Mir-34A Regulates the Invasive Capacity of Canine Osteosarcoma Cell Lines

MiR-34a Regulates the Invasive Capacity of Canine Osteosarcoma Cell Lines

THESIS

Presented in Partial Fulfillment of the Requirements for the Degree Master of Science in the Graduate School of The Ohio State University

Cecilia Montes Lopez

Graduate Program in Comparative and Veterinary Medicine

The Ohio State University

2017

Master's Examination Committee:

Joelle Fenger, DVM, PhD, DACVIM, Advisor

Emma Warry, BVSc (HONS), DACVIM

William Kisseberth, DVM, PhD, DACVIM

Copyrighted by

Cecilia Montes Lopez

2017

Abstract

Osteosarcoma (OSA) is the most common bone tumor in children and dogs; however, no substantial improvement in clinical outcome has occurred in either species over the past 30 years. MicroRNAs (miRNAs) are small non-coding RNAs that regulate gene expression and play a fundamental role in cancer. The purpose of this study was to investigate the potential contribution of miR-34a loss to the biology of canine OSA, a well- established spontaneous model of the human disease.

Real-time PCR demonstrated that miR-34a expression levels were significantly reduced in primary canine OSA tumors and canine OSA cell lines as compared to normal canine osteoblasts. In canine OSA cell lines stably transduced with empty vector or pre- miR-34a lentiviral constructs, overexpression of miR-34a inhibited cellular invasion and migration but had no effect on cell proliferation or cell cycle distribution. Transcriptional profiling of canine OSA8 cells possessing enforced miR-34a expression demonstrated dysregulation of numerous genes, including significant down-regulation of multiple putative targets of miR-34a. Moreover, gene ontology analysis of down-regulated miR-

34a target genes showed enrichment of several biological processes related to cell invasion and motility. Lastly, we validated changes in miR-34a target gene expression, including decreased expression of KLF4, SEM3A, and VEGFA transcripts in canine OSA cells

ii overexpressing miR-34a. Concordant with these data, primary canine OSA tumor tissues demonstrated increased expression levels of putative miR-34a target genes.

These data demonstrate that miR-34a contributes to invasion and migration in canine OSA cells and suggest that loss of miR-34a may promote a pattern of gene expression contributing to the metastatic phenotype in canine OSA.

iii

Dedicated to Aaron Lopez for his incredible support, love, and companionship throughout this journey; to Veronica, Kevin, and Kathleen Montes for their love

throughout my life; and in loving memory of Papi, siempre.

Acknowledgments

I would like to sincerely thank my incredible mentor, Dr. Joelle Fenger, for her patient guidance and teaching ability in this project. Thank you for making me excited about science! My gratitude goes out to the veterinary oncology faculty at The Ohio

State University, including Drs. Cheryl London, William Kisseberth, Emma Warry,

Megan Brown, Eric Green, Noopur Desai, and Vincent Wavreille who have shared their experience and abilities and helped to shape me professionally throughout my residency training. These individuals have helped to make these three years some of the most formative, educational, and progressive of my life.

The contribution of the following collaborators was indispensable to this work:

Peter Wu for his assistance with miR-34a target identification, and Xiaoli Zhang and

Ayse Selen Yilmaz for their efforts with statistical analysis. I would also like to express my gratitude to Tim Vojt of the OSU College of Veterinary Medicine Biomedical Media

Services for his assistance in figure preparation, Feng Xu for her technical assistance, and the Veterinary Clinical Research Shared Resource Biospecimen Repository for their assistance in tumor sample acquisition.

I would like to thank my past and present residentmates, Pachi Clemente, Megan

Brown, Heather Gardner, Rachel Kovac, and Kyle Renaldo, for their encouragement, friendship, and humor. And importantly I would like to thank the veterinary technicians v of the Integrated Oncology Service, Heather Young, Krista Cebulskie, Nicole Pastorek,

Matt Kerzie, and Erin Mash, for their good-natured support.

Vita

2006...... Lake Dallas High School

2010...... B.S. Biomedical Sciences, Texas A&M

University

2013...... D.V.M, Texas A&M University

2013-2014 ...... Rotating Small Animal Intern, VCA

Veterinary Specialty Center of Seattle

2014 to present ...... Combined Medical Oncology Residency and

Masters Program, The Ohio State University

Fields of Study

Major Field: Comparative and Veterinary Medicine

vii

Table of Contents

Abstract ...... ii

Acknowledgments...... v

Vita ...... vii

Table of Contents ...... viii

List of Tables ...... xi

List of Figures ...... xii

Chapter 1: Literature Review ...... 1

Chapter 1.1 Functions of MicroRNAs in Health and Disease ...... 1

Chapter 1.2 MicroRNA Dysregulation in Cancer ...... 2

Chapter 1.3 The miR-34 Family of MicroRNAs in Cancer ...... 4

Chapter 1.4 MiR-34 Dysregulation in Human Osteosarcoma ...... 6

Chapter 1.5 Comparative Aspects of Human and Canine Osteosarcoma ...... 8

Chapter 2: MiR-34a Regulates the Invasive Capacity of Canine Osteosarcoma Cell Lines

...... 11

2.1 Abstract ...... 11

viii

2.2 Introduction ...... 12

2.3 Materials and Methods ...... 15

2.3.1 Cell lines and primary tumor samples ...... 15

2.3.2 RNA isolation, cDNA synthesis, and quantitative real-time PCR ...... 16

2.3.3 MiR-34a lentivirus infection ...... 17

2.3.4 Assessment of cell proliferation ...... 18

2.3.5 Cell-cycle analysis ...... 18

2.3.6 Matrigel invasion assay ...... 19

2.3.7 Wound healing assay ...... 19

2.3.8 RNA sequencing ...... 20

2.3.9 Statistics ...... 21

2.4 Results ...... 21

2.4.1 MiR-34a expression is decreased in primary canine OSA tumors and canine

OSA cell lines ...... 21

2.4.2 MiR-34a expression is decreased in paired canine OSA metastatic lesions

compared to primary OSA tumors ...... 22

2.4.3 Ectopic expression of miR-34a does not influence cell proliferation or cell

cycle distribution in canine OSA cell lines ...... 22

2.4.5 RNA sequencing identifies miR-34a-induced gene alterations in canine OSA8

cells ...... 24 ix

2.4.6 Putative miR-34a target gene expression is increased in primary canine OSA

tumors ...... 25

2.5 Discussion ...... 26

2.6 Future Directions ...... 31

2.7 Conclusions ...... 33

References ...... 46

Appendix A: Supplemental Tables ...... 64

List of Tables

Table 1. Primer sequences for quantitative reverse-transcriptase polymerase chain reaction ...... 34

Table 2: Differentially expressed transcripts (fold change ≥ 1.5 or ≤ -1.5) in canine OSA8 cells overexpressing miR-34a ...... 65

Table 3: Down-regulated transcripts (fold change ≤ -1.4) containing consensus binding sites for miR-34a ...... 68

Table 4: ToppGene Suite gene ontology analysis of putative miR-34a targets ...... 72

List of Figures

Figure 1: MiR-34a expression is reduced in primary canine OSA tumors and OSA cell lines...... 36

Figure 2: Ectopic expression of miR-34a in canine OSA cells does not affect cell proliferation or cell cycle distribution...... 38

Figure 3: Expression of miR-34a in canine OSA cell lines suppresses cell invasion and migration...... 40

Figure 4: Overexpression of miR-34a in canine OSA8 cells significantly alters gene expression...... 42

Figure 5: Identification of putative miR-34a target genes dysregulated by miR-34a overexpression in canine OSA cell lines...... 43

Figure 6: Evaluation of miR-34a target expression in primary canine OSA tumors...... 45

xii

Chapter 1: Literature Review

Chapter 1.1 Functions of MicroRNAs in Health and Disease

MicroRNAs (miRNAs) are small (~15-25 nucleotides), non-protein coding RNAs that alter gene expression at the post-transcriptional level by binding to conserved sequences within the 3’ untranslated (3’ UTR) of target mRNAs, thereby inducing the degradation of the target mRNA or inhibiting translation. The majority of identified miRNAs are highly evolutionarily conserved among many distantly related species suggesting that miRNAs play an important role in essential biological processes [1]. Although miRNAs account for only 1-3% of the human genome, they are thought to regulate up to one-third of the human genes and they have been found to play key roles in a wide variety of biological processes, including cell proliferation, regulation of cell differentiation, lineage determination and homeostasis and their dysregulation is implicated in a variety of pathologic conditions, including cancer [1-3].

Mature miRNAs are typically processed from a 60 to 110 nucleotide hairpin precursor (fold-back) RNA (pre-miRNA) structure that is transcribed from a larger primary transcript (pri-miRNA) [4]. The genes encoding miRNAs are located within introns, intergenic regions or within exons of protein-coding genes. A key feature of all miRNAs

1 is the presence of a stem-loop structure in the primary and precursor miRNA transcripts, in which one strand of the stem is preferentially processed into the functionally mature miRNA [4]. Primary miRNAs are transcribed by RNA polymerase II and processed by the nuclear enzyme Drosha to form a shorter, pre-miRNA hairpin RNA. Following export of the pre-miRNA to the cytoplasm, pre-miRNAs undergo subsequent processing by the endonuclease Dicer, which cleaves and removes the hairpin portion, resulting in a ~15-25 nucleotide double-stranded miRNA. Further processing and unwinding of the double- stranded miRNA ultimately results in one strand (or in some cases, both strands) of the mature miRNA being incorporated into the RNA-induced silencing complex (RISC), which helps guide the interaction of the miRNA strand with its target mRNA. MiRNAs function as regulators of gene expression by binding through imperfect complimentarity to conserved sequences within the 3’ UTR of their target mRNAs. The actual miRNA-mRNA binding site is composed of a small (5-8 base pairs) ‘seed sequence’, which allows for a broad range of mRNA targets. Upon binding to the seed sequence of their target transcripts, miRNAs reduce the levels of target mRNAs through degradation or translational repression [5].

Chapter 1.2 MicroRNA Dysregulation in Cancer

Dysregulation of miRNAs is now a well-described phenomenon in human cancers where they function as either suppressors or promoters of tumor proliferation, differentiation and metastasis. The main mechanism of miRNA dysregulation in cancer cells seems to be represented by aberrant gene expression, characterized by abnormal levels 2 of expression for mature and/or precursor miRNA transcripts in comparison with the corresponding normal tissues [6]. The molecular aberrations underlying dysregulated expression of miRNAs in cancer include the location of miRNAs at cancer associated genomic regions, alterations in miRNA processing genes and proteins, and epigenetic regulation of miRNA expression [7].

Expression profiling of human miRs has demonstrated that poorly differentiated tumors can be identified via unique expression signatures [8, 9]. These studies have demonstrated a general down regulation of miRNAs in tumors compared to normal tissues, suggesting a correlation between the expression levels of certain miRNAs and the degree of cellular differentiation, whereas messenger RNA profiles were highly inaccurate when applied to the same samples. MiRNA profiling of 6 different human solid tumors has identified a solid tumor cancer miRNA signature [10]. Furthermore, the predicted targets for the differentially expressed miRNAs were significantly enriched for essential protein- coding tumor suppressor genes and oncogenes and support their function as either dominant or recessive genes involved in tumorigenesis. Indeed, it is now well-established that miRNAs can play a causal role in the development of cancer by inhibiting the expression of various tumor suppressor genes (thus being considered oncogenic miRNAs, or “oncomiRs”), or they can inhibit tumorigenesis by targeting genes involved in tumor progression or metastasis (referred to as tumor suppressor miRNAs). For example, the tumor suppressor miRNA let-7 [11] is frequently downregulated in human lung cancers and is correlated with shortened survival times. Furthermore, functional studies have shown that let-7 miRNAs repress cell proliferation pathways in human lung cancer cells in

3 vitro and inhibit lung tumor growth in vivo by targeting c-MYC and the RAS family of oncogenes [12, 13]. In contrast, overexpression of miR-21 has been documented in several human cancers, including hepatocellular, lung, and head and neck carcinomas [14-17].

Several studies have shown that miR-21 functions as an oncomiR as evidenced by enhanced cell proliferation and chemoresistance in cell lines overexpressing miR-21. The pro-tumorigenic functions of miR-21 in carcinomas are thought to be mediated, in part through direct repression of the tumor suppressor, PTEN by miR-21 and subsequent downstream activation in the AKT signaling pathway [18]. Taken together, these data suggest a fundamental role for miRNAs in cancer and highlight the potential of miRNA profiling as a diagnostic and prognostic tool. Efforts are now underway to routinely use miRNA profiles for diagnostic and prognostic purposes and explore the clinical utility of tumor-specific miRNAs in the serum of cancer patients as non-invasive biomarkers for early cancer detection and monitoring response to treatment [19, 20].

Chapter 1.3 The miR-34 Family of MicroRNAs in Cancer

The miR-34 family consists of three evolutionarily conserved miRNAs: MiR-34a,

MiR-34b and MiR-34c. The mature miR-34a sequence is located within the second exon of its non-coding host gene whereas miR-34b and miR-34c are co-transcribed and located within a single non-coding precursor (miR-34b/c) [21]. The miR-34 family has been intensively studied in human cancer [21, 22] and data indicate family members are frequently downregulated in a number of human malignancies including neuroblastoma, glioma, breast cancer, non-small cell lung cancer, colorectal cancer, and osteosarcoma [23- 4

28]. The molecular mechanisms underlying the observed loss of miR-34 genes in cancer involve deletions of the gene regions harboring these transcripts or CpG promoter methylation with miR-34 gene silencing. Additionally, p53-mediated transcriptional regulation of the miR-34 family is highly conserved across different cell types [29-33].

Indeed, the miR-34 family members function as tumor suppressors, inducing apoptosis, cell cycle arrest and senescence, in part, through their interaction with the p53 tumor suppressor network [29, 33-35]. For example, recent studies demonstrated that miR-34a contributes to cell cycle arrest and apoptosis through its repression of several p53 target genes, including CDK4, CDK6, and Bcl-2 [28, 36, 37].

Interestingly, studies using miR-34-null transgenic mouse models have shown that mice deficient in miR-34 do not display increased susceptibility to spontaneous, irradiation-induced, or c-Myc-initiated tumorigenesis [38]. Additionally, complete inactivation of miR-34 function is compatible with normal murine development. These findings are in agreement with prior studies using miRNA engineered animal models, few of whom have demonstrated that dysregulation of a single oncogenic or tumor suppressor miRNA can lead to the initiation and development of a malignant tumor. As many miRNAs ‘fine tune’ the expression of most targets, with only a small number of targets experiencing a substantial change in mRNA or protein abundance, the phenotypic consequences of altered miR-34 expression may depend on the nature of the environmental conditions or genetic context of a cell, or alternatively, may emerge during the later stages of tumor progression.

As miR-34 family members are frequently lost or downregulated in cancer and functional studies have confirmed that their dysregulation is causal in many cases of human malignancy, several miRNA mimics and miRNA-targeted therapeutics to have been developed and have shown promise in preclinical development. A few miRNA-targeted therapeutics, including MRX34, a liposomal miR-34a mimic have reached clinical development and are being evaluated for safety and toxicity in humans. A phase I clinical trial using MRX34 was recently completed in patients with advanced solid tumors and data from this trial demonstrated that MRX34 had an acceptable safety and toxicity profile and that antitumor activity was observed [39, 40]. Efforts involving the development and evaluation of novel miRNA mimics and molecules targeted at miRNAs (antimiRs) in cancer and in other diseases are ongoing with the ultimate goal of identifying the most promising and efficacious candidates and achieving safe and targeted delivery of miRNA therapeutically.

Chapter 1.4 MiR-34 Dysregulation in Human Osteosarcoma

While the field of miRNA biology and the contribution of miRNA dysregulation in human osteosarcoma (OSA) has expanded considerably, the mechanisms by which miRNAs contribute to the pathogenesis of OSA remains poorly understood. Unique miRNA expression signatures in human OSA are associated with risk of metastasis and response to chemotherapy, providing evidence to support a role for altered miRNA expression in contributing to the pathogenesis of OSA [41]. In addition, the discovery of miRNA signatures associated with tumor cell origin and/or specific molecular alterations 6 highlights miRNA profiling as a potential molecular diagnostic and prognostic tool for complex karyotype sarcoma. For example, low expression of the microRNAs, miR-34a and miR-192 in osteosarcoma tumor tissues has been shown to be associated with a shorter disease-free survival in human OSA patients [42]. Furthermore, recent studies demonstrate that circulating tumor-specific miRNAs in body fluids such as serum and plasma are capable of mediating tumorigenesis, microenvironment modulation, and metastasis in osteosarcoma. For example, serum levels of miR-27a have been evaluated as a diagnostic marker in osteosarcoma [43], and high plasma expression levels of miR-196a and miR-

196b have been shown to be associated with a poorer prognosis [44].

