Pursuing the Roles of Non-Invasive Biomarkers in Chronic Renal Allograft Dysfunction

Virginia Commonwealth University VCU Scholars Compass

Theses and Dissertations Graduate School

2014

Mba Uzoma U. Mba Virginia Commonwealth University

Follow this and additional works at: https://scholarscompass.vcu.edu/etd

Downloaded from https://scholarscompass.vcu.edu/etd/4080

This Thesis is brought to you for free and open access by the Graduate School at VCU Scholars Compass. It has been accepted for inclusion in Theses and Dissertations by an authorized administrator of VCU Scholars Compass. For more information, please contact [email protected].

Pursuing the Roles of Non-Invasive Biomarkers in Chronic Renal Allograft Dysfunction

A dissertation submitted In partial fulfillment of the requirements for the degree of Doctor of Philosophy at Virginia Commonwealth University

Mba Uzoma Udo Mba B.Sc. Chemistry, Duke University (2004) M.S., Pharmacology & Toxicology, Virginia Commonwealth University (2008)

Directors: Vladimir I. Vladimirov, M.D., Ph.D. Assistant Professor, Department of Psychiatry, Virginia Institute for Psychiatric and Behavioral Genetics

Catherine I. Dumur, Ph.D. Associate Director of Molecular Diagnostics Division & Associate Professor of Pathology

Virginia Commonwealth University Richmond, Virginia August, 27, 2014

ACKNOWLEDGEMENTS

I would like to start by thanking my advisors, Vladimir Vladimirov and Catherine Dumur, and my committee for their guidance and support. Their advocacy for me has been invaluable in producing this work, and achieving my Ph.D. I also thank Dr. Valeria Mas for introducing me to this field of study, and exposing me to the possibilities of genomics research and biomarker discovery. I thank my former lab mates: Mariano Scian, Ricardo

Gehrau, Ryan Fassnacht, and Lacey Suh, for allowing me the opportunity to learn, teach, and grow as a student and scientist. I thank my current lab mates: Dr. Vernell

Williamson, Mohammed Mamdani, and G. Omari McMichael, for giving me a lab home and a place to thrive and learn the finer points of molecular and computational biology.

A special thanks goes to Omari, for without his aptitude, skills, and tireless effort, very little of this work would have been possible. He is the big brother that I didn’t know I needed, and can’t imagine accomplishing any of this work without him.

Thanks to the members of the Virginia Commonwealth Uinversity Department of

Physiology & Biophysics and the M.D./Ph.D. program for their efforts in allowing me to continue working through any and all adversity. Thank you my friends and family for their support and being a persistent buttress to all my endeavors. Finally, a very special thanks goes to Chadia Robertson. Her material, mental, and emotional support has meant more to me than I can properly articulate in this space, and I am eternally grateful to be in her life.

TABLE OF CONTENTS

LIST OF TABLES ...... iv LIST OF FIGURES ...... v LIST OF ABBREVIATIONS ...... vii ABSTRACT ...... x INTRODUCTION ...... 1 Normal Renal Function, Glomerular Filtration Rate & Renal Failure ...... 1 Transplantation, Donor Quality, DGF & CAD ...... 5 Post Transplant Monitoring & Gene Expression Profiles...... 11 MicroRNAs and Renal Dysfunction ...... 16 MicroRNAs and Chronic Allograft Dysfunction ...... 23 Materials and Methods ...... 30 Urine Sample miR and mRNA Expression ...... 30 Cell Culture miR and mRNA Expression ...... 33 Cloning Expression Plasmids ...... 36 Transfection ...... 40 Luciferase Assays ...... 41 Results & Discussion...... 43 Donor Quality and Gene Expression in Delayed Graft Function ...... 43 Independent validation of CAD-related microRNA expression in urine samples ...... 53 Assessment of the Novel Targets of CAD-Related microRNAs ...... 60 Conclusions ...... 75 Appendix A ...... 93 Appendix B ...... 96 Appendix C ...... 98 Appendix D ...... 101 VITA ...... 106

iii

LIST OF TABLES

Table 1. Expanded Criteria Donor Characteristics ...... 7

Table 2. miR-142-3p negatively correlated targets ...... 25

Table 3. miR-204 negatively correlated targets ...... 27

Table 4. miR-211 negatively correlated targets ...... 28

Table 5. Primer sequences for miR-142-3p & miR-211...... 36

Table 6. pre-miR-204 sequence with cohesive ends...... 37

Table 7. Target 3'-UTR Sequences...... 39

Table 8. DGF diagnoses amongst 147 DDK recipients according to donor quality and pump use...... 44

Table 9. Sample cohort classifications...... 55

Table 10. Demographic and clinical data from patients in miR expression analysis. .... 58

LIST OF FIGURES

Figure 1. A schematic representation of the nephron and its functions ...... 2

Figure 2. Histological sections of normal kidney biopsy tissue, and chronic renal failure biopsy tissue ...... 5

Figure 3. Diagnosis and rejection timelines...... 14

Figure 4. Two dimensional hierarchical clustering of differentially expressed mRNA profiles of patient samples comparing CAD to normal allograft kidneys and normal kidney tissue...... 16

Figure 5. Ingenuity generated network of DGF associated genes ...... 45

Figure 6. ToppGene analysis of differentially expressed genes in DGF...... 47

Figure 7. Gene ontology for delayed graft function comparing pre-implantation vs. post- reperfusion ...... 49

Figure 8 – IngenuityTM PPAR signaling canonical pathway, annotated with differential expression from IRI study...... 50

Figure 9. Gene ontology for K2 samples (n=59) DGF vs. no-DGF ...... 51

Figure 10. Ingenuity™ network from K2 array data comparing DGF to no-DGF...... 52

Figure 11 - Relative expression of CAD-related miRs in urine samples of transplant recipients...... 57

Figure 12 – CD44 Expresssion in Urine Samples...... 59

Figure 13 - Schematic of pmR-ZsGreen and DNA gel picture of pre-miR inserts and open pmR-ZsGreen used for cloning...... 61

Figure 14 - Transfection efficiencies...... 62

Figure 15 - MicroRNA overexpression in HEK cells...... 63

Figure 16 - mRNA expression of proposed CAD-related miR targets following miR overexpression in HEK293 cells ...... 65

Figure 17 - Luciferase expression of CAD-related miR target genes...... 68

Figure 18 - miR-142-3p and target expression following TNF-α treatment...... 71

Figure 19 - miR-204 and target expression following TNF-α treatment...... 73

Figure 20 - miR-211 and target expression following TNF-α treatment...... 74

Figure 21 - CD44 control of miR-21 expression...... 78

LIST OF ABBREVIATIONS

Abbreviation Definition Page

ACE Angiotensin I Converting Enzyme 22

ACR Acute Cellular Rejection 9

Ac-SDKP N-acetyl-seryl-aspartyl-lysyl-proline 22

AKI Acute Kidney Injury 19

ANOVA Analysis of Variance 31

BUN Blood Urea Nitrogen 11

CAD Chronic Allograft Dysfunction 5

DCD Donation after Cardiac Death 6

DDK Deceased Donor Kidney 40

DGF Delayed Graft function 5

DGKA Diacylglycerol kinase alpha 27

DMEM Dulbecco’s Modified Eagle’s Medium 38

DNA Deoxyribonucleic acid 13

ECD Expanded Criteria Donor 6

EDTA Ethylenediaminetetraacetic acid 31

EGF Epidermal Growth Factor 14

EGFR Epidermal Growth Factor Receptor 15 eGFR Estimated Glomerular Filtration Rate 3

EMT Epithelial-Mesenchymal Transition 10

ESRD End-Stage Renal Disease 1

vii

Abbreviation Definition Page

EV Empty Vector 59

FBS Fetal bovine serum 38

GAPDH Glyceraldehyde-3-phosphate dehydrogenase 30

GW182 Trinucleotide repeat containing 6A

HEK Human Embryonic Kidney 60

IFN-β Interfereon-beta 26

IFTA Interstitial Fibrosis & Tubular Atrophy 8

IL Interleukin 14

IRI Ischemia Reperfusion Injury 19

LURD Living Unrelated Donation 6

MAP Mitogen Activated Protein 27

MDRD Modification of Diet in Renal Disease 3

MEM Minimum Essential Media 38 miRNA/miR Micro-Ribonucleic Acid 17 mRNA Messenger Ribonucelic Acid 13

NA Normal Allograft 50

NARF Nucleic Acid Research Facility 35

NCBI National Center for Biotechnology Information 33

NEB New England Biolabs 34

PBS Phosphate-Buffered Saline 28

PCR Polymerase Chain Reaction 29

PPARα Peroxisome-Proliferator Activated Receptor-alpha 47

PPIA Peptidylprolyl Isomerase 30

PTEN Phophatase and tensin homolog 48

viii

Abbreviation Definition Page qPCR Qualitative Polymerase Chain Reaction 15

RISC RNA-induced silencing complex 18

RMA Robust Multiarray Average 15

RNA Ribonucleic Acid 13

RQ Relative Quantitation 30

RT-qPCR Reverse Transcription-Quantitative Polymerase Chain 15 Reaction

SCD Standard Criteria Donor 23

SDS Sequence Detection System 30

TE Tris-EDTA 31

TGF-β Transforming Growth Factor-Beta 14

TGF-βR2 Transforming Growth Factor-Beta Receptor 2 35

TNFα Tumor Necrosis Factor alpha 66

TOPO Topoisomerase 58

TSP-1 Thrombospondin-1 15

UMOD Uromodulin 25

UNG Uracil-N-glycosylase 30

UTR Untranslated Region 18

VCUHS Virginia Commonwealth University Health Systems 28

ABSTRACT

PURSUING THE ROLES OF NON-INVASIVE BIOMARKERS IN CHRONIC RENAL ALLOGRAFT DYSFUNCTION

By Mba Uzoma U. Mba, M.S.

A dissertation submitted In partial fulfillment of the requirements for the degree of Doctor of Philosophy at Virginia Commonwealth University

Virginia Commonwealth University, 2014

Directors:

Vladimir I. Vladimirov, M.D., Ph.D. Assistant Professor, Department of Psychiatry, Virginia Institute for Psychiatric and Behavioral Genetics

Catherine I. Dumur, Ph.D. Associate Director of Molecular Diagnostics Division & Associate Professor of Pathology

Kidney failure affects over 600,000 people in the United States with over 100,000 new cases of kidney failure diagnosed every year. Chronic allograft dysfunction (CAD), a disease state highlighted by a decreased eGFR and a histo-pathological diagnosis interstitial fibrosis and renal tubular atrophy, has emerged as one of the leading causes of kidney transplantation. Gene expression analysis can identify reproducible patterns of disease in ribonucleic acid (RNA) expression and may provide insights into the quality of kidney donations heretofore overlooked by current clinical analytical methods.

Additionally, genetic expression profiles can be obtained from patient blood or urine samples, diminishing the necessity of costly and invasive tissue biopsies. Micro-RNA

(miR) profiles have been obtained from non-invasive sources and shown to act as biomarkers of CAD. Specifically, miR-142-3p, miR-204, and miR-211 are shown to be

x differentially expressed in both tissue and urine samples of patients with CAD. The role of these CAD-specific miRs in disease development is not well understood, but given the nature of miR function, they may play a significant role in CAD development.

An independent validation of CAD-specific miR expression was done in 155 deceased donor transplant recipients using RT-qPCR. This new patient cohort delineated the importance of both decreased clinical renal function as measure by GFR, and diagnosed histopathology in identifying CAD. Previous work delineating a messenger

RNA signature of CAD was used to produce a list of potential targets for CAD-specific miRs via in silico analysis. Briefly, targets for each miR were consolidated from 11 miR- targeting networks: PITA, miRanda, TargetScan 5.1, DNA-microT, NBmiRTar,

MicroInspector, PicTar, MirTarget2, RNA22, miTarget, and RNA hybrid (all can be found at: http://c1.accurascience.com/miRecords/prediction_query.php). Targets that appeared in at least 3 of databases were considered for further study. Experimental targets were then chosen according to negative Pearson correlations of miR/mRNA pairings and potential roles in apoptosis or fibrogenesis. Urine samples showed differential expression of transmembrane protein 14A (miR-142-3p target), CD44 molecule (miR-204), basic helix-loop-helix family member e41 (miR-204), and dual specificity kinase 6 (miR-211) between patients with CAD and those with normal allografts. Human embryonic kidney (HEK) cell lines were used to validate these targets via transfection of the CAD-specific miRs and observing the modulation of their selected gene targets. Overexpression the CAD-specific miRs in HEK cells modulated their respective target gene expression. The relationship between CAD-miRs and their targets was validated using luciferase expression. This research verifies the differential

xi expression of these CAD-specific miRs from non-invasive sources, and validates gene targets for these miRs that modulate significant biological pathways in the development of CAD.

xii

INTRODUCTION

Normal Renal Function, Glomerular Filtration Rate & Renal Failure

End stage renal disease (ESRD) is defined as renal failure that requires hemodialysis and kidney transplant (1-7). Normally, the kidneys receive about 25% of cardiac output and filter the blood of waste, maintain electrolyte balance in the body, and produce urine as a means of controlling fluid balance (8). The nephron, an interface of renal tubules and blood vessels, is the functional unit of the kidney. Blood flows into the kidney via the renal artery and passes through the arterial tree to the afferent arteriole. From the afferent arteriole, blood enters the glomerulus, the functional beginning of the nephron, which is made up of highly fenestrated glomerular capillaries and the Bowman’s space, the tubular tissue comprised of podocytes that surrounds and interfaces with the capillaries and vaguely resembles a wrench in cross-sectional views (Fig 1) (8). The glomerular capillaries fuse on the other end of the glomerulus to form the efferent arteriole which continues to travel with and around the nephron as the peritubular capillary in order to aid fluid and electrolyte balance (8).

Figure 1. A schematic representation of the nephron and its functions; from: http://www.wickersham.us/anne/peelab.htm Filtration from the glomerular capillaries into the Bowman’s space is governed chiefly by

Starling forces that result in a net push of fluid and nutrients from the capillaries into the

Bowman’s space (8). In healthy functioning kidneys, water, electrolytes, oligosaccharides, oligopeptides, and soluble metabolites are filtered into the nephron’s tubules and over 99% of the filtrate is reabsorbed in the nephron (8). Filtrate flows from the Bowman’s space into the proximal (convoluted) tubule where the majority of water, sodium, and simple sugars and oligopeptides/amino acids are reabsorbed (8). The filtrate then enters the Loop of Henle (LoH) that descends into, and ascends from, the renal medulla. In the descending portion of the LoH, water is reabsorbed into the hyperosmotic renal medulla (8). The ascending LoH is impermeable to water and is a major site of electrolyte regulation via the 3-ion (Na+/K+/Cl-) co-transporter (8). The distal tubule continues from the ascending LoH and controls Ca2+, Na+, and water reabsorption as well as acid secretion via pharmacologic or hormonal control of ion/water channel proteins (8). Water reabsorption can continue beyond the distal tubule in the collecting ducts as needed (8). Collecting ducts are the functional end of

2 the nephron as they coalesce into the renal calices that form the renal pelvis and ureter that transmit the remaining filtrate, now urine from the kidney into the bladder for excretion.

Waste products that do not enter the renal tubules via the glomerulus can alternatively be secreted into the tubules from the peritubular capillaries for excretion in the urine (8).

Clinical monitoring of renal function is based on the urine and/or serum levels of chemical markers. Renal markers are generally freely filtered but not reabsorbed by, or secreted into the nephron (8). They can be exogenous for optimal study, as done with the plant carbohydrate inulin; or endogenous for easy measurement, as done with creatinine (8). Measuring the serum levels of creatinine is the most expedient, and thus, widely used method of monitoring renal function. Direct measurement of creatinine can be used singularly to assess renal health, but the glomerular filtration rate (GFR), a creatinine-based estimation of renal function calculated below (Eq. 1), is the current clinical standard for monitoring renal function (9-12).

-1.154 -0.203 Equation 1. MDRD Estimated GFR = 175*(sCr ) x (Age ) x [1.212, if Black] x [0.742, if Female] The equation uses a correction coefficient (175) that accounts for the units of serum creatinine (sCr), mg/dL, and the manner in which sCr is measured (11;12). The exponential coefficient for serum creatinine, -1.154, indicates that a 1% increase in sCr results in a 1.154% drop in eGFR (11;12). At any sCr level, increasing age and being female are linked to lower GFR, while being Black is associated with higher GFR

(11;12). The equation ignores the individual’s height and weight by normalizing the results to an average body surface area of 1.73m2, making over estimations in underweight persons and underestimations for people who are overweight or have large

3 muscle mass (12). In the clinical setting, an eGFR >60 ml/min/1.73m2 is considered normal post-transplant function (9-12), and renal failure is marked by a sudden (acute) or progressive (chronic) decrease in GFR. Functionally, the drop in GFR is a reflection of the nephron’s inability to efficiently filter the blood due to a variety of reasons, otherwise known as renal failure. Histologically, renal failure is typified by some combination of the thickening of glomerular basement membranes, obliteration of renal tubules, a decrease/loss of blood flow through the peritubular capillaries, and in some cases, infiltration of immune cells & lymphatic fluids (Fig.2). These tissue changes produce the picture of the physical alterations leading to the characteristic loss of filtration associated with renal failure. Molecularly, renal failure is combination of fibrogenesis, inflammation and apoptosis working alone or in concert to destroy the normal functioning kidney.

Figure 2. Histological sections of (A) normal kidney biopsy tissue, and (B) chronic renal failure biopsy tissue

Transplantation, Donor Quality, DGF & CAD

By the end of 2011, 616,000 U.S. residents were being treated for ESRD with over

115,000 new cases in that same year (13). There are various etiologies of renal failure including, but not limited to: diabetes mellitus, hypertension, trauma, autoimmune diseases, drug toxicity, genetic disorders, and graft failure (1;6). As of 2012, the most prevalent causes of primary renal failure among the over 90,000 adults awaiting transplant are diabetes mellitus (34.2%), hypertension (25.0%), glomerulonephritis

(14.1%), polycystic kidney diease (8.2%), and other/idiopathic causes (18.5%) (6).

Hemodialysis is the most common treatment measure for ESRD as over two-thirds of

ESRD patients (approx. 430,000) are actively undergoing this treatment (13). In 2009, there were over 90,000 deaths among persons being treated for ESRD (14). The

ESRD mortality rate, coupled with patient costs cresting $42 billion, highlight the need for more effective treatments beyond hemodialysis which accounts for the bulk of treatment costs (14). Renal transplantation is a cheaper and more permanent treatment for ESRD (14). Presently, there are approximately 18,000 renal transplantations per year in the United States (14), but the pace of renal transplantation only covers 20% of the 90,000 persons who are currently on the national waiting list for a new kidney (1).

Rising ESRD patient populations and a limited supply of useable organs have made optimizing the process and results of renal transplantation of the utmost importance in treating the disease. Ideally, all ESRD patients would receive living related donor kidneys (LRDKs), as those organs are associated with the best long term outcomes

(6;15). LRDKs are not exposed to ischemia and are far less likely to exhibit any immune-reactivity in the recipient compared to living unrelated donor (LURD) kidneys or deceased donor kidneys (DDKs). Despite the higher quality of living donor transplants, the dearth of organs is such that approximately 2/3 of the nearly 18,000 renal transplants performed in 2011 were from deceased donors (6). Improvements in organ handling and perfusion pump use have allowed for both better assessment of deceased donor organs (16-20) and the use of donations previously believed to be too aged or damaged for use as a graft (4;16-21). Efforts to increase the number of donations have led to use of organs donated after cardiac death (DCD) and a class of donations from patient populations that were previously ignored as donors (21;22). These expanded criteria donors (ECD) have specific criteria seen in table 1.

Table 1. Expanded Criteria Donor Characteristics (22;23) Donor Age Comorbidities (2 or more) >60 None 50-59 Elevated Creatinine (>1.5 mg/dL) Hypertension history Death by cerebrovascular accident (Stroke) Use of 2+ vasopressors

Lack of suitable organs has kept the prevalence of transplanted kidneys low

(approximately 172,000 as of 2009) (14), but improvements in type matching, immunosuppression, and effective prophylaxis have significantly reduced hyperacute and acute graft rejection as significant causes of graft failure (3;5;23). Graft failure, measured independently of the previously mentioned causes of ESRD, accounts for a full 14.5% of patients awaiting a transplant (6). From 2002-2012, the prevalence of graft failure dropped as a percentage of adults awaiting transplant, from 17.9% to the aforementioned 14.5% (6); however, the number of patients increased grossly, from approximately 9,000 to over 13,000 adults (6), keeping graft failure among the top 5 causes of renal failure and highlighting the necessity of continued improvement of post- transplant monitoring and management.

The decrease in acute rejection events and expansion of the donation pool has changed the clinical picture of acute and chronic adverse events surrounding renal transplantation. Acutely, delayed graft function (DGF) has supplanted hyperacute and acute rejection as the chief post-operative adverse event during transplant. Delayed graft function (DGF) is a negative short-term outcome of renal transplantation. Delayed graft function is a purely clinical diagnosis defined by consensus as requiring more than

1 hemodialysis treatment in the first week post-transplant (4;7;18-20;22;24-27), typically

7 during the post-operative hospital stay. More quantitative metrics of DGF are defined by the severity of oliguria (low urine output) or changes in levels of serum creatinine.

These changes are measured in many ways including direct measurements of serum creatinine, ratios of pre- and post-transplant serum creatinine levels, such as creatinine reduction ratio on day 2 (CCR2%), or increases in GFR (4;22;26;28). The mechanism of

DGF development is poorly understood, but there are several hypotheses on DGF development including DGF as a consequence of: ischemia reperfusion injury (21;24), inflammation and apoptosis (21;29), hypoxia-induced fibrogenesis (7), and poor donor quality (22;24-27;29). Despite its loose definition and the varied ideas of its origin, DGF is an important clinical outcome because it is independently related to the development of chronic rejection also known as chronic allograft dysfunction (CAD) (27;28). Chronic allograft dysfunction (CAD), a state of chronic rejection typified by the histological diagnosis of interstitial fibrosis and tubular atrophy (IFTA), is among the leading causes of graft failure (3;5;23). The combination of chronic inflammation, increased fibrogenesis, and decreased initial nephron mass in patients with DGF create a milieu for histological IFTA diagnosis during the first year post-transplant, making DGF an independent risk factor of CAD (22;23;26;27). Additionally, the use of DCD and ECD organs is associated with greater rates of DGF and development of CAD. With limited organs available for transplant and the lower costs of post transplant management compared to hemodialysis (14), novel methods for addressing the quality of kidney donations and post-transplant management are essential to extending the viability of renal transplants and improving the quality of life of renal transplant recipients.

