Unraveling the Genetic Mysteries of Müllerian Anomalies: Research Approaches and Clinical Significance

ABSTRACT

Müllerian anomalies are a collection of heterogeneous anatomical disorders of the female genital tract that present with complex clinical features of which severe subtypes like congenital aplasia of the vagina and uterus, may present with primary amenorrhea and dyspareunia, while mild cases like septate uterus, are often asymptomatic. Regardless of the types, the Müllerian anomalies impose both psychological and physical burdens on patients. Currently, the etiology of Müllerian anomalies remains largely unclear, which hinders early diagnosis and intervention. Although the advent of next‐generation sequencing technologies has promoted a more comprehensive depiction of genetic features of Müllerian anomalies, there is still a lack of experimental validation for the functions of these genes, where some novel preclinical models having been applied in cancer fields may provide potentially available strategies. Thus, in this review, we aim to summarize the genetic defects and novel validation techniques associated with Müllerian anomalies. Elucidating the genetic mechanisms involving Müllerian anomalies can pave the way for the development of early diagnostic strategies and preventional measures in the future.

This review primarily summarizes the genetic defects in Müllerian anomalies, the tools used to validate these genetic defects, and the future clinical significance of identifying the precise genetic etiology of Müllerian anomalies.

1. Introduction

Müllerian anomalies are derived from abnormal development, fusion, and absorption of the Müllerian ducts during the embryonic stage, affecting the fallopian tubes, uterus, cervix, and vagina, leading to various types of congenital tract abnormalities in females [1]. The prevalence of Müllerian anomalies is significantly underestimated, with an incidence of about 7% in the general population and as high as 25% in women with infertility and recurrent pregnancy loss [2]. The symptoms and severity of Müllerian anomalies vary among types of anomalies. Patients with Mayer‐Rokitansky‐Küster‐Hauser (MRKH) syndrome often present with amenorrhea, primary infertility, and difficulty in sexual intercourse [3]. Uterine malformations may result in recurrent miscarriage while vaginal malformations may cause dysmenorrhea, dyspareunia, and hematocolpos [4]. The health, self‐esteem, and sense of wholeness can be adversely and seriously affected in patients with Müllerian anomalies [5]. For better diagnosis and prevention, there is an urgent need to clarify the etiology of Müllerian anomalies. Accumulated studies have shown that the genetic factors are a crucial cause of Müllerian anomalies. The advent of next‐generation sequencing technologies and the use of novel validation techniques have offered effective tools for investigating the genetic basis of Müllerian anomalies, leading to the identification of several candidate genes associated with these conditions. This review aims to systematically summarize the genetic defects and novel validation techniques associated with Müllerian anomalies, laying the theoretical foundation for genetic counseling and early diagnosis of these anomalies in the future.

2. Classification of Müllerian Anomalies

There are various classification systems for Müllerian anomalies, including the classification system proposed by the American Fertility Society in 1988 (also called the AFS classification system), the VCUAM (Vagina Cervix Uterus Adnex‐associated Malformation) classification system, the ESHRE/ESGE consensus, and the new American Society for Reproductive Medicine Müllerian Anomalies Classification 2021 (MAC2021). The AFS classification system is primarily based on female genital tract anatomy and has become the most widely accepted system due to its simplicity and relevance to clinical reproductive outcomes [6, 7]. Subsequently, the VCUAM classification system was proposed, which is based on the principles of the TNM (tumor‐node‐metastasis) cancer staging system and can comprehensively cover all phenotypes of female genital tract anomalies [8]. In 2013, the ESHRE/ESGE consensus utilized both the anatomical and embryonic origins of the female genital tract, separately defining anomalies of the uterus, cervix, and vagina [9]. MAC2021 was proposed in 2021 by extending and updating the AFS classification system. It retained the AFS classification system terminology, incorporated cervical and vaginal anomalies, and categorized Müllerian anomalies into nine major classes [10].

Müllerian anomalies can be broadly classified into four main categories based on the existing classification system, including Mayer‐Rokitansky‐Küster‐Hauser (MRKH) syndrome, uterine anomalies, vaginal anomalies, and complex malformations. To facilitate the study of the genetic basis of Müllerian anomalies, we will discuss genetic factors according to this basic classification. It is anticipated that a more comprehensive understanding of the genetic mechanisms underlying Müllerian anomalies will pave the way for an improved classification system that integrates both structural features and genetic etiology.

3. The Genetic Defects of MRKH Syndrome

MRKH syndrome, characterized by congenital absence of the uterus, cervix, and upper vagina in 46,XX individuals, is divided into type I (isolated Müllerian agenesis) and type II (presence of extragenital malformations) [4]. Researches on the etiology of MRKH syndrome primarily focus on two aspects: DNA copy number variations (CNVs) and the detection of single nucleotide variants (SNVs).

3.1. DNA Copy Number Variants in MRKH Syndrome

CNVs, referring to large‐scale genomic alterations involving insertions, deletions, or duplications typically ranging from kilobases to megabases in size, are frequently observed in individuals with MRKH syndrome [11]. To date, CNVs have been reported across all chromosomes in MRKH patients, with recurrent regions such as 17q21, 22q11.21, 16p11.2, and 1q21 being most relevant to the condition (Table 1) [31]. The chromosomal regions with CNVs discovered repeatedly is not only highly correlated with genetic causes of MRKH syndrome, but also one marker for recognizing candidate genes. And these regions contain some genes may be candidates for MRKH syndrome, such as LHX1 (located in the chromosomal region 17q21), HNF1B (located in the chromosomal region 17q21), TBX6 (located in the chromosomal region 16p11.2), and RBM8A (located in the chromosomal region 1q21). However, in subsequent sequencing studies, researchers observed that point mutations in genes such as LHX1 and HNF1B occur at low frequency in the MRKH syndrome population [12]. This suggests that MRKH syndrome may result from the other genes yet to be identified or the cooperative effects of genes in this region.

TABLE 1.

Recurrent copy number variants in MRKH syndrome.

