EM, Dadsetan M, Wiggins MJ, Anderson JM, Hiltner A.

Poly(carbonate urethane) and poly(ether urethane) biodegradation: In vivo

studies. Journal Biomedical Materials Research 2004;69A(407-416).

15. Christenson EM, Dadsetan M, Anderson JM, Hiltner A. Biostability and

macrophage-mediated foreign body reaction of silicone modified

polyurethanes. Journal Biomedical Materials Research 2005;74A:141-155.

16. Schubert MA, Wiggins MJ, Schaefer MP, Hiltner A, Anderson JM.

Oxidative biodegradation mechansms of biaxially strained

poly(etherurethane urea) elastomers. Journal Biomedical Materials

Research 1995;29(3):337-347.

17. Wiggins MJ, Anderson JM, Hiltner A. Effect of strain and strain rate on

fatigue-accelerated biodegradation of polyurethane. Journal Biomedical

Materials Research Part A 2003;66(3):463-475.

18. Wiggins MJ, Wilkoff B, Anderson JM, Hiltner A. Biodegradation of

polyether polyurethane inner insulation in bipolar pacemaker leads.

Journal of Biomedical Materials Research (Applied Biomaterials)

2001;58:302-307.

19. Vroman L. The life of an artificial device in contact with blood: initial events

and their effects on its final state. Bulletin of the New York Academy of

Medicine 1988;64:352-357.

- 31 - Chapter I

20. Shen M, Horbett TA. The effects of surface chemistry and adsorbed

proteins on monocyte/macrophage adhesion to chemically modified

polystyrene surfaces. Journal Biomedical Materials Research

2001;57:336-345.

21. Kao WJ. Evaluation of protein-modulated macrophage behavior on

biomaterials: designing biomimetic materials for cellular engineering.

Biomaterials 1999;20:2213-2221.

22. Freij-Larsson C, Jannasch P, Wesslen B. Polyurethane surfaces modified

by amphiphilic polymers: effects on protein adsorption. Biomaterials

2000;21:307-315.

23. Woodhouse KA, Brash JL. Adsorption of plasminogen from plasma to

lysine-derivatized polyurethane surfaces. Biomaterials 1992;13(15):1103-

1108.

24. Tziampazis E, Kohn J, Moghe PV. PEG-variant biomaterials as selectively

adhesive protein templates: model surfaces for controlled cell adhesion

and migration. Biomaterials 2000;21:511-520.

25. Chen Z, Ward R, Tian Y, Malizia F, Gracias DH, Shen YR. Interaction of

fibrinogen with surface of end-group-modified polyurethanes: a surface-

specific sum-frequency generation vibrational spectroscopy study. Journal

Biomedical Materials Research 2002;62:254-264.

26. Gombotz WR, Guanghui W, Horbett TA, Hoffman AS. Protein adsorption

to poly(ethylene oxide) surfaces. Journal Biomedical Materials Research

1991;25:1547-1562.

- 32 - Chapter I

27. Jenney C, Anderson JM. Effects of surface-coupled polyethylene oxide on

human macrophage adhesion and foreign body giant cell formation in

vitro. Journal Biomedical Materials Research 1999;44:206-216.

28. Jenney C, Anderson JM. Adsorbed IgG: a potent adhesive substrate for

human macrophages. Journal Biomedical Materials Research

2000;50:281-290.

29. Maheshwari G, Brown G, Lauffenburger DA, Wells A, Griffith LG. Cell

Adhesion and motility depend on nanoscale RGD clustering. Journal of

Cell Science 2000;113:1677-1686.

30. Jahangir A, McClung W, Cornelius R, McCloskey C, Brash J, Santerre J.

Fluorinated surface-modifying macromolecules: modulating adhesive

protein and platelet interactions on a polyether-urethane. Journal of

Biomedical Material Research 2002;60:135-147.

31. McNally A, Anderson JM. Beta-1 and beta-2 integrins mediate adhesion

during macrophage fusion and multinucleated foreign body giant cell

formation. American Journal Pathology 2002;160(2):621-630.

32. Chang CC, Rosenson-Schloss RS, Bhoj TD, Moghe PV. Leukocyte

chemosensory migration on vascular prosthetic biomaterial is medicated

by an integrin B2 receptor chain. Biomaterials 2000;21:2305-2313.

33. McNally AK, Anderson JM. Complement C3 participation in monocyte

adhesion to different surfaces. Proceedings of the National Academy of

Science 1994;91:10119-10123.

- 33 - Chapter I

34. McNally A, Anderson JM. Interleukin-4 induces foreign body giant cells

from human monocytes/macrophages. American Journal Pathology

1995;147:1487-1499.

35. Kao WJ, McNally A, Hiltner A, Anderson JM. Role for interleukin-4 in

foreign body giant cell formation on a poly(etherurethane urea) in vivo.

Journal Biomedical Materials Research 1995;29:1267-1275.

36. DeFife K, Jenney C, McNally A, Colton E, Anderson JM. Interleukin-13

induces human monocyte/macrophage fusion and macrophage mannose

receptor expression. Journal of Immunology 1997;158(3385-3390).

37. Stein M, Keshav S, Harris N, Gordon S. Interleukin-4 potently enhances

murine macrophage mannose receptor activity: a marker of alternative

immunologic macrophage activation. Journal of Experimental Medicine

1992;176(282-292).

38. Collier T, DeFife K, Jenney C, McNally A. Macrophage mannose receptor

inhibitors prevent interleukin-13 induced multinucleated giant cell

formation. 1998; Dallas, Texas.

39. McNally A, Anderson JM. Interleukin-4 induces foreign body giant cells

from human monocytes/macrophages. Amer J Path 1995;147:1487-1499.

40. Te Velda AA, Klomp JP, Yard BA, Vries JE, Figdor CG. Modulation of

phenotypic and fucional properties of human peripheral blood monocytes

by IL-4. Journal of Immunology 1988;140:1548-1554.

- 34 - Chapter I

41. Te Velda AA, Huijbens RJF, Heije K, de Vries JE, Figdor CG. Interleukin-4

inhibits secretion of interleukine-1b, tumor necrosis factor-1 and

interleukin-6 by human monocytes. Blood 1990;76:1392-1397.

42. Donnelly RP, Fenton MJ, Finbloom DS, Gerrard TL. Differential regulation

of IL-1 production in human monocytes by IFN-gamma and IL-4. Journal

of Immunology 1990;145:569-575.

43. Minty A, Chalon P, Derocq JM, Dumont X, Guillemot JC, Kaghad M, Labit

C, Leplatois P, Liauzun P, Miloux B and others. Interleukin-13 is a new

human lymphokine regulating inflammatory and immune responses.

Nature 1993;362:248-250.

44. De Waal Malefyt R, Figdor CG, Huijbens S, Mohan-Peterson S, Bennett

B, Culpepper J, Dang W, Zurawski G, de Vries JE. Effects of IL-13 on

phenotype, cytokine production, and cytotoxic function of human

monocytes. Journal of Immunology 1993;151:6370-6381.

45. DeFife K, Jenney C, Colton E, Anderson JM. Cytoskeletal and Adhesive

Structural Polarization Accompany IL-13-induced Human Macrophage

Fusion. Journal of Histochemistry & Cytochemistry 1999;47(1):65-74.

46. DeFife K, Jenney C, Colton E, Anderson JM. Disruption of filamentous

actin inhibits human macrophage fusion. FASEB J 1999;13:823-832.

47. Huppertz B, Frank HG, Kaufmann P. The apoptosis cascade -

morphological and immunohistochemical methods for its visualization.

Anatomy and Embryology 1999;200:1-18.

- 35 - Chapter I

48. Martin SJ, Reutelingsperger CPM, McGahon AJ, Rader JA, van Schie

RCAA, LaFace DM, Green DR. Early redistribution of plasma membrane

phosphatidylserine is a general feature of apoptosis regardless of initiating

stimulus. Journal of Experimental Medicine 1995;182:1545-1556.

49. Brodbeck W, Shive M, Colton E, Nakayama Y, Matsuda T, Anderson JM.

Influence of biomaterial surface chemistry on the apoptosis of adherent

cells. Journal Biomedical Materials Research 2001;55:661-668.

50. Mayadas TN, Cullere X. Neutrophil B2 integrins: moderators of life or

death decisions. Trends in Immunology 2005;26(7):388-395.

51. Bonfield TL, Colton E, Anderson JM. Protein adsorption of biomedical

polymers influences activated monocytes to produce fibroblast stimulating

factors. Journal Biomedical Materials Research 1992;26:457-465.

52. DeFife K, Colton E, Nakayama Y, Matsuda T, Anderson JM. Spatial

regulation and surface chemistry control of monocyte/macrophage

adhesion and foreign body giant cell formation by photochemically

micropatterned surfaces. Journal Biomedical Materials Research

1999;45(148-154).

53. Anderson JM, Ziats NP, Azeez A, Brunstedt MR, Stack S, Bonfield TL.

Protein adsorption and macrophage activation on polydimethylsiloxane

and silicone rubber. Journal of Biomaterial Science Polymer Edition

1995;7(2):159-169.

54. Abramson ST, Gallin JI. IL-4 inhibits superoxide production by human

mononuclear phagocytes. Journal of Immunology 1990;144(625-530).

- 36 - Chapter I

55. Brodbeck W, Voskerician G, Ziats NP, Nakayama Y, Matsuda T,

Anderson JM. In vivo leukocyte cytokine mRNA responses to biomaterials

are dependent on surface chemistry. Journal Biomedical Materials

Research 2003;64A:320-329.

56. Brodbeck W, Nakayama Y, Matsuda T, Colton E, Ziats NP, Anderson JM.

Biomaterial surface chemistry dictates adherent monocyte/macrophage

cytokine expression in vitro. Cytokine 2002;18(6):311-319.

57. Visse R, Nagase H. Matrix metalloproteinases and tissue inhibitors of

matrix metalloproteinases: structure, function, and biochemistry.

Circulation Research 2003;92:827-839.

58. Meznarich N, Kyriakides T, Donaldson E, Foster M, Schrom B, Ratner B,

Hauch K, Bornstein P. Matrix-metalloproteinase (MMP-9) and its role in

wound healing and the foreign body response. Journal of Undergraduate

Research in Bioengineering 2004;4(2):84-89.

59. Klein S, Anderson G, Kennedy A, Bond S. The effects of broad-spectrum

matrix metalloproteinase inhibitor on characteristics of wound healing.

Journal of Investigative Surgery 2002;15:199-207.

60. Lafleur M, Handsley M, Edwards D. Metalloproteinases and their inhibitors

in angiogenesis. Expert Reviews in Molecular Medicine 2003;5(22):1-39.

61. Durum SK, Hirano T, Vilcek T, Nicola NA. Cytokine Reference: Academic

Press; 2000.

- 37 - Chapter I

62. Nagase H, Visse R, Murphy G. Structure and function of matrix

metalloproteinases and TIMPs. Cardiovascular Research 2006;69:562-

573.

63. Sternlicht M, Werb Z. How matrix metalloproteinases regulate cell

behavior. Annual Review of Cell Developmental Biology 2001;17:463-516.

64. Borden P, Heller R. Transcriptional control of matrix metalloproteinases

and tissue inhibitors of matrix metalloproteinases. Critical Reviews in

Eukaryotic Gene Expression 1997;7:159-178.

65. Newby A. Matrix metalloproteinases regulate migration, proliferation, and

death of vascular smooth muscle cells by degrading matrix and non-matrix

substrates. Cardiovascular Research 2006;69:614-624.

66. Imai K, Hiramatsu A, Fukushima D, Pierschbacher M, Okada Y.

Degradation of decorin by matrix metalloproteinases: identification of the

cleavage sites, kinetic analyses, and transforming growth factor beta-1

release. Biochemistry Journal 1997;322:809-814.

67. Oltmanns J, Fietz I, Witt C, Jung K. Increased production of matrix-

metalloproteinase-2 in alveolar macrophages and regulation by

interleukin-10 in patients with acute pulmonary sarcoidosis. Experimental

Lung Research 2002;28(55-68).

68. Reunanen N, Westermarck J, Hakkinen L, Holmstrom T, Elo I, Eriksson J,

Kahari V. Enhancement of fibroblast collagenase (matrix

metalloproteinase-1) gene expression by ceramide is mediated by

- 38 - Chapter I

extracellular signal-regulated and stress-activated protein kinase

pathways. The Journal of Biological Chemistry 1998;273(9):5137-5145.

69. Vincenti M, Brickerhoff C. Transcriptional regulation of collagenase (MMP-

1, MMP-13) genes in arthritis: integration of complex signaling pathways

for the recruitment of gene-specific transcription factors. Arthritis Research

2002;4(157-164).

70. Munshi H, Stack M. Reciprocal interactions between adhesion receptor

signaling and MMP regulation. Cancer Metastasis Reviews 2006;25:45-

56.

71. Uria J, Jimenez M, Balbin M, Freije J, Lopez-Otin C. Differential effects of

transformation growth factor beta on the expression of collagenase-1 and

collagenase-3 in human fibroblasts. Journal of Biological Chemistry

1998;273(16):9769-9777.

72. Liacini A, Sylvester J, Li W, Zafarullah M. Inhibition of interleukin-1-

stimulated MAP kinases, activating protein-1, nuclear factor kappa B

transcription factors down-regulates matrix metalloproteinase gene

expression in articular chondrocytes. Matrix Biology 2002;21:251-262.

73. Baker A, Edwards D, Murphy G. Metalloproteinase inhibitors: biological

actions and therapeutic opportunities. Journal of Cell Science

2002;115:3719-3727.

74. Owen MR, Sherratt JA. Pattern formation and spatiotemporal irregularity

in a model for macrophage-tumor interactions. Journal of Theoretical

Biology 1997;189:63-80.

- 39 - Chapter I

75. Turner S. Using cell potential energy to model the dynamics of adhesive

biological cells. Physical Review 2005;71:041903.

76. Levine HA, Sleeman BD, Nilsen-Hamilton M. A mathematical model for

the roles of pericytes and macrophage sin the intiation of angiogenesis. I.

The role of protease inhibitors in preventing angiogenesis. Mathematical

Biosciences 2000;168:77-115.

77. Coombs D, Dembo M, Wofsy C, Goldstein B. Equilibrium

Thermodynamics of Cell-Cell AdhesionMediated by Multiple Ligand-

Receptor Pairs. Biophysical Journal 2004;86:1408-1423.

78. Qazi S, Beltukov A, Trimmer BA. Simulation modeling of ligand receptor

interactions at non-equilibrium conditions: processing of noisy inputs by

ionotropic receptors. Mathematical Biosciences 2004;187:93-110.

79. Zhao QH, Anderson JM, Hiltner A, Lodoen GA, Payet CR. Theoretical

analysis on cell size distribution and kinetics of foreign body giant cell

formation in vivo on polyurethane elastomers. Journal Biomedical

Materials Research 1992;26(8):1019-1038.

80. Kao WJ, Zhao QH, Hiltner A, Anderson JM. Theoretical analysis of in vivo

macrophage adhesion and foreign body giant cell formation on

polydimethylsiloxane, low density polyethylene, and polyetherurethanes.

Journal Biomedical Materials Research 1994;28(1):73-9.

- 40 - Chapter II

Chapter II: Proteomic Analysis and Quantification of Cytokines and Chemokines from Biomaterial Surface-Adherent Macrophages and Foreign Body Giant Cells

Abstract

Implantation of biomaterial devices results in the well-known foreign body

reaction consisting of monocytes, macrophages, and foreign body giant cells

(FBGCs) at the material/tissue interface. We continue to address the hypothesis

that material surface chemistry modulates the phenotypic expression of these

cells. Utilizing our human monocyte culture system, we have used surface-

modified polymers displaying hydrophobic, hydrophilic, and/or ionic chemistries

to determine the cytokines/chemokines released from biomaterial-adherent

macrophages/FBGCs. This study broadens our approach by using proteomic

analysis to identify important factors expressed by these cells and further

quantifies these molecules with ELISAs. Proteomic profiles changed over time

suggesting that the adherent macrophages underwent a phenotypic switch.

Macrophage/FBGC-derived pro-inflammatory cytokines, IL-1β and IL-6,

decreased with time, while the anti-inflammatory cytokine, IL-10, gradually

increased with time. Resolution of the inflammatory response was also demonstrated by a decrease in chemoattractant IL-8 and MIP-1β production with

time.

Material-dependent macrophage/FBGC activation was analyzed using

cytokine/chemokine production and cellular adhesion. Monocyte/macrophage

adhesion was similar on all surfaces, except for the hydrophilic/neutral surfaces

that showed a significant decrease in cellular density and minimal FBGC

- 41 - Chapter II

formation. Normalizing the ELISA data based on the adherent cell population

provided cytokine/chemokine concentrations produced per cell. This analysis

showed that although there were fewer cells on the hydrophilic/neutral surface,

these adherent cells were further activated to produce significantly greater

amounts of each cytokine/chemokine tested than the other surfaces.

This study clearly presents evidence that material surface chemistry can differentially affect monocyte/macrophage/FBGC adhesion and cytokine/chemokine profiles derived from activated macrophages/FBGCs

adherent to biomaterial surfaces.

Introduction

Monocytes, macrophages, and foreign body giant cells (FBGC) are known

to have key roles in inflammation, wound healing, fibrous encapsulation, and the foreign body reaction after the implantation of a biomedical device.1,2

Macrophages are recruited to the implantation site, adhere to the biomaterial,

and can fuse to form foreign body giant cells (FBGCs), a classic identifying

feature of the foreign body reaction. These monocyte-derived cells can produce

a variety of components to control their environment including but not limited to

reactive oxygen species to degrade the biomaterial, chemokines to direct

additional inflammatory and wound healing cells (i.e. lymphocytes, neutrophils,

macrophages, fibroblasts) to the site of injury, and cytokines to further activate or deactivate the surrounding inflammatory and wound healing cells as depicted in

Figure 2.1.3-8 Theoretically, if material surface chemistry can affect the

- 42 - Chapter II

production of the macrophage/FBGC derived cytokines and chemokines, then

the subsequent biological responses to biomaterial implants could be directed

accordingly.

GM-CSF I-309 GCSF IL-1 Lymphocytes GM-CSF IL-6 Fibroblasts I-309 IL-8 IL-6 Neutrophils IL-10 IL-8 IL-12 IL-10 IL-15 IL-12 MCP-1,2,3,4 MCP-1,2,3,4 MIG MCSF MIP-1α MIP-1α MIP-1β IL-1 MIP-1β ENA-78 MIP-1δ IL-8 MIP-1δ IL-6 MIP-3 IL-10 MIP-3 IL-8 RANTES MIP-1α RANTES IL-10 TGF-β MIP-1β TGF-β MIP-1α TNF-α TGF-β TNF-α MIP-1β Eotaxin-2 TNF-α IP-10 MIP-3 IP-10 MIF LIGHT TGF-β MIF MIF TNF-α PARC PARC GCP-2

Polymer

Monocytes/Macrophages/FBGC

Figure 2.1: Cellular/Biomaterial/Cytokine Interactions of Inflammation and Wound Healing.

Cytokines are a complex network of proteins released by numerous inflammatory and wound healing cells (i.e. lymphocytes, macrophages, monocytes, fibroblasts, and neutrophils) that bind to surface receptors influencing cellular behaviors in an overlapping and redundant manner. Pro-inflammatory cytokines, such as interleukin-1 (IL-1), IL-6, and interferon gamma (IFN-γ) can

- 43 - Chapter II

increase neutrophil production, stimulate lymphocyte proliferation and

maturation, and further promote pro-inflammatory cytokine expression.9-11 Anti- inflammatory cytokines (i.e. IL-10, IL-4, and IL-13) can modulate the effects of pro-inflammatory cytokines by inhibiting cytokine production and/or block cell receptor binding.9,11-15

Macrophages have been shown to produce numerous chemokines, a

family of low molecular weight cytokines, that chemoattract other inflammatory

and wound healing cells.16,17 CXC subgroup chemokines (i.e. IL-8) chemoattract

neutrophils to sites of inflammation, induce granule exocytosis, and induce respiratory burst directing the early stages of wound healing.16-18 Other

chemokines are known to attract monocytes in addition to leukocytes, basophils,

and/or eosinophils such as the monocyte chemotactic proteins (MCP-1, 2, 3, & 4)

and monocyte inflammatory protein (MIP-1β).16-18 Both chemokines and

chemokine receptors can be up- or down-regulated upon cell maturation and

activation.19

The activation stimulus has been shown to determine the phenotypic

expression of an activated macrophage. Lipopolysaccharide (LPS) or interferon

gamma (IFN-γ) activated macrophages, coined classically activated

macrophages, up-regulate pro-inflammatory cytokines (i.e. IL-6, tumor necrosis

factor (TNF) and IL-1), inhibit anti-inflammatory cytokines (i.e IL-10), variably

promote or inhibit production of a variety of chemokines, produce nitric oxide,

and down-regulate mannose receptors and arginase production.9,11,18-21 In contrast, IL-4 or IL-13 alternatively activated macrophages inhibit pro-

- 44 - Chapter II

inflammatory cytokines (i.e. IL-6, TNF, and IL-1), promote IL-10 and IL-1ra

(receptor antagonist) cytokine production, modulate chemokine expression in a

manner usually different than classically activated macrophages, up-regulate

mannose receptors, and produce arginase.9,11,14,15,19,21,22 An increasing number

of stimuli (i.e. IL-10 and glucocorticoid hormones) have been shown to activate

macrophages.9

Previous studies have demonstrated that cytokine production (i.e IL-1, IL-

6, and TNF-α from monocytes/macrophages were disparate on various commonly used biomaterials (i.e. polydimethylsiloxane, polytetrafluoroethylene, poly-L-Lactide, polyurethanes, Dacron®, and Biomer®) and on modified copolymers. 4,23-27 Recent studies in our laboratory have shown that

hydrophilicity and charge can modulate IL-10, IL-1ra, and IL-8 mRNA expression

both in vitro and in vivo.28,29

A proteomic approach to investigating cytokine and chemokine production is beneficial in that it directly analyzes the final functional products secreted from a cell and presented in the surrounding milieu. Although, a genomic approach measuring mRNA and DNA expression is indicative of cellular up-regulation, it is limited by potential transcriptional regulations and post-translational modifications that decrease or block protein synthesis. Ultimately, it does not measure protein

secretions or their functionality. Techniques such as cytokine arrays, ELISAs,

and two-D gel/mass spectroscopy are capable of detecting numerous proteins in

the same sample, easy to manufacture and use, and continue to be available for

- 45 - Chapter II

the detection of more proteins.30,31 A combinatorial study provides the most

insight into protein production from adherent cells.

This study utilizes a proteomic approach to investigate the effect of

material surface chemistry on monocyte, macrophage, and FBGC production of

cytokines and chemokines building upon previous studies in our laboratory on

cytokine and chemokine mRNA expression. Cytokine arrays were utilized to

detect the inflammatory cytokine and chemokine profiles produced in cell cultures

containing monocytes, macrophages, and FBGCs adherent to photografted

polymers of distinct surface chemistries. Select cytokines and chemokines were

further quantified with enzyme-linked immunosorbent assays (ELISAs). Material-

dependent macrophage activation was also analyzed based on cytokine and

chemokine production and cellular adhesion.

Materials and Methods

Biomaterials

Photografted surfaces were synthesized and obtained from the laboratory

of Dr. Takehisa Matsuda (Kyushu University, Fukuoka, Japan). These surfaces

displaying distinct surface chemistries were manufactured using a custom

designed, semiautomatic device for large-scale laboratory production described

previously.32,33 Briefly, polyethylene terephthalate film (Mylar®, PET) is coated with poly(styrene-co-benzyl N,N-dimethyldithiocarbamate) (BDEDTC).

Photolysis via UV irradiation of the BDEDTC initiated graft polymerization of

BDEDTC with acrylamide, AAm; the sodium salt of acrylic acid, AANa; or the

- 46 - Chapter II

methyl iodide salt of N-[3-(dimethylamino) propyl] acrylamide, DMAPAAmMeI.

In addition, to the PET polymer control, four distinct surfaces were utilized within

this study, specifically, BDEDTC, hydrophobic; PAAm, hydrophilic and neutral;

PAANa, hydrophilic and cationic; and DMAPAAmMEI, hydrophilic and anionic.

All surfaces were cut into 15-mm diameter disks and sterilized using 100%

ethanol. Silicone rings (approximately 8mm in height) were section from tubing

(Cole-Parmer, Vernon Hills, IL), sonicated in 100% ethanol for 5 minutes, and autoclaved for 1 hour. Surfaces were secured with the silicone rings into 24-well tissue culture polystyrene (TCPS) plate wells (Fisher Scientific, Pittsburgh, PA) under sterile conditions. Any residual ethanol was removed via a wash with warmed, sterile Dulbecco’s phosphate buffered saline (Gibco, Grand Island, NY) containing magnesium chloride and calcium chloride (PBS++). Each surface has a resultant surface area of 0.71 cm2.

Surface Characterization

Surface hydrophobicity/hydrophilicity was determined via advancing water contact angles measured using the sessile drop method and a goniometer

(Edmund Scientific, Barrington, NJ) at 22oC room temperature. Surface charge

was detected by staining the surfaces with a dilute solution of toluidine blue

(PAANa) or rose bengal (DMAPAAmMeI) (Sigma-Aldrich, St. Louis, MO).

Surfaces were incubated in 1mL of 1.0 w/v% solution of toluidine blue or 1.0

w/v% solution of rose bengal for 4 hours at room temperature. Samples were

rinsed thrice with deionized, distilled water for 5 minutes and dried overnight prior

- 47 - Chapter II

to imaging. In addition, surface uniformity and surface modification of the PET

Mylar® substrate with the four distinct chemistries were further confirmed using

attenuated total reflectance Fourier transform infrared analysis (ATR-FTIR) as

described later.

In Vitro Cell Culture

Human monocytes and serum were isolated from the whole, venous blood

of healthy, unmedicated donors by a density gradient centrifugation method

using Ficoll and Percoll as described previously.34 Isolated monocytes were

cultured in 0.5mL of macrophage serum-free media (SFM) (Gibco, Grand Island,

NY) with 20% autologous serum (AS) at a concentration of 5x105cells/mL under

o sterile conditions. All cultures were incubated at 37 C with a 5% CO2

environment. Monocytes were allowed to adhere for 2 hours, after which all non- adherent cells were removed via a 1mL wash with warmed, sterile PBS++.

Adherent cells continued to be cultured in fresh media containing SFM with 20% heat-treated (1 hour at 56oC), autologous serum (HAS) for 3, 7, and 10 days. At

these timepoints, appropriate cultures were terminated and the remaining

cultures were continued with fresh media. Interleukin-4 (IL-4), a fusion-inducing

cytokine, was added to the fresh media of select cultures at a concentration of

15ng/mL at days 3 and 7 in order to promote fusion. At days 3, 7, and 10,

supernatants were collected via pipetting, non-adherent cells were rinsed twice

with warmed PBS++, and adherent cells were fixed with 100% methanol for 5

minutes for adherent cell density analysis. Duplicates of the collected

- 48 - Chapter II

supernatants were combined, centrifuged at 10,000 min-1 for 5 minutes to

remove all non-adherent cells, aliquoted to minimize freezing/thawing cycles, and

stored at -80oC until ELISA analysis could be performed.

Adherent Cell Density Analysis

Fixed adherent cells were stained with May-Grünwald for 5 minutes, rinsed with PBS++ twice, incubated in Giemsa for 15 minutes, washed with

distilled water thrice, and allowed to air-dry overnight. The total number of

adherent nuclei, adherent cells, number of FBGCs, and nuclei within FBGCs

were counted over 5-20x objective areas (463 x 463 μm) using optical

microscopy. Data is presented as the average total number of adherent cells

and the average percentage of nuclei within FBGCs of the total number of

adherent nuclei.

Cytokine Screening

A protein array system was utilized to determine the presence or absence

of a total of 77 cytokines/chemokines in cell culture supernatants collected at

days 3 and 10. Human cytokine array membrane kits (catalog # AAH-CYT-5)

were purchased from RayBiotech, Inc. (Norcross, GA) and used according to the

manufacturer’s instructions. Qualitative positive results were determined based

on the presence of a weak (+) to strong (+++) signal in the location for the

particular protein of interest. Cell cultures from a total of three donors were

investigated and trends were compared to confirm a positive detection between

- 49 - Chapter II

donors. Results from this analysis in concert with previous positive findings for a

particular protein determined the cytokines and chemokines to be quantified

using ELISA.

Cytokine Quantification using ELISA

Levels of cytokine production in sample supernatants at days 3, 7 and 10

were quantified using commercially available ELISA kits (R&D Systems,

Minneapolis, MN) in accordance with manufacturer’s instructions. SFM was the

diluent for the both standards and diluted samples to ensure consistency. The

validity of the results for each sample quantified in each ELISA kit was analyzed

based upon the standard curve of calculated values compared to known

standard protein concentrations. For any invalid data points, testing of the

sample was repeated at a different dilution and/or duplicated to further test for

validity. Quantified data presented is the total concentration (pg/mL) of a given cytokine produced.

Statistical Analysis

Triplicate samples were tested for each surface characterization technique. Donor variability was accounted for using multiple (n=4) human donors. Statistical significance between samples was determined using a split- plot general factorial ANOVA analysis and Tukey’s tests with a 95% confidence interval. Statistical analysis was conducted using Minitab 14 statistical software

- 50 - Chapter II

package (Minitab, Inc., State College, PA). All data is presented as the average

± the standard error of the mean (SEM) (n=3-4).

Results

Biomaterial Surface Chemistry Confirmation

The advancing water contact angles for PAAm (46.3 o ± 12.5 o), PAANa

(24.6 o ± 3.9 o), and DMAPAAmMeI (31.4 o ± 6.6 o) indicated the hydrophilic

nature of the surfaces (Table 2.1). BDEDTC and PET were found to be

hydrophobic with advancing water contact angles of 68.1o ± 6.7 o and 70.7 o ± 2.8 o, respectively. An anionic surface charge was confirmed for PAANa via positive

staining with toluidine blue, a monocationic dye, while the cationic surface,

DMAPAAmMeI, stained positively with rose bengal, a dianionic dye. No positive

staining was observed for any of the other material surfaces with either acid or

base reacting dye.

Table 2.1 Surface Chemistries of the Photografted Polymers Advancing Biomaterial Base Coating Photografted Chemical Contact Surface Polymer Property Angle (o)

PET PET - - Hydrophobic 70.7 ± 2.8

BDEDTC PET BDEDTC - Hydrophobic 68.1 ± 6.7 Hydrophilic PAAm PET BDEDTC PAAm & Neutral 46.3 ± 12.5 Hydrophilic PAANa PET BDEDTC PAANa & Anionic 24.6 ± 3.9 Hydrophilic DMAPAAmMeI PET BDEDTC DMAPAAmMeI & Cationic 31.4 ± 6.6 PET, Polyethylene terephthalate; BDEDTC, Poly(styrene-co-benzyl N,N- diethyldithiocarbamate); PAAm, Polyacrylamide; PAANa, Sodium salt of poly(acrylic acid); and DMAPAAmMEI, Methyl iodide of poly[3- (dimethylamino)propyl[acrylamide.

- 51 - Chapter II

Attenuated total reflectance Fourier transform infrared (ATR-FTIR) spectra

(not shown) were acquired using a Nexus 870 FTIR bench coupled to a

Continuum microscope (Thermo Nicolet, Madison, WI) with an attenuated total

reflectance (ATR) slide-on attachment containing a germanium crystal. Spectra

were collected at a resolution of 2.0 cm-1 for 32 scans over an area of 150 μm x

150 μm. Three samples of each material were analyzed over five distinct areas

to investigate sample variability and surface uniformity.

The ATR-FTIR spectra of PET contained the following characteristic IR

peaks with the respective correlating characteristic bonds: 1715 cm-1 (C=O

stretch), 1400-1600 cm-1 (C=C stretch), 1248 cm-1 (C-O stretch), 1100-1125 cm-1

(C-O stretch), 790-1023 cm-1 (C-H bend), and 724 cm-1 (C-H out-of-plane

bend).35,36 BDEDTC-coated PET surface spectra displayed peaks at 3026 cm-1,

1516 & 1410 cm-1, 1210 cm-1, and 700 cm-1 corresponding to the aromatic C-H

stretch, aromatic C=C stretch, C-N stretch, and C-H rock, respectively, in addition

to the above characteristic PET peaks indicating the presence of the

modification. The addition of the polyacrylamide in PAAm was confirmed by the characteristic peaks for the grafted amide in addition to the spectral peaks of

PET and BDEDTC: 3361 cm-1 and 3189 cm-1 (N-H stretch with two peaks for an unsubstituted amide), 1681 cm-1 (C=O stretch), and 1650 cm-1 (N-H bend). FTIR

spectra for PAANa and MEI were similar to the BDEDTC spectra. This is

expected for PAANa, because these surfaces only contain the sodium salt of

- + poly(acrylic acid) (CH2CHCOO Na )n, which does not contribute any new bonds

- 52 - Chapter II

aside from C=O and C-O bonds that were previously contained in the PET base

substrate.35,37 However, MEI is expected to contain additional peaks correlating

to the amide functional group bonds (C=O, C-N, and N-H).

Consistent spectral peaks and peak intensity over multiple areas (5 areas)

on multiple PET and BDEDTC samples (n=3) confirmed the uniformity of these

surface chemistries. For PAAm, PAANa, and MEI, the spectral peak intensities

varied slightly, however the spectral peak frequencies and corresponding bonds

were consistent over multiple areas on multiple samples.

This investigation of the photograft polymers using multiple surface

characterization techniques confirmed the presence of the modifications on each photografted copolymer presenting either a hydrophobic, hydrophilic/neutral, hydrophilic/cationic, and hydrophilic/anionic surface chemistry in agreement with previous studies.32,38-40

Material-Dependent Macrophage Adhesion and FBGC Formation

Adherent cell densities were determined at each timepoint in order to determine the material surface chemistry dependent nature of adhesion and fusion and to explore a relationship between the adherent cell density and the collective cytokine concentration produced. Cell adhesion was equally supported on PET, BDEDTC, PAANa, and DMAPAAmMeI at days 3, 7 and 10 regardless of

IL-4 media content, as shown in Figure 2.2.A. In contrast, a lower number of cells were adherent to PAAm, the hydrophilic/neutral surface, in comparison to

- 53 - Chapter II

the other materials at all timepoints and conditions. At day 3, macrophages predominated the adherent cell population. The adherent macrophages matured

2500 A PET BDEDTC PAAm PAANa DMAPAAmMeI

2000

1500

1000 cells/mm^2 * ofNumber Adherent Cells, 500 * * * * 0

Day 3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4

70% 60% B

o 50% i

us F 40%

t 30% ercen

P 20% 10% * * * 0% Day 3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4 Time

Figure 2.2: Adherent Cell Densities (A) and FBGC Formation (B). Data represents the mean ± the standard error of the mean (n=3). “*” indicates statistical significance in comparison to PET (p<0.05).

between days 3 and 7 as indicated by a progression from a rounded morphology

towards a more spread morphology with an increase in size and the development

of FBGCs by day 7. A mixture of macrophages and FBGCs were present on the

material surfaces at days 7 and 10. FBGC formation is determined based upon

- 54 - Chapter II the percentage of nuclei within FBGCs of the total number of nuclei on a given surface area (Figure 2.2.B). Percent fusion was found to be greater on PAANa than any of the other materials at day 7 with or without IL-4. All other materials supported equally low amounts of fusion (0-15%) at day 7. Similarly, PAANa promoted a greater amount of fusion in comparison to the other materials at day

10 in the absence of IL-4. With the addition of IL-4, fusion increased on

BDEDTC and DMAPAAmMeI. FBGC formation was equally promoted on

BDEDTC, PAANa, and DMAPAAmMeI and was inhibited on PAAm in the presence of IL-4 at day 10.

Cytokine and Chemokine Qualitative Production Profiles

Antibody-bound membrane protein arrays detected the presence of 24 cytokines and chemokines of the total 77 proteins tested at days 3 and 10 as shown in Table 2.2. Qualitative positive results are indicated by the strength of the detected signal (weak, medium, strong). Material and time dependency trends were investigated. Differences in signal strength were seen over time

(day 3 to day 10) and between materials for the majority of the proteins detected.

