It was reported that AT is equipped with a steroidogenic machinery for de novo synthesis from cholesterol, initiated by STAR and CYP11A1 [3]. Not only the total amount of steroids within AT was found to be 40–400 times greater than in plasma, but there also was a positive gradient between tissue and plasma [7]. It was suggested that both STAR and CYP11A1 are rate-limiting factors for steroidogenesis as they produce crucial precursors [3, 18]. Previous studies have shown the association between CYP11A1 and PCOS [19]. For the first time in 2008, MacKenzie et al. identified STAR and CYP11A1 gene expression in paired visceral and subcutaneous AT from the lower abdominal region were taken from eight women undergoing caesarean section [12]. Likewise, Wang et al. (2012) showed expression of CYP11A1 in abdominal subcutaneous AT taken from non-pregnant women [7]. Although Alizadeh et al. reported that CYP11A1 mRNA was not detectable in retroperitoneal and subcutaneous dairy cattle AT [20], they showed that the amount of STAR mRNA in subcutaneous AT on the day of delivery was 3-fold higher than before calving. The up-regulation of genes involved in initial steps of steriodogenes on the day of delivery in PCOS mothers indicate an increase capacity for cholesterol uptake to the inner mitochondrial membrane and steroidogenesis stimulation. These data provide evidence that PCOS women may have more functional and effective sources of steroid metabolism pathways in subcutaneous AT on the day of delivery than age- and BMI-matched non-PCOS women.
CYP17A1 is a key branch point in steroid biosynthesis, driving the pathway to the direction of either mineralocorticoid and glucocorticoid production or sex steroid metabolism [21]. Wickenheisser et al. reported that CYP17A1 mRNA was more abundant in ovarian theca cells of PCOS than non-PCOS women [21, 22]. In the current study, CYP17A1 was transcribed higher in subcutaneous AT of PCOS than non-PCOS women. MacKenzie et al did not detect this gene in subcutaneous AT on the day of delivery, and neither did Valle et al in non-pregnant women and men [12, 23]. However, the presence of CYP17A1 gene in PCOS women’s subcutaneous AT was reported by Wang et al in 2012. Uniquely, Kinoshita et al. using an innovative liquid chromatography–mass spectrometry (LC-MS/MS)-based method could detect CYP17A1 at the protein level [24]. A potential for up-regulation of CYP17A1 expression in AT was proposed, which may contribute to hyperandrogenism in women suffering from PCOS [25–27]. Therefore, our data for CYP17A1 support the hypothesis [27] that confirms the overexpression of this gene that plays a crucial role in PCOS pregnant women, which may be related to the regulation of glucocorticoid as well as mineralocorticoids metabolism .
Among fourteen genes that we detected and compared among both groups, relative expression of CYP21 was 4-fold higher in non-PCOS than in PCOS. Similarly, Azziz et al. [28] and Witchel and Aston [29] showed the deficiency of 21-hydroxylase in hirsute women with PCOS. One of the main steps in adrenal and ovarian steroidogenesis is the conversion of 17-hydroxyprogesterone into 11-deoxycortisol, which is catalyzed by the 21-hydroxylase enzyme encoded by CYP21. The deficiency of this enzyme, which is inherited by an autosomal recessive trait, is responsible for most cases of congenital adrenal hyperplasia and increased serum 17-hydroxyprogesterone levels are correlated with its deficiency. Furthermore, patients suffering from heterozygote CYP21 mutations as well as clinical symptoms exhibit PCOS-like phenotype [30]. Therefore, the changes in amounts of CYP21 mRNA support the above reasoning.
Our data confirmed the findings of previous studies that HSD11B1 and HSD11B2 mRNA are present in both subjects and are also more in subcutaneous AT of PCOS than non-PCOS .In the same study, 42 women with PCOS with a broad range of BMI were studied and it was shown that in these women, there was increased HSD11B1 enzyme activity with increasing central fat distribution [31]. This could indicate impaired HSD11B1 activity in women with PCOS compared with control.
