Relationship of digit ratio with sexual steroid hormone receptor related genes - single nucleotide polymorphisms in a sample from Northern China

Abstract Background Digit ratio, especially 2D:4D, is hypothesised as a potential biological marker of exposure to intrauterine sex hormones. The aim of this study was to investigate the association between 10 SNPs of sex steroid hormone receptor (SSHR) related genes and 2D:4D. Subjects and methods 814 college students were randomly selected as research participants. After taking pictures of both hands of the participants, Image Pro Plus (IPP) software was used to measure 2D:4D. ESR1 (rs2228480 and rs3798758), ESR2 (rs944459, rs8006145, rs928554, and rs8018687), GPER1 (rs10269151 and rs12702047), and PGR (rs1042839 and rs500760) were genotyped using multiplex PCR. Results Females had significantly higher 2D:4D in both hands than male students (p < 0.05), and the R2D:4D of the Han population was significantly higher than that of the Hui population (p < 0.05). The number of females carrying the GPER1 G allele of rs12702047 was significantly higher than that of males (p < 0.05). The L2D:4D in males was significantly different in rs1042839, and the R2D:4D in the Han ethnicity was significantly different in rs3798758. Logistic regression analysis showed that rs12702047 was significantly associated with 2D:4D in both hands (p < 0.05). Conclusions GPER1 rs12702047 may be involved in the formation of digit ratio by affecting phalanx development in the Chinese population.


Introduction
The digit ratio refers to the ratio of the lengths of different digits or fingers on a hand, with the ratio of the index finger to the ring finger (2D:4D) being the most extensively studied.The fact that the digit ratio of females is significantly higher than that of males in different populations is hypothesised as a potential biological marker of intrauterine sex hormone exposure levels before birth (Phelps 1952;Manning et al. 1998;Hönekopp and Watson 2010) .Although 2D:4D varies across ethnic groups, European, Non-Chinese Asians, and Middle Easterners always exhibit higher ratios than Chinese and 'Black' participants; nevertheless, all samples consistently demonstrate higher 2D:4D in females than in males (Manning et al. 2007).
Studies have indicated that the digit ratio is established during the early stages of embryo development, and its formation and final determination are closely linked to the levels of androgen and oestrogen in utero during pregnancy.However, the findings are still contentious (Berenbaum et al. 2009;Zheng and Cohn 2011;Barrett et al. 2021).Presently, a number of studies have validated that 2D:4D is negatively associated with prenatal androgen levels and positively associated with oestrogen levels.Nonetheless, previous research on the subject has produced conflicting results (Hönekopp and Watson 2010;Richards et al. 2020;2022;Banyeh et al. 2023).Once established, the digit ratio exhibits considerable stability and is largely unaffected by sex hormone levels during adolescence and adulthood (Manning and Fink 2018;Butovskaya et al. 2021).
The conclusion has been confirmed in multiple populations.Some studies have shown that human 2D:4D is associated with various sex-related physical anthropological characteristics, such as muscle strength, grip strength, menarche, and penis length (Oberg and Villamor 2012;Schulter et al. 2012;Zhao et al. 2012;Halil et al. 2013).Other studies have found abnormal 2D:4D related to several sex hormone-related diseases, such as azoospermia, prostate cancer, mental disorders, and some metabolic syndromes (Swaddle 2002;Fink et al. 2006;Malas et al. 2006;Tegin et al. 2019).Interestingly, Okten et al. reported that the digit ratio in women with congenital adrenal hyperplasia exhibited a masculine trend, with a lower 2D:4D than that of normal controls (Okten et al. 2002).However, other similar studies have not consistently found these effects (Richards et al. 2020).In contrast, Berenbaum et al. found that 2D:4D in males with androgen insensitivity syndrome showed a feminine tendency, being higher than that of the control group (Berenbaum et al. 2009).These findings were later confirmed, especially in the right hand, but the conclusions are not stable, suggesting that the variation in digit ratio reflects not only changes in prenatal androgen action but also other factors (van Hemmen et al. 2017).Although the accuracy of these results is still open to question, and further verification of different genetic backgrounds is needed, they provide us with some information that the digit ratio is related to the androgen/oestrogen ratio in the body.Given the effects of embryonic steroids on the development of digit ratio and the Developmental Origins of Health and Disease (DOHaD), it is reasonable to believe that 2D:4D may be one of the potential markers of adult diseases caused by exposure to sex steroid hormones in embryonic life (Williams et al. 2000).
Although many studies have suggested that digit ratio formation may depend on intrauterine sex hormone levels during pregnancy, direct evidence is still lacking.Since sex hormone levels during pregnancy are dynamic and influenced by multiple factors, it is not clear which developmental stage is involved in the formation of 2D:4D.Zheng and Cohn's research provided some insights into this question.They found that the gestational age of 14.5 days may be a critical window for the formation of 2D:4D in mice.At this time, male foetal mice have high expression of androgen receptors (AR), while female foetal mice have high expression of oestrogen receptors (ESR), and the expression level of AR in the 4th finger is significantly higher than that in 2nd finger.This may be the basis for the formation of lower 2D:4D in males compared to higher 2D:4D in females.Zheng and Cohn's work suggested that 2D:4D is determined by the balance of prenatal testosterone (PT) and prenatal oestrogen (PE) signals during the narrow window of foetal finger development, specifically the fourth finger (Zheng and Cohn 2011).
For genes related to sexual steroid hormone receptors (SSHRs), the association between AR gene polymorphism and 2D:4D has been studied the most.Manning's initial study showed a positive correlation between the number of CAG repeats in the first exon of the AR gene and male right 2D:4D (Manning et al. 2003).However, two later studies did not replicate these results (Hönekopp 2013;Voracek 2014).Recently, Pearce et al. found a significant positive correlation between AR gene rs6152 and female left 2D:4D (Pearce et al. 2018).Few studies have examined the correlation between ESR and digit ratio, but Vaillancourt et al. found a significant association between the ESR1 gene TA repeat polymorphism and male left 2D:4D (Vaillancourt et al. 2012), and Nishimura et al. found a significant association between ESR1 gene rs9340799 and 2D:4D in male children in Hokkaido, Japan (Nishimura et al. 2019).Genes involved in sex hormone synthesis and metabolism have also been found to be significantly related to digit ratio.For example, we found that the T allele of CYP19A1 rs4775936 is related to an increase in Dr-l (right minus left 2D:4D) in females, while the rs2486758 C allele of the CYP17A1 gene is associated with an increased 2D:4D in the male left hand.Additionally, the rs1004467 G allele is associated with a decreased 2D:4D in the male right hand (Zhanbing et al. 2019;Zhang et al. 2021).
Although some studies have investigated the association between SSHR gene polymorphisms and 2D:4D, limitations in the selection of genes and SNPs, as well as differences in population genetic backgrounds, have prevented a full understanding of the biological mechanism behind digit ratio.Given the crucial role of SSHR in regulating sex hormone levels, it may play a role in mediating the involvement of sex hormones in the formation of 2D:4D.Therefore, it is necessary to expand the research on SSHR SNPs in different populations to further explore their association with digit ratio.This study aims to confirm the contribution of sex steroid hormones to 2D:4D by analysing the correlation between 10 SNPs of 4 sex hormone receptor-related genes and digit ratio among college students in Northern China.

