Molecular and Cellular Mechanisms of the Angiogenic Effect of Poly(Methacrylic Acid-Co-Methyl Methacrylate) Beads

Molecular and Cellular Mechanisms of the Angiogenic Effect of Poly(methacrylic acid-co-methyl methacrylate) Beads

Lindsay Elizabeth Fitzpatrick

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Institute of Biomaterials and Biomedical Engineering University of Toronto

Molecular and Cellular Mechanisms of the Angiogenic Effect of Poly(methacrylic acid-co-methyl methacrylate) Beads

Lindsay Elizabeth Fitzpatrick

Doctorate of Philosophy

Institute of Biomaterials and Biomedical Engineering University of Toronto

2012 Abstract

Poly(methacrylic acid -co- methyl methacrylate) beads were previously shown to have a therapeutic effect on wound closure through the promotion of angiogenesis. However, it was unclear how this polymer elicited its beneficial properties. The goal of this thesis was to characterize the host response to MAA beads by identifying molecules of interest involved in

MAA-mediated angiogenesis (in comparison to poly(methyl methacrylate) beads, PMMA).

Using a model of diabetic wound healing and a macrophage-like cell line (dTHP-1), eight molecules of interest were identified in the host response to MAA beads. Gene and/or protein expression analysis showed that MAA beads increased the expression of Shh, IL-1β, IL-6, TNF-

α and Spry2, but decreased the expression of CXCL10 and CXCL12, compared to PMMA and no beads. MAA beads also appeared to modulate the expression of OPN. In vivo, the global gene expression of OPN was increased in wounds treated with MAA beads, compared to PMMA and no beads. In contrast, dTHP-1 decreased OPN gene expression compared to PMMA and no beads, but expressed the same amount of secreted OPN, suggesting that the cells decreased the expression of the intracellular isoform of OPN. Interestingly, MAA beads had no effect on the expression of pro-angiogenic growth factors VEGF, bFGF and PDGF-B in vivo or in vitro,

suggesting that MAA beads do not induce angiogenesis by simply increasing the expression of pro-angiogenic factors, but use more subtle mechanisms. It was hypothesized that these mechanisms may involve modulation of toll-like receptor signaling in macrophages interacting with the protein layer adsorbed on to MAA beads, in a manner distinct from PMMA beads and no beads.

Taken together, the results suggest that MAA beads promote angiogenesis through increased expression of Shh, decreased expression of CXCL10 and modulation of the expression of OPN, but not through increased expression of typical pro-angiogenic growth factors. The resulting vessel-rich “alternative foreign body reaction” has exciting clinical implications as the polymer itself was found to exert a therapeutic effect in the absence of bioactive components or transplanted cells. Understanding the mechanism could lead to new applications for this material and others designed on similar principles.

iii

Acknowledgments

I thank my supervisor, Prof. Michael Sefton, for the guidance, support and encouragement he has provided me over the past six years. His patience and steady encouragement was very much appreciated and it has been a privilege to have his mentorship. I thank my committee members, Yu-Ling Cheng and Dan Dumont, for their time, advice and encouragement. I also thank Brad Saville, Milica Radisic and Peter Zandstra for their time and advice as part of my supervisory and PhD transfer exam committees. I would also like to thank Heather Sheardown, from McMaster University, for the research opportunities she gave me as an undergraduate student and her mentorship over the past eleven years.

I thank Sasha, Josephine and Laura for their help and shared ideas while working together on the MAA project. I would also like to thank everyone in the Sefton group for their help over the years, Rohini, Mark, Brendan, Omar, Dean, Ema, and Derek. In particular, I would like to thank Chuen Lo for his invaluable skill and help with the wound healing model and venipuncture, and Carl Walkey and Derek Voice for repeatedly donating blood for my primary macrophage experiments.

I acknowledge funding from the Canadian Institutes of Health Research (CIHR), National Science and Engineering Research Council (NSERC), Ontario graduate scholarship (OGS), and the University of Toronto (Doctoral Completion Award).

Finally, I would like to thank my family and friends for their support. My heart-felt thanks go out to my parents, Norma & Wayne, for being such wonderful people whom I will always look up to, and for always being there to support and encourage me. Thank you to Erin & Mike, Jamie, Jaclene & Max, and John, Maya & Kyle for brightening my days and the constant support; you are the best family one could want. Thank you Chris & Dave for all your support, and for being such great examples of living life to the fullest. Finally and most of all, I want to thank my husband Scott; going through graduate school with you has made every part of it better. Thank you for understanding the highs and lows of research, pushing me to succeed and keeping me balanced. Congratulations on completing you PhD.

Table of Contents

Abstract ...... ii

Acknowledgments ...... iv

Table of Contents ...... v

List of Tables ...... xi

List of Figures ...... xiii

List of Abbreviations ...... xxii

Chapter 1: Introduction ...... 1 1 Poly(methacrylic acid-co-methyl methacrylate) beads ...... 1 2 Scope of thesis ...... 2 3 Specific Aims ...... 2 4 Thesis overview ...... 3 5 References ...... 5

Chapter 2: Angiogenesis and scaffold vascularization ...... 6 1 Introduction ...... 6 1.1 The native vascular network ...... 6 1.2 Angiogenesis: the formation of new blood vessels ...... 6 2 Vascularization of biomaterial scaffolds ...... 10 2.1 Growth factor delivery ...... 12 2.1.1 Single growth factor delivery ...... 12 2.1.2 Multiple growth factor delivery ...... 14 2.2 Cell delivery and transplantation ...... 15 2.2.1 Scaffolds seeded with cells ...... 15 2.2.2 Modular tissue engineering ...... 18 2.2.3 Genetically Modified Growth Factor-Producing Cells ...... 20 2.2.4 Scaffold-free cell delivery approaches ...... 21 2.3 In situ vascularization with endogenous cells ...... 23 2.4 Scaffold prevascularization ...... 25 2.5 Decellularized scaffolds ...... 27

2.6 Angiogenic biomaterials ...... 29 2.7 Microfabrication methods ...... 31 3 Limitations and the road ahead ...... 34 4 Summary ...... 35 5 References ...... 36

Chapter 3: Wound healing, diabetes and the foreign body reaction ...... 45 1 Cutaneous wound healing ...... 45 2 Wound healing and diabetes ...... 50 3 The foreign body reaction ...... 51 4 Macrophage activation/polarization and its role in wound healing and foreign body reaction ...... 56 5 Summary ...... 58 6 References ...... 59

Chapter 4: The expression of sonic hedgehog in diabetic wounds following treatment with poly(methacrylic acid-co-methyl methacrylate)beads ...... 68

Abstract ...... 68 1 Introduction ...... 69 2 Materials and methods ...... 69 2.1 Bead preparation ...... 69 2.2 Wound assay ...... 70 2.3 Histology and immunohistochemistry ...... 71 2.4 Vessel counts ...... 71 2.5 Total RNA preparation ...... 71 2.6 Quantitative real-time PCR ...... 72 2.7 Statistical analysis ...... 72 3 Results ...... 73 3.1 Effect of MAA beads on vessel density ...... 73 3.2 Gene expression analysis ...... 74 3.3 Effect of MAA beads on sonic hedgehog expression ...... 76 3.4 Effect of MAA beads on cytokine expression ...... 78 3.5 Effect of MAA beads on genes involved in regulating angiogenesis ...... 81 4 Discussion ...... 81

4.1 The effect of MAA beads on the vascularization of small, excisional cutaneous wounds in diabetic mice ...... 83 4.2 The role of sonic hedgehog in wound healing ...... 84 4.3 The effect of MAA beads on the expression of cytokines and growth factors involved with wound healing and angiogenesis ...... 85 4.4 Insight into the mechanism of MAA-mediated angiogenesis ...... 88 5 Conclusion ...... 90 6 Acknowledgments ...... 90 7 References ...... 90

Chapter 5: On the mechanism of poly(methacrylic acid-co-methyl methacrylate)- induced angiogenesis: gene expression analysis of dTHP-1 cells ...... 96

Abstract ...... 96 1 Introduction ...... 97 2 Materials and methods ...... 98 2.1 Bead preparation ...... 98 2.2 Cell culture ...... 98 2.3 Bead treatment of dTHP-1 and HUVEC cells ...... 99 2.4 RNA isolation and cDNA synthesis ...... 99 2.5 DNA microarray ...... 100 2.6 Quantitative real-time PCR ...... 100 2.7 HUVEC migration assay ...... 102 2.8 Statistical analysis ...... 102 3 Results ...... 103 3.1 Effect of the MAA beads on dTHP-1 mRNA expression ...... 103 3.2 The effect of MAA beads on osteopontin expression in HUVEC ...... 109 3.3 The effect of MAA-treated dTHP-1 conditioned medium on HUVEC migration ...... 109 4 Discussion ...... 111 4.1 Gene expression ...... 113 4.2 Osteopontin, IL-1β, IL-6 and TNF-α ...... 114 4.3 HUVEC migration ...... 115 4.4 The biomaterial problem ...... 117 5 Conclusions ...... 118 6 Acknowledgements ...... 118

vii

7 References ...... 118

Chapter 6: dTHP-1 expression of CXCL10 and CXCL12 following treatment with poly(methacrylic acid –co– methylmethacrylate) beads ...... 123 1 Introduction ...... 123 2 Materials and methods ...... 124 2.1 Bead preparation ...... 124 2.2 Cell culture ...... 125 2.3 Bead treatment of dTHP-1 cells ...... 125 2.4 Immunoassays ...... 126 2.5 HUVEC culture in dTHP-1 conditioned medium ...... 126 2.6 RNA isolation ...... 126 2.7 Quantitative real-time PCR ...... 127 2.8 Statistical analysis ...... 129 3 Results ...... 129 3.1 dTHP-1 secretion profile ...... 129 3.2 Effect of MAA beads on dTHP-1 protein secretion ...... 130 3.3 Effect of MAA beads on dTHP-1 protein secretion – macrophage phenotype ...... 134 3.4 Effect of MAA-treated dTHP-1 conditioned medium on HUVEC gene expression .. 135 4 Discussion ...... 136 4.1 dTHP-1 without bead treatment ...... 138 4.2 The effect of MAA beads on dTHP-1 protein expression ...... 139 4.3 Effect of MAA beads on macrophage phenotype ...... 141 4.4 Effect of MAA-treated dTHP-1 conditioned medium on HUVEC ...... 142 5 Future Directions ...... 143 5.1 Effect of MAA beads on intracellular OPN ...... 143 5.2 Effect of MAA-reduced CXCL12 concentration on HUVEC migration ...... 144 6 Conclusions ...... 144 7 References ...... 145 8 Supplemental data ...... 150

Chapter 7: Summary ...... 152 1 Summary of results ...... 152 2 Toll-like receptors ...... 154 3 Hypothesis for host response to MAA beads and MAA-mediated angiogenesis ... 158

viii

3.1 Hypothesis ...... 158 3.2 Possible role of intracellular OPN ...... 159 3.3 Limitations ...... 160 4 Future directions ...... 162 5 Clinical relevance ...... 164 6 Summary ...... 164 7 References ...... 165

Chapter 8: Conclusions ...... 170

Appendix I: Microarray analysis of MAA-treated diabetic wounds ...... 173 1 Introduction ...... 173 2 Materials and methods ...... 173 2.1 Animal procurement ...... 173 2.2 Bead preparation ...... 173 2.3 Wounding assay ...... 174 2.4 Total RNA preparation ...... 174 2.5 Microarray analysis ...... 174 2.6 Data Analysis ...... 175 2.7 Cluster analysis and pathway analysis (gene ontology) ...... 175 3 Results ...... 175 3.1 Sample preparation ...... 175 3.2 Microarray data analysis ...... 177 3.2.1 Gene lists selected using multiple-test corrected false-discover rates ...... 177 3.2.2 Gene lists selected using unadjusted p-values ...... 179 4 Summary ...... 183 5 References ...... 183 6 Supplemental Tables ...... 188

Appendix II: Effectiveness factor and diffusion limitations in collagen gel modules containing HepG2 cells ...... 216

Abstract ...... 216 1 Introduction ...... 217 2 Theoretical analysis ...... 218 2.1 Parameters ...... 219

3 Materials and methods ...... 220 3.1 Cell culture ...... 220 3.2 Module fabrication ...... 220 3.3 Western blotting ...... 221 3.4 Alamar blue reduction assay ...... 222 3.5 Human albumin ELISA ...... 222 3.6 Histology ...... 222 3.7 Statistics ...... 223 4 Results ...... 223 4.1 Theoretical analysis ...... 223 4.2 Module fabrication, size and effectiveness factors (based on seeding density) ...... 224 4.3 Effect of module diameter on cell density, metabolism and albumin ...... 225 4.4 Cell distribution, morphology and viability within modules ...... 227 4.5 Effectiveness factor based on actual cell density ...... 230 5 Discussion ...... 230 5.1 Experimental protocol ...... 230 5.2 Effect of module diameter ...... 232 5.3 Theoretical model ...... 233 6 Conclusions ...... 234 7 Acknowledgements ...... 234 8 References ...... 235

List of Tables

Chapter 4: The expression of sonic hedgehog in diabetic wounds following treatment with poly(methacrylic acid-co-methyl methacrylate)beads

Table 4-1. Quantitative real-time PCR primer sequences...... 73

Table 4-2. Mouse age, weight and blood glucose levels at time of surgery...... 74

Chapter 5: On the mechanism of poly(methacrylic acid-co-methyl methacrylate)- induced angiogenesis: gene expression analysis of dTHP-1 cells

Table 5-1. Primer sequences for real-time PCR...... 101

Table 5-2. Mean log2NRQ values and ANOVA for real-time PCR analysis of gene expression in dTHP-1 cells treated with no bead, PMMA beads or MAA beads for 24, 48 and 96 h. p < 0.05 (bolded) were considered significant, n=4...... 105

Chapter 6: dTHP-1 expression of CXCL10 and CXCL12 following treatment with poly(methacrylic acid –co– methylmethacrylate) beads

Table 6-1. Primer sequences for genes involved in endothelial activation and angiogenesis. ... 128

Table 6-S1. Concentration of analytes in dTHP-1 conditioned medium at 24 and 96 h (pg/ml, unless otherwise noted) ...... 150

Chapter 7: Summary

Appendix I: Microarray analysis of MAA-treated diabetic wounds

Table AI-1. Gene list of interest for "MAA/No bead" contrast, using fold-change > ± 1.5 and FDR < 0.05 as selection criteria...... 177

Table AI- 2. Gene list of interest for "MAA/PMMA" contrast, using fold-change > ± 1.5 and FDR < 0.05 as selection criteria...... 178

Table AI- 3. Gene list of interest for "MAA/PMMA" contrast, using fold-change > ± 1.5 and FDR < 0.05 as selection criteria...... 178

Table AI-4. Day 4 genes of interest. Genes of interest were selected from the lists of genes differentially regulated in MAA treated tissue samples (p < 0.05, FC > 1.5)...... 180

Table AI-S1. List of genes for Day 4, MAA/No bead (FC > 1.5 and p-value < 0.05)...... 188

Table AI-S2. List of genes for Day 4, MAA/PMMA (FC > 1.5 and p-value < 0.05)...... 194

Table AI-S3. List of genes for Day 4, MAA/PMMA (FC > 1.5 and p-value < 0.05)...... 197

Table AI-S4. List of genes for Day 7, MAA/No bead (FC > 1.5 and p-value < 0.05)...... 201

Table AI-S5. List of genes for Day 7, MAA/PMMA (FC > 1.5 and p-value < 0.05)...... 204

Table AI-S6. List of genes for Day 7, PMMA/No bead (FC > 1.5 and p-value < 0.05)...... 205

Table AI-S7. Gene lists corresponding to Day 4 (Fig AI-3, left panel) Venn diagram sections. Lists were generated by GeneVenn (http://genevenn.sourceforge.net/index.htm)...... 209

Table AI-S8. Gene lists corresponding to Day 4 (Fig AI-3, right panel) Venn diagram sections. Lists were generated by GeneVenn (http://genevenn.sourceforge.net/index.htm)...... 213

Appendix II: Effectiveness factor and diffusion limitations in collagen gel modules containing HepG2 cells

Table AII-1. Effectiveness factors calculated from experimental cell densities and diameters . 230

xii

List of Figures

Chapter 2: Angiogenesis and scaffold vascularization

Figure 2-1. Structure of nascent and mature blood vessels. (a) EC alone assembled into tube-like structures (nascent blood vessel). (b) Capillaries are the smallest type of blood vessels (only 5 - 10 µm in diameter), and are composed of a single layer of EC supported by a basement membrane of matrix proteins, and an incomplete covering of pericytes that stabilize the vessel. (c) Compared to capillaries, arterioles and venules are invested with a higher number of mural cells. Arterioles have a thinner media layer than arteries, and regulate the blood flow between the arteries and the capillaries. Venules (and veins) have thin walls and fewer SMC, relative to arteries. (d) Larger blood vessels have three specialized layers: intima (containing EC), media (containing SMC) and adventitia (containing fibroblasts). Arteries have a thick media composed mainly of SMC, which provide mechanical stability to the vessel wall, and allow the vessels to vasodilate and vasoconstrict in order to regulate blood flow and pressure. Reprinted by permission from Macmillan Publishers Ltd: Nature Medicine [6], Copyright 2003...... 7

Figure 2-2. Molecular basis of vessel branching, and normal vs. abnormal vessel formation. (a) Vessel branching is tightly regulated by a variety of key factors. The principal molecular signals are depicted for each of the consecutive steps of a branching process. (b) In normal angiogenesis, nascent tube-like structures become invested with mural cells and form a mature, organized and stable vascular network. (c) In abnormal angiogenesis seen in established tumours, the newly formed blood vessels are generally abnormal in structure and function. Note: Blood vessel networks formed in tissue engineered constructs are often times leaky, tortuous and unstable. Adapted by permission from Macmillan Publishers Ltd: (a) Nature [15], Copyright 2011, and (b, c) Nature Reviews Drug Discovery [20], Copyright 2011...... 9

Figure 2-3. Scaffold vascularization strategies. (a) Delivery of single or multiple angiogenic GF to stimulate the ingrowth of host vasculature into the scaffold. (b) Transplantation of EC (alone or in combination with support cells). The EC are expected to re-assemble in vivo, participate in blood vessel formation, and connect to the host vasculature. Support cells stabilize the nascent capillary-like structures. (c) Bio-functionalized materials are used to mobilize and capture endogenous cells at the implant site and stimulate de novo blood vessel formation. (d) In vivo prevascularization strategies use the host body as a bioreactor to vascularize the scaffold, prior to relocating the construct to its target implant site. In the case of arteriovenous loops, the prevascularization step takes place in a chamber enclosing an arteriovenous graft. (e) Decellularized scaffolds provide a natural template to recreate the architecture of the vasculature network. After decellularization, organs are reseeded with EC to repopulate the vascular tree and with functional cells to create a functional tissue engineered construct. (f) Upon implantation, some biomaterials induce angiogenesis by themselves, without the addition of any cells or GF. (g) Using microfabrication techniques, predefined, hierarchical pseudo-vasculatures are created (although mostly for in vitro applications thus far)...... 11

Figure 2-4. Modular tissue engineering. Sub-millimeter sized collagen gels containing therapeutic cells are coated with EC and packed together into a larger structure. The network of void spaces formed in between the endothelialized modules is amenable to perfusion...... 19

xiii

Figure 2-5. Cell sheet stacking method. (a) Cells (myoblasts) are cultured on a temperature- responsive culture dish. Once confluent, they are harvested as a cell sheet using a hydrogel- coated plunger and overlaid on top of a second cell sheet. After allowing the two cell sheets to attach, the plunger is lifted and the double-layered construct is placed on top of a third cell sheet. The procedure is repeated several times until a multi-layered construct is formed. (b) To create a vascularized construct, layers of sparsely seeded EC are inserted in between the myoblast cell sheets, using a similar stacking method. Reprinted from [66], Copyright 2010, with permission from Elsevier...... 22

Figure 2-6. Microfabrication of capillary networks. (a) A silicon micromold coated with sucrose was first used to transfer a capillary network pattern to a PGS layer. (b) The patterned PGS layer was then placed on top of a flat PGS layer and the two were bonded together to close the opened capillary channels in the patterned PGS. (c) The PGS microchannels were seeded with HUVEC and a confluent layer of EC was observed in some sections of the capillaries within the first 14 days of culture. Reprinted with permission from[111], Copyright 2005, Mary Ann Liebert, Inc...... 33

Chapter 3: Wound healing, diabetes and the foreign body reaction

Figure 3-1. The inflammatory phase of wound healing is characterized by an influx of neutrophils and macrophages into the hemostatic fibrin clot that fills the wound. Neutrophils phagocytize wound debris (necrotic cells, bacteria) then undergo apoptosis and are cleared by macrophages or are extruded in the eschar (non-viable fibrin clot). Macrophages phagocytize wound debris and secrete cytokines, chemokines and growth factors that propagate inflammation and direct the subsequent tissue formation phase. Reproduced with permission from [4], Copyright 1999, Massachusetts Medical Society...... 47

Figure 3-2. As the inflammation phase resolves, tissue formation begins. Fibroblasts migrate in the wound bed and begin to secrete a collagen-rich extracellular matrix that forms the granulation tissue. Ingrowth of blood vessels (angiogenesis) supports this process. Epidermal keratinocytes proliferate and migrate along the wound surface, dissecting the fibrin-rich clot from the underlying granulation tissue. Contractile cells, called myofibroblasts, remodel the collagen-rich matrix and contract the wound bed. Reproduced with permission from [4], Copyright 1999, Massachusetts Medical Society...... 48

Figure 3-3. Foreign body reaction to biomaterial implants occurs through a sequence of events involving inflammation and wound healing responses. Upon implantation of the biomaterial, an inflammatory response is initiated, which recruits inflammatory cells to the biomaterial and surrounding tissue. Infiltrating monocytes differentiate into macrophages and adhere to the biomaterial surface via the adsorbed protein layer. Macrophages fuse to form multinucleated foreign body giant cells (FBGC) in an attempt to degrade and clear the material from the body. In the surrounding tissue, wound healing progresses through the typical phases of inflammation and granulation tissue formation. In most cases, the biomaterial cannot be degraded and cleared by FBGC, and the implant is encapsulated in a fibrous capsule. Reprinted from [66] Copyright 2008, with permission from Elsevier...... 52

xiv

Figure 3-4. Schematic of alternative complement activation pathway. Hydrolysis of C3 gives rise to C3b, formation of the alternative C3 and C5 convertases and eventual formation of the terminal complement complex (TCC)...... 53

Figure 3-5. M1 (classical) and M2 (alternative) macrophage phenotypes. Distinct macrophage phenotypes can arise in response to specific inducers. M1 macrophages are induced by INF-γ, LPS or TNF-α, while M2 macrophages are divided into three sub-phenotypes, M2a, M2b and M2c based on the route of activation. Abbreviations: DTH, delayed-type hypersensitivity; IC, immune complexes; MR, mannose receptor; PTX3, the long pentraxin PTX3; RNI, reactive nitrogen intermediates; ROI, reactive oxygen intermediates; SLAM, signaling lymphocytic activation molecule; SRs, scavenger receptors; TLR, Toll-like receptor[128]. Reprinted from [128] Copyright 2004, with permission from Elsevier...... 57

Chapter 4: The expression of sonic hedgehog in diabetic wounds following treatment with poly(methacrylic acid-co-methyl methacrylate)beads

Figure 4-1. Histology sections of day 7 wound tissue in diabetic animals treated with MAA beads, PMMA beads or no beads stained with H&E (A). High magnification images of granulation tissue stained with H&E (B) and CD31 (C). Scale bars represent 1 mm (A) and 100 µm (B, C)...... 75

Figure 4-2. Blood vessel density within the granulation tissue at post-operative day 7. Wounds treated with MAA beads had a significantly higher density of CD31+ vessel-like structures within the granulation tissue, compared to PMMA treated wounds and untreated wounds. *, p < 0.05...... 75

Figure 4-3. Sonic hedgehog gene expression in diabetic wounds treated with MAA beads. On post-operative day 4, Shh expression was increased approximately 5-fold in MAA-treated wounds compared to PMMA-treated and untreated wounds (p < 0.05). By day 7, the expression of Shh in all wounds was similar (FC < 1.5). PMMA did not affect the expression of Shh compared to no beads. Circles represent individual animals (n =4) and the bar represents the median. Dashed lines mark +1.5 and -1.5 fold change. *, p < 0.05...... 78

Figure 4-4. Gene expression of sonic hedgehog pathway proteins. On day 4, MAA-treated wounds increased (to a small extent) Gli1, Gli3, Ptch1 and Ptch2 expression compared to untreated wounds and increased all pathway-associated proteins compared to PMMA treated wounds. Shh transcription factor Gli3 had the largest fold increase in MAA wounds compared to PMMA (FC = 2.5, p = 0.052) and untreated (FC = 2.1, p = 0.157) wounds. On day 7, wounds treated with MAA, PMMA or no beads had similar mRNA expression for all the Shh pathway proteins, with the exception of Gli1, which was decreased 1.8-fold (p > 0.05) in MAA wounds compared to PMMA wounds. Circles represent individual animals (n =4) and the bar represents the median. Dashed lines mark +1.5 and -1.5 fold change. *, p < 0.05...... 79

Figure 4-5. Gene expression of IL-1β, TNF-α, OPN and IL-6 in diabetic wound granulation tissue. MAA beads increased the expression (FC > 1.5) of all four genes on day 7 compared to PMMA beads and no beads, although only the increases in IL-1β, TNF-α and OPN compared to PMMA were significant (p < 0.05). PMMA also decreased the expression of IL-1β and IL-6 compared to untreated on day 4 (p > 0.05) and increased the expression of IL-1β compared to

untreated wounds on day 7 (p > 0.05). Circles represent individual animals (n =4) and the bar represents the median. Dashed lines mark +1.5 and -1.5 fold change.*, p < 0.05...... 80

Figure 4-6. Gene expression of proteins with pro- or anti-angiogenic effects. MAA beads significantly increased (FC > 1.5, p < 0.05) the expression of Spry2 compared to PMMA and no beads on day 4, and compared to PMMA on day 7. Changes were also observed in bFGF, PDGF- B, CXCL10 and TSP-1, however these were not significant. Briefly, MAA beads increased the expression bFGF 1.5-fold compared to PMMA on day 4, and PDGF-B was 2.3-fold and 1.6-fold higher in MAA treated wounds than PMMA and no bead. CXCL10 expression was decreased by MAA beads compared to untreated (no bead, day 4 and 7) and PMMA-treated (day 4) wounds. Neither MAA nor PMMA beads affected the expression of VEGF compared to no bead. Conversely, MAA beads increased the expression TSP-1 compared to PMMA-treated and untreated wounds on day 4 and 7 (except TSP-1, MAA / no bead, day 4). PMMA treated wounds had slightly less (1.6-fold decrease) TSP-1 on day 4 than untreated wounds. Circles represent individual animals (n =4) and the bar represents the median. Dashed lines mark +1.5 and -1.5 fold change.*, p < 0.05...... 82

Chapter 5: On the mechanism of poly(methacrylic acid-co-methyl methacrylate)- induced angiogenesis: gene expression analysis of dTHP-1 cells

Figure 5-1. Volcano plot of microarray data comparing gene expression of dTHP-1 cells treated MAA or PMMA beads for 24 h to untreated cells. Data was filtered to remove genes with fold- change less than ± 1.25 or p-values greater than 0.05 [-log (0.05) = 1.3]. Only two genes, OPN (decreased 1.8-fold, p < 0.001) and IGFBP3 (increased 1.6-fold, p = 0.01), were expressed differently in MAA-treated dTHP-1 than in untreated cells. PMMA-treated and untreated cells had similar gene expression profiles...... 104

Figure 5-3. qPCR analysis of IL-6 mRNA expression in dTHP-1 cells cultured with MAA beads, PMMA beads or no bead. MAA treatment of dTHP-1 cells resulted in a three-fold increase in IL- 6 expression (relative to no bead or MMA beads) over the first 48 h. By 96 h, the differential effect of MAA treatment on dTHP-1 IL-6 mRNA expression had disappeared. PMMA beads had no effect on IL-6 expression compared to no bead. Significant changes in expression,*, have a relative ratio less than 0.67 or greater than 1.5 with corresponding p-value less than 0.05. n = 4, +/- SE; p-values are provided in Table 5-2 for NRQ values...... 107

Figure 5-4. qPCR analysis of TNF-α mRNA expression in dTHP-1 cells treated with MAA beads, PMMA beads or no bead. MAA-treated dTHP-1 cells have increased expression of TNF- α mRNA at 96 h of treatment compared to PMMA-treated dTHP-1 and untreated dTHP-1 cells. PMMA beads increased the expression of TNF-α also, but not to the same extent as MAA beads. Significant changes in expression,*, have a relative ratio less than 0.67 or greater than 1.5 with

xvi

corresponding p-value less than 0.05. n = 4, +/- SE; p-values are provided in Table 5-2 for NRQ values...... 108

Figure 5-5. qPCR analysis of IL-1β mRNA expression in dTHP-1 cells receiving different bead treatments. Expression of IL-1β was similar for all treatments during the first 48 h, but was significantly increased at 96 h in MAA treated dTHP-1 cells. IL-1β expression was also increased by PMMA beads at 96 h compared to no bead, however, this increase was approximately half of that for MAA. Significant changes in expression,*, have a relative ratio less than 0.67 or greater than 1.5 with corresponding p-value less than 0.05. n = 4, +/- SE; p- values are provided in Table 5-2 for NRQ values...... 109

Figure 5-6. qPCR analysis of the mRNA expression of six pro-angiogenic genes in dTHP-1 cells treated with MAA beads, PMMA beads or no bead. MAA treatment did not affect the expression of bFGF, CXCL12, HIF-1α or TGF-β over 96 h, compared to PMMA treatment or no treatment. PDGF-B and VEGF were slightly elevated for both MAA and PMMA beads at 96 h but unaffected at earlier times. Significant changes in expression,*, have a relative ratio less than 0.67 or greater than 1.5 with corresponding p-value less than 0.05. n = 4, +/- SE; p-values are provided in Table 5-2 for NRQ values...... 110

Figure 5-7. qPCR analysis of MAA bead effect on OPN expression in HUVEC at 24 and 91 h. OPN expression was similar for all treatments at 24 h. At 91 h, MAA beads increased the expression of OPN 3.5 fold and 4.0 fold compared to no bead and PMMA beads, respectively. PMMA did not significantly effect the expression of OPN at either time points. Significant changes in expression,*, have a relative ratio less than 0.67 or greater than 1.5 with corresponding p-value less than 0.05. n = 4, +/- SE; p-values are provided in Table 3...... 111

Figure 5-8. Effect of culture medium conditioned by dTHP-1 cells treated with MAA, PMMA or no bead for 96 h on HUVEC migration in Transwell migration assays. The upper and lower chambers of the transwell system were filled with different conditioned medium (as depicted) and the HUVEC migrated from the upper chamber to the lower chamber (e.g. from “no bead” medium to “MAA” medium). At 96 h, HUVEC had significantly reduced migration toward MAA conditioned medium (p < 0.001), compared to that which migrated to PMMA- or no bead- conditioned medium. At 24 and 48 h (not shown) the source of the conditioned medium had no observable effect (p > 0.062) on HUVEC migration. For *, p < 0.001 for either MAA (lower) compared to any of the other pairs...... 112

Chapter 6: dTHP-1 expression of CXCL10 and CXCL12 following treatment with poly(methacrylic acid –co– methylmethacrylate) beads

Figure 6-1. Concentration of eleven secreted proteins in dTHP-1 supernatant (without beads), in order of increasing concentration. OPN had the highest concentration at 24 h and 96 h (~ 2000 ng/ml). At 24 h, the next most abundant proteins were TNF-α and IL-1β (> 1600 pg/ml), followed by VEGF-A, CXCL10, CXCL12 and IL-6 (in descending order, 300 – 1000 pg/ml). By 96 h, VEGF-A and CXCL10 had increased in concentration to greater than 900 pg/ml, while the concentration of TNF-α and IL-6 decreased significantly (< 50 pg/ml). IL-1β decreased and CXCL12 increased, relative to 24 h concentrations (500 – 600 pg/ml). The least abundant proteins were IL-10, IL-12, INF-γ and bFGF for both time points (< 20 pg/ml). The expression of

xvii

PDGF-BB was below the detection limit. Results presented as mean ± SE, n = 4. RPMI/FBS medium (no cells) had concentrations below the detection limit for all analytes...... 131

Figure 6-2. Concentration of CXCL10 in conditioned medium from dTHP-1 cultured with MAA, PMMA or no beads. At both time points, culture with MAA beads significantly decreased the expression of CXCL10, compared to PMMA beads and no beads (p < 0.05). PMMA beads had no effect (p > 0.05) on CXCL10 concentration in the supernatant, compared to no beads. (*, p < 0.05) ...... 132

Figure 6-3. CXCL12 concentration in the supernatant of dTHP-1 cells cultured with MAA, PMMA or no beads for 24 – 96 h. At 24 h, bead treatment had no effect on the concentration of CXCL12 in the dTHP-1 supernatant (p > 0.05). At 96 h, MAA treated dTHP-1 supernatant had decreased concentration of CXCL12, compared to PMMA and no beads (p < 0.05). PMMA beads also decreased the concentration of CXCL12 compared to no bead (p < 0.05), but not to the same extent as MAA beads. (*, p < 0.05)...... 132

Figure 6-4. Secreted osteopontin expression in dTHP-1 treated with MAA, PMMA or no beads. Secreted OPN was detected in the supernatant of dTHP- 1 cells at 24 and 96 h, using an ELISA. Bead treatment had no effect on OPN concentration in dTHP-1 supernatant (p > 0.05), and the concentration did not change over time (p > 0.05, 24 h vs 96 h)...... 133

Figure 6-5. The concentration of a) IL-1β, b) IL-6 and c) TNF-α in the cell culture medium of dTHP-1 treated with MAA, PMMA or no beads. At 24 h, bead treatment had no effect on the concentration of IL-1β, IL-6 or TNF-α (p > 0.05). However, the concentration of these cytokines in the MAA-treated dTHP-1 conditioned medium was higher (p < 0.05) at 96 h, compared to PMMA-treated and untreated dTHP-1 conditioned medium. No difference between the concentration of IL-1β, IL-6 or TNF-α in PMMA-treated and untreated concentration medium was observed (p > 0.05). (*, p < 0.05, n = 4)...... 133

Figure 6-6. The concentration of a) VEGF-A and b) bFGF in the supernatant of dTHP-1 beads cultured with MAA, PMMA or no beads for 24 and 96 h. Generally, dTHP-1 secreted more VEGF-A than bFGF at both time points. Exposure to MAA beads caused a small, but significant decrease in the concentration of VEGF-A and bFGF in the supernatant at 24 h, compared to dTHP-1 cells treated with PMMA or no beads (p < 0.05). At 96 h, dTHP-1 with MAA or PMMA beads had slightly higher concentrations of VEGF-A, compared to no bead (p < 0.05). The concentration of bFGF was similar among all treatments at 96 h. (*, p < 0.05)...... 134

Figure 6-7. The concentration of a) IL-10 and b) IL-12 in the supernatant of dTHP-1 cells treated with MAA, PMMA or no beads. MAA beads secreted less IL-10 and IL-12(p70) into the supernatant than untreated cells at both time points (p < 0.05). The concentration of both cytokines in the PMMA/dTHP-1 supernatant was lower than untreated dTHP-1 supernatant at 96 h, but higher than MAA treated dTHP-1 supernatant (p < 0.05)...... 135

Figure 6-8. The concentration of INF-γ in bead-treated dTHP-1 supernatant. dTHP-1 secreted low amounts of INF-γ, relative to other analytes, and bead treatment affected the concentration of INF-γ at both time points. MAA beads had the lowest concentration, while untreated cells had the highest (p < 0.05). (*, p < 0.05)...... 136

xviii

Figure 6-9. Effect of 96 h untreated dTHP-1 conditioned medium on HUVEC. Incubation (for 6 h) in dTHP-1 conditioned medium increased HUVEC expression of multiple genes involved in endothelial cell activation and angiogenesis. Many genes (*, ratio > 1.5 or < 0.67, p < 0.05) increased in expression compared to HUVEC cultured in fresh RPMI 1640/FBS medium, with transcription factor ETS1 and VEGF (i.e. VEGF-A) having the largest increases in expression. VEGF-C mRNA was decreased (p < 0.05) while expression of MMP1, integrin β5 and β3, and VEGF-B were not changed by the conditioned medium. MMP-9 expression did increase (ratio = 2.1), but was not significant (p > 0.05). All ratios represent the HUVEC gene expression in dTHP-1 conditioned medium relative to fresh RPMI/FBS, using GAPDH as an endogenous reference gene...... 137

Figure 6-10. HUVEC gene expression after 6 h incubation in medium conditioned previously for 96 h by dTHP-1 treated with MAA or PMMA beads, relative to untreated dTHP-1 conditioned medium. Of the 19 genes whose expression was analyzed, only two were significantly changed by MAA treated dTHP-1 conditioned medium. The expression of CXCR4 was significantly decreased by MAA/dTHP-1 supernatant (ratio = 0.55, p = 0.001), while VCAM1 was significantly increased by MAA/dTHP-1 (ratio = 1.8, p = 0.008). PMMA treated dTHP-1 conditioned medium had no effect on HUVEC gene expression relative to untreated dTHP-1 conditioned medium. (*, ratio > 1.5 or < 0.67, p < 0.05)...... 138

Figure 6-S1. Gene expression in dTHP-1 cells treated with MAA, PMMA or no beads for 96 h. As expected, MAA beads increased the expression of IL-1β, IL-6 and TNF-α, and decreased the expression of OPN in dTHP-1, relative to no beads. However, PMMA beads elicited a stronger effect than previously reported, increasing the expression of IL-1β and TNF-α almost to the level of MAA beads. Significant changes in expression,*, have a relative ratio less than 0.67 or greater than 1.5 with corresponding p-value less than 0.05. n = 4, ± SE...... 151

Chapter 7: Summary

Figure 7-1. Toll-like receptor signaling pathway. Red stars highlight molecules that were identified in host-response to MAA beads. Molecules in blue were manually added to the KEGG pathway schematic as TLR target genes (Adapted from KEGG[11,12])...... 155

Figure 7-2. Proposed model of MAA-induced TLR signaling. TLR interaction with adsorbed protein on MAA surface modulates TLR signaling, such that NF-κB target genes (IL-1β, IL-6, TNF-α, Shh) are increased, and the expression of IRF and AP-1 target genes (CXCL10, OPN) are decreased, compared to PMMA surfaces or untreated wound tissue. NF-κB signaling can occur through the MyD88-dependent and TRIF-dependent pathways, although the TRIF pathway is delayed relative to MyD88...... 160

Appendix I: Microarray analysis of MAA-treated diabetic wounds

Figure AI-1. Grouping scheme for microarrays. (a) Day 7 samples were pooled such that the wound tissue from three mice of a given treatment was pooled into one RNA sample (n=3). (b) Day 4 samples were not pooled, and the number of samples per treatment was increase to n=4...... 176

xix

Figure AI-2. Overview of gene regulation in treated and untreated diabetic wounds. Genes with p-values less than 0.05 and fold-changes greater than 1.5 from ANOVAs contrasting the three wound treatments...... 179

Figure AI-3. Venn diagrams showing the number of transcripts that are differentially expressed in response to MAA beads and PMMA beads. Refer to Supplemental Tables AI-S7 and AI-S8 for the gene list corresponding to each region in the Venn diagrams. Diagrams created using GeneVenn (http://genevenn.sourceforge.net/index.htm)[6]...... 180

Appendix II: Effectiveness factor and diffusion limitations in collagen gel modules containing HepG2 cells

Figure AII-1. Modular tissue engineering. Small collagen rods containing embedded functional cells and coated with endothelial cells are randomly packed to form a ‘packed bed’ construct. Interstitial spaces between modules allow the construct to be perfused ...... 218

Figure AII-2. Theoretical analysis of Thiele modulus and effectiveness factor for tissue engineering modules. Modules with very small diameters (<0.40 mm) or very low cell densities (1 × 106 cells/ml) are not expected to experience significant mass transfer limitations (η> 0.9). However, as cell density increases (1 × 10−7 cells/ml) the negative effect of module diameter on effectiveness factor becomes more pronounced...... 223

Figure AII-3. Module fabrication and contraction. Module contraction occurred during the 3 days following HUVEC-C seeding. (a) Following contraction (day 3), modules made with 1.40 mm i.d. tubing were significantly larger in diameter and length than modules made with 0.76 mm i.d. tubing (p < 0.05). (b) Embedded HepG2 cells were uniformly distributed within modules at the time of fabrication and retained high viability. Scale bars: 250 µm. Green, live cells; red, dead cells ...... 224

Figure AII-4. Six module fabrication conditions were selected for experimental analysis. Modules made from either 0.76 mm i.d. tubing or 1.40 mm i.d. tubing were seeded with 2 × 106 - 1 × 107 cells/ml. Following module contraction, two module sets (*) were predicted to experience mass transfer restrictions (i.e. η< 0.9) ...... 225

Figure AII-5. (a) The number of cells per module was estimated from the amount of GAPDH per module, using western blots. No differences were found in the number of cells per module among large and small modules, regardless of cell number and time (p > 0.88) (b) Cell density was calculated from the cell number per module using the module volumes. Small-diameter modules had a significantly higher (p < 0.05) cell density compared to large-diameter modules, although these were not different from days 3 to 7. The results are average ± SD (n = 2) ...... 226

Figure AII-6. (a) Alamar blue (AB) reduction per module was higher for large modules (p < 0.05), even though, according to Figure AII-5a, there was a similar number of cells in small and large modules. (b) The large modules also had a higher AB reduction rate when normalized on a per-cell basis. The results are average ± SD, regardless of initial seeding density (n = 9) ...... 227

Figure AII-7. Albumin secretion from modules was measured using an enzyme-linked immunosorbent assay. No significant difference in secretion rate per cell was seen between large

and small modules on days 3 or 7. The results are average ± SD, regardless of initial seeding density (n = 16) ...... 227

Figure AII-8. Confocal microscopy images of small (left) and large (right) modules at day 3. At day 3, a large number of dead cells had formed within the core of the large modules (right panel), leaving only a thin rim (∼200 µm thick) of viable cells. Conversely, the small modules retained a uniform and high distribution of live cells. Green, live; red, dead ...... 228

Figure AII-9. Histology sections of modules stained with trichrome. Cells are distributed evenly on day 0 for both large and small modules. On day 3, the cells in the small modules had assembled into spheroids within the module. Some cells in the large modules had also formed spheroids; however, many cells did not aggregate and instead remained suspended in the collagen module as single cells. By day 7, many of the cells in the core of the large module appeared to have died, while the edges of the modules are densely populated with spheroids. The entire volume of the small modules are densely populated with HepG2 spheroids at day 7 and do not show the dead core seen in the large modules ...... 229

xxi

List of Abbreviations

adMSC adipose-derived mesenchymal stromal cells Ang angiopoietin ANOVA analysis of variance AP-1 activator protein 1 AVL arteriovenous loop BCP biphasic calcium phosphate bFGF basic fibroblast growth factor BMA poly(butyl methacrylate) BMA-MAA poly(butyl methacrylate –co– methacrylic acid) BMP-2 bone morphogenic protein 2 BREC bovine retinal microvascular endothelial cells BRP bovine retinal pericytes BSA bovine serum albumin CCL1 chemokine (C-C motif) ligand 1 CCL2 chemokine (C-C motif) ligand 2 CM carboxymethyl Cp crossing point CXCL10 C-X-C motif chemokine 10 CXCL12 C-X-C motif chemokine 12 (also called SDF-1) CXCR4 C-X-C chemokine receptor type 4 Cy3 cyanine 3 Cy5 cyanine 5 DAMP danger associated molecular patterns DAVID Database for annotation, visualization and integrated discovery DC dendritic cells DC-STAMP dendritic cell-specific transmembrane protein dCTP deoxycytidine triphosphate DEAE diethylaminethanol DLL-4 delta-like ligand 4 dTHP-1 differentiated THP-1 (macrophage-like cell line) EC endothelial cells ECM extracellular matrix EGDMA ethylene glycol dimethacrylate EGF epidermal growth factor ELISA enzyme-linked immunosorbent assay EPC endothelial progenitor cells ERK extracellular signal related kinase FAK focal adhesion kinase FBGC foreign body giant cells FBR foreign body reaction FBS fetal bovine serum FC fold change Fizz1 found in inflammatory zone 1

xxii

G-CSF granulocyte colony-stimulating factor GAG glycosaminoglycans GF growth factor GM-CSF granulocyte-macrophage colony stimulating factor GOI gene of interest H&E hematoxylin and eosin HAF human artery-derived fibroblasts HB-EGF heparin-binding EGF-like growth factor HBPA heparin-binding peptide amphiphiles hESdC human embryonic stem cell-derived cells HIF-1α hypoxia inducible factor 1 alpha HS heparan sulfate HSD (Tukey's) honestly significant difference HSGAG heparan sulfate-like glycosaminoglycans HUVEC human umbilical vein endothelial cells ICAM-1 inter-cellular adhesion molecule 1 IFR3 interferon regulatory factor 3 IGF-1 insulin-like growth factor 1 IGFBP-3 insulin growth factor binding protein 3 IL-10 interleukin 10 IL-12 interleukin 12 IL-13 interleukin 13 IL-1β interleukin 1 beta IL-23 interleukin 23 IL-4 interleukin 4 IL-6 interleukin 6 IL-8 interleukin 8 INF-γ interferon gamma iNOS inducible nitric oxide synthase KEGG Kyoto Encyclopedia of Genes and Genomes KGF keratinocyte growth factor LDPI Laser Doppler Perfusion Imaging LN-5 laminin 5 (also called laminin 332) LPS lipopolysaccharide M1 classical macrophage phenotype M2 alternative macrophage phenotype MA methacrylic acid MAA poly(methacrylic acid -co- methyl methacrylate) MAPK mitogen-activated protein kinases MCP-1 monocyte chemoattractant protein 1 MEMS MicroElectroMechanical Systems MIP-2 macrophage inflammatory protein 2 MMP matrix metalloproteinase mRNA messenger ribonucleic acid MyD88 myeloid differentiation factor 88 NF-κB nuclear factor kappa B NO nitric oxide NRQ normalized relative quantity

xxiii

OCI Ontario Cancer Institute OPN osteopontin PA peptide amphiphiles PAI-1 plasminogen activator inhibitor 1 PAMP pathogen-associated molecular pattern PBAE poly(β-amino esters) PBS phosphate buffered saline PCR polymerase chain reaction PDGF-B platelet-derived growth factor PDMS poly(dimethyl siloxane) PEG polyethylene glycol PEGDA polyethylene glycol diacrylate PGS poly(glycerol sebacate) pHEMA poly(2-hydroxyethyl methacrylate) pHEMA-co-MAA poly(2-hydroxyethyl methacrylate-co-methacrylic acid) PLG poly(lactide-co-glycolide) PLGA poly(lactic-glycolic acid) PLLA poly(L-lactic acid) PMA phorbol myristate acetate PMMA poly(methyl methacrylate) Ptch1 patched 1 Ptch2 patched 2 qPCR quantitative real-time polymerase chain reaction R relative expression ratio RAEC rat aortic endothelial cells RAGE receptor of advanced glycation end-products RNA ribonucleic acid ROS reactive oxygen species RQ relative quantity SCID severe combined immunodeficient SDF-1 stromal derived factor 1 (also called CXCL12) SDS Sequence Detection Systems SEM standard error of the mean Shh sonic hedgehog SMA smooth muscle actin SMC smooth muscle cells Smo smoothened Spry2 sprouty homolog 2 STAT1 signal transducer and activator of transcription 1 TCC terminal complement complex TCPS tissue culture treated polystyrene TGF-β transforming growth factor beta TLR toll-like receptor TNF-α tumor necrosis factor alpha tPA tissue-type plasminogen activator TRIF Toll/IL-1 receptor domain-containing adaptor inducing interferon B TSP-1 thrombospondin 1 TSP-2 thrombospondin 2

xxiv

uPA urokinase-type plasminogen activator VCAM-1 vascular cell adhesion molecule 1 VEGF vascular endothelial growth factor VEGF-PA VEGF-mimetic peptide amphiphiles VSMC vascular smooth muscle cells vWF von Willebrand factor Ym1/ECF eosinophil chemotactic factor

xxv 1

Chapter 1 Introduction 1 Poly(methacrylic acid-co-methyl methacrylate) beads

The modern field of biomaterials originated in the 1940’s and 1950’s when physicians began using common household and industrial materials (e.g. Teflon, poly(methyl methacrylate), stainless steel) to achieve innovative and ground breaking reparative medicine[1]. Since that time, the field has developed from an era when biomaterial inertness was considered ideal to the current aim of creating new biologically interactive materials that can integrate with host tissue and provide intrinsic beneficial effects in directing tissue repair and regeneration. This aim is being achieved through modification of synthetic biomaterials to include conjugated and adsorbed proteins, peptides or other extracellular matrix components, protease-cleavable sites, surfaces porosity and topography that induce tissue ingrowth and blood vessel formation (angiogenesis), and through the development of natural or decellularized matrices[2-8]. These biomaterial designs are aimed at minimizing or abrogating the host response to foreign materials, termed the foreign body reaction (FBR). The FBR is a cascade of processes that occur following implantation of a biomaterial, which ultimately results in the biomaterial being walled off from the body by encapsulation in a fibrous matrix[9]. Upon implantation, proteins adsorb to the surface of the biomaterial and macrophages, a phagocytic cell of the innate immune system, are recruited to the implant site and adhere to the adsorbed protein layer. Macrophages and multinucleated foreign body giant cells (FBGC), which form through fusion of macrophages, accumulate at the surface of the implant and direct the extent of the FBR and fibrous capsule formation. Biomaterial and device failures are often caused by the FBR, either through degradation of the biomaterial by macrophages and FBGC, or through the fibrous capsule formation that creates a barrier to the surrounding tissue and can interfere with the functioning of a device (e.g. an implanted glucose sensor)[9].

Small beads (~200 µm diameter) made of poly(methacrylic acid-co-methyl methacrylate) (MAA), a synthetic copolymer, have been shown to have beneficial effects in animal models. Incorporating MAA beads in skin graft models in healthy rats, and cutaneous wound healing models in diabetic (db/db) mice improved graft health and wound closure by increasing blood vessel density of the graft and wound bed, respectively[10,11]. In contrast, 100% poly(methyl

2 methacrylate) (PMMA) beads, did not have any effect on graft survival, wound closure or neovascularization compared to untreated animals. These results raised the question of how MAA beads elicit a positive biological response, when PMMA beads, like so many other biomaterials, have minimal or negative biological responses in vivo. Unlike recent biomaterials that have been carefully designed to have biological activity, the angiogenic host response to MAA beads was discovered serendipitously, when the beads were being used as a control material in other, unrelated animal studies. The biomaterial properties of MAA beads that are responsible for inducing an angiogenic outcome, and the mechanism through which the host cells respond to the MAA beads are unknown.

2 Scope of thesis

This study was intended to characterize the cellular response to MAA beads using a diabetic wound healing model and in vitro cell culture models. Specifically the focus was to identify signaling pathways, specific genes/proteins and/or cell types that may play a role in the host response to MAA beads and to develop hypotheses as to the mechanism of MAA-mediated angiogenesis for subsequent study (beyond this thesis). The intent was not to determine what material characteristics are critical in inducing the angiogenic response to the beads, or to characterize the blood protein-material surface interactions that dictate the downstream host response.

3 Specific Aims

The aim of the research was to characterize the host response to MAA beads and to identify molecules upon which hypotheses regarding the mechanism of MAA-mediated angiogenesis could be built.

Specific Aim 1: Identify genes whose expression was changed in MAA-treated wound tissue, compared to PMMA-treated or untreated wound tissue. Small excisional cutaneous wounds on diabetic mice were treated with MAA, PMMA or no beads, and the gene expression in the granulation tissue of the wounds was analyzed at day 4 and 7 using quantitative real-time PCR (qPCR) to identify genes of interest.

Specific Aim 2: Identify genes whose expression was changed in MAA-treated macrophage-like dTHP-1 cells in vitro, compared to PMMA treated cells and untreated cells. dTHP-1 cells were

3 cultured with MAA, PMMA or no beads, and the gene expression was analyzed at 24, 48 and 96 h using qPCR.

Specific Aim 3: Validate the expression of genes of interest identified in Specific Aims 1 and 2 in dTHP-1 cells through analysis of protein expression. Immunoassays were used to quantify the secreted protein expression of dTHP-1 treated with MAA, PMMA and no beads.

Specific Aim 4: Generate a hypothesis of how MAA beads mediate a beneficial, angiogenic response in vivo using the list of genes/protein generated in Specific Aims 1 – 3. A hypothesis was generated based on the list of genes/proteins of interest, bioinformatics resources and the current body of literature on wound healing, angiogenesis and the foreign body reaction.

4 Thesis overview

The first chapters review the current literature on angiogenesis and scaffold vascularization (Chapter 2) and wound healing, diabetic wound healing and the foreign body reaction (Chapter 3). Chapter 2 defines the current methods for inducing scaffold vascularization using various approaches, including growth factor delivery, cell transplantation, biomaterial properties and structure. Chapter 3 focuses on identifying the key cell types and molecules in the processes of wound healing and the foreign body reaction to biomaterials, the pathophysiology of diabetic wound healing and reviews macrophage activation, a central cell type in these processes.

Chapter 4 (Fitzpatrick et al. Biomaterials 2012;33(21):5297-5307) explores the in vivo response to MAA beads using a model of diabetic wound healing. Small, bilateral cutaneous wounds were treated with MAA, PMMA or no beads and the gene expression within the granulation tissue of the wounds was studied at day 4 and day 7. Histological analysis of the wound tissue showed that MAA beads increased the vessel density within the granulation tissue at day 7, consistent with previous studies. The most significant result was the increase in gene expression of Sonic hedgehog (Shh) in wounds treated with MAA beads at day 4, compared to wounds treated with PMMA beads and untreated wounds. Shh is a developmental morphogen reported to induce angiogenesis in adult tissue. Osteopontin, interleukin (IL)-1β, and tumor necrosis factor (TNF)-α were also slightly increased in MAA-treated wound tissue at day 7.

Chapter 5 (Fitzpatrick et al. Biomaterials 2011;32(34):8957-8967) focuses on the in vitro response of human macrophage-like dTHP-1 cells to MAA beads. Macrophages were premised to be a main effector of the MAA response, due to the major role this cell type plays in wound healing and the foreign body reaction. Similar to the in vivo results, dTHP-1 cells cultured with MAA beads expressed higher amounts of IL-1β, IL-6 and TNF-α, while pro-angiogenic genes were unaffected, compared to PMMA beads and no beads. However, the mRNA expression of OPN was significantly decreased in dTHP-1 cultured with MAA beads, while human umbilical vein endothelial cells (HUVEC) increased the expression of OPN, compared to PMMA and no beads.

Chapters 4 and 5 were published as original research articles, and have been reproduced in this thesis verbatim.

The in vitro gene expression study in Chapter 5 was followed up with protein expression data for dTHP-1 treated with MAA, PMMA and no beads in Chapter 6, and further probed the response of HUVEC to the conditioned medium from dTHP-1/bead cultures. In this chapter, we confirmed the increased expression of IL-1β, IL-6, and TNF-α at the protein level for MAA- treated dTHP-1 cells, but found that MAA beads did not affect the expression of secreted OPN. Furthermore, we found that MAA beads significantly decreased the expression of CXCL10, a response observed in the diabetic wound model, and the expression of CXCL12. HUVEC decreased the expression of CXCR4 in response to MAA/dTHP-1 conditioned medium.

Chapter 7 summarizes the findings of each chapter and builds a hypothesis for the mechanism of MAA-mediated angiogenesis, based on these results and the current literature. Differential activation of the toll-like receptor signaling pathway was hypothesized to mediate the angiogenic response to MAA beads. This chapter also suggests future directions for testing this hypothesis and further exploring the host response to MAA beads. Chapter 8 summarizes the major conclusions from chapters 4 through 7.

Appendix I summarizes preliminary microarray data for the diabetic wound model. Appendix II (Corstorphine and Sefton. J Tissue Eng Regen 2011:5(2);119-129; reproduced verbatim) is a supplemental study that comprised research that fell outside the scope of this thesis. The work was performed in the first two years of my Masters degree before reclassifying into the PhD program and changing the focus of my thesis to MAA beads. The focus of this study was

5 theoretical and experimental models of oxygen transport within collagen-based microtissues used in our laboratory for modular tissue engineering. The major findings from this study was that the size of modules used in our lab for tissue engineering applications did not experience transport limitations, while larger modules did, and that the density of cells within the modules spontaneously adjusted to a sustainable level.

5 References [1] Reichert WM, Ratner BD, Anderson J, Coury A, Hoffman AS, Laurencin CT, et al. 2010 Panel on the biomaterials grand challenges. J Biomed Mater Res A 2011;96(2):275-87. [2] Ehrbar M, Zeisberger SM, Raeber GP, Hubbell JA, Schnell C, Zisch AH. The role of actively released fibrin-conjugated VEGF for VEGF receptor 2 gene activation and the enhancement of angiogenesis. Biomaterials 2008;29(11):1720-9. [3] Perets A, Baruch Y, Weisbuch F, Shoshany G, Neufeld G, Cohen S. Enhancing the vascularization of three-dimensional porous alginate scaffolds by incorporating controlled release basic fibroblast growth factor microspheres. J Biomed Mater Res A 2003;65(4):489-97. [4] Phelps EA, Landázuri N, Thulé PM, Taylor WR, García AJ. Bioartificial matrices for therapeutic vascularization. Proc Natl Acad Sci USA 2010;107(8):3323-8. [5] Chen RR, Silva EA, Yuen WW, Mooney DJ. Spatio-temporal VEGF and PDGF delivery patterns blood vessel formation and maturation. Pharm Res 2007;24(2):258-64. [6] Ott HC, Matthiesen TS, Goh S-K, Black LD, Kren SM, Netoff TI, et al. Perfusion- decellularized matrix: using nature's platform to engineer a bioartificial heart. Nat Med 2008;14(2):213-21. [7] Brauker JH, Carr-Brendel VE, Martinson LA, Crudele J, Johnston WD, Johnson RC. Neovascularization of synthetic membranes directed by membrane microarchitecture. J Biomed Mater Res 1995;29(12):1517-24. [8] Madden LR, Mortisen DJ, Sussman EM, Dupras SK, Fugate JA, Cuy JL, et al. Proangiogenic scaffolds as functional templates for cardiac tissue engineering. Proc Natl Acad Sci USA 2010;107(34):15211-6. [9] Anderson JM, Rodriguez A, Chang DT. Foreign body reaction to biomaterials. Semin Immunol 2008;20(2):86-100. [10] Eckhaus AA, Fish JS, Skarja G, Semple JL, Sefton MV. A preliminary study of the effect of poly(methacrylic acid-co-methyl methacrylate) beads on angiogenesis in rodent skin grafts and the quality of the panniculus carnosus. Plast Reconstr Surg 2008;122(5):1361- 70. [11] Martin DC, Semple JL, Sefton MV. Poly(methacrylic acid-co-methyl methacrylate) beads promote vascularization and wound repair in diabetic mice. J Biomed Mater Res A 2010;93(2):484-92.

Chapter 2 Literature Review: Angiogenesis and scaffold vascularization1 1 Introduction

Evolution of multi-cellular organisms required the development of a circulatory system capable of distributing gases, nutrients and signaling molecules. Due to the poor diffusional distance of oxygen and nutrients, all cells within the human body are within 100 – 200 µm of a blood vessel with few exceptions (e.g. cells in the cornea, cartilage, ligaments)[1]. Like the evolution of larger organisms, engineered tissues of clinically relevant sizes require the incorporation of a vascular network capable of supplying seeded or recruited cells deep within the scaffold a means of exchanging gases, nutrients, signaling molecules and waste. This chapter focuses on the methods currently being developed to vascularize biomaterial scaffolds.

1.1 The native vascular network

The vascular network in mammals is a hierarchical network of vessels that circulate the blood to and from the heart through the tissues. For most vessels, there are three layers (Figure 2-1). The intima is the innermost layer and is a monolayer of endothelial cells (EC) supported by a basement membrane. EC interface with blood and have a number of functions, which include the regulation of capillary permeability, hemostasis, and leukocyte recruitment and translocation[2,3]. The middle layer is the media, which is composed of layers of contractile vascular smooth muscle cells (SMC) and elastin. The adventitia is the outermost layer and contains fibroblasts, mast cells, nerve endings and, in large vessels such as the aorta, small capillaries[4].

1.2 Angiogenesis: the formation of new blood vessels

In adults, new blood vessels are typically formed through a process called angiogenesis, in which new microvessels sprout from pre-existing vessels. While angiogenesis is not the only mechanism of vascularization, it is the most extensively studied. Two other mechanisms have

1 Fitzpatrick LE*, Lisovsky A*, Ciucurel EC and Sefton MV (2011) Scaffold vascularization. In Migliaresi C and Motta A (Eds.) Scaffolds for tissue engineering: biological design, materials and fabrication. Pan Stanford Publishing (in press). [*Authors contributed equally]

7 been described: post-natal vasculogenesis, in which vessels are formed de novo by recruited bone marrow-derived progenitor cells, and intussusception, in which the lumen of an existing blood vessel divides to form two separate vessels[5].

Figure 2-1. Structure of nascent and mature blood vessels. (a) EC alone assembled into tube-like structures (nascent blood vessel). (b) Capillaries are the smallest type of blood vessels (only 5 -10 µm in diameter), and are composed of a single layer of EC supported by a basement membrane of matrix proteins, and an incomplete covering of pericytes that stabilize the vessel. (c) Compared to capillaries, arterioles and venules are invested with a higher number of mural cells. Arterioles have a thinner media layer than arteries, and regulate the blood flow between the arteries and the capillaries. Venules (and veins) have thin walls and fewer SMC, relative to arteries. (d) Larger blood vessels have three specialized layers: intima (containing EC), media (containing SMC) and adventitia (containing fibroblasts). Arteries have a thick media composed mainly of SMC, which provide mechanical stability to the vessel wall, and allow the vessels to vasodilate and vasoconstrict in order to regulate blood flow and pressure. Reprinted by permission from Macmillan Publishers Ltd: Nature Medicine [6], Copyright 2003.

Numerous conditions can induce angiogenesis including tissue hypoxia, low pH, hypoglycemia and inflammation[7]. The angiogenic process is divided into three phases: 1) vasodilation and increased vessel permeability, 2) EC proliferation and migration, 3) vessel maturation. The first stage of angiogenesis is marked by nitric oxide (NO)-mediated vasodilation, and increased permeability of the blood vessels, which are stimulated by increased levels of vascular endothelial growth factor (VEGF)[8]. The increased permeability allows plasma proteins to extravasate into the surrounding tissue to form a provisional matrix for EC migration. This

8 process is tightly regulated by the angiopoietin (Ang)/Tie system. Ang-1 and Ang-2 both bind the Tie2 receptor, however Ang-1 acts to stabilize blood vessels, while Ang-2 antagonizes Ang-1 and is associated with vessel destabilization and either vessel growth (in the presence of VEGF) or vessel regression[9-11]. Prior to migration, EC loosen their inter-endothelial contacts and release from the supporting mural cells, further destabilizing the blood vessel[12]. Proteinases, such as matrix metalloproteinases (MMPs) and plasminogen activators, degrade the extracellular matrix (ECM) allowing EC migration, and liberate matrix-bound pro-angiogenic growth factors (GF) including VEGF and basic fibroblast growth factor (bFGF)[12].

To prevent en masse migration of EC toward an angiogenic signal, and potential dissolution of a pre-existing vessel, EC compete to lead the migrating sprout along the gradient of angiogenic factors[13]. This leading cell, called a tip cell, is selected through a spatial cell patterning mechanism involving VEGF, notch and notch pathway ligand delta-like 4 (DLL-4) (Figure 2- 2a). Adjacent to the tip cell are stalk cells, which proliferate and elongate the stalk. An in depth look at tip and stalk cell selection and function is beyond the scope of this chapter, however it is reviewed in detail elsewhere[14-16]. As the tip cell migrates and the stalk elongates, the stalk cells form a lumen and stabilize the sprouting vessel by reestablishing tight inter-endothelial junctions, laying down a basement membrane and recruiting pericytes (mediated by platelet- derived growth factor B (PDGF-B), Ang-1, transforming growth factor beta (TGF-β) and other factors)[13,15]. To become a functional, perfused vessel, the tip cell of one sprout must fuse with a neighboring sprout, a process called anastomosis, that is thought to be facilitated by macrophages[17]. As the new vessel matures, the EC resume their quiescent phalanx phenotype[15].

Angiogenesis is tightly regulated in vivo by maintaining a balance between pro-angiogenic and angiostatic (also called anti-angiogenic) stimuli[12]. Angiogenesis inhibitors (e.g. angiostatin, endostatin, anti-thrombin III, interferon-β, leukemia inhibitory factor and platelet factor 4, thrombospondin-1) can suppress EC proliferation and migration, or lumen formation[12,18]. Other molecules (TGF-β, tumor necrosis factor alpha (TNF-α)) can both stimulate or inhibit angiogenesis[12]. In order to form and maintain mature, functional vessels, a balance must exist between factors that induce and inhibit angiogenesis. For example, in cancer the balance is typically shifted too far in favour of angiogenesis, which results in the formation of tortuous, leaky, immature vessels[5,15,19] (Figure 2-2c). Often times, the vascular network formed within

9 tissue engineered constructs is also leaky and lacks functionality, due to the absence of an adequate balance or spatio-temporal coordination between signals initiating angiogenesis and signals promoting vessel maturation in the implant area.

An alternative mechanism for forming new blood vessels is vasculogenesis. Previously, it was thought that vasculogenesis occurred only during development, however recent studies have demonstrated the role of vasculogenesis in adult vascularization (although the extent of contribution is unclear)[7,21-23]. During vasculogenesis, blood vessels are formed de novo through the recruitment and differentiation of a heterogeneous (and controversial) family of bone marrow-derived cells called endothelial progenitor cells (EPC). Once recruited to a tissue, EPC proliferate, differentiate and incorporate themselves into immature vessels or form new vessels[7,24].

2 Vascularization of biomaterial scaffolds

Vascularization of biomaterial scaffolds is currently a major barrier to creating viable, complex engineered tissues, such as liver, heart, bone and adipose, at a clinically relevant scale. There are a few key challenges in creating a vascularized tissue. The first is creating the vascular network within the scaffold. The second is connecting (anastomosing) the scaffold vasculature with the host vasculature so that the scaffold is perfused in vivo. Finally, the newly formed vascular network has to mature, be functional, and persist over time. Researchers are currently using a variety of approaches to achieve scaffold vascularization (Figure 2-3). GF delivery (section 2.1) from biomaterial scaffolds was one of the first methods used to induce vessel formation in vivo. Single or multiple GF can be released from the scaffold in a controlled and sustained manner to encourage the ingrowth of host vessels within the scaffold. Another approach is vascular cell transplantation (section 2.2), where EC alone or in combination with support cells are delivered in a bulk scaffold or in small microtissues (modular tissue engineering). Gene delivery to overexpress angiogenic factors in transplanted cells has also been used. Cells can be cultured and transplanted using scaffold-free approaches or, conversely, scaffolds alone can be used to mobilize and recruit endogenous cells to vascularize the scaffold in situ (section 2.3). A common issue with both GF and cell transplantation approaches is the lack of immediate connection to the host vasculature and perfusion. Scaffold prevascularization (section 2.4) and decellularization strategies (section 2.5) aim to provide a solution to this problem. In some cases, biomaterials themselves (without the addition of GF or cells) were also shown to induce angiogenesis (section 2.6) and microfabrication methods can be used to create pre-defined, hierarchical vascular networks in vitro (section 2.7).

2.1 Growth factor delivery

2.1.1 Single growth factor delivery

One of the primary approaches to vascularize tissue engineering scaffolds is delivering angiogenic GF, such as VEGF, bFGF, and PDGF, to stimulate the ingrowth of host vasculature into the scaffold. While initial attempts to induce scaffold vascularization used single GF delivery systems, recent efforts have focused on dual or multiple GF delivery to form and maintain a vascular network.

GF have narrow therapeutic windows and short half-times in vivo, and in order to induce and maintain vascularization, the supply of exogenous GF must be sustained over several weeks to prevent vessel regression and to allow vessel maturation[25]. Bolus injection of soluble GF into the systemic circulation or directly into an ischemic zone have generally failed to achieve the prolonged, therapeutically relevant concentrations necessary to produce a lasting vascular network[26,27]. In addition, systemic exposure may lead to adverse side effects such as hemorrhage, hypotension and the vascularization of undesired sites[27,28]. Consequently, to ensure patient safety and efficacy of treatment, the dose, location and duration of exogenously administered GF must be tightly controlled. The incorporation of GF into slow-releasing biomaterial scaffolds provides the ability to have sustained local delivery of GF to the site of interest. The pharmacokinetics of protein release can be controlled through the polymer properties (e.g. polymer composition, porosity, degradation kinetics) and the method by which the GF is incorporated into the scaffold (e.g. adsorbed, encapsulated or tethered).