Chromosome	Type of MRKH	Type of imbalance	Number of reported	Size range (Mb)	Genes of interest	References
17q12	MRKH I and MRKH II	Del	17	1.2–1.9	LHX1 and HNFB1	[12, 13, 14, 15, 16, 17, 18, 19, 20, 21]
	MRKH I	Dup	1	0.32	/	[19]
16p11.2	MRKH I and MRKH II	Del	12	0.53–0.746	TBX6	[12, 14, 16, 19, 22, 23, 24, 25]
1q21.1	MRKH I and MRKH II	Del	3	0.399–4.6	RBM8A	[12, 13, 21]
		Dup	2	2.7		[18, 21]
22q11	MRKH I and MRKH II	Del	8	0.387–4	/	[13, 16, 18, 26, 27, 28, 29]
		Dup	3	0.6–3.5		[13, 18, 30]

3.2. DNA Single Nucleotide Variations in MRKH Syndrome

In addition to CNVs, SNVs in candidate genes represent another important etiological genetic factor for MRKH syndrome. Current research identifies candidate genes mainly through the following approaches. First, by investigating chromosomal regions where CNVs have been recurrently reported in MRKH syndrome. Second, by examining genes known to regulate Müllerian duct development during embryogenesis. What's more, by considering genes involved in urinary system development, which may also participate in female reproductive tract formation. To systematically summarize the SNVs identified in MRKH syndrome, this review classifies and interprets these genetic variants according to the 2015 ACMG guidelines, and documents them following the HGVS nomenclature standards (Table 2) [45, 46].

TABLE 2.

Pathogenic single nucleotide variants of the candidate genes in MRKH syndrome.

Genes	Phenotype	Evidence	ACMG classification	cDNA change	Amino acid substitutions	Mutation type	References
LHX1	MRKH II	PVS1 + PM2	LP	NM_005568.2:c.25dup	NP_005559.2:p.Arg9LysfsTer25	frameshift	[32]
ZNHIT3	MRKH II	PVS1 + PS3 + PM2	P	NM_004773.4:c.406C>T	NP_004764.1:p.Gln136X	nonsense	[33]
	MRKH II	PVS1 + PS3 + PM2	P	NM_004773.4:c.255_258del	NP_004764.1:p.Asp85fs	frameshift	[33]
TBX6	MA	PSV1 + PM2 + PP3	P	NM_004608.3:c.622‐2A>T	/	splice site	[14]
	MRKH I	PSV1 + PS4	P	NM_004608.3:c.434G>A	NP_004599.2:p.(=)	frameshift	[34]
	MRKH I	PSV1 + PS4	P	NM_004608.3:c.839+5G>A	/	splice site	[34]
	MRKH I	PS3 + PS4 + PP3	P	NM_004608.3:c.422T>C	NP_004599.2:p.Leu141Pro	missense	[34]
	MRKH I	PS3 + PS4 + PP3	P	NM_004608.3:c.745G>A	NP_004599.2:p.Val249Met	missense	[34]
	MRKH I	PS4 + PS4	P	NM_004608.3:c.10C>T	NP_004599.2:p.Pro4Ser	missense	[34]
	MRKH II	PS3 + PS4 + PP3	P	NM_004608.3:c.356G>A	NP_004599.2:p.Arg119His	missense	[34]
	MRKH II	PS4 + PS4	P	NM_004608.3:c.400G>A	NP_004599.2:p.Glu134Lys	missense	[34]
	MRKH II	PSV1 + PM2	LP	NM_004608.3:c.1146C>A	NP_004599.2:p.Tyr382Ter	nonsense	[35]
	MRKH II	PSV1 + PS4	P	NM_004608.3:c.621+1G>A	/	splice site	[34, 36]
WNT4	MA with hyperandrogenism	PS3 + PS4	P	NM_030761.5:c.647A>G	NP_110388.2:p.Glu216Gly	missense	[37]
	MA with hyperandrogenism	PS3 + PS4	P	NM_030761.5:c.247C>T	NP_110388.2:p.Arg83Trp	missense	[38]
	MA with hyperandrogenism	PS3 + PS4	P	NM_030761.5:c.35T>C	NP_110388.2:p.Leu12Pro	missense	[39]
	MA with hyperandrogenism	PS3 + PS4	P	NM_030761.4:c.697G>A	NP_110388.2:p.Ala233Thr	missense	[40]
WNT9B	MRKH I	PVS1 + PS4	P	NM_003396.3:c.1029C>A	NP_003387.1:p.Cys343Ter	nonsense	[41]
	MRKH I	PS4 + PP3	LP	NM_003396.3:c.665G>A	NP_003387.1:p.Arg222His	missense	[41]
	MRKH I	PS4 + PP3	LP	NM_003396.3:c.722G>A	NP_003387.1:p.Arg241His	missense	[41]
	MRKH I	PS4 + PP3	LP	NM_003396.3:c.919C>T	NP_003387.1:p.Arg307Trp	missense	[41]
PAX8	MRKH I	PS4 + PS3 + PP3	P	NM_003466.3:c.236C>G	NP_003457.1:p.Ser79Cys	missense	[36]
	MRKH I	PS4 + PS3 + PP3	P	NM_003466.3:c.136G>A	NP_003457.1:p.Asp46Asn	missense	[36]
BMP7	MRKH I	PSV1 + PS4	P	NM_001719.2:c.275_287dup	NP_001710.1:p.Pro98GlyfsTer31	frameshift	[36]
TBC1D1	MRKH II	PVS1 + PM2 + PP3	P	NM_015173:c.2553delC	NP_055988.2:p.Arg854GlufsTer24	frameshift	[42]
DLG5	MRKH I	PVS1 + PM2 + PP3	P	NM_004747:c.418C>T	NP_004738.3:p.Gln140Ter	nonsense	[42]
GREB1L	MRKH I	PVS1 + PM2 + PP3	P	NM_001142966.3:c.2787_2788del	NP_001136438.1:p.Asp930fs	frameshift	[43]
	MRKH II	PVS1 + PM2 + PP3	P	NM_001142966.3:c.2227del	NP_001136438.1:p.Gln743fs	frameshift	[43]
	MRKH II	PVS1 + PS4	P	NM_001142966.3:c.5396_5397del	NP_001136438.1:p.Lys1799fs	frameshift	[44]
	MRKH II	PVS1 + PM2 + PP3	P	NM_001142966.3:c.2173C>T	NP_001136438.1:p.Arg725Te	nonsense	[15]

Abbreviations: LP, Likely pathogenic; P, Pathogenic.

