Genome-Wide association study (GWAS) identified PCOS susceptibility variants and replicates reported risk variants

doi:10.21203/rs.3.rs-3584882/v1

Download PDF

Research Article

Genome-Wide association study (GWAS) identified PCOS susceptibility variants and replicates reported risk variants

https://doi.org/10.21203/rs.3.rs-3584882/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 29 Feb, 2024

Read the published version in Archives of Gynecology and Obstetrics →

You are reading this latest preprint version

Background: Genetic predisposition and environmental factors are considered as risk factors for polycystic ovary syndrome (PCOS). Genome-wide association studies (GWAS) have been reported from various subpopulations to evaluate SNPs associated with PCOS risk. As no PCOS associated GWAS study has been reported from India.

Purpose: Present study was conducted to identify the PCOS-susceptible loci among the North Indian population and validation of significant loci reported by previous GWAS studies.

Methods: A total of 272 age-matched participants with 134 PCOS patients and 138 healthy controls were recruited. Genomic DNA was isolated and genotyped by using Infinium Global Screening Array v3.0 microchip considering HWE 10e^-5 statistically significant.

Results: A total of fifteen markers have been identified as candidate PCOS risk factors. Only two SNPs, namely rs17186366 and rs11171739 have been identified through replication analysis while comparing the previously reported PCOS GWAS data. In-silico analysis was performed to study the functional impact of identified gene variants in terms of gene ontology, pathways related to gene set, and cluster analysis to determine protein-protein interaction among genes or gene products.

Conclusion: Study suggests that multiple variants play an important role in PCOS pathogenesis and emphasize the importance of further genetic studies among Indian subpopulations.

Genetics

GWAS

PCOS

phenotype

replication

Study summarizes the importance of potential gene markers validated by replication and in-silico functional study, significantly involved in PCOS pathogenesis in studied population. These markers can be used in future as diagnostic markers for clinical phenotype identification.

Polycystic ovary syndrome (PCOS) is a complex endocrinopathy characterized by diverse metabolic manifestations and reproductive disturbance[1]. Based upon different ancestral segregation, ethnicity, and diagnostic criteria, its prevalence varies in a wide range from 4–21.27% worldwide[2]. The syndrome can be identified by following any of the three diagnostic criteria National Institute of Health (NIH,1990), Androgen Excess Society (AES, 2006), and Rotterdam Criteria 2003, the last one being most commonly followed accordingly to which presence of any two conditions out of three: oligo-ovulation/anovulation; biochemical hirsutism, and polycystic morphology in ovaries[3] indicates PCOS. This multifaceted syndrome has major comorbidities beyond just affecting the reproductive system such as metabolic disorders, diabetes, and cardiovascular disorders[4]. The etiology is still not clear but it is generally considered that its phenotypic heterogeneity is not affected by a single gene, but by gene-gene interaction, and interaction between gene-environmental factors[5]. Modifiable factors include includes dietary intake of trace elements such as Copper (Cu), Selenium (Se), Iron (Fe), and high-fat diet. Among non-modifiable genetic factors, various genetic variants have been studied previously which are involved in metabolic pathways[6], steroidogeneses pathways and immunological pathways. Over past decades, genome wide association studies (GWAS) which are considered as advanced approach to understand the implication of several gene variants have been attempted[7]. GWAS-identified risk variants (rs13429458 in THADA, rs10818854 in DENND1A, and rs2059807 in INSR) have been implicated in metabolic syndrome by a study among PCOS women[8]. Although no GWAS study for PCOS susceptibility have been reported on Indian population but some attempts have been made to validate GWAS findings among PCOS subjects in Indian population[9–11]. The present study focuses on identification of significant genetic variants involved in PCOS pathogenesis among North Indian population and their functional impact through in-silico analysis.

Study population

A total of 286 participants in case control study were included out of which 138 PCOS women and 148 healthy controls were present. The selection of PCOS patients was done based on the Rotterdam criteria having 2 out of 3 following characteristics: oligo/an-ovulation (having > 35 days menstrual cycle/ <8 menstrual cycles per year), hyperandrogenism (presence of acne or hirsutism), and transvaginal ultrasonography for appearance of polycystic ovarian morphology (presence of more than 12 follicles per ovary with diameter 2–9 mm and/or ovarian volume > 10 ml). Women with hyper/hypothyroidism and hypo/hyperprolactinemia was excluded. Control group women were healthy women who had regular menstrual cycle. Controls were recruited during field surveys and from hospital who were mostly companion of patients, and who were without pregnancy or having history of any related metabolic disorders such as diabetes, hypo/hyperthyroidism, adrenal hyperplasia, Cushing’s syndrome, hyperprolactinemia. Women with infertility or related complications were also excluded.

Sample collection and processing

The study is approved by Institutional ethics committee (IEC) ref number CUPB/CC/RO/18/2323 dated 26 December 2018. We confirm that all experiments were performed in accordance with relevant guidelines and regulations. After informed consent was signed by participants, 5ml peripheral venous blood sample was collected, out of which 2ml was preserved in non-coagulative EDTA coated vacutainer and 3ml of blood was taken in vacationer with clot activator for serum separation. Samples were collected from AIIMS Bathinda and Kamra Hospital, Malout.