IL-4 affected the signal intensity of approximately 50% of the proteins detected at day 10. Media containing 20% serum were also analyzed to determine the presence of any serum proteins and are included in Table 2.2. The following were not present in the cell culture supernatants at either day 3 or 10: cytokines:

GCSF, GM-CSF, IL-1α, IL-2, IL-3, IL-5, IL-7, IL-12 p40p70, IL-13, IL-15, IL-16,

IFN-γ, MCSF, SCF, TGF-β1, TNF-α, TNF-β, IGF-1, Oncostatin M,

- 55 - Chapter II

Thrombopoietin, VEGF, BDNF, Ck 88-1, FGF-4,-6,-7,-9, Flt-3 Ligand, GDNF,

IGFBP-3, IGFBP-4, LIF, LIGHT, MIF, NT-3, NT-4, Osteoprotegerin, and PIGF;

and chemokines: MCP-2, MCP-3, MCP-4, MIP-3α, GRO-α, GCP-2, MIG, IP-10,

SDF-1, BLC, Eotaxin, Eotaxin-3, I-309, and Fractalkine. Select cytokines (IL-1β,

IL-6, and IL-10) and chemokines (IL-8 and MIP-1β) were chosen to be quantified

using ELISA based on their presence in the supernatants and potential material

and time dependencies.

Serum Containing Media Controls

Cytokine production was quantitatively detected using ELISAs in the

supernatants of cell cultures at days 3, 7 and 10. The cell culture media

contained autologous serum. To account for serum cytokine levels, media

controls were measured for cytokine concentrations at each timepoint after

incubation at 37oC. Media controls did not contain IL-6 and IL-8. However, the following cytokines were present in the culture media at the given concentrations:

IL-1β (24±7 pg/mL), IL-10 (4±1 pg/mL), and MIP-1β (15±2 pg/mL). These

media/serum concentrations were subtracted from the total supernatant

concentrations at that timepoint in order to isolate the cellular-derived cytokine

concentrations for each supernatant sample.

- 56 - Chapter II

Table 2.2 Potential Differences in Cytokines/Chemokines Due to Material, Time, or IL-4 Addition as Analyzed by Cytokine Arrays Average Signal Strength Affected By Presence Protein Class Signal Material Time IL-4 in Media Detected Strength (Min./Max. (Signal at Day 3 Signal Strength) Strength) IL-1β Cytokine + Yes (+ / ++) Yes Yes - IL-6 Cytokine +++ No Yes Yes - IL-10 Cytokine - No No No - IL-8 Chemokine +++ Yes (+ / +++) Yes Yes + MIP-1β Chemokine - No No No - TGF-β2 Cytokine + Yes (- / +) Yes No + EGF Cytokine ++ Yes (+ / ++) Yes No ++ Angiogenin Cytokine ++ Yes (+ / ++) Yes No ++ PDGF-BB Cytokine ++ Yes (+ / ++) Yes Yes ++ Leptin Cytokine + Yes (- / +) Yes Yes + GDNF Cytokine + Yes (- / +) Yes Yes + HGF Cytokine + No No No + IGFBP-1 Cytokine + Yes (- / +) Yes Yes + IGFBP-2 Cytokine + No No No + TGF-β3 Cytokine + Yes (- / +) Yes No + GRO Chemokine + No Yes Yes - MCP-1 Chemokine ++ Yes (+ / ++) Yes Yes + MCP-4 Chemokine + Yes (- / +) No Yes - MDC Chemokine + Yes (- / +) Yes No - RANTES Chemokine + Yes (- / +) Yes Yes ++ Eotaxin-2 Chemokine + Yes (- / +) Yes No - MIP-3α Chemokine + Yes (- / +) Yes No - NAP-2 Chemokine + Yes (- / +) No No + PARC Chemokine + Yes (- / +) Yes Yes ++ TARC Chemokine + No Yes Yes - ENA-78 Chemokine + Yes (- / +) Yes No - ‘-’ indicates no signal or no difference. ‘+’ indicates a weak signal. ‘++’ indicates a medium strength signal. ‘+++’ indicates a strong signal. IL = interleukin; MIP = macrophage inflammatory protein-1beta; TGF = transforming growth factor; EGF = epidermal growth factor; PDGF = platelet-derived growth factor; GDNF = glial cell-derived neurotrophic factor; HGF = hepatocyte growth factor; IGFBP = insulin-like growth factor binding protein; GRO = growth related oncogene; MCP = macrophage chemotactic protein; MDC = macrophage-derived chemokine; RANTES = regulated upon activation, normal T-cell expressed, and presumably secreted; NAP = neutrophil activating protein; PARC = pulmonary and activation-regulated chemokine a.k.a. MIP-4; TARC = thymus and activation regulated chemokine; ENA = epithelial-derived neutrophil activating protein.

- 57 - Chapter II

Adherent Cell Cytokine Production

Pro-inflammatory cytokines (IL-1β and IL-6) and anti-inflammatory cytokine (IL-10) were quantified using ELISA in the supernatants of monocyte- derived cell cultures adherent to distinct biomaterials at days 3, 7, and 10. IL-1β was produced on all surfaces at day 3 (Figure 2.3.A). Concentrations of IL-1β were greatest on PAAm (264 ± 98 pg/mL) and least on BDEDTC (68 ± 30 pg/mL). IL-6 concentrations were significantly greater on PAAm (5,050 ± 1,150 pg/mL) in comparison to all other materials except DMAPAAmMeI (2,970 ± 1,760 pg/mL) at day 3 (Figure 2.3.B). Between days 3 and 7, both IL-1β and IL-6 concentrations significantly decreased to an average concentration of 11 ± 1 pg/mL and 177 ± 37 pg/ml, respectively. At day 7 and 10, both cytokine concentrations were comparable on all materials. IL-10 was produced on all surfaces at all timepoints (Figure 2.3.C). At day 3, IL-10 concentrations were least on PAANa (5 ± 2 pg/mL) in comparison to all other materials. This trend continued through day 10 and concentrations did not significantly increase or decrease over time. The addition of IL-4 to culture at day 3 significantly increased IL-10 concentrations on DMAPAAmMeI from 27 ± 8 pg/mL to 102 ± 31 pg/ml at day 7 and 30 ± 1 pg/mL to 100 ± 33 pg/ml at day 10 (Figures 2.4.A &

2.4.B). IL-10 concentrations were also slightly increased on PET, BDEDTC, and

PAAm at day 7 and day 10 in the presence of IL-4; however IL-4 addition did not increase or decrease IL-1β and IL-6 concentrations at day 7 or 10 (data not shown).

- 58 - Chapter II

400 PET BDEDTC PAAm PAANa DMAPAAmMeI 350 A 300

250 200 , pg/mL β 150

IL-1 IL-1 100

50 0 Day 3 Day 7 Day 10

7000

6000 B

5000

4000

3000

IL-6, pg/mL IL-6, 2000

1000 0

Day 3 Day 7 Day 10 50 45 C 40

L 35 30 25 20

IL-10, pg/m IL-10, 15 10 5 0 Day 3 Day 7 Day 10

Time Figure 2.3: IL-1β (A), IL-6 (B), and IL-10 (C) Cytokine Production from Macrophages and FBGCs Adherent to Photograft Polymers. Mean ± SEM, n=4.

- 59 - Chapter II

Day 7 160 Day 10 160 140 140 B A 120 120 100 100 80 80

IL-10, pg/mL IL-10, pg/mL 60 60

40 40

20 20

0 0 PET BDEDTC PAAm PAANa DMAPAAmMeI PET BDEDTC PAAm PAANa DMAPAAmMeI 35000 35000 30000 C 30000 D 25000 25000 20000 20000

IL-8, pg/mL 15000 IL-8, pg/mL 15000

10000 10000

5000 5000

0 0 PET BDEDTC PAAm PAANa DMAPAAmMeI PET BDEDTC PAAm PAANa DMAPAAmMeI 14000 14000

12000 12000 E F 10000 10000

, pg/mL 8000 , pg/mL 8000 β β 6000 6000 MIP-1 MIP-1 4000 4000 2000 2000 0 0 PET BDEDTC PAAm PAANa DMAPAAmMeI PET BDEDTC PAAm PAANa DMAPAAmMeI Material Surface Material Surface

Figure 2.4: Effect of IL-4 on IL-10 (A&B), IL-8 (C&D), and MIP-1β (E&F) Production at days 7 and 10. Closed: without IL-4. Open: with added IL-4. Mean ± SEM, n=4.

Chemokine Secretion from Adherent Cells

Chemokine concentrations for IL-8 and MIP-1β are shown in Figures

2.5.A & 2.5.B, respectively. IL-8 concentrations were greater than any other

cytokine/chemokine tested. At day 3, IL-8 levels were higher on PAAm (126,000

± 17,000 pg/mL), PAANa (142,000 ± 23,000 pg/mL), and DMAPAAmMeI

- 60 - Chapter II

(114,000 ± 31,000 pg/mL) in comparison to PET (63,000 ± 22,000 pg/mL) and

BDEDTC (41,000 ± 23,000 pg/mL). Concentrations of IL-8 were comparable on all materials at days 7 and 10. MIP-1β concentrations were greatest on PAANa

(21,900 ± 8,700 pg/mL) at day 3. PAAm (2,940 ± 1,130 pg/mL) and

DMAPAAmMeI (813 ± 415 pg/mL) promoted MIP-1β levels in comparison to PET

(285 ± 77 pg/mL) and BDEDTC (240 ± 80 pg/mL). At day 7, MIP-1β

concentrations were greatest on PAANa (2,150 ± 1,030 pg/mL) and least on

DMAPAAmMeI (324 ± 79 pg/mL). MIP-1β concentrations were promoted on

PAAm (1,190 ± 450 pg/mL) and PAANa (926 ± 330 pg/mL) in comparison to all

other materials at day 10. The addition of IL-4 to cultures did not significantly

affect IL-8 concentrations (Figures 2.4.C & 2.4.D). However, IL-4 addition increased MIP-1β concentrations on PET (5,260 ± 1,610 pg/mL), BDEDTC

(7,460 ± 4,420 pg/mL), PAAm (6,750 ± 1,690 pg/mL), and DMAPAAmMeI (3,170

± 1,420 pg/mL) resulting in comparable MIP-1βconcentrations on all materials at

day 7 (Figures 2.5.E & 2.5.F). An increase in MIP-1β levels with IL-4 addition

also occurred at day 10.

- 61 - Chapter II

PET BDEDTC PAAm PAANa DMAPAAmMeI 180000 160000 A 140000

120000

100000 80000

pg/mL IL-8, 60000 40000

20000

0 Day 3 Day 7 Day 10 35000 * 30000 B

25000 20000 , pg/mL

β 15000 10000 MIP-1 5000 * ^ 0 Day 3 Day 7 Day 10 Time Figure 2.5: Chemokine IL-8 (A) and MIP-1β (B) Expression over Time. “*” indicates a statistical difference in comparison to all other materials, while “^” indicates a statistical difference in comparison to PET. Mean ± SEM, n=4 (p<0.05).

- 62 - Chapter II

3500 PET BDEDTC PAAm PAANa DMAPAAmMeI 3000 A

β 2500

2000 1500

1000

Picograms of IL-1 IL-1 of Picograms perCell, Produced x10-5 500 0 Day3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4 120000

100000 B 80000

60000

40000

Picograms of IL-6 20000 per Cell,Produced x10-5

0 Day3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4

4000 3500 C 3000 ll, x10-5 2500

2000 1500

1000 ofIL-10 Picograms Produced per Ce 500 0 Day3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4 Time

Figure 2.6: Cellular activation as a Function of IL-1β (A), IL-6 (B) and IL-10 (C) Production per Cell. Mean ± SEM, n=3.

- 63 - Chapter II

Adherent Cell Activation

Cellular activation on each sample was analyzed based on the amount of

cytokines and chemokines produced by the adherent cells. This was calculated by dividing the cytokine or chemokine concentration in a sample (pg/mL) by the total number of adherent cells (cells / (mm2 * area)). Cells adherent to PAAm

produced the greatest amounts of IL-6, IL-1β, IL-10, IL-8 and MIP-1β per cell at

all timepoints as shown graphically in Figures 2.6 & 2.7. Activated cellular

expression of IL-6, IL-1β, and IL-10 increased on DMAPAAmMeI variably over

time, and MIP-1β production per cell was promoted on PAANa at day 3.

PET BDEDTC PAAm PAANa DMAPAAmMeI 2500000 A 2000000

1500000

1000000

500000 IL-8 of Picograms

Produced per Cell, x10-5 0 Day3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4

80000 70000 B 60000 β 50000

40000 30000 20000 10000 Produced per Cell, x10-5 Picograms of MIP-1 0 Day3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4 Time

Figure 2.7: IL-8 (A) and MIP-1β (B) Production per Cell. Mean ± SEM, n=3.

- 64 - Chapter II

Discussion

The objective of this study was to elucidate the cytokine and chemokine profiles produced by adherent macrophages and FBGCs on surfaces of distinct

surface chemistries. Macrophages and FBGC are known to secrete a variety of

cytokines and chemokines that can direct inflammation, wound healing,

angiogenesis, and fibrous encapsulation. Few studies have investigated the

material dependent nature of cytokine and/or chemokine production.4,6,23-25,28,41-43

The majority of these studies utilized an LPS stimulus, which is known to induce

a unique set of cytokines associated with classically activated macrophages.4,6,23-

25,42,43 None of these studies investigated the material dependent nature of

cytokine and chemokine proteomic profiles with time.

Material surface chemistry has been shown to play a role in dictating

macrophage adhesion and FBGC formation. The hydrophobic surfaces and the

hydrophilic/charged surfaces all promoted adhesion (see Figure 2.2)

substantiating the hypothesis that macrophages are adaptive with regard to

cellular adhesion on different surface chemistries. The hydrophilic/neutral

surfaces inhibited cellular adhesion confirming previous research in our

laboratory.28,39 As expected with the low cellular densities on the

hydrophilic/neutral surfaces, macrophage fusion was minimal to none. In

contrast, even though the remaining surfaces promoted equivalent amounts of

adhesion, percent fusion was variable between surface chemistries. The

hydrophilic/anionic surfaces promoted FBGC formation significantly more so at later timepoints, while the hydrophilic/cationic surfaces and the hydrophobic

- 65 - Chapter II surfaces supported comparable amounts of fusion. These findings demonstrate that macrophage fusion is more discriminatory than adhesion.

The cytokine array profile results show that the biomaterial adherent macrophages produce unique cytokine/chemokine profiles (Table 2.2). Initially, adherent macrophages produced cytokines (IL-6 and IL-1β) and chemokines (IL-

8, RANTES, ENA-78, and MCP-1) and down-regulated TARC similarly to classically activated macrophages. This suggests that the biomaterial induced activation state of the adherent cells at initial timepoints may be comparable to classically activated macrophages.9,19,20 However, these adherent macrophages did not secrete all the cytokines and chemokines (i.e. Mig, IP-10, GRO, and IL-

12) expected to be produced by classically activated macrophages and did produce cytokines and chemokines (i.e. IL-10, MDC, and Eotaxin-2) generally produced by alternatively activated macrophages.9,19-21 Collectively, this indicates that the biomaterial-adherent macrophages are not completely in a classically activated state initially. At later timepoints, adherent macrophages phenotypically switch to an activation state more comparable to alternatively activated macrophages by down-regulating IL-1β IL-6, IL-8, MIP-1β IL-12, Mig, and IP-10 while producing IL-10, MDC, and Eotaxin-2. Once again, the activation state of the biomaterial adherent macrophages is not completely similar to alternatively activated macrophages because it does not produce

TARC and continues to produce RANTES and MCP-1 as do classically activated macrophages.9,19,21 A phenotypic switch from an activation state similar to classically activated macrophages to one similar to alternatively activated

- 66 - Chapter II

macrophages is possible. Numerous studies have shown that IL-4, IL-13 and IL-

10 inhibit cytokine and chemokine production in macrophages previously activated by classical stimuli, IFN-γ and LPS.9,11-15,19,21,22,44,45 IL-10 is produced

by the adherent macrophages in this study and allows for a change in macrophage activation. The converse phenotypic switch is less likely due to the ability of IL-10 to inhibit the generation of classically activated macrophages.44

Collectively, these cytokine profiles do not clearly identify the biomaterial activated adherent macrophages as being comparable solely to classically or alternatively activated macrophages. These findings suggest that a phenotypic switch may occur over time and that activation is unique in biomaterial adherent macrophages.

A limitation of the cytokine array arises from each cytokine and chemokine being detected over a concentration range specific to that protein. Therefore, direct signal comparisons between cytokine/chemokine “spot” intensities are not indicative of a difference in production levels. Comparisons between the same cytokine and chemokine spot on multiple arrays can be utilized to provide an indication of potential sample concentration differences qualitatively. Differences in signal intensities were found between samples collected at different timepoints

(day 3 and 10), on different materials (PET, BDEDTC, PAAm, PAANa, and

DMAPAAmMeI), and in the presence vs. absence of IL-4 at day 10. These profiles prompted a proteomic analysis to identify the cytokine and chemokine levels with ELISA quantification.

- 67 - Chapter II

Initially, biomaterial adherent macrophages promoted pro-inflammatory cytokines IL-1β and IL-6 and chemokine IL-8, each of which was significantly down-regulated between days 3 and 7 and minimally produced by day 10

(Figures 2.3 & 2.5). This trend holds for all materials regardless of hydrophobic,

hydrophilic, or ionic characteristics. MIP-1β was initially promoted on the

hydrophilic/anionic surfaces and the hydrophilic/neutral surfaces while minimally

produced on the remaining surfaces with a similar decrease in concentration over

time. One potential explanation of this down-regulation of pro-inflammatory

cytokines and chemokines arises in the resolution of inflammation that occurs

naturally with time. A second arises in the production of the anti-inflammatory

cytokine IL-10, an alternatively activating macrophage stimulus. IL-10 was

initially produced and maintained over time by these adherent macrophages on

all material surfaces. IL-10 alternatively activates macrophages to suppress IL-6,

TNF-α, IL-1β, IL-8, MIP-1β, MIP-1α, Mig, RANTES, and IP-10.9,12,19,21,45,46 The

continued production of IL-10 over time may alternatively activate the adherent macrophages explaining the resulting decrease in production of IL-1β, IL-6, IL-8, and MIP-1β.

IL-4, a TH2 cytokine known to enhance macrophage fusion, inhibit pro-

inflammatory cytokine (i.e. IL-1, IL-6, TNF) and MIP-1β production, and promote

IL-10 production, was added to select adherent cell cultures.9,19,20 IL-4

stimulated the adherent cells to produce more IL-10 on the hydrophilic/cationic

surfaces and to promote MIP-1β on the hydrophobic surfaces, the

hydrophilic/neutral surfaces, and the hydrophilic/cationic surfaces (Figure 2.4).

- 68 - Chapter II

IL-4 did not significantly affect the production of any other cytokine or chemokine

tested as expected. Also, an equal addition of IL-4 would be expected to

stimulate the adherent cells on all materials equally; however this was not the case. These findings suggest that biomaterial surface chemistry plays a strong role in macrophage activation.

This study was carried out to determine the presence of a material dependency in cytokine and/or chemokine expression from adherent

macrophages and FBGCs. The cytokine and chemokine concentrations

quantified using ELISAs did not show distinct patterns of material dependency

and were for the most part comparable on all surfaces at each timepoint for the

majority of the proteins investigated. In contrast, the cellular densities were

significantly lower on the hydrophilic/neutral surface in comparison to the other

surfaces. An analysis of the amount of cytokine/chemokine produced per

adherent cell showed that although there were fewer macrophages on the

hydrophilic/neutral surface, these adherent cells produced significantly greater

amounts of cytokines and chemokines than the other surfaces (Figures 2.6 &

2.7). At earlier timepoints, activated macrophages adherent to the

hydrophilic/neutral surfaces produced on average 30 times the amount of IL-6

than adherent cells on the other surfaces. This promotion of cellular activity on

the hydrophilic/neutral surfaces in comparison to the other surfaces was also

seen in the increased production of IL-1β (12x), IL-10 (9x), IL-8 (12x), and MIP-

1β (52x) at day 3. Over time, the adherent cells on the hydrophilic/neutral

surfaces continued to be activated to produce greater amounts of IL-6 (20x), IL-

- 69 - Chapter II

10 (30x), IL-8 (26x), and MIP-1β (58x) by day 10 than the other material

surfaces. Hwang et. al. found similar results that showed monocytes adhered

less to a poly-L-lysine multilayer film and a hyaluronic acid multilayer film in

comparison to tissue culture polystyrene, yet still promoted more TNF-α production47. Collectively, this study shows that although the surrounding milieu

experienced a similar expression profile on each of the material surfaces; the

adherent macrophages on the hydrophilic/neutral surfaces were consistently

more actively producing these proteins.

This study confirms that material surface chemistry can differentially affect

monocyte/macrophage/FBGC adhesion, macrophage fusion into FBGCs,

macrophage activation, and cytokine/chemokine profiles produced by biomaterial

adherent macrophages and FBGCs. Particularly, hydrophilic/neutral surfaces

inhibit macrophage adhesion and fusion, while causing the adherent cells to

produce greater amounts of cytokines and chemokines than their hydrophobic

and hydrophilic/ionic counterparts. It also demonstrates the pleiotrophic nature

of macrophages and suggests that adherent macrophages may undergo a

phenotypic switch in culture over time. The identification of diverse cytokine and

chemokine profiles secreted by biomaterial adherent macrophages/FBGCs

further substantiates the complexity of the cytokine and chemokine network

secreted by macrophages/FBGCs. By using a broad proteomic analysis

approach, we have ascertained numerous adherent macrophage/FBGC-derived

cytokines and chemokines to further investigate for their roles in macrophage

- 70 - Chapter II

adhesion, fusion, and recruitment/activation of other inflammatory and wound

healing cells (i.e. fibroblasts, neutrophils, and endothelial cells).

References

1. Anderson JM. Inflammatory response to implants. ASAIO 1988;11(2):101-

107.

2. Anderson JM. Mechanisms of Inflammation and Infection with Implanted

Devices. Cardiovascular Pathology 1993;2(3):33S-41S.

3. Luttikheizen D, Harmsen M, Van Luyn M. Cellular and molecular dynamics

in the foreign body reaction. Tissue Engineering 2006;12(7):1955-1969.

4. Miller K, Rose-Caprara V, Anderson JM. Generation of IL1-like activity in

response to biomedical polymer implants: A comparison of in vitro and in

vivo models. Journal of Biomedical Materials Research 1989;23:1009-

1026.

5. Miller K, Anderson JM. In vitro stimulation of fibroblast activity by factors

generated from human monocytes activated by biomedical polymers.

Journal of Biomedical Materials Research 1989;23:911-930.

6. Bonfield T, Colton E, Anderson JM. Fibroblast stimulation by monocytes

cultured on protein adsorbed biomedical polymers. I. Biomer and

polydimethylsiloxane. Journal of Biomedical Materials Research

1991;25:1650175.

- 71 - Chapter II

7. Song E, Ouyang N, Horbelt M, Antus B, Wang M, Exton M. Influence of

alternatively and classically activated macrophages on fibrogenic activities

of human fibroblasts. Cellular Immunology 2000;204:19-28.

8. Gretzer C, Emanuelsson L, Liljensten E, Thomsen P. The inflammatory

cell influx and cytokine changes during transition from acute inflammation

to fibrous repair around implanted materials. Journal of Biomaterial

Science Polymer Edition 2006;17(6):669-687.

9. Gordon S. Alternative activation of macrophages. Nature Reviews:

Immunology 2003;3(1):23-35.

10. Gosain A, Gamelli R. A primer in cytokines. Journal of Burn Care

Rehabilitation 2005;26(7):7-12.

11. Donnelly R, Fenton M, Finbloom D, Gerrard T. Differential regulation of IL-

1 production in human monocytes by IFN-g and IL-4. Journal of

Immunology 1990;145(2):569-575.

12. de Waal Malefyt R, Abrams J, Bennett B, Figdor C, de Vries J. Interleukin-

10 (IL-10) inhibits cytokine synthesis by human monocytes: an

autoregulatory role of IL-10 produced by macrophages. Journal of

Experimental Medicine 1991;174:1209-1220.

13. de Waal Malefyt R, Figdor C, Huijbens R, Mohan-Peterson S, Bennett B,

Culpepper J, Dang W, Zurawski G, de Vries J. Effects of IL-13 on

phenotype, cytokine production, and cytotoxic function of human

monocytes: comparison with IL-4 and modulation by IFN-g or IL-10.

Journal of Immunology 1993;151(11):6370-6381.

- 72 - Chapter II

14. Bonecchi R, Sozzani S, Stine J, Luini W, D'Amico G, Allavena P, Chantry

D, Mantovani A. Divergent effects of interleukin-4 and interferon-gamma

on macrophage-derived chemokine production: an amplification circuit of

polarized T helper 2 responses. Blood 1998;92(8):2668-2671.

15. Fenton M, Buras J, Donnelly R. IL-4 reciprocally regulates IL-1 and IL-1

receptor antagonist expression in human monocytes. Journal of

Immunology 1992;149(4):1283-1288.

16. Rollins B. Chemokines. Blood 1997;90(3):909-928.

17. Baggiolini M, Dewald B, Moser B. Human chemokines: an update. Annual

Review of Immunology 1997;15:675-705.

18. Charo I, Ransohoff R. The many roles of chemokines and chemokine

receptors in inflammation. The New England Journal of Medicine

2006;354(6):610-754.

19. Mantovani A, Sica A, Sozzani S, Allavena P, Vecchi A, Locati M. The

chemokine system in diverse forms of macrophage activation and

polarization. Trends in Immunology 2004;25(12):677-685.

20. Mosser D. The many faces of macrophage activation. Journal of

Leukocyte Biology 2003;73:209-212.

21. Mantovani A, Sozzani S, Locati M, Allavena P, Sica A. Macrophage

polarization: tumor-associated macrophages as a paradigm for polarized

M2 mononuclear phagocytes. Trends in Immunology 2002;23(11):549-

555.

- 73 - Chapter II

22. Bonder C, Finlay-Jones J, Hart P. Interleukin-4 regulation of human

monocyte and macrophage interleukin-10 and interleukin-12 production.

Role of a functional interleukin-2 receptor gamma-chain. Immunology

1999;96:529-536.

23. Bonfield T, Colton E, Marchant R, Anderson JM. Cytokine and growth

factor production by monoyctes/macrophages on protein preadsorbed

polymers. Journal of Biomedical Materials Research 1992;26:837-850.

24. Yun J, DeFife K, Colton E, Stack S, Azeez A, Cahalan L, Verhoeven M,

Cahalan P, Anderson JM. Human monocyte/macrophage adhesion and

cytokine production on surface-modified

poly(tetrafluoroethylene/hexafluoropropylene) polymers with and without

protein preadsorption. Journal of Biomedical Materials Research

1995;29:257-268.

25. Bonfield T, Anderson JM. Functional versus quantitative comparison of IL-

1beta from monocytes/macrophages on biomedical polymers. Journal of

Biomedical Materials Research 1993;27:1195-1199.

26. Gretzer C, Gisselfalt K, Liljensten E, Ryden L, Thomsen P. Adhesion,

apoptosis and cytokine release of human mononuclear cells cultured on

degradable poly(urethane urea), polystyrene, and titanium in vitro.

Biomaterials 2003;24:2843-2852.

27. Marques A, Reis R, Hunt J. Cytokine secretion from mononuclear cells

cultured in vitro with starch-based polymers and poly-L-lactide. Journal of

Biomedical Materials Research 2004;71A:419-429.

- 74 - Chapter II

28. Brodbeck W, Nakayma Y, Matsuda T, Colton E, Ziats N, Anderson JM.

Biomaterial surface chemistry dictates adherent monocyte/macrophage

cytokine expression in vitro. Cytokine 2002;18(6):311-319.

29. Brodbeck W, Voskerician G, Ziats N, Nakayama Y, Matsuda T, Anderson

JM. In vivo leukocyte cytokine mRNA responses to biomaterials are

dependent on surface chemistry. Journal of Biomedical Materials

Research 2003;64A:320-329.

30. Zhu H, Snyder M. 'Omic' approaches for unraveling signaling networks.

Current Opinion in Cell Biology 2002;14:173-179.

31. Walduck A, Rudel T, Meyer T. Proteomic and gene profiling approaches to

study host responses to bacterial infection. Current Opinion in

Microbiology 2004;7:33-38.

32. Nakayama Y, Anderson JM, Matsuda T. Laboratory-scale mass

production of a multi-micropatterned grafted surface with different polymer

regions. Journal of Biomedical Materials Research 2000;53(5):584-591.

33. Nakayama Y, Matsuda T. Surface macromolecular artchitectural designs

using photograft copolymerization based on photochemistry of benzyl

N,N-diethyldithiocarbamate. Macromolecules 1996;29:8622-8630.

34. McNally AK, Anderson JM. Complement C3 participation in monocyte

adhesion to different surfaces. Proceedings of the National Academy of

Science 1994;91:10119-10123.

- 75 - Chapter II

35. Mirzadeh H, Dadsetan M, Sharifi-Sanjani N. Platelet adhesion on laser-

induced acrylic acid-grafted polyethylene terephthalate. Journal of Applied

Polymer Science 2002;86:3191-3196.

36. Kalasinsky K, Kalasinsky V. Infrared and raman microspectroscopy of

foreign materials in tissue specimens. Spectrochimica Acta Part A

2005;61:1707-1713.

37. Yang J, Huang M, Yeh T. Preparation of poly(acrylic acid) modified

polyurethane membrane for biomaterial by UV radiation without

degassing. Journal of Biomedical Materials Research 1999;45:133-139.

38. Brodbeck W, Patel J, Voskerician G, Christenson E, Shive M, Nakayama

Y, Matsuda T, Ziats N, Anderson JM. Biomaterial adherent macrophage

apoptosis is increased by hydrophilic and anionic substrates in vivo.

Proceedings of the National Academy of Science 2002;99(16):10287-

10292.

39. Brodbeck W, Shive M, Colton E, Nakayama Y, Matsuda T, Anderson JM.

Influence of biomaterial surface chemistry on the apoptosis of adherent

cells. Journal of Biomedical Materials Research 2001;55:661-668.

40. Nakayama Y, Matsuda T. Surface macromolecular architectural designs

using photograft copolymerization based on photochemistry of benzyl

N,N-diethyldithiocarbamate. Macromolecules 1996;29:8622-8630.

41. Miller K, Huskey R, Bigby L, Anderson JM. Characterization of biomedical

polymer-adherent macrophages: interleukin-1 generation and scanning

electron micrscopy studies. Biomaterials 1989;10:187-196.

- 76 - Chapter II

42. DeFife K, Yun J, Azeez A, Stack S, Ishihara K, Nakabayashi N, Colton E,

Anderson JM. Adhesion and cytokine production by monocytes on poly(2-

methacryloyoxyethyl phosphorylcholine-co-alkyl methacrylate)-coated

polymers. Journal of Biomedical Materials Research 1995;29:431-439.

43. Miller K, Anderson JM. Human monocyte/macrophage activation and

interleukin 1 generation by biomedical polymers. Journal of Biomedical

Materials Research 1988;22:713-731.

44. Katakura T, Miyazaki M, Kobayaski M, Herndon D, Suzuki F. CCL17 and

IL-10 as effectors that enable alternatively activated macrophages to

inhibit the generation of classically activated macrophages. Journal of

Immunology 2004;172:1407-1413.

45. Oswald I, Wynn T, Sher A, James S. Interleukin 10 inhibits macrophage

microbicidal activity by blocking the endogenous production of tumor

necrosis factor alpha required as a costimulatory factor for interferon-

gamma induced activation. Proceedings of the National Academy of

Science 1992;88:8676-8680.

46. Kopydlowski K, Salkowski C, Cody M, van Rooijen N, Major J, Hamilton T,

Vogel S. Regulation of macrophage chemokine expression by

lipopolysaccharide in vitro and in vivo. Journal of Immunology

1999;163:1537-1544.

47. Hwang J, Jelacic S, Samuel N, Maier R, Campbell C, Castner D, Hoffman

A, Stayton P. Monocyte activation on polyelectrolyte multilayers. Journal

of Biomaterial Science Polymer Edition 2005;16(2):237-251.

- 77 - Chapter III

Chapter III: Matrix Metalloproteinases and Their Inhibitors in the Foreign Body Reaction on Biomaterials

Abstract

Matrix metalloproteinases (MMPs) can degrade structural components within the extracellular matrix and at the cellular surface producing changes in cellular behavior (i.e. adhesion and migration) and subsequent pathological responses (i.e. the foreign body reaction and wound healing). We continue to study the foreign body reaction that occurs following biomaterial implantation by investigating secretory responses of biomaterial-adherent macrophages and foreign body giant cells (FBGCs) as directed by material surface chemistry and further this research by determining if secreted MMPs play a role in macrophage adhesion and fusion. We have identified numerous MMPs and their tissue inhibitors (TIMPs) in in vitro cell culture supernatants using antibody arrays and quantified select MMP/TIMPs with ELISAs. MMP-9 concentrations were significantly greater than both TIMP-1 and TIMP-2 on all materials. The ratios of

MMP-9/TIMP-1 and MMP-9/TIMP-2 increased with time due to an increase in

MMP-9 concentrations over time, while the TIMP concentrations remained constant. Total MMP-9 concentrations in the supernatants were comparable on all materials at each timepoint, while TIMP-1 and TIMP-2 concentrations tended to be greater on hydrophilic/anionic surfaces. Analysis of the MMP/TIMP quantities produced per cell revealed that the hydrophilic/neutral surfaces, which inhibited macrophage adhesion, activated the adherent macrophages/FBGCs to produce a greater quantity of MMP-9, TIMP-1, and TIMP-2 per cell.

- 78 - Chapter III

Pharmacological inhibition of MMP-1,-8,-13, and MMP-18 reduced macrophage fusion without affecting adhesion, while inhibitors of MMP-2,-3,-9, and MMP-12 did not affect adhesion or fusion. These findings demonstrate that material surface chemistry does modulate macrophage/FBGC-derived MMP/TIMP secretion and implicates MMP involvement in macrophage fusion.

Introduction

Matrix metalloproteinases (MMP) are a family of zinc-dependent proteolytic enzymes capable of degrading the structural protein components within extracellular matrix (ECM) and at the cellular surface.1 MMPs are involved in numerous biological processes including wound healing, the foreign body reaction, inflammation, angiogenesis, and bone remodeling.2-4 Implantation of biomedical devices, biomaterials, and prostheses induces the foreign body reaction including multinucleated foreign body giant cells (FBGCs) and their predecessors, monocyte-derived macrophages. Proteolytic MMPs may play a role in modulating macrophage adhesion and subsequent fusion. In addition, the active secretion of MMPs by adherent macrophages and FBGCs may modulate wound healing, the foreign body reaction, angiogenesis, and fibrous encapsulation surrounding a biomaterial implant.

There are over 25 known MMPs. The majority are classified into four secreted classes (collagenase, gelatinase, stromelysin, and matrilysins), while six MMPs are classified into one of the two membrane-bound classes (type 1 transmembrane bound and glycosylphosphatidylinositol (GPI) anchored).1

- 79 - Chapter III

These endopeptidases perform multiple and sometimes overlapping functions as

shown in Table 3.1. MMPs cleave structural components of the ECM (i.e.

collagen, gelatins, elastin, laminin) breaking physical barriers and allowing for

cellular migration, growth, proliferation, and matrix remodeling.1,5-8 In addition,

MMPs have been shown to regulate the activities of cytokines and growth factors

such as tumor necrosis factor-alpha (TNF-α), interleukin-1beta (IL-1β), and

transformating growth factor-beta (TGF-β) by cleaving the surrounding ECM

components that physically sequester them.6,9 By cleaving cell surface markers,

MMPs are capable of disrupting cell-cell adhesions and/or cell-surface

adhesions, release molecules (i.e. TNF-α via the cleavage of pro-TNF-α) and

other proteases (MMP-2) bound to the cellular surface, and/or modulate

apoptosis by releasing surface-bound soluble Fas ligand to induce apoptosis or

survival factors (i.e. EGF and PDGF) to inhibit apoptosis.5,8

Regulation of MMPs occurs at both the transcriptional level and after

secretion from the variety of cells (i.e. macrophage, fibroblasts, smooth muscle

cells) that produce MMPs. Expression of MMPs is up- or down-regulated by

numerous factors, including cytokines (i.e. TNF, IL-1, IL-10, IL-4, TGF-β, EGF)

and integrin clustering further complicating the interactions of MMPs.7,10-15 After secretion, MMPs can be regulated by tissue inhibitors of MMPs (TIMPs), proteolytic inactivation, or by clearance via binding to a receptor or protein (i.e.