With respect to OSA, miR-34a expression is decreased in human primary OSA tumor samples when compared to adjacent normal tissues, suggesting a direct role for miR-

34a dysregulation in disease pathogenesis [26]. In human OSA tumor tissues, miR-34 genes exhibited minimal deletions, loss of heterozygosity (LOH), and epigenetic inactivation, demonstrating that other genetic and epigenetic mechanisms may account for the observed decrease in expression [26]. Furthermore, restoration of miR-34a in human

OSA cell lines reduced cell proliferation and migration in vitro and attenuated OSA tumor xenograft growth and metastasis in vivo, in part through targeting of the receptor tyrosine kinase Met [45]. Effects of miR-34 on OSA cell line migration and invasiveness appear to be at least partially mediated through repression of CD44, the receptor for hyaluronic acid and a well-established marker of cancer cell stemness [46]. In concordance with the potential role of miR-34a in malignant osteoblast behavior, decreased expression of miR-

34a in spontaneous human OSA tumors and low levels of circulating miR-34a in OSA

7 patient plasma are associated with shorter disease-free survival and poor prognosis [42,

47].

Chapter 1.5 Comparative Aspects of Human and Canine Osteosarcoma

Osteosarcoma (OSA) is the most common form of malignant bone cancer in dogs and adolescent humans. OSA in both species shares many features including the presence of microscopic tumor spread at diagnosis, the development of resistance to chemotherapy, and dysregulation of several key molecular pathways. Despite aggressive therapy including radical surgery and multi-agent chemotherapy, approximately 30% of children and over 90% of dogs will ultimately succumb to metastatic disease [48, 49]. However, the development of new, more effective therapies are critically needed as little improvement in clinical outcome have occurred in dogs or children in over 30 years.

The similarities between pediatric and canine OSA with regard to histology, biological behavior and molecular genetic alterations indicate that canine OSA is a relevant, spontaneous, large animal model of the pediatric disease. [50] Importantly, OSA in both species exhibit overlapping gene expression profiles and shared DNA copy number aberrations, providing evidence to support the idea that these disease are similar at the molecular level [51]. For example, comparative genomic approaches demonstrate that human and canine OSA are frequently characterized by mutations in the p53 tumor suppressor gene, loss of RB1, and activation of STAT3 signaling [50, 52, 53]. To this end, preclinical studies performed in dogs with naturally occurring OSA have contributed to the development of novel therapeutic approaches such as aerosolized chemotherapy delivery, 8

IGF-1 targeted therapy, and validated the activity of immunotherapy in the metastatic disease setting.

The ability to rapidly advance therapeutics in children with OSA is limited by the low incidence of this disease and need for long follow-up times to assess impact. OSA in pet dogs occurs approximately 10 times more frequently than children, and pet owners are often motivated to seek advanced care for their dogs to alleviate the pain associated with this disease and improve quality of life. This presents a unique opportunity to study the biology of OSA in a spontaneous disease setting and evaluate new treatment strategies in a significantly larger population of patients, thereby accelerating the pace at which clinical trials in client-owned dogs can be completed and informing the drug-development pathway before translation into new human trials [50].

Recent studies have confirmed the similar to their human counterpart, altered miRNA expression is frequently observed in canine cancers. MiRNA expression profiles have been characterized in several canine malignancies, including hemangiosarcoma, urogenital tumors, osteosarcoma, mast cell tumors, and lymphoma, and unique miRNA expression signatures have been shown to correlate with tumor grade and prognosis [54-

61]. Significantly, several miRNAs reported to be aberrantly expressed in canine OSA and lymphoma/leukemia are known to be similarly altered in the human disease suggesting that altered expression of miRNAs may play a critical role in mediating the development and progression of these cancers in both humans and dogs [57, 58, 62]. The use of circulating tumor-specific miRNAs as non-invasive biomarkers is similarly being explored in dogs; for example in Bernese mountain dogs, low circulating levels of let-7g has been identified

9 as a potentially useful biomarker to detect and monitor disease burden in disseminated histiocytic sarcoma [63].

With respect to the contribution of miRNAs to the biology of canine and human

OSA, our laboratory has recently identified a unique miRNA expression profile associated with canine OSA tumors. Members of this miRNA signature are known to be dysregulated in human OSA tumors, including miR-34a which is also downregulated in canine OSA.

These data provide support for the notion that loss of miR-34a promotes the aggressive biologic behavior of canine OSA and suggest that dysregulation of miRNAs may be fundamental to the disease process in both species. As such, the purpose of this body of work was to investigate the potential contribution of miR-34a loss to the biology of canine

OSA, a well-established spontaneous model of the human disease, and evaluate the functional consequences of altered miR-34a expression in canine OSA cell lines.

Chapter 2: MiR-34a Regulates the Invasive Capacity of Canine Osteosarcoma Cell Lines

2.1 Abstract

34a target genes showed enrichment of several biological processes related to cell invasion and motility. Lastly, we validated changes in miR-34a target gene expression, including

11 decreased expression of KLF4, SEM3A, and VEGFA transcripts in canine OSA cells overexpressing miR-34a. Concordant with these data, primary canine OSA tumor tissues demonstrated increased expression levels of putative miR-34a target genes.

2.2 Introduction

Osteosarcoma (OSA) is the most common form of malignant bone cancer in dogs and children, although the incidence of disease in dogs is approximately ten times higher than that in people [50, 64, 65]. OSA in both species shares many features including the presence of microscopic tumor spread at diagnosis, the development of resistance to chemotherapy, and dysregulation of several key molecular pathways. Indeed, the similarities between pediatric and canine OSA with regard to histology, biological behavior and molecular genetic alterations indicate that canine OSA is a relevant, spontaneous, large animal model of the pediatric disease [50]. OSA in both species exhibit overlapping gene expression profiles and shared DNA copy number aberrations, providing evidence to support the idea that these disease are similar at the molecular level

[50, 66]. For example, comparative genomic approaches demonstrate that human and canine OSA are frequently characterized by mutations in the p53 tumor suppressor gene, loss of RB1, and activation of STAT3 signaling [50, 52, 53]. To this end, preclinical studies performed in dogs with naturally occurring OSA have contributed to the 12 development of novel therapeutic approaches such as aerosolized chemotherapy delivery,

IGF-1 targeted therapy, and validated the activity of immunotherapy in the metastatic disease setting. However, despite aggressive therapy including radical surgery and multi- agent chemotherapy, approximately 30% of children and over 90% of dogs will ultimately succumb to metastatic disease [48, 49]. As such, the development of new, more effective therapies are critically needed as little improvement in clinical outcome have occurred in dogs or children in over 30 years.

MicroRNAs (miRNAs) are small noncoding RNAs that regulate gene expression at the post-transcriptional level through either mRNA cleavage and/or translational repression. Although miRNAs account for only 1-3% of the human genome, they are thought to regulate up to one-third of the human genes and they have been found to play key roles in a wide variety of biological processes, including cell proliferation, regulation of cell differentiation, and their dysregulation is implicated in a variety of pathologic conditions, including cancer [67, 68]. Accumulating evidence suggests that miRNA dysregulation is causal in many cases of cancer and that miRNAs can function as tumor suppressors or oncogenes, making them relevant targets for therapeutic intervention [7, 9,

69-71]. In support of this, functional studies using chemically modified oligonucleotides that mimic or molecularly target miRNAs have confirmed that restoring or suppressing the function of miRNAs in malignant cells can alter cancer phenotypes in vitro [72-76].

Among the miRNAs implicated in cancer development and progression, the miR-

34 family has been intensively studied and data indicate family members function as tumor suppressors in a variety of human cancers [21, 22]. The miR-34 family consists of three

13 evolutionarily conserved miRNAs: MiR-34a, MiR-34b and MiR-34c. The mature miR-

34a sequence is located within the second exon of its non-coding host gene whereas miR-

34b and miR-34c are co-transcribed and located within a single non-coding precursor

(miR-34b/c) [21]. Deletions of the gene regions harboring these transcripts, or CpG promoter methylation with miR-34 gene silencing, are frequently observed in human malignancies including neuroblastoma, glioma, breast cancer, non-small cell lung cancer, colorectal cancer, and osteosarcoma [23-28]. Furthermore, p53-mediated transcriptional regulation of the miR-34 family is conserved across different cell types [29-33]. Indeed, the miR-34 family members function as tumor suppressors, inducing apoptosis, cell cycle arrest and senescence, in part, through their interaction with the p53 tumor suppressor network [29, 33-35]. For example, recent studies demonstrated that miR-34a contributes to cell cycle arrest and apoptosis through its repression of several p53 target genes, including CDK4, CDK6, and Bcl-2 [28, 36, 37].

With respect to OSA, miR-34a expression is decreased in human primary OSA tumor samples when compared to adjacent normal tissues, suggesting a role for miR-34a dysregulation in disease pathogenesis [26]. MiR-34 genes exhibited minimal deletions, loss of heterozygosity (LOH), and epigenetic inactivation in human OSA tumor tissues, demonstrating that other genetic and epigenetic mechanisms may account for the observed decrease expression [26]. Furthermore, restoration of miR-34a in human OSA cell lines reduced cell proliferation and migration in vitro and attenuated OSA tumor xenograft growth and metastasis in vivo, in part through targeting of the receptor tyrosine kinase Met

[45]. Effects of miR-34 on OSA cell line migration and invasiveness appear to be at least

14 partially mediated through repression of CD44, the receptor for hyaluronic acid and a well- established marker of cancer cell stemness [46]. In concordance with the potential role of miR-34a in malignant osteoblast behavior, decreased expression of miR-34a in spontaneous human OSA tumors and low levels of circulating miR-34a in OSA patient plasma are associated with shorter disease-free survival and poor prognosis [42, 47].

Studies evaluating miRNA expression in spontaneously occurring canine OSA demonstrate that similar to human cancers, alteration of miRNAs are consistently observed.

Our laboratory has recently identified a unique miRNA expression signature associated with primary canine OSA tumors that is distinct from that of normal canine osteoblasts

[57]. Significantly, several members of this miRNA signature are known to be aberrantly expressed in human OSA tumors, including miR-34a which is also downregulated in canine OSA. The purpose of this study was to investigate the potential contribution of miR-34a loss to the biology of canine OSA, a well-established spontaneous model of the human disease, and evaluate the functional consequences of altered miR-34a expression in canine OSA cell lines.

2.3 Materials and Methods

2.3.1 Cell lines and primary tumor samples

Canine OSA cell lines OSA8 and Abrams were kindly provided by Dr. Jaime

Modiano (University of Minnesota, Minneapolis, MN, USA) and Dr. Doug Thamm

(Colorado State University, Fort Collins, CO, USA), respectively. Canine osteoblast cells were purchased from Cell Applications (Cell Applications Inc.) and cultured in Canine 15

Osteoblast Growth Medium (Cat. # Cn417-500, Cell Applications Inc.) according to manufacturer’s instructions. The canine OSA8 and Abrams cell lines were maintained in

RPMI-1640 (Gibco Life Technologies, Grand Island, NY, USA) or DMEM (Gibco Life

Technologies) supplemented with 10% fetal bovine serum, non-essential amino acids, sodium pyruvate, penicillin, streptomycin, L-glutamine, and HEPES (4-(2-dydroxethyl)-

1-piperazineethanesulfonic acid) at 37ºC, supplemented with 5% CO2 (media supplements from Gibco). Canine OSA tumor tissues were obtained from dogs treated at The Ohio

State University College of Veterinary Medicine (OSU CVM) in compliance with established hospital policies regarding sample collection as part of the Veterinary Clinical

Research Shared Resource Biospecimen Repository (IACUC #2010A0015), Tumors were evaluated and confirmed as OSA by board certified veterinary pathologists at the OSU

CVM.

2.3.2 RNA isolation, cDNA synthesis, and quantitative real-time PCR

RNA was extracted from OSA cell lines and primary OSA tumors using the RNeasy

Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer’s instructions. Real- time PCR was performed using the Applied Biosystems StepOne Plus Detection System

(Applied Biosystems, Foster City, CA, USA). Human Taqman miRNA assays (Applied

Biosystems) were used according to manufacturer’s instructions to quantify mature miR-

34a expression in canine cell lines and tissues (mature miR-34a shares 100% sequence homology between dogs and humans) [57]. 50 ng total RNA was converted to first-strand cDNA with miRNA-specific primers, followed by real-time PCR with TaqMan probes.

All samples were normalized to U6 snRNA. Real-time PCR was performed to validate changes in mRNA expression for selected genes affected by miR-34a over expression. cDNA was made from 1 μg of total RNA using Superscript III (Invitrogen). KLF4,

SEMA3E, VEGFA, and GAPDH transcripts were detected using Fast SYBR green PCR master mix (Applied Biosystems) according to the manufacturer’s protocol; primer sets are detailed in Table 1. Normalization was performed relative to GAPDH. All reactions were performed in triplicate and included no-template controls for each gene. Relative gene expression for all real-time PCR data was calculated using the comparative threshold cycle method [77]. Experiments were repeated 3 times using samples in triplicate.

2.3.3 MiR-34a lentivirus infection

Lentiviral constructs were purchased from Systems Biosciences (Mountain View,

CA, USA). Packaging of the lentiviral constructs was performed using the pPACKH1

Lentivector Packaging KIT (catalog no. LV500A-1) according to the manufacturer’s instructions. Canine OSA8 and Abrams cells were transduced with empty lentivirus

(catalog no. CD511B-1) or pre-miR-34a lentivirus (catalog no. PMIRH34aPA-1). Briefly,

5 x 105 cells were plated and left overnight in 10% serum-containing medium. The following day, the medium was changed to Stemline (Gibco) with transfection agent

TransDux (Systems Biosciences) and either pre-miR-34a or control empty virus was added to cells according to manufacturer’s protocol. Cells were sorted based on GFP expression

72 hour post-transduction and miR-34a expression was evaluated by real-time PCR

(Applied Biosystems).

2.3.4 Assessment of cell proliferation

Changes in cell proliferation were assessed using a commercially available bromodeoxyuridine (BrdU) incorporation assay (Roche, Basel, CH). Briefly, 2 x 103

Abrams or OSA8 cells transduced with either pre-miR-34a or negative control empty virus were plated in complete medium for 24, 48, or 72 hours in 96-well plates. The BrdU reagent was added, cells were incubated for another 3 hours and then harvested, fixed, and digested by nuclease for 30 minutes at 37°C. Cells were then incubated with conjugate for

1 hour and washed, and the plates were developed by adding 100 µL of substrate for 30 minutes. The absorbance was measured using an enzyme-linked immunosorbent assay

(ELISA) plate reader (Spectra Max; Molecular Devices, Sunnyvale, CA, USA).

2.3.5 Cell-cycle analysis

The propidium iodide staining method was used to assess the effects of miR-34a on cell cycle distribution [78]. In brief, OSA8 or Abrams cells (5.0 x 106 cells/well) transduced with pre-miR-34a or empty vector control lentivirus were seeded in 6-well plates in 3 mL of RPMI-1640 (OSA8) or DMEM (Abrams) with 10% fetal bovine serum and incubated overnight at 37°C and 5% CO2. Cells were then collected, washed in 0.1% glucose/PBS, and then fixed in cold 70% ethanol at 4° overnight. Cells were incubated with 200 µL of propidium iodide working solution consisting of propidium iodide (50

µg/mL) and RNAse (10 µg/mL) in PBS containing 0.1% glucose, and analyzed by means of flow cytometry (BD Accuri C6, BD Biociences, San Jose, CA, USA). Data were

18 analyzed by means of standard software (BD Accuri™ C6 Software). Each experiment was repeated 3 times with each cell line.

2.3.6 Matrigel invasion assay

A Matrigel invasion assay was performed to evaluate the impact of miR-34a on invasion. Cell culture inserts (8-μm pore size; Falcon) were coated with 100 μL of diluted

(1:10) Matrigel (BD Bioscience, San Jose, CA, USA) and allowed to solidify at 37°C for

1 h to form a thin continuous layer. Canine OSA8 (5 x 104) or Abrams cells (1 × 105) transduced with empty vector or pre-miR-34a lentivirus were prepared in serum-free medium and seeded into each insert (upper chamber) and media containing 10% fetal bovine serum was placed in the lower chamber. Cells were incubated for 24 hours to permit invasion through the Matrigel layer. Inserts were processed and cells counted as previously described [57]. Samples were run in triplicate and cells from ten independent 20x high powered fields from each replicate were counted. Experiments were repeated 3 times for each cell line.