Demographic and Biochemical analysis

Anthropometric variables (height and weight) were measured with measuring tape and electronic weighing machine. Body mass index (BMI), waist and hip circumference, waist hip ratio (WHR) was calculated using standard formulas. Serum samples were analysed for biochemical parameters including triglycerides, glucose, cholesterol, high density cholesterol (HDL-C), low density cholesterol (LDL-C), total protein, c-reactive protein (CRP), creatinine and albumin by colorimetry method using standard kits (ERBA) in automated clinical chemistry analyser (EM-200, Transasia) and following standard protocols. Different quality control standards (Erba norm or path) and calibrators (Multical, CRP calibrator and HDL calibrator) were used for calibration and quality assurance.

DNA extraction and Genotyping

Genomic DNA was extracted from whole blood using standard organic phenol-chloroform method. The purity and quality of isolated DNA was determined using Qubit assay and 0.8% gel electrophoresis stained with ethidium bromide dye. Genotyping was performed by using Infinium Global Screening Array v3.0 microchip (Illumina Inc.) covering 640k markers. Sample processing with 200ng of genomic DNA and hybridization was performed according to manufacturer protocol (Illumina).

QC and Data Analysis

Genotype cluster data was analysed and processed with Genome Studio. Internal quality control (QC) was performed based upon final sample call rate 99.95% and reproducibility 99.99%. Therefore, nine samples were removed having genotype rate per variant < 99%. Five related samples based upon pi hat cutoff 0.4 were excluded from further analysis. After filtering marker with HWE 1e^− 6 including 4, 38,203 markers among 272 included samples analysed for genotyping. The subsequent data was calculated for odds ratio and minor allele frequency using R version 4.2.3.

Statistical Analysis

Data was analysed by IBM SPSS statistics 22 software (SPSS Inc., Chicago, IL) and R vs 4.0. Comparison between categories was performed using chi –square test (χ²) and all continuous and non-parametric variables were calculated by independent-t test or Mann Whitney test based upon data distribution estimated using Kolmogorov-Smirnov test. Only weight and cholesterol level followed normal distribution. Anthropometric data and data on biochemical parameters were represented as mean ± SD. P < 0.05 was considered as statistically significant. The association between genotype and phenotype was analysed by logistic regression adjusted by age using PLINK software. Manhattan plot and quantile-quantile (Q-Q) plot were generated by R package. All GSA analysis considered statistically significant with p < 10^− 5.

Replication analysis

To assess the replication of previous reported variants in the current data, a comparison was performed with previous reported GWAS genotype data on the genetic risk of PCOS. Data was extracted from twenty-two PCOS GWAS studies^12–25 which include ninety-five associations (NHGRI-EBI Catalogue; EFO-0000660). After removing duplications and analysis of seventy-nine SNPs, twenty-one replicated SNPs were identified further. For other SNPs, which were not replicated, search for its proxy SNP was carried out. To search for functional SNPs being in Linkage disequilibrium (LD) with markers (r² ≥ 0.8) in Asian Punjabi population in Lahore (PJL) populations’ data was extracted from Ensembl database. There are sixteen replicated SNPs identified, of which one SNPs were significantly associated. Overall, only two SNPs were found to be significantly replicated in the current study. The protocol for identification of replicated SNPs is summarised in Fig. 1.

Functional analysis

Following in silico tools were utilized a) Gene ontology analysis: Enrichr tool (https://maayanlab.cloud/Enrichr/enrich) has been used to characterise gene set and product of gene set and to determine the quality of gene sets. It includes three divisions such as cellular component, molecular function and biological process represented b) Pathway analysis: KEGG, Reactome 2022, and BioPlanet 2019 databases have been used for gene enrichment and pathway analysis. c) Protein-protein interaction analysis: Cluster analysis for group of significant genes were analysed by String, and functional enrichment analysis were performed using GeneMANIA to study protein-protein interaction.

The characteristics of 272 age matched participants including 134 women with PCOS and 138 healthy subjects selected in the GSA analysis are represented in Table 1. There are significant differences of weight, BMI and WHR among cases and control. There are also significant differences of triglycerides, HDL-C, LDL-C, total protein, albumin and CRP (p < 0.05).

Table 1

Characteristics including anthropometric, demographic, and biochemical variables of studied participants
Variables	PCOS (n = 134)	Controls (n = 138)	p-value
Anthropometric characteristics
Age (years)	27.22 ± 5.37	27.27 ± 5.30	0.953
Height (cm)	158.37 ± 5.41	159.14 ± 6.87	0.097
Weight (kg)	62.68 ± 13.65	56.90 ± 10.80	0.000*
BMI (kg/m²)
Underweight	13	27	0.000*
Normal	37	57
Overweight	19	24
Obese I	43	21
Obese II	22	9
WHR
Normal	24	43	0.008*
Average risk	52	57
Elevated risk	58	38
Demographic characteristics
Marital status
Married	96	65	0.000*
Unmarried	38	73	0.000*
Occupation
Home Maker	88	30	0.000*
Student	26	42
Government employed	3	7
Private employed	14	55
Self employed	3	4
Education
Primary	11	6	0.000
Secondary	5	10
Higher Secondary	56	27
Graduate	41	39
Post Graduate	20	44
Other (higher)	1	12
Age at menarch
< 8	0	0	0.574
8–10	3	3
11–12	25	36
13–14	63	64
15–16	37	29
> 16	6	6
Biochemical variables
Triglyceride (mg/dl)	151.78 ± 91.37	121.09 ± 67.27	0.001*
Glucose (mg/dl)	80.11 ± 32.13	78.54 ± 25.48	0.557
Cholesterol (mg/dl)	174.79 ± 42.57	168.57 ± 35.07	0.097
HDL-C (mg/dl)	52.16 ± 15.41	58.62 ± 16.75	0.001*
LDL-C (mg/dl)	104.96 ± 30.57	93.26 ± 24.44	0.001*
Total protein (g/dl)	7.45 ± 0.72	7.26 ± 0.66	0.012*
Creatinine (mg/dl)	0.92 ± 0.25	0.92 ± 0.15	0.351
Albumin (g/dl)	4.51 ± 0.55	4.31 ± 0.48	0.001*
CRP (mg/l)^a	2.96 ± 3.30 (n = 122)	1.74 ± 2.68 (n = 102)	0.000*
Data presented as mean ± SD. *p-value < 0.05 value significant. p values calculated by Independent Student’s t test and chi square test. PCOS: Polycystic Ovary Syndrome, BMI: Body mass index, WHR: Waist hip ratio; HDL-C: High density cholesterol; LDL-C: Low density cholesterol; CRP: C-reactive protein
^a Few samples could be analysed for CRP

