ANGIOGENESIS-ASSOCIATED GENE EXPRESSION CHANGES IN

SURGICAL SKIN FLAPS OF DIABETIC RATS

By

Miao Xian (Cindy) Zhou, MB, DMD

Submitted to the Faculty of the School of Graduate Studies

of the Medical College of Georgia in partial fulfillment

of the Requirements of the Degree of

Master of Science

July

2010 ANGIOGENESIS-ASSOCIATED GENE EXPRESSION CHANGES IN

SURGICAL SKIN FLAPS OF DIABETIC RATS

This thesis is submitted by Miao Xian (Cindy) Zhou and has been examined and approved by an appointed committee ofthe faculty of the School of Graduate Studies of the Medical College of Georgia.

The signatures which appear below verify the fact that all required changes have been incorporated and that the thesis has received final approval with reference to content, form and accuracy of presentation.

This thesis is therefore in partial fulfillment of the requirements for the degree of

Master of Science.

Date Major Advisor

Department Chairperson

Dean, School of Graduate Studies iv

To my director, Colonel Frederick C. Bisch, for your remarkable leadership and extraordinary dedication to the program. We are very fortunate to have you as our director as I could not have asked for anything more.

To Colonel Robert C. Gerlach for your endless optimism and cheerful demeanor that never failed to encourage me to do better.

To Colonel Robert W. Herold for your concern in my professional and personal growth. Your balance of family and career is truly admirable and I hope I can learn even half your skills. To Lieutenant Colonel (P) Douglas R. Dixon for all the expertise and time you've put into shaping me to become a better researcher and clinician. I truly appreciate all your effort. Thank you for believing in us.

To Major Thomas Johnson for introducing me to this project and for all the work and effort you've put into it to pave the way for me. I certainly would never have been able to tackle this challenging project otherwise.

To my assistant, Mrs. Idella Harris, for all your hard work, patience and mentorship. You are truly a delight to work with.

To Ms. Vickie Browning for always being there to support us no matter what and for providing outstanding support throughout our Master's and residency program.

To Mrs. Lynne McTier for holding all of us together and for exemplifYing such excellent teamwork skills. To my awesome classmates, Dr. Brian Stancoven, Dr. Kimberly Inouye and Dr. Yat-Ho Ma for being such wonderful friends and great individuals! I value the program so much more because of you guys! Thank you!

To all the other current and prior residents for your friendship, advice and expertise.

To the wonderful staff and assistants at Tingay for making this residency all the more enriching with the vast experience and laughter you bring forth.

To Mrs. Millar and the library staff at Eisenhower for the support you've provided and for always responding to my endless requests for article.

Last but not least, to my family, for always being there for me. Words cannot express how grateful I am for your unconditional love and support. Thank you so much for all that you have done for me. I know I would not be where I am today without you all. v

Major Advisor: Dr. Jill B. Lewis

Committee Members: Dr. Douglas R. Dixon Dr. Robert C. Gerlach Dr. Stephen Hsu Dr. Carol A. Lapp Dr. Jose M. Pizarro Dr. UlfM. Wikesjo TABLE OF CONTENTS

Acknowledgements iii

List of Tables viii

List at: Figures ix

Introduction

Statement of the Problem

Significance 2

Review of the Literature 4

Diabetes mellitus, an overview 4

Diabetes and its complications 8

Association between diabetes mellitus and the periodontium II

Altered immune function in diabetic patients 13

Investigation of genetic biomarkers for periodontal disease 14

Angiogenesis in diabetes mellitus 14

Molecular mechanisms for altered angiogenesis in diabetes mellitus 16

Surgical wound healing in skin vs. oral soft tissues 19

Angiogenesis in the modified McFarlane skin flap model 20

Purpose 22

Hypothesis 22

Specific Aims 22

Materials and Methods 23

Overview of Methods 23

RNA Isolation 25 RNA Amplification and Labeling 26

vi vii

Orthogonal Polarization Spectral Imaging 32

Results 35

Discussion 61

Monitoring of blood flow in diabetic and control rats 61

Temporal and spatial characterization of flaps 62

Genes displaying the greatest up- and down-regulation 63

Baseline Changes 66

Genes changes that may be pertinent to disease entity 66

Correlating gene expression with flap physiology 69

Limitations 73 Future directions 74

Conclusions 75

Appendix A: Cumulative Changes of Significant Gene Changes 78

Appendix B: Volcano plots and scatter plots across all time points 80

Appendix C: Table of angiogenesis-associated genes grouped by function 84

Appendix D: Genes significantly modulated & RNA Checklist 86 Appendix E: Gene modulation across all zones over 7 days 89 Appendix F. Surgical flap model and microarray layout 92 References 93 viii

LIST OF TABLES

Table I. Microarray gene distribution 30

Table 2. Gene regulation in the A vascular Zone (Zone I) 41

Table 3. Gene regulation in the Static Zone (Zone II) 43

Table 4. Gene regulation in Flow Zone I (Zone III) 46

Table 5. Gene regulation in Flow Zone II (Zone N) 52

Table 6. Significantly under-expressed genes across all zones. 54

Table 7. Significantly over-expressed genes across all zones 54

Table 8. Significantly altered genes across all zones (p<0.05; Fold-change: 2::2 or 9 55

Table 9. Significantly down-regulated genes in Zone I & II but not in Zone III & N 56

Table I 0. Genes significantly down-regulated in only the flow zones (Zone III & IV) 56

Table II. Genes significantly up-regulated in only the flow zones 57

Table 12. Significantly altered genes during baseline 58 ix

LIST OF FIGURES

Figure I. Overview of methods 23

Figure 2. Surgical skin flap model measuring 3 x I 0 em 25

Figure 3. Division of the surgical flap into 4 equal zones measuring 3 x 2.5 em 25

Figure 4. Isolation ofmRNA 26

Figure 5. Angiogenic genes and their positions on a microarray 29

Figure 6. Sample of a developed microarray. 30

Figure 7. Determination of gene significance with a volcano plot 32

Figure 8. Cytoscan analysis 34

Figure 9. Determination of avascular and static lengths 34

Figure 10. Bar graph ofthe avascular, static and flow zones in diabetics vs. controls 35

Figure II. Mean total affected lengths over time measured) 36

Figure 12. Mean total affected lengths over time measured 37

Figure 13. Control vs. Diabetic Flaps on Day 0, I, 3 and 7 38

Figure 14. Significant gene regulation per zone across all zones 39

Figure 15. Significant gene regulation per day across all zones 40

Figure 16. Real-Time PCR results of significantly up-regulated genes 59

Figure 17. Pooled and unpooled gene expressions ofVEGFA, TGFBRI and ANGPTI 60 INTRODUCTION

Statement of the Problem

Despite current medical advances, diabetes mellitus remains one of the leading causes of death. Millions of people are affected by diabetes and millions more are either undiagnosed or pre-diabetic. It is well-documented that individuals with this condition have delayed wound healing and an increased susceptibility to infections. Implicated in the pathogenesis of impaired wound healing of diabetics is an abnormal angiogenic response. 1 Diabetic patients have been shown to have less coronary collateral vessel circulation compared to non-diabetic controls? One of the key factors associated with successful wound healing in regenerative periodontal therapy is the ability to achieve an adequate and functional blood supply. Much research has gone into analyzing factors that may enhance the wound healing potential in diabetic patients. However, the underlying genetic basis of these differences in wound healing is still poorly understood.

Important questions to be answered in this study are 1) Is angiogenesis altered in diabetic skin flap healing compared to non-diabetic controls; 2) Is the direction of change toward greater or lesser angiogenesis, and 3) Importantly, what genes are most affected? The modified McFarlane flap helps to generate a predictable skin flap model characterized by

1 2 areas of avascularity, which may also include avascular tissue, an area of stasis and an area with blood flow, thereby allowing us to study the pattern of gene expression in these three physiological zones over time, using a streptozotocin-induced type I diabetic rodent model.

Significance

Angiogenesis, defined as the formation of new blood vessels from the pre-existing blood supply, is a well-orchestrated event fundamental to many physiological and pathological processes such as bone and tissue regeneration, diabetes, ischemic diseases and chronic inflarnmation?· 4 Excessive angiogenesis is a part of the pathology of certain diseases such as diabetic retinopathy and rheumatoid arthritis. Inadequate angiogenesis, on the other hand, occurs in cardiac failure and tissue damage after reperfusion of ischemic tissue. 5 The former group may benefit from therapeutic inhibition of angiogenesis, whereas the latter may see improvement with an enhancement of factors that promote angiogenesis. Anti-angiogenesis therapy is a promising focus of today's research to battle pro-angiogenic diseases, such as in diabetic retinopathy, and is viewed as a desperately needed breakthrough in treatment therapy. Recent studies have shown that angiogenesis is genetically pre-determined. 6 Furthermore, it has been purported that epigenetic changes caused by aberrant DNA methylation, or histone acetylation of anti­ angiogenic genes, may play an important role not only in angiogenesis but also in the

6 9 development of diabetes. " However, to date, no study has examined expression of genes associated with angiogenesis in skin flap healing of diabetics and compared this to 3 controls. Elucidation of the temporal and spatial expression of skin flap healing at the genetic level may provide valuable information that will impact the effect of these genes on a physiological level. By examining the differential pattern of expression of angiogenesis-associated genes in streptozotocin-induced diabetic rats compared to non­ diabetic control rats, we identified key genes regulated in angiogenesis and this knowledge can be used to propel future trials in drug-targeted therapies toward enhanced wound healing. REVIEW OF THE LITERATURE

Diabetes mellitus, an overview

Diabetes mellitus, commonly referred to as diabetes, is a metabolic disorder affecting metabolism of carbohydrates, lipids and proteins. It is a major concern in this

world due to the widespread morbidity and mortality associated with this disease.

Diabetes affects about 21 million Americans and includes over 9% of the adult

10 11 population with greater than 6 million more undiagnosed and even more on the rise. •

Diabetes has been significantly associated with a negative prognosis for free flap head

and neck reconstructive operations (p<0.01). 12 Type 2 diabetes constitutes about 85-90%

of diabetic cases; type 1 diabetes makes up 5-10% of the group, whereas other forms of

diabetes, such as gestational diabetes, and those resulting from pancreatic disease, drug

therapies and endocrine disorders account for the remaining cases. 11

Diabetes is currently classified according to the pathophysiology of the disease.

The currently accepted diagnostic categories provided by the American Diabetes

Association in 2001 include: type 1 diabetes, type 2 diabetes, impaired glucose tolerance,

impaired fasting glucose, gestational diabetes, and other types of diabetes secondary to

diseases of the pancreas, drug therapy, endocrinopathies, infections and genetic disorders.

4 5

Type 1 diabetes, formerly known as insulin-dependent diabetes, is due to cellular mediated autoimmune destruction of pancreatic ~-cells, triggered by an event such as viral infection, culminating in reduced, or loss of, insulin secretion. The exact mechanism by which genetic and environmental factors precipitate the development of type 1 and type 2 diabetes is still largely unknown, although several candidate gene polymorphisms have been linked to disease susceptibility. These polymorphisms have been associated with regulatory regions of a gene sequence, and have served as a starting point for investigation of epigenetic gene regulation in insulin signaling and

8 13 14 inflammation in the shared etiology of both type 1 and type 2 diabetes. • •

As its former name implies, type 1 patients require insulin to survive, and without it, a life-threatening condition, known as ketoacidosis, could develop. Ketoacidosis is characterized by increased acid in the bloodstream, accompanied by an accumulation of ketone bodies that may lead to coma or death. Biomarkers of the autoimmune destruction are usually present and include autoantibodies to pancreatic islet cells, insulin, glutamic acid decarboxylase, or tyrosine phosphatases IA-2/IA-2~. Type 1 diabetes usually has an abrupt onset in genetically predisposed individuals; it is often present in children and adolescents, although 15% to 30% of all cases were diagnosed after 30 years of age. The presence of a family history is more common and occurs predominantly in Whites; these individuals are usually thin or of normal stature. Typical treatment regimens include insulin, diet and exercise.

Type 2 diabetes, also known as non-insulin-dependent diabetes, is characterized by insulin resistance, impaired insulin secretion, or increased liver glucose production.

Unlike in type 1 diabetes, autoimmune destruction of ~-cells does not occur in patients 6 who have type 2 diabetes. Hence, these patients still retain the ability to produce insulin, thereby lowering their chance for ketoacidosis unless the patient's immune system is compromised, as with an infection or stress. Type 2 diabetes usually has a more gradual onset compared to type 1, often being undiagnosed for many years because hyperglycemia appears slowly and frequently without symptoms. In the early stages, pancreatic insulin production is actually increased, to compensate for insulin resistance, before diminishing over time due to the body's secretory demand caused by insulin resistance. Type 2 diabetes more frequently affects the Black, Hispanic, American Indian or Pacific Islanders populations. Obesity, or an increased percentage of body fat, is often seen in patients with type 2 diabetes. Resistance to insulin is in part due to an elevated level of circulating free fatty acids (FFA) derived from adipocytes, which inhibits glucose uptake, glycogen synthesis and glycolysis by increasing hepatic glucose production.

Exercise, pharmacological therapy and diet modifications are seen to be beneficial in lowering insulin resistance. Despite the fact that insulin is not required for survival in type 2 patients, it is often administered as part of usual medical management in conjunction with diet, exercise and oral agents.

The impaired glucose tolerance and impaired fasting glucose categories of diabetes refer to metabolic states that fall between normal glycemia and diabetes and are considered risk factors for the development of diabetes. Those who have impaired glucose tolerance usually have normal blood glucose levels most of the time, and manifest hyperglycemia only after a challenge with a large glucose load. Individuals with impaired fasting glucose have elevated fasting glucose levels but may be normal in a fed state. The production of insulin is normal in a majority of the patients; however, 7 about 30-40% of patients with either impaired fasting glucose or glucose tolerance develop type 2 diabetes within 10 years after initial diagnosis. 15

Gestational diabetes refers to glucose intolerance during pregnancy, and is strongly correlated with insulin resistance. It complicates about 4% of all pregnancies in

the United States and usually occurs during the 3'd trimester of pregnancy. For the

majority of women afflicted with gestational diabetes, a return to normoglycernic levels

takes place after parturition, but their risk for type 2 diabetes is significantly increased. 15

The prevalence of gestational diabetes is higher in women who have a family history of

diabetes, are over 25 years old, obese and members of the ethnic group showing a higher

prevalence for type 2 diabetes (Black, Hispanic, American Indian).

The remaining categories of diabetes may be associated with injuries to the

pancreas, and/or pancreatic disease induced by infection. They may also be related to

genetic defects in the function of pancreatic ~-cells, defects in insulin action or endocrine

disorders. Secondary diabetes may occur due to use of certain drugs such as

corticosteroids.

Classical signs and symptoms of diabetes mellitus are polyuria, polydipsia,

polyphagia and unexplained weight loss. According to the American Diabetes

Association, there presently are 3 ways to establish a diagnosis of diabetes: 1) symptoms

of diabetes plus casual plasma glucose concentration ?:200 mg/dl (?:11.1 mmol/1), with

casual defined as any time of day without regard to last meal intake; 2) fasting plasma

glucose ?:126 mg/dl (?:7.0 mmol/1), with fasting defined as no caloric intake for at least 8

hours; and 3) 2-hour post-load glucose ?:200 mg/dl (?:11.1 mmol/1) during an oral glucose 8 tolerance test, containing an equivalent of 75 g anhydrous glucose dissolved in water.

Normal fasting glucose level is <100 mg/dl (5.6 mmolll). Impaired fasting glucose is diagnosed when the fasting plasma glucose level is ::::100 mg/dl but ::;125 mg/dl (5.6 and

6.9 mmol/1). Impaired glucose tolerance can usually be diagnosed only after an oral glucose tolerance test, when the 2-hour post-load plasma glucose concentration is ::0:140 mg/dl but ::;199 mg/dl (between 7.8 and 11.1 mmol/l). The normal glucose level after an oral glucose tolerance test is <140 mg/dl (7.8 mmol/l).