Subjects
First, sample labels (school number) were randomly selected from the study population through SPSS, and then individual finger length ratio and genotyping data were obtained based on the labels.Finally, a total of 814 unrelated Chinese college students, consisting of 410 boys and 404 girls, were enrolled from Ningxia Medical University located in the Ningxia province of China.Of the participants, 401 were Hui, 383 were Han, and the remaining 30 belonged to other ethnic groups (Table 1).All participants had ancestors living in Ningxia for at least three generations and were healthy, with no hand disability or fine structure loss.
Written informed consent was obtained from all participants, and the study received approval from the Ningxia Medical University Ethics Committee (No. 2019-057).All hand photos, peripheral blood, and DNA samples collected from the participants were kept in the Ningxia Medical University Sample Bank of Human Genetic Resources for proper storage.We guarantee that no information or study data about the participants was or will be shared with any third party.

Selection of polymorphisms
Based on a review of relevant literature, we selected five sex steroid hormone receptor-related genes as the research objects, which were AR (androgen receptor), ESR1 (oestrogen receptor 1), ESR2 (oestrogen receptor 2), GPER1 (G protein-coupled oestrogen receptor 1), and PGR (progesterone receptor).
SNPs were screened using Haploview 4.2 software, and the selected loci were required to meet the following criteria: (1) r 2 ≥0.8 in the preliminary data from HapMap using Haploview; (2) minor allele frequency (MAF)≥0.05(Xing et al. 2008); and (3) SNPs were reported or clinically significant.A total of 12 SNPs of 5 genes were initially selected for the study.However, during the genotyping process, the detection rates of rs12394583 and rs9332969 of the AR gene did not meet the requirements and were therefore excluded from the study.Finally, we selected 10 SNPs from 4 genes.The information and location of the associated SNPs are listed in Table 2.