2.3.7 Wound healing assay

To evaluate the effects of miR-34a on cell migration, 5 x 106 canine OSA8 or

Abrams cells transduced with control lentivirus or pre-miR-34a lentivirus were seeded in complete medium and grown until confluent in 6-well plates. A gap was introduced in the cells by scraping with a P200 pipette tip and cells were placed in fresh medium containing

10% fetal bovine serum. After 24 hours, cells were stained with crystal violet stain and

19 migration across the gap was evaluated by digital photography. Each experiment was repeated 3 times.

2.3.8 RNA sequencing

Total RNA was extracted from OSA8 cells transduced with either empty lentivirus

(n=3) or pre-miR-34a lentivirus (n=3) using the RNeasy Mini Kit (QIAGEN) and RNA sequencing was performed at the OSU Comprehensive Cancer Center Genomics Shared

Resource as previously described [57]. Total RNA was subject to ribosomal reduction and

RNA-seq libraries were constructed using TruSeq stranded total RNA (Illumina; RS-122-

2201) according to the manufacturer’s instructions. Sequencing was performed on an

Illumina HiSEq. 2500 instrument at a depth of ∼40 million paired-end, 50 bp long, strand- specific reads per sample. AdapterRemover was used to trim adapter sequences and the remaining reads were aligned using STAR to the canFam3 genome. Quality control was performed using RNA-SeQC, to identify samples requiring additional sequencing reads.

Differential expression was determined using DESeq2, which compared sample groups using a likelihood-ratio test (LRT). Statistical analysis for mRNA expression data was determined by student’s t-test and p-values of less than 0.05 were considered statistically significant.

Prediction of miR-34a binding to the 3’-UTR of genes down-regulated by miR-34a was performed with computer-aided algorithms obtained from TargetScan

(http://www.targetscan.org), PicTar (http://pictar.mdc-berlin.de), miRanda

(http://www.microrna.org), and miRWalk (http://www.umm.uni-

20 heidelberg.de/apps/zmf/mirwalk). Down-regulated genes (fold change ≤ -1.4) that contained consensus binding sites for miR-34a in their 3’UTR were identified and gene ontology enrichment analysis was performed using the ToppGene database

(https://toppgene.cchmc.org; FDR correction and p < 0.05).

2.3.9 Statistics

Experiments were performed in triplicate and repeated 3 times. Data were presented as mean plus or minus standard deviation. Real-time PCR miRNA or gene expression data was first normalized to internal control (snU6 and GAPDH, respectively) and the ΔΔCt method was used to compare miRNA expression [77]. For analysis of all cell proliferation, invasion, and real-time PCR expression data, a linear mixed effects model was used to take account of the correlations among observations run in the same biological replicate. P-values of less than 0.05 were considered statistically significant.

2.4 Results

2.4.1 MiR-34a expression is decreased in primary canine OSA tumors and canine OSA cell lines

We previously reported a miRNA expression signature that was associated with canine OSA using the nanoString nCounter platform, including down-regulation of miR-

34a in primary canine OSA tumors. Concordant with these data, real-time PCR confirmed that miR-34a expression is substantially decreased in canine OSA tumors as compared to normal canine osteoblasts (Fig 1). Although levels of miR-34a expression varied among 21

OSA tumor samples, this may be attributed, in part, to heterogeneity in tumor stroma and/or inflammatory cells present in the tumor microenvironment. Furthermore, in a panel of canine OSA cell lines, miR-34a expression was found to be significantly decreased in all of the malignant OSA cell lines compared to that observed in normal canine osteoblasts.

These findings are consistent with published data demonstrating that miR-34a expression levels are significantly decreased in human OSA tumor tissues and OSA cell lines [26, 45].

2.4.2 MiR-34a expression is decreased in paired canine OSA metastatic lesions compared to primary OSA tumors

Given the highly metastatic behavior of OSA and data demonstrating that miR-34a influences OSA metastasis in vivo, miR-34a expression levels were evaluated in paired primary OSA tumors and metastatic lung lesions from eight canine patients. Real-time

PCR showed that 50% (4/8) of paired specimens showed reduced miR-34a in metastasis compared to the primary tumor (Fig 1B). MiR-34a levels were unchanged in two of the paired samples (two already had reduced miR-34a levels in the primary tumor compared to normal osteoblasts), whereas two metastatic tissues exhibited higher miR-34a expression as compared to the primary OSA tumor biopsy. Therefore, down-regulation of miR-34a may be associated with the metastatic phenotype in canine OSA.

2.4.3 Ectopic expression of miR-34a does not influence cell proliferation or cell cycle distribution in canine OSA cell lines

To investigate the functional consequences of miR-34a expression on OSA cell behavior, we stably expressed miR-34a in the OSA8 and Abrams cells lines. OSA8 and

Abrams cell lines were transduced with empty or pre-miR-34a expressing lentivirus vector and 72 hours post transduction, cells were sorted based on GFP-positivity and overexpression of miR-34a was confirmed by real-time PCR (Fig 2A).

The impact of miR-34a overexpression on the proliferative capacity of OSA cells was assessed using a bromodeoxyuridine (BrdU) incorporation assay. No effects of miR-

34a expression on proliferation of the OSA8 or Abrams cell line were observed at multiple time points when compared to cells expressing empty vector (Fig 2B). Concordant with this finding, miR-34a expression did not alter cell cycle status in either the OSA8 or

Abrams cells as assessed by propidium iodide (PI) staining (Fig 2C).

2.4.4 MiR-34a expression decreases cell invasion and migration in canine osteosarcoma cell lines

Prior studies indicated that miR-34a expression alters the invasive capacity of tumor cells [45]. Similarly, overexpression of miR-34a significantly inhibited cell invasion through Matrigel in the OSA8 and Abrams cells as compared to cells expressing empty vector (Fig 3A). Furthermore, miR-34a decreased cell motility and scattering as compared to control cells in both cell canine OSA cell lines (Fig 3B). Taken together, these findings demonstrate that miR-34a inhibits the invasive capacity and migratory behavior of canine

OSA cell lines.

2.4.5 RNA sequencing identifies miR-34a-induced gene alterations in canine OSA8 cells

To understand the molecular mechanisms underlying the miR-34a associated decrease in cell motility and invasion in canine OSA cell lines, RNA sequencing was performed on the OSA8 cells expressing either control or pre-miR-34a lentiviral constructs. Overexpression of miR-34a significantly altered the gene expression profile of

OSA8 cells, including substantial up-regulation (using a cut-off for fold change of ≥ 1.5) of 15 genes and downregulation (using a cut-off for fold change of ≤ -1.5) of 109 gene transcripts when compared to control OSA8 cells (Fig 4, p ≤ 0.0001; S1 Table).

Computer-aided bioinformatics algorithms (TargetScan, miRanda. MiRWalk, miRDB) were used to identify predicted miR-34a target genes containing putative miR-

34a binding sites within their 3’-UTR. Out of the 773 genes in total that were underexpressed in cells over-expressing miR-34a, we identified 150 genes containing consensus binding sites for miR-34a that were significantly down-regulated (using a cut- off for fold change of ≤ -1.4) following enforced expression of miR-34a in OSA8 cells.

This demonstrates that the miR-34a-downregulated genes are enriched for direct targets of this miRNA (S2 Table). Furthermore, we analyzed the Gene Ontology (GO) classifications of these genes using ToppGene Suite (http://toppgene.cchmc.org) and found that putative miR-34a target genes are highly enriched for biological processes related to cell migration, including GO categories “regulation of cell projection organization” (p-value 7.55 x 10-5),

“regulation of locomotion” (p-value 1.96 x 10-4), and “cell projection organization” (p- value 3.90 x 10-4), consistent with the role of miR-34a in mediating cellular invasion and motility in OSA cells (S3 Table).

We identified putative miR-34a target genes that were common to all three GO categories related to cell migration including Krüppel-like factor 4 (KLF4), Semaphorin

3E (SEMA3E), and Vascular endothelial growth factor A (VEGFA). Real-time PCR was performed on OSA8 cells expressing empty vector or miR-34a to validate changes in mRNA expression following miR-34a overexpression. Consistent with our RNA sequencing results, transcripts for KLF4, SEMA3E, and VEGFA were significantly downregulated following overexpression of miR-34a in OSA8 cells, suggesting that miR-34a may alter the invasive capacity and migratory behavior of OSA8 cells, in part, by regulating the expression of these genes (Fig 5). To investigate whether similar alterations in transcript expression occurred following overexpression of miR-34a in other canine OSA cell lines, real-time PCR was performed for KLF4, SEMA3E, and VEGFA in the canine

Abrams OSA cell line expressing either empty control or miR-34a lentivectors (Fig 5). In agreement with our findings in the OSA8 cell line, transcript levels for KLF4, SEMA3E, and VEGFA were substantially decreased following enforced miR-34a expression in the

Abrams cell line.

2.4.6 Putative miR-34a target gene expression is increased in primary canine OSA tumors

To determine whether expression levels of putative miR-34a target genes are increased in spontaneous canine OSA tumors, we performed real-time PCR to evaluate gene transcript expression (KLF4, SEMA3E, and VEGFA) in primary canine OSA tumors and normal canine osteoblast cells. We first confirmed that basal levels of miR-34a were significantly lower in the primary canine OSA tumors as compared to normal canine

25 osteoblasts (Fig 6A). As shown in Fig 6, real-time PCR demonstrated that primary canine

OSA tumor tissues express high levels of putative miR-34a target genes, KLF4 and

SEMA3E. In contrast, normal canine osteoblasts possessing high basal levels of miR-34a expressed significantly lower levels of KLF4 and SEMA3E compared to OSA tumor samples. Interestingly, transcript levels of VEGFA varied among tumor samples and differential expression of VEGFA was not statistically significant in OSA tumors compared to osteoblast cells; however, this may have been due, in part, to variations in stroma/inflammatory cells within the tumor microenvironment or baseline necrosis within the primary tumor specimen that influenced the proportion of tumor cells. Taken together, these findings support the assertion that loss of miR-34a may promote a pattern of gene expression contributing to the metastatic phenotype in canine OSA.

2.5 Discussion

The contribution of miRNAs to cancer has been widely confirmed by numerous studies demonstrating aberrant expression of miRNAs in almost all types of human cancer.

The miR-34 family, most notably miR-34a, is frequently lost or down-regulated in human malignancies including neuroblastoma, breast, lung, and colorectal carcinomas, and osteosarcoma. Furthermore, the introduction of miR-34a gene constructs into cell lines derived from solid tumors (lung, liver, colon, pancreatic, brain, prostate, bone, ovary) and hematopoietic malignancies (lymphoma, leukemias) induces apoptosis and cell cycle arrest, and inhibits migration and invasion, providing support for a tumor suppressor function of miR-34a [21, 22, 36]. Given the prevalence of miR-34a dysregulation in 26 cancer and its ability to down-regulate the expression of numerous oncogenes across multiple oncogenic pathways, there is ongoing interest in developing miR-34a mimic therapy as a novel therapeutic strategy for cancer. Indeed, a variety of miRNA formulations and target-specific delivery strategies have accelerated the clinical development of miR-34 mimics, notably Miravirsen (Santaris Pharma) and MRX34 (Mirna

Therapeutics) which recently entered first-in-human phase I clinical trials (NCT01829971) in patients with advanced solid tumors [39, 79, 80].

With respect to the potential role of miR-34a dysregulation in osteosarcoma (OSA), several studies have demonstrated decreased expression in human OSA. Importantly, miR-

34a loss in OSA is associated with enhanced expression of a several targets known to contribute to tumorigenesis including MET, SIRT1, and CDK6 [45, 81, 82] although its contribution to the biology of OSA has not yet been fully elucidated. In our prior work we evaluated miRNA expression signatures in canine OSA and found that canine OSA tumors similarly express low levels of miR-34a. Given that canine OSA is a well validated spontaneous large animal model of the human disease, the purpose of this study was to characterize the impact of miR-34a expression in canine OSA tumor lines to better understand its potential role in the biology of this disease.

Our data demonstrate that expression of miR-34a is significantly decreased in primary canine OSA tumor tissues and cell lines compared to normal canine osteoblasts.

These findings are consistent with previous studies demonstrating that miR-34a expression is reduced or absent in human OSA tumor samples compared to paired non-cancerous bone and suggest that loss of miR-34a is a common event in both species [26]. We analyzed

27 miR-34a expression levels in primary canine OSA tumors and paired metastases and found that half of metastatic lesions had reduced levels of miR-34a compared with their primary tumor, supporting the assertion that decreased expression of miR-34a is associated with a metastatic phenotype.

Studies have shown that ectopic expression of miR-34a through various methods such as chemically modified oligonucleotide miR-34a mimics and bioengineered miR-34a prodrugs results in decreased viability, migration and invasion of human OSA cell lines

[46, 81, 82]. Similarly, in the canine OSA cell lines, stable overexpression of miR-34a resulted in significant inhibition of cellular invasion and migratory capacity. Interestingly, enforced expression of miR-34a had no effect on proliferation or cell cycle distribution in the cell lines evaluated. However, this may be explained by differences in the expression of mRNA targets in distinct cell types/tissues that influence the effect of miR-34a on cell behavior. Similarly, differences in the capacity of miR-34a to inhibit cell viability or suppress cellular migration and invasion have been observed following miR-34a overexpression in different human OSA tumor cell lines [45, 46]. These cross-species data support the notion that miR-34a plays a role in regulating the invasive capacity of OSA cells.

Enforced expression of miR-34a decreases the invasive capacity of human OSA cell lines in vitro and modulates the progression of OSA lung metastasis in vivo, in part, through direct targeting of oncogenes such as c-Met, sirtuin 1 (SIRT1), and CD44 [45, 46,

82]. However, given the broad capacity for miRNAs to regulate multiple genes and it is likely that miR-34a affects a range of genes that participate in metastasis [68]. In order to

28 begin to understand the underlying mechanisms mediating the observed miR-34a- dependent effects in canine OSA cells, RNA sequencing was performed on canine OSA8 cells expressing control or miR-34a lentiviral constructs. Notably, enforced expression of miR-34a significantly altered the transcriptome of canine OSA8 cells. We found that 150 of 773 down-regulated genes (19%) contained consensus binding sites for miR-34a, demonstrating enrichment for direct targets of this miRNA. Furthermore, gene ontology analysis of these miR-34a target genes demonstrated associations with biological processes related to cell migration. We validated the expression profiles of select putative miR-34a target genes including Krüppel-like factor 4 (KLF4), Semaphorin 3E (SEMA3E), and vascular endothelial growth factor receptor A (VEGFA) and confirmed significant down- regulation of these genes in OSA8 and Abrams cells overexpressing miR-34a.

Semaphorins comprise a family of secreted and membrane-associated proteins that play multifunctional roles in embryonic development, immune regulation, angiogenesis, cancer, and bone formation [83, 84]. Accumulating evidence suggests that signaling cascades triggered by class 3 semaphorins play a crucial role in regulating bone homeostasis, osteoblast differentiation and osteodeposition and their dysregulation is implicated in malignant bone pathology. In support of this idea, SEMA4D and SEMA6D transcripts were found to be highly expressed in human OSA tumors compared to normal osteoblasts and over 50% of tumors on an OSA tissue microarray were shown to express high levels of these proteins [85]. While the function of Sema3E has not been evaluated in osteosarcoma, the importance of Sema3E in promoting the metastatic phenotype has been explored in murine mammary adenocarcinoma cells, which show enhanced cellular

29 invasion and rapid experimental lung metastasis formation in vivo in response to Sema3E overexpression [86, 87]. Moreover, the expression of Sema3E in human tumor samples correlates with metastatic progression and in support of these findings, knockdown of

Sema3E in human carcinoma cell lines reduced their metastatic potential when injected into mice [88-90].

Krüppel-like factor 4 (KLF4) is a zinc-finger transcription factor that is often overexpressed in human primary OSA tumors compared normal bone and is highly expressed in OSA stem cell populations [91, 92]. ShRNA-mediated knockdown of KLF4 in human MG63 and SaOS2 cell lines inhibited cell clonogenicity and abrogated cell migration in wound-healing and Matrigel migration assays and these effects were directly related to downregulation of the KLF4 transcriptional target, CRYAB [91]. Furthermore, inhibition of KLF4 in human OSA xenografts reduced tumor growth in vivo, providing further support for its role in OSA biology.

