The role of CYP19A1 and ESR2 gene polymorphisms in female androgenetic alopecia in the Polish population.

Introduction: Androgenetic alopecia is the most common type of non-cicatricial alopecia both in male and female patients. The mechanism that leads to hair loss is similar in both sexes, but the underlying cause, and especially the role of genes and sex hormones in the pathogenesis of the disease in women has not fully been explained as of yet. So far, a few attempts have been made to assess selected SNPs for CYP19A1 and ESR2 genes, but their results are not unequivocal and fully reproducible. Aim: To investigate the association of 13 CYP19A1 and 11 ESR2 gene SNPs with female androgenetic alopecia (FAGA) in a population of Polish patients, including some already genotyped SNPs of possible importance for FAGA pathophysiology in other populations. Material and methods: Twenty-four genetic polymorphisms were analysed for the ESR2 and CYP19A1 genes in 117 patients with FAGA and 128 healthy subjects treated at the Department of Dermatology in Krakow.
Results: In the studied Polish population, none of the selected SNPs, frequently detected in the Caucasian population and linked with the transformation pathway of sex hormones, showed a significant association with FAGA. Conclusions: Further studies into the genetic background of androgenetic alopecia are needed. Ethnic differences as well as the size of the studied population may be of great significance for the obtained results. Copyright:

Introduction

Androgenetic alopecia is the most common type of non-cicatricial alopecia both in female and male patients. In both cases, the mechanism that leads to hair miniaturization is similar and involves a shortened anagen phase, and prolonged telogen and kenogen phases [1, 2]. The disease is chronic and progressive, leading to a gradual and diffuse hair loss in the central, parietal, and frontal scalp regions.

Amongst the numerous variables, including hormonal factors, that may impact the hair growth cycle in female androgenetic alopecia (FAGA), the role of androgens and oestrogens in the development of the disease remains unclear, especially given the fact that hormone levels are usually within normal ranges in FAGA patients [3-5].

Apart from the synthesis in the gonads and other organs, androgens and oestrogens are independently produced in hair follicles, where they undergo mutual interactions, impacting gene transcription and expression [6].

Presently, more and more studies focus on the role of oestrogens and oestrogen receptors located in the skin and its structures, and in particular on ER-β that is prevalent in hair follicles [7].

Both the CYP19A1 and ESR2 genes are involved in the transformation of sex hormones.

CYP19A1 is a gene located on chromosome 151, responsible for the production of aromatase, which is involved in the conversion of androgens (testosterone and androstenedione) into oestrogens [8]. Aromatase, similarly to 17β-HSD and 5α-reductase, is an enzyme that is also located in the pilosebaceous units [9]. The higher levels of aromatase in the occipital regions, described by Sawaya et al., may be indicative of the protective role of oestrogens [10]. The ESR2 gene, located on chromosome 14, encodes the β subtype (ER β) of intracellular nuclear oestrogen receptors.

To date, only a few studies have been carried out on the Chinese, Australian and European (UK, Germany) populations looking into the possible links between SNPs in the CYP19A1 and ESR2 genes and FAGA in women [11-15]. Some of the studies indicated significant correlations only with respect to a few selected SNPs: rs4646, rs6493497 and rs7176005 for the CYP19A1 gene, and rs10137185, rs17101774 and rs2022748 for the ESR2 gene.

So far, no attempt has been made to carry out a similar study involving a population of Polish patients.

Aim

The aim of the study was to investigate single nucleotide polymorphisms (SNPs) in the oestrogen-related genes ESR2 and CYP19A1 in order to examine whether they might be susceptible genes for FAGA development. We tried to replicate some findings of other authors and search for new SNPs of possible importance.

Material and methods

Subjects

117 FAGA patients and 128 healthy subjects, treated at the Department of Dermatology in 2017–2019, were included in the study. The diagnosis of androgenetic alopecia was confirmed by 2 dermatology specialists based on the typical history of the disease, its clinical picture, trichoscopic and histological examination.