Genotyping
3 mL of peripheral blood were collected in tubes coated with EDTA-Na2.DNA was extracted strictly in accordance with the instructions of whole blood cell DNA Extraction Kit (DP304 TIANamp Genomic DNA Kit, TIANGEN Biotech Co. Ltd., Beijing, China), and stored at −20 °C.Multiple polymerase chain reaction (PCR) technique (Novogene Biotech Co., Ltd., Beijing, China) was used for genotyping, and PCR primer information of each SNP site is shown in Table 3.
The PCR procedure was carried out in two rounds.The first round was predenaturation at 95 °C for 15 min, 94 °C for 30 s, 60 °C for 10 min and 72 °C for 30 s, 4 cycles in total; then, 94 °C for 30 s, 60 °C for 1 min and 72 °C for 30 s, a total of 24 cycles.The second round of PCR: predenaturation at 95 °C for 15 min; 94 °C, 30 s; 60 °C, 4 min; 72 °C, 30 s, 5 cycles in total; 94 °C, 30 s; 65 °C, 1 min; 72 °C, 30 s, a total of 10 cycles.After PCR, the amplified products were mixed in a centrifuge tube and shaken overnight.The PCR products were purified using a CA2 absorption column.The purified PCR products were loaded into an Illumina X-10 sequencer for sequencing.Illumina RTA software was used for quality control and raw data analysis.

Digit ratio collection
The data on digit ratio are generally obtained through indirect measurement, such as scanning or taking photos of the palm and then analysing the measurements.Common measurement methods include the vernier calliper measurement method, computer vision feature extraction method (Manning et al. 2005), and the digital measurement method based on image software, such as Image-Pro Plus and Photoshop (Zhanbing et al. 2019).However, due to the significant measurement errors in the first two methods, this study uses the measurement methods previously used by our group.
Participants were asked to place their hands on a flat table surface with their fingers straight and their palms facing up.Photos of the hands were taken with a digital camera (DSCW520; SONY, Tokyo, Japan) mounted on a tripod, maintaining a constant 35 cm distance of the camera lens for all participants.We measured the distance of all 10 fingers on both hands from the tip of the finger to the midpoint of the fold closest to the palm.Therefore, individuals with abnormal or damaged fingers, such as ectrodactylia, polydactylism, syndactylia, swollen knuckles, finger contracture deformity, etc., were excluded.Next, the hand photos were imported into a computer, and Image-Pro Plus 6.0 (Media Cybernetics, Inc., USA) was used to measure the digit lengths of both hands (from the fingertip to the middle point of the most proximal crease to the palm).Each digit's measurement was performed three times by three different individuals who were well-trained in the procedure, and the mean value was used for further analysis.Finally, we calculated the average 2D:4D and Dr-l for each hand, which is the absolute value of the right-hand 2D:4D minus the left-hand 2D:4D.

Bioinformatics analysis
The human function SNPs were annotated using RegulomeDB (http://regulomedb.org)and 3DSNP 2.0 (https://omic.tech/3dsnpv2/).RegulomeDB classifies SNPs based on the presence/absence of functional categories, such as transcription factor binding sites, DNAase hypersensitivity regions, and promoter regions, and assigns a RegulomeDB score (ranging from 1 to 6) to indicate their potential function.In 3DSNP 2.0, SNPs are scored based on six functional categories, including 3D interacting genes, enhancer state, promoter state, transcription factor binding sites, altered sequence motifs, and conservation score, using a quantitative scoring system to measure their function.The potential function of promising variants and their related SNPs (r 2 >0.80) in Asia from the 1000 Genomes Project (Siva 2008) was queried using HaploReg v4.3 (Ward and Kellis 2016).The evolutionary conservation of mammalian species was determined using PhyloP scores (Hubisz et al. 2011), which were calculated by Phylopbase data from GENCODE genes.The Evolutionary Conservation Analysis Line Chart shows the PhyloP scores of 46 vertebrates and 33 mammals across the ± 10 bp region surrounding the SNP.