α2-macroglobin and thrombospondin) that is endocytosed through the low- density lipoprotein receptor-related protein (LRP) scavenger receptor.5,6,16

- 80 - Chapter III

Table 3.1 Matrix Metalloproteinases and their Functions* Classifications Proteins Functions^ Collagenases MMP-1, MMP-8, - Cleaves collagen type I, II, and III MMP-13, & MMP-18 - Digests other ECM and non-ECM components Gelatinases MMP-2 & MMP-9 - Digest gelatins and other ECM components - Binds gelatin, collagen, and laminin - MMP-2 digests collagen type I, II, and III Stomelysins MMP-3 & MMP-10 - Digests stromal proteoglycans and other ECM components - Activates proMMPs Matrilysins MMP-7 & MMP-26 - Digests numerous ECM components - Processes cell surface molecules (i.e. pro-a-defensin, Fas-ligand, pro-TNF-a, and E-cadherin) Membrane- Transmembrane - Activates proMMP-2 Type MMPs MMP-14, MMP-15, (except MMP-17) MMP-16, & MMP- - Digests numerous ECM 24 components GPI Anchored - MMP-14 digests collagen type I, II, MMP-17 & MMP-25 and III *Adapted from references [1,3,7,8] MMP = matrix metalloproteinase; GPI = glycosylphosphatidylinositol ^ECM components can include: collagen, gelatin, aggrecan, laminin, and proteoglycans

Numerous cells including MMP secreting cells are capable of up-or down- regulating the production of TIMPs that bind to MMPs in a 1:1 stoichiometry inhibiting their proteolytic activities in the local environment.1 There are four known TIMPs (TIMP-1, -2, -3, -4) that share structural homology and are individually capable of binding the majority of known MMPs.1 A balance between

MMPs and TIMPs has been hypothesized to maintain the homeostasis of the surrounding ECM, while an inbalance would modulate the remodeling of the

- 81 - Chapter III surrounding ECM and may further be implicated in formation of a fibrotic capsule surrounding biomaterial implants.

Our laboratory continues to address the hypothesis that material surface chemistry modulates adherent macrophage and FBGC behavior. Recently, we have demonstrated that material surface chemistry affects macrophage/FBGC secretion of cytokines and chemokines in our in vitro monocyte-derived macrophage and FBGC culture system.17 We aim to advance this research by investigating macrophage/FBGC secretion of MMPs and TIMPs and their role in macrophage adhesion and fusion using the same in vitro cell culture system.

The initial objective of our study was to identify the MMPs and TIMPs produced by biomaterial-adherent macrophages and FBGCs using our in vitro monocyte- derived macrophage and FBGC culture system, model polymers of distinct surface chemistry, and proteomic analysis techniques such as human antibody arrays and ELISAs. Concurrently, a second goal of this research was to determine the effects of material surface chemistry on these secreted MMP/TIMP profiles. Our final goal was to determine if MMPs play a role in macrophage adhesion and IL-4-induced FBGC formation on RGD-modified substrates through the use of four pharmacological inhibitors of MMPs.

Material and Methods

RGD-Modified Surfaces

Proteomic analysis of cell culture supernatants and the pharmacological inhibition experiments utilized RGD-modified surfaces that have been shown to support macrophage adhesion and fusion.18 The RGD-modification is a

- 82 - Chapter III

fibronectin-like engineered protein polymer (Sigma, F-5022) consisting of 13

identical copies of the ten amino acid sequence (VTGRGDSPAS) utilized for cell

attachment in human fibronectin.19 This sequence is flanked on either side by

repeated structural peptide sequences (GAGAGS)9 that form a hairpin structure

containing the cell attachment epitope for presentation at the material surface in

the quaternary structure. Sterile, 24-well tissue culture polystyrene (TCPS)

plates pre-coated with RGD were utilized in the proteomic analysis experiments, while the inhibition studies used pre-coated 96-well tissue culture polystyrene

(TCPS) plates.

Biomaterial Surfaces of Distinct Chemistries

Photografted polymerized surfaces were provided by Dr. Takehisa

Matsuda at Kyushu University in Fukuoka, Japan. The synthesis and mass production of these materials previously has been described in depth elsewhere.20,21 Briefly, polyethylene terephthalate (PET) film is coated with

poly(styrene-co-benzyl N,N-dimethyldithiocarbamate) (BDEDTC), which is then

photopolymerized with acrylamide, AAm; the sodium salt of acrylic acid, AANa;

or the methyl iodide salt of N-[3-(dimethylamino) propyl] acrylamide,

DMAPAAmMeI. A total of five surfaces of distinct chemistry were utilized in the proteomic analysis experiments in this study: PET, hydrophobic; BDEDTC, hydrophobic; PAAm, hydrophilic and neutral; PAANa, hydrophilic and anionic;

DMAPAAmMeI, hydrophilic and cationic. Material films were cut into 15-mm

diameter discs and sterilized using 100% ethanol just prior to insertion into 24-

- 83 - Chapter III

well tissue culture polystyrene (TCPS) plates. Pre-cut, sterile discs were secured

within TCPS plates with silicone rings, previously sonicated in 100% ethanol for 5

minutes and autoclaved for 1 hour. Surfaces were incubated in warmed, sterile

Dulbecco’s phosphate buffered saline (Gibco, Grand Island, NY) containing magnesium chloride and calcium chloride (PBS++) to remove any residual

ethanol from material sterilization. The resultant surface area was 0.71 cm2.

Human Monocyte/Macrophage/FBGC In Vitro Culture System

Human monocytes and serum were obtained from the whole blood of non-

medicated healthy donors using a centrifugation, Ficoll, and Percoll method

described previously by McNally, et. al.22 In the experiments addressing

MMPs/TIMPs secretion, freshly isolated monocytes were plated onto each

surface at a concentration of 5x105 cells in 0.5mL of macrophage serum-free

media (SFM) (Gibco, Grand Island, NY) containing 20% autologous serum (AS).

After 2 hours, non-adherent cells were removed using a 1mL wash of warm,

sterile PBS++. Fresh media of SFM with 20% heat-treated, autologous serum

o (HAS) were added and cultures were continued at 37 C and 5% CO2 for 3, 7,

and 10 days. Serum was heat-treated at 56oC for 1 hour. At each timepoint,

supernatants containing non-adherent cells were collected via pipetting.

Adherent cells in cultures to be terminated were fixed using 100% methanol for 5

minutes. Any remaining cultures continued to be cultured in fresh media

containing 20% HAS. Non-adherent cells were removed from the supernatants

via centrifugation at 10x103 rpm. Cell-free supernatants were aliquoted to

- 84 - Chapter III

minimize thermal cycles and stored at -20oC for future analysis with protein arrays or ELISAs.

Detection of MMPs and TIMPs using Antibody Arrays

Eleven MMPs and TIMPs were investigated for their presence or absence

in the cell culture supernatants at day 7 using a human antibody array system

purchased from RayBiotech, Inc. (Catalog #AAH-MMP-1, Norcross, GA) in

accordance with the manufacturer’s instructions and were imaged using a

BioRad chemiluminescence imaging system (Hercules, CA). Qualitative data was determined based upon the signal intensity at the location or “spot” of a given MMP/TIMP. Signal intensity was rated as either no signal “-”, weak “+”, medium “++”, or strong “+++”. The detectable concentration range was specific to each MMP/TIMP analyzed; therefore a direct comparison between different

MMPs/TIMPs on the same membrane analyzed was not feasible. Comparisons between multiple membranes for the sample MMP/TIMP were made to determine potential differences between materials. MMP/TIMPs that had positive results in the protein arrays were selected to be further quantified with ELISA assays.

Quantification of MMP/TIMP Concentrations using ELISA

MMP/TIMP concentrations in cell culture supernatants were quantitatively determined using human ELISA assays (R&D Systems, Norcross, GA).

Supernatants tested were collected after 3, 7 and 10 days of culture. ELISA kits

- 85 - Chapter III

were used in accordance with manufacturer’s instructions with SFM as the

diluent for both cell culture samples and standard samples. The ELISA kit sensitivities (minimum detectable dose) were 0.156 ng/mL for MMP-9, 0.08

ng/mL for TIMP-1, and 0.011 ng/mL for TIMP-2. Quantified data is presented as

the total concentration (pg/mL) of the tested MMP/TIMP.

Pharmacological Inhibition of MMPs

Human monocytes were isolated and cultured as previously described.22,23

Briefly, for inhibitor studies, 2 x 105 monocytes were added per well to 96-well

RGD-modified culture plates (Smart Plastic, ICN, Irvine, CA) in 0.l ml of serum free medium for macrophages (SFM, Invitrogen, Grand Island, NY) supplemented with 20% autologous serum. After 1.5 h, nonadherent cells were removed by washing with PBS containing Ca++ and Mg++ (PBS++, Invitrogen) at

37oC, recovered with 0.2 ml of unsupplemented SFM, and incubated for 3 days

o at 37 C in a humidified atmosphere of 95% air and 5% CO2. On day 3, the

medium was replaced with 0.2 ml of SFM without or with inhibitors and with 15

ng/ml recombinant human IL-4 (R&D Systems, Minneapolis, MN). The

macrophages were then incubated until day 8, when the plates were washed

twice with PBS++ at 37oC, fixed with methanol, and stained with May-

Grünwald/Giemsa for evaluation of macrophage/FBGC morphology, % adhesion,

and % macrophage fusion. The pharmological inhibitors used were (1)

epigallocatechin gallate (ECG), which inhibits MMP-2, -9 and-12; (2) N-Isobutyl-

N-(4-methoxyphenylsulfonyl)-glycylhydroxamic acid (NNGH), an inhibitor of

- 86 - Chapter III

MMP-3 and MMP-12; (3) CL-82198, known to inhibit MMP-13 but not MMP-1 or

MMP-9; and (4) actinonin, an inhibitor of aminopeptidases such as the

collagenase MMPs (MMP-1,-8,-13, and -12).

Determination of Adherent Cell Surface Densities

Adherent cell surface density analysis was conducted for each

experiment. Methanol-fixed adherent cells were stained serially with May-

Grünwald and Giemsa.24 Adherent cell densities were determined in 5-20x objective areas (463 x 463 μm) using optical microscopy. Data is presented as the average total number of adherent cells in mm2 and the average percentage of nuclei in FBGCs of the total number of adherent nuclei (% fusion).

Statistical Analysis

Each experiment was repeated using cells isolated from four different human donors to account for donor variability (n=4). Statistical significance was determined using a two-way ANOVA and Tukey’s pairwise test using a 95% confidence level using Minitab 14 statistical software package (Minitab, Inc.,

State College, PA). All data is presented as the mean ± standard error of the mean (SEM) with an n=4.

- 87 - Chapter III

Results

MMP/TIMP Detected within In Vitro Cell Culture using Antibody Arrays

As shown in Table 3.2, MMP-9, TIMP-1 and TIMP-2 were strongly detected in cell cultures on all biomaterials. The detection signal was weak for

MMP-8 on BDEDTC and RGD and for MMP-10 on RGD; however MMP-8 and

MMP-10 were not detected in the cell culture supernatants from any of the other surfaces. MMP-1, MMP-2, MMP-3, MMP-13, TIMP-3, and TIMP-4 were not detected in the cell culture supernatants. MMP-9, TIMP-1, and TIMP-2 were chosen to be quantified using ELISA based on their strong presence in the supernatants. No MMPs/TIMPs were detected in serum free media (SFM) only.

TIMP-2, TIMP-1, MMP-9, MMP-8, and MMP-10 were detected in the SFM with

20% heat-treated, autologous serum (HAS).

Table 3.2 MMPs/TIMPs Detected in Cell Cultures at Day 7 using Antibody Arrays Signal Intensity in Cell Culture Supernatants Protein Biomaterials Media Tested Controls BDEDTC PAAm PAANa DMAPAAmMeI RGD SFM SFM + only serum MMP-1 ------MMP-2 ------MMP-3 ------MMP-8 + - - - + - + MMP-9 +++ +++ +++ +++ +++ - ++ MMP-10 - - - - + - + MMP-13 ------TIMP-1 ++ ++ ++ ++ +++ - ++ TIMP-2 +++ +++ +++ +++ +++ - +++ TIMP-3 ------TIMP-4 ------‘-’ indicates no signal. ‘+’ indicates a weak signal. ‘++’ indicates a medium strength signal. ‘+++’ indicates a strong signal.

- 88 - Chapter III

Production of Matrix Metalloproteinases and their Inhibitors

In initial experiments analyzing MMP/TIMP profiles on RGD-substrates

only (Table 3.3), MMP-9 levels were significantly greater than both TIMP-1 and

TIMP-2 levels, significantly increased between days 3 and 7, and decreased with

IL-4 addition at both days 7 and 10. Both TIMP-1 and TIMP-2 concentrations on

RGD-substrates remained comparable over time.

Table 3.3 MMP/TIMP Concentrations in Cell Cultures with RGD Substrate Over Time (ng/mL) Protein Day 3 Day 7 Day 7 Day 10 Day 10 with IL-4 with IL-4 MMP-9 720±286* 4413±925* 1414±632* 3307±470* 386±68*^ TIMP-1 104±19 167±112 114±26 74±15 74±14 TIMP-2 0±1 25±10 26±10 39±4 30±8 “*” indicates statistical difference between MMP-9 and both TIMP-1 and TIMP-2 at that timepoint and condition, while “^” denotes a statistical difference between with and without IL-4 at day 10 (p<0.05).

Similarly, MMP-9 concentrations were significantly greater than TIMP-1 or

TIMP-2 concentrations on all biomaterial surface chemistries at all timepoints

(Figure 3.1). MMP-9 levels were comparable on all materials at each timepoint

and collectively increased between days 3 and 7 (Figure 3.1.A). TIMP-1 concentrations were comparable on all materials except for PAANa, which promoted TIMP-1 at day 3 (80 ± 37 ng/mL) and day 10 (88 ± 61 ng/mL) as shown in Figure 3.1.B. TIMP-2 levels (Figure 3.1.C) were comparable on all materials at days 3 and 10 and were greatest on PAANa at day 7 (42 ± 7 ng/mL).

The media controls contained 38±10 ng/mL of MMP-9, 18±2 ng/mL of TIMP-1, and 29±4 ng/mL of TIMP-2, which were subtracted from the total concentrations to yield the presented results.

- 89 - Chapter III

4000 PET BDEDTC PAAm PAANa DMAPAAmMeI 3500 A

3000 L 2500 2000

1500

MMP-9, ng/m 1000 500 0 Day 3 Day 7 Day 10

160 140 B

L 120

100

80 60 TIMP-1, ng/m 40

20 0 Day 3 Day 7 Day 10 60 C 50

20 TIMP-2, ng/mL 10

0 Day 3 Day 7 Day 10 Time

Figure 3.1: Macrophage/FBGC Secretion of Matrix Metalloproteinase-9 (A) and Tissue Inhibitors of Matrix Metalloproteinases, TIMP-1 (B) and TIMP-2 (C). Mean ± SEM, n=4.

- 90 - Chapter III

Day 7 Day 10 * 4000 4000 * 3500 A * 3500 B 3000 3000

2500 2500 *

2000 2000

ng/mL MMP-9, 1500 ng/mL MMP-9, 1500

1000 1000

500 500

0 0 PET BDEDTC PAAm PAANa DMAPAAmMeI PET BDEDTC PAAm PAANa DMAPAAmMeI

160 160 140 140 C D 120 120 100 100 80 80 60

60 ng/mL TIMP-1, TIMP-1, ng/mL 40 40 20 20 0 0 PET BDEDTC PAAm PAANa DMAPAAmMeI PET BDEDTC PAAm PAANa DMAPAAmMeI

70 70

60 E 60 F

50 50 L 40 40 30 30 TIMP-2, ng/m TIMP-2, 20 TIMP-2, ng/mL 20 10 10 0 0 PET BDEDTC PAAm PAANa DMAPAAmMeI PET BDEDTC PAAm PAANa DMAPAAmMeI Material Surface Material Surface

Figure 3.2: Effect of IL-4 on MMP-9 (A&B), TIMP-1 (C&D), TIMP-2 (E&F) Production at Days 7 and 10. Open: with added IL-4. Closed: without IL-4. “*” indicates statistical difference between with and without IL-4. Mean ± SEM, n=4.

- 91 - Chapter III

Addition of IL-4 to cultures tended to decrease MMP-9 concentrations on the majority of the materials at day 7, with a statistical decrease in MMP-9 concentrations on PAANa (Figure 3.2.A). This decrease in MMP-9 concentration with IL-4 addition was statistically significant by day 10 on

BDEDTC, PAAm and PAANa (Figure 3.2.B). IL-4 addition did not significantly affect TIMP-1 (Figures 3.2.C & 3.2.D) or TIMP-2 (Figures 3.2.E & 3.2.F) concentration levels at the later timepoints of days 7 and 10.

Cellular Activation

Previous studies have shown that the hydrophilic/neutral surface PAAm inhibits cellular adhesion in comparison to the hydrophobic or hydrophilic/ionic surface chemistries.25,26 Concentrations of MMP and TIMPs produced per cell

were analyzed to determine the material surface chemistry effects on cellular

activation as measured by MMP and TIMPs secretion. The production of MMP

or TIMP per cell was calculated by dividing the total concentration of MMP or

TIMP in a sample (ng/mL) by the total number of adherent cells on the material

surface ((cell/mm2)*area). Non-adherent cells were not accounted for in this calculation because the non-adherent cell population is potentially reduced over time by the removal of the supernatants containing non-adherent cells at each timepoint. As shown in Figure 3.3, PAAm greatly promoted MMP-9, TIMP-1, and TIMP-2 quantities produced per cell in comparison to all other materials at regardless of time or IL-4 addition.

- 92 - Chapter III

70000000 PET BDEDTC PAAm PAANa DMAPAAmMeI

60000000 A x10-5 50000000 ll,

40000000

30000000 20000000

Picograms of MMP-9 Produced per Ce 10000000

0 Day3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4

1200000 B 1000000 800000

600000 400000

Picograms of TIMP-1 200000 x10-5 Cell, per Produced 0 Day3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4

1200000 C 1000000

800000

600000

400000

Picograms of TIMP-2 Produced per Cell, x10-5 Cell, per Produced 200000

0 Day3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4 Time

Figure 3.3: MMP-9 (A), TIMP-1 (B), and TIMP-2 (C) Production per Cell. Mean ± SEM, n=3.

- 93 - Chapter III

Effects of Pharmacological Inhibition of MMPs on Adhesion and FBGC Formation

Pharmacological studies of IL-4-induced macrophage fusion and FBGC formation were carried out to investigate the roles, if any, of matrix metalloproteinases in this phenomenon. We found that epigallocatechin gallate, which inhibits MMP-2, MMP-9, and MMP-12, and NNGH, which inhibits MMP-3 and MMP-12, did not affect either macrophage adhesion or fusion at a range of

concentrations up to 50 μM (Figures 3.4.A & 3.4.B). However, CL-82198,

reported to be a selective inhibitor of MMP-13 that does not affect MMP-1 or

MMP-9, strongly inhibited fusion by approximately 60% between 1 and 5 μM with no effect on adhesion (Figure 3.4.C). In addition, actinonin, which inhibits aminopeptidases such as collagenase MMPs (MMP-1, MMP-8, MMP-13, and

MMP-18), blocked approximately 60% of macrophage fusion at 50 μM, but not at

25 μM, nor did it affect macrophage adhesion (Figure 3.4.D).

- 94 - Chapter III

A B

Figure 3.4: The Effects of MMP Pharmalogical Inhibitors ECG (A), NNGH (B), CL-82198 (C), and Actinonin (D) on Macrophage Adhesion (■) and Fusion (●). Mean ± SEM, n=3.

Discussion

This study was carried out to identify and quantify the presence of MMPs

and TIMPs in monocyte-derived macrophage and FBGC cell cultures and to

determine if MMPs and TIMPs played a role in macrophage adhesion and fusion

into FBGCs. A broad investigation of MMPs and TIMPs was conducted using

- 95 - Chapter III

human antibody array detection of the cell culture supernatants. These results

showed that numerous MMPs/TIMPs are produced by macrophages/FBGCs into

the cell culture supernatants suggesting macrophages may modulate MMP and

TIMP production potentially affecting the structural components of the ECM in the surrounding milieu and those components involved in cellular adhesion, migration, and fusion (Table 3.2). In addition, MMP/TIMP production may be

material-dependent based on the varying signal intensities detected between cell

culture supernatants.

An initial analysis of quantifying the MMP/TIMP concentrations in cultures

containing RGD-surface-modified substrates or photografted polymers showed

that MMP-9 is produced at significantly greater concentrations than TIMP-1 and

TIMP-2 on all materials at all timepoints (Table 3.3). MMP/TIMPs continued to

be produced over time with an increase in MMP production between days 3 and

7 correlating to the advancement of the biomaterial implant site into the foreign

body reaction, fibrous capsule formation, and ultimate wound healing. These

results indicate that macrophages/FBGCs can potentially tip the balance

between MMPs/TIMPs in the surrounding milieu furthering the remodeling

process of the ECM in angiogenesis and wound healing and potentially inhibiting

fibrous capsule formation.

With respect to the effects of material surface chemistry on MMP/TIMP

production, MMP-9 concentrations were comparable on all surface chemistries

over time, while TIMP-1 and TIMP-2 concentration tended to be increased on the

hydrophilic/anionic surface (Figure 3.1). This suggests that TIMP production

- 96 - Chapter III

may be material dependent and prompts further investigation into this

phenomenon. Previous research as well as the adherent cell densities on these

photografted polymer surfaces (data not shown) demonstrated that the

hydrophilic/neutral surface inhibits macrophage adhesion, while the hydrophobic

and hydrophilic/ionic surfaces promote comparable amounts of adherent

cells.17,25,26 This difference in adherent cell densities with comparable

MMP/TIMP concentrations in the cell culture supernatants prompted an investigation into the amount of MMP/TIMP produced per cell to determine if

material surface chemistry dictated adherent cell activity levels via MMP/TIMP

production. The hydrophilic/neutral surface was found to promote the quantity of

MMP/TIMP produced per cell for MMP-9, TIMP-1, and TIMP-2 regardless of timepoint or IL-4 addition (Figure 3.3). These findings indicate that although the hydrophilic/neutral surfaces inhibit macrophage adhesion, it may activate these macrophages to produce a greater amount of MMP/TIMP in comparison to the other surface chemistries. A previous study in our laboratory found a similar inverse relationship between the production of cytokines and chemokines and macrophage adhesion on hydrophilic/neutral surfaces.17 Future analysis of

various hydrophilic/neutral surface chemistries will elucidate this phenomenon.

In the pharmacological inhibition experiments, it was shown that ECG,

which inhibits MMP-2, MMP-9, and MMP-12, and NNGH, an inhibitor of MMP-3 and MMP-12, did not affect macrophage adhesion or fusion indicating that these

MMPs are not involved in these cellular processes (Figure 3.4). In contrast, macrophage fusion was inhibited significantly by CL-82198, an inhibitor of MMP-

- 97 - Chapter III

13, but not MMP-1 or MMP-9, and actinonin, which inhibits aminopeptidates such

as the collagenase MMPs (MMP-1, MMP-8, MMP-13, and MMP-18). The

inhibition of fusion in these cultures indicates that potentially MMP-1, MMP-8,

MMP-13, and MMP-18 are present in the macrophage/FBGC cell culture system

despite their low to minimal detection in the human antibody array, which is

possible due to the set sensitivities of the array system for each MMP.

Biologically, these findings implicate MMP-1, MMP-8, MMP-13, and MMP-18 involvement in macrophage fusion and FBGC formation potentially via one or more of the numerous activities of these proteolytic enzymes including the degradation of structural molecules and/or cell surface bound molecules required for cell adhesion, migration, cell-cell adhesion, and/or cell-cell fusion. Additional research into the mechanisms of MMP involvement in macrophage fusion is warranted.

In summary, this research demonstrates that biomaterial-adherent macrophages produce a variety of MMPs and TIMPs including MMP-9, TIMP-1, and TIMP-2 and can potentially modulate the surrounding ECM and cellular behaviors by secreting disproportionate amounts MMPs and TIMPs. The overall production of these molecules, particularly TIMP-1 and TIMP-2, is promoted on hydrophilic/anionic surfaces, while hydrophilic/neutral surfaces further activate macrophages to secrete greater amounts of MMPs and TIMPs on a cellular level.

These phenomena need additional analysis to further elucidate the material- dependent nature. MMP inhibition studies revealed that MMP-13 and possibly

MMP-1, MMP-8, and MMP-18 are involved in macrophage fusion and require

- 98 - Chapter III

further analysis. By using a multiple-angle approach, we have ascertained clear

evidence that material chemistry has an effect on macrophage and FBGC

production of MMPs and TIMPs and that these MMPs play a crucial role in

macrophage fusion, FBGC formation, and the resulting foreign body reaction.

References

1. Visse R, Nagase H. Matrix Metalloproteinases and Tissue Inhibitors of

Matrix Metalloproteinases: Structure, Function, and Biochemistry.

Circulation Research 2003;92:827-839.

2. Meznarich N, Kyriakides T, Donaldson E, Foster M, Schrom B, Ratner B,

Hauch K, Bornstein P. Matrix-Metalloproteinase (MMP-9) and Its Role in

Wound Healing and the Foreign Body Response. Journal of

Undergraduate Research in Bioengineering 2004;4(2):84-89.

3. Klein S, Anderson G, Kennedy A, Bond S. The effects of broad-spectrum

matrix metalloproteinase inhibitor on characteristics of wound healing.

Journal of Investigative Surgery 2002;15:199-207.

4. Lafleur M, Handsley M, Edwards D. Metalloproteinases and their inhibitors

in angiogenesis. Expert Reviews in Molecular Medicine 2003;5(22):1-39.

5. Nagase H, Visse R, Murphy G. Structure and function of matrix

metalloproteinases and TIMPs. Cardiovascular Research 2006;69:562-

573.

6. Sternlicht M, Werb Z. How matrix metalloproteinases regulate cell

behavior. Annual Review of Cell Developmental Biology 2001;17:463-516.

- 99 - Chapter III

7. Borden P, Heller R. Transcriptional control of matrix metalloproteinases

and tissue inhibitors of matrix metalloproteinases. Critical Reviews in

Eukaryotic Gene Expression 1997;7:159-178.

8. Newby A. Matrix metalloproteinases regulate migration, proliferation, and

death of vascular smooth muscle cells by degrading matrix and non-matrix

substrates. Cardiovascular Research 2006;69:614-624.

9. Imai K, Hiramatsu A, Fukushima D, Pierschbacher M, Okada Y.

Degradation of decorin by matrix metalloproteinases: identification of the

cleavage sites, kinetic analyses, and transforming growth factor beta-1

release. Biochemistry Journal 1997;322:809-814.

10. Oltmanns J, Fietz I, Witt C, Jung K. Increased production of matrix-

metalloproteinase-2 in alveolar macrophages and regulation by

interleukin-10 in patients with acute pulmonary sarcoidosis. Experimental

Lung Research 2002;28(55-68).

11. Reunanen N, Westermarck J, Hakkinen L, Holmstrom T, Elo I, Eriksson J,

Kahari V. Enhancement of Fibroblast Collagenase (Matrix

Metalloproteinase-1) Gene Expression by Ceramide is Mediated by

Extracellular Signal-regulated and Stress-activated protein Kinase

Pathways. The Journal of Biological Chemistry 1998;273(9):5137-5145.

12. Vincenti M, Brickerhoff C. Transcriptional regulation of collagenase (MMP-

1, MMP-13) genes in arthritis: integration of complex signaling pathways

for the recruitment of gene-specific transcription factors. Arthritis Research

2002;4(157-164).

- 100 - Chapter III

13. Munshi H, Stack M. Reciprocal interactions between adhesion receptor

signaling and MMP regulation. Cancer Metastasis Reviews 2006;25:45-

56.

14. Uria J, Jimenez M, Balbin M, Freije J, Lopez-Otin C. Differential effects of

transformation growth factor beta on the expression of collagenase-1 and

collagenase-3 in human fibroblasts. Journal of Biological Chemistry

1998;273(16):9769-9777.

15. Liacini A, Sylvester J, Li W, Zafarullah M. Inhibition of interleukin-1-

stimulated MAP kinases, activating protein-1, nuclear factor kappa B

transcription factors down-regulates matrix metalloproteinase gene

expression in articular chondrocytes. Matrix Biology 2002;21:251-262.

16. Baker A, Edwards D, Murphy G. Metalloproteinase inhibitors: biological

actions and therapeutic opportunities. Journal of Cell Science

2002;115:3719-3727.

17. Jones J, Chang D, Colton E, Kwon I, Matsuda T, Anderson J. Proteomic

analysis and quantification of cytokines and chemokines from biomaterial

surface-adherent macrophages and foreign body giant cells. Journal of

Biomedical Materials Research 2006;Submitted.

18. Anderson JM, DeFife K, McNally A, Collier T, Jenney C. Monocyte,

macrophage and foreign body giant cell interactions with molecularly

engineered surfaces. Journal of Materials Science: Materials in Medicine

1999;10:579-588.

- 101 - Chapter III

19. Anderson JP, Cappello J, Martin DC. Morphology and primary crystal

structure of a silk-like protein polymer synthesized by genetically

engineered E. Coli bacteria. Biopolymers 1994;34(8):1049-1057.

20. Nakayama Y, Anderson JM, Matsuda T. Laboratory-scale mass

production of a multi-micropatterned grafted surface with different polymer

regions. Journal of Biomedical Materials Research 2000;53(5):584-591.

21. Nakayama Y, Matsuda T. Surface macromolecular architectural designs

using photograft copolymerization based on photochemistry of benzyl

N,N-diethyldithiocarbamate. Macromolecules 1996;29:8622-8630.

22. McNally AK, Anderson JM. Complement C3 participation in monocyte

adhesion to different surfaces. Proceedings of the National Academy of

Science 1994;91:10119-10123.

23. McNally A, Anderson JM. Beta-1 and Beta-2 Integrins Mediate Adhesion

during Macrophage Fusion and Multinucleated Foreign Body Giant Cell

Formation. American Journal of Pathology 2002;160(2):621-630.

24. Jones J, Dadsetan M, Collier T, Ebert M, Stokes K, Ward R, Hiltner A,

Anderson JM. Macrophage behavior on surface modified biomaterials.

Journal of Biomaterial Science Polymer Edition 2004;15(5):567-564.

25. Brodbeck W, Shive M, Colton E, Nakayama Y, Matsuda T, Anderson JM.

Influence of biomaterial surface chemistry on the apoptosis of adherent

cells. Journal of Biomedical Materials Research 2001;55:661-668.

- 102 - Chapter III

26. Brodbeck W, Nakayma Y, Matsuda T, Colton E, Ziats N, Anderson JM.

Biomaterial surface chemistry dictates adherent monocyte/macrophage

cytokine expression in vitro. Cytokine 2002;18(6):311-319.

- 103 - Chapter IV

Chapter IV: The Effects of Hydrophobic, Hydrophilic, & Ionic Surface Chemistries on Cellular Behaviors

The effects of surface chemistry on adherent cellular behaviors have been

a key area of research for years. This research is driven by the notion that

defining these relationships will aid in establishing criteria for designing

biomaterials utilized in future applications. In the past, the accepted dogma was

that hydrophobic surfaces inhibit cellular adhesion while hydrophilic surfaces

promote cellular adhesion. Contradictions arose with other hydrophilic and

hydrogel surfaces (i.e. polyethylene oxide and pHEMA), which adsorb significant

amounts of water creating a barrier to protein adsorption and subsequent cellular adhesion. Complexities in defining these relationships arise in the not fully understood complex manner in which adherent cells interact with the substrate involving protein adsorption, integrin expression, ligand-integrin binding, cell signaling and the subsequent effects these behaviors have on the resulting cellular behaviors. The concept that minimizing cellular adhesion minimizes cellular activity on a biomaterial surface has also been an accepted tenet prompting countless studies to investigate ways to minimize cellular adhesion.

Recent research has focused on understanding how biomaterial surface chemistry directs adherent macrophage activity and behaviors including cytokine and chemokine production as a means to direct subsequent biological responses

(i.e. inflammation and wound healing) to implanted biomaterials. The following discussion explores the new hypothesis (resulting from Chapters II and III) that

- 104 - Chapter IV

surface chemistry directs macrophage adhesion and activation in manners

opposing previous dogma.

Cellular Adhesion and Fusion on Hydrophobic and Hydrophilic Surfaces

The studies within this thesis have demonstrated that hydrophobic surfaces support macrophage adhesion and fusion, while hydrophilic and neutral surfaces markedly inhibited macrophage adhesion and fusion. This support of adhesion and fusion was seen in Chapter II on hydrophobic polyethylene terephthalate (PET) and on PET surfaces coated with hydrophobic poly(styrene- co-benzyl N,N-dimethyldithiocarbamate) (BDETDC). See Figure 4.1 for

BDEDTC cell adhesion and fusion results. On surfaces in which the PET was rendered hydrophilic using a photografted acrylamide modification (PAAm), macrophage adhesion and fusion was significantly inhibited to values equal to or less than a third of the values seen on BDEDTC. These findings were supported by previous research in our laboratory by Brodbeck et.al. and confirmed with the increase in macrophage fusion seen in Chapter V with the incorporation of hydrophobic silicone modifications to polyurethanes.1,2 Interestingly, when the

PET surface was modified with a hydrophilic and anionic (PAANa) or hydrophilic and cationic modification (DMAPAAmMeI), cellular adhesion levels were

markedly greater than the hydrophilic/neutral surfaces (2 to 45 fold greater) as

well, at levels comparable to the hydrophobic surfaces (Figure 4.2).

- 105 - Chapter IV

BDEDTC - Hydrophobic 2000 1800 PAAm - Hydrophilic, Neutral A 0.3X 1600 0.3X 0.04X 1400 1200 1000 800 600 (cells/mm^2) ** 400 200

Number ofNumber Adherent Cells * 0 45% Day 3 Day 7 Day 10 B <0.04X 40% 35% 30% 25% 20% 0.3X 15% Percent FusionPercent 10%

* 5% 0% Day 3 Day 7 Day 10 Time (Days)

Figure 4.1: Disparate Effects of Hydrophobic and Hydrophilic Surfaces on Macrophage Adhesion (A) and Fusion (B). Mean ± SEM, n=3. “*” indicates that these values for the hydrophilic surface are statistically less than the hydrophobic surface values (p<0.05).

- 106 - Chapter IV

3.6X 2.4X 2000 A 1800 4.7X 13X* 4.6X * 1600 * 19X* 1400 1200 1000 800 (cells/mm^2) 600 400 * Number of Adherent Cells 200 0 100% Day 3 Day 7 Day 10 PAAm - 90% B Hydrophilic, 5.9X 2.3X 80% Neutral 70% 45X * PAANa - 33X * 60% Hydrohilic, 50% Anionic 40% DMAPAAmMeI - 30% Hydrohilic,

Percent Fusion 20% Cationic 10% 0% Day 3 Day 7 Day 10 Time (Days)

Figure 4.2: Disparate Effects of Hydrophilic Non-Ionic and Ionic Surfaces on Macrophage Adhesion (A) and Fusion (B). Mean ± SEM, n=3. “*” indicates that these values for the hydrophilic/ionic surfaces are statistically greater than the hydrophilic/neutral surface values (p<0.05).

Inverse Relationship of Cellular Adhesion and Activation

Generally speaking, macrophages adhere to a biomaterial, become activated, and then fuse to form multinucleated giant cells. Naturally, the term cellular “activation” is very indistinct. There are no specific hallmarks of activation nor does an active macrophage display a distinct set of “x” traits which could be used to distinguish an activated cell from an inactive one. Also, a given cell can be “activated” to varying degrees and produce varying responses.

- 107 - Chapter IV

Nevertheless, macrophage activation has been investigated in-depth for well

over 40 years and researchers have differentiated inactive cells from active cells

based upon the up- or down-regulation of gene expression, protein production,

biological surface molecules (i.e. receptors, integrins, and protein markers), and reactive oxygen species secretion in addition to the resulting behaviors (i.e.

phagocytosis or fusion).3,4 Some of these components (i.e. integrin activity and

cytokines) that mark a cell as “active” can also up- or down-regulate intracellular

processes, in turn, activating these cells.