One of the simplest methods to incorporate a GF into a polymer scaffold is to impregnate the scaffold with an aqueous solution containing the desired protein. Gelatin hydrogels loaded with bFGF were used to promote vascularization of the thigh muscle following the excision of the femoral artery in a rabbit model of hind limb ischemia[29]. Angiographic assessment with Laser Doppler Perfusion Imaging (LDPI) demonstrated that single intramuscular injection of the bFGF-loaded hydrogel was sufficient to improve tissue blood flow 4 weeks post surgery, suggesting the therapeutic potential of the bFGF-containing gelatin scaffold.

To prolong the release of a GF of interest, the protein can be encapsulated into microspheres that are then embedded into the bulk of the scaffold, and this method has been used to deliver a

13 variety of GF [30,31]. For example, Ennet and colleagues used this method to deliver VEGF by pre-encapsulating it in poly(lactide-co-glycolide) (PLG) microspheres prior to incorporation into PLG scaffolds[30]. The encapsulation method resulted in a smaller burst and a slower, prolonged protein release, and the release kinetics were further adjusted by varying the polymer composition and microsphere size. The concentration of VEGF within the tissue infiltrating and surrounding the VEGF-releasing scaffold was present at physiologically relevant concentrations (more than 10 ng/mL) for up to 21 days, with negligible release into the systemic circulation[30]. Similarly, when implanted in vivo, composite alginate scaffolds containing encapsulated bFGF increased the von Willebrand factor (vWF)-positive microvessel density, compared to control scaffolds[31].

Alternatively, GF can be immobilized onto the scaffold to promote desired cell-material interactions[32-34]. GF can be modified to facilitate covalent binding to the scaffold, thereby prolonging their biological activity in the tissue. Ehrbar and collaborators created a variant of VEGF (TG-VEGF121) containing a transglutaminase (TG) substrate sequence that allowed it to be directly bound to a fibrin scaffold by spontaneously cross-linking with fibrinogen during factor XIII-mediated fibrin polymerization[25,32]. The TG-VEGF121 was released with the gradual degradation of the fibrin gel by enzymes (e.g. plasmin or MMPs) secreted by local cells; providing a sustained, local, low-level release of VEGF. In comparison, the passive release of freely diffusible wild-type VEGF from fibrin resulted in an initial burst release and a shorter sustained delivery (the fibrin gel was almost completely depleted two weeks post-implant). The active release of TG-VEGF121 from the fibrin gels induced the formation of blood vessels that were stabilized by SMC at 3 weeks post-implant. However, the newly formed vessels had regressed by 6 weeks despite the presence of the SMC at 3 weeks[32]. The regression was attributed to the absence of a physiological demand, although it could also indicate that administration of a single GF was insufficient for stable vessel development. This is a common issue when exogenously promoting vessel formation.

Other ways to immobilize GF include use carbodiimide chemistry, which covalently links the protein to the scaffold through an amide bond [34], or physical entrapment of the GF within the crosslinked scaffold[35]. Regardless of the method, immobilization is expected to provide persistent guidance cues for cell behaviour and migration[35], similar to what is observed in the ECM during angiogenesis. However, it also prevents the internalization of the GF upon binding

14 to its receptor, which can eventually lead to the desensitization of the cell to the GF signal[34]. Moreover, certain GF/receptor complex signaling cascades are dependent on internalization[36] and consequently are not suitable for this method of delivery.

More complex scaffold designs exploit synthetic components to direct cell invasion, matrix degradation and controlled GF release to induce vasculature growth in vivo. Phelps et al. created a polyethylene glycol (PEG) hydrogel using photo-crosslinkable PEG diacrylate (PEGDA) monomers conjugated to MMP-sensitive peptides, integrin binding peptides (arginine-glycine- aspartic acid; RGD) and VEGF[26]. This elegant synthetic system encouraged vascularized tissue invasion by allowing cells to bind and remodel the scaffold, while presenting pro- angiogenic signals throughout the matrix. There were higher microvessel densities at 2 and 4 weeks, compared to hydrogels that did not contain all three components (i.e. MMP-cleavage sites, RGD and VEGF), by microCT (indicating also that vessels were connected to the host). Furthermore, in a mouse hind-limb ischemia model, injection of the hydrogel in ischemic muscle improved the perfusion of the leg and foot of the ischemic limb 7 days after vessel ligation, compared to the negative control[26].

2.1.2 Multiple growth factor delivery

While single factor delivery typically results in the formation of a branched network of vessels, the vessels often fail to mature and stabilize. A reason for this outcome is that a single factor is likely incapable of inducing the full cascade of events that occur in angiogenesis from sprouting to maturation of the newly formed vessel. Hence, a strategy for promoting mature vessels is to initiate angiogenesis with a pro-angiogenic factor such as VEGF, then promote vessel stabilization by PDGF-mediated SMC and pericyte recruitment[37-39].

In one study, VEGF and PDGF were included in a calcium phosphate-crosslinked alginate gel to induce mature vessel formation. The alginate scaffold released PDGF more slowly than VEGF, likely due to the difference in affinity of the two GF to the alginate. VEGF had a high release rate during the first 12 days after which it diminished, and PDGF had a slower but sustained release up to 30 days[38]. In vivo, the dual delivery system did not increase the vessel density in a myocardial infarct model, compared to VEGF alone. However, it did increase the density of vessels staining positive for smooth muscle actin (SMA; a marker for SMC and pericytes),

15 indicative of greater vessel remodeling and maturation. Cardiac function was improved compared to single GF therapy[38].

A more controlled, spatially compartmentalized system for sequential VEGF and PDGF delivery was developed in the Mooney laboratory to pattern blood vessel formation and maturation[37]. A bilayered, porous PLG scaffold was designed to deliver the GF with a spatio-temporal gradient. The outer layer (first to be degraded) contained only VEGF, while the second, inner layer contained both VEGF (at a lower concentration) and microencapsulated PDGF. The VEGF was released quickly and created a concentration gradient in the surrounding tissue and within the scaffold that induced the infiltration of sprouting vessels. Encapsulation of PDGF within PLG microspheres in the inner layer delayed its release, and created a localized gradient of PDGF within the scaffold, concentrated mainly within the inner layer. The porous structure of the scaffold allowed for tissue invasion in vivo. When implanted into mouse ischemic hind limbs, the bilayered, dual delivery system yielded a slightly lower blood vessel density, but a significantly more mature vascular network, characterized by large vessels and association of SMA-positive SMC (as compared to single GF systems or empty scaffolds)[37].

One caveat that is relevant to these studies is that there is an antagonistic relationship between VEGF and PDGF; when both VEGF and PDGF are used VEGF activation of VEGFR2 suppresses PDGF-Rβ signaling in SMC, disrupting SMC/pericyte function during neovascularization[40]. How these observations relate to the vascularization results reported above is not clear.

2.2 Cell delivery and transplantation

2.2.1 Scaffolds seeded with cells

Transplantation of vascular cells (EC alone or in combination with support cells) to facilitate the vascularization of the scaffold is another common approach. Typically, EC are seeded onto the scaffold, which can be either a natural or synthetic biomaterial (or a decellularized scaffold, see section 2.5), and these endothelialized constructs are then used for transplantation. The EC are expected to re-assemble in vivo, participate in blood vessel formation (i.e. vasculogenesis), and connect to the host vasculature. Other cell types can be co-cultured and transplanted along with EC to support EC survival and/or to stabilize the nascent capillary-like structures.

A quiescent EC layer is needed to create a non-thrombogenic surface[41,42]. EC produce antithrombotic factors such as thrombomodulin, heparan sulphate, NO and prostacyclin[41]. Moreover, EC express tissue-type plasminogen activator (tPA), urokinase-type plasminogen activator (uPA) and plasminogen activator inhibitor 1 (PAI-1), and play a key role in balancing fibrinolytic versus coagulation events. An intact, non-activated endothelium also provides a non- adherent surface for the platelets and leukocytes present in the blood, whereas activated EC upregulate the expression of adhesion molecules such as E-selectin, inter-cellular adhesion molecule 1(ICAM-1) and vascular cell adhesion molecule 1(VCAM-1), and facilitate the binding of circulating leukocytes to the endothelialized surface[41].

Initial implant studies revealed that EC rapidly undergo apoptosis when transplanted without a supportive scaffold. In a study by Kraehenbuehl and collaborators, only 3% of the EC injected directly into the site of ischemia engrafted, due to extensive cell death[43]. Seeding the EC onto a biomaterial substrate (or embedding the EC within the biomaterial scaffold) can help prevent anoikis (programmed cell death induced by lack of adhesion sites), but the choice of the biomaterial is a factor. For example, seeding human umbilical vein EC (HUVEC) on collagen eliminated anoikis, but the cells still had limited survival in vivo after subcutaneous implantation in an immunodeficient mouse model. On the other hand, when the EC were seeded onto fibronectin (another ECM protein)-coated collagen instead of collagen alone, HUVEC survival was significantly improved (30 – 45% over the first three weeks) and the number of HUVEC- lined blood vessels increased by 100% at day 7 and 14 post-implantation[44].

As an alternative to the use of ECM-based substrates, synthetic polymers have also been designed to promote EC adhesion, survival, migration and organization into tube-like structures. MMP-responsive PEG hydrogels with thymosin β4 (a small bioactive peptide with pro- angiogenic properties) and RGD sequences, created a 3D environment conducive for HUVEC attachment and survival, and induced vascular-like network formation in vitro, although the potential of this system to promote EC survival and increased vascularization remains to be demonstrated in vivo[43]. Apoptosis in the absence of supporting cells is an issue, regardless of the scaffold.

Angiogenesis does not only involve EC, but is the result of a well-coordinated sequence of interactions among EC, supporting cells, and the surrounding ECM. Consequently, co-culture

17 systems with one or multiple support cell types in addition to EC are a closer mimic of the in vivo environment. Several groups reported the development of a functional microcirculation in vivo when using this approach, as reviewed elsewhere[45,46]. Seminal work by the Jain group demonstrated the formation of blood vessels and their persistence over at least 1 year in vivo when HUVEC and 10T1/2 cells were embedded together in a fibronectin-collagen gel and implanted in a transparent window, immune-compromised mouse model. The constructs prepared with HUVEC alone showed minimal perfusion and no cells were present beyond 60 days. In contrast, when HUVEC and mesenchymal precursor cells were implanted together, the 10T1/2 cells provided an abundant source of mural cells to stabilize the HUVEC-lined blood vessels that were connected to the host vasculature[47].

The West group also exploited the cellular interactions between EC and 10T1/2 cells in a synthetic PEG hydrogel containing cell attachment (RGD) and protease (MMP)-sensitive substrates. The co-cultured cells actively remodeled the hydrogel, secreted their own ECM proteins, and formed tube-like structures that were stable for at least 28 days in vitro[48].

Levenberg and collaborators used a tri-culture system to create a vascularized muscle tissue, with EC, myoblasts and embryonic fibroblasts seeded on porous biodegradable polymer scaffolds composed of 50% poly(L-lactic acid) (PLLA) and 50% polylactic-glycolic acid (PLGA)[49]. In vitro, this tri-culture system spontaneously formed tubular structures within the scaffold, with some of the embryonic fibroblasts becoming SMA-positive over time (suggestive of their differentiation into SMC), and co-localizing with the EC to stabilize the newly formed vascular structures. The skeletal muscle constructs were cultured in vitro for 2 weeks and then implanted either subcutaneously or intramuscularly in immunocompromised mouse and rat animal models. Evaluation of the implants 2 weeks after surgery showed the formation of blood vessels within the construct, with 41% of the blood vessels of human origin being perfused with lectin following lectin injection through the tail vein of the animals, thus confirming their connection to the host vasculature. Control injections with EC alone did not result in vessel formation, again presumably due to apoptosis. Furthermore, the implanted muscle construct continued to differentiate and mature over time, with implanted myotubes elongating, becoming multinucleated and expressing myogenin, a muscle-specific marker[49].

2.2.2 Modular tissue engineering

Modular tissue engineering was pioneered by the Sefton group, and provides an alternative platform to creating endothelialized bulk scaffolds. Rather than seeding EC within one large scaffold, modular tissue engineering uses small, EC-coated constructs (modules, ~ 1 mm long and 0.5 mm in diameter, typically made out of collagen) that randomly pack together within a larger chamber to form a three-dimensional construct (Figure 2-4)[50]. Vascular support cells or therapeutic cells of interest (e.g. hepatocytes) can be embedded during fabrication, prior to EC seeding[50]. The network of EC-lined void spaces formed in between the modules is amenable to perfusion, forming a pseudo-vasculature in vitro[51]. Furthermore, the small dimension of the modules (less than 400 µm following contraction by EC) is within the diffusional distance of oxygen and consequently, in individual modules even high densities of embedded cells did not experience the hypoxia that is often observed in thick cell constructs[52]. The three-dimensional tissues assembled using this method have an intrinsic vasculature, are scalable, have uniform cell density throughout the construct, and allow for the mixing of different cell types within the modules or by using modules with different types of embedded cells.

When collagen modules coated with rat aortic EC (RAEC) were transplanted into the omental pouch of allogeneic rats with immunosuppressive drug treatment, the RAEC started to form blood vessels within the first 7 days[53]. Over time, the blood vessels matured, were supported by endogenous SMC, and anastomosed with the host vasculature as shown by the accumulation of erythrocytes. Although these vessels persisted for at least 60 days, some of them were leaky, as shown by microCT perfusion studies. Moreover, a robust inflammatory response was noted around the implant[53]. When bone marrow-derived MSC were added as support cells to the endothelialized modules, the number of blood vessels did not change, but the resulting blood vessels were more stable (i.e. less leaky) compared to vessels formed by RAEC alone[54]. Encapsulated MSC became SMA positive and lined the EC layer suggesting their differentiation into pericyte-like cells. In addition, the presence of MSC decreased the total number of macrophages and shifted macrophages to what was perhaps a more tissue repair phenotype, which further improved RAEC survival and consequently, vascularization[54].

Modular tissue engineering was also tested in islet transplantation and in cardiac tissue engineering[55,56]. In the case of islet transplantation, vessel density in the implant area was significantly increased, and a trend toward increased islet viability was observed for islets

19 implanted in modules coated with EC as compared to islets implanted in collagen modules without EC at 21 days. However, the endothelialization of the modules did not bring any significant benefit in terms of islet function[55]. In the case of the modular cardiac constructs, RAEC-lined collagen modules supplemented with MatrigelTM were embedded with a neonatal rat heart cell population enriched in cardiomyocytes, and the modules were gathered to form a sheet- like, porous structure. The constructs were electrically responsive and presented cardiac markers at inter-module junctions, suggesting the potential of this approach for cardiac tissue engineering[56].

Materials other than collagen have been used to fabricate the modules. For example, the surface composition of the modules was altered by coating them with other ECM proteins, such as

20 fibronectin[44], or the modules were made mechanically stronger by replacing collagen with a photo-crosslinkable, cell adhesive synthetic polymer, such as poloxamine-polylysine acrylate[57]. In addition, other methods such as micromolding were developed to produce similar modular constructs, demonstrating the ease of module manufacturing and the potential of this technology for scale-up and construction of larger 3D tissues[58].

Although fabrication of modular tissues from different cell types is easily feasible, fabricating modular tissues with specific micro-architectures preserved into larger tissue constructs remains a challenge. The Khademhosseini group introduced a sequential assembly method that allows for better control over the relative spatial arrangement of the different building blocks[59]. The group used PEG micro-gels fabricated by sequential photo-crosslinking through two overlaying masks to generate a concentric design that emulated the EC and SMA vascular layers, with HUVEC encapsulated in the internal layer and SMC encapsulated in the external layer of the micro-gels. Several micro-gels were then reassembled into a tube-like structure upon immersion in mineral oil[59]. However, the cells were immobilized in a non-degradable PEG gel within which cells cannot proliferate or migrate. To make this approach applicable to tissue engineering, other scaffolds with chemical and mechanical properties that better mimic the natural ECM, need to be considered. The remodeling of the initial tissue architecture in vivo also needs to be explored for this system.

2.2.3 Genetically Modified Growth Factor-Producing Cells

Transplanted cells (EC or other) can be genetically modified to overexpress angiogenic factors. This approach obviates some limitations associated with the classical scaffold-GF delivery strategy, such as the short half-life of GF in vivo and GF degradation during scaffold manufacturing, while, ideally, the genetically modified cells can continuously produce angiogenic factors at constant rates. However, obtaining pharmacologically useful rates is problematic[60,61].

The Laurencin group combined EC transplantation with ex vivo gene transfer for bone tissue engineering applications[62]. Human adipose-derived mesenchymal stromal cells (adMSC) were isolated and transduced with an adenovirus to express VEGF. The co-transplantation of EC and adMSC overexpressing VEGF resulted in significant vascularization of the three-dimensional PLGA sintered microsphere scaffolds implanted subcutaneously in severe combined

21 immunodeficient (SCID) mice. The scaffolds seeded with VEGF-transduced adMSC and EC resulted in the highest number of blood vessels 2 and 3 weeks post implantation, compared to the blank scaffolds, as well as the scaffolds seeded with either EC or adMSC alone, or with co- cultured EC and non-transduced adMSC. The group speculated that VEGF-transduced adMSC increased migration of endogenous EC into the scaffold and proliferation of transplanted EC, and enhanced the differentiation of EC progenitors present within the adMSC population[62].

Soker and colleagues suspended VEGF-transfected myoblasts (using Lipofectamine) in collagen and injected them subcutaneously into nude mice[63]. While VEGF expression was noted at one week but was absent at 3 weeks, increased microvessel density (by vWF staining) was observed. The resulting muscle tissue progressively increased in volume up to 4 weeks post-implant, a volume that was preserved for the duration of the study (8 weeks). This tissue expressed typical muscle markers (desmin and sarcomeric tropomyosin) and lacked connective tissue formation[63].

Yang and colleagues used a biodegradable polymer-DNA nanoparticle system based on poly(β- amino esters) (PBAE) to deliver VEGF cDNA into human MSC and hESC-derived cells (hESdC)[64]. Scaffolds seeded with VEGF-transfected MSC and implanted subcutaneously into athymic mice led to 2 - 4 fold increase in blood vessel density at 2 and 3 weeks after the implant compared to controls (acellular scaffolds, scaffolds seeded with luciferase transfected cells, and scaffolds seeded with cells transfected with VEGF using Lipofectamine as transfection reagent). A similar trend was noted for the transfected hESdC group. The same study reported enhanced angiogenesis 4 weeks after intramuscular injection of the VEGF-transfected MSC in a mouse ischemic hind limb model, suggesting that transient, non-viral gene delivery of angiogenic factors has great potential for therapeutic angiogenesis applications (at least in hind limb muscle), while providing a safer alternative to viral gene delivery[64].

2.2.4 Scaffold-free cell delivery approaches

Scaffold-based tissue engineering strategies involve seeding of cells onto scaffolds designed to act as ECM substitutes. An alternative strategy uses the cells' ability to create their own ECM and implant the intact cell-ECM construct[65]. The cell-generated ECM may represent the optimal scaffold, because it recapitulates the native instructive signals to promote tissue growth and maturation. The Okano group cultured cells on temperature-responsive surfaces (cell culture

22 dishes grafted with poly(N-isopropylacrylamide), a temperature-responsive polymer) that allowed cell attachment and proliferation at 37°C, then promoted spontaneous cell detachment without enzymatic digestion when the temperature was lowered below 32°C[66]. Once confluent, the cell sheets were detached from the culture surface and stacked together to build a three-dimensional multilayered tissue. When a sparse layer of HUVEC was inserted in between two myoblast cell sheets, EC sprouted and formed capillary-like structures (Figure 2-5).

In a five-sheet construct, EC capillary-like structures connected through all five layers of the construct within 4 days. The five-layer myoblast sheets, with or without HUVEC, were then placed on a fibrin gel and implanted subcutaneously in nude rats. Histological examination showed that HUVEC-free myoblast sheets had no vessels visible within the graft after 7 days. In contrast, the presence of HUVEC resulted in the formation of blood vessels containing

23 endogenous erythrocytes, suggesting these blood vessels were connected to the host vasculature. One can envision that more complex tissues could be made by incorporating micro-fabrication techniques to integrate patterning and micro-texture, as discussed in section 2.7.

Another scaffold-free approach was introduced by Kelm and collaborators[67]. They used a microtissue self-assembly method to create small diameter tissue-engineered blood vessels. First, a hanging drop method resulted in microtissues (spheroids) of human artery-derived fibroblasts (HAF), with 10,000 cells/drop seeded in an inverted 60-well plate. After 2 days, HUVEC were added to coat the spheroids and the final microtissues were harvested after five additional days of culture. The microtissues were collected in a bioreactor and cultured under either static or dynamic conditions (under pulsatile flow and circumferential mechanical stimulation). Accumulation of capillary-like structures occurred within 14 days. Under dynamic conditions, although a layered tissue structure was observed after 14 days of culture, it was not characteristic of native vasculature since EC were present inside the wall while SMA-positive fibroblasts lined the lumen[67].

2.3 In situ vascularization with endogenous cells

Instead of seeding scaffolds with EC prior to implantation, bio-functionalized materials have been used to mobilize endogenous cells such as EPC in situ. This approach differs from the use of GF-containing scaffolds in that it aims to stimulate de novo blood vessel formation through mobilization of vessel forming cells, instead of promoting GF-induced sprouting from existing blood vessels. This alternative in situ strategy exploits the patient's own cells, eliminating any immune response issues.

However, the definition of what constitutes an EPC cell is controversial, since there is no specific marker that is unique to this cell population, and different methods are currently used to isolate, identify and culture these cells. As reviewed by Yoder and collaborators, there are currently several approaches to isolate EPC, which typically involve some form of in vitro cell culture of peripheral monocytes and the distinguishing of particular phenotypes as say early outgrowth or late outgrowth EPC[68,69]. Various cell marker combinations have been used (CD34+AC133+KDR+ is a common one)[70], but none can specifically and uniquely identify EPC. In one study, two types of EPC were suspended in a mixture of collagen type I and fibronectin, and the gelled constructs were implanted in the flank of NOD-SCID mice. CD31

24 staining showed donor (human) derived blood vessels were only seen with EPC that had produced colonies of cobblestone shaped cells after in vitro culture. While both EPC expressed EC markers (such as CD31, VWF, UEA-1 and others), the EPC that did not form vessels also expressed monocyte/macrophage markers (CD14, CD115), as well as the hematopoietic marker CD45. The abundance of the two cell populations was also different, with only ~ 0.02 vessel forming EPC colonies/106 MNC, instead of ~ 4 colonies/106 cells of the non-functioning EPC [71]. In a different study, these functioning EPC cells were used to create a construct for bone tissue engineering applications using a porous biphasic calcium phosphate (BCP)/bone morphogenetic protein-2 (BMP-2) scaffold. Vasculogenesis by the EPC was shown 4 weeks after the implant, with donor (human) derived blood vessels containing host red blood cells[22].

During ischemic injury, bone marrow derived ("hematopoietic") EPC have the ability to home to the site of injury and re-endothelialize the injured blood vessels. Designing a system that can mobilize and capture EPC is an attractive concept, but in addition to further clarifying what cells are actually being mobilized, several key issues must be addressed before its successful implementation. In order to increase the number of circulating EPC, a strategy will need to be devised to mobilize the EPC from the bone marrow, since only very few EPC are thought to be circulating in peripheral blood. Several factors are known to increase the number of circulating EPC, including SDF-1, IL-8, Ang-1, and granulocyte colony-stimulating factor (G-CSF) [72-76]. The main challenge lies in creating a signal that is strong enough to attract EPC and that is above the “background noise” associated with inflammation and tissue injury associated with biomaterial implantation. Moreover, since the implanted biomaterial will not be lined with EC, strategies to inhibit thrombosis and prevent undesirable protein adsorption and cell adhesion will be required.

Alobaid and collaborators evaluated the potential of a nanocomposite biomaterial based on polyhedral oligomeric silsesquioxane nanocages with incorporated RGD sequences to promote capture of circulating EPC in vitro. When peripheral blood mononuclear cells containing 1-2% CD34+ cells were plated onto nanocomposite sheets incorporating the RGD sequence, or on control nanocomposite sheets without the RGD sequence, the RGD-nanocomposite biomaterial showed a significant increase in the number of attached EPC colonies, thus showing the potential of this approach for endothelialization with EPC[77]. However, in order for this method to be applicable in vivo, bioactive sequences other than RGD will need to be considered, since RGD is

25 a universal cell adhesion peptide, and cells more abundant than EPC, such as monocytes, would most likely dominate the surface attachment sites.

SDF-1, a potent chemotactic cytokine, promotes angiogenesis (directly or by modulating the production of GF such as VEGF), and vasculogenesis by modulating cell homing to areas of neovascularization. In a study performed by Simock et al. using a nude rat animal model, a vascularized tissue was first created by placing an arteriovenous loop (AVL) inside a polycarbonate chamber in the groin area, leading to the formation of a highly vascular fibrous tissue (see section 2.4)[78]. SDF-1 was administered postoperatively into the chamber via a catheter connected to a pump, for a total of 1 mg of SDF-1 over 1 week. Human CD34+ cells (collected by leukapheresis from patients after up to 3 days stimulation with G-CSF, sorted by magnetic affinity for CD34, and stored frozen until used) were injected intracardially 6 days after creating the AVL chamber and the tissue was explanted and studied 2 and 7 days following cell injection. This approach was shown to be successful in delivering cells to the implant site, as the number of fluorescently pre-labeled CD34+ cells was 8 times higher for groups receiving SDF-1 treatment compared to the untreated group. However, these cells did not line the blood vessels at the AVL implant site, and there was no significant difference in the extent of vessel density or vascular maturity in the control versus the SDF-1 treated group. An increase in the number of leukocytes, including neutrophils and macrophages, was also observed when SDF-1 was administered. The concomitant influx of leukocytes and CD34+ cells raised the concern over the non-specificity of EPC homing signals. Moreover, the increased number of leukocytes in the treated group also presents a challenge for further studies to decouple the effects due to the homing of CD34+ cells from the effects potentially caused by the presence of other infiltrating cells[78].

2.4 Scaffold prevascularization

A relevant concern of scaffold vascularization schemes via GF and/or cell delivery is that the length of time it takes to form a functional vasculature (> 1 week) is often too long maintain viability of cells co-implanted within the scaffold. Consequently, a vasculature may eventually form, but the therapeutic cells within the engineered construct will have already died from lack of oxygen and nutrients in the interim. To address this issue, scaffolds have been prevascularized prior to implantation at the target site in vivo, so the scaffold vessels need only anastomose with

26 the host vasculature. The prevascularization can occur in vitro[49,79] or in vivo[80,81]. However, the co-culture of multiple cell types in vitro prior to implantation is not trivial, since different cell types typically have different culture requirements. Different cell culture media, as well as different cell ratios between the co-cultured cells can significantly impact the angiogenic activity and the maintenance of cell phenotypes[41,49].

Hiscox and colleagues investigated the potential of a prevascularized collagen construct to enhance the survival of transplanted islets[79,82]. Islets are highly vascularized microtissues. However, the isolation process destroys much of the islet microvasculature, and prompt revascularization and perfusion of transplanted islets is essential to islet survival and function[83]. Freshly isolated microvessel fragments from rat adipose tissue were cultured within collagen gels and were shown to form a vascular network in vitro, and rapidly integrated with the host vasculature in vivo[84]. Isolated islets were embedded within a thin collagen gel, which was then sandwiched between two layers of this prevascularized collagen gel. This construct was implanted subcutaneously into SCID mice, and was shown to improve islet survival and insulin production over the 28 days of the study, compared to islets implanted without the preformed vasculature, which had no viable islets or insulin production[79]. Furthermore, insulin-producing cells were detected and were co-localized with intra-islet EC illustrating an association between islet viability and the presence of intra-islet vasculature[79]. While further in vivo work is required to demonstrate the therapeutic potential of this approach, this prevascularization strategy significantly improved the viability of the implanted cells, and could be used to create prevascularized constructs for other target tissues.

In vivo prevascularization strategies use the host as a bioreactor to vascularize the scaffold, prior to implanting the construct at its target site. One strategy uses an AVL and vascular pedicle to generate a vascularized matrix within an enclosing chamber. In this approach, an AVL or a pedicle, which contains both an artery and a vein, is enclosed within a polymer chamber that contains a matrix (such as collagen or Matrigel) or can be left empty[85]. Within the chamber, a microvascular bed is formed within the enclosed matrix, or in the case of empty chambers, new vascularized granulation tissue is formed. Functional cells are generally incorporated once the prevascularized tissue is formed[80,86]. The newly vascularized tissue is then relocated to the desired target site or left in place[87]. This technique has been used to create a variety of vascularized tissues including adipose[87,88], skeletal muscle[89], pancreatic (islet)[80,86], and

27 cardiac tissues[90,91]. An example of this strategy utilized prevascularized chambers to support the survival and function of transplanted islets[80,86]. A chamber containing GF-reduced Matrigel™ and supplemental bFGF was placed around the epigastric pedicle in the groin of a diabetic mouse and formed a highly vascularized adipose tissue[80]. After 21 days, isolated islets were transplanted into the prevascularized chamber and the implant was left in place. After another 3 weeks, there was a significant reduction in blood glucose levels and improved glycemic control as measured by a glucose tolerance test[80]. The improvement in blood glucose regulation was due to the transplanted islets, as the mice returned to the initial hyperglycemic state after the chambers were removed[80,86].

The highly vascularized omentum is a natural “bioreactor” that has been used to vascularize tissue engineering scaffolds, such as prevascularized cardiac patches for the repair of the myocardial tissue following an infarct[81]. A mixture of neonatal cardiac cells with GF-reduced Matrigel™ and pro-survival and angiogenic factors (insulin-like growth factor 1 (IGF-1), SDF-1, and VEGF) were seeded into an alginate scaffold containing alginate-sulfate groups to enhance binding and sustained release of the GF. After 48 hour in vitro, the patch was transplanted into the omentum and was allowed to mature and remodel. After 7 days in vivo, the patch contained cardiac muscle and a network of perfused blood vessels with associated perivascular cells. The prevascularized cardiac patches were then transplanted onto infarcted rat hearts 7 days after induction of myocardial infarction. The evaluation at 28 days post implant showed that prevascularized cardiac patches were structurally and electrically integrated into the host myocardium. Moreover, the presence of the vascularized cardiac patch induced thicker tissue formation, prevented further dilation of the chamber and ventricular dysfunction. This study provided evidence that prevascularization of the cardiac patch resulted in better grafting of the patch and led to improved cardiac function after myocardial infarction.

2.5 Decellularized scaffolds

While still having elements of cell transplantation (see section 2.2), decellularized scaffolds take advantage of the preexisting architecture of a tissue that has been rendered acellular. The scaffolds are produced by perfusing an organ with detergents and buffers that effectively remove all the immunogenic cellular components from the tissue while preserving the tissue architecture and matrix, including the vascular basement membrane. The advantage of using an organ is that

28 there is a vascular inlet and outlet. This allows the organ to be connected to a perfusion system that can distribute the decellularizing agents throughout the entire organ. As the vascular architecture has been preserved, it is hypothesized that the entire vascular tree of the organ, including the capillaries, is re-endothelialized by perfusing the decellularized organ with EC. The scaffold is then seeded with functional cells, such as hepatocytes or cardiomyocytes, to create a vascularized tissue engineering construct.

In 2007, the Mertsching group engineered liver-like tissue using a decellularized porcine jejunal segment [92]. After removing the original cells, the preserved acellular vasculature was seeded with porcine microvascular EC, which were allowed to attach and grow for one week. The luminal surface of the decellularized intestine was then seeded with primary porcine hepatocytes suspended in a collagen gel, and the cells grew in multiple layers around the endothelialized capillaries. The three-dimensional liver tissue construct was maintained for 3 weeks in vitro, and the hepatocytes retained the capacity to perform liver specific functions, as measured by urea synthesis[92]. In other studies, decellularized arterial[93] and ureter[94] scaffolds have been used to engineer vascular grafts.

Taylor and colleagues presented a seminal study in which they engineered a bioartificial heart using an intact, decellularized cadaveric heart[95]. The cadaveric heart was decellularized by perfusing an SDS-based detergent through the heart chambers for 12 hours. The optimized decellularization scheme fully removed all cellular components from the matrix, while maintaining the micro and macro structures within the heart, including, presumably, a perfusable vasculature. The heart tissue was repopulated with neonatal cardiac cells through intramural injections and was maintained in a specialized bioreactor that simulated the physiological perfusion of the heart and electrical stimulation. The construct was allowed to mature under simulated physiological conditions and after 8 days, developed into a contractile tissue capable of approximately 2 % of the pump action of an adult heart[95]. The group also demonstrated that re-endothelialization of the decellularized whole-heart could be achieved by perfusion of EC into the vascular conduits. After 7 days, EC formed single layers in both larger and smaller vessels and within ventricular cavities. In a similar study, Petersen et al. decellularized adult rat lungs, then repopulated the airways with pulmonary epithelial cells and the vasculature with EC[96]. The lungs were successfully transplanted, perfused, ventilated for 2 hours, and were shown to be functional and effective at gas exchange[96].

2.6 Angiogenic biomaterials

In some cases, biomaterials themselves may induce angiogenesis without the addition of bioactive components such as GF or cells. By comparison to other vascularization strategies, using biomaterials as agonists of angiogenesis may be associated with lower cost, ease of storage, and increased versatility (the same biomaterial can be used to induce vascularization in various circumstances). The Sefton group has demonstrated the angiogenic effect of two synthetic methacrylic acid-based copolymers: poly(methacrylic acid –co– methyl methacrylate) (MAA) beads and poly(butyl methacrylate –co– methacrylic acid) (BMA-MAA) scaffolds. The MAA beads (150 – 200 µm diameter; 45 mol% methacrylic acid, 64 mol% methyl methacryle and 1% ethylene glycol dimethacrylate (EGDMA)) are non-biodegradable, and have a negative surface charge and rough surface topography[97]. In vivo, MAA beads significantly improved the vascularization of skin grafts in rats[97], and diabetic wounds in mice[98], compared to control poly(methyl methacrylate) (PMMA, 100% methyl methacrylate, 150 - 200 µm diameter) beads or no biomaterial. For tissue engineering applications, a porous BMA-MAA scaffold (45 mol% methacrylic acid, 54 mol% n-butyl methacrylate, 1% EGDMA) was developed[99]. BMA- MAA scaffolds were implanted subcutaneously in mice and promoted aggressive tissue penetration. The tissue invading the BMA-MAA scaffolds had a higher microvessel density compared to control poly(butyl methacrylate) (BMA) scaffolds at 21 and 30 days.

The mechanisms of action of MAA beads and BMA-MAA scaffold are unclear. The pro- angiogenic effect of the MAA beads was attributed to the methacrylic acid content, as no increased angiogenesis was observed in response to PMMA beads (same diameter, no methacrylic acid), and in early studies MAA beads and gels with low methacrylic acid content had a reduced effect of angiogenesis. It had been presumed that the anionic charge of MAA beads and BMA-MAA scaffold might bind endogenous cationic GF then release them slowly, extending the duration of the pro-angiogenic signaling. Also, in the context of beads, the surface topography may also play a role; MAA beads are rough while PMMA beads are smooth.

To clarify the biological mechanism of MAA-induced angiogenesis, the group has used quantitative real-time PCR to identify the effect of MAA beads on HUVEC and macrophage-like cells (dTHP-1)[100]. The MAA beads did not modify the gene expression of typical angiogenic genes (such as VEGF), but did modulate the expression of cytokines important in wound

30 healing[100]. Gene expression analysis in MAA-treated cutaneous wounds in diabetic mice showed a significant increase in the amount of sonic hedgehog (Shh; pro-angiogenic morphogen) mRNA at day 4 post-wound, which was followed by increased microvessel density at day 7[101]. Despite these insights, the mechanism remains unclear.

Based on earlier work by Brauker and collaborators[102], the Ratner group has used spherical, interconnected pores to enhance vascularization in poly(2-hydroxyethyl methacrylate) (pHEMA) scaffolds[103]. They extended the use of this scaffold microtemplating strategy for cardiac tissue engineering by adding parallel channels to encourage the organization of aligned cardiomyocyte bundles in poly(2-hydroxyethyl methacrylate-co-methacrylic acid) (pHEMA-co-MAA) scaffolds[104]. Histological examination demonstrated that the scaffold pores and the surrounding interface were filled with vascular granulation tissue, and that an optimal pore size of 30 – 40 µm maximized neovascularization while minimizing fibrous encapsulation. Other groups have developed self-assembling, peptide-based nanofiber scaffolds (reviewed by [105]) that enhance EC survival and angiogenesis, despite using peptide sequences that are not naturally occurring and are presumed to be biological inactive[105-109]. Narmoneva and colleagues demonstrated regulation of EC activation and angiogenesis in vivo using injectable RAD16-II (AcN-RARADADARARADADA-CNH2) peptide nanofibers[106,110]. In a diabetic wound healing mouse model, wound treatment with RAD16-II nanofibers significantly enhanced angiogenesis and improved healing at day 7 post-wounding[106]. The presence of nanofibers resulted in significant increases of VEGF protein levels in the wound tissue. Biotinylated lectin injections demonstrated that most of the vessels formed within the granulation tissue were anastomosed with the host vasculature by day 7. Addition of β1 and β3 integrin inhibitors completely abrogated both in vitro and in vivo effects caused by the nanofibers suggesting that nanofiber-induced angiogenesis was, at least in part, mediated by integrins despite the low- affinity binding kinetics of the integrin-nanofiber interactions[106].

Nanofiber gels formed by self-assembled peptide amphiphiles (PA) have also demonstrated angiogenic activity in vivo[107-109]. PA consist of a hydrophobic alkyl tail and a hydrophilic peptide head, which contains a sequence to promote self-assembly and bioactive domains that can be customized for specific applications[107-109]. Heparin-binding PA (HBPA) were engineered to self-assemble in the presence of heparan sulfate or heparan sulfate-like glycosaminoglycans (HSGAG). These HBPA-HS nanostructures were originally designed to

31 stimulate extensive vascularization by binding minute amounts of VEGF and bFGF[108]. In a serendipitous discovery during an in vivo biocompatibility study, the control nanofibers (without the addition of any GF) were found to promote angiogenesis and the formation of vascularized tissue in both murine subcutaneous implant and dorsal skinfold chamber models, in the absence of a fibrotic response[107]. However, by day 60 macrophages had completely degraded the nanofiber gel and the neovasculature had regressed. The observed vascularization at earlier time points (day 10 and 30) was not noted in nanofiber gels prepared without heparan sulfate or for only heparan sulfate treatment, suggesting that the angiogenic effect was attributed specifically to the presence of heparan sulfate within the nanofiber gel. It is possible that the heparan sulfate presented on the surface of HBPA-HS binds, stabilizes and gradually releases GF secreted by infiltrating endogenous cells, thereby enhancing angiogenesis[107].

Another example of an angiogenic biomaterial is a PA nanostructure designed to mimic the activity of VEGF[109]. The PA incorporated a synthetic oligopeptide that was designed to mimic the α-helical receptor-binding domain of native VEGF, and was capable of binding and inducing the phosphorylation of the VEGF receptor[109]. During in vitro culture, the VEGF-mimetic PA (VEGF-PA) enhanced EC proliferation, survival and migration compared to untreated control and to a higher degree than recombinant VEGF. In a chicken chorioallantoic membrane assay, VEGF-PA nanogels elicited a strong angiogenic response, with a > 2x increase in microvessel density at 3 days post implantation, compared to the initial blood vessel density (day 0). To evaluate the therapeutic efficacy of the VEGF-mimetic nanofibers, they were injected into the ischemic muscle 3 days following the induction of critical hind-limb ischemia. The nanofibers decreased tissue necrosis, and improved tissue perfusion, functional recovery, and limb salvage at day 21 and 28 compared to controls (VEGF peptide, mutant PA and saline), and the nanofibers performed as well or better than the administration of recombinant VEGF protein. The most significantly affected measure was the increased CD31-positive capillary density that VEGF-PA induced in the ischemic muscle, compared to a 20-µg bolus injection of VEGF165 and a saline control[109].

2.7 Microfabrication methods

Many of the strategies for creating vascularized scaffolds rely on manipulating the biology of vessel formation, using angiogenic factors, transplanted cells or bioreactors to induce vessel

32 formation within the scaffold. Another vascularization paradigm focuses on generating predefined, hierarchical pseudo-vasculatures in vitro, with a high degree of spatial control using technologies such as MicroElectroMechanical Systems (MEMS) and cell/protein printing. These approaches allow for the vascular network to be designed, ensuring optimal blood flow and mass transport characteristics[111,112].

The Vacanti group pioneered the concept of engineering a vasculature in vitro using micromachining technologies to create a blueprint for a microvascular network within the scaffold[113]. While early experiments aimed at growing spatially patterned cell sheets that were layered to form a three-dimensional construct[113], the group eventually focused on forming polymer films from non-degradable poly(dimethyl siloxane) (PDMS)[114,115] and degradable poly(glycerol sebacate) (PGS)[111]. These contained branching, vessel-like channels throughout the construct that were partially lined with EC to form a pseudo-vasculature in vitro. Master molds of the vascular blueprint were created using MEMS, then used to cast the PDMS or PGS films[114]. The patterned films were then bonded to flat films, to create an enclosed, perfusable microfluidics network[111] (Figure 2-6). The constructs were either seeded under static conditions with EC directly, or were coated with cell adhesive molecules prior to seeding. Once attached, the cells were exposed to flow for 2 to 4 weeks and formed confluent endothelial monolayers that covered some sections of the microchannels[111,115].

Micropatterning techniques can also be used to pattern surfaces with cell instructive ligands, to spatially direct a desired cell response. The West group used this approach to regulate the formation of capillary-like structures[116]. The surface of nonadhesive PEGDA hydrogels was patterned with a cell adhesive ligand, Arg-Gly-Asp-Ser (RGDS), in stripes of varying width and ligand concentration. Both stripe width and ligand concentration (µg/cm2) modulated the endothelial morphogenesis, with the EC forming cord-like structures on 50 µm-wide stripes with a ligand concentration of 20µg/cm2. This response was inhibited on wider stripes, and stripes containing a higher concentration of ligand, suggesting that both geometrical and biochemical cues were controlling EC morphogenesis[116]. A micropatterning technique was also used to direct the fate of progenitor cells in vitro[117]. Angiogenic progenitors obtained from differentiating embryoid bodies were seeded on 100 µm-wide lanes micropatterned with either collagen alone or collagen displaying immobilized VEGF. Endothelial progenitors attached to the VEGF-collagen surface differentiated into mainly EC, while cells grown on collagen surfaces

33 differentiated mainly into SMC, resulting in the formation of EC stripes lined with SMC, roughly mimicking a blood vessel [117].

Zheng et al. used a microfabrication technique to create a scaffold with a well-defined microchannel network capable of modulating scaffold vascularization in vivo[118]. A PDMS micromold with patterned microstructures (cylindrical pillars with 100 to 400 µm diameters and connected slots with 100 µm width and 600 µm length) was assembled within a larger PDMS macromold and used to cast 2% collagen microstructured tissue templates. The microstructured collagen templates were implanted subcutaneously and the void spaces within the collagen gels were rapidly infiltrated with endogenous cells (compared to the 2% collagen volume). After 14 days, vascularized host tissue had invaded the patterned pores, with vessels running along the length of the pores. The large vessels (20 – 30 µm diameter) surrounded by SMA-positive cells and containing erythrocytes were found along the axis of the pores, while smaller capillaries were found penetrating the deepest regions of the pores and the lateral branches of invaded tissue [118]. While over time, the collagen scaffold would likely undergo remodeling and degradation, at early time points it demonstrated the potential to guide the initial cellular invasion and vascularization.

While the technology is still in its infancy, three-dimensional bioprinting may offer an interesting means to creating thick tissue engineered constructs with pre-defined architecture. Nakamura and colleagues used an inkjet-based platform capable of printing cells and biomolecules, layer by layer to form three-dimensional constructs (reviewed by [119]). However, to date no efforts have been made to attempt to print the complex three-dimensional architecture that would be required for a vascular network.

3 Limitations and the road ahead

Much effort has been devoted in the last decade to develop functional, vascularized tissue engineered constructs. Despite significant progress, especially in terms of expanding our current understanding of the design requirements for creating a functional vasculature within the construct, there is still much more required. The vascularization process is typically too slow to maintain the viability of the therapeutic cells of interest, and only tissues with low vascularization requirements, such as thin skin, tissue or cartilage, are currently available clinically.

Most of the current vascularization approaches have proven successful in inducing blood vessel formation in vivo, but many of these blood vessels are not fully functional (leaky or not connected to the host vasculature), and many regress over time. Where some of these approaches have resulted in a functional, persistent blood vessel network, the time required to achieve such functional vascularization is still too long to be clinically relevant (several weeks for a millimeter sized implant)[120]. Without an immediate vascular supply, the majority of the therapeutic cells (ex. liver cells, islets) within the construct will not survive. Going forward, much effort will need to be directed towards accelerating the formation of a perfused (i.e. anastomosed) vascular network within the construct and ensuring the new vasculature matures and persists.

Another key lesson learned from previous attempts to generate vascularized tissues is that any vascular structures pre-formed in vitro prior to implantation will most likely be remodeled in vivo. For example, when EC-coated modular constructs were implanted in mouse or rat animal models, the EC migrated off the surface of the modules and formed blood vessels in the channels between the modules instead of staying on the modules and providing non-thrombogenic surface to the channels randomly formed between the modules, as was originally expected[50]. The implanted cells were presumed to be the drivers of the remodeling process, with endogenous

35 cells invading and remodeling the construct[53-55,121]. Finding ways to predict and guide the remodeling process in vivo will likely emerge as a key feature for the next generation of vascularized scaffolds, and the focus of vascularization research will likely switch from building pre-vascularized structures to controlling how the vascular structures will be remodeled and integrated with the host vasculature in vivo.

4 Summary

The tissue engineered constructs that are being developed currently have limited clinical usefulness due to the lack of inherent vasculature, which critically constrains the construct size. A vascular supply is needed to maintain the viability of a clinically relevant number of therapeutic cells (in clinically relevant tissue densities) in order to achieve a functional benefit upon transplantation. A broad range of vascularization approaches are being explored to promote blood vessel formation within the construct. Initial studies using single GF delivery from tissue engineering scaffolds were successful in inducing blood vessel formation at early time points, but failed to produce a lasting vascular network. Hence, it was reasoned that the use of multiple GF, with different functions in angiogenesis, would be more suitable to recapitulate all stages of vessel formation. Multiple GF delivery improved vessel maturation and functionality, although issues still remain in terms of delivering physiological doses of GF, in a spatio-temporal controlled manner. Scaffolds are also used to deliver (transplant) cells to a target site, or to mobilize a specific cell type in situ. Transplanted or mobilized vascular cells contribute to the formation of a vascular network both directly, by re-assembling into vessel-like structures, as well as indirectly, by secreting various GF and driving the construct remodeling process and host response. However, the time frame required to develop a perfusable, functional blood vessel network within the construct is on the order of weeks, which significantly hinders the clinical benefit of this approach since most functional cells will not be able to survive without a vascular supply for such an extended period of time.

To circumvent this last issue, scaffold prevascularization approaches have been developed. The scaffold can be seeded with vascular cells and cultured in vitro to create tube-like structures within the scaffold prior to implantation (in vitro prevascularization). Alternatively, avascular scaffolds can be implanted in a highly vascularized site in vivo (such as the omentum) to facilitate construct vascularization, followed by relocation and connection to the host vasculature

36 at the desired target site once the construct has been fully vascularized. Decellularized scaffolds have been used to provide a natural template for recreating the vascular tree architecture. Tissues or whole organs are first decellularized (while preserving the tissue architecture), and then repopulated with new cells (EC to repopulate the vascular tree and the desired functional cell populations). Another approach to induce vascularization is the use of biomaterials that act as angiogenic agonists. The use of angiogenic biomaterials provides a simple, cost effective and versatile method of inducing angiogenesis, as the same biomaterial can potentially be used to induce vascularization of various tissues with minimal modification. Finally, microfabrication techniques are currently being explored to create predefined, hierarchically organized vascular structures in vitro. The remodeling that occurs upon implantation of these structures in vivo remains to be further clarified.

Overall, significant progress has been made in expanding our understanding of the design criteria for fabricating vascularized tissue engineered constructs. However, there is still a limited number of clinically relevant tissue engineered constructs available, and further work is required in resolving the issue of poor tissue construct vascularity. Moving forward, vascularization strategies need focus on accelerating the vascularization process and on improving the functionality of the vascular network formed.

5 References [1] Carmeliet P, Jain RK. Angiogenesis in cancer and other diseases. Nature 2000;407(6801):249-57. [2] Dejana E. Endothelial cell-cell junctions: happy together. Nat Rev Mol Cell Biol 2004;5(4):261-70. [3] Ramcharan KS, Lip GYH, Stonelake PS, Blann AD. The endotheliome: a new concept in vascular biology. Thromb Res 2011;128(1):1-7. [4] Tortora G, Derrickson B. Principles of Anatomy and Physiology. 12th ed. USA: Wiley; 2009. [5] Goel S, Duda DG, Xu L, Munn LL, Boucher Y, Fukumura D, et al. Normalization of the vasculature for treatment of cancer and other diseases. Physiol Rev 2011;91(3):1071- 121. [6] Jain RK. Molecular regulation of vessel maturation. Nat Med 2003;9(6):685-93. [7] Käβmeyer S, Plendl J, Custodis P, Bahramsoltani M. New insights in vascular development: vasculogenesis and endothelial progenitor cells. Anat Histol Embryol 2009;38(1):1-11.

[8] Cooke JP, Losordo DW. Nitric oxide and angiogenesis. Circulation 2002;105(18):2133- 5. [9] Maisonpierre PC, Suri C, Jones PF, Bartunkova S, Wiegand SJ, Radziejewski C, et al. Angiopoietin-2, a natural antagonist for Tie2 that disrupts in vivo angiogenesis. Science 1997;277(5322):55-60. [10] Holash J, Maisonpierre PC, Compton D, Boland P, Alexander CR, Zagzag D, et al. Vessel cooption, regression, and growth in tumors mediated by angiopoietins and VEGF. Science 1999;284(5422):1994-8. [11] Reiss Y, Droste J, Heil M, Tribulova S, Schmidt MHH, Schaper W, et al. Angiopoietin-2 impairs revascularization after limb ischemia. Circ Res 2007;101(1):88-96. [12] Carmeliet P. Mechanisms of angiogenesis and arteriogenesis. Nat Med 2000;6(4):389- 95. [13] Gerhardt H. VEGF guides angiogenic sprouting utilizing endothelial tip cell filopodia. J Cell Biol 2003;161(6):1163-77. [14] Geudens I, Gerhardt H. Coordinating cell behaviour during blood vessel formation. Development 2011;138(21):4569-83. [15] Carmeliet P, Jain RK. Molecular mechanisms and clinical applications of angiogenesis. Nature 2011;473(7347):298-307. [16] Adams RH, Alitalo K. Molecular regulation of angiogenesis and lymphangiogenesis. Nat Rev Mol Cell Biol 2007;8(6):464-78. [17] Fantin A, Vieira JM, Gestri G, Denti L, Schwarz Q, Prykhozhij S, et al. Tissue macrophages act as cellular chaperones for vascular anastomosis downstream of VEGF- mediated endothelial tip cell induction. Blood 2010;116(5):829-40. [18] Davis GE, Stratman AN, Sacharidou A, Koh W. Molecular basis for endothelial lumen formation and tubulogenesis during vasculogenesis and angiogenic sprouting. Int Rev Cell Mol Biol 2011;288:101-65. [19] Schuch G, Kisker O, Atala A, Soker S. Pancreatic tumor growth is regulated by the balance between positive and negative modulators of angiogenesis. Angiogenesis 2002;5(3):181-90. [20] Carmeliet P, Jain RK. Principles and mechanisms of vessel normalization for cancer and other angiogenic diseases. Nat Rev Drug Discov 2011;10(6):417-27. [21] Tongers J, Roncalli JG, Losordo DW. Role of endothelial progenitor cells during ischemia-induced vasculogenesis and collateral formation. Microvasc Res 2010;79(3):200-6. [22] Levengood SKL, Poellmann MJ, Clark SG, Ingram DA, Yoder MC, Wagoner Johnson AJ. Human endothelial colony forming cells undergo vasculogenesis within biphasic calcium phosphate bone tissue engineering constructs. Acta Biomater 2011;7(12):4222- 8. [23] Velazquez OC. Angiogenesis and vasculogenesis: inducing the growth of new blood vessels and wound healing by stimulation of bone marrow-derived progenitor cell mobilization and homing. J Vasc Surg 2007;45 Suppl A:A39-47.

[24] Masuda H, Asahara T. Post-natal endothelial progenitor cells for neovascularization in tissue regeneration. Cardiovasc Res 2003;58(2):390-8. [25] Ehrbar M, Metters A, Zammaretti P, Hubbell JA, Zisch AH. Endothelial cell proliferation and progenitor maturation by fibrin-bound VEGF variants with differential susceptibilities to local cellular activity. J Control Release 2005;101(1-3):93-109. [26] Phelps EA, Landázuri N, Thulé PM, Taylor WR, García AJ. Bioartificial matrices for therapeutic vascularization. Proc Natl Acad Sci USA 2010;107(8):3323-8. [27] Simons M, Bonow RO, Chronos NA, Cohen DJ, Giordano FJ, Hammond HK, et al. Clinical trials in coronary angiogenesis: issues, problems, consensus: An expert panel summary. Circulation 2000;102(11):E73-86. [28] Zisch AH, Lutolf MP, Hubbell JA. Biopolymeric delivery matrices for angiogenic growth factors. Cardiovasc Pathol 2003;12(6):295-310. [29] Doi K, Ikeda T, Marui A, Kushibiki T, Arai Y, Hirose K, et al. Enhanced angiogenesis by gelatin hydrogels incorporating basic fibroblast growth factor in rabbit model of hind limb ischemia. Heart Vessels 2007;22(2):104-8. [30] Ennett AB, Kaigler D, Mooney DJ. Temporally regulated delivery of VEGF in vitro and in vivo. J Biomed Mater Res A 2006;79(1):176-84. [31] Perets A, Baruch Y, Weisbuch F, Shoshany G, Neufeld G, Cohen S. Enhancing the vascularization of three-dimensional porous alginate scaffolds by incorporating controlled release basic fibroblast growth factor microspheres. J Biomed Mater Res A 2003;65(4):489-97. [32] Ehrbar M, Zeisberger SM, Raeber GP, Hubbell JA, Schnell C, Zisch AH. The role of actively released fibrin-conjugated VEGF for VEGF receptor 2 gene activation and the enhancement of angiogenesis. Biomaterials 2008;29(11):1720-9. [33] Zisch AH, Zeisberger SM, Ehrbar M, Djonov V, Weber CC, Ziemiecki A, et al. Engineered fibrin matrices for functional display of cell membrane-bound growth factor- like activities: study of angiogenic signaling by ephrin-B2. Biomaterials 2004;25(16):3245-57. [34] Shen YH, Shoichet MS, Radisic M. Vascular endothelial growth factor immobilized in collagen scaffold promotes penetration and proliferation of endothelial cells. Acta Biomater 2008;4(3):477-89. [35] Moore K, MacSween M, Shoichet M. Immobilized concentration gradients of neurotrophic factors guide neurite outgrowth of primary neurons in macroporous scaffolds. Tissue Eng 2006;12(2):267-78. [36] Ganeff C, Remouchamps C, Boutaffala L, Benezech C, Galopin G, Vandepaer S, et al. Induction of the Alternative NF-κB Pathway by Lymphotoxin αβ (LTαβ) Relies on Internalization of LTβ Receptor. Mol Cell Biol 2011;31(21):4319-34. [37] Chen RR, Silva EA, Yuen WW, Mooney DJ. Spatio-temporal VEGF and PDGF delivery patterns blood vessel formation and maturation. Pharm Res 2007;24(2):258-64. [38] Hao X, Silva EA, Månsson-Broberg A, Grinnemo K-H, Siddiqui AJ, Dellgren G, et al. Angiogenic effects of sequential release of VEGF-A165 and PDGF-BB with alginate hydrogels after myocardial infarction. Cardiovasc Res 2007;75(1):178-85.

[39] Richardson TP, Peters MC, Ennett AB, Mooney DJ. Polymeric system for dual growth factor delivery. Nat Biotechnol 2001;19(11):1029-34. [40] Greenberg JI, Shields DJ, Barillas SG, Acevedo LM, Murphy E, Huang J, et al. A role for VEGF as a negative regulator of pericyte function and vessel maturation. Nature 2008;456(7223):809-13. [41] Khan OF, Sefton MV. Endothelialized biomaterials for tissue engineering applications in vivo. Trends Biotechnol 2011;29(8):379-87. [42] de Mel A, Jell G, Stevens MM, Seifalian AM. Biofunctionalization of biomaterials for accelerated in situ endothelialization: a review. Biomacromolecules 2008;9(11):2969-79. [43] Kraehenbuehl TP, Ferreira LS, Zammaretti P, Hubbell JA, Langer R. Cell-responsive hydrogel for encapsulation of vascular cells. Biomaterials 2009;30(26):4318-24. [44] Cooper TP, Sefton MV. Fibronectin coating of collagen modules increases in vivo HUVEC survival and vessel formation in SCID mice. Acta Biomater 2011;7(3):1072-83. [45] Kaully T, Kaufman-Francis K, Lesman A, Levenberg S. Vascularization - the conduit to viable engineered tissues. Tissue Eng Part B Rev 2009;15(2):159-69. [46] Lovett M, Lee K, Edwards A, Kaplan DL. Vascularization strategies for tissue engineering. Tissue Eng Part B Rev 2009;15(3):353-70. [47] Koike N, Fukumura D, Gralla O, Au P, Schechner JS, Jain RK. Tissue engineering: creation of long-lasting blood vessels. Nature 2004;428(6979):138-9. [48] Moon JJ, Saik JE, Poché RA, Leslie-Barbick JE, Lee S-H, Smith AA, et al. Biomimetic hydrogels with pro-angiogenic properties. Biomaterials 2010;31(14):3840-7. [49] Levenberg S, Rouwkema J, Macdonald M, Garfein ES, Kohane DS, Darland DC, et al. Engineering vascularized skeletal muscle tissue. Nat Biotechnol 2005;23(7):879-84. [50] McGuigan AP, Sefton MV. Vascularized organoid engineered by modular assembly enables blood perfusion. Proc Natl Acad Sci USA 2006;103(31):11461-6. [51] Khan OF, Sefton MV. Perfusion and characterization of an endothelial cell-seeded modular tissue engineered construct formed in a microfluidic remodeling chamber. Biomaterials 2010;31(32):8254-61. [52] Corstorphine L, Sefton MV. Effectiveness factor and diffusion limitations in collagen gel modules containing HepG2 cells. J Tissue Eng Regen Med 2011;5(2):119-29. [53] Chamberlain MD, Gupta R, Sefton MV. Chimeric vessel tissue engineering driven by endothelialized modules in immunosuppressed Sprague-Dawley rats. Tissue Eng Part A 2011;17(1-2):151-60. [54] Chamberlain MD, Gupta R, Sefton MV. Bone marrow-derived mesenchymal stromal cells enhance chimeric vessel development driven by endothelial cell coated microtissues. Tissue Eng Part A 2011;in press. [55] Gupta R, Sefton MV. Application of an endothelialized modular construct for islet transplantation in syngeneic and allogeneic immunosuppressed rat models. Tissue Eng Part A 2011;17(15-16):2005-15.

[56] Leung BM, Sefton MV. A modular approach to cardiac tissue engineering. Tissue Eng Part A 2010;16(10):3207-18. [57] Ciucurel EC, Sefton MV. A poloxamine-polylysine acrylate scaffold for modular tissue engineering. J Biomater Sci Polym Ed 2010;22(18):2515-28. [58] Bruzewicz DA, McGuigan AP, Whitesides GM. Fabrication of a modular tissue construct in a microfluidic chip. Lab Chip 2008;8(5):663-71. [59] Du Y, Ghodousi M, Qi H, Haas N, Xiao W, Khademhosseini A. Sequential assembly of cell-laden hydrogel constructs to engineer vascular-like microchannels. Biotechnol Bioeng 2011;108(7):1693-703. [60] Cheng D, Lo C, Sefton MV. Effect of mouse VEGF164 on the viability of hydroxyethyl methacrylate-methyl methacrylate-microencapsulated cells in vivo: bioluminescence imaging. J Biomed Mater Res A 2008;87(2):321-31. [61] Cheng D, Sefton MV. Dual delivery of placental growth factor and vascular endothelial growth factor from poly(hydroxyethyl methacrylate-co-methyl methacrylate) microcapsules containing doubly transfected luciferase-expressing L929 cells. Tissue Eng Part A 2009;15(8):1929-39. [62] Jabbarzadeh E, Starnes T, Khan YM, Jiang T, Wirtel AJ, Deng M, et al. Induction of angiogenesis in tissue-engineered scaffolds designed for bone repair: a combined gene therapy-cell transplantation approach. Proc Natl Acad Sci USA 2008;105(32):11099- 104. [63] De Coppi P, Delo D, Farrugia L, Udompanyanan K, Yoo JJ, Nomi M, et al. Angiogenic gene-modified muscle cells for enhancement of tissue formation. Tissue Eng 2005;11(7- 8):1034-44. [64] Yang F, Cho S-W, Son SM, Bogatyrev SR, Singh D, Green JJ, et al. Genetic engineering of human stem cells for enhanced angiogenesis using biodegradable polymeric nanoparticles. Proc Natl Acad Sci USA 2010;107(8):3317-22. [65] L'Heureux N, Paquet S, Labbe R, Germain L, Auger FA. A completely biological tissue- engineered human blood vessel. Faseb J 1998;12(1):47-56. [66] Sasagawa T, Shimizu T, Sekiya S, Haraguchi Y, Yamato M, Sawa Y, et al. Design of prevascularized three-dimensional cell-dense tissues using a cell sheet stacking manipulation technology. Biomaterials 2010;31(7):1646-54. [67] Kelm JM, Lorber V, Snedeker JG, Schmidt D, Broggini-Tenzer A, Weisstanner M, et al. A novel concept for scaffold-free vessel tissue engineering: self-assembly of microtissue building blocks. J Biotechnol 2010;148(1):46-55. [68] Richardson MR, Yoder MC. Endothelial progenitor cells: quo vadis? J Mol Cell Cardiol 2011;50(2):266-72. [69] Critser PJ, Voytik-Harbin SL, Yoder MC. Isolating and defining cells to engineer human blood vessels. Cell Prolif 2011;44 Suppl 1:15-21. [70] Peichev M, Naiyer AJ, Pereira D, Zhu Z, Lane WJ, Williams M, et al. Expression of VEGFR-2 and AC133 by circulating human CD34(+) cells identifies a population of functional endothelial precursors. Blood 2000;95(3):952-8.

[71] Yoder MC, Mead LE, Prater D, Krier TR, Mroueh KN, Li F, et al. Redefining endothelial progenitor cells via clonal analysis and hematopoietic stem/progenitor cell principals. Blood 2007;109(5):1801-9. [72] Aoki J, Serruys PW, van Beusekom H, Ong AT, McFadden EP, Sianos G, et al. Endothelial progenitor cell capture by stents coated with antibody against CD34: the HEALING-FIM (Healthy Endothelial Accelerated Lining Inhibits Neointimal Growth- First In Man) Registry. J Am Coll Cardiol 2005;45(10):1574-9. [73] Blindt R, Vogt F, Astafieva I, Fach C, Hristov M, Krott N, et al. A novel drug-eluting stent coated with an integrin-binding cyclic Arg-Gly-Asp peptide inhibits neointimal hyperplasia by recruiting endothelial progenitor cells. J Am Coll Cardiol 2006;47(9):1786-95. [74] Pislaru SV, Harbuzariu A, Gulati R, Witt T, Sandhu NP, Simari RD, et al. Magnetically targeted endothelial cell localization in stented vessels. J Am Coll Cardiol 2006;48(9):1839-45. [75] Suuronen EJ, Zhang P, Kuraitis D, Cao X, Melhuish A, McKee D, et al. An acellular matrix-bound ligand enhances the mobilization, recruitment and therapeutic effects of circulating progenitor cells in a hindlimb ischemia model. Faseb J 2009;23(5):1447-58. [76] Avci-Adali M, Ziemer G, Wendel HP. Induction of EPC homing on biofunctionalized vascular grafts for rapid in vivo self-endothelialization--a review of current strategies. Biotechnol Adv 2010;28(1):119-29. [77] Alobaid N, Salacinski HJ, Sales KM, Ramesh B, Kannan RY, Hamilton G, et al. Nanocomposite containing bioactive peptides promote endothelialisation by circulating progenitor cells: an in vitro evaluation. Eur J Vasc Endovasc Surg 2006;32(1):76-83. [78] Simcock JW, Penington AJ, Morrison WA, Thompson EW, Mitchell GM. Endothelial precursor cells home to a vascularized tissue engineering chamber by application of the angiogenic chemokine CXCL12. Tissue Eng Part A 2009;15(3):655-64. [79] Hiscox AM, Stone AL, Limesand S, Hoying JB, Williams SK. An islet-stabilizing implant constructed using a preformed vasculature. Tissue Eng Part A 2008;14(3):433- 40. [80] Hussey AJ, Winardi M, Han X-L, Thomas GPL, Penington AJ, Morrison WA, et al. Seeding of pancreatic islets into prevascularized tissue engineering chambers. Tissue Eng Part A 2009;15(12):3823-33. [81] Dvir T, Kedem A, Ruvinov E, Levy O, Freeman I, Landa N, et al. Prevascularization of cardiac patch on the omentum improves its therapeutic outcome. Proc Natl Acad Sci USA 2009;106(35):14990-5. [82] Hoying JB, Boswell CA, Williams SK. Angiogenic potential of microvessel fragments established in three-dimensional collagen gels. In Vitro Cell Dev Biol Anim 1996;32(7):409-19. [83] Jansson L, Carlsson PO. Graft vascular function after transplantation of pancreatic islets. Diabetologia 2002;45(6):749-63.

[84] Shepherd BR, Chen HY, Smith CM, Gruionu G, Williams SK, Hoying JB. Rapid perfusion and network remodeling in a microvascular construct after implantation. Arterioscler Thromb Vasc Biol 2004;24(5):898-904. [85] Cronin KJ, Messina A, Knight KR, Cooper-White JJ, Stevens GW, Penington AJ, et al. New murine model of spontaneous autologous tissue engineering, combining an arteriovenous pedicle with matrix materials. Plast Reconstr Surg 2004;113(1):260-9. [86] Forster NA, Penington AJ, Hardikar AA, Palmer JA, Hussey A, Tai J, et al. A prevascularized tissue engineering chamber supports growth and function of islets and progenitor cells in diabetic mice. Islets 2011;3(5):271-83. [87] Dolderer JH, Thompson EW, Slavin J, Trost N, Cooper-White JJ, Cao Y, et al. Long- term stability of adipose tissue generated from a vascularized pedicled fat flap inside a chamber. Plast Reconstr Surg 2011;127(6):2283-92. [88] Wiggenhauser PS, Muller DF, Melchels FP, Egana JT, Storck K, Mayer H, et al. Engineering of vascularized adipose constructs. Cell Tissue Res 2011. [89] Borschel GH, Dow DE, Dennis RG, Brown DL. Tissue-engineered axially vascularized contractile skeletal muscle. Plast Reconstr Surg 2006;117(7):2235-42. [90] Birla RK, Borschel GH, Dennis RG, Brown DL. Myocardial engineering in vivo: formation and characterization of contractile, vascularized three-dimensional cardiac tissue. Tissue Eng 2005;11(5-6):803-13. [91] Morritt AN, Bortolotto SK, Dilley RJ, Han X, Kompa AR, McCombe D, et al. Cardiac tissue engineering in an in vivo vascularized chamber. Circulation 2007;115(3):353-60. [92] Linke K, Schanz J, Hansmann J, Walles T, Brunner H, Mertsching H. Engineered liver- like tissue on a capillarized matrix for applied research. Tissue Eng 2007;13(11):2699- 707. [93] Zhao Y, Zhang S, Zhou J, Wang J, Zhen M, Liu Y, et al. The development of a tissue- engineered artery using decellularized scaffold and autologous ovine mesenchymal stem cells. Biomaterials 2010;31(2):296-307. [94] Derham C, Yow H, Ingram J, Fisher J, Ingham E, Korrosis SA, et al. Tissue engineering small-diameter vascular grafts: preparation of a biocompatible porcine ureteric scaffold. Tissue Eng Part A 2008;14(11):1871-82. [95] Ott HC, Matthiesen TS, Goh S-K, Black LD, Kren SM, Netoff TI, et al. Perfusion- decellularized matrix: using nature's platform to engineer a bioartificial heart. Nat Med 2008;14(2):213-21. [96] Petersen TH, Calle EA, Zhao L, Lee EJ, Gui L, Raredon MB, et al. Tissue-engineered lungs for in vivo implantation. Science 2010;329(5991):538-41. [97] Eckhaus AA, Fish JS, Skarja G, Semple JL, Sefton MV. A preliminary study of the effect of poly(methacrylic acid-co-methyl methacrylate) beads on angiogenesis in rodent skin grafts and the quality of the panniculus carnosus. Plast Reconstr Surg 2008;122(5):1361- 70. [98] Martin DC, Semple JL, Sefton MV. Poly(methacrylic acid-co-methyl methacrylate) beads promote vascularization and wound repair in diabetic mice. J Biomed Mater Res A 2010;93(2):484-92.

