背景

46,XY性发育异常疾病（disorder /differences of sex development，DSD）是一组表型性别和/或性腺性别与其46,XY类型染色体性别不匹配的疾病谱系。随着第二代测序技术的应用，已经能在约50%的46,XY DSD进行明确的基因诊断。常见致病基因如AR、SRD5A2、NR5A1、MAP3K1可占到总检出变异数的50%~70%左右，不同国家地区的队列中检出率和基因谱系有所差异。尽管常见致病基因的致病分子机制较为明确，但发现的新变异位点有待于体外功能实验评估其致病性。新近报道的46,XY DSD致病基因尚缺乏充分的临床表型描述和致病机制研究，尚待充分明确新致病基因与46,XY DSD的因果关系。

目的

1.描述本中心46,XY DSD患者的致病基因变异谱系和检出率，扩增46,XY DSD中国人群基因变异谱系，提供新变异位点和新报道致病基因的见解。

2.对常见致病基因AR、SRD5A2、MAP3K1进行临床表型总结和体外细胞实验，为变异位点的致病性评价、分子机制研究和表型基因型关联提供新的证据。

3.对新报道致病基因PPP2R3C进行系统表型评估和动物模型研究，挖掘其调控性腺发育的机制。

方法

1.基于全外显子组测序技术，描述70例46,XY DSD患者的基因变异谱系。

2.从本中心性发育异常队列（n=180）中选取9名携带AR错义突变（既往无功能验证）的雄激素不敏感患者，回顾临床资料，按照ACMG指南进行变异致病性评估，并进行体外功能验证。体外功能验证具体包括构建突变型质粒，转染进入HEK293T细胞；提取RNA和蛋白质评估表达水平改变；共聚焦显微镜观察细胞亚定位情况；双荧光素酶报告系统观察对雄激素反应元件的转录活性是否改变。

3.从本中心性发育异常队列（n=180）中选取38例5α-还原酶缺乏症患者，总结临床特征，对其中15种SRD5A2变异进行体外功能实验。构建突变型质粒转染HEK293T细胞后通过质谱技术测定培养基中的双氢睾酮浓度，分析变异对5α-还原酶活性影响，最后进行变异的表型基因型关联分析。

4.选取本中心两例MAP3K1错义突变患者及一例墨西哥裔MAP3K1错义突变患者，收集临床资料并分析变异致病性。构建突变型质粒转染HEK293T后，以TOP/FOP Flash质粒观察β-catenin信号改变。

5.对本中心诊断的PPP2R3C复合杂合突变患者进行临床系统评估，并构建了PPP2R3C^+/-及PPP2R3C^c.684_686del杂合子小鼠(C57BL/6J)。通过组化染色观察睾丸及附睾的组织形态；光镜观察精子形态；扫描电镜观察精子尾部畸形和头部畸形比例；透射电镜观察精子中心体结构。对模式动物的一般情况、精子畸形情况、中心体结构及繁育力进行了评估。

结果

1.在56例患者中共鉴定出57种罕见变异（RVs），涉及9个基因（AR、SRD5A2、NR5A1、DHX37、SRY、HSD17B3、MYRF、PPP2R3C、GATA4）。其中，21种为新发现的RVs，36种为已报道的RVs。根据ACMG指南，43种变异被分类为致病性（P）或可能致病性（LP）变异，14种变异被定义为意义未明（VUS）。64.3%（45/70）的患者携带P或LP变异。39种RVs涉及雄激素合成与作用过程，14种RVs涉及睾丸决定与发育过程，4种RVs与综合征型46,XY DSD相关。AR、SRD5A2和NR5A1是导致46,XY DSD的最常见基因。7例患者携带近年来报道的46,XY DSD致病基因的RVs，其中4例患者携带DHX37基因变异，2例患者携带MYRF基因变异，1例患者携带PPP2R3C基因变异。