Using the notion that an active cell produces greater amounts of given

cytokines and/or chemokines, which is common in the study of

biomaterial/cellular interactions, a cell can be defined as being in a more

activated state than at previous timepoints or in comparison to other cells. The

production of these proteins by activated cells may influence the behaviors of

other cells advancing a biological response (i.e. inflammation, the foreign body

reaction, wound healing). Adherent macrophages were investigated in this study

for a material-dependency in the production of cytokines, chemokines, matrix

metalloproteinases (MMPs), and tissue inhibitors of MMPs (TIMPs), which was

then utilized to draw conclusion about the activation state of these cells (detailed

in Chapter II and III). Previously, cellular activation was considered to correlate

with cellular adhesion. The fewer number of cells that adhere to a particular

surface, the less active are these cells. This has been shown to be true in some

instances in which decreasing cellular adhesion decreased the resulting effects

of these cells (i.e. reactive oxygen species production, production of a particular

- 108 - Chapter IV

cytokine, and fusion) concluding that these adherent cell populations are less active.

This dissertation research suggests that macrophage adhesion does not correlate with or indicate a level of macrophage activation. As discussed above, it was shown that macrophage populations on the hydrophilic/neutral PAAM surfaces were significantly fewer in number than the macrophage populations on the hydrophobic BDEDTC surfaces (Figure 4.3.A). One would expect the concentrations of cytokines/chemokines to be greater on the hydrophobic

BDEDTC surface in comparison to the hydrophilic/neutral PAAm surface following the above train of thought. This was not the case. As shown in Figure

4.3.B, the IL-10 cytokine concentrations on the hydrophilic/neutral PAAm

surfaces were equal to the hydrophobic BDEDTC surfaces. To further analyze

this relationship, the concentration of cytokine production was normalized to the

adherent cell density providing the quantity of each

cytokine/chemokine/MMP/TIMP produced by an adherent cell (Figure 4.3.C).

Analysis of the concentrations for IL-1β, IL-6, IL-8, MIP-1β, MMP-9, TIMP-1 and

TIMP-2 (left columns of Figures 4.4 & 4.5) on the hydrophilic/neutral PAAm

surfaces revealed that these concentrations were also equal to or greater than

the concentrations for the hydrophobic BDEDTC surfaces. The results of the

normalization of the concentrations to the total adherent cell number are shown

in the right columns of Figures 4.4 & 4.5. For every protein analyzed, the

amount of

- 109 - Chapter IV

BDEDTC - Hydrophobic 2000 PAAm - Hydrophilic, Neutral 1800 A - Cellular Adhesion 0.3X 1600 0.3X 0.04X 1400 1200 1000 800 (cells/mm^2) 600 ** 400 Number ofNumber Adherent Cells 200 * 0 Day 3 Day 7 Day 10 70 B - Cytokine Concentration 1.0X 60 1.1X 50 40 1.4X 30 IL-10, pg/mL IL-10, 20 10 0 Day 3 Day 7 Day 10 2.00E-02 37X* 1.80E-02 C- Cellular Activation 1.60E-02 1.40E-02 1.20E-02 1.00E-02 8.00E-03 4.1X 6.00E-03 5.6X IL-10, pg/cell 4.00E-03 2.00E-03 0.00E+00 Day 3 Day 7 Day 10 Time (Days) Figure 4.3: Direct Comparison of Cellular Adhesion (A), Cytokine Concentration (B), and Cellular Activation (C) on Hydrophobic and Hydrophilic/Neutral Surfaces. Cellular activation equals cytokine concentration measured in 1mL of media normalized to the total number of adherent cells. Mean ± SEM, n=3,4. “*” indicates that these values for the hydrophilic surface are statistically less than the hydrophobic surface values (p<0.05).

- 110 - Chapter IV

Concentration Activation 3.9X 20X 400 * 3.50E-02 * BDEDTC - Hydrophobic 350 3.00E-02 PAAm - Hydrophilic, Neutral 300 2.50E-02 250 2.00E-02 , pg/mL A , pg/cell 200 E β β 1.50E-02 150 IL-1 IL-1

IL-1 IL-1 1.3X 1.00E-02 3.4X 100 0.6X 50 5.00E-03 0 0.00E+00 Day 3 Day 7 Day 10 Day 3 Day 7 Day 10 7000 5.6X* 1.20E+00 46X* 6000 B 1.00E+00 F 5000 8.00E-01 4000 6.00E-01 3000 IL-6, pg/mL IL-6,

IL-6, pg/cell IL-6, 4.00E-01 2000 3.0X 1.8X 0.7X 16X 1000 2.00E-01 0 0.00E+00 Day 3 Day 7 Day 10 Day 3 Day 7 Day 10 2.50E+01 22X 300000 3.1X * 250000 C 2.00E+01 G 200000 1.50E+01 150000

IL-8, pg/mL IL-8, 1.5X 1.00E+01 3.7X 100000 1.5X pg/cell IL-8, 43X* 50000 5.00E+00 0 0.00E+00 Day 3 Day 7 Day 10 Day 3 Day 7 Day 10 4500 12X * 9.00E-01 4000 8.00E-01 D 7.00E-01 77X H 3500 83X * 3000 6.00E-01 * 2.8X 2500 3.3X 5.00E-01 2000 4.00E-01 1500 3.00E-01 10X MIP-1b, pg/mL *

1000 pg/cell MIP-1b, 2.00E-01 500 1.00E-01 0 0.00E+00 Day 3 Day 7 Day 10 Day 3 Day 7 Day 10 Time (Days) Time (Days)

Figure 4.4: Cytokine/Chemokine Concentration (A-D) and Cellular Activation (E-H) on Hydrophobic and Hydrophilic/Neutral Surfaces. Cellular activation equals cytokine/chemokine concentration measured in 1mL of media normalized to the total number of adherent cells. Notated numbers indicate the x fold increase between the values for BDEDTC and PAAm. Mean ± SEM, n=3. “*” indicates that values statistically greater on the hydrophilic surface compared to the hydrophobic values (p<0.05).

- 111 - Chapter IV Concentration Activation 37X 3500 1.0X 0.7 BDEDTC - Hydrophobic * L 3000 1.1X 0.6 PAAm - Hydrophilic, Neutral D 2500 0.5 4.1X 2000 0.4 1500 0.3 1.8X

MMP-9, pg/m A 1000 0.2 19X

MMP-9, pg/cell 500 0.1 0 0 Day 3 Day 7 Day 10 Day 3 Day 7 Day 10 90 0.012 1.0X 80 11X 0.01 * L 70 B E 1.0X 3.5X 60 0.5X 0.008 20X 50 * 0.006 40 30 0.004 TIMP-1, pg/m 20 TIMP-1, pg/cell 0.002 10 0 0 Day 3 Day 7 Day0.5X 10 Day 3 Day 7 Day 10 30 0.9X 0.006 18X

L * 25 C 0.005 F 20 0.004 2.9X 15 0.003 0.9X 10 0.002 5.4X TIMP-2, pg/m 5 TIMP-2, pg/cell 0.001 0 0 Day 3 Day 7 Day 10 Day 3 Day 7 Day 10 Time (Days) Time (Days)

Figure 4.5: MMP/TIMP Concentration (A-C) and Cellular Activation (D-F) on Hydrophobic and Hydrophilic/Neutral Surfaces. Cellular activation equals cytokine concentration measured in 1mL of media normalized to the total number of adherent cells. Notated numbers indicate the x fold change between the values for BDEDTC and PAAm. Mean ± SEM, n=3. “*” indicates that values statistically greater on the hydrophilic surface compared to the hydrophobic values (p<0.05).

cytokine/chemokine/MMP/TIMP produced by a cell adherent to PAAm was

greater than on BDEDTC by a 2 to 77 fold increase (average increase of 20 fold).

Analysis of the hydrophilic/neutral PAAm surface in comparison to the hydrophilic/ionic surfaces, PAANa and DMAPAAmMeI, revealed the same

inverse relationship between cellular adhesion and activation. As shown

previously in Figure 4.2, the hydrophilic/neutral surface (PAAm) inhibited cellular

adhesion, while the hydrophilic/ionic surfaces (PAANa and DMAPAAmMeI)

supported significantly greater amounts of cellular adhesion. Cytokine (IL-1β, IL-

- 112 - Chapter IV

6, and IL-10) and chemokine (IL-8 and MIP-1β) concentrations were comparable

between the hydrophilic neutral and ionic surfaces for all cytokines/chemokines

at every timepoint (left column of Figure 4.6). The only exceptions were for MIP-

1β at day 3 when the MIP-1β concentrations were significantly greater on the

hydrophilic/anionic surface (PAANa) than the neutral and cationic surfaces and

for IL-6 at day 3 when IL-6 concentrations were greatest on the

hydrophilic/neutral surface (PAAm). Similarly, MMP-9, TIMP-1, and TIMP-2

concentrations were either comparable between materials or greater on the

hydrophilic/anionic surface (PAANa) as shown in the left column of Figure 4.7.

The quantity of each cytokine/chemokine/MMP/TIMP produced per cell were

determined and revealed that the cell adherent to the hydrophilic/neutral surface

(PAAm) produced a greater quantity of each protein analyzed than cell adherent

to the hydrophilic/ionic surfaces (PAANa and DMAPAAmMeI). Once again, the hydrophilic/neutral surfaces with significantly lower adherent cellular densities were activated more so than the cell adherent to the adhesion promoting hydrophilic/ionic surfaces contrary to previous dogma.

One explanation of this inverse relationship could be that the adherent

FBGCs seen on the hydrophobic (BDEDTC) and hydrophilic/ionic (PAANa and

DMAPAAmMeI) surfaces were not as active or did not produce as great of an amount of cytokines/chemokines as adherent macrophages. Distinct surface molecules (i.e. mannose receptors and integrin) are present in FBGCs and macrophages, therefore it is not improbably for the activation states of these cells to be different.5-7 However, the concentrations were usually comparable at later

- 113 - Chapter IV timepoints when the FBGC population increased in number on these surfaces and would have been a factor. Another mechanism that may direct this disparate protein production could result from the increased release of certain cytokine/chemokines/MMPs/TIMPs in cells undergoing apoptosis. Bzowska et.al. demonstrated that changes in the cytokine production profiles of monocyte cultures occurred as monocyte apoptosis increased.8 Specifically, IL-10 levels increased with apoptosis while IL-1β and TNF-α levels were unaffected. We know from previous research that there are a greater percentage of apoptotic cells on the hydrophilic/neutral surfaces (PAAm) in comparison to the other surfaces.1 This alteration in cytokine production profiles associated with apoptosis could explain the significant increases in production of some cytokines/chemokines/MMP/TMPs on the PAAm surface versus the others.

Collectively, this study redefines the relationship cell adhesion and activation. In addition, it implicates material surface chemistry in the production of cytokines, chemokines, MMPs and TIMPs and on cellular activation via multiple mechanisms and prompts in-depth research into this area.

- 114 - Chapter IV

Concentration Activation 400 ~1.5X 0.045 3.7X PAAm - Hydrophilic, Neutral 350 0.04 14X* PAANa - Hydrophilic, Anionic 300 A 0.035 DMAPAAmMeI - Hydrophilic, Cationic 0.03 250 0.025 , pg/mL 200 , pg/cell β β 0.02 150 F

IL-1 IL-1 0.015 IL-1 IL-1 8.3X 100 ~1.3X 0.01 ~0.8X 49X 50 0.005 * 0 0 8000 Day1.7 X3 Day 7 Day 10 1.4 5.4DayX 3 Day 7 Day 10 7000 3.5X B 1.2 29X* G 6000 1 5000 0.8 4000 3000 0.6 pg/mL IL-6,

IL-6, pg/cell IL-6, 4.0X 2000 0.4 5.5X ~1.3X 11X 1000 ~0.6X 0.2 * 27X* 0 0 140 Day 3 Day 7 Day 10 1.3X 0.025 Day 3 Day 7 Day 10 9.9X 120 1.8X C H 0.02 49X 100 1.1X 80 0.8X 2.3X 0.015 60 3.9X 0.01 2.0X 2.5X IL-10, pg/mL IL-10,

40 pg/cell IL-10, 24X 12X 20 0.005 0 0 Day 3 Day 7 Day 10 300000 Day 3 Day 7 Day 10 ~1.0X 30 4.5X 250000 D 25 7.1X I 200000 20 150000 15 4.9X

IL-8, pg/mL IL-8, 100000 1.9X pg/cell IL-8, 17X ~1.1X 10 8.0X 50000 0.8X 11X 5 * 0 0 40000 Day3.6 3X Day 7 Day 10 0.9 Day 3 Day 7 Day 10 35000 0.1X 35X 0.8 20X * 30000 * EJ 0.7 * 26X 1.2X * 25000 0.6

, pg/mL 0.5 β 20000 15000 , pg/cell 0.4 9.2X 3.6X 3.5X β 0.3 MIP-1 3.4X 10000 0.5X 1.3X 0.2 5000 MIP-1 0.1 0 0 Day 3 Day 7 Day 10 Day 3 Day 7 Day 10

Figure 4.6: Cytokine/Chemokine Concentration (A-E) and Cellular Activation (E-J) on Hydrophilic/Neutral and Hydrophilic/Ionic Surfaces. Cellular activation equals cytokine/chemokine concentration measured in 1mL of media normalized to the total number of adherent cells. Notated numbers indicate the x fold increase between the values for PAAm and PAANa or DMAPAAmMeI. “*” indicates that values statistically greater on the hydrophilic /neutral surface compared to the hydrophilic/ionic surface values (p<0.05).

- 115 - Chapter IV

Concentration Activation 5000 0.8X 0.9 PAAm - Hydrophilic, Neutral 4500 0.8 L 1.0X 0.5X PAANa - Hydrophilic, Anionic ~12X D 4000 A 0.7 * DMAPAAmMeI - Hydrophilic, Cationic 3500 0.7X 0.6 3000 4.3X 0.5 2500 5.3X 2000 0.4 4.3X

~0.7X MMP-9, pg/cell MMP-9, pg/m MMP-9, 1500 0.3 6.9X 1000 0.2 500 0.1 0 0 Day 3 Day 7 Day 10 200 1.4X 0.014 Day 3 Day 7 Day 10 180 1.6X 0.2X 0.012 160 0.7X ~5.0X E B 0.6X 4.0X 140 0.4X 0.01 13X 120 4.7X * 0.008 4.0X 100 80 c 0.006

TIMP-1, pg/mL

60 TIMP-1, pg/cell 0.004 40 20 0.002 0 0 70 Day 3 Day1.2 7X Day 10 0.008 Day 3 Day 7 Day 10 60 0.4X 0.007 40X L C * 12X 50 0.006 * F 1.4X 0.005 40 4.7X 0.7X 0.5X 0.004 30 2.8X 0.4X 0.003

20 TIMP-2, pg/cell TIMP-2, pg/m ~3.6X 0.002 10 0.001 0 0 Day 3 Day 7 Day 10 Day 3 Day 7 Day 10 Figure 4.7: MMP/TIMP Concentration (A-C) and Cellular Activation (D-F) on Hydrophilic/Neutral and Hydrophilic/Ionic Surfaces. Cellular activation equals cytokine/chemokine concentration measured in 1mL of media normalized to the total number of adherent cells. Notated numbers indicate the x fold increase between the values for PAAm and PAANa or DMAPAAmMeI. Mean ± SEM, n=3. “*” indicates that values statistically greater on the hydrophilic/neutral surface compared to the hydrophilic/ionic surface values (p<0.05).

Biomaterials and Cellular Activation

In immunological or biological studies of macrophage behaviors,

macrophage activation has been classified based upon specific stimuli that have

similar resulting biological events (i.e. cytokine/chemokine production, integrin and surface molecule expression). “Classical” activation was initially seen in vivo by Mackaness et.al. when the antimicrobial activities of macrophages were enhanced in mice infected with Mycobaterium bovis bacillus Calmette-Guerin

(BCG). These cells are currently classified as active macrophages stimulated by lipopolysaccharide (LPS) or interferon gamma (IFN-γ) that up-regulate pro-

- 116 - Chapter IV inflammatory cytokines (i.e. IL-6, tumor necrosis factor (TNF) and IL-1), inhibit anti-inflammatory cytokines (i.e IL-10), variably up- or down-regulate chemokines, produce nitric oxide, and down-regulate mannose receptors and arginase production.3,4,9-12 Macrophages were also found to be activated by IL-4 and IL-13 producing an distinctly different set of biological results including the inhibition of pro-inflammatory cytokines (i.e. IL-6, TNF, and IL-1), promotion of IL-

10 and IL-1ra (receptor antagonist) cytokine production, opposing regulation of chemokines, up-regulation of mannose receptors, and production of arginase.3,9,10,12-15 This class of macrophages were coined “alternatively” activated macrophages. Additional classes stimulated by IL-10 and other factors have since been identified and are also referred to as alternatively activated macrophages, but are usually denoted to be distinct from IL-4 and IL-13 alternatively activated macrophages when discussed.3,16 The production profiles of numerous cytokines, chemokines, and MMPs by classically, IL-4/IL-13 alternatively, and IL-10 alternatively activated macrophages are shown in Table

4.1.

In Chapters II and III, antibody arrays and ELISAs were utilized to confirm the presence of numerous cytokines, chemokines, MMPs, and TIMPs and to determine any increases or decreases in concentration between materials and over time as shown in Table 4.1. Analysis of these production profiles revealed similarities between the classically activated macrophages and the biomaterial- adherent macrophages at day 3, while at later timepoints the biomaterial- adherent macrophages produced profiles similar to IL-4/IL-13 alternatively

- 117 - Chapter IV

activated macrophages. Based upon these activation state classes, these

findings suggests that the adherent macrophages undergo a phenotypic switch

over time from an activation state similar to classically activated macrophages

into one that is comparable to an alternatively activated macrophage. Material

surface chemistry did not appear to be a factor in this switch; however it is possible, the interaction of the macrophage with the biomaterial surface initiated

this phenotypic switch. Future analysis is necessary.

- 118 -

Table 4.1: A Comparison of Alternatively Activated and Classically Activated Macrophages Parameter Classical Alternative Biomaterial Effects (Day 3 / Day 10) LPS & IFN-γ IL-4 & IL-13 IL-10 Hydrophobic Hydrophilic Hydrophilic Hydrophilic & Neutral & Anionic & Cationic Cytokines IL-1 ↑3,4,10 ↓3,9 ↓ 10,15,17 ↑ / ↓ ↑ / ↓ ↑ / ↓ ↑ / ↓ IL-6 ↑3,4 ↓3,9 ↓ 17 ↑ / ↓ ↑ / ↓ ↑ / ↓ ↑ / ↓ IL-10 ↓9 ↑3,4,9 ↑ / ↑ ↑ / ↑ ↑ / ↑ ↑ / ↑ IL-12 ↑3,4,9 ↓9 ↔ ↔ ↔ ↔ TNF ↑3,4,9,18 ↓3,9 ↓ 17,18 ND ND ND ND Chemokines IL-8 ↑9 ↓ 17 ↑ / ↓ ↑ / ↓ ↑ / ↓ ↑ / ↓ MIP-1β ↑9 ↓9 ↓ 19 ↓ / ↓ ↓ / ↓ ↑ / ↓ ↓ / ↓ 3,9,12 3,12

- 119 MDC ↓ ↑ ↑ / ↑ ↑ / ↑ ↔ / ↑ ↔ / ↑ TARC ↓3,9 ↑3 ↔ ↔ ↔ ↔ Mig ↑9,12 ↓9,12 ↓ 9,12 ↔ ↔ ↔ ↔ RANTES ↑9 ↓9 ↓ 9,19 ↑ / ↑ ↑ / ↑ ↑ / ↑ ↑ / ↑ IP-10 ↑4,9,12 ↓9,12 ↓ 9,12 ↔ ↔ ↔ ↔ ENA-78 ↑9 ↑ / ↔ ↑ / ↔ ↔ ↔ MCP-1 ↑9 ↑ / ↑ ↑ / ↑ ↑ / ↑ ↑ / ↑ MIP-1α ↑4,9 ↓9 ↓ 19 ND ND ND ND Eotaxin ↑9 ↔ ↔ ↔ ↔ Eotaxin-2 ↓9,12 ↑9,12 ↑/↑ ↑/↑ ↔ / ↑ ↔ / ↑ GRO ↑9 ↔ ↔ ↔ ↔ MMP/TIMP MMP-9 ↓3,20 ↑ / ↑ ↑ / ↑ ↑ / ↑ ↑ / ↑ TIMP-1 ↑ ↑ / ↑ ↑ / ↑ ↑ / ↑ ↑ / ↑ TIMP-2 ↑ / ↑ ↑ / ↑ ↑ / ↑ ↑ / ↑ Bold: measured by ELISA and cytokine array; otherwise: only by cytokine array; “↑”/”↓”: level of production increased or decreased.

Chapter IV

Conclusions

This research is pivotal in that provides evidence that macrophage

activation does not proportionally correlate to macrophage adhesion as previous

dogma has indicated. In this case, amount of a given protein measured in the

surrounding milieu was not always distinct between materials. Factors

controlling the adherent cell density (i.e. surface chemistry and cytokine stimuli)

and the potential mechanisms involved (i.e. apoptosis, fusion, cytokine binding and intracellular signaling) would elucidate how these activation levels could be

directed. Additional research needs to be conducted to examine other key

cytokines and chemokines that could be potentially modulated in this same

manner. Controlling this cellular activation and potential phenotypic switch could

create provides an additional means by which macrophages can be induced to

regulate particular secretory proteins that direct inflammation, the foreign body

reaction, and/or wound healing.

References

1. Brodbeck W, Shive M, Colton E, Nakayama Y, Matsuda T, Anderson JM.

Influence of biomaterial surface chemistry on the apoptosis of adherent

cells. Journal of Biomedical Materials Research 2001;55:661-668.

2. Brodbeck W, Nakayma Y, Matsuda T, Colton E, Ziats N, Anderson JM.

Biomaterial surface chemistry dictates adherent monocyte/macrophage

cytokine expression in vitro. Cytokine 2002;18(6):311-319.

- 120 - Chapter IV

3. Gordon S. Alternative activation of macrophages. Nat Rev Immunology

2003;3(1):23-35.

4. Mosser D. The many faces of macrophage activation. Journal of

Leukocyte Biology 2003;73:209-212.

5. DeFife K, Jenney C, McNally A, Colton E, Anderson JM. Interleukin-13

Induces Human Monocyte/Macrophage Fusion and Macrophage Mannose

Receptor Expression. J Immunology 1997;158(3385-3390).

6. McNally A, Anderson JM. Beta-1 and Beta-2 Integrins Mediate Adhesion

durin Macrophage Fusion and Multinucleated Foreign Body Giant Cell

Formation. Amer J Path 2002;160(2):621-630.

7. Stein M, Keshav S, Harris N, Gordon S. Interleukin-4 potently enhances

murine macrophage mannose receptor activity: a marker of alternative

immunologic macrophage activation. J Exp Med 1992;176(282-292).

8. Bzowska M, Guzik K, Barczyk K, Ernst M, Flad HD, Pryjma J. Increased

IL-10 production during spontaneous apoptosis of monocytes. European

Journal of Immunology 2002;32(7):2011-2020.

9. Mantovani A, Sica A, Sozzani S, Allavena P, Vecchi A, Locati M. The

chemokine system in diverse forms of macrophage activation and

polarization. Trends in Immunology 2004;25(12):677-685.

10. Donnelly R, Fenton M, Finbloom D, Gerrard T. Differential regulation of IL-

1 production in human monocytes by IFN-g and IL-4. Journal of

Immunology 1990;145(2):569-575.

- 121 - Chapter IV

11. Charo I, Ransohoff R. The many roles of chemokines and chemokine

receptors in inflammation. The New England Journal of Medicine

2006;354(6):610-754.

12. Mantovani A, Sozzani S, Locati M, Allavena P, Sica A. Macrophage

polarization: tumor-associated macrophages as a paradigm for polarized

M2 mononuclear phagocytes. Trends in Immunology 2002;23(11):549-

555.

13. Bonder C, Finlay-Jones J, Hart P. Interleukin-4 regulation of human

monocyte and macrophage interleukin-10 and interleukin-12 production.

Role of a functional interleukin-2 receptor gamma-chain. Immunology

1999;96:529-536.

14. Bonecchi R, Sozzani S, Stine J, Luini W, D'Amico G, Allavena P, Chantry

D, Mantovani A. Divergent effects of interleukin-4 and interferon-gamma

on macrophage-derived chemokine production: an amplification circuit of

polarized T helper 2 responses. Blood 1998;92(8):2668-2671.

15. Fenton M, Buras J, Donnelly R. IL-4 reciprocally regulates IL-1 and IL-1

receptor antagonist expression in human monocytes. Journal of

Immunology 1992;149(4):1283-1288.

16. Katakura T, Miyazaki M, Kobayaski M, Herndon D, Suzuki F. CCL17 and

IL-10 as effectors that enable alternatively activated macrophages to

inhibit the generation of classically activated macrophages. Journal of

Immunology 2004;172:1407-1413.

- 122 - Chapter IV

17. de Waal Malefyt R, Abrams J, Bennett B, Figdor C, de Vries J. Interleukin-

10 (IL-10) inhibits cytokine synthesis by human monocytes: an

autoregulatory role of IL-10 produced by macrophages. Journal of

Experimental Medicine 1991;174:1209-1220.

18. Oswald I, Wynn T, Sher A, James S. Interleukin 10 inhibits macrophage

microbicidal activity by blocking the endogenous production of tumor

necrosis factor alpha required as a costimulatory factor for interferon-

gamma induced activation. Proceedings of the National Academy of

Science 1992;88:8676-8680.

19. Kopydlowski K, Salkowski C, Cody M, van Rooijen N, Major J, Hamilton T,

Vogel S. Regulation of macrophage chemokine expression by

lipopolysaccharide in vitro and in vivo. Journal of Immunology

1999;163:1537-1544.

20. Chizzolini C, Rezzonico R, De Luca C, Burger D, Dayer J. Th2 cell

membrane factors in association with IL-4-enhance matrix

metalloproteinase-1 (MMP-1) whil decreasing MMP-9 production by

granulocyte-macrophage colony-stimulating factor differentiated human

monocytes. Journal of Immunology 2000;164(11):5952-5960.

- 123 - Chapter V

Chapter V: In vitro and in vivo Macrophage Behaviors on Surface Modified Polyurethanes

Abstract

Adherent macrophages and foreign body giant cells (FBGCs) are known

to release degradative molecules that can be detrimental to the long-term

biostability of polyurethanes. The modification of polyurethanes using surface

modifying endgroups (SMEs) and/or the incorporation of silicone into the polyurethane soft segments may alter macrophage adhesion, fusion, and apoptosis resulting in improved long-term biostability. An in vitro study of macrophage adhesion, fusion, and apoptosis was performed on polyurethanes modified with fluorocarbon SMEs, polyethylene oxide (PEO) SMEs, or poly(dimethylsiloxane) (PDMS) co-soft segment and SMEs. The fluorocarbon

SME and PEO SME modifications were shown to have no effect on macrophage adhesion and activity, while silicone modification had varied effects.

Macrophages were capable of adapting to the surface and adhering in a similar manner to the silicone modified and unmodified polyurethanes. In the absence of IL-4, macrophage fusion was comparable on the modified and unmodified polyurethanes, while macrophage apoptosis was promoted on the silicone modified surfaces. In contrast, when exposed to IL-4, a cytokine known to induce foreign body giant cell (FBGC) formation, silicone modification resulted in more macrophage fusion to form foreign body giant cells. An in vivo investigation into macrophage behavior on the silicone modified and unmodified polyurethanes was performed to further clarify the effects of silicone modification. Similarly to

- 124 - Chapter V

the in vitro study, macrophage adhesion was comparable on the silicone

modified and unmodified surfaces. Also, the silicone modification promoted

macrophage fusion. Macrophage apoptosis in vivo was only increased on the

silicone modified polyurethanes initially. At later timepoints, macrophage

apoptosis was comparable on the all of the materials. In conclusion,

fluorocarbon SME and PEO SME modification does not affect macrophage

adhesion, fusion, and apoptosis, while silicone modification is capable of

mediating macrophage fusion and apoptosis. Silicone modification may be

utilized to direct the fate of adherent macrophages.

Introduction

When a biomaterial is implanted into the body, the inflammatory and

wound healing responses are initiated causing monocytes to infiltrate the injured

site, adhere to the implant surface, and differentiate into macrophages.1,2,3

These macrophages, adherent to the biomaterial, can subsequently fuse to form multi-nucleated foreign body giant cells (FBGCs).3,4 Upon activation, the

adherent macrophages and FBGCs can release degrading molecules such as

surface-damaging oxygen radicals (i.e. superoxide anions) that will cause

oxidative degradation of the polyurethane.5,6,7 Over time, it has been shown that

the oxidation of the polyether segments can lead to environmental stress

cracking and permanent damage of the polyurethane material.1,8,9

Modulation of macrophage adhesion, formation into FBGCs, and

apoptosis activity at the implant/tissue interface provides a reasonable attempt to

- 125 - Chapter V

increase the in vivo lifetime of biomaterials. Induction of apoptosis may be the

most efficient mechanism for minimizing adherent cell activities, because it

results in the removal of cells without inducing an inflammatory response in the

surrounding tissue.10 Apoptosis is a specialized programmed cell death, which

requires active gene expression and is mediated by signal transduction unlike

normal cell death, i.e. necrosis. It is easily characterized by distinct chemical and

morphological features such as the translocation of phosphatidylserine to the

outer leaflet of the cell membrane, cell shrinkage, condensed chromatin,

denatured DNA, and the presence of apoptotic bodies.

Polyether and polycarbonate urethanes have been used for years in a

number of biomedical applications due to their favorable biocompatibility and

mechanical properties.11,12 However, it has been shown that polyurethane

materials are prone to degradation via surface oxidation, environmental stress

cracking (ESC), and metal ion oxidation (MIO).6,13,14 Polyurethanes have the

potential to degrade via oxidative and hydrolytic chain cleavage. Specifically,

oxidative chain cleavage can be reduced by a reduction in reactive oxygen

radicals through the minimization of adherent macrophages.

Polymer Technology Group, Inc. (PTG) has developed a series of

polyurethanes chemically designed to be more oxidative and hydrolytically

biostable. In one system, surface modifying endgroups (SMEs) of materials that are known to be more biostable or less biointeractive are covalently bonded to the ends of the polyurethane chains. SME segments are capable of migrating to the surface of the material in order to produce a more energetically favorable

- 126 - Chapter V

surface without diffusing from the surface of the biomaterial. This chain

rearrangement can affect cellular-biomaterial interface interactions; while having

no effect on the bulk properties of the polyurethane.

Surfaces rich in polyethylene oxide (PEO) have been shown to reduce

protein adsorption and cellular adhesion, in vitro and in vivo.15,16,17,18 It is

believed that the inhibition of these cellular-biomaterial interactions is due to the

high association of PEO with water molecules that provide a barrier to protein

adsorption. Research by Jenny et. al. indicates that long term macrophage

adhesion and IL-4 induced FBGC formation can be significantly reduced by PEO

SME modified surfaces.15

Due to their extremely hydrophobic nature and ability to inhibit the

adsorption of blood clotting proteins, fluorocarbon based materials have been

used in various biomedical applications, such as vascular grafts and other anti-

thrombogenic applications.19 It has been shown that although the initial

adsorption of proteins commonly involved in macrophage adhesion was high on

PTFE surfaces, the strength of adhesion was low and proteins were easily

removed with a SDS wash.20 Jahangir et. al. showed that a system of PEUs

containing a fluorine surface modifying macromolecules (SMMs) created a more hydrophobic surface chemistry, decreased fibrinogen protein adhesion, and

altered platelet adhesion.21 The fluorine SMMs materials were also shown to

inhibit enzyme-induced biodegradation.22 These material characteristics suggest

that the addition of PEO and fluorocarbon SMEs to polyurethanes may alter

- 127 - Chapter V

macrophage adhesion, fusion, and apoptosis and ultimately improve the

biostability of polyurethanes.

A second approach that has been investigated to produce a more

oxidative and hydrolytically biostable polyurethane involves the addition of

poly(dimethylsiloxane) (PDMS) to the soft segment and SME unit.23,24,25,26

PDMS materials have excellent flexibility, good blood biocompatibility, low

toxicity, and an anti-adhesive nature. It has been shown that the modification of

polyurethanes with PDMS resulted in a silicone-rich surface due to the high

mobility and low surface energy of the PDMS segment.27 A previous study in our

group has shown an enhanced in vivo biostability of PDMS end-capped polyurethanes in comparison to the uncapped polyurethanes.14 In addition,

Martin et al reported that the incorporation of PDMS into polyurethane soft

segment increased the in vivo biostability of polyurethane.26 These findings

suggest that the PDMS incorporation into both the soft segment and SME unit

may alter the cellular-biomaterial interactions and may be beneficial to

polyurethane biostability.

In this study, we investigated the effect of the addition of fluorocarbon and

PEO SME to polyurethanes on monocyte/macrophage adhesion, fusion and

apoptosis. Also, we have examined the simultaneous effect of PDMS in the soft

segments and SME unit on macrophage adhesion and activity. These

parameters were investigated in vitro initially to investigate the effects of the SME

and silicone modified polyurethanes. In vitro results revealed little to no

difference in macrophage behavior on the fluorocarbon or PEO SMEs modified

- 128 - Chapter V

verses the unmodified polyurethanes. However, silicone modification did affect

macrophage fusion and apoptosis. Further investigation into the fate of

macrophage adherent to silicone modified polyurethanes was conducted in vivo.

Materials and Methods

Biomaterials

Polyurethane films listed in Table 5.1 were synthesized by Polymer

Technology Group, Inc. (Berkeley, CA) and extruded by Medtronic, Inc.

(Minneapolis, MN).28 Elasthane 80A is a polyether urethane with an aromatic

hard segment of 4, 4’-methylene bisphenyl diisocyanate (MDI) chain extended

with 1, 4-butanediol (BD), and a polytetramethylene oxide (PTMO) soft segment.

Elasthane 80A with surface modifying endgroup (SME) modification have either

polyethylene oxide (PEO) or polytetrafluoroethylene (PTFE) SMEs covalently

bonded to the ends of the polyether urethane chain. The SME average

molecular weights for PEO and PTFE are 2000 and 500, respectively. PurSil 20

80A has the same composition as Elasthane 80A; however, it contains a 20%

PDMS co-soft segment and 0.2wt% PDMS SMEs. Bionate 80A is a polycarbonate urethane with a MDI hard segment chain extended with BD and a polycarbonate soft segment of poly(1, 6 hexyl 1, 2-ethyl carbonate) diol

(PHECD). CarboSil 20 90A is similar to Bionate 80A; however, it has a PHECD and 20wt% PDMS co-soft segment, 0.2wt% PDMS SMEs, and a different hard to soft segment ratio. All polyurethane films were extruded as previously described by Patel et. al.29

- 129 - Chapter V

Table 5.1 Chemical Description of Polyurethane Materials* Molecular Material Description Soft Hard SME Mass of Segment Segment Content SME (Da)

Elasthane 80A E80A PTMO MDI/BD - -

Bionate 80A B80A PHECD MDI/BD - -

Elasthane 80A E80A-F PTMO MDI/BD 6 wt% 500 with FC SME FC

Elasthane 80A E80A-P PTMO MDI/BD 6 wt% 2000 with PEO SME PEO

PurSil 20 80A PSil80A PTMO, MDI/BD 0.2 wt% 2000 20wt%PDMS PDMS

CarboSil 20 90A CSil90A PHECD, MDI/BD 0.2 wt% 2000 20wt%PDMS PDMS *Abbreviations: FC=fluorocarbon; MDI, 4, 4’-methylene bisphenyl diisocyanate; BD, 1, 4-butanediol; PTMO, polytetramethylene oxide; PHECD, poly (1,6 hexyl 1, 2-ethyl carbonate) diol; PDMS, polydimethylsiloxane. SME molecular mass as reported by Medtronic, Inc.

Polydimethyl siloxane (PDMS) films were prepared from SILASTIC®

BioMedical Grade ETR Elastomer Q7-4765 (Dow Coring, Midland, MI) according to manufacturer’s instruction. PDMS film samples were extracted overnight with two changes of hexane to remove silicone oil residues after vulcanization. Prior to cultures, air bubbles were removed from the PDMS surfaces using the method previously described by Kalman et al.30

The polyurethane films were cut into 15 mm diameter disks, wiped with

100% ethanol and a kimwipe, sonicated in 100% ethanol for 10 minutes, rinsed

in distilled water twice, air dried, and then gas sterilized with ethylene oxide.