[99] Butler MJ, Sefton MV. Poly(butyl methacrylate-co-methacrylic acid) tissue engineering scaffold with pro-angiogenic potential in vivo. J Biomed Mater Res A 2007;82(2):265- 73. [100] Fitzpatrick LE, Chan JWY, Sefton MV. On the mechanism of poly(methacrylic acid -co- methyl methacrylate)-induced angiogenesis: Gene expression analysis of dTHP-1 cells. Biomaterials 2011;32:8957-67. [101] Fitzpatrick LE, Lisovsky A, Sefton MV. The expression of sonic hedgehog in diabetic wounds following treatment with poly(methacrylic acid-co-methyl methacrylate) beads. Biomaterials 2012;33(21):5297-307. [102] Brauker JH, Carr-Brendel VE, Martinson LA, Crudele J, Johnston WD, Johnson RC. Neovascularization of synthetic membranes directed by membrane microarchitecture. J Biomed Mater Res 1995;29(12):1517-24. [103] Fukano Y, Usui ML, Underwood RA, Isenhath S, Marshall AJ, Hauch KD, et al. Epidermal and dermal integration into sphere-templated porous poly(2-hydroxyethyl methacrylate) implants in mice. J Biomed Mater Res A 2010;94(4):1172-86. [104] Madden LR, Mortisen DJ, Sussman EM, Dupras SK, Fugate JA, Cuy JL, et al. Proangiogenic scaffolds as functional templates for cardiac tissue engineering. Proc Natl Acad Sci USA 2010;107(34):15211-6. [105] Zhang S. Fabrication of novel biomaterials through molecular self-assembly. Nat Biotechnol 2003;21(10):1171-8. [106] Cho H, Balaji S, Sheikh AQ, Hurley JR, Tian YF, Collier JH, et al. Regulation of endothelial cell activation and angiogenesis by injectable peptide nanofibers. Acta Biomater 2011. [107] Ghanaati S, Webber MJ, Unger RE, Orth C, Hulvat JF, Kiehna SE, et al. Dynamic in vivo biocompatibility of angiogenic peptide amphiphile nanofibers. Biomaterials 2009;30(31):6202-12. [108] Rajangam K, Behanna HA, Hui MJ, Han X, Hulvat JF, Lomasney JW, et al. Heparin binding nanostructures to promote growth of blood vessels. Nano Lett 2006;6(9):2086- 90. [109] Webber MJ, Tongers J, Newcomb CJ, Marquardt K-T, Bauersachs J, Losordo DW, et al. Supramolecular nanostructures that mimic VEGF as a strategy for ischemic tissue repair. Proc Natl Acad Sci USA 2011;108(33):13438-43. [110] Narmoneva DA, Oni O, Sieminski AL, Zhang S, Gertler JP, Kamm RD, et al. Self- assembling short oligopeptides and the promotion of angiogenesis. Biomaterials 2005;26(23):4837-46. [111] Fidkowski C, Kaazempur-Mofrad M, Borenstein J, Vacanti J, Langer R, Wang Y. Endothelialized microvasculature based on a biodegradable elastomer. Tissue Eng 2005;11(1/2):302-9. [112] Kaazempur-Mofrad M, Vacanti J, Kamm RD. Computational modeling of blood flow and rheology in fractal microvascular networks. In: Bathe K-J, editor. Computational Fluid and Solid Mechanics. Kidlington, UK: Elsevier Science Ltd; 2001. p. 864 - 7.

[113] Kaihara S, Borenstein J, Koka R, Lalan S, Ochoa ER, Ravens M, et al. Silicon micromachining to tissue engineer branched vascular channels for liver fabrication. Tissue Eng 2000;6(2):105-17. [114] Borenstein J, Terai H, King K, Weinberg E, Kaazempur-Mofrad M, Vacanti J. Microfabrication technology for vascularized tissue engineering. Biomedical Microdevices 2002;4(3):167-75. [115] Shin M, Matsuda K, Ishii O, Terai H, Kaazempur-Mofrad M, Borenstein J, et al. Endothelialized networks with a vascular geometry in microfabricated poly(dimethyl siloxane). Biomed Microdevices 2004;6(4):269-78. [116] Moon JJ, Hahn MS, Kim I, Nsiah BA, West JL. Micropatterning of poly(ethylene glycol) diacrylate hydrogels with biomolecules to regulate and guide endothelial morphogenesis. Tissue Eng Part A 2009;15(3):579-85. [117] Chiang CK, Chowdhury MF, Iyer RK, Stanford WL, Radisic M. Engineering surfaces for site-specific vascular differentiation of mouse embryonic stem cells. Acta Biomater 2010;6(6):1904-16. [118] Zheng Y, Henderson PW, Choi NW, Bonassar LJ, Spector JA, Stroock AD. Microstructured templates for directed growth and vascularization of soft tissue in vivo. Biomaterials 2011;32(23):5391-401. [119] Arai K, Iwanaga S, Toda H, Genci C, Nishiyama Y, Nakamura M. Three-dimensional inkjet biofabrication based on designed images. Biofabrication 2011;3(3):034113. [120] Rouwkema J, Rivron NC, van Blitterswijk CA. Vascularization in tissue engineering. Trends Biotechnol 2008;26(8):434-41. [121] Gupta R, van Rooijen N, Sefton MV. Fate of endothelialized modular constructs implanted in an omental pouch in nude rats. Tissue Eng Part A 2009;15(10):2875-87.

Chapter 3 Literature Review: Wound healing, diabetes and the foreign body reaction 1 Cutaneous wound healing

Skin is the largest and most easily injured organ in the body. Its primary function of protecting against infection and excessive water loss requires an intact barrier. Consequently, repairing sites of injury is essential to maintaining its function[1,2]. Skin is composed of two main layers, the epidermis and dermis, separated by a basement membrane. The epidermis is a thin cellular layer composed of stratified keratinocytes and a layer of undifferentiated basal cells that contacts the basement membrane[3]. The epidermis is highly impermeable and acts as a barrier against infection and dehydration. The dermis is situated beneath the epidermis, and is composed mainly of collagen, with some elastin and glycosaminoglycans. The dermis is populated by fibroblasts, which secrete extracellular matrix (ECM) proteins and remodel the matrix through the production of proteases and collagenases. This layer provides strength and flexibility to skin, and is highly vascularized and innervated[3]. Dermal appendages, including hair follicles, sweat glands and sebaceous glands, help maintain homeostasis, sensory detection and thermoregulation[2].

Cutaneous wound healing has four overlapping phases: hemostasis, inflammation, tissue formation and tissue remodeling. Injury disrupts blood vessels and causes the extravasation of blood constituents into the wound site. Activation of platelets and coagulation cascades act to reestablish hemostasis and prevent excessive blood loss. The hemostatic fibrin clot also provides a temporary matrix that enables the migration of inflammatory cells into the wound bed. Platelets within the clot release growth factors and cytokines, such as vascular endothelial growth factor (VEGF), platelet-derived growth factor (PDGF) and transforming growth factor beta (TGF-β), that recruit neutrophils, macrophages and fibroblasts to the wound[4,5]. Neutrophils begin to accumulate at the wound site within hours of injury and begin the debridement of the dead tissue and removal of foreign particles and bacteria by phagocytosis. During this initial inflammatory phase (Figure 3-1), neutrophils also release pro-angiogenic factors, such as VEGF and interleukin-8 (IL-8)[5]. Typically, neutrophils undergo apoptosis within 24 – 48 h and are extruded in the eschar (a non-viable fibrin-rich matrix/clot) or phagocytosed by macrophages,

46 which replace neutrophils as the predominant cell type within the wound[5]. Monocytes are recruited from the blood and differentiate into macrophages upon entering the tissue. Macrophages remove necrotic debris, including apoptotic neutrophils, and release additional proinflammatory cytokines and chemokines, such as tumor necrosis factor alpha (TNF-α), IL-1β, IL-6 and monocyte chemoattractant protein 1 (MCP-1), which propagate the inflammatory state[6]. In addition to their phagocytic and inflammatory roles, macrophages also release a host of growth factors, including VEGF, basic fibroblast growth factor (bFGF), PDGF and TGF-β, which induce fibroblast and endothelial cell migration, proliferation and matrix production[7].

Unlike neutrophils, which in the absence of infection are not essential to wound healing, the depletion of macrophages in the wound significantly delays wound repair in adults[8,9]. Consequently, macrophages are considered a crucial component of successful adult wound healing. A recent study using a mouse model that allows conditional depletion of macrophages reported that depletion of macrophages during early (inflammation) and mid (tissue formation) stages of repair significantly impaired and altered wound healing, albeit through different roles[10]. The initial wave of macrophages recruited during the inflammatory phase played a critical role in inducing subsequent granulation tissue formation, vascularization, reepithelialization and eventual scar formation. Conversely, macrophages recruited during tissue formation (i.e. a second wave of infiltrating macrophages) appeared to promote blood vessel stability, as depletion of macrophages during this phase led to severe hemorrhage and endothelial apoptosis within the granulation tissue, potentially due to the reduction in VEGF and TGF-β expression within the macrophage-depleted wound beds[10]. This study also highlighted the role of inflammation in setting up subsequent tissue formation; however, excessive or prolonged inflammation can negatively affect wound healing and the resolution of the inflammatory phase is a critical step in successful tissue repair[11]. The mechanisms through which inflammation is resolved are not as well defined as mechanisms that perpetuate inflammation, but are thought to include the expression of anti-inflammatory factors, such as IL-10, TGF-β and soluble TNF-α receptor, or receptor downregulation in response to high ligand concentration[6,12]. Temporal profiles of macrophage phenotype suggest this transition from inflammation to tissue repair involved a shift in macrophage phenotype (reviewed in section 4) from a mixture of M1 (classical) and M2 (alternative) macrophages during early stages of wound healing, to a predominantly M2 population during later stages[13].

As inflammation subsides, fibroblasts migrate into the provisional matrix and begin to secrete new matrix proteins (i.e. tissue formation phase) (Figure 3-2). Gradually, the fibrin-rich provisional matrix is replaced by collagen-rich granulation tissue. The formation of granulation tissue relies heavily on neovascularization, which is required to support the increasing number of cells within the wound bed[4]. Once the wound has been filled with new tissue, many of the cells within the granulation tissue undergo apoptosis and the newly formed blood vessels regress, resulting in a relatively acellular, collagen-rich scar[4,14]. Thrombospondin (TSP)-1 and TSP- 2[4,15], Sprouty homologue 2 (Spry2)[16], C-X-C motif chemokine 10 (CXCL10; also called IP-10; interferon-inducible protein-10)[17] and other anti-angiogenic factors (e.g. angiostatin, endostatin) facilitate vessel regression[18,19].

Wound closure is achieved through two mechanisms: reepithelialization and wound contraction. Reepithelialization is a critical component in wound healing, and acts to reestablish the epidermal barrier to protect against infection and dehydration (See Santoro et al 2005[20] for an in-depth review). In healthy skin, keratinocytes form a quiescent epidermal tissue that is tightly anchored by desmosomes (cell-cell contacts) and hemidesmosomes (cell-substrate contacts)[21]. Within hours of injury, the process of reepithelialization is initiated. Keratinocytes along the border of the wound release the anchoring desmosomes and hemidesmosomes, and transition to a flattened, elongated phenotype with lamellipodia-like projections[20]. Migrating keratinocytes deposit unprocessed laminin-5 (LN-5; also called laminin-332) along the leading edge and upregulate the expression of integrins that bind unprocessed LN-5, fibronectin, vitronectin and tenascin (i.e. α3β1, α5β1, αvβ6 and αvβ5) to enable migration on the provisional

49 matrix[20,22,23]. Matrix metalloproteinase (MMP)-1, MMP-2, MMP-9 and MMP-10 are also secreted by the migrating keratinocytes to facilitate remodeling of the provisional matrix and migration[20]. As the keratinocytes migrate, they dissect the fibrin-rich clot (eschar) from the underlying granulation tissue[24]. Behind the migrating edge of keratinocytes, a zone of hyperproliferative keratinocytes feeds the growing epithelial tongue as it reestablishes the epidermal barrier. As the epithelial gap closes, the keratinocytes restore the laminin-rich basement membrane and reform the desmosome and hemidesmosome adhesions. A number of growth factors are key modulators of the migration and proliferation of keratinocytes. One of the most important factors in initiating the migration of keratinocytes is TGF-β, which stimulates the expression integrin subunits, ECM molecules and MMPs that promote keratinocyte migration [25-27]. Granulocyte-macrophage colony stimulating factor (GM-CSF) and epidermal growth factor (EGF)-receptor ligands (EGF; TGF-α; keratinocyte growth factor, KGF; and heparin- Binding EGF-Like Growth Factor; HB-EGF) play important roles in stimulating keratinocyte proliferation[27-29].

Contraction of the wound bed by myofibroblasts accelerates reepithelialization by bringing the wound edges closer together. As granulation tissue is formed, the collagen-rich matrix is remodeled by contractile myofibroblasts. The conventional view is that TGF-β (among other myofibroblast-inducing factors) induces fibroblast differentiation into myofibroblasts, which is marked by the expression of α-smooth muscle actin (αSMA) and enhanced ECM production[30,31]. However, other sources of myofibroblasts have been postulated, including pericytes or vascular smooth muscle cells surrounding vessels[32] and a subpopulation of bone marrow-derived cells called fibrocytes[31,33,34]. In renal fibrogenesis, injured epithelial cells were reported to undergo an epithelial-mesenchymal transition to give rise to myofibroblasts[35,36], however recent genetic fate mapping studies have disputed this idea[37- 39].

Once the epithelium is reestablished and the wound is closed, remodeling of the newly formed underlying tissue continues for weeks to months, as wound strength is gained and collagen is slowly deposited and remodeled into larger bundles with increased intermolecular crosslinks[4,40].

2 Wound healing and diabetes

There are many diseases and pathological conditions that disrupt the progression of wound healing, resulting in chronic wounds[11]. Diabetes mellitus is one of the most common comorbidities of chronic wounds [41], and impaired wound healing is now considered a hallmark of diabetes, with the lifetime risk of a person with diabetes developing a foot ulcer estimated to be between 15 – 25%[42]. Chronic diabetic wounds often result in diminished quality of life, limb loss and increased morbidity and mortality[41,43-48]. Amputation is a common consequence for patients with non-healing diabetic foot ulcers, with the risk increasing significantly with age[49]. As diabetes mellitus currently affects 366 million people worldwide[50], diabetic wound healing also represents a significant socio-economic burden now and in the future, as the incidence of diabetes rises.

In chronic diabetic wounds (e.g. foot ulcers), all the stages of wound healing are compromised. The underlying cause of wound healing impairment in diabetes remains unclear, although hyperglycemia and impaired insulin activity have been implicated[51-54]. An array of physiological processes and functions are altered in diabetic wound healing, including impaired angiogenesis, abnormal macrophage function and cytokine expression, prolonged inflammation, decreased collagen synthesis and increased levels of proteolytic enzymes[11,13]. The wound healing process is regulated by the expression of many molecular factors, many of which have multiple and overlapping functions. However, the number of growth factors, cytokines and receptors with altered expression in diabetic wounds, and the altered cellular response to these factors act to overwhelm the redundancy and impair wound healing through numerous mechanisms (reviewed by Blakytny, 2009)[48].

Diabetic wounds typically experience an extended, chronic inflammation phase. Excessive infiltration of neutrophils that release excessive MMPs and reactive oxygen species (ROS), impaired macrophage phagocytosis of apoptotic cells, and dysregulation of cytokine, growth factor and MMP expression act to prolong the inflammatory phase and impede the wound from proceeding through the subsequent phases of wound healing[55]. Keratinocytes from diabetic wounds have impaired migration, proliferation and differentiation, which delay the reepithelialization of the wound[48,53]. Fibroblasts from diabetic foot ulcers have impaired proliferation, altered morphology, an attenuated response to growth factors and decreased

51 collagen deposition[56], leading to impaired tissue formation. Aberrant cytokine expression within the diabetic wounds impairs essential downstream processes, including angiogenesis, collagen deposition and reepithelialization. The expression of multiple factors which direct these processes, including VEGF, insulin-like growth factor (IGF)-1 and -2, TGF-β and PDGF are decreased in diabetic wounds, while the expression of MMPs, TNF-α, TGF-β2, TGF-β3, and NOS is increased[48,57-59].

Another contributing factor to impaired diabetic wound healing is the reduced mobilization and engraftment of bone marrow progenitor cells[60] due, in part, to the reduced expression of SDF- 1 within the granulation tissue of diabetic wounds compared to non-diabetic wounds[61]. SDF-1 is one of the primary cytokines responsible for mobilizing and homing bone-marrow progenitor cells. Bone marrow progenitor cells are normally recruited to the wound site, where they contribute to tissue repair primarily by releasing paracrine factors that enhance wound healing and vascularization[62]. Administration of exogenous SDF-1 and intradermal injection of bone marrow derived progenitor cells have been reported to improve wound closure in diabetic animal models[61,62].

3 The foreign body reaction

The foreign body reaction (FBR) is a well-documented host response to implanted biomaterials, and has been the focus of a number of recent reviews[63-66]. Despite being inert and nontoxic, biomaterials can trigger inflammatory and fibrotic responses upon implantation, which usually result in the implant being walled-off from the host tissue by a fibrous capsule. The FBR follows a sequence of events similar to wound healing (hemostasis, inflammation, granulation tissue formation etc) with several important and distinct steps, which are highlighted below (Figure 3- 3).

Upon implantation, the biomaterial contacts blood leaked from damaged vessels, immediately leading to protein adsorption, activation of the complement and coagulation cascades and the activation and adhesion of platelets and leukocytes[67]. The resulting thrombus at the tissue/material interface acts as a provisional matrix, similar to that formed in wound healing. The layer of adsorbed protein modulates the interaction between the host inflammatory cells and the biomaterial, and the composition and configuration of the adsorbed proteins can dictate the

52 subsequent host response[68-73]. However, the composition of the protein layer can change over time as proteins adsorb and desorb (Vroman Effect)[66,74].

An important component of the host response to a foreign element is the complement system. The complement system consists of over 20 plasma proteins and can be activated by the classical, alternative and mannose-binding lectin pathways, all of which lead to the enzymatic cleavage of complement proteins C3 and C5, and the formation of the terminal complement complex (TCC)[75]. Biomaterials are conventionally thought to activate the complement system via the alternative activation pathway (Figure 3-4), in which the spontaneous hydrolysis of complement protein C3 in the fluid phase eventually gives rise to C3b, which is deposited covalently on the surface, and the formation of the C3 and C5 convertase complexes. Eventually,

C5b initiates the assembly of the TCC (also called the membrane attack complex) by binding C6, then C7, C8 and up to twelve C9 molecules (C5b-9). If attached to a cell surface, this complex can insert itself into the cell membrane and cause cell lysis or damage. TCC can also activate leukocytes, leading to a significant inflammatory response[67]. Certain surfaces, referred to as “non-activating” surfaces, increase the affinity of surface-bound C3b for Factor H, a complement regulatory protein that promotes C3b inactivation. Conversely, “activating” surfaces favour surface-bound C3b uptake of Factor B, and consequently promote the alternative pathway C3 amplification loop[76]. The complement system interacts with other important processes in the early host response to biomaterials, including platelet activation and the coagulation cascade, which stimulate an inflammatory response to the biomaterial[67].

As in wound healing, acute inflammation is characterized by the infiltration of neutrophils. Mast cells also mediate the acute response by degranulating and releasing range of molecules including histamine, fibrinogen, MCP-1 and IL-1β at the implant site[77-79]. The chemoattractants released by platelets and mast cells recruit monocytes/macrophages to the implant surface, which in turn propagate the chemoattractive signals[66]. The accumulation of macrophages at the implant signals the progression to chronic inflammation. Macrophages adhere to the protein layer on the surface of the biomaterial, and can fuse to form multinucleated foreign body giant cells (FBGC), a hallmark of the FBR. As fibroblasts infiltrate the surrounding wound site and granulation tissue formation begins, a layer of macrophages and FBGC forms around the implant, separating the biomaterial from the granulation tissue. Adherent macrophages and FBGC secrete a multitude of degradative factors (reactive oxygen

54 intermediates, enzymes and acid) into an isolated microenvironment between the cell and biomaterial surfaces in an attempt to break down the foreign material[66,80,81]. Within two to three weeks, the biomaterial and layer of FBGC and macrophages become encased in a fibrous capsule, which can persist for the life of the implant[81]. The onslaught of degradative factors released from adhered macrophages and FBGC, and the formation of the fibrous capsule can ultimately lead to device failure[66,81-83].

Many factors can influence the extent and duration of FBR and associated inflammation, including biomaterial surface properties. The composition of proteins in the adsorbed protein layer and their conformation are intimately linked to the biomaterial surface properties, and it is likely that the subsequent events of the FBR are dictated by the cells’ interaction with this unique protein layer[66,70,84,85]. Macrophage adhesion to the adsorbed proteins occurs primarily through integrin surface receptors. The β1 and β2 integrin families have been identified as critical mediators of macrophage adhesion to biomaterials[86,87]. Initial macrophage adhesion occurs through β2 integrins αMβ2 (also called Mac-1, CD11b/CD18) and αXβ2, which bind various adsorbed proteins including complement product iC3b, fibrinogen/fibrin and IgG, and induce the expression of integrin β1 subunit[88-90]. The β1 integrin family mediates subsequent macrophage adhesion and can bind multiple proteins, including fibronectin, laminin, collagen, thrombospondin and vitronectin[86,87,91]. Integrin binding leads to activation of focal adhesion kinase (FAK), src-family kinases and extracellular signal related kinase (ERK 1/2) signaling pathways. Differences in biomaterial surface properties, and consequently, their protein layers, can modulate cell adhesion, FAK and ERK phosphorylation, and downstream integrin signaling[68,92,93]. Biomaterial-adherent macrophages often fuse to form multinucleated FBGC to facilitate the degradation and resorption of the offending biomaterial (reviewed by Brodbeck and Anderson 2009[94]). While this process is not fully understood, it is well-established that macrophage fusion can be induced by IL-4 and IL-13 in vitro and in vivo, and requires the expression of mannose receptors, dendritic cell-specific transmembrane protein (DC-STAMP), chemokine (C-C motif) ligand 2 (CCL2), (CD44) and downregulation of osteopontin (OPN)[95- 100]. Vitronectin has also been identified as a critical ligand during IL-4 induced macrophage fusion in vitro[91].

The relationship between biomaterial surface chemistry and macrophage adhesion and fusion has been studied, however data indicating which surface chemistries promote adhesion and fusion,

55 and which do not, varies slightly among studies[101,102]. These variations were possibly due to differences in the selection of polymer coating to generate the desired surface chemistry (hydrophilic, hydrophobic, ionic, zwitterionic etc) and the methods of analysis (in vivo cage implant models, in vitro peripheral blood monocyte-derived macrophage fusion models). Generally, decreased cell adhesion and FBGC formation was reported for anionic, hydrophilic and zwitterionic surfaces, compared to neutral and hydrophobic surfaces[101,103-106]. Differences in cytokine and chemokine expression in adherent macrophages were also observed with different surfaces chemistries, however the interpretations of these data were highly dependent on what molecules were chosen for analysis[102,107].

The effect of biomaterial surface chemistry on wound healing has been studied using charged beads (positively charged diethylaminethanol (DEAE), negatively charge carboxymethyl (CM) and neutral cross-linked dextrose beads). Increased wound breaking strength and increased presence of macrophages, mast cells and FBGC were associated with positively charged beads[108-111]. No differences were observed among negatively charged or uncharged beads and untreated wounds[109]. Polymer fibers coated with methacrylic acid (MA) were associated with a trend of increased microvessel density in single fiber implant models, compared among positively charged fibers (N,N-dimethylaminoethyl methacrylate) and neutral fibers (hexafluoropropylene), however this increase was not statistically significant[112].

Surface topography and pore size also have a significant effect on the host response to a biomaterial, independent of surface chemistry[113]. In 1995, Brauker et al. studied the FBR towards more than 150 commercially available membranes and reported that pore size played a significant role in directing the host response; membranes with 0.8 - 8 µm pores exhibited enhanced neovascularization at the membrane-tissue interface[114]. Subsequent studies have demonstrated improved tissue integration and neovascularization of three-dimensional scaffolds with interconnected pores, although the range of ideal pore sizes varies among studies[115-118]. Similarly, surface topography has been reported to modulate macrophage activation and FBR[119-122].

4 Macrophage activation/polarization and its role in wound healing and foreign body reaction

Recently, macrophage activation has received considerable attention due to the positive and negative roles macrophages play in innate immune responses, disease modulation, tissue remodeling, and host responses to biomaterials (reviewed in [123-125]). Macrophages are monocyte-derived myeloid cells that originate from the bone marrow. Upon entering the bloodstream, monocytes may circulate for several days before entering tissues and differentiating into macrophages. This process may be part of the steady-state turn over of tissue resident macrophages or may be driven by chemotactic factors released during inflammation. Macrophages are a heterogeneous population of cells, with tissue-resident macrophages having specific phenotypes related to their particular microenvironment and role[126]. Heterogeneity is also observed in the context of inflammation, where macrophages enter the tissue and become activated in response to signals present at the site of inflammation.

In an oversimplified nomenclature, activated macrophages are broadly classified as M1 (classically activated) or M2 (alternatively activated), emulating the nomenclature used to describe T-cell activation (i.e. Th1 and Th2). Perhaps a more realistic depiction of macrophage activation was proposed by Mosser et al, which views macrophage phenotypes along a continuum between the M1 and M2 extremes and depicts them using a “colour wheel” metaphor[127]. The M1/M2 classification system represents two extremes of macrophage activation, with M1 often viewed as detrimental and M2 as beneficial in the context of wound healing and biomaterials. Although useful, it is a simplistic characterization of what occurs in vivo.

Ignoring these subtleties, M1 macrophages can be induced by interferon gamma (INF-γ), lipopolysaccharide (LPS) or TNF-α, and, as effectors and inducers of a Th1 immune response, have enhanced microbicidal capacity and a pro-inflammatory secretion profile (Figure 3-5) [123]. In general, characteristic M1 macrophages express IL-12high, IL-23high, IL-10low, metabolize arginine to form nitric oxide (NO) via inducible nitric oxide synthase (iNOS) and express IL-1β, IL-6 and TNFα[123,128]. The pro-inflammatory cytokines (IL-1, IL-6, TNF-α, IL-12) and nitric oxide (NO) produced by M1 macrophages are important in host defense against

57 infection, however they can also be very damaging to the host tissue in the case of a prolonged inflammatory response.

Figure 3-5. M1 (classical) and M2 (alternative) macrophage phenotypes. Distinct macrophage phenotypes can arise in response to specific inducers. M1 macrophages are induced by INF-γ, LPS or TNF-α, while M2 macrophages are divided into three sub- phenotypes, M2a, M2b and M2c based on the route of activation. Abbreviations: DTH, delayed-type hypersensitivity; IC, immune complexes; MR, mannose receptor; PTX3, the long pentraxin PTX3; RNI, reactive nitrogen intermediates; ROI, reactive oxygen intermediates; SLAM, signaling lymphocytic activation molecule; SRs, scavenger receptors; TLR, Toll-like receptor[128]. Reprinted from [128] Copyright 2004, with permission from Elsevier.

Again ignoring the subtleties, the M2 designation is used to describe all non-classically activated macrophages. In contrast to M1 macrophages, M2 macrophages generally express IL-12low, IL- 23low, IL-10high, arginase (which converts arginine into polyamines and ornithine rather than NO), have high expression of mannose receptors, scavenger receptors and galactose receptors; and are active in Th2-type responses[123]. M2 macrophages are considered to play a greater role in wound healing and tissue repair due to high endocytotic activity to facilitate clearance of extracellular pathogens, enhanced anti-inflammatory cytokine production and reduced pro- inflammatory cytokine and NO production[123]. To account for variations among M2 macrophages, this group is further divided into three sub-groups: M2a (induced by IL-4 or IL-

13), M2b (induced by exposure to immune complexes with IL-1β or LPS) and M2c (induced by IL-10, TGF-β or glucocorticoids)(reviewed in [128,129]).

Macrophage phenotype is also plastic, transitioning between M1-like and M2-like phenotypes in response to environmental signals and signal crosstalk produced during inflammatory responses[130,131]. In vitro, macrophage phagocytosis of apoptotic cells led to a decrease in the expression of TNF-α following stimulation with INF-γ or LPS compared to stimulated macrophages that had not participated in phagocytosis, suggesting that phagocytosis may mediate the shift towards a less inflammatory phenotype[55]. This finding may have important implications for diabetic wound healing, as the authors suggested that the impairment of macrophage phagocytic activity in diabetic wounds may delay the transition from M1 to M2 phenotypes, thereby delaying the resolution of inflammation.

In a recent review, Brown et al. discuss the role of macrophage polarization in disease pathogenesis, tissue injury and host response to biomaterials[123]. Retaining the simple nomenclature, the authors suggest that tumor associated macrophages often have a predominately M1-like phenotype early in tumor formation, then transition to a more M2-like phenotype as the tumor matures. Conversely, atherosclerotic plaques possess a heterogeneous, but predominately M2-like population initially, which are thought to contribute to the lesion progression and growth[123]. The phenotype then shifts to a predominately M1-like phenotype, a process thought to be caused by phagocytosis of low-density lipoproteins and INF-γ releasing- Th1 lymphocytes[123]. In wound healing, a similar phenotypic switch is required to resolve the inflammatory phase and progress to a tissue formation and remodeling phase[13]. As described earlier, gene expression in skin biopsies taken from cutaneous wounds showed that the early, inflammatory phase had an upregulated cluster of genes associated with both M1 and M2 genes, while during the tissue formation phase a predominately M2 gene cluster was upregulated[13].

5 Summary

Cutaneous wound healing is a complex, highly orchestrated series of events that is critical to maintaining homeostasis. Macrophages play a central role in directing the wound healing response, particularly during inflammation, tissue formation and angiogenesis. When a foreign object, such as an implanted biomaterial, is present at the site of injury, the wound healing

59 response is altered to include a host response called the foreign body reaction. Macrophages and FBGC drive the foreign body reaction, which often results in the implanted biomaterial being encapsulated in a fibrous tissue and walled off from the body. Material surface properties, and the surface adsorbed protein layer play an important role in dictating the extent of the foreign body reaction. An underlying theme to wound healing and the foreign body reaction is the phenotype of activated macrophages at the wound or implant site. While the M1/M2 classification system of macrophage phenotype is a simplistic representation of the broad range of macrophage phenotypes observed in vivo, the resolution of inflammation and the progression to a “more” M2-like phenotype is an important step in successful wound healing.

6 References [1] Metcalfe AD, Ferguson MW. Tissue engineering of replacement skin: the crossroads of biomaterials, wound healing, embryonic development, stem cells and regeneration. J R Soc Interface 2007;4(14):413-37. [2] Adolphe C, Wainwright B. Pathways to improving skin regeneration. Expert Rev Mol Med 2005;7(20):1-14. [3] Zhong SP, Zhang YZ, Lim CT. Tissue scaffolds for skin wound healing and dermal reconstruction. Wiley Interdiscip Rev Nanomed Nanobiotechnol 2010;2(5):510-25. [4] Singer AJ, Clark RA. Cutaneous wound healing. N Engl J Med 1999;341(10):738-46. [5] Eming S, Brachvogel B, Odorisio T, Koch M. Regulation of angiogenesis: wound healing as a model. Prog Histochem Cytochem 2007;42(3):115-70. [6] Eming SA, Krieg T, Davidson JM. Inflammation in wound repair: molecular and cellular mechanisms. J Invest Dermatol 2007;127(3):514-25. [7] Li J, Chen J, Kirsner R. Pathophysiology of acute wound healing. Clin Dermatol 2007;25(1):9-18. [8] Simpson DM, Ross R. The neutrophilic leukocyte in wound repair a study with antineutrophil serum. J Clin Invest 1972;51(8):2009-23. [9] Leibovich SJ, Ross R. The role of the macrophage in wound repair. A study with hydrocortisone and antimacrophage serum. Am J Pathol 1975;78(1):71-100. [10] Lucas T, Waisman A, Ranjan R, Roes J, Krieg T, Muller W, et al. Differential roles of macrophages in diverse phases of skin repair. J Immunol 2010;184(7):3964-77. [11] Menke NB, Ward KR, Witten TM, Bonchev DG, Diegelmann RF. Impaired wound healing. Clin Dermatol 2007;25(1):19-25. [12] Sato Y, Ohshima T, Kondo T. Regulatory role of endogenous interleukin-10 in cutaneous inflammatory response of murine wound healing. Biochem Biophys Res Commun 1999;265(1):194-9.

[13] Deonarine K, Panelli MC, Stashower ME, Jin P, Smith K, Slade HB, et al. Gene expression profiling of cutaneous wound healing. J Transl Med 2007;5:11. [14] Desmouliere A, Redard M, Darby I, Gabbiani G. Apoptosis mediates the decrease in cellularity during the transition between granulation tissue and scar. Am J Pathol 1995;146(1):56-66. [15] Ilan N, Mahooti S, Madri JA. Distinct signal transduction pathways are utilized during the tube formation and survival phases of in vitro angiogenesis. J Cell Sci 1998;111 ( Pt 24):3621-31. [16] Wietecha MS, Chen L, Ranzer MJ, Anderson K, Ying C, Patel TB, et al. Sprouty2 downregulates angiogenesis during mouse skin wound healing. Am J Physiol Heart Circ Physiol 2011;300(2):H459-H67. [17] Bodnar RJ, Yates CC, Rodgers ME, Du X, Wells A. IP-10 induces dissociation of newly formed blood vessels. J Cell Sci 2009;122(12):2064-77. [18] Carmeliet P. Angiogenesis in health and disease. Nat Med 2003;9(6):653-60. [19] Folkman J. Angiogenesis and angiogenesis inhibition: an overview. Exs 1997;79:1-8. [20] Santoro MM, Gaudino G. Cellular and molecular facets of keratinocyte reepithelization during wound healing. Exp Cell Res 2005;304(1):274-86. [21] Green KJ, Jones JC. Desmosomes and hemidesmosomes: structure and function of molecular components. Faseb J 1996;10(8):871-81. [22] Haapasalmi K, Zhang K, Tonnesen M, Olerud J, Sheppard D, Salo T, et al. Keratinocytes in human wounds express alpha v beta 6 integrin. J Invest Dermatol 1996;106(1):42-8. [23] Reynolds LE, Conti FJ, Silva R, Robinson SD, Iyer V, Rudling R, et al. Alpha3beta1 integrin-controlled Smad7 regulates reepithelialization during wound healing in mice. J Clin Invest 2008;118(3):965-74. [24] Kubo M, Van de Water L, Plantefaber LC, Mosesson MW, Simon M, Tonnesen MG, et al. Fibrinogen and fibrin are anti-adhesive for keratinocytes: a mechanism for fibrin eschar slough during wound repair. J Invest Dermatol 2001;117(6):1369-81. [25] Weber CE, Li NY, Wai PY, Kuo PC. Epithelial-mesenchymal transition, TGF-beta, and osteopontin in wound healing and tissue remodeling after injury. J Burn Care Res 2012;33(3):311-8. [26] Zambruno G, Marchisio PC, Marconi A, Vaschieri C, Melchiori A, Giannetti A, et al. Transforming growth factor-beta 1 modulates beta 1 and beta 5 integrin receptors and induces the de novo expression of the alpha v beta 6 heterodimer in normal human keratinocytes: implications for wound healing. J Cell Biol 1995;129(3):853-65. [27] Wei W, Barron PD, Rheinwald JG. Modulation of TGF-beta-inducible hypermotility by EGF and other factors in human prostate epithelial cells and keratinocytes. In Vitro Cell Dev Biol Anim 2010;46(10):841-55. [28] Mann A, Breuhahn K, Schirmacher P, Blessing M. Keratinocyte-derived granulocyte- macrophage colony stimulating factor accelerates wound healing: stimulation of keratinocyte proliferation, granulation tissue formation, and vascularization. J Invest Dermatol 2001;117(6):1382-90.

[29] Stoll SW, Rittie L, Johnson JL, Elder JT. Heparin-binding EGF-like growth factor promotes epithelial-mesenchymal transition in human keratinocytes. J Invest Dermatol 2012. [30] Hinz B, Phan SH, Thannickal VJ, Prunotto M, Desmouliere A, Varga J, et al. Recent developments in myofibroblast biology: paradigms for connective tissue remodeling. Am J Pathol 2012;180(4):1340-55. [31] Reilkoff RA, Bucala R, Herzog EL. Fibrocytes: emerging effector cells in chronic inflammation. Nat Rev Immunol 2011;11(6):427-35. [32] Juniantito V, Izawa T, Yuasa T, Ichikawa C, Yamamoto E, Kuwamura M, et al. Immunophenotypical analyses of myofibroblasts in rat excisional wound healing: possible transdifferentiation of blood vessel pericytes and perifollicular dermal sheath cells into myofibroblasts. Histol Histopathol 2012;27(4):515-27. [33] Kao HK, Chen B, Murphy GF, Li Q, Orgill DP, Guo L. Peripheral blood fibrocytes: enhancement of wound healing by cell proliferation, re-epithelialization, contraction, and angiogenesis. Ann Surg 2011;254(6):1066-74. [34] Wada T, Sakai N, Matsushima K, Kaneko S. Fibrocytes: a new insight into kidney fibrosis. Kidney Int 2007;72(3):269-73. [35] Iwano M, Plieth D, Danoff TM, Xue C, Okada H, Neilson EG. Evidence that fibroblasts derive from epithelium during tissue fibrosis. J Clin Invest 2002;110(3):341-50. [36] Kalluri R, Weinberg RA. The basics of epithelial-mesenchymal transition. J Clin Invest 2009;119(6):1420-8. [37] Humphreys BD, Valerius MT, Kobayashi A, Mugford JW, Soeung S, Duffield JS, et al. Intrinsic epithelial cells repair the kidney after injury. Cell Stem Cell 2008;2(3):284-91. [38] Humphreys BD, Lin SL, Kobayashi A, Hudson TE, Nowlin BT, Bonventre JV, et al. Fate tracing reveals the pericyte and not epithelial origin of myofibroblasts in kidney fibrosis. Am J Pathol 2010;176(1):85-97. [39] Koesters R, Kaissling B, Lehir M, Picard N, Theilig F, Gebhardt R, et al. Tubular overexpression of transforming growth factor-beta1 induces autophagy and fibrosis but not mesenchymal transition of renal epithelial cells. Am J Pathol 2010;177(2):632-43. [40] Pierce GF, Vande Berg J, Rudolph R, Tarpley J, Mustoe TA. Platelet-derived growth factor-BB and transforming growth factor beta 1 selectively modulate glycosaminoglycans, collagen, and myofibroblasts in excisional wounds. Am J Pathol 1991;138(3):629-46. [41] Escandon J, Vivas AC, Tang J, Rowland KJ, Kirsner RS. High mortality in patients with chronic wounds. Wound Repair Regen 2011;19(4):526-8. [42] Singh N, Armstrong DG, Lipsky BA. Preventing foot ulcers in patients with diabetes. JAMA 2005;293(2):217-28. [43] Moss SE, Klein R, Klein BE. The 14-year incidence of lower-extremity amputations in a diabetic population. The Wisconsin Epidemiologic Study of Diabetic Retinopathy. Diabetes Care 1999;22(6):951-9.

[44] Ramsey SD, Newton K, Blough D, McCulloch DK, Sandhu N, Reiber GE, et al. Incidence, outcomes, and cost of foot ulcers in patients with diabetes. Diabetes Care 1999;22(3):382-7. [45] Vileikyte L. Diabetic foot ulcers: a quality of life issue. Diabetes Metab Res Rev 2001;17(4):246-9. [46] Nabuurs-Franssen MH, Huijberts MS, Nieuwenhuijzen Kruseman AC, Willems J, Schaper NC. Health-related quality of life of diabetic foot ulcer patients and their caregivers. Diabetologia 2005;48(9):1906-10. [47] Valensi P, Girod I, Baron F, Moreau-Defarges T, Guillon P. Quality of life and clinical correlates in patients with diabetic foot ulcers. Diabetes Metab 2005;31(3 Pt 1):263-71. [48] Blakytny R, Jude EB. Altered molecular mechanisms of diabetic foot ulcers. Int J Low Extrem Wounds 2009;8(2):95-104. [49] Beckman JA, Creager MA, Libby P. Diabetes and atherosclerosis: epidemiology, pathophysiology, and management. JAMA 2002;287(19):2570-81. [50] Whiting DR, Guariguata L, Weil C, Shaw J. IDF diabetes atlas: global estimates of the prevalence of diabetes for 2011 and 2030. Diabetes Res Clin Pract 2011;94(3):311-21. [51] Brownlee M. Biochemistry and molecular cell biology of diabetic complications. Nature 2001;414(6865):813-20. [52] Thangarajah H, Yao D, Chang EI, Shi Y, Jazayeri L, Vial IN, et al. The molecular basis for impaired hypoxia-induced VEGF expression in diabetic tissues. Proc Natl Acad Sci U S A 2009;106(32):13505-10. [53] Velander P, Theopold C, Hirsch T, Bleiziffer O, Zuhaili B, Fossum M, et al. Impaired wound healing in an acute diabetic pig model and the effects of local hyperglycemia. Wound Repair Regen 2008;16(2):288-93. [54] Aghdam SY, Eming SA, Willenborg S, Neuhaus B, Niessen CM, Partridge L, et al. Vascular endothelial insulin/IGF-1 signaling controls skin wound vascularization. Biochem Biophys Res Commun 2012. [55] Khanna S, Biswas S, Shang Y, Collard E, Azad A, Kauh C, et al. Macrophage dysfunction impairs resolution of inflammation in the wounds of diabetic mice. PLoS One 2010;5(3):e9539. [56] Hansen SL, Myers CA, Charboneau A, Young DM, Boudreau N. HoxD3 accelerates wound healing in diabetic mice. Am J Pathol 2003;163(6):2421-31. [57] Rafehi H, El-Osta A, Karagiannis TC. Genetic and epigenetic events in diabetic wound healing. Int Wound J 2011;8(1):12-21. [58] Zykova SN, Jenssen TG, Berdal M, Olsen R, Myklebust R, Seljelid R. Altered cytokine and nitric oxide secretion in vitro by macrophages from diabetic type II-like db/db mice. Diabetes 2000;49(9):1451-8. [59] Brown DL, Kane CD, Chernausek SD, Greenhalgh DG. Differential expression and localization of insulin-like growth factors I and II in cutaneous wounds of diabetic and nondiabetic mice. Am J Pathol 1997;151(3):715-24.

[60] Fiorina P, Pietramaggiori G, Scherer SS, Jurewicz M, Mathews JC, Vergani A, et al. The mobilization and effect of endogenous bone marrow progenitor cells in diabetic wound healing. Cell Transplant 2010;19(11):1369-81. [61] Gallagher KA, Liu Z-J, Xiao M, Chen H, Goldstein LJ, Buerk DG, et al. Diabetic impairments in NO-mediated endothelial progenitor cell mobilization and homing are reversed by hyperoxia and SDF-1α. J Clin Invest 2007;117(5):1249-59. [62] Suh W, Kim KL, Kim JM, Shin IS, Lee YS, Lee JY, et al. Transplantation of endothelial progenitor cells accelerates dermal wound healing with increased recruitment of monocytes/macrophages and neovascularization. Stem Cells 2005;23(10):1571-8. [63] Morais JM, Papadimitrakopoulos F, Burgess DJ. Biomaterials/tissue interactions: possible solutions to overcome foreign body response. AAPS J 2010;12(2):188-96. [64] Franz S, Rammelt S, Scharnweber D, Simon JC. Immune responses to implants - a review of the implications for the design of immunomodulatory biomaterials. Biomaterials 2011;32(28):6692-709. [65] Anderson JM, McNally AK. Biocompatibility of implants: lymphocyte/macrophage interactions. Semin Immunopathol 2011;33(3):221-33. [66] Anderson JM, Rodriguez A, Chang DT. Foreign body reaction to biomaterials. Semin Immunol 2008;20(2):86-100. [67] Gorbet MB, Sefton MV. Biomaterial-associated thrombosis: roles of coagulation factors, complement, platelets and leukocytes. Biomaterials 2004;25(26):5681-703. [68] Diekjurgen D, Astashkina A, Grainger DW, Holt D, Brooks AE. Cultured primary macrophage activation by lipopolysaccharide depends on adsorbed protein c omposition and substrate surface chemistry. J Biomater Sci Polym Ed 2011;(in press). [69] Lindon JN, McManama G, Kushner L, Merrill EW, Salzman EW. Does the conformation of adsorbed fibrinogen dictate platelet interactions with artificial surfaces? Blood 1986;68(2):355-62. [70] Sivaraman B, Latour RA. The relationship between platelet adhesion on surfaces and the structure versus the amount of adsorbed fibrinogen. Biomaterials 2010;31(5):832-9. [71] Jenney CR, Anderson JM. Adsorbed serum proteins responsible for surface dependent human macrophage behavior. J Biomed Mater Res 2000;49(4):435-47. [72] Hu WJ, Eaton JW, Ugarova TP, Tang L. Molecular basis of biomaterial-mediated foreign body reactions. Blood 2001;98(4):1231-8. [73] Wilson CJ, Clegg RE, Leavesley DI, Pearcy MJ. Mediation of biomaterial-cell interactions by adsorbed proteins: a review. Tissue Eng 2005;11(1-2):1-18. [74] Vroman L, Adams AL, Fischer GC, Munoz PC. Interaction of high molecular weight kininogen, factor XII, and fibrinogen in plasma at interfaces. Blood 1980;55(1):156-9. [75] Nilsson B, Ekdahl KN, Mollnes TE, Lambris JD. The role of complement in biomaterial- induced inflammation. Mol Immunol 2007;44(1-3):82-94. [76] Kazatchkine MD, Carreno MP. Activation of the complement system at the interface between blood and artificial surfaces. Biomaterials 1988;9(1):30-5.

[77] Zdolsek J, Eaton JW, Tang L. Histamine release and fibrinogen adsorption mediate acute inflammatory responses to biomaterial implants in humans. J Transl Med 2007;5:31. [78] Thevenot PT, Baker DW, Weng H, Sun M-W, Tang L. The pivotal role of fibrocytes and mast cells in mediating fibrotic reactions to biomaterials. Biomaterials 2011;32(33):8394-403. [79] Orenstein SB, Saberski ER, Klueh U, Kreutzer DL, Novitsky YW. Effects of mast cell modulation on early host response to implanted synthetic meshes. Hernia 2010;14(5):511-6. [80] 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. J Biomed Mater Res 1991;25(2):177-83. [81] Kotanen CN, Moussy FG, Carrara S, Guiseppi-Elie A. Implantable enzyme amperometric biosensors. Biosens Bioelectron 2012. [82] Anderson JM. Biological responses to materials. Ann Rev Mater Res 2001;31:81-110. [83] Shanbhag AS, Jacobs JJ, Glant TT, Gilbert JL, Black J, Galante JO. Composition and morphology of wear debris in failed uncemented total hip replacement. J Bone Joint Surg Br 1994;76(1):60-7. [84] Kamath S, Bhattacharyya D, Padukudru C, Timmons RB, Tang L. Surface chemistry influences implant‐mediated host tissue responses. J Biomed Mater Res 2008;86A(3):617-26. [85] Xia Z, Triffitt JT. A review on macrophage responses to biomaterials. Biomed Mater 2006;1(1):R1-R9. [86] McNally AK, Anderson JM. Beta1 and beta2 integrins mediate adhesion during macrophage fusion and multinucleated foreign body giant cell formation. Am J Pathol 2002;160(2):621-30. [87] McNally AK, Macewan SR, Anderson JM. Alpha subunit partners to beta1 and beta2 integrins during IL-4-induced foreign body giant cell formation. J Biomed Mater Res A 2007;82(3):568-74. [88] Berton G, Lowell CA. Integrin signalling in neutrophils and macrophages. Cell Signal 1999;11(9):621-35. [89] Werr J, Eriksson EE, Hedqvist P, Lindbom L. Engagement of beta2 integrins induces surface expression of beta1 integrin receptors in human neutrophils. J Leukoc Biol 2000;68(4):553-60. [90] McNally AK, Anderson JM. Complement C3 participation in monocyte adhesion to different surfaces. Proc Natl Acad Sci U S A 1994;91(21):10119-23. [91] McNally AK, Jones JA, MacEwan SR, Colton E, Anderson JM. Vitronectin is a critical protein adhesion substrate for IL‐4‐induced foreign body giant cell formation. J Biomed Mater Res 2008;86A(2):535-43. [92] Allen LT, Tosetto M, Miller IS, O’Connor DP, Penney SC, Lynch I, et al. Surface- induced changes in protein adsorption and implications for cellular phenotypic responses to surface interaction. Biomaterials 2006;27(16):3096-108.

[93] Keselowsky BG, Collard DM, García AJ. Integrin binding specificity regulates biomaterial surface chemistry effects on cell differentiation. Proc Natl Acad Sci U S A 2005;102(17):5953-7. [94] Brodbeck WG, Anderson JM. Giant cell formation and function. Curr Opin Hematol 2009;16(1):53-7. [95] Tsai AT, Rice J, Scatena M, Liaw L, Ratner BD, Giachelli CM. The role of osteopontin in foreign body giant cell formation. Biomaterials 2005;26(29):5835-43. [96] DeFife KM, Jenney CR, McNally AK, Colton E, Anderson JM. Interleukin-13 induces human monocyte/macrophage fusion and macrophage mannose receptor expression. J Immunol 1997;158(7):3385-90. [97] Yagi M, Miyamoto T, Sawatani Y, Iwamoto K, Hosogane N, Fujita N, et al. DC-STAMP is essential for cell-cell fusion in osteoclasts and foreign body giant cells. J Exp Med 2005;202(3):345-51. [98] Kyriakides TR, Foster MJ, Keeney GE, Tsai A, Giachelli CM, Clark-Lewis I, et al. The CC chemokine ligand, CCL2/MCP1, participates in macrophage fusion and foreign body giant cell formation. Am J Pathol 2004;165(6):2157-66. [99] Helming L, Gordon S. Macrophage fusion induced by IL-4 alternative activation is a multistage process involving multiple target molecules. Eur J Immunol 2007;37(1):33- 42. [100] Cui W, Ke JZ, Zhang Q, Ke HZ, Chalouni C, Vignery A. The intracellular domain of CD44 promotes the fusion of macrophages. Blood 2006;107(2):796-805. [101] Jones JA, Dadsetan M, Collier TO, Ebert M, Stokes KS, Ward RS, et al. Macrophage behavior on surface-modified polyurethanes. J Biomater Sci Polym Ed 2004;15(5):567- 84. [102] Khandwekar A, Rho CK. Modulation of cellular responses on engineered polyurethane implants. J Biomed Mater Res A 2012. [103] Brodbeck WG, Patel J, Voskerician G, Christenson E, Shive MS, Nakayama Y, et al. Biomaterial adherent macrophage apoptosis is increased by hydrophilic and anionic substrates in vivo. Proc Natl Acad Sci U S A 2002;99(16):10287-92. [104] Shen M, Garcia I, Maier RV, Horbett TA. Effects of adsorbed proteins and surface chemistry on foreign body giant cell formation, tumor necrosis factor alpha release and procoagulant activity of monocytes. J Biomed Mater Res A 2004;70(4):533-41. [105] Desai NP, Hubbell JA. Tissue response to intraperitoneal implants of polyethylene oxide- modified polyethylene terephthalate. Biomaterials 1992;13(8):505-10. [106] Sawada S, Sakaki S, Iwasaki Y, Nakabayashi N, Ishihara K. Suppression of the inflammatory response from adherent cells on phospholipid polymers. J Biomed Mater Res A 2003;64(3):411-6. [107] Jones JA, Chang DT, Meyerson H, Colton E, Kwon IK, Matsuda T, et al. Proteomic analysis and quantification of cytokines and chemokines from biomaterial surface- adherent macrophages and foreign body giant cells. J Biomed Mater Res A 2007;83(3):585-96.

[108] Burgess E, Hollinger J, Bennett S, Schmitt J, Buck D, Shannon R, et al. Charged beads enhance cutaneous wound healing in rhesus non-human primates. Plast Reconstr Surg 1998;102(7):2395-403. [109] Wu L, Mockros NE, Casperson ME, Gruskin EA, Ladin DA, Roth SI, et al. Effects of electrically charged particles in enhancement of rat wound healing. J Surg Res 1999;85(1):43-50. [110] Sasaki A, Mueller RV, Xi G, Sipe R, Buck D, Hollinger J. Mast Cells: An unexpected finding in the modulation of cutaneous wound repair by charged beads. Plast Reconstr Surg 2003;111(4):1446-53. [111] Mustoe TA, Weber DA, Krukowski M. Enhanced healing of cutaneous wounds in rats using beads with positively charged surfaces. Plast Reconstr Surg 1992;89(5):891-7; discussion 8-9. [112] Sanders JE, Lamont SE, Karchin A, Golledge SL, Ratner BD. Fibro-porous meshes made from polyurethane micro-fibers: effects of surface charge on tissue response. Biomaterials 2005;26(7):813-8. [113] Chen S, Jones JA, Xu Y, Low H-Y, Anderson JM, Leong KW. Characterization of topographical effects on macrophage behavior in a foreign body response model. Biomaterials 2010;31(13):3479-91. [114] Brauker JH, Carr-Brendel VE, Martinson LA, Crudele J, Johnston WD, Johnson RC. Neovascularization of synthetic membranes directed by membrane microarchitecture. J Biomed Mater Res 1995;29(12):1517-24. [115] Fukano Y, Usui ML, Underwood RA, Isenhath S, Marshall AJ, Hauch KD, et al. Epidermal and dermal integration into sphere-templated porous poly(2-hydroxyethyl methacrylate) implants in mice. J Biomed Mater Res A 2010;94(4):1172-86. [116] Chiu Y-C, Cheng M-H, Engel H, Kao S-W, Larson JC, Gupta S, et al. The role of pore size on vascularization and tissue remodeling in PEG hydrogels. Biomaterials 2011;32(26):6045-51. [117] Bai F, Wang Z, Lu J, Liu J, Chen G, Lv R, et al. The correlation between the internal structure and vascularization of controllable porous bioceramic materials in vivo: a quantitative study. Tissue Eng Part A 2010;16(12):3791-803. [118] Madden LR, Mortisen DJ, Sussman EM, Dupras SK, Fugate JA, Cuy JL, et al. Proangiogenic scaffolds as functional templates for cardiac tissue engineering. Proc Natl Acad Sci USA 2010;107(34):15211-6. [119] Cao H, McHugh K, Chew SY, Anderson JM. The topographical effect of electrospun nanofibrous scaffolds on the in vivo and in vitro foreign body reaction. J Biomed Mater Res A 2010;93(3):1151-9. [120] Collie AMB, Bota PCS, Johns RE, Maier RV, Stayton PS. Differential monocyte/macrophage interleukin-1β production due to biomaterial topography requires the β2 integrin signaling pathway. J Biomed Mater Res A 2010;96A(1):162-9. [121] Bota PC, Collie AM, Puolakkainen P, Vernon RB, Sage EH, Ratner BD, et al. Biomaterial topography alters healing in vivo and monocyte/macrophage activation in vitro. J Biomed Mater Res A 2010;95(2):649-57.

[122] Saino E, Focarete ML, Gualandi C, Emanuele E, Cornaglia AI, Imbriani M, et al. Effect of electrospun fiber diameter and alignment on macrophage activation and secretion of proinflammatory cytokines and chemokines. Biomacromolecules 2011;12(5):1900-11. [123] Brown BN, Ratner BD, Goodman SB, Amar S, Badylak SF. Macrophage polarization: an opportunity for improved outcomes in biomaterials and regenerative medicine. Biomaterials 2012;33(15):3792-802. [124] Biswas SK, Mantovani A. Orchestration of metabolism by macrophages. Cell Metab 2012;15(4):432-7. [125] Martinez FO, Helming L, Gordon S. Alternative activation of macrophages: an immunologic functional perspective. Annu Rev Immunol 2009;27(1):451-83. [126] Gordon S, Taylor PR. Monocyte and macrophage heterogeneity. Nat Rev Immunol 2005;5(12):953-64. [127] Mosser DM, Edwards JP. Exploring the full spectrum of macrophage activation. Nat Rev Immunol 2008;8(12):958-69. [128] Mantovani A, Sica A, Sozzani S, Allavena P, Vecchi A, Locati M. The chemokine system in diverse forms of macrophage activation and polarization. Trends Immunol 2004;25(12):677-86. [129] Kou PM, Babensee JE. Macrophage and dendritic cell phenotypic diversity in the context of biomaterials. J Biomed Mater Res 2010;96A(1):239-60. [130] Porcheray F, Viaud S, Rimaniol AC, Leone C, Samah B, Dereuddre-Bosquet N, et al. Macrophage activation switching: an asset for the resolution of inflammation. Clin Exp Immunol 2005;142(3):481-9. [131] Stout RD, Jiang C, Matta B, Tietzel I, Watkins SK, Suttles J. Macrophages sequentially change their functional phenotype in response to changes in microenvironmental influences. J Immunol 2005;175(1):342-9.

Chapter 4 The expression of sonic hedgehog in diabetic wounds following treatment with poly(methacrylic acid-co-methyl methacrylate)beads1 Abstract

The expression of native sonic hedgehog (Shh) was significantly increased in poly(methacrylic acid-co-methyl methacrylate) bead (MAA) treated wounds at day 4 compared to both poly(methyl methacrylate) bead (PMMA) treated and untreated wounds in diabetic db/db mice.

MAA beads also increased the expression of the Shh transcription factor Gli3 at day 4.

Previously, topical application of MAA beads (45 mol % methacrylic acid) improved wound closure and blood vessel density in excisional wounds in these mice, while PMMA beads did not.

Gene expression within the granulation tissue of healing wounds was studied to provide insight into the mechanism of vessel formation and wound healing in the presence of MAA beads. In addition to the increased expression of Shh, MAA-treated wounds had increased expression of osteopontin (OPN), IL-1β and TNF-α, (at day 7) similar to the previously reported MAA response of macrophage-like and endothelial cells in vitro.

1 Reprinted from Biomaterials, 33(21), Fitzpatrick LE, Lisovsky A and Sefton MV, The expression of sonic hedgehog in diabetic wounds following treatment with poly(methacrylic acid-co-methyl methacrylate) beads, 5297- 307, Copyright 2012, with permission from Elsevier.

1 Introduction

The formation of new blood vessels (angiogenesis) is an essential process for tissue regeneration and tissue engineering. In wound healing, impaired angiogenesis is associated with the delayed healing of chronic wounds, including diabetic wounds[1,2]. Poly(methacrylic acid-co- methyl methacrylate) (MAA) beads have beneficial effects in skin graft and diabetic wound healing models by promoting neovascularization of the surrounding tissue[3,4]. In vitro, macrophage- like cells (dTHP-1) increased the expression of selected cytokines, while having little to no effect on the expression of pro-angiogenic growth factors[5], suggesting that MAA beads exerts their effect, in part, by modulating the inflammatory response in macrophages.

In order to better understand the effect of MAA beads in vivo, and connect the in vitro response to the angiogenic effect MAA beads have in wound healing models, the expression of seventeen genes within the granulation tissue of diabetic wounds treated with MAA beads was analyzed, including genes involved in wound healing [interleukin(IL)-1β, IL-6, tumor necrosis factor alpha (TNF-α) and osteopontin (OPN)] and the regulation of angiogenesis [vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF), platelet-derived growth factor B (PDGF-B), thrombospondin 1 (TSP-1), Sprouty 2 (Spry2) and C-X-C motif chemokine 10 (CXCL10)]. The expression of sonic hedgehog (Shh) and its pathway proteins [transmembrane receptors Patched (Ptch)1, Ptch2, and Smoothened (Smo); and transcription factors Gli1, Gli2 and Gli3] were also analyzed as preliminary microarray data implicated Shh involvement in the host response to MAA beads (unpublished data). While best known as a developmental morphogen, Shh also contributes to post-natal angiogenesis[6] and upon delivery or overexpression (via gene therapy) has been shown to accelerate wound healing in diabetic mice[7,8]. The present study suggests that MAA beads modulate the inflammatory response and the Shh pathway, resulting in increased wound vascularization.

2 Materials and methods 2.1 Bead preparation

MAA beads (150 - 250 µm in diameter, 45 mol % methacrylic acid) were obtained from Rimon Therapeutics (Toronto, Canada) and PMMA control beads (same diameter, 100 mol % methyl methacrylate) were acquired from Polysciences (Warrington, PA, USA). MAA and PMMA bead

70 preparation[5] and characterization[3,4] are described elsewhere. Briefly, beads were washed in a sonicating water bath for 20 minutes in 95% ethanol (MAA beads) or 1N HCl (PMMA beads) ten times, then rinsed in endotoxin-free water five times. After drying in a vacuum, beads were tested for endotoxin using LAL Pyrochrome Endotoxin Kit (Associates of Cape Cod, Falmouth, MA). Only beads with less than 0.25 EU/ml (100 mg beads/ml) were used in vivo for wound treatment.

2.2 Wound assay

7-9 week old, genetically diabetic, male BKS.Cg-m+/+ Leprdb/J mice (db/db) were ordered from Jackson Laboratories (Bar Harbor, ME, USA) with the approval of the University of Toronto Animal Care Committee and were housed under sterile conditions in the Department of Comparative Studies (DCM) Animal facility. Mice were allowed to acclimatize for one week prior to surgery. Mouse age (8 – 12 weeks), weight and blood glucose levels (OneTouch® UltraMini Blood Glucose Monitoring System, LifeScan Canada Ltd., BC) were recorded at time of surgery.

Mice were anaesthetized with isoflurane. The dorsal surface of the mouse was shaved with an electric shaver, followed by treatment with a hair removal cream (Veet). The surgical site was sterilized with Betadine and 70% ethanol. A sterile acetate template was used to mark two wound areas, approximately 4 cm caudal to the ears on the dorsum. Scissors were used to create the mid-dorsal, 7.5 mm x 7.5 mm full-thickness, bilateral wounds by excising the epidermis and dermis, including the panniculus carnosus. Either 7 mg of dry MAA beads or 7 mg of dry PMMA beads were applied topically to the wound beds, or they were left untreated. Upon hydration, MAA beads swell while PMMA beads do not. Each mouse received the same treatment for both wounds (e.g., left and right wounds were both treated with MAA beads); one wound was used for histology and the other was used for gene expression analysis. The wounds were left undressed and scabs were allowed to form. Following the surgical procedure, buprenorphine (0.03 mg/kg) was given subcutaneously for analgesia. Thereafter, the mice were caged individually. Mice were sacrificed using CO2 asphyxiation, followed by cervical dislocation, on day 4 and 7 post-wounding. A total of 24 animals were used, with four animals per treatment per time point.

2.3 Histology and immunohistochemistry

The wound tissue, including approximately 2 mm of healthy skin surrounding the wound and the underlying fat layer, was excised from the mice following euthanasia, and was fixed in formalin. Tissue samples were embedded in paraffin blocks and then cut into 4 µm sections. Sections were processed and stained with hematoxylin and eosin (H&E; Fisher Scientific) and for CD31 (SC- 1506-R, 1/2000, rabbit polyclonal; Santa Cruz Biotechnology, USA). Histology slides were scanned (20x) using the Aperio ScanScope XT (Aperio Technologies, USA) at the Advanced Optical Microscopy Facility (AOMF, Toronto, Canada).

2.4 Vessel counts

The images of scanned CD31 slides were analyzed by a blinded investigator, using the Aperio ImageScope (Version 10). Six vessel “hot spots” within the granulation tissue were selected for each sample, with the criteria that three hotspots be located at the edge of the granulation tissue and three be within the middle of the granulation tissue. Counting windows (300 µm x 300 µm) were draw around the hot spots and the number of CD31 positive vessel-like structures were counted and the area of granulation tissue was calculated for each window. Criteria for counting a vessel-like structure were that it had to (1) be positive for CD31, (2) have a lumen (tissue section perpendicular to the vessel) or an elongated structure (tissue section parallel to the vessel), and (3) be within the granulation tissue. The vessel density was calculated by dividing the number of vessels by the area of the granulation tissue within the window (i.e. the area of “non-granulation” tissue was subtracted from the total area of the window). Tissue samples from four mice were analyzed for each treatment. Results are presented as the average vessel density ± standard deviation.

2.5 Total RNA preparation

The wound beds were excised and trimmed of any visible fat and healthy skin. The tissue was then quartered and placed in RNAlater RNA Stabilization Reagent (Qiagen Inc, Mississauga, ON, Canada) and kept at 4 oC for no more than 10 days. Total RNA was isolated using TRIzol reagent (Invitrogen, Burlington, ON) according to manufacturer’s protocol. The isolated RNA was further purified using the Qiagen RNeasy Mini kit (Qiagen) according to the manufacturer’s Clean-up protocol with on-column DNAse digestion. The concentration of RNA was determined

72 by measuring the absorbance at 260 nm using a Nanodrop ND-1000 spectrophometer (NanoDrop Technologies, Inc., USA). All RNA samples had a 260/280 ratio greater than 2.0.

2.6 Quantitative real-time PCR

Gene-specific primer sequences (Table 4-1) were either selected from Primerbank (http://pga.mgh.harvard.edu/primerbank/) or designed using Primer-BLAST (http://www.ncbi.nlm.nih.gov/tools/primer-blast/) and were synthesized by Sigma Genosys (Sigma-Aldrich Canada Ltd, Oakville, ON, Canada). All primer pairs were analyzed by Primer- BLAST to ensure specificity for the intended target gene within the human genome. The rationale for assaying these molecules is given in the discussion.

Total RNA was used for cDNA synthesis, which was performed using a Superscript III First- Strand Synthesis SuperMix (Invitrogen Canada Inc, Burlington, ON, Canada) according to manufacturer’s protocol. Quantitative real-time PCR (qPCR) reactions were carried out in a 7900HT Fast Real-Time PCR system (Applied Biosystems Canada, Streetsville, ON, Canada) using the following thermal profile: 50 oC for 2 minutes, 95 oC for 10 minutes, then 40 cycles of 95 oC for 15 second, 60 oC for 30 seconds and 72 oC for 30 seconds. A dissociation curve was generated for each plate following the amplification cycles. qPCR reactions were in a final volume of 10 µL, containing 5 µL of SYBR Green Master Mix (Applied Biosystems) 1 µL of 10µM forward and reversed primers and 10 ng of cDNA template. 7900HT Sequence Detection Systems (SDS) Software (Version 2.3, Applied Biosystems) was used to determine the crossing point values (Cp). The PCR efficiencies were calculated using LinRegPCR software (Version 11.0, download: http://LinRegPCR.HFRC.nl). The relative fold change in gene expression was calculated for each animal using the equation described by Pfaffl (2001)[9], and results were presented as the individual fold changes for each animal relative to a control group (PMMA or no bead) and the median fold changes (n = 4).

2.7 Statistical analysis

Statistical analysis of the data was performed using SPSS PASW Statistics 18.0 (release 18.0.0). Statistical comparisons among treatment groups (MAA, PMMA and no bead) were performed using analysis of variance (ANOVA), with α = 0.05. Levene’s test for homogeneity was used to test for equal variance among samples. Tukey’s honestly significant difference (HSD) post-hoc

73 analysis was used to compare the mean vessel density among treatments. Games-Howell post- hoc analysis was used for gene expression data, as many genes had unequal variances.