Clinicians’ steadily improving ability to select and maintain renal grafts have lowered the risk of hyperacute and acute graft failure (4;5;30). Hyperacute graft failure occurs within hours of transplant. It is mediated by the recipient’s preformed antibodies attacking the graft’s endothelial antigens resulting in thrombotic occlusions and ischemic necrosis of the graft (5;31-33). Acute graft failure, also known as acute cellular rejection (ACR), develops in the first week post-transplant as a result of T cells and macrophages infiltrating the kidney parenchyma and causing necrosis (5;31-33). Likewise, a lack or extended interruption of immunosuppressive treatment at any point post-transplant can cause ACR in graft recipients. Immunosuppression, organ handling and short-term management improvements have made the progressive loss of function known as CAD the leading cause of graft failure with approximately 5,000 renal grafts failing every year

(4;5;7;21;23;30;34-39). Interstitial fibrosis and tubular atrophy (IFTA) typifies the progressive loss of function that leads to CAD (5;30;38;40). The development of IFTA is poorly understood; it has been linked to poor donor quality, recipient disease, immunosuppressant toxicity, and most recently to certain gene products

(5;30;38;40;41). IFTA is primarily a histo-pathological diagnosis based on the evaluation of serial graft biopsies and it is most notably marked by a steady loss of nephron mass beyond 6 months post-transplant leading to graft loss (5;30-33;38;40).

The onset of IFTA makes it a useful clinical identifier of chronic rejection (5;31-

33;42;43); however, because it develops slowly, IFTA cannot be used as a predictive tool for diagnosing CAD (3;5;31-33;42;43). Instead, IFTA exists clinically as a negative long term outcome of renal transplantation.

IFTA is a process that begins with tubulointerstitial damage that progresses into fibrotic scarring of renal vasculature and resulting in renal tissue dominated by fibrotic lesions that obliterate nephron tubules and microvasculature (2;3;37;44;45). Initial post- transplant renal damage appears to be secondary to ischemia reperfusion injury, drug nephrotoxicity, acute tubular necrosis and subclinical rejection (2;3;37;44;45). These factors work in conjunction with any present donor abnormalities and recipient comorbidities to potentially damage the kidney (2;3). The kidney undergoes a process similar to epithelial-to-mesenchymal transition (EMT) during which tubular epithelial cells become myofibroblasts through a loss of cell-cell adhesion and E-cadherin expression coupled with tubular basement membrane disruption and increased production of fibrogenic proteins including collagen types I, III and V as well as fibronectin (2;3;40;46-50). As this fibrotic process occurs, renal function remains clinically viable until fibrotic lesions dominate the organ resulting in glomerulosclerosis from lack of circulation to glomeruli and thickening of basement membranes (2;3;40;45-

51). Areas of the kidney affected by fibrosis also are surrounded by inflammatory infiltrate and invading immune cells (lymphocytes and macrophages/monocytes) that contribute to both the overall histo-pathological picture of IFTA and the chronic inflammatory state that provides the milieu for fibrosis to occur (2;3;45;51). Finding methods to target and mitigate the molecular processes important to IFTA development can provide a new means of monitoring grafts for signs of IFTA and even developing treatments for the condition.

Post Transplant Monitoring & Gene Expression Profiles

Post-transplant monitoring of renal graft function remains a major challenge in optimizing renal graft management. Current clinical metrics for monitoring graft function range in effectiveness and invasiveness, but are limited to 3 major categories: plasma levels of drugs & waste products, urine electrolyte & protein levels, and the tissue biopsy (2;38;39;42;52;53). Tracking immunosuppressant plasma levels allow physicians to modify treatments when coupled with information about patient side effects and other metrics of renal function (54;55). Clinicians use plasma levels of the renal waste products creatinine and urea nitrogen (BUN, blood urea nitrogen) to monitor the filtration and secretion capacity of the kidney (38;52-58). The main advantage of measuring plasma levels of drugs and waste products is the ease of drawing and processing blood

(53-55;57;58). The chief disadvantage of measuring these plasma levels is their inability to properly reflect subacute changes in function (53-55;57;58). The plasticity of the kidney will mitigate acute increases in the levels of waste products in the blood (53-

55;57;58). This plasticity masks subclinical and chronic rejection, where there may be histological evidence of rejection in the absence of functional indices (53-55;57;58).

Clinicians are left to rely on rapid increases in creatinine and BUN, indicative of significant renal damage, to diagnose graft dysfunction (53). Furthermore, while acute changes in creatinine levels suggest rejection or organ failure, any creatinine-based diagnoses of renal pathology require a tissue biopsy for confirmation (31;53;59-61).

The inability to identify subclinical rejection via normal clinical surveillance makes managing renal allograft function more difficult than other organs with more direct links

11 between plasma biomarker expression and histological function (31;53;59-61).

Similarly, clinical urine evaluation exhibits the same advantages as plasma analysis in that both methods are minimally invasive (42;53;59;61). Urine electrolytes and protein measurements are effective reporters of the metabolic state of the kidney; however, the accuracy of measurements is limited to the 24hrs surrounding the time the sample was taken and levels measured can vary with several factors including renal disease etiology and recipient dietary management (42;53;59;61).

Histo-pathological analysis of the renal tissue biopsy is presently the gold standard for monitoring graft function (5;38-40;42;43;52;53;59;61). The great advantage of the tissue biopsy is its ability to diagnose inflammation, infection, fibrosis, glomerulosclerosis, and tubular atrophy before the standard clinical renal function markers at any point post- transplant (5;32;38;42;62). The disadvantages of the tissue biopsy are the invasiveness of the procedure, the subjectivity of the analysis (5;42;52;61), and the current clinical approach for using the procedure. Needle-core biopsies are procedures requiring ultrasound guidance and local anesthesia to be most successful (5;42;52;61). The prospect of needle use and the necessary hospital stay following the procedure lowers patient compliance, and subsequently the number of patients who receive the procedure. Single sample variability between pathologists in analyzing renal tissue ranges between 25-50% (5;42;61). Additionally, in most clinical settings, tissue biopsies are done to confirm signs of renal failure from serum-based clinical indicators of renal dysfunction (GFR, creatinine and/or blood urea nitrogen, BUN) further lowering the biopsy’s utility as prospective marker of transplant function (42;53;61). Histologically,

IFTA is not ascribed to any specific etiology, and is noted as Banff category 5 with

12 grades ranging from mild (grade 1) to severe (grade 3) (31-33;63). Grade 1 IFTA is defined as interstitial fibrosis and tubular atrophy in <25% of the renal cortical area sampled (31-33;63). Grade 2 is IFTA between 25 & 50% of the sampled renal cortex, and grade 3 is IFTA greater than 50% of the sampled renal cortex (31-33;63).

Additionally, there is an ungraded IFTA where each element is observed alongside nonspecific vascular and glomerular sclerosis (63).

The challenges with current clinical metrics reveal the necessity for less invasive and more accurate methods of monitoring renal grafts. Gene expression profiling is an increasingly viable means for meeting the challenges of graft monitoring. Gene expression profiles can utilize RNA retrieved from tissue, blood or urine (2-

5;21;30;34;36;52;55;56;64-67), making it an easily accessible and, particularly in the case of urine collection, non-invasive method of evaluating graft function

(30;38;40;52;66;67). RNA expression is preferable because of the relative dynamism compared to DNA and the ease of isolation from many sources. More importantly, RNA expression changes can inform us about protein expression changes and the resultant function of the cell. The direct link between mRNA and protein expression is understood in relation to the central dogma, with studies explicitly illustrating that mRNA and protein levels are correlated as long as the mRNA and corresponding protein have the same cellular stability (68). These profiles represent a snapshot of cellular and, by extension, tissue function that can potentially be the earliest point that IFTA can be identified (64)

(Fig 3).

Figure 3. Diagnosis and rejection timelines. Earlier diagnosis allows greater time for delivering effective treatments. Molecular diagnosis of rejection gives clinicians the most time to effectively treat patients. (9) Presently, messenger RNA (mRNA) profiles have been used to identify acute rejection

(ACR), BK virus nephopathy, and chronic rejection with interstitial fibrosis and tubular atrophy (52). Patterns of mRNA expression linked to IFTA are linked to specific cellular processes including the immune response, inflammation, fibrogenesis, and apoptosis

(2-5). Genes that hallmark these processes including, but not limited to interleukin-10

(IL-10), transforming growth factor beta (TGF-β), and epidermal growth factor (EGF), have been shown to be differentially expressed in IFTA compared to normal allografts

(5;69). Previous work done by Mas and colleagues (2008) has shown a specific pattern of mRNA expression associated with IFTA (2;3;5;38) (Fig 4). Briefly, in that study RNA was isolated from tissue biopsy samples from 23 kidney transplant patients, 17 with histo-pathologically diagnosed IFTA, and 24 normal donor kidneys and prepared for

Affymetrix gene array analysis through a process of reverse transcription followed by in vitro transcription (2). The gene array was analyzed using the Affymetrix Gene Chip

14 operating software and Bioconductor packages in the R programming environment (2).

The gene expression summaries for normal donor (n=24), normal allograft (n=6), and

IFTA (n=17) tissues were estimated from probe-level data using the RMA (Robust

Multiarray Average) method. Genes with significant expression changes across the three groups were identified using the Jonckheere-Terpstra test for trends in gene expression. This test identified genes that increased or decreased in expression from normal donor to normal allograft to IFTA. Gene ontology and interaction analyses were done to identify which cellular processes are most represented, and by extension altered, by the genes that are differentially expressed in IFTA. Reverse transcription quantitative polymerase chain reaction (RT-qPCR) experiments were used to validate the effect sizes and directionality of expression among genes of interest including: transforming growth factor beta (TGF-β), thrombospondin-1 (TSP-1) and the epidermal growth factor receptor (EGFR). More broadly, differentially expressed genes in IFTA cases are involved in the immune response, fibrogenesis, and apoptosis (2;3;5).

Moreover, monitoring gene expression changes has led to the identification of epithelial- mesenchymal transition (EMT), a process once only linked to tumor metastasis, as, potentially, a key process in renal fibrogenesis (70-73). Modulating mRNA function that leads to IFTA can be a potential mechanism to treat or prevent IFTA development post transplant.

Figure 4. Two dimensional hierarchical clustering of differentially expressed mRNA profiles of patient samples comparing CAD to normal allograft kidneys and normal kidney tissue. Green cells represent down-regulated genes and red cells represent up-regulated genes The bars below the dendrogram represent the different patient populations: CAD (light blue), normal renal allograft (dark blue), normal kidney (pink). The bar below the heatmap represents the colors associated with the degree of down- regulation (green) and up-regulation (red). (2)

MicroRNAs and Renal Dysfunction

Messenger RNA from non-invasive sources like urine present a unique challenge for evaluating renal allografts for IFTA. Large single stranded mRNA transcripts are unstable above room temperature, in alkaline pH, and in the presence of RNA nucleases (RNases). Fresh urine samples, especially those from renal transplant recipients, will have all 3 mRNA destabilizing factors. Consequently, the circumstances

16 of urine sample collection in a clinical setting make obtaining and storing urine samples in a manner that would preserve the integrity of any mRNA therein extremely difficult.

Typically, isolating RNA from a standard urine sample collection results in nucleic acids that are short in length (<100nt) and are difficult to assess for gene expression due to problems probing for complete transcripts that span multiple exons. The challenges presented by assessing mRNA expression in urine samples call for alternative methods of using urine to molecularly evaluate renal allograft status. MicroRNAs (miRNA or miRs) provide a less complex and more consistent source of nucleic acids from non- invasive sources, like urine (38;40;52).

MicroRNA is a potential source of nucleic acid biomarkers for a variety of diseases, including renal disease. These short (17-25nt) non-coding RNA were first discovered in

1993, with the isolation and characterization of lin-4 in C. elegans (74). MicroRNA form heterologous pairs with mRNA to modulate gene expression by destabilizing mRNA and repressing translation(40;65;75-80). MicroRNA are differentially expressed in different tissues, found in a variety of body fluids including blood and urine, and their expression is associated with various diseases and cancers (79-81). Their short length means miRs are far more likely to form stabilizing duplexes and remain present in and isolatable in urine where most RNA is in the oligonucleotide form (52). The increased stability compared to mRNA make miRNA a consistently expressed source nucleic acids in the urine (52). The small size makes the miRNA sequences easier to probe and track expression than mRNAs, as miRs have fewer splice variants and no exons to fuse in order to have a complete transcript (52).

MicroRNAs “genes” are typically found in intronic DNA sequences. The majority of miRs are transcribed by RNA polymerase II (75) in response to specific cellular conditions like inflammation, infection, and cell cycle arrest (48;80;82;83). MicroRNA gene transcription produces the primary-miRNA (pri-miR) which is up to 2kb long and contains a hairpin loop with 5’ and 3’ extensions that are capped and polyadenylated respectively (75;83).

These extensions are cleaved by the enzyme Drosha in process called cropping, leaving a hairpin loop 60-120nt long (75;83). The remaining hairpin loop, now called the precursor-miRNA (pre-miR), contains the mature miR sequence, but remains inactive.

The pre-miR is then transported from the nucleus into the cytosol by Exportin-5 to finish maturation and activation (75;83). In the cytosol, the loop of the hairpin is cleaved by the RNase Dicer followed by Argonaute (Ago) protein selecting the strand from the resulting duplex that will be the mature miR (75;83). In some cases, Ago will select either strand in the duplex resulting in the miR having a -3p or -5p suffix depending on its proximity to the 3’- or 5’-end of the pre-miR loop (80;84). Because of this lack of selectivity, there are over 2500 identified miRs originating from just under 1900 miRNA- gene sequences, according to the latest release of miRbase (85).

The mature miR will remain bound to the Ago protein completing the RNA-induced silencing complex (RISC) which also includes Dicer and the GW182 protein, a glycine and tryptophan rich protein that is the main mediator of gene silencing upon activation

(80;83). Once complete, the RISC finds complementary mRNA sequences to bind in the cytoplasm. The complementarity between miR-bound RISC and its target mRNA is largely imperfect except for the 2nd-8th bases on the 5’ end of the mature miR sequence, called the seed sequence (83;84). Canonically, perfect complementarity between the

18 miR seed sequence and the target mRNA is required for miR activity (83;84). The majority of RISC-mRNA duplexes form either in the coding sequence of the mRNA transcript or its 3’-untranslated region (3’-UTR) resulting in mRNA translational repression or degradation (83;84). The transcript remains stable during translational repression, but the activity and recruitment of ribosomes is blocked (80;83). Messenger

RNA degradation involves deadenylating and uncapping the transcript allowing cytosolic exonucleases to break down the mRNA (80;83).

The study of miRs in transplanted kidneys has been mostly limited to biomarker discovery. Many studies have looked at differential expression of miRs in ischemia reperfusion injury, acute rejection, renal fibrosis, and CAD with IFTA. Ischemia reperfusion injury (IRI) is the leading cause of acute kidney injury post transplant

(40;86). IRI stems from the loss of blood supply to the organ followed by reperfusion and subsequent microvascular damage, tubule & interstitial inflammation, and cell death

(36;40;86). Functionally, miRs-126 & -296 protect against microvascular damage in rat models of IRI (40;87). Many other miRs are differentially expressed in several models of

IRI. Chief amongst these is miR-21, which is shown to be up-regulated in murine kidney tissue and tubular epithelial cell models of IRI as well as patients with acute kidney injury (AKI) (40;86;88;89). Other IRI-up-regulated miRs in murine and cell models include miR-20a, miR-146a, miR-199a-3p, miR-214, miR-223, miR-142-5p and miR-

142-3p (40;88;89). Down-regulated miRs in these same models include miR-101a, miR-

187, miR-192, miR-193, miR-194, miR-218 and miR-805 (40;86;88;89). Although these miRs are among the most differentially expressed in different models of IRI, they have not been shown to be the only biomarkers of IRI (86). Additionally, there are no defined

19 roles of these miRs in the molecular responses to IRI that produce the histologically identified AKI.

Acute rejection (ACR) events post transplant are the result of the recipient’s innate and adaptive immune responses to the donor organ. ACR is managed by pharmacological immunosuppression, but miRs can exert control over modulators of both innate and adaptive immunity including immune cell differentiation, cytokine production, antigen processing & presentation, and Toll-like receptor signaling (40;65). The miRNA- biomarkers associated with ACR have come largely from studying human renal allograft biopsy tissue, as well as urine or blood samples from transplant recipients (40;65). In

ACR, miR-142-5p, miR-155, miR-223 and miR-320 all exhibit up-regulation in biopsy tissue compared to controls, while let-7c, miR-30a-3p, and miR-324 exhibited down- regulation in the same biopsies (40;65). Urine cell pellets from patients with ACR showed an up-regulation of miR-10a and a down-regulation of miR-10b and miR-210

(40). Peripheral blood mononuclear cells (PBMCs) taken from patients with ACR showed an up-regulation of miR-223 (40). These studies show how efficacious miRs can be diagnostically for ACR, but they shed little light on the role these miRs have on

ACR development and progression.

From a functional standpoint, miRs role in renal fibrosis is much better delineated than both IRI and ACR; largely because fibrogenesis is a molecular process that is not unique to organ transplantation or renal dysfunction, and is comparatively a far larger field of study. Canonically, both beneficial and pathological fibrosis is mediated by the transforming growth factor-beta-1 (TGF-β1) (40;46-50). TGF-β1 binds the TGF-β type II receptor which phosphorylates the TGF-β type I receptor to induce the function of Smad

20 proteins, the intracellular effectors of TGF-β1 signaling (40;46-50). Several miRs in many tissues have been shown to regulate different portions of the TGF-β signaling system.

Renal fibrosis specifically is a consequence of chronic inflammation that triggers the activation of myofibroblasts that deposit extra cellular matrix proteins into the renal interstitium. These myofibroblasts are both native to the interstitium and can differentiate from tubular epithelium in a process called epithelial-to-mesenchymal transition (EMT). The hallmark of EMT is a loss of E-cadherin and other epithelial proteins as well as a loss of the basal-apical cellular morphology indicative of epithelial cells. The increased deposition of collagens and the loss of tubuloepithelial cells result in an obliteration of nephrons and loss of overall renal function over time.

MicroRNAs affect fibrosis on nearly every level of the canonical fibrosis pathway as well as in renal fibrosis specifically. miR-192 has been shown to up-regulate the expression of E-cadherin and subsequent repression of the E-cadherin regulator, ZEB2 (40;48).

Both of these factors are important in EMT, and miR-192 prevents the loss of E- cadherin that is indicative of EMT (40;48). miR-192 is down-regulated in renal biopsies of persons with diabetic nephropathy which correlates with concurrent increases in tubulointerstitial fibrosis and loss of renal function (40;48). Other miRs that regulate renal fibrosis via EMT modulation are in the miR-200 family (miRs-200a, 200b, 200c, miR-141 & miR-429), which like miR-192 inhibit ZEB2 to decrease fibrogenesis (40;48).

Additionally, the miR-200 family also inhibits ZEB1 resulting in increased E-cadherin expression (40;48). Two members of the miR-200 family, miR-200a and miR-141,

21 directly regulate the TGF-β signaling pathway through inhibiting the TGF-β2 ligand

(40;48).

Another regulator of the TGF-β signaling pathway is the miR-29 family (-29a, b, & c) which is mostly associated with inhibiting fibrosis in the heart, liver, and lungs (40;48). In the kidney, miR-29 inhibits collagen expression and modulates matrix protein function

(40;48). In vitro, TGF-β1 expression will reduce miR-29 expression in tubular epithelium and podocytes; however, the activity of the miR-29 family is not completely uniform

(40;48). Murine and in vitro models of diabetic nephropathy show that miR-29c may up- regulate matrix protein expression and fibrosis (40;48). These findings make delineating the role of miR-29 more difficult and require observing miR-expression in different etiologies of renal fibrosis/failure. miR-21 has several observed roles in renal fibrosis.

Originally observed as cardio-fibrogenic (40;41;48;89;90), miR-21 has exhibited pro- fibrotic renal expression in both human and murine models of fibrosis (41). Within the canonical fibrosis signaling pathway, miR-21 targets smad7, an inhibitory regulator of

TGF-β mediated fibrosis (40;41;48;89-91). miR-21 also has demonstrated effects on peroxisome proliferator-activated receptor-α (PPARα), a regulator of lipid metabolism that mitigates renal injury and fibrogenesis, as well as MPV171, an inhibitor of mitochondrial reactive oxygen species production (40;41;48;90). Having both direct and indirect ties to renal fibrogenesis makes miR-21 a prime molecule for further study and a potential treatment target for treating or preventing fibrosis. Another miR that regulates non-canonical fibrogenic pathways is miR-324. miR-324 is up-regulated in a murine model of renal fibrosis and targets prolyl endopeptidase (48;92), an enzyme that metabolizes angiotensin and produces the anti-fibrotic peptide, Ac-SDKP (92). The pro-

22 fibrotic effects of miR-324 are mitigated by ACE inhibitors and illustrate a TGF-β and

EMT independent mechanism of renal fibrosis.

MicroRNAs and Chronic Allograft Dysfunction

Unlike renal fibrosis, very little is understood about how miRs affect the development of

CAD with IFTA. Differentially expressed miRs in transplant recipients with IFTA compared to those without provide insight into what miRs can be used as prospective diagnostic markers. MicroRNA sequence profiles comparing tissue biopsies from transplant recipients with IFTA to those without show an up-regulation of miR-21, miR-

142-3p &-5p, and miR-506 and a down-regulation of miR-30b and miR-30c (93). These results were from a limited patient population (n=26) that included both living and deceased donor kidneys (93). An array-based miRNA expression profile comparing tissue biopsies from deceased donor transplant recipients with IFTA to those without

IFTA showed an up-regulation of miR-142-3p & miR-32 and a down-regulation of miR-

107, miR-204, and miR-211 (38). Studies that looked at differential miRNA-expression in urine samples either exclusively, or in addition to tissue biopsy miRNA expression showed differential expression of miR-125 (52), miR-203 (52), miR-142-3p (38;52), miR-

204 (38;52), and miR-211 (38;52). These array based studies looked exclusively at deceased donor kidney of varying quality (DCD, ECD, & standard criteria donors, SCD) and validated the array results using RT-qPCR (38;52). The homogenization of donor type (living vs. deceased) is an important factor to consider when evaluating gene expression because the living donor kidneys are not exposed to the same caliber of stresses as deceased donor kidneys. Consequently, living donor kidneys have a much

23 lower potential for ischemia reperfusion injury, acute kidney injury, and delayed graft function (94-96); all of which are correlated to the development of chronic rejection.

Recent work by Scian et. al in 2011 has shown that 3 miRs; miR-142-3p, miR-204, and miR-211; are differentially expressed in both tissue and urine samples from renal transplant patients diagnosed with CAD with IFTA and patients who would later develop

CAD with IFTA (38;52). While these results show some promise for using miRs as part of a new paradigm for prospective graft management, little is understood about how these miRs affect renal function in a manner that would steer the kidney toward CAD with IFTA. The same study that identified miRs-142-3p, -204, & -211 as prospective

CAD with IFTA biomarkers, also used in-silico statistical analysis of the mRNA expression signature from the same study cohort to identify potential targets for these miRs (38). These targets include genes that play a role in the immune, inflammatory, fibrotic, and apoptotic regulatory processes and could provide some insight into the manner in which these IFTA miRs effect the development of IFTA (38).