The trichoscopic criteria of FAGA in videodermoscopy were based on the presence of 2 major criteria or 1 major and 2 minor criteria. Major criteria included an increased number of yellow dots (more than 4 yellow dots in the frontal area in 4 fields of vision at 70-fold magnification) and thin hairs, decreased average hair thickness in the frontal area and more than 10% of thin hairs (< 0.03 mm) in the frontal area. Minor criteria included an increased frontal area to occiput ratio of single-hair units (> 2 : 1), vellus hairs (> 1.5 : 1) and follicles with perifollicular discoloration (> 3 : 1) [16].

Exclusion criteria included co-existing diseases and medicines that could influence the hair loss.

The assessment of alopecia intensity was based on the 5-stage Sinclair scale. Table 1 presents the number of patients with individual stages of disease progression.

Table 1

Number of patients with particular grades of hair loss according to the Sinclair scale

Sinclair scale	Number of FAGA patients (n = 117)
1	1
2	38
3	33
4	30
5	15

The patients’ average age was 49.2 in the FAGA group, and 47.8 in the control group. In the FAGA study arm, 35 patients started suffering from FAGA before the age of 40.

Approval for the study was obtained from the Bioethics Committee of the Jagiellonian University (decision no. 1072.6120.203.2017). All patients gave written consent for genetic studies.

SNP selection

SNP database “dbSNP” of the National Centre for Biotechnology Information was used to select the SNPs for analysis in the study [17]. The available polymorphisms, most frequently detected in the Caucasian population, with the genotype prevalence of at least 0.15, were eventually chosen for the study.

With respect to the CYP19A1 gene, 13 SNPs were selected, and with respect to the ESR2 gene, 11 SNPs were selected for analysis in the study (Table 2) [18-27]. The SNPs were selected based on literature data. 4 SNPs were chosen for analysis based on their significance demonstrated in other studies carried out to date (rs4646, rs6493497, rs7176005 and rs10137185).

Table 2

Data for the SNP positions under study

SNP	Gene	GRCh38	Allele variants	MAF	Role	Gene consequence	Assay ID
rs1152578	ESR2	chr14:64230319	T/C	T 0.424	Not reported	Intron variant	C___1436977_10
rs1256065	ESR2	chr14:64232214	G/T	G 0.431	Associated with bone mineral density [18]	Intron variant	C___1436975_10
rs8006145	ESR2	chr14:64232732	C/A	A 0.245	Prostate volume [19]	Intron variant	C__29383989_10
rs867443	ESR2	chr14:64234324	G/A	A 0.238	Not reported	Intron variant	C___1436972_20
rs17766755	ESR2	chr14:64249055	G/A	A 0.327	Benign prostatic hyperplasia [20]	Intron variant	C__34495232_10
rs4365213	ESR2	chr14:64253546	T/C	C 0.412	Alzheimer disease in women with Down syndrome [21]	Intron variant	C__32395442_20
rs6573549	ESR2	chr14:64254931	T/C	C 0.422	Not reported	Intron variant	C__32091355_10
rs61984409	ESR2	chr14:64263303	A/C	C 0.333	Not reported	Intron variant	C__90266873_10
rs7154455	ESR2	chr14:64269942	G/C	C 0.290	Not reported	Intron variant	C__29383994_10
rs960069	ESR2	chr14:64278284	C/T	C 0.443	Not reported	Intron variant	C___1436935_10
rs10137185	ESR2	chr14:64309058	C/T	T 0.100	Associated with FPHL [22]	Intron variant	C__29621308_10
rs934634	CYP19A1	chr15:51208341	C/T	T 0.190	Not reported	Intron variant, 3 prime UTR variant	C___8794656_10
rs2255192	CYP19A1	chr15:51208638	C/T	T 0.189	Not reported	Intron variant, 3 prime UTR variant	C__15798398_10
rs4275794	CYP19A1	chr15:51208920	T/C	C 0.190	Not reported	Intron variant, 3 prime UTR Variant	C__32394041_10
rs12148604	CYP19A1	chr15:51209207	C/T	C 0.424	Associated with sex hormone levels (estrone) [23]	Intron variant, 3 prime UTR variant	C__32071398_10
rs4646	CYP19A1	chr15:51210647	A/C	A 0.230	Associated with FPHL, premature ovarian failure [12, 24]	Intron variant, 3 prime UTR variant	C___8234730_1_
rs10046	CYP19A1	chr15:51210789	G/A	G 0.422	Associated with miscarriages [25]	Intron variant, 3 prime UTR variant	C___8234731_30
rs2899470	CYP19A1	chr15:51211480	T/G	T 0.348	Serum oestrogen and oestrogen/testosterone ratio [26]	Intron variant	C___8234732_10
rs12591172	CYP19A1	chr15:51211530	G/A	G 0.420	Not reported	Intron variant	C___8234742_10
rs8029120	CYP19A1	chr15:51212737	T/G	T 0.410	Not reported	Intron variant	C___8234756_10
rs749292	CYP19A1	chr15:51266534	G/A	A 0.500	Associated with ovarian cancer [27]	Intron variant	C___8801261_20
rs6493497	CYP19A1	chr15:51338638	G/A	A 0.092	Associated with FPHL [11]	Upstream variant	C__29374681_10
rs7176005	CYP19A1	chr15:51339082	C/T	T 0.092	Associated with FPHL [11]	Upstream variant	C_189237142_10
rs752760	CYP19A1	chr15:51339282	C/T	T 0.385	Not reported	Upstream variant	C____798312_10