Table 4-1. Quantitative real-time PCR primer sequences. Probe Forward Primer Sequence Reverse Primer Sequence GenBank Accession # bFGF GCGACCCACACGTCAAACTA TCCCTTGATAGACACAACTCCTC NM_008006 CXCL10 CCAAGTGCTGCCGTCATTTTC GGCTCGCAGGGATGATTTCAA NM_021274 GAPDH AGGTCGGTGTGAACGGATTTG TGTAGACCATGTAGTTGAGGTCA NM_008084 Gli1 CCAAGCCAACTTTATGTCAGGG AGCCCGCTTCTTTGTTAATTTGA NM_010296 Gli2 CAACGCCTACTCTCCCAGAC GAGCCTTGATGTACTGTACCAC BC031171 Gli3 CACAGCTCTACGGCGACTG CTGCATAGTGATTGCGTTTCTTC NM_008130 IL-1β CTGCTGGTGTGTGACGTTCCCAT GGTCCGACAGCACGAGGCTTT NM_008361 IL-6 TAGTCCTTCCTACCCCAATTTCC TTGGTCCTTAGCCACTCCTTC NM_031168 OPN AGCAAGAAACTCTTCCAAGCAA GTGAGATTCGTCAGATTCATCCG NM_009263 PDGF-B AAGTGTGAGACAATAGTGACCCC CATGGGTGTGCTTAAACTTTCG NM_011057 Ptch1 AAAGAACTGCGGCAAGTTTTTG CTTCTCCTATCTTCTGACGGGT NM_008957 Ptch2 CTCCGCACCTCATATCCTAGC TCCCAGGAAGAGCACTTTGC NM_008958 Shh AAAGCTGACCCCTTTAGCCTA TTCGGAGTTTCTTGTGATCTTCC NM_009170 Smo GAGCGTAGCTTCCGGGACTA CTGGGCCGATTCTTGATCTCA NM_176996 Spry2 TCCAAGAGATGCCCTTACCCA GCAGACCGTGGAGTCTTTCA NM_011897 TSP-1 GGGGAGATAACGGTGTGTTTG CGGGGATCAGGTTGGCATT NM_011580 TNF-α CCCTCACACTCAGATCATCTTCT GCTACGACGTGGGCTACAG NM_013693 VEGF GCACATAGAGAGAATGAGCTTCC CTCCGCTCTGAACAAGGCT NM_009505 3 Results 3.1 Effect of MAA beads on vessel density

Previous studies using large 1.5 cm x 1.5 cm cutaneous wounds in diabetic mice demonstrated that topical application of MAA beads improved wound closure and increased wound bed vessel density, compared to treatment with PMMA or no beads[4]. Preliminary microarray data (data not shown), from day 14 wounds, suggested that the mechanism of MAA-mediated angiogenesis may be further elucidated through measurement of differential gene expression within the wound tissue. Gene expression during early wound healing was of particular interest as this was when angiogenic signaling and vessel ingrowth was thought likely to be occurring. However, obtaining sufficient amount of RNA from the large diabetic wound beds at early times (e.g. day 4) was problematic. Consequently, a new wound model was employed, which used small, bilateral

74 wounds. The small (7.5 mm x 7.5 mm) wounds used in this study were not as severe as the large wounds, and consequently, had enough granulation tissue formed at day 4 to enable excision of the wound bed and isolation of RNA. The weights and blood glucose levels of the mice were recorded at the time of surgery (Table 4-2).

Table 4-2. Mouse age, weight and blood glucose levels at time of surgery. Age (weeks) Weight (g) RBG (mmol/L)a Day 4 Animals 10 - 12 46.2 ± 2.3 32.1 ± 2.5 Day 7 Animals 8 - 10 37.7 ± 2.1 28.1 ± 5.9 aMice with RBG above 33.3 mmol/L registered as “high” on the glucometer. In average RBG calculation, “high” was given a numerical value of 34 mmol/L.

The day 4 explant group was approximately two weeks older than the day 7 explant group, and weighed significantly more (p < 0.001). Consistent with the age, the glucose levels were also higher for the day 4 explant group (p = 0.043), with seven mice registering as HIGH (> 33.3 mmol/L, assigned value of 34 mmol/L for calculated average and statistics) while day 7 mice had only two mice registering as HIGH. The implication of these differences is discussed below.

Histological sections stained with H&E or for CD31 showed the presence of vascularized granulation tissue 7 days post-wounding (Figure 4-1). Vessel counts from day 7 histology sections (Figure 4-2) showed that MAA-treated wounds had significantly higher (p < 0.05) CD31+ vessel density within the granulation tissue, compared to wounds treated with PMMA beads or no beads. The average vessel density for PMMA treated wounds was lower than untreated wounds, although the difference was not significant (p = 0.085). The total counting area (counting window area minus non-granulation tissue area) was similar for each treatment (p = 0.353). Histology sections from the day 4 mice (not shown) indicated that vessel ingrowth into the granulation tissue had not occurred by day 4 and consequently, the microvessel density in the granulation tissue was not quantified.

3.2 Gene expression analysis

Gene expression within the granulation tissue was analyzed using qPCR. The amplification efficiencies for all primer sets were above 0.935 and the Cp values, as determined by SDS software, were between cycle 13 and 36. As described previously[5], changes in gene expression associated with the presence of either MAA or PMMA beads were determined using two

75 metrics: the normalized relative quantity (NRQ) and the relative fold change (FC) in gene expression.

The NRQ is the quantity of a given gene (amount of cDNA) relative to the quantity of GAPDH present in the same sample. The NRQ was log-transformed to a linear scale (log2NRQ) and a one-way ANOVA was performed to determine if the treatment (MAA, PMMA, no bead) had a statistically significant effect on the expression of each gene. The log2NRQ values also give a general sense of the relative abundance of different genes (corrected for amplification efficiency). For example, the average log2NRQ of Shh is approximately -14.5, while the average log2NRQ for OPN is -0.4, indicating that there are more cDNA copies of OPN than Shh. The log2NRQ data and the p-values from the ANOVA for each gene are summarized in Table 4-3

The fold change (FC) is the ratio of an NRQ of the “treated” group to the mean NRQ of the

“control” group (i.e. NRQMAA, animal 1 / mean_NRQno bead). The fold changes describe the effect MAA beads (MAA/no bead) or PMMA beads (PMMA /no bead) had on the gene expression within the granulation tissue of the diabetic mice. To highlight the difference in response to the methacrylic acid-containing beads, the effect of MAA beads on gene expression relative to PMMA genes (MAA / PMMA) was also reported. Genes with a fold change greater than 1.5 (or less than -1.5) and a p-value less than 0.05 were considered significant (marked by asterisk in figures). In certain cases, high fold changes were not significant (p > 0.05) due to high variability within a treatment group (see Table 4-3 for standard error of log2NRQ values). However, the high fold changes indicate these genes may still be of biological importance, despite the conservative statistical test. Since this study is an initial look at the molecular events in the MAA bead associated angiogenesis process, we chose to include these high p-value but high fold change molecules in our discussion.

3.3 Effect of MAA beads on sonic hedgehog expression

Motivated by earlier microarray results from MAA-treated diabetic wounds, the expression of Shh mRNA was analyzed using qPCR (Figure 4-3). In day 4 MAA-treated wounds, Shh was expressed 4.6–fold higher and 5.5–fold higher than in PMMA-treated (p = 0.001) and untreated (p < 0.001) wounds, respectively. PMMA beads did not affect Shh expression at day 4 (p = 0.306) compared to no beads. None of the treatments had an effect on Shh expression on day 7

(FC < 1.5, p > 0.494). The average Shh log2NRQ was approximately -14.5, which was the lowest value of all the genes (i.e. the least abundant mRNA).

Table 4-3. Mean log2(NRQ) values and ANOVA for real-time PCR analysis of gene expression in the granulation tissue of wounds treated with MAA beads, PMMA beads or no beads. p < 0.05 (bolded) was considered significant, n = 4. Gene aDaya MAA No bead PMMA p-values (Games-Howell) MAA MAA PMMA compared to compared to compared to log2NRQ ± SE log2NRQ ± SE log2NRQ ± SE No bead PMMA No bead bFGF 4 -8.73 ± 0.16 -9.47 ± 0.18 -9.40 ± 0.44 0.192 0.086 0.980 7 -11.58 ± 0.10 -11.59 ± 0.22 -11.24 ± 0.25 1.000 0.459 0.568 CXCL10 4 -7.34 ± 0.46 -6.92 ± 0.44 -6.51 ± 0.45 0.781 0.348 0.671 7 -7.21 ± 0.39 -6.77 ± 0.34 -7.07 ± 0.23 0.694 0.949 0.768 Gli1 4 -12.48 ± 0.25 -13.13 ± 0.11 -13.17 ± 0.18 0.328 0.156 0.993 7 -11.44 ± 0.31 -11.27 ± 0.34 -10.89 ± 0.25 0.930 0.424 0.677 Gli2 4 -12.23 ± 0.32 -12.67 ± 0.06 -12.68 ± 0.08 0.696 0.493 1.000 7 -10.26 ± 0.13 -10.19 ± 0.26 -10.12 ± 0.19 0.967 0.835 0.981 Gli3 4 -8.37 ± 0.29 -9.27 ± 0.16 -9.46 ± 0.29 0.157 0.052 0.848 7 -7.82 ± 0.05 -7.98 ± 0.47 -7.90 ± 0.11 0.941 0.838 0.982 IL-1β 4 -1.36 ± 0.15 -0.98 ± 0.14 -1.36 ± 0.20 0.758 1.000 0.752 7 -2.55 ± 0.24 -3.85 ± 0.78 -3.50 ± 0.14 0.359 0.047 0.901 IL-6 4 -5.91 ± 0.37 -6.18 ± 0.37 -6.42 ± 0.56 0.896 0.513 0.899 7 -6.50 ± 0.67 -7.56 ± 0.33 -7.91 ± 0.26 0.410 0.240 0.704 OPN 4 -0.96 ± 0.15 -1.05 ± 0.22 -1.07 ± 0.25 0.952 0.869 0.999 7 0.56 ± 0.17 0.06 ± 0.39 -0.12 ± 0.14 0.513 0.049 0.904 PDGF-B 4 -7.96 ± 0.30 -8.49 ± 0.11 -8.89 ± 0.27 0.510 0.105 0.598 7 -6.18 ± 0.13 -6.23 ± 0.16 -6.52 ± 0.11 0.957 0.200 0.365 Ptch1 4 -10.42 ± 0.22 -10.97 ± 0.04 -10.99 ± 0.16 0.353 0.168 0.998 7 -8.29 ± 0.15 -8.68 ± 0.40 -8.36 ± 0.18 0.656 0.942 0.762 Ptch2 4 -9.73 ± 0.28 -10.46 ± 0.14 -10.45 ± 0.33 0.216 0.172 0.999 7 -10.39 ± 0.15 -10.70 ± 0.31 -10.15 ± 0.12 0.673 0.488 0.335 Shh 4 -12.07 ± 0.17 -14.68 ± 0.55 -14.11 ± 0.83 0.002 0.005 0.408 7 -14.96 ± 0.33 -15.29 ± 0.32 -15.61 ± 0.50 0.763 0.559 0.852 Smo 4 -7.72 ± 0.17 -8.08 ± 0.05 -8.20 ± 0.11 0.544 0.128 0.907 7 -6.90 ± 0.17 -6.81 ± 0.21 -6.96 ± 0.15 0.945 0.958 0.836 Spry2 4 -6.85 ± 0.14 -7.69 ± 0.21 -7.87 ± 0.32 0.009 0.007 0.591 7 -8.29 ± 0.14 -7.13 ± 1.99 -9.36 ± 0.17 0.840 0.007 0.567 TSP-1 4 -0.54 ± 0.30 -1.06 ± 0.27 -1.52 ± 0.52 0.599 0.081 0.591 7 -0.62 ± 0.30 -3.17 ± 1.98 -1.58 ± 0.13 0.495 0.091 0.727 TNF-α 4 -4.88 ± 0.14 -4.71 ± 0.10 -4.65 ± 0.10 0.870 0.561 0.988 7 -7.03 ± 0.24 -7.87 ± 0.75 -8.29 ± 0.14 0.589 0.017 0.855 VEGF 4 -6.67 ± 0.12 -6.34 ± 0.18 -6.60 ± 0.17 0.210 0.888 0.335 7 -6.95 ± 0.34 -7.39 ± 0.22 -6.78 ± 0.33 0.555 0.935 0.347

Upon finding an increase in Shh expression at day 4, the gene expression of six proteins involved in the Shh signaling pathway was also studied: Gli1, Gli2, Gli3, Ptch1, Ptch2 and Smo (Figure 4- 4). All proteins had increased expression (FC > 1.5) in MAA-treated wounds compared to PMMA-treated or untreated wounds on day 4, except Gli2 and Smo compared to no bead. Transcription factor Gli3 had the highest fold change (greater than 2) for MAA compared to both PMMA (p = 0.052) and no bead (p = 0.157). Although the increase in Gli3 expression compared to PMMA fell just outside our criterion for significance (p = 0.052) using the Games-Howell post-hoc test, the change was significant using Tukey HSD post-hoc test (p = 0.033). No other gene expression differences were significant, by either test. On day 7, the wounds treated with MAA, PMMA or no beads had similar expression of all six pathway proteins except Gli1 which was expressed 1.8-fold lower in MAA treated wounds than PMMA treated wounds (p = 0.677).

Figure 4-3. Sonic hedgehog gene expression in diabetic wounds treated with MAA beads. On post-operative day 4, Shh expression was increased approximately 5-fold in MAA- treated wounds compared to PMMA-treated and untreated wounds (p < 0.05). By day 7, the expression of Shh in all wounds was similar (FC < 1.5). PMMA did not affect the expression of Shh compared to no beads. Circles represent individual animals (n =4) and the bar represents the median. Dashed lines mark +1.5 and -1.5 fold change. *, p < 0.05.

3.4 Effect of MAA beads on cytokine expression

MAA beads increased the expression of IL-1β, IL-6, TNF-α and OPN on day 7 compared to PMMA beads and no beads (fold changes > 1.5, Figure 4-5). However, only increases in IL-1β, TNF-α and OPN compared to PMMA were significant (p < 0.05), despite large fold changes in these genes in MAA compared to no bead.

For example, IL-1β expression was increased 3.3-fold or 1.9-fold in MAA treated wounds, compared to no bead (p = 0.359) and PMMA (p = 0.049), respectively. This is likely due to the high variability in IL-1β expression within the no bead treatment group at day 7 (See Table 4-3). For the other molecules, fold changes were about 2 on day 7 with MAA beads, but there was no effect on the expression of IL-1β, IL-6, TNF-α and OPN on day 4. PMMA beads decreased the expression of IL-1β and IL-6 1.6-fold compared to no beads on day 4, although these changes also had high p-values.

3.5 Effect of MAA beads on genes involved in regulating angiogenesis

The expression of three growth factors with pro-angiogenic effects (bFGF, VEGF and PDGF-B) and three proteins with anti-angiogenic effects (CXCL10, Spry2 and TSP-1) were studied to determine if the balance between pro- and anti-angiogenic factors in the wound tissue favoured vessel formation (i.e. more pro-angiogenic signaling) (Figure 4-6). Unexpectedly, the only significant (p < 0.05) change was the expression of Spry2, which was 1.7-fold and 1.8-fold higher in MAA treated wounds than in PMMA-treated wounds or untreated wounds respectively at day 4, and 1.6-fold higher than PMMA on day 7. The fold increase relative to untreated on day 7 was similar (FC = 1.9) but p > 0.05 (again, likely due to high variability in expression in the untreated group). Changes in bFGF, PDGF-B, TSP-1 and CXCL10 were also observed, however these changes were not significant (refer to Table 4-3 for p-values). Briefly, on day 4 MAA beads increased the expression of bFGF 1.5-fold compared to PMMA, and PDGF-B 2.3-fold and 1.6-fold, compared to both PMMA and no beads, respectively. Again with p > 0.05, the expression of CXCL10 was 1.7-fold lower in MAA-treated wounds compared to untreated wound on day 4 and day 7, and 2.1-fold lower compared to PMMA-treated wounds on day 4. The expression of TSP-1 in MAA-treated wounds was 1.7-fold higher than PMMA on both days, and 1.5-fold higher in untreated wounds on day 7 only. Bead treatment had no effect on VEGF expression at either time point.

4 Discussion

The motivation for this study was the question of how the synthetic methacrylic acid-containing beads were altering the cellular response within diabetic wounds to give rise to the improved wound healing and increased angiogenesis previously reported by Martin et al.[4]. In earlier studies, we also showed that the beads were angiogenic in a skin graft model in healthy rats[3] and that in culture, macrophage-like cells and endothelial cells have a distinct response at the gene expression level to MAA beads, compared to PMMA beads and no beads[5]. The next step in elucidating the cellular mechanism through which MAA beads elicit their effect was to see if we could detect changes in gene expression within the granulation tissue of excisional cutaneous wounds in diabetic mice, similar to those reported on before[4] and then relate the detected changes to the in vitro results observed with MAA and macrophage-like cells.

Figure 4-6. Gene expression of proteins with pro- or anti-angiogenic effects. MAA beads significantly increased (FC > 1.5, p < 0.05) the expression of Spry2 compared to PMMA and no beads on day 4, and compared to PMMA on day 7. Changes were also observed in bFGF, PDGF-B, CXCL10 and TSP-1, however these were not significant. Briefly, MAA beads increased the expression bFGF 1.5-fold compared to PMMA on day 4, and PDGF-B was 2.3-fold and 1.6-fold higher in MAA treated wounds than PMMA and no bead. CXCL10 expression was decreased by MAA beads compared to untreated (no bead, day 4 and 7) and PMMA-treated (day 4) wounds. Neither MAA nor PMMA beads affected the expression of VEGF compared to no bead. Conversely, MAA beads increased the expression TSP-1 compared to PMMA-treated and untreated wounds on day 4 and 7 (except TSP-1, MAA / no bead, day 4). PMMA treated wounds had slightly less (1.6-fold decrease) TSP-1 on day 4 than untreated wounds. Circles represent individual animals (n =4) and the bar represents the median. Dashed lines mark +1.5 and -1.5 fold change.*, p < 0.05.

The initial series of experiments used cDNA microarrays to identify genes of interest (data not shown). Although a small number of key genes (e.g. Shh, CXCL10) were identified using this high throughput analysis method, the approach did not yield the expected large number of potential genes, presumably due to the subtle effect the beads have on gene expression (i.e. most fold changes were very small, FC < 3). Consequently, further gene expression analysis was performed using qPCR to validate genes of interest from the microarrays (Shh, CXCL10) and to analyze the expression of genes highlighted in in vitro studies (IL-1β, IL-6, TNF-α, OPN). The list was also expanded to include genes encoding proteins involved in the regulation of angiogenic activity (VEGF, bFGF, PDGF-B, Spry2 and TSP-1) to determine if gene expression indicated a more angiogenesis-inducing environment: i.e. an increase in expression of angiogenesis inducers or a decrease in inhibitors, which would cause the overall balance of angiogenesis regulators to favour new blood vessel formation.

The db/db mouse strain used in this study is a commonly used murine model for diabetic wound healing due to their delayed wound healing, severe hyperglycemia and obesity [10-14]. The rationale for selecting day 4 and day 7 as explant time points was that we were interested in the gene expression early in wound healing when the angiogenic signaling was commencing, and a slightly later time when vessels would be present in the wound bed. Based on the diabetic db/db wound healing literature[1,13], day 4 and day 7 were deemed appropriate time points.

The differences in weight and random blood glucose levels between the day 4 and 7 explant groups were due to the small increase (~2 week) in age. Consequently, the older day 4 group had noticeably larger fat deposits underlying the skin at the time of explant compared to the day 7 mice. As adipose tissue is a known source of stem cells and factors that have been shown to enhance wound healing and angiogenesis[15-17], it is necessary to be cautious in directly comparing the day 4 and day 7 time points and inferring a time course of activation. We are following up on these results in other models to address the question of time.

4.1 The effect of MAA beads on the vascularization of small, excisional cutaneous wounds in diabetic mice

The topical application of MAA beads to small (7.5 mm x 7.5 mm) cutaneous wounds increased the density of CD31+ vessels at day 7 roughly 25% and 60% compared to untreated and PMMA- treated wounds, respectively. These results were consistent with the 35% increase in vessel

84 density at day 7 and 14 in large (1.5 cm x 1.5 cm) MAA treated wounds, compared to PMMA and no bead groups[4]. The lack of vascular ingrowth in the granulation tissue four days after wounding was consistent with other excisional wound studies in this mouse strain[1] .

4.2 The role of sonic hedgehog in wound healing

The most dramatic effect MAA beads had on gene expression was the roughly 5-fold increase in the expression of Shh at day 4, compared to PMMA beads and no beads. This was accompanied by a modest increase in the expression of Shh-pathway proteins, of which only transcription factor Gli3 was significant (when using Tukey’s HSD). While Shh is best known for its roles in cancer and tissue patterning during embryogenesis (reviewed by [18-20]), recent work has highlighted the role of Shh in post-natal angiogenesis and tissue repair [8,21,22]. Disruption of the Shh signaling cascade profoundly impaired wound healing in excisional cutaneous wound models[22] and angiogenesis in hind-limb ischemia models[21], while exogenous addition of Shh improved wound healing in diabetic animals[7,8]. There is evidence that Shh indirectly induces angiogenesis by promoting fibroblasts to express pro-angiogenic cytokines, such as VEGF, angiopoietin-1, insulin-like growth factor 1 (IGF-1) and stromal cell-derived factor-1 (SDF-1)[7,21,23,24], although Shh-induced VEGF expression is contentious[25]. It can also promote angiogenic activity in endothelial cells (EC) directly via non-classical Rho kinase signaling pathways to promote in vitro capillary morphogenesis and migration, and induce EC expression of matrix metalloproteinase 9 (MMP-9) and OPN[26].

The classical Shh signaling cascade is activated when Shh binds its receptor Ptch, which releases the Ptch-mediated repression of Smo. Smo is then free to transduce the Shh signal, which eventually results in the stabilization of Gli transcription factors and the expression of Gli transcriptional target genes (reviewed by [20]). Gli1 and Gli2 act as Shh activators, while Gli3 is known to have Shh repressor activity in developmental processes[27]. However, there is evidence that Gli3 is a positive regulator of post-natal angiogenesis[6]. Gli3-encoding adenovirus injections increased the microvessel density and perfusion in hind-limb ischemia models[6]. Furthermore, Gli3-deficient mice have reduced microvessel density in ischemic models and impaired angiogenesis in corneal angiogenesis models; suggesting Gli3 is a regulator of post- natal angiogenesis[6].

4.3 The effect of MAA beads on the expression of cytokines and growth factors involved with wound healing and angiogenesis

Presumably, the increased expression of Shh at day 4 played a role in the increased vascularization in mice treated with MAA beads, compared to PMMA beads or no beads. However, it is unclear how Shh leads to new vessels in the context of the foreign body reaction to MAA beads. Some clues are provided by the other molecules that are differentially produced in the wound.

As macrophages are a principal responder to biomaterial implants[28] and also play an important role in wound healing[29], they were hypothesized to be principle mediators of the pro- angiogenic effect of MAA beads in vivo. There was some similarity in macrophage gene expression in vitro and the global gene expression in the granulation tissue in vivo, suggesting macrophages are involved in the host response to MAA beads. However, differences (in vitro, MAA beads decreased OPN expression in dTHP-1 cells but increased expression of OPN in endothelial cells[5]; while in vivo there was an increase in OPN expression) reflect the greater complexity of the in vivo situation and that gene expression is the summation of the expression of multiple cell types in the complex and dynamic wound healing environment.

There are multiple cases where MAA beads increased (or decreased) the expression of a gene (FC > 1.5), but the change was not statistically significant (p > 0.05). The lack of statistical significance is likely a reflection of high variability that occurs in wound healing models, and increasing the number of animals may decrease the p-values for some of the genes. For many genes which had relatively high fold changes but p > 0.05, there was a high degree of variability within one or more treatment group (see Table 4-3 for the standard error of the mean log2NRQ). Consequently, we did not completely discount the involvement of factors with increased (or decreased) gene expression, but large p-values, as we believe they merit further investigation. The remaining discussion speculates on the role the four cytokines and six angiogenic regulators may play in MAA-induced angiogenesis, including genes having non-significant changes (FC > |1.5| and p > 0.05).

The MAA beads increased the expression of IL-1β, IL-6, TNF-α and OPN at day 7, compared to PMMA beads and no beads. This suggests that the inflammatory response may be involved in

86 the host response to MAA beads; however, considering the increase in microvessel density at day 7, this is not necessarily an undesirable outcome. As previously discussed by Fitzpatrick et al.[5], IL-1β, IL-6 and TNF-α are all involved in the regulation of angiogenesis, although they are typically referred to simply as pro-inflammatory cytokines. In brief, IL-1β induces macrophages to express pro-angiogenic factors[30], IL-6 induces endothelial cell migration in vitro [31], and TNF-α is a key molecule inducing “tip cell” phenotype in endothelial cells, which leads to rapid sprouting following inflammation[32].

The increased expression of OPN in the wound bed is of particular interest due to the interplay between Shh and OPN signaling[26], and it is possible that the increased OPN expression was a downstream effect of the increased Shh expression in MAA-treated mice. Endothelial cells express OPN in response to numerous factors, including Shh[26] and IL-1β[33], and OPN can induce angiogenesis via αvβ3 integrin signaling[34]. Experimental studies have shown that OPN can interact with many factors, including cell surface receptors, proteases and growth factor signaling pathways, and plays important roles in cancer, wound healing, inflammation and immune responses (as reviewed in [35]). However, OPN has been shown to have a varied effect in wound healing. In one study, blocking OPN expression accelerated wound healing[36]; while other studies showed OPN expression was concomitant with angiogenesis[1] and was necessary for normal wound healing in mice[37]. In vitro studies have shown that macrophages downregulate iNOS and NO in response to OPN[38] and downregulate OPN when undergoing macrophage fusion to form foreign body giant cells[39], which may imply that OPN is involved in promoting M2 (i.e., alternative) macrophage phenotype. OPN is also essential for TGF-β induced fibroblast to myofibroblasts differentiation[40].

Six factors with direct roles in regulating angiogenesis were selected for study, to determine if MAA beads promoted a more pro-angiogenic environment (i.e. the balance of inducers and inhibitors is shifted in favour of angiogenesis). VEGF, bFGF and PDGF-B were selected, as they are some of the best-documented inducers of angiogenesis and PDGF-B plays an important role in vessel maturation[41,42]. CXCL10 is a potent inhibitor of angiogenesis and promotes vessel regression in vivo via CXCR3 signaling, even in the presence of angiogenic factors[43]. It is also chemotactic for activated T-cells, macrophages and NK cells[44-46]. Spry2 is ubiquitously expressed by endothelial cells and vascular smooth muscle cells[47] and is a negative feedback

87 loop modulator of MAPK activation, thereby modulating growth factor signaling cascades active in wound healing and angiogenesis (reviewed by[48]). There is also evidence that Spry2 is involved in regulating angiogenesis in wound healing as its expression increases during the wound resolution when vessels begin to regress and exogenous application of Spry2 reduced vascularization of the wound bed in mice[49]. TSP-1 inhibits angiogenesis by numerous mechanisms including inducing apoptosis in endothelial cells via CD36 activation[50] and by blocking VEGFR2 signaling through binding to CD47[51], however its role in wound healing is not restricted to angiostatic behavior. As described in a recent review by Lopez-Dee and contributors [52], TSP-1 also modulates inflammatory processes including apoptosis[53], modulation IL-10 expression and dendritic cell tolerance[54,55], and T-cell activation[56,57]. Moreover, expression of TSP-1 is required for normal dermal wound healing in mice, despite its anti-angiogenic activity[58].

MAA beads were expected to increase the expression of pro-angiogenic factors (VEGF, PDGF- B, bFGF) and/or decrease in the expression of angiostatic factors (CXCL10, Spry2, TSP-1). However, the data (taken at “face value”) indicated that the mechanism of the MAA response is more complex. Surprisingly, MAA beads did not increase the expression of VEGF, despite the increased Shh expression. While there were small increases in the expression of bFGF and PDGF-B and a small decrease in the expression of CXCL10, the expression of TSP-1 and Spry2 was increased, with only the expression of Spry2 being significant. However, angiogenesis is a tightly regulated process that requires a balance of inducers and inhibitors to prevent the excessive sprouting and vascular malformations commonly observed in cancer (reviewed in [59]). Studies of gene expression in diabetic and non-diabetic wound tissue show that, accompanying the increase in expression of pro-angiogenic factors, some anti-angiogenic factors also have increased gene expression (CD36, TSP-1, TSP-2), while others have decreased expression (restin), compared to unwounded skin[1,58]. It is possible that, while the balance of these six angiogenic regulators is arguably shifted slightly in favour of angiogenesis, there is also an increase in anti-angiogenic activity within the wound bed to temper some of the pro- angiogenic activity.

The lack of effect of MAA on VEGF expression was not entirely unexpected. VEGF expression is attenuated in diabetic tissues and cells isolated from diabetic animals[60,61], and the high glucose concentration within diabetic tissues impairs hypoxia-induced VEGF expression[2].

Furthermore, there is evidence that aberrant insulin pathway signaling may reduce VEGF expression in diabetic animals during wound healing[61]. Consequently, VEGF expression may already be “maxed out” in the diabetic granulation tissue.

Currently, there is considerable emphasis in the literature on the activation state of macrophages (i.e. M1 / M2 polarization)[12,62,63]. It is possible that the MAA beads encourage an M2 phenotype, however we have not studied this in depth. In the diabetic wound model (unpublished data), the expression of two M1 markers (Nos2 and IL-12b) were slightly decreased by MAA and PMMA beads, compared to no beads, while the expression of two M2 markers (IL-10 and Arginase 1) was similar among all the wound treatments (data not shown). However, this data is confounded by potential differences in macrophage numbers among the various samples. Different methods are needed to assess in vivo macrophage phenotype, in which the macrophage population is isolated prior to analysis by flow cytometry and qPCR[62].

An important caveat when interpreting these results is that the data is only reporting the relative amounts of mRNA of genes of interest, which we assume are representative of protein levels. While it is possible to have a sense of the relative level of expression of these genes (i.e. Shh is less abundant, Cp ~ 30, than OPN, Cp ~ 18), the actual amount of transcribed protein, and the activity of these proteins are unknown, and will need to be investigated in the future. It is also possible that genes critical to the response to MAA beads were not within the group examined in this study, and the role of mechanisms beyond gene expression must be tested as our understanding of how MAA beads promote angiogenesis improves.

4.4 Insight into the mechanism of MAA-mediated angiogenesis

Based on the results presented here, we hypothesize that the MAA beads are modulating the expression of Shh, likely in cells from the bordering hair follicles or nerves (few cells involved in wound healing are known to express Shh), which enhances the cascade of pro-angiogenic signaling that normally occurs during wound healing. This leads to increased vessel formation and increased OPN expression within the wound bed, as observed in the MAA-treated wounds at day 7. It is unlikely that the beads are directly causing Shh expression; cells present in the wound bed during the acute phase (mononuclear and polymorphonuclear cells) are not known to express Shh. T-cells can be manipulated in vitro to release Shh-containing microparticles when undergoing activation and apoptosis, but this is not considered highly relevant here. While the

89 injection of these microparticles improved vessel density and limb-perfusion in hind-limb ischemic models[64], Shh-containing microparticles, to our knowledge, have not been shown to occur in vivo, and may be an artifact of in vitro manipulation.

A more plausible scenario is that macrophages mediate the effect of MAA beads. Macrophages have principal roles in wound healing and the host response to biomaterial implants (foreign body response), and similar increases in expression of IL-1β, IL-6 and TNF-α were observed in wound tissue and in vitro macrophage-like cells.

Two critical pieces that remain unknown are what cell type(s) is producing Shh within the wound bed, and what factors are up-stream regulators of Shh expression. Only a small subset of cells is known to express this protein in adults. In the context of cutaneous wound healing, there is evidence that adult vascular smooth muscle cells[65,66], and cutaneous nerves[67] express Shh. Recently, Shh-signaling in the perineural niche of hair follicles was shown to be an important factor in maintaining a multipotent stem cell population that contributes to re-epithelialization, and possibly neovascularization of the dermis during wound healing[67]. Immunohistochemical staining for Shh is underway to determine candidate sources of Shh in the wound tissue. As a caveat, it is also conceivable that other molecules are of critical importance especially given the low amount of mRNA (lowest average log2NRQ) that was found even with MAA beads.

From the biomaterial perspective, many biomaterial properties are known to influence the host response, including surface chemistry, surface topography, porosity, particle shape and size (reviewed by [68]). While we have not addressed this question thus far, we may speculate as to how MAA beads differ from “non-bioactive” PMMA beads. While the general shape and size of the beads are similar, the MAA beads are more negatively charged and have a rougher surface topography than PMMA beads. MAA beads also require equilibration in buffered medium when used in vitro, as MAA beads lower the pH of the incubating medium, which can lead to cell toxicity without equilibration, whereas PMMA beads do not.

Presumably the negative surface charge of MAA beads affects the composition (and conformation) of the adsorbed protein layer, which in turn modulates the macrophage response, leading eventually to Shh expression and improved neovascularization. The beads may also preferentially bind positively charged proteins or factors, and release them slowly over time (prolonging their presence within the wound), or they may sequester harmful factors, dampening

90 their negative effects. This hypothesis is currently under study and beyond the scope of this paper. It is also possible that the beads lower the pH of the microenvironment within the wound bed, affecting the activity of growth factors, cytokines, proteases and wound healing in general[69]. Further studies focusing on the biological response to the beads and the properties of the beads themselves are required.

5 Conclusion

MAA beads increased the microvessel density in excisional dermal wounds in diabetic mice. The increase in vascularization was accompanied by a significant increase in the expression of Shh and modest increases in Shh pathway proteins, which suggests Shh signaling may play an important role in MAA-mediated angiogenesis. MAA beads also increased the expression of factors involved in wound healing (IL-1β, IL-6, TNF-α and OPN), similar to the in vitro response of dTHP-1 cells (macrophage-like cells) to MAA beads, suggesting macrophages may modulate the host response to MAA beads in vivo. While MAA beads did not increase the expression of VEGF, relative to PMMA beads or no beads, they slightly increased the expression of bFGF and PDGF-B, and decreased the expression of CXCL10. However, expression of Spry2 and TSP-1 were also increased, suggesting the mechanism of action of the MAA beads is complex and not simply enhancing the expression of pro-angiogenic factors.

6 Acknowledgments

We acknowledge the financial support of the Natural Sciences and Engineering Research Council, the Canadian Institutes of Health Research and the US National Institutes of Health (EB006903). L.E. Fitzpatrick acknowledges scholarship support from the Canadian Institutes of Health Research.

7 References [1] Sharma A, Singh AK, Warren J, Thangapazham RL, Maheshwari RK. Differential regulation of angiogenic genes in diabetic wound healing. J Investig Dermatol 2006;126(10):2323-31. [2] Thangarajah H, Yao D, Chang EI, Shi Y, Jazayeri L, Vial IN, et al. The molecular basis for impaired hypoxia-induced VEGF expression in diabetic tissues. Proc Natl Acad Sci U S A 2009;106(32):13505-10. [3] Eckhaus AA, Fish JS, Skarja G, Semple JL, Sefton MV. A preliminary study of the effect of poly(methacrylic acid-co-methyl methacrylate) beads on angiogenesis in rodent skin

grafts and the quality of the panniculus carnosus. Plast Reconstr Surg 2008;122(5):1361- 70. [4] Martin DC, Semple JL, Sefton MV. Poly(methacrylic acid-co-methyl methacrylate) beads promote vascularization and wound repair in diabetic mice. J Biomed Mater Res A 2010;93(2):484-92. [5] Fitzpatrick LE, Chan JWY, Sefton MV. On the mechanism of poly(methacrylic acid -co- methyl methacrylate)-induced angiogenesis: Gene expression analysis of dTHP-1 cells. Biomaterials 2011;32:8957-67. [6] Renault MA, Roncalli J, Tongers J, Misener S, Thorne T, Jujo K, et al. The hedgehog transcription factor Gli3 modulates angiogenesis. Circ Res 2009;105(8):818-26. [7] Asai J. Topical sonic hedgehog gene therapy accelerates wound healing in diabetes by enhancing endothelial progenitor cell-mediated microvascular remodeling. Circulation 2006;113(20):2413-24. [8] Luo JD, Hu TP, Wang L, Chen MS, Liu SM, Chen AF. Sonic hedgehog improves delayed wound healing via enhancing cutaneous nitric oxide function in diabetes. Am J Physiol Endocrinol Metab 2009;297(2):E525-E31. [9] Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 2001;29(9):e45. [10] Brem H, Tomic-Canic M. Cellular and molecular basis of wound healing in diabetes. J Clin Invest 2007;117(5):1219-22. [11] Werner S, Breeden M, Hubner G, Greenhalgh DG, Longaker MT. Induction of keratinocyte growth factor expression is reduced and delayed during wound healing in the genetically diabetic mouse. J Invest Dermatol 1994;103(4):469-73. [12] Mirza R, Koh TJ. Dysregulation of monocyte/macrophage phenotype in wounds of diabetic mice. Cytokine 2011;56(2):256-64. [13] Michaels J, Churgin SS, Blechman KM, Greives MR, Aarabi S, Galiano RD, et al. db/dbmice exhibit severe wound-healing impairments compared with other murine diabetic strains in a silicone-splinted excisional wound model. Wound Repair Regen 2007;15(5):665-70. [14] Hoffmann DC, Textoris C, Oehme F, Klaassen T, Goppelt A, Romer A, et al. Pivotal role for 1-antichymotrypsin in skin repair. J Biol Chem 2011;286(33):28889-901. [15] Rehman J. Secretion of angiogenic and antiapoptotic factors by human adipose stromal cells. Circulation 2004;109(10):1292-8. [16] Sarkanen JR, Kaila V, Mannerstrom B, Raty S, Kuokkanen H, Miettinen S, et al. Human adipose tissue extract induces angiogenesis and adipogenesis in vitro. Tissue Eng Part A 2012;18(1-2):17-25. [17] Cherubino M, Rubin JP, Miljkovic N, Kelmendi-Doko A, Marra KG. Adipose-derived stem cells for wound healing applications. Ann Plast Surg 2011;66(2):210-5. [18] Byrd N, Grabel L. Hedgehog signaling in murine vasculogenesis and angiogenesis. Trends Cardiovasc Med 2004;14(8):308-13.

[19] Niswander L. Pattern formation: old models out on a limb. Nat Rev Genet 2003;4(2):133-43. [20] Chari NS, McDonnell TJ. The sonic hedgehog signaling network in development and neoplasia. Adv Anat Pathol 2007;14(5):344-52. [21] Pola R, Ling LE, Aprahamian TR, Barban E, Bosch-Marce M, Curry C, et al. Postnatal recapitulation of embryonic hedgehog pathway in response to skeletal muscle ischemia. Circulation 2003;108(4):479-85. [22] Le H, Kleinerman R, Lerman OZ, Brown D, Galiano R, Gurtner GC, et al. Hedgehog signaling is essential for normal wound healing. Wound Repair Regen 2008;16(6):768- 73. [23] Pola R, Ling LE, Silver M, Corbley MJ, Kearney M, Blake Pepinsky R, et al. The morphogen Sonic hedgehog is an indirect angiogenic agent upregulating two families of angiogenic growth factors. Nat Med 2001;7(6):706-11. [24] Kusano KF, Pola R, Murayama T, Curry C, Kawamoto A, Iwakura A, et al. Sonic hedgehog myocardial gene therapy: tissue repair through transient reconstitution of embryonic signaling. Nat Med 2005;11(11):1197-204. [25] Fujii T, Kuwano H. Regulation of the expression balance of angiopoietin-1 and angiopoietin-2 by Shh and FGF-2. In Vitro Cell Dev Biol Anim 2010;46(6):487-91. [26] Renault MA, Roncalli J, Tongers J, Thorne T, Klyachko E, Misener S, et al. Sonic hedgehog induces angiogenesis via Rho kinase-dependent signaling in endothelial cells. J Mol Cell Cardiol 2010;49(3):490-8. [27] Alvarez-Medina R, Cayuso J, Okubo T, Takada S, Martí E. Wnt canonical pathway restricts graded Shh/Gli patterning activity through the regulation of Gli3 expression. Development 2008;135(2):237-47. [28] Xia Z, Triffitt JT. A review on macrophage responses to biomaterials. Biomed Mater 2006;1(1):R1-R9. [29] Singer AJ, Clark RA. Cutaneous wound healing. N Engl J Med 1999;341(10):738-46. [30] Carmi Y, Voronov E, Dotan S, Lahat N, Rahat MA, Fogel M, et al. The role of macrophage-derived IL-1 in induction and maintenance of angiogenesis. J Immunol 2009;183(7):4705-14. [31] Rosen EM, Liu D, Setter E, Bhargava M, Goldberg ID. Interleukin-6 stimulates motility of vascular endothelium. Exs 1991;59:194-205. [32] Sainson RC, Johnston DA, Chu HC, Holderfield MT, Nakatsu MN, Crampton SP, et al. TNF primes endothelial cells for angiogenic sprouting by inducing a tip cell phenotype. Blood 2008;111(10):4997-5007. [33] Singh K, Balligand JL, Fischer TA, Smith TW, Kelly RA. Glucocorticoids increase osteopontin expression in cardiac myocytes and microvascular endothelial cells. Role in regulation of inducible nitric oxide synthase. J Biol Chem 1995;270(47):28471-8. [34] Dai J, Peng L, Fan K, Wang H, Wei R, Ji G, et al. Osteopontin induces angiogenesis through activation of PI3K/AKT and ERK1/2 in endothelial cells. Oncogene 2009;28(38):3412-22.

[35] Wang KX, Denhardt DT. Osteopontin: Role in immune regulation and stress responses. Cytokine Growth Factor Rev 2008;19(5-6):333-45. [36] Mori R, Shaw TJ, Martin P. Molecular mechanisms linking wound inflammation and fibrosis: knockdown of osteopontin leads to rapid repair and reduced scarring. J Exp Med 2008;205(1):43-51. [37] Liaw L, Birk DE, Ballas CB, Whitsitt JS, Davidson JM, Hogan BL. Altered wound healing in mice lacking a functional osteopontin gene (spp1). J Clin Invest 1998;101(7):1468-78. [38] Wai PY, Guo L, Gao C, Mi Z, Guo H, Kuo PC. Osteopontin inhibits macrophage nitric oxide synthesis to enhance tumor proliferation. Surgery 2006;140(2):132-40. [39] Tsai AT, Rice J, Scatena M, Liaw L, Ratner BD, Giachelli CM. The role of osteopontin in foreign body giant cell formation. Biomaterials 2005;26(29):5835-43. [40] Lenga Y, Koh A, Perera AS, McCulloch CA, Sodek J, Zohar R. Osteopontin expression is required for myofibroblast differentiation. Circ Res 2008;102(3):319-27. [41] Cao R, Brakenhielm E, Pawliuk R, Wariaro D, Post MJ, Wahlberg E, et al. Angiogenic synergism, vascular stability and improvement of hind-limb ischemia by a combination of PDGF-BB and FGF-2. Nat Med 2003;9(5):604-13. [42] Carmeliet P, Jain RK. Molecular mechanisms and clinical applications of angiogenesis. Nature 2011;473(7347):298-307. [43] Bodnar RJ, Yates CC, Rodgers ME, Du X, Wells A. IP-10 induces dissociation of newly formed blood vessels. J Cell Sci 2009;122(12):2064-77. [44] Taub DD, Sayers TJ, Carter CR, Ortaldo JR. Alpha and beta chemokines induce NK cell migration and enhance NK-mediated cytolysis. J Immunol 1995;155(8):3877-88. [45] Cuenca AG, Wynn JL, Kelly-Scumpia KM, Scumpia PO, Vila L, Delano MJ, et al. Critical role for CXC ligand 10/CXC Receptor 3 signaling in the murine neonatal response to sepsis. Infect Immun 2011;79(7):2746-54. [46] Taub DD, Longo DL, Murphy WJ. Human interferon-inducible protein-10 induces mononuclear cell infiltration in mice and promotes the migration of human T lymphocytes into the peripheral tissues and human peripheral blood lymphocytes-SCID mice. Blood 1996;87(4):1423-31. [47] Antoine M, Wirz W, Tag CG, Mavituna M, Emans N, Korff T, et al. Expression pattern of fibroblast growth factors (FGFs), their receptors and antagonists in primary endothelial cells and vascular smooth muscle cells. Growth Factors 2005;23(2):87-95. [48] Cabrita MA, Christofori G. Sprouty proteins, masterminds of receptor tyrosine kinase signaling. Angiogenesis 2008;11(1):53-62. [49] Wietecha MS, Chen L, Ranzer MJ, Anderson K, Ying C, Patel TB, et al. Sprouty2 downregulates angiogenesis during mouse skin wound healing. Am J Physiol Heart Circ Physiol 2011;300(2):H459-H67. [50] Jiménez B, Volpert OV, Crawford SE, Febbraio M, Silverstein RL, Bouck N. Signals leading to apoptosis-dependent inhibition of neovascularization by thrombospondin-1. Nat Med 2000;6(1):41-8.

[51] Kaur S, Martin-Manso G, Pendrak ML, Garfield SH, Isenberg JS, Roberts DD. Thrombospondin-1 inhibits VEGF receptor-2 signaling by disrupting its association with CD47. J Biol Chem 2010;285(50):38923-32. [52] Lopez-Dee Z, Pidcock K, Gutierrez LS. Thrombospondin-1: multiple paths to inflammation. Mediators Inflamm 2011;2011:1-10. [53] Manna PP, Dimitry J, Oldenborg PA, Frazier WA. CD47 augments Fas/CD95-mediated apoptosis. J Biol Chem 2005;280(33):29637-44. [54] Doyen V, Rubio M, Braun D, Nakajima T, Abe J, Saito H, et al. Thrombospondin 1 is an autocrine negative regulator of human dendritic cell activation. J Exp Med 2003;198(8):1277-83. [55] Torres-Aguilar H, Aguilar-Ruiz SR, Gonzalez-Perez G, Munguia R, Bajana S, Meraz- Rios MA, et al. Tolerogenic dendritic cells generated with different immunosuppressive cytokines induce antigen-specific anergy and regulatory properties in memory CD4+ T cells. J Immunol 2010;184(4):1765-75. [56] Vallejo AN, Mugge LO, Klimiuk PA, Weyand CM, Goronzy JJ. Central role of thrombospondin-1 in the activation and clonal expansion of inflammatory T cells. J Immunol 2000;164(6):2947-54. [57] Li SS, Liu Z, Uzunel M, Sundqvist KG. Endogenous thrombospondin-1 is a cell-surface ligand for regulation of integrin-dependent T-lymphocyte adhesion. Blood 2006;108(9):3112-20. [58] DiPietro LA, Nissen NN, Gamelli RL, Koch AE, Pyle JM, Polverini PJ. Thrombospondin 1 synthesis and function in wound repair. Am J Pathol 1996;148(6):1851-60. [59] Goel S, Duda DG, Xu L, Munn LL, Boucher Y, Fukumura D, et al. Normalization of the vasculature for treatment of cancer and other diseases. Physiol Rev 2011;91(3):1071- 121. [60] Chou E, Suzuma I, Way KJ, Opland D, Clermont AC, Naruse K, et al. Decreased cardiac expression of vascular endothelial growth factor and its receptors in insulin-resistant and diabetic States: a possible explanation for impaired collateral formation in cardiac tissue. Circulation 2002;105(3):373-9. [61] Goren I, Müller E, Schiefelbein D, Gutwein P, Seitz O, Pfeilschifter J, et al. Akt1 controls insulin-driven VEGF biosynthesis from keratinocytes: implications for normal and diabetes-impaired skin repair in mice. J Investig Dermatol 2008;129(3):752-64. [62] Hu Y, Zhang H, Lu Y, Bai H, Xu Y, Zhu X, et al. Class A scavenger receptor attenuates myocardial infarction-induced cardiomyocyte necrosis through suppressing M1 macrophage subset polarization. Basic Res Cardiol 2011;106(6):1311-28. [63] Martinez FO, Gordon S, Locati M, Mantovani A. Transcriptional profiling of the human monocyte-to-macrophage differentiation and polarization: new molecules and patterns of gene expression. J Immunol 2006;177(10):7303-11. [64] Benameur T, Soleti R, Porro C, Andriantsitohaina R, Martínez MC. Microparticles carrying sonic hedgehog favor neovascularization through the activation of nitric oxide pathway in mice. PLoS One 2010;5(9):e12688.

[65] Li F, Duman-Scheel M, Yang D, Du W, Zhang J, Zhao C, et al. Sonic hedgehog signaling induces vascular smooth muscle cell proliferation via induction of the G1 cyclin-retinoblastoma axis. Arterioscler Thromb Vasc Biol 2010;30(9):1787-94. [66] Morrow D, Sweeney C, Birney YA, Guha S, Collins N, Cummins PM, et al. Biomechanical regulation of hedgehog signaling in vascular smooth muscle cells in vitro and in vivo. Am J Physiol Cell Physiol 2007;292(1):C488-C96. [67] Brownell I, Guevara E, Bai CB, Loomis CA, Joyner AL. Nerve-derived sonic hedgehog defines a niche for hair follicle stem cells capable of becoming epidermal stem cells. Cell Stem Cell 2011;8(5):552-65. [68] Morais JM, Papadimitrakopoulos F, Burgess DJ. Biomaterials/tissue interactions: possible solutions to overcome foreign body response. AAPS J 2010;12(2):188-96. [69] Schneider LA, Korber A, Grabbe S, Dissemond J. Influence of pH on wound-healing: a new perspective for wound-therapy? Arch Dermatol Res 2007;298(9):413-20.

Chapter 5 On the mechanism of poly(methacrylic acid-co-methyl methacrylate)-induced angiogenesis: gene expression analysis of dTHP-1 cells1 Abstract

Identifying the critical molecules associated with “biocompatibility” is a grand challenge.

Poly(methacrylic acid -co- methyl methacrylate) (MAA) beads improve wound closure and wound vascularity in vivo, but the mechanism of this phenomenon is unknown. We used quantitative real-time PCR to identify the subtle changes in the expression of a small selection of molecules involved in wound healing and angiogenesis in a macrophage-like cell (dTHP-1) treated with the MAA beads (45 mol% methacrylic acid). MAA beads decreased the expression of osteopontin (OPN) compared to poly(methyl methacrylate) (PMMA) and untreated cells, and increased the expression of IL-1β, IL-6 and TNF-α over the 24 – 96 h of the experiment.

Interestingly, molecules associated with angiogenesis, such as bFGF, CXCL12, HIF1α, PDGFB,

TGFβ and VEGF, were not significantly affected by MAA beads over the course of the study.

MAA beads also increased the gene expression of OPN in HUVEC compared to untreated cells, while PMMA beads did not. MAA beads modified the phenotype (gene expression) of dTHP-1 cells in a subtle yet distinct manner that was different than PMMA. It remains to connect the changes in OPN in dTHP-1 (and HUVEC) and other molecules to the enhanced vascularity seen in vivo with this polymer.

1 Reprinted from Biomaterials, 32(34), Fitzpatrick LE, Chan JW and Sefton MV, On the mechanism of poly(methacrylic acid –co– methyl methacrylate)-induced angiogenesis: Gene expression analysis of dTHP-1 cells, 8957-67, Copyright 2011, with permission from Elsevier.

1 Introduction

Despite >40 years of active research into biocompatibility[1], our understanding of the molecular events in the host response to biomaterials remains rather limited. While proteomic studies[2,3] highlight a large number of genes that are involved in the foreign body response, it is difficult to determine which are the key molecules giving rise to particular aspects of the host response. We focus here on the subtle but important difference in response to a biomaterial that induces angiogenesis and a control material that does not. Because the effect was subtle, great care was needed to distinguish molecules that were differentially involved in the angiogenic biomaterial response.

The host response is directed, in large part, by macrophages interacting with the adsorbed protein layer that forms at the surface of the biomaterial upon implantation[4]. Angiogenesis is a necessary element of the granulation tissue that forms as part of this foreign body reaction. In this, endothelial cells at the wound site are switched from a quiescent to an angiogenic state, a process that is orchestrated by the macrophages, which also direct the recruitment of fibroblasts and myofibroblasts [5-8]. Eventually the newly formed vessels regress to be replaced, in the case of many biomaterials, by the familiar fibrous capsule, by action of the macrophages that have been shown to mediate capillary regression in vivo[9] and that are also a source of many angiogenic inhibitors[3].

Previously, our group has demonstrated that poly(methacrylic acid -co- methyl methacrylate) (MAA) beads significantly improved healing in a skin graft model in rats[10] and a full- thickness cutaneous wound model in diabetic (db/db) mice[11] by increasing angiogenesis in the grafts or wound beds, respectively. However, the mechanism of this form of biomaterial- mediated angiogenesis is unknown. Because macrophages are a principal responder to biomaterial implantation[12] and a main mediator of wound healing[13], we undertook this study to examine the response of macrophage-like cells (dTHP-1 cells) to MAA beads in vitro, focusing on changes in gene expression of a few select molecules, in an effort to better understand the biological response to this pro-angiogenic material. We also chose to examine the effect of conditioned medium from the bead-exposed cells on the migration of human umbilical vein endothelial cells (HUVEC) since endothelial cells are key cells in angiogenesis.

As is described below, a preliminary microarray experiment identified osteopontin (OPN) as being differentially expressed in the presence of MAA beads. Quantitative real-time PCR (qPCR) was used to validate the differential expression of OPN in response to MAA beads, as this molecule is highly involved in wound healing[14-16], angiogenesis[17] and the foreign body response[18]. The expression of nine other genes involved in wound healing and angiogenesis were also studied using qPCR. Vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF), platelet-derived growth factor B (PDGF-B), transforming growth factor beta, (TGF-β) and hypoxia inducible factor 1 alpha (HIF-1α) are well-known pro-angiogenic molecules that are expressed by macrophages during wound healing[13,19]. TNF-α, IL-1β and IL-6 are pro-inflammatory cytokines secreted by activated macrophages and other cells to drive the inflammatory response during wound healing[20] and during the foreign body response[4]. The chemokine CXCL12 (stromal derived factor 1, SDF-1) is expressed by activated macrophages[21], and has been shown to improve wound healing in diabetic models[22] and mediate endothelial progenitor cell homing and incorporation into injured tissue[23].

2 Materials and methods 2.1 Bead preparation

MAA beads (150 - 250 µm in diameter, 45 mol % methacrylic acid) were obtained from Rimon Therapeutics (Toronto, Canada) and PMMA control beads (same diameter) were obtained from Polysciences (Warrington, PA). Bead characteristics are presented elsewhere[11,24]. Endotoxin was removed from beads by repeatedly sonicating beads in 1N hydrochloric acid (PMMA beads) or 95% ethanol (MAA beads) then washing beads in endotoxin-free water. Beads were tested for endotoxin using LAL Pyrochrome Endotoxin Kit (Associates of Cape Cod, Falmouth, MA) to confirm that beads contained less than 0.25 EU per 100 mg beads.

2.2 Cell culture

THP-1 cells (TIB-202, ATCC, Manassas, VA) were maintained in RPMI 1640 with 25 mM HEPES and 2 mM L-glutamine (Gibco, Invitrogen Canada Inc, Burlington, ON, Canada) o supplemented with 10% fetal bovine serum (FBS; Invitrogen) at 37 C, 5% CO2. THP-1 cells were differentiated into a macrophage-like cell type (dTHP-1) by treatment with 100 nM phorbol myristate acetate (PMA; Sigma-Aldrich Canada Ltd., Oakville, ON) for 24 - 48 h on tissue

99 culture treated polystyrene (TCPS) 6-well plates (Nunclon, Thermo Fisher Scientific Inc., Mississauga, ON). Upon differentiation, dTHP-1 cells attached to the TCPS and spread. After the PMA incubation, dTHP-1 cells were washed with PBS and cultured for 1 - 2 h in THP-1 culture medium before the addition of beads.

Human umbilical vein endothelial cells (HUVEC, Lonza, Basel, Switzerland) were cultured in EGM-2 medium (Lonza) in 75 cm2 TCPS flasks (Falcon, Becton Dickinson Canada, o Mississauga, ON) at 37 C, 5% CO2. HUVEC were passaged every 2-3 days using a splitting ratio of 1:4. Only HUVEC at passage 3 - 5 were used for migration and gene expression experiments.

2.3 Bead treatment of dTHP-1 and HUVEC cells

Beads were weighed into pyrogen-free microcentrifuge tubes (Axygen, VWR International, Mississauga, ON) and equilibrated by repeated washes in unsupplemented RPMI 1640 (Gibco, Invitrogen) containing phenol red, until no colour change was detected in the medium. MAA beads (0.3 mg/cm2) or PMMA beads (0.9 mg/cm2) were added to the dTHP-1 culture in 0.5 ml of unsupplemented RPMI 1640; the 3:1 ratio in bead amount is explained below. Untreated dTHP-1 cells received 0.5 ml of unsupplemented RPMI 1640 containing no beads (referred to as “untreated” or “no bead” in the text). Cells were incubated with the beads for 24, 48 and 96 h in 6-well plates. At the end of each time point, the culture medium was either discarded or used immediately in the HUVEC migration assay, described below, and RNA was isolated from the dTHP-1 cells. The bead treatment of HUVEC was performed using the same method as dTHP-1, except that the experiment was performed in EGM-2 medium and only two time points were used (24 and 91 h).

2.4 RNA isolation and cDNA synthesis

RNA was extracted from the dTHP-1 cells (24 h, 48 h and 96 h) using Trizol reagent (Invitrogen), according to manufacturer’s protocol, and was further purified using the Qiagen RNeasy Mini Kit (Qiagen Inc, Mississauga, ON) clean-up protocol with on-column DNAse digestion. For HUVEC, RNA was isolated at 24 and 91 h using the Qiagen Shredder and RNeasy Mini Kit, according to manufacturer’s instructions.

100

RNA content was quantified by measuring the absorbance at 260 nm using a Nanodrop ND-1000 spectrophometer (NanoDrop Technologies, Inc., Wilmington, DE). All RNA samples had a 260/280 ratio greater than 2.0. Total RNA (1 µg) was used for cDNA synthesis, which was performed using a Superscript III First-Strand Synthesis SuperMix (Invitrogen) according to manufacturer’s protocol.

2.5 DNA microarray

RNA samples (at least 4 per treatment) to be used in the DNA microarray were fluorescently labeled using a direct method by adding Cyanine 3 (Cy3) or Cyanine 5 (Cy5) dCTP (PerkinElmer, Woodbridge, ON) to the cDNA synthesis reaction mixture. Labeled cDNA was then hybridized at 37°C overnight onto an Ontario Cancer Institute (OCI; Toronto, ON) 1.7k human cDNA array. Each chip compared cDNA from either MAA to no bead, or PMMA to no bead. Chips were washed and then scanned using a Packard GSI Scanner (Packard GSI Lumonics, Billerica, MA). The chip images were processed using QuantArray Analysis Software (GSI Lumonics) to obtain the corrected spot intensities, which were exported to Excel for further analysis. The expression profile of MAA- or PMMA-treated dTHP-1 cells were compared to that of untreated dTHP-1 cells, and the results were filtered such that any ratios between 0.8 and 1.25 or with a p-value greater than 0.05 were discarded. A volcano plot was generated from the filtered data, in which expression ratios were presented as fold change.

2.6 Quantitative real-time PCR

Gene-specific primer sequences (Table 5-1) were either selected from Primerbank (http://pga.mgh.harvard.edu/primerbank/) or designed using Primer-blast (http://www.ncbi.nlm.nih.gov/tools/primer-blast/) and were synthesized by Sigma Genosys (Sigma-Aldrich). All primer pairs were analyzed by Blast to ensure specificity for the intended target gene within the human genome. The rationale for the choice of genes was described in the introduction.

Quantitative real-time PCR (qPCR) reactions were carried out in a 7900HT Fast Real-Time PCR system (Applied Biosystems Canada, Streetsville, ON) using the following thermal profile: 50 oC for 2 minutes, 95 oC for 10 minutes, then 40 cycles of 95 oC for 15 second, 60 oC for 30 seconds and 72oC for 30 seconds. A dissociation curve was generated for each plate following

101

the amplification cycles by slowly cooling from 95 oC at a 2% ramp rate. PCR reactions occurred in a final volume of 10 µL, containing 5 µL of SYBR Green Master Mix (Applied Biosystems), 1 µL of 10µM forward and reverse primers and 10 ng of cDNA template in 4 µl of RNase/DNase free water.

Table 5-1. Primer sequences for real-time PCR. GenBank Probe Forward Reverse Accession

GAPDH ATGGGGAAGGTGAAGGTCG GGGGTCATTGATGGCAACAATA NM_002046

OPN ACTCGAACGACTCTGATGATGT GTCAGGTCTGCGAAACTTCTTA NM_000582

bFGF ATCAAAGGAGTGTGTGCTAACC ACTGCCCAGTTCGTTTCAGTG NM_002006

CXCL12 ATGCCCATGCCGATTCTTCG GCCGGGCTACAATCTGAAGG NM_00609

HIF1α GGCGCGAACGACAAGAAAAAG CCTTATCAAGATGCGAACTCACA NM_001530

IL1β CTCGCCAGTGAAATGATGGCT GTCGGAGATTCGTAGCTGGAT NM_000576

IL6 AAATTCGGTACATCCTCGACGG GGAAGGTTCAGGTTGTTTTCTGC NM_000600

PDGF-B TCTCTGCTGCTACCTGCGT CAAAGGAGCGGATCGAGTGG NM_002608

TGF-β GGCCAGATCCTGTCCAAGC GTGGGTTTCCACCATTAGCAC NM_000660

TNF-α ATGAGCACTGAAAGCATGATCC GAGGGCTGATTAGAGAGAGGTC NM_000594

VEGF CAACATCACCATGCAGATTATGC GCTTTCGTTTTTGCCCCTTTC NM_113376

The normalized reporter signal was baseline corrected using the 7900HT Sequence Detection Systems (SDS) Software (Version 2.3, Applied Biosystems). The qPCR data analysis was performed as described in Rieu and Power (2009)[25]. Briefly, individual crossing point (Cp) values and mean amplicon efficiencies were estimated using LinRegPCR (Version 11.0, download: http://LinRegPCR.HFRC.nl). The relative quantity (RQ) of each template was normalized to the RQ of the endogenous control, GAPDH, for each sample to give the normalized relative quantity (NRQ) for every gene of interest (GOI)[25]. The relative expression ratio (R) was calculated as the ratio of the mean NRQ(treated) to the mean NRQ(control)[25]:

Cpref NRQ E ref R treated where NRQ () = = Cpgoi NRQcontrol ()E goi

! 102

Results were expressed as the mean relative expression ratio ± standard error (n = 4). Statistical analysis of variance (ANOVA) was performed on the log transformed NRQ to determine if treatment significantly effected the mean NRQ for each gene of interest[25]. Genes were considered differentially expressed if 0.67 ≤ R ≥ 1.5 and p < 0.05.

2.7 HUVEC migration assay

Two days before use, the HUVEC growth medium was gradually changed from 100% EGM-2 medium to 100% THP-1 culture medium (RPMI 1640 with 10% FBS). HUVEC (50 000 cells/transwell insert) were seeded in the top well of 24-well plate polycarbonate (8.0 µm pore) transwell insert (Corning, Sigma-Aldrich) in RPMI 1640 with 0.1% bovine serum albumin (BSA; Sigma-Aldrich) for 2 h. Following adhesion, the 0.1% BSA medium was removed and the conditioned medium from dTHP-1 cells treated with beads (or no beads) for 24, 48 or 96 h was added to the upper and lower wells of the transwell, such that HUVEC migrated from one conditioned medium (e.g. no bead-conditioned medium = untreated medium) to a different conditioned medium (e.g. MAA-conditioned medium). Cells were allowed to migrate for 16 to 18 h, at which time the conditioned medium was removed from the upper and lower wells, and the top side of the membranes was gently swabbed with cotton to removed non-migrated cells. The migrated cells (on the bottom side of the membrane) were fixed in 4% paraformaldehyde for 10 minutes, and then rinsed in PBS. The membranes were stained with 1 µg/ml Hoechst 33258 dye (Molecular Probes, Invitrogen) for 20 minutes, rinsed in PBS and imaged on a Zeiss Axiovert 135 inverted fluorescence microscope (Carl Zeiss Canada Ltd., Toronto, ON, Canada). The number of nuclei of the migrated HUVEC was counted in three high power (10x) fields and averaged for each migration condition. The experiment was run in triplicate.

2.8 Statistical analysis

All data is presented as the mean ± SEM (standard error of the mean). Statistical comparisons among treatments (MAA, PMMA, no beads) were performed using a one-way analysis of variance (ANOVA). Post hoc analysis between multiple groups was performed using Tukey’s honestly significant difference. Statistically significant differences were defined as having p < 0.05.

103

3 Results 3.1 Effect of the MAA beads on dTHP-1 mRNA expression

To study the effect of MAA beads on macrophage-like cells, the THP-1 monocytic cell line was differentiated into macrophage-like cells (dTHP-1) using 100nM PMA incubation. Over the 24 - 48 h PMA incubation, the THP-1 cells began to adhere to the TCPS. The PMA-containing medium was removed, and the adherent dTHP-1 cells were washed with PBS to remove any loosely adherent cells and cultured in RPMI 1640 with 10% FBS.

The dTHP-1 cells were then treated with MAA and PMMA beads. PMMA beads were applied in an amount three times greater (per weight) than MAA to account for the swelling that MAA beads undergo upon hydration[10]; the intent was to test equal volumes of beads. Untreated dTHP-1 cells were used as a negative control (i.e. no bead). The results from the OCI 1.7k human microarray analysis of dTHP-1 cells at 24 h are shown in a volcano plot (Figure 5-1). Only two genes were differentially expressed (fold-change greater than 1.5, p < 0.05) in MAA- treated cells compared to untreated cells. Osteopontin (OPN) expression was decreased 1.8-fold by MAA beads (p < 0.001); while MAA beads increased the expression of insulin growth factor binding protein (IGFBP-3) 1.6-fold (p = 0.01). No genes were differentially expressed by PMMA-treated dTHP-1, compared to untreated dTHP-1.

The microarray results prompted further gene expression analysis using qPCR to validate the decreased expression of OPN and also investigate the expression of other genes involved in wound healing and angiogenesis, using GAPDH as an endogenous control: OPN, bFGF, CXCL12, HIF1α, IL1β, IL6, PDGFB, TGFβ, TNFα and VEGF. All templates were successfully amplified for all treatment conditions and time points. Two metrics were used to determine significance in gene expression differences: the normalized relative quantity (NRQ) and the relative expression ratio (R). The NRQ is the relative quantity of a given gene (amount of cDNA), normalized to the relative quantity of GAPDH present in the same sample. However, the qPCR data are nonlinear on the NRQ scale and consequently, a log transformation was applied to the NRQ data before performing a one-way ANOVA to determine if the treatments (MAA, PMMA or no beads) had a statistically significant effect of the expression of each gene. The log-

2NRQ data for each gene and p-values from the ANOVA are summarized in Table 5-2. There

104 were a few statistically significant differences (i.e. p < 0.05), as is noted in the paragraphs below, for some molecules and none for several others.

Figure 5-1. Volcano plot of microarray data comparing gene expression of dTHP-1 cells treated MAA or PMMA beads for 24 h to untreated cells. Data was filtered to remove genes with fold-change less than ± 1.25 or p-values greater than 0.05 [-log (0.05) = 1.3]. Only two genes, OPN (decreased 1.8-fold, p < 0.001) and IGFBP3 (increased 1.6-fold, p = 0.01), were expressed differently in MAA-treated dTHP-1 than in untreated cells. PMMA- treated and untreated cells had similar gene expression profiles.

The relative expression ratio (R) is the ratio of two NRQ values. These ratios determine the effect MAA beads (MAA/no bead) or PMMA beads (PMMA/no bead) had on dTHP-1 gene expression. The gene expression in MAA-treated cells was also compared to PMMA-treated cells (MAA/PMMA) to highlight the response to the methacrylic acid containing beads. A change in gene expression was considered to be relevant only if the relative expression ratio was greater than 1.5 (increased expression) or less than 0.67 (decreased expression) and the corresponding p-value from the log2NRQ data was less than 0.05 (i.e. R ≥ 1.5 OR R ≤ 0.67,

AND p < 0.05). It should be noted that in some cases log2NRQ values (i.e. MAA compared to PMMA for gene x) were significantly different (p < 0.05), but the expression ratio was less than 1.5. This contradiction of metrics represented cases in which the ANOVA analysis was able to resolve the difference between two mean log2NRQ values due to the small amount of variation

105 within the replicates of each treatment group, but the difference in means was very small (less than a 1.5 fold change) so that the change in gene expression was not considered to be biologically important.

