Clinical characterization and analysis of the SRD5A2 gene in 16 Chinese patients with 5α-reductase type 2 deficiency: a single-center retrospective study

Abstract

Background

5α-reductase type 2 deficiency(5α-RD2) is a rare autosomal recessive disorder primarily caused by variants in the SRD5A2 gene. This condition is characterized by a spectrum of undervirilization in 46, XY individuals, ranging from ambiguous genitalia to predominantly male external genitalia with minimal undervirilization. Early diagnosis and effective management of 5α-reductase type 2 deficiency patients are major challenges for healthcare providers. The mutational analysis of SRD5A2 gene is regarded as the gold standard for the diagnosis of the disease. This study aimed to summarize our center’s experience with patients diagnosed with 5α-reductase type 2 deficiency due to variations in the SRD5A2 gene and the associated clinical manifestations to enhance clinicians’ awareness of this condition.

Methods

We conducted a retrospective single-center study by reviewing the medical records of patients diagnosed with 5α-reductase type 2 deficiency after genetic testing between April 2018 and July 2025.

Results

All 16 patients presented with ambiguous genitalia at birth and received diagnoses between the ages of 3 months and 11 years and 3 months. Of these patients, ten exhibited micropenis accompanied by hypospadias, six displayed an isolated micropenis, and six had either bilateral or unilateral cryptorchidism. Additionally, six patients had an underdeveloped scrotum. All affected individuals had a 46, XY karyotype and tested positive for the SRY gene, with two patients initially raised as female. The testosterone/dihydrotestosterone (T/DHT) ratios after human chorionic gonadotropin (hCG) stimulation ranged from 1.5- to 38.6-fold above the baseline values. This study identified ten distinct pathogenic variants of the SRD5A2 gene, including nine previously reported variants p.Gln6Ter, p.Gly66Arg, p.(Tyr128Cys), p.Asn193Ser, p.Gly196Ser, p.Gly203Ser, p.Phe219Serfs*60, p.Arg227Gln, and p.Arg246Gln and one novel variant p.(Cys119Arg), which expands the spectrum of known SRD5A2 variants.

Conclusion

This study provides insights into genotype–phenotype correlations in a cohort of 16 Chinese patients diagnosed with 5α-reductase type 2 deficiency and summarized the enzymatic activity of the identified genetic variants. Identifying variants associated with 5α-reductase type 2 deficiency will facilitate genetic counseling for at-risk families and enable prenatal diagnosis, thus empowering parents to make informed decisions and allowing healthcare providers to tailor treatment strategies effectively.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12887-026-07012-0.

Introduction

5α-reductase type 2 deficiency (5α-RD2, #264600) is a rare autosomal recessive 46, XY disorder of sexual development (46, XY DSD) caused by mutations in the steroid 5α-reductase type 2 (SRD5A2) gene, which encodes the steroid 5α-reductase (testosterone 5α-reductase or 3-oxo-5-α-steroid 4-dehydrogenase; EC 1.3.99.5) protein. This condition is responsible for 7–25% of all 46, XY DSD [1, 2] and was initially described in patients with hypospadias, micropenis, and cryptorchidism by Walsh et al. and Imperato-McGinley et al. in 1974 [3, 4]. Steroid 5α-reductase is predominantly expressed in the epididymides, seminiferous vesicles, prostate, and genital skin and plays a crucial role in catalyzing the conversion of testosterone (T) into dihydrotestosterone (DHT) [5]. Consequently, variations in SRD5A2 resulting in reduced enzymatic activity can reduce DHT production, triggering a wide range of genital ambiguities in patients with 5α-RD2, from mild underdevelopment of male sexual characteristics to complete female external genitalia [6].

SRD5A2, which includes five exons and four introns, is located on chromosome 2p23.1. This gene encodes the 5α-reductase protein, which comprises a 254 amino acid sequence containing androgen-binding (N-terminal) and NADPH cofactor-binding domains (C-terminal) [7, 8]. To date, more than 216 variants of the SRD5A2 gene (http://www.hgmd.cf.ac.uk) have been identified, of which 72% are missense/nonsense, 16% are small deletions/insertions, and 5% are splice variants. Homozygous variants account for about 60–70% of cases, while heterozygous variants account for approximately 30–40% [8, 9].

The conventional approach for diagnosing 5α-RD2 deficiency involves the assessment of serum testosterone-to-dihydrotestosterone (T/DHT) ratios, which poses challenges in distinguishing this condition from other DSDs [10]. But this ratio can be particularly challenging to interpret in prepubertal children, potentially resulting in misdiagnoses. Consequently, molecular genetic analysis has emerged as a crucial method for accurately identifying 5α-RD2 deficiency, particularly in cases of partial enzyme deficiency and prepubertal patients [11].