- 130 - Chapter V

Silicone tubing (Cole-Parmer, Vernon Hills, IL) was sectioned into rings, 8 mm in

length, sonicated, and autoclaved for 1 hour. Under sterile conditions, the material samples were inserted into 24-well tissue culture polystyrene (TCPS) plates (Fisher Scientific, Pittsburgh, PA) and secured with the silicone rings. The resultant surface area for each culture well was 0.71 cm2.

Surface Characterization via Contact Angle Analysis

Advancing contact angle measurements were used to determine the

hydrophobic/hydrophilic nature of the materials using a goniometer (Edmund

Scientific Co. Barrington, NJ) at 22oC, room temperature. Contact angle measurements were performed in order to understand the surface effect of the modified chemistries and the surface effect of an aqueous environment on the material. Two systems were utilized in the contact angle analysis to study these effects, respectively; A) in air with a deionized, distilled water bubble probe and

B) under aqueous conditions in a deionized, distilled water bath with a hydrophobic methylene iodide probe. It is important to note that more hydrophobic surfaces produce a greater contact angle degree in the first system, while more hydrophilic surfaces produce a greater contact angle degree in the second system.

In Vitro Cell Culture

Human peripheral monocytes and serum were obtained from the whole, venous blood of healthy, unmedicated donors as described by McNally et.al.31

- 131 - Chapter V

Isolated monocytes were cultured at a concentration of 5x105 cells/well in 0.5ml

of culture media containing 20% autologous serum and RPMI 1650 cell media

o (Gibco BRL Gaithersburg, MD). Cultures were incubated at 37 C with a 5% CO2 environment. After the initial two hours, non-adherent cells were removed by rinsing with warmed Dulbecco’s Phosphate-Buffered Saline (PBS) with Ca2+ and

Mg2+ (Gibco BRL Gaithersburg, MD), and fresh culture media was added. At days 3 and 7, the remaining cultures were refed with a culture media of RPMI and 20% autologous, heat inactivated serum. Serum was heat inactivated at

56oC for 1 hour. Initially at day 3, IL-4 was added to specific cultures at a

concentration of 15 ng/ml to induce FBGC formation.32 Cultures were terminated

at days 0 (2 hours), 3, 7, and 10. Each sample was rinsed with warm PBS and

treated according to the procedure for adhesion, fusion, and apoptosis analysis.

In Vivo Procedure

Materials were implanted into 3-month-old, 250-300g Sprague-Dawley

rats as previously described.33 34,35 To summarize, sonicated samples were

placed within surgical-stainless steel cages (3.5 cm in length, 1.2 cm in diameter,

0.25 mm thick wire diameter) and sterilized using ethylene oxide. Sterile cages

were then implanted subcutaneously into the backs at the level of the panniculus

carnosus of anesthetized rats. The implant and explant procedures were in

agreement with the National Institute of Health’s guidelines with special care and

attention to the animals. At days 4, 7 and 21, the rats underwent euthanasia and

the samples/cages were removed. All samples were gently rinsed with PBS and

- 132 - Chapter V

then staining according to the procedures described below with May Grunwald

and Giemsa, Annexin V-FITC, or TUNEL.

Adhesion and Fusion Analysis

For optical microscopy (OM), samples were rinsed with PBS, fixed in

100% methanol for 5 minutes, and air-dried. Adherent cells were stained with

May-Grunwald for 1 minute, rinsed with PBS twice, stained with Giemsa for 5

minutes, rinsed twice with distilled water, and air dried overnight. Total number

of adherent cells and foreign body giant cells (FBGC) were counted from 5-20x

objective fields (463μm x 463μm) on each sample. These cell densities were

averaged to determine the adherent macrophage density and percentage of

nuclei within FBGC at all timepoints.

Apoptosis Analysis

Annexin V fluorescein isothiocyanine (FITC) staining was used to detect

early stage macrophage apoptosis for both in vitro and in vivo samples. This staining technique detects early stage apoptosis using a fluorescently labeled antibody specific for negatively charged phospholipids (i.e. phosphatidyl serine).

In a viable cell, these phospholipids are found located predominately on the cytosollic side of the membrane, due to mechanisms that continually “flip” these phospholipids inward. However, in an apoptotic cell, these mechanisms fail and the negatively charged phospholipids can be detected on the outer side of the membrane. At the appropriate timepoints, each sample was rinsed twice with

- 133 - Chapter V

PBS and then stained using an Annexin V (FITC) conjugate regent kit according to the instructions provided by the manufacturer (R&D Systems, Minneapolis,

MN).36

A TUNEL (TdT-mediated dUTP nick end labeling) technique staining kit

was used to detect late stage macrophage apoptosis for the in vivo samples.

The TUNEL technique fluorescently labels nicked DNA in apoptotic cells only.

Texas Red X (Molecular Probes, Eugene, OR) was used as a co-staining label to determine adherent macrophage morphology. At appropriate timepoints, samples were explanted, gently rinsed with PBS, and stained with the TUNEL in

situ cell death detection kit-fluorescein according to the manufacturer’s

instructions (Roche Applied Sciences, Indianapolis, IN). In brief, rinsed samples

were fixed with 4% paraformaldehyde in PBS for 1 hour at 25oC, rinsed twice with PBS, permeablized with 0.1% Triton X-100 in 0.1% sodium citrate for 2 minutes on ice, stained with the kit’s TUNEL reaction mixture and enzyme solution, rinsed again, and stained with Texas Red X for 30 minutes at 37oC.

Stained samples were then removed from the 24-well plate, mounted to

glass slides with gel mount (Biomeda, Foster City, CA), and imaged using CSLM.

A minimum of 200 cells were counted per sample to obtain the total number of

adherent cells per millimeter squared. Positively stained cells were considered to

be apoptotic. The data is presented as a percentage of apoptotic macrophages

of the total number of adherent macrophages and as the total number of

apoptotic and viable cells per square millimeter.

- 134 - Chapter V

Separate samples were stained with May-Grunwald and Giemsa as

previously described and examined for the classical morphological features of

apoptosis within 5-20x objective fields (463μm x 463μm) using OM.

Statistical Analysis

For the macrophage adhesion, fusion, and apoptosis studies,

measurements were taken from 5-8 fields on duplicate samples, yielding 10-16

datapoints per experiment and a total of 30-48 datapoints. The data was

collected from three experiments and analyzed using a one-way factorial ANOVA

post-hoc test (Bonfferoni/Dunn) with a 95% confidence level (5% alpha value).

For contact angle analysis, three fields on triplicate samples were

measured for advancing and receding contact angles. The advancing contact

angle measurements were obtained from an incremental addition of 5-2 l droplets. The 5 measurements were averaged together for the 3 fields to produce the mean advancing water contact angle of the single experiment. The presented data is the mean of 3 experiments in which each experiment is the

result of the mean of 15 datapoints. The receding contact angles were obtained

in a similar manner. Statistical analysis was performed using an unpaired t-test

with a 95% confidence level.

All statistical analysis was performed using the StatView 5.0.1.0 software

(SAS Institute Inc. Cary, NC). All data is presented as the mean ± standard error

of the mean (SEM) (n=3).

- 135 - Chapter V

Results

Surface Characterization

Contact angle measurements performed in both systems: (A) air

surrounding with a water probe and (B) water surrounding with a methylene

iodide probe; are shown in Table 5.2. System A was used as a standard contact

angle analysis system. Under dry conditions, the fluorocarbon SME modification

(109 ± 0) caused the surface chemistry of the unmodified Elasthane 80A (61 ± 2)

to be more hydrophobic, while the PEO SME modification (62 ± 3) had no effect

on the surface chemistry. After hydrating the sample for 1 hour, the fluorocarbon

SME modified surfaces became more hydrophilic (96 ± 5), but still statistically

different (p<0.0001) from the Elasthane 80A (70 ± 4). This increase in

hydrophilicity may be due to chain rearrangement and the migration of the

hydrophobic fluorocarbon SMEs away from the surface and into the bulk of the material. Hydration of the PEO SME modified polyurethanes for 1 hour resulted in a more hydrophilic surface chemistry (56 ± 1). Chain rearrangement of the

hydrophilic PEO SMEs towards the water-surface interface may have caused the

increase in hydrophilicity. The increase in the contact angle of Elasthane 80A

after 1 hour of hydration (61 ± 2 vs. 70 ± 4) may have resulted from an

interference of the water molecules with the hydrogen bonds within the polyether

urethane resulting in chain reorientation. Silicone modification provided a more

hydrophobic surface chemistry for PurSil 20 80A (90 ± 5) and CarboSil 20 90A

(86 ± 8) when unhydrated. Hydrating the silicone modified materials for 1 hour

did not affect the hydrophobicity of either silicone modified polyurethane.

- 136 - Chapter V

Table 5.2 Water Contact Angle Measurements on Polyurethane Materials@ Water Probe in Air^ Methylene Iodide Probe (System A) in Water+ Materials and (System B) Conditions₫ Advancing Receding Advancing Receding Contact Contact Contact Contact Angle (o) Angle (o) Angle (o) Angle (o) Elasthane 80A - Unhydrated 61 ± 2 50 ± 1 68 ± 5 57 ± 7 - Hydrated for 1 hr 70 ± 4 # 60 ± 3 # 74 ± 3 55 ± 4 - Hydrated for 24 hrs - - 77 ± 5 54 ± 6 PurSil 20 80A - Unhydrated 90 ± 5 # ¥ 74 ± 2 # ¥ - -

- Hydrated for 1 hr 88 ± 5 # ¥ 73 ± 9 # ¥ - - Elasthane 80A – Fluorocarbon - Unhydrated 109 ± 0 ¥ 95 ± 2 ¥ 77 ± 2 56 ± 4

- Hydrated for 1 hr 96 ± 5 # ¥ 77 ± 11 # ¥ 71 ± 4 53 ± 4 - Hydrated for 24 hrs - - 73 ± 3 51 ± 5 Elasthane 80A – PEO - Unhydrated 62 ± 3 42 ± 1 ¥ 96 ± 4 ¥ 89 ± 4 ¥

- Hydrated for 1 hr 56 ± 1 # 38 ± 9 # ¥ 98 ± 4 ¥ 98 ± 6 ¥ - Hydrated for 24 hrs - - 103 ± 3 ¥ 96 ± 6 ¥ Bionate 80A - Unhydrated 62 ± 1 48 ± 0 66 ± 5 46 ± 4 - Hydrated for 1 hr 69 ± 4 # 59 ± 3 # 75 ± 2 55 ± 6 - Hydrated for 24 hrs - - 80 ± 2 55 ± 7 CarboSil 20 90A - Unhydrated 86 ± 8 # * 71 ± 4 # * - - - Hydrated for 1 hr 91 ± 6 # * 73 ± 13 # * - - @ The data represents the mean ± SEM of three experiments. ₫ Surfaces were hydrated in PBS for 0, 1, and/or 24 hours. ^ For system A, more hydrophobic surfaces have a greater contact angle degree, while more hydrophilic surfaces have a lower contact angle degree. + For system B, more hydrophilic surfaces have a greater contact angle degree, while more hydrophobic surfaces have a lower contact angle degree. ¥ Statistically significant from Elasthane 80A (p=<0.05) * Statistically significant from Bionate 80A (p<0.05) # Data referenced from past studies in our laboratory by Patel et.al.29

- 137 - Chapter V

System B, the methylene iodide probe in water, was utilized to investigate

the effect of an aqueous environment on the materials as seen in culture. After

hydrating the samples for 1 hour, Elasthane 80A-F (71 ± 4), Elasthane 80A (74 ±

3), and Bionate 80A (75 ± 2) had similar hydrophilicity as seen by the contact

angles, while Elasthane 80A-P (98 ± 4) was more hydrophilic. This indicates that

in culture the fluorocarbon SME modification no longer increases the

hydrophobicity of Elasthane 80A, while the PEO SME modification resulted in a

more hydrophilic surface. The greatest change in contact angle was seen

between the unhydrated and hydrated for 1-hour samples, while no difference

was found between the hydrated for 1 and 24 hours samples for any of the materials. This suggests that the majority of the chain rearrangements occurred

during the first hour of hydration.

Macrophage Adhesion – In Vitro

As in the normal progression of adherent monocytes and macrophages,

cell morphology progressed from the expected rounded monocyte morphology

into the spread cytoplasm morphology of macrophages between days 0 and 10.

Figure 5.1 shows monocyte/macrophage adhesion on the silicone modified,

SME modified, and unmodified polyurethanes. Adherent cell densities

decreased from day 0 to day 10 as expected 37,15. The total number of adherent

cells was greatest at day 0. A large decrease in the adherent cell density was

seen between days 0 and 3. After day 3, the adherent cell densities decreased

linearly.

- 138 - Chapter V

6000 E80A PSil80A E80A-F ) 2 E80A-P 5000 B80A CSil90A PDMS 4000

3000

2000

1000 Number of Adherent Cells (cells/mm

0 Day 0Day 3Day 7Day 10

Figure 5.1: Monocyte/Macrophage Adhesion on Modified and Unmodified Polyurethane Materials, In Vitro. Mean ± SEM, n=3.

Initially at day 0, the fluorocarbon and PEO SME modified Elasthane 80A appeared to support slightly more macrophage adhesion than Elasthane 80A; however, one-way factorial ANOVA statistical analysis revealed that the materials were not significantly different on a 5% confidence level. PurSil 20

80A and CarboSil 20 90A also did not promote or inhibit macrophage adhesion in comparison to Elasthane 80A and Bionate 80A, respectively. After day 3, all materials supported similar numbers of adherent macrophages. Macrophage adhesion was similar on PDMS, silicone modified polyurethane, and the unmodified polyurethanes indicating that silicone did not have an effect on macrophage adhesion.

- 139 - Chapter V

6000

) E80A 2 PSil80A B80A 5000 CSil90A B90A PDMS

4000

3000

2000

1000 Number of Adherent Cells (cells/mm

0 Day 4 Day 7 Day 21

Figure 5.2: Monocyte/Macrophage Adhesion on Silicone Modified and Unmodified Polyurethane Materials, In Vivo. Mean ± SEM, n=4.

Macrophage Adhesion – In Vivo

Cell adhesion followed the normal progression between days 4 and 21.

Cell densities were greatest at day 4 and significantly decreased until day 21 on all materials. As shown in Figure 5.2, all materials supported similar cell densities at each timepoint. At day 4, optical microscopy and TUNEL dual staining revealed two layers of cells: one layer of adherent cells with a spread cellular morphology and aligned actin filaments; and a second layer of non- adherent cells. In between the two layers of cells appeared to be a layer of filaments, possibly fibrin. Except for B80A, all materials supported similar minimal percentages of non-adherent cells, where Bionate 80A had 0% ± 0% non-adherent cells. Differential cell analysis with optical microscopy showed that

- 140 - Chapter V

there were some PMNs on the material surface; particularly on CarboSil 20 90A

and PDMS. Table 5.3 shows the percentage of non-adherent cells and the percentage of PMNs on each surface at day 4 as revealed by TUNEL/Texas Red

X dual staining and optical microscopy respectively. No non-adherent cells or

PMNs were observed on any of the surfaces at days 7 and 21.

Table 5.3 Differential Cell Analysis of Cell Layers at Day 4* Material PMNs ª Non-adherent Cells ^ Apoptotic Non-adherent Cells (%) (%) (%) E80A 1% ± 1% 39% ± 39% 16% ± 16% PSil 80A 0% ± 0% 18% ± 22% 11% ± 14% B80A 0% ± 0% 0% ± 0% 0% ± 0% CSil 90A 8% ± 4% 19% ± 19% 0% ± 0% B90A 0% ± 0% 25% ± 25% 0% ± 0% PDMS 6% ± 7% 17% ± 8% 0% ± 0% * All values are the mean ± SEM of 3 experiments. ª The percentage of PMNs determined on each surface from OM cell differentiation analysis. PMNs may be a part of the upper layer of cells seen at day 4. ^ The percentage of non-adherent cells on each surface as revealed by dual TUNEL and Texas Red V staining and CSLM. The non-adherent cells did not have the typical organized actin filaments associated with adherent macrophages. These cells may also be a part of the upper layer of cells seen at day 4; however the non-adherent cells are not necessarily PMNs. ° The percentage of non-adherent cells that are apoptotic on each surface at day 4. These apoptotic, non-adherent cells did not stain for the typical organized actin filaments, but did stain positive for apoptosis with TUNEL.

- 141 - Chapter V

Macrophage Fusion – In Vitro

The capacity for macrophages to fuse into FBGC was examined on each biomaterial surface (Figure 5.3). In the presence of IL-4, macrophages aggregated on the surfaces of all materials by days 7 and 10. At day 7, little to no fusion was seen on all polyurethane materials (less than 5%). PDMS promoted 13% of the macrophages to fuse into FBGCs. Percent fusion tended to be greater on PurSil 20 80A and CarboSil 20 90A in comparison to Elasthane

80A and Bionate 80A, respectively. The trend for FBGC formation at day 10 is as follows: E80A-F = E80A-P ≤ E80A ≤ B80A < PSil80A < CSil90A << PDMS.

These significant differences in FBGC formation on PDMS and the silicone modified and unmodified polyurethanes at day 10 can be seen in the photomicrographs in Figure 5.4.

Macrophage Fusion – In Vivo

The percentage of FBGC formation in vivo is shown in Figure 5.5. At day

4, FBGC fusion is minimal on all surfaces (less than 2%). By day 21, PurSil 20

80A (70%±12%) promoted a greater percentage of FBGCs than Elasthane 80A

(47%±15%); as did CarboSil 20 90A (87%±9%) in comparison to Bionate 80A

(42%±17%) and Bionate 90A (36%±16%).

- 142 - Chapter V

25% without IL-4 A with IL-4

20%

15%

10%

0% E80A PSil80A E80A-F E80A-P B80A CSil90A PDMS

25% B

20%

15% Percentage of Fused Cells (%)

10%

0% E80A PSil80A E80A-F E80A-P B80A CSil90A PDMS

Figure 5.3: Foreign Body Giant Cell Formation on Modified and Unmodified Polyurethane Materials at Days 7 (A) and 10 (B), In Vitro. Surfaces with missing datapoints (i.e. E80A, E80A-F, E80A-P at day 7) had less than 1 percentage of fused cells on the surface. Mean ± SEM, n=3.

- 143 - Chapter V

a b

100 μm 100 μm

d c

100 μm 100 μm

e f

100 μm 100 μm

g h

100 μm 100 μm

i j

100 μm 100 μm

Figure 5.4: Optical Micrographs of Adherent Macrophages and IL-4 Induced FBGCs on PDMS and Silicone Modified and Unmodified Polyurethanes at Day 10, In Vitro. Adherent cells are shown on the surfaces of Elasthane 80A (a,b); PurSil 20 80A (c,d); Bionate 80A (e,f); CarboSil 20 90A (g,h); and PDMS (i,j). Scale bars represent 100μm.

- 144 - Chapter V

100% E80A PSil80A B80A CSil90A 80% B90A PDMS

60%

40%

20% Percentage of FBGCs (%)

0% Day 4 Day 7 Day 21

Figure 5.5: FBGC Formation on Silicone Modified and Unmodified Polyurethane Materials, In Vivo. Surfaces with missing data points (i.e. E80A and B80A at day 4) had less than 1 percentage of fused cells on the surface. Mean ± SEM, n=3.

Macrophage Apoptosis – In Vitro

Figure 5.6 shows the percentage of apoptotic macrophages, while Figure

5.7 shows the surface densities of apoptotic and viable adherent macrophages for the polyurethane materials. E80A, E80A-F, E80A-P, and B80A materials promoted similar minimal amounts of macrophage apoptosis at days 7 and 10 in the presence and absence of IL-4. At day 7, PurSil 20 80A promoted a similar percentage of apoptotic macrophages as Elasthane 80A, while CarboSil 20 90A promoted more apoptosis than Bionate 80A. By day 10, PurSil 20 80A and

CarboSil 20 90A promoted more apoptotic cells than Elasthane 80A

- 145 - Chapter V

and Bionate 80A, respectively, in the absence of IL-4. Also, the percentage of

apoptotic cells decreased on PurSil 20 80A and CarboSil 20 90A at day 10 due

to the addition of IL-4. This decrease in apoptotic cells on PurSil 20 80A and

CarboSil 20 90A is correlated to higher level of fusion on silicone modified

polyurethanes.

Macrophage Apoptosis – In Vivo

The percentage of apoptotic cells in comparison to the total number of individual

cells is shown in Figure 5.8. At day 4, PurSil 20 80A (44% ± 6%) and CarboSil

20 90A (32% ± 0%) promoted more cell apoptosis than Elasthane 80A (21% ±

2%), Bionate 80 (25% ± 4%), and Bionate 90A (13% ± 1%). As discussed earlier, some of the cells were non-adherent at day 4. Of the non- adherent cells

on Elasthane 80A (39% ± 39%) and PurSil 20 80A (18% ± 22%), the following

percentage of non-adherent cells were apoptotic: 16% ± 16% and 11% ± 14%,

respectively (Table 5.3). The non-adherent cells on all other materials were

viable. By day 7, only adherent macrophages were found. Percent apoptosis

was similar on all materials at day 7 and 21.

- 146 - Chapter V

50% without IL-4 A with IL-4

40%

30%

20%

10%

0% E80A PSil80A E80A-F E80A-P B80A CSil90A

50% B

40% Percentage of Apoptotic Cells (%)

30%

20%

10%

0% E80A PSil80A E80A-F E80A-P B80A CSil90A

Figure 5.6: Annexin V Apoptosis Analysis of Cells Adherent to Unmodified and Modified Polyurethane Materials, In Vitro. Mean ± SEM, n=3.

- 147 - Chapter V

2000 Viable Apoptotic

1500

1000 ) 2

500

0 E80A w/ IL-4 E80A PSil80A IL-4 w/ PSil80A E80A-F IL-4 w/ E80A-F E80A-P IL-4 w/ E80A-P B80A w/ IL-4 B80A CSil90A IL-4 w/ CSil90A

2000

1500 Number of Adherent Cells (cells/mm 1000

500

0 E80A w/ IL-4 E80A PSil80A IL-4 w/ PSil80A E80A-F IL-4 w/ E80A-F E80A-P IL-4 w/ E80A-P B80A w/ IL-4 B80A CSil90A IL-4 w/ CSil90A

Figure 5.7: Viable and Apoptotic Adherent Cell Densities on the Unmodified and Modified Polyurethane Materials, In Vitro. Mean ± SEM, n=3.

- 148 - Chapter V

60% E 80A PSil 20 B-80A CSil 20 50% B90A PDMS

40%

30%

20%

10% Percentage of Apoptotic Cells (%) 0% Day 4 Day 7 Day 21

Figure 5.8: Annexin V Apoptosis Analysis of Cells Adherent to Silicone Modified and Unmodified Polyurethane Materials, In Vivo. Surfaces with missing data points (PDMS at days 7 and 21) are not available. Mean ± SEM, n=3.

- 149 - Chapter V

Discussion

Adherent macrophages can be detrimental to a polyurethane surface by

releasing degradative enzyme and oxygen radicals that result in chain scission

and the eventual failure of the device. Modulation of cell-biomaterial interactions

such as adhesion, fusion, and apoptosis could be a possible mechanism of

reducing these degradative effects. Incorporation of more biostable and less

biointeractive materials into the polyurethane polymer chain may provide a

means to modulate the cell-biomaterial interactions without compromising the

mechanical properties of the polyurethanes.

Previous studies in our laboratory have shown that the fluorocarbon SME,

PEO SME, and silicone modifications are present on the polyurethane surface

using ATR-FTIR.29 Specifically, the presence of the fluorocarbon SME at the

surface was noted by the presence of a small peak at 1150cm-1, which is the

characteristic peak for the C-F bond. The characteristic peak for PEO overlaps

with the polyurethane ether peak; therefore, the small increase observed in the

ether peak is indicative of the presence of PEO SME segments at the surface.

Three peaks (800 cm-1, 1253 cm-1, 1018 cm-1) indicate the presence of the

PDMS groups at the polyurethane surface.

Although past studies have demonstrated that fluorocarbon surfaces

inhibit the adsorption of proteins required for macrophage adhesion and PEO- rich surfaces inhibit protein adsorption and cellular adhesion, Jenny et. al.

showed that macrophages are capable of adhering to a wide variety of surface

chemistries in a similar manner.15,16,17,19 Therefore, it was not surprising to find

- 150 - Chapter V

no significant difference in macrophage adhesion between Elasthane 80A-F,

Elasthane 80A-P, and Elasthane 80A by days 7 and 10. However, it was unique

to find that macrophage fusion and apoptosis was not significantly different on these materials. It is important to note that the only difference seen between either of the SME modified and the unmodified polyurethanes was the smaller, more rounded morphology of macrophages adherent to Elasthane 80A-P in comparison to Elasthane 80A.

Contact angle analysis performed in air under hydrophobic, dry conditions showed that the surface chemistry of Elasthane 80A-F was more hydrophobic than Elasthane 80A (Table 5.2). This confirmed Patel’s findings that the fluorocarbon SMEs were initially present at the material surface and capable of influencing the surface chemistry. However, under aqueous conditions, the surface chemistries of the fluorocarbon SME modified and unmodified polyurethanes were comparable. This suggests that in culture media fluorocarbon SMEs rearrange away from the surface in order to avoid the energetically unfavorable interface with the aqueous environment. As a result, this allows the polyurethane hard and soft segments to dominate the surface producing a surface similar to the unmodified polyurethane. This explains our similar macrophage adhesion, fusion, and apoptosis findings.

Unlike with the fluorocarbon SMEs, contact angle analysis in both air and aqueous conditions indicate an increase in hydrophilicity on Elasthane 80A-P in comparison to Elasthane 80A. In aqueous conditions, the PEO SMEs migrate to the polyurethane surface to produce a more energetically favorable interface.

- 151 - Chapter V

The confirmed presence of the PEO SMEs at the surface both in air and aqueous

environments further suggest that the PEO SMEs would affect macrophage adhesion and activity on the polyurethane surface; however, no significant differences were seen between the PEO SME modified and unmodified polyurethanes. Past studies by Jenney et. al. demonstrated that not only was the

presence of PEO segments important to macrophage adhesion but the molecular

mass of the PEO segments.15 They found that PEO segments with molecular masses between 200 and 4600Da were capable of shifting the surface chemistry while not altering macrophage adhesion or FBGC formation. Macrophage adhesion was reduced only when the PEO segments had a molecular mass of

4600 Da or 18.5kDa, while FBGC formation was reduced with the longer chains of 18.5kDa. In our study, the PEO SMEs have a molecular mass of 2kDa, which is sufficient to alter surface chemistry; however not sufficient enough to reduce macrophage adhesion or FBGC formation. Possibly an increase in the molecular mass of the PEO SME segments to at least 18.5 kDa would reduce the cell- biomaterial interactions.

Research by Patel et. al. in our laboratory confirmed the presence of silicone on the surface of polyurethanes with ATR-FTIR while contact angle measurements indicated an increase in hydrophobicity of the silicone modified verses unmodified polyurethanes.29 Water contact angles on silicone modified samples did not change following 1 h hydration indicating surface reorientation in water is minimal to very slow. These results are in agreement with Ward’s

- 152 - Chapter V

findings that silicone modification of polyurethanes promotes a silicone-rich

surface due to high mobility and low surface energy of PDMS.

Macrophages have been shown to be adaptable to various surface

chemistries allowing for similar adhesion to a wide range of surface

chemistries.37 Previous studies in our laboratory on alkyl-silane surfaces have

revealed that monocyte/macrophage adhesion over a 10-day, in vitro culture was

unaffected by surface chemistry, while the density of IL-4 induced FBGCs was

correlated to carbon content. Surfaces with shorter alkyl chains (DM and C1)

appeared to promote FBGC formation. As expected, macrophage adhesion was

not affected by silicone modification after day 3 in vitro and after day 7 in vivo

(Figure 5.1 & 5.2). At day 4 in vivo, some PMNs and non-adherent cells were found on the material surface (Table 5.3). This is indicative of normal acute inflammation and was resolved by day 7. FBGC formation was promoted by the silicone modified polyurethanes in vitro and in vivo similarly to dimethyl silane- modified glass (DM) (Figures 5.3, 5.4, & 5.5).38 These results are also

consistent with Brodbeck’s findings that hydrophobic surfaces promote FBGC

formation.10

Finally, in vitro, the percentage of apoptotic cells on the silicone modified

polyurethanes (PurSil 20 80A and CarboSil 20 90A) tended to be greater than on

the unmodified polyurethanes (Elasthane 80A and Bionate 80A, respectively) as

shown in Figures 5.6, 5.7 & 5.8. Also, a greater number of apoptotic cells were

seen on silicone modified polyurethane after 10 days in vitro. This confirms

previous results suggesting that macrophage apoptosis is surface dependent.10

- 153 - Chapter V

Further investigations in vivo revealed that initially silicone modification increased

macrophage apoptosis; however, at later timepoints, no effect from the silicone modified polyurethanes is seen on macrophage apoptosis. The in vitro and early

in vivo apoptosis results contradict Brodbeck’s findings that hydrophobic surfaces

inhibit apoptosis; however, the increase in apoptotic macrophages may instead

be due to differences in surface compositions. For later timepoints in vivo, the

same differences in surface composition may keep the more hydrophobic

silicone modified polyurethanes from inhibiting apoptosis.

An investigation into the effect of PDMS on macrophage adhesion, fusion,

and apoptosis as compared to the silicone modified and unmodified polyurethane

was performed to further understand the effect of PDMS modification (apoptosis

data not shown for PDMS, in vitro). In vitro and in vivo PDMS results indicated

that macrophage adhesion was similar on PDMS, PurSil 20 80A and Elasthane

80A. Also, macrophage fusion was significantly greater on PDMS in comparison

to PurSil 20 80A and Elasthane 80A, explaining the promotion of fusion seen on

the silicone modified polyurethanes.

The relationship between macrophage fusion and apoptosis has been

considered. In the presence of IL-4 in vitro, macrophage fusion decreased on

PDMS and the silicone modified polyurethanes. The decrease in apoptosis

correlates to an increase in fusion seen in the presence of IL-4. Also, in vivo, at

day 4 macrophage apoptosis was greater on the silicone modified verses the

unmodified polyurethanes. By day 7, macrophage fusion increased on the

silicone modified polyurethanes in comparison to the unmodified polyurethanes,

- 154 - Chapter V

while macrophage apoptosis became comparable between the materials. These

correlations support Brodbeck’s hypothesis that macrophages fuse into FBGC in

order to escape apoptosis.

Conclusions

Modulation of macrophage adhesion, fusion, and apoptosis is an effective

mechanism to ultimately control the biostability of polyurethanes over long

periods of time. It was hypothesized that the incorporation of more biostable

and/or less biointeractive chemistries (i.e. fluorocarbon, PEO, and PDMS) into

the polymer chain may modify the surface chemistry of the material and

ultimately have an effect on macrophage adhesion, fusion, and apoptosis. In this

study, it was found that none of the modifications altered macrophage adhesion

and only the silicone modification promoted fusion in comparison to the

unmodified polyurethane. The fluorocarbon SMEs modified polyurethanes did

not have an effect on macrophage adhesion, fusion, or apoptosis in vitro due to

the rearrangement of the hydrophobic SME segments away from the hydrophilic

surroundings into the bulk of the polyurethane. In contrast, the PEO SME

segments were still present at the polyurethane surface in the culture media in vitro; however, the PEO molecular mass was insufficient to induce a cellular response as confirmed by past studies in this laboratory. Finally, macrophages were adaptive enough to still adhere to the silicone-modified surfaces; however the silicone modification was effective at modifying macrophage fusion and apoptosis, in vitro. In the absence of IL-4, macrophage apoptosis was promoted

- 155 - Chapter V with little FBGC formation. When induced with IL-4, FBGC formation increased significantly and macrophage apoptosis was inhibited.

It was hypothesized from in vitro results that silicone modification may provide a possible means for modulation the ultimate fate of adherent macrophages. If left unstimulated, adherent macrophages will undergo apoptosis on silicone modified surfaces; however if stimulated by the proper cytokines, macrophages will be more inclined to fuse into FBGCs. The in vivo analysis of the materials showed that macrophage fusion was promoted on the silicone modified polyurethanes over time. In conclusion, polyurethanes incorporated with surface modifying endgroups may still affect macrophage adhesion and activity with a few alterations in the SME segments. Silicone modification does affect macrophage adhesion and activity and may be an effective method of controlling the fate of macrophages in vivo.

References

1. Anderson JM. Mechanisms of Inflammation and Infection with Implanted

Devices. Cardiovascular Pathology 1993;2(3):33S-41S.

2. Anderson JM, DeFife K, McNally A, Collier T, Jenney C. Monocyte,

macrophage, and foreign body giant cell interactions with molecularly

engineered surfaces. Journal of Materials Science: Materials in Medicine

1999;10:579-588.

3. Anderson JM. Biological responses to materials. Annual Reviews of

Material Research 2001;31:81-110.

- 156 - Chapter V

4. Anderson JM. Multinucleated giant cells. Hematology 2000;7:40-47.

5. Kao WJ, Hiltner A, Anderson JM, Lodoen GA. Theoretical analysis of in

vivo macrophage adhesion and foreign body giant cell formation on

strained poly (etherurethane urea) elastomers. Journal of Biomedical

Material Research 1994;28:819-829.

6. Stokes KS, Coury A, Urbanski P. Autooxidative degradation of implanted

polyether polyurethane devices. Journal of Biomaterial Applications

1987;1:411-448.

7. Zhao Q, Topham N, Anderson JM, Hiltner A, Lodoen G, Payet CR.

Foreign-body giant cells and polyurethane biostability: In vivo correlation

of cell adhesion and surface cracking. Journal of Biomedical Material

Research 1991;25:177-183.

8. Wiggins MJ, Wilkoff B, Anderson JM, Hiltner A. Biodegradation of

polyether polyurethane inner insulation in bipolar pacemaker leads.

Journal of Biomedical Material Research 2001;58:302-307.

9. Schubert MA, Wiggins MJ, Anderson JM, Hiltner A. Comparsion of two

antioxidants for poly(etherurethane urea) in an accelerated in vitro

biodegradation system. Journal of Biomedical Material Research

1997;34:493-505.

10. Brodbeck W, Shive M, Colton E, Nakayama Y, Matsuda T, Anderson JM.

Influence of biomaterial surface chemistry on the apoptosis of adherent

cells. Journal of Biomedical Material Research 2001;55:661-668.

- 157 - Chapter V

11. Brown JM. Polyurethane and silcone: mythes and misconceptions. Journal

of Intravenous Nursing 1995;18(3):120-122.

12. Hsu SH, Tseng HJ, Wu MS. Comparative in vitro evaluation of Two

Different Preparations of Small Diameter Polyurethane Vascular Grafts.

Artificial Organs 2000;24:119-128.

13. Stokes KS, McVenes R, Anderson JM. Polyurethane elastomer

biostability. Journal of Biomaterial Applications 1995;9:331-353.

14. Mathur AB, Collier TO, Kao WJ, Wiggins MJ, Schubert MA, Hiltner A,

Anderson JM. In vivo biocompatibility and biostability of modified

polyurethanes. Journal of Biomedical Material Research 1997(36):246-

257.

15. Jenney C, Anderson JM. Effects of surface-coupled polyethylene oxide on

human macrophage adhesion and foreign body giant cell formation in

vitro. Journal of Biomedical Material Research 1999;44:206-216.

16. Cima LG. Polymer substrates for controlled biological interactions. Journal

of Cell Biochemistry 1994;56:155-161.

17. Gombotz WR, Guanghui W, Horbett TA, Hoffman AS. Protein adsorption

to poly(ethylene oxide) surfaces. Journal of Biomedical Material Research

1991;25:1547-1562.

18. Lee JH, Jeong BJ, Lee HB. Plasma protein adsorption and platelet

adhesion onto comb-like PEO gradient surfaces. Journal of Biomedical

Material Research 1997;34:105-114.

- 158 - Chapter V

19. Boffa GA, Lucien N, Faure A, Boffa MC. Polytetrafluorethylene-N-

vinylpyrrolidone graft copolymers affinity with plasma proteins. Journal of

Biomedical Material Research 1977;11(3):317-337.

