Introduction

Coronary artery disease (CAD), a leading cause of death in the world, is a complex multifactorial disease.1 It is thought to arise from an interaction between genetic susceptibility loci and various environmental determinants.2 The genetic basis of CAD has attracted much attention in recent years. With the development of CAD genome-wide association study (GWAS), more and more CAD susceptibility genes have been found, and many common, rare, and functional CAD variants have been identified in large-scale association analyses.3,4 Among these, the CYP17A1 gene, located on chromosome 10q24.3, is primarily expressed in the adrenal glands and gonads. It encodes a member of the cytochrome P450 superfamily. The resulting P450c17 protein possesses both 17α-hydroxylase and 17,20-lyase activities, serving as a key enzyme in the steroidogenic pathway for the biosynthesis of progesterone, corticosteroids, glucocorticoids, androgens, and estrogens.5 Recent research suggests that CYP17A1 may be involved in the development of hypertension,6,7 ischemic stroke,8 myocardial hypertrophy,9 myocardial infarction10 and premature coronary heart disease.11

In two large-scale association analyses the same year, researchers found that the CYP17A1 gene locus as one of 13 novel susceptibility loci for CAD.12,13 For instance, Butterworth et al14 selected a cohort of 15596 patients and 34992 controls to assess 2100 genes and 49094 variants, suggesting that the CYP17A1 gene is one of the susceptibility genes to CAD. However, it is crucial to note that these seminal GWAS were predominantly conducted in populations of European ancestry. Due to distinct linkage disequilibrium patterns and allele frequencies across ethnicities, genetic markers identified in Western populations may not carry the same risk magnitude or relevance in East Asian populations. Although subsequent studies have identified specific SNPs within CYP17A1, such as rs4409766 and rs1004467, as susceptibility variants, data regarding their specific association with CAD in the Chinese population remain sparse and inconsistent.11,15 This “ancestry gap” necessitates dedicated validation studies in specific ethnic groups to ensure the applicability of genetic findings.16

Furthermore, a significant gap exists between genetic discovery and clinical application. While polygenic risk scores (PRS) demonstrate significant value in evaluating the risk of cardiovascular diseases, they often require extensive genomic data that may not be feasible in routine clinical settings.17,18 In contrast, a visual statistical tool known as a nomogram can integrate specific genetic variants with routine clinical variables to provide a personalized and easy-to-interpret probability of disease risk.19 Therefore, this study aims to assess the correlation between CYP17A1 polymorphisms and CAD in a Chinese population and to develop a novel predictive nomogram incorporating these variants for early risk stratification in clinical practice.

Methods Ethical Approval of the Study Protocol

The present study was conducted in accordance with the principles outlined in the Helsinki Declaration and was approved by the Ethics Committee of the First Affiliated Hospital of Xinjiang Medical University (No. K201909-02). Written informed consent was obtained from each participant, explicitly granting permission for DNA analysis and the collection of relevant clinical data.

Subjects

This case-control study enrolled participants between December 2017 and October 2021. Genotyping was performed in 2023. All participants were hospitalized patients undergoing coronary angiography. Coronary angiography was performed by highly skilled physicians using the Judkins method. Results were interpreted by at least two experienced radiologists. CAD was diagnosed based on angiography reports, defined as the presence of at least one significant coronary artery stenosis with a luminal diameter reduction of > 50%. Subjects with vasospastic angina, active infection within the preceding 2 weeks, heart failure, adrenal dysfunction, or thyroid dysfunction were excluded. Controls were recruited from individuals with angiographically normal coronary arteries and no history of CAD. Finally, 1363 CAD patients (1007 men, 356 women) and 858 controls (439 men, 419 women) were included. Hypertension was defined as a systolic blood pressure ≥ 140 mmHg and/or a diastolic blood pressure ≥ 90 mmHg on at least three separate occasions. Subjects with T2DM (Type II diabetes mellitus) were defined as those who had fasting plasma glucose (FPG) ≥ 7.0 mmol/L, or 2-hour post-load plasma glucose (2hPG) ≥ 11.1 mmol/L, or had hypoglycemic therapy history.

Blood Collection and DNA Extraction

Fasting venous blood samples were collected in the morning into EDTA-containing tubes. Genomic DNA was isolated from peripheral leukocytes using standard phenol-chloroform extraction. DNA samples were stored at −80°C. Prior to genetic analysis, the DNA concentration was adjusted to 50 ng/μL.