In this study, we identified ten distinct SRD5A2 variants among 16 patients from 15 unrelated families and assessed the correlation between clinical phenotype and genotype in patients wih 5α-RD2 deficiency. A previously unreported variant, p.(Cys119Arg), was identified, which enlarged the spectrum of SRD5A2 variants. These results will provide clinicians with valuable information for identifying and treating this rare disorder, enabling parents to make well-informed choices regarding gender selection and allowing healthcare providers to customize treatment strategies.

Materials and methods

Ethics approval

This study was approved by the Ethics Committee of the Children’s Hospital Affiliated to Zhengzhou University (Approval No. 2022-K-055). In accordance with the principles outlined in the Declaration of Helsinki, written informed consent was obtained from the parents or legal guardians of the 16 patients with 5α-RD2 deficiency involved in this retrospective study. Data from the medical records were anonymized and maintained confidential. All patients visited our center between April 2018 and July 2025. All the information analyzed in this study was retrieved from the hospital’s archived database. The investigations conducted had no impact on the patients involved.

Evaluation of patients’ clinical information

None of the patients were offspring of consanguineous marriages, and all 16 had a 46, XY karyotype and were positive for the SRY gene. Clinical data, including the chief complaint and physical examination of the external genitalia at birth were gathered and evaluated. A pediatric endocrinologist performed the physical examination of external genitalia. Serum hormone levels, including follicle-stimulating hormone (FSH), luteinizing hormone (LH), estradiol (E2), progesterone (P), testosterone (T), dihydrotestosterone (DHT), prolactin (PRL), 17-OH progesterone (17α-OHP), androstenedione, cortisol, and adrenocorticotrophic hormone (ACTH), were all measured using an automated electrochemiluminescence immunoassay (Roche, Cobas e801).

With the exception of patient 13(the brother of patient 5) and patient 16, who did not undergo the human chorionic gonadotropin (hCG) stimulation test, all patients were administered hCG. Patients 1–8 received a daily injection of 500 IU of hCG for three consecutive days, while patients 9–12,14–15 were administered with a dosage of 1000 IU/d. Serum samples were collected at baseline on day one before the initial hCG injection, followed by post-stimulation sample collection on day four to measure T and DHT levels.

Molecular genetic analyses

Molecular genetic analyses were conducted as follows: Considering the significant phenotypic overlap between 5α-RD2 deficiency and other disorders of sex development (e.g., androgen insensitivity syndrome), whole-exome sequencing (WES) was employed as the primary diagnostic approach. Candidate variants were subsequently identified and validated using Sanger sequencing, while parental testing was performed to determine the hereditary source of the variants. Data analysis was conducted using the GRCh37/hg19 human genome assembly, incorporating the reference sequence for SRD5A2 sourced from GenBank (NM_000348.3).

Genomic DNA was extracted from the peripheral blood of the sixteen affected patients and their parents (excluding the mother of patient 12, who declined to participate) using a commercial kit (QIAamp DNA Blood Kit, Qiagen), following the manufacturer’s protocol. For targeted Sanger sequencing, PCR amplification of the exons was performed in a 25 µl reaction volume containing 20–50 ng genomic DNA, 10 µM of each primer, 250 µM of each dNTP, 1×PCR buffer (containing 1.5mM Mg²⁺), and 1U Taq DNA polymerase. Initial denaturation was performed for 5 min at 94 °C, followed by 35 cycles of denaturation at 94 °C for 30 s, annealing at 60 °C for 30 s, extension at 72 °C for 45 s, and a final extension at 72 °C for 10 min. The primers used have been previously described [12]. The amplified PCR products were analyzed using 2.0% agarose gel electrophoresis to determine band size. Bidirectional sequencing was performed using an ABI 3730 automated sequencer.

The HiSeq 3000 (Illumina, USA) high-throughput sequencing platform was used for WES, following standard procedures. Sequencing was completed with a minimum target sequence coverage of 99%. Candidate variants were further confirmed by Sanger sequencing as described above. The Human Gene Mutation (http://www.hgmd.cf.ac.uk/), NCBI dbSNP (http://www.ncbi.nlm.nih.gov/SNP), 1000 Genomes Project (http://browser.1000genomes.org/), and ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/) databases were searched to ascertain whether the detected variants had been previously reported.

Protein function predictions of novel mutation

Novel SRD5A2 missense variants were subjected to multiple sequence alignments to confirm amino acid conservation using the UCSC Genome Bioinformatics Database. The potential pathogenicity of these novel SRD5A2 variants was evaluated through in silico analysis using various software tools, including Polyphen-2 (http://genetics.bwh.harvard.edu/pph2), PROVEAN (http://provean.jcvi.org/), and AlphaMissense (https://aigenetics.com.cn/). Furthermore, PyMOL software was applied to predict the 3D protein structure, as well as to assess the impact of the novel missense variants on protein structure. The American College of Medical Genetics and Genomics criteria were used for variant classification [13].