20. Jenney CR, Anderson JM. Adsorbed serum proteins responsible for

surface dependent human macrophage behavior. Journal of Biomedical

Material Research 2000;49:435-447.

21. Jahangir A, McClung W, Cornelius R, McCloskey C, Brash J, Santerre J.

Fluorinated suface-modifying macromolecules: modulating adhesive

protein and platelet interactions on a polyether-urethane. Journal of

Biomedical Material Research 2002;60:135-147.

22. Tang Y, Santerre J, Labow R, Taylor D. Use fo surface-modifying

macromolecules to enhance the biostability of segmented polyurethanes.

Journal of Biomedical Material Research 1997;35:371-381.

23. Ward R. Surface modification prior to surface formation: control of polymer

surface properties via bulk composition. Medical Plastics Biomaterials

1995;2(Spring):34-41.

24. Hergenrother R, Yu X, Cooper S. Blood contacting properties of

polydimethylsiloxane polyurea-polyurethanes. Biomaterials

1994;15(8):635-640.

25. Lim F, Yang C, Cooper S. Synthesis, characterization and ex vivo

evaluationof polydimethylsiloxane polyurea-urethane. Biomaterials

1994;15(6):408-416.

- 159 - Chapter V

26. Martin D, Poole W, Gunatillake P, McCarthy S, Meijs G, Schindhelm K.

Polydimethylsiloxane/polyether-mixed macrodiol-based polyurethane

elastomers: biostability. Biomaterials 2000;21:1021-1024.

27. Ward R, White K, Gill R, Lim F. The effect of phase separation and end

group chemistry on in vivo biostability of polyurethanes. 1996; Washington

DC. p 17-25.

28. Ward R, White K; Surface-modifying end-groups for biomedical polymers.

USA patent 5,589,563. 1996.

29. Patel J, Ebert M, Stokes K, Ward R, Anderson JM. Inhibition of bacterial

and leukocyte adhesion under shear stress conditions by material surface

chemistry. Journal of Biomaterial Science Polymer Edition

2003;14(3):279-295.

30. Kalman PG, Ward CA, McKeown NB, McCullough D, Romaschin AD.

Improved biocompatibility of silicone rubber by removal of surface

entrapped air nuclei. Journal of Biomedical Material Research

1991;25:199-211.

31. McNally AK, Anderson JM. Complement C3 participation in monocyte

adhesion to different surfaces. Proceedings of the National Academy

Science 1994;91:10119-10123.

32. McNally A, Anderson JM. Interleukin-4 induces foreign body giant cells

from human monocytes/macrophages. American Journal of Pathology

1995;147:1487-1499.

- 160 - Chapter V

33. Marchant RE, Hiltner A, Hamlin C, Rabinovich A, Slobodkin R, Anderson

JM. In vivo biocompatibility studies. I. The cage implant system and a

biodegradable hydrogel. Journal of Biomedical Material Research

1983;17(2):301-325.

34. Suggs LJ, Shive MS, Garcia CA, Anderson JM, Mikos AG. In vitro

cytotoxicity and in vivo biocompatibility of poly(propylene fumarate-co-

ethylene glycol) hydrogels. Journal of Biomedical Material Research

1999;46(1):22-32.

35. Kao WJ, Anderson JM. The Cage Implant Testing System. In: von Recum

AF, editor. Handbook of Biomaterials Evaluation. Philadelphia, PA: Taylor

and Francis; 1998. p 659-669.

36. Shive M, Brodbeck W, Colton E, Anderson JM. Shear stress and material

surface effects on adherent human monocyte apoptosis. Journal of

Biomedical Material Research 2002;60(148-158).

37. DeFife K, Shive MS, Hagen KM, Clapper DL, Anderson JM. Effects of

photochemically immobilized polymer coatings on protein adsorptoin, cell

adheison, and the foreign body reaction to silicone rubber. Journal of

Biomedical Material Research 1999;44:298-307.

38. Jenney C, DeFife KM, Colton E, Anderson JM. Human

monocyte/macrophage adhesion, macrophage motility and IL-4 induced

foreign body giant cell formation on silane modified surfaces in vitro.

Journal of Biomedical Material Research 1998;41:171-184.

- 161 - Chapter VI

Chapter VI: Dynamic Systems Model of Cellular Interactions with Biomaterial Surfaces

Introduction and Significance

Macrophage Behavior in Biological Responses

Inflammation, wound healing, and the foreign body reaction take place

following the implantation of any biomedical implant. These processes have

been assiduously characterized with respect to the general time course, cells

involved, cell-cell interactions, and cell-biomaterial interactions.1-3 In brief,

inflammation is characterized by the infiltration of neutrophils and monocytes to

the injury site, which can subsequently adhere to the biomaterial surface. Within

a few days, the neutrophil population at the implant site decreases. In contrast,

the monocytes differentiate into adherent macrophages and remain as a key

player in inflammation, the foreign body reaction, and wound healing. Adherent

macrophages release degradative enzymes and oxygen radicals in an attempt to

degrade the biomaterial surface.

Due to size disparity, macrophages undergo what has been termed

“frustrated phagocytosis” and fuse to form foreign body giant cells (FBGCs).4

Adherent FBGCs further release these degradative agents and are capable of degrading a biomaterial surface as experimentally shown by the resulting pitting and cracking of a polyurethane surface directly beneath once-adherent FBGCs.

As part of the foreign body reaction, FBGCs have been clinically shown to be present at the biomaterial surface and in the surrounding milieu for years after implantation. Monocytes, macrophages, and foreign body giant cells also

- 162 - Chapter VI

produce a variety of cytokines (i.e. IL-1β, IL-6, IL-10, TNF-α,and TGF-β) and

chemokines (i.e. IL-8 and MIP-1β) that can direct other cells in the inflammation

(i.e. PMNs, lymphocytes, and endothelial cells) and wound healing responses

(i.e. fibroblasts).5-8 In addition, these monocyte-derived cells can produce matrix

metalloproteinases (MMPs) and tissue inhibitor of matrix metalloproteinase

(TIMPs) that can further inhibit or promote the formation of a surrounding

extracellular matrix.9-11 Theoretically, via these secreted proteins, the

monocytes, macrophages, and FBGCs can either maintain the implant site in a

state of inflammation or cause it to progress toward granulation tissue formation, fibrosis, and wound healing.

Directing Cell-Biomaterial Interactions

Significant research has been done investigating the effects of biomaterials on cellular behavior. Surface properties, such as texture and

chemistry, have been studied and shown to affect cellular behaviors; in general,

cell adhesion, activation, proliferation, and viability. With respect to monocytes and macrophages, it has been shown that monocyte/macrophage adhesion, activation, apoptosis, and fusion into foreign body giant cells are modulated with biomaterial surface chemistry. The relationship between surface chemistry and cellular behaviors has led to a dogma or commonly accept belief that hydrophobic surfaces promote fast and high levels of protein adsorption and inhibit cellular adhesion, while hydrophilic surfaces inhibit protein adsorption and allow for variable to high amounts of cellular adhesion. Recent studies

- 163 - Chapter VI

investigating the effect of surface chemistry on monocyte and macrophage

adhesion and apoptosis have begun to refute these relationships suggesting

hydrophobic surfaces promote monocyte/macrophage adhesion and inhibit apoptosis while hydrophilic surfaces inhibit monocyte/macrophage adhesion and promote apoptosis.12,13

Due to the significant roles of monocytes, macrophages, and FBGCs in inflammation, wound healing, and the foreign body reaction, cellular events such as adhesion, activation, fusion, apoptosis, and cytokine production can aid in directing these biological responses. Intuitively, increased levels of monocyte/macrophage adhesion and macrophage fusion with decreased levels of apoptosis leads to increased formation of FBGCs further advancing the foreign body reaction. Conversely, decreasing monocyte and macrophage adhesion, decreasing macrophage fusion, and/or increasing apoptosis inhibits the foreign body reaction. Similar statements could be made in regards to the production of specific cytokines, chemokines, MMPs, and/or TIMPs with respect to inflammation and wound healing. A quantitative kinetic understanding of these processes (i.e. adhesion, activation, fusion, apoptosis, and cytokine production) would be useful in further clarifying the above cause and effect statements. One or more of these processes may outweigh the others in their effect on the ultimate system response. A kinetic mathematical model investigating monocyte and macrophage adhesion, differentiation, apoptosis, and FBGC formation can be designed that will determine the relative rates of these processes, show the affect of surface chemistry on the kinetics of these processes, and be utilized to

- 164 - Chapter VI

further design polymers that modulate monocyte, macrophage, and FBGC

behavior in a way to direct inflammation, wound healing, and the foreign body

reaction.

The desired ultimate response to a biomaterial varies with application. A

suppression of the foreign body reaction may be ideal for an application in which

polymer degradation is a concern for it can significantly degrade the polymer

rendering the device ineffective. An example of this is found with the foreign

body giant cell degradation and resulting full thickness cracking of polyurethanes

used for pacemaker lead insulation, which can ultimately short the lead circuit and require the lead to be replaced to restore the function of the pacemaker.

Following the new hypothesis that hydrophilic materials inhibit monocyte/macrophage adhesion and fusion, utilizing a hydrophilic polymer or modulating the surface chemistry of the existing polymer via a coating, incorporation into the polymer, or another technique would be beneficial in prolonging the life of the pacemaker lead.

Another application (i.e. vascular graft) may utilize adherent macrophages and foreign body giant cells to provide enzymes needed to degrade an enzymatically susceptible polymer scaffold at a rate allowing for biomimetic replacement of the polymer scaffold with extracellular matrix (ECM) and the cells of interest for that site (i.e. endothelial cells, smooth muscle cells, and fibroblasts). For these scaffolds, a hydrophobic surface chemistry would promote monocyte/macrophage adhesion and FBGC formation. Many tissue engineering constructs utilize hydrolytically degradable polymers to breakdown the polymer

- 165 - Chapter VI

scaffold and utilizes a hydrophilic chemistry to increase the amount of water

associated with the polymer scaffold; therefore, a trade-off between the

degradation mechanisms would need to be considered.

Other applications (i.e. a drug delivery device or scaffold) may require the

suppression of the fibrotic capsule that can result in wound healing response. A

polymer that causes adherent monocytes, macrophages, and foreign body giant

cells to minimize the recruitment of fibroblasts and/or produce MMPs to degrade

the fibrotic capsule would be useful. In short, the criteria for an ideal surface can

vary and have numerous trade-offs that need to be considered and accounted for. A predictive mathematical model could aid in determining which polymer surface chemistries promote the above desired responses. Despite numerous investigators utilizing in vitro cell culture systems containing monocytes, macrophages, and FBGCs to elucidate the in vivo cellular responses, few mathematical models have been developed regarding cellular adhesion.

Previous Cellular Mathematical Models

Studies have developed models to investigate cell-cell adhesion, the spatial interactions between tumor-associated macrophages, tumor cells, and normal tissue cells, and the role of macrophages in angiogenesis.14-20 The cell-

cell adhesion model by Coombs et.al. analyzes cell-cell adhesion based on the

interactions of multiple receptor-ligand interactions with respect to the physical

properties of the cellular membranes and concentration and physical dimensions

of the ligand-receptor pairs.20 This model’s focus is not applicable to cell-

- 166 - Chapter VI

biomaterial interactions. Another model by Bigerelle and Anselme addresses the

kinetics of cell adhesion with biomaterials with respect to material nature, surface

topography, and surface chemistry; however the cells of interest are osteoblasts,

which undergo proliferation, unlike monocytes/macrophages, and do not fuse. In addition, this model focuses on adhesion in terms of strength not cell number and does not account for apoptosis or cell detachment. Only one model addressed macrophage adhesion and fusion. Previous studies in our laboratory by Zhao, et.al. mathematically modeled the probability of the FBGCs size distribution (i.e. number of nuclei in each FBGC) at a given time in vivo on biomaterial surfaces.17,19 The parameters of the system were the initial density of FBGCs

measured, the initial density of cells or macrophages measured, time, and a

value that accounts for the cell area covered by 2-5 macrophages. The model

was capable of predicting FBGC size distribution and showed it to be material-

dependent. The model however determines the probability of the size

distribution, without accounting for cell differentiation, detachment, and

apoptosis.

A New Dynamic Model of Our Macrophage In Vitro Cell Culture System

A predictive mathematical model of the in vitro cell culture system would

be useful in defining the relationship between surface chemistry and the resulting

monocyte, macrophage, and FBGCs cellular behaviors discussed above.

Therefore, this ongoing research aims to develop a quantitative, predictive

mathematical model of our laboratory’s in vitro monocyte cell culture system.

- 167 - Chapter VI

The dynamic systems model examined the processes of monocyte adhesion, differentiation in macrophages, apoptosis, detachment, fusion into FBGCs, and cytokine production using ordinary differential equations to determine the numbers of adherent monocytes, adherent macrophages, adherent FBGCs of varying sizes, apoptotic cells, and detached cells. Experiments were performed to determine the observable variables of the collective average number of adherent monocytes and macrophages, the average number of adherent FBGC and FBGC cell sizes, the average number of detached cells, the average number of adherent apoptotic cells, and the average concentration of a particular cytokine produced with respect to one material. Parameter estimation will be conducted to calculate the rate constants for the above processes. Subsequent simulations and model analysis will occur. The model’s results, conclusions and predications will be interpreted and compared to the real cell culture system and revisions to the model will be made as needed to minimize error.

The resulting dynamic systems model will be utilized to study the effects of surface chemistry; specifically, hydrophobicity, hydrophilicity, and surface charge; on monocyte, macrophage, and FBGC behavior in order to determine the material-dependent nature of the model parameters. A relationship between the material chemistries and cellular behaviors will be formulated based upon experimental quantitative data to further develop a mechanistic model based on these relationships. Parameter estimation and model fitting will be utilized to determine surface chemistry dependent rate constants. Ideally, the resulting rate constants will indicate that one or two of the above mentioned processes are

- 168 - Chapter VI

material surface chemistry dependent, while the remaining processes are

dependent on the cell’s nature and not the surface chemistry. This situation will

elucidate the cellular processes that are directly affected by surface chemistry

and further aid in the prediction of the optimum application of surfaces for

biomedical material applications.

In Vitro Cell Culture Experiments

The in vitro cell system consists of adherent monocyte, macrophages, and

foreign body giant cells as described below. Monocytes are isolated from the

whole, human blood of healthy unmedicated donors at day 0 and adhere to the

material surface within 2 hours. In vivo, the blood monocyte is defined as a

tissue macrophage upon exiting the vasculature into the surrounding tissue. By

that distinction, monocytes do not undergo fusion, while tissue macrophages do

fuse to form FBGCs. In vitro, the blood-isolated adherent cells are defined as

monocytes until differentiation into macrophages as noted by a change in

morphology from a more rounded cell to a more spread morphology.21,22 This

cell differentiation occurs within approximately 3 days. By day 7 and 10,

adherent macrophages fuse together to form foreign body giant cells (FBGCs).

A FBGC is defined as a multinucleated cell with three or more nuclei. FBGCs

are defined as a cell having three or more nuclei versus a fused cell containing two nuclei based upon past precedence and the increased accuracy of

determining a fused versus non-fused cell associated with a cell containing three

nuclei. A microscopically observed cell with two nuclei may be fused, in the

- 169 - Chapter VI process of fusing, or merely associated, while an observed cell containing three nuclei is more likely to be fused verses not fused. Therefore, a cell containing two nuclei is considered to be a macrophage despite its fused state. Adherent monocytes and macrophages may undergo programmed cell death (apoptosis) or become detached from the material surface instead of progressing towards

FBGCs at anytime.23,24 A distribution of these cell types may be present at each timepoint; traditionally, at day 0, monocytes predominate; day 3, macrophages predominate; day 7, macrophages with some FBGCs; and at day 10 macrophages with an increasing number and size of FBGCs. The above is depicted in Figure 6.1.

Cytokine Cytokine Cytokine Production Production Production

Plating Adhesion Differentiation Fusion t=0 t~2 hrs t~3 days t~7&10 days Adherent Blood- Adherent Adherent Isolated Foreign Body Monocyte Macrophage Monocyte Giant Cell (MO) (Mφ) (FBGC) (MO) 1 nucleus 1-2 nuclei 1 nucleus ≥ 3 nuclei

Apoptotic Apoptotic Monocytes Macrophages

Detached Detached Monocytes Macrophages

Figure 6.1: Schematic of In Vitro Cell Culture System.

At each timepoint, fixed adherent cells are stained with May-Grünwald and

Giemsa for cell density analysis with optical microscopy or with Annexin-V or

- 170 - Chapter VI

TUNEL for apoptotic cell analysis using fluorescence microscopy. The

experimentally observed variables for the adherent cells are as follows: 1) the

total number of FBGC precursor cells (P) defined as the summation of the

adherent monocytes (C0), macrophages (C1), and macrophages containing two

nuclei (C2); 2) the total number of FBGCs (Cj, j=3,4,….N), where N is the maximum number of nuclei in one FBGC; 3) the total number of nuclei within

FBGCs (F_Nuclei), 4) the number of FBGCs of each size (C3 thru CN), and 5) the

total number of apoptotic adherent cells (A). The total number of detached cells

(D) is quantified using immunohistochemistry and flow cytometry. Exudates

containing cytokines and chemokines (S) secreted by monocytes, macrophages,

and FBGCs are collected and quantified for particular cytokines using enzyme-

linked immunosorbent assays (ELISAs). The cellular processes and

corresponding rate constants considered within the kinetic model, which are

indicated with ovals in Figure 6.2, are adhesion (η), differentiation (γ), monocyte

apoptosis (α0), macrophage apoptosis (α1), monocyte detachment (δ), macrophage detachment (δ), fusion (β), FBGC precursor cells cytokine

production (χP), and FBGC cytokine production (χF).

Quantitative results from cells cultured onto hydroxyl containing

alkanethiol self-assembling monolayer (SAM) surfaces were utilized for the

mathematical modeling project initially. In summary, these surfaces consist of a

glass substrate pre-coated with gold and modified using self-assembling

alkanethiols (SH(CH)2-OH) that present a homogenous surface of the hydroxyl

- 171 - Chapter VI

C0(0): Monocytes Non-adherent (1 nucleus) M(plated)

Adhesion: η

Apoptosis: α0 Detachment: δ P C0: Monocytes Adherent (1 nucleus)

Differentiation: γ A: Apoptotic D: Detached Cells Cells

Production: χP

C: Macrophages (1 nuclei) Apoptosis: α1 Detachment: δ

Fusion: β

C : Macrophages (2 nuclei)

S: Cytokine Concentration Fusion: β

F C3: FBGCs (3 nuclei)

Fusion: β Production: χF Legend:

CN: FBGCs State Variables (N nuclei) Processes & Rate Constants Observed Variables

Figure 6.2: System Diagram.

- 172 - Chapter VI

functional group. All of the observable variables of interest were quantified within

the hydroxyl SAM studies allowing for very accurate parameter estimation.

Additional studies containing polyethylene terephthalate (PET) modified

photografted polymers that present hydrophobic or hydrophilic surface

chemistries will be analyzed to elucidate the effects of surface chemistry on the

cellular processes and corresponding rate constants. Only select observable

variables (i.e. adherent monocyte, macrophage, FBGC populations) were

quantified with these experiments. Although this appears limiting, most

experiments conducted within our laboratory using the in vitro cell culture system

do not analyze all of the observed variables (i.e. apoptotic cells and detached cell populations). Developing the model with respect to these limited observed

variables expands the experimental pool to which the model can be applied.

Model Development

A dynamic systems model (Figure 6.2) was designed to analyze cell

interaction with biomaterial surfaces in a cell culture. Initially, only monocytes

adhere to the surface with an initial number density C0 )0( = Mplated *η

determined experimentally, where Mplated is the number of monocytes plated.

The number density of adherent monocytes 0 tC )( changes with time as:

dC 0 []++−= δαγ C (1) dt 000

where the rates coefficients γ ,α0 ,δ0 relate respectively to cell differentiation,

apoptosis, and detachment. The number density of adherent macrophages with

- 173 - Chapter VI

one nucleus 1 tC )( increases from differentiated monocytes (γ ) and decreases by

apoptosis (α1) and detachment (δ1 ). In addition, an adherent macrophage with 1

or 2 nuclei CC 21 ),( can fuse with each other and with a FBGC containing 3 or

more nuclei j = njtC ),....4,3()( . For simplicity, we assume that these processes

have the same rate coefficient(β) :

n dC1 C []−+−= βδαγ 11110 ∑CCC j (2) dt j=1

The number density of apoptotic cells A(t) on the surface increases as:

dA += αα CC (3) dt 1100

and the concentration of detached cells Cδ (t) in solution increases as:

dD += δδ CC (4) dt 1100

Macrophages containing two nuclei and FBGCs are not included in eqs. (3) and

(4) because it is assumed that fused cells do not undergo apoptosis or

detachment within the typical timeframe of a monocyte/macrophage in vitro cell

culture systems (10-14 days). For a macrophage with 2 nuclei, we assume that

fusion is the only significant process. Therefore, the number density of this

macrophage 2 tC )( changes as:

n dC2 ⎡ ⎤ β ⎢ −= 211 ∑CCCC j ⎥ (5) dt ⎣ j=1 ⎦

A FBGC is formed by the fusion of two cells, macrophage or FBGC, whose total

nuclei is greater that 2. Therefore, the FBGC number density k = njtC ),....4,3()(

- 174 - Chapter VI

increases by fusing cells of with fewer nuclei and decreases by fusing with cells

of any number of nuclei:

p n dCk ⎡ ⎤ = β ⎢∑ wwk +− 1 − ∑CCCC jk ⎥ (6) dt ⎣ w=1 j=1 ⎦

where

⎧k − 1 ⎫ for k = 3,5,7... ⎪ 2 ⎪ p = ⎨ ⎬ k ⎪ for k = 4,6,8... ⎪ ⎩⎪ 2 ⎭⎪

The concentration of cytokines tS )( in solution increases as:

dS n χ P ( CCC 210 ) +++= χ ∑CkF (7) dt k=3

where (χP, χF) are rate coefficients for monocytes/macrophages and FBGC,

respectively.

Model variables related to experimental measurements

Model outputs are computed for comparison with observed variables

obtained from in vitro experiments. Adherent monocytes and macrophages, with

no more than 2 nuclei are enumerated together as a single group of FBGC

precursors:

= + + CCCP 210 (8)

The number density of all FBGC is

n = ∑CF k (9) k=3

The total number of adherent cells is

- 175 - Chapter VI

T = + FPC (10)

In the precursor group, the number density of nuclei is

P = + + 2CCCN 210 (11)

and the number density of FBGC nuclei is

n F = ∑ kCN k (12) k =3

The number density of nuclei of all adherent cells is

= + NNN FPT (13)

Numerous percentages are commonly reported and determined using the

experimental data collected. Percent adhesion is defined as the percentage of

the initially adherent cells remaining at a given timepoint. Percent fusion is

N %Fusion = F (14) NT

Percent apoptosis is

A %Apoptosis = (14) NT

Percent detachment is

D %Detachment = (14) NT

With an initial working model, materials of interest can be tested within the model. Considering that material surface chemistry (hydrophilicity,

hydrophobicity, and charge) dictate cell adhesion, differentiation, and fusion, it

- 176 - Chapter VI

can be assumed that these chemistries ultimately change the adhesion,

differentiation, apoptosis, and fusion rate constants.

Model Simulation

The stiff, ordinary differential equations shown above were solved using

Matlab’s ode15s function. Initial rate constants were selected to correlate with

the experimental trends and data for each cell population. Experimental data

was collected at day 0, day 0 at 2 hours, day 3, day 7, and day 10 and was

utilized to determine the percent adhesion, percent fusion, percent detachment

and percent apoptosis for parameter estimation. As stated earlier, cells were

plated at t=0 and allowed to adhere for 2 hours. The rate of cellular adhesion (η) was determined based on the percentage of cells initially plated at t=0 that adhered to the biomaterial surface. The initial adherent cell population (C1(0)) studied within the model at t=0 correlates to the experimental adherent cell population at t=2 hours. The rate constants directing the adherent, detached, and apoptotic populations were determined based upon general experimental trends.

Knowledge of the experimental cell culture system allows for specific trends or limitation of the outputs to be expected. The monocyte population experimentally decreases within three days mainly due to differentiation (γ) into macrophages and in part due to cell detachment (δ) and apoptosis (α). As a

result of this differentiation, the macrophage population increases rapidly by day

3 and slowly decreases as macrophages either fuse (β), detach (δ), or apoptose

- 177 - Chapter VI

(α). FBGCs are usually not seen experimentally until day 7 and increase with time. In some fusion promoting cultures, FBGCs are found as early as 3 days, so flexibility within the model is allowed. In contrast to other cell types, monocytes, macrophages, and FBGCs do not undergo proliferation; therefore, the total number of adherent cells or nuclei should never increase above the initially plated number of cells. Finally, the total number of adherent cells tends to decline significantly over the ten day period.

In order to relate the experimental data to the computational model results, the percentages of total initial adherent cells (Co(0)) that remain adherent, fuse, become detached or undergo apoptosis were determined from the experimental data. Within the model simulations, 100 cells were plated with an adhesion rate constant (η) of 1 resulting in Co(0)=100. Results for the subsequent cellular populations were therefore percentages of the original adherent cell population.

Initial Model Results

The following mathematical model results were computed using the estimated rate constants shown in Table 6.1. These parameters were selected to produce values comparable to quantitative empirical data derived from cell cultures containing the –OH SAM surface. Experimental values are indicated with point markers in Figure 6.3. The percentage of the original adherent cell population (C0(0)) that remained adherent, underwent fusion, detached, and became apoptotic were analyzed and compared for the computed and

- 178 - Chapter VI

empirically-derived data sets. Percent adhesion decreased over time to 71% at

day 3, 49% at day 7, and 39% at day 10. These values correlated to the

experimental percent adhesion values of 74%, 48.5%, and 37.8% respectively.

Table 6.1 Estimated Rate Constants Utilized in Model Simulations Symbol Rate Constant Value η: Monocyte Adhesion 1 γ  Monocyte Differentiation 0.8 α0  Monocyte Apoptosis 0.005 α1 Macrophage Apoptosis 0.002 δ0 Monocyte Detachment 0.05 δ1 Macrophage Detachment 0.16 β Fusion 0.0008

The estimated percent fusion increased beginning at 1% at day 3 to values of

5.4% at day 7 and 9.7% at day 10. Experimentally, percent fusion increased from 3% at day 3 to 10% at day 7 and 9% at day 10. Percent detachment increased within the computation model from 26% to 56% to 69% at days 3, 7 and 10 correlating almost exactly to the experimental results of 26%, 47% and

69%. Finally, the computed percent apoptosis increased from 7% at day 3 to

8.6% and 9% at days 7 and 10 correlating to the experimental data of 1%, 4% and 15%.

In addition to estimating the experimental data, these results correlate to expected trends regarding cellular adhesion, detachment, and apoptosis. Based upon the assumptions in this system, a cell can remain adherent, remain adherent and undergo apoptosis, or become detached suggesting that the percentage of adherent and detached cells sum to the total initial adherent cell population (100%). This was found to be true for most timepoints.

- 179 - Chapter VI

Experimentally, these values oftentimes do not sum to 100% partly due to the removal of additional adherent cells during the washing and staining procedures and partly due to some of the non-adherent cells remaining in culture during media collection. This experimental discrepancy will far surpass the less than

10% difference computed within the model.

Figure 6.3: Comparison of Experimental and Computed Percentages.

Experimentally, monocytes and macrophages are normally not distinguished when collecting the adherent cell density data; therefore the expected trends for monocyte and macrophage population discussed under the

Model Simulation section are not experimentally analyzed. Model simulation provides a means by which the populations of monocytes, macrophages with 1 nucleus, and macrophages with 2 nuclei change over time can be analyzed.

- 180 - Chapter VI

These results are shown in Figure 6.4. As expected, the monocyte population

decreased significantly within the first three days. Within the same time period,

the macrophage with 1 nucleus population increased corresponding to monocyte

differentiation into macrophages as confirmed with the percent detachment

results discussed above. Subsequently, the macrophage with 1 nucleus

population decreased further as macrophage fusion occurred resulting in an

increase in the populations of macrophages containing two nuclei and FBGCs.

At each timepoint, the computed values for the monocyte, macrophage, and

FBGC populations summed to the total adherent cellular population as expected.

Figure 6.4: Additional Analysis of Adherent Cell Populations over Time.

- 181 - Chapter VI

Effect of Parameters and Parameter Sensitivity

All of the above results were computed using the rate constants listed in

Table 6.1 and provide the initial values for parameter estimation. Modification of

these rate constants change the resulting data as listed in Table 6.2. Sensitivity

ranges were determined by systematically modifying these parameters while

fixing other parameters (Table 6.3). Increasing or decreasing the values beyond

these ranges causes significant error in the computed data. It was found that the

maximum number of nuclei in a FBGC (N) proved to have a minimal effect on the

resulting data. Therefore, N will be fixed at 20 for future simulations thus

minimizing the number of parameters to estimate to seven and increasing the

accuracy of the remaining parameter estimates. The sensitivity ranges shown in

Table 6.3 will be utilized as upper and lower bounds in future parameter

estimation.

Table 6.2 Effects of Material-Dependent Parameters on Cellular Behavior Cellular FBGC FBGC Parameters Adhesion Number Size Apoptosis η: Adhesion ↓ ↓ ↓ ↔ γ: Differentiation ↓ ↓ ↓ ↔ α0: Monocyte Apoptosis ↑ ↑ ↑ ↓ α1: Macrophage Apoptosis ↑ ↑ ↑ ↓ δ0: Monocyte Detachment ↑ ↑ ↑ ↔ Parameter Decreasing δ1: Macrophage Detachment ↑ ↑ ↑ ↔ β: Fusion ↔ ↓ ↓ ↔ η: Adhesion ↑ ↑ ↑ ↔ γ: Differentiation ↑ ↑ ↑ ↔ α0: Monocyte Apoptosis ↓ ↓ ↓ ↑ α1: Macrophage Apoptosis ↓ ↓ ↓ ↑ δ0: Monocyte Detachment ↓ ↓ ↓ ↔ Parameter Increasing δ1: Macrophage Detachment ↓ ↓ ↓ ↔ β: Fusion ↔ ↑ ↑ ↔ Symbol representations: ↑ enhance, ↓diminish, ↔ no direct effect on indicated behavior.

- 182 - Chapter VI

Table 6.3 Sensitivity Ranges for Each Parameter Symbol Rate Constant Sensitivity Range η: Monocyte Adhesion -* γ  Monocyte Differentiation 0.05-0.9 α0  Monocyte Apoptosis 0.00005-0.5 α1 Macrophage Apoptosis 0.00005-0.05 δ0 Monocyte Detachment 0.00005-0.5 δ1 Macrophage Detachment 0.1-0.5 β Fusion 0.000008-0.0009 * Monocyte adhesion is fixed at 1 for these simulations.

Future Work

Formal Parameter Estimation

Formal parameter estimation using the lsqcurvefit in Matlab needs to be continued to determine the optimal values of these rate constant parameters that produce the minimal error between the computed and experimental results. The above sensitivity ranges provided lower and upper bounds for each parameter.

The above model simulations did not estimate true values for monocyte adhesion

(η) because all the cells that were plated (100) were assumed to adhere using

η=1. Analysis of experimental data on multiple surfaces (C0 )0( = Mplated *η ) showed that the monocyte adhesion (η) values ranged from 0.4 to 0.7 providing the sensitivity range for η that will be utilized parameter estimation. During parameter estimation, the parameters or rate constants will be ranked according to which values have the most and least effect on the resulting data and in turn which processes most affect these cellular populations. Parameter values that do not significantly affect the computed data will be fixed further minimizing the parameters to estimate and increasing the accuracy of the remaining estimated parameters.

- 183 - Chapter VI

Analysis of Material Effects on Cellular Behaviors

The next phase of model simulation will be conducted comparing data

collected from two biomaterials that produce distinct cellular behaviors. The

hydrophobic photografted polymer surface supports adhesion and fusion, while

the hydrophilic/neutral photograft polymer surface significantly inhibits both adhesion and fusion. Experimental data for these studies did not collect day 0

adherent cell density results; therefore, modifications to the model will need to be

made in order to simulate the experimentally-derived day 3, 7, and 10 data. As

stated above, formal parameter estimation needs to be conducted utilizing the

monocyte adhesion range described. Most likely, the parameter estimation

process will determine that the monocyte adhesion parameter has a limited sensitivity range based upon the ability of comparable adherent cell densities seen at day 0 on other surfaces; therefore a specific rate constant (η) value can

be utilized thus simplifying the model and eliminating the need for day 0

experimental data. Upon completion of formal parameter estimation for the two

distinct biomaterials, the estimated parameters will be compared and analyzed to

determine which parameters and therefore processes are greatly or minimally

affect by biomaterial surface chemistry. Additional analysis will be conducted to

determine which parameters/processes have the most and least influence on

determining the final adherent cell densities. Ultimately, the mathematical model

will be further developed and utilized to analyze experimental data collected from

cell cultures on a variety of surface chemistries thus creating a database of rate

- 184 - Chapter VI

constant parameters that could provide insight into the relationships of

biomaterial surface chemistry and adherent cellular behaviors and potentially

allow for predicts of cellular/biomaterial interactions on future materials.

References

1. Anderson JM. Inflammatory reaction: The nemesis of implants. In: Zilla P,

Greisler HP, Landes RG, editors. Tissue Engineering of Vascular

Prosthetic Grafts; 1999. p 197-206.

2. Anderson JM. Inflamatory response to implants. ASAIO 1988;11:101-107.

3. Anderson JM. Biological responses to materials. Annual Reviews Material

Research 2001;31:81-110.

4. Anderson JM. Multinucleated giant cells. Current Opinion in Hematology

2000;7:40-47.

5. De Waal Malefyt R, Figdor CG, Huijbens S, Mohan-Peterson S, Bennett

B, Culpepper J, Dang W, Zurawski G, de Vries JE. Effects of IL-13 on

phenotype, cytokine production, and cytotoxic function of human

monocytes. Journal of Immunology 1993;151:6370-6381.

6. Gordon S. Alternative activation of macrophages. Nature Reviews:

Immunology 2003;3(1):23-35.

7. Gosain A, Gamelli R. A primer in cytokines. Journal of Burn Care

Rehabilitation 2005;26(7):7-12.

8. Rollins B. Chemokines. Blood 1997;90(3):909-928.

9. Meznarich N, Kyriakides T, Donaldson E, Foster M, Schrom B, Ratner B,

Hauch K, Bornstein P. Matrix-metalloproteinase (MMP-9) and its role in

- 185 - Chapter VI

wound healing and the foreign body response. Journal of Undergraduate

Research in Bioengineering 2004;4(2):84-89.

10. Nagase H, Visse R, Murphy G. Structure and function of matrix

metalloproteinases and TIMPs. Cardiovascular Research 2006;69:562-

573.

11. Sternlicht M, Werb Z. How matrix metalloproteinases regulate cell

behavior. Annual Review of Cell Developmental Biology 2001;17:463-516.

12. Brodbeck W, Shive M, Colton E, Nakayama Y, Matsuda T, Anderson JM.

Influence of biomaterial surface chemistry on the apoptosis of adherent

cells. Journal Biomedical Materials Research 2001;55:661-668.

13. Brodbeck W, Nakayama Y, Matsuda T, Colton E, Ziats NP, Anderson JM.

Biomaterial surface chemistry dictates adherent monocyte/macrophage

cytokine expression in vitro. Cytokine 2002;18(6):311-319.

14. Owen MR, Sherratt JA. Pattern formation and spatiotemporal irregularity

in a model for macrophage-tumor interactions. Journal of Theoretical

Biology 1997;189:63-80.

15. Turner S. Using cell potential energy to model the dynamics of adhesive

biological cells. Physical Review 2005;71:041903.

16. Levine HA, Sleeman BD, Nilsen-Hamilton M. A mathematical model for

the roles of pericytes and macrophage sin the initiation of angiogenesis. I.

The role of protease inhibitors in preventing angiogenesis. Mathematical

Biosciences 2000;168:77-115.

- 186 - Chapter VI

17. Kao WJ, Zhao QH, Hiltner A, Anderson JM. Theoretical analysis of in vivo

macrophage adhesion and foreign body giant cell formation on

polydimethylsiloxane, low density polyethylene, and polyetherurethanes.