Genotyping

The Haploview 4.2 software and the International HapMap Project website were used in the Phase II data base, we selected five tag SNPs (rs11191548, rs17115100, rs4409766, rs6162, rs6163) based on a minor allele frequency (MAF) ≥ 0.05 and a linkage disequilibrium threshold of r2 ≥ 0.8. The linkage disequilibrium (LD) map for these SNPs is shown in Figure 1. Genotyping for the case-control study was performed using an improved multiplex ligation detection reaction (iMLDR) technique (Digena Diagnostics Technology Co., Zhejiang, China), a highly specific and accurate method based on ligation and fluorescent signal detection. Stringent negative and positive controls were included in each experiment, and any ambiguous results were re-genotyped for confirmation. The accuracy of the iMLDR genotyping was further validated by an independent TaqMan assay, with 100% concordance (see Supplementary Material: Genotyping Protocol and Validation Data).

Figure 1 Genetic variation and linkage disequilibrium (LD) in the human CYP17A1 gene. An LD heatmap displays the pairwise associations among five genotyped single-nucleotide polymorphisms (SNPs). The analysis was performed using the “LD heat map” package in R (version 4.2.1), with LD blocks defined by the solid spline method. The strength of LD is represented by the color intensity of each square according to the r2 statistic, ranging from white (r2 = 0) through shades of pink (0 < r2 < 1) to red (r2 = 1). The physical length of the genomic region shown is 254.8 kb.

Biochemical Analysis

Fasting blood glucose (FBG), glycosylated serum protein (GSP), glycosylated hemoglobin (GHb), white blood cell count (WBC), total cholesterol (TC), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), blood urea nitrogen (BUN), creatinine (Cr), and uric acid (UA) were measured using standard methods at the Central Laboratory of the First Affiliated Hospital of Xinjiang Medical University.

Statistical Analysis Sample Size and Statistical Power

Prior to the study, an a priori sample size calculation was performed. Based on a standard genetic association study design (additive model, two-sided α=0.05, power= 80%), we aimed to detect an odds ratio (OR) of 1.4 for a common variant in the CYP17A1 gene, such as rs17115100 with a minor allele frequency (MAF) of 0.18 in the general population (gnomAD). This calculation indicated that approximately 853 subjects per group (1706 total) would be required. Our final sample size of 2221 participants (1363 cases, 858 controls) provided adequate statistical power for the primary analysis.

Statistical Methods

Data analysis was conducted using SPSS26.0 (SPSS Inc., Chicago, IL, USA) and R software (version 4.2.1). Continuous variables are presented as mean ± SD and were compared using Student’s t-test. Categorical variables are presented as frequencies (percentages) and were compared using the χ2 test or Fisher’s exact test, as appropriate. Hardy-Weinberg equilibrium was assessed using χ2 analysis. The genotype and allelic model (dominant, recessive, additive) distributions for the five SNPs were compared between CAD cases and controls using χ2 tests. We evaluated these three genetic models (additive: per-allele effect; dominant: effect allele carriers vs non-carriers; recessive: effect allele homozygotes vs others) to comprehensively assess association patterns without prior assumption of the underlying mode of inheritance. To control the FDR across these multiple comparisons (5 SNPs × 3 models), we applied the Benjamini-Hochberg procedure. The resulting FDR-adjusted P-values (denoted as q-values) are reported alongside the nominal P-values in the relevant results; a q-value < 0.05 was considered statistically significant after correction. Associations between genetic variants and CAD risk were further evaluated using multivariate logistic regression, with results expressed as ORs and 95% confidence intervals (CIs). A nomogram was used to build a multi-factor regression model for predicting the probability of the final event by using Regression Modeling Strategies (RMS) package in R. A receiver operation characteristic curve (ROC), c-index and calibration plots were used to evaluate the discrimination ability, calibration ability, and clinical effectiveness of the nomogram model.

Results

Table 1 presents the demographic and clinical characteristics of the study participants. Concentrations of FBG, GSP, GHb, Cr, TG and WBC were significantly higher in CAD group, while HDL-C was lower. The CAD group was older and had a higher prevalence of hypertension and diabetes. However, concentrations of TC, LDL-C were lower in the CAD, which might be attributable to the common use of lipid-lowering drugs for secondary prevention among high-risk cardiovascular disease subjects. Similar results were observed in both male and female subgroups.

