Human Fertility, 2014; Early Online: 1–8 © 2014 The British Fertility Society ISSN 1464-7273 print/ISSN 1742-8149 online DOI: 10.3109/14647273.2014.956811

ORIGINAL ARTICLE

Circulating microRNAs in patients with polycystic ovary syndrome

Hum Fertil (Camb) Downloaded from informahealthcare.com by QUT Queensland University of Tech on 10/15/14 For personal use only.

Cai-Fei Ding, Wang-Qiang Chen, Yu-Tian Zhu, Ya-Li Bo, Hui-Min Hu & Ruo-Heng Zheng Reproductive Department, Integrated Chinese and Western Medicine Hospital of Zhejiang Province, Hangzhou, P. R. China Abstract Aim: To explore the pattern of expression of circulating miRNAs in patients with polycystic ovary syndrome (PCOS). Materials and methods: Microarray and qRT-PCR were used to investigate circulating miRNAs in PCOS during clinical diagnosis. The targets of dys-regulated miRNAs were predicted using bioinformatics, followed by function and pathway analysis using the databases of Gene Ontology and the KEGG pathway. Results: BMI, triglyceride, HOMA-IR, Testosterone and CRP levels were significantly higher, while estradiol was significantly lower in PCOS than in control groups. After SAM analysis, 5 circulating miRNAs were significantly up-regulated (let-7i-3pm, miR-5706, miR-4463, miR-3665, miR-638) and 4 (miR-124-3p, miR-128, miR-29a-3p, let-7c) were down-regulated in PCOS patients. Hierarchical clustering showed a general distinction between PCOS and control samples in a heat map. After joint prediction by different statistical methods, 34 and 41 genes targeted were up-and down-regulated miRNAs, in PCOS and controls, respectively. Further, GO and KEGG analyses revealed the involvement of the immune system, ATP binding, MAPK signaling, apoptosis, angiogenesis, response to reactive oxygen species and p53 signaling pathways in PCOS. Conclusions: We report a novel non-invasive miRNA profile which distinguishes PCOS patients from healthy controls. The miRNA-target database may provide a novel understanding of PCOS and potential therapeutic targets.

Keywords: Polycystic ovary syndrome, circulating miRNA, microarray, qRT-PCR

Introduction

region of target mRNAs with imperfect complementarity (Ambros, 2004). MiRNAs are processed from 70- to 100-nt double-stranded hairpin precursors by the RNaseIII Dicer, and endogenously expressed in the RNA-induced silencing complex in the cytoplasm (Lau et al., 2001). The importance of miRNA largely relates to its conservative expression in all multi-cellular organisms with certain biological functions (Pasquinelli et al., 2000). Recently, studies on specific miRNA expression patterns in various diseases have been reported, including cancer (Calin & Croce, 2006) and nonalcoholic fatty liver disease (Jin et al., 2009). In addition, an altered miRNA expression profile has been reported in the ovaries in a dihydrotestosterone-induced rat PCOS model (Hossain et al., 2013), but a study in humans is lacking. Recently, it has been reported that serum and other body fluids contain stable miRNA signatures (Chen et  al., 2008) while expression patterns of various miRNAs have been revealed in different diseases (Brase et  al., 2010). Although specific patterns of expression of miRNAs have been identified in the follicular fluid of PCOS patients (Sang et al., 2013; Roth et al., 2014)

Polycystic ovary syndrome (PCOS) is a common endocrine and metabolic disorder characterized by ovulation failure, hyperandrogenemia and insulin resistance (Glintborg & Andersen, 2010). PCOS affects up to 5–10% women of reproductive age and accounts for approximately 75% of anovulatory infertility (Ehrmann, 2005). Though under intensive investigation, the pathogenesis of PCOS is still unclear, since metabolic influences, insulin resistance, genetic and environmental factors may act synergistically (Goodarzi et al., 2011). Various reports have revealed significantly altered gene expression in the ovaries of PCOS patients (Diao et al., 2004). Further in-depth analysis categorized these into different functional groups, involving steroidogenesis, apoptosis, Wnt signaling and ovarian folliculogenesis (Hughes et al., 2006). Nevertheless, how these genes are transcriptionally and post-transcriptionally regulated is poorly understood. MicroRNAs (miRNAs) belong to a 19–25 nucleotide (nt) non-coding family and cause translational repression or mRNA cleavage by recognizing the 3′-untranslated

Correspondence: Ruo-Heng Zheng, Reproductive Department, Integrated Chinese and Western Medicine Hospital of Zhejiang Province, 208 East Road, Hangzhou Ring, Hangzhou 310003, P. R. China. Tel:  86-571-56108767. E-mail: [email protected] (Received 30 January 2014; revised 2 May 2014; accepted 8 May 2014)

1

2  C.-F. Ding et al. together with miR-21, miR-27b, miR-103 and miR-155 levels (Murri et al., 2013), the global serum profile of miRNAs in PCOS patients is unknown. In the present study, we have prospectively examined the relationship between the expression of different circulating miRNAs in PCOS and control patients with screening and verification, with an explanation of what the screening and verification settings mean, with the aim of forming a specific miRNA profile with which to distinguish PCOS from control patients.