Association analysis between 15 SNPs variants and PCOS

There were fifteen SNPs which surpassed threshold genome wide significance p < 10^− 5 Details of allele frequency distribution along with odds ratio (OR) is illustrated in Supplementary Fig. 1a, Table 2. Out of these SNPs, ten SNPs are located within intronic region. Details include rs17724508 and rs36001636 located in the intronic region of gene MAF (MAF BZIP Transcription Factor). Three SNPs namely rs77876641, rs57799830 and rs1808529 are located in the intronic region of FGF18 (Fibroblast growth factor 18), PNPLA1 (Patatin like phospholipase domain containing 1) and ADAMTS16 (A disintegrate metallopeptidase With Thrombospondin Type 1 Motif 16) respectively. While rs4859756 is located at 33kb downstream CCNG2 (cyclin G2); rs5766088 at18kb upstream SHISAL1 (Shisa like 1); rs6491147 at 23kb downstream in gene GPR12 (G-protein coupled receptor 12 ); rs9977253 at 113kb nearby LINC00158 locus; and rs73163625 at 43kb downstream of RBM11 (RNA-binding protein 11) locus. Two variants were located nearby uncharacterised gene locus which are rs6857792, located at 24kb downstream to LOC107986292 locus and rs11750797 located at 183kb nearby LOC102467226 locus.

Table 2

The significant genetic variants associated with PCOS risk in the present study
SNP	Nearest gene	Chromosome location	ALT	MAF		OR (95% CI)	P-value
SNP	Nearest gene	Chromosome location	ALT	PCOS	Control	OR (95% CI)	P-value
rs17724508	MAF	chr16:79316307	C	0.3481	0.2036	2.61307 (1.6872, 4.0471)	1.68E-05
GSA-rs2575337	LINC02124_HS3ST6	chr16:1893417	A	0.4191	0.2518	2.49942 (1.6361, 3.8182)	2.26E-05
rs77876641	FGF18	chr5:171427753	C	0.0556	0.15	0.20046 (0.0953, 0.4217)	2.28E-05
rs36001636	MAF	chr16:79315909	C	0.3419	0.2036	2.54767 (1.6475, 3.9397)	2.61E-05
rs4859756	CCNG2	chr4:77203195	G	0.5478	0.3669	2.26434 (1.5416, 3.3260)	3.09E-05
rs1283168	NR4A1_ LOC12490234_TAMALIN	chr12:52020311	G	0.3088	0.4857	0.448199 (0.3062, 0.6559)	3.62E-05
GSA-rs57799830	PNPLA1	chr6:36274392	G	0.0809	0.2071	0.287897 (0.1589, 0.5215)	3.99E-05
GSA-rs1808529	ADAMTS16	chr5:5183587	G	0.2978	0.1571	2.62787 (1.6519, 4.1802)	4.51E-05
rs5766088	SHISAL1	chr22:44349814	C	0.2978	0.4393	0.441043 (0.2967, 0.6556)	5.18E-05
rs9977253	LINC00158	chr21:25272769	C	0.263	0.1286	2.73983 (0.2511, 1.6749)	5.97E-05
rs73163625	RBM11	chr21:14271908	T	0.1801	0.06071	3.75934 (0.3330, 1.9573)	6.98E-05
GSA-rs11750797	LOC102467226	chr5:121508701	C	0.2463	0.1107	2.78272 (1.6672, 4.6447)	9.02E-05
rs6491147	GPR12	chr13:26732414	C	0.2941	0.4357	0.454489 (0.3059, 0.6753)	9.51E-05
rs6857792	LOC107986292	chr4:77208160	G	0.3346	0.4856	0.471303 (0.3230, 0.6877)	9.54E-05
JHU_X.32325853		Chr23.32325853	A	0.4185	0.5714	0.441733 (0.2930, 0.6659)	9.59E-05
SNP: single nucleotide polymorphism; ALT: Altered allele, MAF: Minor allele frequency; MAF: MAF BZIP Transcription Factor; ; LINC: Long Intergenic Non-Protein Coding RNA; HS3ST6: Heparan sulfate-glucosamine 3-sulfotransferase 6; FGF18: Fibroblast growth factor 18; CCNG2: Cyclin G2; LOC: Uncharacterized; NR4A1: Nuclear Receptor Subfamily 4 Group A Member 1; TAMALIN: Trafficking regulator and scaffold protein tamalin; PNPLA1: Patatin like phospholipase domain containing 1; ADAMTS16: ADAM Metallopeptidase With Thrombospondin Type 1 Motif 16; SHISAL1: Shisa like 1; RBM11: RNA-binding protein 11; GPR12: G-protein coupled receptor 12; OR: Odds ratio; 95% CI: 95% confidence interval