MAF – minor allele frequency, SNP – single nucleotide polymorphism, FAGA – female androgenetic alopecia.

Polymorphisms rs12148604 and rs2899470 in the CYP19A1 gene are linked with the levels of sex hormones. The remaining SNPs have not been examined or described in the literature yet.

Extraction of DNA

Peripheral blood for molecular tests was collected in EDTA anticoagulant tubes. A 6% solution of high-molecular weight dextran (Dextran T500, Pharmacia) was added to the whole blood for erythrocyte sedimentation. Leukocyte-rich plasma was drawn off and centrifuged for 10 min/20000 x g. DNA was isolated from white blood cells with the Chomczynski and Sacchi method (DNAzol, Gibco). 0.4 ml of the DNAzol reagent was added to the centrifuged sediment and shaken. Following the dissolution of leukocyte cellular membranes, DNA was precipitated with 0.2 ml of 95% ethyl alcohol. Afterwards, the isolated DNA was washed in 70% ethanol, and dried in an incubator at the temperature of 37°C. Later on, the DNA was solubilized in redistilled water, and in that form used for further studies.

Real-time PCR

The isolated DNA was amplified and genotyped with the use of the real-time polymerase chain reaction (real-time PCR) method.

The real-time PCR reaction was carried out on 96-well plates in the 7900 HT Real-Time PCR System (Applied Biosystems). A standard reaction mix included polymerase and substrate solutions; TaqMan Universal PCR Master Mix, starters, probes, water, and DNA based on the manufacturer’s recommendations (TaqMan Universal PCR Master Mix Protocol – Applied Biosystems). The probes and starters were purchased from Applied Biosystems in the form of ready-to-use TaqMan Gene Assays.

Statistical analysis

The study analysed 24 single nucleotide polymorphisms (SNP) located in the ESR2 and CYP19A1 genes (Table 2). Each genotype for 24 studied polymorphisms was coded in an additive way as 0, 1, or 2 considering the number of minor alleles. Genotype–phenotype association analyses were conducted using univariate binary logistic regression for FAGA defined as 1 (case) vs. 0 (control). Allelic odds ratios (ORs) with respective 95% confidence intervals (CIs) and p-values for the minor alleles classified in an additive manner were calculated (Table 3). Recessive and dominant modes of inheritance were also assessed where genotypes were coded as 0 and 1 vs. 2, and 0 vs. 1 and 2, respectively (Tables 4 and 5). Associations were tested using PS IMAGO PRO 6.0 (IBM SPSS Statistics 26).