Hum Fertil (Camb) Downloaded from informahealthcare.com by QUT Queensland University of Tech on 10/15/14 For personal use only.

Methods Patients During 2012, subjects from the local community who visited the Reproductive Department of the Integrated Chinese and Western Medicine Hospital of Zhejiang Province, for infertility evaluation, were recruited to the study. The diagnosis of PCOS was defined by the Rotterdam criteria (Dewailly et  al., 2010), which mainly includes: menstrual cycle anomalies, amenorrhoea, oligomenorrhoea or long cycles, clinical and/ or biochemical hyperandrogenism and ultrasound appearance of polycystic ovaries. Nine patients had PCOS and there were nine healthy control women without evidence of androgen excess or ovulatory dysfunction. Recruitment to the study was randomized at the initial screening stage, and serum miRNA expression pattern was investigated through microarray analysis. A further cohort of PCOS patients (n  9) and healthy controls (n  9) were enrolled, whose previously selected serum miRNAs were verified through real-time reverse transcription-polymerase chain reaction (RT-PCR). None of the subjects had undergone previous treatment with oral contraceptives, antiandrogens, insulin sensitizers or other drugs known to interfere with sex hormone secretion and metabolism, during the previous 6 months. The study was carried out according to the declaration of Helsinki and approved by the Ethics Committee of the Reproductive Department of the Integrated Chinese and Western Medicine Hospital of Zhejiang Province. The study design and manuscript preparation were based on guidelines from the STROBE statement (von Elm et al., 2007). Written informed consent was obtained from all the participants. Anthropometric and laboratory analysis Common anthropometric parameters were collected, including age, weight, height, body mass index (BMI) and waist circumference. Samples for laboratory analysis were collected on the third day of the menstrual period. The biomarkers measured were plasma glucose, triglyceride, total cholesterol, neutrophils, C-reactive protein (CRP), total testosterone, estradiol, follicle stimulating hormone (FSH), luteinizing hormone (LH), progesterone, prolactin, fasting glucose and fasting insulin. In addition, the homeostasis model assessment of insulin

resistance (HOMA-IR) was calculated according to Matthews et al. (1985). RNA isolation, miRNA microarray and Real-time RT-PCR Total serum miRNA was harvested using TRIzol (Invitrogen) and miRNeasy mini kit (QIAGEN), according to the manufacturer’s instructions. After RNA quantity measurement by NanoDrop 1000, the samples were labeled by the miRCURY™ Hy3™/Hy5™ Power labeling kit followed by hybridization on the miRCURY™ LNA Array (v.16.0). Slides were scanned using the Axon GenePix 4000B microarray scanner after several washes, followed by importation into GenePix Pro 6.0 software (Axon) for grid alignment and data extraction. Replicated miRNAs were averaged and miRNAs with intensities over 50 in all samples were chosen for further median normalization. Thereafter, differentially expressed miRNAs were identified through Volcano Plot filtering. Real-time RT-PCR was carried out to verify significantly changed miRNAs revealed by microarray, by using the stem-loop antisense primer mix and AMV transcriptase (TaKaRa, China), as previously reported (Chen et  al., 2005). All primers were purchased from Wo Seng Biotechnology Company (Hangzhou, China) based on miRNA sequences released by the Sanger Institute (Griffiths-Jones, 2004). Since U6 was found to be stable in serum, it was selected as the endogenous reference control for all miRNAs assessed in this study. The relative amount of each miRNA to U6 RNA was calculated using the equation 2ΔCT, where ΔCT  CTmiRNA CTU6. Bioinformatics analysis of significantly dys-regulated miRNAs Initially, the PCR-retrieved -ΔCT data were normalized, mean-centered, log2- transformed and sequentially analyzed by the algorithm of significance analysis of microarrays (SAM, http://www-stat.stanford.edu/∼tibs/ SAM/). SAM calculates a score for each gene as the change of expression relative to the standard deviation of all measurements (Olson, 2006). Furthermore, in SAM, a False Discovery Rate (FDR)  5% was selected and miRNAs with an over two-fold change in expression were considered significantly dys-regulated. To visualize numeric changes through graphic representation of the raw -ΔCT, hierarchical clustering was used to generate both miRNA (from SAM results) and sample (from PCOS and control) trees based on the algorithm of average linkage and Euclidian distance (Eisen et al., 1998). Target gene prediction and statistics Several methods of bioinformatics were applied to analyze the function and pathways of miRNA-targeted genes. Firstly, an open access platform for miRNA target prediction (http://www.microrna.org/microrna/home.do) Human Fertility