Two significant genetic variants were intergenic, namely rs1283168 located at 2kb upstream of NR4A1 (Nuclear Receptor Subfamily 4 Group A Member 1) and 5kb downstream of TAMALIN (Trafficking regulator and scaffold protein lambda), and rs257533 located at 44kb downstream to LINC02124 (Long Intergenic Non-Protein Coding RNA 2124) and 180kb downstream to the gene HS3ST6 (Heparan sulfate-glucosamine 3-sulfotransferase 6. While, one variant is present on the X chromosome at location of X.32325853.

Nine variants found to be significantly associated with higher risk of PCOS as indicated by OR > 1. Six variants namely rs1283168, rs57799830, rs5766088, rs6857792, rs6491147 and X.32325853 are potentially protective genetic variant with OR < 1. To evaluate proportion of heterogeneity among samples, Q-Q plot was derived and data distribution was observed against expected distribution (Supplementary Fig. 1b).

The principal component analysis (PCA) score plot PC1 vs. PC2 displayed clear differentiation of studied GSA population from other ethnic groups while, GSA group completely overlapping with south asian group as indicated by colour differentiation (Fig. 2). Population overlapping showed evidence of genomic data conservation and sharing among these lineages. Percentage of variation was also explained by each of the PC axis.

Replication analysis between SNPs reported in previous GWAS and PJL population

A stratified search was performed among 22 previous GWAS studies reporting 95 SNPs associated with PCOS risk. After comparing the previous GWAS gene variants, 21 found to be replicated, out of which only one SNP rs17186366, 10kb downstream in uncharacterised gene locus LOC10192142 were significantly associated with PCOS in present study also. Locus LOC10192142 variant rs17186366 although had significant association, but not observed with increased ORs in studied population, while higher PCOS risk as reported in previous GWAS.

Previous GWAS gene variants which were not found in present studied markers were further analysed for their proxy SNPs under linkage disequilibrium considering r² < 0.08 to discover replication in Punjabi population in Lahore (PJL). The population was selected based upon ethnicity to know functional SNPs. Total 720 proxy SNPs for GWAS reported SNPs have been found in PJL GWAS data. Only 16 common SNPs were replicated in the present study, out of which one rs11171739 intergenic variant at 2kb upstream in nearby gene ERBB3 (Erb-B2 Receptor Tyrosine Kinase 3) on 12q13.2 chromosome significantly associated with PCOS with ORs > 1 with its alteration of C allele. Details of the thirty-seven SNPs identified during replication analysis of previous GWAS risk variants with PCOS and also observed in present study is summarized in Table 3.

Table 3

Replication analysis of previous GWAS risk variants with PCOS among thirty-seven SNPs identified in present study
Previous GWAS data reported						Present Study
GWAS SNP	SNP (PJL population)	Variant allele	Chromosomal location	ORs	p-value	ORs	p-value
rs13429458	-	C	2:43411699	1.49	4 x 10^− 13	1.3689	0.20234
rs2479106	-	G	9:123762933	1.35	5 x 10^− 10	1.4227	0.06055
rs13405728	-	G	2:48751020	1.35	4 x 10^− 9	0.9875	0.95989
rs2268361	-	T	2:48974473	1.15	1 x 10^− 12	0.7531	0.11403
rs3802457	-	A	9:94979054	1.3	5 x 10–¹⁴	0.6776	0.35633
rs1894116	-	G	11:102199908	1.27	1 x 10–²²	0.8076	0.35795
rs705702	-	G	12:55996852	1.27	9 x 10–²⁶	1.1132	0.63059
rs2272046	-	C	12:65830681	1.43	2 x 10^− 21	0.3759	0.12389
rs4784165	-	G	16:52313907	1.15	4 x 10^− 11	1.0066	0.97427
rs2059807	-	A	19:7166098	1.14	1 x 10^− 8	0.9207	0.67732
rs6022786	-	G	20:53830764	1.13	2 x 10^− 9	1.1659	0.39979
rs10818854	-	A	9:123684499	1.51	9 x 10^− 18	0.9182	0.78717
rs12478601	-	C	2:43494369	1.39	3 x 10^− 23	1.0746	0.68886
rs10505648	-	G	8:136144207	1.92	5 x 10^− 8	1.0324	0.86975
rs11031002	-	A	11:30193714	1.24	9 x 10^− 9	1.4764	0.10031
rs17186366	-	C	6:159477229	1.37	3 x 10^− 8	0.4611	0.01506
rs1784692	-	C	11:114078510	-	2 x 10^− 10	0.8176	0.38482
rs8050136	-	A	16:53782363	-	2 x 10^− 85	1.0720	0.71260
rs1509097	-	T	2:164881421	-	3 x 10^− 8	1.2128	0.42120
rs4135247	-	A	3:12355089	-	6 x 10^− 9	1.3162	0.13279
rs1265564	-	T	12:111270654	-	3 x 10^− 13	1.1325	0.82462
rs55762028	rs1976490	G	2:163323008	-	-	1.0960	0.84636
rs2178575	rs1351593	T	2:212529881	-	-	1.3984	0.14412
rs10938398	rs13130484	T	4:45173674	-	-	0.8357	0.32676
rs10938398	rs10938397	G	4:45180510	-	-	0.8074	0.24189
rs10993397	rs10821415	A	9:94951177	-	-	0.7803	0.19928
rs1355471	rs10843210	T	12:28575800	-	-	1.0660	0.72564
rs2271194	rs11171739	C	12:56076841	-	-	1.7588	0.00775
rs2271194	rs1131017	C	12:56042145	-	-	1.4560	0.06425
rs1795379	rs10879928	A	12:75531996	-	-	0.9596	0.82181
rs8043701	rs11075468	T	16:52344261	-	-	1.3762	0.11977
rs8043701	rs12918945	G	16:52325558	-	-	1.4103	0.09765
rs9930501	rs9941349	T	16:53791576	-	-	1.0894	0.64605
rs9930501	rs9922619	T	16:53797859	-	-	1.1028	0.58876
rs9930501	rs9930506	G	16:53796553	-	-	1.1028	0.58876
rs8047587	rs112198012	T	3:129352662	-	-	0.6726	0.28723
rs8047587	rs9940128	A	16:53766842	-	-	1.1062	0.58546
GWAS: Genome wide association study; SNP: Single nucleotide polymorphism; ORs: Odds ratio; -:Not available