Table 1 Demographic and Clinical Characteristics in CAD and Controls

Table 2 shows the distribution of genotypes of the CYP17A1 gene. All five tested SNPs were in Hardy-Weinberg equilibrium in the control group (all P > 0.05). The detailed allele frequencies and exact HWE P values for the three SNPs showing significant association (rs11191548, rs17115100, and rs4409766) are provided in Supplementary Table S1. There was no significant difference in the distribution of SNPs rs6162 and rs6163 genotypes, or in their dominant, recessive, and additive models between the CAD group and the control group (see Supplementary Tables S2 and S3). After applying FDR correction for multiple testing across five SNPs and three genetic models, the recessive models of rs11191548, rs17115100 and rs4409766 remained significantly different between CAD patients and controls (all q = 0.035), with the CC genotype of rs11191548, the CC genotype of rs4409766 and the TT genotype of rs17115100 being less frequent in the CAD group. A similar result was observed in males. Additionally, the dominant model (TT vs CC + CT) of rs4409766 showed a nominally significant difference, suggesting the TT genotype could increase the susceptibility to CAD in males.

Table 2 Genotype Distribution of rs11191548, rs17115100, rs4409766 of CYP17A1 Between CAD and Controls

Among the female subjects, the additive model of rs11191548 (CT vs CC + TT) showed a nominally significant difference between CAD patients and controls (P = 0.043). For rs17115100, both the TG genotype and its additive model (TG vs GG + TT) were found at nominally higher frequencies in the CAD group (P = 0.013). For rs4409766, its additive model (CT vs CC + TT) showed a marginal statistical difference (P = 0.055). Additionally, a trend toward statistical significance was observed in the recessive models across these four SNPs (rs11191548, rs17115100, rs4409766, rs6163) in females (see Table 2).

Tables 3–5 show multivariate logistic regression analysis of these three SNPs identified in the initial FDR-corrected screen. After adjustment for hypertension, age, T2DM, dyslipidemia, FBG, TG, Cr, WBC, and HDL-C, a significant protective effect was confirmed under the recessive model, indicating that the CC genotype of rs11191548 (OR = 0.507, 95% CI: 0.352–0.730, P < 0.001), the CC genotype of rs4409766 (OR = 0.557, 95% CI: 0.394–0.786, P = 0.001) and the TT genotype of rs17115100 (OR = 0.632, 95% CI: 0.473–0.844, P = 0.002) were independently associated with a reduced risk of CAD.

Table 3 Multiple Logistic Regression Analysis of Recessive Model for CAD (rs11191548)

Table 4 Multiple Logistic Regression Analysis of Recessive Model for CAD (rs17115100)

Table 5 Multiple Logistic Regression Analysis of Recessive Model for CAD (rs4409766)

For men, after adjusting for the main confounding factors, nominally significant protective effects were observed in the recessive model of rs11191548 (OR = 0.546, 95% CI: 0.349–0.852, P = 0.008), rs17115100 (OR = 0.688, 95% CI: 0.479–0.988, P = 0.043) and rs4409766 (OR = 0.602, 95% CI: 0.395–0.918, P = 0.018). Furthermore, under the dominant model, the TT genotype of rs4409766 was associated with an increased risk of CAD at a nominal significance level (OR = 1.323, 95% CI: 1.045–1.676, P = 0.020) (see Supplementary Table S4).

For women, after adjusting for the main confounding factors, additive models showed that the heterozygous genotype of rs11191548 and rs17115100 were nominally associated with an increased susceptibility to CAD (OR = 1.477, 95% CI: 1.070–2.039, P = 0.018; OR = 1.615, 95% CI: 1.176–2.216, P = 0.003) (see Supplementary Table S5).

Figures 2–4 show the nomograms based on the results of the logistic regression analysis. Confounders including FBG, hypertension, TG, Age, Gender, HDL-C and TT genotype of rs11191548 were incorporated into the final model to construct the nomogram (shown in Figure 2A). Based on the identification and calibration, the heavy samples were internally validated using the bootstrap-squared method. We applied this method to estimate the model’s performance by creating multiple datasets through resampling with replacement. The resampling was performed 1000 times, which allowed for the evaluation of variability and the assessment of the model’s robustness. As shown in Figure 2B, the AUC was 0.728 (95% CI: 0.706–0.749, P < 0.001). This indicates that the model has good discriminatory power. The calibration curve of the model is shown in Figure 2C. A similar analysis was performed for rs17115100 (see Figure 3, AUC = 0.727, 95% CI: 0.705–0.748, P = 0.001) and rs4409766 (see Figure 4, AUC = 0.727, 95% CI: 0.705–0.749, P = 0.001).

Figure 2 A nomogram for predicting the risk of CAD and its validation. (A) Nomogram based on seven variables (FBG, EH, TG, Age, Gender, HDL-C, and rs11191548.CC). The total points from all variables correspond to a predicted probability on the bottom scale. (B and C) Validation of the nomogram. The receiver operating characteristic (ROC) curve shows an area under the curve (AUC) of 0.728 (95% CI: 0.706–0.749). The calibration plot (generated from 1000 bootstrap replicates, n = 2221) shows agreement between predicted and observed risk (mean absolute error = 0.023).