No Bead PMMA MAA p-values (Tukey's HSD test)

MAA MAA PMMA Time Gene log NRQ SE log NRQ SE log NRQ SE compared to compared to compared to (h) 2 2 2 No Bead PMMA No Bead bFGF 24 -5.56 0.03 -5.61 0.08 -5.17 0.01 0.057 0.034 0.941 48 -6.55 0.16 -6.95 0.04 -6.55 0.02 1.000 0.335 0.326 96 -6.25 0.03 -6.16 0.05 -5.87 0.02 0.012 0.048 0.663 CXCL12 24 -8.39 0.03 -8.27 0.01 -8.28 0.06 0.592 0.995 0.537 48 -7.89 0.14 -8.46 0.10 -7.90 0.03 0.999 0.164 0.157 96 -8.38 0.02 -8.48 0.03 -8.57 0.04 0.117 0.528 0.523 HIF -1α 24 -6.28 0.09 -6.05 0.05 -6.23 0.09 0.979 0.719 0.604 48 -6.23 0.17 -6.61 0.07 -6.44 0.01 0.794 0.842 0.471 96 -6.00 0.06 -5.72 0.10 -6.11 0.04 0.859 0.183 0.377 IL -1β 24 2.16 0.03 2.18 0.08 2.62 0.04 0.042 0.049 0.995 48 2.30 0.27 2.05 0.09 2.20 0.03 0.975 0.944 0.853 96 -3.26 0.09 -1.88 0.15 -0.71 0.08 0.000 0.012 0.005 IL -6 24 -10.59 0.04 -10.54 0.04 -9.01 0.03 0.000 0.000 0.888 48 -12.28 0.16 -12.41 0.08 -10.83 0.04 0.002 0.001 0.897 96 -11.96 0.12 -12.07 0.18 -11.56 0.14 0.620 0.463 0.959 OPN 24 2.14 0.04 2.04 0.02 1.58 0.02 0.000 0.001 0.487 48 1.39 0.17 1.28 0.10 0.43 0.03 0.032 0.010 0.916 96 0.92 0.03 1.17 0.06 0.30 0.08 0.016 0.002 0.365 PDGF -B 24 -11.58 0.05 -11.27 0.08 -11.49 0.09 0.906 0.579 0.357 48 -14.02 0.13 -14.42 0.08 -13.96 0.03 0.968 0.201 0.282 96 -12.24 0.04 -11.60 0.10 -11.51 0.04 0.009 0.895 0.017 TGF -β 24 -7.35 0.06 -7.35 0.04 -7.58 0.04 0.291 0.291 1.000 48 -9.07 0.13 -9.41 0.10 -9.63 0.07 0.188 0.731 0.508 96 -8.97 0.05 -9.06 0.07 -9.06 0.08 0.902 1.000 0.901 TNF -α 24 -2.15 0.03 -2.30 0.06 -1.49 0.05 0.003 0.001 0.531 48 -3.37 0.10 -3.61 0.07 -2.57 0.04 0.010 0.002 0.507 96 -6.49 0.05 -5.91 0.08 -4.67 0.04 0.000 0.000 0.018 VEGF 24 -6.38 0.03 -6.28 0.03 -6.16 0.03 0.055 0.315 0.494 48 -6.03 0.20 -6.62 0.09 -6.18 0.06 0.921 0.505 0.316 96 -7.34 0.04 -6.80 0.06 -6.59 0.05 0.001 0.326 0.009

The inhibitory effect of MAA on OPN expression, observed in the microarray, was confirmed with qPCR (Figure 5-2), and, although the magnitude of the relative expression ratio is not large (e.g. R = 0.65 at 48 h, compared no bead); the trend was consistent for MAA-treated dTHP-1 cells for all time-points and the log2NRQ were significantly different (p ≤ 0.032) compared to

106

PMMA and the untreated cells for all time points, except MAA compared to the untreated cells at 96 h. PMMA had no effect on OPN expression compared to untreated cells.

Figure 5-2. Osteopontin mRNA expression in dTHP-1 cells treated with MAA beads, PMMA beads or no bead. MAA beads consistently decreased the expression of OPN mRNA in dTHP-1 cells, compared with PMMA beads and no bead. PMMA caused a slight increase in OPN expression at 96 h, however, the ratio was less than 1.5. Significant changes in expression,*, have a relative ratio less than 0.67 or greater than 1.5 with corresponding p-value less than 0.05. n = 4, +/- SE; p-values are provided in Table 5-2 for NRQ values. dTHP-1 expression of IL-6 (Figure 5-3) was increased more than 2.5-fold with MAA treatment compared to the PMMA treatment and no bead control during the first 48 h (p < 0.002). By 96 h, the IL-6 expression in MAA-treated dTHP-1 cells decreased to a level similar to the other groups (p > 0.463). PMMA beads did not significantly affect the expression of IL-6 over the duration of the experiment (p >0.89) indicating that the increase observed with the MAA beads was not simply due to the presence of a biomaterial in the cell culture.

TNF-α (Figure 5-4) mRNA expression gradually increased over 96 h with MAA treatment, compared to PMMA-treated and untreated dTHP-1 cells. MAA-treated cells expressed TNF-α 1.8-fold higher than PMMA (p = 0.001) and 1.6-fold higher than untreated cells (p = 0.003) after 24 h of incubation. At 96 h, TNF-α expression increased to be 3.5-fold (p < 0.001) and 2.3-fold

107

(p <0.001) higher in MAA-treated cells than PMMA-treated and untreated cells, respectively. Again, PMMA did not significantly effect the expression of TNF-α compared to no bead (p > 0.51), except at 96 h when a small increase in expression (R = 1.5, p = 0.018) was observed.

Figure 5-3. qPCR analysis of IL-6 mRNA expression in dTHP-1 cells cultured with MAA beads, PMMA beads or no bead. MAA treatment of dTHP-1 cells resulted in a three-fold increase in IL-6 expression (relative to no bead or MMA beads) over the first 48 h. By 96 h, the differential effect of MAA treatment on dTHP-1 IL-6 mRNA expression had disappeared. PMMA beads had no effect on IL-6 expression compared to no bead. Significant changes in expression,*, have a relative ratio less than 0.67 or greater than 1.5 with corresponding p-value less than 0.05. n = 4, +/- SE; p-values are provided in Table 5-2 for NRQ values.

Although the log2NRQ values for IL-1β were significantly different (p < 0.05) in MAA compared to PMMA and no bead at 24 h, the expression ratio (Figure 5-5) was less than 1.5 during the first 48 h of treatment. Therefore, the difference in expression was not considered to be biologically important. At 96 h, IL-1β expression was 2.2-fold (p = 0.012) and 5.8-fold (p < 0.001) higher in MAA-treated cells than PMMA-treated and untreated cells, respectively. IL-1β expression also increased 2.7-fold (p = 0.005) in the PMMA group (compared to untreated cells) at 96 h.

108

Figure 5-4. qPCR analysis of TNF-α mRNA expression in dTHP-1 cells treated with MAA beads, PMMA beads or no bead. MAA-treated dTHP-1 cells have increased expression of TNF-α mRNA at 96 h of treatment compared to PMMA-treated dTHP-1 and untreated dTHP-1 cells. PMMA beads increased the expression of TNF-α also, but not to the same extent as MAA beads. Significant changes in expression,*, have a relative ratio less than 0.67 or greater than 1.5 with corresponding p-value less than 0.05. n = 4, +/- SE; p-values are provided in Table 5-2 for NRQ values.

MAA treatment did not affect the mRNA expression ratio (relative to PMMA or no bead) of the pro-angiogenic factors CXCL12 (SDF-1), HIF-1α and TGF-β at any of the three times (Figure 5- 6). There was a statistically significant difference in bFGF expression at 96 h (p = 0.012) compared to no bead and at 24 h (p = 0.03) and 96 h (p = 0.048) compared to PMMA-treated dTHP-1, however these differences were not considered to be of biological importance as the changes were small (R < 1.5). MAA-treated cells did not affect the expression of VEGF and PDGF-B at 24 and 48 h (p > 0.055), but did increase their expression 1.69 times (p = 0.001) and 1.66 times (p = 0.009) at 96 h, respectively (relative to PMMA or no bead). PMMA did not have a significant effect (p ≥ 0.19) on expression of pro-angiogenic genes compared to the no bead treatment, with the exception of PDGFB at 96 h (R = 1.60, p = 0.009).

109

3.2 The effect of MAA beads on osteopontin expression in HUVEC

HUVEC expression of OPN is shown in Figure 5-7 and the log2NRQ and p-values are given in Table 3. At 91 h, MAA beads increased OPN expression 3.5-fold and 4.0-fold compared to PMMA (p = 0.019) and no bead (p = 0.033). At 24 h all treatments lead to similar expression of OPN (R < 1.5, p > 0.74). No difference was observed between PMMA and no bead at either time point.

3.3 The effect of MAA-treated dTHP-1 conditioned medium on HUVEC migration

The conditioned media from dTHP-1 cultures of the three treatment groups (MAA, PMMA, no bead) were tested in a HUVEC transwell migration assay to determine the effect of MAA- conditioned medium on HUVEC migration. At 96 h, significantly fewer HUVEC (p < 0.001)

110 migrated towards the MAA-conditioned medium in the lower chamber (with PMMA- or no bead-conditioned medium in the upper chamber) than for any other combination of media (Figure 5-8). When the upper and lower chambers were filled with no bead-conditioned medium and PMMA-conditioned medium, the same number of HUVEC migrated through the membrane independent of which medium was in the upper or lower chamber (p = 0.990). No effect of material was seen at 24 or 48 h (p > 0.062) (data not shown).

111

The motivation for this study came from in vivo experiments which demonstrated that MAA beads, but not PMMA beads, increased skin graft survival in rats[10], and increased wound closure rate and blood vessel density in a diabetic wound healing model in mice[11]. The positive effect MAA beads had on wound healing, particularly increasing wound bed vascularity, was exciting but mysterious since no “bioactive” molecules (growth factors, drugs, cells) had been incorporated into the material to induce such a beneficial host response. The effect was attributed to the presence of methacrylic acid in the copolymer since angiogenesis was not seen with PMMA beads of a similar diameter. In the initial development of the MAA beads, low MAA containing beads and gels had less of an angiogenic response.

112

Figure 5-8. Effect of culture medium conditioned by dTHP-1 cells treated with MAA, PMMA or no bead for 96 h on HUVEC migration in Transwell migration assays. The upper and lower chambers of the transwell system were filled with different conditioned medium (as depicted) and the HUVEC migrated from the upper chamber to the lower chamber (e.g. from “no bead” medium to “MAA” medium). At 96 h, HUVEC had significantly reduced migration toward MAA conditioned medium (p < 0.001), compared to that which migrated to PMMA- or no bead-conditioned medium. At 24 and 48 h (not shown) the source of the conditioned medium had no observable effect (p > 0.062) on HUVEC migration. For *, p < 0.001 for either MAA (lower) compared to any of the other pairs.

The sub-cutaneous implant of poly(methacrylic acid -co- butyl methacrylate) scaffolds (45 mol % methacrylic acid) also resulted in increased vessel density in the tissue invading the porous scaffold, compared to scaffolds made of butyl methacrylate alone[26], further supporting that the vascularity was induced by the biomaterial and particularly the MAA component. The challenge was to understand why this material resulted in increased vessel density. In vitro angiogenesis sprouting studies using endothelial cells alone or portions of viable aortic walls showed no difference between MAA and PMMA beads (unpublished results), leading us to conclude that the angiogenic response was a manifestation of the in vivo host response: the MAA beads were causing more vessels to form as part of the “granulation phase” or the vessels that were forming were not regressing as fast or to as great an extent as with PMMA beads. Interestingly, in some

113 of the very first studies focusing on MAA-containing materials, vessels persisted for as long as 12 weeks (unpublished).

We have never seen increased numbers of macrophages near the MAA beads and hence, we set out to determine whether there were changes in macrophage phenotype in response to the presence of MAA beads in vitro. We report here the results with a monocyte cell-line (dTHP-1, ref [27]), which was used because of the need for large, homogeneous cell population to generate sufficient mRNA for analysis. Although dTHP-1 cells closely resemble primary monocyte- derived macrophages in morphology, antigen and receptor expression, and secretory products[27], we are familiar with the differences between dTHP-1 and primary monocytes[28].

As described above, preliminary microarray results for dTHP-1 cells showed only two “hits”: decreased expression of osteopontin (OPN) and increased expression of insulin growth factor binding protein 3 (IGFBP-3) in cells treated with MAA beads compared to cells cultured with PMMA beads or untreated cells. Perhaps as interesting as the two hits was that there was no difference in response for the large multitude of other genes: the effect of MAA beads (relative to PMMA) was subtle. Unfortunately, the limited availability of the custom microarrays at the time limited the number of replicates that could be done and the absence of the “conventional” angiogenesis genes on this particular microarray limited the utility of these microarray results. Nonetheless, these findings motivated this more detailed study of gene expression in these cells, looking beyond these two “hits” while also assessing global effects on HUVEC migration. We have not yet followed up on IGFBP-3 since it does not appear to have much connection to wound healing.

4.1 Gene expression

The presence of PMMA beads, considered here as a “non-bioactive” material, induced very little response from the dTHP-1 cells compared to untreated cells, based on the genes studied. No difference in expression was observed in any genes after 24 or 48 h of treatment. By 96 h, four genes (IL-1b, PDGF-B, TNF-α and VEGF), but not the others, had significantly different (p <

0.05) log2NRQ values and a relative ratio greater than or equal to 1.5 (R = 2.7, 1.6, 1.5, 1.5, respectively), compared to untreated dTHP-1. These results suggest the presence of PMMA beads had a small, delayed effect on dTHP-1 cells growing on TCPS, consistent with a mild

114 activation response of the cells to a biomaterial. Furthermore, these results are consistent with previous data that suggests the PMMA beads did not cause a notable effect in vivo.

Conversely, beads composed of 45 mol% MAA affected the gene expression of dTHP-1 cells, compared to both untreated and PMMA-treated cells. Considering the angiogenic effect observed in vivo, it was surprising to find the beads did not increase the expression of the genes typically associated with angiogenesis (bFGF, VEGF, PDGF-B, TGF-β, CXCL12), until 96 h when VEGF and PDGF-B were slightly elevated. Instead, the beads decreased the expression of OPN and increased the expression of IL1β, IL-6 and TNF-α.

4.2 Osteopontin, IL-1β, IL-6 and TNF-α

The consequences of decreasing OPN expression in macrophages are unknown as the specific role of OPN in wound healing is not clear. Mori et al. found that blocking OPN expression in mouse skin wounds (via antisense oligonucleotides) accelerated wound healing and reduced granulation tissue formation and fibrosis in normal mice[29]. However, Sharma et al. showed that OPN expression in full-thickness wounds in diabetic and normal mice was concomitant with the onset of angiogenesis[16]. Incisional wounds in OPN-null mice revealed decreased debridement, indicative of abnormal macrophage function, altered collagen fibrillogenesis and greater matrix disorganization[14].

During inflammation, OPN acts as a chemotactic molecule and recruits inflammatory cells to the site of injury[30]. It also functions as an adhesive molecule with multiple binding motifs[31], including a cryptic αvβ3 binding site that has been shown to be anti-apoptotic for endothelial cells[32]. While absent on circulating monocytes, OPN is elevated during macrophage maturation and highly expressed by activated macrophages[33]. Interestingly, TNF-α, IL-1β, and IL-6 have all been shown to induce OPN in macrophages[34-36], although here the increased expression of these cytokines in the MAA-treated dTHP-1 cells was associated with a reduction in OPN expression (relative to PMMA beads or no beads). In vitro studies have shown that OPN downregulates iNOS and NO in macrophages[37], is essential for normal TGF-β-induced myofibroblast differentiation[15] and that downregulation of OPN is required for macrophage fusion when forming foreign body giant cells[18]. We have not explored foreign body cell formation in vitro nor have we seen it in vivo. The wide range and variability in results imply that

115 the role of OPN in wound healing and inflammation is complex, involves multiple cell types, and is likely dependent on expression level and timing.

The effect of OPN in endothelial cell biology is better studied, due to the role of OPN in mediating angiogenesis in cancer and inflammatory responses. A number of factors have been shown to induce endothelial cell expression of osteopontin, including VEGF[17], sonic hedgehog[38], IL-1β[39] and angiotensin II[40]. OPN has been shown to directly promote endothelial cell survival and migration, and to induce angiogenesis through the αvβ3-mediated PI3K/AKT and ERK pathway[17]. The increase in HUVEC OPN expression in response to MAA (and decrease in dTHP-1 cells) seen here highlights the difficulty of parsing complex in vivo phenomena into particular in vitro analyses of single cell types.

MAA beads also increased the expression of IL-1β , IL-6 and TNF-α in dTHP-1 cells, in vitro. These cytokines were selected because they are some of the best-studied pro-inflammatory cytokines, and are commonly used to assess the level of inflammatory response in cells and tissues[3,20,41-44]. They are actively involved in wound healing and inflammation, but also play a role in mediating angiogenesis. Carmi et al. showed that IL-1β did not directly activate angiogenesis in endothelial cells, but induced supporting cells (e.g. macrophages) to produce angiogenic factors (e.g. VEGF) and that angiogenesis induced by LPS-activated macrophages was dependent on IL-1β[21]. IL-6, which had been highly expressed in MAA treated dTHP-1 cells at 24 and 48 h, stimulated endothelial cell (EC) migration in vitro[45], while the expression of TNF-α, which blocked EC migration and proliferation in some in vitro studies[46,47], was increased in MAA-treated cells (compared to PMMA and no bead-conditioned medium). Sainson et al. (2008) has also shown that TNF-α primes EC to sprout rapidly upon resolution of inflammation by inducing EC “tip-cell” phenotype, while blocking VEGFR2 signaling and delaying VEGF-induced angiogenesis[48].

4.3 HUVEC migration

The aim of this experiment was to compare the chemoattractive potential of the three conditioned media. It was expected that the differences in gene expression caused by treatment with MAA beads would lead to differences in protein expression that would be evident in a migration assay. All three media were similarly chemoattractive relative to RPMI containing 0.1% BSA, drawing

116 similar numbers of HUVEC from the upper to lower chambers of the Transwell system (data not shown). Like the gene expression differences, there was a subtle difference among the conditioned media that became evident only when the conditioned media were compared against each other. Thus fewer HUVEC were drawn from the PMMA- (or no bead-) conditioned medium towards the MAA-conditioned medium.

Consistent with the absence of gene expression differences, the PMMA- and no bead- conditioned media appeared to have similar chemotactic activity to each other since the direction of migration (e.g. no bead to PMMA; or PMMA to no bead) did not affect the number of HUVEC migrating through the membrane.

We had expected however, that the MAA conditioned medium would be more chemoattactive than the other media given the array of genes that were expressed by the dTHP-1 cells and the pro-angiogenic effect of MAA beads have in vivo. This decreased migration was only observed with the 96 h conditioned medium and at this time point, MAA-treated dTHP-1 cells had decreased OPN and elevated VEGF, TNF-α, IL-1β mRNA expression compared to untreated and PMMA-treated cells. It is presumed, that the decrease in osteopontin expression contributed to the decrease in cell migration. However, decreased migration would have been expected at earlier time points as well (which was not seen) if this were the only factor contributing to the effect. We also note that the conditioned media accumulate all proteins secreted over the 96 h of conditioning while the gene expression results reflect a smaller interval of time. Furthermore it is unclear whether the MAA-conditioned medium actively inhibited migration, or if the HUVEC found this medium less chemoattractive than the other media and were not stimulated to migrate towards the MAA-conditioned medium. All that can be said is that the balance of factors that induce or inhibit EC migration (including factors not examined here) was shifted in the MAA- conditioned medium at 96 h, such that fewer HUVEC migrated through the membrane.

We have examined only a small subset of the genes likely involved in the host response to MAA beads. Expanding this list to include, for example, genes associated with vessel regression as well as looking at protein levels is warranted. Connecting osteopontin down regulation to the expression of other molecules more directly associated with angiogenesis also needs to be a focus of future studies. We also have not yet considered the complexity of the wound environment, where many other cells at comparable times (4 days) and longer times (say 2

117 weeks) play a role in the observed angiogenesis. That MAA beads increase the expression of osteopontin in HUVEC and decrease it in dTHP1 cells point to the complexity of this analysis. Such studies are underway and will be reported separately.

4.4 The biomaterial problem

The increased blood density observed in the wound tissue of diabetic mice treated with 45 mol% MAA beads was evidence that a synthetic biomaterial (without exogenous growth factors) can induce a pro-angiogenic response in vivo. This lead to the delineation of a new class of biomaterials (Therapeutic Polymers, Theramers™) – biomaterials with biologic activity but without bioactive components[11,24,49,50]. The focus in this study was the changes in the cells associated with the host response because of interaction with the MAA beads. An important related question is the biomaterial problem: what is it about this material that induced such a host response?

The effect of biomaterial surface chemistry on host response in the context of wound healing has mainly been studied using charged beads and linear incision models[51-55]. Increased wound breaking strength[51-53], and the increased presence of macrophage[53], and FBGC[56] were associated with positively charged diethylaminethanol (DEAE) beads. Conversely, negatively charged carboxy methyl (CM) Sephadex beads either lowered the wound breaking strength [52] or had no effect[53]. Increased mast cell recruitment was observed for both DEAE and CM Sephadex beads[55]. No effect on angiogenesis was observed for any of the charged beads[53]. In further studies, small diameter polymer fibers coated with methacrylic acid were associated with increased microvessel density in single fiber implant models, compared among positively charged fibres (N,N-dimethylaminoethyl methacrylate) and neutral fibres (hexafluoropropylene), however this increase was not statistically significant[57,58]. While we attribute the observed angiogenesis to the difference in chemistry between MAA and PMMA beads, we also note that PMMA beads are smooth while the MAA beads are rough. Surface topography has been considered a pro-angiogenic factor in 5 µm pore size membranes[59] and recently, the Ratner group described the pro-angiogenic properties of a porous scaffold composed of poly(2- hydroxyethyl methacrylate-co-methacrylic acid) (pHEMA-co-MAA) (5% MAA), designed to contain spherical, interconnected pores (30-40 µm in diameter)[60]. We present no evidence here to distinguish the effect of surface chemistry (and charge) from topography. This topic, as is the

118 related question of what proteins are adsorbed and how they influence the host response, are outside the scope of this study.

5 Conclusions

MAA beads had a distinct effect on the gene expression profile of macrophage-like dTHP-1 cells; an effect that was not observed with PMMA beads. While MAA beads had minimal to no effect on typical angiogenic genes (bFGF, CXCL12, HIF1α, PDGFB, TGFβ and VEGF), they did increase the expression of three cytokines (IL-1β, IL6 and TNF-α) compared to PMMA beads and a no beads control. MAA beads also decreased the expression of OPN in these cells. However the MAA beads increased the expression of OPN in HUVEC suggesting that the pro- angiogenic response observed in vivo will be a product of the interaction among multiple cell types in vivo. The next step to better understand the angiogenic effect of MAA beads is to relate the observed changes in expression in vitro to the complexity of the angiogenic response in vivo.

6 Acknowledgements

7 References [1] Williams DF. On the nature of biomaterials. Biomaterials 2009;30(30):5897-909. [2] Li Y, Schutte RJ, Abu-Shakra A, Reichert WM. Protein array method for assessing in vitro biomaterial-induced cytokine expression. Biomaterials 2005;26(10):1081-5. [3] Jones JA, Chang DT, Meyerson H, Colton E, Kwon IK, Matsuda T, et al. Proteomic analysis and quantification of cytokines and chemokines from biomaterial surface- adherent macrophages and foreign body giant cells. J Biomed Mater Res A 2007;83(3):585-96. [4] Anderson JM, Rodriguez A, Chang DT. Foreign body reaction to biomaterials. Semin Immunol 2008;20(2):86-100. [5] Anand S, Cheresh DA. MicroRNA-mediated regulation of the angiogenic switch. Curr Opin Hematol 2011;18(3):171-6. [6] Fantin A, Vieira JM, Gestri G, Denti L, Schwarz Q, Prykhozhij S, et al. Tissue macrophages act as cellular chaperones for vascular anastomosis downstream of VEGF- mediated endothelial tip cell induction. Blood 2010;116(5):829-40.

119

[7] Lucas T, Waisman A, Ranjan R, Roes J, Krieg T, Muller W, et al. Differential roles of macrophages in diverse phases of skin repair. J Immunol 2010;184(7):3964-77. [8] Goren I, Allmann N, Yogev N, Schurmann C, Linke A, Holdener M, et al. A transgenic mouse model of inducible macrophage depletion: effects of diphtheria toxin-driven lysozyme M-specific cell lineage ablation on wound inflammatory, angiogenic, and contractive processes. Am J Pathol 2009;175(1):132-47. [9] Diez-Roux G, Lang RA. Macrophages induce apoptosis in normal cells in vivo. Development 1997;124(18):3633-8. [10] Eckhaus AA, Fish JS, Skarja G, Semple JL, Sefton MV. A preliminary study of the effect of poly(methacrylic acid-co-methyl methacrylate) beads on angiogenesis in rodent skin grafts and the quality of the panniculus carnosus. Plast Reconstr Surg 2008;122(5):1361- 70. [11] Martin DC, Semple JL, Sefton MV. Poly(methacrylic acid-co-methyl methacrylate) beads promote vascularization and wound repair in diabetic mice. J Biomed Mater Res A 2010;93(2):484-92. [12] Xia Z, Triffitt JT. A review on macrophage responses to biomaterials. Biomed Mater 2006;1(1):R1-R9. [13] Singer AJ, Clark RA. Cutaneous wound healing. N Engl J Med 1999;341(10):738-46. [14] Liaw L, Birk DE, Ballas CB, Whitsitt JS, Davidson JM, Hogan BL. Altered wound healing in mice lacking a functional osteopontin gene (spp1). J Clin Invest 1998;101(7):1468-78. [15] Lenga Y, Koh A, Perera AS, McCulloch CA, Sodek J, Zohar R. Osteopontin expression is required for myofibroblast differentiation. Circ Res 2008;102(3):319-27. [16] Sharma A, Singh AK, Warren J, Thangapazham RL, Maheshwari RK. Differential regulation of angiogenic genes in diabetic wound healing. J Invest Dermatol 2006;126(10):2323-31. [17] Dai J, Peng L, Fan K, Wang H, Wei R, Ji G, et al. Osteopontin induces angiogenesis through activation of PI3K/AKT and ERK1/2 in endothelial cells. Oncogene 2009;28(38):3412-22. [18] Tsai AT, Rice J, Scatena M, Liaw L, Ratner BD, Giachelli CM. The role of osteopontin in foreign body giant cell formation. Biomaterials 2005;26(29):5835-43. [19] Carmeliet P. Mechanisms of angiogenesis and arteriogenesis. Nat Med 2000;6(4):389- 95. [20] Grimstad O, Sandanger O, Ryan L, Otterdal K, Damaas JK, Pukstad B, et al. Cellular sources and inducers of cytokines present in acute wound fluid. Wound Repair Regen 2011;19(3):337-47. [21] Carmi Y, Voronov E, Dotan S, Lahat N, Rahat MA, Fogel M, et al. The role of macrophage-derived IL-1 in induction and maintenance of angiogenesis. J Immunol 2009;183(7):4705-14.

120

[22] Restivo TE, Mace K, Harken AH, Young D. Application of the chemokine CXCL12 expression plasmid restores wound healing to near normal in a diabetic mouse model. J Trauma 2010;69(2):392-8. [23] Gallagher KA, Liu ZJ, Xiao M, Chen H, Goldstein LJ, Buerk DG, et al. Diabetic impairments in NO-mediated endothelial progenitor cell mobilization and homing are reversed by hyperoxia and SDF-1 alpha. J Clin Invest 2007;117(5):1249-59. [24] Eckhaus AA, Fish JS, Skarja G, Semple JL, Sefton MV. A preliminary study of the effect of poly(methacrylic acid-co-methyl methacrylate) beads on angiogenesis in rodent skin grafts and the quality of the panniculus carnosus. Plast Reconstr Surg 2008;122(5):1361- 70. [25] Rieu I, Powers SJ. Real-time quantitative RT-PCR: design, calculations, and statistics. Plant Cell 2009;21(4):1031-3. [26] Butler MJ, Sefton MV. Poly(butyl methacrylate-co-methacrylic acid) tissue engineering scaffold with pro-angiogenic potential in vivo. J Biomed Mater Res A 2007;82(2):265- 73. [27] Auwerx J. The human leukemia cell line, THP-1: a multifacetted model for the study of monocyte-macrophage differentiation. Experientia 1991;47(1):22-31. [28] Vaddi K, Newton RC. Comparison of biological responses of human monocytes and THP-1 cells to chemokines of the intercrine-beta family. J Leukoc Biol 1994;55(6):756- 62. [29] Mori R, Shaw TJ, Martin P. Molecular mechanisms linking wound inflammation and fibrosis: knockdown of osteopontin leads to rapid repair and reduced scarring. J Exp Med 2008;205(1):43-51. [30] Giachelli CM, Lombardi D, Johnson RJ, Murry CE, Almeida M. Evidence for a role of osteopontin in macrophage infiltration in response to pathological stimuli in vivo. Am J Pathol 1998;152(2):353-8. [31] Scatena M, Liaw L, Giachelli CM. Osteopontin: a multifunctional molecule regulating chronic inflammation and vascular disease. Arterioscler Thromb Vasc Biol 2007;27(11):2302-9. [32] Scatena M, Almeida M, Chaisson ML, Fausto N, Nicosia RF, Giachelli CM. NF-kappaB mediates alphavbeta3 integrin-induced endothelial cell survival. J Cell Biol 1998;141(4):1083-93. [33] Krause SW, Rehli M, Kreutz M, Schwarzfischer L, Paulauskis JD, Andreesen R. Differential screening identifies genetic markers of monocyte to macrophage maturation. J Leukoc Biol 1996;60(4):540-5. [34] Bruemmer D, Collins AR, Noh G, Wang W, Territo M, Arias-Magallona S, et al. Angiotensin II-accelerated atherosclerosis and aneurysm formation is attenuated in osteopontin-deficient mice. J Clin Invest 2003;112(9):1318-31. [35] Nakamachi T, Nomiyama T, Gizard F, Heywood EB, Jones KL, Zhao Y, et al. PPARalpha agonists suppress osteopontin expression in macrophages and decrease plasma levels in patients with type 2 diabetes. Diabetes 2007;56(6):1662-70.

121

[36] Ogawa D, Stone JF, Takata Y, Blaschke F, Chu VH, Towler DA, et al. Liver x receptor agonists inhibit cytokine-induced osteopontin expression in macrophages through interference with activator protein-1 signaling pathways. Circ Res 2005;96(7):e59-67. [37] Wai PY, Guo L, Gao C, Mi Z, Guo H, Kuo PC. Osteopontin inhibits macrophage nitric oxide synthesis to enhance tumor proliferation. Surgery 2006;140(2):132-40. [38] Renault MA, Roncalli J, Tongers J, Thorne T, Klyachko E, Misener S, et al. Sonic hedgehog induces angiogenesis via Rho kinase-dependent signaling in endothelial cells. J Mol Cell Cardiol 2010;49(3):490-8. [39] Singh K, Balligand JL, Fischer TA, Smith TW, Kelly RA. Glucocorticoids increase osteopontin expression in cardiac myocytes and microvascular endothelial cells. Role in regulation of inducible nitric oxide synthase. J Biol Chem 1995;270(47):28471-8. [40] Xie Z, Pimental DR, Lohan S, Vasertriger A, Pligavko C, Colucci WS, et al. Regulation of angiotensin II-stimulated osteopontin expression in cardiac microvascular endothelial cells: role of p42/44 mitogen-activated protein kinase and reactive oxygen species. J Cell Physiol 2001;188(1):132-8. [41] Popova A, Kzhyshkowska J, Nurgazieva D, Goerdt S, Gratchev A. Pro- and anti- inflammatory control of M-CSF-mediated macrophage differentiation. Immunobiology 2011;216(1-2):164-72. [42] Mosser DM. The many faces of macrophage activation. J Leukoc Biol 2003;73(2):209- 12. [43] Antonov AS, Antonova GN, Munn DH, Mivechi N, Lucas R, Catravas JD, et al. αVβ3 integrin regulates macrophage inflammatory responses via PI3 Kinase/ Akt-dependent NF-κB activation. J Cell Physiol 2011;226(2):469-76. [44] Xue Y, Liu X, Sun J. PU/PTFE-stimulated monocyte-derived soluble factors induced inflammatory activation in endothelial cells. Toxicol In Vitro 2010;24(2):404-10. [45] Rosen EM, Liu D, Setter E, Bhargava M, Goldberg ID. Interleukin-6 stimulates motility of vascular endothelium. Exs 1991;59:194-205. [46] Fràter-Schröder M, Risau W, Hallmann R, Gautschi P, Böhlen P. Tumor necrosis factor type alpha, a potent inhibitor of endothelial cell growth in vitro, is angiogenic in vivo. Proc Natl Acad Sci U S A 1987;84(15):5277-81. [47] Sato N, Goto T, Haranaka K, Satomi N, Nariuchi H, Mano-Hirano Y, et al. Actions of tumor necrosis factor on cultured vascular endothelial cells: morphologic modulation, growth inhibition, and cytotoxicity. J Natl Cancer Inst 1986;76(6):1113-21. [48] Sainson RC, Johnston DA, Chu HC, Holderfield MT, Nakatsu MN, Crampton SP, et al. TNF primes endothelial cells for angiogenic sprouting by inducing a tip cell phenotype. Blood 2008;111(10):4997-5007. [49] Butler MJ, Sefton MV. Poly(butyl methacrylate-co-methacrylic acid) tissue engineering scaffold with pro-angiogenic potential in vivo. J Biomed Mater Res A 2007;82(2):265- 73. [50] Skarja GA, Brown AL, Ho RK, May M, Sefton M. The effect of a hydroxamic acid- containing polymer on active matrix metalloproteinases. Biomaterials 2009;30(10):1890- 7.

122

[51] Burgess E, Hollinger J, Bennett S, Schmitt J, Buck D, Shannon R, et al. Charged beads enhance cutaneous wound healing in rhesus non-human primates. Plast Reconstr Surg 1998;102(7):2395-403. [52] Tawil NJ, Connors D, Gies D, Bennett S, Gruskin E, Mustoe T. Stimulation of wound healing by positively charged dextran beads depends upon clustering of beads and cells in close proximity to the wound. Wound Repair Regen 1999;7(5):389-99. [53] Wu L, Mockros NE, Casperson ME, Gruskin EA, Ladin DA, Roth SI, et al. Effects of electrically charged particles in enhancement of rat wound healing. J Surg Res 1999;85(1):43-50. [54] Connors D, Gies D, Lin H, Gruskin E, Mustoe TA, Tawil NJ. Increase in wound breaking strength in rats in the presence of positively charged dextran beads correlates with an increase in endogenous transforming growth factor-beta1 and its receptor TGF-betaRI in close proximity to the wound. Wound Repair Regen 2000;8(4):292-303. [55] Sasaki A, Mueller RV, Xi G, Sipe R, Buck D, Hollinger J. Mast cells: an unexpected finding in the modulation of cutaneous wound repair by charged beads. Plast Reconstr Surg 2003;111(4):1446-53. [56] Mustoe TA, Weber DA, Krukowski M. Enhanced healing of cutaneous wounds in rats using beads with positively charged surfaces. Plast Reconstr Surg 1992;89(5):891-7; discussion 8-9. [57] Sanders JE, Cassisi DV, Neumann T, Golledge SL, Zachariah SG, Ratner BD, et al. Relative influence of polymer fiber diameter and surface charge on fibrous capsule thickness and vessel density for single-fiber implants. J Biomed Mater Res A 2003;65(4):462-7. [58] Sanders JE, Lamont SE, Karchin A, Golledge SL, Ratner BD. Fibro-porous meshes made from polyurethane micro-fibers: effects of surface charge on tissue response. Biomaterials 2005;26(7):813-8. [59] Brauker JH, Carr-Brendel VE, Martinson LA, Crudele J, Johnston WD, Johnson RC. Neovascularization of synthetic membranes directed by membrane microarchitecture. J Biomed Mater Res 1995;29(12):1517-24. [60] Madden LR, Mortisen DJ, Sussman EM, Dupras SK, Fugate JA, Cuy JL, et al. Proangiogenic scaffolds as functional templates for cardiac tissue engineering. Proc Natl Acad Sci USA 2010;107(34):15211-6.

123

Chapter 6 dTHP-1 expression of CXCL10 and CXCL12 following treatment with poly(methacrylic acid –co– methylmethacrylate) beads 1 Introduction

Macrophages are phagocytic, innate immune cells that play principal roles in wound healing and the host response to biomaterials. Due to their prominent role directing the foreign body reaction (FBR), macrophage response to different biomaterials has been studied extensively. Many traditional biomaterials elicit a FBR that culminates in the implant being walled off from the body by a fibrous capsule[1]. However, a new paradigm for biomaterial science aims to overcome the traditional FBR and develop biomaterials that integrate with host tissue and have therapeutic effects[2-4]. Previously, we reported on the effect of poly(methacrylic acid –co– methyl methacrylate) (MAA) beads on wound healing in diabetic mice. Contrary to the conventional foreign body reaction, topical application of MAA beads improved wound closure and vascularization of diabetic wounds[5]. Further studies showed MAA beads increased the gene expression of sonic hedgehog (Shh), interleukin (IL)-1β, tumor necrosis factor α (TNF-α) and osteopontin (OPN) in the granulation tissue, while slightly decreasing the expression of CXCL10 (C-X-C motif chemokine 10; also called inducible protein 10, IP-10)[6]. However, MAA beads did not appear to affect the expression of typical pro-angiogenic growth factors, such as vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF), or platelet derived growth factor B (PDGF-B).

To assess the effect MAA beads had on macrophages, macrophage-like cells (dTHP-1) were cultured with MAA beads in vitro[7]. Gene expression analysis revealed that dTHP-1 cells increased the expression of IL-1β, IL-6 and TNF-α, but decreased the expression of OPN in response to MAA beads. MAA beads also had minimal affect on dTHP-1 expression of typical angiogenic growth factors (VEGF, bFGF, PDGF-B and C-X-C motif chemokine 12 (CXCL12; also called stromal derived factor 1 (SDF-1))[7]. Poly(methyl methacrylate) (PMMA) beads, which did not affect wound healing and vascularization in vivo (relative to a no bead, “blank” treatment), had no effect on dTHP-1 gene expression in vitro[5,7], again relative to the same blank and for the genes measured. The PMMA (100 mol % methyl methacrylate) beads were the same size as MAA beads (45 mol % methacrylic acid), but lacked the methacrylic acid

124

component and, seemingly, the therapeutic effect of MAA beads in vivo. Consequently, PMMA beads were used as a biomaterial “control” for the bioactive MAA beads.

The present study continues to focus on the dTHP-1 response to MAA beads by characterizing a profile of twelve secreted proteins, including OPN, CXCL10, IL-1β, IL-6, TNF-α, CXCL12, VEGF, PDGF-BB and bFGF. Two markers of macrophage polarization were also analyzed (alternative activation marker IL-10 and classical activation marker IL-12(p70)), in addition to interferon γ (INF-γ), an inducer of classical activation. The expression of Shh was not analyzed, as dTHP-1 cells did not express detectable levels of Shh mRNA (unpublished data). The effect of dTHP-1 supernatant on endothelial cells was also studied using human umbilical vein endothelial cells (HUVEC). The expression of genes involved in endothelial activation and angiogenesis were analyzed in response to medium conditioned by dTHP-1 treated with MAA, PMMA or no beads for 96 h. Overall, the results suggested that CXCL10 and CXCL12, in addition to previously identified IL-1β, IL-6, TNF-α and Shh, may have a role in modulating the macrophage (and host) response to MAA beads.

2 Materials and methods 2.1 Bead preparation

Methacrylic acid bead (45 mol% methacrylic acid, 1 mol% ethylene glycol dimethacrylate; remainder methyl methacrylate) synthesis and characterization were performed as previously described [7,8]. Briefly, MAA beads were produced by free radical suspension polymerization of the monomers (Sigma-Aldrich, Canada Ltd., Oakville, ON, Canada) in an aqueous calcium chloride solution (0.2 g/ml, Sigma-Aldrich) with hydroxyapatite (0.5 g/g monomers, Sigma- Aldrich) as a dispersing agent, and benzoyl peroxide as the initiator (Sigma-Aldrich). The reaction proceeded under agitation for 5 h at 70 oC. The beads were washed by serial incubations in aqueous solutions (hydrochloric acid, sodium hydroxide and water) and organic solvents (acetone and dimethylformamide) to remove any unreacted materials and low-molecular weight impurities, and then dried under vacuum at 65 oC.

The bulk methacrylic acid content of MAA beads was determine by titration[8]. Beads were incubated in sodium hydroxide for 3 h and the reduction of sodium hydroxide was determined by

125

back-titration with potassium hydrogen phthalate, and used to determine the molar concentration of methacrylic acid.

Poly(methyl methacrylate) beads were acquired from Polysciences (Warrington, PA, USA). Dried MAA and PMMA beads were sieved to obtain beads 150 – 250 µm in diameter. To remove contaminating endotoxin, beads were washed for 20 min in 95% ethanol (MAA beads) or 1 N HCl (PMMA beads) in a sonicating water bath ten times, then rinsed in endotoxin-free water five times. After drying in a vacuum at room temperature, beads were tested for endotoxin contamination using LAL Pyrochrome Endotoxin Kit (Associates of Cape Cod, Falmouth, MA, USA). Only beads with less than 0.25 EU/ml (100 mg beads/ml) were used for experiments.

2.2 Cell culture

As described elsewhere[7], THP-1 cells (TIB-202, ATCC, Manassas, VA, USA) were maintained in THP-1 culture medium (referred to as RPMI/FBS)[RPMI 1640 with 25 mM HEPES and 2 mM L-glutamine (Invitrogen Canada Inc, Burlington, ON, Canada) supplemented with 10% fetal bovine serum (FBS; Invitrogen), 1% penicillin/streptomycin (Invitrogen, final o concentration 100 units/ml of penicillin and 100 mg/ml streptomycin)] at 37 C, 5% CO2. THP-1 cells (1x107 cells/dish) were differentiated into a macrophage-like cell type (dTHP-1) by treatment with 100 nM phorbol myristate acetate (PMA; Sigma-Aldrich) for 48 h in 6 cm tissue culture treated polystyrene (TCPS) dishes (Falcon, Becton Dickinson Canada, Mississauga, ON, Canada). Upon differentiation, dTHP-1 cells attached to the TCPS dish and spread. The dTHP-1 cells were washed with PBS to remove the PMA, then cultured for 1-2 h in fresh RPMI/FBS before the addition of beads.

Human umbilical vein endothelial cells (HUVEC, Lonza, Basel, Switzerland) were cultured in 2 o EGM-2 medium (Lonza) in 75 cm TCPS flasks (Falcon) at 37 C, 5% CO2. HUVEC were passaged every 3-4 days using a splitting ratio of 1:10. HUVEC at passage 3 – 5 were used in the experiments.

2.3 Bead treatment of dTHP-1 cells

Beads were weighed into pyrogen-free microcentrifuge tubes (Axygen, VWR International, Mississauga, ON) and equilibrated by repeated washes in unsupplemented RPMI 1640 (Invitrogen) containing phenol red, until no colour change was detected in the medium[7]. MAA

126

beads (0.3 mg/cm2) or PMMA beads (0.9 mg/cm2) were added to the dTHP-1 cells. The 3:1 ratio in bead amount was used to account for the swelling MAA beads undergo upon hydration. dTHP-1 cells that received no beads were used as controls (referred to as “untreated” or “no bead”). The cells were incubated with the beads for 24 and 96 h in RPMI/FBS. At the end of each time point, the culture medium was centrifuged, filtered to remove any cell or bead debris, and was either used immediately for HUVEC culture or frozen until analyzed for protein content. The dTHP-1 cells were washed with PBS and RNA was isolated. The conditioned media from dTHP-1 treated with MAA, PMMA or no beads were referred to as MAA/dTHP-1, PMMA/dTHP-1 or untreated dTHP-1, respectively.

2.4 Immunoassays

The concentrations of eleven proteins (CXCL12, IL-1β, CXCL10, IL-6, IL-10, IL-12(p70), PDGF-BB, INF-γ, TNF-α, bFGF, VEGF-A) in the 24 and 96 h conditioned medium were quantified using a bead-based Procarta 12-plex immunoassay (Affymetrix, Santa Clara, CA, USA) and analyzed using a Luminex 200 system (Luminex, Toronto, ON, Canada), according to the manufacturer’s instructions (n = 4). Three technical replicates were used for standards and samples. The concentration of secreted OPN was quantified using a human OPN enzyme-linked immunosorbent assay (ELISA) (R&D Systems, Minneapolis, MN, USA), according to the manufacturer’s instructions (n = 4, run in duplicate). Due to the high concentration of OPN in the supernatant, all samples were diluted (1/1000) to ensure they fell within the range of the standard curve.

2.5 HUVEC culture in dTHP-1 conditioned medium

HUVEC were seeded in 6 well TCPS plates (5x105 cells/well) in fresh RPMI/FBS and were allowed to adhere for 2 h. The medium was then replaced with 96 h conditioned RPMI/FBS medium from dTHP-1 cultured with MAA, PMMA or no beads. HUVEC were incubated in the conditioned medium for 6 h, then the cells were washed with PBS and RNA was isolated. HUVEC incubated in fresh, unconditioned RPMI/FBS for 6 h were used as a control.

2.6 RNA isolation

Total RNA was isolated using an RNeasy Mini Kit with on-column DNase digestion (Qaigen Inc, Mississauga, ON, Canada), according to manufacturer’s instructions. RNA content was

127

quantified by absorbance at 260 nm using a Nanodrop ND-1000 spectrophotometer (NanoDrop Technologies, Inc., Wilmington, DE, USA). All RNA samples had a 260/280 ratio greater than 2.0. Total RNA was used for cDNA synthesis, which was performed using a Superscript III First-Strand Synthesis SuperMix (Invitrogen) according to the manufacturer’s instructions.

2.7 Quantitative real-time PCR

Gene-specific primer sequences (Table 6-1) were designed using Primer-blast (http://www.ncbi.nlm.nih.gov/tools/primer-blast/) or were selected from Primerbank (http://pga.mgh.harvard.edu/primerbank/), and all primer pairs were analyzed by Blast to ensure specificity for the intended target. Primers were synthesized by Sigma Genosys (Sigma-Aldrich).

Quantitative real-time PCR (qPCR) was performed using a 7900HT Fast PCR system (Applied Biosystems Canada, Streetsville, ON, Canada) and the following thermal profile: 50 oC for 2 min, 95 oC for 10 min, then 40 cycles of 95 oC for 15 s, 60 oC for 30 s and 72 oC for 30 s. PCR reactions occurred in a final volume of 10 µl, containing 5 µl of SYBR Green Master Mix (Applied Biosystems), 1 µL of 10 mM forward and reverse primers and 10 ng of cDNA template in 4 µl of RNase/DNase free water. A dissociation curve was generated for each plate following the amplification cycles by slowly cooling from 95 oC at a 2% ramp rate.

The qPCR data analysis was performed as described elsewhere[7]. Briefly, the crossing point values (Cp) were determined using the 79000HT Sequence Detection System (SDS) Software (Version 2.3, Applied Biosystems) and the PCR efficiencies (E) were calculated using LinRegPCR software (Version 11.0, download: http://LinRegPCR.HFRC.nl). The relative quantity (RQ = [E^Cp]-1) of each amplicon was normalized to the RQ of GAPDH, the endogenous reference, to obtain the normalized relative quantity (NRQ) for every gene of interest[9]. The relative expression ratio (R) was calculated as the ratio of NRQ(treated) /

NRQ(control)[10]. The results were expressed as the mean relative expression ratio ± standard error (n = 4). Statistical analysis of variance (ANOVA) was performed on the log transformed NRQ

(log2NRQ) to determine if conditioned medium treatment had a significant effect on the mean NRQ for each gene of interest[9]. Genes were considered differentially expressed if the ratio was greater than 1.5 (or less than 0.67) and p < 0.05.

128

Table 6-1. Primer sequences for genes involved in endothelial activation and angiogenesis. GenBank Gene Primer Sequence Accession References Growth factors VEGF-A fwd: 5'-CAACATCACCATGCAGATTATGC-3' NM_113376 [11] rev: 5'-GCTTTCGTTTTTGCCCCTTTC-3' VEGF-B fwd: 5'-GCCACGCATGGGAGTCAGGG-3' NM_003377.4 [11] rev: 5'-CCCGAGGACGCCAACCACAG-3' VEGF-C fwd: 5'-GGTGTTCTGGTGTCCCCCGC-3' NM_005429.2 [11] rev: 5'-CATTCGGAGCCCGCGAGGTG-3' Transcription factors HIF-1A fwd: 5'-TGGGCCCTGACAAGCCACCT-3' NM_001530.3 [12] rev: 5'-AGAGAAGCGGGCGGCAATCG-3' ETS1 fwd: 5'-GGGAAAGAAAGGCAGCGGGAATTTG-3' NM_005238.3 [13] rev: 5'-TCCGGGGAGGGGAAAAGCTCCA-3' HoxD3 fwd: 5'-AGAGTCTCGACAGAACTCCAAG-3' NM_006898 [14] rev: 5'- GCGTTCCGTGAGATTCAGC-3' TR3 fwd: 5'-TTCCATGCCTACGGCCTTCCCA-3' NM_002135.4 [15] rev: 5'-ACACACAGCACAGCGGCCTT-3' Integrins and adhesion molecules beta3 fwd: 5'-ATACCTGGCCCTGTGCCTTGGT-3' NM_000212.2 [14,16] rev: 5'-AGGCCACACGTGCTGATACAACTG-3' beta5 fwd: 5'-CAGGTGGAGGACTATCCTGTG-3' NM_002213 [14,16] rev: 5'-GTGCCGTGTAGGAGAAAGGAG-3' ICAM1 fwd : 5'-GAGCTGAAGCGGCCAGCGAG-3' NM_000201.2 [17] rev: 5'-CCAAGGGGCGGTGCTGCTTT-3' VCAM1 fwd: 5'-GGCACACACAGGTGGGACACA-3' NM_001078.3 [18] rev: 5'-ACAGCCTGTGGTGCTGCAAGT-3' VE-Cadherin fwd: 5'-GACGCCCGGCCTTCCCTCTA-3' NM_001795.3 [19] rev: 5'-TCGTGGTCCGCCTCGTCCTT-3' CD34 fwd: 5'-GCGCTTTGCTTGCTGAGTTT-3' NM_001025109.1 [20] rev: 5'-TCCAAGGGTACTAGGTGTTGTAG-3' Proteases MMP2 fwd: 5'-GGAAAGCCAGGATCCATTTT-3' NM_004530 [21] rev: 5'-ATGCCGCCTTTAACTGGAG-3' MMP9 fwd: '5-AGACGGGTATCCCTTCGACG-3' NM_004994 [21] rev: 5'-AAACCGAGTTGGAACCACGAC-3' uPA fwd: 5'-AGCGACTCCAAAGGCAGCAATGAA-3' NM_002658.3 [16,22] rev: 5'-TTCTTTGGGCAGTTGCACCAGTGA-3' Receptors CXCR4 fwd: 5'-CTTGTCCGTCATGCTTCTCA-3' NM_001008540 [23,24] rev: 5'-GAACCCTGTTTCCGTGAAGA-3' FLT1 fwd: 5'-TCCCTTCCTTCAGTCATGTGT-3' NM_002019 [11,22] rev: 5'-AAGAAGGAAACAGAATCTGCAA-3' VEGFR2 fwd: 5'-GTCCGTCTGGCAGCCTGGATATCC-3' NM_002253.2 [11,22] rev: 5'-TCCCAGCGCCTGTCTAGAGAAGG-3'

129

2.8 Statistical analysis

Statistical analysis of the data was performed using SPSS PASW Statistics 18.0 (release 18.0.0). Statistical comparisons among treatment groups (MAA, PMMA and no bead) were performed using analysis of variance (ANOVA), with α = 0.05. Post hoc analysis between multiple groups was performed using Tukey’s honestly significant difference.

3 Results

Previously, we reported the effect of MAA beads on the gene expression of dTHP-1 cells in vitro, and showed that MAA beads increased the expression of three cytokines (IL-1β, IL-6, TNF-α) while having a minimal effect of the gene expression of VEGF, PDGF-B, bFGF, TGF-β, CXCL12 and HIF-1α[7]. MAA beads also decreased the expression of OPN mRNA in dTHP-1 cells, but increased OPN expression in HUVEC. The current study is focused on the protein secretion profile from dTHP-1 cells treated with MAA, PMMA and no beads. MAA beads were synthesized using free-radical suspension polymerization and were determined to have a bulk methacrylic acid content of 41% by titration. MAA and PMMA beads used for this study had less than 0.25 EU/100 mg.

3.1 dTHP-1 secretion profile

The concentration of eleven proteins (bFGF, IL-10, IL-12(p70), IL-1β, IL-6, INF-γ, CXCL10, PDGF-BB, CXCL12, TNF-α, VEGF-A) in the dTHP-1 supernatant was quantified using a Luminex bead-based immunoassay. A four-fold, 7-point standard curve was generated for each analyte using the 5-parameter logistics (5-PL) algorithm in the Luminex analysis software. The concentration of OPN in the conditioned medium was measured using a commercially available ELISA. A 2-fold, 7-point standard curve was generated using a 5-PL algorithm in Prism GraphPad software. The concentration range of the standards and the coefficients of determination (R2) for each analyte are given in Table 6-2.

Following PMA differentiation, adherent dTHP-1 were washed to remove residual PMA and were cultured in fresh RPMI/FBS for 24 and 96 h. For untreated dTHP-1, OPN had the highest concentration of the selected proteins at 24 h (1761 ± 54 ng/ml) and 96 h (2019 ± 95 ng/ml) (Figure 6-1). TNF-α (9308 ± 125 pg/ml) and IL-1β (1660 ± 81 pg/ml) had the next highest

130

concentration at 24 h, while VEGF-A, CXCL10, CXCL12 and IL-6 were present at moderately high concentrations (300 – 1000 pg/ml). By 96 h, the relative order of protein concentration changed, such that VEGF-A (1479 ± 25 pg/ml) and CXCL10 (928 ± 37 pg/ml) had the highest concentrations (after OPN) and the concentration of TNF-α dropped to approximately 25 pg/ml. CXCL12 (576 ± 7 pg/ml) and IL-1β (504 ± 27 pg/ml) were present at moderate concentrations. At both time points, IL-10, IL-12, INF-γ and bFGF were present at very low concentrations (less than 20 pg/ml). PDGF-BB was below the detection limit (< 5 pg/ml). Fresh RPMI/FBS (i.e. unconditioned medium) had concentrations below the detection limit for all the analytes. The concentrations for all analytes are listed in supplemental Table 6S-1.

Table 6-2. The standard curve range and the coefficient of determination, R2, for secreted proteins. Standard curves were prepared as a 4-fold, 7 point standard dilution and were fit using the 5-PL algorithm, with the exception of OPN, which used a 2-fold, 7 point standard dilution. Standard Curve Range (pg/ml) Analyte R2 Upper Lower bFGF 51900 3.2 0.990 IL-10 20100 1.2 0.992 IL-12(p70) 34400 2.1 0.992 IL-1β 46300 2.8 0.993 IL-6 54400 3.3 0.992 INF-γ 13300 0.8 0.970 CXCL10 25600 1.6 0.985 OPN 20000 312 0.998 PDGF-BB 49700 3.0 0.976 CXCL12 81200 5.0 0.948 TNF-α 28400 1.7 0.975 VEGF-A 47300 2.9 0.981

3.2 Effect of MAA beads on dTHP-1 protein secretion

MAA beads significantly decreased the concentration of CXCL10 in the dTHP-1 supernatant, compared to PMMA and no beads (p < 0.001) at both time points (Figure 6-2). At 24 h, the CXCL10 concentration was reduced by approximately 45% in MAA/dTHP-1 supernatant and was approximately 65% lower at 96 h, compared to PMMA/dTHP-1 and untreated dTHP-1. PMMA had no effect on dTHP-1 protein secretion relative to untreated dTHP-1 (p > 0.05).

131

MAA beads also decreased the expression of a second chemokine, CXCL12 (Figure 6-3); although previous studies had not found any differences in CXCL12 at the gene expression level for dTHP-1 cells treated with MAA beads. The concentration of CXCL12 in the MAA/dTHP-1 supernatant (341 ± 19 pg/ml) was significantly lower than in PMMA/dTHP-1 (449 ± 17 pg/ml, p = 0.002) and untreated dTHP-1 (576 ± 7 pg/ml, p < 0.001) at 96 h. PMMA also decreased CXCL12 concentration in the supernatant compared to untreated dTHP-1 (p = 0.001). No differences in CXCL12 concentration were observed at 24 h.

132

OPN was the most highly expressed protein in the dTHP-1 supernatant (~ 2000 ng/ml) and bead treatment had no effect the concentration of secreted OPN at 24 or 96 h (p > 0.05) (Figure 6-4). dTHP-1 cells also expressed high levels of TNF-α and IL-1β, and moderate levels of IL-6 at 24 h; however the presence of MAA or PMMA beads did not affect the concentration (p > 0.05) (Figure 6-5). By 96 h, the concentration of the three cytokines had decreased (relative to 24 h),

133

and differences were observed among the dTHP-1 treatment groups. Similar to previously reported gene expression data[7], MAA/dTHP-1 supernatant had increased IL-1β, IL-6 and TNF- α concentrations (p < 0.05) at 96 h, compared to PMMA/dTHP-1 and untreated dTHP-1 at 96 h. PMMA beads had no effect at 96 h, compared to no beads.

134

MAA/dTHP-1 supernatant had slightly lower concentrations of VEGF-A and bFGF at 24 h (p < 0.05), compared to PMMA/dTHP-1 and untreated dTHP-1 (Figure 6-6). However at 96 h, the concentration of VEGF-A increased for all treatments, relative to 24 h (p < 0.001), and both MAA/dTHP-1 and PMMA/dTHP-1 had slightly higher concentrations of VEGF-A, compared to untreated dTHP-1 (p < 0.05). No differences among bead treatments were observed for bFGF at 96 h. The biological significance of these small differences is discussed below.

Figure 6-6. The concentration of a) VEGF-A and b) bFGF in the supernatant of dTHP-1 beads cultured with MAA, PMMA or no beads for 24 and 96 h. Generally, dTHP-1 secreted more VEGF-A than bFGF at both time points. Exposure to MAA beads caused a small, but significant decrease in the concentration of VEGF-A and bFGF in the supernatant at 24 h, compared to dTHP-1 cells treated with PMMA or no beads (p < 0.05). At 96 h, dTHP-1 with MAA or PMMA beads had slightly higher concentrations of VEGF- A, compared to no bead (p < 0.05). The concentration of bFGF was similar among all treatments at 96 h. (*, p < 0.05).

3.3 Effect of MAA beads on dTHP-1 protein secretion – macrophage phenotype

The concentrations of IL-10 and IL-12 were analyzed to gain insight into whether MAA beads induced an M2-like phenotype in dTHP-1 cells (Figure 6-7). M1 macrophages generally express IL-12high and IL-10low, while M2 macrophage generally express IL-12low and IL-10high[25]. All dTHP-1 had relatively low concentrations of IL-10 and IL-12, compared to the concentration of other secreted proteins (e.g. IL-1β). When compared among bead treatments, IL-10 and IL-12 concentrations were decreased in MAA treated dTHP-1 supernatant, compared to no bead at 24 h

135

and compared to PMMA and no bead at 96 h (p < 0.05). PMMA treated dTHP-1 also had a decreased concentration of both interleukins at 96 h, relative to untreated dTHP-1. Despite the differences in concentration among treatment groups, the ratio of IL-12 to IL-10 (i.e. IL-12/IL- 10) was similar among all conditioned media (p (ratio) > 0.05).

The concentration of INF-γ, an inducer of the classical macrophage phenotype, was also quantified. INF-γ had low expression (< 10 pg/ml) at 24 and 96 h for untreated dTHP-1 cells (Figure 6-8). dTHP-1 treatment with MAA and PMMA beads decreased the concentration of INF-γ relative to untreated cells (p < 0.05), with MAA/dTHP-1 supernatant having the lowest concentration at both time points, relative to PMMA/dTHP-1 and untreated dTHP-1 supernatants (p < 0.05).

3.4 Effect of MAA-treated dTHP-1 conditioned medium on HUVEC gene expression

The effect of MAA beads on protein concentration in dTHP-1 supernatant prompted the question of how the different conditioned media would affect endothelial cell gene expression in vitro, in particular the expression of CXCL12 receptor CXCR4, given the changes in CXCL12 shown in Figure 6-3. In a previous study, we reported that MAA/dTHP-1 conditioned RPMI/FBS medium reduced the migration of HUVEC in a transwell migration assay, indicating the MAA treated

136

dTHP-1 cells were either secreting a factor(s) that inhibited EC migration, or had a lower amount of chemoattractive factors relative to the PMMA/dTHP-1 and untreated dTHP-1 conditioned medium (i.e. MAA conditioned medium was less chemoattractive than the other media).

The gene expression of HUVEC was analyzed following a 6 h incubation in 96 h MAA/dTHP-1, PMMA/dTHP-1 or untreated dTHP-1 conditioned media, or in fresh RPMI/FBS. Nineteen genes involved in endothelial cell activation and angiogenesis were studied (refer to Table 6-1 for primer sequences). Based on the gene expression profile, incubation with the dTHP-1 conditioned media activated the HUVEC, relative to HUVEC cultured in unconditioned RPMI/FBS (Figure 6-9) and all media had the same effect on 17 of the 19 genes. The exceptions were CXCR4 and VCAM-1 (Figure 6-10). CXCR4 expression was decreased in HUVEC cultured in MAA/dTHP-1 conditioned medium, compared to untreated dTHP-1 cells (ratio = 0.55, p = 0.001), while VCAM-1 had increased expression after MAA/dTHP-1 supernatant incubation compared to untreated dTHP-1 supernatant (ratio = 1.8, p = 0.008). PMMA/dTHP-1 did not affect the expression of any genes, relative to untreated dTHP-1 (p > 0.05).

4 Discussion

The purpose of this study was to characterize the protein expression of several proteins of interest that had been identified though gene expression analysis of dTHP-1 cells and diabetic

137

wounds treated with MAA beads. While changes in mRNA levels were noted in a prior study, here we saw significant changes in the cumulative amount of protein secreted over 24 or 96 h by dTHP-1 cells. Not surprisingly, there were differences between the effect of the beads on protein and mRNA levels. One caveat however, is that both the beads and cells used in this set of experiments were from different lots than those used in earlier studies[6,7]. Generally, the new lot of MAA beads elicited a similar response compared to the previous MAA bead lot, but the new lot of PMMA beads had a more pronounced effect on dTHP-1 gene expression than previously reported (refer to supplemental Figure 6-S1 for dTHP-1 gene expression for this study).

138

4.1 dTHP-1 without bead treatment

OPN was the most highly secreted protein, exceeding the concentration of other proteins by an order of magnitude. There is considerable variability in OPN expression reported in the literature, based on the type of cell, method and duration of activation, and cell density, however, the amount of secreted OPN is reasonably consistent with previously reported values for macrophages, macrophage-like cells and in plasma[26-29]. Also, PMA is known to be a particularly strong inducer of OPN expression [27,30,31]: here strong OPN expression was seen at the mRNA level (Cp ~ 18), compared to TNF-α, IL-1β, IL-6 and VEGF (Cp ~ 25 – 31) (data not shown).

The remaining protein profile in the dTHP-1 supernatant at 24 h is consistent with the effect of PMA activation (a “pro-inflammatory response”), with the high expression of TNF-α, IL-1β, and to a lesser extent IL-6. This is consistent with other studies that induced THP-1 differentiation with PMA[32]. By 96 h, the “pro-inflammatory response” has subsided, as characterized by the decrease in TNF-α, IL-1β, IL-6 and INF-γ concentrations, and the increase in proteins involved in regulating angiogenesis, VEGF and CXCL10. IL-12, IL-10, bFGF, PDGF-BB and INF-γ were present at very low concentrations in the supernatant at both time points.

139

4.2 The effect of MAA beads on dTHP-1 protein expression

The most striking effect of MAA beads on protein expression was the reduction in CXCL10 secreted into the supernatant, compared to both PMMA-treated and untreated dTHP-1 conditioned medium (Figure 6-2). A decrease in CXCL10 mRNA had been noted in the diabetic wounds, although it was not significant[6]. The results suggest that decreased CXCL10 may play an important role in MAA-induced angiogenesis, as CXCL10 is a potent anti-angiogenic factor that inhibits neovascularization and promotes vessel regression in vivo, even in the presence of VEGF[33,34]; its reduction is consistent with an increase in vascularization. CXCL10 is also a chemoattractant for monocytes, macrophages and lymphocytes, and is highly expressed in M1 macrophages[35,36]. It is induced by Toll-like receptor (TLR) activation of INF regulatory factor 3 (IFR3) and subsequently STAT1, (signal transducer and activator of transcription 1) [37,38]. During wound healing, expression of CXCL10 is significantly upregulated within 12 h of wounding, and loss of macrophages (in PU.1 null mice) abrogates its expression within the wound[39], suggesting macrophages are the main source of CXCL10 during wound healing.

An interesting response of dTHP-1 cells to MAA beads was the decreased expression of CXCL12 at 96 h (but not 24 h). Although recent research has focused on the role CXCL12 plays in recruiting and homing bone marrow derived progenitor cells involved in angiogenesis and tissue repair, it was originally isolated as a chemokine that potently induced lymphocyte and monocyte chemotaxis[40]. It is possible that reduction of CXCL12 reduces the leukocyte recruitment to the wound bed in MAA-treated wounds. However, this idea is contrary to evidence obtained in previous diabetic wound models and air-pouch models which reported equal or greater recruitment of inflammatory cells in animals treated with MAA beads, compared to PMMA beads and no beads[5,41]. However, the expression of CXCL12 in small diabetic wounds (chapter 4) was similar for all treatment groups at day 4 and day 7 (unpublished data) and we have not measured the amount of CXCL12 in the wound tissue or in the systemic circulation following wounding and bead treatment.

Previous in vitro evidence suggested that MAA/dTHP-1 supernatant was less chemotactic for HUVEC than PMMA/dTHP-1 or untreated dTHP-1 supernatant[7], although this was contrary to the expected effect based on the increased vascularization observed in vivo with MAA beads[5,6]. The decrease in HUVEC migration was observed with 96 h conditioned medium, but

140

not 24 h conditioned medium, which is consistent with the decrease in CXCL12 protein expression. Assuming a causative relationship, these results suggested that the decreased CXCL12 played a role in the decrease in HUVEC migration in response to MAA beads. This hypothesis was supported by decreased gene expression of CXCR4 in HUVEC cultured in dTHP-1/MAA conditioned medium (Figure 6-10). CXCL12 and CXCR4 contribute directly to angiogenesis through their role in endothelial cell biology, but also play a central role in the mobilization and recruitment of bone marrow derived progenitor cells, and impairment of this latter process is typically associated with impaired wound healing[23,42-44]. At this point, it is not known if MAA beads reduce CXCL12 expression in vivo. Further studies are required to address these questions, and to test the hypothesis that decreased CXCL12 in MAA/dTHP-1 conditioned medium actually contributed to the reduced migration in HUVEC.

Although there were statistically significant differences in the concentration of VEGF (and bFGF) among the difference supernatants, the differences was less than 10% for both time points, and it is unclear if such small differences would have any notable biological effect. The strong reduction of CXCL10 expression, coupled with the lack of effect on VEGF, bFGF and PDGF-B gene expression in the wound model, suggest that MAA beads exert their angiogenic effect through modulation of CXCL10 (and other factors such as Shh[6]) rather than modulating the expression of conventional angiogenic growth factors. This concept is discussed further, along with an overall picture of MAA bead angiogenesis, in the summary chapter (Chapter 7).

The small increases in TNF-α, IL-1β and IL-6 concentration in MAA/dTHP-1 supernatant at 96 h are consistent with gene expression data from dTHP-1 cells in vitro and MAA treated diabetic wound tissue in vivo[6,7]. The implication of increased expression of these cytokines have been discussed in detail elsewhere[7]. In brief, all three cytokines have potent pro-inflammatory activity and are expressed by activated macrophages (particularly M1-like macrophages). However, they also have roles in promoting angiogenesis, through direct and indirect mechanisms. As inflammation typically has negative connotations when referred to with respect to biomaterial responses and chronic wounds, it is important to emphasize that inflammation plays an important role in setting up subsequent stages of normal wound healing, and that it is excessive/chronic inflammation that garners concern[45].

141

MAA beads had no effect on the expression of secreted OPN by dTHP-1, contrary to the decrease in OPN mRNA consistently observed in dTHP-1 cells cultured with MAA beads[7]. A possible explanation for this discrepancy is that OPN has a secreted and an intracellular isoform, both of which are generated from the same mRNA sequence[46]. Consequently, the gene expression data represents the expression of both isoforms, suggesting that the intracellular isoform is decreased in MAA-treated dTHP-1. The secreted OPN isoform (sOPN) was discovered first and has been studied extensively, while the role of intracellular OPN (iOPN) is still being elucidated (reviewed by Inoue et al.[47]). However, there is increasing evidence that iOPN is a critical component in the downstream signaling of multiple TLR through association with Myd88, an adaptor molecule[48,49]. The intracellular isoform is also involved in cell migration and cytoskeletal rearrangement through association with CD44[50-52]. Macrophages constitutively express both isoforms of OPN, and there is evidence that macrophages produce higher levels of iOPN, while lymphocytes produce more sOPN upon activation[47,49].

However, distinguishing the relative amounts of each isoform of OPN within cells remains challenging. The amount of OPN inside the cell represents iOPN within the cytoplasm and peri- membrane area, and sOPN in the endoplasmic reticulum and Golgi[47]. The murine iOPN isoforms have been identified using western blotting techniques, although it has a very similar size to sOPN and appears to be difficult to distinguish[46]. The precise location of human iOPN has not been identified on SDS-PAGE gel[47], however OPN has been identified in the cytoplasm of human cells using histological and confocal techniques[53,54]. Future experiments will address the effect of MAA beads on iOPN expression, or at a minimum, quantify the amount of total OPN within MAA-treated cells.