Due to cortisol peak and critical roles of cortisol around delivery day, more attention must be paid to cortisol metabolism. Previously, the expression of HSD11B1 in human AT, particularly in visceral fat, has been confirmed [32]. Makenzie et al suggested high expression of HSD11B1 gene in both subcutaneous and visceral AT on the day of delivery [12]. Based on the evidence for HSD11B1 in human AT and bovine AT from different visceral and subcutaneous depots, AT likely affects glucocorticoid metabolism with consequences for both endocrine as well as auto/paracrine glucocorticoid effects locally within AT, the latter being supported by the presence of glucocorticoid receptors, also demonstrated in bovine AT [12, 33]. In this regard, Alizadeh et al (2016 b) studied the presence of HSD11B1 mRNA in dairy cattle AT around calving day. They showed that the highest expression of HSD11B1 in bovine AT was shown on the day of delivery rather the days before and after calving [34]. Therefore, the elevation of cortisol metabolism around parturition is normal; but our study clearly showed 1.5-fold elevation of HSD11B1 mRNA levels in PCOS over non-PCOS. Similarly, Corto´n et al. reported that non-pregnant PCOS women had up regulation of HSD11B1 expression in abdominal subcutaneous AT compared to non-PCOS ones, which is consistent with our data [35]. Although the elevation of HSD11B1 in obesity has been reported, it was surprising to note that the alarming increase in BMI-matched subject incidence of AT importance is associated with an array of metabolic pathologies, including PCOS.
The presence of CYP19 (aromatase) mRNA in human AT has already been well established by Corbould et al [36] and Makenzi et al [12]. Although it has been proposed that PCOS may result from reduced aromatase activity and subcutaneous AT have higher expression of Cyp19a1 compared to visceral AT in morbidly obese men and premenopausal women [37], we cannot find difference in mRNA levels of CYP19 between PCOS and non-PCOS women. Previously, the conversion of androstenedione to estrone in subcutaneous AT from the lower body of women (i.e., thighs and buttocks) was shown to be higher than that of in the subcutaneous AT of upper body fat (i.e., breast and abdomen), along with the highest level of Cyp19a1 gene expression [38]. Altogether, the absence of changes in Cyp19a1 suggest that delivery day or sampling area may influence CYP19 expression than PCOS.
Enzymes with activities associated with various 17BHSD isoforms are widespread, not only in classic steroidogenic tissues such as the testis, ovary, and placenta but also in a large number of peripheral sites including AT [39]. The estrogenic isoforms of 17BHSD (types 1, 7 and 12) catalyze the conversion of estrone to estradiol [3] and they are similar in PCOS and non-PCOS groups in the current study. The 17BHSD family genes involved in the estrogenic process in AT are much fewer than those of the androgenic one, inferring that subcutaneous AT may serve as a dominant activation site for androgens than estrogens. Bellemare et al. (2009) suggested that type 12 17BHSD is probably involved in the local conversion of estrone into estradiol in differentiated adipocytes and it has more activity than other isoforms [40]. Uniquely, Quinkler et al showed that 17BHSD5 has pivotal roles in women androgen metabolism in subcutaneous buttock AT [11]. Therefore, we expected that 17BHSD5 and type 12 17BHSD may differ between experimental groups; but our findings did not support the hypotheses regarding different sex steroid related gene expression in subcutaneous AT.
It appears that sampling area (buttock AT vs. abdominal AT) may influence these results; but the mRNA levels of 17BHSD5 were 2-folds higher in abdominal subcutaneous AT of PCOS women compared to the non-PCOS women [7]. So, this result may affect the status on the day of delivery, as shown by Mackenzie et al. [12]. They suggested that de novo synthesis of sex steroids from cholesterol is not possible in pregnant women on the day of delivery because of the absence of key steroidogenic mRNAs in the AT.
Finally, the survey on PCOS mother and offspring health is an exciting and emerging area for research and clinical studies. Uniquely, Kosidou et al. [41] reported that children of mothers with PCOS have an increased risk of developing autism spectrum disorder and their findings support that early life androgen exposure may be important for the development of autism in both sexes which it was supported by Katsigianni et al [42]. These reports alongside our data suggest further studies are warranted in order to shed more light on of AT’s roles in PCOS mother.
Our study has a few limitations that must be considered when interpreting our results. AT samples are not easily available. Therefore, small sample size was the limitation of this study for age- and BMI- matched PCOS women. It would have been interesting to compare expression with non-pregnant women or across gestation and additional studies with larger sample sizes could confirm these results in pregnant and non- pregnant PCOS women alongside measured several hormone levels. Other factors such as fat distribution may be related to the differences observed.