Abbreviations: FGB, fasting blood glucose; TG, triglycerides; EH, hypertension; HDL, high-density lipoprotein cholesterol.

Figure 3 A nomogram for predicting CAD risk incorporating the TT genotype of rs17115100. (A) Nomogram based on seven variables (FBG, EH, TG, Age, Gender, HDL-C, and rs17115100.TT). The total points from all variables correspond to a predicted probability on the bottom scale. (B and C) Validation of the nomogram. The ROC curve shows an AUC of 0.727 (95% CI: 0.705–0.748). The calibration plot (generated from 1000 bootstrap replicates, n = 2221) shows agreement between predicted and observed risk (mean absolute error = 0.025).

Abbreviations: FGB, fasting blood glucose; TG, triglycerides; EH, hypertension; HDL, high-density lipoprotein cholesterol.

Figure 4 A nomogram for predicting CAD risk incorporating the CC genotype of rs4409766. (A) Nomogram based on seven variables (FBG, EH, TG, Age, Gender, HDL-C, and rs4409766.CC). The total points from all variables correspond to a predicted probability on the bottom scale. (B and C) Model validation. The discriminative ability is shown by the ROC curve (AUC = 0.727, 95% CI: 0.705–0.749). The calibration plot (1000 bootstrap repetitions, n = 2221) indicates good agreement between predictions and observations (mean absolute error = 0.023).

Abbreviations: FGB, fasting blood glucose; TG, triglycerides; EH, hypertension; HDL, high-density lipoprotein cholesterol.

Discussions

The relationship between polymorphisms in the cytochrome P450 (CYP) gene superfamily and cardiovascular disease (CVD) susceptibility has been established, linking genetic variation to metabolic phenotypes. In our study, we identified a significant association between specific polymorphisms in the CYP17A1 gene and the risk of CAD in a Chinese population. Notably, our findings withstood stringent statistical correction. After applying the Benjamini-Hochberg FDR correction for multiple testing across five SNPs and three genetic models, the recessive models of rs11191548, rs17115100, and rs4409766 remained significantly associated with CAD (q = 0.035). This statistical robustness reinforces the validity of these loci. Specifically, the CC genotype of rs11191548, the TT genotype of rs17115100, and the CC genotype of rs4409766 demonstrated independent protective effects against CAD after adjusting for confounding factors such as hypertension, diabetes, and lipid profiles. These results align with recent multi-ancestry GWAS, which have increasingly pinpointed the CYP17A1 as a hotspot for pleiotropic effects on blood pressure and steroid metabolism, suggesting a shared genetic architecture between hypertension and coronary atherosclerosis.9,11,20

A pivotal finding of our study is the distinct sexual dimorphism observed in the genetic association patterns. Our exploratory sex-stratified analyses revealed distinct patterns: while the recessive models showed nominally significant protective effects in males, and TT genotype of rs4409766 tended to be an independent risk factor for CAD in males; additive models indicated that heterozygous genotypes of rs11191548 and rs17115100 were nominally associated with increased CAD susceptibility in females. This divergence likely stems from the fundamental sex-specific reliance on CYP17A1-mediated steroidogenesis. CYP17A1 encodes P450c17, an enzyme with both 17α-hydroxylase and 17, 20-lyase activities, serving as a critical gatekeeper for the synthesis of sex steroids (androgens and estrogens) versus mineralocorticoids. In males, variants affecting 17, 20-lyase efficiency could subtly shift the androgen-to-estrogen ratio, a biomarker increasingly linked to endothelial function and plaque stability.21–24 Recent analyses of the UK Biobank have confirmed that genetic predictors of CAD exhibit significant sex-interactions, particularly in loci regulating hormonal homeostasis.25–27 Conversely, in females, where cardiovascular protection is historically attributed to 17β-estradiol, variants impairing substrate availability for aromatization could disproportionately elevate CAD risk.28,29 This hypothesis is supported by recent animal models showing that Cyp17a1-deficient female mice develop accelerated atherosclerosis due to estrogen deficiency-mediated metabolic dysregulation.11

Beyond individual SNPs, the integration of genetic and clinical data is crucial for precision medicine. We developed an early warning nomogram model integrating clinical characteristics with CYP17A1 genotypes. The model achieved an AUC of 0.728. While an AUC of 0.73 indicates moderate discrimination, it is comparable to the performance of traditional risk scores, such as the Framingham Risk Score (FRS), when applied to East Asian populations, where FRS often yields AUCs in the range of 0.65–0.75.30,31 The clinical utility of our nomogram lies in its potential to refine risk stratification for individuals in the “intermediate” clinical risk category. Recent studies suggest that integrating polygenic risk scores (PRS) with standard clinical metrics significantly improves net reclassification improvement (NRI), effectively identifying high-risk individuals who would be missed by clinical factors alone.32,33 By incorporating CYP17A1 variants, our model captures a portion of the heritable risk component that traditional factors like LDL-C and blood pressure fail to account for.