Functional Analysis of Significant SNPs

Eleven significant gene variants (MAF rs17724508, rs36001636; HS3ST6 rs2575337; FGF18 rs77876641; CCNG2 rs4859756; ADAMTS16 rs1808529; NR4A1_TAMALIN rs1283168; PNPLA1 rs57799830; SHISAL1 rs5766088; RBM11 rs73163625 and GPR12 rs6491147) were further evaluated to determine its functional impact using in silico tools. Results by Enrichr are represented in Fig. 3a-c, according to which most of these genes are involved in cellular response to fibroblast growth factor and cell proliferation. Significantly associated pathways are represented in Fig. 3d-f. PPI network analysis indicate 41 nodes and 256 edges, and the protein interaction was enriched (P < 1.0e − 16), with an average local clustering coefficient of 0.687 (Fig. 3g). Functional analysis of 11 selected genes was done using GeneMANIA which suggested that the selected genes are significantly associated with 20 genes. A total of 31 genes network showing gene-gene association was represented in Fig. 3h. The thickness of edges between two nodes shows interaction strength between associated genes was significant among all enlisted genes and highly significant weighted by 0.044 score shared between NR4A1 and VPS25 genes. GeneMANIA (Fig. 3h) represents list of gene-gene interaction networks weighted by their ability to associate its related gene and, its phenotype predicted by multiple association network integration algorithms. Thickness of edges between two nodes shows interaction strength of gene and it’s associated other gene or phenotype.

PCOS associated genetic variants can vary among subpopulations due to different ethnicity. As per literature review, current study is the first attempt to identify associated variants among north Indian by GWAS. Fifteen SNPs are suggested to be significantly associated with PCOS development risk. Out of which ten SNPs are in the intronic region of gene (MAF, FGF18, CCNG2, SHISAL1, PNPLA1, ADAMTS16, GPR12, LINC00158, RBM11), two are in intergenic region (LINC02124-HS3ST6 and NR4A1_ LOC12490234_TAMALIN) and two SNPs are located in nearby uncharacterised gene (LOC107986292 and LOC102467226) region.

GWAS studies reported earlier had identified PCOS susceptible multiple loci among European[15] and Chinese[21, 26]. The significantly associated SNPS were checked in the current data and two significant SNPs are being suggested further.

MAF gene acts as transcription factor and is expressed in adipose tissue and pancreas and plays major role in food uptake, energy consumption and metabolism[27]. Deregulation and deficiency of this gene expression in adipose tissue leads to decreased expression of AIM which is associated with obesity development[28]. Increased FGF18 gene in stromal cell induced by estrogen, further stimulate endometrial stromal cells (ESCs) through ERK and Akt signalling[29]. Elevated FGF18 act as proapoptotic signal to inhibit follicular growth and cause follicular atresia determined in experimental model[30]. HS3ST6 encode enzyme heparan sulfate 3-o- sulfotransferase, responsible for the formation of 3-o- sulfated heparan sulfate required during anticoagulant expression. Modulation in its structure cause inhibition of heapin synthesis and its binding with glycosaminoglycan present on endothelial surface is helpful for inflammation response [31, 32]. PCOS women with amenorrhea condition had significant higher endometrial concentration of heparan sulphate, also associated with BMI[33]. CCNG2 exerts its metabolic function through decreased expression indicated by glucose in visceral adipose tissue and had inverse correlation with insulin resistance among obese patients[34]. In vitro study suggests down regulation of liver specific gene IDE accompanied by decreased CCNG2 expression among patients with type 2 diabetes mellitus (T2DM)[35]. NR4A1 which is differentially expressed in ovarian granulosa cells participate in androgen receptor (AR) signalling through binding to AR response element NGFI-B[36]. Androgen excess or hyperandrogenism is considered as underlying cause of many clinical complications among women with PCOS. ADAMTS1 is located on chromosome 21q21.3, a secretory protein expressed in various organs according to its receptor specificity of extracellular matrix to perform biological functions[37]. Its increased expression was observed in primary granulosa cells to affect oocyte maturation and estrogen synthesis in PCOS subjects[38]. Long intergenic non coding RNA is associated with cell proliferation, migration and epithelial mesenchymal transition by regulating other transcription factors expression such as LINGO2, NFIA, OTUD7B, SEC22C, and VAMP3[39]. Therefore, it is also considered as oncogene. In the ovarian granulosa cells, EMT is major process to induce ovulation and it has been reported that α-1,3/ 1,6-mannosyltransferase (ALG2) inhibit the Wnt/ β- catenin signaling pathway for ultimately inhibition of EMT among PCOS patients[40]. RBM11 gene represented as member of RNA binding protein has functional impact in ovarian cancer development and invasion through the activation of Akt/mTOR signaling pathway[41]. Several G-protein coupled receptors (GPCRs) have been reported among adipose tissues, but GPR12 detected only in white adipose tissue in a vitro study and its significant expression in adipocytes to impact obesity RNA sequence and adipose metabolism observed[42]. Another study performed in model organism to understand its mechanism evidenced that deficiency of GPR12 responsible for development of high body mass index, obesity and low energy expenditure which is valuable for its therapeutic target in PCOS patients[43]. GPR12 RNA also expressed in human ovary and maintain oocyte meiotic arrest by providing cAMP source and gap junction which is regulated by other GPRs consecutive signalling[44].