Journal Biomedical Materials Research 1994;28(1):73-9.

18. Qazi S, Beltukov A, Trimmer BA. Simulation modeling of ligand receptor

interactions at non-equilibrium conditions: processing of noisy inputs by

ionotropic receptors. Mathematical Biosciences 2004;187:93-110.

19. Zhao QH, Anderson JM, Hiltner A, Lodoen GA, Payet CR. Theoretical

analysis on cell size distribution and kinetics of foreign body giant cell

formation in vivo on polyurethane elastomers. Journal Biomedical

Materials Research 1992;26(8):1019-1038.

20. Coombs D, Dembo M, Wofsy C, Goldstein B. Equilibrium thermodynamics

of cell-cell adhesion mediated by multiple ligand-receptor pairs.

Biophysical Journal 2004;86:1408-1423.

21. Jones J, Chang D, Colton E, Kwon I, Matsuda T, Anderson J. Proteomic

analysis and quantification of cytokines and chemokines from biomaterial

surface-adherent macrophages and foreign body giant cells. Journal of

Biomedical Materials Research 2006;Submitted.

22. Jones J, Dadsetan M, Collier T, Ebert M, Stokes K, Ward R, Hiltner A,

Anderson JM. Macrophage behavior on surface modified biomaterials.

Journal of Biomaterial Science Polymer Edition 2004;15(5):567-564.

- 187 - Chapter VI

23. Huppertz B, Frank HG, Kaufmann P. The apoptosis cascade -

morphological and immunohistochemical methods for its visualization.

Anatomy and Embryology 1999;200:1-18.

24. Brodbeck W, Patel J, Voskerician G, Christenson E, Shive M, Nakayama

Y, Matsuda T, Ziats N, Anderson JM. Biomaterial adherent macrophage

apoptosis is increased by hydrophilic and anionic substrates in vivo.

Proceedings of the National Academy of Science 2002;99(16):10287-

10292.

- 188 - Chapter VII

Chapter VII: Conclusions

Summary of Findings and their Implications

This work set out to elucidate the relationships between surface chemistry and adherent monocyte, macrophage, and foreign body giant cell (FBGC) behaviors for the goal of ultimately directing the foreign body reaction, inflammation, and wound healing responses to implanted biomedical devices. To this end, three model material systems were utilized to analyze various adherent cellular behaviors. The general material properties of hydrophobicity, hydrophilicity, and ionic chemistry were investigated in part due to the limited number of studies directly comparing these properties and in part due to the broad range of materials that can be implicated with this study’s findings.

Macrophage adhesion, fusion into FBGCs, and apoptosis were investigated with respect to material surface chemistry with the aim of potentially being able to control the adherent cell population. In addition, a mechanistic mathematical model of our in vitro human monocyte cell culture system was developed based on experimentally-derived quantitative metrics.

Using photografted polyethylene terephthalate (PET) based polymers displaying hydrophobic, hydrophilic/neutral, and hydrophilic/ionic surface chemistries, it has been demonstrated that material surface chemistry influences macrophage adhesion and fusion ultimately modulating the cytokines/chemokines/MMPs/TIMPs released from biomaterial-adherent macrophages/FBGCs (Chapters II, III, and IV). Hydrophilic/neutral surfaces significantly inhibited adhesion and fusion in comparison to the hydrophobic and

- 189 - Chapter VII

hydrophilic/ionic surfaces investigated. Pro-inflammatory cytokines

concentrations (IL-1β and IL-6) and chemokines (IL-8 and MIP-1β) decreased

with time, while the anti-inflammatory cytokine, IL-10, remained relatively

constant. This potentially correlates with the resolution of inflammation that

occurs naturally with time or it could implicate IL-10 stimulation of the adherent

macrophages, which is known to down-regulate pro-inflammatory cytokines and chemokines, prompting additional analysis into IL-10 stimulation on biomaterial- adherent macrophages. This change in cytokine/chemokine concentrations over time along with qualitative analysis of over 70 cytokines, chemokines, MMPs, and TIMPs using antibody arrays revealed a phenotypic switch in adherent macrophages from a cytokine/chemokine profile that is similar to classically activated macrophage at earlier timepoints to an alternatively activated state at later timepoints.

The pivotal finding in this research provided evidence contradicting previous dogma that cellular activation proportionally correlates with adhesion.

Instead, the cells adherent to the cell-limiting hydrophilic/neutral surfaces produced greater quantities of each protein analyzed in comparison to the adhesion supporting surfaces indicating an increased activation state in these cells. This prompts further investigation whose results could potentially be exploited in the design of future biomaterial applications.

The ability to minimize the adherent cell densities by inhibiting adhesion and fusion or promoting apoptosis can also minimize macrophage/FBGC interactions with implanted biomaterials, thus reducing their detrimental effects

- 190 - Chapter VII

(i.e biodegradation resulting from the release of reactive oxygen species, acid, and degradative enzymes by adherent macrophages and FBGCs). To this end,

numerous companies and researchers have been developing surface

modification techniques using surface chemistries and polymers known to direct cellular adhesion. An example of a surface modification of clinically-utilize polyurethanes designed by Polymer Technology Group is presented in Chapter

V. The incorporation of hydrophobic polydimethylsiloxane (PDMS) (a polymer more resistant to biodegradation) into the polyurethane backbone and via covalently attached surface modify-endgroups (SMEs) initially produced greater amounts of apoptosis in vivo and in in vitro cell cultures and yet ultimately promoted FBGC formation at later timepoints in vivo and in the presence of interleukin -4 (IL-4), a fusion-inducing cytokine, in vitro. In contrast, the fluorocarbon or polyethylene oxide (PEO) SME modifications which were expected to minimize adhesion produced no effect. In the case of the fluorocarbon SME surfaces, this was due to the inherent chain flexibility of the polyurethane and SME allowing for hysteresis of the hydrophobic fluorocarbons from the aqueous interface into the bulk of the polymer, while in the second case the length of the PEO SME chains were insufficient to elicit an effect. These findings have a tri-fold set of implications. Related to the hypothesis of this research, these results confirm the findings in Chapter II that hydrophobic surface chemistry promotes macrophage fusion. Secondly, it supports the ongoing hypothesis of our laboratory that macrophages undergo fusion as a means by which to escape apoptosis. Finally, it suggests criteria for the

- 191 - Chapter VII

improvement of these modifications in polyurethanes: 1) hydrophobic

modifications need to be incorporated throughout the polymer chemistry similar

to the case of the PDMS modification or needs to be sequestered to the material

surface in order to be effective and 2) the hydrophilic SME units need to be

sufficiently long enough to inhibit macrophage adhesion and subsequent fusion.

This technology provides an effective technique by which to alter the material

surface chemistry of polyurethanes.

It was the intention of this research to determine how specific commonly

utilized functional groups (–CH3, –COOH, and –OH) direct adherent cellular

behaviors using an alkanethiol self-assembled monolayer (SAM) material

system. Unfortunately, in the process of this investigation it was determined that

the SAM surfaces were unstable because the alkanethiols began to desorb after

synthesis, a phenomenon that was expedited by immersion in an aqueous, 37oC

environment similar to culture condition. Under these conditions, changes in the

material surface chemistry were seen with contact angle analysis and confirmed using AFM as early as two hours after incubation (shown in the Appendix). This

research sheds light upon the downplayed instabilities of SAMs and adds doubt

to the use of novel surfaces as a viable system for biological analysis.

In summary, this research demonstrates that hydrophobic and hydrophilic

surface chemistries can be utilized to direct macrophage and FBGC behavior.

Previous research has classified hydrophobic and hydrophilic materials as being

conditionally (non)adhesive hydrophilic surfaces exemplified by highly hydrated

surfaces (i.e. PEO and carbohydrates), adhesive hydrophilic surfaces (TCPS),

- 192 - Chapter VII

adhesive surfaces that demonstrate intermediate wetting (polyurethanes and

PET), and nonadhesive hydrophobic surfaces (PTFE, PE, and PP), see Figure

1.2. Modification of a hydrophobic PET mylar substrate with conditionally

(non)adhesive hydrophilic polyacrylamide derivative produced a

hydrophilic/neutral surface (PAAm) that significantly inhibited adhesion in

comparison to cell adhesion supporting hydrophobic PET. The incorporation

hydrophobic PDMS into polyurethanes did not significantly affect cellular

adhesion, however it did inhibit apoptosis and promote fusion over time in

comparison to the less hydrophobic polyurethane. Although classification of

biomaterials as hydrophobic or hydrophilic has a relative nature, research such

as this can be useful in establishing relationships between these variables

providing a means by which to modify future surfaces thus directing the next

stage of research and biomaterial development.

Future Work

Considering the countless known and unknown intricacies that are

involved in the cellular responses to a biomaterial surface, oftentimes

investigations will focus on how one, key modifiable variable affects the cellular

behavior outcomes. This is useful in that it provides a distinct input-output

relationship for biomaterial design. Also, it elucidates mechanisms for future

avenues of analysis. Future work will need to be conducted along two fronts.

One front will focus on gaining a further understanding of the mechanism(s) and natures of the above findings. The second will need to exploit multiple

- 193 - Chapter VII

biomaterial surface chemistries in order to ascertain which surface chemistries

and clinically available surfaces can be utilized in the design of future biomaterial

applications.

Little is known of the effects of biomaterial surface chemistry on

macrophage activation or whether the mere interaction of a macrophage with a

biomaterial is capable of activating the cell in a specific manner. This research

provides a starting point by which to determine the material-dependent markers

involved in the phenotypic switch (classically-activated to alternatively-activated)

of adherent inflammatory cells. Investigations of cytokines and chemokines

specifically known to be associated with alternatively- or classically-activated macrophages will accomplish this goal. In addition, investigation of cell surface receptors (i.e. cytokine/chemokine receptors and the mannose receptor, which is required for macrophage fusion and are modulated by material surface chemistry) will aid in relating these phenotypic profiles to subsequent FBGC formation and material surface chemistry. Multiple hydrophilic/neutral surfaces will need to be utilized in future studies in order to understand the disparate response of low adherent cell surface densities with high cytokine/chemokine production levels. Analysis of integrin expression and subsequent signaling may also shed light on the relationships of material surface chemistry, cellular activation, and cytokine/chemokine/MMP/TIMP production.

Finally, continued development of the mechanistic mathematical model of the in vitro human monocyte culture system will further elucidate the relationships between the cellular behaviors and biomaterial surface chemistries. Full

- 194 - Chapter VII parameter estimation of the rate constants for monocyte adhesion, monocyte differentiation into macrophages, monocyte/macrophage detachment and apoptosis, and macrophage fusion will determine the impact each of these factors have on the overall adherent cell population and function. In addition, analysis of estimated parameters based on experimental-derived quantitative metrics will shed light upon which processes are most affected by material surface chemistry. Ultimately, this model could be developed to permits biologically-based predictions of reduced or enhanced inflammatory cell/material surface interactions and, ultimately, the engineering of appropriate surfaces for specific biomedical material applications.

- 195 - Appendix

Appendix: Self-Assembled Monolayers as a Model Material System for Material-Dependant Analysis of Macrophage/FBGC Cellular Behaviors

Abstract

Novel self-assembled monolayers (SAMs) designed to present

homogenous surface chemistries were utilized to further investigate the material

surface chemistry dependent macrophage and foreign body giant cell (FBGC) behaviors including macrophage adhesion, fusion, apoptosis, and protein

production. Contact angle analysis revealed instabilities in the –CH3, –COOH,

and –OH terminated SAM surfaces upon incubation in serum-free media at 37oC or under dry, room temperature conditions. Further analysis indicated that the –

CH3 terminated SAM surface degraded rapidly within 2 hours and loss sufficient

SAM units to be comparable to the Au control surface within 24 hours of incubation in serum-free media at 37oC. By day 5, the contact angles for the –

COOH and –OH terminated SAM surfaces increased markedly. Cellular

adhesion decreased more rapidly on the Au control and –CH3 terminated SAM

surfaces in comparison to the other surfaces. However by day 10, cellular

adhesion, fusion, and apoptosis were comparable on all surfaces. Production of

various cytokines, chemokines, matrix metalloproteinase, and tissue inhibitors of

matrix metalloproteinases were comparable between materials at each timepoint

and condition. This research prompts in depth analysis into the stability of SAM

for use in biological applications.

- 196 - Appendix

Introduction

Self-assembled monolayer (SAM) surfaces are a unique type of surface

that can present a variety of distinct surface chemistries offering a novel means

for studying cellular/biomaterial interactions. SAMs consist of long-chained

alkanethiols that spontaneously adsorb onto a gold surface from a solution and

self-assemble to form a homogeneous monolayer. These surfaces are very

advantageous in that they can be quickly synthesized and easily modified to

present a variety of homogenous, mixed, or pattern surface chemistries. It is well

understood that inflammation, wound healing, and the foreign body reaction

occur following implantation of a biomaterial. Significant research has been

conducted to gain insight into the cellular/biomaterial interactions involved in and

directing these responses. Theoretically, an in-depth understanding of the

relationships between surface chemistry and the behaviors of key inflammatory

and wound healing cells (i.e. monocytes, macrophages, foreign body giant cells,

and fibroblasts) would provide a basis for establishing criteria for designing

biomaterials for use in specific applications. This study utilizes homogeneous

SAM surfaces to investigate the effects of particular surface chemistries

(hydrophobic and hydrophilic) and functional groups (CH3, COOH, and OH) on

the behavior of macrophages and foreign body giant cells (FBGCs), two key cells

involved in inflammation, wound healing, and the foreign body reaction.

Self-assembling alkanethiols have the chemistry of HS-(CH2)n-R, where

n=10-18 and R is a terminal group and self-assemble into a well-defined aligned structure shown in Figure A.1. SAM surfaces are an advantageous model

- 197 - Appendix

material system in that the composition and properties of the surface are

controlled by the synthesis parameters, distinct alkanethiol units and terminal groups (CH3, COOH, OH, NH2), post-synthesis modifications adding specific ligands or additional chemistries, and micropatterning techniques that pattern the functional groups within the monolayer.1-6 Using these techniques, tailored SAM

surfaces have been created to study the effects of specific surface chemistries

on protein adsorption and numerous cellular behaviors including cellular

adhesion, recruitment, orientation, migration, and cytokine release.2,7-17 In

addition, these surfaces are being utilized in other applications to examine ligand

interactions, improve the reproduction of microarrays, and alter nanostructures or

the surfaces of thin metal films.1,6,18,19 Significant research has been conducted

with these surfaces since their inception in the 1980s; however more recently,

the stability of these surfaces has been questioned.6,20-29

α X X X X X X X X X

S S S S S S S S S

Figure A.1: Self-Assembled Monolayer Structure. X= Terminal functional o group: CH3, COOH, or OH. α= molecular tilt, typically 30 .

- 198 - Appendix

Recent studies in our laboratory have shown that hydrophobic,

hydrophilic/neutral, and hydrophilic/ionic surface chemistries affect macrophage

adhesion, fusion, apoptosis, and activation in opposing manners.30-36

Particularly, hydrophobic and hydrophilic/ionic surfaces promote macrophage

adhesion and fusion and inhibit macrophage apoptosis, while hydrophilic/neutral

surfaces inhibit macrophage adhesion and fusion and promote macrophage

apoptosis and activation.30,35,36 These findings prompted a more intensive

investigation into the effects of particular surface chemistries and functional groups on macrophage behavior.

Barbosa et.al investigated the acute inflammatory cell interactions with select SAM surfaces. This work showed that hydrophobic CH3-terminated SAMs

supported significantly greater amounts of adherent PMNs and mononuclear

cells in comparison to the hydrophilic OH- and COOH- terminated SAM surfaces

37,38 in vitro. In contrast, in vivo, the CH3-terminated SAMs were shown to support

low levels of leukocyte adhesion in comparison to the gold substrate and the OH-

and COOH-terminated SAMs and a greater number of inflammatory cells within

the surrounding subcutaneous pouch. Particularly, the CH3-terminated SAMs

attracted inflammatory leukocyte population predominately of Mac-1+ phagocytic

cells.39 It was suggested that hydrophobicity modulates both local inflammatory

cell response and leukocyte adhesion.40 These contrasting results allow room for further investigation.

In this study, we employ these unique SAM surfaces to evaluate the effects of the commonly utilized functional groups (CH3, COOH, OH) and

- 199 - Appendix hydrophobic/hydrophilic surface chemistries on macrophage adhesion, fusion, detachment, apoptosis, and cytokine production. As the study progressed, the

SAM material stability became uncertain; therefore, the focus of this research was revised to include an evaluation of these surfaces as a stable material for use in our 10 day, in vitro cell culture system.

Materials and Methods

Self-Assembled Monolayer Synthesis

Self-assembled monolayers (SAMs) are long chain, alkanethiols

(HS(CH2)mX, where m ≥ 10, and X = a functional group) that spontaneously absorb from solution onto pre-coated gold substrates to form homogeneous, well-ordered monolayers. Glass coverslips (no. 2, d=18mm, Scientific,

Pittsburgh, PA) were primed with a 100Å thick layer of titanium and pre-coated with a subsequent layer of gold (1000Å thick) using a sputter coating technique

(2M/torr deposition rate, 3x10-6 torr power). The titanium primer was utilized to improve the adhesion of the gold coating. The pre-coated glass substrates were cleaned to remove any debris or contaminants using a series of washes with acetone, absolute ethanol, and deionized, distilled water. Clean substrates were immersed in a 1.0 mM ethanolic, alkanethiol solution under a nitrogen environment for 18-20 hours at room temperature to allow the monolayers to spontaneously form and assemble. Subsequently, the alkanethiol surfaces were rinsed in absolute ethanol (3x) to remove any non-adsorbed alkanethiol molecules. Surfaces were then dried in a nitrogen stream and equilibrated in

- 200 - Appendix

Dulbecco’s Phosphate Buffer Saline with Mg+2 and Ca+2 (PBS++) for 15 minutes

at room temperature. SAM surfaces were placed into culture within 2 hours of

synthesis. Any remaining surfaces were stored under dry, dark conditions.

The SAM surfaces investigated within this study have a chemistry of

HS(CH2)11X with either a –CH3 (hydrophobic), –COOH (hydrophilic), or –OH

(hydrophilic) functional group. Unmodified gold surfaces were utilized as a

control within these studies.

Studies were conducted using SAM surfaces synthesized in house,

denoted with “in-house”, or in the laboratory of Dr. Takehisa Matsuda (Kyushu

University, Fukuoka, Japan) denoted with “out-of-house”. The synthesis

procedure for out-of-house SAM surfaces was similar to that for those produced in-house, except a chromium primer layer (200nm thick) deposited using electron

beam evaporation was utilized instead of the titanium primer. The experiments

that were conducted using the out-of-house SAM surfaces were carried out

months after synthesis and receipt of materials, while the in-house SAM surfaces

were made fresh for each experiment and used within 2 hours of synthesis.

Surface Characterization

SAM surfaces were characterized for hydrophobicity and hydrophilicity

using contact angle analysis via the sessile drop method and a goniometer

(Edmund Scientific, Barrington, NJ) at 22oC room temperature. Contact angles

were measured immediately upon synthesis at day 0 and subsequently on

untreated samples at day 40 to determine modification long-term stability. In

- 201 - Appendix addition, select samples were treated via incubation in serum-free media (SFM)

(Gibco, Grand Island, NY) at 37oC to determine modification stability under aqueous culture conditions. Treated samples were measured for contact angle after 2 hours, 6 hours, 24 hours, 5 days, 10 days, and 18 days of treatment.

Contact angles were taken on two areas of each of the surfaces to indicate surface homogeneity. Both advancing and receding water contact angles were measured and are reported as the mean ± standard deviation to further detect surface inhomogeneities.

Attenuated force microscopy was conducted using a Nanoscope IIIa

(Digital Instruments, Santa Barbara, CA). Measurements were taken using a silicon nitride cantilever tip (spring constant = 0.58 N/m) in air. Dry, freshly synthesized surfaces were imaged at t=0 and after 2hours of treatment in serum- free media at 37oC to determine changes in the topography of the surfaces within

2 hours in an aqueous, 37oC environment. A total of 3 samples were analyzed at

2 areas on each sample. AFM 3-D images and roughness measurements are presented.

Macrophage/FBGC In Vitro Cell Culture System

Within two hours of synthesis, equilibrated SAM surfaces were sterilized with 100% ethanol, placed into 12-well tissue culture polystyrene (TCPS) plates

(Fisher Scientific, Pittsburgh, PA) and secured with silicone rings (ID=1.59cm,

OD=2.22cm, h≈8mm). Silicone rings were cut from silicone tubing (Cole-Parmer,

Vernon Hills, IL), sonicated for 5 minutes in 100% ethanol, rinsed with distilled

- 202 - Appendix

water, air-dried, and sterilized using ethylene oxide. The resultant surface area

was 198mm2.

Human blood monocytes and serum were isolated using a Ficoll and

density gradient centrifugation technique as described previously.41 Freshly

isolated monocytes were plated at a concentration of 7.5 x 105 cells in 0.5 mL of

serum-free media (SFM) with 20% autologous serum and were allowed to

adhere for 2 hours. After 2 hours, supernatants containing non-adherent cells

were collected via pipetting for non-adherent cell analysis, adherent cells were

rinsed with warmed PBS++ and refed fresh media containing SFM with 20% heat-

treated (56oC for 1 hour) autologous serum. Cell cultures were incubated at

o 37 C in a 5% CO2 environment for 3, 7 and 10 days. At each timepoint, supernatants containing non-adherent cells were collected via pipetting for non- adherent cell analysis, select cultures of adherent cells were terminated for

adherent cell analysis, and remaining adherent cell cultures were refed fresh

media. At days 3 and 7, IL-4, a fusion inducing cytokine, was added to the fresh

media (15ng/mL) for select samples. Adherent and non-adherent cell cultures

were examined using the following procedures.

Adherent Cell Density

Any remaining non-adherent cells were removed from the terminated

adherent cell cultures using a wash with warmed PBS++. Adherent cells were fixed with 4% paraformaldehyde for 1 hour and were subsequently permeabilized with Triton-X-100 (5 minutes). Adherent cell cultures obtained at all timepoints

- 203 - Appendix

(day 0 (2 hours), day 3, day 7, and day 10) were stained with DAPI, a nuclear

fluorescent stain, (Roche, Indianapolis, IN, USA) and Alexa Flour 594, a

phalloidin stain for actin filaments in the cytoskeleton, in accordance with

manufacturer’s instructions. Stained cells were visualized using a fluorescence

microscope (Olympus IX71, Olympus Microscopes, Center Valley, PA). The

combination of nuclear and cytoskeletal fluorescent stains allowed for the visual

identification of both single and multinucleated cells. A total of 10 fields (40x)

were analyzed for the total number of adherent cells, the number of nuclei in

FBGCs, and the number of FBGCs. The adherent cell density is reported in

cells/mm2, while percent fusion was calculated to be the number of nuclei in

FBGCs divided by the total number of nuclei. Both are reported as the mean ±

the standard error of the mean (SEM).

Adherent Cell Apoptosis

For cultures obtained at days 3, 7, and 10, adherent cells were rinsed with

PBS++, fixed, and permeabilized as described above; however prior to

immunostaining with DAPI and Alexa Flour 594, the adherent cells were stained

for late stage apoptosis using TUNEL (Roche, Indianapolis, IN, USA), which

binds to fragmented DNA, in accordance with the manufacturer’s instructions.

Two positive and negative controls were generated and analyzed at each

timepoint. The adherent cells of the positive controls were treated with DNase I

(Roche, Indianapolis, IN, USA) to fragment all DNA prior to staining with TUNEL.

The adherent cells of the negative controls were only stained with the TUNEL

- 204 - Appendix kit’s labeling solution without the addition of the TUNEL enzymatic solution that binds to fragmented DNA. Any visible TUNEL staining in the negative controls was considered to be background.

Fluorescently labeled samples were simultaneously analyzed using fluorescence microscopy within the 500 – 540 nm (FITC) light range for TUNEL,

420 – 480 nm for DAPI, and 590 – 617 nm for Alexa Fluor 594. Adherent cells in ten-40x objective fields were analyzed for apoptosis. Cells that stained positive for TUNEL were considered to be apoptotic. The data is presented as the percentage of apoptotic cells of the total number of adherent cells.

Non-adherent Cell Analysis

Non-adherent cells were counted using trypan blue and a hemocytometer.

In addition, the non-adherent cells were immunoflourescently stained with CD14 for monocytes, CD3 for lymphocytes, CD45 for hematopoetic cells, and FITC for non-viable cells and then analyzed using flow cytometry. All staining and flow cytometry was conducted by Howard Meyerson, M.D. in the Hematopathology

Laboratories of University Hospitals of Cleveland. Data is presented as the total number of non-adherent cells and the percentages of the monocytes, lymphocytes, and non-viable cells in that population (mean ± SEM).

Cytokine/Chemokine/MMP/TIMP Quantification with ELISA

Supernatants were collected via pipetting at 3, 7 and 10 days as mentioned above for specific protein quantification using commercially available

- 205 - Appendix

ELISA kits (R&D Systems, Minneapolis, MN) in accordance with manufacturer’s

instructions. The non-adherent cells in the supernatants were removed using

microcentrifugation at 2000min-1. Cell-free supernatants were stored at -20oC until ELISA analysis could be performed. SFM was utilized as the diluent for both the standards and diluted samples to ensure consistency. Media samples containing serum-free media and 20% heat-treated, autologous serum as used in the cell cultures were measured. Any concentration found in the media controls were subtracted from the concentration measured in the cell culture samples as done in previous studies.35,36 Quantified data is presented as the total

concentration (pg/mL) of a given cytokine/chemokine/MMP/TIMP produced.

Results

Surface Characterization

Contact angle analysis was measured to determine the presence of the

SAM modifications on both the in-house and out-of-house SAM surfaces. For

the in-house SAM surfaces at day 0, the advancing water contact angles for the

o Au control and the -CH3, -OH, and -COOH terminated SAM surfaces were 68 ±

o o o o o o o 11 , 97 ± 2 , 49 ± 3 , and 35 ± 4 , respectively (Figure A.2.A). The -CH3

terminated SAM surface contact angle decreased significantly to 58o ± 1o after incubation in an aqueous environment of serum-free media at 37oC for 5 days

(Figure A.2.B). The contact angle for the COOH-terminated SAM surface

increased slightly to 49o ± 2o, while the -OH terminated SAM surface contact

- 206 - Appendix

# 100 100 Advancing 90 Day 0 90 Day 40 A D Receding 80 80 70 70 60 # 60 # 50 # 50 # 40 40 30 30 20 20

10 10 0 0 Gold Methyl Carboxyl Hydroxyl Gold Methyl Carboxyl Hydroxyl

100 100 90 90

80 B Day 5 80 C Day 18 70 70 # 60 # # 60 # 50 50 40 40 30 30 20 20 10 10 0 0 Gold Methyl Carboxyl Hydroxyl Gold Methyl Carboxyl Hydroxyl

Figure A.2: Advancing and Receding Contact Angles on SAM (in-house) Surfaces. Untreated, dry samples were examined at days 0 (A) and 40 (D) after synthesis. Treated samples were incubated in serum-free media at 37oC for 5 (B) and 18 (C) days post synthesis. Mean ± standard deviation (n=6,3). “#” indicates statistical difference between advancing and receding value sets and advancing and receding values for gold (p<0.05).

angle remained unchanged. After 18 days of treatment in serum-free media at

37oC (Figure A.2.C), the contact angles for the Au control (55o ± 3o) and the –

o o o o o o CH3 (60 ± 2 ), –COOH (50 ± 2 ), and –OH terminated SAM (52 ± 2 ) surfaces

were relatively unchanged. When samples were left untreated and dry for 40+

days at room temperature in the dark (Figure A.2.D), the contact angles

decreased for the CH3-terminated SAM surfaces over time to levels relatively

- 207 - Appendix

comparable to the treated samples at days 5 and 18. The –COOH and –OH

terminated SAM surfaces remained relatively comparable to the values seen at

day 0 (40o ± 4o and 50o ± 2o).

100 Methyl - Advancing ) * o 90 Methyl - Receding 80 * Gold - Advanci ng 70 Gold - Recedi ng 60 50 40 30 20

Water Contact Angle ( 10 0 024681012141618202224

Time (hours)

Figure A.3: Changes in Contact Angles on Methyl SAM (in-house) Surfaces within the 1st day in an Aqueous Environment. Samples were incubated in serum-free media at 37oC. Mean ± Standard deviation (n=3). “*” indicates contact angles for the –CH3 terminated surface that are statistically different than angles for the Au surface (p<0.05).

A time course experiment on the in-house CH3-terminated SAM and Au control surfaces was conducted to determine changes in contact angle values within the first 24 hours of treatment in serum-free media at 37oC (Figure A.3).

The advancing and receding contact angles for the CH3-terminated SAM surface

decreased significantly from t=0 (98o ± 2o) to 2 hours of treatment (81o ± 13o), decreased more by 6 hours (65o ± 4o) and 24 hours (52o ± 2o). The contact

angles for the Au control surfaces were comparable to one another over time and

- 208 - Appendix

ultimately comparable to the CH3-terminated SAM surface by 24 hours of treatment (56o ± 3o).

100

) Advancing o 90 * Recedi ng 80 70 60 50 40 30 20

Water Contact Angle ( 10 0 Methyl Carboxyl Hydroxyl

Material & Functional Group

Figure A.4: Advancing and Receding Contact Angles on SAM (out-of- house) Surfaces. Mean ± Standard deviation. n=3. “*” Indicates a statistical difference between this and the other materials (p<0.05).

Contact angle analysis on the out-of-house SAM surfaces after months of storage at room temperature revealed advancing contact angles of 83o ± 4o, 57o

o o o ± 3 , and 56 ± 3 for the –CH3, –COOH, and –OH terminated SAM surfaces,

respectively (Figure A.4). The values for the –CH3 terminated SAM surface were statistically greater (p<0.05) than the –COOH and –OH terminated SAM surfaces. The values for each of the out-of-house SAM surfaces were distinct from the day 0 contact angles measured for the in-house surfaces.

- 209 - Appendix

AFM Analysis

AFM analysis was conducted on the in-house SAM surfaces at t=0 following synthesis and after 2 hours of treatment in serum-free media at 37oC in

order to further investigate the stability of the SAM alkanethiols. AFM images are

shown in Figure A.5 and the roughness and thickness values are shown in

Table A.1. The gold surface remained unchanged between t=0 and t=2hours with roughness values of 0.40 ± 0.06 and 0.44 ± 0.03 respectively (Figures

A.5.A & A.5.E). The topography for the –CH3 SAM surface was relatively

smooth at t=0 with a roughness of 0.25 ± 0.10 and a thickness of 0.58 ± 0.39 nm

(Figure A.5.B). As suspected, this topography changed after 2 hours of

incubation in serum-free media at 37oC increasing significantly in roughness

(0.76 ± 0.34) and thickness (1.89 ± 0.58 nm) (Figure A.5.F). This change in

topography supports the hypothesis that the alkanethiols desorbed from the surface leaving pits where the desorbed alkanethiols were and peaks where the alkanethiols remain. In Figures A.5.C & A.5.G, strong peaks with average heights of 5.00 ± 2.91 nm were present on the –COOH SAM surfaces following synthesis at t=0 and were reduced after 2 hours of treatment to 3.49 ± 1.72 nm.

The taller peaks may correlate with regions of bi-planar H-bonding between the

COOH terminated alkanethiol chains that desorbed during incubation in serum- free media. This bi-layer formation will be addressed further later in the

- 210 - Appendix

t=0 t=2 hours A E

B F

C G

D H

Figure A.5: AFM Images of the Gold Control (A,E) and the Methyl (B,F), Carboxyl (C,G), and Hydroxyl (D,H) SAM Surfaces Immediately following Synthesis (A-D) and After 2 hours of Treatment (E-H). Samples treated were incubated in serum-free media at 37oC for 2 hours.

- 211 - Appendix

discussion. The –OH SAM surfaces remained relatively comparable before and after treatment in serum-free media at 37oC suggesting that these surfaces are stable up through 2 hours and that the instabilities seen by day 5 occur at time periods longer than 2 hours.

Table A.1 AFM Results for in-house SAM Surfaces Prior to and After Treatment^ Parameter Surface At t=0 After 2 hours of Treatment* Gold 0.40 ± 0.06 0.44 ± 0.03 Roughness Methyl 0.25 ± 0.10 0.76 ± 0.34 Carboxyl 1.97 ± 1.15 1.23 ± 0.52 Hydroxyl 0.61 ± 0.03 0.65 ± 0.30 Thickness/ Gold - - Peak Methyl 0.58 ± 0.39 1.89 ± 0.58 Height Carboxyl 5.00 ± 2.91 3.49 ± 1.72 (nm) Hydroxyl 1.67 ± 0.12 1.77 ± 1.16 * Treated samples were incubated in serum-free media at 37CC for 2 hours ^ Results are the mean ± standard deviation (n=3).

Cellular Adhesion

The adherent cell densities were determined on both the in-house and out-of-house SAM surfaces (Figure A.6). For the in-house SAM surfaces, cellular adhesion was comparable on all surfaces at day 0 averaging 2,250±90 cells/mm2 (Figure A.6.A). The adherent cell density decreased significantly on

2 the –CH3 terminated SAM (691±279 cells/mm ) and Au control (1,120±240 cells/mm2) surfaces by day 3 to values statistically less than the –COOH

(2,030±160 cells/mm2) and –OH terminated SAM (1830±170 cells/mm2) surfaces

(p<0.05). A decrease in adherent cell densities continued on all surfaces by days

- 212 - Appendix

7 and 10 in the presence and absence of IL-4; however there were no significant

differences between surfaces at these timepoints.

On the out-of-house SAM surfaces, cellular adhesion was comparable on

all surfaces at each timepoint and IL-4 condition (Figure A.6.B). Cellular

adhesion decreased over time as expected; except at day 10, when increased

densities comparable to day 0 were seen. McNally et.al. showed initially monocyte adhesion is mediated by β2 integrins, while β1 integrins are

subsequently expressed with the induction of fusion.42 The greater levels of

cellular adhesion seen at day 10 in comparison to earlier timepoints may result

from the increased expression of both β2 and β1 integrins in the adherent cells

over time creating stronger cellular-biomaterial adhesive interactions that reduce

the number of cells washed away during the fixation and staining processes.

In general, the adherent cellular densities on the out-of-house SAM surfaces were greater than those seen on the in-house SAM surfaces.

Preparation and storage conditions may have lead to the loss of the alkanethiols and/or loss or oxidation of the metallic surface potentially accounting for these differences.

- 213 - Appendix

3000 Gold Methyl 2500 ^ * A Carboxyl ^ * Hydroxyl 2000 ^

1500 ^ 1000

500 Adherent Cell Density (Cells/mm^2) Density Cell Adherent 0 Day 0 Day 3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4 9000 Methyl 8000 Carboxyl B 7000 Hydroxyl

6000

5000

4000

3000

2000

1000 Adherent Cell Density (Cells/mm^2) Density Cell Adherent

0 Day 0 Day 3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4 Time (days) & Condition

Figure A.6: Cellular Adhesion on SAM Surfaces in-house (A) and out-of- house (B) over Time. Mean ± SEM, n=3. “^” indicates a significant decrease in adhesion occurs after this timepoint (p=0.002). “*” indicates that the values for these materials are statistically different from the methyl SAM surface (p=0.03).

- 214 - Appendix

Macrophage Fusion into FBGCs

Macrophage fusion was analyzed on both the in-house and out-of-house

SAM surfaces (Figure A.7). On the in-house SAM surfaces, macrophage fusion into FBGCs was equivalently minimal on all materials at day 3 (Figure A.7.A).

Percent fusion increased over time and remained comparable on the majority of the surfaces. Macrophage fusion was significantly greater on the –OH terminated SAM surface at day 7 without (10%±6%) or with (19%±9%) IL-4 and at day 10 with IL-4 (33%±12%). Also, fusion was greater on the –COOH

terminated SAM surface than on the Au control or –CH3 terminated SAM surface at day 10 with IL-4 (16%±6%).

For the out-of-house SAM surfaces, no macrophage fusion was present at day 3 on any surface, but was found at days 7 and 10 (Figure A.7.B). Percent fusion was greater on the –CH3 terminated SAM surface at days 7 without

(17%±14%) and with (16%±8%) IL-4 and at day 10 without IL-4 (17%±4%) in comparison to the comparable –COOH and –OH terminated SAM surfaces, which were less than 10%. At day 10 in the presence of the fusion-inducing cytokine IL-4, percent fusion increased on all surfaces to a comparable level of

27%±2%.