4.3 Effect of MAA beads on macrophage phenotype

Regardless of bead treatment, dTHP-1 cells secreted large quantities of three cytokines associated with classical activation (IL-6, IL-1β and TNF-α), and MAA-treated cells had the highest concentrations of these cytokines in the supernatant at 96 h. This suggests that MAA beads polarized dTHP1 cells in a classical manner. However, MAA treated dTHP-1 also secreted considerably less CXCL10 and CXCL12 (than untreated medium); these are two chemokines associated with classical and alternative activation, respectively. The expression of IL-10, IL-12 and INF-γ did not clarify the phenotype question, as all three had very low expression, and all

142

dTHP-1 cells had a similar ratio of IL-12 / IL-10 in the supernatant. MAA beads lowered the concentration of IL-10, IL-12 and INF-γ in the supernatant, indicating the presence of the MAA beads modulated the expression of these factors. However this did not enable classification of the dTHP-1 cells as being polarized towards either classical or alternative activation. Further characterization of macrophage phenotype using cell surface receptors and additional macrophage polarization markers, such as mannose receptor (CD206), CC-chemokine ligand 1 (CCL1), inducible nitric oxide synthase (iNOS), found in inflammatory zone 1 (Fizz1) and eosinophil chemotactic factor (Ym1/ECF), may provide additional insight, however is it likely that MAA-treated macrophages fall within the grey-zone between M1- and M2-like phenotypes.

In a separate study, macrophages in the exudate surrounding implanted gelatin discs with MAA beads showed a similar mixed response, in that the MAA-exudate macrophages had increased gene expression of mannose receptor C type lectin (an alternative marker) but also expressed high levels of IL-6 (typically a classical marker, and highly expressed by MAA-treated dTHP-1 cells)[41], relative to PMMA-exudate macrophages. However, the overall gene expression of MAA and PMMA-exudate macrophages suggested a more classical activation profile initially (day 4) and then switched to a more alternative profile at day 10. The concentration of IL-10 and IL-12 in the exudate (contributed to by multiple cell types, including macrophages) supported the shift from classical-like to alternative-like activation from day 4 to 10 for MAA-exudate, while PMMA-exudate contained similar amounts of IL-10 and IL-12 at day 4 and day 10[41]. In contrast to the in vitro dTHP-1 experiment, the MAA-exudate contained higher amounts of IL-12 (although not significant) and lower amounts of IL-10 than PMMA-exudate. However the exudate represents the amount of secreted interleukins from all the cell types present at an implant site containing beads and a gelatin disc, which makes it challenging to compare to the supernatant of an in vitro monoculture of macrophage-like cells.

4.4 Effect of MAA-treated dTHP-1 conditioned medium on HUVEC

Incubation in (untreated) dTHP-1 conditioned medium increased the expression of multiple genes involved in angiogenesis and endothelial cell activation. This result was expected as the dTHP-1 conditioned medium contained pro-angiogenic and inflammatory factors. However, despite the differences in growth factor and cytokine concentrations among the different

143

conditioned media (MAA-treated, PMMA-treated and untreated), only two of the genes we studied were differentially expressed. The increased expression of VCAM-1 suggests that the HUVEC cultured in MAA-treated dTHP-1 conditioned medium were more activated than HUVEC cultured in PMMA-treated or untreated dTHP-1 conditioned medium. VCAM-1 is an adhesion molecule that enables leukocyte extravasation and is expressed by activated endothelial cells[18]. This is consistent with increased expression of numerous cytokines and growth factors in the MAA/dTHP-1 conditioned medium. However, it is not clear why other markers of endothelial cell activation (e.g. ICAM-1) were not increased by MAA/dTHP-1.

As discussed earlier, the decrease in CXCR4 expression in HUVEC treated with MAA/dTHP-1 conditioned medium likely reflects the decreased concentration of CXCR4 ligand, CXCL12, in MAA/dTHP-1 supernatant. The reduction in CXCL12/CXCR4 signaling in HUVEC was thought to contribute to the decreased migration of HUVEC observed in response to MAA/dTHP-1 conditioned medium in earlier experiments (Chapter 5, [7]). In endothelial cells, CXCL12/CXCR4 signaling contributes to angiogenesis through activation of tip cells, and induces endothelial cell migration and proliferation in vitro[23,24]. Consequently, the decrease in CXCL12/CXCR4 is expected to contribute to the decreased HUVEC migration observed with MAA/dTHP-1 supernatant.

5 Future Directions

This chapter is an incomplete manuscript, currently under preparation for peer review and publication. The data presented here raised two major questions that will be addressed as part of the final version of this manuscript.

5.1 Effect of MAA beads on intracellular OPN

The discrepancy between the effect of MAA beads on OPN mRNA and extracellular sOPN raised the question of whether MAA beads modulate the expression of iOPN. As mentioned above, the human iOPN-associated band has not been identified using immunoblotting techniques[47]. Quantification of total OPN in the cell lysate using immunoblotting may provide insight into iOPN levels: MAA-treated cells are expected to have decreased total OPN in the lysate, since there is reduced gene expression and there were similar amounts of extracellular sOPN among the bead treatments (reflecting a similar amount of sOPN inside the cells).

144

However, if no differences are observed, iOPN may be qualitatively studied using confocal microscopy[53]. Alternatively, the use of murine macrophages may allow visualization of sOPN and iOPN bands from cell lysates using immunoblotting techniques. However the difference in band molecular weight is approximately 5 kDa, which may make quantification challenging [46].

As mouse iOPN uses an alternative initiation site within the 40 – 48 nucleotide region of the mRNA sequence, a mutant OPN lacking the first 39 nucleotides has been used by Shinohara et al. to express only the iOPN isoform[46]. Transfecting macrophages isolated from OPN-null mice with Δ1-39nt mutant OPN plasmid may provide a model to study the MAA-effect on iOPN expression in vitro. While is it beyond the scope of this particular thesis, the in vivo role of iOPN in MAA-mediated angiogenesis and wound healing could be studied using Opn-/- mice[48,55] and delivery of exogenous full-length OPN (i.e. sOPN).

5.2 Effect of MAA-reduced CXCL12 concentration on HUVEC migration

The hypothesis that the decreased CXCL12 concentration in MAA/dTHP-1 conditioned medium contributed to the reduced HUVEC migration in the transwell system (described elsewhere[7]) will be tested by blocking the CXCR4 receptor in HUVEC using AMD3100, a CXCR4-specific peptide antagonist and repeating the migration assay[56]. Future studies may also include studying bone marrow progenitor cell mobilization and engraftment at the wound site, and CXCL12 expression in the diabetic wound healing model, although this set of experiments is beyond the scope of this thesis.

6 Conclusions

The purpose of this study was to characterize the dTHP-1 protein expression in response to MAA beads, following up on previous gene expression data. MAA beads drastically reduced the concentration of CXCL10, a potent anti-angiogenic chemokine, but also reduced the concentration of the angiogenic chemokine CXCL12. When exposed to MAA/dTHP-1 conditioned medium, HUVEC had decreased expression of the CXCL12 receptor, CXCR4 (compared to HUVEC cultured in PMMA/dTHP-1 and untreated dTHP-1 conditioned medium), which may explain the reduced HUVEC migration in response to MAA/dTHP-1 conditioned medium observed in an earlier study (Chapter 5). Protein analysis confirmed that MAA beads

145

increased the expression of IL-1β, IL-6 and TNF-α in 96 h (but not 24 h) dTHP-1 conditioned medium consistent with gene expression. Interestingly, MAA beads did not affect the concentration of secreted OPN, despite consistently decreasing the expression of OPN mRNA, suggesting that MAA beads may modulate the expression of the intracellular OPN isoform. Further studies are required to determine the effect of MAA beads on intracellular OPN, and to test if MAA-decreased CXCL12 concentration contributed to the reduced HUVEC migration in vitro.

7 References [1] Anderson JM, Rodriguez A, Chang DT. Foreign body reaction to biomaterials. Semin Immunol 2008;20(2):86-100. [2] Madden LR, Mortisen DJ, Sussman EM, Dupras SK, Fugate JA, Cuy JL, et al. Proangiogenic scaffolds as functional templates for cardiac tissue engineering. Proc Natl Acad Sci USA 2010;107(34):15211-6. [3] Butler MJ, Sefton MV. Poly(butyl methacrylate-co-methacrylic acid) tissue engineering scaffold with pro-angiogenic potential in vivo. J Biomed Mater Res A 2007;82(2):265- 73. [4] Fleckman P, Usui M, Zhao G, Underwood R, Maginness M, Marshall A, et al. Cutaneous and inflammatory response to long-term percutaneous implants of sphere-templated porous/solid poly(HEMA) and silicone in mice. J Biomed Mater Res A 2012;100(5):1256-68. [5] Martin DC, Semple JL, Sefton MV. Poly(methacrylic acid-co-methyl methacrylate) beads promote vascularization and wound repair in diabetic mice. J Biomed Mater Res A 2010;93(2):484-92. [6] Fitzpatrick LE, Lisovsky A, Sefton MV. The expression of sonic hedgehog in diabetic wounds following treatment with poly(methacrylic acid-co-methyl methacrylate) beads. Biomaterials 2012;33(21):5297-307. [7] Fitzpatrick LE, Chan JWY, Sefton MV. On the mechanism of poly(methacrylic acid -co- methyl methacrylate)-induced angiogenesis: Gene expression analysis of dTHP-1 cells. Biomaterials 2011;32:8957-67. [8] Eckhaus AA, Fish JS, Skarja G, Semple JL, Sefton MV. A preliminary study of the effect of poly(methacrylic acid-co-methyl methacrylate) beads on angiogenesis in rodent skin grafts and the quality of the panniculus carnosus. Plast Reconstr Surg 2008;122(5):1361- 70. [9] Rieu I, Powers SJ. Real-time quantitative RT-PCR: design, calculations, and statistics. Plant Cell 2009;21(4):1031-3. [10] Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 2001;29(9):e45.

146

[11] Carmeliet P, Jain RK. Molecular mechanisms and clinical applications of angiogenesis. Nature 2011;473(7347):298-307. [12] Toffoli S, Roegiers A, Feron O, Van Steenbrugge M, Ninane N, Raes M, et al. Intermittent hypoxia is an angiogenic inducer for endothelial cells: role of HIF-1. Angiogenesis 2009;12(1):47-67. [13] Heo SH, Choi YJ, Ryoo HM, Cho JY. Expression profiling of ETS and MMP factors in VEGF-activated endothelial cells: role of MMP-10 in VEGF-induced angiogenesis. J Cell Physiol 2010;224(3):734-42. [14] Boudreau NJ, Varner JA. The homeobox transcription factor Hox D3 promotes integrin alpha5beta1 expression and function during angiogenesis. J Biol Chem 2004;279(6):4862-8. [15] Zeng H, Qin L, Zhao D, Tan X, Manseau EJ, Van Hoang M, et al. Orphan nuclear receptor TR3/Nur77 regulates VEGF-A-induced angiogenesis through its transcriptional activity. J Exp Med 2006;203(3):719-29. [16] Boudreau N, Andrews C, Srebrow A, Ravanpay A, Cheresh DA. Induction of the angiogenic phenotype by Hox D3. J Cell Biol 1997;139(1):257-64. [17] Lester EA, Babensee JE. Proinflammatory phenotype of endothelial cells after coculture with biomaterial-treated blood cells. J Biomed Mater Res A 2003;64(3):397-410. [18] Fries JW, Williams AJ, Atkins RC, Newman W, Lipscomb MF, Collins T. Expression of VCAM-1 and E-selectin in an in vivo model of endothelial activation. Am J Pathol 1993;143(3):725-37. [19] Bach TL, Barsigian C, Chalupowicz DG, Busler D, Yaen CH, Grant DS, et al. VE- Cadherin mediates endothelial cell capillary tube formation in fibrin and collagen gels. Exp Cell Res 1998;238(2):324-34. [20] Siemerink MJ, Klaassen I, Vogels IM, Griffioen AW, Van Noorden CJ, Schlingemann RO. CD34 marks angiogenic tip cells in human vascular endothelial cell cultures. Angiogenesis 2012;15(1):151-63. [21] Dao Thi MU, Trocme C, Montmasson MP, Fanchon E, Toussaint B, Tracqui P. Investigating metalloproteinases MMP-2 and MMP-9 mechanosensitivity to feedback loops involved in the regulation of in vitro angiogenesis by endogenous mechanical s tresses. Acta Biotheor 2012;60(1-2):21-40. [22] Witmer AN, van Blijswijk BC, van Noorden CJ, Vrensen GF, Schlingemann RO. In vivo angiogenic phenotype of endothelial cells and pericytes induced by vascular endothelial growth factor-A. J Histochem Cytochem 2004;52(1):39-52. [23] Unoki N, Murakami T, Nishijima K, Ogino K, van Rooijen N, Yoshimura N. SDF- 1/CXCR4 contributes to the activation of tip cells and microglia in retinal angiogenesis. Invest Ophthalmol Vis Sci 2010;51(7):3362-71. [24] Kuhlmann CR, Schaefer CA, Reinhold L, Tillmanns H, Erdogan A. Signalling mechanisms of SDF-induced endothelial cell proliferation and migration. Biochem Biophys Res Commun 2005;335(4):1107-14.

147

[25] Brown BN, Ratner BD, Goodman SB, Amar S, Badylak SF. Macrophage polarization: an opportunity for improved outcomes in biomaterials and regenerative medicine. Biomaterials 2012;33(15):3792-802. [26] Li X, O'Regan AW, Berman JS. IFN-gamma induction of osteopontin expression in human monocytoid cells. J Interferon Cytokine Res 2003;23(5):259-65. [27] Atkins K, Berry JE, Zhang WZ, Harris JF, Chambers AF, Simpson RU, et al. Coordinate expression of OPN and associated receptors during monocyte/macrophage differentiation of HL-60 cells. J Cell Physiol 1998;175(2):229-37. [28] Nakamachi T, Nomiyama T, Gizard F, Heywood EB, Jones KL, Zhao Y, et al. PPARalpha agonists suppress osteopontin expression in macrophages and decrease plasma levels in patients with type 2 diabetes. Diabetes 2007;56(6):1662-70. [29] Sato T, Nakai T, Tamura N, Okamoto S, Matsuoka K, Sakuraba A, et al. Osteopontin/Eta-1 upregulated in Crohn's disease regulates the Th1 immune response. Gut 2005;54(9):1254-62. [30] Oyama Y, Akuzawa N, Nagai R, Kurabayashi M. PPARgamma ligand inhibits osteopontin gene expression through interference with binding of nuclear factors to A/T- rich sequence in THP-1 cells. Circ Res 2002;90(3):348-55. [31] Suzuki H, Forrest AR, van Nimwegen E, Daub CO, Balwierz PJ, Irvine KM, et al. The transcriptional network that controls growth arrest and differentiation in a human myeloid leukemia cell line. Nat Genet 2009;41(5):553-62. [32] Daigneault M, Preston JA, Marriott HM, Whyte MK, Dockrell DH. The identification of markers of macrophage differentiation in PMA-stimulated THP-1 cells and monocyte- derived macrophages. PLoS One 2010;5(1):e8668. [33] Bodnar RJ, Yates CC, Wells A. IP-10 blocks vascular endothelial growth factor-induced endothelial cell motility and tube formation via inhibition of calpain. Circ Res 2006;98(5):617-25. [34] Bodnar RJ, Yates CC, Rodgers ME, Du X, Wells A. IP-10 induces dissociation of newly formed blood vessels. J Cell Sci 2009;122(12):2064-77. [35] Mosser DM, Edwards JP. Exploring the full spectrum of macrophage activation. Nat Rev Immunol 2008;8(12):958-69. [36] Martinez FO, Gordon S, Locati M, Mantovani A. Transcriptional profiling of the human monocyte-to-macrophage differentiation and polarization: new molecules and patterns of gene expression. J Immunol 2006;177(10):7303-11. [37] Toshchakov V, Jones BW, Perera PY, Thomas K, Cody MJ, Zhang S, et al. TLR4, but not TLR2, mediates IFN-beta-induced STAT1alpha/beta-dependent gene expression in macrophages. Nat Immunol 2002;3(4):392-8. [38] Rivieccio MA, John GR, Song X, Suh HS, Zhao Y, Lee SC, et al. The cytokine IL-1beta activates IFN response factor 3 in human fetal astrocytes in culture. J Immunol 2005;174(6):3719-26. [39] Cooper L, Johnson C, Burslem F, Martin P. Wound healing and inflammation genes revealed by array analysis of 'macrophageless' PU.1 null mice. Genome Biol 2005;6(1):R5.

148

[40] Bleul CC, Fuhlbrigge RC, Casasnovas JM, Aiuti A, Springer TA. A highly efficacious lymphocyte chemoattractant, stromal cell-derived factor 1 (SDF-1). J Exp Med 1996;184(3):1101-9. [41] Patel RJ, Sefton MV. Some aspects of the host response to methacrylic acid containing beads in a mouse air pouch. J Biomed Mater Res A 2012;100(8):2054-62. [42] Fiorina P, Pietramaggiori G, Scherer SS, Jurewicz M, Mathews JC, Vergani A, et al. The mobilization and effect of endogenous bone marrow progenitor cells in diabetic wound healing. Cell Transplant 2010;19(11):1369-81. [43] Gallagher KA, Liu Z-J, Xiao M, Chen H, Goldstein LJ, Buerk DG, et al. Diabetic impairments in NO-mediated endothelial progenitor cell mobilization and homing are reversed by hyperoxia and SDF-1α. J Clin Invest 2007;117(5):1249-59. [44] Morris LM, Klanke CA, Lang SA, Pokall S, Maldonado AR, Vuletin JF, et al. Characterization of endothelial progenitor cells mobilization following cutaneous wounding. Wound Repair Regen 2010;18(4):383-90. [45] Lucas T, Waisman A, Ranjan R, Roes J, Krieg T, Muller W, et al. Differential roles of macrophages in diverse phases of skin repair. J Immunol 2010;184(7):3964-77. [46] Shinohara ML, Kim HJ, Kim JH, Garcia VA, Cantor H. Alternative translation of osteopontin generates intracellular and secreted isoforms that mediate distinct biological activities in dendritic cells. Proc Natl Acad Sci U S A 2008;105(20):7235-9. [47] Inoue M, Shinohara ML. Intracellular osteopontin (iOPN) and immunity. Immunol Res 2011;49(1-3):160-72. [48] Inoue M, Moriwaki Y, Arikawa T, Chen YH, Oh YJ, Oliver T, et al. Cutting edge: critical role of intracellular osteopontin in antifungal innate immune responses. J Immunol 2011;186(1):19-23. [49] Shinohara ML, Lu L, Bu J, Werneck MB, Kobayashi KS, Glimcher LH, et al. Osteopontin expression is essential for interferon-alpha production by plasmacytoid dendritic cells. Nat Immunol 2006;7(5):498-506. [50] Zohar R, Suzuki N, Suzuki K, Arora P, Glogauer M, McCulloch CA, et al. Intracellular osteopontin is an integral component of the CD44-ERM complex involved in cell migration. J Cell Physiol 2000;184(1):118-30. [51] Zhu B, Suzuki K, Goldberg HA, Rittling SR, Denhardt DT, McCulloch CA, et al. Osteopontin modulates CD44-dependent chemotaxis of peritoneal macrophages through G-protein-coupled receptors: evidence of a role for an intracellular form of osteopontin. J Cell Physiol 2004;198(1):155-67. [52] Suzuki K, Zhu B, Rittling SR, Denhardt DT, Goldberg HA, McCulloch CA, et al. Colocalization of intracellular osteopontin with CD44 is associated with migration, cell fusion, and resorption in osteoclasts. J Bone Miner Res 2002;17(8):1486-97. [53] Junaid A, Moon MC, Harding GE, Zahradka P. Osteopontin localizes to the nucleus of 293 cells and associates with polo-like kinase-1. Am J Physiol Cell Physiol 2007;292(2):C919-26.

149

[54] Chabas D, Baranzini SE, Mitchell D, Bernard CC, Rittling SR, Denhardt DT, et al. The influence of the proinflammatory cytokine, osteopontin, on autoimmune demyelinating disease. Science 2001;294(5547):1731-5. [55] Sato I, Yamamoto N, Rittling SR, Denhardt DT, Hino M, Morimoro J, et al. Osteopontin is dispensable for protection against high load systemic fungal infection. Int Immunopharmacol 2008;8(10):1441-8. [56] Zheng H, Fu G, Dai T, Huang H. Migration of endothelial progenitor cells mediated by stromal cell-derived factor-1alpha/CXCR4 via PI3K/Akt/eNOS signal transduction pathway. J Cardiovasc Pharmacol 2007;50(3):274-80.

150

8 Supplemental data Table 6-S1. Concentration of analytes in dTHP-1 conditioned medium at 24 and 96 h (pg/ml, unless otherwise noted) Analyte Time No bead PMMA MAA 24 h 653.7 ± 28.8 637.1 ± 27.7 361.8 ± 8.2 CXCL10 96 h 928 ± 36.8 1038.8 ± 48.2 241.9 ± 42.2 24 h 16.1 ± 0.5 16.1 ± 0.4 14.4 ± 0.3 bFGF 96 h 13.1 ± 0.3 11.9 ± 0.5 11.1 ± 0.8 24 h 4.8 ± 0.2 4.4 ± 0.1 4 ± 0.1 IL-10 96 h 3.1 ± 0 2.6 ± 0.1 2 ± 0 24 h 8.9 ± 0.3 8.2 ± 0.1 7.4 ± 0.2 IL-12(p70) 96 h 4.1 ± 0.1 3.1 ± 0.2 2.4 ± 0.1 24 h 1660 ± 81.2 1849.5 ± 100.5 1679.3 ± 104 IL-1β 96 h 503.9 ± 27.3 602.5 ± 20.1 867.1 ± 49.5 24 h 300.7 ± 4.1 320.3 ± 11.4 329.6 ± 14.6 IL-6 96 h 49.4 ± 2 65 ± 5.8 128.4 ± 14.7 24 h 6.7 ± 0.2 6.1 ± 0.1 5.5 ± 0.1 INF-γ 96 h 3.8 ± 0 3 ± 0.1 2.3 ± 0.1 24 h 1761 ± 54 ng/ml 1742 ± 94 ng/ml 1674 ± 39 ng/ml OPN 96 h 2019 ± 95 ng/ml 1656 ± 73 ng/ml 1805 ± 120 ng/ml 24 h 350.1 ± 3.9 356.6 ± 5.2 347.4 ± 3.8 CXCL12 96 h 575.7 ± 7.3 449 ± 16.8 340.9 ± 18.8 24 h 9308.4 ± 125.3 10904.4 ± 566.5 11034.8 ± 679.1 TNF-α 96 h 25.3 ± 3 37.1 ± 7.4 71.1 ± 11.6 24 h 948.7 ± 17 947.3 ± 3.1 876.3 ± 5.1 VEGF-A 96 h 1479.3 ± 25.5 1604.2 ± 21.2 1611.2 ± 19.2

151

152

Chapter 7 Summary 1 Summary of results

The aim of this thesis was to identify molecules of interest in the host response to MAA beads, and generate a hypothesis regarding the mechanism through which MAA beads induce a therapeutic effect in vivo. Using gene and protein expression analysis, eight molecules of interest involved in wound healing and angiogenesis were found to be changed in response to exposure to MAA beads in a model of diabetic wound healing and/or in macrophage-like dTHP-1 cells (summarized in Table 7-1). Briefly, MAA beads increased the global gene expression of sonic hedgehog (Shh), interleukin (IL)-1β, tumor necrosis factor α (TNF-α), osteopontin (OPN) and sprouty homolog 2 (Spry2) and decreased expression of C-X-C motif chemokine 10 (CXCL10) in the granulation tissue of diabetic wounds, compared to no beads and poly(methyl methacrylate) beads. A similar profile was observed in the gene expression of macrophage-like dTHP-1 cells treated with MAA beads in vitro, where IL-1β, IL-6 and TNF-α were increased at the gene and protein level and CXCL10 was decreased at the protein level.

In contrast to the global gene expression in vivo, dTHP-1 decreased the expression of OPN mRNA compared to PMMA beads and no beads, but expressed the same amount of secreted OPN protein. These results suggest that the cells had decreased the expression of the intracellular isoform of the protein, iOPN, but had no effect on the secreted form, sOPN in dTHP-1. Although no difference was observed at the gene expression level for CXCL12 in vivo (data not shown in thesis) or in vitro, protein expression analysis revealed that MAA beads decreased the concentration of CXCL12 in the supernatant of dTHP-1 cells. In both in vivo and in vitro models, minimal to no changes were observed in the expression of VEGF, PDGF-B or bFGF in response to MAA bead treatment.

As macrophages play a major role in directing wound healing and foreign body reactions (FBR), they were presumed to be a principal responder to MAA beads. Phorbol myristate acetate (PMA) differentiated THP-1 cells (dTHP-1) were used to model the macrophage response to MAA beads in vitro. The change in the gene expression of cytokines and chemokines observed in MAA-treated wounds was recapitulated by the gene and protein expression of macrophage-like

153 cells treated with MAA beads in vitro, supporting the premise that macrophages are a principal responder to MAA beads. However, the difference in OPN expression highlights the fact that the in vivo gene expression represents the global expression of multiple cell types, and presents an intriguing question regarding the role of intracellular OPN in the macrophage response to MAA beads.

Table 7-1. List of genes/proteins identified as potential mediators of the biological response to MAA beads. These proteins had changes in gene or protein expression in response to the presence of MAA beads in two model systems. In vivo Diabetic Wound Tissue (global) In vitro dTHP-1 (monoculture) Day 4 Day 7 24 h 96 h Shh é mRNA NC N/A N/A

OPN NC mRNA é mRNA ê mRNA ê mRNA NC sOPN NC sOPN CXCL10 ê mRNA (NS) ê mRNA (NS) ê protein ê protein

IL-1β NC mRNA é mRNA NC mRNA é mRNA NC protein é protein IL-6 NC mRNA é mRNA (NS) é mRNA é/NC mRNA NC protein é protein TNF-α NC mRNA é mRNA é mRNA é mRNA NC protein é protein Spry2 é mRNA é mRNA N/A N/A

CXCL12 NC mRNA NC mRNA NC mRNA NC mRNA (unpublished) (unpublished) NC protein ê protein é, increased expression; ê, decreased expression; NC, no change in expression; N/A, no data available; NS, change not significant (p > 0.05)

The expression of Shh in the wound bed also raises questions regarding its source, as the source of Shh in adult wound healing (or other pathological conditions in adults) has not been extensively studied. Recently, peripheral blood mononuclear cells were reported to express Shh in response to NF-κB activation by LPS in vitro[1]. The same study also reported Shh expression in LPS-treated THP-1 cells (using 0.1 – 1 ug/ml LPS). Human umbilical vein endothelial cells (HUVEC), bovine retinal microvascular endothelial cells (BREC) and bovine retinal pericytes (BRP) were also reported to express Shh mRNA in vitro[2-4]. Other possible sources of Shh include mesenchyme-derived fibroblasts[5], Schwann cells from injured peripheral nerves[6] and hair follicle associated sensory neurons along the wound edge[7]. Apoptotic/activated T-cells (treated with phytohemagglutinin, PMA and actinomycin D in vitro) have also been reported to

154 release Shh-containing microparticles[8]. Experiments to identify what cells produce Shh in MAA-treated cutaneous wounds are being conducted using transgenic mouse models, however no conclusive results are currently available.

The results generated in this thesis have identified a small group of molecules that represent individual pieces of the puzzle in the story of how MAA beads mediate host-response. However, it is not clear how these pieces fit together to form a coherent story. In an effort to generate a hypothesis of how these molecules relate to each other, the eight genes of interest were annotated using DAVID (Database for annotation, visualization and integrated discovery) bioinformatics resource (http://david.abcc.ncifcrf.gov/home.jsp, version 6.7)[9,10]. Functional annotation pathway analysis (Kyoto Encyclopedia of Genes and Genomes, KEGG-pathway[11,12]) was used to identify enriched pathways (using count threshold of 2 and EASE threshold of 0.1). The most enriched pathway was the Toll-like receptor (TLR) signaling pathway (p = 4.9 x 10-6, Bonferroni = 1.7 x 10-4), which was associated with 5 of the 8 genes of interest (pathway proteins: OPN; pathway target genes: IL-6, TNF-α, IL-1β, CXCL10)(Figure 7-1). Further literature searches revealed that OPN and Shh were also possible targets of TLR signaling[13- 15], suggesting that this pathway may play an important role in modulating the host-response to MAA beads.

2 Toll-like receptors

TLR are a family of pattern-recognition receptors that play a critical role in pathogen recognition and innate immune responses. Ten TLR (TLR1 – TLR10) have been identified in humans, with additional receptors in other mammalian species (including TLR11 – TLR13 in mice). TLR can be divided into two categories, based on cellular location. TLR1, TLR2, TLR4, TLR5 and TLR6 are expressed and bind ligands on the cell surface, while TLR3, TLR7, TLR8 and TLR9 are located within the endosome, and bind nucleic acid-based ligands (reviewed by Lee et al. 2012[16]). While TLR bind a diverse range of microbial molecular structures that were initially thought to enable cells to distinguish between self and non-self (Table 7-2), it has become increasingly evident that TLR also bind numerous endogenous ligands that are generated during tissue injury and cell death[17,18].

155

With the exception of TLR3, ligation of all TLR activates myeloid differentiation factor 88 (MyD88), which initiates signal transduction pathways leading to the activation of NF-κB and AP-1, and expression of pro-inflammatory target genes (IL-6, IL-1β, TNF-α, IL-8)[16]. A MyD88-independent pathway has been identified for TLR3, TLR4 and TLR9, which signals via TRIF (Toll/IL-1 receptor domain-containing adaptor inducing interferon-B) to eventually activate interferon response factors (such as IFR3) and induce expression of Type I interferons, interferon-inducible genes (CXCL10) and OPN[14,19,20]. The TRIF-dependent (MyD88- independent) pathway also leads to late-phase NF-κB activation, although it is a delayed response compared to the MyD88-dependent pathway[20]. Although cell-surface TLR bind their ligands at the cell surface, some are internalized following ligand recognition and signal from within the cell[21]. For example, TLR4 binds its ligand at the cell surface and initiates MyD88 signaling pathway, then is internalized into the early endosome where MyD88 signaling ceases and TRIF-dependent (MyD88 independent) signaling begins[22,23].

All TLR have specific microbial-derived ligands, which drive acute innate immune responses. In addition to exogenous microbial-derived pathogen-associated molecular patterns (PAMPs), TLR also respond to a number of host-derived non-microbial molecules, collectively referred to as

156 damage-associated molecular patterns (DAMPs) that are released by damaged or necrotic tissues, or in specific pathological conditions (summarized in Table 7-2[24]). Cutaneous wounds and tissue injury that occurs during implantation of biomaterials create environments that would produce numerous DAMPs capable of activating TLR signaling (e.g. heat shock proteins, high mobility group box 1, S100 proteins, tenascin-C) [25-30]. The protein layer acquired by implanted materials would also contain TLR4 ligands fibronectin and fibrinogen (among others)[31,32].

TLR2 Peptidoglycan, lipoteichoic acid (Gram- Biglycan, carboxyalkylpyrrole, endoplasmin, positive bacteria); Lipoarabinomannan high mobility group box-1 protein (Mycobacteria); lipopeptide (HMGB1), HSP60, HSP70, human cardiac (Mycoplasma); Zymosan, myosin, hyaluronan, monosodium urate Glucuronoxylomannan, crystals, pancreatic adenocarcinoma Phospholipomannan (fungi); Porins upregulated factor (PAUF), and versican (Neisseria) and LPS (Leptospira)

TLR3 Double strand RNA [dsRNA] (virus) mRNA

TLR4 LPS (Gram-negative bacteria); Mannan Biglycan, CD138, a-crystallin A chain, b- (Candida albicans); Envelope proteins defensin 2, endoplasmin, fibrinogen, (RSV, MMTV) and fibronectin, heparin sulphate, HMGB1, Glycoinositolphospholipids HSP22, HSP60, HSP70, HSP72, hyaluronan, (Trypanosoma) monosodium urate crystals, oxidized phospholipids (OxPAPC), PAUF, peroxiredoxin 1 (Prx1), resistin, S100 proteins, serum amyloid A3 (SAA3), surfactant protein A, and tenascin-C

TLR5 Flagellin (Bacteria)

TLR6 Diacyl lipopeptides (Mycoplasma); Versican peptidoglycan (Gram-positive bacteria) and Zymosan (fungi)

TLR7 Single strand RNA (RNA virus) RNA and small interfering RNA (siRNA)

TLR8 Single stranded RNA (RNA virus) Human cardiac myosin and siRNA

TLR9 Unmethylated CpG DNA (Microbes) DNA and HMGB1

157

TLR signaling in response to DAMPs are capable of activating three transcription factors that are reported to induce expression of six molecules identified in the MAA-host and cell response. NF- κB signaling induces the expression of pro-inflammatory cytokines IL-1β, IL-6 and TNF-α, and has been reported to induce Shh expression, while AP-1 regulates the expression of OPN[13- 15,24]. The TRIF-dependent pathway leads to expression CXCL10 and other interferon- inducible genes[19,20]. Consequently, it is plausible that MAA-mediated angiogenesis is modulated by TLR signaling in response to the composition or conformation of DAMPs (i.e. TLR 4 ligands fibronectin and fibrinogen) in the adsorbed protein layer on MAA beads, which is presumed to be distinct from PMMA beads and the untreated wound environment.

There is growing evidence that TLR play a central role in acute macrophage and dendritic cell (DC) responses to biomaterials. Deletion of TLR4 or MyD88 strongly impaired the expression of IL-1β, IL-6, IL-12 and TNF-α in biomaterial activated DC[33], indicating that these cytokines were induced through the MyD88-dependent TLR4 signaling cascade [33]. There is also evidence that TLR2 and TLR6 modulate DC response to certain biomaterials, however the effect does not appear to be as robust as that of TLR4 [33]. Pearl et al. found that PMMA wear particle- induced production of TNF-α was inhibited in macrophages from MyD88 knockout mice but amplified in macrophages from TRIF-knockout mice[34]. Furthermore, the increase in TNF-α in response to loss of TRIF signaling was due to a compensatory increase in the TLR MyD88- dependent signaling[34]. In vivo implant models using PET and silicone discs demonstrated that TLR4 mediates the FBR to implanted materials[35,36]. Interestingly, TLR4-/- mice had significantly reduced blood vessel density in the tissue surrounding the implant, but also had decreased fibrous capsule thickness[35], demonstrating that biomaterial-induced TLR4 signaling can promote angiogenesis in vivo.

There is also evidence of crosstalk between TLR signaling and the complement system, which is also a key mediator of innate immune responses and inflammation (reviewed in Chapter 3), however the reported effects are often contradictory[37]. In monocytes and macrophages, complement anaphylatoxin C5a inhibited LPS-induced production of IL-12 family cytokines[38- 42]. Zaal et al. also reported C5a inhibition of LPS-induced IL-6 and TNF-α expression[37]. Other studies reported a synergistic effect of C5a on TLR-induced expression of IL-1β, IL-6, macrophage inflammatory protein 2 (MIP-2) and monocyte chemoattractant protein 1 (MCP-1)

158 in neutrophils, endothelial cells and mononuclear cells[43-46]. The effect of C5a on TLR signaling appears to occur through a complex regulatory mechanism and is highly context dependent; with C5a inhibiting TLR-induced expression IL-12 family cytokines, but synergistically increasing other pro-inflammatory cytokines. However, as the majority of this research has focused on LPS-induced TLR signaling, it is unclear what effect complement and TLR crosstalk would have in the context of biomaterial/DAMP-induced inflammation.

3 Hypothesis for host response to MAA beads and MAA-mediated angiogenesis

An hypothesis describing the cellular response to MAA beads was formulated based on the molecules identified in this thesis and the current body of literature regarding wound healing, angiogenesis, FBR, inflammation and macrophage biology.

3.1 Hypothesis

First, it is hypothesized that macrophages play a key role in MAA-mediated therapeutic effects observed in vivo, and that MAA-activated macrophages direct an alternative FBR that promotes angiogenesis instead of fibrosis (despite increased expression of pro-inflammatory cytokines). This hypothesis is supported by the consistent profiles observed in the global gene expression in MAA-treated diabetic wound tissue in vivo and the response of MAA-treated macrophage-like cells in vitro (with the exception of OPN). Furthermore, the responses to MAA beads were distinctly different than the expression induced by PMMA, which is known to induce a conventional FBR[47]. Of particular interest among the differences was the strong inhibition MAA beads had on CXCL10 expression in dTHP-1 cells, compared to PMMA beads (this trend was also observed in vivo, but was not significant). This decrease in CXCL10 promotes angiogenesis through the reduction of anti-angiogenic signaling, thus contributing to the alternative FBR associated with MAA beads. The contradiction in the expression of OPN is likely due to the fact that the gene expression in vivo reflects overall (or net) expression of multiple cell types, and our in vitro data shows that the effect of MAA beads on OPN expression is cell type dependent.

Second, the macrophage response to MAA beads is modulated by TLR signaling, in response to the composition and conformation of the protein layer adsorbed to the MAA surface (Figure 7-

159

2). Possible TLR ligands include fibronectin and fibrinogen, but likely involve other molecules as well. The protein layer on MAA beads modulates the TLR signaling pathway in macrophages in a manner distinct from PMMA beads (or DAMPs present in untreated wounds), leading to the increased expression of Shh, IL-1β, IL-6 and TNF-α (NF-κB pathway) and decreased expression of CXCL10 and intracellular OPN (transcriptional target of AP-1 and IRF pathways).

Thirdly, the increase in pro-angiogenic Shh and decrease in anti-angiogenic CXCL10 promote angiogenesis, in addition to the downstream pro-angiogenic effects that IL-1β, IL-6 TNF-α and other TLR signaling target molecules can also elicit[48]. The increased Shh expression also stimulates increased expression of (secreted) OPN in endothelial cells and fibroblasts present in the wound bed, further promoting wound healing and angiogenesis[49].

Sprouty proteins are intracellular proteins that act as negative feedback loop modulators of receptor tyrosine kinase signaling (Raf/Mek/Erk) associated with growth factor activity, and inhibit angiogenesis through inhibiting Raf/Mek/Erk and MAPK signaling[50,51]. Sprouty proteins are also ubiquitously expressed in endothelial cells and vascular smooth muscle cells, and are upregulated during angiogenesis[50-52]. Consequently, the increased expression of Spry2 may reflect increased growth factor-receptor binding, or increased number of endothelial cells (and VSMC) present in MAA-treated wounds.

3.2 Possible role of intracellular OPN

Decreased expression of iOPN in macrophages (if shown to be occurring) would have interesting implications for TLR signaling. iOPN acts as a TLR adaptor molecule that associates with MyD88 upon TLR ligation, and is essential for TLR9 activation of IRF-7[53]. However, it is not required for TLR-induced NF-κB signaling[53]. iOPN is also involved in TLR2/6 signaling, through association with MyD88, and enhances MAPK signaling. Consequently, it is possible that decreased iOPN in MAA-treated macrophages attenuates TLR signaling via the IRF and MAPK/AP-1 pathways, but not the NF-κB pathway.

160

3.3 Limitations

It is evident that there are multiple limitations in the hypothesis. First, it is unlikely that modulation of TLR signaling is the only pathway contributing to MAA-mediated angiogenesis and wound healing. However, the importance of TLR in directing innate immune responses supports the premise that it may be a major contributor to the MAA response. Other important pathways in the FBR, sterile (i.e. non-microbial) inflammation and the expression of pro-

161 inflammatory cytokines include the complement system, IL-1/IL-1R and other cytokine signaling cascades, the receptor for advanced glycation end-products (RAGE), interferon signaling and the NRLP3 inflammasome[47,54,55]. Consequently, it is unlikely that TLR signaling is solely responsible for modulating the expression of IL-1β, IL-6, TNF-α and other pro-inflammatory cytokines. Therefore, future studies should not be constrained to the role of TLR signaling in the MAA response, but should continue to explore multiple pathways.

It is also unknown if or how MAA-TLR signaling would be able to increase the activation of one signaling transduction pathway (i.e. NF-κB) while decreasing the activation of other pathways (i.e. IRF, MAPK/AP-1). Furthermore, many of the molecules of interest identified in this thesis can be induced or inhibited by multiple signals, in addition to TLR ligands; these additional possibilities are not accounted for in this model.

The use of bioinformatics resources, such as DAVID and KEGG, to gain insight into pathways that may play a role in the alternative FBR to MAA beads can be useful, however there are inherent limitations that must be considered. These resources are particularly useful in distilling large data sets generated by high-throughput methodologies (e.g. gene expression microarrays) that are (ideally) free of the investigators' assumptions or biases. However, the gene list generated by this thesis is biased, as it predominately contains genes that were selected for analysis based on their involvement in inflammation, wound healing and angiogenesis. Also small gene lists have limited statistical power and consequently, only highly enriched or very broad terms/pathways are highlighted[56]. Other limitations of bioinformatics resources are reviewed in detail by Khatri et al[56]. Briefly, the majority of annotation databases, including KEGG, are maintained and updated manually by curators, and certain facts (gene function, pathway targets etc) can be overlooked, often due to backlog. Also, as annotation databases are based on what is known, this type of analysis cannot be used to discover new functions or relationships. Furthermore, functional annotation relies heavily on statistical analysis, which can overlook weak but biologically interesting results, especially with small gene lists. Finally, the analysis does not take into account any information about the system being analyzed and often generate results that are not contextually relevant or helpful. However, when used with caution, bioinformatics can be useful in identifying pathways of interest.

162

This hypothesis also presumes that macrophages are the source of Shh, despite dTHP-1 cells not expressing detectable levels of Shh mRNA in vitro (unpublished data). Currently, the source of Shh in adult tissue has not been studied extensively; however there is evidence that macrophages are capable of expressing this protein. MAA-induced expression of Shh in macrophages should therefore be revisited, using primary macrophages and protein expression, in conjunction with current efforts to identify Shh-producing cells in wound healing models. If macrophages are not the source of Shh in this model, other cell types will need to be considered.

This hypothesis also does not account for the temporal order in which the molecules appear to be increased or decreased by MAA beads in vivo, with the exception of Shh-induced OPN expression. For example, Shh was increased at day 4, but IL-6, IL-1β and TNF-α were not increased until day 7. However, all four proteins are suggested to be induced by the same signaling pathway.

The decreased expression of CXCL12 was not included in the hypothesis, as there is no evidence currently that CXCL12 is decreased in vivo. However, there is evidence that TLR signaling can induce the expression of CXCL12 and that CXCL12 is a transcription target of AP-1[57,58]. Furthermore, increased expression of IL-1β and TNF-α has been found to reduce CXCL12 expression in dermal fibroblasts[59].

4 Future directions

The scope of this thesis was to identify molecules of interest in the host response to MAA beads. Eight proteins were identified as having changed expression in the presence of MAA beads, and a hypothesis has been generated regarding the mechanism of MAA-mediated host response. The increased expression of IL-1β, IL-6 and TNF-α, although useful in characterizing the MAA- response, are not compelling molecules to study as actual drivers of the response as they likely represent the inflammatory response. Instead, it is recommended that the future direction of research validate the role of Shh, CXCL10, OPN and CXCL12 in mediating the host response to MAA beads.

Currently studies are underway to identify the cell types that express Shh in MAA-treated wounds using histological analysis in non-diabetic transgenic mouse models. This work will hopefully give insight into the source of MAA-induced Shh and will also verify if the pro-

163 angiogenic effect of MAA beads occurs in models of normal (non-diabetic) wound healing. In parallel, the role of Shh in MAA-mediated angiogenesis should be tested by inhibiting the Shh pathway (e.g. with cyclopamine) in the db/db diabetic mouse wound healing model used in this thesis. It is also of interest to revisit Shh expression in dTHP-1 or primary macrophage response to MAA beads. The role of CXCL10 can be studied in a similar manner, using siRNA to knockdown CXCL10 expression or using blocking antibodies for CXCL10[60] or CXCL10- receptor CXCR3[61] in vitro and in the db/db mouse model. CXCL10 knockout mice may also prove to be a useful tool in this matter[62].

As discussed in detail in Chapter 6, the effect of MAA beads on iOPN and the effect of MAA- reduced CXCL12 on HUVEC migration remain to be addressed. Briefly, the effect of MAA beads on iOPN could be studied using western blotting techniques in dTHP-1 cells or primary mouse macrophages. Furthermore, the in vivo role of iOPN in MAA-mediated angiogenesis and wound healing could be studied using Opn-/- mice[63,64] and delivery of exogenous full-length OPN (i.e. sOPN), if desired. The role of CXCL12 in inhibiting HUVEC migration in MAA/dTHP-1 conditioned medium could be tested by blocking CXCR4 in HUVEC during the transwell migration assay, described in Chapter 5. Although no differences were observed in the mRNA expression of CXCL12 in vivo, it would be interesting to study the effect of MAA beads on the concentration of CXCL12 in circulation following injury, and the recruitment and homing of bone-marrow progenitor cells.

The hypothesis that TLR mediate the host (and macrophage) response to MAA beads should be tested in vitro by knocking down TLR using siRNA[65], blocking TLR signaling with by TLR neutralizing antibodies or small molecules (e.g. TAK-242), or using macrophages isolated from TLR knockout mice strains (specifically TLR4-/- or TLR2-/- mice)[66]. In vivo, this hypothesis can be tested using TLR-null or MyD88-null mice strains and studying the effect on gene expression and ultimately, wound bed vascularization. Studying macrophage response to MAA- beads pre-adsorbed with known DAMPs, compared to PMMA beads, may also provide insight into the role of TLR in macrophage-response to MAA beads.

Furthermore, our limited knowledge regarding what cells are responding to MAA beads calls for the examination of MAA-response in other cells types, such as dendritic cells, fibroblasts, neutrophils and endothelial cells. As it is unlikely that only one cell population is responding to

164

MAA beads, co-culture systems will be developed to look at the interaction of cell types in response to MAA beads. Also, the use of gene expression microarrays should be revisited now that models and techniques for studying the host/cell responses to MAA beads have been established. This approach may generate further (non-biased) insight into MAA-host interactions.

Finally, the biomaterial question must be addressed to determine what component(s) of MAA beads is inducing the angiogenesis response. We have proposed that methacrylic acid content may be an essential factor in the therapeutic response, or the ability of the beads to bind or sequester important molecules in the wound bed[67-69]. Others would point to the topography[70]. These questions should be addressed through pertubing the material properties (i.e. methacrylic acid content, form, surface topography) using the expression profile of proteins identified in this thesis as benchmarks, and studying the adsorbed protein layer using proteomics. Identifying what is making the beads angiogenic would enable our lab to optimize the MAA bead properties and direct the development of other therapeutic polymers (TheramersTM).

5 Clinical relevance

The motivation for this thesis was to begin to understand how a synthetic, “inert” biomaterial was able to promote angiogenesis and wound healing in vivo. As MAA beads appear to elicit an alternative FBR (i.e. angiogenesis instead of fibrosis), understanding the mechanism through which this process occurs is expected to be useful in designing and testing new generations of biomaterials capable of modulating the FBR. Indeed, one could envision using molecules deemed essential in MAA-mediated angiogenesis as a molecular profile for identifying candidate Theramers in high throughput screening methods. Identifying molecular components of the MAA host-response also provides benchmarks that can be used when optimizing characteristics of MAA beads (i.e. methacrylic acid content, bead size, surface roughness) for inducing angiogenesis.

6 Summary

MAA beads are hypothesized to promote angiogenesis and wound healing through the modulation of TLR signaling in macrophages, thereby increasing expression of NF-κB target genes (Shh, IL-1β, IL-6, TNF-α) and decreased AP-1 and IRF target genes (OPN, CXCL10).

165

The composition and conformation of the adsorbed proteins, including TLR DAMP ligands, to MAA beads is thought be distinct from what is present on PMMA beads and in the untreated wound bed. The downstream affect of MAA/TLR signaling would increase angiogenesis primarily in response to the increase in pro-angiogenic Shh, and the decreased in anti-angiogenic CXCL10.

7 References [1] Yamasaki A, Kameda C, Xu R, Tanaka H, Tasaka T, Chikazawa N, et al. Nuclear factor kappaB-activated monocytes contribute to pancreatic cancer progression through the production of Shh. Cancer Immunol Immunother 2010;59(5):675-86. [2] Sekiguchi H, Ii M, Jujo K, Renault MA, Thorne T, Clarke T, et al. Estradiol triggers sonic-hedgehog-induced angiogenesis during peripheral nerve regeneration by downregulating hedgehog-interacting protein. Lab Invest 2012;92(4):532-42. [3] Bhardwaj G, Murdoch B, Wu D, Baker DP, Williams KP, Chadwick K, et al. Sonic hedgehog induces the proliferation of primitive human hematopoietic cells via BMP regulation. Nat Immunol 2001;2(2):172-80. [4] Walshe TE, Connell P, Cryan L, Ferguson G, Gardiner T, Morrow D, et al. Microvascular retinal endothelial and pericyte cell apoptosis in vitro: role of hedgehog and Notch signaling. Invest Ophthalmol Vis Sci 2011;52(7):4472-83. [5] Pola R, Ling LE, Aprahamian TR, Barban E, Bosch-Marce M, Curry C, et al. Postnatal recapitulation of embryonic hedgehog pathway in response to skeletal muscle ischemia. Circulation 2003;108(4):479-85. [6] Hashimoto M, Ishii K, Nakamura Y, Watabe K, Kohsaka S, Akazawa C. Neuroprotective effect of sonic hedgehog up-regulated in Schwann cells following sciatic nerve injury. J Neurochem 2008;107(4):918-27. [7] Brownell I, Guevara E, Bai CB, Loomis CA, Joyner AL. Nerve-derived sonic hedgehog defines a niche for hair follicle stem cells capable of becoming epidermal stem cells. Cell Stem Cell 2011;8(5):552-65. [8] Martinez MC, Larbret F, Zobairi F, Coulombe J, Debili N, Vainchenker W, et al. Transfer of differentiation signal by membrane microvesicles harboring hedgehog morphogens. Blood 2006;108(9):3012-20. [9] Huang da W, Sherman BT, Lempicki RA. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic acids research 2009;37(1):1-13. [10] Huang da W, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 2009;4(1):44-57. [11] Kanehisa M, Goto S, Sato Y, Furumichi M, Tanabe M. KEGG for integration and interpretation of large-scale molecular data sets. Nucleic Acids Res 2012;40(Database issue):D109-14.

166

[12] Kanehisa M, Goto S. KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res 2000;28(1):27-30. [13] Yang IV, Alper S, Lackford B, Rutledge H, Warg LA, Burch LH, et al. Novel regulators of the systemic response to lipopolysaccharide. Am J Respir Cell Mol Biol 2011;45(2):393-402. [14] Zhao W, Wang L, Zhang L, Yuan C, Kuo PC, Gao C. Differential expression of intracellular and secreted osteopontin isoforms by murine macrophages in response to toll-like receptor agonists. J Biol Chem 2010;285(27):20452-61. [15] Zhao W, Wang L, Zhang M, Wang P, Zhang L, Yuan C, et al. NF-kappaB- and AP-1- mediated DNA looping regulates osteopontin transcription in endotoxin-stimulated murine macrophages. J Immunol 2011;186(5):3173-9. [16] Lee CC, Avalos AM, Ploegh HL. Accessory molecules for Toll-like receptors and their function. Nat Rev Immunol 2012;12(3):168-79. [17] Marshak-Rothstein A. Toll-like receptors in systemic autoimmune disease. Nat Rev Immunol 2006;6(11):823-35. [18] Medzhitov R, Janeway C, Jr. Innate immune recognition: mechanisms and pathways. Immunol Rev 2000;173:89-97. [19] Proost P, Vynckier AK, Mahieu F, Put W, Grillet B, Struyf S, et al. Microbial Toll-like receptor ligands differentially regulate CXCL10/IP-10 expression in fibroblasts and mononuclear leukocytes in synergy with IFN-gamma and provide a mechanism for enhanced synovial chemokine levels in septic arthritis. Eur J Immunol 2003;33(11):3146-53. [20] Selvarajoo K, Takada Y, Gohda J, Helmy M, Akira S, Tomita M, et al. Signaling flux redistribution at toll-like receptor pathway junctions. PLoS One 2008;3(10):e3430. [21] Blasius AL, Beutler B. Intracellular toll-like receptors. Immunity 2010;32(3):305-15. [22] Kagan JC, Su T, Horng T, Chow A, Akira S, Medzhitov R. TRAM couples endocytosis of Toll-like receptor 4 to the induction of interferon-beta. Nat Immunol 2008;9(4):361-8. [23] Tanimura N, Saitoh S, Matsumoto F, Akashi-Takamura S, Miyake K. Roles for LPS- dependent interaction and relocation of TLR4 and TRAM in TRIF-signaling. Biochem Biophys Res Commun 2008;368(1):94-9. [24] Yu L, Wang L, Chen S. Exogenous or endogenous Toll-like receptor ligands: which is the MVP in tumorigenesis? Cell Mol Life Sci 2012;69(6):935-49. [25] Curtin JF, Liu N, Candolfi M, Xiong W, Assi H, Yagiz K, et al. HMGB1 mediates endogenous TLR2 activation and brain tumor regression. PLoS Med 2009;6(1):e10. [26] Apetoh L, Ghiringhelli F, Tesniere A, Obeid M, Ortiz C, Criollo A, et al. Toll-like receptor 4-dependent contribution of the immune system to anticancer chemotherapy and radiotherapy. Nat Med 2007;13(9):1050-9. [27] Asea A, Rehli M, Kabingu E, Boch JA, Bare O, Auron PE, et al. Novel signal transduction pathway utilized by extracellular HSP70: role of toll-like receptor (TLR) 2 and TLR4. J Biol Chem 2002;277(17):15028-34.

167

[28] Vabulas RM, Ahmad-Nejad P, Ghose S, Kirschning CJ, Issels RD, Wagner H. HSP70 as endogenous stimulus of the Toll/interleukin-1 receptor signal pathway. J Biol Chem 2002;277(17):15107-12. [29] Vogl T, Tenbrock K, Ludwig S, Leukert N, Ehrhardt C, van Zoelen MA, et al. Mrp8 and Mrp14 are endogenous activators of Toll-like receptor 4, promoting lethal, endotoxin- induced shock. Nat Med 2007;13(9):1042-9. [30] Midwood K, Sacre S, Piccinini AM, Inglis J, Trebaul A, Chan E, et al. Tenascin-C is an endogenous activator of Toll-like receptor 4 that is essential for maintaining inflammation in arthritic joint disease. Nat Med 2009;15(7):774-80. [31] Smiley ST, King JA, Hancock WW. Fibrinogen stimulates macrophage chemokine secretion through toll-like receptor 4. J Immunol 2001;167(5):2887-94. [32] Okamura Y, Watari M, Jerud ES, Young DW, Ishizaka ST, Rose J, et al. The extra domain A of fibronectin activates Toll-like receptor 4. J Biol Chem 2001;276(13):10229- 33. [33] Shokouhi B, Coban C, Hasirci V, Aydin E, Dhanasingh A, Shi N, et al. The role of multiple toll-like receptor signalling cascades on interactions between biomedical polymers and dendritic cells. Biomaterials 2010;31(22):5759-71. [34] Pearl JI, Ma T, Irani AR, Huang Z, Robinson WH, Smith RL, et al. Role of the Toll-like receptor pathway in the recognition of orthopedic implant wear-debris particles. Biomaterials 2011;32(24):5535-42. [35] Auquit-Auckbur I, Caillot F, Arnoult C, Menard JF, Drouot L, Courville P, et al. Role of toll-like receptor 4 in the inflammation reaction surrounding silicone prosthesis. Acta biomaterialia 2011;7(5):2047-52. [36] Rogers TH, Babensee JE. Altered adherent leukocyte profile on biomaterials in Toll-like receptor 4 deficient mice. Biomaterials 2010;31(4):594-601. [37] Zaal A, Lissenberg-Thunnissen SN, van Schijndel G, Wouters D, Ham SM, Ten Brinke A. Crosstalk between Toll like receptors and C5a receptor in human monocyte derived DCs suppress inflammatory cytokine production. Immunobiology 2012. [38] Braun MC, Lahey E, Kelsall BL. Selective suppression of IL-12 production by chemoattractants. J Immunol 2000;164(6):3009-17. [39] Hawlisch H, Belkaid Y, Baelder R, Hildeman D, Gerard C, Kohl J. C5a negatively regulates toll-like receptor 4-induced immune responses. Immunity 2005;22(4):415-26. [40] la Sala A, Gadina M, Kelsall BL. G(i)-protein-dependent inhibition of IL-12 production is mediated by activation of the phosphatidylinositol 3-kinase-protein 3 kinase B/Akt pathway and JNK. J Immunol 2005;175(5):2994-9. [41] Okazaki N, Hazeki K, Izumi T, Nigorikawa K, Hazeki O. C5a controls TLR-induced IL- 10 and IL-12 production independent of phosphoinositide 3-kinase. J Biochem 2011;149(3):265-74. [42] Wittmann M, Zwirner J, Larsson VA, Kirchhoff K, Begemann G, Kapp A, et al. C5a suppresses the production of IL-12 by IFN-gamma-primed and lipopolysaccharide- challenged human monocytes. J Immunol 1999;162(11):6763-9.

168

[43] Okusawa S, Dinarello CA, Yancey KB, Endres S, Lawley TJ, Frank MM, et al. C5a induction of human interleukin 1. Synergistic effect with endotoxin or interferon-gamma. J Immunol 1987;139(8):2635-40. [44] Laudes IJ, Chu JC, Huber-Lang M, Guo RF, Riedemann NC, Sarma JV, et al. Expression and function of C5a receptor in mouse microvascular endothelial cells. J Immunol 2002;169(10):5962-70. [45] Riedemann NC, Guo RF, Hollmann TJ, Gao H, Neff TA, Reuben JS, et al. Regulatory role of C5a in LPS-induced IL-6 production by neutrophils during sepsis. Faseb J 2004;18(2):370-2. [46] Fang C, Zhang X, Miwa T, Song WC. Complement promotes the development of inflammatory T-helper 17 cells through synergistic interaction with Toll-like receptor signaling and interleukin-6 production. Blood 2009;114(5):1005-15. [47] Malik AF, Hoque R, Ouyang X, Ghani A, Hong E, Khan K, et al. Inflammasome components Asc and caspase-1 mediate biomaterial-induced inflammation and foreign body response. Proc Natl Acad Sci U S A 2011;108(50):20095-100. [48] Grote K, Schutt H, Schieffer B. Toll-like receptors in angiogenesis. ScientificWorldJournal 2011;11:981-91. [49] Renault MA, Roncalli J, Tongers J, Thorne T, Klyachko E, Misener S, et al. Sonic hedgehog induces angiogenesis via Rho kinase-dependent signaling in endothelial cells. J Mol Cell Cardiol 2010;49(3):490-8. [50] Wietecha MS, Chen L, Ranzer MJ, Anderson K, Ying C, Patel TB, et al. Sprouty2 downregulates angiogenesis during mouse skin wound healing. Am J Physiol Heart Circ Physiol 2011;300(2):H459-H67. [51] Ozaki K, Kadomoto R, Asato K, Tanimura S, Itoh N, Kohno M. ERK pathway positively regulates the expression of Sprouty genes. Biochem Biophys Res Commun 2001;285(5):1084-8. [52] Antoine M, Wirz W, Tag CG, Mavituna M, Emans N, Korff T, et al. Expression pattern of fibroblast growth factors (FGFs), their receptors and antagonists in primary endothelial cells and vascular smooth muscle cells. Growth Factors 2005;23(2):87-95. [53] Shinohara ML, Lu L, Bu J, Werneck MB, Kobayashi KS, Glimcher LH, et al. Osteopontin expression is essential for interferon-alpha production by plasmacytoid dendritic cells. Nat Immunol 2006;7(5):498-506. [54] Chen GY, Nuñez G. Sterile inflammation: sensing and reacting to damage. Nat Rev Immunol 2010;10(12):826-37. [55] O'Shea JJ, Murray PJ. Cytokine signaling modules in inflammatory responses. Immunity 2008;28(4):477-87. [56] Khatri P, Draghici S. Ontological analysis of gene expression data: current tools, limitations, and open problems. Bioinformatics 2005;21(18):3587-95. [57] Sadik CD, Hunfeld KP, Bachmann M, Kraiczy P, Eberhardt W, Brade V, et al. Systematic analysis highlights the key role of TLR2/NF-kappaB/MAP kinase signaling for IL-8 induction by macrophage-like THP-1 cells under influence of Borrelia burgdorferi lysates. Int J Biochem Cell Biol 2008;40(11):2508-21.

169

[58] Florin L, Hummerich L, Dittrich BT, Kokocinski F, Wrobel G, Gack S, et al. Identification of novel AP-1 target genes in fibroblasts regulated during cutaneous wound healing. Oncogene 2004;23(42):7005-17. [59] Fedyk ER, Jones D, Critchley HO, Phipps RP, Blieden TM, Springer TA. Expression of stromal-derived factor-1 is decreased by IL-1 and TNF and in dermal wound healing. J Immunol 2001;166(9):5749-54. [60] Cuenca AG, Wynn JL, Kelly-Scumpia KM, Scumpia PO, Vila L, Delano MJ, et al. Critical role for CXC ligand 10/CXC receptor 3 signaling in the murine neonatal r esponse to sepsis. Infect Immun 2011;79(7):2746-54. [61] He S, Cao Q, Qiu Y, Mi J, Zhang JZ, Jin M, et al. A new approach to the blocking of alloreactive T cell-mediated graft-versus-host disease by in vivo administration of anti- CXCR3 neutralizing antibody. J Immunol 2008;181(11):7581-92. [62] Zhai Y, Shen XD, Gao F, Zhao A, Freitas MC, Lassman C, et al. CXCL10 regulates liver innate immune response against ischemia and reperfusion injury. Hepatology 2008;47(1):207-14. [63] Inoue M, Moriwaki Y, Arikawa T, Chen YH, Oh YJ, Oliver T, et al. Cutting edge: critical role of intracellular osteopontin in antifungal innate immune responses. J Immunol 2011;186(1):19-23. [64] Sato I, Yamamoto N, Rittling SR, Denhardt DT, Hino M, Morimoro J, et al. Osteopontin is dispensable for protection against high load systemic fungal infection. Int Immunopharmacol 2008;8(10):1441-8. [65] Kandadi MR, Frankel AE, Ren J. Toll-like receptor 4 knockout rescues against anthrax lethal toxin-induced cardiac contractile dysfunction: role of autophagy. Br J Pharmacol 2012. [66] Megias J, Yanez A, Moriano S, O'Connor JE, Gozalbo D, Gil ML. Direct toll-like receptor-mediated stimulation of hematopoietic stem and progenitor cells occurs in vivo and promotes differentiation toward macrophages. Stem Cells 2012;30(7):1486-95. [67] Fitzpatrick LE, Chan JWY, Sefton MV. On the mechanism of poly(methacrylic acid -co- methyl methacrylate)-induced angiogenesis: Gene expression analysis of dTHP-1 cells. Biomaterials 2011;32:8957-67. [68] Fitzpatrick LE, Lisovsky A, Sefton MV. The expression of sonic hedgehog in diabetic wounds following treatment with poly(methacrylic acid-co-methyl methacrylate) beads. Biomaterials 2012;33(21):5297-307. [69] Martin DC, Semple JL, Sefton MV. Poly(methacrylic acid-co-methyl methacrylate) beads promote vascularization and wound repair in diabetic mice. J Biomed Mater Res A 2010;93(2):484-92. [70] Bota PC, Collie AM, Puolakkainen P, Vernon RB, Sage EH, Ratner BD, et al. Biomaterial topography alters healing in vivo and monocyte/macrophage activation in vitro. J Biomed Mater Res A 2010;95(2):649-57.

170

Chapter 8 Conclusions

The goal of this thesis was to identify molecules involved in MAA-mediated angiogenesis and wound healing, and to develop a hypothesis describing the mechanism of MAA-mediated angiogenesis. The effect of MAA beads on gene and protein expression was analyzed in a diabetic wound model and in macrophage-like dTHP-1 cells in vitro. Our conclusions are as follows:

General

1. MAA beads elicited gene and protein expression profiles in diabetic wounds and dTHP-1 cells that were distinct from PMMA beads and no bead controls.

Host-response to MAA beads in diabetic wounds

1. MAA beads increased the density of CD31-positive vessel-like structures in the granulation tissue of small (7.5 mm x 7.5 mm) cutaneous wounds, while PMMA beads did not.

2. MAA beads may promote angiogenesis by increasing the expression of pro-angiogenic sonic hedgehog (Shh), and Shh transcription factor Gli3 in cutaneous wounds in diabetic (db/db) mice, compared to PMMA beads and no beads.

3. MAA beads increased the expression of interleukin 1 β (IL-1β), tumor necrosis factor α (TNF-α) and osteopontin (OPN), indicating that the host-response to MAA beads involves modulation of the inflammatory response.

4. MAA beads increased the expression of sprouty homolog 2 (Spry2), a negative regulator of MAPK signaling with angiostatic properties, compared to PMMA bead and no beads. This may reflect increased numbers of endothelial and vascular smooth muscle cells (which constitutively express Spry2) in the wound bed. As Spry2 expression is also upregulated during later phases of wound healing, it may also reflect that despite increased expression of inflammatory factors, the MAA-treated wounds are at a more advanced stage of wound healing than PMMA treated or untreated wounds.

171

5. MAA beads did not affect the expression of pro-angiogenic growth factors (vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF) and platelet-derived growth factor B (PDGF-B)) in diabetic wounds. The lack of effect on typical pro-angiogenic growth factors suggests that the mechanism of action of the MAA beads is complex (i.e. through Shh signaling) and does not simply occur through enhancing the expression of pro- angiogenic factors.

The effect of MAA beads on macrophage-like dTHP-1

1. MAA beads increased the expression of IL-1β, IL-6 and TNF-α, reflecting the expression profile observed in diabetic wounds, supporting the conclusion that MAA beads modulate the inflammatory response.

2. MAA beads decreased the expression of C-X-C motif chemokine 10 (CXCL10), a chemokine with potent anti-angiogenic activity, compared to PMMA beads and no beads. The decrease in CXCL10 would promote angiogenesis and likely contributes to MAA- mediated angiogenesis observed in vivo.

3. MAA beads did not affect the expression of VEGF, bFGF and PDGF-B in dTHP-1 cells, supporting the conclusion that MAA beads do not promote angiogenesis by simply increasing the expression of pro-angiogenic growth factors.

4. MAA beads modulated the expression of OPN, but the effect was dependent on the cell type. OPN mRNA was increased in HUVEC treated with MAA beads, but decreased in dTHP-1 cells. However, MAA beads had no effect on the dTHP-1 expression of secreted OPN.

5. MAA beads decreased the expression of CXCL12 protein in dTHP-1 cells, compared to PMMA and no beads. Conditioned medium from MAA-treated dTHP-1 cells decreased the expression of C-X-C chemokine receptor type 4 (CXCR4) in HUVEC and decreased HUVEC migration, compared to HUVEC cultured in PMMA-treated dTHP-1 conditioned medium and untreated (i.e. no bead) dTHP-1 conditioned medium.

172

Hypothesis of MAA-mediated angiogenesis

1. Bioinformatics analysis of the proteins highlighted in the host/cell-response to MAA beads showed enrichment for toll-like receptor (TLR) signaling.

2. MAA beads are hypothesized to modulate (TLR) in macrophages, resulting in increased expression of NF-κB target genes (Shh, IL-1β, IL-6, TNF-α), and decreased expression of activator protein 1 (AP-1) and interferon regulatory transcription factor target genes (OPN, CXCL10). Angiogenesis is likely promoted by the increase in Shh and the decrease in CXCL10. Shh signaling subsequently induces OPN expression in a paracrine fashion (i.e. in non-macrophage cell types), further promoting angiogenesis and wound healing.

173

Appendix I Microarray analysis of MAA-treated diabetic wounds 1 Introduction

Gene expression microarrays were used in an initial effort to identify important pathways or processes in the host response to MAA beads. The following chapter is a brief summary of the methods and results from this analysis. While a small number of genes related to wound healing, angiogenesis and inflammation were highlighted in this experiment; this experiment did not highlight the involvement of particular pathways of interest. Consequently, other techniques for analyzing gene expression were pursued.

2 Materials and methods 2.1 Animal procurement

7-9 week old, genetically diabetic, male BKS.Cg-m+/+ Leprdb/J mice (db/db) were ordered from Jackson Laboratories (Bar Harbor, ME, USA) with the approval of the University of Toronto Animal Care Committee and were housed under sterile conditions in the Department of Comparative Studies (DCM) Animal facility. Mice were allowed to acclimatize for one week prior to surgery. Mice were 8-10 weeks old at the time of surgery. No weights or blood glucose levels were recorded.

2.2 Bead preparation

MAA beads (150 - 250 µm in diameter, 45 mol % methacrylic acid) were obtained from Rimon Therapeutics (Toronto, Canada) and PMMA beads (same diameter, 100 mol % methyl methacrylate) were acquired from Polysciences (Warrington, PA, USA). As described elsewhere[1], MAA and PMMA beads were washed in a sonicating water bath for 20 minutes in 95% ethanol (MAA beads) or 1N HCl (PMMA beads) ten times, then rinsed in endotoxin free water five times. After drying in a vacuum, beads were tested for endotoxin using LAL Pyrochrome Endotoxin Kit (Associates of Cape Cod, Falmouth, MA). Only beads with less than 0.25 EU/ml (100mg beads/ml) were used in vivo for wound treatment.