Obesity, insulin resistance, energy metabolism, dyslipidaemia are common metabolic complexities observed among affected women. Beside the wide spectrum of genetic variants found in present study, seven genes which have been discussed above are involved in pathways associated with oligo-ovulation, follicular atresia. Therefore, those genetic variants must be further evaluated to understand their qualitative role in PCOS trait.

Strength of the study includes association study and replication study in the North Indian population based on GWAS for the first time. Although the study involves smaller sample size but the results indicates that high throughput technology and unhindered sharing of data among researchers can significantly reduce cost of such studies.

In the summary, study suggest that fifteen gene variants are significantly associated with PCOS risk among the north Indian population. Two variants which had been reported earlier in GWAS study carried out on populations of different associated have been replicated. Gene variants such as MAF rs17724508, rs36001636, CCNG2 rs4859756, LOC102467226 rs11750797, RBM11 rs73163625, LINC00158 rs9977253, ADAMTS16 rs1808529 significantly increased risk of PCOS susceptibility (OR > 2) while PNPLA1 rs57799830, SHISAL1 rs5766088, FGF18 rs77876641, GPR12 rs6491147 and LOC107986292 rs6857792 are significant genetic protective marker for PCOS development among Indian population.

Funding

The authors would like to thank the financial support for this work by ICMR New Delhi [54/08/2019/HUM-BMS, Project id 2019-1384].

Author contributions

P.S.: Protocol development, data collection, data analysis and manuscript writing. L.D.G., and B.K. and P Kamra sample collection, subject enrollment and manuscript writing, S.S. Protocol development, data analysis, supervised the study and analysis. P.K. Project development, protocol development, supervised the study and analysis, manuscript writing and fund acquisition. All the authors and reviewed the manuscript.

Availability of materials and data

The datasets generated and analysed during the current study are available with corresponding author on reasonable request.

Competing interests
The authors declare no competing interests.

Consent to participants

Informed consent was obtained from all the included participants in the present study.