2.九种AR错义突变均未显著影响AR mRNA和蛋白质的表达水平。所有突变型AR蛋白在睾酮刺激下均无法有效进入细胞核，滞留在细胞质中。九种AR突变均显著降低了AR蛋白对两种雄激素反应元件（MMTV和GE2E1b）的转录激活活性。其中，与轻微性雄激素不敏感（MAIS）相关的p.S176R突变对转录活性的影响最小，而与完全性雄激素不敏感(CAIS)相关的突变（如p.Y572S、p.D696V、p.L723S、p.M750V、p.M776I、p.I899T）则导致转录活性大幅降低（1%-22%）。

3.在38例5α-还原酶缺乏症患者中共鉴定出24种SRD5A2罕见变异，其中15种变异既往无功能实验。十五种SRD5A2变异均导致5α-还原酶的活性显著降低（<8%），表明这些变异几乎完全丧失了酶活性。38例患者表现出三种不同的外生殖器表型（完全女性化、阴蒂肥大、尿道下裂），但携带相同变异的患者外生殖器表型存在显著差异。携带相同变异的患者睾丸位置（阴囊/阴蒂和隐睾）也存在差异。暂未发现5α-还原酶缺乏症患者中明确的基因型表型联系。

4.三名携带MAP3K1错义突变的46,XY性腺发育不全患者，均为女性表型，存在子宫发育不良、性腺发育异常、促性腺激素水平升高等情况，且有罹患性腺肿瘤风险。三种错义变异在人群数据库中均无报道，ACMG评级为致病意义不明，但Revel 和AlphaMissense 评分提示可能影响蛋白功能，且均可能改变MAP3K1蛋白结构，影响其功能。在HEK293T细胞中，三种突变型质粒未激活β-catenin信号通路，与野生型相比无统计学差异，可能原因是MAP3K1突变并非直接激活该通路，还可能通过影响p38、JNK、ERK等信号通路发挥作用，后续可更换细胞系重复验证。

5.携带PPP2R3C复合杂合变异的46,XY性腺发育不全患者，有面部畸形、身材矮小、骨龄延迟、免疫细胞亚群异常等症状，子宫发育不全且双侧性腺为条索状，性腺切除术后接受激素替代治疗。鉴定出两个变异分别遗传自父母，在人群数据库中低频或未出现，REVEL预测均有害，且会改变蛋白结构和氢键数量。成功获得杂合PPP2R3C^+/-小鼠与杂合PPP2R3C^c.684_686del 小鼠，未获得纯合子，推测纯合致死。杂合小鼠体重、睾丸及附睾称重与野生型无差异；精子密度和活力与野生型相比无统计学差异，但KI小鼠精子尾部和头部畸形比例增加，中心体结构存在异常；繁育力与正常小鼠无明显差异。

结论

围绕46,XY性发育异常患者展开系列研究，通过多方法探究致病基因变异谱系、功能及机制。第一章基于全外显子组测序，明确70例患者的基因变异谱，鉴定出多个致病基因及新变异。第二章对AR基因错义突变研究发现，突变不影响表达但阻碍核转运和转录激活，导致不同程度雄激素不敏感综合征。第三章研究SRD5A2基因变异，发现15种变异使酶活性几乎完全丧失，且基因型与表型无明显关联。第四章对MAP3K1错义突变进行功能验证，未发现其激活β-catenin 信号通路。第五章对PPP2R3C复合杂合变异患者临床评估并构建小鼠模型，发现患者有多系统异常，小鼠模型存在精子畸形和中心体结构异常。

本研究明确了中国人群46,XY DSD患者的基因变异谱系，发现新变异和致病基因；对常见致病基因AR、SRD5A2、MAP3K1进行功能验证，揭示其突变影响；对PPP2R3C基因进行临床表型描述，构建小鼠模型初步探索其调控性腺发育机制，为46,XY性发育异常疾病的诊断和治疗提供了重要依据。

Background

46,XY disorders of sex development (DSD) represent a spectrum of conditions in which the phenotypic and/or gonadal sex does not match the 46,XY chromosomal sex. With the application of next-generation sequencing (NGS) technologies, a definitive genetic diagnosis can now be made in approximately 50% of 46,XY DSD cases. Common pathogenic genes such as AR, SRD5A2, NR5A1, and MAP3K1 account for about 50%-70% of the identified variants, with variations in detection rates and gene spectra observed across different national and regional cohorts. Although the molecular mechanisms of these common pathogenic genes are relatively well understood, newly identified variants require functional validation to assess their pathogenicity. Additionally, the clinical phenotypes and pathogenic mechanisms of recently reported 46,XY DSD genes remain insufficiently characterized, necessitating further clarification of their causal relationships with 46,XY DSD.