During the process of hematogenous metastasis, VEGFA is suggested to play a major role in regulating angiogenesis. Ectopic expression of miR-34a in human head and neck squamous cell carcinoma (HNSCC) cells was found to inhibit tumor growth and angiogenesis that was partially reversed by blocking VEGFA production by tumor cells

[93]. Concordant with these data, stable overexpression of miR-34a in canine OSA cell lines reduced VEGFA expression with a concomitant decrease in cell invasion and migration. Higher levels of VEGFA expression in human OSA tumors have been shown to correlate with the presence of lung metastasis and VEGFA levels have predictive value for survival of OSA patients [94]. With respect to the canine disease counterpart, higher

30 levels of plasma VEGFA were found in more aggressive neoplasms in a survey of spontaneously occurring tumors in dogs, including OSA [95]. Furthermore, activation of

∆Np63 has been shown to promote cellular invasion, migration and angiogenesis of canine

OSA cell lines, in part through its effects on VEGFA, IL-8, and STAT3 phosphorylation

[96]. Collectively, these data support the notion that enhanced VEGFA production in canine OSA cells may promote angiogenesis, thereby contributing to the metastatic cascade.

Concordant with the observation that miR-34a decreased the expression of

SEM3A, KLF4, and VEGFA in canine OSA cell lines, we found that in primary canine

OSA tumors expressing low basal levels of miR-34a, expression of these target genes was significantly increased compared to normal osteoblasts possessing high miR-34a levels.

These data provide a further association between miR-34a and regulation of cellular pathways that promote a metastatic phenotype. To draw firm conclusions regarding direct regulation of target gene expression by miR-34a, a functional approach for each gene would be required to confirm direct targeting of transcripts by miR-34a. Furthermore, loss and gain of function studies would further elucidate the contribution of these targets to normal and malignant osteoblast invasion and as such, this represents an ongoing area of investigation.

2.6 Future Directions

Our data indicate that loss of miR-34a contributes to the aggressive biologic behavior of canine OSA, possibly through the promotion of an invasive phenotype as 31 demonstrated by enhanced invasion and migration of canine OSA cell lines in the presence of enforced miR-34a expression. Furthermore, overexpression of miR-34a in

OSA cell lines altered their gene expression profile and downregulated a number of predicted miR-34a target genes that were enriched for biological processes related to cell motility. While we have confirmed that several of these putative miR-34a gene targets

(KLF4, SEMA3E, VEGFA) are significantly downregulated in OSA cells overexpressing miR-34a, further investigation of individual targets merit further investigation.

Specifically, demonstration of direct binding of miR-34a to the seed sequence of its target genes using dual-luciferase reporter assay approaches would provide further convincing evidence of direct targeting and regulation of individual gene targets by miR-34a.

Further assessment of the contribution of miR-34a predicted target genes (KLF4,

SEMA3E, VEGFA) using molecular silencing techniques (siRNA) in OSA cell lines overexpressing miR-34a would define their roles in mediating the miR-34a-induced invasion and migration phenotype in OSA cell lines. Lastly, the function of miR-34a in contributing to normal canine osteoblast biology remain largely unclear. To this end, inhibition of miR-34a in normal canine osteoblast cells is necessary in order to further evaluate the functional consequences of miR-34a silencing in normal osteoblast biology.

The findings of decreased miR-34a expression in primary and metastatic canine osteosarcoma tumor tissues merit further investigation as to the role of loss of miR-34a in the biology of OSA metastasis. Given the heterogeneity of primary OSA tumor tissues, evaluating miR-34a expression levels using miRNA in situ hybridization techniques or performing real time PCR to detect miR-34a expression levels on microdissected tumor

32 samples would allow more accurate determination of tumor tissue-specific miR-34a expression in OSA. The limited number of samples analyzed in this study precluded the ability to draw clinical conclusions and increasing the number of paired primary and metastatic samples may afford an opportunity to identify possible prognostic factors related to miR-34a expression. Additionally, further evaluation of paired and metastatic tumor samples would provide a unique opportunity to identify pathways altered by miR-

34a that are associated with the metastatic phenotype.

2.7 Conclusions

Our data demonstrate that miR-34a expression is significantly decreased in primary canine OSA tumor tissues and canine OSA cell lines compared to normal canine osteoblasts. Our findings are concordant with data generated in human OSA tumors, suggesting that loss of miR-34a may be common in this disease. Enforced expression of miR-34a in canine OSA cell lines reduced cellular invasion and migration and significantly altered the transcriptional profile of OSA cells, including down-regulation of several genes implicated in promoting the metastatic phenotype. This work serves as the foundation for future work to dissect the molecular mechanisms by which miR-34a regulates the invasive capacity of canine OSA cells, with the ultimate goal of identifying new targets for therapeutic intervention in this disease.

Gene/primers Primer sequences

Canine KLF4 457F 5’-GCC ACC TGG CGA GTC TGA CAT G-3’

Canine KLF4 606R 5’-CGC TTC ATG TGC GAG AGC TCC TC-3’

Canine SEMA3E 766F 5’-GCT CTG TAC CAG GAA TGA ATG-3’

Canine SEMA3E 935R 5’-CCG GAT GCT AGA CAT ATG ATA C-3’

Canine VEGFA 299F 5’-GTG CAT TGG AGC CTT GCC TTG CTG-3’

Canine VEGFA 429R 5’-CAG TAG CTG CGC TGG TAG ACG TC-3’

Canine GAPDH F 5’-GTC CAT GCC ATC ACT GCC ACC CAG-3’

Canine GAPDH R 5’-CTG ATA CAT TGG GGG TGG GGA CAC-3’

Table 1. Primer sequences for quantitative reverse-transcriptase polymerase chain reaction

Figure 1. MiR-34a expression is reduced in primary canine OSA tumors and OSA cell lines.

(A) Real time PCR was used to assess mature miR-34a levels in primary canine OSA tumors (N=20) and canine OSA cell lines (N=8) relative to normal canine osteoblasts.

Total RNA was reverse-transcribed and human Taqman miRNA assays were used to detect miR-34a expression. All results were normalized to snU6 control. All reactions were performed in triplicate and each experiment was repeated 3 times. Raw data were log transformed to reduce variance and skewness. Linear mixed effects models were applied to expression data to account for the correlation of the observations from the same batch.

Real time PCR results showed significant down-regulation of miR-34a in canine OSA tumors (p < 0.01) and OSA cells lines (p < 0.05) as compared to normal canine osteoblasts.

(B) Metastasis-associated changes in miR-34a expression in canine OSA patients using paired primary OSA tumor biopsy and metastatic tumor tissue. MiR-34a expression was evaluated in paired primary canine OSA tumor biopsy and lung metastasis from 8 patients using Human Taqman miRNA assays to detect miR-34a as described previously. All results were normalized to snU6 control and data are shown relative to normal canine osteoblasts. While 2/8 samples showed increased miR-34a expression in metastatic tissues compared to primary tumor biopsies, 4/8 (50%) had decreased miR-34a expression. MiR-

34a levels were unchanged in 2 patients.

Figure 1

Figure 2. Ectopic expression of miR-34a in canine OSA cells does not affect cell proliferation or cell cycle distribution.

(A) Canine OSA8 and Abrams cells were transduced with pre-miR-34a lentiviral constructs or empty control vector and sorted to greater than 95% purity based on GFP expression 72 hr following infection. Total RNA was isolated and real time PCR was performed as described above to confirm transduction efficiency miR-34a levels in wild- type (WT), empty vector (EV), and miR-34a expressing cells (*p < 0.05). (B) OSA8 and

Abrams cells transduced with either empty control or pre-miR-34a lentiviral constructs were plated in complete media and the cell proliferation was assessed at 24, 48, and 72 hours using the BrdU incorporation assay. Cell proliferation was measured at 490 nm.

Values of optical density (OD) are expressed as means +/- SD of 3 independent experiments. (C) OSA8 and Abrams cells expressing empty vector or pre-miR-34a lentivirus vector were incubated in complete media for 24 hours. Cells were then evaluated for effects on cell cycle using propidium iodide (PI) staining and flow cytometry. Three independent experiments were performed, and 1 representative result is presented. Linear mixed effects models were applied to OSA8 and Abrams cell line miR-34a expression, proliferation and cell cycling data to take account of the correlation among observations from the same replicates. No statistically significant difference in cell proliferation or cell cycle distribution was detected in OSA8 or Abrams cells expressing empty vector or pre- miR-34a vector for any of the tested time points.

Figure 2

Figure 3. Expression of miR-34a in canine OSA cell lines suppresses cell invasion and migration.

(A) Canine OSA8 and Abrams cells expressing control or pre-miR-34a lentiviral constructs were plated in serum free media in the upper wells of plates for Matrigel invasion assays.

The lower chamber of each well contained complete media supplemented with 10% fetal bovine serum. Cells were allowed to invade for 24 hours through a layer of Matrigel basement membrane before removal of Matrigel and media from the upper chambers.

Cells that had migrated to the lower surface of the insert membrane were stained with crystal violet and the number of invaded cells were counted in ten random fields in triplicate replicates. Experiments were performed in triplicate and data are represented as the mean ± standard deviation (*p ≤ 0.05). (B) Migration behavior was assessed in OSA8 and Abrams cell lines expressing empty vector or pre-miR-34a lentiviral constructs using standard wound-healing assays. Cells were seeded in 6-well plates in complete growth medium and allowed to grow to 70% confluency. A scratch was introduced using a P200 pipet tip and after 24 hours, cells were fixed, stained with crystal violet and evaluated by digital photography.

Figure 3 40

Figure 4: Overexpression of miR-34a in canine OSA8 cells significantly alters gene expression.

Total RNA was isolated from canine OSA8 cells expressing control or pre-miR-34a lentiviral constructs from three separate transduction experiments and next-generation sequencing was performed to identify differences in gene transcript expression.

Hierarchical cluster analysis of 110 genes differentially expressed in OSA8 cells expressing either empty vector (EV) or miR-34a (miR34a) as determined by one-way

ANOVA comparison test (p ≤ 0.0001). Color areas indicate relative expression of each gene after log2 transformation with respect to the gene median expression (red and green colors denote high and low expression, respectively).

REEP2 PLCB2 C11orf73 12 CYP2S1 GRIK5 HSPB6 GNB3 HHIP 10 ENSCAFG00000015625 THRB ENSCAFG00000029992 DNM3 8 AMER1 SEC14L4 FAM69B SYT4 PICALM 6 ATAD1 ENSCAFG00000019046 ENSCAFG00000018844 MEG3_2 LTBP2 4 LIPA MAP3K7CL ZNF280B BRINP2 ENSCAFG00000004384 2 FIGF CASP14 PRAME PRSS23 PAPSS2 0 SI TMEM60 RSBN1L ENSCAFG00000019160 FAS MAGEB18 ENSCAFG00000031663 SGCE PHTF2 CREBZF MAGEB16 LUZP4 PTEN cfa-mir-34a TNS4 GPRIN1 SELP CNTN2 GPR85 ZNF793 ITGA4 FAM111A GATA3 DNAJC1 C11orf24 SLC2A10 CHAC1 ASAH1 SLC5A6 STK38L FAM3C STRA6 TINAGL1 ANGPTL4 PDE4B MINPP1 ENSCAFG00000014034 GABRA3 TEX13A ENSCAFG00000014436 ENSCAFG00000014043 TMEM126B ADAMTS14 ADCY2 LHFPL3 RIPK3 DEF6 GUCY1A3 SLC13A5 PRTFDC1 CTPS2 EFNB3 TCP11L1 ZNF385B CUBN RNF167 SLC24A3 TAGLN3 MFAP5 LOXL4 CRTAC1 COL9A2 ARMCX2 P3H3 PEG10 PDPN GAP43 ENSCAFG00000020060 TFPI2 P3H2 ENSCAFG00000029405 EYA4 ENSCAFG00000032762 ENSCAFG00000005922 ENSCAFG00000024834 NIPAL2 TMEM200A PCDHB2 PCDHB15

EV3 EV1 EV2 miR34a5 miR34a4 miR34a6

Figure 4 42

Figure 5: Identification of putative miR-34a target genes dysregulated by miR-34a overexpression in canine OSA cell lines.

Transcriptional profiling of canine OSA8 cells expressing pre-miR-34a (miR-34a) or empty vector (EV) control was performed by next-generation sequencing to identify genes showing differential expression with miR-34a overexpression. Real-time PCR was performed to independently validate changes in gene expression for putative miR-34a targets (A) KLF4, (B) SEM3AE, and (C) VEGFA altered by miR-34a overexpression in both canine OSA8 and Abrams cell lines.

Figure 6: Evaluation of miR-34a target expression in primary canine OSA tumors.

(A) Real-time PCR evaluating mature miR-34a expression in primary canine OSA tumors

(N=9) and normal canine osteoblasts demonstrated that the mean expression of miR-34a was significantly reduced in OSA tumors compared to osteoblasts (p < 0.05). All reactions were performed in triplicate and results were normalized to snU6. Transcript levels of putative miR-34a target genes (B) KLF4, (C) SEM3AE, and (D) VEGFA were assessed in primary canine OSA tumors and normal canine osteoblasts using real-time PCR. All reactions were performed in triplicate and results were normalized to GAPDH. Raw data were log transformed to reduce variance and skewness. Linear mixed effects models were applied to expression data to account for the correlation of the observations from the same batch.

Figure 6

References

1. Bartel DP. MicroRNAs: genomics, biogenesis, mechanism, and function. Cell.

2004;116(2):281-97. Epub 2004/01/28. PubMed PMID: 14744438.

2. Du T, Zamore PD. microPrimer: the biogenesis and function of microRNA.

Development (Cambridge, England). 2005;132(21):4645-52. Epub 2005/10/15. doi:

10.1242/dev.02070. PubMed PMID: 16224044.

3. Griffiths-Jones S, Grocock RJ, van Dongen S, Bateman A, Enright AJ. miRBase: microRNA sequences, targets and gene nomenclature. Nucleic acids research.

2006;34(Database issue):D140-4. Epub 2005/12/31. doi: 10.1093/nar/gkj112. PubMed

PMID: 16381832; PubMed Central PMCID: PMCPMC1347474.

4. Hammond SM. An overview of microRNAs. Adv Drug Deliv Rev. 2015. doi:

10.1016/j.addr.2015.05.001. PubMed PMID: 25979468.

5. Ling H, Fabbri M, Calin GA. MicroRNAs and other non-coding RNAs as targets for anticancer drug development. Nat Rev Drug Discov. 2013;12(11):847-65. doi:

10.1038/nrd4140. PubMed PMID: 24172333; PubMed Central PMCID:

PMCPMC4548803.

6. Bracken CP, Scott HS, Goodall GJ. A network-biology perspective of microRNA function and dysfunction in cancer. Nature reviews Genetics. 2016;17(12):719-32. doi:

10.1038/nrg.2016.134. PubMed PMID: 27795564.

7. Iorio MV, Croce CM. Causes and consequences of microRNA dysregulation.

Cancer journal (Sudbury, Mass). 2012;18(3):215-22. Epub 2012/06/01. doi:

10.1097/PPO.0b013e318250c001. PubMed PMID: 22647357; PubMed Central PMCID:

PMCPMC3528102.

8. Lu J, Getz G, Miska EA, Alvarez-Saavedra E, Lamb J, Peck D, et al. MicroRNA expression profiles classify human cancers. Nature. 2005;435(7043):834-8. Epub

2005/06/10. doi: 10.1038/nature03702. PubMed PMID: 15944708.

9. Calin GA, Croce CM. MicroRNA signatures in human cancers. Nature reviews

Cancer. 2006;6(11):857-66. Epub 2006/10/25. doi: 10.1038/nrc1997. PubMed PMID:

17060945.

10. Volinia S, Calin GA, Liu CG, Ambs S, Cimmino A, Petrocca F, et al. A microRNA expression signature of human solid tumors defines cancer gene targets.

Proceedings of the National Academy of Sciences of the United States of America.

2006;103(7):2257-61. Epub 2006/02/08. doi: 10.1073/pnas.0510565103. PubMed PMID:

16461460; PubMed Central PMCID: PMCPMC1413718.

11. Masliah-Planchon J, Garinet S, Pasmant E. RAS-MAPK pathway epigenetic activation in cancer: miRNAs in action. Oncotarget. 2016;7(25):38892-907. Epub

2016/10/23. doi: 10.18632/oncotarget.6476. PubMed PMID: 26646588; PubMed Central

PMCID: PMCPMC5122439.

12. Esquela-Kerscher A, Trang P, Wiggins JF, Patrawala L, Cheng A, Ford L, et al.

The let-7 microRNA reduces tumor growth in mouse models of lung cancer. Cell cycle

(Georgetown, Tex). 2008;7(6):759-64. Epub 2008/03/18. doi: 10.4161/cc.7.6.5834.

PubMed PMID: 18344688.