Another measure of glycemic control comes in the form of the hemoglobin Ale

(HbAlc) test. In erythrocytes, the non-enzymatic reaction between glucose and the hemoglobin protein forms glycohemoglobin. The binding of glucose to hemoglobin is very stable and therefore, once bound, the red blood cell stays in its glycated form during its entire life span of -123±23. days. Normal HbAlc level is <6%, reflecting the average blood glucose level over the preceding 30 to 90 day period of the erythrocyte. The treatment goal for diabetic patients is to attain a HbAlc level of <7%. Physician intervention is recommended when the HbAlc level is >8%. 11

Diabetes and its complications

Insulin is a hormone produced by the beta cells of the Islets of Langerhans in the pancreas; this hormone enables glucose to be absorbed and converted to energy. In the presence of diabetes, insulin resistance and hyperglycemia, inflammation is notably more pronounced, as indicated by higher levels of inflammatory indices and adhesion molecules. 16 Various studies have shown that an upregulation of pro-inflammatory cytokines, such as tumor necrosis factor-a, interleukin-1, and interleukin-6, and vascular 9 adhesion molecules, such as vascular cell adhesion molecule 1 and E-selectin, are what account for the amplified inflammatory response. 17 Tumor necrosis factor-a exerts its effect on inducing insulin resistance by preventing autophosphorylation of the insulin receptor, which subsequently causes second messenger inhibition of the enzyme tyrosine kinase.

Complications of diabetes are thought to arise from the formation of advanced glycation end-products (AGE), or glycated proteins in individuals with sustained hyperglycemia. At the macrovascular level, the deposition of advanced glycation end­ products on collagen, that comprises a large component of the extracellular matrix in the body, causes an accumulation in the blood vessel wall, resulting in a thickening of blood vessels and a narrowing of the lumen. Modified vascular collagen can also contribute to atheroma formation by immobilizing and covalently cross-linking circulating low-density lipoproteins. At the microvascular level, collagen alteration by AGE also occurs in the basement membrane of small blood vessels, causing both an increase in the thickness of the basement membrane and a disruption in the normal homeostatic transport across the membrane. These microvascular complications may manifest as retinopathy, nephropathy or neuropathy. Formation of AGE has been associated with increased production of vascular endothelial growth factor (VEGF), a multifunctional cytokine that induces neovascularization and plays a role in the microvascular complications of

18 19 diabetes, in serum levels and major organ tissues of diabetics. • In human gingival tissues, elevated VEGF expression was also found in diabetic subjects compared to non­ diabetics, illustrating the similarity between the periodontium and other organs affected by diabetes.20 10

At the cellular level, AGE may affect cell-cell, cell-matrix and matrix-matrix

interactions through the receptors for advanced glycation end-products (RAGE). RAGE is found at various locations, such as the surface of smooth muscle cells, endothelial cells, neurons, monocytes and macrophages. The mRNA levels for RAGE showed a

50% increase in gingival tissues of type 2 diabetic subjects compared to non-diabetic

controls.Z1 In a hyperglycemic environment, an increase in vascular permeability and

thrombus formation is thought to be due to both the increased expression of RAGE and

the increased interactions between these receptors and their substrates on endothelium.

In addition to its effect on the endothelium, these interactions may also cause a change on

the cell surface of monocytes by induction of cellular oxidant stress and activation of the

transcription factor, nuclear factor kappa B (NF-KB). Consequently, tlie monocytes and

macrophages are altered; and pro-inflanunatory cytokines such as interleukin-1 {IL-l)

and tumor necrosis factor-a (TNF-a) are produced, contributing to the chronic

inflanunatory processes in formation of atheromatous lesions. A decreased alveolar bone

loss in response to Porphyromonas gingiva/is was shown upon blocking RAGE, resulting

in lower levels of TNF-a, IL-6, and matrix metalloproteinase (MMP) levels in the

gingival tissues and decreasing AGE accumulation in periodontal sites.22 11

Association between diabetes mellitus and the periodontium

The magnitude of tissue destruction and inflammation caused by bacteria is determined not only by the bacterial species themselves, but also can be attributed to the patient's genetic and envirornnental factors. 23 Although it has long been known that diabetics are more difficult to treat, it was only recently that evidence emerged implicating diabetes as a significant risk factor for greater severity of periodontal disease

24 26 with a less predictable response to therapy. " Periodontitis is an infection that has been reported to be twice as prevalent in diabetic individuals compared to non-diabetics.27 In fact, the strength of evidence linking diabetes as a risk factor for periodontal disease is so strong that it has been suggested to be the sixth complication of diabetes, along with the five classic complications which include retinopathy, nephropathy, neuropathy, macrovascular disease and altered wound healing.28

The association between type 1 diabetes and gingivitis/periodontal disease has been demonstrated in several studies. In a study of 85 French adolescents, aged 12-18 years, with type 1 diabetes, and a similar age group serving as controls, it also was demonstrated that in spite of similar plaque levels among the diabetic and the non­ diabetic group, there were twice as many sites with gingival inflammation in children with diabetes, compared to non-diabetic controls.29 In another study of 263 type 1 diabetic patients, compared to 59 non-diabetic siblings and 149 non-diabetic unrelated controls, 13.6% of the diabetic group between 13 and 18 years of age had periodontitis, with an even higher rate of 39% among those between 19 to 32 years of age. In contrast, prevalence of periodontitis in the non-diabetic control subjects was <3%. A separate 12 study showed that the prevalence of gingivitis was also higher in children with type 1 diabetes than in non-diabetic children with similar plaque levels. 30

Not only is type 1 diabetes purported to be linked to periodontitis, but type 2 diabetes also shows a similar association. A multivariate risk analysis showed that type 2 diabetics displayed 2.8- to 3.4-fold increased odds of having periodontitis compared to non-diabetic subjects.31 The Pima Indians from the Gila River Indian Community of

Arizona have the highest incidence of type 2 diabetes in the world. A longitudinal research study found similar odds values in 2,273 Pima Indian subjects aged 15 years of age or older. The prevalence of periodontitis in this group was 60% in diabetic individuals, compared to 36% in those without diabetes. The subjects were followed for

2.5 years, and the incidence of periodontitis was deemed to be 2.6-fold higher in the diabetic subjects than in non-diabetic patients. 32 In a separate 2-year longitudinal study, the odds of having an increased risk of progressive alveolar bone loss was four-fold higher in subjects with type 2 diabetes compared to the non-diabetic group. 33

Studies pertaining to glycemic control and periodontitis have been largely inconclusive, although one large epidemiologic study in the United States showed that adults with poorly controlled diabetes had a 2.9-fold increased risk of having periodontitis compared to non-diabetic adult subjects. In contrast, individuals with their diabetes well-controlled had no significant increased risk of periodontitis.34 However, there are also well-controlled diabetics who maintain excellent periodontal health and others who develop periodontitis, as well as some poorly controlled diabetics without any major complications.35 In the 2-year longitudinal study of Pima Indians, individuals with poor glycemic control of their type 2 diabetes had an 11-fold increased risk of 13 progressive bone loss compared to non-diabetic controls whereas well-controlled diabetic subjects did not show a significant increase in risk. 11

Altered immune function in diabetic patients

Despite the higher risk of periodontal disease in the diabetic population, the subgingival microflora between the diabetic and non-diabetic patients with periodontitis

36 37 is remarkably similar. • The similarity in the microflora suggests that the severity and prevalence of periodontal disease in individuals with diabetes may instead be attributed to differences in the host immunoinflarnmatory response rather than to differences in bacterial species.

In diabetes, neutrophils, monocytes and macrophages have altered immune cell function. 36 Thus, the usual defense mechanisms of neutrophil adherence, chemotaxis, and phagocytosis, that would normally help combat periodontal destruction, are often impaired. The altered function of these immune cells hinders bacterial killing in the

38 39 periodontal pocket and significantly increases periodontal destruction. • Although neutrophil function is suppressed in diabetes, the monocyte/macrophage cell line shows a hyper-responsiveness, resulting in significantly increased production of pro-inflammatory cytokines and mediators.40 The level of tumor necrosis factor-alpha (TNF-u) was elevated in peripheral blood monocytes from diabetic subjects in response to antigens from Porphyromonas gingivalis compared to monocytes from non-diabetic control subjects.40 14

Investigation of genetic biomarkers for periodontal disease

The relationship between genetics and the risk for periodontal disease is under investigation although some studies show promise of using genotype to predict future tooth loss. For instance, specific IL-l genotype positive individuals were shown to have

2.7 times greater likelihood of tooth loss than genotype negatives.41 Suggestive of such studies is the potential to incorporate genetics and biomarkers into diagnosis to better assess periodontal disease susceptibility in order to render appropriate treatment therapy.

Considering the association between diabetes and periodontal disease, a study of the genetic profile of angiogenesis in diabetics may help us better understand why diabetes may be a risk factor for periodontal disease.

Angiogenesis in diabetes mellitus

Angiogenesis is defined as the formation of new blood vessels from the pre­ existing blood supply. It is essential for supplying growing tissue with adequate blood flow and oxygenation during wound healing, menstruation and embryogenesis. Aberrant angiogenesis may lead to a variety of diseases. For example, excessive angiogenesis contributes to diabetic retinopathy and nephropathy, whereas inhibited angiogenesis contributes to impaired wound healing, impaired collateral vessel development, embryonic vasculopathy in pregnancies complicated by maternal diabetes, as well as in transplant rejection in diabetic recipients. In circumstances of deficient angiogenesis, as

in diabetic neuropathy, inducing blood vessel growth may help to improve symptoms. 5

The process of angiogenesis is a well-orchestrated event regulated by numerous

growth factors and cytokines. Enzymatic degradation of the basement membrane occurs 15 first followed by chemotactic and mitogenic factors causing endothelial cells (EC) to extend through the gaps, elongate, align and proliferate to form a capillary sprout.

Joining of two sprouts initiates blood flow in the newly formed loop, and the interaction between EC and pericytes (single cell layers present on the walls of capillaries and fine vessels) trigger the formation of a new basement membrane and leads to vessel

5 42 maturation and stabilization. • Maintaining a delicate balance of stimulatory and inhibitory factors is essential for the normal progression of angiogenesis. Pathological consequences due to. aberrant vessel formation may occur with up- or down-regulation of any of these factors.

Excessive angiogenesis, as occurs in diabetic retinopathy and diabetic nephropathy, is detrimental and based on a well-established "ischemia-angiogenesis hypothesis." This hypothesis states that hypoxia triggers release of vasoproliferative substances in response to reduced blood supply and inadequate nourishment. The resulting newly formed vessels are malformed and show increased permeability to albumin and fluorescein, due to a fragile basement membrane and the absence of tight junctional proteins and pinocytes. Moreover, due to the fragility of the vessels, there is a greater propensity for vessels to rupture, further conferring an enhanced angiogenic response, almost in a vicious cycle. In a similar manner, excessive angiogenesis in diabetic nephropathy causes an enlarged glomerular filtration surface, giving rise to an increased glomerular filtration rate. Nyengaard et al. has shown that there was a 30-50% increase in the surface area of capillaries per glomerulus in streptozotocin-induced diabetic rats compared to age-matched controls, indicative of a heightened angiogenic 16 response in the diabetic state, as both capillary diameter or glomeruli numbers remain unchanged.43

Diminished wound healing capacity has been demonstrated in both clinical and experimental diabetes. Normal healing of a wound proceeds through five phases: hemostasis, inflammation and debridement, proliferation, epithelialization and remodeling. The non-healing nature of a diabetic wound was attributed to the second and third phases: inflammation and debridement together with proliferation.44 In a study involving incisional skin-wound models made on the backs of db/db mice, delayed repair was accompanied by reduced angiogenesis, delayed formation of granulation tissue, decreased collagen content, and low breaking strength as compared with normal littermates. The aberrant angiogenesis was further characterized as having decreased arteriolar number and density, loss of vascular tone and a reduction in the cross-sectional area of new vessel walls. 45

Molecular mechanisms for altered angiogenesis in diabetes mellitus

VEGF, a family of homodimeric heparin-binding proteins that plays a role in regulating the proliferation and migration of endothelial cells forming the basis of any vessel, is reported to have potent proangiogenic activity both in vitro and in vivo. 42 There are four distinct isoforms ofVEGF generated by an alternative splicing mechanism of the

VEGF gene.46 VEGF is regarded as an endothelial-cell-specific mitogen, a chemotactic agent for endothelial cells and monocytes with a permeability factor that is 5 x 104 times as strong as histamine. Intraocular injection of human recombinant VEGF in non-human primates was reported to induce pathological vascular symptoms such as these seen in 17 diabetic retinopathy characterized by vascular tortuosity, microaneurysm-like capillary abnormalities, flame hemorrhages, capillary closure and leakage during angiography.47

In hypoxic conditions, upregulation of the hypoxia-inducible factor (HIF), a transcription

factor that binds to the promoter region of the VEGF gene, was observed.48 When VEGF

binds to its receptors FLT-1 and FLK-1/KDR on the surface of endothelial cells and to a

lesser extent on monocytes and macrophages, a signal transduction pathway is initiated

that culminates in an increase in vascular permeability and endothelial cell proliferation

and migration.3 The expression of VEGF is also increased by epidermal growth factor

(EGF), transforming growth factors alpha and beta (TGFa and TGF~), insulin-like

growth factor 1 (IGF-1), fibroblast growth factor (FGF), platelet-derived growth factors

(PDGFs) as well as others by autocrine or paracrine regulation.42

Upregulation of FGF, a heparin-binding protein with mitogenic, chemotactic and

proangiogenic properties similar to VEGF has also been seen in diabetics. FGF

activation is triggered by endothelial injury and exists in two forms: FGF-1 and FGF-2,

also known as aFGF and bFGF, respectively.49 Inactive FGF is localized in capillary

basement membrane (BM) in a heparin-bound inactive state and can be rapidly released

in response to injury to stimulate neovascularization without the need to undergo a

lengthy transcription and translation process. Integrins are also essential in angiogenesis

as they mediate cell attachment to extracellular components (fibronectin, vitronectin,

laminin, collagen) and promote cellular migration and mobilization of growth factors

needed for angiogenesis. Factors that serve as positive regulators of angiogenesis include

FGF-1, FGF-2, TGFa and TGF~, hepatocyte growth factor (HGF), tumor necrosis factor

alpha (TNFa), angiogenin, interleukin-I and angiopoietins.42 18

Angiopoietin (ANGPT) ligands and their receptor, tyrosine kinase receptor,

TIE2ffEK also participates in the interaction between endothelium and surrounding cells.

Similar to VEGF and its receptor, the TIE/ANGPT signaling system is needed for vascular system development during embryogenesis. Mice deficient in the TIE2 genes through targeted gene mutation introduced by homologous recombination failed to survive beyond day 9.5 to 10.5 of embryogenesis.50 There are two angiopoietin molecules, ANGPTI and ANGPT2, and although they show structural similarity, each

51 52 exhibits a different action on the TIE2-associated signaling cascade. • ANGPTI stimulates TIE2 phosphorylation, but ANGPT2 does not result in activation of its receptor, indicating that ANGPT2 is a competitive inhibitor of ANGPTI. 52

A deficiency in angiogenesis may be due to improper degradation of the basement membrane, alterations in the balance of growth factors and cytokines that regulate vascular stability or problems in signal transduction. Decreased levels of urokinase plasminogen activator (uP A) appear to contribute to impaired degradation of the basement membrane/extracellular matrix. uP A helps to convert plasminogen to plasmin, which promotes angiogenesis by degrading fibronectin, laminin and the proteoglycan protein core by activating matrix metalloproteinases (MMP) and by mobilizing bFGF

3 from the ECM pool. 5 In the diabetic state, there is a decreased level of uP A and supranormal levels of plasminogen activator inhibitor (PAI-l). As a result, collateral circulation is impaired by hindering ECM degradation, consequently preventing the outgrowth of capillaries from pre-existing ones. 54 Thrombospondins (THBS)-1 and -2 are known as anti-angiogenic agents that inhibit endothelial cell proliferation. 55 THBS-2 19 was shown to help limit the extent and duration of edema formation, angiogenesis and inflammatory cell infiltration during acute and chronic inflammation. 56

Surgical wound healing in skin vs. oral soft tissues

The healing of both surgical skin wounds and oral soft tissue injuries includes a

series of highly reproducible and rigidly controlled biologic cascades of events beginning with the chemoattraction of cells that accumulate and debride the wound of injured tissue, foreign material and microbial cells, and culminating with the formation and maturation

of a new extracellular matrix. The extracellular matrix serves to bridge wound margins,

support cells and regenerate vasculature and restore tissue resistance to functional stress.