Statistical analysis
Statistical Package for The Social Sciences (SPSS) 22.0 software (IBM, Armonk, NY, USA) was used for statistical analysis.Haploview 4.2 software (Broad Institute of Harvard and MIT, MA, USA) and SNPStats (https://snpstats.net/)online software were used for descriptive analysis, single-SNP analysis, haplotype, and linkage disequilibrium analysis.Age and ethics were used as covariates to correct for the logistic analysis between left and right-digit ratios with 10 SNPs.Three independent measurements, each repeated 3 times, were performed to evaluate the reliability of digit ratio measurement, the consistency of digit ratio measurement was then determined using intraclass correlation coefficient (ICC) tests with type of the mean of three measurements.T-tests were used for comparing two groups, and One-way ANOVA was used for the comparison of differences for more than 2 groups.
The Bonferroni method was used to correct the error detection rate when the group whose column proportions differed significantly from each other at the 0.05 level by selecting the dbSNP wild-type loci as the reference group.Fisher's exact test was used to evaluate the genotype and allele frequency distribution between groups.The statistical test was set at α = 0.05, a 2-tailed test, and p < 0.05 was considered statistically significant.Using the SNPstats (Solé et al. 2006) web software, regression analysis was used to establish the association between positive SNPs and digit ratios by ANOVA.All 2D:4D data are presented as mean ± standard deviation (SD).The genotypic association characteristics of 2D:4D were visualised using a box scatter composite map drawn by R-4.1.0.

Description and distribution of 10 SNPs of sex steroid hormone receptor related genes in our population
The distribution of genotype frequencies for 10 SNPs was analysed using a Chi-square test, and the results were consistent with the Hardy-Weinberg equilibrium in our population (p > 0.05, Table 2), indicating that the subjects were from the same population and had good representation.After applying the Bonferroni correction, the results showed that only the genotype frequency of GPER1-rs12702047 was significantly different between males and females (χ2 = 5.924, p = 0.040).The frequency of GG genotype in females was significantly higher than that in males (Table 4).There was also a significant difference in allele frequency between males and females (χ2 = 4.437, p = 0.035), with males carrying the A allele of rs12702047 significantly more than females.However, there was no significant difference between the other nine SNPs in different sexes (p > 0.05).
The Han and Hui groups are mainly mixed in Ningxia Hui Autonomous Region, China.Therefore, we analysed the distribution of genotypes and alleles for the 10 SNPs in these two ethnic groups and found no significant difference between them (p > 0.05).

Haplotype analysis
The haplotypes of the 10 SNPs were analysed using Haploview 4.2 software.Good linkage disequilibrium was found among rs8006145-rs8018687-rs928554-rs944459 in both different sexes and different ethnic groups (Figure 1).Haplotypes were then constructed among these four SNPs.The statistical results showed that there was no significant difference in the   five main haplotypes between male and female college students or between Hui and Han ethnic groups in Ningxia (Table 5) (p > 0.05).

Digit ratio (2D:4D) in our Chinese college population
The measurement results of digit ratio (2D:4D) are consistent with statistical requirements (ICC Left =0.978, F = 44.620,p < 0.001; ICC Right =0.976, F = 42.484,p < 0.001).Table 6 shows the 2D:4D distribution results of the left and right hands among different sexes in our population.The results indicate that, regardless of the left hand (L) or right hand (R), male 2D:4D was significantly lower than that of females (p < 0.05), which is consistent with previous reports.However, the difference in digit ratio between the right and left hands (D r-l ) between males and females showed no significant difference (p > 0.05).In an inter-ethnic comparison, we found a significant difference in the 2D:4D of the right hand between the Hui and Han ethnic groups.The right-hand 2D:4D (R 2D:4D ) of the Han ethnic group was significantly higher than that of the Hui ethnic group (p = 0.006).