13. He XY, Chen JX, Zhang Z, Li CL, Peng QL, Peng HM. The let-7a microRNA protects from growth of lung carcinoma by suppression of k-Ras and c-Myc in nude mice. Journal of cancer research and clinical oncology. 2010;136(7):1023-8. Epub

2009/12/25. doi: 10.1007/s00432-009-0747-5. PubMed PMID: 20033209.

14. Gao L, Ren W, Zhang L, Li S, Kong X, Zhang H, et al. PTENp1, a natural sponge of miR-21, mediates PTEN expression to inhibit the proliferation of oral squamous cell carcinoma. Molecular carcinogenesis. 2017;56(4):1322-34. Epub 2016/11/20. doi:

10.1002/mc.22594. PubMed PMID: 27862321.

15. Darido C, Georgy SR, Wilanowski T, Dworkin S, Auden A, Zhao Q, et al.

Targeting of the tumor suppressor GRHL3 by a miR-21-dependent proto-oncogenic network results in PTEN loss and tumorigenesis. Cancer cell. 2011;20(5):635-48. Epub

2011/11/19. doi: 10.1016/j.ccr.2011.10.014. PubMed PMID: 22094257.

16. He C, Dong X, Zhai B, Jiang X, Dong D, Li B, et al. MiR-21 mediates sorafenib resistance of hepatocellular carcinoma cells by inhibiting autophagy via the PTEN/Akt pathway. Oncotarget. 2015;6(30):28867-81. Epub 2015/08/28. doi:

10.18632/oncotarget.4814. PubMed PMID: 26311740; PubMed Central PMCID:

PMCPMC4745697.

17. Xue X, Liu Y, Wang Y, Meng M, Wang K, Zang X, et al. MiR-21 and MiR-155 promote non-small cell lung cancer progression by downregulating SOCS1, SOCS6, and

PTEN. Oncotarget. 2016;7(51):84508-19. Epub 2016/11/05. doi:

10.18632/oncotarget.13022. PubMed PMID: 27811366.

18. Meng F, Henson R, Wehbe-Janek H, Ghoshal K, Jacob ST, Patel T. MicroRNA-

21 regulates expression of the PTEN tumor suppressor gene in human hepatocellular cancer. Gastroenterology. 2007;133(2):647-58. Epub 2007/08/08. doi:

10.1053/j.gastro.2007.05.022. PubMed PMID: 17681183; PubMed Central PMCID:

PMCPMC4285346.

19. Schwarzenbach H, Nishida N, Calin GA, Pantel K. Clinical relevance of circulating cell-free microRNAs in cancer. Nature reviews Clinical oncology.

2014;11(3):145-56. Epub 2014/02/05. doi: 10.1038/nrclinonc.2014.5. PubMed PMID:

24492836.

20. Hayes J, Peruzzi PP, Lawler S. MicroRNAs in cancer: biomarkers, functions and therapy. Trends in molecular medicine. 2014;20(8):460-9. doi:

10.1016/j.molmed.2014.06.005. PubMed PMID: 25027972.

21. Agostini M, Knight RA. miR-34: from bench to bedside. Oncotarget.

2014;5(4):872-81. Epub 2014/03/25. doi: 10.18632/oncotarget.1825. PubMed PMID:

24657911; PubMed Central PMCID: PMCPMC4011589.

22. Misso G, Di Martino MT, De Rosa G, Farooqi AA, Lombardi A, Campani V, et al. Mir-34: a new weapon against cancer? Molecular therapy Nucleic acids. 2014;3:e194.

Epub 2014/09/24. doi: 10.1038/mtna.2014.47. PubMed PMID: 25247240; PubMed

Central PMCID: PMCPMC4222652.

23. Chim CS, Wong KY, Qi Y, Loong F, Lam WL, Wong LG, et al. Epigenetic inactivation of the miR-34a in hematological malignancies. Carcinogenesis.

2010;31(4):745-50. Epub 2010/02/02. doi: 10.1093/carcin/bgq033. PubMed PMID:

20118199.

24. Cole KA, Attiyeh EF, Mosse YP, Laquaglia MJ, Diskin SJ, Brodeur GM, et al. A functional screen identifies miR-34a as a candidate neuroblastoma tumor suppressor gene. Mol Cancer Res. 2008;6(5):735-42. Epub 2008/05/29. doi: 10.1158/1541-

7786.mcr-07-2102. PubMed PMID: 18505919; PubMed Central PMCID:

PMCPMC3760152.

25. Gao H, Zhao H, Xiang W. Expression level of human miR-34a correlates with glioma grade and prognosis. J Neurooncol. 2013;113(2):221-8. Epub 2013/03/27. doi:

10.1007/s11060-013-1119-1. PubMed PMID: 23529798.

26. He C, Xiong J, Xu X, Lu W, Liu L, Xiao D, et al. Functional elucidation of MiR-

34 in osteosarcoma cells and primary tumor samples. Biochem Biophys Res Commun.

2009;388(1):35-40. Epub 2009/07/28. doi: 10.1016/j.bbrc.2009.07.101. PubMed PMID:

19632201.

27. Siemens H, Neumann J, Jackstadt R, Mansmann U, Horst D, Kirchner T, et al.

Detection of miR-34a promoter methylation in combination with elevated expression of c-Met and beta-catenin predicts distant metastasis of colon cancer. Clinical cancer research : an official journal of the American Association for Cancer Research.

2013;19(3):710-20. Epub 2012/12/18. doi: 10.1158/1078-0432.ccr-12-1703. PubMed

PMID: 23243217.

28. Li L, Yuan L, Luo J, Gao J, Guo J, Xie X. MiR-34a inhibits proliferation and migration of breast cancer through down-regulation of Bcl-2 and SIRT1. Clinical and experimental medicine. 2013;13(2):109-17. Epub 2012/05/25. doi: 10.1007/s10238-012-

0186-5. PubMed PMID: 22623155.

29. He L, He X, Lim LP, de Stanchina E, Xuan Z, Liang Y, et al. A microRNA component of the p53 tumour suppressor network. Nature. 2007;447(7148):1130-4. Epub

2007/06/08. doi: 10.1038/nature05939. PubMed PMID: 17554337; PubMed Central

PMCID: PMCPMC4590999.

30. He L, He X, Lowe SW, Hannon GJ. microRNAs join the p53 network--another piece in the tumour-suppression puzzle. Nature reviews Cancer. 2007;7(11):819-22. Epub

2007/10/05. doi: 10.1038/nrc2232. PubMed PMID: 17914404; PubMed Central PMCID:

PMCPMC4053212.

31. He X, He L, Hannon GJ. The guardian's little helper: microRNAs in the p53 tumor suppressor network. Cancer research. 2007;67(23):11099-101. Epub 2007/12/07. doi: 10.1158/0008-5472.can-07-2672. PubMed PMID: 18056431.

32. Hermeking H. p53 enters the microRNA world. Cancer cell. 2007;12(5):414-8.

Epub 2007/11/13. doi: 10.1016/j.ccr.2007.10.028. PubMed PMID: 17996645.

33. Okada N, Lin CP, Ribeiro MC, Biton A, Lai G, He X, et al. A positive feedback between p53 and miR-34 miRNAs mediates tumor suppression. Genes Dev.

2014;28(5):438-50. Epub 2014/02/18. doi: 10.1101/gad.233585.113. PubMed PMID:

24532687; PubMed Central PMCID: PMCPMC3950342.

34. Iqbal N, Mei J, Liu J, Skapek SX. miR-34a is essential for p19(Arf)-driven cell cycle arrest. Cell cycle (Georgetown, Tex). 2014;13(5):792-800. Epub 2014/01/10. doi:

10.4161/cc.27725. PubMed PMID: 24401748; PubMed Central PMCID:

PMCPMC3979915.

35. Bommer GT, Gerin I, Feng Y, Kaczorowski AJ, Kuick R, Love RE, et al. p53- mediated activation of miRNA34 candidate tumor-suppressor genes. Current biology :

CB. 2007;17(15):1298-307. Epub 2007/07/28. doi: 10.1016/j.cub.2007.06.068. PubMed

PMID: 17656095.

36. Ghandadi M, Sahebkar A. MicroRNA-34a and its target genes: Key factors in cancer multidrug resistance. Current pharmaceutical design. 2016;22(7):933-9. Epub

2015/12/10. PubMed PMID: 26648462.

37. Sun F, Fu H, Liu Q, Tie Y, Zhu J, Xing R, et al. Downregulation of CCND1 and

CDK6 by miR-34a induces cell cycle arrest. FEBS letters. 2008;582(10):1564-8. Epub

2008/04/15. doi: 10.1016/j.febslet.2008.03.057. PubMed PMID: 18406353.

38. Concepcion CP, Han YC, Mu P, Bonetti C, Yao E, D'Andrea A, et al. Intact p53- dependent responses in miR-34-deficient mice. PLoS Genet. 2012;8(7):e1002797. doi:

10.1371/journal.pgen.1002797. PubMed PMID: 22844244; PubMed Central PMCID:

PMCPMC3406012.

39. Beg MS, Brenner AJ, Sachdev J, Borad M, Kang YK, Stoudemire J, et al. Phase I study of MRX34, a liposomal miR-34a mimic, administered twice weekly in patients

52 with advanced solid tumors. Investigational new drugs. 2016. Epub 2016/12/06. doi:

10.1007/s10637-016-0407-y. PubMed PMID: 27917453.

40. Rupaimoole R, Slack FJ. MicroRNA therapeutics: towards a new era for the management of cancer and other diseases. Nat Rev Drug Discov. 2017;16(3):203-22.

Epub 2017/02/18. doi: 10.1038/nrd.2016.246. PubMed PMID: 28209991.

41. Ram Kumar RM, Boro A, Fuchs B. Involvement and Clinical Aspects of

MicroRNA in Osteosarcoma. Int J Mol Sci. 2016;17(6). doi: 10.3390/ijms17060877.

PubMed PMID: 27271607; PubMed Central PMCID: PMCPMC4926411.

42. Wang Y, Jia LS, Yuan W, Wu Z, Wang HB, Xu T, et al. Low miR-34a and miR-

192 are associated with unfavorable prognosis in patients suffering from osteosarcoma.

Am J Transl Res. 2015;7(1):111-9. Epub 2015/03/11. PubMed PMID: 25755833;

PubMed Central PMCID: PMCPMC4346528.

43. Tang J, Zhao H, Cai H, Wu H. Diagnostic and prognostic potentials of microRNA-27a in osteosarcoma. Biomedicine & pharmacotherapy = Biomedecine & pharmacotherapie. 2015;71:222-6. Epub 2015/05/12. doi: 10.1016/j.biopha.2015.01.025.

PubMed PMID: 25960240.

44. Zhang C, Yao C, Li H, Wang G, He X. Combined elevation of microRNA-196a and microRNA-196b in sera predicts unfavorable prognosis in patients with osteosarcomas. Int J Mol Sci. 2014;15(4):6544-55. Epub 2014/04/22. doi:

10.3390/ijms15046544. PubMed PMID: 24747591; PubMed Central PMCID:

PMCPMC4013646.

45. Yan K, Gao J, Yang T, Ma Q, Qiu X, Fan Q, et al. MicroRNA-34a inhibits the proliferation and metastasis of osteosarcoma cells both in vitro and in vivo. PLoS One.

2012;7(3):e33778. Epub 2012/03/30. doi: 10.1371/journal.pone.0033778. PubMed

PMID: 22457788; PubMed Central PMCID: PMCPMC3310405.

46. Zhao H, Ma B, Wang Y, Han T, Zheng L, Sun C, et al. miR-34a inhibits the metastasis of osteosarcoma cells by repressing the expression of CD44. Oncol Rep.

2013;29(3):1027-36. Epub 2013/01/15. doi: 10.3892/or.2013.2234. PubMed PMID:

23314380.

47. Ouyang L, Liu P, Yang S, Ye S, Xu W, Liu X. A three-plasma miRNA signature serves as novel biomarkers for osteosarcoma. Medical oncology (Northwood, London,

England). 2013;30(1):340. Epub 2012/12/28. doi: 10.1007/s12032-012-0340-7. PubMed

PMID: 23269581.

48. Stephen J. Withrow D, DACVS, DACVIM, David M. Vail D, DACVIM, Rodney

L. Page D, DACVIM. Small Animal Clinical Oncology. 5th ed. St. Louis, Missouri:

Elsevier; 2013.

49. Isakoff MS, Bielack SS, Meltzer P, Gorlick R. Osteosarcoma: Current Treatment and a Collaborative Pathway to Success. J Clin Oncol. 2015;33(27):3029-35. Epub

2015/08/26. doi: 10.1200/jco.2014.59.4895. PubMed PMID: 26304877; PubMed Central

PMCID: PMCPMC4979196.

50. Fenger JM, London CA, Kisseberth WC. Canine osteosarcoma: a naturally occurring disease to inform pediatric oncology. ILAR journal / National Research

Council, Institute of Laboratory Animal Resources. 2014;55(1):69-85. doi:

10.1093/ilar/ilu009. PubMed PMID: 24936031.

51. Paoloni M, Davis S, Lana S, Withrow S, Sangiorgi L, Picci P, et al. Canine tumor cross-species genomics uncovers targets linked to osteosarcoma progression. BMC genomics. 2009;10:625. Epub 2009/12/24. doi: 10.1186/1471-2164-10-625. PubMed

PMID: 20028558; PubMed Central PMCID: PMCPMC2803201.

52. Kirpensteijn J, Kik M, Teske E, Rutteman GR. TP53 gene mutations in canine osteosarcoma. Veterinary surgery : VS. 2008;37(5):454-60. Epub 2008/11/07. doi:

10.1111/j.1532-950X.2008.00407.x. PubMed PMID: 18986312.

53. Levine RA, Fleischli MA. Inactivation of p53 and retinoblastoma family pathways in canine osteosarcoma cell lines. Veterinary pathology. 2000;37(1):54-61.

Epub 2000/01/22. doi: 10.1354/vp.37-1-54. PubMed PMID: 10643981.

54. Grimes JA, Prasad N, Levy S, Cattley R, Lindley S, Boothe HW, et al. A comparison of microRNA expression profiles from splenic hemangiosarcoma, splenic nodular hyperplasia, and normal spleens of dogs. BMC Vet Res. 2016;12(1):272. Epub

2016/12/04. doi: 10.1186/s12917-016-0903-5. PubMed PMID: 27912752; PubMed

Central PMCID: PMCPMC5135805.

55. Kobayashi M, Saito A, Tanaka Y, Michishita M, Kobayashi M, Irimajiri M, et al.

MicroRNA expression profiling in canine prostate cancer. J Vet Med Sci. 2017. Epub

2017/02/28. doi: 10.1292/jvms.16-0279. PubMed PMID: 28239051.

56. Fenger JM, Bear MD, Volinia S, Lin T-Y, Harrington BK, London CA, et al.

Overexpression of miR-9 in mast cells is associated with invasive behavior and spontaneous metastasis. BMC Cancer. 2014;14.

57. Fenger JM, Roberts RD, Iwenofu OH, Bear MD, Zhang X, Couto JI, et al. MiR-9 is overexpressed in spontaneous canine osteosarcoma and promotes a metastatic phenotype including invasion and migration in osteoblasts and osteosarcoma cell lines.

BMC Cancer. 2016;16(1):784. doi: 10.1186/s12885-016-2837-5. PubMed PMID:

27724924; PubMed Central PMCID: PMCPMC5057229.

58. Uhl E, Krimer P, Schliekelman P, Tompkins SM, Suter S. Identification of altered

MicroRNA expression in canine lymphoid cell lines and cases of B- and T-Cell lymphomas. Genes, chromosomes & cancer. 2011;50(11):950-67. Epub 2011/09/13. doi:

10.1002/gcc.20917. PubMed PMID: 21910161.

59. Vinall RL, Kent MS, deVere White RW. Expression of microRNAs in urinary bladder samples obtained from dogs with grossly normal bladders, inflammatory bladder disease, or transitional cell carcinoma. Am J Vet Res. 2012;73(10):1626-33. Epub

2012/09/28. doi: 10.2460/ajvr.73.10.1626. PubMed PMID: 23013190.

60. Albonico F, Mortarino M, Avallone G, Gioia G, Comazzi S, Roccabianca P. The expression ratio of miR-17-5p and miR-155 correlates with grading in canine splenic lymphoma. Vet Immunol Immunopathol. 2013;155(1-2):117-23. Epub 2013/07/23. doi:

10.1016/j.vetimm.2013.06.018. PubMed PMID: 23871213.

61. Noguchi S, Mori T, Hoshino Y, Yamada N, Maruo K, Akao Y. MicroRNAs as tumour suppressors in canine and human melanoma cells and as a prognostic factor in

56 canine melanomas. Veterinary and comparative oncology. 2013;11(2):113-23. Epub

2013/05/04. doi: 10.1002/vco.306. PubMed PMID: 23638671.