Circulating miRNAs in PCOS patients  3 Table I. Anthropometric, metabolic and hormonal variables in PCOS patients and healthy controls of two cohorts. Screening cohort Variables

Control (n  9)

Age (years) BMI (kg/m2) Testosterone (ng/dl) Estradiol (pg/ml) Cholesterol (mg/dl) Triglycerides(mg/dl) HOMA-IR CRP (mg/L)

28.7  5.2 23.5  2.1 57.1  11.7 75.4  18.7 181.3  27.7 69.1  15.6 1.5  0.4 6.3  2.7

Verification cohort

PCOS (n  9) 27.9  4.3 29.1  4.3* 83.2  13.1* 43.9  10.3* 195.6  47.7 131.4  25.1* 2.9  0.7* 22.9  6.2*

Control (n  9)

PCOS (n  9)

29.1  4.8 22.7  3.2 55.6  9.8 82.1  16.8 169.8  35.3 71.3  13.9 1.3  0.5 5.1  2.2

28.3  5.6 31.5  4.1* 88.5  16.9* 55.1  12.7* 175.9  41.5 145.4  21.7* 3.1  0.9* 33.7  9.1*

Hum Fertil (Camb) Downloaded from informahealthcare.com by QUT Queensland University of Tech on 10/15/14 For personal use only.

*P  0.05 compared with control group.

and TargetScan (http://www.targetscan. org/) were jointly performed to predict miRNA-targeted genes. Secondly, gene ontology and KEGG pathway were jointly analyzed to explore the functional annotation and pathway enrichment of those predicted genes, as previously described (Shen et al., 2010). Each experiment was performed in triplicate and data were expressed as means  SD (standard deviation). The student’s t-test for two unpaired groups was executed by SPSS 17.0. The differences were considered statistically significant at p  0.05. Results Anthropometric and laboratory analysis of PCOS patients The differences in anthropometric, metabolic and hormonal variables between PCOS patients and healthy controls are summarized in Table I. In both screening and verification cohorts, all participants were of Han ethnicity. There were no significant differences in age and cholesterol levels between the PCOS and control groups. Nevertheless, the BMI, triglycerides, HOMAIR levels, testosterone, and inflammatory factor CRP were significantly higher in the PCOS groups, while estradiol was significantly decreased. Unique expression pattern of serum miRNAs in PCOS patients At the screening stage, we compared the differences of serum miRNAs between nine PCOS patients and nine control subjects by microarray. Generally, there were 11 up-regulated and 8 down-regulated serum miRNAs revealed by this method. These miRNAs were initially verified by the stem loop RT-PCR re-examination, a high- throughput method regarded as the gold standard of RNA quantification, with less technical noise and greater reproducibility. Except for three miRNAs (hsamiR-27b, hsa-miR-25 and hsa-miR-4490) that showed a change within 0.5–2 fold, the other miRNAs had the same degree of expression as in the microarray results, reinforcing the effectiveness of using microarray for serum miRNA screening. The remaining 9 up-regulated and 7 down-regulated miRNAs were re-examined in a verification cohort © 2014 The British Fertility Society

consisting of 9 PCOS patients and 9 control subjects, by the stem loop RT-PCR. In this cohort, 75% of the miRNAs showed the same significant expression changes, where two increased miRNAs (hsa-miR-125 and hsa-miR-449) and one decreased miRNA (hsa-miR-371) were further excluded. Thereafter, the SAM method was employed to identify miRNAs that were significantly associated with PCOS, and a group of 5 up- and 4 down-regulated miRNAs was revealed (Table II). Finally, according to the results of hierarchical clustering, these SAM-revealed miRNAs could be distinguished between the PCOS and control samples in the column as well as being up- and down-regulated miRNAs in the row (Figure 1). On the whole, the pattern of expression of miRNAs reconfirmed at the verification stage showed their potential capacity as non-invasive biomarkers in PCOS diagnosis. Functions and pathway analysis of miRNAs targets The data from the SAM results were further analyzed to explore the functions and pathways of the miRNAtargets. Firstly, downstream 1515 and 4851 genes of 5 up- and 4 down-regulated miRNAs were predicted by the algorithm of TargetScan, further supporting the hypothesis that a single miRNA may regulate hundreds of targets. Using another gene prediction method (www.microrna.org), downstream 9707 and 30145 genes of 5 up- and 4 down-regulated miRNAs were predicted. These two results were compared to reduce redundancy, and only genes predicted by both two Table II. Differentially expressed serum miRNAs in PCOS patients revealed by SAM. MiRNAs Up-regulated hsa-let-7i-3p hsa-miR-5706 hsa-miR-4463 hsa-miR-3665 hsa-miR-638 Down-regulated hsa-miR-124-3p hsa-miR-128 hsa-miR-29a-3p hsa-let-7c

Fold change (PCOS/control)

p value

10.62 10.23 1.77 1.18 1.14

1.55E-02 9.44E-03 2.19E-02 8.07E-02 6.98E-02

0.10 0.40 0.67 0.72

4.09E-02 5.92E-02 9.21E-02 4.01E-02

4  C.-F. Ding et al.

Hum Fertil (Camb) Downloaded from informahealthcare.com by QUT Queensland University of Tech on 10/15/14 For personal use only.