Results

Clinical characteristics

The clinical characteristics of the 16 patients and external masculinization scores (EMS) are presented in Table 1. These individuals all had ambiguous genitalia at birth and were diagnosed between the ages of 3 months and 11 years and 3 months. Among them, patients 1–6,10,12,15 and 16 presented with micropenis and hypospadias, while the remaining six patients had isolated micropenis. Patients 10 and 11 presented with bilateral cryptorchidism, with the testes located in the groin area, while patients 1, 6, 9, and 12 had unilateral cryptorchidism. Additionally, patients 1,3,6,10,12 and 15 exhibited an underdeveloped scrotum. Karyotype analysis and SRY gene testing were conducted to elucidate the potential etiology of ambiguous genitalia, revealing a 46, XY karyotype and a positive SRY gene in all 16 patients. At the time of initial consultation at our hospital, patients 1 through 8 were all under the age of three. Patients 6 and 10 were assigned female at birth, while the remaining 14 were assigned male. Eight out of the 16 patients (50%) sought guidance from our urology department, and patient 1 was incidentally identified during an evaluation for pneumonia.

Table 1.

Clinical characteristics and serum hormone profiles of the 16 Chinese patients with 5α-RD2 deficiency

Patient	Age at hormonal investigation	chief complaint	cryptorchidism	External genitalia EMS Score	Sex of rearing	Baseline					Post-hCG stimulation			T/DHT ratio (post-hCG: baseline)
						LH	FSH	T	DHT	T/DHT	T	DHT	T/DHT
1	3 months	micropenis and hypospadias	yes(located in left groin)	3	Male	6.22	2.91	2.36	117.08	20.2	7.62	193.92	39.3	1.9
2	4 months	micropenis and hypospadias	No	6	Male	6.27	3.68	1.69	120.889	14	7.33	164.7	44.5	3.2
3	4 months	micropenis and hypospadias	No	6	Male	3.18	2.07	2.09	57.05	36.6	11.88	217.31	54.7	1.5
4	9months	micropenis and hypospadias	No	6	Male	0.38	2.68	< 0.025	<9.4	-	13	326.8	25.1	25.1
5	10 months	micropenis and hypospadias	No	6	Male	14.1	6.83	< 0.025	<9.4	-	7.29	261.17	27.9	27.9
6	2.5years	micropenis and hypospadias	yes(located in right groin)	2.5	Female	< 0.1	0.256	0.087	48.86	-	4.72	162.66	29.0	16.3
7	2.10years	micropenis	No	9	Male	< 0.3	0.605	< 0.025	31.15	-	3.62	146.74	24.6	24.6
8	3years	micropenis	No	9	Male	< 0.3	0.937	< 0.025	< 1.55	-	6.53	91.26	38.6	38.6
9	3.9years	micropenis	yes(located in left groin)	8.5	Male	0.102	0.84	< 0.025	3.99	-	5.39	162.12	33.2	33.2
10	6.11years	micropenis and hypospadias	yes(located in bilateral groin)	2	Female	< 0.1	5.2	< 0.025	22.09	-	1.71	47.89	35.7	35.7
11	7 years	micropenis	yes(located in bilateral groin)	8	Male	< 0.1	0.997	< 0.025	31.31	-	1.14	39.36	28.9	28.9
12	7.1years	micropenis and hypospadias	yes(located in right groin)	5.5	Male	< 0.1	0.873	< 0.025	240.59	-	2.89	250.83	11.5	11.5
13	7.8years	micropenis	No	9	Male	0.85	13.3	< 0.69	31.55	-	-	-	-	-
14	8.2years	micropenis	No	9	Male	< 0.3	0.865	< 0.025	<9.4	-	2.54	77.09	30.4	30.4
15	9.7years	micropenis and hypospadias	No	6	Male	0.406	2.36	< 0.025	121.03	-	2.08	250.46	8.3	8.3
16	11.3years	micropenis and hypospadias	No	6	Male	1.34	3.23	< 0.07	0.132	-	-	-	-	-

- not done, EMS external masculinization score, Units are as follows: luteinizing hormone (LH) and follicle-stimulating hormone (FSH), mIU/mL, testosterone (T) ng/mL, dihydrotestosterone (DHT), pg/mL

To further clarify the differential diagnosis of 5ɑ-RD2 deficiency, serum levels of sexual hormones and other parameters, such as 17ɑ-OHP, ACTH, androstenedione, and cortisol, were also measured. All patients tested negative for these parameters. The baseline T levels ranged from < 0.025 to 2.36ng/mL, while baseline DHT levels ranged from < 1.55 to 240.59pg/mL. Following hCG stimulation, the T levels ranged from 1.14 to 11.88 ng/mL, while the DHT levels ranged from 39.36 to 250.83pg/mL. The T/DHT ratios ranged from 1.5 to 38.6-fold the baseline values after hCG stimulation (Table 1). The identification of SRD5A2 gene variants by molecular genetic analyses in all 16 patients confirmed the diagnosis of 5ɑ-RD2 deficiency (Table 2).