The epithelial cells on the surface of the wound migrate from the wound margins over the

maturing fibrin clot in the wound space. When the tissue is fully healed, the new

epithelium forms a protective and mechanically resistant barrier, not unlike its

predecessor.

It has been reported that a periodontal wound heals at a similar pace as a

epidermal wound, despite having a calcified, avascular and rigid root surface. However,

wound healing in a periodontal defect following flap surgery is more involved than in

healing of a dermal or mucosal wound. The reason for this is that the gingival flap is

seated against another vascularized wound margin which includes gingival connective

tissue and the alveolar process. Moreover, the avascular, calcified and rigid root surface

opposes the flap. The other complicating factor is the presence of the tooth itself.

Healing of such a wound has been described as maturation of the gingival connective

tissue, regeneration of alveolar bone and cementum and most importantly, 20 epithelialization of root surface. Thus, it has been deduced that for successful healing of a surgical mucosal flap, the epithelium must be prevented from early access to the root surface to allow for connective tissue repair at the root surface-gingival flap interface.57

In spite of these differences, in both locations the healing of skin and oral soft tissue wounds requires the presence of an adequate blood supply.

Angiogenesis in the modified McFarlane skin flap model

Many researchers have sought to uncover effective drug therapy to enhance the salvage of skin flaps and grafts. Topical anti-ischemic agents such as nitroglycerin, trolarnine salicylate and nifedipine, when applied on random skin-flaps of male Sprague­

Dawley rats, produced significant improvements in flap survival.58 Similarly, topical drug application of two angiogenic substances, Nonivarnide and Nicoboxil, tested on a random-pattern dorsal skin flap in Wistar rats also showed a significant decrease in skin flap necrosis.59 Tsai et a!. demonstrated that subdermal injections of Sildenafil given alone results in a significant reduction in the avascular and stasis areas of treated rats.60

Slower, but longer-releasing, site-specific delivery has also been explored with agents such as fibrin glue in combination with Sildenafil, resulting in a highly significant improvement in flap viability. 61 Calcium sulfate, a well-tolerated, biodegradable, osteoconductive bone graft substitute was also observed to promote greater blood vessel density compared to controls in rabbits. 62 Although certain angiogenic factors have been shown to be regulated differently in diabetic individuals, the endogenous expression patterns for these angiogenic factors in the diabetic skin flap model have not been well characterized. 21

The rat dorsal model was originally described by McFarlane for the study of flap necrosis and its prevention. The modified McFarlane model is a caudally based, dorsal flap model which is perfused through two constant sacral axial vessels. 63 The flap consists of the skin, panniculus carnosus muscle and the submuscular areolar tissue.

Inadequate blood supply to the distal part of the flap, away from the pedicle base, creates varying degrees of ischemia. The initial 6 to 12 hours of the flap's interaction with its bed was deemed critical for optimal survival. 64 The distal necrosis rate using the

McFarlane model was found to fall between 25-50%.65

The value of using this flap model to study angiogenesis is based on the fact that the observed survival pattern seems to be based on flap circulation and not support from the bed. 64 Hence, in an attempt to enhance flap survival, many growth factors have been

66 examined. For example, FGF and EGF have been shown to enhance skin flap survival. -

68 Previous studies have found that skin flap survival is significantly reduced in the

12 69 diabetic compared to the non-diabetics rats. • By identifYing the pattern of expression of angiogenesis-associated genes in streptozotocin-induced diabetic rats, we aim to identifY key regulators in angiogenesis that are involved in the diminished wound healing capacity of diabetics and use the knowledge to propel future trials in drug-targeted therapies to enhance wound healing. PURPOSE

The purpose of this study was to determine if a differential expression of angiogenesis-associated genes exists between diabetic and non-diabetic rats as well as to characterize those differences.

HYPOTHESIS

Within a surgically created skin flap model in rats, diabetic animals will show a greater area of skin flap avascularity, which would be reflected by a significant down-regulation of pro-angiogenic genes and a significant up-regulation of anti-angiogenic genes both temporally and spatially.

SPECIFIC AIMS

1. To determine if there is a differential regulation of angiogenesis-associated genes

between diabetic and non-diabetic animals both temporally and spatially within a

surgically created skin flap model.

2. To characterize the differential gene expression between the two groups over time

and with respect to different regions of the flap, reflecting blood vessel flow,

stasis or avascularity.

3. To confirm the gene expression data of several candidate genes from microarray

analysis using Real-Time PCR.

22 MATERIALS AND METHODS

Figure 1. Overview ofmethods

23 24

All procedures were reviewed and approved by the Eisenhower Army Medical

Center Animal Care and Use Committee and were performed in a facility accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care,

International. Research was conducted in compliance with the Animal Welfare Act and

other federal statutes and regulations. An overview of the methods and materials used is

provided in Figure 1. A total of 30 male Streptozotocin-induced diabetic and 38 non­

diabetic control Sprague-Dawley rats (Harlan Sprague-Dawley) weighing approximately

200-370 grams was used for this study (Taconic Labs, Hudson, NY). Dr. Thomas

Johnson performed the experiment with 30 non-diabetic control rats that were previously

analyzed and 8 additional rats were used in this study to make up the baseline group. The

rats were verified to be diabetic based on a blood glucose test of greater than 250 mg/dl.

The rats were housed in an environmentally controlled room, on a 12-hour light-dark

cycle, and were fed standard rat chow and water ad libitum. Animals were housed in

pairs throughout a one-week acclimation period until surgery. Following surgery, the

animals were housed individually to prevent injury to skin flaps from cannibalism and

normal socialization activities.

Modified McFarlane flaps measuring 3 x 10 em were created on the dorsal skin of

the rats (Figure 2). The myocutaneous flaps were elevated from the underlying fascia,

then returned to their native positions and secured with staples. The rats were divided

into four groups based on a terminal evaluation interval of 0, 1, 3, or 7 days. Following

euthanasia, flaps were harvested, photographed and sectioned transversely into four 3 x

2.5 em zones and stored at -80°C (Figure 3). 25

A.

B.

Figure 2. Surgical skin flap model measuring 3 x 10 em on a model rat (A) and with actual surgery shown (B). Adapted from Yang and Morris 1998. 70

Figure 3. Division ofthe surgical flap into 4 equal zones measuring 3 x 2. 5 em

RNA isolation

Total RNA was isolated from the skin flaps, using a monophase phenol and guanidine thiocyanate isolation reagent according to the manufacturer's instructions

(TriPure™, Boerhringer Mannheim: Indianapolis, IN). Briefly, 100 mg of tissue, was immersed in 1 mL ofTriPure, homogenized, and centrifuged at 12,000 x g for 15 minutes at 4°C (Figure 4a). Homogenized samples were incubated for 5 minutes at room 26 temperature, followed by an addition of 0.5 mL of chloroform (Figure 4b). The mixture was vigorously shaken and incubated for another 15 minutes. Samples were then centrifuged at 12,000 x g for 15 minutes at 4°C. After centrifugation, the solution contained three phases (Figure 4c ). The upper aqueous phase contained the RNA, while the lower organic phase and cloudy interphase contained DNA and proteins. Total RNA was recovered from the upper phase by isopropanol precipitation (Figure 4d). The RNA was resuspended in molecular biology grade water, and quantity and purity of the isolated

RNA was analyzed utilizing Bioanalyzer 2100 (Agilent Technologies), then stored at

-80°C until processed. Only RNA samples displaying two sharp peaks on the

Bionanalyzer, without shoulders, for the 18S and 28S ribosomal RNA and showing high grade RNA, were used for microarray analysis.

Figure 4. A. Small pieces of a 3 x 2. 5 em flap being harvested for RNA. B. RNA isolation C. RNA separation into 3 distinct layers and D. RNA pellet

RNA amplification and labeling

Total RNA (25-100 ng) was reverse transcribed into eDNA using the target Amp

1-Round Biotin-aRNA Amplification Kit 104 (Qiagen, Valencia, CA). The resulting eDNA was then used to generate Biotin-aRNA (biotin-labeled cRNA) as instructed by 27 the manufacturer (Epicentre, Madison, WI). The Biotin-aRNA was purified using

Qiagen RNeasy Mini Kit according to the manufacturer's protocol (Qiagen, Valencia, eA) and 2 J.lg of biotin-labeled cRNA was used to hybridize to the Angiogenesis specific oligoDNA arrays.

oligoDNA microarray

Rat angiogenesis arrays (ORN24.2) containing 128 positions and 113 unique angiogenesis-associated oligonucleotide probes, were obtained from SuperArray

(Frederick, MD), in the hybtube format, and were processed according to the manufacturer's instructions (Figure 5). Briefly, the membrane was pre-wet with 5 mL molecular biology grade water. The water was then replaced with 2 mL of pre-warmed

(60°C) GEAhyb hybridization™ solution. The membrane was incubated at 60°e for two hours in a hybridization oven with gentle rotation (8 rpm). After two hours, the pre­ hybridization solution was removed, and 0.75 mL of hybridization solution mix, containing 2 J.lg of the biotin-labeled cRNA, was added. The membrane was returned to the hybridization oven and incubated overnight (approximately 16 hours) at 60°e with continuous rotation (8 rpm). The following day, the membrane was washed at 60°e for

15 min each, with 15 mL of wash solution I (2X sse, 1% SDS), followed by high stringency, wash solution II (O.lX sse, 0.5% SDS). The membrane was then blocked with GEAblocking Solution Q™, at room temperature for 40 min with continuous agitation (30 rpm). After 40 min, the blocking buffer was replaced with an alkaline phosphatase-conjugated streptavidin solution, and the membrane was incubated at room temperature for 10 min with gentle rotation (8 rpm). The membrane was then washed 28 four times for 5 min each, with 5 mL of Wash Buffer F™, then rinsed twice with 3 mL

Buffer G™. One mL of chemiluminescent substrate was then added to the membrane and allowed to incubate for 5-7 min with gentle rotation (8 rpm). The resulting images were captured using the Kodak ImageStation 2000MM system, using the time-lapse image acquisition setting with an image capture every I 0 mins for a total of 40 mins.

(Kodak, New Haven, CT). 29

A. LOi 2 It 3 It 4 II s li_!i=:\i 1 li s I CLJLio_jl li:JC!L:::::JI l,LJLlLJL15_JL1CJ t 11 IL1LJI 19 JLjg:::::JI 21 II 22 IL~L2LJ DCJL_26__jL27 JI=:20L29___ ]Uo=:JO-:LJL32=:J 001 34 IOLJI 36 II 37 1Gs:JL3L]I 40 l t_ 41 JL42_jL 43 _][:=M=:JDOL 46__jc::;iL:=JL4B=:J GOLs.ct=JL~L::JLs_L]I 53 II 54 ILULsLJ LsLJCsLJLs9_ JL~ JL61 .JL 62__j! 113::::::JLGL] L~L]t 66 II 67 II '68 II 69 IL19:::::::JI n II n ! C:::ZCJ[ 74 IUUDCJC 77 'l[.__z---cJL~9=:lL8P=.l I u II ~ II ~ II ~ II ~ II ~ II 67 II ~ I OCJC®!L!!1 lL9LJU"'LJL9LJL9~_J!_ 9_L] I 97 IOU! 99 IL1w:JI 101 11 102 IL]jB:=:JlJQLJ CiQs=::JLi_o0C1.9l_jDQS.::JLJoLJLiiLJLuCJLm:::J LiiL:J 114 ILULJI u6 II m IL:i1UI 11.9 II 120 l ,---m-u m i! m IC 124 u m I026 11 m 11 128 J

B. __ E~gl _j Efnal _• Efna~_ ~ Efna3 lgf Eng Epasl Ereg F2 _ "_ F~l ___ Fgn_ ~gf6_ Fgfr3 _ -~i&f Fl!l ____ FzdS Gna13 ; Hand2 • Hgf _ • Hilla lfnal lfng lgfl 1110 1112a 1118 lllb . 116 ltgav__p .ltgb3 Jagl Kdr ------4'--- ' -'"--+'-''""---,-,....-----"·----,.."'~------~------lamaS ~- lectl _ lep • l~_C299339__ Mapk14_ _ r;1dk_ _Mlllpl9_p "- _MJ!IP2 _ Mmp~ ____ t'!_o!d1_4 Npp~ _ Nprl _ N(Jll _ _ Nrp2 Nudt6 Pdgfa _ Pd&fb _ +_ Pecam __ Cxcl4 _ Pgf Piau Pig Pofutl Prok2 Pten · Ptgsl Ptgs2 Ptn Sapsl__p . Serpinfl Sh2d2a SmadS -- --.... +- . ------~4 ., - , ...."~""--- ,_ - -·~ ~ -"·- Sphkl . Stabl..P Stab2 Tbxl__p Tbx4__p Tek Tgfa Tgfbl - " .. --;:-- - ~------.- -- - "" ...... T&fb2 Tgfb3 Tgfbrl Thbs2 Tiel iimpl • Timp2 __ Timp3 _ 1- --·- ..._ - , T - • rlmyrss6_1l_: __Tn~--. ~nfrsfl~~-- _Tn!~f!2 _TnfsflS Tnntl _ _vegfa Vegfb. y~gtc --+ _Wasf2_ : PUC!t ,__ }l_?nJ , Blank ~1R2 ASlRl A_~l _ Rpl32 ' ldha Aldoa ' Aldoa Gapdh Gapdh BAS2C BAS2C

Figure 5. A. The 128 angiogenic gene positions on an array B. Their corresponding genes, where (_y) denotes a predicted gene.

Microarray distribution is shown in Table I and a sample of a developed microarray is shown in Figure 6. 30

II

Ill

IV

Table 1. Microarray distribution showing 5 microarrays for Zone I-IV on days 1, 3 and 7 and 3 microarrays for Zone I-IV on day 0. Each zone represents 3 x 2.5 em ofthe flap. (DB = diabetics; C = controls)

Figure 6. Sample of a developed microarray that was uploaded onto the GEArray website for gene expression analysis.

Real-Time PCR

To validate the microarray analysis, quantitative two-step real time PCR was performed on 3 genes using the IQ5 Real Time PCR System (Biorad, Hercules, CA). All primers for the genes were purchased from Qiagen (Valencia, CA.). The primers used were Angiopoietin 1 (ANGPTl), Thrombospondin 2 (THBS2) and Vascular endothelial 31 growth factor A (VEGF A). One 11g of total RNA was reversed transcribed into eDNA

using the QuantiTech reverse transcription kit (Qiagen, Valencia, CA). A ten-fold

dilution series of eDNA was done for each gene to determine PCR efficiency with a

standard curve. One hundred ng of the eDNA was used for each primer pair in the real­

time PCR experiments, using the Quantifast SYBR Green PCR kit, according to the

manufacturer's protocol (Qiagen, Valencia, CA). Each sample was run in triplicate. The

cycling conditions were as follows: 5 minute initial PCR activation step at 95°C,

followed by an amplification step which included denaturation at 95°C for 10 sec

followed by annealing at 60°C for 30 sec (40 cycles). To verify that a single specific

product was amplified, melting curve analysis was done. Results were presented as a

ratio of each experimental gene to the expression of Hypoxanthine-guanine

phosphoribosyltransferase (HPRT) housekeeping gene.

Data and Statistical Analysis

Web-based software (GEArray Analysis Suite 2.0, SuperArray BioScience Corp.:

Frederick, MD) was used to analyze the signal intensity produced for each gene. To

account for background differences among microarrays, the lowest average density spot

on each array was determined and the average intensity across that spot was used as a

background correction factor. Interquartile normalization of the signal intensity was

utilized, ranking the absolute intensity of each signal then dividing it by the mean

intensity of the genes that fell between the 25% and 75% quartile range.

