Genes in that group mainly encode transcriptional regulators and have been identified as important for the development of the embryonic intermediate mesoderm and the initial formation of the ducts [86]. It is well known that the development of the Müllerian duct is induced and dependent on the presence of the Wolffian duct [82]; hence genes involved in the formation of the latter are also of great importance. Genes essential for the initial, biphasic process of Müllerian duct development are the Lim Homeobox 1 (LHX1), Paired box 2 (Pax2), Empty spiracles homeobox 2 (Emx2), Dachshund homologs 1 and 2 (Dach1, Dach2) and the GATA binding protein 3 (GATA3) [67]. Those genes have been suggested as candidates for genital tract malformations on the basis of similar phenotypes observed mainly in mutant mice.

LHX1 (chromosome 17q12) encodes a transcription factor which plays an important role in early mesoderm formation and later in lateral mesoderm differentiation. Absent Wolffian and Müllerian duct derivatives in LHX1-null mice reveal that the particular gene is required for the formation of both sexual ducts [55, 102]. LHX1 disruption results in reduced expression of Pax2, another transcription factor acting as marker of the Wolffian duct. In human, heterozygous LHX1 mutations have been reported in sporadic cases of MRKH syndrome [59]. Those findings however were not confirmed by other researchers studying similar anomalies [101].

Pax2 (chromosome 10q24) is involved in the formation of the epithelial components derived from the intermediate mesoderm. Homozygous mutant mice lack kidneys, ureters and genital tracts in both males and females [92]. In human however genetic analysis did not demonstrate any significant association between molecular variants at this locus and the occurrence of MRKH syndrome [97] or other Müllerian anomalies [99].

EMX2 (chromosome 10q26) encodes a transcription factor and is homolog to the “empty spiracles” gene in Drosophila. In humans, apart from its main expression on the developing dorsal telencephalon, it is also expressed on epithelial tissues of the developing urogenital system. Homozygous mutant mice completely lack the urogenital tract in both males and females. No such defects have been observed in heterozygous mice [67, 72]. However, there are no studies suggesting a similar association of Müllerian anomalies with mutations of the particular gene in human.

Dach1 (chromosome 13q22) and Dach2 (chromosome Xq21) encode transcriptional factors which participate in the molecular cascade of Müllerian duct development. Inactivation of each corresponding gene does not affect genital development. Combined knock-out mice however demonstrate drastic defects of Müllerian derivatives (hypoplastic oviducts, severely hypoplastic uterine horns, aplastic vagina) [32]. It is possible that those two genes act redundantly to control development of the female genital tract.

Insulin-like growth factors (IGFs) encode proteins similar to insulin in function and structure, involved in mediating growth and development. IGF1 (chromosome 12q23) is believed to have a role in the developing rat uterus, as loss of gene function in mice results in severe uterine hypoplasia [7, 44]. Retinoic Acid Receptors (RARs) regulate gene expression in several biological processes. Null mutations lead to various developmental anomalies, including severe urogenital defects. In particular, RARαβ2 double mutant mice lack Müllerian ducts [69]. Similar defects (cervical and vaginal aplasia) have also been described in Disks large homolog 1 (Dlgh1) mutant mice, though no relation has been established between those genes [52].

GATA3 encodes a transcription which is a regulator of T-cell development and plays an important role in endothelial cell biology. Mutations of this gene have been reported in women with Hypoparathyroidism-Deafness-Renal dysplasia (HDR syndrome) and Müllerian duct fusion defects (didelphys or septate uterus) or vaginal atresia [51, 77]. It is not clarified though, if the mutation is the aetiology of the HDR syndrome or the aetiology of the genital tract anomaly.

Although a number of cases with severe anomalies of the reproductive tract have been attributed to mutations in the genes involved in the early development of the ducts in mice, molecular progress in similar malformations has been disappointing in human. Indeed, no association has been identified with most of those genes.

Scientists were prompted to investigate the role of such genes based mainly on the association of MRKH syndrome with galactosemia and cystic fibrosis. The most well studied genes in this group involve the galactose-1-phosphate uridyl transferace (GALT), the cystic fibrosis transmembrane regulator (CFTR), the transcription factor 2 (TCF2) gene [formerly known as HNF1 homeobox b or hepatocyte nuclear factor 1-beta (HNF-1β)] and the wilms tumor 1 (WT1) gene.