Discussion

Figure 1. Cluster analysis of PCOS patients and control subjects. The heat map showed the separation of PCOS patients from controls, based on 9 serum miRNA signatures. Red, increased expression; green, decreased expression; black, median expression level equal to 1; DN, PCOS patients, Con, control.

algorithms with high confidence (evaluated from novel algorithm miRWalk, http://www.umm.uni- heidelberg. de/apps/zmf/mirwalk/) were selected for further analysis (Table III). These included those involved in the functions of ATP binding (ABCC5), transmembrane signal transduction (TSPAN2), RAS activation (RASGRP1), proteasome regulation (PSME3), nuclear RNA exportation (NXF1), insulin release and resistance (IGF2BP1), cell signaling regulation in response to insulin and growth factors (AKT3), transcriptional regulation (HDAC5), apoptosis regulation (PDCD6, BCL2L11, BCL2L10), ATP metabolism (ABCF2, ATP9A, PARP3) and secretion processes (SCAMP2, SCAMP3). More importantly, several previously identified disease-related genes were also found in the prediction list of dys-regulated miRNAs, such as DSCR3 (related to Down’s syndrome) and WASF2 (WiskottAldrich syndrome), To reduce nonspecific genes and explore in-depth biological information, a functional annotation and over-representation analyses of the GO Biological Process and KEGG pathway on those genes were performed through searching on the website “DAVID”. According to the GO analysis, genes predicted by upand down-regulated miRNAs were respectively related to cell motion, vacuolar transport, proteolysis, immune system development, transcription and ATP binding (Figure 2A) and intracellular signaling cascade, apoptosis, angiogenesis, response to reactive oxygen species and hypoxia and cell adhesion (Figure 2B). According to the KEGG pathway analysis, genes predicted by upand down- regulated miRNAs were respectively found to participate in MAPK signaling, calcium signaling, purine metabolism, and antigen processing and presentation (Figure 3A) and cell cycle, oocyte meiosis and the p53 signaling pathway (Figure 3B).

Currently, the pathogenesis of PCOS is still unclear and the rise in interest of genetics (Barber & Franks, 2013) and proteomics (Insenser & Escobar-Morreale, 2013) has promoted its rapid development. MiRNA expression analysis belongs to the category of transcriptomics, which has already shown its power in cancer diagnosis and potential therapeutic target investigation (Di Leva & Croce, 2013). Moreover, recent evidence implicates the regulatory function of miRNAs in oocyte maturation, ovarian follicular development and ovarian cancer progression (Toloubeydokhti et  al., 2008). Although miRNA expression and the function of specific miRNAs in PCOS have been partially revealed in animal and human studies (Hossain et al., 2013; Murri et al., 2013; Sang et al., 2013), evidence for serum miRNAs as a non-invasive biomarkers is lacking. Our studies on serum miRNA profiles in PCOS patients might provide potential non-invasive molecular biomarkers for PCOS diagnosis, and expand our understanding of PCOS at the level of post-transcriptional regulation. In addition, the hundreds of miRNA-target genes revealed from this study may provide data reservoir for functional studies to explore therapeutic targets of PCOS. In this study, serum miRNA expression patterns were sequentially screened and verified in two independent cohorts of PCOS patients. Specifically, microarray technology was used to select dys-regulated miRNAs in the PCOS screening cohort, followed by qRT-PCR verification. The verified miRNAs were further tested in an independent PCOS cohort by qRT-PCR and ensuing bioinformatics analysis. Finally, 5 increased and decreased miRNAs were collected to distinguish the PCOS patients from the controls (Table II). It is well known that bioinformatics is a double-edged sword, whose capacity for reducing redundancy from high-flux data is at the cost of increased possibility of enrolling false positives or redundant data. Searching for biological functions of these miRNAs from literature provides one option to tackle this dilemma. In our SAM results, although miR-5706, miR-4463 and miR3665 were miRNAs already reported with unknown biological functions, several miRNAs have been found to participate in various pathophysiological processes of the genital system, consistent with our results. For example, serum let-7i was revealed as a novel biomarker and therapeutic target in human epithelial ovarian cancer (Yang et al., 2008) while miR-128 and its target genes were involved in regulating ovarian cancer cell behavior (Woo et  al., 2012). In addition, miR-214 was reported to inhibit the migration and invasion of ovarian cancer cells (Zhang et  al., 2013) and control male reproductive success in Drosophila (Weng et al., 2013). Finally, Let-7c may be involved in auto-immune activity by regulating dendritic cell (Kim et  al., 2013) and macrophage polarization (Banerjee et al., 2013), potentially influencing the immune status of the reproductive tract. Human Fertility

Circulating miRNAs in PCOS patients  5 Table III. Genes jointly predicted by algorithms of Target Scan and miRanda.