Recurrent CNVs regions have highlighted several candidate genes in MRKH syndrome, including LHX1, TBX6, ZNHIT3, and RBM8A. Deletions encompassing the 17q12 region, which harbors LHX1, are relatively common in patients, whereas SNVs in the LHX1 gene appear rare [12]. Currently, five heterozygous LHX1 mutations (four missenses and one frameshift) have been reported in seven individuals, only the latter (NM_005568.2:c.25dup) being defined as likely pathogenic according to ACMG criteria [13, 14, 32]. Also, as a gene located in 17q12, HNF1B has been implicated by in vitro validation that mice with Hnf1b gene ablation in the Müllerian duct epithelium exhibited symptoms resembling those of MRKH syndrome type II [47]. However, the meaningful SNVs of HNF1B were not found in two large‐sample cohorts of MRKH syndrome until now [12, 13]. Conversely, two pathogenic variants (one frameshift, one nonsense) in ZNHIT3 that are also located in the chromosomal 17q12 region have been experimentally proved to reduce protein expression, providing direct supporting evidence for its etiological role [33]. TBX6, located within the chromosomal 16p11.2 region and encoding a transcription factor related to mesodermal development, has emerged as a strong candidate gene for MRKH syndrome, supported by converging lines of evidence. Firstly, a large‐scale association study reported a significant burden of TBX6 variants in patients (p = 0.0004, OR = 5.25) [34]. Furthermore, the gene has repeatedly presented CNVs in the chromosomal 16p11.2 region of MRKH syndrome and a significant number of specific pathogenic mutations have also been identified [14, 34, 35, 36]. Most recently, functional validation from a major study which utilized western blotting, dual‐luciferase reporter assays, and immunofluorescence staining, has provided multiple rigorous experimental evidence for the pathogenicity of TBX6 mutations. Together, genetic association, recurrent genomic involvement, and functional data have jointly confirmed TBX6 as a compelling candidate gene in the etiology of MRKH syndrome [34]. In contrast, early interest in RBM8A (within the 1q21.1 region) has not been substantiated. A follow‐up study of 116 patients found no conclusive evidence that three intronic variants were common polymorphisms and a missense change (NM_005105.5:c.371A>T) was predicted to be benign. No subsequent studies have further implicated RBM8A in MRKH syndrome [48].

Investigating genes that regulate embryonic Müllerian duct development represents a complementary strategy for identifying candidate genes for MRKH syndrome. These genes mainly include WNT4, WNT9B, PAX8, BMP4, BMP7, and so on, of which WNT4 is the most important according to current studies and supporting evidence. Early studies identified and functionally validated four distinct pathogenic WNT4 mutations in vitro, confirming its role as a significant candidate gene [37, 38, 39, 40]. However, subsequent large‐scale cohort studies failed to detect any pathogenic WNT4 mutations in patients [12, 49, 50, 51]. Motivated by the MRKHS‐like phenotype observed in Wnt9b ^−/− female mice, WNT9B has emerged as a candidate gene in human MRKH syndrome [41]. To date, four pathogenic or likely pathogenic variants in WNT9B have been reported, including one pathogenic nonsense mutation (NM_003396.3:c.1029C>A) and three likely pathogenic missense mutations (NM_003396.3:c.665G>A; NM_003396.3:c.722G>A; NM_003396.3:c.919C>T) [41]. This translational evidence has greatly strengthened the potential role of WNT9B in the development of MRKH. Notably, WNT9B mutations in patients often co‐occur with alterations in other genes, such as LHX1 deletions or TBX6 missense mutations, suggesting that MRKHS etiology may involve the combined effects of multiple genetic variants [41, 52]. Recent large‐sample analyses further underscore the importance of identifying pathogenic genes among those regulating the development of Müllerian duct anomalies. A study of 442 patients versus 941 controls examined 19 Müllerian duct‐related genes and identified PAX8, BMP4 and BMP7 (OR = 3.77, 10.77, and 8.62, respectively) significantly differentially expressed. Two PAX8 likely gene‐disrupting (LGD) variants (NM_003466.3:c.236C>G and NM_003466.3:c.136G>A) were classified as pathogenic through silico prediction and functional validation [36]. Beyond PAX8, BMP4, and BMP7, the key regulators of early Müllerian duct formation, also exhibited a significant enrichment of variants in patients. However, two BMP4 variants (NM_001202.3:c.−132−1G>A and NM_001202.3:c.367G>T) were predicted to lie within the non‐nonsense‐mediated mRNA Decay (non‐NMD) region, implying minimal effect on mRNA stability. In contrast, a microduplication (NM_001719.2:c.275_287dup) in BMP7 was predicted to locate within the NMD region and may potentially disrupt transcription. Collectively, these observations implicate that dysregulation of the BMP signaling pathway may contribute to the pathogenesis of MRKH syndrome [36].

Given the shared developmental origin and close embryological association between the urinary and reproductive systems, it is plausible that genes implicated in urinary tract development may also contribute to the pathogenesis of MRKH syndrome. Chu et al. conducted whole‐exome sequencing on a cohort of 10 patients with MRKH syndrome, identifying nine potential candidate genes. Of these, only the variants in TBC1D1 and DLG5 were defined as pathogenic. Notably, both genes have previously been associated with renal developmental disorders: TBC1D1 with kidney and urinary tract malformations, and DLG5 with renal agenesis. Their first reported involvement in MRKH syndrome suggests a possible shared genetic etiology, implicating pathways essential for both urinary and reproductive system development [42]. GREB1L also emerges as a strong candidate gene, with several studies reporting pathogenic mutations in individuals with MRKH syndrome. These variants have been documented in both familial and isolated cases [15, 43, 44, 53]. In familial presentations, the mutations follow an autosomal dominant inheritance pattern with incomplete penetrance, indicating the complex etiology of MRKH syndrome [53]. Despite accumulating genetic evidence, the functional pathogenicity of GREB1L mutations lacks validation in vivo or in vitro, constituting a major knowledge gap that requires further investigation.