All results were presented as mean values ± SEM. Differences among groups

were assessed by using Student's t-test or one-way analysis of variance (ANOVA). 32

Measurements were analyzed usmg SigmaStat 3.5 (Systat, Point Richards, CA).

Statistical significance was accepted at p

Group 19 vs. Group 1

+ P= 0.05 ++ + + + + +1-

+

-s ~ -l -2 2-Fold Decrease Lov2 <&roup 19/ croup 1)

Figure 7. Determination ofgene significance: Genes in the green area are significantly under-expressed, and genes in the red area are significantly over-expressed in the diabetic, byafold-change of2 or more (p<0.05).

Orthogonal Polarization Spectral Imaging

Determination of the distance from the distal end of the flap to the areas of stasis and avascularity on days 1, 3 and 7 was done through orthogonal polarization spectral imaging using a cytoscan (Figure 8). The procedure has been previously described by

Tsai et al. 60 To better facilitate visualization of the blood vessels, a chemical depilatory 33

(Nair) was used to remove hair on the dorsal skin of the rats prior to measurements. The mean avascular lengths were measured from the base of the flap to the area devoid of blood vessels, averaged over 3 measurements, 1 measured from the center of the flap and

2 measured from the left- and rightmost borders of the flap (Figure 9). Similarly, the mean static length was calculated as the distance from where the avascular area ends to where blood vessels were intact but failed to display flow, averaged over the 3 regions.

The mean affected length was the sum of the mean avascular plus the mean static length, otherwise known as the area without flow. 34

A. Avascular B. Stasis C . Flow (Necrotic)

Figure 8. Cytoscan analysis showing the images seen for a section ofthe flap displaying avascularity (a), stasis (b) or flow (c).

Figure 9. The avascular and static lengths for each zone were determined by taking the average of three measurements - two from the lateral and one from the centermost portion ofthe flap using cytoscan analysis. For example, A, B and C represent the three avascular lengths. Flow lengths were calculated by deducting the sum of the avascular and static lengths from 10 as the remaining portion of the 10 em flap was comprised of flow. RESULTS

Visual inspection of all flaps showed that the mean avascular lengths increased over time in both diabetic and non-diabetic rats. An overall view of the three zones over

7 days is shown in Figure 10. The mean avascular lengths increased over time from 1.81 to 2.91 em of the flap in the control animal over 7 days (Figure 11). In the diabetic rats, the mean avascular lengths increased from 0.14 to 3.76 em of the flap. On baseline, day

0, no avascularity was noted among the two groups.

Figure 10. Bar graph ofthe avascular, static andflow zones in diabetics and controls on days 1, 3 and 7 with a representative flap oriented to the right. Significant differences were noted on days 1 and 7 (P < 0. 05; Fold-change 2::: 2).

35 36

Mean Avascular Lengths

4.50 * * e 4.00 ~ 3.50 ;.. "0 ";.. = 3.00 =.. 2.50 ~;.. .. 2.00 z" E 1.50 ....=... ".. 1.00 ~= .~ 0.50

= 0.00 Day 1 Day3 Day? Time

Controls SEM Diabetics SEM Day 1 1.81 0.28 0.14 0.10 Day3 2.31 0.16 2.17 0.30 Dav7 2.91 0.18 3.76 0.31

Figure 11. Mean total avascular lengths over time measured from the avascular border (em) (Note : '*'indicates significant difference (p<0.05) between control and diabetic).

The mean total affected length (sums of avascular plus static lengths) was consistently greater in the diabetic than in the controls (Figure 12). From day 1 to day 7, the total affected length increased from 2.72 em to 3.43 em in the controls and from 3.90 em to 4.55 em in the diabetics. The changes were significant on both days 1 and 7 between the two groups. 37

Mean Affected Lengths

6.00 * * ,-.. E 5.00 ~.. 'E"" Q 4.00 .... =..! =... 3.00 ..."'..s < E 2.00 ~.. ...c s 1.00 i:S"'

0.00 Day 1 Day 3 Day7 Time

Controls SEM Diabetics SEM Day 1 2.72 0.26 3.90 0.10 Day3 3.42 0.19 3.97 0.24 Day 7 3.43 0.14 4.55 0.35

Figure 12. Mean total affected lengths (the sum ofavascular and static lengths) over time measured from the avascular border (em) (Note: * indicates significant difference (p

Photographs of representative skin flaps on days 0, 1, 3, and 7 are shown in

Figure 13 for both the diabetic and controls groups. 38

Day7 'E ,...,-! ~i c " I ·~ 0 (.) ,~ - I

u i .CII Ill 6

Figure 13. Control vs. Diabetic Flaps on days 0, 1, 3 and 7.

Angiogenesis-associated gene expression profiles were generated for zones I-IV at 0, 1, 3, and 7 days (Appendix A). Flow zone I is the zone closest to the avascular area whereas flow zone IV is the zone closest to the pedicle flap. Volcano plots and scatter plots were generated to illustrate the change in gene expression in the diabe~ic group compared to the control group, at specified time intervals and tissue zones (Appendix B).

Angiogenesis-associated gene expression profiles for the diabetic and control groups show unique expression profiles for all experimental groups compared to the control, at all 4 time points following surgery.

Genes were separated into 5 maJor categories according to their function: adhesion molecules, cytokines and chemokines, growth factors and receptors, protease inhibitors and matrix proteins and transcription factors and other genes (Appendix C).

Only genes with a statistically significant (p<0.05) change in expression on days 0, 1, 3 or 7 days, and a 2-fold change in either direction post-surgery were examined. In all 4 zones, statistically significant genes were shown to be up-regulated and down-regulated across all time points. The greatest number of significantly modulated genes occurred in

Flow Zone III (73 genes modulated), and the least in the avascular zone, Zone I (34 genes 39 modulated) (Figure 14). The greatest number of up-regulated genes also occurred in

Zone III (44 genes) and the least number in Zone I (14) (Figure 14).

Gene regulation across zones

80

70

60

50

2::.... 40 r::: Up/Down ~ It: 30 • Down-regulated only ~ 20 • Up-regulated only 0"' I 10 ~ 0 Zone I Zone II Zone Ill Zone IV

Zone L::::l ------' Zone I Zone II Zone Ill Zone IV Up-regulated only 14 19 44 28 Down-regulated only 20 19 26 22 Up/Down 0 1 1 1 Down/Up 0 0 0 0 Total 34 39 71 51

Figure 14. Significant gene regulation seen across all zones in graphical and tabular form. Order ofgreatest gene changes by zone: Zone Ill> Zone IV> Zone II> Zone 1

Further delineation showed that the greatest gene modulation occurred on day 3 in the flow zone (Zone III), and the least on day 0 in the avascular zone (Zone I) (Figure

15). 40

Significant gene regulation per day across each zone 40 35 30 25 20 • Zone I 15 • Zone II 10 5 • Zone III 0 ...A-.-._.___._ .__ Zone IV I Up 1 Down Up joown l Up loown Up loown I DayO Day 1 I Day 3 Day7 I Day

DevO O.Vl oaV3 O.V7 Up Down Up Down Up Down Up Down Zone I 1 3 11 18 1 2 3 3 Zone II 0 1 13 11 3 9 3 4 Zone Ill 1 3 8 10 36 21 6 15 Zone IV 0 0 9 5 19 5 13 18 Total 2 7 41 44 59 ~ 25 40

Figure 15. Significant gene regulation per day across zones I-IV according to whether the gene was significantly up- or down-regulated in both graphical and tabular form. Greatest up-regulation ofgenes was seen on day 3 Zone III (36 genes).

Angiogenesis-associated gene regulation in the Avascular Zone (Zone I)

A total of 34 genes were significantly modulated in Zone I (p<0.05 and fold- change of:::::2 in either direction) (Table 2). Expressions of six of the 34 modulated genes were significantly changed on more than one day. Genes significantly down-regulated on more than one day included the following: thrombospondin 2 (THBS2) and interferon- alpha 1 (IFNA1) on days 0 and 1, and tumor necrosis factor ligand superfamily member 41

12 (TNFSF12) and troponin Tl, skeletal, slow (TNNTI) on days 1, 3 and 7. The two genes significantly up-regulated on more than one day included: angiopoietin 1

(ANGPTI) on days 3 and 7 and connective tissue growth factor (CTGF) on days 1 and 7

(Table 2).

Chemokine motif) ligand 9 i i

lnterleukin 12a

I II I

·2.28 7 U II

Table 2. Angiogenesis-associated gene regulation in the Avascular Zone (Zone I) 42

Angiogenesis-associated gene regnlation in the Static Zone (Zone II)

Expression of a total of 39 genes were significantly modulated in Zone II (p

Eight genes were significantly modulated on more than one day. Genes significantly down-regulated on both days 1 and 3 include: cadherin 5 (CDHS), THBS2, interleukin 12a (IL12a), neuropilin (NRPl), TNFSF12 and TNNTl, the latter which was significant on 3 days - days 1, 3 and 7. The one significantly up-regulated gene that occurred on more than one day was endothelial differentiation sphingolipid G-protein­ coupled receptor 1 (EDGl) on days 1 and 3. One gene, platelet/endothelial adhesion molecule (PECAM), made a switch from being significantly up-regulated on day 1 to later being significantly down-regulated on day 7. The greatest gene fold-change in this study was seen in this zone with TNFSF12 having a value of -44.64 down-regulation on day 1. 43

Table 3. Angiogenesis-associated gene regulation in the Static Zone (Zone II)

.. Angiogenesis-associated gene regulation in Flow Zone (Zone Ill)

A total of 71 genes, the greatest out of all 4 zones, were significantly modulated in Zone III (p

There were four genes in this zone that were down-regulated on all 3 days, days 1, 3 and

7 and these include: IL12A. NRPI, TNNTI and SH2 domain protein 2A (SH2D2A).

Nine genes were significantly down-regulated on more than one day: THBS2 on day 3 and 7, epiregulin (EREG) on days 3 and 7, jagged I (JAGI), natriuretic peptide receptor 1

(NPRI) on days 1 and 7, TNFSF12 on days 3 and 7, matrix metallopeptidase 9 (MMP9), tissue inhibitor of metalloproteinase 2 (TIMP2) on days 3 and 7, tissue inhibitor of

metalloproteinase 3 (TIMP3) on days 0 and 3 and MAD homolog 5 (Drosophila)

(SMAD5) on days 3 and 7.

Three genes were significantly up-regulated on all3 days, days 1, 3, and 7, which

included: chemokine (C-X-C motif) ligand 10 (CXCLIO), fibroblast growth factor 1

(FGFI), and procollagen, type IV, alpha 3 (COL4A3). Genes significantly down­

regulated on greater than one day include: chemokine (C-X-C motif) ligand 2 (CXCL2)

on days 3 and 7, interferon gamma (IFNG) on days 1 and 3, angiopoietin (ANGPT) 2 on

days 1 and 3, EDGI on days 3 and 7, fibroblast growth factor 6 (FGF6) on days 3 and 7,

hepatocyte growth factor (HGF) on days 1 and 3, thymoma viral proto-oncogene 1

(AKTI) on days 1 and 3, heart and neural crest derivatives expressed transcript 2

(HAND2) on days 3 and 7 and WAS protein family, member 2 (W ASF2) on days 3 and

7. The greatest gene up-regulation was seen in this zone with ANGPT2 at a 16.86 fold­ change on day 3.

ANGPTI showed a switch from baseline in being significantly down-regulated to

later being significantly up-regulated. A gene labeled as similar to tumor necrosis factor, 45 alpha-induced protein 2 (LOC299339) was seen to be significantly down-regulated in this zone across all 4 days. 46

supetfamily, member 21

Table 4. Angiogenesis-associated gene regulation in Flow Zone I (Zone III) 47

Angiogenesis-associated gene regulation in the Flow Zone II (Zone IV)

A total of 51 genes were significantly modulated in Zone IV (p<0.05 and fold­ change of 2 or greater) (Table 5). From all the genes that were down-regulated in Flow

Zone II (Zone IV), none were from baseline, 5 of them were from day 1, 5 from day 3 and 18 from day 7.

There were four genes in this zone that were down-regulated across on all 3 days, days 1, 3 and 7, IL12a, EREG, TNFSF12 and TNNTl. Three genes were down­ regulated on both days 3 and 7 which include the following: NRPl, tumor necrosis factor-like, alpha induced protein 2 (LOC299339) and SAPSl. Two genes were down­ regulated on both days 1 and 7: THBS2 and TIMP2. No genes were found to be significantly down-regulated on days 1 and 3.

From the genes that were up-regulated in Flow Zone II (Zone IV), none were from day 0, nine were from day 1, 19 from day 3 and 13 from day 7. Four genes were up-regulated on all3 days, days 1, 3 and 7 which included: chemokine (C-C motif) ligand

2 (CCL2), VEGFA, COL4A3, AND EFNA2. Five genes were up-regulated on both days

3 and 7 which were IFNG, BAil, fibroblast growth factor receptor 3 (FGFR3), VEGFC and WASF2. One gene was up-regulated on both days 1 and 3 which include: CXCL2.

Two genes were significantly up-regulated on days 1 and 7 and these were CXCL10 and

AK.Tl.

The pattern of expression of Transforming growth factor, beta receptor 1

(TGFBRl) shifted from being significantly up-regulated on day 1 to being significantly down-regulated on day 7 in this zone. However, on day I, TGFBRl was also shown to 48 alpha-induced protein 2 (LOC299339) was seen to be significantly down-regulated in this zone across all 4 days. 49

~II II .I ».1~09 Aggf1 Angiogenic factor with G patch and FHA dom.11ins 1 3.47 3 -~ NM_033230 Akt1 lhymoma VIral proto-oncogene 1 2.40, 5.60 :1.. 3 NM_053599 E:tna1 Ephli"nA1 -· 6.4 _3__ XM_234903 ~------EPhrinA2 ------u.--39--- _____ 3__ - ,q.,.,_574979------Etna3 Ephrln A3 ------7.48-- - 3 NM_023090 Epas1 EndotheitaiPAS

Table 4. Angiogenesis-associated gene regulation in Flow Zone I (Zone III) so

Angiogenesis-associated gene regulation in the Flow Zone II (Zone IV)

A total of 51 genes were significantly modulated in Zone IV (p

Zone II (Zone IV), none were from baseline, 5 of them were from day 1, 5 from day 3 and 18 from day 7.

There were four genes in this zone that were down-regulated across on all 3 days, days 1, 3 and 7, IL!2a, EREG, TNFSF12 and TNNTI. Three genes were down­ regulated on both days 3 and 7 which include the following: NRPI, tumor necrosis factor-like, alpha induced protein 2 (LOC299339) and SAPS!. Two genes were down­ regulated on both days 1 and 7: THBS2 and TIMP2. No genes were found to be significantly down-regulated on days 1 and 3.

From the genes that were up-regulated in Flow Zone II (Zone IV), none were from day 0, nine were from day 1, 19 from day 3 and 13 from day 7. Four genes were up-regulated on all 3 days, days 1, 3 and 7 which included: chemokine (C-C motif) ligand

2 (CCL2), VEGFA, COL4A3, AND EFNA2. Five genes were up-regulated on both days

3 and 7 which were IFNG, BAil, fibroblast growth factor receptor 3 (FGFR3), VEGFC and WASF2. One gene was up-regulated on both days 1 and 3 which include: CXCL2.

Two genes were significantly up-regulated on days 1 and 7 and these were CXCLIO and

AKTl.

The pattern of expression of Transforming growth factor, beta receptor 1

(TGFBRl) shifted from being significantly up-regulated on day 1 to being significantly down-regulated on day 7 in this zone. However, on day I, TGFBRI was also shown to 51 be significantly down-regulated in the avascular zone but up-regulated in the two flow zones on day 1 and 3. 52

II II I

Table 5. Angiogenesis-associated gene regulation in Flow Zone II (Zone IV) 53

Genes significantly modulated across all 4 zones

Seven genes were identified to be significantly down-regulated across all zones at various time points and 4 genes were identified to be significantly up-regulated across all time points (p<0.05; fold-change: 2 or greater) (Tables 6 and 7).