Hum Fertil (Camb) Downloaded from informahealthcare.com by QUT Queensland University of Tech on 10/15/14 For personal use only.

Gene symbol Predicted by up-regulated miRNAs KCNE3 SH2B3 HMG2L1 GJC1 ABCC5 DPP3 PTPRU UST ACTR2 TSPAN2 TRIM10 ENAM ACTR1B RASGRP1 FARP1 MPHOSPH6 NME6 GNPDA1 ZBTB33 CELA3A CDK2 PSME3 SPEG DSCR3 CORO2B HOXB13 NXF1 CRTAP POLR3G IGF2BP1 ADCY1 ZNF266 STARD10 ZNF275 Predicted by down-regulated miRNAs CDH2 AKT3 HDAC5 PDCD6 BCL2L11 TOM1L1 AP1M2 ABCF2 SCAMP2 IL18BP CDH9 ATP9A ARPC5 CTDSP2 HIPK3 NAMPT GNPDA1 BCL2L10 HCN4 CDH5 PARP3 SH2D3C SCAMP3 SNUPN TSPAN1 PPIF RAD50 LRPPRC ADA NAALADL1 ABI1 SLC17A4 ABCB6 RWDD2B © 2014 The British Fertility Society

Full name of the gene Potassium voltage-gated channel, Isk-related family, member SH2B adaptor protein 3 High mobility group box domain containing 4 Gap junction protein, gamma 1 ATP-binding cassette, sub-family C (CFTR/MRP), member 5 Dipeptidyl-peptidase 3 Protein tyrosine phosphatase, receptor type, U Uronyl-2-sulfotransferase ARP2 actin-related protein 2 homolog Tetraspanin 2 Tripartite motif containing 10 Enamelin ARP1 actin-related protein 1 homolog B RAS guanyl releasing protein 1 (calcium and DAG-regulated) FERM, RhoGEF (ARHGEF) and pleckstrin domain protein 1 (chondrocyte-derived) M-phase phosphoprotein 6 NME/NM23 nucleoside diphosphate kinase 6 Glucosamine-6-phosphate deaminase 1 zinc finger and BTB domain containing 33 Chymotrypsin-like elastase family, member 3A Cyclin-dependent kinase 2 Proteasome (prosome, macropain) activator subunit 3 Spermidine N1-acetyltransferase Down syndrome critical region gene 3 Coronin, actin binding protein, 2B Homeobox B13 Nuclear RNA export factor 1 Cartilage associated protein Polymerase (RNA) III (DNA directed) polypeptide G Insulin-like growth factor 2 mRNA binding protein 1 Adenylate cyclase 1 Zinc finger protein 2 StAR-related lipid transfer (START) domain containing 10 Zinc finger protein 275 Cadherin 2, type 1, N-cadherin v-akt murine thymoma viral oncogene homolog 3 Histone deacetylase 5 Programmed cell death 6 BCL2-like 11 (apoptosis facilitator) Target of myb1 (chicken)-like 1 Adaptor-related protein complex 1, mu 2 subunit ATP-binding cassette, sub-family F (GCN20), member 2 Secretory carrier membrane protein 2 Interleukin 18 binding protein Cadherin 9, type 2 ATPase, class II, type 9A Actin related protein 2/3 complex, subunit 5 CTD (carboxy-terminal domain, RNA polymerase II, polypeptide A) small phosphatase 2 Homeodomain interacting protein kinase 3 Nicotinamide phosphoribosyltransferase Glucosamine-6-phosphate deaminase 1 BCL2-like 10 (apoptosis facilitator) Hyperpolarization activated cyclic nucleotide-gated potassium channel 4 Cadherin 5, type 2 Poly (ADP-ribose) polymerase family, member 3 SH2 domain containing 3C Secretory carrier membrane protein 3 Snurportin 1 Tetraspanin 1 Peptidylprolyl isomerase F (cyclophilin F) RAD50 homolog (S. cerevisiae) Leucine-rich pentatricopeptide repeat containing Adenosine deaminase N-acetylated alpha-linked acidic dipeptidase-like 1 Abl-interactor 1 Solute carrier family 17, member 4 ATP-binding cassette, sub-family B (MDR/TAP), member 6 RWD domain containing 2B (Continued)

6  C.-F. Ding et al. Table III. (Continued) Gene symbol

Hum Fertil (Camb) Downloaded from informahealthcare.com by QUT Queensland University of Tech on 10/15/14 For personal use only.