Objectives

1.To describe the spectrum and detection rate of pathogenic gene variants in 46,XY DSD patients from our center, expanding the genetic variant spectrum of the Chinese population and providing insights into novel variants and newly reported pathogenic genes.

2.To summarize the clinical phenotypes and conduct in vitro functional experiments for common pathogenic genes (AR, SRD5A2, MAP3K1), providing new evidence for variant pathogenicity evaluation, molecular mechanism exploration, and genotype-phenotype correlations.

3.To systematically evaluate the clinical phenotypes and conduct animal model studies for the recently reported pathogenic gene PPP2R3C, exploring its regulatory mechanisms in gonadal development.

Methods

1.Based on whole-exome sequencing (WES), the genetic variant spectrum of 70 patients with 46,XY DSD was described.

2.Nine patients with androgen insensitivity syndrome (AIS) carrying AR missense mutations (previously unvalidated) were selected from our center's DSD cohort (n=180). Clinical data were retrospectively reviewed, and variant pathogenicity was assessed according to the American College of Medical Genetics and Genomics (ACMG) guidelines. In vitro functional validation included constructing mutant plasmids, transfecting HEK293T cells, and evaluating changes in RNA and protein expression levels. Subcellular localization was observed using confocal microscopy, and transcriptional activity on androgen-responsive elements was assessed using a dual-luciferase reporter system.

3.Thirty-eight patients with 5α-reductase deficiency were selected from our center's DSD cohort (n=180). Clinical features were summarized, and in vitro functional experiments were performed for 15 variants. Mutant plasmids were transfected into HEK293T cells, and dihydrotestosterone (DHT) concentrations in the culture medium were measured using mass spectrometry to analyze the impact of variants on 5α-reductase enzyme activity. Genotype-phenotype correlations were then analyzed.

3.Two patients with MAP3K1 missense mutations from our center and one international patient（Mexican）with a MAP3K1 missense mutation were selected. Clinical data were collected, and variant pathogenicity was analyzed. Mutant plasmids were transfected into HEK293T cells, and changes in β-catenin signaling were observed using TOP/FOP Flash plasmids.

4.A patient with PPP2R3C compound heterozygous mutations from our center underwent systematic clinical evaluation. PPP2R3C^+/- and PPP2R3C^c.684_686del heterozygous mice were generated. Testicular and epididymal tissue morphology was observed using histochemical staining; sperm morphology was examined under light microscopy; sperm tail and head abnormalities were assessed using scanning electron microscopy; and sperm centrosome structure was evaluated using transmission electron microscopy. General conditions, sperm abnormalities, centrosome structure, and fertility were assessed in the mouse models.

Results

1.A total of 57 rare variants (RVs) were identified in 56 patients, involving nine genes (AR, SRD5A2, NR5A1, DHX37, SRY, HSD17B3, MYRF, PPP2R3C, GATA4). Among these, 21 were novel RVs, and 36 were recurrent RVs. According to ACMG guidelines, 43 variants were classified as pathogenic (P) or likely pathogenic (LP), and 14 were classified as variants of uncertain significance (VUS). P or LP variants were identified in 64.3% (45/70) of patients. Thirty-nine RVs were involved in androgen synthesis and action, 14 in testicular determination and development, and four in syndromic 46,XY DSD. AR, SRD5A2, and NR5A1 were the most frequently affected genes. Seven patients carried RVs in recently reported 46,XY DSD pathogenic genes, including DHX37 in four patients, MYRF in two patients, and PPP2R3C in one patient.