MicroRNA-142-3p is on chromosome 17 and is significant in the development of acute cellular rejection and may inhibit the immune-suppressing functions of regulatory T

(TReg) cells (40;65). TReg cell inhibition is linked to miR-142-3p’s down-regulation of adenylate cyclase 9 (AC9) (97). This action is also implicated in the oncogenesis of human T-cell acute lymphoblastic leukemia (98). Additionally, miR-142-3p down- regulation is associated with, and potentially diagnostic for human acute myeloid leukemia . miR-142-3p expression is up-regulated in patients with IFTA (40;52). In an effort to identify potential targets for miR-142-3p and the other miRs that are differentially expressed in IFTA, correlation analyses were done comparing the IFTA

24 miR expression profiles to the independently generated mRNA gene expression profile seen previously (Fig. 4). Pearson correlation coefficients and P-values were calculated for each miR/mRNA pairing using differentially expressed IFTA miRs (n = 15) and mRNA gene probe sets (n = 2223). These calculations were followed by permutation analyses between the miR/mRNA pairings to determine the statistical significance of the pairings. Such analyses have the added benefit of developing a null distribution for the data in question. The mRNA IFTA signature genes that are most negatively correlated with miR-142-3p up-regulation play roles in microbial defense, IP3 receptor degradation, and apoptosis (38) (Table 2). Specifically, transmembrane protein 14A (TMEM14A,

Table 2) is involved in apoptosis inhibition through stabilization of the mitochondrial membrane potential (99). Decreased TMEM14A activity as a result of increased miR-

142-3p expression could be a potential cause of Bcl-2 mediated apoptosis in renal cells, a process that can be reasonably linked to IFTA development.

Table 2. miR-142-3p negatively correlated targets. Genes are according to their negative correlation to miR expression. The significance of the negative correlation is delineated by the p-value. The empirical p-value measures the statistical significance of the miR/mRNA pairings compared to a null distribution of randomized miR/mRNA pairings from the same data set (38). Gene Pearson Empirical MicroRNA Probeset p-value Symbol Coefficient p-value hsa-miR-142-3p 210397_at DEFB1 -0.977 1.15E-06 4.68E-05 hsa-miR-142-3p 218477_at TMEM14A -0.966 5.33E-06 4.99E-04 hsa-miR-142-3p 206716_at UMOD -0.965 6.33E-06 6.25E-04 hsa-miR-142-3p 221542_s_at ERLIN2 -0.964 7.20E-06 7.09E-04 hsa-miR-142-3p 204485_s_at TOM1L1 -0.962 8.69E-06 8.34E-04

MicroRNA-204 is down regulated in IFTA (40;52). Found on chromosome 9, miR-204 is associated with maintenance of epithelial integrity when its function is coupled with miR-

211, whose expression is also down-regulated in IFTA (40;100). Both miRs-204 & 211

25 target mRNAs in the TGF-β signaling cascade relating to different diseases (100-102). miR-204 has also been linked to targets involved in the endoplasmic reticulum’s stress response such as Bcl-2-L2, an apoptosis inhibitor (103). Moreover, miR-204 stimulates an anti-autophagy signal in cardiomyocytes and suppresses apoptotic signaling in pulmonary arterial smooth muscle and HeLa cells (40). These targets implicate miR-

204’s down-regulation of both apoptotic and fibrogenic processes. The mRNA CAD-

IFTA signature genes that are most negatively correlated with miR-204 expression play roles in inflammation, apoptosis and fibrogenesis (Table 3). Chief among these are the basic helix-loop-helix family member e41 (BHLHE41) and CD44-standard molecule

(CD44s) (38) (Table 3). BHLHE41is a transcription factor that can act as a transcriptional repressor (104-106). It has several roles including the negative feedback of the interferon-beta (IFN-β) inflammatory response (104). Additionally, BHLHE41 is a demonstrated apoptosis inhibitor in cancer cell models (105;106). CD44s is a cell surface glycoprotein that is involved with cell adhesion and migration. It is not highly expressed in normal renal tissue, but its levels increase following renal injury (107-112).

CD44s acts to up-regulate collagen degradation (107-112), and Rouschop and associates showed that mice lacking CD44 have increased tubular atrophy but decreased fibrosis following obstruction nephropathy (107). CD44s also plays a key role in TGF-β1 induced apoptosis and fibrosis that degrades the peritubular capillary network in mice following obstruction nephropathy (111;112). CD44s’s expression in damaged renal tissue is such that Kers et al. (2004) have used tubular CD44s expression as a surrogate marker for CAD-IFTA and estimated glomerular filtration rate

(eGFR) at 12 months post transplant (113). The functions of miR-204 targets implicate miR-204 down-regulation in increased levels of inflammation, apoptosis, and fibrosis.

Table 3. miR-204 negatively correlated targets. Genes are according to their negative correlation to miR expression. The significance of the negative correlation is delineated by the p-value. The empirical p-value measures the statistical significance of the miR/mRNA pairings compared to a null distribution of randomized miR/mRNA pairings from the same data set (38). Gene Pearson Empirical MicroRNA Probeset p-value Symbol Coefficient p-value hsa-miR-204 219694_at FAM105A -0.873 9.64E-04 6.53E-04 hsa-miR-204 221565_s_at CALHM2 -0.846 2.02E-03 1.44E-03 hsa-miR-204 212796_s_at TBC1D2B -0.842 2.24E-03 1.67E-03 hsa-miR-204 221530_s_at BHLHE41 -0.823 3.42E-03 2.58E-03 hsa-miR-204 203320_at SH2B3 -0.822 3.53E-03 2.75E-03 hsa-miR-204 212063_at CD44 -0.821 3.60E-03 2.88E-03 hsa-miR-204 218870_at ARHGAP15 -0.817 3.88E-03 3.21E-03 miR-211’s link to IFTA is not well understood, but like miR-204, miR-211 expression is down-regulated in IFTA. miR-211 is found on chromosome 15 and is associated with decreased invasion and metastasis of melanoma in addition to its epithelial functions and TGF-β modulation like miR-204 as stated earlier (101;114-117). The mRNA IFTA signature genes that are most negatively correlated with miR-211 expression play roles in mitosis and mitogenic processes (38). Most interesting among these is dual specificity phosphatase 6 (DUSP6), a tumor suppressor that inhibits cytoplasmic ERK2, a key member of the MAP kinase cascade (118). Increased levels of DUSP6 impair cell invasion and EMT (epithelial-mesenchymal transition) while decreasing mitogenic activity and cell survival in cancer cells (118). It is reasonable to hypothesize that rising levels of DUSP6 secondary to decreased miR-211 expression will act as a

27 countermeasure to IFTA development in the chronically injured kidney, further complicating the relationship between these miRs and CAD with IFTA progression.

Table 4. miR-211 negatively correlated targets. Genes are according to their negative correlation to miR expression. The significance of the negative correlation is delineated by the p-value. The empirical p-value measures the statistical significance of the miR/mRNA pairings compared to a null distribution of randomized miR/mRNA pairings from the same data set (38). Gene Pearson Empirical p- MicroRNA Probeset p-value Symbol Coefficient value hsa-miR-211 219544_at C13orf34 -0.831 2.92E-03 1.65E-03 hsa-miR-211 208891_at DUSP6 -0.814 4.15E-03 2.85E-03 hsa-miR-211 203086_at KIF2A -0.79 6.54E-03 5.84E-03 hsa-miR-211 203385_at DGKA -0.786 7.06E-03 6.78E-03 hsa-miR-211 202241_at TRIB1 -0.783 7.45E-03 7.42E-03

The relationship between these CAD with IFTA associated miRs and their in silico- derived gene targets from mRNA CAD-IFTA signature paint a picture of these miRs as more than simple biomarkers, but potentially key factors in disease development. These relationships help shape the hypothesis that CAD-related miR expression directly alters renal cell function and viability in a manner consistent with CAD-IFTA development.

This research seeks to evaluate the indicators of the renal allograft dysfunction in two ways: First, by affirming the pattern of miR expression in CAD-IFTA amongst an expanded and independent cohort of DDK recipients. Second, by evaluating the role miRs -142-3p, -204, & -211 play in the development of CAD-IFTA through studying the interactions between the CAD-IFTA associated miRs and their biologically relevant targets. Looking at CAD-related miR expression in a larger, more inclusive DDK recipient cohort will give greater insight into how these miRs reflect different clinical metrics of renal function. The established pattern of CAD-related miR expression in a

28 wider cohort will be a reflection of both poor clinical function (eGFR<60) and the presence of IFTA pathology. Studying the relationship between the CAD-related miRs and mRNA target expression in patients with CAD-IFTA will show that miR-142-3p, miR-204, & miR-211 are not just markers of CAD-IFTA, but also modulate genes that are important to the disease’s development.

Materials and Methods

Urine Sample miR and mRNA Expression

Urine cell collection and RNA isolation

Urine samples (up to 100mL) were collected from the VCUHS Hume-Lee transplant clinic and centrifuged in 50mL Falcon® tubes at 3000 rpm for 20 minutes at room temperature to isolate the urine cell pellet. The supernatant was discarded, the pellets were re-suspended in 1mL of phosphate buffered saline (PBS) each, and transferred to

1.5mL-microcentrifuge tubes. The re-suspended pellet was spun for 5 minutes at

10,000 rpm and the supernatant was completely removed from the washed pellet in preparation for RNA isolation or storage. For storage, the pellet was then re-suspended in 150μL of PBS and 500μL of RNAlater. The pellet was then stored at either 4oC for up to 24 hours or at -20oC for long term storage. Upon thawing or removal from storage,

500μL of cold (4-8oC) PBS was added to the urine cells in RNAlater before pelleting in a table top centrifuge at 13,300rpm for 5 minutes. The miRNeasy (Qiagen) RNA isolation kit was used to isolate RNA from both fresh and thawed urine cell pellets, according to their total RNA isolation protocol. RNA was eluted in 35μL of nuclease free water and stored at -80oC. RNA was quantified prior to use with the Qubit® RNA Assay kit and fluorometer (Life Technologies).

Reverse transcription: miR & mRNA

Single tube miR reactions were performed to synthesize cDNA using the TaqMan

MicroRNA reverse transcription kit (Life Technologies), according to the manufacturer’s protocol. Briefly, 10 ng of RNA (2 ng/μl) were combined with 3.0 μl 5X RT primer,

0.15 μl 100 mM dNTPs, 1.0 μl MultiScribe Reverse Transcriptase (50 U/μl), 1.5 μl 10X

RT buffer and 0.19 μl RNase inhibitor (20 U/μl) in a total reaction volume of 15 μl.

Reactions were run on a 96-well plate at 16 °C for 30 min, 42 °C for 30 min and 85 °C for 5 min.

Messenger RNA-cDNA synthesis were performed according to manufacturer's recommendation (Life Technologies). Briefly, RNA was reverse transcribed using the high capacity cDNA reverse transcription kit (Life Technologies) . 10μl of total RNA (2-

50 ng/μl) was combined with 2μl RT buffer (10X), 1μl MultiScribe Reverse Transcriptase

(50 U/μl), 0.8μl dNTP mixture (25X), 2μL Random Hexamer primer (10X), and 1 μl

RNase inhibitor (0.25 U/μl) in a total reaction volume of 20 μl. Reactions were run in a

96-well plate at 25 °C for 10 min, 37 °C for 120 min, and 85 °C for 5 min.

miR & mRNA Expression

The single tube miR expressions were detected using TaqMan chemistry according to the manufacturer's protocol (Life Technologies). 1.3 μl of cDNA were combined with

0.25 μl 20X TaqMan MicroRNA assay and 5.0 μl 2X TaqMan Universal PCR MasterMix,

No AmpErase UNG in a final volume of 10 μl. The reactions were run in triplicate in 384- well plates on the 7900HT Real-Time PCR System according to manufacturer's

31 recommendations (Life Technologies). The raw fluorescence values were recorded by the SDS (Sequence Detection System) 2.3 software and converted into Cq values using the RQ (Relative Quantitation) Manager 1.2 software. PCR efficiency for each reaction was calculated by the LinRegPCR program (119). Expression of each miR was normalized against the geometric mean of 3 housekeeping small RNAs, RNU44,

RNU48, and RNU66 using the Efficiency-ΔΔCt algorithm. These three RNAs were chosen because of their respective high, moderate, and low levels of stable expression in renal tissue and urine. Samples that failed to amplify these housekeepers (n=7) were excluded from analyses.

Prior to gene expression analysis cDNA was diluted 1:3 and genes of interest were detected using TaqMan chemistry according to the manufacturer's protocol (Life

Technologies). Each reaction included 1.5 μL diluted cDNA combined with 0.25 μL of

20X TaqMan gene expression assay, and 5μL of TaqMan Gene Expression Master Mix

(Life Technologies). The reactions were run in triplicate in 384-well plates on the

7900HT Real-Time PCR System according to manufacturer's recommendations (Life

Technologies). The raw fluorescence values were recorded by the SDS 2.3 software and converted into Cq values using the RQ Manager 1.2 software. PCR efficiency for each reaction was calculated by the LinRegPCR program. Gene expression was normalized against PPIA (peptidylprolyl isomerase) and GAPDH (glyceraldehyde-3- phosphate dehydrogenase) using the Efficiency-ΔΔCt algorithm.

Statistical Approaches

The expression values were set to a normal distribution by exponential transformation, and the cohort average expression (approx. 1) was subtracted from the average expression in each patient group to plot the expression as an increase or decrease from the cohort average. Post-hoc F-tests were done to determine significant expression differences between CAD and normal allograft patient groups. One-way ANOVA was done to test the significant differences in expression between all the groups. All tests were considered significant at or below the Bonferroni corrected p=0.0125, adjusting for the number of tests (0.05/4).

Cell Culture miR and mRNA Expression

RNA isolation

HEK293 cells were loosened from culture dishes using trypsin, and pelleted at 2,000 rpm for 5 minutes. The trypsin and media were removed from the cell pellet; and the pellet was either resuspended in RNAlater for storage, or immediately isolated with the

Ambion® mirVana RNA isolation kit (Life Technologies) using their total RNA isolation protocol.

Reverse transcription: miR & mRNA

In higher abundance (>90ng/ul) RNA samples, typically isolated from HEK293 cell culture, 10μL of each of the 5X RT primers tested (RNU44, RNU48, RNU66, miR-142-

3p, miR-204, & miR-211) were pooled into 60μL of nuclease free water and 880μL of

Tris-EDTA (TE) buffer (pH 8.0) to make 1mL of pooled primers for up to 150 reverse

33 transcriptions. At least 350ng (approx. 90ng/μL) were added to 6μL of pooled 5X RT primers, 0.30 μl 100 mM dNTPs, 3.0 μl MultiScribe Reverse Transcriptase (50 U/μl),

1.5 μl 10X RT buffer and 0.19 μl RNase inhibitor (20 U/μl) in a total reaction volume of

15 μl. The reaction plating and temperature were the same as the low concentration miR reverse transcription above.

100 ng/μl) was combined with 2μl RT buffer (10X), 1μl MultiScribe Reverse

Transcriptase (50 U/μl), 0.8μl dNTP mixture (25X), 2μL Random Hexamer primer (10X), and 1 μl RNase inhibitor (0.25 U/μl) in a total reaction volume of 20 μl. Reactions were run in a 96-well plate at 25 °C for 10 min, 37 °C for 120 min, and 85 °C for 5 min.

miR & mRNA Expression

The pooled primer miR expressions were detected using TaqMan chemistry according to the manufacturer's protocol (Life Technologies). For each reaction 0.08μl of cDNA were combined with 0.25μl 20X TaqMan MicroRNA assay and 5.0μl 2X TaqMan

Universal PCR MasterMix, No AmpErase UNG in a final volume of 10μl. The reactions were run in triplicate in 384-well plates on the 7900HT Real-Time PCR System according to manufacturer's recommendations (Life Technologies). The raw fluorescence values were recorded by the SDS 2.3 software and that data was converted into Cq values using the RQ Manager 1.2 software. PCR efficiency for each

34 reaction was calculated by the LinRegPCR program. Expression of each miR was normalized against the geometric mean of 3 housekeeping small RNAs, RNU44,

RNU48, and RNU66 using the Efficiency-ΔΔCt algorithm. Samples that failed to amplify these housekeepers were excluded from analyses.

Random hexamer cDNA was diluted 1:5 and genes of interest were detected using

TaqMan chemistry according to the manufacturer's protocol (Life Technologies). Each reaction comprised 1.5 μL diluted cDNA combined with 0.25 μL of 20X TaqMan gene expression assay, and 5μL of TaqMan Gene Expression Master Mix (Life

Technologies). The reactions were run in triplicate in 384-well plates on the 7900HT

Real-Time PCR System according to manufacturer's recommendations (Life

Technologies). The raw fluorescence values were recorded by the SDS 2.3 software and that data was converted into Cq values using the RQ Manager 1.2 software. PCR efficiency for each reaction was calculated by the LinRegPCR program. Gene expression was normalized against PPIA (peptidylprolyl isomerase) using the Efficiency-

ΔΔCq algorithm.

Statistical Approaches

The expression values were plotted as either increased or decreased from the cohort median, typically a value of 1. Paired t-tests between miR-transfected and empty vector control conditions were used to establish the significance of expression results, with p<0.05 as the threshold for significance.

Cloning Expression Plasmids

miR sequences

Primers were designed using the Ensembl browser

(http://useast.ensembl.org/index.html) to identify the pre-miR sequence and flanking

DNA, and NCBI Primer-Blast (http://www.ncbi.nlm.nih.gov/tools/primer-blast/) was used to probe the DNA surrounding pre-miR sequence for primers. The primers were selected for specificity and size of gene product (Table 5).

Table 5 - Primer sequences for miR-142-3p & miR-211. Primer Sequence Product size Forward 5'-CATGCTGAGTCACCGCCC-3' miR-142-3p 484 bp Reverse 5'-TCAGGGTTCCACATGTCCAG-3' Forward 5'-TGATGCTGCAGAGTGGGTAG-3' miR-211 475 bp Reverse 5'-GACATGCATTGAGGGTCTGC-3'

Restriction enzyme sequences were added to the primers to allow for sequence selection and efficient cloning. A Sac-I site (5’-GAGCTC-3’) was added to the 5’ end of the forward primer, while a BamH-I restriction site (5’-GGATCC-3’) was added to the 3’ end of the reverse primer. The 25 nanomoles of lyophilized and desalted primers were made by Invitrogen. The primers were diluted to 20μM with TE (pH = 8.0) before being used in a gradient PCR using the Immolase PCR kit (Bioline) according to the manufacturer’s recommended protocol. Upon finding the optimum PCR temperature, genomic DNA was probed to produce the pre-miR sequence. The pre-miR sequence was size selected on an agarose gel and the DNA was digested using Sac-I (New

England Biolabs, NEB) and BamH-I (NEB) to produce cohesive ends. The same cohesive ends were produced on the pmR-ZsGreen expression plasmid, and the pre-

36 miR sequence was ligated into the pmR-ZsGreen (4714bp) plasmid using T4 Ligase

(NEB) according to their 3:1 molar ratio formula and recommended experimental conditions. Challenges with cloning the pre-miR-204 sequence necessitated ordering the 162bp pre-miR-204 sequence with the Sac-I and BamH-I cohesive ends from

Eurofins/Operon in the PCR-2.1 vector (Table 6). The pre-miR-204 was cut out of the

PCR-2.1 plasmid and ligated into the pmR-ZsGreen in the same manner as the PCR products mentioned above.

Table 6. pre-miR-204 sequence with cohesive ends. Restriction sites are underlined and mature sequence in bold. Name Sequence Size 5’-TATAGAGCTCTGATCGCGTACCCATGGCTACAGTCTTTC TTCATGTGACTCGTGGACTTCCCTTTGTCATCCTATGCCTG miR-204 AGAATATATGAAGGAGGCTGGGAAGGCAAAGGGACGTTCA (162 bp) Sac+Bam ATTGTCATCACTGGCATCTTTTTTGATCATTGGGATCCTATA

All cloned plasmids were transformed into E. Coli by heat shocking the bacteria in medium seeded with the newly ligated pre-miR-ZsGreen. The transformed bacteria were grown overnight on kanamycin-selective agar plates. Colonies from the selective plate were amplified in liquid medium for 16-24 hours, after which their DNA was collected using the Qiagen Mini-Prep DNA Isolation Kit (Qiagen) according to their recommended protocol. DNA concentration was measured by nanodrop. The DNA was then digested and gel purified to verify the presence of the cloned insert. All cloned sequences were verified by the Nucleic Acid Research Facility (NARF) and Virginia

Commonwealth University.

3’-UTR sequences

Eight 3’-UTR sequences were cloned into the pmiR-Glo dual luciferase vector

(Promega), the 4 miR targets of interest: TMEM14A (142-3p), CD44 (miR-204),

BHLHE41 (miR-204), & DUSP6 (miR-211); 3 genes that are known targets for each miR: ADCY9, BCL2L2, & TGF-βR2; and one nonsense sham sequence (Table 7).

These sequences contain the miR seed sequence and 70bp flanks up- and downstream of the seed sequence. Each of these sequences was ordered from Eurofins/Operon in the PCR-2.1 expression plasmid. The plasmids were amplified by the same mini-prep procedure described above, and digested with Sac-I and Xba-I (NEB) to make the 3’-

UTR insert with cohesive ends. These inserts were individually ligated into the pmiR-Glo plasmid and grown using a mini-prep procedure. The DNA was then digested to verify the presence of the insert and the cloned sequences were verified by the NARF at

Virginia Commonwealth University.

Table 7. Target 3'-UTR Sequences. Seed complements in red miR Target Sequence

5'-GCGATCGCCAGTTATGTCTGGGATGTTTTAATCCT CCAAGGGGTTGAGCTGGGAGAGCCCCCAAGCCAAG TMEM14A GATTAGTAACACTACTGGGTAGGCGAGTGCATGGCC CTTTCTGTCTTATGCCAGATAAAAGTGACTCTCCCTTT GTCTTTGCTCGAG-3' miR-142-3p 5'-GCGATCGCGGTTGTGACAATACCTCTTGCTTCTAA AGAATGTATTATAAAACACCGCAGATTTTTTTTTTTCC ADCY9 TTAAAAAACACTACCTGATGCTTTCCTTGTTCGTGGG GATTGTGGTCACATGAAGCTCTTTCTGCATCAGTATT AAGGTGTATATTCTCGAG-3' 5'-GCGATCGCAGAGTGAGAAGTTAAAATACCCTTAAG GAGGTTCAAGCAGAGTGAGAAGTTAAAATACCCTTAA BHLHE41 GGTCTTTAAGGGAGGAAGTGTAATAGATGCACGACA GGCATAAACAAGAACAACAAAACAGGTGTTATGTGTA CATTCGGAGTTCTCGAG-3' 5'-GCGATCGCCAGTGTCTGTTCTTGATGCAGTTGCTA TTTAGGATGAGTTAAGTGCCTGGGGAGTCCCTCAAAA miR-204 CD44 GGTTAAAGGGATTCCCATCATTGGAATCTTATCACCA GATAGGCAAGTTTATGACCAAACAAGAGAGTACTGGC TTTATCCTCTAACTCGAG-3' 5'-GCGATCGCAAATAAGGGAAATAAATGTAATTGCCA TTTTTCAAAGATTAAGTAGGAGGAGAGGGGTTTCTTG BCL2L2 CTCTCCAGAGCCCAAAGGGACAAATAGGGACTTTGT TTAGGCCAAGGAAGGAGCGGAAGTAGGGCAACTCG GTCCTGCGATTATTAATCCCACTCCCTCGAG-3' 5'-GCGATCGCCTTGAATCACTTGACAGTGTTTGTTTG AATTGTGTTTGTTTTTTCCTTTGATGGGCTTAAAAGAA DUSP6 ATTATCCAAAGGGAGAAAGAGCAGTATGCCACTTCTT AAAACAGAACAAAACAAAAAAAGAAAATTGTGCTCTT TTCTAATCCAACTCGAG-3' miR-211 5'-GCGATCGCTTATTCAAGAAAAAAGACCAAGGAATA ACATTCTGTAGTTCCTAAAAATACTGACTTTTTTCACT TGFBR2 ACTATACATAAAGGGAAAGTTTTATTCTTTTATGGAAC ACTTCAGCTGTACTCATGTATTAAAATAGGAATGTGAA TGCTATATACTCTTCTCGAG-3' 5’-GCGATCGCCCTGGCCCTCTTCCATAACTGCGGTTA GACCTCCGCATAGATCTCCGTCCCCGCTCCCCTCTG Sham GAGTACGATTCAAGTGTGCGCAGATTAAGTCATTATC GAACCTAGAGCATACTGGTTGCGTAGGGGTTATTGCA GTAAGCCAAAAGTCTCGAG-3’

Transfection

Six-well plates

Transfections for miR and gene expression assays were done in 6-well plates.