FEM1B ARL4C ARFRP1 CLEC3A WASF2 ZNF197 DHRS2

Full name of the gene Feminization 1 homolog b ADP-ribosylation factor-like 4C ADP-ribosylation factor related protein 1 C-type lectin domain family 3, member A WAS protein family, member 2 Zinc finger protein 197 Dehydrogenase/reductase (SDR family) member 2

Previously, two researchers reported the miRNA profile of follicular fluid from PCOS patients. In the study from Sang et al. (2013), miR-132 and miR-320 were expressed at significantly lower levels in the follicular fluid of PCOS patients than in healthy controls. In another study, from Roth et al. (2014), 5 miRNAs (32, 34c, 135a, 18b, and 9) showed significantly increased expression in the PCOS group. These two results showed that significantly down- and up- regulated miRNAs could be seen as complementarity to each other. In another study investigating selected circulating miRNAs, Murri et al. (2013) revealed a tendency for increased circulating miR-21, miR-27b, miR-103, and miR-155 in PCOS patients. Nevertheless, none of the above mentioned studies explored the whole serum miRNA profile in PCOS patients whereas our

SAM analysis identified 9 significantly changed serum miRNAs (Table II). Intriguingly, there were no overlaps in changes to miRNAs between previous studies and our own, which may be due to different tissues (serum vs. follicular fluid) and the race of subjects (Han vs. Caucasus). It is therefore important to investigate miRNA profiles from different races in a larger study. Except for revealing the possibility of using the serum miRNA profile as a non-invasive biomarker for PCOS diagnosis, our results may shed light on the mechanisms of PCOS by predicting miRNA-regulated genes. Currently, several genes have been reported to participate in various reproductive diseases. For instance, SH2B3 was discovered as a link between immune and inflammatory signaling (Devalliere & Charreau, 2011) while RASGRP1

Figure 2. Gene ontology analysis of genes regulated by significantly up-regulated (A) and down-regulated (B) miRNAs.

Figure 3. KEGG pathway analysis of genes regulated by significantly up-regulated (A) and down-regulated (B) miRNAs. Human Fertility

Hum Fertil (Camb) Downloaded from informahealthcare.com by QUT Queensland University of Tech on 10/15/14 For personal use only.

Circulating miRNAs in PCOS patients  7 was involved in invariant NKT cell development (Shen et  al., 2011), which may further influence the immune status of the genital system. In addition, IGF2BP1 was found to be a post-transcriptional “driver” of tumor cell migration (Stohr & Huttelmaier, 2012) while increased ABCC5 expression was observed in human cervical cancer cells during growth (Eggen et  al., 2012), the SNP of PDCD6 was associated with increased risk of endometriosis (Shi et  al., 2013) and BCL2L10 was regarded as the survival factor expressed in oocytes and early embryos (Guillemin et al., 2009). More importantly, FEM1B was reported as a candidate gene for PCOS (Goodarzi et  al., 2008). Nevertheless, the functions of many predicted genes in Table III are still to be discovered and to overcome this redundancy, we used GO functional analysis and the KEGG pathway to explore the pathogenesis of PCOS (Figures 2 and 3). According to the categories of OMIM disease, predicted genes were found to be involved in various disorders, including diabetes, celiac disease, osteogenesis imperfecta, hypokalemic periodic paralysis, adenosine deaminase deficiency, sick sinus syndrome and Leigh syndrome. There are several limitations to this research that should be acknowledged. Firstly, for the miRNAregulated genes found to be predictive, we did not carry out related experiments to verify these novel findings. However, we did use enrichment analysis on the GO category and KEGG pathway as a compensation, to derive maximal information from the information retrieved. Secondly, other than miRNA regulation, processes such as protein methylation and ubiquitination may also modulate target gene expression and the invoking the miRNA-gene axis should be done with caution. Finally, although the serum miRNA profile was screened and verified in two sequential cohorts of PCOS patients, the limited number of subjects has decreased the validity of the results and a larger human cohort will be required in the future. In summary, we have, for the first time, reported the use of a non- invasive serum miRNA profile which can potentially differentiate PCOS patients from healthy controls. Furthermore, the miRNA-target reservoir may provide novel understanding of PCOS pathogenesis and provide potential therapeutic targets.­­­­­­ Declaration of interest:  The authors report no conflict of interest. The authors alone are responsible for the content and writing of the paper. Supported by the 2012 Foundation of Health Bureau of Zhejiang Province (no. 2012KYA159) and the 2012 Foundation of the Science and Technology Department of Zhejiang Province (no.2012C33103). References Ambros, V. (2004). The functions of animal microRNAs. Nature, 431, 350–355. Banerjee, S., Xie, N., Cui, H., Tan, Z., Yang, S., Icyuz, M., et  al. (2013). MicroRNA let-7c regulates macrophage polarization. Journal of Immunology, 190, 6542–6549. © 2014 The British Fertility Society