Difference analysis of 10 SNPs of sex steroid hormone receptor related genes to 2D: 4D in our Chinese college population
Table 7 shows the difference analysis results of 10 SNPs and the digit ratio between males and females.In the total population, all three different genotypes of GPER1 rs12702047 were significantly different in both left (χ2 = 3.325, p = 0.038) and right 2D:4D (χ2 = 3.076, p = 0.050).Individuals with rs12702047 AG genotype had significantly higher left and right 2D:4D than those with AA and GG genotype.In addition, there was a significant correlation between PGR rs1042839 and left 2D:4D in the male population, and the left 2D:4D in the GG genotype was always significantly lower than that in the AG genotype (χ2 = 4.772, p = 0.029).
In comparison between ethnic groups, we also found a difference in rs12702047 and 2D:4D, but this difference only occurred in the total population, which also showed that the 2D:4D of individuals carrying AG genotype was significantly higher than that of individuals carrying AA or GG genotype (χ2 = 3.253, p = 0.044) (Table 8).Additionally, we found that the 2D:4D of the right hand in the rs3798758 (ESR1) genotype was significantly different in the Han ethnic group.For rs3798758, 2D:4D in the CC genotype was significantly higher than in those carrying AC and AA genotypes (χ2 = 3.764, p = 0.028).
Furthermore, we used R language to draw boxplot superimposed scatter plots to show the correlation between different genotypes of 10 SNPs and 2D:4D both in sexes and ethnic groups.The groups with statistical differences of GPER1 rs12702047 in different sexes are shown in Figure 2(A), while Figure 2(B) shows the statistical differences in different ethnic groups.Both-hand 2D:4D for different ethnic groups for different genotypes of rs3798758 (Figure S1-A) and rs1042839 (Figure S1-B) are shown in Supplementary Figure S1.
Logistic regression analysis was conducted to further clarify whether these two different loci were correlated with digit ratio (Table 9).As rs1042839 lacks the AA genotype, a logistic regression model cannot be constructed.The results were adjusted for age and sex in different genetic models, and only the genotype of rs12702047 had a significant correlation to both right (p = 0.031) and left digit ratio (p = 0.039) in the codominant and recessive model (right p = 0.011; left p = 0.037).

Function prediction of 12702047 through bioinformatics
Through previous studies, we found that the distribution of genotypes and alleles of rs12702047 showed significant sex differences, and there was a significant correlation between rs12702047 and digit ratio in the total population.Therefore, we next wanted to investigate the potential biological function of rs12702047 and further explore the possible mechanism of its influence on 2D:4D.A detailed bioinformatics analysis of this locus was performed.
We first performed an interspecies conservation analysis of rs12702047 based on Phylop scores for 46 vertebrates and 33 mammals, which were computed across the ±10bp region surrounding the SNP.The results showed that rs12702407 is located in the evolutionarily conserved region and has high conservation scores (Phylop Score=-3.125),which suggested that rs12702407 is highly conserved in evolution (Figure 3(A)).Next, we used HaploReg4.1 to explore the potential function of rs12702407 in gene expression regulation.Through the analysis, the results showed that rs12702407 is located within the histone markers of the promoter and enhancer regions, and it can influence 7 bound proteins, including PAX5C20, CEBPB, HDAC2, HEY1, HNF4A, HNF4G, and POL2.It may also change the regulatory motifs of chromodomain DNA helicase protein 2 (CHD2) (Supplementary Table 1).Furthermore, the HaploReg prediction revealed that rs12702407 falls into the DNase peak region and the region of proteins bound, including PAX5C20, CEBPB, HDAC2, HEY1, HNF4A, HNF4G and POL2, which are crucial regulators of mammalian gene expression (Figure 3(B)).Additionally, using the 3DSNP databases and eQTL analysis, we showed that rs12702407 could increase GPER expression in human muscle and skeletal tissue (Figure 3(C)).
By cross-prediction and a combination of miRanda and Sanger microRNA databases, we found that rs12702407 may be the binding site of miRNA, which may affect gene expression by altering the binding of has-miR-744 (Supplementary Table 2).We found that the binding sequence region is ATCCCCAAAGCA/GGCCCCGCA.Rs12702407 does not change the binding score of has-miR-744 but can affect the binding energy value of has-miR-744.The binding energy value of the A allele is Δ = −23.94,and that of the G allele is Δ = −25.99,which means that when A changes to G, the binding ability of rs12702407 to has-miR-744 is enhanced.