Collectively, these findings highlight the complex and polygenic nature of MRKH syndrome. In 2020, Li et al. provided the first evidence for a dual‐gene mechanism in Müllerian anomalies, demonstrating that combined mutation of Gen1 and Wnt9b in mice produces more severe uterine defects than single‐gene perturbations [54]. The finding by Li et al. once again reinforces the notion that Müllerian anomalies (including MRKH syndrome) involve a complex etiology and necessarily entail the contribution of multiple genes. Despite this, current research remains predominantly focused on SNVs of single genes, with few studies exploring digenic or oligogenic interactions. Furthermore, methodological limitations are common. Many reports merely list variants without sufficient supporting analysis, often precluding definitive ACMG classification. While technological advances have increased the volume of clinical sequencing data, simply enlarging cohort sizes is inadequate. It is also crucial to provide sufficient clinical evidence for the identified variants. Such evidence primarily encompasses the following aspects, such as: parental segregation studies, computational predictions, functional validation, and population frequency data. Based on the current research on MRKH syndrome, the scale of investigation continues to expand and substantial evidence has been provided for variants in certain genes. What's more, functional validation of multiple gene mutations and their interactions remains scarce. It is hoped that future studies will place greater emphasis on providing ample clinical evidence and mechanistic investigations into the combined effects of co‐occurring genetic variants on molecular and cellular function.

4. The Genetic Defects in Other Müllerian Anomalies

Besides MRKH syndrome, other genetic factors associated with Müllerian anomalies have been summarized. The genetic etiologies of other Müllerian anomalies are also predominantly categorized as CNVs (Table 3) and SNVs (Table 4). And the anomalies are described according to the following three categories of malformations including uterine anomalies, complex anomalies, and vaginal anomalies. Within the category of complex anomalies, this review focuses on HWWS syndrome, while for vaginal anomalies; it primarily discusses distal vaginal atresia. According to the predominant theory, the upper two‐thirds are derived from the Müllerian ducts, and the lower third originates from the urogenital sinus [60]. Although, from a strict embryological perspective, distal vaginal atresia may not be classified as a type of Müllerian anomalies, it is nevertheless summarized in this review due to its significant impact on female reproductive health.

TABLE 3.

Recurrent copy number variants in other Müllerian Anomalies.

Phenotype	Chromosome	Alternation	Reported times	Size range (Mb)	Genes of interest	References
Septate uterus	22q11.21	Gain	1	2.5	TBX1	[55]
	7q31.1	Gain	1	0.22	ZNF277	[55]
	7q31.1	Loss	1	0.22	ZNF277	[55]
	9q33.1	Loss	1	0.098	TRIM32	[55]
Bicornuate uterus	22q11.21	Loss	1	2.57	LZTR1, CRKL	[15]
Uterus didelphys	9q21.2	Loss	1	0.2	VPS13A	[55]
	9q21.2	Loss	1	0.2	VPS13A	[55]
	9q33.1	Loss	1	0.098	TRIM32	[55]
Distal Vaginal Atresia	17q12	Dup	1	1.93	HNF1B and LHX1	[56]
	17q12	Dup	1	1.90	HNF1B and LHX1	[56]
	17q12	Dup	1	1.87	HNF1B and LHX1	[56]

TABLE 4.

Single nucleotide variants of the candidate genes in other Müllerian anomalies.