Barber, T.M. & Franks, S. (2013). Genetics of polycystic ovary syndrome. Frontiers of Hormone Research, 40, 28–39. Brase, J.C., Wuttig, D., Kuner, R., & Sultmann, H. (2010). Serum microRNAs as non-invasive biomarkers for cancer. Molecular Cancer, 9, 306. Calin, G.A. & Croce, C.M. (2006). MicroRNA signatures in human cancers. Nature Reviews Cancer, 6, 857–866. Chen, C., Ridzon, D.A., Broomer, A.J., Zhou, Z., Lee, D.H., Nguye, J.T., et al. (2005). Real-time quantification of microRNAs by stem-loop RT-PCR. Nucleic Acids Research, 33, e179. Chen, X., Ba, Y., Ma, L., Cai, X., Yin, Y., Wang, K., et  al. (2008). Characterization of microRNAs in serum: a novel class of biomarkers for diagnosis of cancer and other diseases. Cell Research, 18, 997–1006. Devalliere, J. & Charreau, B. (2011). The adaptor Lnk (SH2B3): an emerging regulator in vascular cells and a link between immune and inflammatory signaling. Biochemical Pharmacology, 82, 1391–1402. Dewailly, D., Hieronimus, S., Mirakian, P., & Hugues, J.N. (2010). Polycystic ovary syndrome (PCOS). Annales d’Endocrinologie, 71, 8–13. Diao, F.Y., Xu, M., Hu, Y., Li, J., Xu, Z., Lin, M., et al. (2004). The molecular characteristics of polycystic ovary syndrome (PCOS) ovary defined by human ovary cDNA microarray. Journal of Molecular Endocrinology, 33, 59–72. Di Leva, G. & Croce, C.M. (2013). miRNA profiling of cancer. Current Opinion in Genetics & Development, 23, 3–11. Eggen, T., Sager, G., Berg, T., Nergaard, B., Moe, B.T., & Orbo, A. (2012). Increased gene expression of the ABCC5 transporter without distinct changes in the expression of PDE5 in human cervical cancer cells during growth. Anticancer Research, 32, 3055–3061. Ehrmann, D.A. (2005). Polycystic ovary syndrome. The New England Journal of Medicine, 352, 1223–1236. Eisen, M.B., Spellman, P.T., Brown, P.O., & Botstein, D. (1998). Cluster analysis and display of genome-wide expression patterns. Proceedings of the National Academy of Sciences USA, 95, 14863–14868. Glintborg, D. & Andersen, M. (2010). An update on the pathogenesis, inflammation, and metabolism in hirsutism and polycystic ovary syndrome. Gynecological Endocrinology, 26, 281–296. Goodarzi, M.O., Dumesic, D.A., Chazenbalk, G., & Azziz, R. (2011). Polycystic ovary syndrome: etiology, pathogenesis and diagnosis. Nature Reviews Endocrinology, 7, 219–231. Goodarzi, M.O., Maher, J.F., Cui, J., Guo, X., Taylor, K.D., & Azziz, R.(2008). FEM1A and FEM1B: novel candidate genes for polycystic ovary syndrome. Human Reproduction, 23, 2842–2849. Griffiths-Jones, S. (2004). The microRNA Registry. Nucleic Acids Research, 32, D109–111. Guillemin, Y., Lalle, P., Gillet, G., Guerin, J.F., Hamamah, S., & Aouacheria, A. (2009). Oocytes and early embryos selectively express the survival factor BCL2L10. Journal of Molecular Medicine, 87, 923–940. Hossain, M.M., Cao, M., Wang, Q., Kim, J.Y., Schellander, K., Tesfaye, D., & Tsang, B.K. (2013). Altered expression of miRNAs in a dihydrotestosterone-induced rat PCOS model. Journal of Ovarian Research, 6, 36. Hughes, C., Elgasim, M., Layfield, R., & Atiomo, W. (2006). Genomic and post-genomic approaches to polycystic ovary syndrome–progress so far: Mini Review. Human Reproduction, 21, 2766–2775. Insenser, M. & Escobar-Morreale, H.F. (2013). Proteomics and polycystic ovary syndrome. Expert Review of Proteomics, 10, 435–447. Jin, X., Ye, Y.F., Chen, S.H., Yu, C.H., Liu, J., & Li, Y.M. (2009). MicroRNA expression pattern in different stages of nonalcoholic fatty liver disease. Digestive and Liver Disease, 41, 289–297. Kim, S.J., Gregersen, P.K., & Diamond, B. (2013). Regulation of dendritic cell activation by microRNA let-7c and BLIMP1. Journal of Clinical Investigation, 123, 823–833.