Sanchez-Garrido MA, Tena-Sempere M (2020) Metabolic dysfunction in polycystic ovary syndrome: Pathogenic role of androgen excess and potential therapeutic strategies. Mol Metab 35:100937. https://doi.org/10.1016/j.molmet.2020.01.001
Deswal R, Narwal V, Dang A, Pundir CS (2020) The Prevalence of Polycystic Ovary Syndrome: A Brief Systematic Review. J Hum Reprod Sci 13:261–271. https://doi.org/10.4103/jhrs.JHRS_95_18
Bani Mohammad M, Majdi Seghinsara A (2017) Polycystic Ovary Syndrome (PCOS), Diagnostic Criteria, and AMH. Asian Pac J Cancer Prev 18:17–21. https://doi.org/10.22034/APJCP.2017.18.1.17
Gilbert EW, Tay CT, Hiam DS et al (2018) Comorbidities and complications of polycystic ovary syndrome: An overview of systematic reviews. Clin Endocrinol (Oxf) 89:683–699. https://doi.org/10.1111/cen.13828
Bhandary P, Shetty PK, Manjeera L, Patil P (2022) Hormonal, genetic, epigenetic and environmental aspects of polycystic ovarian syndrome. Gene Rep 29:101698. https://doi.org/10.1016/j.genrep.2022.101698
Sharma P, Bilkhiwal N, Chaturvedi P et al (2021) Potential environmental toxicant exposure, metabolizing gene variants and risk of PCOS-A systematic review. Reprod Toxicol 103:124–132. https://doi.org/10.1016/j.reprotox.2021.06.005
Day F, Karaderi T, Jones MR et al (2018) Large-scale genome-wide meta-analysis of polycystic ovary syndrome suggests shared genetic architecture for different diagnosis criteria. PLoS Genet 14:e1007813. https://doi.org/10.1371/journal.pgen.1007813
Tian Y, Li J, Su S et al (2020) PCOS-GWAS Susceptibility Variants in THADA, INSR, TOX3, and DENND1A Are Associated With Metabolic Syndrome or Insulin Resistance in Women With PCOS. Front Endocrinol (Lausanne) 11:274. https://doi.org/10.3389/fendo.2020.00274
Dakshinamoorthy J, Jain PR, Ramamoorthy T et al (2020) Association of GWAS identified INSR variants (rs2059807 & rs1799817) with polycystic ovarian syndrome in Indian women. Int J Biol Macromol 144:663–670. https://doi.org/10.1016/j.ijbiomac.2019.10.235
Irgam K, Reddy BM, Annamaneni S, Rayabarapu P (2019) The genetic susceptibility profile of the South Indian women with polycystic ovary syndrome and the universality of the lack of association of type 2 diabetes genes. Gene 701:113–120. https://doi.org/10.1016/j.gene.2019.03.042
Dadachanji R, Sawant D, Patil A, Mukherjee S (2021) Replication study of THADA rs13429458 variant with PCOS susceptibility and its related traits in Indian women. Gynecol Endocrinol 37:716–720. https://doi.org/10.1080/09513590.2021.1906854
Zhang Y, Ho K, Keaton JM et al (2020) A genome-wide association study of polycystic ovary syndrome identified from electronic health records. Am J Obstet Gynecol 223. https://doi.org/10.1016/j.ajog.2020.04.004. :559.e1-559.e21
Day FR, Hinds DA, Tung JY et al (2015) Causal mechanisms and balancing selection inferred from genetic associations with polycystic ovary syndrome. Nat Commun 6:8464. https://doi.org/10.1038/ncomms9464
Lee H, Oh J-Y, Sung Y-A et al (2015) Genome-wide association study identified new susceptibility loci for polycystic ovary syndrome. Hum Reprod 30:723–731. https://doi.org/10.1093/humrep/deu352
Dapas M, Lin FTJ, Nadkarni GN et al (2020) Distinct subtypes of polycystic ovary syndrome with novel genetic associations: An unsupervised, phenotypic clustering analysis. PLoS Med 17:e1003132. https://doi.org/10.1371/journal.pmed.1003132
Day F, Karaderi T, Jones MR et al (2018) Large-scale genome-wide meta-analysis of polycystic ovary syndrome suggests shared genetic architecture for different diagnosis criteria. PLoS Genet 14:e1007813. https://doi.org/10.1371/journal.pgen.1007813
Chen Z-J, Zhao H, He L et al (2011) Genome-wide association study identifies susceptibility loci for polycystic ovary syndrome on chromosome 2p16.3, 2p21 and 9q33.3. Nat Genet 43:55–59. https://doi.org/10.1038/ng.732
Hwang J-Y, Lee E-J, Jin Go M et al (2012) Genome-wide association study identifies GYS2 as a novel genetic factor for polycystic ovary syndrome through obesity-related condition. J Hum Genet 57:660–664. https://doi.org/10.1038/jhg.2012.92
Shi Y, Zhao H, Shi Y et al (2012) Genome-wide association study identifies eight new risk loci for polycystic ovary syndrome. Nat Genet 44:1020–1025. https://doi.org/10.1038/ng.2384
Sh K, Hs ML J, S P (2019) High Genetic Risk Scores of ASIC2, MACROD2, CHRM3, and C2orf83 Genetic Variants Associated with Polycystic Ovary Syndrome Impair Insulin Sensitivity and Interact with Energy Intake in Korean Women. Gynecol Obstet Invest 84. https://doi.org/10.1159/000493131
Chen L, Hu L-M, Wang Y-F et al (2017) Genome-wide association study for SNPs associated with PCOS in human patients. Exp Ther Med 14:4896–4900. https://doi.org/10.3892/etm.2017.5113
Mg H, M U, Da E et al (2015) Genome-wide association of polycystic ovary syndrome implicates alterations in gonadotropin secretion in European ancestry populations. Nat Commun 6. https://doi.org/10.1038/ncomms8502
Jd B, Ah L, A M, et al (2021) Exome sequencing and analysis of 454,787 UK Biobank participants. Nature 599. https://doi.org/10.1038/s41586-021-04103-z
Tyrmi JS, Arffman RK, Pujol-Gualdo N et al (2022) Leveraging Northern European population history: novel low-frequency variants for polycystic ovary syndrome. Hum Reprod 37:352–365. https://doi.org/10.1093/humrep/deab250
Liu Q, Tang B, Zhu Z et al (2022) A genome-wide cross-trait analysis identifies shared loci and causal relationships of type 2 diabetes and glycaemic traits with polycystic ovary syndrome. Diabetologia 65:1483–1494. https://doi.org/10.1007/s00125-022-05746-x
Hong S, Hong YS, Jeong K et al (2020) Relationship between the characteristic traits of polycystic ovary syndrome and susceptibility genes. Sci Rep 10:10479. https://doi.org/10.1038/s41598-020-66633-2
Haller S, Spiegler J, Hemmelmann C et al (2013) Polymorphisms in FTO and MAF Genes and Birth Weight, BMI, Ponderal Index, Weight Gain in a Large Cohort of Infants with a Birth Weight below 1500 Grams. PLoS ONE 8:e66331. https://doi.org/10.1371/journal.pone.0066331
Tran MTN, Hamada M, Nakamura M et al (2016) MafB deficiency accelerates the development of obesity in mice. FEBS Open Bio 6:540–547. https://doi.org/10.1002/2211-5463.12058
Wu J, Tao X, Zhang H et al (2020) Estrogen-Induced Stromal FGF18 Promotes Proliferation and Invasion of Endometrial Carcinoma Cells Through ERK and Akt Signaling. Cancer Manag Res 12:6767–6777. https://doi.org/10.2147/CMAR.S254242
Portela VM, Dirandeh E, Guerrero-Netro HM et al (2015) The Role of Fibroblast Growth Factor-18 in Follicular Atresia in Cattle1. Biol Reprod 92. https://doi.org/10.1095/biolreprod.114.121376
Miyata T, Horiuchi T (2023) Biochemistry, molecular genetics, and clinical aspects of hereditary angioedema with and without C1 inhibitor deficiency. Allergology Int 72:375–384. https://doi.org/10.1016/j.alit.2023.04.004
Farrugia BL, Lord MS, Melrose J, Whitelock JM (2018) The Role of Heparan Sulfate in Inflammation, and the Development of Biomimetics as Anti-Inflammatory Strategies. J Histochem Cytochem 66:321. https://doi.org/10.1369/0022155417740881
Giordano MV, Giordano LA, Gomes RCT et al (2015) The evaluation of endometrial sulfate glycosaminoglycans in women with polycystic ovary syndrome. Gynecol Endocrinol 31:278–281. https://doi.org/10.3109/09513590.2014.989980
Garrido-Sánchez L, Roca-Rodríguez MDM, Fernández-Veledo S et al (2014) CCNG2 and CDK4 is associated with insulin resistance in adipose tissue. Surg Obes Relat Dis 10:691–696. https://doi.org/10.1016/j.soard.2013.12.011
Pivovarova O, von Loeffelholz C, Ilkavets I et al (2015) Modulation of insulin degrading enzyme activity and liver cell proliferation. Cell Cycle 14:2293–2300. https://doi.org/10.1080/15384101.2015.1046647
Dai A, Yan G, He Q et al (2012) Orphan Nuclear Receptor Nur77 Regulates Androgen Receptor Gene Expression in Mouse Ovary. PLoS ONE 7:e39950. https://doi.org/10.1371/journal.pone.0039950
Porter S, Clark IM, Kevorkian L, Edwards DR (2005) The ADAMTS metalloproteinases. Biochem J 386:15–27. https://doi.org/10.1042/BJ20040424
G Y, G Y, Z X et al (2021) Expression Level of ADAMTS1 in Granulosa Cells of PCOS Patients Is Related to Granulosa Cell Function, Oocyte Quality, and Embryo Development. Front cell Dev biology 9. https://doi.org/10.3389/fcell.2021.647522
Barbagallo C, Caltabiano R, Broggi G et al (2020) LncRNA LINC00518 Acts as an Oncogene in Uveal Melanoma by Regulating an RNA-Based Network. https://doi.org/10.3390/cancers12123867. Cancers 12:
Zheng Q, Liu M, Fu J (2022) ALG2 inhibits the epithelial-to-mesenchymal transition and stemness of ovarian granulosa cells through the Wnt/β-catenin signaling pathway in polycystic ovary syndrome. Reprod Biol 22:100706. https://doi.org/10.1016/j.repbio.2022.100706
Fu C, Yuan M, Sun J et al (2021) RNA-Binding Motif Protein 11 (RBM11) Serves as a Prognostic Biomarker and Promotes Ovarian Cancer Progression. https://doi.org/10.1155/2021/3037337. Disease Markers 2021:
Al Mahri S, Okla M, Rashid M et al (2023) Profiling of G-Protein Coupled Receptors in Adipose Tissue and Differentiating Adipocytes Offers a Translational Resource for Obesity/Metabolic Research. Cells 12:377. https://doi.org/10.3390/cells12030377
Bjursell M, Gerdin A-K, Jönsson M et al (2006) G protein-coupled receptor 12 deficiency results in dyslipidemia and obesity in mice. Biochem Biophys Res Commun 348:359–366. https://doi.org/10.1016/j.bbrc.2006.07.090
DiLuigi A, Weitzman VN, Pace MC et al (2008) Meiotic Arrest in Human Oocytes Is Maintained by a Gs Signaling Pathway1. Biol Reprod 78:667–672. https://doi.org/10.1095/biolreprod.107.066019