Approximately 700,000 HEK293 cells were plated in each well in Dulbecco’s Modified

Eagle’s Medium (DMEM) with 10% fetal bovine serum (FBS) without antibiotics 24 hours before transfection. The cells were over 90% confluent at the time of transfection.

The pre-miR plasmid or empty vector controls were diluted into 250μL of Opti-MEM without serum while Lipofectamine 2000 (Invitrogen) was diluted into 250μL of Opti-

MEM without serum. The volume of lipofectamine was determined by a 1:3 ratio of micro-liters of lipofectamine to micrograms of DNA transfected. The individual Opti-

MEM solutions are incubated at room temp for 5-10 minutes before being combined to make 500uL of DNA and lipofectamine in Opti-MEM. The new solution is incubated at room temperature for 30 minutes to allow the lipofectamine to complex with the DNA.

The 500μL of DNA-Lipofectamine solution is added to 2.5mL of media without antibiotics and the reactions were incubated at 37oC for 24 hours, after which the cells were collected and their RNA isolated.

96-well Plates

Transfections for luciferase assays were done in 96 well plates. Each reaction condition was transfected in triplicate. In each reaction, 75,000 HEK293 cells were seeded in each well 24 hours prior to transfection. For each reaction, 10ng of pre-miR plasmid or empty vector control, 6.67ng of pmiR-Glo from a target of interest or sham control, and

40 where applicable, 10pmol of anti-miR inhibitor or negative control (Ambion) were diluted in Opti-MEM without serum to 25 μL. Likewise 0.3μL of Lipofectamine 2000 was diluted into 24.7 μL of Opti-MEM without serum. After room temperature incubations of 5 minutes individually, and 30 minutes together, 50 μL of the DNA-Lipofectamine solution is plated onto cells in 100 μL of media without antibiotics and the reaction goes on for

24 hours at 37oC after which, cells are examined for the fluorescent marker of the pre- miR plasmid and harvested.

Luciferase Assays

The luciferase assays were done according to the Dual-Glo® Luciferase Assay

(Promega) microplate procedure. First, the medium on the cells was removed from each well and replaced with 75 μL of medium without antibiotics. After equilibrating to room temperature, 75 μL of the Dual-Glo® Luciferase reagent is added to the cells in media.

The entire mixture is agitated by pipetting to aid lysis and the mixture is transferred to the Lumitrac 200 96-well luminometer plate (Greiner® Bio-One). The lysed cells were incubated in the luciferase reagent for 10 minutes and then firefly luciferase activity was measured in a Wallac Victor II luminometer (PerkinElmer). Next, 75 μL of Dual-Glo®

Stop & Glo® Reagent (1 μL substrate : 100 μL buffer) is added to each well and mixed by pipetting. After a 10 minute incubation, the renilla luciferase activity was measured.

Background correction was done by subtracting the average firefly luciferase reading of mock transfections (Lipofectamine and cells only) from the raw firefly luciferase measurements and doing the same with the raw renilla readings. The ratio of firefly to renilla (L/R) was taken for each well and the average of the triplicates for each condition

41 was taken. The ratios were corrected to a percentage of empty vector control expression, or in the case of the anti-miR, negative control expression. The normalized ratios were then plotted relative to the luminescence of the corrected miR-sham control pairing. A paired Student’s t-test was used to calculate significant mean differences between each condition and its sham control.

Results & Discussion

Donor Quality and Gene Expression in Delayed Graft Function

DGF is a purely clinical diagnosis with no chemical or molecular measure. DGF is loosely defined as the recipient’s need for dialysis during the 1st post-operative week. Its development is linked to hypoxia and ischemia reperfusion injury. An examination of a

VCUHS patient cohort of deceased donor kidney (DDK) recipients (n=147) showed that poor donor quality (Table 8) is also associated with DGF. Our patient cohort had 54

DDK recipients (36.7%) diagnosed with DGF. Grossly, the distribution of DGF diagnoses was split evenly between standard criteria donors (SCD) and extended criteria donor/donations after cardiac death (ECD/DCD) (Table 8). However, as a percentage of patients within the specific donor quality groups, only 27.1% of SCD recipients were diagnosed with DGF, 50% of ECD, and nearly 60% of DCD patients within our patient cohort were diagnosed with DGF (Table 5). Donor quality is also significantly associated with use of perfusion pumps, where 70.8% & 81.5% of poor quality kidneys received pump treatment compared to less than half of SCD kidneys

(Table 5). Perfusion pumps are used to both preserve and, in some cases, test the quality of the donor organs (18;19). The use of perfusion pumps in our patient cohort was not associated with a statistically significant decrease in DGF diagnosis.

Table 8. DGF diagnoses amongst 147 DDK recipients according to donor quality and pump use. Significance was determined by Chi-squared analyses of Donor Quality vs DGF & Pump use >360 minutes. *p-value <0.0028 to account for multiple testing no DGF DGF Donor Quality Totals %DGF %Pump no pump Pump no pump SCD 33 37 9 17 96 27.1* 47.9* ECD 4 8 9 3 24 50.0* 70.8* DCD 0 11 11 5 27 59.3* 81.5* Totals 37 56 29 25 147 36.7 57.8 %SCD 89.2 66.1 31.0* 68.0 α = 0.0028 %ECD 10.8 14.3 31.0* 12.0 df = 6 %DCD 0.0 19.6 37.9* 20.0

Clinical data on donor quality affirms a link between donor quality and DGF development. Gene expression analysis of DGF was done to compare expression in

DGF vs. normal function, assess the effects of reperfusion, and look at post operative changes that may result in DGF. A total of 118 biopsy tissue samples from 59 DDKs taken pre-implantation (K1) and 60 minutes post-perfusion (K2) shows differential expression of 146 genes in patients with DGF compared to those without DGF.

IngenuityTM pathway analysis of these genes show a picture of DGF as both an anti- inflammatory response, with up-regulation of genes like SOCS1 (suppressor of cytokine signaling 1), and anti-proliferative with increased expression of tumor suppressor MZF1

(myeloid zinc-finger protein 1) & PDE4B (cAMP-specific phosphodiesterase); and decreased expression of the mitogenic signal transducer, PI3K (phosphoinositol-3- kinase) (Fig. 5).

Figure 5. Ingenuity generated network of DGF associated genes, down-reglated genes in green, up- regulated genes in red. Down-regulated genes include PI3-K complex, RTK2, & ABCA1. Up-regulated genes include SOCS1, PDE4B, & APOBEC3A Further analysis of genes expressed in both the K1 and K2 samples was done to develop a perioperative gene expression pattern associated with DGF. Comparing array data from 91 K1 samples of patients with DGF to those without showed 190

45 differentially expressed genes. A similar comparison of DGF vs. no-DGF expression in

37 K2 samples resulted in 127 differentially expressed genes. A ToppGene analysis of the 190 genes differentially expressed in K1 samples with DGF and the 127 genes differentially expressed in K2 samples with DGF showed 18 genes with common gene expression in the 2 cohorts (Fig. 6A). While these 18 genes were not mapped to a particular biological process, the gene expression signature (Fig. 6B) of this analysis shows the potential to be either predictive or diagnostic for DGF perioperatively.

2.5 Pre-implantation Post-reperfusion 2.0

1.5

1.0

0.5

0.0

-0.5

-1.0 Fold ChangeFold Expression -1.5

-2.0 SST HPN GPX3 GPX3 APOE MPST SMG5 MAGI1 BARD1 FCGRT TRADD TRADD TSEN34 ALDH6A1 ANGPTL2 CSNK2A2 RAPGEF3 EIF4EBP2 FAM108A1 Figure 6. ToppGene analysis of differentially expressed genes in DGF. (A) Diagram of commonly expressed genes in K1 DGF vs. no-DGF (190) and K2 DGF vs. non-DGF (127). (B) Differential expression of DGF vs. non-DGF genes commonly expressed in K1 (pre-implantation, blue) and K2 (post-reperfusion, yellow)

To analyze the effect on reperfusion and examine the genes implicated in ischemia reperfusion injury, the differential gene expression was compared between the K1 and

K2 samples from patients that developed DGF. IngenuityTM gene ontology analysis of the differentially expressed genes showed an up-regulation of several inflammatory processes (Fig. 7). The activation of these cellular processes present an interesting juxtaposition with the miRs that have altered expression in ischemia reperfusion injury.

Of particular interest is miR-21, which is both up-regulated in IRI models (40;86;88;89), and directly targets PPARα in murine models of renal fibrosis (41). The abrogation of the PPARα signaling leads to higher levels of COX2 (cyclooxygenase 2) and production of pro-inflammatory prostaglandins (Fig. 8). Alterations of inflammatory signaling seen in DGF may be attributed to miR-21 activity as well as other miRs indicated in IRI, including miR-142-3p (40;86;88;89), meaning that these miRs could be activating the inflammation of IRI or perpetuating a positive feedback loop resulting in the chronic inflammation that sets the stage for CAD with IFTA. Overall, these results further characterize the kidney’s reaction to recipient reperfusion, and situate ischemia reperfusion injury as a hyper-inflammatory state that can potentially trigger tissue death and fibrogenesis.

Figure 7. Gene ontology for delayed graft function comparing pre-implantation vs. post-reperfusion. Statistically significant (q < 0.05) differentially expressed genes were analyzed for their presence in canonical pathways. A Fisher’s exact test was used to calculate the p-value of the comparison between the proportion of differentially expressed genes within a pathway in the K2 dataset to the proportion of differentially expressed pathway genes in the K1 dataset. Blue bars represent the negative logarithmic values of the significance level beyond the threshold of p = 0.05. Orange points represent the log- transformed ratio of the number of differentially expressed genes in the pathway to the total number of genes in the pathway.

Figure 8 – IngenuityTM PPAR signaling canonical pathway, annotated with differential expression from IRI study. Down-reglated genes in green, up-regulated genes in red. Note the up-regulation of the inducible inflammatory enzyme COX2 and its transcription factors, c-Fos & c-Jun. Array data from post-perfusion (K2) biopsy tissue from the 59 DDK recipients comparing patients with DGF to those without DGF, showed 673 differentially expressed genes.

Gene ontology shows the differentially expressed genes from this cohort are associated with both pro- and anti-apoptotic processes (Fig. 9). Network analysis shows the expression pattern of these genes is consistent with increased cell survival, but decreased proliferation (Fig. 10). Up-regulation of MAP-kinases, purine metabolism enzymes (adenosine deaminase & adenosine phosphoribosyl transferase), translation initiators (EIF4EBP2), transcription activators (CDK8) are all consistent with cell

50 survival. However, up-regulation of PTEN (phosphatase and tensin homolog), which initiates cell cycle arrest, coupled with down-regulation of actin depolymerizers (LIMK1-

2), glucose transporters, and anti-apoptotic proteins like Bcl-3, shows a cell that is decidedly not mitogenic. The impression from these expression patterns illustrates that the post-reperfusion kidney tissue in DGF is mounting a response to the stress of IRI, but that response is not robust, or directed toward cell survival and normal proliferation enough to stave off post-operative dysfunction.

Figure 9. Gene ontology for K2 samples (n=59) DGF vs. no-DGF. Statistically significant (q < 0.05) differentially expressed genes were analyzed for their presence in canonical pathways. A Fisher’s exact test was used to calculate the p-value of the comparison between the proportion of differentially expressed genes within a pathway in the K2-DGF dataset to the proportion of differentially expressed pathway genes in the K2-non-DGF dataset. Blue bars represent the negative logarithmic values of the significance level beyond the threshold of p = 0.05.

Figure 10. Ingenuity™ network from K2 array data comparing DGF to no-DGF. Down-reglated genes in green, up-regulated genes in red.

In summary, the gene expression patterns validate the clinical findings resulting from the organ transplantation at a molecular level, thus having important early prognostic value, i.e. a kidney without blood flow experiences inflammatory and pro-apoptotic stresses that can be mitigated upon reperfusion. The ability of the tissue to mount and then abrogate and appropriate inflammatory response followed by cell growth and proliferation might be the driving forces behind whether or not the implanted kidney tissue will begin to function properly, or will scar and eventually lose its functionality.

While not completely predictive of graft failure, the molecular changes seen in the early modes of transplant dysfunction indicate the beginnings of the chronic dysfunction results in renal transplant loss.

Independent validation of CAD-related microRNA expression in urine samples

Research previously published work by Scian and colleagues (2011) had shown that 56 tissue-specific miRs are associated with the development of IFTA. (38). Specifically, genome wide miR expression profiles of IFTA were generated on Illumina BeadChip arrays using total RNA isolated from a biopsy tissue from 18 deceased donor transplant recipients, 13 of whom were diagnosed with IFTA, during the first year of their cadaveric renal transplants. In this study, the authors found that miR-32, miR-107, miR-142-3p, miR-204, and miR-211 were the top differentially expressed miRs. Our microarray results were validated on the same samples using the real time PCR approach.

Furthermore, the expression of these miRs was tested in an additional 27 biopsies, 19 of which had diagnosed IFTA, and paired urine samples from 7 patients with IFTA and

7 with normal allograft function (NA). We observed excellent agreement in the

53 expression of miR-142-3p, miR-204, and miR-211 between biopsies and urine samples

(38). The validation of miR expression was analyzed using RNU48 as an endogenous control for normalization. Fold change was calculated from threshold cycle (Ct) values using the ΔΔCt method with the equations below:

Equation 2. Fold change = 2-ΔΔCt

Equation 3. ΔΔCt = ΔCt of IFTA – ΔCt of NA (38). Hierarchical clustering of urine samples ΔCt values was done in a prospective validation set of 108 urine samples from 36 DDK transplant recipients where each patient provided a urine sample at 3,9, and 12 months post transplant (38). The clustering validated the pattern of miR expression associated with IFTA. The approaches done in previous work evaluating CAD-related miR expression assessed miR-142-3p, miR-204, and miR-211 as biomarkers of CAD-IFTA, but did not truly delve into their role in disease development. An independent validation of CAD-related miR expression in a larger more varied DDK recipient cohort, using more stringent methods of evaluating differential expression, is the first step in elucidating the role these miRs play in the development or progression of CAD-IFTA.

An independent verification of CAD-related miR expression was done using RT-qPCR of 155 urine samples from 134 patients collected from discarded urine samples. The sample cohort was divided into 4 categories based on renal dysfunction and histo- pathologically diagnosed IFTA (Table 9). Renal dysfunction is defined as a creatinine- based eGFR < 60 ml/min/1.73m2 at the time of sample collection because that is the threshold for sufficient single-kidney function post transplant (11;12). IFTA is defined as

Banff Category 5 grades 0-3, where the presence of any Banff Category 5 diagnosis qualifies as present (Table 9).

Table 9 - Sample cohort classifications. The cohort was divided into 4 groups based on their renal function as assessed by eGFR and any diagnosed interstitial fibrosis and tubular atrophy (IFTA). Group 1 (CAD-IFTA) is associated with the worst renal function and graft prognosis, and is the focal diagnosis of these studies. Grou IFTA Renal Name eGFR p pathology Function/Prognosis <60 1 CAD-IFTA Present Worst ml/min/1.73m2 Renal <60 2 Absent Poor Dysfunction ml/min/1.73m2 >60 3 IFTA Present Adequate-Poor ml/min/1.73m2 >60 4 Normal Allograft Absent Good-Adequate ml/min/1.73m2

The geometric mean of 3 endogenous controls, RNU44, RNU48, and RNU66 was used for normalization. Relative expression was calculated from threshold cycle (Ct) values using the ΔΔCt method with the equations below:

Equation 4. ΔΔCt = ΔCt of Sample – median ΔCt of cohort Equation 5. Relative Expression = Efficiency-ΔΔCt The PCR efficiency was determined using the LinRegPCR program that utilizes the

Log(fluorescence) per cycle from qPCR experiments to make a linear regression model that calculates the PCR efficiency of each reaction (119). This method proves for a more accurate assessment of measuring relative expression rather than assuming an efficiency of 2. Calculating the relative expression of the miRs using a ΔΔCt value based on the cohort median offers two advantages: 1) minimizes the potential bias introduced by outlying PCR reactions and 2) provides an individual expression value for each sample. The relative expression values for each gene in the sample were further

55 assessed for normality by the Kolmogorov-Smirnov test and genes whose expression deviated from normal distribution, were exponentially transformed to approximate normal distribution. Prior to the actual analyses, the normalized expressions of miR-

142-3p, -204, and -211 were fitted into a forward step-wise regression model to assess the impact of potential demographic covariates (i.e. age, sex, race, donor quality) on the expression of these molecules. These covariates had no significant effect on the expression of miR-142 (ANOVA df=7, F=1.67, p=0.12), miR-204 (ANOVA df=7, F=1.72, p=0.11) and miR-211 (ANOVA df=7, F=0.46, p=0.86) and therefore were not included in the main analysis.

The urine samples from group 1 (CAD with IFTA) showed up-regulation of miR-142-3p and down regulation of miR-204 and miR-211 compared to normal allograft (Fig. 11).

The samples from group 2 (eGFR<60 ml/min/1.73m2) showed decreased expression of all the CAD-related miRs. The samples from group 3 (IFTA pathology) showed an up- regulation of miR-204 and miR-211, and a down-regulation of miR-142. The samples from the normal allograft group (group 4) showed a down-regulation of miR-142-3p, and up-regulation of miR-204 and miR-211 compared to the CAD-IFTA group (Fig. 11).

Each miR’s expression was analyzed by one-way ANOVA to assess the significance of the differential expression between groups. miR-142-3p (df=3, F=2.91, p=0.036) showed a significant expression differences between the patient groups. Post-hoc F- tests were done to assess significant differences between the patient groups. A

Bonferroni correction for multiple testing lowered the threshold for significance, α, to

0.0125 (0.05/4), and we observed differential miR-204 expression between group 1

(CAD-IFTA) and group 3 (IFTA only) (p=0.009). A similar difference in miR-204

56 expression was observed between the CAD-IFTA group and the normal allograft group

(p=0.0104). A difference in miR-211 expression was also seen between the CAD-IFTA group and normal allograft group (p=0.0037)

Urine Sample miR Expression 0.25

0.2

0.15

0.1

0.05 miR-142

0 miR-204 miR-211 -0.05 Adj. Relative Expression Relative Adj. -0.1

-0.15

-0.2 CAD+IF/TA, n=31 GFR<60, n=57 Banff 5, n=20 NA, n=46

Figure 11 - Relative expression of CAD-related miRs in urine samples of transplant recipients. Relative expression values were exponentially transformed to approximate a normal distribution. The normally distributed expression values were then plotted using the cohort average as the origin to show how each miR in each condition deviated from the cohort average.

Table 10 - Demographic and clinical data from patients in miR expression analysis. Percentages are in parentheses CAD+IFTA NA eGFR<60 IFTA Patients 31 46 57 20 52.94 + 53.19 + 50.73 + Age 47.00 + 12.43 12.77 12.56 12.02 Sex

Male 11 (35.5) 27 (58.7) 35 (61.4) 13 (65.0) Female 20 (64.5) 19 (41.3) 22 (38.6) 7 (35.0) Race

Black 23 (74.2) 37 (80.4) 35 (61.4) 11 (55.0) Caucasian 5 (16.1) 8 (17.4) 16 (28.1) 6 (30.0) Hispanic 2 (6.5) 0 4 (7.0) 3 (15.0) Asian 1 (3.2) 1 (2.2) 2 (3.5) 0 Donor Quality

SCD 19 (61.3) 34 (73.9) 27 (47.4) 11 (55.0) ECD 8 (25.8) 1 (2.2) 15 (26.3) 2 (10.0) DCD 3 (9.7) 8 (17.4)) 12 (21.0) 3 (15.0) Other/Unknown 1 (3.2) 3 (6.5) 3 (5.3) 4 (20.0) DGF?

Yes 13 (41.9) 20 (43.5) 33 (57.9) 11 (55.0) No 18 (58.1) 22 (47.8) 23 (40.4) 9 (45.0) Unknown 0 4 (8.7) 1 (1.7) 0 IFTA Grade 18/4/4/5 8/12/0/0 n/a n/a 0/1/2/3 (58.1/12.9/12.9/16.1) (40.0/60.0/0/0) Avg. Serum 2.33 + 1.01 1.03 + 0.21 1.76 + .051 1.10 + 0.24 Creatinine (mg/dL) Avg. Estimated 82.22 + 74.03 + 35.26 + 13.65 43.96 + 9.53 GFR 19.69 12.28

The expression of the selected CAD-related miR targets: TMEM14A (for miR-142-3p),

CD44 & BHLHE41 (for miR-204), and DUSP6 (for miR-211), was assessed in the urine samples of 57 DDK patients via RT-qPCR. Of the 4 genes tested, CD44 showed a significantly different expression (one-way ANOVA, df=3, p=0.007) in patients with

CAD-IFTA compared to normal allograft patients (Fig. 12). The CD44 expression affirms the findings of Kers et al. who showed that CD44 expression is a marker of IFTA and decreased graft survival (109-113). Our findings indicate that CD44 expression is

58 highest in the combination renal dysfunction, but also elevated in the presence of IFTA pathology alone. This pattern is particularly the case in simple renal dysfunction (eGFR

<60 ml/min/1.73m2), whose CD44 expression is not significantly different from patients with normal allografts (F-test, p=0.26), while patients with only IFTA pathology have some increased CD44 expression, but not to the degree of patients with CAD-IFTA. The other gene target-miR pairings showed a trend of anti-correlation characteristic of canonical miR function, but their expression differences did not meet statistical significance.