174

2.3 Wounding assay

Mice were anaesthetized with isoflurane. The dorsal surface of the mouse was shaved with an electric shaver, followed by treatment with a hair removal cream (Veet). The surgical site was sterilized with Betadine and 70% ethanol. Scissors were used to create mid-dorsal, 7.5 mm x 7.5 mm full-thickness, bilateral wounds by excising the epidermis and dermis including the panniculus carnosus. Either 7 mg of dry MAA beads or 7 mg of dry PMMA beads were applied topically to the wound beds, or they were left untreated. Each mouse received the same treatment for both wounds (i.e. left and right wounds were both treated with MAA beads). The wounds were left undressed and scabs were allowed to form. Following the surgical procedure, buprenorphine (0.03 mg/kg) was given subcutaneously for analgesia. Thereafter, the mice were caged individually. Mice were sacrificed using CO2 asphyxiation, followed by cervical dislocation, on day 4 and 7 post-wounding.

2.4 Total RNA preparation

The wound beds were excised and trimmed of any visible fat and healthy skin. The tissue was then quartered and placed in RNAlater RNA Stabilization Reagent (Qiagen, Mississauga, ON) and kept at 4oC for no more than 10 days. Total RNA was isolated from pooled (day 7) or unpooled (day 4) tissue samples using TRIzol reagent (Invitrogen, Burlington, ON) according to manufacturer’s protocol. The isolated RNA was further purified using the Qiagen RNeasy Mini kit (Qiagen) according to the manufacturer’s Clean-up protocol with on-column DNAse digestion. The concentration of RNA was determined by measuring the absorbance at 260 nm using a Nanodrop ND-1000 spectrophometer (NanoDrop Technologies, Inc., USA). All RNA samples had a 260/280 ratio greater than or equal to 2.0.

2.5 Microarray analysis

Total RNA samples were submitted to The Center of Applied Genomics (TCAG), who first tested the RNA quality using a Bioanalyzer (Agilent Technologies) before performing the microarray. Only RNA samples with RNA integrity number (RIN) greater than 7 were used for microarray analysis. TCAG staff performed a Mouse Gene 1.0 ST microarray (Affymetrix, Santa Clara, CA, USA), using Ambion WT Expression Kit (Ambion Inc, Austin, TX, USA) for sample

175 preparation. Affymetrix GeneChip Scanner 3000 and Affymetrix Expression Console software were used to scan the array and create .CEL files.

2.6 Data Analysis

The microarray data (.CEL) files were analyzed using Partek Genomic Suite software (Partek Inc., St Louis, MO) using the Gene expression workflow. All array data were normalized using the Robust Multi-array Average (RMA) method. Analysis of variance (ANOVA) and linear contrasts were used to create gene lists of interest (based on fold-change and p-value criteria). Three contrasts were used to describe the effect MAA beads (MAA/No bead) or PMMA beads (PMMA /No bead) had on the gene expression within the granulation tissue of the diabetic mice and to highlight the response to the methacrylic-acid containing beads, relative to PMMA genes (MAA / PMMA). The adjustment for multiple testing was carried out using the Benjamini-Hochberg False Discovery Rate (FDR) Step-up multiple test correction (FDR). Gene lists were created by selecting for genes with expression fold-changes greater than ± 1.5, and FDR < 0.05 (more conservative) or an unadjusted p-values < 0.05 (less conservative).

2.7 Cluster analysis and pathway analysis (gene ontology)

Once gene lists were created, gene ontology (Functional Annotation Clustering) and pathway analyses (Pathways) were conducted using Database for Annotation, Visualization and Integrated Discovery (DAVID)[2,3]. As not all genes are fully annotated, manual searches on gene/protein function, with specific emphasis on keywords “inflammation, wound healing, angiogenesis” were performed on Pubmed (http://www.ncbi.nlm.nih.gov/pubmed/), Entrez Gene (www.ncbi.nlm.nih.gov/entrez/ query.fcgi?db=gene) and Gene Cards (http://www.genecards.org/) databases.

3 Results 3.1 Sample preparation

For day 7 explants, wound tissues from three mice treated with MAA beads were pooled and homogenized, and the total RNA was isolated to yield one sample (Figure AI-1a). This process was repeated for PMMA-treated wounds and untreated wounds, in triplicate, such that a total of 9 RNA samples were prepared (3 samples per treatment, with 3 mice pooled per sample), Figure

176

3a. The tissue was pooled to ensure sufficient amounts of RNA were extracted from the tissue samples. Pooling also has been shown to reduce variation in gene expression among samples of the same treatment[4]. The RNA isolated from the Day 7 wound tissue had high purity (A260/A280 > 2.0) and quality (RIN > 9.0).

"##!$%&!!!!!!!!!!!!!!!!!!!"##!$'&!!!!!!!!!!!!!!!!!!!!"##!$(&! "##0!!!!$%&!!!!!!$'&!!!!!!$(&!!!!!!$1&!

)""#!$%&!!!!!!!!!!!!!!!)""#!$'&!!!!!!!!!!!!!!!!)""#!$(&! )""#0!!!!$%&!!!!!!$'&!!!!!!$(&!!!!!!$1&!

*+,-.+/!$%&!!!!!!*+,-.+/!$'&!!!!!!*+,-.+/!$(&! *+,-.+/0!!!!$%&!!!!!!$'&!!!!!!$(&!!!!!!$1&! !"# # # # # # # # $"# !

The RNA yield from the tissue samples on day 7 was sufficiently high that it was evident pooling tissue samples was unnecessary. Consequently, day 4 samples were not pooled, to minimize the loss of subtle effect among samples due to pooling. The number of arrays was increased to 4 arrays per treatment (i.e. 12 mice: 4 per treatment, one mouse per array) (Figure AI-1b). The RNA quality and purity was acceptable (A260/A280 ≥ 2.0, RIN >7.9).

The mice used in the day 4 microarray group were approximately 2 weeks younger than the day 4 qPCR group, discussed in Chapter 4. Although the mice used for the microarray study were not weighed, it is likely that the day 4 microarray mice were less obese than the day 4 qPCR mice. The day 7 mice were same age in both the microarray and qPCR studies.

177

3.2 Microarray data analysis

3.2.1 Gene lists selected using multiple-test corrected false-discover rates

The microarray .CEL files for day 4 and 7 samples were analyzed using Partek Software as described in the methods. Following RMA normalization, ANOVA and linear contrasts were used to compare the gene expression among the three treatment groups: MAA, PMMA and no bead control (untreated). Gene lists were created by filtering for genes that had FC > 1.5 and FDR < 0.05. At day 4, MAA treatment significantly changed the expression of 25 genes (Table AI-1), compared to untreated wounds and 24 genes (Table AI-2), compared to PMMA-treated (FDR < 0.05, FC > ±1.5). Twelve genes [1700029F09Rik, 1700034H14Rik, 1810020D17Rik, 2410019A14Rik, Bxdc2, Commd6, Cysltr1, Dnajc12, Faim, Mphosph6, Mtcp1, Olfr471, Rabggtb] were common to both gene lists (MAA vs. untreated; and MAA vs PMMA). PMMA treatment changed the expression of 2 genes at day 4, compared to untreated wound samples (Table AI-3). No differences in gene expression were found among the three treatment groups (MAA, PMMA, No bead) for the day 7 samples (FDR < 0.05, FC > 1.5), possibly due to sample pooling and/or fewer replicates (n=3).

Table AI-1. Gene list of interest for "MAA/No bead" contrast, using fold-change > ± 1.5 and FDR < 0.05 as selection criteria. Official Gene Symbol RefSeq p-value FDR Fold-change 1700029F09Rik BC119032 3.86E-04 0.044 -1.65 1700034H14Rik BC034297 1.50E-04 0.030 -1.56 1810020D17Rik BC026557 1.45E-04 0.030 -1.55 2410019A14Rik ENSMUST00000098534 2.63E-04 0.038 -1.63 2410129H14Rik BC117005 3.14E-06 0.012 -1.51 Ankrd1 NM_013468 1.45E-04 0.030 -1.51 Bxdc2 NM_026396 4.41E-05 0.019 -1.54 Commd6 NM_001033132 1.03E-04 0.026 -1.60 Cysltr1 NM_021476 2.36E-05 0.019 -1.81 Dnajc12 NM_013888 2.08E-04 0.035 -1.51 Eml5 NM_001081191 4.80E-04 0.046 -1.56 Faim NM_001122851 2.00E-04 0.035 -1.61 Fastkd2 NM_172422 4.20E-04 0.045 -1.57 Gm10305 ENSMUST00000094956 4.66E-04 0.046 -1.78 LOC100134980 BC016578 4.39E-04 0.046 -1.53 Med18 NM_026039 3.70E-04 0.043 -1.52 Mphosph6 NM_026758 1.13E-04 0.027 -1.92 Nufip1 NM_013745 2.50E-04 0.037 -1.50 Olfr471 ENSMUST00000084758 1.50E-04 0.030 1.63 Ppil4 NM_026141 2.51E-04 0.037 -1.57

178

Official Gene Symbol RefSeq p-value FDR Fold-change Psmb9 NM_013585 1.83E-04 0.034 1.56 Rabggtb NM_011231 9.50E-05 0.026 -1.81 Sfrs6 NM_026499 9.26E-05 0.025 -1.58 Tas2r126 NM_207028 1.48E-04 0.030 -1.82 Zh2c2 NM_026250 4.58E-04 0.046 -1.52

Table AI- 2. Gene list of interest for "MAA/PMMA" contrast, using fold-change > ± 1.5 and FDR < 0.05 as selection criteria. Official Gene Symbol RefSeq p-value FDR Fold-change 1700029F09Rik BC119032 1.21E-04 0.035 -1.55 1700034H14Rik BC034297 2.47E-04 0.042 -1.61 1810020D17Rik BC026557 2.58E-04 0.043 -1.61 Akr1c18 NM_134066 3.37E-04 0.048 -2.22 Bxdc2 NM_026396 3.46E-05 0.029 -1.52 C430014K11Rik ENSMUST00000092304 2.82E-05 0.026 -1.52 Commd6 NM_001033132 3.34E-05 0.028 -1.52 Cysltr1 NM_021476 1.68E-05 0.024 -1.77 Dnajc12 NM_013888 2.04E-04 0.041 -1.51 Dpp4 NM_010074 1.17E-04 0.035 -1.55 Eno2 NM_013509 3.43E-04 0.049 -1.59 Gm9958 ENSMUST00000068250 2.92E-05 0.026 -1.54 Erh NM_007951 2.21E-04 0.041 -1.52 Faim NM_001122851 2.70E-04 0.043 -1.64 Fkbp3 NM_013902 1.65E-04 0.038 -1.52 Gpm6a NM_153581 2.02E-04 0.041 -1.78 Mphosph6 NM_026758 1.44E-05 0.024 -1.70 Mtcp1 NM_001039373 1.36E-04 0.035 -1.50 Olfr471 ENSMUST00000084758 1.42E-04 0.036 1.62 Rab12 NM_024448 5.56E-06 0.020 -1.51 Rabggtb NM_011231 4.03E-06 0.019 -1.57 Trim13 NM_023233 2.04E-04 0.041 -1.55 Zcchc7 NM_138590 1.38E-04 0.035 -1.50 2410019A14Rik ENSMUST00000098534 5.83E-05 0.033 -1.51

Table AI- 3. Gene list of interest for "MAA/PMMA" contrast, using fold-change > ± 1.5 and FDR < 0.05 as selection criteria. Gene Symbol RefSeq p-value FDR Fold-Change Eml5 NM_001081191 9.19E-07 0.013 -1.33047 Lats1 NM_010690 4.78E-07 0.013 -1.28761

Gene lists from Tables AI-1, AI-2 and AI-3 were analyzed with DAVID functional annotation. RNA degradation pathway (mmu03018) was enriched (p-value = 0.0006; Bonferroni 0.0032;

179

FDR 0.277) in the MAA/No bead gene list (Table A1-2). However, no other pathway or process was clearly annotated at either time point.

3.2.2 Gene lists selected using unadjusted p-values

Unadjusted p-values are less conservative than FDR because they do not control the family-wide false-positive rate, which is potentially very high in microarray data. However, controlling the FDR can lead to a high false negative rate (FNR), which can cause less significant, but interesting gene expression changes to be overlooked[5]. Consequently, a second set of gene lists for day 4 and 7 microarrays were created by filtering for genes that had unadjusted p-values < 0.05 and fold-changes > ± 1.5 (refer to Table AI-S1 through AI-S6). These criteria yielded much larger gene lists. On day 4, the majority of genes in the filtered lists (p < 0.05, FC > ± 1.5) had decreased expression for each contrast (MAA/PMMA, MAA/No bead and PMMA/No bead), while most highlighted genes were increased on Day 7, Figure AI-2. Venn diagrams, highlighting any overlap among the gene lists were also generated using GeneVenn[6] (Figure AI-3).

&#!"

/,0-12345(62/" &!!" 4.12345(62/" %#!"

%!!"

$#!" !"#$%&'()'*%+%,' $!!"

#!"

!" '((")*"+,-" '((")*".''(" .''(")*"+,-" '((")*"+,-" '((")*".''(" .''(")*"+,-" -./'0'''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''-./'1' Figure AI-2. Overview of gene regulation in treated and untreated diabetic wounds. Genes with p-values less than 0.05 and fold-changes greater than 1.5 from ANOVAs contrasting the three wound treatments.

180

Figure AI-3. Venn diagrams showing the number of transcripts that are differentially expressed in response to MAA beads and PMMA beads. Refer to Supplemental Tables AI- S7 and AI-S8 for the gene list corresponding to each region in the Venn diagrams. Diagrams created using GeneVenn (http://genevenn.sourceforge.net/index.htm)[6].

From the lists generated using unadjusted p-values (p < 0.05) and FC > ± 1.5, genes involved in wound healing, inflammation and angiogenic pathways were selected, using DAVID annotation or by manually performing literature searches on individual genes within the Day 4 and Day 7 lists, Table AI-4 and AI-5 respectively. Overall, the microarray results at day 4 and day 7 did not suggest a specific target pathway to focus on, however the collected data did give some insight onto the activity occurring within the wound beds.

Table AI-4. Day 4 genes of interest. Genes of interest were selected from the lists of genes differentially regulated in MAA treated tissue samples (p < 0.05, FC > 1.5).

MAA VS. MAA VS. GENE DESCRIPTION PMMA NO BEAD DAY 4 DAY

CD48 p = 0.0043 p=0.00046 Antigen expressed by T-cells, NK-cells and APC FC= -1.56 FC= -1.29 (macrophages)[7,8]

CD209 p = 0.03 DC-SIGN/CD209 FC = -1.54 M2 macrophage marker; Expressed by dendritic cells and macrophages, high affinity for ICAM2 and ICAM3[9]

MARKERS Mtcp1 p= 0.00014 p = 0.0071 Mature T-cell proliferation 1

IMMUNECELL FC = -1.5 FC = -1.8 Mtcp-1 directly activate AKT signaling pathways (anti-apoptotic)[10]

181

MAA VS. MAA VS. GENE DESCRIPTION PMMA NO BEAD DAY 4 DAY

CysLTR1 p = 1.7e-5 p = 2.4e-5 Cysteinyl leukotriene receptor 1 FC = -1.76 FC = -1.8 Potent pro-inflammatory factors; CysTR1 signaling induces vascular permeability; Involved in activated T-cells recruitement, potent macrophage activator[11-13] FAIM p = 0.0003 p = 0.0002 Fas apoptotic inhibitory molecule FC = -1.64 FC = -1.61 Blocks FAS activation (and FAS-induced apoptosis)[14] Perp p = 0.021 p53 apoptosis effector related to PMP-22 FC = -1.74 p53 target gene, mediates apoptosis[15] Lgals7 p = 0.003 Galectin-7 FC = -2.09 associated with p53 apoptosis and epithelial cell migration[16]

APOPTOSIS Trp63 p = 0.003 transformation related protein 63 FC = -1.65 encodes p63 protein; p63 is involved in p53 apoptosis pathways; Regulates anti- and pro-apoptotic pathways in keratinocytes[17] CXCL15 p = 0.047 CXCL15/Lungkine

FC = -2.3 Inflammatory chemokines; involved in neutrophils trafficking in the lungs [18] CXCL10 p = 0.022 CXCL10/IP-10 (10 kDa interferon-gamma- FC = 1.62 induced protein) Anti-angiogenic chemokines[19,20] Expression is highly induced by IFN-γ; Recruites Th1 and NK cells expressing CXCR3 while antagonizes recruitement of CCR3- expressing Th2 cells [21] CCL12 p = 0.022 Potent chemoattractant for macrophages and binds FC = -1.55 to same receptor (CCR2) as CCL2 (MCP-1)[22]

CYTOKINES & CHEMOKINES & CYTOKINES Chi3l3 p = 0.027 chitinase 3-like 3 (Ym1) FC = -1.7 Highly upregulated in M2 macrophage [23] MARCO p = 0.036 macrophage receptor with collagenous structure FC = -1.57 Macrophage antigen, response to bacterial and inflammation response; Innate scavenger receptor, involved in clearance of apoptotic cells [24] MMP1b p = 0.003 Matrix metalloproteinase 1b FC = -1.73 Peptidase activity[25]

OTHER Sfrp2 p = 0.0014 p = 0.4 secreted frizzled-related protein 2 FC = -1.53 FC = -2.1 Modulates Wnt signaling; Sfrp2 protects cells from apoptosis by blocking the effect of canonical Wnt3a [26] Stimulates angiogenesis via a calcineurin/NFAT signaling pathway by modulating non-canonical

182

MAA VS. MAA VS. GENE DESCRIPTION PMMA NO BEAD DAY 4 DAY

Wnt signaling[27] Expression is repressed by SHh[28] OGN p = -0.02 p = 0.047 Osteoglycin/Mimecan FC = -1.63 FC = -1.77 Produced mainly by smooth muscle cells (SMC) and perivascular fibroblasts; Downregulation in SMC may be required for arteriogenesis[29] Regulated by NF-κB (with secreted frizzled) in mouse embryonic fibroblasts[30] Processed by BMP-1 and regulates type 1 collagen fibrillogenesis[31] Expression down-regulated by bFGF, TGF-beta, PDGF, and angiotensin-2[32] Nkiras1 p = 0.0005 p = 0.0047 NF-κB inhibitor interacting Ras-like protein 1 FC = -1.51 FC = -1.7 Binds to NF-κB inhibitors and decreases their degradation rate, thereby regulating NF-κB signaling and induced expression of pro- inflammatory cytokines [33]. NF-κB signaling is antiapoptotic and pro- angiogenic in tumours [34,35] Olfr p = 0.045 – 0.0001 Olfactory and Vomeronasal Receptors FC = 1.5 – 1.7 Also upregulated (FC > 1.25) in day 7 data V1r p = 0.47 – 0.009 Unknown function outside of FC = 1.51 – 1.72 olfactory/vomeronasal organs G-coupled Receptors.

Table 1. Day 7 Microarray Results. Genes of interest were selected from the lists of genes differentially regulated in MAA treated tissue samples (p < 0.05, FC > 1.5) .

MAA VS. MAA VS. GENES DESCRIPTION PMMA CONTROL DAY 7 DAY

Ttn p = 0.007 p = 0.047 Titin FC = 2.34 FC = 3.20 Muscle protein, known to be expressed in

myofibroblast cell line BHK-21/C13[36] Mylk4 p = 0.015 p = 0.035 Myosin light chain kinase family, member 4 FC = 1.81 FC = 2.03 muscle marker Jph1 p = 0.010 p = 0.014 Junctophilin 1 FC = 1.69 FC = 1.91 Muscle protein[37] Actn2 p = 0.040 Actinin alpha 2 FC = 2.64 muscle contraction[38]

MUSCLE MARKERS MUSCLE Actn3 p = 0.049 Actinin alpha 3 FC = 2.35 muscle contraction [38]

183

MAA VS. MAA VS. GENES DESCRIPTION PMMA CONTROL DAY 7 DAY

Mef2c p = 0.016 Myocyte enhancer factor 2C FC = 1.82 Induced during myogenic differentiation[39] Involved in Galpha13 embryonic angiogenesis[40] Activated by VEGF in endothelial cells[41] Ogn p = 0.029 See description in Table 1. FC = 2.03 MMP3 p = 0.013 Matrix metalloproteinase 3 FC = -1.63 Expressed in healing wounds by fibroblasts[42] Epr p = 0.043 Epiregulin FC = 1.66 Member of the EGF family; Plays critical role in immune/inflammatory-related responses of keratinocytes and macrophages[43] Shown to enhance wound healing in cutaneous OTHER wounds in mice[44] Expressed at healing wound-edge in excisional murine cutaneous wounds[45] Lcn2 p = 0.033 Lipocalin-2/neutrophil gelatinase-associated FC = -1.66 lipocalin Expressed by host cells in acute response to infection and can induce apoptosis in hematopeitic cells[46,47] Binds mediators of inflammation (leukotrienes, LPS)[48] and MMP-9, preventing its degradation[49] 4 Summary

Gene expression microarrays were used to identify gene of interest in the host response to MAA beads. In general, only subtle changes in gene expression were observed among the different treatment groups at day 4 and day 7; as characterized by the small fold-changes, and FDR/high p-values. Lists of genes that had changes in expression in response to MAA or PMMA beads were generated and analyzed using bioinformatics software. The data did not generate any target pathways to focus on in future experiments; however, a small list of genes of interest was generated for each time point using manual annotation.

5 References [1] Fitzpatrick LE, Lisovsky A, Sefton MV. The expression of sonic hedgehog in diabetic wounds following treatment with poly(methacrylic acid-co-methyl methacrylate) beads. Biomaterials 2012;33(21):5297-307.

184

[2] Huang da W, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 2009;4(1):44-57. [3] Huang da W, Sherman BT, Lempicki RA. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic acids research 2009;37(1):1-13. [4] Affymetrix I. Sample Pooling for Microarray Analysis: A Statistical Assessment of Risks and Biases. Technical Notes 2004. [5] Pawitan Y, Michiels S, Koscielny S, Gusnanto A, Ploner A. False discovery rate, sensitivity and sample size for microarray studies. Bioinformatics 2005;21(13):3017-24. [6] Pirooznia M, Nagarajan V, Deng Y. GeneVenn - A web application for comparing gene lists using Venn diagrams. Bioinformation 2007;1(10):420-2. [7] Abadia-Molina AC, Ji H, Faubion WA, Julien A, Latchman Y, Yagita H, et al. CD48 controls T-cell and antigen-presenting cell functions in experimental colitis. Gastroenterology 2006;130(2):424-34. [8] Lee KM, Forman JP, McNerney ME, Stepp S, Kuppireddi S, Guzior D, et al. Requirement of homotypic NK-cell interactions through 2B4(CD244)/CD48 in the generation of NK effector functions. Blood 2006;107(8):3181-8. [9] Martinez FO, Helming L, Gordon S. Alternative Activation of Macrophages: An Immunologic Functional Perspective. Annu Rev Immunol 2009;27(1):451-83. [10] Auguin D, Barthe P, Royer C, Stern MH, Noguchi M, Arold ST, et al. Structural basis for the co-activation of protein kinase B by T-cell leukemia-1 (TCL1) family proto- oncoproteins. J Biol Chem 2004;279(34):35890-902. [11] Maekawa A, Austen KF, Kanaoka Y. Targeted gene disruption reveals the role of cysteinyl leukotriene 1 receptor in the enhanced vascular permeability of mice undergoing acute inflammatory responses. J Biol Chem 2002;277(23):20820-4. [12] Crowther SD, Rees PJ. Current treatment of asthma--focus on leukotrienes. Expert Opin Pharmacother 2000;1(5):1021-40. [13] Prinz I, Gregoire C, Mollenkopf H, Aguado E, Wang Y, Malissen M, et al. The type 1 cysteinyl leukotriene receptor triggers calcium influx and chemotaxis in mouse alpha beta- and gamma delta effector T cells. J Immunol 2005;175(2):713-9. [14] Rothstein TL. Inducible resistance to Fas-mediated apoptosis in B cells. Cell Res 2000;10(4):245-66. [15] Singaravelu K, Devalaraja-Narashimha K, Lastovica B, Padanilam BJ. PERP, a p53 proapoptotic target, mediates apoptotic cell death in renal ischemia. Am J Physiol Renal Physiol 2009;296(4):F847-58. [16] Saussez S, Kiss R. Galectin-7. Cell Mol Life Sci 2006;63(6):686-97. [17] Borrelli S, Candi E, Alotto D, Castagnoli C, Melino G, Vigano MA, et al. p63 regulates the caspase-8-FLIP apoptotic pathway in epidermis. Cell Death Differ 2009;16(2):253- 63.

185

[18] Rossi DL, Hurst SD, Xu Y, Wang W, Menon S, Coffman RL, et al. Lungkine, a novel CXC chemokine, specifically expressed by lung bronchoepithelial cells. J Immunol 1999;162(9):5490-7. [19] Frangogiannis NG. The role of the chemokines in myocardial ischemia and reperfusion. Curr Vasc Pharmacol 2004;2(2):163-74. [20] Belperio JA, Keane MP, Arenberg DA, Addison CL, Ehlert JE, Burdick MD, et al. CXC chemokines in angiogenesis. J Leukoc Biol 2000;68(1):1-8. [21] Proost P, Vynckier AK, Mahieu F, Put W, Grillet B, Struyf S, et al. Microbial Toll-like receptor ligands differentially regulate CXCL10/IP-10 expression in fibroblasts and mononuclear leukocytes in synergy with IFN-gamma and provide a mechanism for enhanced synovial chemokine levels in septic arthritis. Eur J Immunol 2003;33(11):3146-53. [22] Tsui P, Das A, Whitaker B, Tornetta M, Stowell N, Kesavan P, et al. Generation, characterization and biological activity of CCL2 (MCP-1/JE) and CCL12 (MCP-5) specific antibodies. Hum Antibodies 2007;16(3-4):117-25. [23] Gordon S. Alternative activation of macrophages. Nat Rev Immunol 2003;3(1):23-35. [24] Rogers NJ, Lees MJ, Gabriel L, Maniati E, Rose SJ, Potter PK, et al. A defect in Marco expression contributes to systemic lupus erythematosus development via failure to clear apoptotic cells. J Immunol 2009;182(4):1982-90. [25] Balbin M, Fueyo A, Knauper V, Lopez JM, Alvarez J, Sanchez LM, et al. Identification and enzymatic characterization of two diverging murine counterparts of human interstitial collagenase (MMP-1) expressed at sites of embryo implantation. J Biol Chem 2001;276(13):10253-62. [26] Zhang Z, Deb A, Pachori A, He W, Guo J, Pratt R, et al. Secreted frizzled related protein 2 protects cells from apoptosis by blocking the effect of canonical Wnt3a. J Mol Cell Cardiol 2009;46(3):370-7. [27] Courtwright A, Siamakpour-Reihani S, Arbiser JL, Banet N, Hilliard E, Fried L, et al. Secreted frizzle-related protein 2 stimulates angiogenesis via a calcineurin/NFAT signaling pathway. Cancer Res 2009;69(11):4621-8. [28] Ingram WJ, Wicking CA, Grimmond SM, Forrest AR, Wainwright BJ. Novel genes regulated by Sonic Hedgehog in pluripotent mesenchymal cells. Oncogene 2002;21(53):8196-205. [29] Kampmann A, Fernandez B, Deindl E, Kubin T, Pipp F, Eitenmuller I, et al. The proteoglycan osteoglycin/mimecan is correlated with arteriogenesis. Mol Cell Biochem 2009;322(1-2):15-23. [30] Li X, Massa PE, Hanidu A, Peet GW, Aro P, Savitt A, et al. IKKalpha, IKKbeta, and NEMO/IKKgamma are each required for the NF-kappa B-mediated inflammatory response program. J Biol Chem 2002;277(47):45129-40. [31] Ge G, Seo NS, Liang X, Hopkins DR, Hook M, Greenspan DS. Bone morphogenetic protein-1/tolloid-related metalloproteinases process osteoglycin and enhance its ability to regulate collagen fibrillogenesis. J Biol Chem 2004;279(40):41626-33.

186

[32] Shanahan CM, Cary NR, Osbourn JK, Weissberg PL. Identification of osteoglycin as a component of the vascular matrix. Differential expression by vascular smooth muscle cells during neointima formation and in atherosclerotic plaques. Arterioscler Thromb Vasc Biol 1997;17(11):2437-47. [33] Fenwick C, Na SY, Voll RE, Zhong H, Im SY, Lee JW, et al. A subclass of Ras proteins that regulate the degradation of IkappaB. Science 2000;287(5454):869-73. [34] Tabruyn SP, Memet S, Ave P, Verhaeghe C, Mayo KH, Struman I, et al. NF-kappaB activation in endothelial cells is critical for the activity of angiostatic agents. Mol Cancer Ther 2009;8(9):2645-54. [35] De Martin R, Hoeth M, Hofer-Warbinek R, Schmid JA. The transcription factor NF- kappa B and the regulation of vascular cell function. Arterioscler Thromb Vasc Biol 2000;20(11):E83-8. [36] van der Ven PF, Bartsch JW, Gautel M, Jockusch H, Furst DO. A functional knock-out of titin results in defective myofibril assembly. J Cell Sci 2000;113 ( Pt 8):1405-14. [37] Kakizawa S, Moriguchi S, Ikeda A, Iino M, Takeshima H. Functional crosstalk between cell-surface and intracellular channels mediated by junctophilins essential for neuronal functions. Cerebellum 2008;7(3):385-91. [38] Schiaffino S, Reggiani C. Molecular diversity of myofibrillar proteins: gene regulation and functional significance. Physiol Rev 1996;76(2):371-423. [39] McDermott JC, Cardoso MC, Yu YT, Andres V, Leifer D, Krainc D, et al. hMEF2C gene encodes skeletal muscle- and brain-specific transcription factors. Mol Cell Biol 1993;13(4):2564-77. [40] Liu G, Han J, Profirovic J, Strekalova E, Voyno-Yasenetskaya TA. Galpha13 regulates MEF2-dependent gene transcription in endothelial cells: role in angiogenesis. Angiogenesis 2009;12(1):1-15. [41] Maiti D, Xu Z, Duh EJ. Vascular endothelial growth factor induces MEF2C and MEF2- dependent activity in endothelial cells. Invest Ophthalmol Vis Sci 2008;49(8):3640-8. [42] Li M, Moeen Rezakhanlou A, Chavez-Munoz C, Lai A, Ghahary A. Keratinocyte- releasable factors increased the expression of MMP1 and MMP3 in co-cultured fibroblasts under both 2D and 3D culture conditions. Mol Cell Biochem 2009. [43] Sugiyama S, Nakabayashi K, Baba I, Sasazuki T, Shirasawa S. Role of epiregulin in peptidoglycan-induced proinflammatory cytokine production by antigen presenting cells. Biochem Biophys Res Commun 2005;337(1):271-4. [44] Draper BK, Komurasaki T, Davidson MK, Nanney LB. Topical epiregulin enhances repair of murine excisional wounds. Wound Repair Regen 2003;11(3):188-97. [45] Roy S, Khanna S, Rink C, Biswas S, Sen CK. Characterization of the acute temporal changes in excisional murine cutaneous wound inflammation by screening of the wound- edge transcriptome. Physiol Genomics 2008;34(2):162-84. [46] Persengiev SP, Devireddy LR, Green MR. Inhibition of apoptosis by ATFx: a novel role for a member of the ATF/CREB family of mammalian bZIP transcription factors. Genes Dev 2002;16(14):1806-14.

187

[47] Devireddy LR, Teodoro JG, Richard FA, Green MR. Induction of apoptosis by a secreted lipocalin that is transcriptionally regulated by IL-3 deprivation. Science 2001;293(5531):829-34. [48] Goetz DH, Willie ST, Armen RS, Bratt T, Borregaard N, Strong RK. Ligand preference inferred from the structure of neutrophil gelatinase associated lipocalin. Biochemistry 2000;39(8):1935-41. [49] Yan L, Borregaard N, Kjeldsen L, Moses MA. The high molecular weight urinary matrix metalloproteinase (MMP) activity is a complex of gelatinase B/MMP-9 and neutrophil gelatinase-associated lipocalin (NGAL). Modulation of MMP-9 activity by NGAL. J Biol Chem 2001;276(40):37258-65.

188

6 Supplemental Tables Table AI-S1. List of genes for Day 4, MAA/No bead (FC > 1.5 and p-value < 0.05). Gene Symbol RefSeq p-value Fold-Change (MAA/No bead) (MAA/No bead) Genes with increased expression in MAA treated wounds Olfr121 NM_146629 4.55E-02 2.08 Rtp4 NM_023386 2.50E-03 1.93 Ifit1 NM_008331 2.55E-02 1.92 Irf7 NM_016850 2.33E-02 1.91 Isg15 NM_015783 2.91E-02 1.91 Ly6i NM_020498 4.56E-02 1.84 OTTMUSG00000005723 DQ508487 1.01E-02 1.83 AW112010 EF660528 1.05E-02 1.81 Olfr1199 NM_146458 2.76E-02 1.80 EG668725 DQ386867 2.76E-02 1.80 Olfr197 NM_146484 3.16E-02 1.76 H2-T24 NM_008207 2.68E-02 1.75 EG434460 ENSMUST00000087473 1.39E-02 1.74 Aard NM_175503 3.58E-02 1.74 Dio3 NM_172119 1.79E-02 1.72 Olfr591 NM_001011847 1.43E-02 1.71 V1rg5 NM_134206 3.81E-02 1.69 Vmn2r105 NM_001104567 2.15E-02 1.69 Olfr215 NM_146446 3.18E-02 1.67 Sbpl NM_001077421 4.69E-03 1.67 ENSMUSG00000055419 AK047662 4.84E-02 1.65 Gm1758 ENSMUST00000116654 4.89E-02 1.65 Aldh1a2 NM_009022 2.96E-02 1.65 Tgtp NM_011579 4.66E-02 1.64 Gpx6 NM_145451 1.24E-02 1.63 Olfr471 ENSMUST00000084758 1.50E-04 1.63 Igtp NM_018738 3.17E-02 1.62 Cxcl10 NM_021274 2.22E-02 1.62 Odf3l2 ENSMUST00000095464 1.15E-02 1.61 Klk14 NM_174866 2.41E-02 1.60 Olfr60 NM_146955 2.00E-02 1.60 Ifit3 NM_010501 1.45E-02 1.59 Psmb8 NM_010724 1.60E-02 1.59 Olfr1489 NM_146635 3.81E-02 1.59 Cd52 NM_013706 3.38E-02 1.57 AU021034 BC120488 1.83E-02 1.57 4930529M08Rik ENSMUST00000099283 4.54E-02 1.57 Psmb9 NM_013585 1.83E-04 1.56 Ifit2 NM_008332 1.73E-02 1.56 Olfr1045 NM_147017 3.89E-02 1.55 Lonrf1 ENSMUST00000065297 2.42E-02 1.55 Prss21 NM_020487 1.91E-02 1.54

189

Gene Symbol RefSeq p-value Fold-Change (MAA/No bead) (MAA/No bead) Acox2 NM_053115 2.11E-03 1.54 Tpo NM_009417 3.29E-02 1.54 Csmd3 NM_001081391 1.20E-02 1.53 Olfr893 NM_146336 3.49E-02 1.53 Tmco5 NM_026104 1.66E-02 1.53 Fcgr4 NM_144559 1.98E-02 1.53 Slc22a16 NM_027572 4.06E-02 1.52 Rhox11 NM_198598 3.32E-02 1.51 EG237300 BC125502 9.99E-03 1.51 Olfr684 NM_207249 4.38E-02 1.50 Genes with decreased expression in MAA treated wounds Rfesd NM_178916 8.96E-03 -1.50 Crebzf NM_145151 1.40E-02 -1.50 Hrsp12 NM_008287 1.76E-03 -1.50 Elovl3 NM_007703 1.94E-02 -1.50 Nufip1 NM_013745 2.50E-04 -1.50 Gins1 BC027537 2.54E-03 -1.50 Syne2 NM_001005510 2.08E-02 -1.50 2410129H14Rik BC117005 3.14E-06 -1.51 1110059G10Rik BC034133 1.09E-03 -1.51 Exosc8 NM_027148 3.71E-03 -1.51 Eif2c3 NM_153402 3.51E-03 -1.51 Zim1 NM_011769 3.20E-02 -1.51 4930422I07Rik BC141010 6.64E-03 -1.51 Nudcd2 NM_026023 7.71E-04 -1.51 Chac2 NM_026527 2.76E-02 -1.51 Zfp131 NM_028245 4.93E-02 -1.51 Ankrd1 NM_013468 1.45E-04 -1.51 Dnajc12 NM_013888 2.08E-04 -1.51 Clcc1 NM_145543 3.28E-02 -1.51 Rnu2 NR_004414 3.94E-02 -1.51 Dnajb4 NM_025926 1.38E-03 -1.51 Pcdhb17 NM_053142 9.04E-04 -1.51 ENSMUSG00000066331 ENSMUST00000084967 1.43E-03 -1.52 Ccdc5 NM_146089 1.63E-02 -1.52 Asb11 NM_026853 3.84E-02 -1.52 Kpna2 NM_010655 2.63E-02 -1.52 Upk1b NM_178924 2.29E-02 -1.52 2310003F16Rik NM_026318 1.68E-03 -1.52 Med18 NM_026039 3.70E-04 -1.52 Dennd2c NM_177857 3.65E-02 -1.52 Arpp19 NM_021548 1.94E-02 -1.52 Fusip1 NM_010178 3.32E-02 -1.52 Zh2c2 NM_026250 4.58E-04 -1.52 Sprr2a NM_011468 7.90E-03 -1.52 Kpna2 NM_010655 2.25E-02 -1.53 Rars2 NM_181406 3.34E-02 -1.53

190

Gene Symbol RefSeq p-value Fold-Change (MAA/No bead) (MAA/No bead) A130014H13Rik AK079474 8.60E-03 -1.53 Guf1 NM_172711 7.56E-03 -1.53 Zfp748 NM_001035231 1.17E-03 -1.53 Bet1 NM_009748 1.26E-03 -1.53 LOC100134980 BC016578 4.39E-04 -1.53 2310035C23Rik NM_173187 3.17E-02 -1.53 Rbks NM_153196 3.11E-02 -1.53 Sprr2a ENSMUST00000090872 2.55E-02 -1.53 Sprr2a ENSMUST00000090872 2.55E-02 -1.53 2610203C20Rik ENSMUST00000098868 4.93E-03 -1.53 Tfam NM_009360 8.04E-03 -1.53 Taok1 NM_144825 5.02E-03 -1.53 Gjb6 NM_001010937 4.56E-02 -1.53 Zfand1 NM_025512 7.56E-03 -1.53 LOC626711 XR_004836 2.43E-03 -1.54 Grhl1 NM_145890 3.90E-02 -1.54 2610039C10Rik NM_025642 1.67E-02 -1.54 Rnu2 NR_004414 1.53E-02 -1.54 Syne2 NM_001005510 7.09E-04 -1.54 Kcnk1 NM_008430 3.47E-02 -1.54 Fam175a NM_172405 4.38E-02 -1.54 Slc6a20a NM_139142 3.08E-02 -1.54 Bxdc2 NM_026396 4.41E-05 -1.54 Safb ENSMUST00000095224 2.45E-02 -1.54 Nae1 NM_144931 2.51E-03 -1.54 Tanc2 NM_181071 2.92E-03 -1.54 LOC100039300 ENSMUST00000055309 5.57E-03 -1.54 Bcas2 NM_026602 3.23E-03 -1.54 Pop4 NM_025390 2.05E-02 -1.54 Myo6 NM_001039546 3.11E-02 -1.54 Terf1 NM_009352 6.99E-03 -1.54 Gnpnat1 NM_019425 5.76E-04 -1.54 Chka NM_013490 2.58E-03 -1.54 Ccdc131 NM_001033261 2.52E-03 -1.54 Slc33a1 NM_015728 2.70E-02 -1.54 Fv1 NM_010244 4.42E-02 -1.54 Zfp715 NM_027264 2.73E-02 -1.54 Prkci NM_008857 3.16E-02 -1.54 Zfp85-rs1 NM_001001130 2.18E-02 -1.55 Dpy30 NM_024428 1.58E-03 -1.55 Ptpn13 NM_011204 3.67E-02 -1.55 Ide NM_031156 2.44E-02 -1.55 1810020D17Rik BC026557 1.45E-04 -1.55 2310001H12Rik BC012405 2.87E-02 -1.55 Hmgb1 NM_010439 5.45E-03 -1.55 Calm4 NM_020036 2.69E-02 -1.55 Thumpd3 NM_008188 1.26E-03 -1.55

191

Gene Symbol RefSeq p-value Fold-Change (MAA/No bead) (MAA/No bead) Eml5 NM_001081191 4.80E-04 -1.56 Cetn3 NM_007684 1.86E-02 -1.56 1700034H14Rik BC034297 1.50E-04 -1.56 Pkp1 NM_019645 3.74E-02 -1.56 Polg2 NM_015810 4.85E-02 -1.56 2700007P21Rik NM_173750 1.43E-03 -1.56 ENSMUSG00000062319 ENSMUST00000076071 2.01E-02 -1.56 Fastkd2 NM_172422 4.20E-04 -1.57 Ppil4 NM_026141 2.51E-04 -1.57 Rdh14 NM_023697 2.64E-03 -1.57 D2Ertd750e NM_026412 4.03E-02 -1.57 Zfp677 NM_172486 4.64E-02 -1.57 2610044O15Rik NM_153780 2.07E-02 -1.57 Gas5 NR_002840 2.22E-02 -1.58 Irx5 NM_018826 1.03E-02 -1.58 Igbp1 NM_008784 2.60E-02 -1.58 Prim1 NM_008921 9.80E-03 -1.58 Sfrs6 NM_026499 9.26E-05 -1.58 Tas2r135 NM_199159 5.89E-04 -1.58 4930430F08Rik NM_175128 6.78E-03 -1.58 Stx19 NM_026588 4.80E-02 -1.58 Tcfap2a NM_011547 3.01E-02 -1.58 Gnpnat1 NM_019425 1.04E-03 -1.58 Rab38 NM_028238 2.88E-02 -1.58 4930503L19Rik BC057927 8.18E-03 -1.59 Aim1 NM_172393 1.16E-02 -1.59 Clca5 NM_178697 3.73E-02 -1.59 Mrfap1 NM_026242 5.49E-03 -1.60 Tmem56 NM_178936 3.90E-02 -1.60 2310039E09Rik NM_026509 3.35E-02 -1.60 Commd6 NM_001033132 1.03E-04 -1.60 Faim NM_001122851 2.00E-04 -1.61 Cmbl NM_181588 7.07E-03 -1.61 Gemin6 NM_026053 1.08E-03 -1.61 Wtap NM_001113533 1.33E-02 -1.61 Apoo-ps NR_004438 1.34E-03 -1.61 Myef2 NM_010852 5.87E-03 -1.62 Chek1 NM_007691 3.54E-02 -1.62 Lmod3 NM_001081157 4.02E-03 -1.62 C1galt1c1 NM_021550 2.00E-02 -1.62 Lztfl1 NM_033322 1.89E-03 -1.62 Taf9b NM_001001176 4.78E-02 -1.62 Klf12 NM_010636 1.57E-02 -1.62 Zbtb25 NM_028356 2.73E-02 -1.62 Znhit6 ENSMUST00000098534 2.63E-04 -1.63 Nasp NM_016777 6.03E-03 -1.63 Centd1 NM_178407 2.24E-02 -1.63

192

Gene Symbol RefSeq p-value Fold-Change (MAA/No bead) (MAA/No bead) Snora65 NR_002898 1.34E-03 -1.63 Myo1d NM_177390 4.52E-02 -1.63 Phxr4 BC107288 5.71E-04 -1.63 Naaladl2 XM_975226 1.58E-03 -1.63 Tcp1 AF357405 1.42E-02 -1.63 Itga6 NM_008397 3.43E-02 -1.64 Flrt3 NM_178382 4.93E-03 -1.64 Cdh1 NM_009864 4.40E-02 -1.64 Trp63 NM_001127259 2.94E-03 -1.65 Gpr111 NM_001033493 2.19E-02 -1.65 Akap8 NM_019774 2.79E-02 -1.65 1700029F09Rik BC119032 3.86E-04 -1.65 1700094D03Rik BC048560 1.46E-02 -1.66 Art3 NM_181728 1.70E-02 -1.66 EG214321 NM_001038995 2.91E-02 -1.67 Klf5 NM_009769 1.80E-02 -1.67 Tmem68 NM_028097 2.18E-02 -1.67 C130090K23Rik BC016523 3.40E-02 -1.67 Atp12a NM_138652 3.69E-03 -1.68 Rdh1 NM_080436 6.18E-04 -1.68 Rptn NM_009100 3.90E-02 -1.68 Krt5 NM_027011 1.46E-02 -1.68 BC002059 BC002059 2.49E-02 -1.68 4921533L14Rik BC070446 4.10E-02 -1.69 Elovl6 NM_130450 3.00E-02 -1.69 Tbrg3 BC095996 3.54E-03 -1.69 ENSMUSG00000075538 ENSMUST00000100426 2.26E-02 -1.69 Zfp825 NM_146231 2.09E-02 -1.70 Rras2 NM_025846 6.17E-03 -1.70 Col17a1 NM_007732 3.85E-02 -1.70 Nkiras1 NM_023526 4.70E-03 -1.70 Gemin6 NM_026053 2.76E-03 -1.70 Them5 NM_025416 4.05E-02 -1.71 Tnip3 NM_001001495 3.10E-02 -1.72 Tacstd1 NM_008532 4.41E-02 -1.72 Sgcg NM_011892 3.52E-02 -1.72 Sptlc3 NM_175467 3.09E-02 -1.72 Dsg1b NM_181682 2.77E-02 -1.73 Perp NM_022032 2.07E-02 -1.74 1700081L11Rik AK153679 3.91E-03 -1.74 Fzd6 NM_008056 2.45E-03 -1.74 Itga2 NM_008396 3.47E-02 -1.75 Ogn NM_008760 4.74E-02 -1.77 Tmod4 NM_016712 2.41E-02 -1.77 Trim59 NM_025863 2.15E-02 -1.77 Has3 NM_008217 4.54E-02 -1.77 9530009G21Rik AK156585 1.97E-03 -1.78

193

Gene Symbol RefSeq p-value Fold-Change (MAA/No bead) (MAA/No bead) ENSMUSG00000070892 ENSMUST00000094956 4.66E-04 -1.78 Car12 NM_178396 9.10E-03 -1.79 Cyp2b19 NM_007814 1.22E-02 -1.80 A630033H20Rik NM_175442 3.22E-02 -1.80 Ccdc68 NM_201362 1.65E-03 -1.80 Myf6 NM_008657 2.53E-03 -1.81 Mtcp1 NM_001039373 7.13E-03 -1.81 Cysltr1 NM_021476 2.36E-05 -1.81 Zcchc7 NM_138590 7.00E-03 -1.81 Rabggtb NM_011231 9.50E-05 -1.81 Tas2r126 NM_207028 1.48E-04 -1.82 C430014K11Rik ENSMUST00000092304 1.68E-03 -1.82 Zfp322a NM_001111107 2.02E-02 -1.83 A130010J15Rik ENSMUST00000016334 3.76E-02 -1.84 2700097O09Rik BC056967 2.51E-02 -1.85 Cdcp1 NM_133974 1.65E-02 -1.85 C130079G13Rik ENSMUST00000094227 1.72E-02 -1.85 Dsg3 NM_030596 4.59E-02 -1.87 Dsg1a NM_010079 1.78E-02 -1.88 Spink5 NM_001081180 2.17E-02 -1.90 Dsp NM_023842 5.43E-03 -1.91 Hrnr AY027660 1.44E-03 -1.92 Smpx NM_025357 2.77E-02 -1.92 Mphosph6 NM_026758 1.13E-04 -1.92 Rbm35a NM_194055 3.65E-02 -1.94 Hal NM_010401 1.40E-02 -1.95 Eno2 NM_013509 1.22E-02 -1.95 2310046A06Rik BC089626 4.44E-02 -1.95 Ttn ENSMUST00000099981 6.05E-03 -1.96 Pkia NM_008862 1.19E-02 -1.97 Bnc1 NM_007562 4.63E-02 -1.99 Rny1 NR_004419 3.00E-02 -2.00 Muc15 NM_172979 4.23E-02 -2.02 Rdh9 NM_153133 4.21E-03 -2.02 Gpr115 BC089564 2.76E-02 -2.05 ENSMUSG00000054945 ENSMUST00000068250 3.89E-02 -2.06 Atp13a3 NM_001128096 4.50E-02 -2.09 Lgals7 NM_008496 2.98E-03 -2.09 Ncl AF357416 3.74E-02 -2.13 Abca12 BC158063 2.65E-02 -2.13 Lor NM_008508 3.48E-03 -2.14 Dsc1 NM_013504 2.28E-02 -2.15 Mal2 NM_178920 2.60E-02 -2.16 Ifi205 NM_172648 1.45E-02 -2.17 Elovl7 NM_029001 1.25E-02 -2.18 Far2 NM_178797 4.43E-02 -2.18 C530008M07Rik BC120577 4.03E-02 -2.29

194

Gene Symbol RefSeq p-value Fold-Change (MAA/No bead) (MAA/No bead) Cxcl15 NM_011339 4.71E-02 -2.30 B230325K18Rik ENSMUST00000079045 5.54E-03 -2.39 Gpm6a NM_153581 1.70E-02 -2.40 Dsc3 NM_007882 2.91E-02 -2.47 Krt1 NM_008473 1.37E-02 -2.68 Aqp3 NM_016689 1.25E-02 -2.84 Abi3bp NM_001014423 5.23E-03 -2.86 Ptprz1 NM_001081306 2.11E-02 -3.01 Gsdmc1 NM_031378 3.96E-02 -3.02 Krt10 NM_010660 2.38E-02 -3.30

Table AI-S2. List of genes for Day 4, MAA/PMMA (FC > 1.5 and p-value < 0.05). Gene Symbol RefSeq p-value Fold-Change (MAA/PMMA) (MAA/PMMA) Genes with increased expression in MAA treated wounds EG668725 DQ386867 2.30E-02 1.76 3110040M04Rik AK014162 2.46E-02 1.74 Vmn2r74 NM_001105187 2.28E-02 1.72 100043482 XM_001480057 3.75E-02 1.71 Olfr121 NM_146629 5.22E-03 1.71 Olfr197 NM_146484 1.88E-02 1.66 EG432649 NR_003649 4.38E-02 1.65 V1rg5 NM_134206 2.62E-02 1.62 Olfr1199 NM_146458 8.35E-03 1.62 Olfr471 ENSMUST00000084758 1.42E-04 1.62 Olfr215 NM_146446 2.19E-02 1.61 Olfr591 NM_001011847 6.66E-03 1.60 Olfr671 NM_001011755 2.69E-02 1.59 Olfr1045 NM_147017 4.25E-02 1.57 Vmn2r105 NM_001104567 8.91E-03 1.57 Olfr60 NM_146955 1.62E-02 1.57 Svs3a NM_021363 2.39E-02 1.56 Olfr676 NM_147095 4.46E-02 1.55 Olfr1164 NM_146641 3.87E-02 1.54 Gpx6 NM_145451 5.06E-03 1.53 Vmn2r111 NM_001104573 4.13E-02 1.53 Sbpl NM_001077421 1.12E-03 1.53 Olfr893 NM_146336 3.35E-02 1.52 V1re3 NM_134192 4.57E-02 1.52 V1rd13 NM_206868 4.75E-02 1.51 Genes with decreased expression in MAA treated wounds Mtcp1 NM_001039373 1.36E-04 -1.50 Myom2 NM_008664 1.90E-02 -1.50 Aldh1a1 NM_013467 8.09E-03 -1.50

195

Gene Symbol RefSeq p-value Fold-Change (MAA/PMMA) (MAA/PMMA) Zcchc7 NM_138590 1.38E-04 -1.50 Gstk1 NM_029555 1.61E-02 -1.51 Gfpt2 NM_013529 5.47E-03 -1.51 Gtf3c6 NM_026113 4.51E-04 -1.51 Znhit6 ENSMUST00000098534 5.83E-05 -1.51 Dnajc12 NM_013888 2.04E-04 -1.51 C1galt1c1 NM_021550 7.74E-03 -1.51 Prg4 NM_021400 3.94E-02 -1.51 Ppp1r3a NM_080464 2.33E-02 -1.51 Zfp85-rs1 NM_001001130 1.68E-02 -1.51 Cpe NM_013494 4.24E-04 -1.51 Nkiras1 NM_023526 5.41E-04 -1.51 Rab12 NM_024448 5.56E-06 -1.51 Asb11 NM_026853 3.78E-02 -1.51 Rfc4 NM_145480 8.16E-04 -1.51 Commd6 NM_001033132 3.34E-05 -1.52 Serpinb10 NM_198028 2.85E-02 -1.52 Taf9b NM_001001176 2.34E-02 -1.52 Tspan8 NM_146010 2.75E-03 -1.52 2700097O09Rik BC056967 7.02E-04 -1.52 Bxdc2 NM_026396 3.46E-05 -1.52 Tnip3 NM_001001495 5.46E-03 -1.52 Mrpl50 NM_178603 1.14E-03 -1.52 Erh NM_007951 2.21E-04 -1.52 Nola3 NM_025403 5.42E-04 -1.52 Zfp322a NM_001111107 7.50E-04 -1.52 C430014K11Rik ENSMUST00000092304 2.82E-05 -1.52 Fkbp3 NM_013902 1.65E-04 -1.52 Gemin6 NM_026053 3.66E-04 -1.53 Asb12 NM_080858 3.40E-02 -1.53 A030007L17Rik BC080818 2.34E-02 -1.53 Gnpnat1 NM_019425 5.01E-04 -1.53 Sfrp2 NM_009144 1.44E-03 -1.53 Actr6 NM_025914 1.64E-03 -1.53 C130079G13Rik NM_177661 2.97E-02 -1.53 ENSMUSG00000054945 ENSMUST00000068250 2.92E-05 -1.54 Fhl1 NM_001077361 1.12E-02 -1.54 Calr4 NM_001033226 1.83E-02 -1.54 Rdh14 NM_023697 1.95E-03 -1.54 Gnpnat1 NM_019425 6.64E-04 -1.54 Cd209d NM_130904 2.30E-02 -1.54 Cenpq NM_031863 2.41E-03 -1.54 Igbp1 NM_008784 2.07E-02 -1.54 Trim13 NM_023233 2.04E-04 -1.55 Pkhd1l1 NM_138674 9.31E-03 -1.55 C1s NM_144938 3.91E-02 -1.55 Ccl12 NM_011331 2.19E-02 -1.55

196

Gene Symbol RefSeq p-value Fold-Change (MAA/PMMA) (MAA/PMMA) 1700029F09Rik BC119032 1.21E-04 -1.55 1110059G10Rik BC034133 1.65E-03 -1.55 Cspp1 NM_026493 4.46E-04 -1.55 Upk1b NM_178924 2.88E-02 -1.55 Dpp4 NM_010074 1.17E-04 -1.55 Lmod3 NM_001081157 2.27E-03 -1.56 Trappc2 NM_025432 5.32E-04 -1.56 Klhl6 NM_183390 4.66E-02 -1.56 Lztfl1 NM_033322 1.04E-03 -1.56 Nexn NM_199465 1.98E-02 -1.56 Bet1 NM_009748 1.68E-03 -1.56 Cd48 NM_007649 4.34E-03 -1.56 Dpt NM_019759 1.70E-02 -1.56 Rabggtb NM_011231 4.03E-06 -1.57 Trim63 NM_001039048 1.93E-02 -1.57 Marco NM_010766 3.63E-02 -1.57 Hrsp12 NM_008287 3.23E-03 -1.57 Asb5 NM_029569 4.58E-03 -1.58 Rad54l2 NM_030730 2.80E-03 -1.58 Trim59 NM_025863 5.14E-03 -1.58 Phc3 BC050266 1.65E-03 -1.58 2310046A06Rik BC089626 1.95E-03 -1.58 Epgn NM_053087 5.69E-04 -1.59 Eno2 NM_013509 3.43E-04 -1.59 Tmem100 NM_026433 1.20E-02 -1.59 Cmbl NM_181588 6.20E-03 -1.59 Zdhhc17 BC051527 3.52E-03 -1.60 Zfp825 NM_146231 1.05E-02 -1.60 Rnu2 NR_004414 4.69E-02 -1.60 Rnu2 NR_004414 4.69E-02 -1.60 Rnu2 NR_004414 4.69E-02 -1.60 Rnu2 NR_004414 4.69E-02 -1.60 Ptpn22 NM_008979 5.86E-03 -1.60 Art3 NM_181728 1.13E-02 -1.60 A630033H20Rik NM_175442 8.54E-03 -1.60 1700034H14Rik BC034297 2.47E-04 -1.61 1810020D17Rik BC026557 2.58E-04 -1.61 Sgcg NM_011892 1.90E-02 -1.62 Apoo-ps NR_004438 1.41E-03 -1.62 Ogn NM_008760 2.20E-02 -1.63 Efemp1 NM_146015 4.75E-03 -1.63 Myl1 NM_021285 4.10E-02 -1.63 Fabp3 NM_010174 2.54E-02 -1.64 Gemin6 NM_026053 1.32E-03 -1.64 Faim NM_001122851 2.70E-04 -1.64 Slitrk6 NM_175499 1.04E-02 -1.65 Rnu3a NR_002842 1.79E-02 -1.65

197

Gene Symbol RefSeq p-value Fold-Change (MAA/PMMA) (MAA/PMMA) 2810021G02Rik ENSMUST00000108923 4.59E-02 -1.66 Fabp3 NM_010174 2.93E-02 -1.66 Rnu2 NR_004414 3.02E-02 -1.67 Snora69 NR_002900 1.22E-02 -1.67 Mphosph6 NM_026758 7.42E-03 -1.69 Tmod4 NM_016712 1.56E-02 -1.69 Casq2 NM_009814 1.71E-02 -1.69 Atp13a3 NM_001128096 4.09E-03 -1.70 Mphosph6 NM_026758 1.44E-05 -1.70 Myot NM_001033621 1.66E-02 -1.71 Mmp1b NM_032007 3.05E-03 -1.73 Smpx NM_025357 1.04E-02 -1.73 Ankrd1 NM_013468 7.32E-04 -1.74 Pkia NM_008862 2.83E-03 -1.74 Cysltr1 NM_021476 1.68E-05 -1.77 Chi3l3 NM_009892 2.76E-02 -1.77 Gpm6a NM_153581 2.02E-04 -1.78 Myf6 NM_008657 2.47E-03 -1.80 Gm94 NM_001033280 3.75E-02 -1.87 Ivl NM_008412 3.76E-03 -1.88 Ifi205 NM_172648 6.62E-03 -1.97 Rny1 NR_004419 2.94E-02 -1.99 Akr1c18 NM_134066 3.37E-04 -2.22

Table AI-S3. List of genes for Day 4, MAA/PMMA (FC > 1.5 and p-value < 0.05). Gene Symbol RefSeq p-value Fold-Change (PMMA/No bead) (PMMA/No bead) Genes with increased expression in PMMA treated wounds I830127L07Rik ENSMUST00000100541 3.91E-02 2.21 AW112010 EF660528 2.65E-02 2.14 OTTMUSG00000005723 DQ508487 1.66E-02 1.96 EG240921 ENSMUST00000037976 2.37E-02 1.90 Ly6i NM_020498 4.92E-02 1.87 Ifit1 NM_008331 1.69E-02 1.82 Cxcl10 NM_021274 4.74E-02 1.82 Rtp4 NM_023386 1.23E-03 1.81 Rsad2 NM_021384 1.82E-02 1.73 Ifit2 NM_008332 3.53E-02 1.71 Irf7 NM_016850 6.95E-03 1.70 Ifit3 NM_010501 2.32E-02 1.68 Ms4a4c NM_029499 3.59E-02 1.67 Ifi47 NM_008330 3.29E-02 1.64 EG434460 ENSMUST00000087473 7.03E-03 1.64 Tgtp NM_011579 4.65E-02 1.64 Alas2 NM_009653 2.75E-02 1.63

198

Gene Symbol RefSeq p-value Fold-Change (PMMA/No bead) (PMMA/No bead) Pyhin1 NM_175026 1.85E-02 1.63 Oas3 NM_145226 1.95E-02 1.63 Slamf7 NM_144539 2.64E-02 1.62 Irgm1 NM_008326 3.93E-02 1.62 Igtp NM_018738 3.10E-02 1.62 H2-T24 NM_008207 1.08E-02 1.61 Mx1 NM_010846 2.53E-02 1.60 Isg15 NM_015783 2.11E-03 1.59 Aldh1a2 NM_009022 2.02E-02 1.58 Usp18 NM_011909 1.58E-02 1.58 Fcgr4 NM_144559 2.63E-02 1.57 Dio3 NM_172119 5.20E-03 1.56 Olfr1095 NM_146730 4.91E-02 1.55 5830443L24Rik NM_029509 3.50E-02 1.54 Psmb9 NM_013585 1.19E-04 1.52 Oasl1 NM_145209 1.47E-02 1.51 Psmb8 NM_010724 8.15E-03 1.51 OTTMUSG00000010673 BC085112 3.24E-03 1.51 Genes with decreased expression in PMMA treated wounds Ar NM_013476 3.59E-02 -1.50 2810047C21Rik BC071238 4.48E-02 -1.50 Hmcn1 NM_001024720 3.26E-02 -1.50 2810047C21Rik BC071238 4.15E-02 -1.50 Gsta2 NM_008182 3.43E-02 -1.50 Klf5 NM_009769 3.86E-03 -1.50 Itga2 NM_008396 3.29E-03 -1.51 Tns4 NM_172564 1.88E-02 -1.51 1810011O10Rik NM_026931 1.62E-02 -1.51 Hsn2 NM_001037155 3.27E-04 -1.51 Hpgd NM_008278 6.69E-03 -1.51 Centd1 NM_178407 8.63E-03 -1.51 Lamc2 NM_008485 2.13E-02 -1.51 Tbrg3 BC095996 5.01E-04 -1.51 Tas2r143 NM_001001452 1.27E-03 -1.52 Col17a1 NM_007732 8.89E-03 -1.52 Fgfbp1 NM_008009 3.85E-02 -1.52 Ccnj NM_172839 2.78E-02 -1.52 Rps6ka6 NM_025949 3.18E-03 -1.53 Rbpj AY512934 4.44E-02 -1.53 Serpinb11 NM_025867 4.77E-02 -1.53 Slc6a20a NM_139142 2.98E-02 -1.53 Galnt3 NM_015736 1.88E-02 -1.53 EG214321 NM_001038995 1.06E-02 -1.53 Fbn1 NM_007993 4.97E-02 -1.54 Bcl2l15 BC117780 4.04E-03 -1.54 Dpp4 NM_010074 1.32E-04 -1.54 ENSMUSG00000062319 ENSMUST00000076071 1.76E-02 -1.54

199

Gene Symbol RefSeq p-value Fold-Change (PMMA/No bead) (PMMA/No bead) Cdh19 NM_001081386 8.63E-03 -1.55 Serpinb6c NM_148942 3.37E-02 -1.55 C130090K23Rik BC016523 1.49E-02 -1.55 Il33 NM_133775 4.38E-02 -1.56 Krt14 NM_016958 3.54E-02 -1.56 Clca5 NM_178697 3.01E-02 -1.56 Nrarp NM_025980 4.76E-02 -1.56 Il1f8 NM_027163 2.64E-02 -1.56 Trp63 NM_001127259 1.30E-03 -1.56 Il24 NM_053095 4.98E-02 -1.56 Ccdc68 NM_201362 1.04E-04 -1.56 Fzd6 NM_008056 3.89E-04 -1.56 Mboat2 NM_026037 8.82E-03 -1.56 Lama3 ENSMUST00000092070 1.68E-02 -1.56 Ly6d NM_010742 4.04E-02 -1.56 A130010J15Rik ENSMUST00000016334 4.36E-03 -1.57 4833403I15Rik BC113158 3.37E-03 -1.57 Grhl1 NM_145890 4.75E-02 -1.57 Bnc1 NM_007562 1.05E-03 -1.57 Hook1 NM_030014 9.19E-03 -1.58 Moxd1 NM_021509 5.43E-03 -1.58 Sprr2a ENSMUST00000090872 3.36E-02 -1.58 Sprr2a ENSMUST00000090872 3.36E-02 -1.58 Tmem56 NM_178936 3.53E-02 -1.58 Tas2r135 NM_199159 5.80E-04 -1.58 2310038E17Rik BC117742 2.78E-02 -1.58 Abca8a NM_153145 4.00E-02 -1.59 Gsta1 NM_008181 3.57E-02 -1.59 Col14a1 NM_181277 4.64E-02 -1.59 AU042651 BC107250 1.66E-02 -1.60 Lphn2 NM_001081298 1.41E-02 -1.60 ENSMUSG00000070892 ENSMUST00000094956 7.21E-05 -1.60 Car12 NM_178396 1.95E-03 -1.60 C030002C11Rik BC058715 3.38E-03 -1.60 Ide NM_031156 3.31E-02 -1.61 Nov NM_010930 4.69E-02 -1.61 1700081L11Rik AK153679 1.30E-03 -1.61 C130079G13Rik ENSMUST00000094227 2.59E-03 -1.61 Gsta1 NM_008181 3.53E-02 -1.61 Sprr2a NM_011468 1.41E-02 -1.61 Gpr111 NM_001033493 1.82E-02 -1.62 Scn2a1 NM_001099298 6.13E-03 -1.62 Cyp2b19 NM_007814 3.11E-03 -1.62 Polr3g NM_001081176 1.61E-02 -1.62 Syne2 NM_001005510 1.45E-03 -1.62 Serpinb12 NM_027971 4.70E-02 -1.62 Rptn NM_009100 3.14E-02 -1.63

200

Gene Symbol RefSeq p-value Fold-Change (PMMA/No bead) (PMMA/No bead) Elovl4 NM_148941 4.39E-02 -1.64 Scn2a1 NM_001099298 3.55E-02 -1.64 Calm4 NM_020036 4.10E-02 -1.64 Gpr87 NM_032399 4.35E-03 -1.65 Krt5 NM_027011 1.20E-02 -1.65 A130014H13Rik AK079474 1.79E-02 -1.65 Cdcp1 NM_133974 4.62E-03 -1.66 Scel NM_022886 2.67E-02 -1.66 Rdh1 NM_080436 5.50E-04 -1.66 Clca2 NM_030601 4.90E-02 -1.67 Efemp1 NM_146015 3.52E-03 -1.67 9530008L14Rik BC027755 1.26E-02 -1.68 C130079G13Rik NM_177661 1.20E-02 -1.68 Ncl AF357416 1.59E-03 -1.68 Phxr4 BC107288 8.57E-04 -1.68 Atp12a NM_138652 4.02E-03 -1.69 Prrg4 NM_178695 1.36E-02 -1.69 A030007L17Rik BC080818 7.85E-03 -1.70 Aadac NM_023383 3.60E-02 -1.70 Muc15 NM_172979 7.23E-03 -1.70 2210023G05Rik BC027185 2.11E-02 -1.71 Hal NM_010401 3.24E-03 -1.72 Elovl6 NM_130450 3.42E-02 -1.72 Serpinb7 NM_027548 2.68E-03 -1.72 Sptlc3 NM_175467 3.03E-02 -1.72 Elovl3 NM_007703 4.98E-02 -1.72 Epgn NM_053087 1.68E-04 -1.73 Lipk NM_172837 4.40E-02 -1.73 Perp NM_022032 2.11E-02 -1.74 Tas2r126 NM_207028 8.89E-05 -1.76 9530009G21Rik AK156585 2.27E-03 -1.80 Rbm35a NM_194055 2.23E-02 -1.81 Gpr115 BC089564 1.00E-02 -1.83 Lgals7 NM_008496 5.52E-04 -1.83 Ivl NM_008412 5.01E-03 -1.83 Tmprss11f NM_178730 8.23E-03 -1.83 Rdh9 NM_153133 1.47E-03 -1.83 Lipm NM_023903 3.24E-02 -1.84 Spink5 NM_001081180 1.78E-02 -1.85 2610528A11Rik AK012157 1.08E-02 -1.87 B230325K18Rik ENSMUST00000079045 2.21E-04 -1.87 Akr1c18 NM_134066 1.35E-03 -1.92 Serpinb5 NM_009257 6.55E-03 -1.92 Elovl7 NM_029001 4.52E-03 -1.94 Ttn ENSMUST00000099981 6.57E-03 -1.98 Cxcl15 NM_011339 1.77E-02 -1.99 Mal2 NM_178920 1.73E-02 -2.03