62. Gioia G, Mortarino M, Gelain ME, Albonico F, Ciusani E, Forno I, et al.

Immunophenotype-related microRNA expression in canine chronic lymphocytic leukemia. Vet Immunol Immunopathol. 2011;142(3-4):228-35. Epub 2011/06/15. doi:

10.1016/j.vetimm.2011.05.020. PubMed PMID: 21663977.

63. Borresen B, Nielsen LN, Jessen LR, Kristensen AT, Fredholm M, Cirera S.

Circulating let-7g is down-regulated in Bernese Mountain dogs with disseminated histiocytic sarcoma and carcinomas - a prospective study. Veterinary and comparative oncology. 2016. doi: 10.1111/vco.12196. PubMed PMID: 26792388.

64. Fan TM. Animal models of osteosarcoma. Expert review of anticancer therapy.

2010;10(8):1327-38. Epub 2010/08/26. doi: 10.1586/era.10.107. PubMed PMID:

20735317.

65. Withrow SJ, Powers BE, Straw RC, Wilkins RM. Comparative aspects of osteosarcoma. Dog versus man. Clinical orthopaedics and related research.

1991;(270):159-68. Epub 1991/09/01. PubMed PMID: 1884536.

66. Rodriguez CO, Jr. Using canine osteosarcoma as a model to assess efficacy of novel therapies: can old dogs teach us new tricks? Adv Exp Med Biol. 2014;804:237-56.

Epub 2014/06/14. doi: 10.1007/978-3-319-04843-7_13. PubMed PMID: 24924178.

67. Bartel DP, Chen CZ. Micromanagers of gene expression: the potentially widespread influence of metazoan microRNAs. Nature reviews Genetics. 2004;5(5):396-

400. Epub 2004/05/15. doi: 10.1038/nrg1328. PubMed PMID: 15143321.

68. Ambros V. The functions of animal microRNAs. Nature. 2004;431(7006):350-5.

Epub 2004/09/17. doi: 10.1038/nature02871. PubMed PMID: 15372042.

69. Iorio MV, Croce CM. MicroRNA dysregulation in cancer: diagnostics, monitoring and therapeutics. A comprehensive review. EMBO molecular medicine.

2012;4(3):143-59. Epub 2012/02/22. doi: 10.1002/emmm.201100209. PubMed PMID:

22351564; PubMed Central PMCID: PMCPMC3376845.

70. Calin GA, Dumitru CD, Shimizu M, Bichi R, Zupo S, Noch E, et al. Frequent deletions and down-regulation of micro- RNA genes miR15 and miR16 at 13q14 in chronic lymphocytic leukemia. Proceedings of the National Academy of Sciences of the

United States of America. 2002;99(24):15524-9. Epub 2002/11/16. doi:

10.1073/pnas.242606799. PubMed PMID: 12434020; PubMed Central PMCID:

PMCPMC137750.

71. Garzon R, Fabbri M, Cimmino A, Calin GA, Croce CM. MicroRNA expression and function in cancer. Trends in molecular medicine. 2006;12(12):580-7. Epub

2006/10/31. doi: 10.1016/j.molmed.2006.10.006. PubMed PMID: 17071139.

72. Fabbri M, Garzon R, Cimmino A, Liu Z, Zanesi N, Callegari E, et al. MicroRNA-

29 family reverts aberrant methylation in lung cancer by targeting DNA methyltransferases 3A and 3B. Proceedings of the National Academy of Sciences of the

United States of America. 2007;104(40):15805-10. Epub 2007/09/25. doi:

10.1073/pnas.0707628104. PubMed PMID: 17890317; PubMed Central PMCID:

PMCPMC2000384.

73. Cimmino A, Calin GA, Fabbri M, Iorio MV, Ferracin M, Shimizu M, et al. miR-

15 and miR-16 induce apoptosis by targeting BCL2. Proceedings of the National

Academy of Sciences of the United States of America. 2005;102(39):13944-9. Epub

2005/09/17. doi: 10.1073/pnas.0506654102. PubMed PMID: 16166262; PubMed Central

PMCID: PMCPMC1236577.

74. Garzon R, Heaphy CE, Havelange V, Fabbri M, Volinia S, Tsao T, et al.

MicroRNA 29b functions in acute myeloid leukemia. Blood. 2009;114(26):5331-41.

Epub 2009/10/24. doi: 10.1182/blood-2009-03-211938. PubMed PMID: 19850741;

PubMed Central PMCID: PMCPMC2796138.

75. Li QJ, Chau J, Ebert PJ, Sylvester G, Min H, Liu G, et al. miR-181a is an intrinsic modulator of T cell sensitivity and selection. Cell. 2007;129(1):147-61. Epub 2007/03/27. doi: 10.1016/j.cell.2007.03.008. PubMed PMID: 17382377.

76. Slack F. let-7 microRNA reduces tumor growth. Cell cycle (Georgetown, Tex).

2009;8(12):1823. Epub 2009/04/21. doi: 10.4161/cc.8.12.8639. PubMed PMID:

19377282.

77. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real- time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 2001;25(4):402-8.

Epub 2002/02/16. doi: 10.1006/meth.2001.1262. PubMed PMID: 11846609.

78. Lin TY, Fenger J, Murahari S, Bear MD, Kulp SK, Wang D, et al. AR-42, a novel

HDAC inhibitor, exhibits biologic activity against malignant mast cell lines via down- regulation of constitutively activated Kit. Blood. 2010;115(21):4217-25. Epub

2010/03/18. doi: 10.1182/blood-2009-07-231985. PubMed PMID: 20233974; PubMed

Central PMCID: PMCPMC3398750.

79. Bouchie A. First microRNA mimic enters clinic. Nature biotechnology.

2013;31(7):577. Epub 2013/07/11. doi: 10.1038/nbt0713-577. PubMed PMID: 23839128.

80. Lindow M, Kauppinen S. Discovering the first microRNA-targeted drug. The

Journal of cell biology. 2012;199(3):407-12. Epub 2012/10/31. doi:

10.1083/jcb.201208082. PubMed PMID: 23109665; PubMed Central PMCID:

PMCPMC3483128.

81. Zhao Y, Tu MJ, Yu YF, Wang WP, Chen QX, Qiu JX, et al. Combination therapy with bioengineered miR-34a prodrug and doxorubicin synergistically suppresses osteosarcoma growth. Biochem Pharmacol. 2015;98(4):602-13. doi:

10.1016/j.bcp.2015.10.015. PubMed PMID: 26518752.

82. Zhao Y, Tu MJ, Wang WP, Qiu JX, Yu AX, Yu AM. Genetically engineered pre- microRNA-34a prodrug suppresses orthotopic osteosarcoma xenograft tumor growth via the induction of apoptosis and cell cycle arrest. Sci Rep. 2016;6:26611. doi:

10.1038/srep26611. PubMed PMID: 27216562; PubMed Central PMCID:

PMCPMC4877571.

83. Gurrapu S, Tamagnone L. Transmembrane semaphorins: Multimodal signaling cues in development and cancer. Cell adhesion & migration. 2016;10(6):675-91. Epub

2016/06/14. doi: 10.1080/19336918.2016.1197479. PubMed PMID: 27295627.

84. Tamagnone L. Emerging role of semaphorins as major regulatory signals and potential therapeutic targets in cancer. Cancer cell. 2012;22(2):145-52. Epub 2012/08/18. doi: 10.1016/j.ccr.2012.06.031. PubMed PMID: 22897846.

85. Moriarity BS, Otto GM, Rahrmann EP, Rathe SK, Wolf NK, Weg MT, et al. A

Sleeping Beauty forward genetic screen identifies new genes and pathways driving osteosarcoma development and metastasis. Nature genetics. 2015;47(6):615-24. Epub

2015/05/12. doi: 10.1038/ng.3293. PubMed PMID: 25961939; PubMed Central PMCID:

PMCPMC4767150.

86. Christensen CR, Klingelhofer J, Tarabykina S, Hulgaard EF, Kramerov D,

Lukanidin E. Transcription of a novel mouse semaphorin gene, M-semaH, correlates with the metastatic ability of mouse tumor cell lines. Cancer research. 1998;58(6):1238-44.

Epub 1998/03/27. PubMed PMID: 9515811.

87. Christensen C, Ambartsumian N, Gilestro G, Thomsen B, Comoglio P,

Tamagnone L, et al. Proteolytic processing converts the repelling signal Sema3E into an inducer of invasive growth and lung metastasis. Cancer research. 2005;65(14):6167-77.

Epub 2005/07/19. doi: 10.1158/0008-5472.can-04-4309. PubMed PMID: 16024618.

88. Casazza A, Finisguerra V, Capparuccia L, Camperi A, Swiercz JM, Rizzolio S, et al. Sema3E-Plexin D1 signaling drives human cancer cell invasiveness and metastatic spreading in mice. J Clin Invest. 2010;120(8):2684-98. Epub 2010/07/29. doi:

10.1172/jci42118. PubMed PMID: 20664171; PubMed Central PMCID:

PMCPMC2912191.

89. Yong LK, Lai S, Liang Z, Poteet E, Chen F, van Buren G, et al. Overexpression of Semaphorin-3E enhances pancreatic cancer cell growth and associates with poor patient survival. Oncotarget. 2016;7(52):87431-48. Epub 2016/12/03. doi:

10.18632/oncotarget.13704. PubMed PMID: 27911862.

90. Luchino J, Hocine M, Amoureux MC, Gibert B, Bernet A, Royet A, et al.

Semaphorin 3E suppresses tumor cell death triggered by the plexin D1 dependence receptor in metastatic breast cancers. Cancer cell. 2013;24(5):673-85. Epub 2013/10/22. doi: 10.1016/j.ccr.2013.09.010. PubMed PMID: 24139859.

91. Zhang L, Zhang L, Xia X, He S, He H, Zhao W. Kruppel-like factor 4 promotes human osteosarcoma growth and metastasis via regulating CRYAB expression.

Oncotarget. 2016;7(21):30990-1000. Epub 2016/04/23. doi: 10.18632/oncotarget.8824.

PubMed PMID: 27105535; PubMed Central PMCID: PMCPMC5058733.

92. Martins-Neves SR, Corver WE, Paiva-Oliveira DI, van den Akker BE, Briaire-de-

Bruijn IH, Bovee JV, et al. Osteosarcoma Stem Cells Have Active Wnt/beta-catenin and

Overexpress SOX2 and KLF4. Journal of cellular physiology. 2016;231(4):876-86. Epub

2015/09/04. doi: 10.1002/jcp.25179. PubMed PMID: 26332365.

93. Zhang D, Zhou J, Dong M. Dysregulation of microRNA-34a expression in colorectal cancer inhibits the phosphorylation of FAK via VEGF. Digestive diseases and sciences. 2014;59(5):958-67. Epub 2013/12/29. doi: 10.1007/s10620-013-2983-4.

PubMed PMID: 24370784.

94. Kaya M, Wada T, Akatsuka T, Kawaguchi S, Nagoya S, Shindoh M, et al.

Vascular endothelial growth factor expression in untreated osteosarcoma is predictive of

62 pulmonary metastasis and poor prognosis. Clinical cancer research : an official journal of the American Association for Cancer Research. 2000;6(2):572-7. Epub 2000/02/26.

PubMed PMID: 10690541.

95. Wergin MC, Kaser-Hotz B. Plasma vascular endothelial growth factor (VEGF) measured in seventy dogs with spontaneously occurring tumours. In vivo (Athens,

Greece). 2004;18(1):15-9. Epub 2004/03/12. PubMed PMID: 15011746.

96. Cam M, Gardner HL, Roberts RD, Fenger JM, Guttridge DC, London CA, et al.

DeltaNp63 mediates cellular survival and metastasis in canine osteosarcoma. Oncotarget.

2016;7(30):48533-46. Epub 2016/07/09. doi: 10.18632/oncotarget.10406. PubMed

PMID: 27391430; PubMed Central PMCID: PMCPMC5217036.

Appendix A: Supplemental Tables

Table 2: Differentially expressed transcripts (fold change ≥ 1.5 or ≤ -1.5) in canine OSA8 cells overexpressing miR-34a

Gene Name mir34a vs EV Fold Change mir34a vs EV P-value CFA-MIR-34A 150.15 1.43E-202 PICALM -14.94 7.06E-48 ATAD1 -12.68 9.57E-42 ARMCX2 -7.20 3.99E-39 PEG10 -11.71 8.71E-37 ENSCAFG00000019046 -10.53 4.43E-34 ENSCAFG00000014436 -7.26 2.90E-33 ENSCAFG00000020060 -5.45 3.62E-26 P3H3 -6.09 8.01E-25 REEP2 -6.05 6.10E-22 ZNF280B -6.25 2.62E-21 LUZP4 -6.14 1.03E-19 TFPI2 -4.40 1.59E-19 PTEN -5.98 2.23E-19 MAGEB16 -5.91 4.53E-19 COL9A2 -4.87 1.44E-18 SYT4 -5.60 7.02E-18 TAGLN3 -3.41 9.39E-18 SGCE -5.03 7.65E-16 MINPP1 -4.06 8.29E-16 PHTF2 -4.94 1.54E-15 CREBZF -4.88 2.39E-15 SLC5A6 -2.09 3.51E-14 ENSCAFG00000031663 -4.32 2.97E-13 FAS -4.28 3.79E-13 ENSCAFG00000014034 -3.26 1.71E-11 ENSCAFG00000004384 -3.70 5.40E-11 LTBP2 -3.67 7.10E-11 HSPB6 -3.50 7.63E-11 GNB3 -3.31 3.05E-10 SI -3.47 3.69E-10 TEX13A -3.31 4.30E-10 RSBN1L -3.43 5.42E-10 GABRA3 -3.10 6.81E-10 PDPN -3.16 7.21E-10 MFAP5 -2.63 9.87E-10 ENSCAFG00000019160 -3.37 1.13E-09 GAP43 -3.21 5.02E-09 PLCB2 -3.20 6.30E-09 SLC24A3 -2.57 7.59E-09 TINAGL1 2.70 1.38E-08 ENSCAFG00000029992 -2.93 1.75E-08 ENSCAFG00000014043 -3.03 1.93E-08 Continued 65

Table 2 continued

Gene Name mir34a vs EV Fold change mir34a vs EV P-value CNTN2 2.97 2.17E-08 TMEM60 -2.97 2.87E-08 EFNB3 -2.88 3.01E-08 GPRIN1 2.78 4.05E-08 TCP11L1 -2.44 7.74E-08 PRTFDC1 -2.36 8.41E-08 CASP14 -2.89 1.11E-07 PRSS23 -2.77 1.65E-07 MAGEB18 -2.83 1.69E-07 LOXL4 -2.77 3.74E-07 ADCY2 -2.49 3.91E-07 ENSCAFG00000024834 -2.23 4.34E-07 ANGPTL4 2.33 5.72E-07 CYP2S1 -2.69 5.91E-07 C11ORF24 -2.01 7.05E-07 GRIK5 -2.64 7.31E-07 ENSCAFG00000005922 -1.89 8.87E-07 PAPSS2 -2.58 1.26E-06 NIPAL2 -2.07 1.32E-06 SLC2A10 -1.87 1.34E-06 PCDHB2 -2.01 1.48E-06 CUBN -2.33 1.78E-06 FAM111A -2.15 1.79E-06 LHFPL3 -2.26 1.94E-06 FIGF -2.59 1.94E-06 GPR85 2.50 2.33E-06 ZNF793 2.36 3.26E-06 CTPS2 -2.10 4.04E-06 TMEM126B -2.27 4.37E-06 ITGA4 -2.39 5.70E-06 BRINP2 -2.46 5.76E-06 EYA4 -2.12 6.49E-06 LIPA -2.46 6.91E-06 ENSCAFG00000015625 -2.44 7.17E-06 CHAC1 -2.44 8.22E-06 ENSCAFG00000032762 -1.90 8.28E-06 MAP3K7CL -2.42 9.60E-06 ENSCAFG00000018844 -2.42 1.00E-05 RIPK3 -2.22 1.09E-05 SEC14L4 -2.25 1.30E-05 PRAME -2.33 1.39E-05 AMER1 -2.27 1.49E-05 PCDHB15 -1.95 1.56E-05 GATA3 -1.95 1.61E-05 PDE4B 2.15 1.87E-05 C11ORF73 -2.33 2.19E-05 Continued 66