CD44 Expression in Patient Urine Samples 0.3 0.25 * 0.2

0.15 * 0.1 0.05 0 -0.05

Adjusted Avg. Rel. Exp. Rel. Avg. Adjusted -0.1 -0.15 -0.2 -0.25 NA, n=13 GFR, n=14 IFTA, n=15 CAD, n=15

Figure 12 – CD44 Expresssion in Urine Samples. 57 of 60 patient urine samples passed quality control and relative expression of CD44, BHLHE41 (not shown), DUSP6 (not shown), and TMEM14A (not shown) were assessed via RT-qPCR. Relative expression values were corrected to a normal distribution and average expression for each patient group was adjusted to the cohort average. * = p-value < 0.05 The increased CD44 gene expression coupled with the decreased miR-204 expression

(Fig. 11) in the urine samples of patients with CAD-IFTA further indicate a role for miR-

204 in the development of CAD-IFTA through targeting of the CD44 molecule. The relationship between the expression of CD44 and miR-204 is negatively correlated in all

59 patient groups While the BHLHE41, DUSP6, and TMEM14A expression differences between patient groups was not as pronounced as seen in CD44, they also showed a trend of negative correlation between their expression and miRNA expression. This trend of anti-correlation between CAD-related miRs and their targets indicates a potential interaction between these miRs and their proposed targets. These findings, especially the CD44 expression, provide another potential molecular metric of chronic renal allograft dysfunction, and establish a basis for further study of the direct relationships between the CAD-related miRs and their predicted targets.

Assessment of the Novel Targets of CAD-Related microRNAs

Novel targets for the CAD-related miRs were derived from an in-silico analysis of miR and mRNA arrays comparing patients with and without CAD-IFTA. The differentially expressed miRs and mRNAs were correlated to identify regulation patterns consistent with miR-mediated changes in gene expression (Tables 3-5). Genes that were in silico predicted and anti-correlated with miR-142, -204, and -211 (Appendices A-C) in the urine samples of subjects with CAD-IFTA were selected for experimental validation in

HEK293 cells. The pre-miR sequence of each CAD-related miR was cloned into the pmR-ZsGreen microRNA expression vector and transfected into HEK293 cells to simulate over-expression of each miR (Figs. 13-14). Over-expression of the CAD related miRs in the HEK293 was assessed using RT-qPCR (Figs. 14-15).

Figure 13 - Schematic of pmR-ZsGreen (A) and DNA gel picture of pre-miR inserts and open pmR- ZsGreen (B) used for cloning. The Inserts for miRs-142-3p (484bp), miR-204 (572bp), & -211 (475bp) were cut from TOPO PCR 2.1 vector using SacI & ApaI restriction enzymes, and cloned into the multiple cloning site of pmR-ZsGreen. Cloning was confirmed by selective digestion following bacterial transformation, growth and expansion in selective media

Tranfection Efficiencies 90

cells 60 4 100ng 50 500ng

40 1000ng 1500ng 30 2000ng Percentage GFP/10 Percentage 20

0 142 T-eff 204 T-eff 211 T-eff EV T-eff

Figure 14 - Transfection efficiencies. Average transfection efficiencies for pmR-ZsGreen plasmid transfection by transfected DNA weight. EV-empty vector.

A Relative Expression miR-142-3p Plasmid 1000 * * * * 100 *

10 miR-142-3p 1 EV Relative Expression Relative 0.1 0 ng 100 ng 500 ng 1000 ng 1500 ng 2000 ng Plasmid Amount

B Relative Expression miR-204 plasmid 10000 * * * * * 1000 100 10 miR-204 1 0.1 EV Relative Expression Relative 0.01 0 ng 100 ng 500 ng 1000 ng 1500 ng 2000 ng Plasmid Amount

C Relative Expression miR-211 Plasmid 1000 * * * * * 100

10 miR-211 1 EV Relative Expression Relative 0.1 0 ng 100 ng 500 ng 1000 ng 1500 ng 2000 ng Plasmid Amount

Figure 15 - MicroRNA overexpression in HEK cells. Relative expression measured by RT-qPCR for miR-142-3p (A), miR-204 (B), and miR-211 (C) all compared to their respective empty vector (EV) controls. Student’s t-test used to assess significance (p < 0.05) between miR-transfected and EV control cell groups. These experiments show the cells’ ability to produce mature microRNA from increasing amounts of pre-miR plasmid, and confirm the function of the cells transcriptional machinery in the presence of exogenous DNA. * = p < 0.05

Cells with miR over-expression were also evaluated for mRNA target expression using

RT-qPCR. The miR-target pairings tested were: miR-142-3p with TMEM14A, miR-204 with CD44 molecule & BHLHE41, and miR-211 with DUSP6. Cells over-expressing miR-142-3p showed a decrease in expression of both TMEM14A mRNA relative to untransfected controls (Fig. 16A). Cells over-expressing miR-204 showed a decreased expression of BHLHE41 and CD44 compared to controls (Fig. 16B-C). Cells over expressing miR-211 did not exhibit the same type of decreased expression of DUSP6 as the other miR-target pairings tested via over-expression. The targets for miR-211 consistently showed increases in mRNA expression compared to controls with increasing levels of miR-211 present in the cell (Fig. 16D). This deviation from the expected expression pattern may indicate an unexpected relationship between miR-211 and its targets, or, more likely, give insight to how miR-211’s actions balance between stabilizing mRNA and triggering mRNA degradation.

A B TMEM14A Expression CD44 Expression 1.2 1.2 1 1 * * * * * * * 0.8 0.8 0.6 0.6 0.4 0.4

Relative Expression Relative 0.2 Expression Relative 0.2 0 0

miR-142-3p Transfection Amount miR-204 Transfection Amount

C D BHLHE41 Expression DUSP 6 Relative 1.2 Expression 1 2.5 * * 0.8 2 * * * * * * * * 0.6 1.5 0.4 1

Relative Expression Relative 0.2 0.5 Relative Expresion Relative 0 0

miR-204 Transfection Amount miR-211 Transfection Amount

Figure 16 - mRNA expression of proposed CAD-related miR targets following miR overexpression in HEK293 cells. Gene expression with increasing amounts of transfected miR-142-3p (A), miR-204 (B- C), and miR-211 (D). Expression presented is relative to untransfected controls. miR-142-3p and miR-204 targets show decreased expression compared to cohort median (A-C), while DUSP6 expression increases with miR-expression. * = p < 0.05

The changes in gene expression when following miR overexpression in HEK293 cells indicate that the miRs-142-3p and -204 are decreasing the expression of their prospective target genes, while miR-211 overexpression drives an increase in its prospective target, DUSP6, compared to controls. While the miR-target relationship between miRs-142-3p and -204 and their respective targets follows the canonical model of miR function, the exact nature of the miR-211-DUPS6 hybridization may result in more stabilization or even translational activation rather than mRNA degradation. The gene silencing may trigger a feedback mechanism to increase transcription of DUSP6.

Increased mRNA production coupled with the miR-211-DUSP6-mRNA duplexes not being degraded could result in greater detection of DUSP6 transcripts once those duplexes are melted during the RT-qPCR reactions.

Luciferase assays were performed in order to establish the specificity of the relationship between the CAD-related miRs and their targets. The 3’-untranslated region (3’UTR) for each of the target genes specified above, along with a sham control sequence, and a panel of established gene targets for the CAD-related miRs were individually cloned into the pmiR-Glo dual luciferase plasmid (Promega). The established targets comprised of: adenylate cyclase 9 (ADCY9), a miR-142-3p target (97); BCL2-like2 (BCL2L2), a miR-

204 target (102;103); and TGFβR2, a miR-211 target (100). HEK cells grown on a 96- well plate were co-transfected for 24 hours with the luciferase plasmid containing a

3’UTR sequence and a miR-mimic plasmid. Each miR-target pairing was done in triplicate along with relevant controls including a sham 3’UTR sequence and empty vector controls. Additionally, 10nM of antimiR microRNA-inhibitor (or control) oligonucleotides were also transfected to block the miR-target binding in the cells. The

66 experimental results show that miR-142-3p blocked luciferase production by binding the

3’UTR of its target, TMEM14A. The miR-142-3p interaction with TMEM14A produced a

13.7% decrease in luciferase production compared to a sham sequence control. The inhibition of luciferase production was reversed with the co-transfected anti-miR in culture (Fig. 17A). Similar results were also obtained for miR-204 with both CD44 and

BHLHE41 (31.4% and 29.9% decreases respectively) (Fig. 17C) as well as miR-211 and DUSP6 (12.1% decrease in expression) (Fig. 17B).

A B miR-142-3p miR-211 Luciferase Luciferase Expression expression 1.4 1.8 1.2 1.6 1.4 1 * * 1.2 0.8 1 * * 0.6 0.8 0.4 0.6 0.4 Normalized L/R Normalized Normalized L/R Normalized 0.2 0.2 0 0

C miR-204 Luciferase Expression 1.4 1.2 1 0.8 * * * 0.6 0.4 Normalized L/R Normalized 0.2 0

Figure 17 - Luciferase expression of CAD-related miR target genes. The expression of proposed and established miR targets decreases by approximately, 14% in miR-142-3p (A), 12% in miR-211 (B), and 30% in miR-204 (C). The decreased expression was reversed with co-transfection of the corresponding miR inhibitor which is a oligonucleotide sequence that is a perfect complement to the mature miR sequence, * = p-value < 0.05.

The luciferase assay results confirm our previous results of over-expressing the CAD- related miRs in HEK293 cell culture; providing evidence of a direct relationship between the CAD-related miRs and their proposed targets. The differential expression of the

CAD-related miRs appears to trigger the gene changes observed in the mRNA signature of CAD-IFTA. Factors that trigger the up-regulation of miR-142-3p and repression of miRs-204 &-211 indicative of CAD-IFTA may be related to the chronic inflammatory state of the kidney observed in both delayed graft function and interstitial fibrosis and tubular atrophy (2;5;7;24;64;120). Additionally, TNF-α blockade has been linked to decreased miR-142-3p expression in cases of psoriasis (121), and miRs-204 &

-211 are both associated with the modulations of TGF-β mediated inflammatory responses that lead to fibrosis (100;102). Thus testing the effects of inflammatory stimuli on miR expression may provide some insight into target interaction and miR expression.

To test the effect of inflammatory stimuli on the expression of the CAD-related miRs and their targets, HEK293 cells were transfected with 500ng of pre-miR sequence plasmid from one of the CAD-related miRs and after 24 hours the cells were treated with media containing low (5ng/ml) and high (25ng/ml) concentrations of tumor necrosis factor- alpha (TNF-α), along with untreated vehicle (Veh) controls. Following 8 and 24-hour incubations, the cells were collected, their RNA was isolated and assessed for miR and mRNA expression using RT-qPCR. Inflammatory stimuli increased over-expression in cells that were seeded with pre-miR sequence plasmids prior to the inflammatory stimulus (Figs. 18A, 19A, 20A). The effects of inflammatory stimuli on target gene expression are not neatly coupled to the observed miR expression.

Over expression of miR-142 that was augmented in the inflammatory state (Fig. 18A), a concurrent decrease in expression of the established miR-142-3p target, ADCY9, was observed after both 8 & 24-hour treatments with TNF-α (Fig. 18B). After the 8-hour

TNF-α treatment, ADCY9 expression decreased by 22.4% and 23.4% for the low and high dose treatments respectively. The 24-hour ADCY9 expression increased at every

TNF-α dosage compared to the 8-hour time point, but the low dose showed a 13.8% decrease compared to the untreated control, while the high dose expression decreased by 21.7 percent. TMEM14A expression is consistently decreased in the presence of miR-142-3p over-expression; however, with increasing time and concentration of TNF-α treatments the TMEM14A expression increases (Fig. 18B). TMEM14A expression remains near 50% of median expression during the 8-hour time point at untreated, low dose, and high dose TNF-α treatments. After the 24-hour TNF-α treatment, the

TMEM14A expression pattern changes, with increases of 14.1% and 30.7% from untreated controls at the low and high doses respectively, but TMEM14A expression never rises beyond median expression levels.

A miR-142 Relative Expression 10000 * * * * * * 1000

100

miR-142 10 EV

Relative Expression Relative 1

0.1 Veh 5ng TNF 25ng TNF Veh 5ng TNF 25ng TNF 8-hour 24-hour

B ADCY9 & TMEM14A Expression 1.2

0.8

0.6 ADCY9 0.4 TMEM Relative Expression Relative 0.2

0 142-Veh 142-5ng TNF 142-25ng 142-Veh 142-5ng TNF 142-25ng TNF TNF 8-hour 24-hour

Figure 18 - miR-142-3p and target expression following TNF-α treatment. miR-142-3p expression (A) and ADCY9 & TMEM14A expression (B). The miR expression increased with TNF-α treatments at both 8hr and 24hr timepoints (A). ADCY9 expression falls with rising inflammatory stimuli during both time points. TMEM14A expression remains stably decreased after 8-hour treatment, and increases, but remains down-regulated post 24-hour treatment. * = p < 0.05

Over-expression of miR-204 increases with TNF-α concentration and at all treatment concentrations between the 8-hour and 24-hour timepoints.(Fig. 19A). The inflammatory stimulus also decreased expression of the established miR-204 target BCL2L2, in addition to the proposed targets, BHLHE41 & CD44 molecule (Fig. 19B). The established target, BCL2L2, had a 7.1% decrease in expression from untreated controls following an 8-hour treatment with 5ng/ml TNF-α. BCL2L2 expression decreased 25.8% from untreated controls following an 8-hour treatment with 25ng/ml TNF-α. The decreases in BCL2L2 expression at the 24-hour timepoint were 42.6% and 66.7% for the 5ng/ml and 25ng/ml TNF-α treatments respectively. BHLHE41 and CD44 expression show a similar patterns of expression after 8-hour TNF-α treatments. Both

BHLHE41 and CD44 expressions increase from control expression after the 5ng/ml treatment and decrease 31.2% and 50.5% respectively after 25ng/ml treatment (Fig.

19B). BHLHE41 expression decreases after the 24-hour TNF-α treatments by 75.7% and 61.9% for the 5ng/ml and 25ng/ml treatments respectively. CD44 expression decreases 20.6% after the 24-hour 5ng/ml TNF-α treatment, and increases after the

25ng/ml 24-hour treatment, reversing the pattern seen in the 8-hour time point (Fig.

19B).

A miR-204 Relative Expression 10000 * * * * * * 1000

100

miR-204 10 EV

Relative Expression Relative 1

0.1 Veh 5ng TNF 25ng TNF Veh 5ng TNF 25ng TNF 8-hour 24-hour

B miR-204 Target Expression 1.8 1.6 1.4 1.2 1 BCL2L2 0.8 BHLHE41 0.6

Relative Expression Relative CD44 0.4 0.2 0 204-Veh 204-5ng TNF 204-25ng 204-Veh 204-5ng TNF 204-25ng TNF TNF 8-hour 24-hour

Figure 19 - miR-204 and target expression following TNF-α treatment. miR-204 expression is increased in response to TNF-α treatments at both 8 & 24 hours (A). Decreases in expression of BCL2L2, an established miR-204 target, and the proposed targets, BHLHE41 and CD44, are seen in response to TNF-α treatments, but exhibit different patterns. BCL2L2 falls with increasing TNF-α dosing, while BHLHE41 and CD44 expressions spike at the 8-hr low dose (5ng/ml) TNF-α treatment and decreases at the 24 hours. High dose (25ng/ml) TNF-α treatment decrease BHLHE41 and CD44 expression short term (8hrs) and the expression begins to rebound at 24hrs (B). * = p < 0.05

A miR-211 Relative Expression 100

* * * * * 10

miR-211 1 EV Relative Expression Relative

0.1 Veh 5ng TNF 25ng TNF Veh 25ng TNF 8-hour 24-hour

B DUSP6 Expression 1.4

1.2 * 1 *

0.8

0.6

Relative Expression Relative 0.4

0.2

0 211-08h-Veh 211-08h-5ng 211-08h-25ng 211-24h-Veh 211-24h-25ng 8-hour 24-hour

Figure 20 - miR-211 and target expression following TNF-α treatment. MicroRNA-211 overexpression increases in response to TNF-α treatments and between the 8-hour and 24-hour time points in the untreated controls (A). Expression of the miR-211 target, DUSP6 decreased as TNF-α dosing and incubation time increased. After an 8-hour treatment with low and high dose TNF-α, DUSP6 expression decreased by 15.3% and 20.8% from untreated control levels respectively (B). After a 24-hour incubation, there was a 17% decrease in DUSP6 expression compared to expression levels following the 8-hour incubation. High dose TNF-α treatment resulted in a 12.8% decrease in DUSP6 from untreated controls in the 24-hour TNF-α incubation. These results are in keeping with the established relationship between miR-211 and DUSP6. (Fig. 17D). * = p < 0.05

Using inflammation to precipitate alterations in miR expression resulted in increased expression of miR-142-3p, miR-204, and miR-211 compared to plasmid transfection alone. The DNA transfection weight (500ng) was chosen based on data gathered on transfection efficiency (Fig. 14), miR expression (Fig. 15) and target expression (Fig.

16). The results are in agreement with the relationships between the CAD-related miRs and their targets established by previous data (Fig. 17). The consistency in pattern, and degree, of the changes in miR expression with exposure, albeit short-term, to inflammatory stimuli, indicate that inflammation alone may not be enough to trigger the up-regulation of miR-142-3p and the concurrent down-regulation of miR-204 and miR-

211 that is consistent with renal CAD-IFTA.

Conclusions

The main goal of this work was to find out if metrics that are otherwise seen as markers of disease have some level of functional importance to the development of the diseases which they measure. We were able to identify parallels between organ quality, organ handling and DGF in our patient cohort and then used tissue array data to identify the gene dysregulation and molecular processes that contribute to ischemia reperfusion injury and DGF development. There may be an opportunity to create a panel of gene markers for DGF, and more interestingly the increased inflammation observed in IRI and DGF are linked to molecular processes that can be directly regulated or be regulated by miRs including one of our miRs of interest, miR-142-3p (40;86;88;89).

These findings give some credence to the thought that IRI, DGF, and even CAD-IFTA

75 begin developing in the initial moments post-reperfusion, and that the development of all three may be directed by changes in miR expression and activity.

Because obtaining urine samples from patients with poorly or non-functioning kidneys post-operatively is not feasible, follow-up experimentation from otherwise easily obtainable urine samples would have to necessarily rely on invasive tissue biopsies or less invasive but less informative blood testing to assess miR activity in DGF development. As long as the goal remains to assess function based on information from non-invasive sources, there will be limitations to the effectiveness of studying IRI and

DGF in the clinical setting without access to patients’ biopsy tissue.

The data on miR expression from patient urine samples that we produced showed that there was a difference in miR expression of miR-142-3p, miR-204 and miR-211 between the CAD-IFTA and normal allograft patient groups, which agrees with the previous work on the expression of these miRs in CAD-IFTA (38). Our attempt to test the consistency of the CAD-IFTA miR-expression pattern, led to the inclusion of patients that were not included in our previous work. Specifically, we delineated between those patients without pathology, but with poor function (eGFR <60 ml/min/1.73m2); and those patients with milder IFTA pathology, but no clinical manifestations of dysfunction (IFTA only, Table 10). Including these two extra groups in our analysis, with their classifications being based on a single clinical metric, resulted in a more complex analysis of our patient cohort. The added layer of complexity, and our reliance on gene expression metrics that minimize bias and measure each patient individually, dampened the robust fold change differences observed in our previous work, and in the case of miR-204 and miR-211 expression, practically eliminated it altogether. Post-hoc testing

76 that compared only patients with CAD-IFTA and those with normal allografts was necessary to establish the previously observed patterns of miR expression associated with CAD-IFTA.

The gene expression of the proposed miR targets in the urine samples showed a significant increase in CD44 expression in between the patient groups (Fig. 12). These results coupled with its anti-correlation tomiR-204 expression, the known roles of CD44 as key regulator of TGF-β-mediated fibrogenesis & apoptosis (111;112;122), and

CD44’s role as a regulator of the fibrogenic miR-21 (123-125) (Fig. 21), come together to paint a picture where CD44 is one of the earliest activators of the cell processes seen in IRI, DGF, and CAD-IFTA. Furthermore, the loss of miR-204 function in renal transplant injury (and perhaps renal injury generally) may be a key regulator in development of renal transplant injury and a potential path of exogenous treatment for renal transplant injury. These results, and the expressions of the other 3 miR targets, should be expanded to a well defined groups of patients with CAD-IFTA and normal functioning allografts to validate anti-correlation between the urine sample miR and gene target expression in patients with and without CAD-IFTA.

Figure 21 - CD44 control of miR-21 expression. From Bourguignon et al. (2012), schematic of proposed model of miR-21 expression secondary to CD44 mediated activation of the stem cell marker, Nanog, and STAT3 in head and neck squamous cell carcinoma. The over-expression of miRs-142-3p, -204, & -211 in HEK293 cells illustrated the ability of the cells machinery to process mature, detectable miRs from a pre-miR sequence.

The titratable effect of miR expression with increasing amounts of seeded DNA (Fig.

15), implied that target gene expression would also have a similar effect though anti- correlated to miR expression. The targets for miR-142-3p, TMEM14A, and miR-204,

CD44 and BHLHE41, all showed decreased expression compared to the cohort median

(Fig.16 A-C). TMEM14A expression decreased with increasing miR-142-3p expression

(Fig. 16A), while both the miR-204 targets maintained a less defined pattern of decreased expression (Fig. 16B-C). The observed relationship between miR-211 and its

78 proposed target, DUSP6, was a direct correlation, with DUSP6 expression increasing with miR-211 concentration, rather than the expected anti-correlation (Fig. 16D). It is possible that miR-211 binding DUSP6 is not degrading the mRNA, simulteously triggering a positive feedback response and leaving the gene product viable for detection upon post-transfection RT-qPCR. Another explanation could be that miR-211 is driving DUSP6 production. Some miRs have been shown to trigger transcriptional and translational activation mechanisms (65;75;77;79;80). There is increasing evidence that miRs are also associated with increased translational activity. The difference between miR-induced repression and activation has been linked to miR binding to 5’- untranslated regions of transcripts (82) and more recently to the proliferative state of the cell; where miRs repress translation in actively proliferating cells and increase translation in quiescent cells (75;79;80). The proposed mechanism of miRNA-induced translation activation is not as well elucidated as its inhibitory counterpart, but it may be linked to a decrease in GW182 production seen particularly in states of cellular quiescence (80;126). Without assaying the cells for metrics of viability, it is difficult to say whether the state of the cell affected the miR-target relationship only in the case of miR-211 and DUSP6, but it may warrant further study.

The luciferase assays confirmed the relationships between the CAD-related miRs and their proposed targets (Fig. 17). The decrease in luciferase protein gives some insight into the miR-target relationship of all 3 CAD-related miRs, including miR-211, which appeared to increase DUSP6 expression in miR-211 over-expression assays (Fig.

16D). The decreased luciferase production in the miR-211-DUSP6 interaction shows that miR-211 binds the DUSP6 3’-UTR and inhibits protein production. This outcome,

79 coupled with previous data from miR-211 over-expression, shows that miR-211-DUSP6 interaction results in an isolated decrease in DUSP6 expression that may trigger alternate pathways for DUSP6 expression.

In summary, this work aimed to take miRs that were thought to be possible biomarkers of CAD-IFTA and explore the idea that they may not just be biomarkers, but also be mediators of renal transplant dysfunction. By relating the expression of miRs-142-3p, -

204, & -211 to genes that are differentially expressed in CAD-IFTA, we attempted to identify potential mechanisms for the miR effect beyond their currently understood cellular functions. The increase of miR-142-3p and the loss of miR-204 and miR-211 are linked to genes that play roles in promoting apoptosis, inflammation, and fibrosis.