Discussion
The digit ratio has been hypothesised to reflect individual differences in exposure to androgen/oestrogen levels during intrauterine development and early embryonic ontogeny.Despite continuing controversy and a lack of in vivo experiment results, many studies still believe that the digit ratio can be used as a marker to reflect physiological dysfunction of individuals to a certain extent (Manning and Fink 2018;Manning et al. 2019;Zhang et al. 2020).Therefore, the study of digit ratio has become a hot topic in biological anthropology, developmental biology, psychology, and clinical medicine.Although many studies have confirmed that the digit ratio, especially for 2D:4D, is significantly higher in females than in males in different populations, including the results of this study, no substantial progress has been made regarding the biological basis of digit ratio development in humans (Manning and Fink 2017;Richards 2017).
The formation of the digit ratio may begin at the 13th to 14th week of gestation, during which the level of intrauterine sex steroids secretion of the mother not only plays a key role in determining the outcome of the embryo and foetus but also may contribute to the formation of 2D:4D by influencing *p < 0.05; dbsnP wild-type loci as reference in the first column genotype of each snP, Bonferroni was used to correct p-values when there were significant differences between groups.Table 7. Continued.the development of the phalanges (Garn et al. 1975;Csathó et al. 2003).Therefore, the normal functions and interactions of sex steroid generation, transporters, and receptors are not only closely related to the normal development of embryos but also become key factors affecting the development of the digit ratio (M.McCarthy 2016).The results showed that 2D:4D of Ningxia college students was significantly lower in males than in females, which was consistent with the results of most similar studies reported.Although it is not enough to indicate that the development of the digit ratio is related to sex hormone exposure, it at least indicates that the digit ratio is an indicator of significant sex dimorphism (Hong et al. 2014;Lu et al. 2017;Wang L et al. 2018).
In 2011, Zheng et al. used a mouse model to study the specific gene expression patterns of digit ratio which have been regulated by AR and ESR.A total of 19 bone development-related genes were found to be significantly regulated by AR and/or ESR (Zheng and Cohn 2011).This study confirmed that the developing finger primordia is rich in androgen and oestrogen receptors, particularly in the fourth finger, and that 2D:4D is determined by the balance of androgen and oestrogen on the fourth finger.Although this study successfully confirmed AR and ESR genes were associated with 2D:4D in mice, in humans, the results were inconsistent.
In 2003, Manning et al. reported the positive correlation between CAG repeats in AR gene and the right 2D:4D in males, which has since garnered extensive attention from researchers, leading to further studies on digit ratio and SSHR.In vitro studies have demonstrated that the transcriptional activity of AR is negatively correlated with the number of CAG repeats in the AR gene (Beilin et al. 2000;Ding et al. 2004;Rodríguez-González et al. 2009).The more CAG repeats in the AR gene, the weaker the transcriptional activity of AR, resulting in decreased sensitivity to androgens, which may be one of the factors contributing to increased male 2D:4D.However, subsequent studies in other populations, including our study in the Chinese population (Zhang et al. 2013), have not been able to confirm this correlation.More recently, Zhang et al. conducted a replication study and meta-analysis, which demonstrated that the number of CAG repeats in the AR gene and current testosterone levels are not significantly related to 2D:4D at the individual level (Zhang et al. 2020).Pearce et al. found that rs6152 of the AR gene was significantly positively correlated with female left 2D:4D (Pearce et al. 2018), and Swift-Gallant et al. found that mice with global AR overexpression showed an increase in the right 2D:4D, regardless of sex (Swift-Gallant et al. 2021).Therefore, the relationship between AR and 2D:4D remains inconclusive and requires further investigation using multi-locus detection in different populations.
Besides AR, the oestrogen receptor (ESR) also plays an essential role in regulating sex hormone levels.Although some studies have reported a correlation between ESR and 2D:4D, there is still a lack of verification in different populations, especially Asian populations.In our previous study, we analysed the association of 5 SNPs in the ESR1 gene in Ningxia Hui and Han populations but found no association between these SNPs and 2D:4D (Mengjing 2013).In this study, we selected 6 SNPs in ESR1 (rs2228480 and rs3798758) and ESR2 *p < 0.05; dbsnP wild-type loci as reference in the first column genotype of each snP, Bonferroni was used to correct P values when there were significant differences between groups.Table 8.Continued.
(rs944459, rs8006145, rs928554, and rs8018687) for analysis, but we did not find any correlation between these 6 SNPs and 2D:4D in either sex.However, we did find that the Han ethnic in Ningxia, carrying the CC genotype of rs3798758 (p = 0.028), had the longest 2D:4D in the right hand.Further analysis using logistic regression did not show any significant association between rs3798758 and 2D:4D.Therefore, it remains uncertain whether rs3798758 is related to digit ratio, and more extensive sample sizes and diverse populations are required to confirm our findings.Based on the previous research results, we believe that the human digit ratio may have a more complex biological basis.The functional changes of AR and ESR may only reflect the sensitivity of individuals to androgens, oestrogens, and other sex hormones, and not necessarily the actual concentration of androgen and oestrogen.Additionally, the genetic background and sample size may contribute to the inconsistent findings across studies.Therefore, our present results cannot fully confirm or exclude the association between ESR and 2D:4D in the Chinese population, and further research using larger sample sizes and diverse ethnic groups is necessary to clarify.