As reported in previous studies, the P450c17 protein is a crucial enzyme for catalyzing the formation of all endogenous androgens.34 Mutations in the CYP17A1 gene affect the synthesis of steroids, which are precursors of sex hormones.35 This gene plays an important role in transforming 17OH-pregnenolone to dehydroepiandrosterone (DHEA) and 17OH-progesterone to androstenedione. Thus, mutations in CYP17A1 may result in loss of enzymatic activity of P450c17, potentially reducing the biosynthesis of DHEA.36 In recent years, numerous clinical studies have indicated that DHEA levels play a significant role in the progression of CAD.37–39 Furthermore, functional genomics studies have shown that non-steroidal CYP17A1 inhibitors and genetic variants can alter the structural conformation of the enzyme, thereby modulating its catalytic efficiency and downstream hormone levels.5,40

In addition, evidence suggests that testosterone is also associated with risk factors for CAD, especially in elderly males.41,42 Testosterone acts as a precursor to estrogen, so normal estrogen signaling also relies on CYP17A1. There is evidence that sex hormone levels affect the development of cardiovascular and cerebrovascular diseases, such as estrogen, which can prevent oxidative stress and is thought to have vascular protective effects. Estrogen plays a crucial role in regulating a variety of physiological and pathological processes, including vascular constriction, endothelial repair, and lipid metabolism, and influences cardiovascular function directly or indirectly through pathways such as glucose metabolism and insulin-related signal transduction.43,44

In addition, as a key steroid metabolism gene, CYP17A1 may also be closely related to cardiovascular risk factors, such as lipid metabolism and insulin resistance, that promote CAD.10,11,45,46 Wu et al also confirmed that the SNP rs11191548 near CYP17A1 is associated with HDL-C and leptin in Chinese children.47 Besides, the coagulation and fibrinolysis systems play a crucial role in the pathogenesis of CAD. Physiological levels of testosterone promote the release of nitric oxide (NO) from endothelial cells, thereby improving endothelial function. In turn, low testosterone levels increase the concentration of clotting factor VIII, leading to endothelial dysfunction and vascular inflammation. Ultimately, platelet adhesion occurs on damaged vessels, thereby promoting the occurrence of CAD.48 Insulin resistance and diabetes are significant independent risk factors for CAD. Gucenmez et al confirmed that weak negative correlations were observed between testosterone levels and several anthropometric measures and glucose metabolism parameters.49 Collectively, these mechanisms suggest that CYP17A1 variants contribute to CAD risk through a “multi-hit” pathway involving hemodynamic, metabolic, and hormonal dysregulation.

Limitation

There are several limitations of our study that need to be addressed. Firstly, we detected only a limited number of polymorphisms in CYP17A1 and could not determine whether there were other synergistic genes. Second, our study failed to monitor the levels of important steroid-related hormones, such as DHEA, testosterone, and estrogen, which may be a confounding factor for the results. Third, while we applied FDR correction to the initial genetic screen, the predictive nomogram was developed and validated only in an internal cohort. The lack of an external validation cohort limits the generalizability of our model, particularly across different ethnic groups or geographical regions. Future multi-center studies are required to calibrate the model for broader clinical applicability. Finally, due to the relatively small sample size, larger clinical samples and further studies of other SNPs of CYP17A1 are required in future studies.Specifically, the sample size for female subgroup analyses may have been underpowered to detect smaller genetic effect sizes, warranting cautious interpretation of the sex-specific findings.

Conclusions

In summary, this study provides evidence that specific CYP17A1 gene variants (rs11191548, rs17115100, and rs4409766) are associated with CAD risk in a Chinese population. Their recessive genotypes are linked to reduced CAD risk. Furthermore, our analyses reveal distinct sex-specific association patterns, underscoring the importance of considering gender in genetic studies of CAD. The developed nomogram, which integrates these genetic factors with clinical variables, demonstrates preliminary predictive utility and offers a basis for future refined risk assessment.

Data Sharing Statement

The data that support the findings of this study are available from the corresponding authors, Xiang Xie and Ting-ting Wu, upon reasonable request.