Hum Fertil (Camb) Downloaded from informahealthcare.com by QUT Queensland University of Tech on 10/15/14 For personal use only.

8  C.-F. Ding et al. Lau, N.C., Lim, L.P., Weinstein, E.G., & Bartel, D.P. (2001). An abundant class of tiny RNAs with probable regulatory roles in Caenorhabditis elegans. Science, 294, 858–862. Matthews, D.R., Hosker, J.P., Rudenski, A.S., Naylor, B.A.,Treacher, D.F., & Turner, R.C. (1985). Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia, 28, 412–419. Murri, M., Insenser, M., Fernandez-Duran, E., San-Millan, J.L., & Escobar-Morreale, H.F. (2013). Effects of Polycystic Ovary Syndrome (PCOS), Sex Hormones, and Obesity on Circulating miRNA-21, miRNA-27b, miRNA-103, and miRNA-155 Expression. Journal of Clinical Endocrinology and Metabolism, 98, E1835–1844. Olson, N.E. (2006). The microarray data analysis process: from raw data to biological significance. NeuroRx, 3, 373–383. Pasquinelli, A.E., Reinhart, B.J., Slack, F., Martindale, M.Q., Kuroda,  M.I., Maller, B., et  al. (2000). Conservation of the sequence and temporal expression of let-7 heterochronic regulatory RNA. Nature, 408, 86–89. Roth, L.W., McCallie, B., Alvero, R., Schoolcraft,W.B., Minjarez, D., & Katz-Jaffe, M.G. (2014). Altered microRNA and gene expression in the follicular fluid of women with polycystic ovary syndrome. Journal of Assisted Reproduction and Genetics, 31, 355–362. Sang, Q., Yao, Z., Wang, H., Feng, R., Wang, H., Zhao, X., et  al. (2013). Identification of microRNAs in human follicular fluid: characterization of microRNAs that govern steroidogenesis in vitro and are associated with polycystic ovary syndrome in vivo. Journal of Clinical Endocrinology and Metabolism, 98, 3068–3079. Shen, E., Diao, X., Wei, C., Wu, Z., Zhang, L., & Hu, B. (2010). MicroRNAs target gene and signaling pathway by bioinformatics analysis in the cardiac hypertrophy. Biochemical and Biophysical Research Communications, 397, 380–385.

Shen, S., Chen, Y., Gorentla, B.K., Lu, J., Stone, J.C., & Zhong, X.P. (2011). Critical roles of RasGRP1 for invariant NKT cell development. Journal of Immunology, 187, 4467–4473. Shi, S., Zhou, B., Zhang, K., & Zhang, L. (2013). Association between two single nucleotide polymorphisms of PDCD6 gene and increased endometriosis risk. Human Immunology, 74, 215–218. Stohr, N. & Huttelmaier, S. (2012). IGF2BP1: a post-transcriptional “driver” of tumor cell migration. Cell Adhesion & Migration, 6, 31231–31238. Toloubeydokhti, T., Bukulmez, O., & Chegini, N. (2008). Potential regulatory functions of microRNAs in the ovary. Seminars in Reproductive Medicine, 26, 469–478. von Elm, E., Altman, D.G., Egger, M., Pocock, S.J., Gotzsche, P.C., & Vandenbroucke, J.P. (2007). The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet, 370, 1453–1457. Weng, R., Chin, J.S., Yew, J.Y., Bushati, N., & Cohen, S.M. (2013). miR-124 controls male reproductive success in Drosophila. Elife, 2, e00640. Woo, H.H., Laszlo, C.F., & Greco, S., Chambers, S.K. (2012). Regulation of colony stimulating factor-1 expression and ovarian cancer cell behavior in vitro by miR-128 and miR-152. Molecular Cancer, 11, 58. Yang, N., Kaur, S., Volinia, S., Greshock, J., Lassus, H., Hasegawa, K., et al. (2008). MicroRNA microarray identifies Let-7i as a novel biomarker and therapeutic target in human epithelial ovarian cancer. Cancer Research, 68, 10307–10314. Zhang, H., Wang, Q., Zhao, Q., & Di, W. (2013). MiR-124 inhibits the migration and invasion of ovarian cancer cells by targeting SphK1. Journal of Ovarian Research, 6, 84.

Human Fertility

Circulating microRNAs in patients with polycystic ovary syndrome.

To explore the pattern of expression of circulating miRNAs in patients with polycystic ovary syndrome (PCOS)...
3MB Sizes 0 Downloads 6 Views