Genotype based Risk Predictors for Polycystic Ovary Syndrome in Western Saudi Arabia

Polycystic ovary syndrome (PCOS) is the most common endocrine disease among premenopausal women. The genetic risk of PCOS in the Saudi population is still unclear. Therefore, it is of interest to study the genotype and allele frequency for six gene variants (THADA rs13429458, TOX3 rs4784165, FSHR rs2268361, YAP1 rs1894116, RAB5B rs705702, and HMGA2 rs2272046) in patients with PCOS in western Saudi population. The study included 95 PCOS patients and 94 normal ovulatory females as controls. Genotyping was performed using TaqMan™ real-time polymerase chain reaction assays. There was significant link between the THADA rs13429458 variant and PCOS. Homozygosity in allele A of the rs13429458 variant was correlated with hyperandrogenism (HA) risk. Homozygosity in the T allele of the FSHR rs2268361 variant was associated with normal levels of AMH among non-PCOS women. The THADA rs13429458 and TOX3 rs4784165 variants were significantly associated with the combined oligo/amenorrhea (OA) and polycystic ovarian morphology subgroups while the HMGA2 rs2272046 variant was significantly associated with the combined HA and OA subgroup. Thus, results show the genetic risk of the THADA rs13429458, TOX3 rs4784165, and HMGA2 rs2272046 variants on PCOS patients in the western Saudi population.


Background:
The diagnostic criteria for polycystic ovary syndrome (PCOS), a complex endocrine disorder affecting reproductive-aged women, have evolved since the disorder was first recognized. The National Institutes of Health first defined PCOS as the presence of clinical or biochemical hyperandrogenism (HA) comorbid with oligo/amenorrhea (OA) [1]. The Rotterdam consensus added the polycystic ovarian morphology (PCOM) phenotype, and the diagnosis was redefined as the presence of two out of the three conditions [2]. The Androgen Excess Society subsequently considered HA a key component in PCOS diagnosis [3]. The Rotterdam criteria was endorsed by an Endocrine Society clinical practice guideline [4]. According to the diagnostic criteria, the prevalence of PCOS varies worldwide but is generally 6-20% [5][6][7]. Although there are no prevalence studies that include the entire kingdom of Saudi Arabia, a study conducted in the city of Madinah found a prevalence of 32.5% [8]. Associated with significant multiple clinical manifestations including reproductive, metabolic, and psychological disorders [9][10][11][12][13][14][15], PCOS represents 80% of anovulatory infertility cases [10], and 80-85% of women with clinical HA have PCOS [16,17]. Manifestation of the disorder varies depending upon the particular diagnostic criteria. Patients diagnosed according to the Rotterdam and NIH criteria are at higher risk of developing reproductive and metabolic disorders such as infertility and type-2 diabetes [18 -21]. The etiology of PCOS is not entirely clear; however, the disease is primarily attributed to multiple genetic and environmental factors aggravated by obesity [22].

813
©Biomedical Informatics (2019) , and written informed consent was obtained from participants prior to sample collection. The study was conducted in accordance with the Declaration of Helsinki. The PCOS patients were divided into four subgroups according to their clinical symptoms. Each subgroup was tested individually to investigate whether the associations between SNPs and PCOS were absolute or relative to combined symptoms. After classification of the patients into subgroups, the sample size was 82 as 13 patients were excluded to avoid misleading results as the third symptom (HA) was not investigated (Table 1).

Statistical analysis:
Data analysis was conducted using the IBM SPSS software version 24 (SPSS TM Inc., NY, USA). The participants clinical characteristics were expressed as median ± inter quartile range (IQR), and p-values were calculated using the Mann-Whitney test as the data were nonnormally distributed. The genotype and allele data were expressed as frequencies. The differences between study groups were analyzed using the chi-squared test. Multinomial logistic regression was used to examine the association of the variants with PCOS clinical variables. Values of p < 0.05 were considered statistically significant.

Results:
All participant clinical parameters are listed in Table 2.

Allele and genotype frequency:
The genotype distribution and allele frequencies of the six SNPs are listed in Table 3. There was significant relationship between PCOS and THADA rs13429458 (p = 0.033), but no link was detected with the other genetic variants.

The association of the six variants with PCOS clinical characteristics
There was a significant relationship between THADA variant rs13429458 and HA phenotype in the PCOS group (p = 0.031, Table  4) by the chi-squared test. Multinomial logistic regression revealed that the AA genotype in THADA variant rs13429458 was positively correlated with higher frequency of HA than the AC genotype (OR = 2.9, B = 1.053, p = 0.026). Therefore, homozygosity in allele A in rs13429458 variant is predicted as a risk genotype associated with HA. THADArs13429458 and TOX3rs4784165 variants are correlated to the OA+PCOM subgroup The PCOS group was divided into four subgroups according to the clinical symptoms. There was significant correlation between the THADA rs13429458 and TOX3 rs4784165 SNPs and the OA+PCOM subgroup (p = 0.009, p = 0.028 respectively, Table 6).

HMGA2rs2272046 variant is correlated to HA+OA subgroup
The HMGA2variant rs2272046 showed a significant correlation with the HA+OA subgroup (p = 0.043, Table 6). No other correlations were detected among other variants within PCOS subgroups.

Discussion
In The thyroid adenoma-associated protein encoded by the THADA gene is expressed in many organs [33].One GWAS reported the association of THADA with type 2 diabetes particularly through a probability effect on pancreatic beta-cell function [50]. As a result, such a protein would be expected to affect various body processes, not unlike PCOS, which is characterized by dysfunction in multiple organ systems. In the present study, it was correlated in PCOS women with the HA phenotype. This may provide clues to the role of rs13429458 in the etiology of PCOS, as HA is one of the clinical symptoms of PCOS. Previously, the AA genotype for rs13429458 in THADA was detected in different phenotypes to be associated with increased LH, testosterone levels, and the LH/FSH ratio in subjects with PCOS [51]. Moreover, the genotype frequency distribution of rs13429458 was not influenced by hirsutism or increased metabolic parameters, including fasting glucose and insulin level [48].
It was demonstrated that the AMH receptors expressed in the brain may be involved in initiation of gonadotropin-releasing hormone (GnRH) release from hypothalamic neurons [52]. GnRH

Conclusion:
We report the link between the THADA rs13429458 gene variant and PCOS in western population. We also document the link of THADA rs13429458 and TOX3 rs4784165 variants with combined OA and PCOM phenotype of PCOS patients. It is further noted that the HMGA2 rs2272046 variant is linked with combined HA and PCOM phenotypes. These observations should be further verified using large GWAS to delineate the polygenic risk in PCOS among Saudi population.