201

Gene Symbol RefSeq p-value Fold-Change (PMMA/No bead) (PMMA/No bead) 4833413O15Rik BC125450 1.29E-02 -2.04 Bpil2 NM_177772 4.19E-02 -2.08 Ace2 NM_027286 1.23E-02 -2.09 Aadacl2 NM_001128091 1.50E-02 -2.09 Abi3bp NM_001014423 1.59E-04 -2.11 C530008M07Rik BC120577 2.54E-02 -2.12 4732474O15Rik ENSMUST00000048880 7.70E-03 -2.12 Dsp NM_023842 1.25E-02 -2.17 Dsg1a NM_010079 3.97E-02 -2.21 Far2 NM_178797 4.75E-02 -2.21 Slitrk6 NM_175499 5.43E-04 -2.25 RP23-24J10.7 NM_181989 2.35E-02 -2.26 Hrnr AY027660 4.55E-03 -2.27 Flg2 NM_001013804 1.13E-02 -2.30 Abca12 BC158063 3.85E-02 -2.31 Lor NM_008508 5.81E-03 -2.32 Dsc3 NM_007882 2.05E-02 -2.32 Ptprz1 NM_001081306 2.97E-03 -2.32 Dsc1 NM_013504 3.36E-02 -2.33 Gm94 NM_001033280 9.01E-03 -2.34 Aqp3 NM_016689 3.31E-03 -2.36 Gsdmc1 NM_031378 9.84E-03 -2.40 4833423E24Rik BC120879 2.63E-02 -2.46 Krt1 NM_008473 2.32E-02 -3.06 Krt10 NM_010660 2.12E-02 -3.20

Table AI-S4. List of genes for Day 7, MAA/No bead (FC > 1.5 and p-value < 0.05). Gene Symbol RefSeq p-value Fold-Change (MAA/No bead) (MAA/No bead) Genes with increased expression in MAA treated wounds Ttn ENSMUST00000099980 4.66E-02 3.20 1110028A07Rik NM_026808 3.93E-02 3.18 2810021G02Rik ENSMUST00000108923 4.20E-02 2.59 EG625558 AK136154 3.68E-02 2.54 BC023105 BC023105 3.52E-02 2.52 Kctd4 NM_026214 1.70E-02 2.28 Mylk4 ENSMUST00000057428 3.47E-02 2.03 Ogn NM_008760 2.97E-02 2.03 Hmcn1 NM_001024720 7.17E-03 1.96 Jph1 NM_020604 2.87E-02 1.91 Zfp51 NM_009558 2.98E-02 1.89 Sema3a NM_009152 4.13E-02 1.86 Hmcn1 NM_001024720 4.84E-02 1.84

202

Gene Symbol RefSeq p-value Fold-Change (MAA/No bead) (MAA/No bead) Hmcn1 NM_001024720 4.37E-02 1.81 Ecm2 NM_001012324 3.41E-02 1.80 OTTMUSG00000023442 XM_001480612 4.28E-02 1.78 Angptl1 NM_028333 3.64E-02 1.77 2310001H12Rik BC012405 4.72E-03 1.76 Zfp85-rs1 NM_001001130 3.57E-02 1.76 ENSMUSG00000073005 ENSMUST00000101289 4.85E-02 1.76 Hmcn1 NM_001024720 1.95E-02 1.75 Gm889 BC147386 2.21E-02 1.75 Hmcn1 NM_001024720 4.75E-02 1.74 LOC635086 XM_910030 1.81E-03 1.67 Ankrd32 NM_134071 2.76E-02 1.67 Apod NM_007470 4.40E-02 1.66 Ereg NM_007950 4.28E-02 1.66 EG632964 XM_989174 3.17E-02 1.65 Phf20 BC060121 2.80E-02 1.65 Mrpl47 NM_029017 2.54E-02 1.64 Fbxl7 BC050864 4.65E-02 1.63 Zfp455 NM_001048204 9.55E-03 1.63 A2bp1 NM_021477 4.60E-02 1.63 Prss35 NM_178738 6.27E-03 1.60 Rnu35a NR_000003 2.53E-02 1.60 Rps6ka6 NM_025949 2.48E-02 1.59 Hmcn1 NM_001024720 5.45E-03 1.59 Nrn1 NM_153529 2.17E-02 1.59 Olfr624 NM_001011865 3.43E-02 1.57 Zfp322a NM_001111107 8.66E-03 1.57 100043775 XM_001480931 3.46E-02 1.57 LOC280487 X16670 3.46E-02 1.57 Fcer1a NM_010184 1.10E-02 1.56 Aspn NM_025711 5.24E-03 1.55 Lztfl1 NM_033322 4.93E-02 1.55 Cabc1 NM_023341 4.29E-02 1.54 EG639396 BC147148 3.26E-02 1.54 B930036N10Rik AK047216 3.30E-02 1.53 Zfp442 BC023805 4.93E-02 1.53 Kbtbd8 NM_001008785 8.58E-03 1.53 Zfp617 NM_133358 8.06E-03 1.53 Dnajb4 NM_025926 2.26E-02 1.53 Trim13 NM_023233 1.64E-02 1.52 OTTMUSG00000013370 XR_033267 2.05E-02 1.52 Arhgap5 NM_009706 1.16E-02 1.52 Krt78 NM_212487 3.77E-02 1.51 Nt5c3 NM_026004 3.63E-02 1.51 Genes with decreased expression in MAA treated wounds 3830408D24Rik NM_027511 4.40E-02 -1.50 4933407I18Rik ENSMUST00000086982 4.09E-02 -1.50

203

Gene Symbol RefSeq p-value Fold-Change (MAA/No bead) (MAA/No bead) Capn9 NM_023709 9.01E-03 -1.50 Chrdl2 NM_133709 3.63E-02 -1.50 Pramel7 NM_178250 9.25E-04 -1.51 Snai3 NM_013914 3.71E-02 -1.52 4732444A12Rik ENSMUST00000066154 4.62E-02 -1.52 Gnb1l NM_023120 4.32E-02 -1.52 Olfr544 NM_020289 1.15E-02 -1.53 LOC100045676 XM_001474727 3.17E-02 -1.53 Chst13 XM_978459 9.47E-03 -1.53 ENSMUSG00000064032 ENSMUST00000081739 2.53E-02 -1.54 1700093J21Rik ENSMUST00000023632 2.18E-02 -1.54 Gcm1 NM_008103 3.95E-02 -1.55 Reep6 NM_139292 2.81E-02 -1.55 Acoxl NM_028765 3.80E-02 -1.56 Slc16a5 NM_001080934 2.23E-02 -1.56 Oxct2a NM_022033 3.61E-02 -1.57 Lrrc26 NM_146117 1.71E-02 -1.57 Ntng1 ENSMUST00000051253 2.98E-02 -1.58 Pzp NM_007376 8.42E-03 -1.58 EG544710 ENSMUST00000092369 2.12E-02 -1.58 1600016N20Rik BC119093 3.55E-02 -1.58 Gm414 NM_001018031 4.83E-02 -1.59 Olfr1349 NM_207136 1.78E-02 -1.60 2610318N02Rik BC039993 2.20E-02 -1.60 Olfr1337 NM_146309 3.70E-02 -1.61 ENSMUSG00000074381 ENSMUST00000098818 3.60E-02 -1.62 Gucy2d ENSMUST00000098274 4.39E-02 -1.62 Slc25a42 NM_001007570 4.97E-02 -1.63 Tmem37 NM_019432 2.44E-02 -1.64 Neurog1 NM_010896 1.71E-02 -1.66 6720416L17Rik ENSMUST00000100000 4.64E-02 -1.67 P2rx1 NM_008771 4.96E-05 -1.67 EG266459 NR_003647 8.73E-03 -1.68 Lmtk3 NM_001005511 3.88E-02 -1.69 Doc2a NM_010069 3.10E-02 -1.76 6430598A04Rik NM_175521 3.29E-02 -1.81 Gkn1 NM_025466 3.68E-02 -1.81 D930028M14Rik ENSMUST00000098678 1.09E-02 -1.87 EG624855 NM_001037922 2.49E-02 -1.90 Trav7d-2 ENSMUST00000103637 4.81E-02 -2.00 Smoc1 NM_022316 2.38E-02 -2.54

204

Table AI-S5. List of genes for Day 7, MAA/PMMA (FC > 1.5 and p-value < 0.05). Gene Symbol RefSeq p-value Fold-Change (MAA/PMMA) (MAA/PMMA) Genes with increased expression in MAA treated wounds 1110028A07Rik NM_026808 1.26E-03 2.12 2310002L09Rik BC115483 1.18E-02 1.68 8030451F13Rik NM_175418 4.05E-02 2.86 A2bp1 NM_021477 2.97E-02 1.55 Actn2 NM_033268 4.04E-02 2.64 Actn3 NM_013456 4.92E-02 2.35 Ampd1 NM_001033303 3.00E-02 2.29 Ankrd2 NM_020033 3.45E-02 1.68 Ankrd23 NM_153502 2.81E-02 2.17 Ano5 NM_177694 2.91E-02 2.13 Art1 NM_009710 2.12E-02 1.66 Asb12 NM_080858 2.34E-02 1.98 Asb14 NM_080856 2.24E-02 1.74 Bex1 NM_009052 3.47E-02 2.48 Cacna1s NM_001081023 2.73E-02 1.62 Cacng1 NM_007582 3.89E-02 1.79 Cap2 NM_026056 3.30E-02 2.06 Car3 NM_007606 4.52E-02 1.56 Chrna1 NM_007389 3.64E-02 1.62 Cox8b NM_007751 1.53E-02 1.84 Cryab NM_009964 2.73E-02 1.54 Des NM_010043 2.82E-02 2.24 Dmd NM_007868 4.95E-02 1.79 Genes with increased expression in MAA treated wounds Dppa5a NM_025274 8.40E-03 -1.52 Eda2r NM_175540 2.27E-02 1.51 EG266459 NR_003647 8.19E-03 -1.67 EG435337 NM_001013824 2.19E-02 -1.70 EG625558 AK136154 1.59E-02 2.18 EG628696 XR_004960 8.48E-03 -1.60 ENSMUSG00000074792 AY344585 3.79E-02 1.55 Fhl1 NM_001077361 3.88E-02 2.05 Fsd2 NM_172904 4.33E-03 1.67 Hspb6 NM_001012401 1.50E-02 2.19 Hspb8 NM_030704 4.14E-02 1.56 Jph1 AK081751 1.40E-02 2.32 Jph1 NM_020604 9.65E-03 1.69 Klhl31 NM_172925 4.99E-02 2.33 Klhl31 NM_172925 4.03E-02 1.71 Lcn2 NM_008491 3.31E-02 -1.66 Lrrc39 NM_175413 2.91E-03 1.61 Mb NM_013593 2.56E-02 2.34 Mef2c NM_025282 1.60E-02 1.82 Mmp3 NM_010809 1.37E-02 -1.64

205

Gene Symbol RefSeq p-value Fold-Change (MAA/PMMA) (MAA/PMMA) Mybpc2 NM_178067 2.04E-02 1.97 Myh4 NM_010855 3.62E-02 2.70 Myl3 NM_010859 3.75E-02 1.50 Mylk2 NM_001081044 2.45E-02 1.51 Mylk4 ENSMUST00000057428 1.48E-02 1.81 Myo18b XM_912851 2.81E-02 1.80 Myo18b XM_912851 4.33E-02 1.56 Myot NM_001033621 4.89E-02 2.72 Myoz1 NM_021508 3.20E-02 2.18 Nexn NM_199465 4.05E-02 1.69 Nrap NM_008733 4.12E-02 2.05 Obox6 NM_145710 3.63E-02 -1.71 Olfr624 NM_001011865 2.31E-02 1.51 OTTMUSG00000010956 XR_034527 2.55E-02 -1.56 Pfkm NM_021514 1.49E-02 2.00 Pfn2 NM_019410 1.68E-03 1.50 Pkia NM_008862 4.30E-02 1.89 Popdc3 NM_024286 1.67E-02 1.51 Ppp1r3a NM_080464 7.25E-03 2.61 Prkaa2 NM_178143 3.38E-02 1.61 Prr8 NM_028234 2.51E-02 1.50 Pygm NM_011224 4.40E-02 2.38 Rbm24 NM_001081425 3.13E-02 1.51 Ryr1 NM_009109 3.97E-02 1.64 Saa3 NM_011315 3.30E-02 -1.50 Scn2a1 ENSMUST00000112399 1.75E-02 1.66 Scn2a1 NM_001099298 7.35E-03 1.54 Smyd1 NM_009762 3.74E-02 1.96 Srd5a2l2 NM_153801 4.23E-02 1.75 Tmem182 NM_001081198 1.71E-02 1.71 Tmod1 NM_021883 2.85E-02 1.64 Tomm22 NM_172609 1.71E-03 1.53 Trim72 NM_001079932 4.45E-02 1.67 Ttn ENSMUST00000099980 7.29E-03 2.34 Txlnb NM_138628 1.07E-02 1.78

Table AI-S6. List of genes for Day 7, PMMA/No bead (FC > 1.5 and p-value < 0.05). Gene Symbol RefSeq p-value Fold-Change (PMMA/No bead) (PMMA/No bead) Genes with increased expression in PMMA treated wounds Gpm6a NM_153581 4.82E-02 2.04 2810021G02Rik ENSMUST00000108923 7.17E-03 2.02 Kctd4 NM_026214 6.42E-03 1.99 Ppp1r3a NM_080464 3.54E-02 1.92 Pkia NM_008862 4.66E-02 1.87

206

Gene Symbol RefSeq p-value Fold-Change (PMMA/No bead) (PMMA/No bead) Asb12 NM_080858 3.49E-02 1.84 Spink12 NM_030061 4.74E-02 1.83 Zfp51 NM_009558 2.06E-02 1.79 Ogn NM_008760 1.10E-02 1.79 Lce1i NM_029667 3.78E-02 1.79 BC023105 BC023105 3.97E-04 1.77 Ifit1 NM_008331 2.53E-02 1.76 Prss35 NM_178738 1.18E-02 1.75 Hmcn1 NM_001024720 5.35E-03 1.72 OTTMUSG00000023442 XM_001480612 3.49E-02 1.72 Hist1h2ab NM_175660 2.44E-02 1.72 Klk5 NM_026806 3.71E-02 1.68 Lce1f NM_026394 4.33E-02 1.68 Akr1c14 NM_134072 4.13E-02 1.68 Phf20 BC060121 3.19E-02 1.68 Hmcn1 NM_001024720 8.84E-04 1.65 Apod NM_007470 4.08E-02 1.65 Mx2 NR_003508 1.93E-02 1.64 LOC635086 XM_910030 1.48E-03 1.64 Slitrk6 NM_175499 1.93E-02 1.63 ENSMUSG00000073005 ENSMUST00000101289 2.64E-02 1.63 Itgbl1 NM_145467 2.07E-03 1.62 Tmprss11d NM_145561 2.01E-02 1.62 V1ra7 NM_053222 3.08E-02 1.61 2310002L09Rik BC115483 1.74E-02 1.61 Hmcn1 NM_001024720 3.54E-02 1.60 I830012O16Rik NM_001005858 3.35E-02 1.60 Myo9a ENSMUST00000085572 2.41E-02 1.60 Omd NM_012050 1.52E-02 1.60 Zfp712 ENSMUST00000071628 2.81E-02 1.60 Scn2a1 NM_001099298 5.81E-03 1.58 Zfp455 NM_001048204 6.36E-03 1.57 Rps6ka6 NM_025949 2.16E-02 1.57 Hmcn1 NM_001024720 9.25E-03 1.57 Zfp617 NM_133358 9.98E-03 1.56 Ear1 NM_007894 2.06E-02 1.56 Hmcn1 NM_001024720 7.55E-03 1.56 EG668536 XR_030503 2.11E-02 1.56 Ecm2 NM_001012324 6.24E-03 1.55 2810055G20Rik ENSMUST00000068704 4.74E-02 1.55 Scn2a1 NM_001099298 2.79E-02 1.54 Ereg NM_007950 2.11E-02 1.54 Sema3a NM_009152 3.56E-03 1.54 2310001H12Rik BC012405 6.75E-04 1.54 Hmcn1 NM_001024720 4.03E-02 1.53 Scn2a1 ENSMUST00000112399 3.43E-02 1.53 OTTMUSG00000005148 BC147254 3.69E-02 1.53

207

Gene Symbol RefSeq p-value Fold-Change (PMMA/No bead) (PMMA/No bead) Klhl4 NM_172781 8.40E-03 1.53 Rnu35a NR_000003 1.60E-02 1.52 Hmcn1 NM_001024720 1.13E-02 1.52 Lce1b NM_026822 4.59E-02 1.52 Lztfl1 NM_033322 4.03E-02 1.51 Tomm22 NM_172609 1.98E-03 1.51 2810055G20Rik AK148800 4.71E-03 1.51 OTTMUSG00000013370 XR_033267 1.95E-02 1.51 Krt78 NM_212487 3.65E-02 1.51 666761 XR_032444 2.88E-02 1.50 Fcer1a NM_010184 6.94E-03 1.50 ENSMUSG00000074792 AY344585 4.96E-02 1.50 Hmcn1 NM_001024720 2.36E-03 1.50 Genes with increased expression in PMMA treated wounds Crlf1 NM_018827 1.81E-02 -1.50 ENSMUSG00000075578 ENSMUST00000100504 3.80E-02 -1.50 Mmd NM_026178 4.99E-02 -1.50 ENSMUSG00000074381 ENSMUST00000098818 1.63E-02 -1.50 Olfr1337 NM_146309 1.86E-02 -1.51 Rhcg NM_019799 2.65E-02 -1.51 EG432982 BC147350 1.31E-02 -1.51 EG240549 BC089619 4.53E-02 -1.52 6430598A04Rik NM_175521 3.07E-03 -1.52 Paqr9 NM_198414 3.89E-02 -1.53 Slc22a3 NM_011395 1.33E-02 -1.53 Oxtr NM_001081147 2.93E-02 -1.53 Kcng4 NM_025734 1.08E-02 -1.54 BC022713 BC022713 4.85E-02 -1.54 Doc2a NM_010069 7.29E-03 -1.54 EG544710 ENSMUST00000092369 1.67E-02 -1.54 Nkx2-9 NM_008701 4.83E-02 -1.54 Olfr1499 NM_146796 3.63E-03 -1.55 Foxi2 NM_183193 1.10E-02 -1.55 D930028M14Rik ENSMUST00000098678 8.36E-04 -1.56 EG626058 ENSMUST00000097535 1.03E-02 -1.56 EG668662 AK132630 8.21E-03 -1.56 EG623356 ENSMUST00000097807 3.51E-02 -1.56 Olfr328 BC110465 1.84E-02 -1.56 BC061194 BC049634 3.96E-02 -1.57 Ppyr1 NM_008919 1.15E-02 -1.58 G0s2 NM_008059 1.25E-02 -1.58 Onecut3 NM_139226 2.20E-02 -1.59 Trav2 ENSMUST00000103569 1.99E-02 -1.60 EG624855 NM_001037922 3.74E-03 -1.61 Sp9 NM_001005343 1.61E-02 -1.61 Lcn2 NM_008491 4.01E-02 -1.62 EG435337 NM_001013824 3.24E-02 -1.62

208

Gene Symbol RefSeq p-value Fold-Change (PMMA/No bead) (PMMA/No bead) F830116E18Rik NM_001033981 2.56E-02 -1.63 Trav7d-2 ENSMUST00000103637 6.51E-03 -1.65 Sh2d1b1 NM_012009 4.39E-02 -1.66 Cfd NM_013459 4.56E-02 -1.69 S3-12 NM_020568 2.18E-02 -1.75 Aqp7 NM_007473 3.82E-02 -1.77 Retn NM_022984 3.80E-02 -1.79 Slc36a2 NM_153170 1.62E-02 -1.81 Klb NM_031180 1.74E-02 -1.81 Smoc1 NM_022316 5.26E-04 -1.82 Mosc1 NM_001081361 2.02E-02 -1.89 Cdkl4 NM_001033443 3.33E-02 -1.92 Plin NM_175640 3.28E-02 -2.04 Mrap NM_029844 2.07E-02 -2.33 Cxcl13 NM_018866 2.33E-02 -2.35 Cidec NM_178373 4.77E-02 -2.64 Hpx NM_017371 2.64E-02 -3.14

Table AI-S7. Gene lists corresponding to Day 4 (Fig AI-3, left panel) Venn diagram sections. Lists were generated by GeneVenn (http://genevenn.sourceforge.net/index.htm). M/NB ∩ M/P ∩ M/NB ∩ M/P M/NB ∩ P/NB M/P ∩ P/NB M/NB only M/P only P/NB only P/NB (67) (87) (8) (125) (62) (83) (1)

C130079G13Rik 1110059G10Rik 1700081L11Rik A030007L17Rik 1700094D03Rik 100043482 1810011O10Rik

1700029F09Rik 9530009G21Rik Akr1c18 2310001H12Rik 2810021G02Rik 2210023G05Rik

1700034H14Rik A130010J15Rik Dpp4 2310003F16Rik 3110040M04Rik 2310038E17Rik

1810020D17Rik A130014H13Rik Efemp1 2310035C23Rik Actr6 2610528A11Rik

2310046A06Rik AW112010 Epgn 2310039E09Rik Aldh1a1 2810047C21Rik

2700097O09Rik Abca12 Gm94 2410129H14Rik Asb12 2810047C21Rik

A630033H20Rik Abi3bp Ivl 2610039C10Rik Asb5 4732474O15Rik

Ankrd1 Aldh1a2 Slitrk6 2610044O15Rik C1s 4833403I15Rik

Apoo-ps Aqp3 2610203C20Rik Calr4 4833413O15Rik

Art3 Atp12a 2700007P21Rik Casq2 4833423E24Rik

Asb11 B230325K18Rik 4921533L14Rik Ccl12 5830443L24Rik

Atp13a3 Bnc1 4930422I07Rik Cd209d 9530008L14Rik

Bet1 C130090K23Rik 4930430F08Rik Cd48 AU042651

Bxdc2 C530008M07Rik 4930503L19Rik Cenpq Aadac

C1galt1c1 Calm4 4930529M08Rik Chi3l3 Aadacl2

C430014K11Rik Car12 AU021034 Cpe Abca8a

Cmbl Ccdc68 Aard Cspp1 Ace2

Commd6 Cdcp1 Acox2 Dpt Alas2

Cysltr1 Centd1 Aim1 EG432649 Ar

Dnajc12 Clca5 Akap8 Erh Bcl2l15

EG668725 Col17a1 Arpp19 Fabp3 Bpil2

ENSMUSG000000 Cxcl10 BC002059 Fabp3 C030002C11Rik

54945

Eno2 Cxcl15 Bcas2 Fhl1 Ccnj

Faim Cyp2b19 Ccdc131 Fkbp3 Cdh19

Gemin6 Dio3 Ccdc5 Gfpt2 Clca2

Gnpnat1 Dsc1 Cd52 Gstk1 Col14a1

Gpm6a Dsc3 Cdh1 Gtf3c6 EG240921 209

M/NB ∩ M/P ∩ M/NB ∩ M/P M/NB ∩ P/NB M/P ∩ P/NB M/NB only M/P only P/NB only P/NB (67) (87) (8) (125) (62) (83) (1)

Gpx6 Dsg1a Cetn3 Klhl6 Elovl4

Hrsp12 Dsp Chac2 Marco Fbn1

Ifi205 EG214321 Chek1 Mmp1b Fgfbp1

Igbp1 EG434460 Chka Mrpl50 Flg2

Lmod3 ENSMUSG000000 Clcc1 Myl1 Galnt3

62319

Lztfl1 ENSMUSG000000 Crebzf Myom2 Gpr87

70892

Mphosph6 Elovl3 Csmd3 Myot Gsta1

Mtcp1 Elovl6 D2Ertd750e Nexn Gsta1

Myf6 Elovl7 Dennd2c Nola3 Gsta2

Nkiras1 Far2 Dnajb4 Olfr1164 Hmcn1

Ogn Fcgr4 Dpy30 Olfr671 Hook1

Olfr1045 Fzd6 Dsg1b Olfr676 Hpgd

Olfr1199 Gpr111 Dsg3 Phc3 Hsn2

Olfr121 Gpr115 EG237300 Pkhd1l1 I830127L07Rik

Olfr197 Grhl1 ENSMUSG000000 Ppp1r3a Ifi47

55419

Olfr215 Gsdmc1 ENSMUSG000000 Prg4 Il1f8

66331

Olfr471 H2-T24 ENSMUSG000000 Ptpn22 Il24

75538

Olfr591 Hal Eif2c3 Rab12 Il33

Olfr60 Hrnr Eml5 Rad54l2 Irgm1

Olfr893 Ide Exosc8 Rfc4 Krt14

Pkia Ifit1 Fam175a Rnu3a Lama3

Rabggtb Ifit2 Fastkd2 Serpinb10 Lamc2

Rdh14 Ifit3 Flrt3 Sfrp2 Lipk

Rnu2 Igtp Fusip1 Snora69 Lipm

Rny1 Irf7 Fv1 Svs3a Lphn2

Sbpl Isg15 Gas5 Tmem100 Ly6d 210

Sgcg Itga2 Gins1 Trappc2 Mboat2

M/NB ∩ M/P ∩ M/NB ∩ M/P M/NB ∩ P/NB M/P ∩ P/NB M/NB only M/P only P/NB only P/NB (67) (87) (8) (125) (62) (83) (1)

Smpx Klf5 Gjb6 Trim13 Moxd1

Taf9b Krt1 Gm1758 Trim63 Ms4a4c

Tmod4 Krt10 Guf1 Tspan8 Mx1

Tnip3 Krt5 Has3 V1rd13 Nov

Trim59 Lgals7 Hmgb1 V1re3 Nrarp

Upk1b Lor Irx5 Vmn2r111 OTTMUSG000000

10673

V1rg5 Ly6i Itga6 Vmn2r74 Oas3

Vmn2r105 Mal2 Kcnk1 Zdhhc17 Oasl1

Zcchc7 Muc15 Klf12 Olfr1095

Zfp322a Ncl Klk14 Polr3g

Zfp825 OTTMUSG000000 Kpna2 Prrg4

05723

Zfp85-rs1 Perp Kpna2 Pyhin1

Znhit6 Phxr4 LOC100039300 RP23-24J10.7

Psmb8 LOC100134980 Rbpj

Psmb9 LOC626711 Rps6ka6

Ptprz1 Lonrf1 Rsad2

Rbm35a Med18 Scel

Rdh1 Mrfap1 Scn2a1

Rdh9 Myef2 Scn2a1

Rptn Myo1d Serpinb11

Rtp4 Myo6 Serpinb12

Slc6a20a Naaladl2 Serpinb5

Spink5 Nae1 Serpinb6c

Sprr2a Nasp Serpinb7

Sptlc3 Nudcd2 Slamf7

Syne2 Nufip1 Tas2r143

Tas2r126 Odf3l2 Tmprss11f

Tas2r135 Olfr1489 Tns4 211 Tbrg3 Olfr684 Usp18

M/NB ∩ M/P ∩ M/NB ∩ M/P M/NB ∩ P/NB M/P ∩ P/NB M/NB only M/P only P/NB only P/NB (67) (87) (8) (125) (62) (83) (1)

Tgtp Pcdhb17

Tmem56 Pkp1

Trp63 Polg2

Ttn Pop4

Ppil4

Prim1

Prkci

Prss21

Ptpn13

Rab38

Rars2

Rbks

Rfesd

Rhox11

Rras2

Safb

Sfrs6

Slc22a16

Slc33a1

Snora65

Stx19

Tacstd1

Tanc2

Taok1

Tcfap2a

Tcp1

Terf1

Tfam

Them5

Thumpd3 212 Tmco5

M/NB ∩ M/P ∩ M/NB ∩ M/P M/NB ∩ P/NB M/P ∩ P/NB M/NB only M/P only P/NB only P/NB (67) (87) (8) (125) (62) (83) (1)

Tmem68

Tpo

Wtap

Zbtb25

Zfand1

Zfp131

Zfp677

Zfp715

Zfp748

Zh2c2

Zim1

Table AI-S8. Gene lists corresponding to Day 4 (Fig AI-3, right panel) Venn diagram sections. Lists were generated by GeneVenn (http://genevenn.sourceforge.net/index.htm). M/NB ∩ M/P M/NB ∩ P/NB M/P ∩ P/NB M/NB only M/P only P/NB only (8) (33) (9) (54) (59) (64)

1110028A07Rik 2310001H12Rik 2310002L09Rik 100043775 8030451F13Rik 2810055G20Rik

A2bp1 2810021G02Rik Asb12 1600016N20Rik Actn2 2810055G20Rik

EG266459 6430598A04Rik EG435337 1700093J21Rik Actn3 666761

EG625558 Apod ENSMUSG000000747 2610318N02Rik Ampd1 Akr1c14

Jph1 BC023105 Lcn2 3830408D24Rik Ankrd2 Aqp7

Mylk4 D930028M14Rik Pkia 4732444A12Rik Ankrd23 BC022713

Olfr624 Doc2a Ppp1r3a 4933407I18Rik Ano5 BC061194

Ttn EG544710 Scn2a1 6720416L17Rik Art1 Cdkl4

EG624855 Tomm22 Acoxl Asb14 Cfd

ENSMUSG00000073 Angptl1 Bex1 Cidec

005 213

M/NB ∩ M/P M/NB ∩ P/NB M/P ∩ P/NB M/NB only M/P only P/NB only (8) (33) (9) (54) (59) (64)

ENSMUSG00000074 Ankrd32 Cacna1s Crlf1

381

Ecm2 Arhgap5 Cacng1 Cxcl13

Ereg Aspn Cap2 EG240549

Fcer1a B930036N10Rik Car3 EG432982

Hmcn1 Cabc1 Chrna1 EG623356

Kctd4 Capn9 Cox8b EG626058

Krt78 Chrdl2 Cryab EG668536

LOC635086 Chst13 Des EG668662

Lztfl1 Dnajb4 Dmd ENSMUSG000000755

OTTMUSG00000013 EG632964 Dppa5a Ear1

370

OTTMUSG00000023 EG639396 EG628696 F830116E18Rik

442

Ogn ENSMUSG00000064 Eda2r Foxi2

032

Olfr1337 Fbxl7 Fhl1 G0s2

Phf20 Gcm1 Fsd2 Gpm6a

Prss35 Gkn1 Hspb6 Hist1h2ab

Rnu35a Gm414 Hspb8 Hpx

Rps6ka6 Gm889 Klhl31 I830012O16Rik

Sema3a Gnb1l Klhl31 Ifit1

Smoc1 Gucy2d Lrrc39 Itgbl1

Trav7d-2 Kbtbd8 Mb Kcng4

Zfp455 LOC100045676 Mef2c Klb

Zfp51 LOC280487 Mmp3 Klhl4

Zfp617 Lmtk3 Mybpc2 Klk5

Lrrc26 Myh4 Lce1b

Mrpl47 Myl3 Lce1f

Neurog1 Mylk2 Lce1i

Nrn1 Myo18b Mmd 214

M/NB ∩ M/P M/NB ∩ P/NB M/P ∩ P/NB M/NB only M/P only P/NB only (8) (33) (9) (54) (59) (64)

Nt5c3 Myo18b Mosc1

Ntng1 Myot Mrap

Olfr1349 Myoz1 Mx2

Olfr544 Nexn Myo9a

Oxct2a Nrap Nkx2-9

P2rx1 OTTMUSG00000010 OTTMUSG00000005

956 148

Pramel7 Obox6 Olfr1499

Pzp Pfkm Olfr328

Reep6 Pfn2 Omd

Slc16a5 Popdc3 Onecut3

Slc25a42 Prkaa2 Oxtr

Snai3 Prr8 Paqr9

Tmem37 Pygm Plin

Trim13 Rbm24 Ppyr1

Zfp322a Ryr1 Retn

Zfp442 Saa3 Rhcg

Zfp85-rs1 Smyd1 S3-12

Srd5a2l2 Sh2d1b1

Tmem182 Slc22a3

Tmod1 Slc36a2

Trim72 Slitrk6

Txlnb Sp9

Spink12

Tmprss11d

Trav2

V1ra7

Zfp712

215

216

Appendix II Effectiveness factor and diffusion limitations in collagen gel modules containing HepG2 cells1 Abstract

A major obstacle in tissue engineering is overcoming hypoxia in thick, three-dimensional (3D) engineered tissues, which is caused by the diffusional limitations of oxygen and lack of internal vasculature to facilitate mass transfer. Modular tissue engineering is a bio-mimetic strategy that forms scalable, vascularized and uniform 3D constructs by assembling small (sub-mm), cell- containing modules. It was previously assumed that mass transfer resistance within the individual modules was negligible, due to their small size. In the present study, this assumption was tested using theoretical analysis of oxygen transport within the module (effectiveness factor) and experimental studies. Small (400 µm diameter post-contraction) and large (700 µm diameter post-contraction) HepG2–collagen modules were made for a range of seeding densities (2 × 105–

7 1 × 10 cells/ml collagen). Cell density, distribution and morphology within the modules showed

7 7 that the small modules were capable of sustaining high cell densities (8.0 × 10 ± 4.4 × 10 cells/cm3) with negligible mass transfer inhibition. Conversely, large modules developed a necrotic core and had significantly (p < 0.05) reduced cell densities (1.5 × 107± 9.2 × 106 cells/cm3). It was also observed that the embedded cells responded quickly to the oxygen availability, by proliferating or dying, to reach a sustainable density of approximately 8000 cells/module. Furthermore, a simple effectiveness factor calculation was successful in estimating the maximum cell density per module. The results gathered in this study confirm the previous

1 Reprinted from Journal of Tissue Engineering and Regenerative Medicine, 5(2), Corstorphine L and Sefton MV, Effectiveness factor and diffusion limitations in collagen gel modules containing HepG2 cells, 119 - 129, Copyright 2011, with permission from John Wiley and Sons.

217 assumption that the small-diameter modules avoid the internal mass transfer limitations that are often observed in larger constructs.

1 Introduction

Engineering of thick, three-dimensional (3D) tissues continues to be limited by the lack of an internal vasculature, which leads to hypoxia and necrosis of cells in the interior of thick tissues. Strategies to address this issue include the incorporation of angiogenic growth factors and/or endothelial cells within the scaffold or the use microfabrication techniques to create a vascular blueprint in the scaffold, which can then be seeded with endothelial cells[1]. Modular tissue engineering is a bio-mimetic strategy that forms scalable, vascularized and uniform three- dimensional (3D) constructs. As implemented, the modules (Figure AII-1) are sub-mm collagen cylinders (∼600 µm in length and ∼400 µm in diameter), embedded with functional cells and coated with endothelial cells. Randomly packing modules together creates a perfusable construct, where interstitial spaces among the modules form interconnected channels lined with endothelial cells. The modular system has the ability to produce perfusable constructs with high cell densities [2], indicating that the interconnected channels between packed modules provides sufficient mass transport to prevent overtly hypoxic conditions.

Previous modelling efforts have focused on design constraints of the modular construct, including pressure drop across the construct, shear stress on the endothelial surface layer and oxygen depletion along the axial length of the construct[3]. The resulting model predicted that for superficial velocities, associated with physiological shear stresses and arteriovenous pressure drops, a modular construct should be capable of supporting cell densities of 106–107 cells/cm3 (roughly 0.3–3% of normal tissue densities) without suffering necrosis due to hypoxia [3]. However, a major assumption in this calculation was that mass transfer resistance within the modules themselves was negligible. The present work has focused on testing this assumption, using an analysis of oxygen transport within the module and related experimental studies. Module contraction, cell growth and heterogeneous cell distribution complicate the theoretical analysis but support the previous conclusion [3], that mass transfer resistance in the modules made previously is negligible.

218

Embedded functional cells

Collagen Endothelial cells

Mass transport within the modules was modelled using the Thiele modulus (the ratio of the rate of reaction and the rate of internal diffusion, ϕ) and the effectiveness factor (the ratio of the actual rate of reaction and the rate of reaction in the absence of internal mass transfer limitations, η). The use of these dimensionless numbers is well-established in the literature for describing transport-reaction kinetics in immobilized cell[4-7] and enzyme systems[8,9]. Oxygen was selected as the target molecule, as it is the primary molecule contributing to cell death, due to insufficient mass transfer[10]. A Thiele modulus expression, describing oxygen transport and reaction in immobilized cell aggregates, was based on the work of Karel et al.[4] and the following assumptions: (a) the aggregate is isothermal; (b) mass transfer is governed by Fick's law; (c) the aggregate is homogeneous (i.e. a uniform cell distribution); (d) the external transport to the particle is fast (i.e. no external mass transfer resistance); (e) the system is at steady state; (f) diffusion occurs in only the radial direction; and (g) a single molecule (e.g. oxygen) can be studied independently. Further assumptions were made to adapt this model to describe a module: (h) the modules were cylindrical in shape and end-effects were absent; (i) the oxygen uptake rate (OUR) was assumed to be independent of cell density; and (j) only one cell type (i.e. HepG2

219 cells) was present. These assumptions were used to generate the following Thiele modulus expression:

2 2 ρcell ()OUR ()dm 4 Thiele modulus φ = (1) ∗DC eff collagen)(

3 where ρcell is the cell density (cells/m ), OUR is the oxygen uptake rate per cell (mol O2/cell/s), 3 dm is the module diameter (m), C* is the molar concentration of oxygen in bulk (mol O2/m ) and 2 Deff is the effective diffusivity of oxygen in the module (m /s).

The effectiveness factor equation (equation 2) represents the analytical solution of the effectiveness factor for particles of cylindrical geometry, assuming first-order reaction kinetics[11]:

2 I1 ()φ Effectiveness Factor η = (2) φ I 0 ()φ

where I0 and I1 are the modified Bessel functions of the 0th and first order, respectively. An effectiveness factor of 1.0 represents a system with no mass transfer resistances. For the purposes of this study, it was assumed that mass transfer effects would not be apparent experimentally unless the theoretical effectiveness factor was < 0.9. Consequently, modules with an effectiveness factor of 0.9–1.0 were not expected to show any signs of mass transfer resistance, such as reduced metabolic activity or cell death.

2.1 Parameters

Values for the oxygen uptake rate (OUR) of HepG2 cells, the concentration of oxygen in bulk fluid (C*) and the diffusivity constants of oxygen in collagen (Deff(collagen)) were estimated from −16 3 −5 2 the literature to be 3.72 × 10 mol O2/cell/s [12-14], 0.13 m/m [15] and 2.99 × 10 cm /s [13], respectively.

220

3 Materials and methods 3.1 Cell culture

Human hepatoma cell line (HepG2; HB-8065), and human umbilical vein endothelial cell line (HUVEC-C; CRL-1730) were obtained from American Type Culture Collection (American Type Culture Collection, Manassaa, VA, USA). As these cell lines were to be used together, a co- culture medium was prepared by combining the suggested growth medium for HepG2 cells [Eagle's minimum essential medium (eMEM), 10% FBS, 1% penicillin–streptomycin] with the HUVEC-C growth medium supplements (ECGS and heparin) suggested by the supplier. HepG2 and HUVEC-C were cultured in 25 mm2 tissue culture flasks in this co-culture medium [eMEM (American Type Culture Collection, Manassas, VA, USA) supplemented with 10% fetal bovine serum (Invitrogen Canada, Burlington, ON, Canada), 1% penicillin-streptomycin (Invitrogen Canada), 0.03 mg/ml endothelial cell growth supplement (ECGS; BD Biosciences, Mississauga, ON) and 0.01 mg/ml heparin (Heparin LEO, LEO Pharma Inc, Thornhill, ON, Canada)] at 37 °C in a 5% CO2 humidified air atmosphere. The culture medium was changed every 2–3 days.

3.2 Module fabrication

Modules containing embedded HepG2 cells were fabricated and then coated with endothelial cells, as described previously[2,16]. A collagen solution was prepared by mixing acidified collagen (type I, bovine dermal, 3.1 mg collagen/ml; Vitrogen, Cohesion technologies, Palo Alto, CA, USA) with 10x minimum essential medium (Invitrogen Canada), followed by neutralization using 0.8 m NaHCO3. Pelleted HepG2 cells were resuspended in the neutralized collagen. The suspension was drawn into a gas-sterilized polyethylene tube through the open end by withdrawing the plunger of a 3 cc syringe attached to the needle at the opposite end of the tube. The tubing was incubated for 30–45 min until the collagen gelled. The tubing was then cut into 2 mm sections, using an automated cutter (FCS Technology, London, ON, Canada). The modules were released from the tubing by vortexing gently in co-culture medium.

Six module fabrication conditions, (i.e. initial module diameter and HepG2 seeding density combinations), were selected for experimental analysis, based on their theoretical effectiveness factors after contraction (see Results). Modules were made using either 0.762 mm i.d. tubing (small modules) or 1.40 mm i.d. tubing (large modules) and contained 1 × 106 - 1 × 107 HepG2

221 cells/ml collagen. Immediately following fabrication, HUVEC-C (7.5 × 105 cells/ml of settled modules) were added to the modules to effect module contraction. The endothelial cells were incubated with modules in a 15 ml centrifuge tube for 1 h, while the tube was gently rocked to ensure full coverage. The modules were then transferred to an untreated polystyrene Petri dish and placed in a 37 °C, 5% CO2 incubator. The medium was changed every 2–3 days. Contraction of the modules by the HUVEC-C occurred within the first 72 h.

Modules were selected randomly and transferred to a 24-well plate for imaging, using an inverted microscope (Zeiss, Axiovert 135) at × 2.5 objective magnification. AxioVision, digital imaging software supplied by Zeiss, was used to measure the module lengths and diameters.

3.3 Western blotting

Western blots probing for the GAPDH housekeeping protein were used to estimate the total number of cells/module (combined HepG2 and HUVEC-C). The rationale for selecting this method is discussed below. On days 3 and 7, 60 modules were transferred to a 1.5 ml Eppendorf tube and centrifuged at 50 000 rpm for 2 min and the supernatant was removed. To prepare protein samples for western blotting, 100 µl Laemmli sample buffer (Biorad; containing 5% β- mercaptoethanol) was added to each module sample and the mixtures were heated in a dry bath at 100 °C for 5 min to lyse the cells and solubilize the proteins, including collagen gel. After heating, the samples were passed through a 23-gauge syringe to further homogenize the modules, which tended to agglomerate. The samples were stored at − 20 °C until use.

The proteins were separated on 10% SDS–PAGE, transferred to a nitrocellulose membrane (Amersham Hybond-ECL, GE Healthcare) and probed with 1 µg/ml anti-GAPDH (rabbit; Santa Cruz Biotechnology) in milk for 1 h at room temperature. An HRP-conjugated IgG goat anti- rabbit antibody was used to visualize the blot with ECL reagents. The bands were quantified using a Versadoc 3000 and Quantity 1-D Analysis Software (Bio-Rad Laboratories). Standard curves for each gel were generated using band density (intensity/mm2) for known numbers of HepG2 cells (processed in the same way as module samples) and were used to calculate the number of cells per module.

222

3.4 Alamar blue reduction assay

On days 3 and 7, 10 modules were transferred to a 24-well plate (in triplicate) in 250 µl fresh co- culture medium. Alamar Blue™ (AB; Biosource, Camarillo, CA, USA) was added to each well such that the final solution (500 µl) contained 10% AB. Samples were incubated for 4–6 h. Supernatant samples were transferred to 96-well plates and the absorbances were measured at 570 nm and 600 nm, using a Sunrise plate reader. Co-culture medium containing 10% AB was used as the negative control. The reduction of AB was calculated from the absorbances according to the supplier's instructions.

3.5 Human albumin ELISA

On days 3 and 7, the rate of albumin secretion from the modules was measured. Ten modules were transferred to a 24-well plate (in triplicate) in 500 µl fresh medium. Fresh co-cultured medium was used as a negative control. The plate was incubated for 24 h, after which samples of supernatant were transferred to 1 ml Eppendorf tubes and stored at 4 °C until the albumin concentration was measured using a Human Albumin ELISA Quantitation Kit (Bethyl Laboratories, TX, USA). The ELISA protocol outlined by the manufacturer was followed. The standard curve was constructed in Statistica 6.1 (StatSoft, Tulsa, OK, USA) using a three- parameter logistics curve-fit.

3.6 Histology

Large and small modules were fabricated with an initial HepG2 cell density of 8 × 106 cells/ml and were seeded with HUVEC-C, as described above. Cell distribution within the modules was assessed using histology and confocal imaging. For histological analysis, modules were fixed in 4% paraformaldehyde, rinsed with PBS and embedded in agar blocks. The blocks were stored in formalin prior to being embedded in paraffin. The paraffin blocks were cut into 4 µm sections and stained with Masson's trichrome. Cell viability in modules was assessed using a Live/Dead® cell viability assay (Molecular Probes, Invitrogen). Modules were stained according to the supplier's instructions, rinsed with PBS and imaged immediately, using a Zeiss LSM510 confocal microscope and software (Advanced Optical Microscopy facility, Princess Margaret Hospital, Toronto, Canada).

223

3.7 Statistics

One-factor or two-factor analyses of variance (ANOVA) were used to analyse the effect of initial module diameter, or module seeding density and initial module diameter, respectively, using α = 0.05. All error bars on graphs represent standard deviation (SD).

4 Results 4.1 Theoretical analysis

The Thiele modulus and effectiveness factor were used to estimate the theoretical availability of oxygen within the modules, with a view to estimating appropriate and attainable conditions for experimental evaluation. For small modules (<0.5 mm diameter), the effectiveness factor was very high (>0.95) and cell density had a negligible effect (Figure AII-2).

1.00 2.5

0.95 2.0

■ 1x106 cells/ml 0.90 1.5 ● 5x106 cells/ml ▲ 1x107 cells/ml 0.85 1.0 Thiele Modulus Effectiveness Factor Effectiveness 0.80 0.5

0.75 0.0 0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 Module Diameter (mm)

As the module diameter increased, the cell density began to affect the effectiveness factor and Thiele modulus to a greater extent, and mass transfer limitations appeared to become increasingly severe. For the purpose of this study, it was assumed that the effects of mass transfer

224 inhibition would not be evident in the experimental data for modules with an effectiveness factor above 0.90. Hence, for example, a module with a contracted diameter of 0.40 mm, approximately the size of modules used earlier [2,16], was expected to be able to support a maximum 9.4 × 107 cells/cm3 before experiencing significant intramodule mass transfer resistance (based on the literature oxygen uptake rate, etc.).

4.2 Module fabrication, size and effectiveness factors (based on seeding density)

In previous studies, modules were made with 0.76 mm tubing. In order to study the effect of module diameter on diffusional limitations, (i.e. have η< 0.9, large modules were prepared using 1.40 mm tubing). The latter were indeed larger than the normal small-diameter tubing modules when freed from the polyethylene tubing (Figure AII-3a, day 0) and live cells were distributed uniformly throughout both large and small modules (Figure AII-3b).

Collagen gel modules containing only HepG2 cells were too soft and weak to be used as they were. Hence, modules were seeded with HUVEC-C immediately following fabrication, which resulted in the contraction of modules by day 3 (Figure AII-3a); these modules are more robust

225 and are similar to what we have used in earlier studies[2]. Following contraction, modules made with 0.76 mm tubing were significantly smaller in both diameter and length than modules made with 1.40 mm tubing; the modules had been the same length at day 0. Using only the seeding density, the change in volume (i.e. both cell density and diameter) was used to recalculate the effectiveness factor (Figure AII-4) for modules post-contraction, assuming no cell proliferation or death had occurred in the 3 days of contraction. This ‘no change in cell number’ assumption was, not surprisingly, found to be incorrect (see below). Nonetheless, this assumption was sufficient to ensure that the selected experimental conditions would highlight conditions under which diffusion limitations should have appeared. From the adjusted effectiveness factor values, it was expected that all the small-diameter modules would not show evidence of mass transfer resistance (i.e. η> 0.9). Only the large modules seeded at 8 × 106 or 1 × 107 cells/ml would be expected to have η< 0.9 and therefore have a reduced, diffusion-limited, cellular metabolism and viability.

1.00 Fabrication

0.95 Post-Contraction

0.90 !" !" 0.85

Effectiveness Factor 0.80

0.75 Seeding 2.0 4.0 10.0 2.0 8.0 10.0 (106 cells/ml)

Initial diameter 0.76 0.76 0.76 1.40 1.40 1.40 (mm)

4.3 Effect of module diameter on cell density, metabolism and albumin

The total number of cells (HUVEC-C and HepG2 combined) per module was determined from the amount of GAPDH present in a sample of 60 modules. The number of cells per module

226

(based on GAPDH) was similar among all module sets (Figure AII-5a). However, small modules had significantly higher (p < 0.05) cell densities (cells/cm3) than the large modules when module volumes were taken into account (Figure AII-5b). Two-way ANOVAs were used to determine whether seeding density and module diameter had significant affects on module cell density and cell number. The cell density (cells/cm3) was found to be significantly affected by module size (p < 0.05) but not by seeding density (p > 0.88) on days 3 and 7 (Figure AII-5b). Neither module size nor seeding density had a significant effect on the number of cells per module (p > 0.67 and p > 0.90; Figure AII-5a). It was concluded that the cellular content of the modules (cells/mod and cells/cm3) was independent of the seeding density (p = 0.88). Consequently, the remaining data have been presented as the mean for small and large modules. For example, small modules contained 8.0 × 107 ± 4.4 × 107 cells/cm3 on day 7, while the average density for large modules was 1.5 × 107 ± 9.2 × 106 cells/cm3, independent of the original seeding density.

14000 1.40E+08

12000 1.20E+08

10000 1.00E+08

8000 8.00E+07

6000 6.00E+07 Cells/ml

Cells/Module 4000 4.00E+07

2000 2.00E+07

0 0.00E+00 Day 3 Day 7 Day 3 Day 7 !" #" Small: 2x106 cells/ml 4x106 cells/ml 1x107 cells/ml Large: 2x106 cells/ml 8x106 cells/ml 1x107 cells/ml Figure AII-5. (a) The number of cells per module was estimated from the amount of GAPDH per module, using western blots. No differences were found in the number of cells per module among large and small modules, regardless of cell number and time (p > 0.88) (b) Cell density was calculated from the cell number per module using the module volumes. Small-diameter modules had a significantly higher (p < 0.05) cell density compared to large-diameter modules, although these were not different from days 3 to 7. The results are average ± SD (n = 2)

Although large and small modules had similar cell numbers (per module), large modules had significantly higher Alamar blue (AB) reduction rates per module on days 3 and 7 compared to small modules (p < 0.05) (Figure AII-6), again combining results by module size and ignoring the small effect of seeding density. The reduction rate per cell was also greater in large modules (p < 0.05) on both days. No significant (p > 0.10) difference was seen in the albumin secretion

227 rates (per cell) of large and small modules at either time point (combining results regardless of seeding density), although there was a high degree of variability among all modules for all time points, as indicated by the relatively large error bars (Figure AII-7).

0.25 0.07

0.06 0.20 0.05

0.15 0.04 Small Large 0.10 0.03 0.02 0.05 0.01 % AB reduction/ 1000 cells / hr % % AB reduction / 10 modules hr % 0.00 0 Day 3 Day 7 a b Day 3 Day 7

0.030

0.025

0.020 Small 0.015 Large 0.010

ng albumin / cell / day / cell / albumin ng 0.005

0.000 Day 3 Day 7

Figure AII-7. Albumin secretion from modules was measured using an enzyme-linked immunosorbent assay. No significant difference in secretion rate per cell was seen between large and small modules on days 3 or 7. The results are average ± SD, regardless of initial seeding density (n = 16)

4.4 Cell distribution, morphology and viability within modules

While cells were uniformly distributed in modules on day 0 (Figure AII-3a), images of modules on day 3 clearly showed a core of dead cells in the large modules and a thick band of viable (green) cells along the perimeter of the module (ca. 150–200 µm deep) (Figure AII-8b). In

228 contrast, the small diameter modules had retained a homogeneous distribution of embedded cells, similar to what was present on day 0, and showed very little cell death (Figure AII-8a). Dye penetration was not sufficient to determine cell viability in the core of modules on day 7.

! Small Modules Large Modules Day 3

Histology sections showed uniformly dispersed embedded cells prior to HUVEC-C seeding for both module sizes (Figure AII-9). By day 3, the HepG2 cells in the small modules had formed spheroids which were still present and well distributed within the module at day 7. Spheroid formation was also seen in the large modules on day 3; however, it appeared that a smaller percentage of HepG2 cells aggregated to form spheroids and the majority of the cells remained isolated from each other. A large mass of cell debris and voids was visible in the centre of the large modules by day 7, indicating the presence of a necrotic core.

229

230

4.5 Effectiveness factor based on actual cell density

Figure 4 was used to calculate effectiveness factors based on initial seeding densities and day 3 experimental volumes, on the assumption that there was no change in cell number over these 3 days. This led to the choice of experimental conditions. Using the experimental cell densities, experimental effectiveness factors were calculated (Table AII-1) based on the oxygen uptake rate −16 (OUR) of 3.7 × 10 mol O2/cell/s that was based on the reported OUR of hepatocytes and HepG2 cells grown in different configurations [12-14]. The experimental effectiveness factor was surprisingly consistent among all groups, in the range 0.91–0.94.

Table AII-1. Effectiveness factors calculated from experimental cell densities and diameters Day 3 Day 7 Small modules 0.92 ± 0.01 0.91 ± 0.01 Large modules 0.94 ± 0.01 0.94 ± 0.01 5 Discussion

The theoretical and experimental analysis supported the assumption that mass transfer limitations do not pose a major design constraint for modular tissue engineering, as very high cell densities were reached in the small modules without the formation of a necrotic core. As expected, mass transfer became a greater concern both theoretically and experimentally as module diameter increased. More interestingly, it appeared that the cell density and cell distribution within the modules ‘self-regulates’ to accommodate to the available nutrient supply, at least in the system investigated here.

5.1 Experimental protocol

HepG2 cells were selected for study here due to their ease of culture, production of unique, liver- specific markers, (e.g. albumin, and high OUR[17]). These cells were studied previously in modules[2,3,16]. Additional benefits lie in the extensive literature on bioartificial livers studying HepG2 and hepatocytes in multiple culture configurations, which provide suitable benchmarks for evaluating hepatic activity and cell density. HepG2 cells were cultured in a co-culture medium because the modules also contained HUVEC-C. In 96-well plate culture, the chosen medium yielded higher metabolic activity for both HepG2 and HUVEC-C (measured by Alamar blue reduction) and higher albumin secretion rates in HepG2 cells (measured by human albumin ELISA) than either normal HepG2 culture medium or normal HUVEC-C culture medium.

231

The highest seeding density that was useable was 1 × 107 cells/ml. At 1.5 × 107 cells/ml, the modules lacked the necessary structural integrity. This constraint prevented the fabrication of modules with very low theoretical effectiveness factors, even following contraction by endothelial cells. The endothelial layer was not required to reduce the thrombogenicity of the modules [2,3,16,18]; rather, the HUVEC-C were used to contract the modules to enhance their robustness for experimental study, and to enable their comparison to earlier studies using the same 0.762 mm i.d. tubing for module fabrication.

The total number of cells per module was calculated from the amount of GAPDH present in a sample of 60 modules. Other cell enumeration methods were attempted, including manual cell counting following collagenase digestion of modules, and total DNA quantification following SDS-based module digestion and cell lysis. However, western blotting was found to produce the most consistent and reproducible results. These other standard means of determining cell number are typically used with larger cell aliquots (and without the contaminating effects of collagen) than we had available.

GAPDH was selected to be the reference protein because its expression is not regulated by hypoxia in HepG2 cells: mRNA and protein GAPDH levels were unchanged in severely hypoxic, normoxic conditions and following reoxygenation[19]. GAPDH is known to be regulated by hypoxia in endothelial cells[20]but this was not expected to impact the results, as the endothelial cells were seeded at low density and being at the surface of the modules they were not expected to experience hypoxic conditions. The cell density for small modules (8 × 107 cells/cm3 or about 8000 cells/module), obtained using the GAPDH western blot, was consistent with previously reported cell densities (3–10 × 107 cells/cm3) for HepG2-collagen modules [16]. Histology sections of small and large modules stained with CD31 antibody showed only 1 or 2 positively stained cells per 4 µm-thick module section (data not shown), indicating that < 1% of the 8000 cells/module were endothelial cells. α-Tubulin was also considered for the western blot cell enumeration assay, but the modules samples did not contain sufficient amounts of α-tubulin to obtain reproducible data (data not shown). It did, however, support the trends seen with GAPDH, in which small modules had significantly higher cell densities (but not number per module) than large modules.

232

5.2 Effect of module diameter

The small modules, the size normally made [2], were capable of supporting significantly higher cell densities than large modules, and these high densities were reached rapidly, with the majority of the proliferation occurring within the first 3 days (Figure AII-5). The cell density (8.0 × 107 ± 4.4 × 107 cells/cm3) of the small modules (ignoring the small fraction of these that are endothelial cells) was higher than hepatocytes or liver cell lines cultured in PMDS-microfluidic devices (4 × 107 cells/cm3) [21], microcarriers (7 × 106 cells/cm3) [22], hollow-fibre bioreactors (2.5 × 107 cells/cm3) [23] and packed-bed bioreactors (4.8 × 107 cells/cm3) [24], but was lower than reported values for radial flow bioreactors (1 × 108 cells/cm3 scaffold) [25]. The consistency of albumin secretion per cell among all modules (Figure 7) implies the viable cells within both large and small modules had similar hepatic activity, independent of cell density (ignoring the small fraction of endothelial cells). The average albumin secretion rate (14 µg/106 cell/day) was comparable to those of both primary hepatocytes and HepG2 spheroids reported in literature[26- 28].

Although cell densities were higher with small modules, the number of cells per module were similar (because large modules had a higher volume than small modules). Initially, the HepG2 cells were well-dispersed and within both large and small modules. The HepG2 cells in the small modules proliferated, formed spheroids and remained uniformly distributed, while in the large modules the cell density increased less or not at all (relative to the seeding density) and the core became necrotic (Figures AII-8, AII-9). This dead core is commonly observed with hypoxia- induced necrosis in large tissue-engineering constructs and tumours [29], and is an excellent example of why vascularized constructs are critical in developing clinically relevant tissues. The shell of viable cells at the surface of the large modules was approximately the same depth as oxygen's diffusion limit (∼200 µm), as expected from the literature [30]. In small modules, whose diameter is ∼400 µm, this necrotic core was not seen and by day 7, a highly dense ‘microtissue’ was observed.

On the other hand, Alamar blue reduction results (Figure AII-6) indicate a greater metabolic activity for the cells in larger modules on a per-module basis (despite similar cell numbers per module) or a per-cell basis. Here we presume that a difference in distribution (spheroidal aggregates in small modules, a viable periphery/necrotic core in large modules) results in

233 phenotypic differences such that metabolic activity is lower in small modules. A similar effect was reported by Cho et al.[31], who found that hepatocytes in low and medium hepatocytes density co-cultures (with 3T3 fibroblasts) had higher oxygen consumption, albumin synthesis and urea formation rates per cell than hepatocytes cultured at high densities.

5.3 Theoretical model

The results from this study highlight the importance of construct thickness when creating engineered tissues. Although three different seeding densities were used for each module diameter when fabricating the modules, only the diameter elicited any obvious effect. Changing the seeding density did not result in different final cell densities. In small-diameter modules, where diffusion limitations for the seeding density were not expected (theoretical effectiveness factors were all > 0.9 after module contraction) cells grew throughout the module until they reached the apparently maximum permissible cell number (about 8000 cells/module) with a uniform (throughout the module) density of 8 × 107 cells/cm3. That is, the module's dimensions can enable support for no more than these 8000 metabolically active cells. At this cell density, the calculated effectiveness factor was 0.91–0.92 (Table AII-1, using the estimated oxygen −16 uptake rate of 3.7 × 10 m O2/cell/s), which may be an upper limit in effectiveness factor. The actual numerical value is dependent on the oxygen uptake rate. Embedded HepG2 cells may have a significantly lower OUR than HepG2 cultured on flat surfaces, perhaps as low as 1.5 × 10−16 m

O2/cell/s [32]. Using this value (and ignoring the small fraction of endothelial cells) would raise the effectiveness factor of the modules to ∼0.97, close to the value of 1 that would be expected if there were indeed no diffusion limitations.

In larger modules, there were also about the same number of cells per module as in small modules, but the cells were not distributed uniformly. The viable cells were clustered at the periphery of the module, leaving a necrotic core. When the seeding density was high, the module's dimensions could not support the viability of all the seeded cells and cell density (especially in the core where hypoxia was greatest) decreased to reach the maximum supportable density. As a result the effectiveness factor again became > 0.9 (actually 0.94) although it had been much lower at seeding (Figure AII-4). It is interesting to note that the effectiveness factors for large and small modules were almost identical, suggesting that the 0.91–0.94 range (for the assumed OUR) is as high as is achievable, and that the cells in the module adjust their number

234

(and distribution) to reach this limit, regardless of module diameter and seeding density. Some caution is needed, however, in applying the Thiele modulus/effectiveness calculation for the large modules, since such modules did not have a homogeneous distribution of cells, an assumption within the model. On the other hand, formally taking into account the non-uniform cell distribution is unlikely to have much impact on the estimated effectiveness factor[4,8].

The simple theoretical model used here was effective as a means of selecting suitable experimental conditions for further study. While this model is limited, due to the numerous assumptions and simplistic kinetics used, it was useful for estimating appropriate seeding densities for modules. Had we not used this model, we would not have known what size of module and seeding conditions would likely suffer from diffusion limitations. Using the model, it is apparent that under most circumstances there would be no diffusion limitations even without cells growing after seeding. When cells, like HepG2 cells, can adjust their cell density to suit their environment, it becomes even harder to make modules that are ‘too big’. Or rather, ‘too big’ means a necrotic core with not too low an effectiveness factor.

6 Conclusions

The data gathered from this study clearly demonstrates the ability of the small diameter modules to support high cell densities (8 × 107 cells/cm3) with negligible mass transfer inhibition. This confirms the assumption made previously by our group, that these modules avoid the internal mass transfer limitations to which larger constructs fall victim. Furthermore, the maximum cell density per module can be estimated using a simple effectiveness factor calculation, based on the oxygen uptake rate of the embedded cell and the module diameter. It was also observed that the cells embedded in the modules quickly responded to oxygen availability, by either proliferating or undergoing necrosis in the module core to reach a sustainable density. The self-regulation of cell density implies that the target cells embedded in modules will (if capable) proliferate to, and maintain, optimal cell densities, so long as the critical depth of oxygen penetration (∼200 µm) is not exceeded.

7 Acknowledgements

This work was supported by funding from the Natural Sciences and Engineering Research Council and the US National Institutes of Health (Grant No. EB 006903) and through an Ontario

235

Graduate Scholarship and National Science and Engineering Council Canadian Graduate Masters Scholarship. The authors would like to thank the Advanced Optical Microscope Facility (AOMF) for the use of their equipment and the Histological Services for Animal Research in the Pathology Research Program at University Health Network (Toronto, Ontario) for the services provided.

8 References

[1] Kaully T, Kaufman-Francis K, Lesman A, Levenberg S. Vascularization - the conduit to viable engineered tissues. Tissue Eng Part B Rev 2009;15(2):159-69. [2] McGuigan AP, Sefton MV. Vascularized organoid engineered by modular assembly enables blood perfusion. Proc Natl Acad Sci USA 2006;103(31):11461-6. [3] McGuigan AP, Sefton MV. Design criteria for a modular tissue-engineered construct. Tissue Eng 2007;13(5):1079-89. [4] Karel SF, Libicki SB, Robertson CR. The Immobilization of Whole Cells: Engineering Principles. Chemincal Engineering Science 1985;40:1321 - 54. [5] Dalili M, Chau PC. Intraparticle diffusional effects in immobilized cell particles. Appl Microbiol Biotechnol 1987;26(6):500 - 6. [6] Chang HN, Moo-Young M. Estimation of oxygen penetration depth in immobilized cells. Appl Microbiol Biotechnol 1988;29(2-3):107 - 12. [7] Annesini MC, Castello G, Conti F, Conti De Virgiliis L, Marrelli L, Miccheli A, et al. Transport and consumption rate of O2 in alginate gel beads entrapping hepatocytes Biotechnol Lett 2000;22 (10):865-70. [8] Do D. Enzyme immobilization in porous solid supports-penetration of immobilized enzyme. Biotechnol Bioeng 1984;26(9):1032 - 937. [9] Watanabe H, Matsuyama T, Yamamoto H. Preparation of immobilized enzyme gel particles using an electrostatic atomization technique. Biochem Eng J 2001;8(2):171-4. [10] Avgoustiniatos ES, Colton CK. Design considerations in immunoisolation. In: Lanza RP, Langer R, Chick WL, editors. Principles of Tissue Engineering. Austin TX: Academic Press; 1997. p. 333-46. [11] Churchill SW. A generalized expression for the effectiveness factor of porous catalyst pellets. AIChE J 1977;23 (2):208-10. [12] Rotem A, Toner M, Tompkins RG, Yarmush ML. Oxygen Uptake Rates in Cultured Rat Hepatocytes. Biotechnol Bioeng 1992;40 1286-91. [13] Shatford RA, Nyberg SL, Meier SJ, White JG, Payne WD, Hu WS, et al. Hepatocyte function in a hollow fibre bioreactor: a potential bioartificial liver. J Surg Res 1992;53 (6):549-57.

236

[14] Smith MD, Smirthwaite AD, Cairns DE, Cousins RB, Gaylor JD. Techniques for measurement of oxygen consumption rates of hepatocytes during attachment and post- attachment. Int J Artif Organs 1996;19(1):36-44. [15] Fournier RL. Basic Transport Phenomena in Biomedical Engineering Philadelphia PA: Taylor & Francis 1998. [16] McGuigan AP, Sefton MV. Design and fabrication of sub-mm-sized modules containing encapsulated cells for modular tissue engineering. Tissue Eng 2007;13(5):1069-78. [17] Hay PD, Veitch AR, Smith MD, Cousins RB, Gaylor JD. Oxygen transfer in a diffusion- limited hollow fiber bioartificial liver. Artif Organs 2000;24(4):278-88. [18] McGuigan AP, Sefton MV. The thrombogenicity of human umbilical vein endothelial cell seeded collagen modules. Biomaterials 2008. [19] Said HM, Polat B, Hagemann C, Anacker J, Flentje M, Vordermark D. Absence of GAPDH regulation in tumor-cells of different origin under hypoxic conditions in - vitro. BMC Res Notes 2009;2:8. [20] Graven KK, Troxler RF, Kornfeld H, Panchenko MV, Farber HW. Regulation of endothelial cell glyceraldehyde-3-phosphate dehydrogenase expression by hypoxia. J Biol Chem 1994;269(39):24446-53. [21] Leclerc E, Sakai Y, Fujii T. Microfluidic PDMS (polydimethylsiloxane) bioreactor for large-scale culture of hepatocytes. Biotechnol Prog 2004;20(3):750-5. [22] Werner A, Duvar S, Muthing J, Buntemeyer H, Lunsdorf H, Strauss M, et al. Cultivation of immortalized human hepatocytes HepZ on macroporous CultiSpher G microcarriers. Biotechnol Bioeng 2000;68(1):59-70. [23] Jasmund I, Langsch A, Simmoteit R, Bader A. Cultivation of primary porcine hepatocytes in an OXY-HFB for use as a bioartificial liver device. Biotechnol Prog 2002;18(4):839-46. [24] Fukuda J, Okamura K, Nakazawa K, Ijima H, Yamashita Y, Shimada M, et al. Efficacy of a polyurethane foam/spheroid artificial liver by using human hepatoblastoma cell line (Hep G2). Cell Transplant 2003;12(1):51-8. [25] Hongo T, Kajikawa M, Ishida S, Ozawa S, Ohno Y, Sawada J, et al. Three-dimensional high-density culture of HepG2 cells in a 5-ml radial-flow bioreactor for construction of artificial liver. J Biosci Bioeng 2005;99(3):237-44. [26] Yamashita Y, Shimada M, Ijima H, Nakazawa K, Funatsu K, Sugimachi K. Hybrid- artificial liver support system. Surgery 2002;131(1 Suppl):S334-S40. [27] Chen JP, Yu SC, Hsu BR, Fu SH, Liu HS. Loofa sponge as a scaffold for the culture of human hepatocyte cell line. Biotechnol Prog 2003;19(2):522-7. [28] Verma P, Verma V, Ray P, Ray AR. Formation and characterization of three dimensional human hepatocyte cell line spheroids on chitosan matrix for in vitro tissue engineering applications. In Vitro Cell Dev Biol Anim 2007;43(10):328-37. [29] Carmeliet P. Mechanisms of angiogenesis and arteriogenesis. Nat Med 2000;6(4):389- 95.

237

[30] Bruzewicz DA, McGuigan AP, Whitesides GM. Fabrication of a modular tissue construct in a microfluidic chip. Lab Chip 2008;8(5):663-71. [31] Cho CH, Park J, Nagrath D, Tilles AW, Berthiaume F, Toner M, et al. Oxygen uptake rates and liver-specific functions of hepatocyte and 3T3 fibroblast co-cultures. Biotechnol Bioeng 2007;97(1):188-99. [32] Mishra A, Starly B. Real time in vitro measurement of oxygen uptake rates for HEPG2 liver cells encapsulated in alginate matrices. Microfluid Nanofluidics 2009;6.