Further exploration of these miRs within these cellular processes is necessary to elucidate a complete functional mechanism, but the results presented here show promise that these miRs play a role in renal transplant injury. The ability to assess the levels of these 3 miRs in the urine presents a clinical picture of a panel of disease mediators and potential pharmaceutical targets that can also be easier to consistently monitor in patients than the current clinical standards of eGFR and tissue biopsies.

Reference List

1. OPTN: Organ Procurement and Transplantation Network. Health Resources and Services Administration 2014.

2. Maluf DG, Mas VR, Archer KJ, Yanek K, Gibney EM, King AL, Cotterell A, Fisher RA, Posner MP. Molecular pathways involved in loss of kidney graft function with tubular atrophy and interstitial fibrosis. Mol.Med. 2008 May;14(5-6):276-85. PMCID:PMC2242778

3. Mas V, Maluf D, Archer K, Yanek K, Mas L, King A, Gibney E, Massey D, Cotterell A, Fisher R, et al. Establishing the molecular pathways involved in chronic allograft nephropathy for testing new noninvasive diagnostic markers. Transplantation 2007 Feb 27;83(4):448-57

4. Mas VR, Archer KJ, Yanek K, Dumur CI, Capparuccini MI, Mangino MJ, King A, Gibney EM, Fisher R, Posner M, et al. Gene expression patterns in deceased donor kidneys developing delayed graft function after kidney transplantation. Transplantation 2008 Feb 27;85(4):626-35

5. Mas VR, Archer KJ, Scian M, Maluf DG. Molecular pathways involved in loss of graft function in kidney transplant recipients. Expert.Rev.Mol.Diagn. 2010 Apr;10(3):269-84

6. Matas AJ, Smith JM, Skeans MA, Thompson B, Gustafson SK, Schnitzler MA, Stewart DE, Cherikh WS, Wainright JL, Snyder JJ, et al. OPTN/SRTR 2012 Annual Data Report: kidney. Am.J.Transplant. 2014 Jan;14 Suppl 1:11-44

7. Wilflingseder J, Kainz A, Muhlberger I, Perco P, Langer R, Kristo I, Mayer B, Oberbauer R. Impaired metabolism in donor kidney grafts after steroid pretreatment. Transpl.Int. 2010 Aug;23(8):796-804

8. Costanzo L. Renal Physiology. In Physiology. 5th ed. Saunders; 2013. p. 239- 302.

9. Matsushita K, Mahmoodi BK, Woodward M, Emberson JR, Jafar TH, Jee SH, Polkinghorne KR, Shankar A, Smith DH, Tonelli M, et al. Comparison of risk prediction using the CKD-EPI equation and the MDRD study equation for estimated glomerular filtration rate. JAMA 2012 May 9;307(18):1941-51. PMCID:PMC3837430

10. Matsushita K, Tonelli M, Lloyd A, Levey AS, Coresh J, Hemmelgarn BR. Clinical risk implications of the CKD Epidemiology Collaboration (CKD-EPI) equation compared with the Modification of Diet in Renal Disease (MDRD) Study equation for estimated GFR. Am.J.Kidney Dis. 2012 Aug;60(2):241-9

11. Levey AS, Coresh J, Greene T, Marsh J, Stevens LA, Kusek JW, Van LF. Expressing the Modification of Diet in Renal Disease Study equation for estimating glomerular filtration rate with standardized serum creatinine values. Clin.Chem. 2007 Apr;53(4):766-72

12. Levey AS, Inker LA, Coresh J. GFR estimation: from physiology to public health. Am.J.Kidney Dis. 2014 May;63(5):820-34. PMCID:PMC4001724

13. U.S. Renal Data System, USRDS 2013 Annual Data Report: Atlas of Chronic Kidney Disease and End-Stage Renal Disease in the United States. National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD, 2013.; 2013.

14. Kidney Disease Statistics for the United States 2012 Nov 15.

15. Schnitzler MA, Lentine KL, Gheorghian A, Axelrod D, Trivedi D, L'Italien G. Renal function following living, standard criteria deceased and expanded criteria deceased donor kidney transplantation: impact on graft failure and death. Transpl.Int. 2012 Feb;25(2):179-91

16. Ciancio G, Gaynor JJ, Sageshima J, Chen L, Roth D, Kupin W, Guerra G, Tueros L, Zarak A, Hanson L, et al. Favorable outcomes with machine perfusion and longer pump times in kidney transplantation: a single-center, observational study. Transplantation 2010 Oct 27;90(8):882-90

17. Hartono C, Suthanthiran M. Transplantation: Pump it up: conserving a precious resource? Nat.Rev.Nephrol. 2009 Aug;5(8):433-4

18. Moers C, Pirenne J, Paul A, Ploeg RJ. Machine perfusion or cold storage in deceased-donor kidney transplantation. N.Engl.J.Med. 2012 Feb 23;366(8):770-1

19. Moers C, Varnav OC, van HE, Jochmans I, Kirste GR, Rahmel A, Leuvenink HG, Squifflet JP, Paul A, Pirenne J, et al. The value of machine perfusion perfusate biomarkers for predicting kidney transplant outcome. Transplantation 2010 Nov 15;90(9):966-73

20. Moers C, Smits JM, Maathuis MH, Treckmann J, van GF, Napieralski BP, van Kasterop-Kutz M, van der Heide JJ, Squifflet JP, van HE, et al. Machine perfusion or cold storage in deceased-donor kidney transplantation. N.Engl.J.Med. 2009 Jan 1;360(1):7-19

21. Muhlberger I, Perco P, Fechete R, Mayer B, Oberbauer R. Biomarkers in renal transplantation ischemia reperfusion injury. Transplantation 2009 Aug 15;88(3 Suppl):S14-S19

22. Jochmans I, Pirenne J. Graft quality assessment in kidney transplantation: not an exact science yet! Curr.Opin.Organ Transplant. 2011 Apr;16(2):174-9

23. Park WD, Griffin MD, Cornell LD, Cosio FG, Stegall MD. Fibrosis with inflammation at one year predicts transplant functional decline. J.Am.Soc.Nephrol. 2010 Nov;21(11):1987-97. PMCID:PMC3014013

24. Mba MU, Maluf DG, Dumur CI, Scian MJ, Posner M, King AL, Gehr TW, Sharma A, Cotterell A, Ren Q, et al. Molecular Biomarkers of Human Kidney Transplantation Ischemia Reperfusion Injury. In American Transplant Congress, National Meeting; 2011.

25. Mueller TF, Solez K, Mas V. Assessment of kidney organ quality and prediction of outcome at time of transplantation. Semin.Immunopathol. 2011 Mar;33(2):185- 99

26. Moreira P, Sa H, Figueiredo A, Mota A. Delayed renal graft function: risk factors and impact on the outcome of transplantation. Transplant.Proc. 2011 Jan;43(1):100-5

27. Peeters P, Terryn W, Vanholder R, Lameire N. Delayed graft function in renal transplantation. Curr.Opin.Crit Care 2004 Dec;10(6):489-98

28. Roels L, Peeters J, Vanrenterghem Y. Delayed graft function as principal correlate of kidney allograft outcome in a single-center multivariate analysis. Transplant.Proc. 1996 Feb;28(1):272-4

29. Israni AK, Li N, Cizman BB, Snyder J, Abrams J, Joffe M, Rebbeck T, Feldman HI. Association of donor inflammation- and apoptosis-related genotypes and delayed allograft function after kidney transplantation. Am.J.Kidney Dis. 2008 Aug;52(2):331-9. PMCID:PMC2562522

30. Mas VR, Mas LA, Archer KJ, Yanek K, King AL, Gibney EM, Cotterell A, Fisher RA, Posner M, Maluf DG. Evaluation of gene panel mRNAs in urine samples of kidney transplant recipients as a non-invasive tool of graft function. Mol.Med. 2007 May;13(5-6):315-24. PMCID:PMC1906687

31. Haas M, Sis B, Racusen LC, Solez K, Glotz D, Colvin RB, Castro MC, David DS, David-Neto E, Bagnasco SM, et al. Banff 2013 meeting report: inclusion of c4d- negative antibody-mediated rejection and antibody-associated arterial lesions. Am.J.Transplant. 2014 Feb;14(2):272-83

32. Mengel M, Sis B, Haas M, Colvin RB, Halloran PF, Racusen LC, Solez K, Cendales L, Demetris AJ, Drachenberg CB, et al. Banff 2011 Meeting report: new concepts in antibody-mediated rejection. Am.J.Transplant. 2012 Mar;12(3):563-70. PMCID:PMC3728651

33. Sis B, Mengel M, Haas M, Colvin RB, Halloran PF, Racusen LC, Solez K, Baldwin WM, III, Bracamonte ER, Broecker V, et al. Banff '09 meeting report: antibody mediated graft deterioration and implementation of Banff working groups. Am.J.Transplant. 2010 Mar;10(3):464-71

34. Mas VR, Mueller TF, Archer KJ, Maluf DG. Identifying biomarkers as diagnostic tools in kidney transplantation. Expert.Rev.Mol.Diagn. 2011 Mar;11(2):183-96. PMCID:PMC3116652

35. Mas VR, Scian MJ, Archer KJ, Suh JL, David KG, Ren Q, Gehr TW, King AL, Posner MP, Mueller TF, et al. Pretransplant transcriptome profiles identify among kidneys with delayed graft function those with poorer quality and outcome. Mol.Med. 2011;17(11-12):1311-22. PMCID:PMC3321827

36. Mas VR, Dumur CI, Scian MJ, Gehrau RC, Maluf DG. MicroRNAs as biomarkers in solid organ transplantation. Am.J.Transplant. 2013 Jan;13(1):11-9. PMCID:PMC3927320

37. Park W, Griffin M, Grande JP, Cosio F, Stegall MD. Molecular evidence of injury and inflammation in normal and fibrotic renal allografts one year posttransplant. Transplantation 2007 Jun 15;83(11):1466-76

38. Scian MJ, Maluf DG, David KG, Archer KJ, Suh JL, Wolen AR, Mba MU, Massey HD, King AL, Gehr T, et al. MicroRNA profiles in allograft tissues and paired urines associate with chronic allograft dysfunction with IF/TA. Am.J.Transplant. 2011 Oct;11(10):2110-22. PMCID:PMC3184368

39. Scian MJ, Maluf DG, Archer KJ, Turner SD, Suh JL, David KG, King AL, Posner MP, Brayman KL, Mas VR. Identification of biomarkers to assess organ quality and predict posttransplantation outcomes. Transplantation 2012 Oct 27;94(8):851-8. PMCID:PMC3927314

40. Scian MJ, Maluf DG, Mas VR. MiRNAs in kidney transplantation: potential role as new biomarkers. Expert.Rev.Mol.Diagn. 2013 Jan;13(1):93-104

41. Chau BN, Xin C, Hartner J, Ren S, Castano AP, Linn G, Li J, Tran PT, Kaimal V, Huang X, et al. MicroRNA-21 promotes fibrosis of the kidney by silencing metabolic pathways. Sci.Transl.Med. 2012 Feb 15;4(121):121ra18. PMCID:PMC3672221

42. Schold JD, Kaplan B. The elephant in the room: failings of current clinical endpoints in kidney transplantation. Am.J.Transplant. 2010 May;10(5):1163-6

43. Scian MJ, Maluf DG, Archer KJ, Suh JL, Massey D, Fassnacht RC, Whitehill B, Sharma A, King A, Gehr T, et al. Gene expression changes are associated with loss of kidney graft function and interstitial fibrosis and tubular atrophy: diagnosis versus prediction. Transplantation 2011 Mar 27;91(6):657-65

44. Peeters J, Roels L, Vanrenterghem Y. Chronic renal allograft failure: clinical overview. The Leuven Collaborative Group for Transplantation. Kidney Int.Suppl 1995 Dec;52:S97-101

45. Rodder S, Scherer A, Raulf F, Berthier CC, Hertig A, Couzi L, Durrbach A, Rondeau E, Marti HP. Renal allografts with IF/TA display distinct expression profiles of metzincins and related genes. Am.J.Transplant. 2009 Mar;9(3):517-26

46. Muller GA, Zeisberg M, Strutz F. The importance of tubulointerstitial damage in progressive renal disease. Nephrol.Dial.Transplant. 2000;15 Suppl 6:76-7

47. Strutz F, Muller GA. Renal fibrosis and the origin of the renal fibroblast. Nephrol.Dial.Transplant. 2006 Dec;21(12):3368-70

48. Wang B, Ricardo S. Role of microRNA machinery in kidney fibrosis. Clin.Exp.Pharmacol.Physiol 2014 May 6;

49. Zeisberg M, Strutz F, Muller GA. Role of fibroblast activation in inducing interstitial fibrosis. J.Nephrol. 2000 Nov;13 Suppl 3:S111-S120

50. Zeisberg M, Bonner G, Maeshima Y, Colorado P, Muller GA, Strutz F, Kalluri R. Renal fibrosis: collagen composition and assembly regulates epithelial- mesenchymal transdifferentiation. Am.J.Pathol. 2001 Oct;159(4):1313-21. PMCID:PMC1850511

51. Rodder S, Scherer A, Korner M, Marti HP. A subset of metzincins and related genes constitutes a marker of human solid organ fibrosis. Virchows Arch. 2011 Apr;458(4):487-96

52. Maluf DG, Dumur CI, Suh JL, Scian MJ, King AL, Cathro H, Lee JK, Gehrau RC, Brayman KL, Gallon L, et al. The urine microRNA profile may help monitor post- transplant renal graft function. Kidney Int. 2014 Feb;85(2):439-49. PMCID:PMC3946645

53. Poggio ED, Batty DS, Flechner SM. Evaluation of renal function in transplantation. Transplantation 2007 Jul 27;84(2):131-6

54. Furness PN. Predicting allograft survival: abundant data, but insufficient knowledge? Transplantation 2007 Mar 27;83(6):681

55. Gillespie A, Lee IJ. Biomarkers in renal transplantation. Biomark.Med. 2008 Dec;2(6):603-12

56. Maluf DG, Dumur CI, Suh JL, Lee JK, Cathro HP, King AL, Gallon L, Brayman KL, Mas VR. Evaluation of molecular profiles in calcineurin inhibitor toxicity post- kidney transplant: input to chronic allograft dysfunction. Am.J.Transplant. 2014 May;14(5):1152-63

57. Maillard N, Mehdi M, Thibaudin L, Berthoux F, Alamartine E, Mariat C. Creatinine-based GFR predicting equations in renal transplantation: reassessing the tubular secretion effect. Nephrol.Dial.Transplant. 2010 Sep;25(9):3076-82

58. Mariat C, Alamartine E, Barthelemy JC, De Filippis JP, Thibaudin D, Berthoux P, Laurent B, Thibaudin L, Berthoux F. Assessing renal graft function in clinical trials: can tests predicting glomerular filtration rate substitute for a reference method? Kidney Int. 2004 Jan;65(1):289-97

59. Furness PN, Philpott CM, Chorbadjian MT, Nicholson ML, Bosmans JL, Corthouts BL, Bogers JJ, Schwarz A, Gwinner W, Haller H, et al. Protocol biopsy of the stable renal transplant: a multicenter study of methods and complication rates. Transplantation 2003 Sep 27;76(6):969-73

60. Rush DN, Jeffery JR, Gough J. Sequential protocol biopsies in renal transplant patients: repeated inflammation is associated with impaired graft function at 1 year. Transplant.Proc. 1995 Feb;27(1):1017-8

61. Rush DN, Nickerson P, Jeffery JR, McKenna RM, Grimm PC, Gough J. Protocol biopsies in renal transplantation: research tool or clinically useful? Curr.Opin.Nephrol.Hypertens. 1998 Nov;7(6):691-4

62. Rush DN. Can urinary monokine induced by interferon-gamma accurately predict acute renal allograft rejection? Nat.Clin.Pract.Nephrol. 2005 Nov;1(1):10-1

63. Solez K, Colvin RB, Racusen LC, Haas M, Sis B, Mengel M, Halloran PF, Baldwin W, Banfi G, Collins AB, et al. Banff 07 classification of renal allograft pathology: updates and future directions. Am.J.Transplant. 2008 Apr;8(4):753-60

64. Anglicheau D, Suthanthiran M. Noninvasive prediction of organ graft rejection and outcome using gene expression patterns. Transplantation 2008 Jul 27;86(2):192-9. PMCID:PMC3595195

65. Anglicheau D, Sharma VK, Ding R, Hummel A, Snopkowski C, Dadhania D, Seshan SV, Suthanthiran M. MicroRNA expression profiles predictive of human renal allograft status. Proc.Natl.Acad.Sci.U.S.A 2009 Mar 31;106(13):5330-5. PMCID:PMC2663998

66. Ben-Dov IZ, Tan YC, Morozov P, Wilson PD, Rennert H, Blumenfeld JD, Tuschl T. Urine microRNA as potential biomarkers of autosomal dominant polycystic kidney disease progression: description of miRNA profiles at baseline. PLoS.One. 2014;9(1):e86856. PMCID:PMC3906110

67. Sir ER, Li JY, Yong TY, Gleadle JM. Dipping your feet in the water: podocytes in urine. Expert.Rev.Mol.Diagn. 2014 May;14(4):423-37

68. Schwanhausser B, Busse D, Li N, Dittmar G, Schuchhardt J, Wolf J, Chen W, Selbach M. Global quantification of mammalian gene expression control. Nature 2011 May 19;473(7347):337-42

69. Hueso M, Navarro E, Moreso F, O'Valle F, Perez-Riba M, Del Moral RG, Grinyo JM, Seron D. Intragraft expression of the IL-10 gene is up-regulated in renal

protocol biopsies with early interstitial fibrosis, tubular atrophy, and subclinical rejection. Am.J.Pathol. 2010 Apr;176(4):1696-704. PMCID:PMC2843461

70. Liu Y. Epithelial to mesenchymal transition in renal fibrogenesis: pathologic significance, molecular mechanism, and therapeutic intervention. J.Am.Soc.Nephrol. 2004 Jan;15(1):1-12

71. Rastaldi MP, Ferrario F, Giardino L, Dell'Antonio G, Grillo C, Grillo P, Strutz F, Muller GA, Colasanti G, D'Amico G. Epithelial-mesenchymal transition of tubular epithelial cells in human renal biopsies. Kidney Int. 2002 Jul;62(1):137-46

72. Strutz F, Okada H, Lo CW, Danoff T, Carone RL, Tomaszewski JE, Neilson EG. Identification and characterization of a fibroblast marker: FSP1. J.Cell Biol. 1995 Jul;130(2):393-405. PMCID:PMC2199940

73. Strutz F. Pathogenesis of tubulointerstitial fibrosis in chronic allograft dysfunction. Clin.Transplant. 2009 Dec;23 Suppl 21:26-32

74. Lee RC, Feinbaum RL, Ambros V. The C. elegans heterochronic gene lin-4 encodes small RNAs with antisense complementarity to lin-14. Cell 1993 Dec 3;75(5):843-54

75. Fabian MR, Sonenberg N, Filipowicz W. Regulation of mRNA translation and stability by microRNAs. Annu.Rev.Biochem. 2010;79:351-79

76. Grigoryev YA, Kurian SM, Hart T, Nakorchevsky AA, Chen C, Campbell D, Head SR, Yates JR, III, Salomon DR. MicroRNA regulation of molecular networks mapped by global microRNA, mRNA, and protein expression in activated T lymphocytes. J.Immunol. 2011 Sep 1;187(5):2233-43. PMCID:PMC3159804

77. Kasinath BS, Feliers D, Sataranatarajan K, Ghosh CG, Lee MJ, Mariappan MM. Regulation of mRNA translation in renal physiology and disease. Am.J.Physiol Renal Physiol 2009 Nov;297(5):F1153-F1165. PMCID:PMC2781325

78. Li JY, Yong TY, Michael MZ, Gleadle JM. Review: The role of microRNAs in kidney disease. Nephrology.(Carlton.) 2010 Sep;15(6):599-608

79. Vasudevan S, Tong Y, Steitz JA. Switching from repression to activation: microRNAs can up-regulate translation. Science 2007 Dec 21;318(5858):1931-4

80. Vasudevan S, Tong Y, Steitz JA. Cell-cycle control of microRNA-mediated translation regulation. Cell Cycle 2008 Jun 1;7(11):1545-9. PMCID:PMC2556257

81. Reid G, Kirschner MB, van ZN. Circulating microRNAs: Association with disease and potential use as biomarkers. Crit Rev.Oncol.Hematol. 2011 Nov;80(2):193- 208

82. Orom UA, Nielsen FC, Lund AH. MicroRNA-10a binds the 5'UTR of ribosomal protein mRNAs and enhances their translation. Mol.Cell 2008 May 23;30(4):460- 71

83. Fabian MR, Sonenberg N. The mechanics of miRNA-mediated gene silencing: a look under the hood of miRISC. Nat.Struct.Mol.Biol. 2012 Jun;19(6):586-93

84. Bartel DP. MicroRNAs: target recognition and regulatory functions. Cell 2009 Jan 23;136(2):215-33. PMCID:PMC3794896

85. Kozomara A, Griffiths-Jones S. miRBase: annotating high confidence microRNAs using deep sequencing data. Nucleic Acids Res. 2014 Jan;42(Database issue):D68-D73. PMCID:PMC3965103

86. Shapiro MD, Bagley J, Latz J, Godwin JG, Ge X, Tullius SG, Iacomini J. MicroRNA expression data reveals a signature of kidney damage following ischemia reperfusion injury. PLoS.One. 2011;6(8):e23011. PMCID:PMC3156120

87. Gatti S, Bruno S, Deregibus MC, Sordi A, Cantaluppi V, Tetta C, Camussi G. Microvesicles derived from human adult mesenchymal stem cells protect against ischaemia-reperfusion-induced acute and chronic kidney injury. Nephrol.Dial.Transplant. 2011 May;26(5):1474-83

88. Godwin JG, Ge X, Stephan K, Jurisch A, Tullius SG, Iacomini J. Identification of a microRNA signature of renal ischemia reperfusion injury. Proc.Natl.Acad.Sci.U.S.A 2010 Aug 10;107(32):14339-44. PMCID:PMC2922548

89. Zarjou A, Yang S, Abraham E, Agarwal A, Liu G. Identification of a microRNA signature in renal fibrosis: role of miR-21. Am.J.Physiol Renal Physiol 2011 Oct;301(4):F793-F801. PMCID:PMC3191802

90. Li YF, Jing Y, Hao J, Frankfort NC, Zhou X, Shen B, Liu X, Wang L, Li R. MicroRNA-21 in the pathogenesis of acute kidney injury. Protein Cell 2013 Nov;4(11):813-9

91. Wang JY, Gao YB, Zhang N, Zou DW, Wang P, Zhu ZY, Li JY, Zhou SN, Wang SC, Wang YY, et al. miR-21 overexpression enhances TGF-beta1-induced epithelial-to-mesenchymal transition by target smad7 and aggravates renal damage in diabetic nephropathy. Mol.Cell Endocrinol. 2014 Jul 5;392(1-2):163- 72

92. Macconi D, Tomasoni S, Romagnani P, Trionfini P, Sangalli F, Mazzinghi B, Rizzo P, Lazzeri E, Abbate M, Remuzzi G, et al. MicroRNA-324-3p promotes renal fibrosis and is a target of ACE inhibition. J.Am.Soc.Nephrol. 2012 Sep;23(9):1496-505. PMCID:PMC3431411