Table 2.

SRD5A2 gene variations in the 16 Chinese patients with 5α-RD2 deficiency

Patient	Genotype	Type of mutation	Location	Base change	Amino acid change	Remaining enzyme activity(%)[ref.]
1	compound heterozygous	Missense	Exon 2	c.355T > C	p.(Cys119Arg)	NR
		Missense	Exon 4	c.680G > A	p.Arg227Gln	3.2 [6]
2	compound heterozygous	Missense	Exon 4	c.578 A > G	p.Asn193Ser	5 [14]
		Missense	Exon 4	c.680G > A	p.Arg227Gln	3.2 [6]
3	compound heterozygous	Missense	Exon 4	c.607G > A	p.Gly203Ser	40 [15]
		Missense	Exon 4	c.578 A > G	p.Asn193Ser	5 [14]
4,7,11,15	compound heterozygous	Missense	Exon 4	c.680G > A	p.Arg227Gln	3.2 [6]
		Missense	Exon 5	c.737G > A	p.Arg246Gln	1.5 [6]
5,13,16	compound heterozygous	stop gained	Exon 1	c.16 C > T	p.Gln6Ter	0 [16]
		Missense	Exon 4	c.680G > A	p.Arg227Gln	3.2 [6]
6	compound heterozygous	stop gained	Exon 1	c.16 C > T	p.Gln6Ter	0 [16]
		Missense	Exon 1	c.196G > A	p.Gly66Arg	6.68 [6]
8	compound heterozygous	Missense	Exon 4	c.586G > A	p.Gly196Ser	5.7 [14]
		Missense	Exon 4	c.680G > A	p.Arg227Gln	3.2 [6]
9	homozygous	Missense	Exon 4	c.680G > A	p.Arg227Gln	3.2 [6]
10	compound heterozygous	stop gained	Exon 1	c.16 C > T	p.Gln6Ter	0 [16]
		Missense	Exon 2	c.383 A > G	p.(Tyr128Cys)	NR
12*	heterozygous	Frameshift	Exon 4	c.656delT	p.Phe219Serfs*60	0 [15]
		Missense	Exon 5	c.737G > A	p.Arg246Gln	1.5 [6]
14	compound heterozygous	Missense	Exon 4	c.607G > A	p.Gly203Ser	40 [15]
		Missense	Exon 4	c.680G > A	p.Arg227Gln	3.2 [6]

NR not reported, ref. reference

* His mother refused genetic testing

Molecular genetic analyses

Collectively, ten variants of SRD5A2 were identified (Table 2; Fig. 1a). The most prevalent genotype was p.Arg227Gln variation(12/16), followed by p.Arg246Gln(5/16) and p.Gln6Ter(5/16). Of these variants, p.(Cys119Arg) was newly identified, while the remaining 9 have been previously reported. The p.Gln6Ter variant is nonsense, the p.Phe219Serfs*60 variant is frameshift, and the remaining variants are all missense variations. These variants predominantly clustered in exons 1, 2,4, and 5 (Fig. 1a).

Fig. 1.

a SRD5A2 gene variants in 16 patients diagnosed with 5α-reductase type 2 deficiency. Most of the variants are located at exons 1 and 4 of the SRD5A2 gene. The red color represents the newly identified variation in this study. b Location of the SRD5A2 variants in the 16 patients with 5α-reductase type 2 deficiency. Protein visualization was generated using the Protter website (http://wlab.ethz.ch/protter/start/)

In our study, compound heterozygous variations were observed in 15 patients, while patient 9 exhibited a homozygous variation p.Arg227Gln. Pedigree analyses confirmed that 15/16 variations were inherited from both parents. Only the mother of patient 12 refused to undergo genetic testing to verify the origin of the variation.

Evolutionary conservation across different species was investigated using multiple sequence alignment analysis. The results demonstrated that the nine known variants identified in our study affect strictly conserved domains across vertebrate orthologs. The novel p.(Cys119Arg) variant also affects a highly conserved amino acid position of the SRD5A2 protein (Fig. 2).

Fig. 2.