Table 3

Results of association analysis between individual SNP positions and FAGA

SNP	Gene	OR (95% CI)	P-value
rs1152578	ESR2	1.092 (0.755–1.581)	0.640
rs1256065	ESR2	1.153 (0.799–1.665)	0.446
rs8006145	ESR2	0.883 (0.561–1.391)	0.592
rs867443	ESR2	0.941 (0.597–1.483)	0.793
rs17766755	ESR2	1.070 (0.713–1.607)	0.743
rs4365213	ESR2	0.992 (0.685–1.436)	0.967
rs6573549	ESR2	1.007 (0.699–1.452)	0.969
rs61984409	ESR2	0.966 (0.648–1.441)	0.865
rs7154455	ESR2	0.870 (0.563–1.346)	0.533
rs960069	ESR2	1.124 (0.776–1.628)	0.535
rs10137185	ESR2	0.892 (0.502–1.584)	0.697
rs934634	CYP19A1	1.151 (0.728–1.820)	0.547
rs2255192	CYP19A1	1.073 (0.678–1.698)	0.763
rs4275794	CYP19A1	1.032 (0.653–1.630)	0.894
rs12148604	CYP19A1	0.917 (0.636–1.323)	0.645
rs4646	CYP19A1	0.860 (0.554–1.335)	0.501
rs10046	CYP19A1	0.970 (0.671–1.401)	0.870
rs2899470	CYP19A1	0.980 (0.660–1.455)	0.919
rs12591172	CYP19A1	0.947 (0.654–1.371)	0.771
rs8029120	CYP19A1	0.893 (0.607–1.314)	0.566
rs749292	CYP19A1	1.171 (0.824–1.663)	0.378
rs6493497	CYP19A1	1.396 (0.762–2.558)	0.280
rs7176005	CYP19A1	1.404 (0.766–2.574)	0.272
rs752760	CYP19A1	0.987 (0.678–1.438)	0.947

CI – confidence interval, OR – odds ratio, SNP – single nucleotide polymorphism.

Table 4

Results of association analysis between individual SNP positions and FAGA for a recessive model of allele classification

SNP	Gene	OR (95% CI)	P-value
rs1152578	ESR2	0.833 (0.424–1.638)	0.597
rs1256065	ESR2	0.948 (0.490–1.836)	0.875
rs8006145	ESR2	0.439 (0.083–2.311)	0.331
rs867443	ESR2	0.444 (0.084–2.340)	0.338
rs17766755	ESR2	1.013 (0.395–2.600)	0.978
rs4365213	ESR2	1.090 (0.544–2.184)	0.808
rs6573549	ESR2	1.048 (0.535–2.053)	0.890
rs61984409	ESR2	0.826 (0.334–2.047)	0.680
rs7154455	ESR2	0.411 (0.105–1.605)	0.201
rs960069	ESR2	0.904 (0.469–1.742)	0.762
rs10137185	ESR2	0.350 (0.036–3.417)	0.366
rs934634	CYP19A1	1.844 (0.430–7.909)	0.410
rs2255192	CYP19A1	1.860 (0.434–7.981)	0.404
rs4275794	CYP19A1	1.844 (0.430–7.909)	0.410
rs12148604	CYP19A1	0.797 (0.408–1.560)	0.509
rs4646	CYP19A1	0.734 (0.202–2.673)	0.640
rs10046	CYP19A1	0.939 (0.479–1.840)	0.854
rs2899470	CYP19A1	0.960 (0.410–2.248)	0.924
rs12591172	CYP19A1	0.887 (0.449–1.754)	0.730
rs8029120	CYP19A1	0.648 (0.308–1.366)	0.254
rs749292	CYP19A1	1.082 (0.610–1.921)	0.787
rs6493497	CYP19A1	2.390 (0.212–26.988)	0.481
rs7176005	CYP19A1	2.390 (0.212–26.988)	0.481
rs752760	CYP19A1	1.621 (0.771–3.410)	0.203

CI – confidence interval, OR – odds ratio, SNP – single nucleotide polymorphism.

Table 5

Results of association analysis between individual SNP positions and FAGA for a dominant model of allele classification