Phenotype	Candidate genes	cDNA change	Amino acid substitutions	Mutation type	Evidence	ACMG classification	References
Septate uterus	TBX6	NM_004608.3:c.1015C>A	NP_004599.2:p.Pro339Thr	missense	PM2 + PP3	VUS	[48]
		NM_004608.3:c.484G>A	NP_004599.2:p.ly162Ser	missense	PP3 + PM2	VUS	[35, 48]
	RBM8A	NM_005105.5:c.‐21G>A	NP_005096.1:p.(=)	Polymorphic site	BS1	LB	[48]
Bicornuate uterus	RBM8A	NM_005105.5:c.‐21G>A	NP_005096.1:p.(=)	Polymorphic site	BS1	LB	[48]
	TBX6	NM_004608.3:c.815G>A	NP_004599.2:p.Arg272Gln	missense	PS4 + PM1 + PM2 + PP3+	LP	[35]
	WNT9B	NM_003396.3:c.919C>T	NP_003387.1:p.Arg307Trp	missense	PM2 + PP3	VUS	[41]
Unicornuate uterus	TBX6	NM_004608.3:c.484G>A	NP_004599.2:p.ly162Ser	missense	PP3 + PM2	VUS	[35]
Uterus didelphys	LHX1	NM_005568.4:c.1070_1081del	NP_005559.2:p.Pro357_Ser360del	inframe deletion	BS1 + BP7	B	[57]
HWWS	TBX6	NM_004608.3:c.484G>A	NP_004599.2:p.ly162Ser	missense	PM2 + PP3	VUS	[35]
		NM_004608.3:c.358A>G	NP_004599.2:p.Met120Val	missense	PM2 + PP3	VUS	[35]
	FRAS1	NM_025074:c.3191G>T	NP_079350.5:p.Arg1064Leu	missense	PM2 + BP1 + BP4	LB	[58]
		NM_025074:c.779G>C	NP_079350.5:p.Arg260Thr	missense	PM2 + BP1 + BP4	LB	[58]
	FAT1	NM_005245:c.249C>G	NP_005236.2:p.Phe83Leu	missense	PP3 + BS1	VUS	[58]
	CHD1L	NM_004284:c.3481G>C	/	splice site	PP3 + BS1	VUS	[58]
	WNT4	NM_030761:c.527G>C	NP_110388.2:p.Ser176Thr	missense	PM2 + PP2 + BP4	VUS	[58]
	FOXF1	NM_001451:c.491A>G	NP_001442.2:p.Asn164Ser	missense	PM2 + PP2 + BP4	VUS	[58]
	TGFBR3	NM_003243:c.1997_1998del	NP_003234.2:p.Val666GlufsTer16	frameshift	PVS1 + PM2 + PP3	P	[58]
	TRIM32	NM_012210:c.1012G>T	NP_036342.2:p.Ala338Ser	missense	PM1 + PM2 + P2 + BP4	VUS	[58]
	PCSK5	NM_001190482:c.5041C>G	NP_001177411.1:p.Pro1681Ala Protein Sequences	missense	BP1	VUS	[58]
	RET	NM_020630:c.1433G>A	NP_065681.1:p.Cys478Tyr	missense	PM2 + PP2 + PP3	VUS	[58]
Distal Vaginal Atresia	TBX6	NM_004608.3:c.266delC	NP_004599.2:p.Pro89ArgfsTer5	frameshift	PSV1 + PM2	LP	[59]
	KMT2C	NM_170606:c.2710C>T	NP_733751.2:p.Arg904Ter	nonsense	PS1 + PM2 + PP3	P	[56]
	AXL	NM_001699.5:c.1316G>T	NP_001690.2:p.Trp439Leu	missense	NA	VUS	[56]
	TBX3	NM_005996.3:c.1447delC	NP_005987.3:p.Phe483GlyfsTer18	frameshift	PSV1 + PM2	LP	[56]
	BRIP1	NM_032043:c.2392C>T	NP_114432.2:p.Arg798Ter	nonsense	PSV1 + PM2 + PP3	P	[56]
	NDUFAF7	NM_001083946.1:c.217‐1580G>A	/	splice site	PSV1 + PM2 + PP3	P	[56]
	PLXNA3	NM_017514.4:c.154C>T	NP_059984.3:p.Arg52Ter	nonsense	PSV1 + PM2 + PP3	P	[56]
	MYLK	NM_001321309.1:c.1876G>T	NP_001308238.1:p.Glu626Ter	nonsense	PSV1 + PM2 + PP3	P	[56]
	MYF5	NM_005593.4:c.418G>T	NP_005584.1:p.Gly140Trp	nonsense	PSV1 + PM2 + PP3	P	[56]
	ELN	NM_000501.3:c.1786+1G>A	/	splice site	PSV1 + PM2 + PP3	P	[56]
	ITGA7	NM_001144996.1:c.3052C>T	NP_001138468.1:p.Arg1018Ter	nonsense	PSV1 + PM2 + PP3	P	[56]
	CEP152	NM_001194998.1:c.3925C>T	NP_001181927.1:p.Arg1309Ter	nonsense	PSV1 + PM2 + PP3	P	[56]
	SLC12A2	NM_001046.2:c.2804‐2A>G	/	splice site	PSV1 + PM2 + PP3	P	[56]
	SPATA7	NM_001040428.3:c.157C>T	NP_001035518.1:p.(Arg53Ter)	nonsense	PSV1 + PM2 + PP3	P	[56]
	SPINK1	NM_003122:c.194+2T>C	/	splice site	PSV1 + PM2 + PP3	P	[56]
	DNAH1	NM_015512:c.2168A>G	NP_056327.4:p.Glu723Gly	nonsense	PM2	VUS	[56]
	RTTN	NM_173630.3:c.350C>A	NP_775901.3:p.Ser117Ter	nonsense	PSV1 + PM2 + PP3	P	[56]
	PLEC	NM_201384.2:c.6715G>T	NP_958786.1:p.Glu2239Ter	nonsense	PSV1 + PM2 + PP3	P	[56]
	ZFPM2	NM_012082.3:c.1015G>A	NP_036214.2:p.Val339Ile	missense	PM1 + PM2	VUS	[56]
	DHX37	NM_032656.3:c.2792C>T	NP_116045.2:p.Ala931Val	missense	PM1 + PM2 + PP3	VUS	[56]
	WNT9B	NM_003396.4:c.938G>A	NP_003387.2:p.Arg313Gln	missense	PM2	VUS	[56]
		NM_003396:c.566G>A	NP_003387.1:p.Arg189Gln	missense	NA	VUS	[56]
	CTNND1	NM_001085458.1:c.2833G>T	NP_001078927.1:p.Glu945Ter	nonsense	PSV1 + PM2 + PP3	P	[56]
	COL4A6	NM_001287760:c.4608G>A	NP_001274689.1:p.Trp1536Ter	nonsense	PSV1 + PM2 + PP3	P	[56]
	CRELD1	NM_015513.4:c.257+1G>T	/	splice site	PSV1 + PM2 + PP3	P	[56]
	ZP1	NM_207341.3:c.199G>T	NP_997224.2:p.Glu67Ter	nonsense	PSV1 + PM2 + PP3	P	[56]
	INPP5E	NM_001318502.1:c.1532G>A	NP_001305431.1:p.Arg511Gln	missense	PM1 + PM2 + BP4	VUS	[56]
	FANCC	NM_001243743.1:c.996+1G>A	/	splice site	PSV1 + PM2 + PP3	P	[56]
	ERBB3	NM_001982.3:c.2900G>A	NP_001973.2:p.Arg967Lys	missense	PM1 + PM2 + PP2	VUS	[56]
		NM_001982.3:c.3637A>T	NP_001973.2:p.Arg1213Trp	missense	PM2 + PP2	VUS	[56]

Abbreviations: B, Benign; LB, Likely Benign; LP, Likely Pathogenic; NA, not available; P, Pathogenic; VUS, Variant of Uncertain Significance.

4.1. DNA Copy Number Variants in Other Müllerian Anomalies

Investigations of CNVs associated with other Müllerian anomalies are quite limited. Although researchers have conducted CNV analysis in HWWS, no significant findings were reported [58]. The pathogenic mutations documented in the literature are predominantly associated with uterine malformations and distal vaginal atresia (Table 3) [55, 56]. It is noteworthy that chromosomal regions recurrently implicated in MRKH syndrome are also frequently observed in these other anomaly subtypes. What's more, Kang et al. detected CNVs in distal vaginal atresia, with only three pathogenic findings, all confined to the 17q12 region [56]. This converging evidence indicates that the etiologies of different Müllerian anomalies may not totally separate but likely involve overlapping pathogenic pathways.