SRD5A2 mRNA Cys119 sequence in 10 mammalian species. The red line and arrow indicate the residue of interest

Prediction of the effects of the novel variant

The novel missense variant p.(Cys119Arg) was predicted to be “probably damaging” by PolyPhen 2 (score:1.000; sensitivity: 0.00; specificity: 1.00) and “Deleterious” by PROVEAN. The AlphaMissense score was 0.9589, which falls within the “likely pathogenic” range (0.906–0.971) and provides PP3_Moderate level evidence. This variant substitutes the polar neutral cysteine at position 119 with a polar positively charged arginine, resulting in alterations of the side chain structure and hydrogen bonding (Fig. 3). According to the ACMG guidelines, the novel variant p.(Cys119Arg), was classified as a variant of uncertain significance (VUS, PM2+PP3_Moderate+PP4).

Fig. 3.

SRD5A2 3D structure before and after introduction of the variation at position 119( 3a:before variation, 3b: after variation). The purplish-red and grey colors indicate amino acid variation sites. The yellow color indicates the hydrogen bonding of the surrounding interactions. CYS, cysteine ARG, arginine

Discussion

Testosterone (T) and its 5α-reduced metabolite, dihydrotestosterone (DHT), are essential for male sexual development, playing distinct roles in embryogenesis. T, the primary androgen synthesized by the testes, plays a crucial role in male development via action on two distinct pathways. First, this steroid directly influences the transformation of Wolffian ducts into essential male reproductive structures, such as the epididymis, vasa deferentia, seminal vesicles, and ejaculatory ducts. Second, T serves as a prohormone for the formation of DHT, a key hormone in the virilization process in males [17, 18]. DHT is also essential for prostate development, induction of midline fusion, and elongation and enlargement of the urogenital tubercle and folds, which collectively drive the development of the penis and scrotum [4, 18]. Individuals with 5α-RD2 deficiency commonly present with ambiguous external genitalia, including a clitoral-like phallus, severely bifid scrotum, or perineoscrotal hypospadias. Despite normal Wolffian duct differentiation, Müllerian structures are typically absent in these patients [19, 20].

Specific SRD5A2 genetic variations and their effect on enzyme activity can result in variability in clinical and biochemical phenotypes [21]. The primary external genital malformation identified in a prior substantial cohort study of 190 Chinese patients with this condition was hypospadias, with micropenis and cryptorchidism present at a slightly lower frequency [1]. In our study, patients with ambiguous genitalia at birth were diagnosed between the ages of 3 months and 11 years and 3 months. Among them, patients 1–6,10,12,15 and 16 presented with micropenis and hypospadias, while the remaining six patients had isolated micropenis. Patients 10 and 11 presented with bilateral cryptorchidism, with the testes located in the groin area, while patients 1, 6, 9, and 12 had unilateral cryptorchidism. Additionally, patients 1,3,6,10,12 and 15 exhibited an underdeveloped scrotum. All affected patients exhibited a 46, XY karyotype with a positive SRY gene. Among these, patients 6 and 10 were assigned female at birth, while the other 14 were assigned male. Out of the 16 patients, eight (50%) initially sought consultation and treatment from our urology department. Patient 1 was incidentally identified during an evaluation for pneumonia. Consequently, these nine patients were promptly referred to our pediatric endocrinology outpatient clinic. The remaining seven patients directly sought consultation at our pediatric endocrinology outpatient clinic.

Variability in the manifestation of 5α-RD2 deficiency is believed to be associated with diverse factors. The presence of the SRD5A2 gene variant plays a significant role in determining the level of residual enzyme activity and degree of undervirilization, with nonsense mutations typically associated with the most severe phenotypes due to the complete absence of enzymatic activity. Conversely, missense and splicing junction variations that do not completely inactivate the enzyme result in a broad range of phenotypic variability [22]. The p.Arg227Gln and p.Arg246Gln variants have further been reported to reduce the enzyme activity to 3.2% and 1.5%, respectively [6, 23]. The variant p.Gln6Ter is commonly observed in Asian patients and is associated with a severe phenotype and reduced external masculinization score [24]. This could likely be attributed to significant protein truncation and complete loss of enzyme activity caused by the p.Gln6Ter variant. The substitution of C to T results in the conversion of the amino acid glutamine to a premature stop codon at position 6 in exon 1 [15]. The resulting lack of T- and NADPH-binding domains causes a total loss of enzymatic activity [16].