Genes significantly down-regulated across all 4 zones

THBS2 was significantly down-regulated across all 4 zones: Avascular zone - days 0 and 1, Static zone- days 1 and 3, Flow zone I- days 1, 3 and 7 and also Flow zone II - days 1 and 7. IL12A was significantly down-regulated in the following:

Avascular- day 1, Static- days 1 and 3, Flow zones I and II- days 1, 3, and 7. EREG was significantly down-regulated during the following times and regions: Avascular - day 1, Static- days 1 and 3, Flow zone I- days 3 and 7, and Flow zone II- days 1, 3 and

7 (Table 6).

TNFSF12 was also significantly down-regulated across all time points: Avascular

-days 3 and 7, Static- days 1 and 3, Flow zone I- days 3 and 7, and Flow zone II­ days 1, 3 and 7. TNNTl was also significantly down-regulated on all 3 days (days 1, 3 and 7) across all zones (avascular, static and both flow zones). 54

-3.1l9 -5.32 1,.3, 7 I rnterleukin 12a -' :10~ 1 ~i-4.z~ 1,3 -5.D1 i.3,7 _,_ _._. MAD homolog 5 (Drosophila) .. 1 -2.07 1 -2.90,-2.10 3,7 7

sH.2 ~airi1aln- pi'ateih 2A .-2.08 1 -3.21 1 -3.85.. -.us. -295 1.3, 7 ,3,;7

Thrombospondin 2 -2..04,-9.06 0,1 -4.01,-3.44 ... -LBB, -3.74, -2.94 1.3, 7 -Lt#.~-3A9 1,7

Traponin tl, s-keletal, slow -3.54,·_-2.53, -3.!5 1.3.7 -.!.89,' -ZBS, -3.19 1,3,7 -2.01. .:.t.18 1.3,7

-25.28,. -&.2S, -8..09 1. 3, 7 -44.64.-ll.l ... -J0.94,-2S.06 3,7

Table 6. Significantly under-expressed genes across all zones (p

Significantly over-expressed genes across all zones

VEGFA was significantly up-regulated across all four zones: Avascular- day 1,

Static - days 1, Flow zone I - day 3 and also Flow zone II - days 1, 3 and 7. EFNA3 was significantly down-regulated in the following: A vascular - day 1, Static - day 1, Flow zones I -day 3 and Flow Zone 2 -day 1. PDGFB was significantly up-regulated during the following time points: Avascular Zone- day 7, Stasis Zone- day 1, Flow Zone- day

3 and Flow Zone II- day 3 (Table 7).

CDagulation faciDr II :1.99 1 3.35 3 3.94 3 4.21 3 . Ephriri A3 2.15 1 1 .7.118 3 3.59'····. . .i

2.92 7 1:99 1 2A2 3 2:80 3 2.37 1 1 * 3 Table 7. Significantly over-expressed genes across all zones (p

Genes with altered modulation

Both endothelial cell growth factor l(ECGFI) and coagulation factor II (F2) were both significantly altered across all zones at various time points. The shift in expression occurred for both of these genes; ECGFI was expressed at a significantly higher level in diabetic rats at baseline, then was significantly lowered after baseline. The opposite occurred with F2, which was expressed at a significantly lower level in diabetic rats at baseline and later up-regulated after baseline (Table 8).

Avascular/Nccroticl . I . I . I zone Stas1szonc How zone (I -3) How zone (11- 4) Name Jl F~ld-Change II Day II FoldChan&e II Day II FoldCliange-Jj Day 11-FoldChange, Jl-oay t f EndOthelialceliiro~·thfactorl j; 2.7! - ·:· -~ ---~: -8.99 ;Lr- ~ ll-5.97,-1.91.~~·p1,3,7 1:1 ~~- - !J 7 J L __ (p~~el~t-d_e]lved} ! _.. .:t ___ j._ ____ •••. J ·----~L-... __ . .JI ____i_ _...... ;L. I Coogulationfactorll II ·1.70,1.99 181 335 ~~~~ 3.94 101 4.21 JI:=J

Table 8. Significantly altered genes across all zones (p<0.05; Fold-change: :::2 or 9).

Genes significantly up-regulated only in Zone I & II but not Zone III & IV

No gene was seen to be significantly up-regulated in only Zone I & II and not in the two flow zones. 56

Genes significantly down-regnlated only in Zone I & ll but not Zone m & IV

The only genes to be significantly down-regulated in only Zone I & II and not in

Zone III & IV were CDH5 (Zone I- day 1, Zone II- days 1 and 3), ITGAV (Zone I­ day 1, Zone II- day 1) and TMPRSS6_PREDICTED (Zone I- day 7, Zone II- day 7)

(Table 9).

Table 9. -Significantly down-regulated genes in Zone I & II but not in Zone III & IV

Genes significantly down-regnlated in only the flow zones (Zone III & IV but not

Zone I & II)

Genes showing significant modulation in only the flow zones but not in the avascular or static zones include: interleukin 18 (1118) (Flow Zone I - day 3 and Flow

Zone II- day 7), NPR1 (Flow Zone I- day 1 and 7, Flow Zone II- day 7}, MMP9 (Flow

Zone I - days 1 and 7 and Flow Zone II - day 7), TIMP2 (Flow Zone I - day 3 and 7;

Flow Zone II- days 1 and 7) and TIMP3 (Flow Zone I- days 0 and 3; Flow Zone II­ day 7) (Table 10).

Table I 0. Genes significantly down-regulated in only the flow zones (Zone III & IV) 57

Genes significantly up-regulated in only the flow zones (Zone III & IV but not Zone

I & II)

The following genes were significantly up-regulated in both the flow zones only and not in the avascular or the static zones: CXCL2 (Flow Zone I - day 3 and 7; Flow

Zone II- day 1 and 3), fibroblast growth factor receptor 3 (FGFR3) (Flow Zone I- day 3 and Flow Zone II- day 3 and 7), VEGFC (Flow Zone I- day 3; Flow Zone II- day 3,

7), alanyl (membrane) aminopeptidase (ANPEP) (Flow Zone I & II - day 3), Tissue inhibitor of metallopeptidase 1 (TIMP1) (Flow Zone I -day 3, Flow Zone II -day 7), angiogenic factor with G patch and FHA domains 1 (AGGF1) (Flow Zone I & II on days

3), AKTl (Flow Zone I - days 1 and 3, Flow Zone II - days 1 and 7), HAND2 (Flow

Zone I- days 3 and 7, Flow Zone II- day 3), protein 0-fucosyltransferase 1 (POFUTl)

(Flow Zone I - day 3 and Flow Zone II - day 7) and WASF2 (Flow Zone I & II on day 3 and 7) (Table 11 ).

2.40, 5.80 3.03.2.15 2.60, 2.39 3.7 ~02 .,, 3 ~07 7

Table 11. Genes significantly up-regulated in only the flow zones 58

Angiogenesis-associated gene regulation differences during baseline

Nine genes were seen to be significantly different amongst the diabetics compared to the controls at baseline (Table 12). Seven were significantly down-regulated and two were significantly up-regulated. These nine genes included: THBS2 (-2.04 in Zone I),

CXCL4 (-2.12 in Zone II), IFNA1 (-2.58 in Zone I), ANGPTl (-2.24 in Zone III),

BAil _pREDICTED (-2.36 in Zone I), frizzled homolog 5 (Drosophila) (FZD5) (2.21 in

Zone III), ECGF1 (2.71 in Zone I), TIMP3 (-2.57 in Zone III) and LOC299339 (-2.28 in

Zone III).

i i i

Table 12. - Significantly altered genes during baseline. Blue highlight = significantly down-regulated genes; Red highlight =significantly up-regulated genes

Real-Time PCR

Real-Time PCR data verified that the diabetic group had a greater expression of

VEGFA, TGFBR1 and ANGPTl on day 3 Zone III as seen with microarrays and a lesser amount of TNFSF12 on day 0 Zone I and IV in the pattern similar to the others time and zones (Figures 16 and 17). 59

Figure 16. Real-Time PCR results of significantly up-regulated genes relative to HPRT housekeeping gene (p<0.05). 60

Average RT-PCR Expression of Average Pooled RT-PCR VEGF Expression ofVEG F -~ .£ c 1.60 "'c 150 "c 1.40 c" 100 - 1.20 >SO - I. CO - c c zoo 0.80 '".5!!' '".5!!' LSD 0.60 "'c "' 1.00 .. 0.40 " 020 ~ 050 ::E o.oo ::E 000 CONTROL DIABETIC P-CON P·OB Average RT-PCR Expression of Average Pooled RT-PCR .£ TGFBRl .£ Expression ofTGFBRl 15 "'c 1.2 "'c " " - - >S -= 0.8 -= c c '".5!!' 0.6 '".5!!' LS 00 "'c 0.4 c .. .. OS 0.2 ::E" ::E" 0 P-

AVERAGES SEM CONTROL DIABETIC CONTROL DIABETIC Average RT ~PCR Expression of VEGF 0.13 1.11 0.00 0.32 Average RT·PCR Expression of TGFBR1 0.36 0.91 0.12 0.20 c. Average RT-PCR Expression of ANGPT1 0 58 0.95 0.08 0.18

AVERAGES SEM CONTROL DIABETIC CONTROL DIABETIC Average RT..PCR Expression of VEGF 0.13 1.11 0.00 0.32 Average RT-PCR Expression of TGFBR1 0.36 0.91 0.12 0.20 D. Average RT-PCR Expression of ANGPT1 0.58 0 95 0.08 0.18

Figure 17. Gene expressions of VEGFA, TGFBRJ and ANGPT1 verified using Real­ Time PCR in an un-pooled (A) and pooled sample (B) with their respective values (C &D). DISCUSSION

Monitoring of blood flow in diabetic and control rats

Diabetics are known to have compromised wound healing, and angiogenesis plays an important role in the healing process. In our study, we attempted to uncover, through microarray analysis, how genes that have been reported to play a role in angiogenesis are modified in uncontrolled diabetics compared to normal controls using a rodent model.

Microarray analysis has served as a very useful tool for helping to dissect the biological mech amsms. o f angiogenesis. . . 71 An understanding of the angiogenesis-related gene profile differences between the two groups would enable us not only to identifY genes of interest that may account for the compromised wound healing state in diabetics but also guide us in a direction toward new therapies aimed at reducing these differences.

Using orthogonal polarization, the three different zones in the flap were characterized: flow (area with red blood cell movement), stasis (area with cessation of red

blood cells) and avascularity (area devoid of blood vessels). It was shown by Olivier that

orthogonal polarization is a direct, non-invasive modality to reliably measure microcirculation in random pattern skin flaps and that the complete absence of blood

flow may predict future necrosis. Hence, the area of stasis may be amenable to therapeutic intervention before necrosis occurs. 72 Tsai et al. showed that the area of stasis

61 62 showed that the area of stasis is responsive to phannacological intervention. 60 If no phannacological intervention was rendered the area of avascularity will increase at the expense of the stasis area.

Temporal and spatial characterization of flaps

In the diabetic modified McFarlane skin flaps, within the first 24 hours, the avascular portion of the flap was significantly smaller than that seen in the normal animal

(0.14 em in diabetics vs. 1.81 em in normal controls). By day 3, the mean length of avascularity in diabetics approached that seen in the controls and surpassed it by day 7

(3.76 em in diabetics vs. 2.91 em in normal controls). The significantly lower amount of avascularity in diabetics within the first 24 hours may be suggestive of diabetics displaying a heightened pro-angiogenic state at baseline that favors blood vessel formation. This observation is supported by a significantly higher expression of genes that promote angiogenesis (VEGFA) and significantly lesser amount of expression in genes that suppress blood vessel formation (THBS2, TNNTI) as seen on day 1, which is at first seen to be protective at the initial stages with lesser avascularity noted but is later shown to be incapable of sustaining flap viability at later stages as demonstrated by the increased amount of avascularity by day 7.

The avascular area located at the distal-most portion of the flap was next to an area of stasis exhibiting vessels without flow, and an unaffected area clearly showing vessels with blood flow. The total affected length, which is the sum of the static plus the avascular length, was greater in the diabetic group on all 3 days, and the difference was significant on both days 1 and 7. The mean total affected length was less than 5 em for 63 both groups. Similarly, the avascular or static areas of the flap, on average, did not exceed more than the first 5 em of the flap. Hence, a large majority of the flap, as depicted by the last two zones (Zone III and IV), exhibited flow across all time points.

Assimilating gene expression data from the microarrays together with flap physiology attained from cytoscans, suggests that the combination seen in the latter half of the flap closest to the pedicle may represent a successful balance of factors that helps to promote angiogenesis and increase flap survival. In contrast, gene expression seen in the first half of the flap may represent a combination of factors that fails to promote flap survival, particularly in the avascular region of Zone I. Factors such as hypoxia and distance of the flap from the blood source must also be taken into consideration as well.

Genes displaying the greatest up- and down-regulation

The greatest up-regulation of a gene was seen with Angiopoietin-2, (ANGPT-2 with 16.86 fold-change on day 3 Zone III; p

ANGPT-1 is a ligand for TIE2 and has been shown to induce formation of capillary sprouts, promote survival of endothelial cells and also be essential for blood vessel maturation and stabilization. 74 ANGPT -1 acts as a paracrine agonist by inducing phosphorylation ofTIE2 and subsequent vessel stabilization.73

ANGPT-2 is produced by endothelial cells and acts as an autocrine antagonist of

ANGPT-1 mediated TIE2 activation. Recent evidence showed that overexpression of 64

ANGPT-2 results in vascular defects, presumably through inhibition of ANGPT-1 function and that ANGPT-2 may in fact be an antagonist for ANGPT-1. 74 Glomerular

ANGPT-2 expression was shown to increase in diabetic glomerulopathy and immune­ mediated glomerulonephritis.75 ANGPT-2 is thought to act by priming the vascular endothelium to exogenous cytokines and to induce vascular destabilization at higher concentration. ANGPT-2 has also been suggested to act synergistically with other cytokines such as VEGF to promote tumor-associated angiogenesis and tumor progression and is currently a target for cancer suppression. 73 A more recent study shows that human ANGPT-2 mRNA was highly overexpressed in retinas of streptozotocin­

induced diabetic animals, indicating that ANGPT-2 overexpression in the retina enhances vascular pathology and may play an essential role in diabetic vasoregression via

destabilization of pericytes. 76

In our study, we have up-regulation of VEGF A through all zones at various time points. At the mRNA level, we have significantly increased expression of ANGPT-2 as well as VEGF A. Hence, at the genetic level, these two combinations would appear to

enhance angiogenesis, but as angiogenesis results from a cascade of cellular and

molecular events, defective events occurring downstream of these genetic signaling

factors may be what contributes to poor angiogenesis in diabetics.

The factor with the greatest significant down-regulation was seen with Tumor

necrosis factor ligand superfamily member 12 (TNFSF12 with -44.64 fold down­

regulation on day 1 Zone II). The significant down-regulation was observed in all zones

on day 1, 3 and 7. TNFSF12, a cytokine of the TNF family, is also known as Tumor

necrosis factor-like weak inducer of apoptosis (TWEAK) and is involved in cellular 65 proliferation, angiogenesis, inflammation and apoptosis. TWEAK was first discovered by Chicheportiche and others in 1997 in a search for erythropoietin-related mRNAs in macrophages. 77 It has been shown to be capable of inducing cell death via endogenous

TNF and TNF receptor.78 TWEAK and its receptor fibroblast growth factor-inducible 14

(FN14) mRNA and protein were expressed in periodontally diseased tissues, suggesting its role in the exacerbation of periodontal disease through induction of proinflammatory cytokines. 79 It has also been found to induce skeletal muscle atrophy through activation of the ubiquitin-proteasome and NF-kB system and inhibition of the PBK/AKT signaling pathway.80 TWEAK has been known to induce a strong angiogenic response when implanted as pellets in rat comeas.81 Another group has found that TWEAK can potentiate FGF-2 and VEGF-A activity in endothelial cell proliferation, suggesting a role

in regulating the physiology and pathological process of angiogenesis. 82

The circulating soluble tumor necrosis factor-like weak inducer of apoptosis

(sTWEAK) was introduced as a potential biomarker that is down-regulated in

atherosclerosis. TWEAK was also found to have a dual role in angiogenic regulation

depending on the angiogenic context in that TWEAK was a potent inducer of endothelial

cell survival and cooperates with basic fibroblast growth factor to induce proliferation and migration of human endothelial cells and morphogenesis of capillary lumens.