4.2. DNA Single Nucleotide Variations in Other Müllerian Anomalies

Congenital uterine malformations, such as septate and bicornuate uterus, primarily arise from abnormalities in Müllerian duct fusion. Researches on the genetic defects associated with uterine malformations are limited. Candidate genes associated with uterine malformations include TBX6, RBM8A, WNT9B, and LHX1 (Table 4) [35, 41, 48, 54, 57]. However, following evaluation using the ACMG guidelines, the majority of variants identified in these genes were classified as variants of uncertain significance (VUS). Notably, only one single TBX6 missense variant (NM_004608.3:c.815G>A), identified in a case of bicornuate uterus, was assessed as likely pathogenic [35]. Collectively, current evidence is insufficient to establish definitive genotype–phenotype correlations for Müllerian duct fusion defects. The relative paucity of research on genetic defects underlying fusion defects, such as septate, bicornuate, and unicornuate uterus, may be attributable to their clinical presentation. These anomalies often do not cause overt symptoms in non‐pregnant women. Although they are associated with adverse reproductive outcomes, including pregnancy loss and preterm birth, these complications are frequently manageable via surgical correction or assisted reproductive technologies [61]. Consequently, the imperative to elucidate their genetic etiology has been less urgent compared to the more profound clinical implications of MRKH syndrome, which necessitates immediate and complex medical and psychological intervention.

HWWS is a complex Müllerian anomaly involving failed absorption of the vaginal septum, often coexisting with fusion defects, which is characterized by a double uterus, double cervix, vaginal obstruction, and often presents with unilateral kidney malformation [62]. In addition to two missense variants of uncertain significance previously reported in TBX6 [35], Li et al. [58] proposed a potential genetic link between HWWS and genes associated with renal agenesis, including CHD1L, TRIM32, RET, and WNT4, based on whole‐exome sequencing of 12 affected individuals. In that study, a total of 11 variants across 9 genes were identified. However, subsequent analysis according to ACMG guidelines classified only one missense variant in TGFBR3 (NM_003243:c.1997_1998del) as pathogenic, while the remaining variants were interpreted as either VUS or likely benign. The aforementioned studies provide some evidence regarding the etiology of HWWS, but the sample size is too small, comprising only 12 patients. Future research needs to investigate the genetic etiology of HWWS using a larger patient sample.

Currently, there is limited genetic research on vaginal atresia. Notably, TBX6 defects represent the only recurrently implicated genetic alteration in distal vaginal atresia reported to date, supporting its role as an established pathogenic gene for this condition [56, 59]. Although a recent study on distal vaginal atresia has identified a substantial number of candidate genes, most represent novel associations reported for the first time. Among these, only TBX3 and AXL have been previously studied in murine models, where homozygous variants were shown to cause vaginal atresia in female mice. Future functional studies are warranted to validate the pathogenicity of these candidate genes in human [56].

Current evidence suggests potential etiological links between MRKH syndrome and other Müllerian anomalies. This is supported by the observation that chromosomal regions harboring recurrent CNVs in MRKH syndrome have also been identified in other Müllerian anomalies. Furthermore, SNVs of established MRKH candidate genes, such as WNT4, TBX6, and WNT9B, have been detected in individuals with other Müllerian anomalies. This overlap implies that, similar to MRKH syndrome, other Müllerian anomalies may also arise from polygenic mechanisms. Nevertheless, research into the genetic etiology of other Müllerian anomalies remains limited. Future studies should aim to expand cohort sizes to enhance statistical power and mirror approaches in MRKH research, prioritizing investigations into polygenic contributions alongside functional validation of candidate genes.

5. Multiple Factors Attribute to Müllerian Anomalies

The etiology of Müllerian anomalies is multifaceted and the genotype–phenotype relationship remains elusive. Researchers observed phenotypic discordance in monochorionic monoamniotic twin sisters: one exhibited normal Müllerian duct development while the other presented with MRKH syndrome [63]. This phenotypic discordance suggests that the genetic defects represent only one aspect of pathogenesis, while postnatal environmental influences, such as divergent intrauterine environments potentially leading to functional genomic alterations and differential gene expression may constitute another contributing etiological factor [64].

First and foremost are the genetic defects in Müllerian anomalies. To date, numerous sequencing studies have been published; the gene mutations reported in patients with Müllerian anomalies are largely sporadic cases characterized by de novo mutations, with a limited number of familial cases also described. Familial case reports have predominantly focused on MRKH syndrome, which indicate that at least a subset of MRKH syndrome may be heritable, following an autosomal dominant inheritance pattern characterized by incomplete penetrance and variable expressivity [65]. What's more, according to current studies, with the help of semi‐cloned mice researches elucidated that Müllerian anomalies tend to result from the combined effects of digenic/oligogenic pathogenic genetic variants rather than a single gene mutation [54].

Müllerian anomalies exhibit a complex genotype–phenotype relationship, whose pathogenesis cannot be simplistically attributed to genetic defects alone; the epigenetic regulation and aberrant gene expression patterns also play significant roles in the development of this disease. For example, in a recent study, Nik‐Zainal et al. reported a high incidence of recurrent CNVs in patients with sporadic and syndromic Müllerian aplasia, while no specific CNV was consistently detected in a larger patient cohort, suggesting that the epigenetic phenomenon may exist in Müllerian anomalies [16]. Epigenetics refers to the regulation of gene expression through chemical modifications of chromatin structure without altering the DNA sequence, with these modifications typically being mitotically heritable [66]. In 2011, Katharina et al. compared gene expression and methylation sites between MRKH syndrome patients and control groups, identifying significant differences. In MRKH patients, overexpression of genes such as GATA4, WT1, ESR1, and HOXA9, along with hypomethylation of specific CpG sites, disrupts Müllerian duct development, ultimately leading to the disease [67]. This underscores that elucidating epigenetic mechanisms and gene expression patterns is essential for clarifying the etiology of Müllerian anomalies.

Beyond the aforementioned, environmental exposures also play a significant role. Certain endocrine‐disrupting chemicals can affect the development of the female reproductive tract by altering gene expression and modulating developmental reprogramming. Prenatal exposure to these substances may lead to abnormal formation of the fetal reproductive tract [68]. Diethylstilbestrol (DES), a xenoestrogen and known endocrine disruptor, can alter the expression of genes critical for uterine patterning in mouse models, including Wnt7a, Hoxa9, Hoxa10, and Hoxa11, thereby disrupting reproductive tract development [69]. There are numerous endocrine‐disrupting chemicals similar to DES, such as Organotins, Phthalate Esters, Methoxychlor, and so on, having the potential to cause female reproductive tract malformations [70].