- 215 - Appendix

60% Gold Methyl 50% Carboxyl A Hydroxyl ^ 40%

30%

20% Percent Fusion (%)

10%

0% Day 3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4 60% Methyl Carboxyl B 50% Hydroxyl * 40%

30%

20% Percent Fusion (%) Fusion Percent

10%

0% Day 3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4 Time (days) & Condition

Figure A.7: Macrophage Fusion into FBGCs on SAM Surfaces in-house (A) and out-of-house (B) over Time. Mean ± SEM, n=3. “^” indicates that the values for this material are statistically greater than gold (p=0.04). “*” indicates that the values at this timepoint are significantly greater than at day 3 (p<0.05).

Monocyte/Macrophage Apoptosis

Monocyte/macrophage apoptosis was examined on the in-house SAM surfaces only (Figure A.8). The percentage of apoptotic cells on the Au control and –CH3 terminated SAM surfaces were comparable (8%±7% and 11%±9%) and slightly greater than the percentages on the –COOH (1%±0%) and –OH

- 216 - Appendix

terminated SAM (1%±0%) surfaces at day 3. This trend remained at day 7

without IL-4, however at day 7 in the presence of IL-4 and by day 10, percent

apoptosis was relatively comparable on all materials.

60% Gold Methyl 50% Carboxyl Hydroxyl 40%

30%

20% Percent Apoptosis (%) 10%

0% Day 3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4 Time (days) & Condition

Figure A.8: Cellular Apoptosis on SAM Surfaces (in-house) over Time. Percent apoptosis is the percentage of total number of adherent nuclei that are apoptotic. Mean ± SEM, n=3.

Cellular Detachment

The numbers of non-adherent cells collected via gentle pipetting were measured using hemocytometry on the in-house SAM surfaces only (Figure

A.9). The numbers of non-adherent cells were comparable for all material at 2 hours, 3 days, and 7 days. Samples at day 10 contained too few cells to measure accurately with similar concentration procedures. Approximately

290,000 ± 110,000 cells or 38% of the ~750,000 cells plated did not adhere

within 2 hours. A total of ~59,000 ± 32,000 cells were non-adherent by day 3

with fewer non-adherent cells at day 7 (~55,000 ± 28,000 cells).

- 217 - Appendix

500000 * Gold 450000 Methyl 400000 Carboxyl Hydroxyl 350000

300000

250000

200000

150000

Non-Adherent Cells (cells)100000

50000

0 2 Hours 3 Days 7 Days Time & Condition

Figure A.9: Cellular Detachment on SAM Surfaces (in-house) at each Timepoint. Mean ± SEM, n=3,4. “*” indicates that the values at 2 hours are significant greater than at subsequent timepoints (p<0.05).

Table A.2 Mixed Monocyte and Lymphocyte Non-Adherent Cell Population Ratio of Non-Adherent Lymphocytes and Monocytes* Materials 2 hours 3 days 7 days Gold 7±3 0±0 4±1 Methyl SAM 6±2 1±0 4±1 Carboxyl SAM 11±5 1±1 7±4 Hydroxyl SAM 13±1 0±0 2±1 *Ratio of non-adherent lymphocytes to monocytes is the number of CD3+ lymphocytes to every single CD14+ monocyte measured in the non-adherent cell population using flow cytometry. Mean ± SEM.

Presence of Lymphocytes in Non-Adherent Cell Populations

Flow cytometry analysis was utilized on the non-adherent cell populations

in cell cultures containing the in-house SAM surfaces to determine the presence

of lymphocytes in culture (Table A.2). At 2 hours, there were approximately 9±3

non-adherent CD3+ lymphocytes for every 1 non-adherent CD14+ monocyte cell

(9:1). This ratio of CD3+ lymphocytes to CD14+ monocytes decreased to 1:1 at

- 218 - Appendix day 3 and 4:1 at day 7. Ratios were comparable on all surfaces at each timepoint.

Cytokine/Chemokine/MMP/TIMP Production

The concentrations of select cytokines, chemokines, MMPs, and TIMPs were quantified using ELISA in the supernatants of cell cultures on the out-of- house SAM surfaces only at days 3, 7, and 10 (Figure A.10 & Figure A.11).

Proteins investigated were selected based on previous investigations in our laboratory showing that these cytokines/chemokines/MMP/TIMPs were present in monocyte/macrophage/FBGC cell cultures.35,36

Pro-inflammatory cytokines, IL-1β and IL-6 concentrations were comparable between materials at day 3 at average values of 39±20 pg/mL and

146±81 pg/mL, respectively, and subsequently decreased by day 7 and 10

(Figure A.10.A & A.10.B). The concentrations of the anti-inflammatory cytokine, IL-10, were also comparable between materials at each timepoint, yet remained comparable in concentration between days 3 (24±10 pg/mL) and 10

(15±7 pg/mL) (Figure A.10.C). By day 7, the addition of IL-4 increased the concentrations of IL-6 on the –CH3 and –OH terminated SAM surfaces and of IL-

10 on the –COOH and –OH terminated SAM surfaces.

Chemokine IL-8 concentrations were comparable between materials at each timepoint and decreased from 23,500±13,100 pg/mL at day 3 to

5,080±1,150 pg/mL at day 10 (Figure A.10.D). Chemokine, MIP-

1β concentrations were comparable between materials at all timepoint and

- 219 - Appendix

conditions except for on the –CH3 terminated SAM surface which was significantly greater than the other surfaces at day 7 in the presence of IL-4

(Figure A.10.E). In addition, the average MIP-1β concentrations increased

between day 3 (192±48 pg/mL) and day 7 (2,730±1,710 pg/mL) and remained

relatively constant at day 10 (1,170±700 pg/mL). IL-8 and MIP-1β concentrations

did not change in the presence of IL-4 except for the increased MIP-1β

concentration of the –CH3 terminated SAM surface at day 7, previously

mentioned.

MMP-9 concentrations were significantly greater than the TIMP-1 and

TIMP-2 concentrations measured (Figure A.11.A-C). For each MMP/TIMP, the concentrations were comparable between materials at each timepoint and condition. MMP-9 concentrations were minimal at day 3 (249±59 pg/mL) and increased significantly at days 7 (3,070±970 pg/mL) and 10 (3,070±550 pg/mL); however, IL-4 addition decreased MMP-9 concentration at both days 7 and 10 to levels more comparable to day 3 (Figure A.11.A). TIMP-1 concentrations also

were minimal at day 3 (13±8 pg/mL) and increased by days 7 (62±25 pg/mL) and

10 (52±13 pg/mL); however, IL-4 addition did not significantly affect on TIMP-1

concentrations (Figure A.11.B). TIMP-2 concentrations increased steadily from

0±0 pg/mL at day 3 to 18±9 pg/mL at day 7 and 31±9 pg/mL at day 10 with a minimal decrease in concentration with IL-4 addition (Figure A.11.C).

- 220 - Appendix

100 90 Methyl 80 Carboxyl 70 * Hydroxyl 60

, pg/mL 50 β A 40 IL-1 IL-1 30 20 10 0 Day 3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4 800 700 B 600 L 500 400 300 pg/m IL-6, 200

100

0 Day 3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4 60

50 C

L 40 30

pg/m IL-10, 20

0 Day 3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4 60000 50000 D

L 40000

30000

pg/m IL-8, 20000 10000 0 Day 3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4 22500 * 20000 E 17500 15000 12500 , pg/mL β 10000 7500 MIP-1 ** 5000 2500 0 Day 3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4 Time & Condition Figure A.10: Cytokine (A-C) and Chemokine (D-E) Production from Cell Cultures on in-house SAM Surfaces. Mean ± SEM, n=3. “*” indicates that the values at this timepoint are statistically greater than at subsequent timepoints (p<0.05).

- 221 - Appendix

4500 * Methyl 4000 A * Carboxyl 3500 Hydroxyl 3000 2500 2000 1500 * pg/mL MMP-9, 1000 500 0 Day 3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4 120 B 100 L 80

60 40 TIMP-1, pg/m 20 0 Day 3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4 45 C * * 40 35

L * 30 * 25 20 15 TIMP-2, pg/m 10 5 0 Day 3 Day 7 Day 7 with IL-4 Day 10 Day 10 with IL-4 Time & Condition

Figure A.11: MMP (A) and TIMP (B-C) Production from Cell Cultures on in-house SAM Surfaces. Mean ± SEM, n=3. “*” indicates values statistically greater than at day 3 (p<0.05).

- 222 - Appendix

Discussion

Analysis of Initial SAM Study Using Out-of-House SAM Surfaces

Initial experiments analyzing macrophage adhesion, fusion, and cytokine

production were conducted using SAM surfaces supplied to us months

previously. Contact angle analysis revealed that the expected surface

chemistries were not present on the SAM surfaces and that the

hydrophobicity/hydrophilicity of these surfaces was comparable (Figure A.3).

Consequently as expected, the cellular adhesion, fusion, apoptosis, detachment, and cytokine production were not statistically different between out-of-house

SAM surfaces at each timepoint and condition (Figures A.6, A.7, A.8, A.9, A.10,

& A.11). Recent studies suggest that the stability of the SAM surface is questionable over time due to various factors explained below; therefore a SAM

synthesis protocol was developed based upon commonly utilized techniques in

order to synthesis fresh SAM surfaces for each experiment (Table A.3).2,7,13,14,43-

46 A secondary study was conducted with the in-house SAM surfaces that

analyzed the stability of the SAM surface over time under dry and aqueous

conditions similar to cell culture in addition to an analysis of macrophage

adhesion, fusion, apoptosis, and detachment.

- 223 -

Table A.3A Commonly Used and Chosen Self-Assembled Monolayer Synthesis Techniques (Part I) Synthesis Laboratory Synthesis Techniques* Procedure/Parameter Our Laboratory T. Matsuda45 G.M. Whitesides2 A.J. Garcia43,44,46 Alkanethiol Utilized HS-(CH2)11-CH3 HS-(CH2)11-CH3 HS-(CH2)15- CH3 HS-(CH2)11-CH3 HS-(CH2)11-OH HS-(CH2)11-OH HS-(CH2)10-OH HS-(CH2)11-COOH HS-(CH2)11-COOH HS-(CH2)11-COOH Substrate Utilized Glass coverslip Glass coverslip Glass coverslip Chamber slides 1. Substrate Cleaning Radio frequency Ethanol & air dry - - Procedure 2 min. @ 125W 2. Coating Technique Sputter coating Electron beam Electron beam Electron beam evaporation evaporation evaporation Specs 4.0 x 10 -6 torr - - 2x10-6 Torr, 2 Å/s 2M/torr 2 min. @ 125 W Gold Coating Thickness 100nm 200nm 12nm or silver 12nm 20nm Base Coating Thickness Titanium 10nm Chromium 200nm Titanium 1.5 nm Titanium 10nm - 224 3. Coated Substrates Acetone, ethanol, Acetone, ethanol, - Used immediately Cleaning Procedure ddH2O (3x), ethanol ddH2O (3x), ethanol following coating process

4. Alkanethiol Solution 1.0mM in ethanol 1.0mM in ethanol 2.0mM in ethanol 1.0mM in ethanol Concentration (N2 environment) 5. Incubation Time/Condition 18 hrs at RT 24 hrs at RT 12 hrs 12 hrs 6. Rinsing Procedure Ethanol (3x) Ethanol, ultrasonication Ethanol Ethanol for 2 min., ethanol 7. Drying Procedure Nitrogen Nitrogen Nitrogen 8. Equilibration PBS for 15 min. 3 baths in sterile PBS - PBS for 15 min. Storage PBS PBS - - Time Until Use 2 hrs 2 hrs - - Characterization Contact angle XPS & Contact angle XPS & Contact angle *Absolute ethanol was utilized in each protocol.

Table A.3B Commonly Used and Chosen Self-Assembled Monolayer Synthesis Techniques (Part II) Synthesis Laboratory Synthesis Techniques Procedure/Parameter B.D. Ratner11 P. Tengvall14 S. Downes13 Alkanethiol Utilized HS-(CH2)15- CH3 HS(CH2)15CH3 HS-(CH2)7-CH3 HS-(CH2)11-OH HS(CH2)16OH HS-(CH2)2-COOH Substrate Utilized Silicon wafers Titanium disc Glass coverslip 1. Substrate Cleaning - - hot piranha solution^, ddH2O, Procedure dry by baking at 80oC for 2 hrs 2. Coating Technique Ion beam sputtering Physical vapor Electron beam deposition evaporation Specs 3.5x10-5 Torr 10-7 torr 0.5 Å/s 99% purity

Gold Coating Thickness 25nm (0.033nm/s) 200nm 20-40nm Base Coating Thickness Chromium 5nm (0.050nm/s) Titanium disc (1mm) Chromium 3. Coated Substrates acetone (2x), piranha solution, None - 225 Cleaning Procedure ethanol, ddH2O, 2min ultrasonication, dry with argon 4. Alkanethiol Solution 1.0mM in ethanol 2mM ethanolic solution 1.0mM in degassed ethanol Concentration (N2 environment) (N2 environment) 5. Incubation Time/Condition 24 hours at RT 18-48 hours 12-18 hours 6. Rinsing Procedure Ethanol, sonication for 2 min. (3x) Ethanol, ultrasonication Degassed ethanol for 1 min, ethanol (3x) 7. Drying Procedure Argon - Nitrogen 8. Equilibration ddH2O (3x) Storage Nitrogen filled container Hank’s balanced salt Use within 2 hours solution Time Until Use Characterization XPS & Contact Angle Contact Angle *Absolute ethanol was utilized in each protocol.

^ Piranha solution contained 7 parts concentrated H2SO4 and 3 parts 30% H2O2.

Appendix I

SAM Surface Instability

Contact angle analysis of each of the in-house SAM surfaces taken

immediately following synthesis confirmed the presence of the alkanethiol on the

material surface post-synthesis. Advancing contact angles on the –CH3

terminated SAM surfaces were comparable to the expected, previously reported

values of 101o, 107o, 117o (Figure A.2).12,47 Contact angles for both –COOH and

–OH terminated SAM surfaces were higher than the expected values, 25o and

20o, respectively, but were within range of other previously reported values. Prior research conducted in numerous laboratories have reported a wide range of contact angle values for the –COOH (28o, 42o, and 72o) and –OH (25o and 82o) terminated SAM surfaces.12,21,47,48 Defects in the SAM surface during synthesis

and/or subsequent alkanethiol degradation would remove or disrupt these self- assembled hydrophilic-terminated alkanethiols thus altering the chemical groups exposed at the material surface and may ultimately account for the observed increased contact angles and hydrophobicity.

The SAM synthesis is considered to occur within a two-step process verses a one-step process, as previous thought. Initially, the alkanethiols adsorb onto the gold surface within minutes forming Au-S bonds, then the alkanethiol chains spontaneously assemble due to van der Waals interactions within hours to a few days. Defects in the self-assembled monolayer surface can result from synthesis parameters including but not limited to debris or defects on the gold substrate, low concentrations or impure alkanethiol solutions, incubation temperatures, insufficient time allowed for assembly (typically need 2-12 hrs for

- 226 - Appendix I

long chain thiols and at least 24 hours for short chain thiols), and high oxygen

content in the alkanethiol solution and environment.6,49 Therefore, the SAM

synthesis parameters utilized in this study were chosen carefully to minimize

these inhibitory factors and based upon commonly utilized SAM synthesis

techniques (Table A.3).2,7,13,14,43-46 In addition, studies have shown that longer

alkanethiol chains (C18) form SAM surfaces with fewer defects than short chain

alkanethiols (n<9) and that fewer defects occur with higher incubation

temperatures.27,49 The alkanethiols utilized within this study were of a moderate

to long chain length (C12) and were incubated at room temperature. Elevated

temperatures were not feasible due to physical constraints of the inflatable,

nitrogen filled Atmosbag® (Sigma-Aldrich, St. Louis, MO). Research has also

shown that unbound alkanethiols can remain at the material surface.49,50 It has

been suggested that these alkanethiols in the solution form interplanar hydrogen

bonds with the –COOH terminal groups of alkanethiols within the monolayer

potentially forming a partial bi-layer. Wang et.al. found that adding small

molecules (i.e. CF3COOH), which can form hydrogen bonds with the alkanethiols

in the solution during synthesis, improves the quality of the SAM surfaces and

decreases the water contact angles.21 AFM analysis revealed the potential

presence of this bi-layer effect on the –COOH SAM surfaces with peak heights

over twice the size of the alkanethiol chains. These peaks diminish after 2 hours of treatment in serum-free media at 37oC suggesting that the bi-layer desorbed

within the aqueous environment. This bi-layer effect may also account for the

slightly greater contact angle initially seen following synthesis of these materials.

- 227 - Appendix I

Variations in the contact angles over time revealed that the SAM surfaces were unstable in both a dry, room temperature environment (untreated condition) and within serum-free media at 37oC (treated condition) over long periods of time

(Figures A.2 & A.3). This instability began as quickly as 2 hours for the –CH3 terminated SAM surface incubated in serum-free media, in agreement with similar findings by Scotchford et.al.13 SAM instability was seen on all of the SAM surfaces after 5 days of incubation in serum-free media at 37oC and within 30

days when left untreated and dry.

Once assembled, alkanethiol SAM surfaces are susceptible to

degradation due to UV, increasing temperatures particularly those greater than

70oC, low pH, direct-desorption, and oxidation-desorption.20,24,29 For these reasons, the samples were kept in a dark environment and were not exposed to temperatures above 37oC.

Research by Silver et.al. investigated the pericellular pH values of murine

peritoneal macrophages and osteoclast giant cells collected from neonatal rat

femurs or osteoporotic chicken wing bones.51 The macrophages and osteoclasts

were seeded onto collagen substratum and the pericellular pH values of 3.6-3.7

for macrophages and 4.0 for osteoclasts were measured between the cell and substratum using microelectrodes. These decreased pH values beneath the adherent macrophages may further promote SAM degradation. Research by

Chen et. al. suggested that SAM degradation was cell-dependent.25 In this

study, bovine pulmonary artery endothelial cells or 3T3-L1 pre-adipocytes were

seeded onto gold substrates coated with non-adhesive hexa(ethylene glycol)-

- 228 - Appendix I

terminated alkanethiols (EG6 SAM) and patterned with hexadecanethiol. After

the cells were seeded and adhered to the hexadecanthiol patterns, the time until

this pattern failed was measured to determine the timeframe in which the non-

adhesive EG6 SAM surface degraded. The time until failure on the EG6 SAM

surface was more rapid for the L1 adipocytes (4 days) versus the endothelial

cells (9 days). Chen et.al. hypothesized that this increase of EG6 degradation in

the presence of adipocytes occurs because the adipocytes potentially produced

alcohol dehydrogenase (ADH) or aldehyde hydrogenase, two enzymes that are

associated with fatty acid metabolism and have been shown to oxidize and

degrade poly(ethylene glycol) chains. This hypothesis was further supported by

an increased in the time until failure when two inhibitors of ADH activity were

included in cell culture. Adherent macrophages are known to secrete reactive

oxygen species and enzymes. Theoretically, these adherent macrophages could

also increase SAM degradation by producing reactive oxygen species and

enzymes that specifically degrade the SAM material of interest. In this current

study, contact angle analysis was conducted on surfaces that were untreated

and treated in media without cells. Therefore, the changes in SAM surface

chemistry measured may potentially be more unstable on the SAM surfaces in

cell culture.

Direct desorption occurs in liquid media in which the alkanethiols desorb

as disulfides.20,22,24 Upon exposure to air, SAM alkanethiols can also oxidize into sulfinates and sulfonates , which quickly desorb from the surface upon exposure to solvents.20,22,24,28 Increasing temperatures further promotes desorption.20,52

- 229 - Appendix I

Significant research has showed that SAM surfaces became unstable over time

as indicated by changes in IR, XPS, SPR, contact angle analysis, and Raman

spectroscopy results13,22,23,25,28. Studies have indicated that SAM degradation

occurs rapidly within hours of exposure to air and that subsequent immersion in

liquids or furthers this degradation.24,28 This SAM instability is further promoted in

cell culture or upon immersion in media or PBS.22,23 In addition, degradation of

SAMs has also been shown to begin at boundaries of defects; therefore,

potential defects resulting from synthesis, discussed above, provide additional

locations for degradation.20,26 A combination of these mechanisms: direct-

desorption, oxidation, and synthesis defects; may explain the instability of these

SAM surfaces over time.

Analysis of Cellular/Biomaterial Interactions

Naturally, following the determination of unstable surfaces, the next

question probes at the validity of results from the cellular/biomaterial interactions analysis and what information can be gained from this component of the

research. For the most part, adhesion, fusion, apoptosis, detachment and

cytokine production were similar on all surfaces.

Interestingly, the higher levels of adhesion on the –COOH vs. the –CH3

terminated SAM surfaces at early timepoints and the instability of the surfaces

within a short period of time mirrors a previous study conducted analyzing the

osteoblasts-like cells and the stability of a HS(CH2)2COOH and HS(CH2)7CH3

containing SAM surfaces within 24 hours.13 This study found that a greater

- 230 - Appendix I

number of osteoblast-like cells adhered to –COOH terminated SAM surfaces at

24 hours and that the contact angles decreased significantly for both the –CH3

and –COOH terminated SAM surfaces incubated in serum-free media or DMEM

containing fetal bovine serum.13 Although it is a different cell type, the studies

were conducted using similar synthesis techniques and conditions suggesting

that the cellular/biomaterial interactions seen in this study are valid with respect

to these particular SAM surfaces. Conclusions regarding the effects of specific

functional groups can not be made due to the unknown nature of the SAM

surface over time.

Contact angle analysis reveals that degradation or loss of the SAM unit

occurs at a greater rate on the –CH3 terminated SAM surface in comparison to

the –COOH and –OH terminated SAM surfaces. Therefore as time continues,

the Au surface area exposed on each surface increases up through day 10 when

the –CH3 terminated SAM surfaces and the Au surface are equivalent as confirmed by the comparable adhesion, fusion, and apoptosis results. By this

time, the contact angles of –COOH and –OH terminated SAM surface have

increased to values closer to, but still statistically significant from the –Au surface

indicating that some degradation has occurred increasing the Au surface area

exposed, but to a lesser extent than on the –CH3 terminated SAM surfaces. An

analysis of the Au surface data shows that as time increases, cellular adhesion

rapidly decreases initially and then continues to decrease over time correlating

with a rapid increase in apoptosis initially followed by a decrease in apoptosis

over time (Figures A.6 & A.8). Macrophage fusion remains minimal over time

- 231 - Appendix I

on the Au surfaces (Figure A.7). Analysis of the Au only surfaces can be utilized

to implicate cellular behaviors on Au electrode biosensors.

Previous research in our laboratory showed that the addition of IL-4 to

culture inhibited concentrations of matrix metalloproteinase-9 (MMP-9) in

macrophage cell cultures while not effecting it’s inhibitors, TIMP-1 and TIMP-2.36

This was confirmed by the significant decreased in concentration levels of MMP-

9 and comparable concentration levels of TIMP-1 and TIMP-2 with the addition of

IL-4 at both days 7 and 10 (Figure A.11). Similarly, this research confirmed

previous findings of an increase in IL-10 and MIP-1β concentrations with the

addition of IL-4 (Figure A.10).35 It also showed over time a decrease in IL-1β, IL-

6, and IL-8 concentrations, comparable IL-10 concentrations, and low MIP-1β concentrations correlating with that previous study.35

Apoptosis levels increased on the Au control and –CH3 terminated

surfaces, while detachment was comparable on all surfaces at timeframes

correlating to significantly decreased adhesion on these particular surfaces

(Figures A.6, A.8, & A.9). This indicates that apoptosis was the primary means

of inhibiting adhesion and that these particular surfaces induced apoptosis.

Conclusions

Self-assembled monolayers surfaces have significant instability issues

making them a questionable model surface system for utilization in studies of

cellular/biomaterial interactions. Yet, numerous studies have been conducted

with these seemingly simplistic materials. Many studies examine these materials

- 232 - Appendix I

over short periods of time (minutes to <24 hours) and begin use of the samples

within 2 hours of synthesis. For some SAMs, this is potentially within the

timeframe during which the SAM surfaces are intact. In order to suggest SAM

stability, many studies investigating cellular and/or protein interactions with

biomaterials characterize their surfaces with contact angle analysis.

Unfortunately, the times at which the SAM surfaces were analyzed post- synthesis were not indicated in many studies. In addition, the surfaces were analyzed only once without an analysis of surface stability over the timeframe of the study. Research investigating SAM stability has led to many advances in the

SAM synthesis techniques that have improved the characteristics of these surfaces; however, long-term analysis of the SAM surface stability is needed.

References

1. Lahiri J, Isaacs L, Tien J, Whitesides GM. A strategy for the generation of

surfaces presenting ligands for studies of binding based on an active ester

as a common reactive intermediate: a surface plasmon resonance study.

Analytical Chemistry 1999;71(4):777-790.

2. Mrksich M, Dike L, Tien J, Ingber D, Whitesides G. Using microcontact

printing to pattern the attachment of mammalian cells to self-assembled

monolayers of alkanethiolates on transparent films of gold and silver.

Experimental Cell Research 1997;235:305-313.

3. Ito Y. Surface micropatterning to regulate cell functions. Biomaterials

1999;20:2333-2342.

- 233 - Appendix I

4. Kane RS, Takayama S, Ostuni E, Ingber DE, Whitesides GM. Patterning

proteins and cells using soft lithography. Biomaterials 1999;20:2363-2376.

5. Whitesides GM, Ostuni E, Takayama S, Jiang X, Ingber DE. Soft

lithography in biology and biochemistry. Annual Reviews of Biomedical

Engineering 2001;3(335-373).

6. Love JC, Estroff LA, Kreibel JK, Nuzzo RG, Whitesides GM. Self-

assembled monolayers as thiolates on metals as a form of

nanotechnology. Chemical Reviews 2005;105:1103-1169.

7. Mrksich M, Chen CS, Xia Y, Dike LE, Ingber DE, Whitesides GM.

Controlling cell attachment on contoured surfaces with self-assembled

monolayers of alkanethiolates on gold. Proceedings of the National

Academy of Science 1996;93:10775-10778.

8. Zhang F, Kang ET, Neoh KG, Huang W. Modification of gold surface by

grafting of poly(ethylene glycol) for reduction in protein adsorption and

platelet adhesion. Journal of Biomaterial Science Polymer Edition

2001;12(5):515-531.

9. Latour RA, Rini CJ. Theoretical analysis of adsorption thermodynamics for

hydrophobic peptide residues on SAM surfaces of varying functionality.

Journal of Biomedical Materials Research 2002;60(564-577).

10. Martins C, Naeemi E, Ratner B, Barbosa M. Albumin adsorption on

Cibaron Blue F3G-A immobilized onto oligo(ethylene glycol)-terminated

self-assembled monolayers. Journal of Material Science: Materials in

Medicine 2003;14:945-954.

- 234 - Appendix I

11. Martins MCL, Fonseca C, Barbosa MA, Ratner BD. Albumin adsorption on

alkanethiols self-assembled monolayers on gold electrodes studied by

chronopotentiometry. Biomaterials 2003;24:3697-3706.

12. Michael KE, Vernekar VN, Keselowsky BG. Adsorption-Induced

conformational changes in fibronectin due to interactions with well-defined

surface chemistries. Langmuir 2003;19:8033-8040.

13. Scotchford CA, Cooper E, Leggett GJ, Downes S. Growth of human

osteoblast-like cells on alkanethiol on gold self-assembled monolayers:

The effect of surface chemistry. Journal of Biomedical Materials Research

1998;41:431-442.

14. Kalltorp M, Oblogina S, Jacobsson S, Karlsson A, Tengvall P, Thomsen P.

In vivo cell recruitment, cytokine release, and chemiluminescence

response at gold, and thiol functionalized surfaces. Biomaterials

1999;20:2123-2137.

15. Lindblad M, Lestelius M, Johansson A, Tengvall P, Thomsen P. Cell and

soft tissue interactions with methyl- and hydroxyl-terminated alkanethiols

on gold surfaces. Biomaterials 1997;18(1059-1068).

16. Chen CS, Mrksich M, Huang S, Whitesides GM, Ingber DE.

Micropatterned surfaces for control of cell shape, position, and function.

Biotechnology Progress 1998;14(356-363).

17. Jiang X, Bruzewicz DA, Wong AP, Piel M, Whitesides GM. Directing cell

migration with asymmetric micropatterns. Proceedings of the National

Academy of Science 2005;102(4):975-978.

- 235 - Appendix I

18. Datwani SS, Vijayendran RA, Johnson E, Biondi SA. Mixed alkanethiol

self-assembled monolayers as substrates for microarraying applications.

Langmuir 2004;20(4970-4976).

19. Slaughter GE, Bieberich E, Wnek GE, Wynne KJ, Guiseppi-Elie A.

Improving neuron-to-electrode surface attachment via alkanethiol self-

assembly: an alternating current impedance study. Langmuir

2004;20(7189-7200).

20. Yang G, Amro NA, Starkwolfe ZB, Liu G. Molecular-level approach to

inhibit degradations of alkanethiol self-assembled monolayers in aqueous

media. Langmuir 2004;20:3995-4003.

21. Wang H, Chen S, Li L, Jiang S. Improved method for the preparation of

carboxylic acid and amine terminated self-assembled monolayers of

alkanethiolates. Langmuir 2005;21(7):2634-2636.

22. Flynn NT, Tran TNT, Cima MJ, Langer R. Long-term stability of self-

assembled monolayers in biological media. Langmuir 2003;19:10909-

10915.

23. Jiang X, Bruzewicz DA, Thant MM, Whitesides GM. Palladium as a

substrate for self-assembled monolayers used in biotechnology. Analytical

Chemistry 2004;76:6116-6121.

24. Kondoh H, Kodama C, Sumida H, Nozoye H. Molecular processes of

adsorption and desorption of alkanethiol monolayers on Au(111). Journal

of Chemical Physics 1999;111(3):1175-1184.

- 236 - Appendix I

25. Nelson CM, Raghavan S, Tan JL, Chen CS. Degradation of

micropatterened surfaces by cell-dependent and -independent processes.

Langmuir 2003;19:1493-1499.

26. Noh J, Hara M. Molecular-scale desorption processes and the alternating

missing-row phase of alkanethiol self-assembled monolayers on Au(111).

Langmuir 2001;17:7280-7285.

27. Prathima N, Harini M, Raj N, Chandrashekara RH, Ayappa KGS, S.,

Biswas SK. Thermal study of accumulation of conformational disorders in

the self-assembled monolayers of C8 and C18 alkanethiols on the Au(111)

surface. Langmuir 2005;21:2364-2374.

28. Schoenfisch MH, Pemberton JE. Air stability of alkanethiol self-assembled

monolayers on silver and gold surfaces. Journal of American Chemical

Society 1998;120:4502-4513.

29. Zhang Y, Terrill RH, Tanzer TA, Bohn PW. Ozonolysis is the primary

cause of UV photooxidation of alkanethiolate monolayers at low

irradiance. Journal of American Chemical Society 1998;120:2654-2655.

30. Brodbeck W, Shive M, Colton E, Nakayama Y, Matsuda T, Anderson JM.

Influence of biomaterial surface chemistry on the apoptosis of adherent

cells. Journal of Biomedical Materials Research 2001;55:661-668.

31. Brodbeck W, Nakayma Y, Matsuda T, Colton E, Ziats N, Anderson JM.

Biomaterial surface chemistry dictates adherent monocyte/macrophage

cytokine expression in vitro. Cytokine 2002;18(6):311-319.

- 237 - Appendix I

32. Brodbeck W, Patel J, Voskerician G, Christenson E, Shive M, Nakayama

Y, Matsuda T, Ziats N, Anderson JM. Biomaterial adherent macrophage

apoptosis is increased by hydrophilic and anionic substrates in vivo.

Proceedings of the National Academy of Science 2002;99(16):10287-

10292.

33. Brodbeck W, Voskerician G, Ziats N, Nakayama Y, Matsuda T, Anderson

JM. In vivo leukocyte cytokine mRNA responses to biomaterials are

dependent on surface chemistry. Journal of Biomedical Materials

Research 2003;64A:320-329.

34. Jones J, Dadsetan M, Collier T, Ebert M, Stokes K, Ward R, Hiltner A,

Anderson JM. Macrophage behavior on surface modified biomaterials.

Journal of Biomaterial Science Polymer Edition 2004;15(5):567-564.

35. Jones J, Chang D, Colton E, Kwon I, Matsuda T, Anderson J. Proteomic

analysis and quantification of cytokines and chemokines from biomaterial

surface-adherent macrophages and foreign body giant cells. Journal of

Biome dial Materials Research 2006;In Press.

36. Jones JA, McNally AK, Chang DT, Qin LA, Meyerson H, Colton E, Kwon

IK, Matsuda T, Anderson JM. Matrix metalloproteinases and their

Inhibitors in the foreign body reaction. Journal of Biomedical Materials

Research 2006;In Press.

37. Barbosa JN, Barbosa MA, Aguas AP. Adhesion of human leukocytes to

biomaterials: An in vitro study using alkanethiolate monolayers with

- 238 - Appendix I

different chemically functionalized surfaces. Journal of Biomedical

Materials Research 2003;65A:429-434.

38. Barbosa JN, Barbosa MA, Aguas AP. Inflammatory Cell Recruitment and

Adhesion to Methyl-Terminated Self-Assembled Monolayers: Effect of

Implantation Time. Microsc. Res. Tech. 2005;66:37-42.

39. Barbosa JN, Madureira P, Barbosa MA, Aguas AP. The attraction of Mac-

1+ phagocytes during acute inflammation by methyl-coated self-assembled

monolayers. Biomaterials 2005;26:3021-3027.

40. Barbosa JN, Barbosa MA, Aguas AP. Inflammatory responses and cell

adhesion to self-assembled monolayers of alkanethiolate on gold.

Biomaterials 2004;25(2257-2263).

41. McNally AK, Anderson JM. Complement C3 participation in monocyte

adhesion to different surfaces. Proceedings of the National Academy of

Science 1994;91:10119-10123.

42. McNally A, Anderson JM. Beta-1 and Beta-2 Integrins Mediate Adhesion

durin Macrophage Fusion and Multinucleated Foreign Body Giant Cell

Formation. American Journal of Pathology 2002;160(2):621-630.

43. Keselowsky B, Collard D, Garcia A. Surface chemistry modulates focal

adhesion composition and signaling through changes in integrin binding.

Biomaterials 2004;25:5947-5954.

44. Keselowsky B, Collard D, Garcia A. Integrin binding specificity regulates

biomaterial surface chemistry effects on cell differentiation. Proceedings of

the National Academy of Science 2005;102(17):5953-5957.

- 239 - Appendix I

45. Tanahaski M, Matsuda T. Surface functional group dependence on apatite

formation on self-assembled monolayers ina simulated body fluid. Journal

of Biomedical Materials Research 1997;34:314-315.

46. Sayre C, Collard D. Electrooxidative deposition of polypyrrole and

polyaniline on self-assembled monolayer modified electrodes. Langmuir

1997;13:714-722.

47. Wang MS, Palmer LB, Schwartz JD, Razatos A. Evaluating protein

attraction and adhesion to biomaterials with the atomic force microscope.

Langmuir 2004;20(7753-7759).

48. Shyue JJ, de Guire MR. Acid-base properties and zeta potentials of self-

assembled monolayers obtained via in situ transformations. Langmuir

2004;20:8693-8698.

49. Li L, Chen S, Jiang S. Molecular-scale mixed alkanethiol monolayers of

different terminal groups on Au(111) by low-current scanning tunneling

microscopy. Langmuir 2003;19:3266-3271.

50. Li L, Chen S, Jiang S. Protein adsorption on alkanethiolate self-assembled

monolayers: nanoscale surface structural and chemical effects. Langmuir

2003;19:2974-2982.

51. Silver IA, Murrills RJ, Etherington DJ. Microelectrode studies on the acid

microenvironments beneath adherent macrophages and osteoclasts.

Experimental Cell Research 1988;175:266-276.

- 240 - Appendix I

52. Garg N, Carrasquillo-Molina E, Lee TR. Self-assembled monolayers

composed of aromatic thiols on gold: structural characterization and

thermal stability in solution. Langmuir 2002;18:2717-2726.

- 241 -