However, TWEAK antagonizes the morphogenic response of endothelial cells to VEGF without inhibiting VEGF-induced survival or proliferation.83 66

Baseline Changes

In terms of changes in gene expression between diabetics compared to controls on

day 0, microarray analysis reveals nine genes to be significantly different (THBS2,

CXCL4, IFNAl, ANGPTl, BAil_FREDICTED, FZDS, ECGF, TIMP3 and

LOC299339). Moreover, none of these genes were up- or down-regulated in greater than

one zone. Hence, further investigation should be performed, perhaps with Real-Time

PCR to determine whether or not these baseline differences are valid.

Genes changes that may be pertinent to disease entity

Three genes were significantly up-regulated across all zones at different time

points: VEGFA, EFNA3 and PDGFB. The source ofblood vessels originate from

endothelial cell tubes which secrete PDGFB. Although PDGFB has been proposed to

drive the formation of surrounding muscular wall by nearby mesenchymal cells, it was

found that targeted inactivation of the PDGFB gene or the PDGF receptor beta gene by

homologous recombination does not prevent the development of normal large arteries

and connective tissue, possibly due to masking of homeostatic mechanism. 84 However, a

more recent study showed that murine (m) VEGF-164 induces malignant and invasive

tumor growth of keratinocytes, which showed ulceration, disorganized epithelium

interrupted by lacunae with limited basement membrane and endothelial cell coverage,

but with the addition of human platelet-derived growth factor (hPDGF)-B the tumor and

vessel morphology improved markedly.85 The authors emphasized the need for

multifactorial therapy in formation of functional vasculature. In addition, it was found 67 that overexpression ofPDGF-B in rats receiving intraperitoneal injection ofadenoviral vector expressing PDGF-b had significant angiogenesis.86

In terms of diabetic status, PDGFB levels were over-expressed in diabetic

87 88 retinopathy and diabetic nephropathy patients. • Some groups have suggested that protective treatment such as ACE inhibitors and inhibitors of advanced glycation end­ product formation on diabetic vascular complications is associated with reduced

89 90 expression ofcytokines such as PDGF. •

VEGF is one of the most powerful angiogenic cytokines and promotes all steps in the cascade of angiogenesis.91 More specifically, it induces degeneration of the extracellular matrix by proteases, causes endothelial cell migration and proliferation, as well as determines tube proliferation of endothelial cells.92 VEGF A is thought to be the single most important angiogenic factor.93

It was shown that hypoxia up-regulates not only HIF-la but also VEGF and ephrin A2 and B subclasses at the gene and protein level in mouse skin.94 Eph receptors have emerged as critical regulators of angiogenesis in vivo to the level ofVEGF and angiopoietins. The Eph receptors are expressed mostly in endothelial cells from embryonic stages to adulthood.95 The ephrins of the A subclass are tethered to the cell membrane through glycosylphosphatidylinositol (GPI) whereas the B ephrins are transmembrane proteins.96 There are many Eph receptors (EphAI-8 and EphBI-6), named according to their affinity to specific ephrin ligands. In the adult, they have been shown to be up-regulated in some tumor tissues and implicated in its pathogenic

97 98 neovascularization. • Further, it was discovered that the EphA2 receptor stimulation 68 by ephrinAl in bovine retinal endothelial cells inhibits VEGF-induced phosphorylation of the principal receptor ofVEGF (VEGFR2) and its subsequent downstream signaling cascade, include PKC-ERKl/2 and. Moreover, inhibition of the receptor-associated signal cascade resulted in reduction of both VEGF-induced angiogenesis and vasopermeability in vivo.99 Gale discussed the role of three different growth factors­

VEGF, ANGPT and ephrins- that act via endothelial cell-specific receptor tyrosine kinases. 95 VEGF was noted to be absolutely required for the early stages of vasculogenesis and critical for subsequent angiogenesis; ANGPTl, unlike VEGF is unable to promote mitogenic responses or tubule formation, rather it comes into play after

VEGF by acting to elicit the branching and remodeling of vessels as well as to promote maturation and vessel stabilization. The natural antagonist of ANGPTl is ANGPT2, which blocks the stabilizing effect of ANGPTl, thereby allowing vessels to revert to a more plastic state where they are more responsive to VEGF but at the same time also subject to regression in the absence ofVEGF. Ephrins also appear to function during the later stages of vascular development. The role of ephrinA3 in the process of angiogenesis has not been very well covered in literature but may be a good future focus and as Gale predicted in 1999, "the precise understanding of the interaction in these various growth factors is certain to keep biologists enthralled for years to come", a prediction that is apparently still holding.

Retrovirally transfected gene expression ofPDGFB was found to accelerate

100 101 diabetic wound healing in fibroblasts as well as when applied to diabetic ulcers. • It

may be that at the genetic level we have an increased expression ofPDGFB mRNA that

is not translated to the protein level. Immunohistochemistry staining of the highlighted 69 markers seen in the study would yield further insight into the molecular mechanism of compromised wound healing in diabetics and guide us in the right direction toward using a combination therapy of significant growth factors to promote better wound healing.

Correlating gene expression with flap physiology

In spite of up-regulation of pro-angiogenic factors (VEGFA, ANGPT-1, ANGPT-

2) and down-regulation of anti-angiogenic factors, diabetic rats appear to have non­ functional angiogenesis as depicted by the increase in stasis area in diabetics over 7 days.

Our observation of increases in VEGFA and ANGPT-2 mRNA is in agreement with another study noting that the angiogenic mRNA expression of these genes was increased in the experimentally induced diabetic rats although the group noticed a decrease in

ANGPT-1 (which was noted during baseline in our study in Zone III). 102 The study found that application of Enalapril prevented an increase in mRNA expression of these

angiogenic factors in the retinas of experimentally induced diabetic rats. Further, in

another study using immunohistochemistry and testing of vitreous samples with Enzyme­

linked immunosorbent assay (ELISA), VEGFA and ANGPT-2 were found to be equally

important in the neovascular process in both type 1 and type 2 diabetes whereas VEGFD

was abundantly present in type 2 diabetes. Nuclear factor kappa B (NFKB) and

macrophages were also found, indicating the presence of an inflammatory process in the

neovascular tissues. It was concluded by the authors that in order to achieve better

control of diabetic retinopathy, it may be beneficial to develop treatment that prevents the

actions of ANGPT-2 and VEGFD. 103 70

Based on our study, the pathology of skin flap healing shares some similarity to that of diabetic retinopathy/nephropathy and research in one area may shed light on the other. The application of pro-angiogenic factors may not be appropriate for skin flap healing in diabetics and instead, administration of anti-angiogenic agents in skin flap healing may promote a better outcome. Our results suggest that significant gene down­ regulation of TWEAK combined with a significant gene up-regulation ofVEGFA may be a novel biomarker for risk of type 1 diabetes and compromised flap healing.

Furthermore, investigation of the role in ANGPT-1/TIE-2 and ANGPT-2 signaling may yield new fmdings in improving wound healing in diabetics. However, investigation of whether changes we see at the genetic level translate to the protein level should be carried out.

Genes that were seen to be significantly up-regulated in the flow zone but not significantly modulated in the control zone consisted of: CXCL2, FGFR3, VEGFC,

ANPEP, TIMP1, AGGFl, HAND2, POFUTl AND WASF2. Genes showing significant down-regulation in only the flow zones but not in the avascular or static zones include:

ILlS, NPR1, MMP9, TIMP2 AND TIMP3.

In the present study, the diabetic group shows a delay in flap avascularity compared to the control group as depicted by the significantly smaller zone of avascularity at 24 hours. However, there was a large increase in total affected area from day 1 to day 3, indicating that although the avascular zone may be smaller, skin flap survival in the diabetic group during this early time period may be due to the significant over-expression of pro-angiogenic factors such as rnidkine, vascular endothelial growth factor A (VEGFA), ephrin A2, ephrin A3, and platelet derived growth factor, B 71 polypeptide (PDGFB), as well as an under-expression of anti-angiogenic factors such as thrombospondin I (THBSl) and troponin I (TNNTl).

Over-expression of midkine, a heparin-binding growth factor, caused an increase in the angiogenic activity of human bladder cancer cells due to vascular endothelial cell growth produced by secretion of interleukin-8. 104 In addition, we observed the under­ expression of anti-angiogenenic genes such as thrombospondin 2 and troponin Tl.

Thrombospondin 2 (THBS2), a matricellular glycoprotein involved in the regulation of proliferation, adhesion and migration of a number of normal and transformed cell types.

Just like thrombospondin-1 (THBS 1), THBS2 is highly expressed in developing blood vessels, indicating a potential role in the regulation of primary angiogenesis. Mice that lacked the THBS2 protein demonstrated an increase in vascular density, abnormal collagen fibrils, and dermal fibroblasts that were defective in adhesion. It was hypothesized that THBS2 may play a critical part in the organization and vascularization of the granulation tissue during healing, possibly by modulating fibroblast-matrix interactions in early wounds and controlling the extent of angiogenesis in late wounds. 105

In another study using human A431 squamous cell carcinoma cells, THBS2 was shown to act as an inhibitor of angiogenesis, as demonstrated in a murine model where the areas of avascularity in THBS2 expressing tumors were significantly reduced despite high levels of VEGF. 106 Furthermore, mice lacking THBS2 displayed a significantly enhanced inflammatory response with increased angiogenesis, edema formation and inflammatory infiltration, with greater microvascular leakage in inflamed skin of THBS2 deficient mice. The study revealed an important role of THBS2 in limiting the extent and the duration of edema formation, angiogenesis and inflammatory cell infiltration during 72

6 acute and chronic inflammation. 5 The mechanism of inhibition by THBS2 may be due to its activity in matrix modulation. 107 The endothelial cell receptor for THBS 1 and -2

was identified as CD36 due to its ability to activate a specific signaling cascade that results in diversion of a pro-angiogenic response to an apoptotic response. 108

Troponin I is a newly discovered cartilage-derived inhibitor which inhibits

endothelial cell proliferation and angiogenesis both in vivo and in vitro. Along with

tropomyosin, Troponin I is responsible for the calcium-dependent regulation of striated

muscles. TNNT1 is capable of inhibiting actomyosin ATPase, which may explain its role

109 110 in angiogenesis, tumor growth and metastasis-inhibitory activity. •

Pulsed electromagnetic fields accelerated wound closure in both diabetic and

control mice through up-regulation ofFGF-2 although the mechanism of its action is still

unknown. 111 FGF-1 and FGF-2 are both potent inducers of endothelial cell migration,

proliferation and tube formation in vitro and are highly angiogenic in vivo. 112

Interestingly, at baseline, the diabetic group had a significantly decreased amount of

FGF-1 in our study (-1.8 fold change). However, after surgery, the diabetic rats

compensated by over-expressing FGF-1 on subsequent days in various zones, illustrating

that factors from the fibroblast growth family may take on an important role early on in

diabetic skin flap healing.

Cellular and molecular mechanisms by which the pedicle skin flaps survive

without collateral circulation from the intact vessels in the flap base are not well

elucidated. 113 Our data suggests that skin flap survival observed during the early time

period may be due to tight regulation of pro- and anti-angiogenic factors. This initial 73 difference between diabetic and· normal skin flaps was no longer observed 3 days following surgery. The increased zone of avascularity observed in diabetic rats at 3 days post-surgery may be due to reduced gene regulation of both of pro- and anti-angiogenic factors. Furthermore, at 7 days post-surgery the diabetic group had a significantly larger area of avascularity. The healing of the area closest to the pedicle may be due to the regulation of pro-angiogenic factors combined with under-expression of anti-angiogenic factors which lead to enhanced survivability in this zone. Enhanced angiogenesis could be initiated and successful in the early stages of wound healing, thus explaining less avascularity at the beginning, but which is unable to be sustained at later time points, giving rise to an increase in avascular and static zones of the flap by day 7.

Limitations

The limitations of this study included being unable to more definitively correlate the specific flap region displaying avascularity, stasis or flow with its gene expression. If we were able to identify the cells that display these traits and harvest mRNA for processing, we would be able to better ascertain that the gene expression is related to the particular trait. Another limitation of this study was that we were unable to determine if the specific gene expression properly translates to protein expression. In addition, rnicroarray analysis takes on more of a shotgun approach to genetic screening and thus may not be as specific in measuring gene expression as with Real-Time PCR. 74

Future directions

Future directions should include further verification of selected candidate genes using Real-Time PCR as well as determining whether gene expression patterns of those candidate genes correlate with protein expression with the goal of being able to enhance skin flap healing in diabetics. CONCLUSIONS

In our study, diabetic rats mounted an exaggerated angiogenic response to skin flap healing manifested by a significant up-regulation of pro-angiogenic genes (VEGFA,

ANGPT) and a significant down-regulation of anti-angiogenic genes (THBS2, TNNTI).

Poor skin flap healing in diabetics may be due to an over-expression of pro-angiogenic

factors and an under-expression of anti-angiogenic genes. Multifactorial therapy aimed

to increase anti-angiogenic mediators and lower pro-angiogenic mediators may lessen

inflannnation, reduce vascular permeability and edema to further enhance skin flap

healing. In closing, the following are the conclusions drawn from this study:

1. Significant differences in angiogenesis-associated gene expression levels were

found between diabetic and non-diabetic animals, both temporally and spatially in

a surgically created skin flap model.

2. Several pro-angiogenic genes were significantly up-regulated in diabetic animals,

such as Vascular endothelial growth factor a (VEGFA), and angiopoietin 1

(ANGPTI). Several anti-angiogenic genes were significantly down-regulated in

diabetic samples, which included thrombospondin 2 (THBS2) and troponin I

(TNNTl).

75 3. RT-PCR data confirmed the results from microarray analysis in a few selected

candidate genes, such as VEGF A, transforming growth factor beta receptor 1

(TGFBRI) and ANGPTI.

4. Diabetic animals appeared to mount a hyper-angiogenic response, characterized

by enhanced but inefficient angiogenesis. Continued investigation into the altered

gene state in diabetic animals will further enhance therapeutic approaches and

provide new directions for adjunctive pharmacologic therapies.

76 APPENDICES

77 78

Appendix A: Cumulative Changes of Significant Gene Changes across all zones and days. Note: yellow =significantly down-regulated across all zones; beige =significantly up-regulated across all zones; blue =significantly up/down regulated across all zones. 79 80

Appendix B: Volcano plots and scatter plots across all time points (Day 0, 1, 3 and 7) DayO

Volcano___ ._ ..plots Sca~-~!:_plots

0 ,:•"" ,. . . ,.. I . ... l.'i--~~'-~-----...,,--~---l l·· ·. ' 0 0 0 :0 0 0 0 - •• . :I 0 0 0 ., I o •• •* •.- ·- o • ·, t,t,:.,?.:~::;.

DayO Zone 1 Diabetic____ .. vs. Controls DayO Zone 1 Diabetic vs. Controls !. ! .. i'' L .. . . j .• .. ,__ .. ,_ ... Day 0 Zone 2 Diabetic vs. Controls Day 0 Zone 2 Diabetic vs. Controls ,._u_._.., ._.,_._ ...... ! ,.. : 0 0 I i. ·-!r ---,.-··••";--,----=------1 • • !·· ·: ·. ~"!"'.: --~:.-:? .. ·· .-... ~....:;~~ -___...... ,, __. .. Day 0 Zone 3 Diabetic____ ... vs. Controls Day 0 Zone 3 Diabetic____ ... vs. Controls

. . .. !·1----..:....:c...:.... ___---,.------i i·

....___ ~=----

Day 0 Zone 4 Diabetic vs. Controls Day 0 Zone 4 Diabetic vs. Controls 81

Dayl

Volcano plots Scatter plots ...... _. .._,._.._, . .. --. . ,.. ··.... i' j ... .. -

Day 1 Zone 1 Diabetic vs. Controls Day 1 Zone 1 Diabetic vs. Controls ..._,s ...... _, -"-"'-'

.~ I j• ...... •,... - ...... # •

.,,..__ ·•n.. ,_ .. ., Day 1 Zone 2 Diabetic vs. Controls Day 1 Zone 2 Diabetic vs. Controls ..._ ...... _ .. _......