In summary, the pathogenesis of Müllerian anomalies involves a complex and continuous process, influenced by multiple factors including genetic alterations, epigenetic phenomena, and differential gene expression as well as environmental influences. Current studies on the etiology of Müllerian duct anomalies remain fragmented, precluding a comprehensive, and systematic understanding of its pathogenic mechanisms. We anticipate that advanced validation tools will pave the way for deeper insights into the etiology of Müllerian anomalies in the future.

6. Approaches to Verify Functions of Genes Related to Müllerian Anomalies

As mentioned above, an increasing number of genetic defects associated with Müllerian anomalies have been identified in recent years, but the functions of these genes remain unvalidated, likely due to both limitations in current validation methods and the complex genetic etiology of Müllerian anomalies. Recently, techniques such as semi‐cloned mouse models and pluripotent stem cell‐based studies have provided novel approaches for validating gene functions. The following sections summarize current approaches for functional validation of genes implicated in Müllerian anomalies (Figure 1).

FIGURE 1.

Strategies to evaluate the functions of genes potentially involving Müllerian anomalies. (a) cKO mice are generated using the Cre/loxP system. In this approach, Cre recombinase excises DNA flanked by parallel loxP sites. “Floxed” mice carry the target gene inserted between two loxP sites, while Cre‐driver mice express Cre recombinase under a tissue‐specific promoter (e.g., *Wnt7a‐Cre*). Crossing these lines produces offspring that carry both alleles. Cre‐mediated excision of the floxed gene in target tissues then achieves tissue‐specific knockout. (b) SC mice are generated using AG‐haESCs. The process begins with injecting a sperm head into an enucleated MII oocyte. Following activation, the embryo is cultured to the blastocyst stage. Due to spontaneous diploidization (> 30% frequency), AG‐haESCs are purified via flow cytometry, and morphological screening. DMRs of *H19/Igf2r* produces DKO‐AG‐haESCs, which improve live birth rates to ~20% and allow iterative gene editing. Finally, edited AG‐haESCs are electrofused with MII oocytes to form reconstructed embryos, which are transferred into pseudo‐pregnant females to yield SC mice. (c) iPSCs and Organoid Technology. Somatic cells can be reprogrammed into iPSCs using specific factors (such as OSKM). These iPSCs can then be cultured in 3D systems under defined conditions, specifically through organoid technology. cKO mice, conditional gene knockout mice, SC mice, Semi‐cloned mice, AG‐haESCs, androgenetic haploid embryonic stem cells, DMRs, deleting differentially methylated regions, iPSCs, pluripotent stem cells, 3D, three‐dimensional (By Figdraw).

6.1. Conditional Knockout Mice in Genetic Function Verification of Müllerian Anomalies

Conditional gene knockout (cKO) mice represent an advanced alternative to traditional germline knockout models, which are characterized by non‐specific gene disruption in all cells and potential embryonic lethality. cKO mice enable tissue‐specific gene knockout and are crucial for verifying gene function [71]. However, studies utilizing cKO mice specifically for Müllerian anomalies remain limited. Huang et al. generated Müllerian duct epithelium‐specific Lhx1 conditional knockout mice (Lhx1 cKO), using a WNT7a‐Cre transgene, and found that Lhx1 cKO female mice presented hypoplastic oviducts, and agenesis of the uterus, cervix, and upper vagina. This phenotype resembles MRKH syndrome, suggesting that LHX1 may play a key role in the pathogenesis of Müllerian duct dysplasia [72]. Recently, Thomson et al. investigated HNF1B and found that its deficiency causes hypoplastic development of the uterus and kidney anomalies, resembling the type II MRKH phenotype. They further analyzed the downstream molecules and pathways of HNF1B using single‐cell RNA sequencing of uterine tissues, revealing dysregulation in processes related to cell proliferation, migration, and differentiation [47]. However, cKO mice have their limitations. Firstly, generating these models requires prolonged breeding schemes and sophisticated genetic engineering techniques [73]. Additionally, due to the complexity of the animal's internal environment, compensatory effects may appear in cKO mice, which may cause phenotypes difficult to observe or explain [74]. Finally, since Müllerian anomalies often involve polygenic mechanisms, generating multi‐gene cKO mice through iterative breeding is prohibitively time‐consuming [54]. Therefore, developing efficient combinatorial gene‐editing platforms is essential for modeling polygenic disorders like Müllerian anomalies.

6.2. Semi‐Cloned Mice in Genetic Function Verification of Müllerian Anomalies

Semi‐cloned (SC) mice represent a novel model system that plays a significant role in verifying gene functions associated with Müllerian anomalies; these mice can be generated through various methods [75]. In 2012, Li et al. proposed using androgenetic haploid embryonic stem cells (AG‐haESCs) to generate SC mice [76]. AG‐haESCs are derived by injecting sperm into enucleated oocytes. The sperm can fuse with metaphase II (MII) oocytes to form semi‐cloned embryos, which develop into SC mice [77]. Through this approach, researchers can generate mice carrying multiple genetic modifications in a single step, significantly improving efficiency over traditional cKO methods [73]. Moreover, removal of the differentially methylated regions (DMRs) of H19 and Igf2r from AG‐haESCs yields DKO‐AG‐haESCs, which not only increase SC mouse birth rates from 4.5% to 20% but also support the SC pup generation even after multiple rounds of gene editing [78]. In 2020, researchers established a Müllerian anomaly mouse model with multiple gene variants using SC technology. Mice with heterozygous variants in both Gen1 and WNT9b exhibited more severe uterine malformations than single‐gene mutants, indicating synergistic roles in uterine development [54]. This provides the first evidence for a digenic etiology in Müllerian anomalies. While SC mice are valuable for validating genetic mechanisms in Müllerian anomalies, their application for identifying causal gene defects remains limited.