SNP	Gene	OR (95% CI)	P-value
rs1152578	ESR2	1.361 (0.790–2.346)	0.267
rs1256065	ESR2	1.421 (0.822–2.456)	0.208
rs8006145	ESR2	0.935 (0.565–1.547)	0.793
rs867443	ESR2	1.010 (0.609–1.674)	0.971
rs17766755	ESR2	1.108 (0.666–1.843)	0.692
rs4365213	ESR2	0.935 (0.549–1.594)	0.806
rs6573549	ESR2	0.985 (0.576–1.685)	0.957
rs61984409	ESR2	1.005 (0.604–1.671)	0.986
rs7154455	ESR2	0.954 (0.576–1.579)	0.855
rs960069	ESR2	1.399 (0.801–2.442)	0.238
rs10137185	ESR2	0.957 (0.500–1.831)	0.894
rs934634	CYP19A1	1.109 (0.654–1.878)	0.702
rs2255192	CYP19A1	1.008 (0.594–1.712)	0.975
rs4275794	CYP19A1	0.958 (0.566–1.624)	0.874
rs12148604	CYP19A1	0.961 (0.561–1.646)	0.885
rs4646	CYP19A1	0.856 (0.513–1.428)	0.552
rs10046	CYP19A1	0.975 (0.570–1.668)	0.926
rs2899470	CYP19A1	0.980 (0.587–1.639)	0.940
rs12591172	CYP19A1	0.960 (0.561–1.644)	0.881
rs8029120	CYP19A1	1.012 (0.590–1.735)	0.965
rs749292	CYP19A1	1.416 (0.792–2.529)	0.240
rs6493497	CYP19A1	1.401 (0.719–2.733)	0.322
rs7176005	CYP19A1	1.410 (0.723–2.750)	0.313
rs752760	CYP19A1	0.765 (0.454–1.292)	0.317

CI – confidence interval, OR – odds ratio, SNP – single nucleotide polymorphism.

Results

Univariate association analyses did not show statistical significance for any of the DNA variants of the ESR2 and CYP19A1 genes in the studied sample population (Table 3). No significant effects were also noted for a recessive and dominant manner of minor allele classification (Tables 4 and 5).

Discussion

Many hormones, including the thyroid, adrenal, pineal and pituitary gland hormones, impact the hair growth cycle [28]. Oestrogens are usually believed to have protective qualities [3, 29, 30], even though some studies indicate that hair growth is slowed under their influence due to the initiation of an earlier catagen phase, and prolongation of telogen phase [31].

With respect to the CYP19A1 gene, Yip et al. demonstrated a statistically significant incidence of the CC genotype for polymorphism rs4646 in their study carried out in Australia [12]. Johansson et al. described higher oestrogen levels in patients with that genotype, what could substantiate the hypothesis that the hair growth cycle is inhibited under the influence of oestrogens [32].

Redler et al. failed to confirm the findings of Yip et al. in a study involving the German and British populations [13]. In a study of Rui et al. involving the Chinese population, a statistically significant difference was observed with respect to the distribution of alleles in polymorphisms rs6493497 and rs7176005, but no correlation was found for rs4646 either [11].

As regards the ESR2 gene, Yip et al. indicated the significance of rs10137185, rs17101774, and rs2022748 polymorphisms in the Australian population [14].

Even though initially Redler et al. did not observe any significant correlations in the German and British populations, they did confirm the significance of rs10137185 in a later study involving the German population only. The authors pointed to potentially important variables such as the number of patients included in the study and a later stage of alopecia progression [22].

That proposition is confirmed by the results obtained in a study conducted by Yip et al., which involved the largest group of subjects (484 FAGA patients and 471 control subjects) [12].

In our study, we searched for correlations that would be true for the entire group of patients, taking into account the severity grade of alopecia, however the small number of patients with advanced grade of hair loss might have impacted the final results, leading to a lack of agreement with Redler’s study results.

The number of patients included in our study was sufficient to assess allele frequency in the Polish population, but it might be too small to reach the threshold of statistical significance.

Conclusions

Numerous studies carried out so far on FAGA etiopathogenesis have not been successful at explicating the role of genes or sex hormones in the development of the disease.

Our study showed no association of CYP19A1 and ESR2 genes with FAGA in the Polish population, although some of the SNPs included in our research showed a statistically significant incidence in Australian, Chinese and German populations.

As the small size of the studied group could be the reason for no findings regarding the selected SNPs, it prompts us to conclude that further research is needed that would include more patients.

It is also possible that another variant in the CYP19A1 and ESR2 gene is associated with FAGA in the Polish population.

Due to the suspected possibility of communication and co-dependence of receptor pathways for steroid hormones like oestrogens and androgens, it is also necessary to look into the functioning of hormones, and a potential impact of genes on hormonal activity [33, 34].

Undoubtedly, the high costs incurred in the course of studies remain one of the limitations involved.

Conflict of interest

The authors declare no conflict of interest.