In addition to AR and ESR, sex hormones also affect local or systemic growth and development through the mediation of GPER and PGR.However, the correlation between GPER and PGR gene polymorphisms and digit ratio has not been reported yet.In this study, we not only found that the genotype and allele frequency of rs12702047 in GPER1 showed a significant difference between males and females, but we also found that different genotypes of rs12702047 were significantly related to both hand 2D:4D, especially for left 2D:4D.Individuals with the AG genotype had significantly longer 2D:4D than those with the AA and GG genotypes (p < 0.05).
GPER1 encodes a multichannel membrane protein receptor located in the endoplasmic reticulum that binds to oestrogen to activate multiple downstream signalling pathways, increase cAMP levels, promote intracellular calcium mobilisation, and nuclear phosphatidylinositol 3,4,5-triphosphate synthesis.Therefore, GPER1 plays an important role in regulating oestrogen on cells and tissues.Furthermore, GPER1 has been shown to play a role in various biological processes, including skeletal and nervous system development, metabolism, and cognition (Fuentes and Silveyra 2019).It has also been shown that GPER1 is involved in the formation of phenotypic sexual dimorphisms, but the specific pathways remain unclear (Dovey and Vasudevan 2020).
This study found that rs12702047 may be involved in the formation of digit ratio, which may be related to the location of this locus.We analysed the function of this locus via bioinformatics.Rs12702047 is located in the 3'UTR of the GPER1 gene, and bioinformatic analysis showed that it may regulate the expression of GPER1 by affecting the binding of micro-RNA: hsa-miR-744.When the allele changes from A to G, the GPER1 gene 3'UTR binds to hsa-miR-744 easily.Therefore, it can be speculated that GPER1 carrying the rs12702047 G allele may bind to hsa-miR-4734 and decrease the expression of GPER1, which then affects the development of digit ratio.
Furthermore, HaploReg prediction revealed that rs12702407 fell into the DNase peak region and the region of proteins bound, including PAX5C20, CEBPB, HDAC2, HEY1, HNF4A, HNF4G, and POL2.These proteins are crucial regulators of mammalian gene expression.Among them, HDAC2 is an important methylated epigenetic modifier and is involved in gene expression regulation through binding to the promoter or enhancer element of GPER1.It may affect GPER1 gene expression through epigenetic modifications.Meanwhile, the change from allele G to A of rs12702407 remarkably affected the CHD2 motif binding according to HaploReg, which might subsequently influence chromatin conformation statuses and may regulate the GPER1 expression in muscle.Another important gene, CEBPB, regulates skeletal stem cell osteogenic differentiation and promotes osteoclastogenesis (Wang J et al. 2022), and may be involved in the formation of finger length ratio.
Through eQTL analysis of muscle and skeletal tissues, it was further shown that the expression of the GPER1 gene in individuals with AA genotype was significantly lower than that in individuals with GG genotype, which may be involved in the formation of digit ratio.Bioinformatics analysis showed that rs12702407 may be a functional SNP involved in the formation of 2D:4D by affecting the expression of GPER1, but this conclusion needs to be confirmed by further in vitro and in vivo experiments.
The receptor proteins encoded by PGR genes are distributed in multiple organs, including limb bones.They mainly mediate the physiological effects of progesterone and play an important role in the early and subsequent processes of pregnancy.It has been shown that rs1042839 of the PGR gene is a risk factor for recurrent abortion (Khan et al. 2021).This site was also found to be closely related to the onset of endometriosis and breast cancer (Treloar et al. 2005;Mao et al. 2015).Notably, these diseases have all been reported to be significantly associated with digit ratio (Schulter et al. 2012).
In this study, different genotypes of rs1042839 showed a significant difference in male left 2D:4D.The left-hand 2D:4D of men with GG genotype was significantly lower than that of individuals with AG genotype, suggesting that this locus may affect phalanx development and digit ratio sex dimorphism.Through bioinformatics analysis, it was found that when the rs1042839 allele changed from G to A, the codon encoding protein at position 770 would change from CAC to CAT.However, the coding amino acid remains unchanged, and this mutation is considered synonymous.Furthermore, this site is not located in key parts, such as α helix, β fold, and corners of protein secondary structure, which have less The analysis shows that rs12702047 is within the histone markers of promoter and enhancer regions and may influence 7 bound proteins (PAX5C20, CEBPB, HDAC2, HEy1, Hnf4A, Hnf4g, and Pol2). it may also change the regulatory motifs of chromodomain DnA helicase protein 2 (CHD2).(C) QTl analysis results of rs12702047 from the gTEx database, AC091729.9represents the locus-specific expression of muscle-skeletal tissues, and AC091729.7 represents other tissues.The analysis shows that rs12702047 has significant tissue-specific expression characteristics, and it could increase gPER expression in human muscle and skeletal tissue.
impact on protein function.Therefore, further investigation is needed to determine whether rs1042839 can affect progesterone metabolism through other ways and then participate in the formation of digit ratio.
Although our study found a correlation between sexual steroid hormone receptor-related gene polymorphisms and 2D:4D in college students in Ningxia, especially for rs12702047, there are still some limitations to this study.First, the number of SNPs included in this study is limited, which may not fully explain the gene's function.Therefore, future studies should include a larger sample size and consider the possible interaction between different loci on gene function.Additionally, all participants in this study were college students at Ningxia Medical University, which may limit the generalizability of the findings to other age groups.Furthermore, while we conducted bioinformatics analysis for some SNPs and predicted their potential function, further in vivo and in vitro experiments are necessary to fully elucidate the key molecular mechanism of how this locus may affect digit ratio development.In the next stage, we will conduct further research to verify the function of these SNPs and their interactions on 2D:4D, providing more evidence for exploring the potential mechanism of digit ratio formation.