6.3. Induced Pluripotent Stem Cells and Organoids in Exploring Pathogenesis of Müllerian Anomalies

In addition to cKO mice and SC mice, induced pluripotent stem cells (iPSCs) hold potential for validating the genetic defects in Müllerian anomalies. iPSCs are reprogrammed cells with self‐renewal capacity and multi‐lineage differentiation potential, similar to embryonic stem cells [79]. Human iPSCs (hiPSCs) derived from patient primary cells retain individual genetic profiles and differentiate into specific lineages (e.g., cardiomyocytes, neurons) under defined conditions [80]. However, applying iPSCs to model tissue‐ and organ‐level diseases requires developing complex 3D multicellular systems [81]. Organoids, derived from pluripotent stem cells or tissue progenitors, are 3D culture systems that recapitulate native tissue architecture, function, and genetics [82, 83]. Combined with hiPSCs, organoids enable precise in vitro modeling of diverse human diseases, such as genetic disorders [83]. To date, diverse organoids have been established, spanning liver, lung, cardiac, and others [84, 85, 86]. Diverse human female reproductive tract organoids have also been generated, including endometrial, fallopian tube, ovarian, endocervical, and vaginal types [87]. These organoids primarily model reproductive tract carcinogenesis or explore fertility preservation strategies [88, 89]. As understanding of reproductive organoids deepens, this technology may address key mechanistic and therapeutic challenges in Müllerian anomalies, especially MRKH syndrome.

7. Future Early Diagnosis and Malformation Screening of Müllerian Anomalies

Müllerian anomalies negatively impact women's physical and mental health and their delayed diagnosis poses serious risks to patients [5]. Unfortunately, it is not currently possible to diagnose this anomaly by imaging techniques during the fetal period. Thus, prenatal genetic testing is a potential tool for early detection of Müllerian anomalies [90]. In theory, prenatal genetic testing allows researchers to detect CNVs or SNVs in the fetal genome, enabling the screening of high‐risk populations for Müllerian anomalies [91]. As more genetic defects are identified in Müllerian anomalies, gene sequencing data could enable AI‐assisted multimodal diagnostics in the future. However, current studies lack both high‐value variants for early fetal diagnosis and sufficient biological interpretation of known variants, hindering prenatal genetic testing.

Except early diagnosis, malformation screening is also crucial for ensuring a healthy child in cases of congenital diseases. This is especially important for MRKH syndrome, which is the most challenging aspect of malformations that require resolution. However, there are currently no guidelines about early diagnosis and malformation screening of MRKH syndrome. Evidence suggests that a family history is present in approximately 60% of individuals with MRKH syndrome, establishing it as a significant high‐risk factor and a valuable criterion for clinicians to identify susceptible populations. Furthermore, chromosomal CNVs are detected at a notably high frequency (up to 20%) in patients with MRKH syndrome. Commonly implicated genomic regions include 17q21, 22q11.21, 16p11.2, and 1q21 [12]. Consequently, prenatal chromosomal microarray analysis is strongly recommended for two groups: unaffected women with a family history of MRKH syndrome and women who conceive following uterine transplantation for MRKH syndrome [92]. The detection of CNVs within these critical regions in the fetus signifies a substantially elevated risk for female offspring to inherit this severe congenital anomaly, warranting prompt, and comprehensive genetic counseling. For other Müllerian anomalies, it remains challenging to identify high‐risk populations based on current research. It is hoped that future advances in elucidating their etiology will pave the way for better prevention and timely treatment of these anomalies.

8. Discussion

Müllerian anomalies place significant burdens on both the physical and mental health of women. Growing evidence indicates genetic defects as a key etiological factor in Müllerian anomalies, which have the potential for improved early diagnosis and genetic counseling of patients. Despite the extensive sequencing studies conducted, the genetic causes of these disorders remain unclear.

In MRKH syndrome, researchers have identified chromosomal regions that frequently exhibit CNVs, specifically 17q12, 22q11, 16p11.2, and 1q21.1. The commonly reported candidate pathogenic genes of MRKH syndrome include LHX1, HNF1B, ZNHIT3, TBX6, RBM8A, WNT4, WNT9B, PAX8, and GREB1L. In contrast to MRKH syndrome, the genetic etiology of other Müllerian anomalies is less explored. Based on a synthesis of genetic defects identified in MRKH syndrome and other Müllerian anomalies, we propose that the etiologies of various Müllerian anomalies cannot be fully discussed in isolation; the pathogenic mechanisms underlying different Müllerian anomalies may overlap. In addition, the etiology of Müllerian anomalies is better explained by a polygenic mode. And the potential roles of other factors, such as epigenetic modifications, differential gene expression, and environmental influences, cannot be excluded in their pathogenesis.

There are some major fields that should be further investigated in genetic research on Müllerian anomalies. Firstly, although sequencing studies have identified numerous variants, most lack functional validation and mechanistic interpretation. Additionally, most sequencing studies concentrate on individual candidate genes for MRKH syndrome, potentially leaving gaps in the genetic etiology research of other Müllerian anomalies. What's more, the complex physiology of the female reproductive tract and lack of parental genetic data hinder validation of inheritance patterns to some extent, especially in MRKH syndrome. Müllerian anomalies exhibit a complex genetic etiology. Therefore, it is essential to identify polygenetic defects associated with these anomalies rather than focusing solely on defects in individual genes. SC mice technology enables efficient construction of mouse models harboring multiple gene variants, facilitating study of polygenic mechanisms. While iPSCs and organoids currently have limited applications in Müllerian anomalies, they may be promising additions to our toolbox to gain a better understanding of Müllerian anomalies. In the future, with clarification of genetic mechanisms underlying Müllerian anomalies, early detection of severe anomalies through prenatal diagnosis and more meaningful genetic counseling are anticipated.

Author Contributions

Xiangyi Ma and Li Li directed the writing and designed the structure of the manuscript. Jingfang Li wrote the original manuscript, arranged the tables, and depicted the figure. Xin Hou made revisions to the review. Xiangyu Wang and Juan Li collected the literatures. All authors have reviewed and approved the manuscript.

Funding

This study is supported by the National Key Research and Development Program of China (grant number 2021YFC2701402) and Huazhong University of Science and Technology Teaching Reforming Program (54000‐3041540016).

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.