2.Nine AR missense mutations did not significantly affect AR mRNA or protein expression levels. All mutant AR proteins failed to enter the nucleus effectively upon testosterone stimulation and remained in the cytoplasm. The nine AR mutations significantly reduced the transcriptional activity of two androgen-responsive elements (MMTV and GE2E1b). The p.S176R mutation, associated with mild androgen insensitivity (MAIS), showed the least reduction in activity, while mutations associated with complete androgen insensitivity (CAIS) (e.g., p.Y572S, p.D696V, p.L723S, p.M750V, p.M776I, p.I899T) resulted in a substantial reduction in transcriptional activity (1%-22%).

3.Twenty-four SRD5A2 rare variants were identified in 38 patients with 5α-reductase deficiency, including 15 variants without prior functional validation. All 15 SRD5A2 variants significantly reduced 5α-reductase 2 activity (<8%), indicating an almost complete loss of enzyme activity. The 38 patients exhibited three different external genital phenotypes (complete feminization, clitoromegaly, and hypospadias), but patients with the same variants showed significant phenotypic variability. Patients with the same variants also exhibited differences in testicular position (scrotum/clitoris and cryptorchidism). No clear genotype-phenotype correlation was observed in 5α-reductase deficiency patients.

4.Three 46,XY patients with gonadal dysgenesis carrying MAP3K1 missense mutations exhibited female phenotypes, uterine hypoplasia, gonadal abnormalities, and elevated gonadotropin levels, with a risk of gonadal tumors. The three missense variants were not reported in population databases and were classified as variants of uncertain significance (VUS) according to ACMG guidelines. However, Revel and AlphaMissense scores suggested potential functional impacts, and all variants were predicted to alter MAP3K1 protein structure and function. In HEK293T cells, the three mutant plasmids did not activate the β-catenin signaling pathway, showing no statistical difference compared to the wild type. This may be because MAP3K1 mutations do not directly activate this pathway but may affect other signaling pathways such as p38, JNK, and ERK. Further validation using different cell lines is needed.

5.A 46,XY patient with gonadal dysgenesis carrying PPP2R3C compound heterozygous mutations exhibited facial dysmorphism, short stature, delayed bone age, abnormal immune cell subsets, uterine hypoplasia, and bilateral streak gonads. Postoperatively, the patient received hormone replacement therapy. Two variants were identified, inherited from each parent, and were either low-frequency or absent in population databases. REVEL predictions indicated deleterious effects, and both variants were predicted to alter protein structure and hydrogen bonding. Heterozygous PPP2R3C^+/- and PPP2R3C^c.684_

^686del  mice were successfully generated, but no homozygous mice were obtained, suggesting embryonic lethality. Heterozygous mice showed no differences in body weight, testicular or epididymal weight compared to wild-type mice. Sperm density and motility were not statistically different, but KI mice exhibited increased sperm tail and head abnormalities and centrosome structural defects. Fertility was not significantly different from that of wild-type mice.

Conclusion

This series of studies on 46,XY DSD patients employed multiple approaches to explore the spectrum, function, and mechanisms of pathogenic gene variants. Chapter 1, based on whole-exome sequencing, identified the genetic variant spectrum in 70 patients and discovered multiple pathogenic genes and novel variants. Chapter 2 investigated AR missense mutations, revealing that mutations do not affect expression but impair nuclear translocation and transcriptional activation, leading to varying degrees of androgen insensitivity syndrome. Chapter 3 studied SRD5A2 gene variants, finding that 15 variants caused near-complete loss of enzyme activity, with no clear genotype-phenotype correlation. Chapter 4 functionally validated MAP3K1 missense mutations but found no activation of the β-catenin signaling pathway. Chapter 5 systematically evaluated a patient with PPP2R3C compound heterozygous mutations and generated a mouse model, revealing multi-system abnormalities in the patient and sperm abnormalities and centrosome structural defects in the mouse model.
This study clarified the genetic variant spectrum of 46,XY DSD patients in the Chinese population, identified novel variants and pathogenic genes, functionally validated common pathogenic genes (AR, SRD5A2, MAP3K1), and the clinical phenotype of patients with PPP2R3C gene variants was systematically described., constructing a mouse model to preliminarily explore its regulatory mechanisms in gonadal development. These findings provide important insights for the diagnosis and treatment of 46,XY DSD.