93. Ben-Dov IZ, Muthukumar T, Morozov P, Mueller FB, Tuschl T, Suthanthiran M. MicroRNA sequence profiles of human kidney allografts with or without

tubulointerstitial fibrosis. Transplantation 2012 Dec 15;94(11):1086-94. PMCID:PMC3541003

94. Khedmat H, Taheri S. Characteristics and prognosis of lymphoproliferative disorders post-renal transplantation in living versus deceased donor allograft recipients. Saudi.J.Kidney Dis.Transpl. 2013 Sep;24(5):903-9

95. Lee S, Kim J, Shin M, Kim E, Moon J, Jung G, Choi G, Kwon C, Joh J, Lee S, et al. Comparison of outcomes of living and deceased donor kidney grafts surviving longer than 5 years in Korea. Transplant.Proc. 2010 Apr;42(3):775-7

96. Naderi GH, Mehraban D, Kazemeyni SM, Darvishi M, Latif AH. Living or deceased donor kidney transplantation: a comparison of results and survival rates among Iranian patients. Transplant.Proc. 2009 Sep;41(7):2772-4

97. Huang B, Zhao J, Lei Z, Shen S, Li D, Shen GX, Zhang GM, Feng ZH. miR-142- 3p restricts cAMP production in CD4+. EMBO Rep. 2009 Feb;10(2):180-5. PMCID:PMC2637310

98. Lv M, Zhang X, Jia H, Li D, Zhang B, Zhang H, Hong M, Jiang T, Jiang Q, Lu J, et al. An oncogenic role of miR-142-3p in human T-cell acute lymphoblastic leukemia (T-ALL) by targeting glucocorticoid receptor-alpha and cAMP/PKA pathways. Leukemia 2012 Apr;26(4):769-77

99. Woo IS, Jin H, Kang ES, Kim HJ, Lee JH, Chang KC, Park JY, Choi WS, Seo HG. TMEM14A inhibits N-(4-hydroxyphenyl)retinamide-induced apoptosis through the stabilization of mitochondrial membrane potential. Cancer Lett. 2011 Oct 28;309(2):190-8

100. Wang FE, Zhang C, Maminishkis A, Dong L, Zhi C, Li R, Zhao J, Majerciak V, Gaur AB, Chen S, et al. MicroRNA-204/211 alters epithelial physiology. FASEB J. 2010 May;24(5):1552-71. PMCID:PMC3231816

101. Xie T, Liang J, Guo R, Liu N, Noble PW, Jiang D. Comprehensive microRNA analysis in bleomycin-induced pulmonary fibrosis identifies multiple sites of molecular regulation. Physiol Genomics 2011 May 13;43(9):479-87. PMCID:PMC3110895

102. Pollari S, Leivonen SK, Perala M, Fey V, Kakonen SM, Kallioniemi O. Identification of microRNAs inhibiting TGF-beta-induced IL-11 production in bone metastatic breast cancer cells. PLoS.One. 2012;7(5):e37361. PMCID:PMC3357420

103. Li G, Luna C, Qiu J, Epstein DL, Gonzalez P. Role of miR-204 in the regulation of apoptosis, endoplasmic reticulum stress response, and inflammation in human trabecular meshwork cells. Invest Ophthalmol.Vis.Sci. 2011 May;52(6):2999- 3007. PMCID:PMC3109013

104. Imaizumi T, Sato F, Tanaka H, Matsumiya T, Yoshida H, Yashiro-Aizawa T, Tsuruga K, Hayakari R, Kijima H, Satoh K. Basic-helix-loop-helix transcription factor DEC2 constitutes negative feedback loop in IFN-beta-mediated inflammatory responses in human mesangial cells. Immunol.Lett. 2011 Apr 30;136(1):37-43

105. Liu Y, Sato F, Kawamoto T, Fujimoto K, Morohashi S, Akasaka H, Kondo J, Wu Y, Noshiro M, Kato Y, et al. Anti-apoptotic effect of the basic helix-loop-helix (bHLH) transcription factor DEC2 in human breast cancer cells. Genes Cells 2010 Apr 1;15(4):315-25

106. Wu Y, Sato F, Bhawal UK, Kawamoto T, Fujimoto K, Noshiro M, Seino H, Morohashi S, Kato Y, Kijima H. BHLH transcription factor DEC2 regulates pro- apoptotic factor Bim in human oral cancer HSC-3 cells. Biomed.Res. 2012 Apr;33(2):75-82

107. Rouschop KM, Sewnath ME, Claessen N, Roelofs JJ, Hoedemaeker I, van der Neut R, Aten J, Pals ST, Weening JJ, Florquin S. CD44 deficiency increases tubular damage but reduces renal fibrosis in obstructive nephropathy. J.Am.Soc.Nephrol. 2004 Mar;15(3):674-86

108. Rouschop KM, Roelofs JJ, Rowshani AT, Leemans JC, van der Poll T, Ten Berge IJ, Weening JJ, Florquin S. Pre-transplant plasma and cellular levels of CD44 correlate with acute renal allograft rejection. Nephrol.Dial.Transplant. 2005 Oct;20(10):2248-54

109. Rouschop KM, Roelofs JJ, Claessen N, da Costa MP, Zwaginga JJ, Pals ST, Weening JJ, Florquin S. Protection against renal ischemia reperfusion injury by CD44 disruption. J.Am.Soc.Nephrol. 2005 Jul;16(7):2034-43

110. Rouschop KM, Roelofs JJ, Sylva M, Rowshani AT, Ten Berge IJ, Weening JJ, Florquin S. Renal expression of CD44 correlates with acute renal allograft rejection. Kidney Int. 2006 Sep;70(6):1127-34

111. Rouschop KM, Claessen N, Pals ST, Weening JJ, Florquin S. CD44 disruption prevents degeneration of the capillary network in obstructive nephropathy via reduction of TGF-beta1-induced apoptosis. J.Am.Soc.Nephrol. 2006 Mar;17(3):746-53

112. Rampanelli E, Rouschop KM, Claessen N, Teske GJ, Pals ST, Leemans JC, Florquin S. Opposite role of CD44-standard and CD44-variant-3 in tubular injury and development of renal fibrosis during chronic obstructive nephropathy. Kidney Int. 2014 Apr 9;

113. Kers J, Xu-Dubois YC, Rondeau E, Claessen N, Idu MM, Roelofs JJ, Bemelman FJ, Ten Berge IJ, Florquin S. Intragraft tubular vimentin and CD44 expression correlate with long-term renal allograft function and interstitial fibrosis and tubular atrophy. Transplantation 2010 Sep 15;90(5):502-9

114. Boyle GM, Woods SL, Bonazzi VF, Stark MS, Hacker E, Aoude LG, Dutton- Regester K, Cook AL, Sturm RA, Hayward NK. Melanoma cell invasiveness is regulated by miR-211 suppression of the BRN2 transcription factor. Pigment Cell Melanoma Res. 2011 Jun;24(3):525-37

115. Levy C, Khaled M, Iliopoulos D, Janas MM, Schubert S, Pinner S, Chen PH, Li S, Fletcher AL, Yokoyama S, et al. Intronic miR-211 assumes the tumor suppressive function of its host gene in melanoma. Mol.Cell 2010 Dec 10;40(5):841-9. PMCID:PMC3004467

116. Mazar J, DeYoung K, Khaitan D, Meister E, Almodovar A, Goydos J, Ray A, Perera RJ. The regulation of miRNA-211 expression and its role in melanoma cell invasiveness. PLoS.One. 2010;5(11):e13779. PMCID:PMC2967468

117. Wang F, Wang XS, Yang GH, Zhai PF, Xiao Z, Xia LY, Chen LR, Wang Y, Wang XZ, Bi LX, et al. miR-29a and miR-142-3p downregulation and diagnostic implication in human acute myeloid leukemia. Mol.Biol.Rep. 2012 Mar;39(3):2713-22

118. Wong VC, Chen H, Ko JM, Chan KW, Chan YP, Law S, Chua D, Kwong DL, Lung HL, Srivastava G, et al. Tumor suppressor dual-specificity phosphatase 6 (DUSP6) impairs cell invasion and epithelial-mesenchymal transition (EMT)- associated phenotype. Int.J.Cancer 2012 Jan 1;130(1):83-95

119. Ramakers C, Ruijter JM, Deprez RH, Moorman AF. Assumption-free analysis of quantitative real-time polymerase chain reaction (PCR) data. Neurosci.Lett. 2003 Mar 13;339(1):62-6

120. Wang Z, Zhu Q, Li PL, Dhaduk R, Zhang F, Gehr TW, Li N. Silencing of hypoxia- inducible factor-1alpha gene attenuates chronic ischemic renal injury in two- kidney, one-clip rats. Am.J.Physiol Renal Physiol 2014 May 15;306(10):F1236- F1242. PMCID:PMC4024731

121. Pivarcsi A, Meisgen F, Xu N, Stahle M, Sonkoly E. Changes in the level of serum microRNAs in patients with psoriasis after antitumour necrosis factor-alpha therapy. Br.J.Dermatol. 2013 Sep;169(3):563-70

122. Bommaya G, Meran S, Krupa A, Phillips AO, Steadman R. Tumour necrosis factor-stimulated gene (TSG)-6 controls epithelial-mesenchymal transition of proximal tubular epithelial cells. Int.J.Biochem.Cell Biol. 2011 Dec;43(12):1739- 46

123. Chen L, Bourguignon LY. Hyaluronan-CD44 interaction promotes c-Jun signaling and miRNA21 expression leading to Bcl-2 expression and chemoresistance in breast cancer cells. Mol.Cancer 2014;13:52. PMCID:PMC3975292

124. Bourguignon LY, Earle C, Wong G, Spevak CC, Krueger K. Stem cell marker (Nanog) and Stat-3 signaling promote MicroRNA-21 expression and

chemoresistance in hyaluronan/CD44-activated head and neck squamous cell carcinoma cells. Oncogene 2012 Jan 12;31(2):149-60. PMCID:PMC3179812

125. Bourguignon LY, Spevak CC, Wong G, Xia W, Gilad E. Hyaluronan-CD44 interaction with protein kinase C(epsilon) promotes oncogenic signaling by the stem cell marker Nanog and the Production of microRNA-21, leading to down- regulation of the tumor suppressor protein PDCD4, anti-apoptosis, and chemotherapy resistance in breast tumor cells. J.Biol.Chem. 2009 Sep 25;284(39):26533-46. PMCID:PMC2785342

126. Lian S, Jakymiw A, Eystathioy T, Hamel JC, Fritzler MJ, Chan EK. GW bodies, microRNAs and the cell cycle. Cell Cycle 2006 Feb;5(3):242-5

Appendix A – negatively correlated targets of miR-142-3p Pearson p- Empirical MicroRNA Probeset Gene Symbol Coefficient value p-value hsa-miR- 210397_at DEFB1 -0.977 1.15E- 4.68E-05 142-3p 06 hsa-miR- 218477_at TMEM14A -0.966 5.33E- 4.99E-04 142-3p 06 hsa-miR- 206716_at UMOD -0.965 6.33E- 6.25E-04 142-3p 06 hsa-miR- 221542_s_at ERLIN2 -0.964 7.20E- 7.09E-04 142-3p 06 hsa-miR- 204485_s_at TOM1L1 -0.962 8.69E- 8.34E-04 142-3p 06 hsa-miR- 212448_at NEDD4L -0.956 1.53E- 1.47E-03 142-3p 05 hsa-miR- 205501_at PDE10A -0.956 1.49E- 1.39E-03 142-3p 05 hsa-miR- 213227_at PGRMC2 -0.955 1.68E- 1.60E-03 142-3p 05 hsa-miR- 211941_s_at PEBP1 -0.954 1.88E- 1.72E-03 142-3p 05 hsa-miR- 206414_s_at ASAP2 -0.952 2.25E- 1.87E-03 142-3p 05 hsa-miR- 211033_s_at PEX7 -0.952 2.18E- 1.82E-03 142-3p 05 hsa-miR- 202209_at LSM3 -0.951 2.43E- 2.04E-03 142-3p 05 hsa-miR- 202300_at HBXIP -0.946 3.48E- 2.99E-03 142-3p 05 hsa-miR- 220329_s_at RMND1 -0.946 3.43E- 2.71E-03 142-3p 05 hsa-miR- 219954_s_at GBA3 -0.945 3.72E- 3.08E-03 142-3p 05 hsa-miR- 205761_s_at DUS4L -0.945 3.65E- 3.03E-03 142-3p 05 hsa-miR- 220999_s_at CYFIP2 -0.944 4.06E- 3.25E-03 142-3p 05 hsa-miR- 203067_at PDHX -0.942 4.74E- 3.91E-03 142-3p 05 hsa-miR- 210544_s_at ALDH3A2 -0.942 4.62E- 3.72E-03 142-3p 05 hsa-miR- 220161_s_at EPB41L4B -0.942 4.61E- 3.68E-03 142-3p 05 hsa-miR- 200718_s_at SKP1 -0.941 5.06E- 4.36E-03 142-3p 05 hsa-miR- 206054_at KNG1 -0.941 4.86E- 4.13E-03

142-3p 05 hsa-miR- 207981_s_at ESRRG -0.941 4.83E- 4.04E-03 142-3p 05 hsa-miR- 214598_at CLDN8 -0.94 5.28E- 4.57E-03 142-3p 05 hsa-miR- 211569_s_at HADH -0.94 5.14E- 4.49E-03 142-3p 05 hsa-miR- 221556_at CDC14B -0.939 5.62E- 4.97E-03 142-3p 05 hsa-miR- 202825_at SLC25A4 -0.939 5.53E- 4.89E-03 142-3p 05 hsa-miR- 202000_at NDUFA6 -0.938 6.01E- 5.25E-03 142-3p 05 hsa-miR- 201007_at HADHB -0.937 6.50E- 5.85E-03 142-3p 05 hsa-miR- 213133_s_at GCSH -0.937 6.39E- 5.66E-03 142-3p 05 hsa-miR- 220197_at ATP6V0A4 -0.936 6.93E- 6.17E-03 142-3p 05 hsa-miR- 202960_s_at MUT -0.935 7.22E- 6.44E-03 142-3p 05 hsa-miR- 208732_at RAB2A -0.935 7.22E- 6.40E-03 142-3p 05 hsa-miR- 200972_at TSPAN3 -0.935 7.18E- 6.35E-03 142-3p 05 hsa-miR- 213933_at PTGER3 -0.935 7.10E- 6.31E-03 142-3p 05 hsa-miR- 204556_s_at DZIP1 -0.935 7.04E- 6.22E-03 142-3p 05 hsa-miR- 221620_s_at APOO -0.933 8.22E- 7.38E-03 142-3p 05 hsa-miR- 200659_s_at PHB -0.932 8.79E- 7.52E-03 142-3p 05 hsa-miR- 218024_at BRP44L -0.932 8.40E- 7.43E-03 142-3p 05 hsa-miR- 201121_s_at PGRMC1 -0.931 9.34E- 7.90E-03 142-3p 05 hsa-miR- 201619_at PRDX3 -0.931 9.08E- 7.68E-03 142-3p 05 hsa-miR- 204547_at RAB40B -0.931 9.03E- 7.64E-03 142-3p 05 hsa-miR- 212445_s_at NEDD4L -0.93 9.83E- 8.39E-03 142-3p 05 hsa-miR- 218285_s_at BDH2 -0.93 9.81E- 8.29E-03 142-3p 05 hsa-miR- 202850_at ABCD3 -0.93 9.77E- 8.16E-03

142-3p 05 hsa-miR- 221142_s_at PECR -0.93 9.69E- 8.12E-03 142-3p 05 hsa-miR- 213657_s_at Affy_213657_s_at -0.93 9.51E- 8.03E-03 142-3p 05 hsa-miR- 203608_at ALDH5A1 -0.929 1.04E- 9.27E-03 142-3p 04 hsa-miR- 218789_s_at C11orf71 -0.929 1.02E- 8.86E-03 142-3p 04 hsa-miR- 213346_at C13orf27 -0.929 1.01E- 8.82E-03 142-3p 04 hsa-miR- 203970_s_at PEX3 -0.929 1.01E- 8.78E-03 142-3p 04 hsa-miR- 217963_s_at NGFRAP1 -0.929 1.01E- 8.68E-03 142-3p 04 hsa-miR- 203790_s_at HRSP12 -0.928 1.10E- 9.54E-03 142-3p 04 hsa-miR- 202959_at MUT -0.927 1.11E- 9.63E-03 142-3p 04

Appendix B – negatively correlated targets of miR-204 Gene Pearson Empirical MicroRNA Probeset p-value Symbol Coefficient p-value hsa-miR- 219694_at FAM105A -0.873 9.64E- 6.53E-04 204 04 hsa-miR- 221565_s_at CALHM2 -0.846 2.02E- 1.44E-03 204 03 hsa-miR- 212796_s_at TBC1D2B -0.842 2.24E- 1.67E-03 204 03 hsa-miR- 219544_at C13orf34 -0.824 3.35E- 2.50E-03 204 03 hsa-miR- 209457_at DUSP5 -0.823 3.44E- 2.63E-03 204 03 hsa-miR- 221530_s_at BHLHE41 -0.823 3.42E- 2.58E-03 204 03 hsa-miR- 203320_at SH2B3 -0.822 3.53E- 2.75E-03 204 03 hsa-miR- 212063_at CD44 -0.821 3.60E- 2.88E-03 204 03 hsa-miR- 218870_at ARHGAP15 -0.817 3.88E- 3.21E-03 204 03 hsa-miR- 44790_s_at C13orf18 -0.815 4.06E- 3.41E-03 204 03 hsa-miR- 220052_s_at TINF2 -0.813 4.28E- 3.67E-03 204 03 hsa-miR- 219471_at C13orf18 -0.812 4.34E- 3.77E-03 204 03 hsa-miR- 204502_at SAMHD1 -0.809 4.58E- 4.18E-03 204 03 hsa-miR- 204563_at SELL -0.803 5.18E- 4.94E-03 204 03 hsa-miR- 212646_at RFTN1 -0.803 5.13E- 4.86E-03 204 03 hsa-miR- 202664_at WIPF1 -0.802 5.29E- 5.11E-03 204 03 hsa-miR- 203086_at KIF2A -0.801 5.39E- 5.25E-03 204 03 hsa-miR- 209360_s_at RUNX1 -0.801 5.38E- 5.20E-03 204 03 hsa-miR- 209683_at FAM49A -0.798 5.62E- 5.80E-03 204 03 hsa-miR- 213113_s_at SLC43A3 -0.797 5.81E- 6.19E-03 204 03 hsa-miR- 203385_at DGKA -0.796 5.87E- 6.23E-03 204 03 hsa-miR- 213241_at PLXNC1 -0.794 6.09E- 6.75E-03

204 03 hsa-miR- 205624_at CPA3 -0.794 6.08E- 6.71E-03 204 03 hsa-miR- 208891_at DUSP6 -0.793 6.25E- 6.88E-03 204 03 hsa-miR- 204780_s_at FAS -0.793 6.20E- 6.80E-03 204 03 hsa-miR- 212959_s_at GNPTAB -0.792 6.29E- 7.02E-03 204 03 hsa-miR- 203409_at DDB2 -0.791 6.48E- 7.29E-03 204 03 hsa-miR- 219014_at PLAC8 -0.791 6.44E- 7.15E-03 204 03 hsa-miR- 204959_at MNDA -0.79 6.59E- 7.65E-03 204 03 hsa-miR- 202663_at WIPF1 -0.789 6.70E- 8.01E-03 204 03 hsa-miR- 210073_at ST8SIA1 -0.787 6.96E- 8.51E-03 204 03 hsa-miR- 210176_at TLR1 -0.787 6.93E- 8.38E-03 204 03 hsa-miR- 202241_at TRIB1 -0.787 6.90E- 8.29E-03 204 03 hsa-miR- 202693_s_at STK17A -0.786 7.05E- 8.55E-03 204 03 hsa-miR- 207075_at NLRP3 -0.778 7.98E- 9.98E-03 204 03

Appendix C – negatively correlated targets of miR-211

MicroRNA Probeset Gene Pearson p-value Empirical Symbol Coefficient p-value hsa-miR- 219544_at C13orf34 -0.831 2.92E-03 1.65E-03 211 hsa-miR- 208891_at DUSP6 -0.814 4.15E-03 2.85E-03 211 hsa-miR- 203086_at KIF2A -0.79 6.54E-03 5.84E-03 211 hsa-miR- 203385_at DGKA -0.786 7.06E-03 6.78E-03 211 hsa-miR- 202241_at TRIB1 -0.783 7.45E-03 7.42E-03 211 hsa-miR- 213101_s_at ACTR3 -0.767 9.68E-03 1.10E-02 211 hsa-miR- 208736_at ARPC3 -0.762 1.04E-02 1.24E-02 211 hsa-miR- 219694_at FAM105A -0.757 1.13E-02 1.36E-02 211 hsa-miR- 201603_at PPP1R12A -0.751 1.22E-02 1.50E-02 211 hsa-miR- 208644_at PARP1 -0.751 1.24E-02 1.54E-02 211 hsa-miR- 204527_at MYO5A -0.746 1.32E-02 1.65E-02 211 hsa-miR- 218319_at PELI1 -0.744 1.36E-02 1.70E-02 211 hsa-miR- 221565_s_at CALHM2 -0.741 1.42E-02 1.80E-02 211 hsa-miR- 205126_at VRK2 -0.739 1.45E-02 1.88E-02 211 hsa-miR- 213168_at SP3 -0.739 1.47E-02 1.92E-02 211 hsa-miR- 209457_at DUSP5 -0.737 1.50E-02 1.96E-02 211 hsa-miR- 208296_x_at TNFAIP8 -0.736 1.52E-02 1.98E-02 211 hsa-miR- 200791_s_at IQGAP1 -0.734 1.57E-02 2.05E-02 211 hsa-miR- 204780_s_at FAS -0.732 1.60E-02 2.12E-02 211 hsa-miR- 203696_s_at RFC2 -0.732 1.61E-02 2.14E-02 211 hsa-miR- 220052_s_at TINF2 -0.73 1.65E-02 2.20E-02 211