Table 2 continued

Gene Name mir34a vs EV Fold change mir34a vs EV P-value P3H2 -2.05 2.73E-05 RNF167 -1.99 2.83E-05 STK38L -1.75 3.00E-05 DNM3 -2.24 3.03E-05 TNS4 2.30 3.21E-05 FAM69B -2.18 3.68E-05 SLC13A5 -2.06 3.97E-05 ENSCAFG00000029405 -2.21 4.07E-05 STRA6 1.79 4.23E-05 DNAJC1 -1.76 4.51E-05 DEF6 -1.95 4.75E-05 ZNF385B -2.17 4.90E-05 ASAH1 -1.90 5.28E-05 TMEM200A -1.98 6.54E-05 HHIP -2.20 6.55E-05 GUCY1A3 -1.92 6.92E-05 SELP 2.18 7.02E-05 ADAMTS14 -2.13 7.74E-05 MEG3_2 -2.20 8.62E-05 CRTAC1 -2.19 8.75E-05 FAM3C 1.58 9.14E-05 THRB -2.18 9.69E-05 ZNF454 1.87 0.00010 ENSCAFG00000016517 -1.71 0.00010 PTPRB -2.11 0.00011 ENSCAFG00000024687 -1.74 0.00011 CHST2 -1.78 0.00011 COA1 -1.61 0.00012 COLEC12 1.88 0.00013 ENSCAFG00000031825 1.73 0.00013 DPT -1.98 0.00014 INHBA -2.09 0.00014 AKAP6 -2.13 0.00014 CREB3L1 -1.75 0.00014

Table 3: Down-regulated transcripts (fold change ≤ -1.4) containing consensus binding sites for miR-34a

Gene Name mir34a vs EV Fold Change PEG10 -11.71 ZNF280B -6.25 LUZP4 -6.14 SYT4 -5.60 LTBP2 -3.67 HSPB6 -3.50 GABRA3 -3.10 EFNB3 -2.88 AMER1 -2.27 DNM3 -2.24 THRB -2.18 PSD4 -2.15 AKAP6 -2.13 PTPRB -2.11 PTPRR -2.09 ASTN1 -2.07 TMEM255A -2.00 SLC26A7 -1.97 PADI2 -1.95 GATA3 -1.95 GUCY1A3 -1.92 KCNH7 -1.92 TRDMT1 -1.81 CA12 -1.79 CREB5 -1.79 DAAM1 -1.79 MCIDAS -1.80 ARHGAP26 -1.76 IL7R -1.76 CREB3L1 -1.75 STK38L -1.75 NPEPL1 -1.74 OAS3 -1.74 SLC1A4 -1.75 ANKRD1 -1.73 DAB1 -1.73 PFKFB3 -1.73 CRY2 -1.71 RELN -1.71 ALDH1L2 -1.71 GRB10 -1.70 SP7 -1.69 PEX11G -1.68 Continued 68

Table 3 continued

Gene Name mir34a vs EV Fold Change PPP2R3A -1.65 RAI14 -1.66 SPON1 -1.63 HNF4G -1.63 IGF2BP2 -1.62 KLHDC8B -1.62 SEMA3E -1.61 B3GAT3 -1.60 MRVI1 -1.60 ST6GAL1 -1.60 CMPK2 -1.59 FAM110B -1.59 GBP5 -1.59 HTR2C -1.60 SAMD4A -1.60 STK17A -1.59 TGFB3 -1.59 KLF12 -1.58 KLF7 -1.59 ZNF831 -1.58 UROS -1.57 DCBLD1 -1.55 DTNA -1.55 TRIM9 -1.56 TWIST2 -1.56 KIAA1217 -1.55 MICAL2 -1.55 PHF21A -1.55 ZEB1 -1.55 AKAP7 -1.54 GALNT15 -1.54 METTL4 -1.53 VEGFA -1.54 ADRA1D -1.53 DDX26B -1.53 DUSP4 -1.53 IL1RN -1.53 MMP25 -1.53 PAX7 -1.52 SLCO3A1 -1.52 ABCA1 -1.52 ACTL8 -1.51 ARHGAP31 -1.52 ARNTL2 -1.52 HIPK2 -1.52 PCLO -1.51 Continued 69

Table 3 continued

Gene Name mir34a vs EV Fold Change PRDM6 -1.51 UBE2QL1 -1.52 ARHGAP36 -1.51 ERG -1.51 HLF -1.50 LMO4 -1.50 QKI -1.51 RGS3 -1.51 ZC3H6 -1.50 FLVCR2 -1.49 EMP1 -1.49 FNDC3B -1.48 MED19 -1.49 RNF152 -1.48 TLDC1 -1.49 XIRP1 -1.48 STK32A -1.47 FZD5 -1.47 HEG1 -1.46 ARID5A -1.46 CYLD -1.45 MYRIP -1.45 TGM2 -1.45 TRMT61A -1.45 CLDN1 -1.44 DDX58 -1.44 GCH1 -1.44 RNF41 -1.45 ALX3 -1.43 CRABP2 -1.44 CYR61 -1.44 FAM76A -1.43 KLF4 -1.44 PTPRD -1.44 RAB7B -1.44 BACH2 -1.42 CPA4 -1.43 FRMD4A -1.42 JAG1 -1.42 NIN -1.43 OLIG2 -1.43 SYT11 -1.43 GSTM3 -1.41 HAVCR2 -1.42 MYLK -1.41 NHS -1.41 Continued 70

Table 3 continued

Gene Name mir34a vs EV Fold Change OSGIN2 -1.41 PPAPDC1B -1.41 R3HCC1L -1.41 SLC38A1 -1.42 SLITRK4 -1.42 TANC2 -1.41 ACSL4 -1.40 AKTIP -1.40 KCNA3 -1.41 PKIA -1.41 RNF144B -1.41 SYT13 -1.40 TNFSF15 -1.41 TRPM3 -1.40 XYLT1 -1.40

Table 4: ToppGene Suite gene ontology analysis of putative miR-34a targets

ID Name p-value Hit in Query List GO:0045595 regulation of cell 4.21E-07 AKAP6,CYLD,FNDC3B,RELN,ABCA1,ZEB1,H differentiation TR2C,DAB1,GATA3,TWIST2,ST6GAL1,LMO4, KLF4,CRABP2,TGFB3,XYLT1,THRB,PTPRD,C YR61,JAG1,ANKRD1,DNM3,RAB7B,SYT4,RN F41,PAX7,QKI,OLIG2,PRDM6,IL7R,SEMA3E,S P7,VEGFA,NIN GO:0045597 positive 2.58E-06 AKAP6,CYLD,FNDC3B,RELN,ZEB1,HTR2C,D regulation of cell AB1,GATA3,CRABP2,TGFB3,THRB,PTPRD,C differentiation YR61,JAG1,ANKRD1,RAB7B,SYT4,QKI,OLIG 2,IL7R,SP7,VEGFA,NIN GO:0051240 positive 3.43E-06 AKAP6,CYLD,RELN,ZEB1,HTR2C,ERG,DDX5 regulation of 8,SLITRK4,DAB1,GATA3,HIPK2,CRABP2,TGF multicellular B3,HAVCR2,ADRA1D,HEG1,THRB,PTPRD,CY organismal R61,GBP5,FZD5,JAG1,ANKRD1,RAB7B,SYT4, process QKI,OLIG2,TNFSF15,IL7R,VEGFA,NIN GO:0030182 neuron 5.36E-06 RELN,ZEB1,SLITRK4,DAB1,GATA3,HIPK2,Z differentiation NF280B,LMO4,KLF4,CRABP2,XYLT1,ACSL4, PPP2R3A,EFNB3,PTPRD,KLF7,PTPRR,FZD5,J AG1,ANKRD1,DNM3,SYT4,PAX7,QKI,OLIG2, SEMA3E,VEGFA,NIN GO:0048699 generation of 9.99E-06 RELN,ZEB1,SLITRK4,DAB1,GATA3,HIPK2,Z neurons NF280B,LMO4,KLF4,CRABP2,XYLT1,ACSL4, PPP2R3A,EFNB3,PTPRD,KLF7,PTPRR,FZD5,J AG1,ANKRD1,DNM3,ASTN1,SYT4,PAX7,QKI, OLIG2,SEMA3E,VEGFA,NIN GO:0022008 neurogenesis 1.21E-05 RELN,ZEB1,SLITRK4,DAB1,GATA3,HIPK2,Z NF280B,LMO4,KLF4,CRABP2,XYLT1,ACSL4, PPP2R3A,EFNB3,PTPRD,KLF7,PTPRR,FZD5,J AG1,ANKRD1,DNM3,ASTN1,SYT4,PAX7,QKI, OLIG2,PRDM6,SEMA3E,VEGFA,NIN GO:2000026 regulation of 1.29E-05 AKAP6,CYLD,RELN,ZEB1,SLITRK4,DAB1,GA multicellular TA3,HIPK2,TWIST2,KLF4,CRABP2,TGFB3,XY organismal LT1,HAVCR2,PPP2R3A,THRB,PTPRD,CYR61, development FZD5,JAG1,ANKRD1,DNM3,RAB7B,SYT4,RN F41,DAAM1,QKI,IL1RN,OLIG2,IL7R,SEMA3E, VEGFA,NIN GO:0006357 regulation of 1.66E-05 KLF12,ZEB1,ERG,DDX58,MCIDAS,BACH2,G transcription from ATA3,HIPK2,TWIST2,ZNF280B,LMO4,KLF4,H RNA polymerase NF4G,ARNTL2,CREB3L1,ARID5A,CREB5,PHF II promoter 21A,CRY2,TGFB3,MED19,THRB,KLF7,CYR61, MICAL2,FZD5,JAG1,ANKRD1,PKIA,PAX7,OLI G2,SP7,VEGFA GO:0045664 regulation of 2.00E-05 RELN,ZEB1,DAB1,GATA3,KLF4,CRABP2,XY neuron LT1,PTPRD,JAG1,ANKRD1,DNM3,SYT4,QKI, differentiation OLIG2,SEMA3E,VEGFA,NIN Continued

Table 4 continued

ID Name p-value Hit in Query List GO:0045893 positive 2.77E-05 KLF12,RELN,ZEB1,DDX58,MCIDAS,GATA3,H regulation of IPK2,LMO4,KLF4,HNF4G,ARNTL2,CREB3L1, transcription, ARID5A,CREB5,TGFB3,THRB,KLF7,CYR61,M DNA-templated ICAL2,FZD5,JAG1,ANKRD1,RAB7B,RNF41,P AX7,PADI2,SP7,VEGFA GO:1903508 positive 2.77E-05 KLF12,RELN,ZEB1,DDX58,MCIDAS,GATA3,H regulation of IPK2,LMO4,KLF4,HNF4G,ARNTL2,CREB3L1, nucleic acid- ARID5A,CREB5,TGFB3,THRB,KLF7,CYR61,M templated ICAL2,FZD5,JAG1,ANKRD1,RAB7B,RNF41,P transcription AX7,PADI2,SP7,VEGFA GO:1902680 positive 3.42E-05 KLF12,RELN,ZEB1,DDX58,MCIDAS,GATA3,H regulation of IPK2,LMO4,KLF4,HNF4G,ARNTL2,CREB3L1, RNA biosynthetic ARID5A,CREB5,TGFB3,THRB,KLF7,CYR61,M process ICAL2,FZD5,JAG1,ANKRD1,RAB7B,RNF41,P AX7,PADI2,SP7,VEGFA GO:0003007 heart 3.63E-05 XIRP1,ERG,GATA3,TGFB3,HEG1,CYR61,MIC morphogenesis AL2,JAG1,ANKRD1,VEGFA GO:0031077 post-embryonic 4.07E-05 KLF4,FZD5,VEGFA camera-type eye development GO:0010975 regulation of 4.20E-05 RELN,DAB1,GATA3,KLF4,CRABP2,XYLT1,PT neuron projection PRD,ANKRD1,DNM3,SYT4,QKI,SEMA3E,VEG development FA,NIN GO:0031328 positive 4.44E-05 KLF12,RELN,ABCA1,ZEB1,HTR2C,DDX58,M regulation of CIDAS,GATA3,HIPK2,LMO4,KLF4,HNF4G,AR cellular NTL2,CREB3L1,ARID5A,CREB5,TGFB3,THRB biosynthetic ,KLF7,CYR61,GUCY1A3,MICAL2,FZD5,JAG1, process ANKRD1,RAB7B,RNF41,PAX7,PADI2,SAMD4 A,SP7,VEGFA GO:0051094 positive 4.86E-05 AKAP6,CYLD,FNDC3B,RELN,ZEB1,HTR2C,S regulation of LITRK4,DAB1,GATA3,HIPK2,CRABP2,TGFB3, developmental THRB,PTPRD,CYR61,JAG1,ANKRD1,RAB7B,S process YT4,QKI,OLIG2,IL7R,SP7,VEGFA,NIN GO:0051254 positive 5.81E-05 KLF12,RELN,ZEB1,DDX58,MCIDAS,GATA3,H regulation of IPK2,LMO4,KLF4,HNF4G,ARNTL2,CREB3L1, RNA metabolic ARID5A,CREB5,TGFB3,THRB,KLF7,CYR61,M process ICAL2,FZD5,JAG1,ANKRD1,RAB7B,RNF41,P AX7,PADI2,SP7,VEGFA GO:0045944 positive 5.91E-05 KLF12,ZEB1,DDX58,MCIDAS,GATA3,HIPK2, regulation of LMO4,KLF4,ARNTL2,CREB3L1,ARID5A,CRE transcription from B5,TGFB3,THRB,CYR61,MICAL2,FZD5,JAG1, RNA polymerase ANKRD1,PAX7,SP7,VEGFA II promoter Continued

Table 4 continued

ID Name p-value Hit in Query List GO:0009891 positive 6.48E-05 KLF12,RELN,ABCA1,ZEB1,HTR2C,DDX58,M regulation of CIDAS,GATA3,HIPK2,LMO4,KLF4,HNF4G,AR biosynthetic NTL2,CREB3L1,ARID5A,CREB5,TGFB3,THRB process ,KLF7,CYR61,GUCY1A3,MICAL2,FZD5,JAG1, ANKRD1,RAB7B,RNF41,PAX7,PADI2,SAMD4 A,SP7,VEGFA GO:0031344 regulation of cell 7.55E-05 CYLD,RELN,DAB1,GATA3,KLF4,CRABP2,T projection GFB3,XYLT1,PTPRD,ANKRD1,DNM3,SYT4, organization QKI,SEMA3E,VEGFA,NIN GO:0010628 positive 8.02E-05 KLF12,RELN,ZEB1,DDX58,MCIDAS,GATA3,H regulation of IPK2,LMO4,KLF4,HNF4G,ARNTL2,CREB3L1, gene expression ARID5A,CREB5,TGFB3,THRB,KLF7,CYR61,M ICAL2,FZD5,JAG1,ANKRD1,RAB7B,RNF41,P AX7,QKI,PADI2,IL7R,SAMD4A,SP7,VEGFA GO:0009790 embryo 1.01E-04 ALX3,TANC2,ZEB1,DUSP4,GATA3,HIPK2,TW development IST2,LMO4,KLF4,CRABP2,TGFB3,ACSL4,PPP 2R3A,HEG1,CYR61,PTPRR,MICAL2,FZD5,KIA A1217,PAX7,IL1RN,VEGFA GO:0045935 positive 1.10E-04 KLF12,RELN,ABCA1,ZEB1,DDX58,MCIDAS,G regulation of ATA3,HIPK2,LMO4,KLF4,HNF4G,ARNTL2,CR nucleobase- EB3L1,ARID5A,CREB5,TGFB3,THRB,KLF7,C containing YR61,GUCY1A3,MICAL2,FZD5,JAG1,ANKRD compound 1,RAB7B,RNF41,PAX7,PADI2,SP7,VEGFA metabolic process GO:0060284 regulation of cell 1.16E-04 AKAP6,RELN,ZEB1,DAB1,GATA3,ST6GAL1, development KLF4,CRABP2,TGFB3,XYLT1,PTPRD,JAG1,A NKRD1,DNM3,SYT4,QKI,OLIG2,SEMA3E,VE GFA,NIN GO:0048729 tissue 1.30E-04 MYLK,GATA3,HIPK2,LMO4,KLF4,TGFB3,TG morphogenesis M2,HEG1,CYR61,MICAL2,FZD5,JAG1,ANKRD 1,PAX7,DAAM1,SEMA3E,VEGFA GO:0003179 heart valve 1.34E-04 ERG,GATA3,CYR61,JAG1 morphogenesis GO:0051173 positive 1.35E-04 KLF12,RELN,ABCA1,ZEB1,DDX58,MCIDAS,G regulation of ATA3,HIPK2,LMO4,KLF4,HNF4G,ARNTL2,CR nitrogen EB3L1,ARID5A,CREB5,TGFB3,THRB,KLF7,C compound YR61,GUCY1A3,MICAL2,FZD5,JAG1,ANKRD metabolic process 1,RAB7B,RNF41,PAX7,PADI2,SAMD4A,SP7,V EGFA GO:0000904 cell 1.35E-04 FNDC3B,RELN,ERG,SLITRK4,DAB1,GATA3,S morphogenesis T6GAL1,ZNF280B,CRABP2,TGFB3,HEG1,EFN involved in B3,PTPRD,KLF7,DNM3,SEMA3E,VEGFA,NIN differentiation GO:0010557 positive 1.56E-04 KLF12,RELN,ZEB1,DDX58,MCIDAS,GATA3,H regulation of IPK2,LMO4,KLF4,HNF4G,ARNTL2,CREB3L1, macromolecule ARID5A,CREB5,TGFB3,THRB,KLF7,CYR61,M biosynthetic ICAL2,FZD5,JAG1,ANKRD1,RAB7B,RNF41,P process AX7,PADI2,SAMD4A,SP7,VEGFA Continued 74