The variant p.Asn193Ser(c.578 A > G) has been widely observed in patients globally [25]. The asparagine residue at position 193 is highly conserved among mammals, and this variant has been identified as both damaging and pathogenic. Site-directed mutagenesis and in vitro assays for experimental validation have revealed that the variants p.Asn193Ser and p.Gly196Ser(c.586G > A) reduce the enzyme’s affinity for NADPH, thereby decreasing the enzyme activity to 5% and 5.7% of the wildtype enzyme, respectively [14]. The variant p.Phe219Serfs*60 (c.del656T) causes a frameshift mutation at amino acid position 219, and has been reported to lead to a complete loss of enzymatic activity [15]. The variant p.Gly66Arg (c.196G > A) has further been reported to reduce enzyme activity to 6.68% [6]. Table 2 summarizes the enzyme activities associated with these variant sites.

In addition to the previously classified variants, we identified a novel variant, p.(Cys119Arg)(c.355T > C), within the SRD5A2 gene, further diversifying the spectrum of the genetic variants of this gene. This newly discovered variant converted cysteine to arginine, resulting in alterations in the side chain structure and hydrogen bonding. Conservation analysis further revealed that Cys119 is evolutionarily conserved, while in silico analysis predicted the variant as “deleterious” and “probably damaging”. The enzyme 5α-reductase utilizes NADPH as a cofactor to catalyze the reduction of double bonds in various steroid substrates, including T, through the 3-oxo-delta (4,5) structures [14]. SRD5A2 further contains a 3-oxo-5-alpha-steroid 4-dehydrogenase domain, which encodes 149 amino acids spanning from position 106 to position 254. The p.(Cys119Arg) variant is situated within this domain, indicating that a mutation at amino acid position 119 may disrupt the affinity of 5α-reductase type 2 for NADPH, ultimately reducing its ability to convert T to DHT.

Patients 4, 7, 11 and 15 in our study were found to have the same variant, p.Arg227Gln/p.Arg246Gln; however, they presented with distinct clinical manifestations. Patient 4 and 15 presented with both micropenis and hypospadias, whereas patients 7 and 11 displayed isolated micropenis. In addition, patient 15 also demonstrated poorly developed scrotum. Moreover, the variants p.Arg227Gln and p.Arg246Gln showed significantly reduced 5α-reductase enzyme activity, resulting in almost complete inhibition of the conversion of T to DHT, ultimately triggering dysplasia of ambiguous external genitalia.

Patients 1, 2, 5, 8, 13, 14, and 16 were identified as carriers of the common variant p.Arg227Gln, each in conjunction with a unique accompanying variant, and patient 9 was found to be homozygous for the p.Arg227Gln variant. Overall, the p.Arg227Gln variant was identified in 12 of the 16 patients (75%) in our study. This high frequency aligns with previous reports in Chinese populations, where it has been consistently documented as the most prevalent SRD5A2 variant: 67.7% (88/130) in a large single‑center cohort [26] and 91.1% (41/45) in a study from South China [27]. To further assess its geographic specificity, we examined population allele frequencies in the gnomAD database (v4.1.0). The allele frequency of p.Arg227Gln was 0.003611 in East Asians, which is markedly higher than that in other populations (0.0001600). Collectively, these data support that p.Arg227Gln represents a mutational hotspot in the Chinese population. Among these patients, phenotypic variations of the external genitalia were observed: patients 1, 2, 5, and 16 exhibited both micropenis and hypospadias, whereas patients 8, 13, and 14 presented with isolated micropenis. Additionally, patient 1 also demonstrated unilateral cryptorchidism and a poorly developed scrotum. Patient 13, the brother of patient 5, was referred to our clinic by his parents following the genetic diagnosis of his sibling; he presented with micropenis, and the diagnosis was subsequently confirmed by genetic testing. The variants p.Arg227Gln and p.Asn193Ser or p.Gly196Ser were found to significantly hinder enzymatic activity, which concurs with the previously published studies mentioned above. Additionally, it was predicted that the p.(Cys119Arg) variant might lead to a reduction in enzyme activity. This may explain why patient 1 presented with a comparatively greater impairment in male sexual characteristics than patient 9, who had the p.Arg227Gln homozygous variation.

Patients 6, 10, 12, and 15 exhibited significant ambiguity in external genitalia, characterized by the presence of micropenis and hypospadias, accompanied by either unilateral or bilateral cryptorchidism. Notably, Patient 10 presented with underdeveloped scrotal structures and carried the compound variants p.(Tyr128Cys) and p.Gln6Ter, has been assigned female at birth. Similarly, a two-year-old Thai child (Patient 3) with severe undervirilization carried the variants p.(Tyr128Cys) and p.Arg227Gln, and was also raised as female [28]. This clinical presentation suggests that the p.(Tyr128Cys) variant may significantly impair enzyme activity. This is further supported by Kumar et al., who reported a homozygous p.(Tyr128Cys) variant in a child with SRD5A2 deficiency and an EMS score of 4.0 [29]. As mentioned above, the variants p.Gln6Ter, p.Phe219Serfs*60, p.Gly66Arg, and p.Arg246Gln exhibited significantly reduced enzyme activity. Patient 6 carried the variants p.Gln6Ter and p.Gly66Arg, and the sex assigned at birth was female. Patient 3, who carried the p.Gly203Ser and p.Asn193Ser variants, presented with micropenis, hypospadias, and underdeveloped scrotal structures. The p.Asn193Ser variant was observed to significantly impair enzymatic activity, while the p.Gly203Ser variant had been reported in the literature to decrease enzyme activity to 40% [15].