98 hsa-miR- 203518_at LYST -0.73 1.66E-02 2.23E-02 211 hsa-miR- 210176_at TLR1 -0.728 1.69E-02 2.27E-02 211 hsa-miR- 204781_s_at FAS -0.728 1.70E-02 2.29E-02 211 hsa-miR- 202664_at WIPF1 -0.725 1.78E-02 2.41E-02 211 hsa-miR- 219402_s_at DERL1 -0.723 1.81E-02 2.46E-02 211 hsa-miR- 203269_at NSMAF -0.723 1.82E-02 2.50E-02 211 hsa-miR- 202693_s_at STK17A -0.719 1.91E-02 2.73E-02 211 hsa-miR- 213102_at ACTR3 -0.718 1.93E-02 2.77E-02 211 hsa-miR- 206991_s_at CCR5 -0.713 2.06E-02 3.04E-02 211 hsa-miR- 217947_at CMTM6 -0.708 2.19E-02 3.28E-02 211 hsa-miR- 202663_at WIPF1 -0.7 2.42E-02 3.64E-02 211 hsa-miR- 210260_s_at TNFAIP8 -0.699 2.46E-02 3.71E-02 211 hsa-miR- 218854_at DSE -0.698 2.47E-02 3.72E-02 211 hsa-miR- 203320_at SH2B3 -0.697 2.50E-02 3.79E-02 211 hsa-miR- 213241_at PLXNC1 -0.696 2.54E-02 3.85E-02 211 hsa-miR- 220266_s_at KLF4 -0.692 2.66E-02 4.07E-02 211 hsa-miR- 205249_at EGR2 -0.691 2.70E-02 4.14E-02 211 hsa-miR- 202672_s_at ATF3 -0.689 2.74E-02 4.22E-02 211 hsa-miR- 210073_at ST8SIA1 -0.688 2.78E-02 4.31E-02 211 hsa-miR- 204563_at SELL -0.686 2.86E-02 4.44E-02 211 hsa-miR- 212377_s_at NOTCH2 -0.685 2.88E-02 4.49E-02 211 hsa-miR- 212577_at SMCHD1 -0.684 2.91E-02 4.57E-02 211 hsa-miR- 205624_at CPA3 -0.681 3.02E-02 4.82E-02 211

99 hsa-miR- 201044_x_at DUSP1 -0.68 3.03E-02 4.83E-02 211 hsa-miR- 212796_s_at TBC1D2B -0.68 3.04E-02 4.86E-02 211 hsa-miR- 201368_at ZFP36L2 -0.68 3.06E-02 4.88E-02 211 hsa-miR- 208892_s_at DUSP6 -0.679 3.07E-02 4.92E-02 211 hsa-miR- 209189_at FOS -0.679 3.08E-02 4.93E-02 211 hsa-miR- 203414_at MMD -0.679 3.10E-02 4.96E-02 211

100

Appendix D – Differentially Expressed Genes in Reperfusion (K1 vs. K2) in DGF

146 K2 K1 Fold Probe Mean Change p- q- sets Gene symbol Mean (Log2 (Geometric value value only in (Log2 ) ) "DGF" ) 202464_ 4.04E 1.83E s_at PFKFB3 8.84 9.98 2.21 -04 -02 220330_ 1.06E 3.27E s_at SAMSN1 6.58 7.45 1.83 -03 -02 210002_ 1.47E 3.84E at GATA6 6.60 7.46 1.81 -03 -02 201417_ 8.92E 3.02E at SOX4 9.03 9.73 1.63 -04 -02 202779_ 1.33E 3.63E s_at LOC731049 /// UBE2S 6.34 6.98 1.56 -03 -02 203471_ 1.11E 3.32E s_at PLEK 6.31 6.90 1.50 -03 -02 203835_ 8.22E 2.90E at LRRC32 8.56 9.08 1.44 -04 -02 221045_ 8.90E 3.02E s_at PER3 6.44 6.94 1.41 -04 -02 205463_ 2.19E 4.83E s_at PDGFA 6.90 7.37 1.38 -03 -02 214721_ 3.06E 1.50E x_at CDC42EP4 7.32 7.79 1.38 -04 -02 203044_ 6.79E 2.59E at CHSY1 8.63 9.09 1.37 -04 -02 213716_ 2.17E 4.82E s_at SECTM1 8.00 8.39 1.31 -03 -02 203634_ 1.57E 3.97E s_at CPT1A 4.86 5.24 1.30 -03 -02 206472_ 2.27E 4.95E s_at TLE3 5.23 5.60 1.29 -03 -02 220027_ 1.06E 3.27E s_at RASIP1 6.92 7.28 1.29 -03 -02 201313_ 7.15E 2.65E at ENO2 5.17 5.53 1.28 -04 -02 210008_ 1.88E 4.41E s_at MRPS12 6.40 6.74 1.27 -03 -02 209652_ 1.23E 8.00E s_at PGF 5.48 5.83 1.27 -04 -03 211289_ 1.64E 4.09E x_at CDC2L1 /// CDC2L2 6.38 6.72 1.26 -03 -02 218612_ 8.31E 2.91E s_at TSSC4 6.56 6.89 1.26 -04 -02 220510_ 1.60E 4.02E at RHBG 7.00 7.32 1.25 -03 -02 213098_ 3.78E 1.75E at RQCD1 6.05 6.36 1.24 -04 -02 221922_ 9.60E 3.13E at GPSM2 5.08 5.38 1.23 -04 -02 219480_ 2.14E 1.18E at SNAI1 5.95 6.25 1.23 -04 -02 209999_ 1.41E 3.74E x_at SOCS1 5.40 5.69 1.23 -03 -02 214742_ AZI1 5.97 6.27 1.23 1.75E 4.22E

101 at -03 -02 215454_ 2.18E 4.83E x_at SFTPC 6.19 6.49 1.22 -03 -02 211180_ 2.89E 2.94E x_at RUNX1 4.94 5.23 1.22 -05 -03 215280_ 6.02E 2.35E s_at PPFIA3 5.62 5.90 1.22 -04 -02 219798_ 7.05E 2.63E s_at MEPCE 8.52 8.80 1.21 -04 -02 209359_ 1.21E 3.51E x_at RUNX1 4.83 5.10 1.21 -03 -02 210336_ 3.05E 1.50E x_at MZF1 7.39 7.66 1.21 -04 -02 212707_ 2.06E 4.67E s_at FLJ21767 /// RASA4 6.29 6.55 1.20 -03 -02 219609_ 1.55E 3.96E at WDR25 6.23 6.48 1.19 -03 -02 40446_a 3.71E 1.73E t PHF1 9.35 9.60 1.19 -04 -02 207876_ 1.60E 4.02E s_at FLNC 6.24 6.49 1.19 -03 -02 206827_ 1.29E 3.60E s_at TRPV6 6.84 7.09 1.19 -03 -02 205223_ 1.84E 4.36E at DEPDC5 6.15 6.38 1.17 -03 -02 214133_ 1.77E 4.26E at MUC6 6.17 6.40 1.17 -03 -02 219518_ 1.07E 3.27E s_at ELL3 5.67 5.89 1.17 -03 -02 214148_ 1.15E 3.39E at 0 6.48 6.70 1.17 -03 -02 212048_ 1.07E 3.27E s_at YARS 8.46 8.68 1.16 -03 -02 212356_ 1.06E 3.27E at KIAA0323 7.48 7.69 1.15 -03 -02 210873_ 3.66E 1.71E x_at APOBEC3A 4.47 4.67 1.15 -04 -02 221994_ 9.30E 3.09E at PDLIM5 4.62 4.82 1.15 -04 -02 204014_ 1.73E 4.18E at DUSP4 4.38 4.57 1.14 -03 -02 210726_ 1.09E 3.29E at CYP3A4 5.46 5.64 1.14 -03 -02 217664_ 1.19E 7.82E at 0 4.35 4.52 1.13 -04 -03 214412_ 1.03E 3.24E at H2AFB1 /// H2AFB3 6.03 6.19 1.12 -03 -02 212860_ 9.83E 3.16E at ZDHHC18 7.03 7.19 1.12 -04 -02 210567_ 5.05E 2.08E s_at SKP2 4.26 4.39 1.10 -04 -02 215671_ 9.63E 3.13E at PDE4B 3.84 3.96 1.09 -04 -02 218691_ 1.66E 1.01E s_at PDLIM4 4.30 4.40 1.07 -04 -02 210599_ 2.11E 4.73E at ZNF614 5.00 5.09 1.06 -03 -02 215803_ 1.71E 4.18E at CXorf57 3.84 3.92 1.06 -03 -02 208955_ 9.35E 3.09E at DUT 6.71 6.56 -1.11 -04 -02 207409_ LECT2 4.38 4.23 -1.11 7.67E 2.78E

102 at -04 -02 216387_ 1.76E 4.24E x_at LOC390411 9.15 8.99 -1.12 -03 -02 217340_ 1.32E 1.61E at LOC645452 4.72 4.54 -1.13 -05 -03 201507_ 1.49E 3.87E at PFDN1 9.14 8.97 -1.13 -03 -02 206037_ 1.92E 4.47E at CCBL1 7.27 7.09 -1.13 -03 -02 210525_ 7.44E 2.72E x_at C14orf143 4.38 4.20 -1.13 -04 -02 211628_ 1.29E 3.60E x_at FTHP1 13.49 13.31 -1.13 -03 -02 203609_ 3.95E 1.81E s_at ALDH5A1 5.66 5.47 -1.14 -04 -02 208364_ 1.52E 3.91E at INPP4A 5.28 5.09 -1.14 -03 -02 209517_ 1.88E 1.08E s_at ASH2L 9.24 9.05 -1.14 -04 -02 216532_ 1.55E 3.96E x_at LOC728344 8.11 7.92 -1.14 -03 -02 218534_ 1.97E 4.54E s_at AGGF1 8.67 8.47 -1.15 -03 -02 219217_ 5.57E 2.24E at NARS2 8.43 8.22 -1.16 -04 -02 211955_ 1.29E 3.60E at RANBP5 8.96 8.75 -1.16 -03 -02 216315_ 1.48E 3.85E x_at Kua-UEV /// LOC730052 /// UBE2V1 5.39 5.18 -1.16 -03 -02 220149_ 1.59E 4.01E at C2orf54 6.48 6.26 -1.16 -03 -02 203630_ 8.27E 2.90E s_at COG5 7.89 7.67 -1.17 -04 -02 206369_ 8.69E 2.98E s_at PIK3CG 4.90 4.66 -1.18 -04 -02 217249_ 4.06E 6.95E x_at 0 11.19 10.94 -1.18 -06 -04 214647_ 1.00E 3.20E s_at HFE 5.11 4.87 -1.18 -03 -02 218185_ 2.09E 4.71E s_at ARMC1 9.50 9.26 -1.19 -03 -02 219385_ 7.38E 2.72E at SLAMF8 6.23 5.98 -1.19 -04 -02 202349_ 1.71E 4.18E at TOR1A 7.99 7.73 -1.20 -03 -02 216383_ hCG_2040224 /// LOC285053 /// 1.28E 3.60E at LOC390354 /// LOC729955 /// RPL18A 8.04 7.78 -1.20 -03 -02 211557_ 2.23E 4.89E x_at SLCO2B1 6.80 6.53 -1.21 -03 -02 203505_ 1.34E 3.63E at ABCA1 9.64 9.36 -1.21 -03 -02 203843_ 1.87E 4.41E at RPS6KA3 8.14 7.86 -1.21 -03 -02 201838_ 3.05E 1.50E s_at SUPT7L 6.87 6.59 -1.22 -04 -02 205402_ 3.63E 1.70E x_at PRSS2 5.37 5.08 -1.22 -04 -02 211014_ 1.53E 3.92E s_at LOC161527 /// LOC652671 /// PML 6.37 6.09 -1.22 -03 -02 207821_ 2.20E 4.85E s_at PTK2 8.64 8.35 -1.22 -03 -02 201856_ ZFR 7.60 7.31 -1.22 1.12E 3.34E

103 s_at -03 -02 200890_ 1.03E 3.24E s_at SSR1 8.72 8.43 -1.22 -03 -02 217135_ 1.72E 1.02E x_at 0 7.28 6.99 -1.22 -04 -02 219671_ 1.88E 4.41E at HPCAL4 6.03 5.73 -1.22 -03 -02 204074_ 2.06E 4.67E s_at KIAA0562 7.24 6.95 -1.23 -03 -02 217939_ 7.23E 2.67E s_at AFTPH 10.40 10.10 -1.23 -04 -02 209476_ 1.60E 4.02E at TXNDC1 10.61 10.31 -1.24 -03 -02 208848_ 1.73E 4.18E at ADH5 10.16 9.86 -1.24 -03 -02 211460_ 4.42E 1.94E at TTTY9A /// TTTY9B 5.26 4.95 -1.24 -04 -02 202132_ 1.51E 3.90E at WWTR1 8.23 7.92 -1.24 -03 -02 220007_ 4.58E 1.98E at METTL8 6.60 6.29 -1.24 -04 -02 211994_ 7.54E 2.75E at WNK1 10.48 10.16 -1.25 -04 -02 202100_ 1.47E 3.84E at RALB 10.35 10.03 -1.25 -03 -02 206770_ 1.69E 4.15E s_at SLC35A3 9.43 9.10 -1.25 -03 -02 201943_ 1.03E 3.24E s_at CPD 8.16 7.83 -1.25 -03 -02 203359_ 6.94E 2.62E s_at MYCBP 8.80 8.46 -1.26 -04 -02 205942_ 1.84E 4.36E s_at ACSM3 8.33 7.98 -1.27 -03 -02 214224_ 2.10E 4.72E s_at PIN4 10.11 9.76 -1.27 -03 -02 219287_ 9.29E 3.09E at KCNMB4 7.31 6.96 -1.28 -04 -02 209484_ 1.17E 3.43E s_at NSL1 9.45 9.09 -1.28 -03 -02 201067_ 1.19E 7.82E at PSMC2 9.40 9.05 -1.28 -04 -03 202053_ 1.15E 7.73E s_at ALDH3A2 10.65 10.29 -1.28 -04 -03 210154_ 1.42E 3.75E at ME2 6.57 6.20 -1.29 -03 -02 211960_ 1.42E 3.75E s_at RAB7A 9.83 9.45 -1.29 -03 -02 203243_ 1.68E 4.13E s_at PDLIM5 9.60 9.22 -1.31 -03 -02 214724_ 9.34E 3.09E at DIXDC1 8.71 8.31 -1.32 -04 -02 210840_ 9.73E 3.14E s_at IQGAP1 8.52 8.12 -1.32 -04 -02 209404_ 1.85E 4.38E s_at TMED7 10.02 9.62 -1.32 -03 -02 200745_ 1.14E 3.37E s_at GNB1 10.11 9.70 -1.32 -03 -02 219558_ 1.46E 3.83E at ATP13A3 7.18 6.78 -1.32 -03 -02 201198_ 1.65E 4.11E s_at PSMD1 9.06 8.65 -1.33 -03 -02 202378_ LEPROT 11.53 11.12 -1.33 2.15E 4.80E

104 s_at -03 -02 221745_ 8.17E 2.90E at WDR68 7.62 7.20 -1.34 -04 -02 216650_ 2.77E 1.41E at LOC650303 /// LOC729105 7.18 6.75 -1.35 -04 -02 202592_ 1.25E 3.56E at BLOC1S1 9.64 9.20 -1.36 -03 -02 200760_ 1.45E 9.10E s_at ARL6IP5 10.05 9.61 -1.36 -04 -03 201732_ 1.32E 3.62E s_at CLCN3 7.50 7.04 -1.37 -03 -02 212582_ 7.00E 2.62E at OSBPL8 10.77 10.32 -1.37 -04 -02 205606_ 1.17E 3.42E at LRP6 7.88 7.43 -1.37 -03 -02 36920_a 1.25E 3.56E t MTM1 7.47 7.02 -1.37 -03 -02 217832_ 1.71E 4.18E at SYNCRIP 8.91 8.45 -1.37 -03 -02 216348_ 8.77E 2.99E at LOC402057 /// RPS17 9.72 9.26 -1.37 -04 -02 205371_ 9.22E 3.09E s_at DBT 7.68 7.20 -1.39 -04 -02 200729_ 8.59E 2.97E s_at ACTR2 9.85 9.36 -1.41 -04 -02 212634_ 1.24E 3.56E at KIAA0776 8.10 7.61 -1.41 -03 -02 216521_ 1.61E 4.02E s_at BRCC3 5.85 5.34 -1.42 -03 -02 205882_ 1.91E 4.46E x_at ADD3 10.48 9.97 -1.42 -03 -02 205369_ 2.56E 1.34E x_at DBT 7.16 6.63 -1.45 -04 -02 220038_ 5.40E 2.19E at SGK3 7.88 7.34 -1.45 -04 -02 218967_ 7.07E 2.63E s_at PTER 10.32 9.77 -1.47 -04 -02 201411_ 1.42E 3.75E s_at PLEKHB2 9.91 9.34 -1.49 -03 -02 201646_ 1.30E 3.62E at SCARB2 8.69 8.10 -1.51 -03 -02 220342_ 1.18E 3.44E x_at EDEM3 6.69 6.08 -1.52 -03 -02 221041_ 1.87E 1.08E s_at SLC17A5 8.61 8.00 -1.53 -04 -02 211992_ 1.06E 3.27E at WNK1 9.99 9.37 -1.53 -03 -02 219948_ 1.05E 3.27E x_at UGT2A3 11.48 10.74 -1.66 -03 -02 218700_ 2.68E 1.38E s_at RAB7L1 9.49 8.73 -1.70 -04 -02 213362_ 8.20E 2.90E at PTPRD 8.00 7.19 -1.76 -04 -02 201971_ 5.51E 2.23E s_at ATP6V1A 9.09 8.17 -1.89 -04 -02

105

VITA

Education

Doctor of Medicine Virginia Commonwealth University, Richmond, VA, May 2016 (expected)

Doctor of Philosophy, major: Physiology & Biophysics Virginia Commonwealth University, Richmond, VA, May 2016 (expected)

Masters of Science, major: Pharmacology & Toxicology Virginia Commonwealth University, Richmond, VA, August 2008

Bachelor of Science, major: Chemistry, Pharmacology Concentration, minor: Sociology Duke University, Durham, NC, May 2004

Teaching Experience

Virginia Commonwealth University MCV Campus Richmond, VA October 2010-September 2014 School of Medicine Basic Sciences (M1/M2) Tutor  Tutored students in Anatomy, Medical Physiology, System Patho-physiology  Worked with both individual student & groups of up to 15 students

Virginia Commonwealth University MCV Campus Richmond, VA June 2012 – July 2012 Summer Academic Enrichment Program (SAEP) Teaching Assistant  Conducted preparatory seminar for Medical College Admission Test (MCAT)  Designed mock examination sections of MCAT

Virginia Commonwealth University MCV Campus Richmond, VA June 2011 – August 2011 Pre-Matriculation Workshop Teaching Assistant  Directed gross anatomy lab instruction for incoming medical students  Assisted and tutored students in Biochemistry, Physiology & Anatomy

Duke University Durham, NC August 2003 – May 2004 General Chemistry Teaching Asssistant  Directed general chemistry lab experimentation  Graded tests, class exercises and lab reports

Research Experience

106

Virginia Commonwealth University Department of Physiology, VCUHS Division of Transplant Surgery Richmond, VA August 2010-Present Graduate Assistant  Experimental focus: molecular mechanisms of chronic allograft nephropathy  Prinicipal Investigators: Vladimir Vladimirov, PhD & Catherine Dumur, PhD.  Coordinating tissue sample collection from consented patients  Conducting nucleic acid isolation and genetic experiments on patient tissue samples for gene expression analysis

Virginia Commonwealth University Department of Pharmacology and Toxicology, Richmond, VA, October 2006 – July 2008 Student Researcher/Lab Technician  Experimental focus: localization of the Sphingosine-1-phosphate type 1 receptor in the brain  Principal Investigator: Laura Sim-Selley, Ph.D.  Conducting autoradiography, cell staining and binding experiments  Perform radiation safety tests, make buffers, and manage laboratory upkeep

Duke University Medical Center Anesthesiology Department, Durham, NC, January 2004 – July 2005 Clinical Trials Assistant  Principal Investigators: Eugene Moretti, MD; Elliot Bennett-Guerrero, MD; Anthony Roche, MD  Experimental foci: Genetic factors of Sepsis; Role of biomarkers in the assessment of Sepsis; Biomarkers and post-surgical treatment; Perioperative tissue factor hypercoagulablity; Changes in blood deformability over time.  Coordinated federally and industry sponsored clinical trials with department physicians.  Constructed databases for collecting and organizing pharmacological patient data.  Drafted paper: Pharmacological treatment of genetic factors of Sepsis  Trained in Phlebotomy (Venipuncture, Intravenous, and Arterial Line)

Duke University Medical Center Department of Biological Psychiatry, Durham, NC, December 2003 – May 2004 Research Assistant  Principal Investigator: Edward Levin, Ph.D.  Experimental focus: Effect of nicotinic agonists on β-2 nicotinic receptor knockout mice  Designed dosage regimens and manufactured aliquots of nicotine and mecamylamine for nicotinic receptor experiments.

107

Duke University Medical Center Department of Neurobiology, Durham, NC, May – November 2003 Lab Assistant  Principal Investigator: George Augustine, Ph.D.  Experimental focus: Effects of calcium on synaptic vesicle formation and function  Assisted in performing procedures designed to observe neurotransmitter function.  Organized lab data with both Windows and Linux operating systems.  Handled lab purchases, bookkeeping, and equipment

Publications

Mba U Mba, Daniel G Maluf, MD, Catherine I Dumur, Mariano Scian, PhD, Marc P Posner, MD, Anne L King, MD, Todd WB Gehr, MD, Amit Sharma, MD, Adrian U Cotterell, MD, Qing Ren, MD and Valeria R Mas, PhD. Molecular Biomarkers of Human Kidney Transplantation Ischemia Reperfusion Injury. (Abstract accepted, ATC National Meeting April 30, 2011. Philadelphia, PA)

Mariano Scian, Daniel Maluf, Kellie Archer, Andre A Williams, Jihee Suh, Mba U Mba, Anne King, Todd Gehr, Marc Posner and Valeria Mas. Time Dependent Changes in Kidney Allograft Biopsies Through the First Year Post- Transplantation Associate with Allograft Function. (Abstract accepted, ATC National Meeting April 30, 2011. Philadelphia, PA)

Jihee L Suh, Mariano Scian, Daniel G Maluf, Kellie J Archer, Sarah Reese, Anne L King, Todd W Gehr, Mba U Mba, Adrian H Cotterell, Marc P Posner and Valeria R Mas. Early Protocol Biopsies and Peripheral Blood Profiles Correlates with Graft Function at 9 Months Post-Kidney Transplantation. (Abstract accepted, ATC National Meeting April 30, 2011. Philadelphia, PA)

Mariano Scian, Daniel Maluf, Krystle David, Kellie Archer, Mba U Mba, Anne King, Amit Sharma, Adrian Cotterell, Dhiren Kumar, Marc Posner and Valeria Mas. miRNA Expression Signature Characterizes Kidney Allografts with Interstitial Fibrosis and Tubular Atrophy. (Abstract accepted, ATC National Meeting April 30, 2011. Philadelphia, PA)

Huong Nguyen, Daniel G Maluf, Mariano Scian, Kellie J Archer, Adrian H Cotterell, Jiayi Hou, Mba U Mba, Anne L King, Marc P Posner and Valeria R Mas. Molecular Insights on Calcineurin Inhibitor Toxicity in Kidney Allografts (Abstract accepted, ATC National Meeting April 30, 2011. Philadelphia, PA).

Sim-Selley LJ, Goforth PB, Mba MU, Macdonald TL, Lynch KR, Milstien S, Spiegel S, Satin LS, Welch SP, Selley DE. Sphingosine-1-phosphate receptors mediate neuromodulatory functions in the CNS.J Neurochem. 2009 Aug;110(4):1191-202. Epub 2009 May 31.

108

Moretti EW, Morris RW, Podegreanu M, Schwinn DA, Newman MF, Bennett E, Moulin VG, Mba MU. APOE polymorphism is associated with risk of severe sepsis in surgical patients. Crit CCare Med 2005; 33 (11)

109