Ethics Approval and Consent to Participate

The study protocol was reviewed and approved by the Medical Ethics Committee of the First Affiliated Hospital of Xinjiang Medical University (Approval No. [20170214-33]). Written informed consent for participation and publication was obtained from all individual participants included in the study. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the Declaration of Helsinki.

Acknowledgments

Yu-Juan Feng and Zhi-Hui Jiang are co-first authors for this study. Xiang Xie and Ting-Ting Wu are co-correspondence authors for this study. The authors acknowledge all nurses and doctors in the Heart Centre of the First Affiliated Hospital of Xinjiang Medical University. In addition, we acknowledge all donors of blood samples. The ethics committee or review committee of the First Affiliated Hospital of Xinjiang Medical University approved the research protocol.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Funding

This work was supported by the Special Funds Program for Central Guiding Local Science and Technology Development (ZYYD2024CG07) and Xinjiang Science and Technology, Tian shan yingcai Qingnian Project (2024TSYCJU0008).

Disclosure

The authors report no conflicts of interest in this work.

References

1. Eberly LA, Shultz K, Merino M, et al. Cardiovascular disease burden and outcomes among American Indian and Alaska Native Medicare Beneficiaries. JAMA Network Open. 2023;6(9):e2334923. doi:10.1001/jamanetworkopen.2023.34923

2. Cho SMJ, Koyama S, Honigberg MC, et al. Genetic, sociodemographic, lifestyle, and clinical risk factors of recurrent coronary artery disease events: a population-based cohort study. Eur Heart J. 2023;44(36):3456–15. doi:10.1093/eurheartj/ehad380

3. López Rodríguez M, Arasu UT, Kaikkonen MU. Exploring the genetic basis of coronary artery disease using functional genomics. Atherosclerosis. 2023;374:87–98. doi:10.1016/j.atherosclerosis.2023.01.019

4. Quaye LNK, Dalzell CE, Deloukas P, et al. The genetics of coronary artery disease: a vascular perspective. Cells. 2023;12(18):2232. doi:10.3390/cells12182232

5. Wróbel TM, Jørgensen FS, Pandey AV, et al. Non-steroidal CYP17A1 inhibitors: discovery and assessment. J Med Chem. 2023;66(10):6542–6566. doi:10.1021/acs.jmedchem.3c00442

6. Diver LA, Mackenzie SM, Fraser R, et al. Common polymorphisms at the CYP17A1 Locus Associate with steroid phenotype: support for blood pressure Genome-Wide Association Study Signals at This Locus. Hypertension. 2016;67(4):724–732. doi:10.1161/HYPERTENSIONAHA.115.06925

7. Lu X, Wang L, Lin X, et al. Genome-wide association study in Chinese identifies novel loci for blood pressure and hypertension. Human Mol Genetics. 2015;24(3):865–874. doi:10.1093/hmg/ddu478

8. Kim SK, Yim SV, Lee BC. Association between cytochrome P450 promoter polymorphisms and ischemic stroke. Exp Ther Med. 2012;3(2):261–268. doi:10.3892/etm.2011.388

9. Huber M, Lezius S, Reibis R, et al. A single nucleotide polymorphism near the CYP17A1 gene is associated with left ventricular mass in hypertensive patients under pharmacotherapy. Int J Mol Sci. 2015;16(8):17456–17468. doi:10.3390/ijms160817456

10. Künstner A, Aherrahrou R, Hirose M, et al. Effect of differences in the microbiome of Cyp17a1-deficient mice on atherosclerotic background. Cells. 2021;10(6):1292. doi:10.3390/cells10061292

11. Wu TT, Zheng YY, Ma X, et al. Mutated CYP17A1 promotes atherosclerosis and early-onset coronary artery disease. Cell Commun Signal. 2023;21(1):155. doi:10.1186/s12964-023-01061-z

12. IBC 50K CAD Consortium. Large-scale gene-centric analysis identifies novel variants for coronary artery disease. PLoS Genetics. 2011;7(9):e1002260. doi:10.1371/journal.pgen.1002260

13. Schunkert H, König IR, Kathiresan S, et al. Large-scale association analysis identifies 13 new susceptibility loci for coronary artery disease. Nat Genet. 2011;43(4):333–338. doi:10.1038/ng.784

14. Wang Y, Wang L, Liu X, et al. Genetic variants associated with myocardial infarction and the risk factors in Chinese population. PLoS One. 2014;9(1):e86332. doi:10.1371/journal.pone.0086332