Table 4 continued

ID Name p-value Hit in Query List GO:0009792 embryo 1.57E-04 ALX3,TANC2,ZEB1,GATA3,TWIST2,LMO4,T development GFB3,ACSL4,PPP2R3A,HEG1,CYR61,PTPRR,F ending in birth or ZD5,KIAA1217,PAX7,VEGFA egg hatching GO:1902531 regulation of 1.62E-04 AKAP6,CYLD,RELN,ABCA1,RNF152,HTR2C, intracellular ARHGAP36,ARHGAP26,DUSP4,DAB1,GATA3, signal HIPK2,KLF4,RGS3,CREB3L1,TGFB3,HAVCR2, transduction TGM2,CYR61,PTPRR,ARHGAP31,FZD5,ANKR D1,RNF41,IL1RN,TNFSF15,AKAP7,PSD4,VEG FA GO:0048585 negative 1.69E-04 CYLD,AMER1,RNF152,PEG10,DUSP4,DAB1,G regulation of ATA3,HIPK2,GRB10,ST6GAL1,KLF4,RGS3,CR response to EB3L1,CRY2,TGFB3,XYLT1,HAVCR2,PPP2R3 stimulus A,PTPRR,RAB7B,RNF41,IL1RN,PADI2,IL7R,S EMA3E,VEGFA GO:0034616 response to 1.71E-04 ABCA1,KLF4,TGFB3 laminar fluid shear stress GO:0050767 regulation of 1.82E-04 RELN,ZEB1,DAB1,GATA3,KLF4,CRABP2,XY neurogenesis LT1,PTPRD,JAG1,ANKRD1,DNM3,SYT4,QKI, OLIG2,SEMA3E,VEGFA,NIN GO:0097477 lateral motor 1.90E-04 RELN,DAB1 column neuron migration GO:0040012 regulation of 1.96E-04 RELN,HTR2C,MYLK,DDX58,GATA3,ST6GA locomotion L1,LMO4,KLF4,PPP2R3A,CYR61,PTPRR,JA G1,RNF41,PAX7,IL1RN,PADI2,SEMA3E,VE GFA GO:0031175 neuron projection 2.20E-04 RELN,SLITRK4,DAB1,GATA3,ZNF280B,KLF4, development CRABP2,XYLT1,ACSL4,EFNB3,PTPRD,KLF7, ANKRD1,DNM3,SYT4,QKI,SEMA3E,VEGFA,N IN GO:0009887 organ 2.36E-04 ALX3,XIRP1,ZEB1,ERG,MYLK,GATA3,HIPK2 morphogenesis ,TWIST2,TGFB3,TGM2,PPP2R3A,HEG1,THRB, CYR61,MICAL2,FZD5,JAG1,ANKRD1,PAX7,D AAM1,VEGFA GO:0051960 regulation of 2.42E-04 RELN,ZEB1,SLITRK4,DAB1,GATA3,KLF4,CR nervous system ABP2,XYLT1,PTPRD,JAG1,ANKRD1,DNM3,S development YT4,QKI,OLIG2,SEMA3E,VEGFA,NIN GO:0061564 axon 2.55E-04 RELN,SLITRK4,DAB1,GATA3,ZNF280B,KLF4, development CRABP2,XYLT1,EFNB3,KLF7,SEMA3E,VEGF A,NIN GO:0030278 regulation of 2.57E-04 FNDC3B,TWIST2,CREB3L1,TGFB3,THRB,CY ossification R61,JAG1,VEGFA GO:0003170 heart valve 3.17E-04 ERG,GATA3,CYR61,JAG1 development GO:0097475 motor neuron 3.78E-04 RELN,DAB1 migration Continued 75

Table 4 continued

ID Name p-value Hit in Query List GO:0097476 spinal cord motor 3.78E-04 RELN,DAB1 neuron migration GO:0030030 cell projection 3.90E-04 CYLD,RELN,MYLK,MCIDAS,SLITRK4,DAB organization 1,GATA3,ZNF280B,KLF4,CRABP2,TGFB3,X YLT1,ACSL4,EFNB3,PTPRD,KLF7,ANKRD1, DNM3,SYT4,EMP1,QKI,SEMA3E,VEGFA,NI N GO:0032989 cellular 4.19E-04 CYLD,FNDC3B,RELN,XIRP1,ERG,MCIDAS,SL component ITRK4,DAB1,GATA3,ST6GAL1,ZNF280B,CRA morphogenesis BP2,TGFB3,HEG1,EFNB3,PTPRD,KLF7,ANKR D1,DNM3,SYT4,IL7R,SEMA3E,VEGFA,NIN GO:0043009 chordate 4.38E-04 ALX3,TANC2,ZEB1,GATA3,TWIST2,LMO4,T embryonic GFB3,PPP2R3A,HEG1,CYR61,PTPRR,FZD5,KI development AA1217,PAX7,VEGFA GO:2000113 negative 5.22E-04 CYLD,KLF12,ZEB1,BACH2,GATA3,HIPK2,GR regulation of B10,TWIST2,ZNF280B,KLF4,ARID5A,PHF21A, cellular CRY2,TGFB3,HAVCR2,IGF2BP2,THRB,ANKR macromolecule D1,PKIA,QKI,OLIG2,PRDM6,SAMD4A,VEGFA biosynthetic process GO:0030155 regulation of cell 5.32E-04 CYLD,ZEB1,DAB1,GATA3,ST6GAL1,KLF4,H adhesion AVCR2,TGM2,EFNB3,CYR61,PTPRR,IL1RN,IL 7R,SEMA3E,VEGFA GO:0048812 neuron projection 5.43E-04 RELN,SLITRK4,DAB1,GATA3,ZNF280B,CRA morphogenesis BP2,EFNB3,PTPRD,KLF7,DNM3,SYT4,SEMA3 E,VEGFA,NIN GO:0021517 ventral spinal 5.83E-04 RELN,DAB1,LMO4,OLIG2 cord development GO:0040013 negative 6.12E-04 HTR2C,GATA3,ST6GAL1,KLF4,PTPRR,RNF41 regulation of ,IL1RN,PADI2,SEMA3E locomotion GO:0031339 negative 6.26E-04 TRIM9,SYT4 regulation of vesicle fusion GO:1902261 positive 6.26E-04 AKAP6,AKAP7 regulation of delayed rectifier potassium channel activity GO:1903818 positive 6.26E-04 AKAP6,AKAP7 regulation of voltage-gated potassium channel activity GO:0060429 epithelium 6.42E-04 FNDC3B,ZEB1,MCIDAS,GATA3,LMO4,KLF4, development CRABP2,GSTM3,TGM2,PPP2R3A,HEG1,THRB ,ACTL8,CYR61,MICAL2,FZD5,JAG1,PAX7,DA AM1,EMP1,SEMA3E,VEGFA Continued 76

Table 4 continued

ID Name p-value Hit in Query List GO:0051241 negative 6.67E-04 CYLD,FNDC3B,DDX58,DAB1,GATA3,TWIST2 regulation of ,MRVI1,KLF4,CRY2,TGFB3,XYLT1,HAVCR2, multicellular ADRA1D,GUCY1A3,PTPRR,JAG1,DNM3,OLIG organismal 2,SEMA3E,VEGFA process GO:0022603 regulation of 6.75E-04 CYLD,RELN,ERG,DAB1,GATA3,HIPK2,ST6G anatomical AL1,KLF4,CRABP2,TGFB3,PPP2R3A,PTPRD,F structure ZD5,DNM3,SYT4,DAAM1,IL1RN,SEMA3E,VE morphogenesis GFA,NIN GO:0010720 positive 6.99E-04 AKAP6,RELN,ZEB1,DAB1,CRABP2,TGFB3,PT regulation of cell PRD,ANKRD1,SYT4,QKI,OLIG2,VEGFA,NIN development GO:0048598 embryonic 7.15E-04 ALX3,ZEB1,DUSP4,GATA3,HIPK2,TWIST2,L morphogenesis MO4,KLF4,CRABP2,TGFB3,CYR61,MICAL2,F ZD5,IL1RN GO:0031346 positive 7.56E-04 RELN,CRABP2,TGFB3,PTPRD,ANKRD1,DNM regulation of cell 3,SYT4,QKI,VEGFA,NIN projection organization GO:0016079 synaptic vesicle 7.80E-04 PCLO,SYT11,TRIM9,SYT4,SYT13 exocytosis GO:0046903 secretion 7.83E-04 ABCA1,HTR2C,PCLO,GATA3,MYRIP,CREB3L 1,CRY2,TGFB3,HAVCR2,ACSL4,GBP5,SLC26 A7,SYT11,TRIM9,ANKRD1,SYT4,LTBP2,IL1R N,TNFSF15,SYT13,VEGFA GO:0048569 post-embryonic 7.90E-04 KLF4,FZD5,VEGFA organ development GO:0048706 embryonic 7.94E-04 ALX3,ZEB1,TWIST2,TGFB3,KIAA1217,PAX7 skeletal system development GO:0072359 circulatory 7.96E-04 AKAP6,XIRP1,ERG,MYLK,GATA3,HIPK2,LM system O4,KLF4,TGFB3,HEG1,PTPRB,CYR61,MICAL development 2,FZD5,JAG1,ANKRD1,QKI,SEMA3E,VEGFA GO:0072358 cardiovascular 7.96E-04 AKAP6,XIRP1,ERG,MYLK,GATA3,HIPK2,LM system O4,KLF4,TGFB3,HEG1,PTPRB,CYR61,MICAL development 2,FZD5,JAG1,ANKRD1,QKI,SEMA3E,VEGFA GO:0010769 regulation of cell 8.73E-04 RELN,DAB1,ST6GAL1,CRABP2,TGFB3,PTPR morphogenesis D,DNM3,SEMA3E,VEGFA,NIN involved in differentiation GO:0010648 negative 8.86E-04 CYLD,AMER1,RNF152,PEG10,DUSP4,DAB1,G regulation of cell ATA3,HIPK2,GRB10,KLF4,RGS3,CREB3L1,CR communication Y2,TGFB3,PPP2R3A,SYT11,PTPRR,RAB7B,RN F41,IL1RN,PADI2,VEGFA Continued

Table 4 continued

ID Name p-value Hit in Query List GO:0051049 regulation of 8.96E-04 AKAP6,CYLD,RELN,ABCA1,KCNH7,HTR2C, transport MYLK,DDX58,PCLO,GATA3,MYRIP,GRB10,B 3GAT3,CRY2,TGFB3,HAVCR2,ACSL4,KCNA3 ,SYT11,TRIM9,FZD5,ANKRD1,PKIA,SYT4,IL1 RN,TNFSF15,SYT13,AKAP7,VEGFA GO:0045666 positive 9.09E-04 RELN,ZEB1,DAB1,CRABP2,PTPRD,ANKRD1, regulation of SYT4,QKI,VEGFA,NIN neuron differentiation GO:0048667 cell 9.28E-04 RELN,SLITRK4,DAB1,GATA3,ZNF280B,CRA morphogenesis BP2,EFNB3,PTPRD,KLF7,DNM3,SEMA3E,VE involved in GFA,NIN neuron differentiation GO:0006868 glutamine 9.34E-04 SLC1A4,SLC38A1 transport GO:0031327 negative 9.36E-04 AKAP6,CYLD,KLF12,ZEB1,BACH2,GATA3,HI regulation of PK2,GRB10,TWIST2,ZNF280B,KLF4,ARID5A, cellular PHF21A,CRY2,TGFB3,HAVCR2,IGF2BP2,THR biosynthetic B,ANKRD1,PKIA,QKI,OLIG2,PRDM6,SAMD4 process A,VEGFA GO:0000902 cell 9.68E-04 CYLD,FNDC3B,RELN,ERG,MCIDAS,SLITRK4 morphogenesis ,DAB1,GATA3,ST6GAL1,ZNF280B,CRABP2,T GFB3,HEG1,EFNB3,PTPRD,KLF7,DNM3,SYT4 ,IL7R,SEMA3E,VEGFA,NIN GO:0090287 regulation of 1.08E-03 ZEB1,PEG10,GATA3,HIPK2,GRB10,TGFB3,CY cellular response R61,VEGFA to growth factor stimulus GO:0032940 secretion by cell 1.13E-03 ABCA1,HTR2C,PCLO,GATA3,MYRIP,CREB3L 1,CRY2,TGFB3,HAVCR2,GBP5,SYT11,TRIM9, ANKRD1,SYT4,LTBP2,IL1RN,TNFSF15,SYT13 ,VEGFA GO:0051962 positive 1.18E-03 RELN,ZEB1,SLITRK4,DAB1,CRABP2,PTPRD, regulation of ANKRD1,SYT4,QKI,OLIG2,VEGFA,NIN nervous system development GO:0009968 negative 1.19E-03 CYLD,AMER1,RNF152,PEG10,DUSP4,DAB1,G regulation of ATA3,HIPK2,GRB10,KLF4,RGS3,CREB3L1,CR signal Y2,TGFB3,PPP2R3A,PTPRR,RAB7B,IL1RN,PA transduction DI2,VEGFA GO:0010558 negative 1.19E-03 CYLD,KLF12,ZEB1,BACH2,GATA3,HIPK2,GR regulation of B10,TWIST2,ZNF280B,KLF4,ARID5A,PHF21A, macromolecule CRY2,TGFB3,HAVCR2,IGF2BP2,THRB,ANKR biosynthetic D1,PKIA,QKI,OLIG2,PRDM6,SAMD4A,VEGFA process GO:0001974 blood vessel 1.19E-03 ERG,TGM2,JAG1,VEGFA remodeling Continued 78

Table 4 continued

ID Name p-value Hit in Query List GO:0045778 positive 1.20E-03 TGFB3,THRB,CYR61,JAG1,VEGFA regulation of ossification GO:0009890 negative 1.22E-03 AKAP6,CYLD,KLF12,ZEB1,BACH2,GATA3,HI regulation of PK2,GRB10,TWIST2,ZNF280B,KLF4,ARID5A, biosynthetic PHF21A,CRY2,TGFB3,HAVCR2,IGF2BP2,THR process B,ANKRD1,PKIA,QKI,OLIG2,PRDM6,SAMD4 A,VEGFA GO:0001654 eye development 1.22E-03 ZEB1,GATA3,HIPK2,TWIST2,KLF4,PPP2R3A, FZD5,JAG1,NHS,VEGFA GO:0007416 synapse assembly 1.23E-03 RELN,PCLO,SLITRK4,PTPRD,FZD5,DNM3 GO:0035295 tube development 1.23E-03 FNDC3B,ZEB1,DAB1,GATA3,LMO4,TGFB3,T GM2,HEG1,THRB,MICAL2,JAG1,PAX7,SEMA 3E,VEGFA GO:0051172 negative 1.27E-03 AKAP6,CYLD,KLF12,ZEB1,HTR2C,BACH2,G regulation of ATA3,HIPK2,TWIST2,ZNF280B,KLF4,ARID5A nitrogen ,PHF21A,CRY2,TGFB3,HAVCR2,IGF2BP2,THR compound B,ANKRD1,PKIA,QKI,OLIG2,PRDM6,SAMD4 metabolic process A,VEGFA GO:0005667 transcription 3.55E-05 ZEB1,GATA3,HIPK2,TWIST2,LMO4,KLF4,HN factor complex F4G,ARNTL2,ARID5A,ANKRD1,PAX7,OLIG2 GO:0030054 cell junction 1.78E-04 AKAP6,XIRP1,MYLK,DDX58,DTNA,ARHGAP 26,PCLO,CLDN1,HAVCR2,FRMD4A,TGM2,HE G1,SYT11,PTPRR,TRIM9,ARHGAP31,FZD5,JA G1,NHS,SYT4,SAMD4A,GABRA3