Figure 1 .
Figure 1.linkage disequilibrium (lD) map of the seven candidate snPs.The measure of lD (D′) in group of (A) male and female and (B) Hui and Han ethnic groups, shown graphically between all possible pairs of snPs in terms of the coloured darker (D′), where white represents very low D′ and dark represents very high D′.The numbers in squares are D' values.

Figure 2 .
Figure 2. Both-hand 2D:4D ratios for different genotypes of rs12702047 in (A) sexes and (B) ethnic groups.Box plots show the interquartile range (iQR), with the median marked by a horizontal line inside the box.Whiskers extend to the highest and lowest observations within 1.5 times the iQR.individual right-hand 2D:4D ratio observations are represented by scatter points.measurements were acquired by three independent observers and data reliability was confirmed using iCC tests.Asterisks (*) on the graph represent statistically significant f-and P-values with AnoVA in different groups.

Figure 3 .
Figure 3.The integrative prediction of rs12702047 function through bioinformatics.(A) Conservative analysis of rs12702047 based on 10 Phylop scores for 46vertebrates and 33 mammals which were computed across the ± 10 bp region surrounding the rs12702047.The results indicate that rs12702047 is located in an evolutionarily conserved region and has high conservation scores (Phylop score=-3.125),suggesting that it is highly conserved in evolution.(B) Prediction of rs12702047 function through 3DsnP database, different coloured bars represent genomic elements that snPs may affect.The analysis shows that rs12702047 is within the histone markers of promoter and enhancer regions and may influence 7 bound proteins (PAX5C20, CEBPB, HDAC2, HEy1, Hnf4A, Hnf4g, and Pol2). it may also change the regulatory motifs of chromodomain DnA helicase protein 2 (CHD2).(C) QTl analysis results of rs12702047 from the gTEx database, AC091729.9represents the locus-specific expression of muscle-skeletal tissues, and AC091729.7 represents other tissues.The analysis shows that rs12702047 has significant tissue-specific expression characteristics, and it could increase gPER expression in human muscle and skeletal tissue.

Table 1 .
Basic information of participants in our population.

Table 3 .
multiple PCR primer information of 10 snP.

Table 4 .
genotype and allele frequency of 10 snPs between genders and ethnic groups in ningxia.
*p < 0.05; P-value were corrected by Bonferroni (Reference dbsnP wild-type loci, with the first column genotype of each loci as the reference value).

Table 5 .
Analysis of sex difference of ESR2 gene haplotype.

Table 6 .
2D:4D distribution of left and right hands among different genders or different ethnic groups in ningxia.

Table 7 .
Difference analysis of 10 snP of sex steroid hormone receptor related genes with 2D:4D between female and male (N = 814).

Table 8 .
Difference analysis of 10 snP of sex steroid hormone receptor related genes with 2D:4D between Hui and Han (N = 784).