15. Dai CF, Xie X, Yang YN, et al. Relationship between CYP17A1 genetic polymorphism and coronary artery disease in a Chinese Han population. Lipids Health Dis. 2015;14:16. doi:10.1186/s12944-015-0007-4

16. Tcheandjieu C, Zhu X, Hilliard AT, et al. Large-scale genome-wide association study of coronary artery disease in genetically diverse populations. Nat Med. 2022;28(8):1679–1692. doi:10.1038/s41591-022-01891-3

17. Patel AP, Wang M, Ruan Y, et al. A multi-ancestry polygenic risk score improves risk prediction for coronary artery disease. Nat Med. 2023;29(7):1793–1803. doi:10.1038/s41591-023-02429-x

18. Vassy JL, Posner DC, Ho YL, et al. Cardiovascular disease risk assessment using traditional risk factors and polygenic risk scores in the million veteran program. JAMA Cardiol. 2023;8(6):564–574. doi:10.1001/jamacardio.2023.0857

19. Selvarajah S, Kaur G, Haniff J, et al. Comparison of the Framingham Risk Score, SCORE and WHO/ISH cardiovascular risk prediction models in an Asian population. Int J Cardiol. 2014;176(1):211–218. doi:10.1016/j.ijcard.2014.07.066

20. Lu Q, L PR, Wang Q, et al. Integrative tissue-specific functional annotations in the human genome provide novel insights on many complex traits and improve signal prioritization in genome wide association studies. PLOS Genetics. 2016;12(4):e1005947. doi:10.1371/journal.pgen.1005947

21. Guengerich FP, Mccarty KD, Tateishi Y, et al. Steroid 17α-hydroxylase/17, 20-lyase (cytochrome P450 17A1). Methods Enzymol. 2023;689:39–63.

22. Xu Y, Jiang S, Yan Z, et al. Phenotypic heterogeneity and fertility potential of patients with 17-Hydroxylase/17,20-lyase deficiency. J Clin Endocrinol Metab. 2022;107(6):e2610–e2618. doi:10.1210/clinem/dgac029

23. Aherrahrou R, Kulle AE, Alenina N, et al. CYP17A1 deficient XY mice display susceptibility to atherosclerosis, altered lipidomic profile and atypical sex development. Sci Rep. 2020;10(1):8792. doi:10.1038/s41598-020-65601-0

24. Kurnaz E, Kartal Baykan E, Türkyilmaz A, et al. Genotypic sex and severity of the disease determine the time of clinical presentation in steroid 17α-Hydroxylase/17,20-Lyase deficiency. Hormone Res Paediatrics. 2020;93(9–10):558–566. doi:10.1159/000515079

25. Jain V, Dabbs-Brown A, Liu C, et al. Genome-Wide European Ancestry Study identifies coronary artery disease-associated loci through gene-sex hormone interaction. J Am Heart Assoc. 2024;13(24):e034132. doi:10.1161/JAHA.123.034132

26. Chen JY, Ardissino M, Reddy RK, et al. Genetically predicted androgenic profiles and adverse cardiac markers: a sex-specific Mendelian randomization study. ESC Heart Fail. 2023;10(6):3525–3537. doi:10.1002/ehf2.14527

27. Flynn E, Tanigawa Y, Rodriguez F, et al. Sex-specific genetic effects across biomarkers. Eur J Human Genet. 2021;29(1):154–163. doi:10.1038/s41431-020-00712-w

28. Dos Santos BR, Casanova G, Da Silva TR, et al. Association between the aromatase (CYP19A1) gene variant rs10046 and cardiovascular risk in postmenopausal women. Arch Endocrinol Metabol. 2024;68(Spec Issue):e240087. doi:10.20945/2359-4292-2024-0087

29. Roszkowska-Gancarz M, Kurylowicz A, Polosak J, et al. The −351A/G polymorphism of ESR1 is associated with risk of myocardial infarction but not with extreme longevity. Int J Clin Chem. 2010;411(23–24):1883–1887. doi:10.1016/j.cca.2010.06.028

30. Vyas P, Vadodariya J, Kalsariya V, et al. Predictive role of novel echocardiographic parameter aortic velocity propagation, QRISK3 and Framingham risk score for presence and severity of CAD in Asian patients. J Cardiovasc Thoracic Res. 2022;14(3):153–158. doi:10.34172/jcvtr.2022.25

31. Kasim SS, Ibrahim N, Malek S, et al. Validation of the general Framingham Risk Score (FRS), SCORE2, revised PCE and WHO CVD risk scores in an Asian population. In: The Lancet Regional Health–Western Pacific. Vol. 35. 2023:100742