Previous studies from different countries have identified variants predominantly in exons 1, 4, and 5, with exon 4 being the most frequently implicated, followed by exons 1 and 5. Furthermore, variants in exon 4 commonly result in the complete absence of enzymatic functions [1, 15, 24]. Consistent with prior studies, the variants p.(Cys119Arg) and p.(Tyr128Cys) were detected in exon 2, while the remaining variants were distributed across exons 1, 4, and 5. Based on the severe clinical phenotypes exhibited by patients 1 and 10, carrying the p.(Cys119Arg) and p.(Tyr128Cys) variants, respectively, it was speculated that the two variants would lead to a significant decline in enzyme activity.

Additionally, it has been reported that approximately two out of every three individuals diagnosed with 5α-RD2 deficiency carry homozygous variants, while the remaining third present with heterozygous variants. Deletions and disruptive variations are infrequent and have been identified in diverse familial lineages worldwide [30]. However, patient 9 in our study exclusively exhibited homozygous variation, while all remaining SRD5A2 variants detected were compound heterozygous variants. This may stem from China’s large population, its diverse ethnic composition, and the cultural taboo against incestuous unions [1], indicating that the carrier frequency of a single mutant allele was likely underestimated due to the rarity of the disease.

Previously, it was believed that an elevated T/DHT ratio indicates 5α-RD2 deficiency. The stimulated T/DHT value was divided by the baseline T/DHT value. In our study, we assessed the T/DHT ratios of 14 patients (excluding patients 13 and 16). Following hCG stimulation test, patients 4 through 12, 14, and 15 demonstrated a significant increase in their T/DHT ratios. In contrast, patients 1, 2, and 3 exhibited T/DHT ratios of less than 5.0 after undergoing the hCG stimulation test. This was performed in infancy within the mini-puberty, when they have a significant endogenous testosterone production (patient 1, 2 and 3; 3–4 month old) and can therefore also produce a small (due to the SRD5A2 defect) amount of DHT, resulting in an elevated T/DHT ratio of 20.2(P1), 14 (P2) and 36.6 (P3). Stimulation with hCG can increase the testosterone level to some extend in these young people and also showed an elevated T/DHT level (P1: 39.3, P2:44.5, P3: 54.7), but if the stimulated T/DHT level is divided by the baseline (mini-puberty) T/DHT value this value is very small and does not reflect the SRD5A2 deficiency. The unstimulated mini-puberty T values (1.69–2.36 ng/ml) are in the range of the stimulated T values of the patients in infancy between 9 month and 11 years (1.14-13 ng/ml). In contrast, the unstimulated T values in infancy between 9 month and 11 years were undetectable. This is the reason why a hCG stimulation is necessary in infancy to measure T and DHT values. During the mini-puberty phase, the baseline T/DHT ratio may serve as a more critical diagnostic indicator than the fold-change. Odame et al. previously described two siblings born with predominantly female external genitalia, including a micropenis, a bifid scrotum, a urogenital sinus, and palpable gonads. The diagnosis of 5α-RD2 deficiency for the older sibling was established at the age of 6 years. However, despite the established diagnosis of the elder sibling, a sex assignment was still difficult for the younger sibling because the external genitalia were extremely poorly virilized, even when compared with the older brother at a similar age, while the plasma T/DHT ratio before hCG stimulation was within normal at 3 days of age [31]. Thus, 5α-RD2 deficiency cannot be ruled out, even in patients with a normal T: DHT ratio, particularly in infants younger than 6 months of age. These findings may highlight the necessity of conducting larger, multi-center cohort studies to validate the clinical interpretation of T/DHT values during this period. The precise etiology of the divergent phenotypes observed in patients 1, 2, and 3 also require differentiation from PAIS. The clinical manifestations of 5α-RD2 deficiency are similar to those of PAIS [32]. Given the potential limitations of biochemical analysis in definitively distinguishing these conditions, WES was performed on all patients based on two principal considerations: (1) the significant phenotypic overlap among various forms of 46, XY DSD (e.g., androgen insensitivity syndrome) necessitates a comprehensive differential diagnostic approach; and (2) this method provides an opportunity to identify potential genetic modifiers or unexpected findings that may inform individual patient management and broaden the molecular understanding of the condition. In our study, genetic testing confirmed biallelic SRD5A2 variants in patients 1, 2, and 3, thereby excluding PAIS and establishing the diagnosis of 5α-reductase 2 deficiency. These findings underscore the value of WES as a preferred tool for differential diagnosis in clinically overlapping DSD cases. Furthermore, the plasma T/DHT ratio was within the normal range in these three patients, highlighting the diagnostic challenge posed by phenotypic variability.