SupplementaryFig1.jpg
Supplementary figure 1. a) Manhattan plot of results of GWAS analysis for PCOS. 17 genome wide significant index variants located above the vertical blue line representing threshold for genome wide significance p < 10 ^-5 on -log10 scale. Y axis represents log (two-sided p-value) of associated SNPs with PCOS risk; b) Q-Q plot from heterogeneity test among all the samples in PCOS GWAS analysis. Observed vs expected GWAS p values on -log10 scale are shown.

Download PDF

Journal Publication

published 29 Feb, 2024

Read the published version in Archives of Gynecology and Obstetrics →

Reviewers agreed at journal
27 Nov, 2023
Reviewers invited by journal
24 Nov, 2023
Editor invited by journal
15 Nov, 2023
Editor assigned by journal
08 Nov, 2023
First submitted to journal
07 Nov, 2023

You are reading this latest preprint version

Genome-Wide association study (GWAS) identified PCOS susceptibility variants and replicates reported risk variants

Status:

Journal Publication

Version 1

Abstract

Figures

What this study adds to clinical work

Introduction

Materials and methods

Study population

Sample collection and processing

Demographic and Biochemical analysis

DNA extraction and Genotyping

QC and Data Analysis

Statistical Analysis

Replication analysis

Functional analysis

Results

Association analysis between 15 SNPs variants and PCOS

Replication analysis between SNPs reported in previous GWAS and PJL population

Functional Analysis of Significant SNPs

Discussion

Conclusions

Declarations

References

Supplementary Files

Status:

Journal Publication

Version 1