32. Busby GB, Kulm S, Bolli A, et al. Ancestry-specific polygenic risk scores are risk enhancers for clinical cardiovascular disease assessments. Nat Commun. 2023;14(1):7105. doi:10.1038/s41467-023-42897-w

33. Omiye JA, Ghanzouri I, Lopez I, et al. Clinical use of polygenic risk scores for detection of peripheral artery disease and cardiovascular events. PLoS One. 2024;19(5):e0303610. doi:10.1371/journal.pone.0303610

34. Martin LL, Kubeil C, Simonov AN, et al. Electrochemistry of cytochrome P450 17α-hydroxylase/17,20-lyase (P450c17). Mol Cellular Endocrinol. 2017;441:62–67. doi:10.1016/j.mce.2016.09.034

35. Tumilovich A, Yablokov E, Mezentsev Y, et al. The multienzyme complex nature of dehydroepiandrosterone sulfate biosynthesis. Int J Mol Sci. 2024;25(4):2072. doi:10.3390/ijms25042072

36. Miller WL. Androgen synthesis in adrenarche. Rev Endocr Metab Disord. 2009;10(1):3–17. doi:10.1007/s11154-008-9102-4

37. Forti P, Maltoni B, Olivelli V, et al. Serum dehydroepiandrosterone sulfate and adverse health outcomes in older men and women. Rejuvenation Res. 2012;15(4):349–358. doi:10.1089/rej.2011.1248

38. Hsu HJ, Yen CH, Chen CK, et al. Low plasma DHEA-S increases mortality risk among male hemodialysis patients. Exp Gerontol. 2012;47(12):950–957. doi:10.1016/j.exger.2012.08.012

39. Kim YM, Kang M, Choi JH, et al. A review of the literature on common CYP17A1 mutations in adults with 17-hydroxylase/17,20-lyase deficiency, a case series of such mutations among Koreans and functional characteristics of a novel mutation. Metabolism. 2014;63(1):42–49. doi:10.1016/j.metabol.2013.08.015

40. Wróbel TM, Grudzińska A, Yakubu J, et al. Pyridine indole hybrids as novel potent CYP17A1 inhibitors. J Enzyme Inhibition Med Chem. 2025;40(1):2463014. doi:10.1080/14756366.2025.2463014

41. Shufelt C, Bretsky P, Almeida CM, et al. DHEA-S levels and cardiovascular disease mortality in postmenopausal women: results from the National Institutes of Health--National Heart, Lung, and Blood Institute (NHLBI)-sponsored Women’s Ischemia Syndrome Evaluation (WISE). J Clin Endocrinol Metab. 2010;95(11):4985–4992. doi:10.1210/jc.2010-0143

42. Zhai G, Shu T, Xia Y, et al. Characterization of sexual trait development in cyp17a1-deficient zebrafish. Endocrinology. 2018;159(10):3549–3562. doi:10.1210/en.2018-00551

43. Lincoff AM, Bhasin S, Flevaris P, et al. Cardiovascular Safety of Testosterone-Replacement Therapy. N Engl J Med. 2023;389(2):107–117. doi:10.1056/NEJMoa2215025

44. Lopez DS, Taha S, Gutierrez S, et al. Association of total and free testosterone with cardiovascular disease in a nationally representative sample of white, black, and Mexican American men. Int J Impotence Res. 2024;36(4):385–393. doi:10.1038/s41443-022-00660-7

45. Zhang N, Chen H, Jia J, et al. The CYP17A1 gene polymorphisms are associated with hypercholesterolemia in Han Chinese. J Gene Med. 2019;21(8):e3102. doi:10.1002/jgm.3102

46. Zhao D, Guallar E, Ouyang P, et al. Endogenous sex hormones and incident cardiovascular disease in post-menopausal women. J Am Coll Cardiol. 2018;71(22):2555–2566. doi:10.1016/j.jacc.2018.01.083

47. Pabbidi MR, Kuppusamy M, P DS, et al. Sex differences in the vascular function and related mechanisms: role of 17β-estradiol. Am J Physiol Heart Circ Physiol. 2018;315(6):H1499–H1518. doi:10.1152/ajpheart.00194.2018

48. Dubey RK, Jackson EK. Cardiovascular protective effects of 17beta-estradiol metabolites. J Appl Physiol. 2001;91(4):1868–1883. doi:10.1152/jappl.2001.91.4.1868

49. Wu L, Gao L, Zhao X, et al. Associations of the hypertension-related single nucleotide polymorphism rs11191548 with high-density lipoprotein cholesterol and leptin in Chinese children. BMC Med Genet. 2018;19(1):9. doi:10.1186/s12881-018-0523-y