. . ~-··· ,I .. • ~· l· . ' ......

...,._,. Day 1 Zone 3 Diabetic vs. Controls Day 1 Zone 3 Diabetic vs. Controls ...... _ ... . /"•/' ,.,_._ .· .• / ,..,. :... ·~,...... -...... ' "'· ..·· -:.-;· .. • .. • .:;;tr):.... :f"' .. .

Day 1 Zone 4 Diabetic vs. Controls Day 1 Zone 4 Diabetic vs. Controls 82

Day3

Volcano plots Scatter...._,,_...._. plots .. .. / :' ··. . !'"t------...,c'---1...... ~~ j"'... ' ...... - ·- ...... -..... -... .. - .. .. Day 3 Zone 1 Diabetic vs. Controls Day 3 Zone 1 Diabetic vs. Controls ...._, ...... _. -··--· .. .. ,r.. 'i

- - •• - •• .....- a--••• •• •• ou •• Day 3 Zone 2 Diabetic vs. Controls Day 3 Zone 2 Diabetic vs. Controls .. _.. _...._,. .. .. I ... ,'•• ...... i- ...... ~-··: ·.·:· ·::~.· ...... :.. ·~ ·~·~·~·-;~: .·- ' ...... ' ·~.._: ..... : .. ..1.• • •• .. • I .._ • :r-··. ~·~-·. .. I~:- ~-:>. . "::: ...,_ .. ,.._ .. Day 3 Zone 3 Diabetic vs. Controls Day 3 Zone 3 Diabetic vs. Controls ...._ ...... _ ...... _ .. , __ ...._,.. . .. : !·· .. j .. l" i ......

Day 3 Zone 4 Diabetic vs. Controls Day 3 Zone 4 Diabetic vs. Controls 83

Day7

Volcano plots Scatter plots -··--· " ... ,.

I ''l--'---~-~------'=-----1 i·· !.,.• !···- ... ·­ ·-

Day 7 Zone 1 Diabetic vs. Controls Day 7 Zone 1 Diabetic vs. Controls _...... _.. .._ ...... _,, ·- . ~- .•... I·· ..... - i- .. ..

,___,., .. _.., Day 7 Zone 2 Diabetic vs. Controls Day 7 Zone 2 Diabetic vs. Controls ...... _ .. "'_ ...... _...... , ! i .. ~· , ··.· j .

..

4"...,_ ... .., ,.._ ..... M Day 7 Zone 3 Diabetic vs. Controls Day 7 Zone 3 Diabetic vs. Controls ..._ ...... -... ..• . .. .-: #

~· .. !'" .. .• ~ ...... ! . ' . ... . ••' . r .· '"1-----,·:--·--·.,....__.~:~~:---1 ... ."".: ,;:·.. :: :: 7:;.,#· 4...__ ...... ,.._...... Day 7 Zone 4 Diabetic vs. Controls Day 7 Zone 4 Diabetic vs. Controls 84

Appendix C: Table ofangiogenesis-associated genes grouped by JUnction Five jUnctional groups groups: I. Adhesion Molecules (A) 2. Cytokines and Chemokines (B) 3. Growth Factors and Receptors (C) 4. Protease Inhibitors and Matrix

A.

B. c. 86

D.

E. 87

Appendix D: Genes significantly modulated across all time points Note: Green colored genes were significantly down-regulated across all 4 zones and red colored genes were significantly up-regulated across all 4 zones at various times points except Ecg/1 and F2 (p<0.05;fold-change 2-fold dif.forence in either direction)

Genes s1gn1'fi 1cant1y I mo d u Iate d across a II t1me. POintS (p< 005; ; F0 ld -c h ange: <: 2 or :!02 ) Adhesion Molecules Thbs2 Thrombospondin 2 Cytokines and Chemokines 1112a lnte~eukin 12a Growth Factors and Receptors Ereg Epiregulin Pdgfb Platelet derived growth factor, B polypeptide Tnfsf12 Tumor necrosis factor.ligand superfamily member 12 Tnnt1 Troponin T1, skeletal, slow Vegfa Vascular endothelial growth factor A Protease Inhibitors and Matrix Proteins Ecgft* Endothelial cell growth factor 1 (platelet-deriwd) F2* Coagulation factor II Transcription Factors and Other Genes Ema2 Ephrin A2 Efna3 Ephrin A3 Pofut1 Protein 0-fucosyltransfemse 1 Sh2d2a SH2 domain protein 2A Smad5 MAD homolog 5 (Drosophila)

* Ecg/1 was significantly up-regulated on Day 0 Zone I but significantly down-regulated across all other zones (//, III, IV) ! F2 was significantly down-regulated on Day 0 Zone I but significantly up-regulated across all other zones (/, II, III, IV) 88

Appendix D: RNA Checklist

Da 0 Da 1 Da 3 Da 7 Normal Diabetic Normal Diabetic Normal Diabetic Normal Diabetic RNA Checklist Sham Sham Sham Sham Sham Sham Sham Sham Z1lDd CZB2 Zl Tll zs ASH2 Z13 AB-Sham FIS CZB4 Z2 T12 ZB ASH9 Z14 A9-Sham F27 AvasruWrlqnc CZBS Z3 T13 Z19 CB-Sham ZIS 51'9 F30 T14 Z20 C9-Sham CZ28 ASO F31 - TIS Z21 ST!O- CZ29 CZ34 F32 Z22

37 I 38 1 19 3 21 5 23 I I I I I I I I I Z!!Ml1 CZB3 Z2 Tll CZB ASHS Z13 51'2-Sham F3 CZB4 Z3 T12 CZ19 ASH6 Z14 AB-Sham F6 Statfr: Zqne CZBS Z4 T13 CZ21 ASH7 ZIS 51'10- F15 T14 CZ20 ASH9 CZ28 CZ34 F27 T15 CZ26 C9 CZ29 CZ35 F30

39 40 7 25 9 27 11 29 I I I I I I I I ZimLill CZB1 Z2 C2 Z7 ASH2 Z13 ST3 F6 CZB2 Z3 C7 ZB ASHS Z14 All F9 Row zone I CZB4 Z4 ST6Sham CZ19 ASH6 Z15 A46 F27 CZBS 51'7-Sham CZ21 ASH9 CZ2B A47 F30 STB Sham CZ22 ST!3 CZ29 CZ34 F31 F32

41 42 13 31 15 33 17 35 I I ! I I I I I z.n!ill' CZB3 Z1 STS CZ6 ASH3 Z13 AB·Sham F3 CZBS Z2 ST6 CZB ASHS Z14 All F6 FlawronrU CZB2 Z4 STBSham CZ19 ASH9 Z15 A46 F9 CZB4 Til 51'4 A47 FIS T12 STll ST! F27 51'12 F30

119 100 120 121 122 123 124 125 89

Appendix E: Gene modulation across all zones over 7 days Figures a-c ofgenes significantly down-regulated/up-regulated and up/down-regulated genes.

a. Bar graphs of significantly down-regulated genes

Epiregulin (Ereg) Thrombospondin 2 (Thbs2) 0 0 .~ .. ..,~i .. -1 ]g -2 .l1! c -=e-~ = -2 II DayO -=8 II DayO .s.s .Sg -4 -3 11 Day1 ...... II Dayl ....c ... -6 ~~ • Day3 11 Day3 .... -4 .=c... .. -Q"" f' .,. E II Day? II Day7 -=§ -5 -8 -a...... ~8... -6 -10 Day and Zone Day and Zone

Interleukin 12a (D12a) Troponin Tl, skeletal, slow (Tnntl) 0 0 -~., -1 Zone I z II z 'l ~ ., -0.5 Zo z ..,~'0 ... -2 ..,'t

MAD homolog 5 (Drosophila) (SmadS) Tumor necrosis factor ligand 0 superfamily member 12 (Tnfsf12) -0.5 Zone I Zone IV -1 II DayO -1.5 II DayO II Dayl -2 • Day1 11 Day3 -2.5 • Day3 11 Day7 1!1 Day7 -3 -3.5 Day and Zone Day and Zone

SH2 domain protein 2A (Sh2d2a) o I -0.5 J Zone I Z e II Z Zo -1 -1.5 I II Day O -2 11 Day1 -2.5 II Day 3 -3 11 Day 7 -3.5 -4 -4.5 Day and Zone 90 b. Bar graphs of significantly up-regulated genes

Coagulation factor II (F2) 4.! j 3.5 3 2.5 • DayO 2 • Day1 1.5 1 • Day3 0.5 L___ _ Day7 0 Zone I Zone II Zone III Zone IV Day and Zone

Ephrin A3 (Efna3) 8 .~ CI'J 7 ~0 ~ ~ 56 "C8 • DayO ·= s 4 3 • oay1 • Day3 Day7

Zone I Zone II Zone III Zone IV Day and Zone

Platelet derived growth factor, B polypeptide (Pdgfb) 3.5 ' ,;! "' 3 1» Q ..CI--..,- ... 2.5 "C8·- = 2 - • Zone! = Q 1.5 - ·--'»"C • Zone II tu:l~a~ 1 - =... • Zone III -="'':' e"'=- 0.5 "C Q Zone IV "8'"' 0 '"" DayO Day1 Day3 Day7 Day and Zone

Vascular endothelial growth factor A (Vegfa) 5

-~"' tl1 'li: "S 4 ~ ~ 3 "C 8 • DayO ·= s 2 • Day1 t,~ :;:;1 , • oay3 ..s=- -b § 0 Day7 ;2 '"' Zone I Zone II Zone III Zone IV Day and Zone 91 c. Bar graphs of significantly up/down-regulated genes

Endothelial cell growth factor 1 (platelet-derived) (Ecgfl) 4 l 2

o I • oayo -2 1 Zone I II Zone • Dayl :: i • Day3 Day? -8 i -10 J Day and Zone 92

Appendix F. A. Surgical skin .flap model, B. Division of.flap into 4 segments, C. View of flap using orthogonal polarization D. Blue region = Total affected length (avascular + static lengths) E. Microarray layout

B.

c.

D.

E. 93

References (Formatted according to the Journal ofPeriodontology)

1. Cho CH, Sung HK, Kim KT, et al. COMP-angiopoietin-1 promotes wound healing through enhanced angiogenesis, lymphangiogenesis, and blood flow in a diabetic mouse model. Proc Nat/ Acad Sci US A 2006;103:4946-4951. 2. Abaci A, Oguzhan A, Kahraman S, et al. Effect of diabetes mellitus on formation of coronary collateral vessels. Circulation 1999;99:2239-2242. ;l. Griffioen AW, Molema G. Angiogenesis: potentials for pharmacologic intervention in the treatment of cancer, cardiovascular diseases, and chronic inflammation. Pharmacal Rev 2000;52:237-268. 4. Scarano A, Iezzi G, Petrone G, et al. Cortical bone regeneration with a synthetic cell­ binding peptide: a histologic and histomorphometric pilot study. Implant Dent 2003;12:318-324. 5. Martin A, Komada MR, Sane DC. Abnormal angiogenesis in diabetes mellitus. Med Res Rev 2003;23:117-145. 6. Buysschaert I, Schmidt T, Ronca I C, Carmeliet P, lambrechts D. Genetics, epigenetics and pharmaco-(epi)genomics in angiogenesis. J Cell Mol Med 2008;12:2533-2551. 7. Szyf M. Epigenetics, DNA methylation, and chromatin modifying drugs. Annu Rev Pharmacal Toxico/2009;49:243-263. 8. Litherland SA. lmmunopathogenic interaction of environmental triggers and genetic susceptibility in diabetes: is epigenetics the missing link? Diabetes 2008;57:3184-3186. 9. Krause B, Sobrevia l, Casanello P. Epigenetics: new concepts of old phenomena in vascular physiology. Curr Vase Pharmaco/2009;7:513-520. 10. American College of Endocrinology and American Diabetes Association Consensus statement on inpatient diabetes and glycemic control. Diabetes Care 2006;29:1955- 1962. 11. Mealey Bl, Oates TW. Diabetes mellitus and periodontal diseases. J Periodontal 2006;77:1289-1303. 12. Valentini V, Cassani A, Marianetti TM, et al. Diabetes as main risk factor in head and neck reconstructive surgery with free flaps. J Craniofac Surg 2008;19:1080•1084. 13. Miao F, Smith DO, Zhang l, Min A, Feng W, Natarajan R. Lymphocytes from patients with type 1 diabetes display a distinct profile of chromatin histone H3 lysine 9 dimethylation: an epigenetic study in diabetes. Diabetes 2008;57:3189-3198. 14. Wren JD, Garner HR. Data-mining analysis suggests an epigenetic pathogenesis for type 2 diabetes. J Biomed Biotechno/2005;2005:104-112. 15. Mealey Bl, Rethman MP. Periodontal disease and diabetes mellitus. Bidirectional relationship. Dent Today 2003;22:107-113. 16. Mealey Bl, Ocampo GL. Diabetes mellitus and periodontal disease. Periodonto/2000 2007;44:127-153. 94

17. Molitch ME, DeFronzo RA, Franz MJ, et al. Nephropathy in diabetes. Diabetes Care 2004;27 Suppi1:S79-83. 18. Chiarelli F, Santilli F, Mohn A. Role of growth factors in the development of diabetic complications. Harm Res 2000;53:53-67. 19. Paques M, Massin P, Gaudric A. Growth factors and diabetic retinopathy. Diabetes Metab 1997;23:125-130. 20. Unlu F, Guneri PG, Hekimgil M, Yesilbek B, Boyacioglu H. Expression of vascular endothelial growth factor in human periodontal tissues: comparison of healthy and diabetic patients. J Periodonto/2003;74:181-187. 21. Schmidt AM, Weidman E, lalla E, et al. Advanced glycation end products (AGEs) induce oxidant stress in the gingiva: a potential mechanism underlying accelerated periodontal disease associated with diabetes. J Periodontal Res 1996;31:508-515. 22. Willershausen-Zonnchen B, Lemmen C, Hamm G. Influence of high glucose concentrations on glycosaminoglycan and collagen synthesis in cultured human gingival fibroblasts. J C/in Periodonto/1991;18:190-195. 23. Kornman KS. Patients are not equally susceptible to periodontitis: does this change dental practice and the dental curriculum? J Dent Educ 2001;65:777-784. 24. perioPapapanou PN. Periodontal diseases: epidemiology. Ann Periodonto/1996;1:1-36. 25. Grossi SG, Genco RJ. Periodontal disease and diabetes mellitus: a two-way relationship. Ann Periodonto/1998;3:51-61. 26. King GL. The role of inflammatory cytokines in diabetes and its complications. J Periodonto/2008;79:1527-1534. 27. Grossi SG. Treatment of periodontal disease and control of diabetes: an assessment of the evidence and need for future research. Ann Periodonto/2001;6:138-145. 28. Loe H. Periodontal disease. The sixth complication of diabetes mellitus. Diabetes Care 1993;16:329-334. 29. de Pommereau V, Dargent-Pare C, Robert JJ, Brion M. Periodontal status in insulin­ dependent diabetic adolescents. J Clin Periodonto/1992;19:628-632. 30. Cianciola U, Park BH, Bruck E, Mosovich L, Genco RJ. Prevalence of periodontal disease in insulin-dependent diabetes mellitus (juvenile diabetes). JAm Dent Assoc 1982;104:653-660. 31. Emrich U, Shlossman M, Genco RJ. Periodontal disease in non-insulin-dependent diabetes mellitus. J Periodonto/1991;62:123-131. 32. Nelson RG, Shlossman M, Budding LM, et al. Periodontal disease and NIDDM in Pima Indians. Diabetes Care 1990;13:836-840. 33. Taylor GW, Burt BA, Becker MP, et al. Non-insulin dependent diabetes mellitus and alveolar bone loss progression over 2 years. J Periodanto/1998;69:76-83. 34. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). UK Prospective Diabetes Study (UKPDS) Group. Lancet 1998;352:837-853. 35. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. The Diabetes Control and Complications Trial Research Group. N Eng/ J Med 1993;329:977-986. 36. Mealey B. Diabetes and periodontal diseases. J Periodonto/1999;70:935-949. 37. Sastrowijoto SH, Hillemans P, van Steenbergen TJ, Abraham-lnpijn L, de Graaff J. Periodontal condition and microbiology of healthy and diseased periodontal pockets in type 1 diabetes mellitus patients. J Clin Periodonto/1989;16:316-322. 95