After confirming the genetic diagnosis of 5α-RD2 deficiency, as patients 6 and 10 were reared as female at birth, the assigned sex was changed from female to male. All 16 patients received treatment with DHT gel applied to the root of the penis, while penile growth was observed to gradually increase during regular outpatient follow-up. Corrective surgery for hypospadias should be considered for patients 1–6, 10, 12, 15, and 16 once their penile length has reached approximately 3 cm. Currently, all 16 patients continue to receive ongoing outpatient follow-up care.

The primary limitation of this study was its small sample size. Regarding biochemical analysis, all DHT measurements were performed using an automated electrochemiluminescence immunoassay (Roche Cobas e801). Because this is a retrospective study, we re-verified the atypical DHT values by cross-checking the original laboratory reports against patient names, hospital identification numbers, and assay results; all values were confirmed to be correctly recorded. The baseline DHT levels in patient 12 (240.59 pg/mL) and patient 15 (121.03 pg/mL) were markedly elevated in the absence of measurable T, a physiologically atypical profile. As both patients are prepubertal males, this warrants further investigation with larger cohorts to explore potential contributing factors (e.g., assay limitations or individual biological variability). However, the definitive diagnosis for all patients in our study, including patients 12 and patient 15, remains firmly established by the genetic confirmation of 5α-RD2 deficiency. Additionally, the novel SRD5A2 variant p.(Cys119Arg) identified in this study has not been previously reported, and further functional studies are required to elucidate its pathogenic mechanism.

5α-RD2 deficiency is a rare genetic condition, the diagnosis of which generally critically depends on molecular analysis, which is essential for distinguishing it from other disorders with similar clinical features. This study identified ten variations in the SRD5A2 gene in 16 distinct Chinese individuals. Among these, we identified a novel variant, p.(Cys119Arg). Based on the clinical phenotypes associated with this variant, we predicted that the enzyme activity was severely impaired. Moreover, this variant is expanding the spectrum of SRD5A2 variations. The p.Arg227Gln variant was the most frequently identified variant in our study, suggesting a mutational hotspot in the Chinese population. Furthermore, early identification through molecular testing enables better clinical management, including hormone replacement therapy and sex assignment support. The identification of variants will enable genetic counseling for families at risk, in addition to facilitating prenatal diagnosis in family members.

Abbreviations

5α-RD2

5α-reductase type 2 deficiency

SRD5A2

steroid 5α-reductase type 2

46,XY DSD

46, XY disorder of sexual development

T

testosterone

DHT

dihydrotestosterone

FSH

follicle-stimulating hormone

LH

luteinizing hormone

E2

estradiol

PRL

prolactin

17α-OHP

17-OH progesterone

ACTH

adrenocorticotrophic hormone

P

progesterone

hCG

human chorionic gonadotropin

WES

whole-exome sequencing

PAIS

partial androgen insensitivity syndrome

EMS

external masculinization scores

ACMG

The American College of Medical Genetics and Genomics

Authors’ contributions

Linghua Shen: designed and performed the research and wrote the manuscript. Dongxia Fu, Wei Yang, Yan Cui, and Huizhen Wang: designed the research and supervised the manuscript. Dongxiao Li, Shiqi Wang, and Xiaocui Ma: collected clinical data and contributed to the analysis. Haiyan Wei and Yongxing Chen: made many suggestions to the revised manuscript. All the authors critically reviewed, revised, and approved the final version of the manuscript.

Funding

This work was supported by Henan Provincial Natural Science Foundation, China (grant number: 222300420351). The funder had no role in conceptualization, design, data collection, analysis, decision to publish, or preparation of the manuscript.

Data availability

The novel variant data have been submitted to the ClinVar database (https://www.ncbi.nlm.nih.gov/clinvar/variation/4686661).

Declarations

Ethics approval and consent to participate

This study was approved by the Medical Ethics Committee of Children’s Hospital Affiliated to Zhengzhou University (Approval No.2022-K-055). The study was performed in accordance with the ethical standards outlined in the 1964 Declaration of Helsinki and its later amendments or other comparable ethical standards. This study was conducted after informed consent to participate was obtained from the research participants and their parents, who signed the consent forms.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.