38. Manouchehr-Pour M, Spagnuolo PJ, Rodman HM, Bissada NF. Comparison of neutrophil chemotactic response in diabetic patients with mild and severe periodontal disease. J Periodonto/1981;52:410-415. 39. McMullen JA, Van Dyke TE, Horoszewicz HU, Genco RJ. Neutrophil chemotaxis in individuals with advanced periodontal disease and a genetic predisposition to diabetes mellitus. J Periodonto/1981;52:167-173. 40. Salvi GE, Valda B, Collins JG, et al. Inflammatory mediator response as a potential risk marker for periodontal diseases in insulin-dependent diabetes mellitus patients. J Periodonto/1997;68:127-135. 41. McGuire MK, Nunn ME. Prognosis versus actual outcome. IV. The effectiveness of clinical parameters and IL-l genotype in accurately predicting prognoses and tooth survival. J Periodonto/1999;70:49-56. 42. Karamysheva AF. Mechanisms of angiogenesis. Biochemistry {Mosc) 2008;73:751-762. 43. Nyengaard JR, Flyvbjerg A, Rasch R. The impact of renal growth, regression and regrowth in experimental diabetes mellitus on number and size of proximal and distal tubular cells in the rat kidney. Diabetologia 1993;36:1126-1131. 44. Loots MA, Lamme EN, Mekkes JR, Bos JD, Middelkoop E. Cultured fibroblasts from chronic diabetic wounds on the lower extremity (non-insulin-dependent diabetes mellitus) show disturbed proliferation. Arch Dermatol Res 1999;291:93-99. 45. Galeano M, Torre V, Deodato B, et al. Raxofelast, a hydrophilic vitamin E-like antioxidant, stimulates wound healing in genetically diabetic mice. Surgery 2001;129:467-477. 46. Mattei MG, BorgJP, Rosnet o, Marme D, Birnbaum D. Assignment of vascular endothelial growth factor (VEGF) and placenta growth factor (PLGF) genes to human chromosome 6pl2-p21 and 14q24-q31 regions, respectively. Genomics 1996;32:168- 169. 47. Tolentino MJ, Miller JW, Gragoudas ES, Chatzistefanou K, Ferrara N, Adamis AP. Vascular endothelial growth factor is sufficient to produce iris neovascularization and neovascular glaucoma in a nonhuman primate. Arch Ophtha/mo/1996;114:964-970. 48. Freedman SB, Vale P, Kalka C, et al. Plasma vascular endothelial growth factor (VEGF) levels after intramuscular and intra myocardial gene transfer ofVEGF-1 plasmid DNA. Hum Gene Ther 2002;13:1595-1603. 49. Zimering MB, Eng J. Increased basic fibroblast growth factor-like substance in plasma from a subset of middle-aged or elderly male diabetic patients with microalbuminuria or proteinuria. J Clin Endocrinol Metab 1996;81:4446-4452. 50. Sa to TN, Tozawa Y, Deutsch U, et al. Distinct roles of the receptor tyrosine kinases Tie-l and Tie-2 in blood vessel formation. Nature 1995;376:70-74. 51. Davis S, Aldrich TH, Jones PF, et al. Isolation of angiopoietin-1, a ligand for the TIE2 receptor, by secretion-trap expression cloning. Ce//1996;87:1161-1169. 52. Suri C, Jones PF, Patan S, et al. Requisite role of angiopoietin-1, a ligand for the TIE2 receptor, during embryonic angiogenesis. Ce//1996;87:1171-1180. 53. Williams B. Angiotensin II, VEGF, and diabetic retinopathy. Lancet 1998;351:837-838. 54. Yarom R, Zirkin H, Stammler G, Rose AG. Human coronary microvessels in diabetes and ischaemia. Morphometric study of autopsy material. J Patho/1992;166:265-270. 55. Bornstein P. Thrombospondins function as regulators of angiogenesis. J Cell Commun Signa/2009. 96

56. lange-Asschenfeldt B, Weninger W, Velasco P, et al. Increased and prolonged inflammation and angiogenesis in delayed-type hypersensitivity reactions elicited in the skin of thrombospondin-2--deficient mice. Blood 2002;99:538-545. 57. Wikesjo UM, Nilveus RE, Selvig KA. Significance of early healing events on periodontal repair: a review. J Periodonto11992;63:158-165. 58. Davis RE, Wachholz JH, Jassir D, Perlyn CA, Agrama MH. Comparison of topical anti­ ischemic agents in the salvage of failing random-pattern skin flaps in rats. Arch Facial Plast Surg 1999;1:27-32. 59. Huemer GM, Wechselberger G, Otto-Schoeller A, Gurunluoglu R, Piza-Katzer H, Schoeller T. Improved dorsal random-pattern skin flap survival in rats with a topically applied combination of nonivamide and nicoboxil. Plast Reconstr Surg 2003;111:1207- 1211. 60. Tsai JW, Ayubi FS, Hart Kl, et al. Evaluation of the effect of sildenafil and vascular endothelium growth factor combination treatment on skin flap survival in rats. Aesthetic Plast Surg 2008;32:624-631. 61. Ulusoy MG, Uysal A, Kocer U, et al. Improved flap viability with site-specific delivery of sildenafil citrate using fibrin glue. Ann Plast Surg 2005;55:292-296. 62. Strocchi R, Orsini G, Iezzi G, et al. Bone regeneration with calcium sulfate: evidence for increased angiogenesis in rabbits. J Oral/mplantal2002;28:273-278. 63. Kaufman T, Angel MF, Eichenlaub EH, levin M, Hurwitz DJ, Futrell JW. The salutary effects of the bed on the survival of experimental flaps. Ann Plast Surg 1985;14:64-73. 64. Angel MF, Kaufman T, Swartz WM, Ramasastry SS, Narayanan K, Futrell JW. Studies on the nature of the flap/bed interaction in rodents--Part 1: Flap survival under varying conditions. Ann Plast Surg 1986;17:317-322. 65. McFarlane RM, Deyoung G, Henry RA. The Design of a Pedicle Flap in the Rat to Study Necrosis and Its Prevention. Plast Recanstr Surg 1965;35:177-182. 66. Hom DB, Baker SR, Graham lM, McClatchey KD. Utilizing angiogenic agents to expedite the neovascularization process in skin flaps. Laryngoscope 1988;98:521-526. 67. Hom DB, Assefa G. Effects of endothelial cell growth factor on vascular compromised skin flaps. Arch Otolaryngol Head Neck Surg 1992;118:624-628. 68. lshiguro N, Yabe Y, Shimizu T, Iwata H, Miura T. Basic fibroblast growth factor has a beneficial effect on the viability of random skin flaps in rats. Ann Plast Surg 1994;32:356- 360. 69. Bozikov K, Arnez ZM. Factors predicting free flap complications in head and neck reconstruction. J Plast Reconstr Aesthet Surg 2006;59:737-742. 70. Yang D, Morris SF. Comparison of two different delay procedures in a rat skin flap model. Plast Recanstr Surg 1998;102:1591-1597. 71. Costouros NG, libutti SK. Microarray technology and gene expression analysis for the study of angiogenesis. Expert Opin Bioi Ther 2002;2:545-S56. 72. Olivier WA, Hazen A, levine JP, Soltanian H, Chung S, Gurtner GC. Reliable assessment of skin flap viability using orthogonal polarization imaging. Plast Reconstr Surg 2003;112:547-555. 73. Thomas M, Felcht M, Kruse K, et al. Angiopoietin-2 stimulation of endothelial cells induces {alpha}v{beta}3 integrin internalization and degradation. J Bioi Chern 2010. 74. Hu B, Cheng SY. Angiopoietin-2: development of inhibitors for cancer therapy. Curr Oncol Rep 2009;11:111-116. 97

75. Yao D, Taguchi T, Matsumura T, et al. High glucose increases angiopoietin-2 transcription in microvascular endothelial cells through methylglyoxal modification of mSin3A. J Bioi Chern 2007;282:31038-3104S. 76. Pfister F, Wang Y, Schreiter K, et al. Retinal overexpression of angiopoietin-2 mimics diabetic retinopathy and enhances vascular damages in hyperglycemia. Acta Diabetol 2010;47:S9-64. 77. Chicheportiche Y, Bourdon PR, Xu H, et al. TWEAK, a new secreted ligand in the tumor necrosis factor family that weakly induces apoptosis. J Bioi Chern 1997;272:32401- 32410. 78. Schneider P, Schwenzer R, Haas E, et al. TWEAK can induce cell death via endogenous TNF and TNF receptor 1. Eur J /mmunol1999;29:1785-1792. 79. Hosokawa Y, Hosokawa I, Ozaki K, Nakae H, Matsuo T. Proinflammatory effects of tumour necrosis factor-like weak inducer of apoptosis (TWEAK) on human gingival fibroblasts. Clin Exp /mmunol2006;146:540-549. 80. Bhatt S, Wehbe C, Dogra VS. Phlegmasia cerulea dolens. J Clin Ultrasound 2007;35:401- 404. 81. lynch CN, Wang YC, lund JK, Chen YW, leal JA, Wiley SR. TWEAK induces angiogenesis and proliferation of endothelial cells. J Bial Chern 1999;274:8455-8459. 82. Donohue PJ, Richards CM, Brown SAN, et al. TWEAK is an endothelial cell growth and chemotactic factor that also potentiates FGF-2 and VEGF-A mitogenic activity. Arterioscl Throm Vas 2003;23:594-600. 83. Jakubowski A, Browning B, lukashev M, et al. Dual role for TWEAK in angiogenic regulation. J Cell Sci 2002;115:267-274. 84. Crosby JR, Seifert RA, Soriano P, Bowen-Pope DF. Chima eric analysis reveals role of Pdgf receptors in all muscle lineages. Nat Genet 1998;18:385-388. 85. lederle W, Linde N, Heusel J, et al. Platelet-derived growth factor-B normalizes micromorphology and vessel function in vascular endothelial growth factor-A-induced squamous cell carcinomas. Am J Patho/2010;176:981-994. 86. Cina D, Patel P, BethuneJC, et al. Peritoneal morphological and functional changes associated with platelet-derived growth factor B. Nephrol Dial Transplant 2009;24:448- 457. 87. Yokota T, MaRC, Park JY, et al. Role of protein kinase Con the expression of platelet­ derived growth factor and endothelin-1 in the retina of diabetic rats and cultured retinal capillary pericytes. Diabetes 2003;52:838-845. 88. Langham RG, Kelly DJ, Maguire J, Dowling JP, Gilbert RE, Thomson NM. Over-expression of platelet-derived growth factor in human diabetic nephropathy. Nephrol Dial Transplant 2003;18:1392-1396. 89. Kelly DJ, Gilbert RE, Cox AJ, Soulis T, Jerums G, Cooper ME. Aminoguanidine ameliorates overexpression of prosclerotic growth factors and collagen deposition in experimental diabetic nephropathy. JAm Soc Nephro/2001;12:2098-2107. 90. Lassila M, Allen TJ, Cao Z, et al. lmatinib attenuates diabetes-associated atherosclerosis. Arterioscler Thromb Vase Biol2004;24:935-942. 91. Altavilla D, Saitta A, Cucinotta D, et al. Inhibition of lipid peroxidation restores impaired vascular endothelial growth factor expression and stimulates wound healing and angiogenesis in the genetically diabetic mouse. Diabetes 2001;50:667-674. 92. Ferrara N, Davis-Smyth T. The biology of vascular endothelial growth factor. Endocr Rev 1997;18:4-25. 98

93. Dvorak HF. VPF/VEGF and the angiogenic response. Semin Perinato/2000;24:75-78. 94. Vihanto MM, PlockJ, Erni D, Frey BM, Frey FJ, Huynh-Do U. Hypoxia up-regulates expression of Eph receptors and ephrins in mouse skin. FASEB J 2005;19:1689-1691. 95. Gale NW, Yancopoulos GD. Growth factors acting via endothelial cell-specific receptor tyrosine kinases: VEGFs, angiopoietins, and ephrins in vascular development. Genes Dev 1999;13:1055-1066. 96. Kullander K, Klein R. Mechanisms and functions of Eph and ephrin signalling. Nat Rev Mol Cell 8/o/2002;3:475-486. 97. Dobrzanski P, Hunter K, Janes-Bolin S, et al. Antiangiogenic and antitumor efficacy of EphA2 receptor antagonist. Cancer Res 2004;64:910-919. 98. Brantley DM, Cheng N, Thompson EJ, et al. Soluble Eph A receptors inhibit tumor angiogenesis and progression in vivo. Oncogene 2002;21:7011-7026. 99. Ojima T, Takagi H, Suzuma K, et al. EphrinA1 inhibits vascular endothelial growth factor­ induced intracellular signaling and suppresses retinal neovascularization and blood­ retinal barrier breakdown. Am J Potho/2006;168:331-339. 100. Breitbart AS, Laser J, Parrett B, et al. Accelerated diabetic wound healing using cultured dermal fibroblasts retrovirally transduced with the platelet-derived growth factor B gene. Ann Pfost Surg 2003;51:409-414. 101. Steed DL. Clinical evaluation of recombinant human platelet-derived growth factor for the treatment of lower extremity diabetic ulcers. Diabetic Ulcer Study Group. J Vase Surg 1995;21:71-78; discussion 79-81. 102. Kim HW, Kim JL, Lee HK, Hur DY, Yun IH, Kim SD. Enalapril alters expression of key growth factors in experimental diabetic retinopathy. CurrEye Res 2009;34:976-987. 103. Kinnunen K, Puustjarvi T, Terasvirta M, et al. Differences in retinal neovascular tissue and vitreous humour in patients with type 1 and type 2 diabetes. Br J Ophthalmol 2009;93:1109-1115. 104. Muramaki M, Miyake H, Hara I, Kamidono S. Introduction of midkine gene into human bladder cancer cells enhances their malignant phenotype but increases their sensitivity to antiangiogenic therapy. Clin Cancer Res 2003;9:5152-5160. 105. Kyriakides TR, Tam JW, Bornstein P. Accelerated wound healing in mice with a disruption ofthe thrombospondin 2 gene. J Invest Dermato/1999;113:782-787. 106. Streit M, Riccardi L, Velasco P, et al. Thrombospondin-2: a potent endogenous inhibitor of tumor growth and angiogenesis. Proc Nat/ Acad Sci US A 1999;96:14888-14893. 107. Krady M M, Zeng J, Yu J, et al. Thrombospondin-2 modulates extracellular matrix remodeling during physiological angiogenesis. Am J Potho/2008;173:879-891. 108. Silverstein RL, Febbraio M. CD36-TSP-HRGP interactions in the regulation of angiogenesis. Curr Pharm Des 2007;13:3559-3567. 109. Nyberg P, Xie L, Kalluri R. Endogenous inhibitors of angiogenesis. Cancer Res 2005;65:3967-3979. 110. Moses MA, Wiederschain D, Wu I, et al. Troponin I is present in human cartilage and inhibits angiogenesis. Proc Nat/ Acad Sci US A 1999;96:2645-2650. 111. Callaghan MJ, Chang El, Seiser N, et al. Pulsed electromagnetic fields accelerate normal and diabetic wound healing by increasing endogenous FGF-2 release. Pfost Reconstr Surg 2008;121:130-141. 112. Klagsbrun M, Moses MA. Molecular angiogenesis. Chem 8io/1999;6:R217-224. 113. Myers B. Understanding flap necrosis. Plast Reconstr Surg 1986;78:813-814.