The findings regarding a possible association of the N314D allele of GALT (chromosome 9p13) with MA have been contradictory. In a study by Cramer et al. [29], 46 % of the MRKH patients exhibited the N314D allele compared to 14 % of the control group. However these results were not confirmed by subsequent studies [12, 54, 103].

Mutations of the CFTR gene (chromosome 7q31) have been associated with congenital bilateral absence of vas deferens in some males. The incidence of such mutations in cases of MRKH syndrome (8 %) was found to be twice as high compared to the general population (4 %), but significantly lower than the incidence of CFTR mutations in men with aplasia of the vas deferens (80 %) [90]. Those results suggest that such mutations do not cause MA in women in a similar way that they cause vas deferens agenesis in some men.

TCF2 mutations have been associated with MODY-type diabetes, diabetes mellitus and renal defects. It is interesting that similar mutations were found in some familial cases of genital tract anomalies, mainly bicornuate uterus, didelphys uterus and MA, co-existing with renal anomalies [15, 62]. The heterogeneous genital malformations and the absence of a direct genotype/phenotype correlation however do not suggest a straight aetiological association with TCF2 defects.

WT1 gene (chromosome 11p13) is involved in the development of both the internal and external genital organs. No mutations or polymorphisms have been found in a number of MRKH patients studied, suggesting that its expression is possibly required at a later stage of development, well after the initial formation of the ducts [96].

HOX genes encode numerous transcription factors (Homeoproteins), which are expressed along various developmental axes of the body and control embryonic morphogenesis. Some of the HOX genes are involved in the formation of the genitourinary tract and their deletions seem to result in renal agenesis and reproductive tract malformations. The embryonic female genital tract could be considered a developmental axis; the initially uniform Müllerian duct will finally form the oviducts, the uterus, the cervix and the upper part of the vagina.

HOX genes belonging to paralogue groups 9–13 seem to provide the axis of the developing paramesonephric duct with a positional identity: HOX A9 is expressed in areas designated to form the future oviduct, HOX A10 is mainly expressed on the developing uterus, HOX A11 is expressed on parts of the Müllerian duct which will form the lower compartment of the uterus and the cervix and HOX A13 is expressed on the upper third of the vagina. There is no HOX A12 gene; this has been possibly lost during evolution [35]. As those genes provide regional identity and specify the segmental body plan, their defects could be involved in the aetiology of severe Müllerian anomalies.

Genital tract malformations have been observed in HOX A10, HOX A11 and HOX A13 mutant mice. Such mutations lead in region-specific defects along the female genital tract. In human though, apart from some non-specific, rare polymorphisms and mutations found in sporadic cases of Müllerian anomalies, most researchers did not find genetic perturbations of HOX A9 to HOX A13 genes in the vast majority of patients with MA or other severe genital tract anomalies studied [17, 36, 57, 60].

HOX A13 is the most well studied gene in anomalies resulting from abnormal fusion of the Müllerian ducts. A variety of HOX A13 mutations (nonsence, missence, polyalanine tract expansions) have been associated with a rare, dominantly inherited condition called Hand-Foot-Genital syndrome (HFGS), which involves skeletal and urogenital (incomplete Müllerian fusion) malformations [34, 42, 75, 94]. Female genital tract defects range from isolated vaginal septum to didelphys (bicorporeal) uterus [41]. Such mutations have only been found in the context of the syndrome and not in sole fusion defects of the paramesonephric ducts [88].

The Pre-B-cell leukemia homeobox1 (PBX1) gene (chromosome 1q23) encodes an essential co-factor for HOX proteins that is expressed on the Müllerian ducts. Inactivation of the gene in mice does not result in congenital anomalies. Similarly in human, no mutations have been found in cases of Müllerian aplasia or other genital malformations studied [17, 63].

Short stature homeobox (SHOX) gene (chromosomes Xp22 and Yp11.3) controls fundamental aspects of growth and development. In contrast to other genes of the HOX family, it is absent in the mice. No obvious role of this gene in the development of the female reproductive tract has been reported in the literature. Although partial duplications of the gene have been reported in sporadic cases of MA [39], no association was confirmed in an extensive cohort of patients with similar anomalies [84].

Wnt genes seem to contribute significantly in the patterning and differentiation of the female genital tract [53]. They encode a number of cysteine-rich secreted growth factors and they guide the epithelial-mesenchymal interactions that direct uterine development. Wnt4, Wnt5a and Wnt7a are mainly expressed on the developing Müllerian duct and deficiency of those genes in mice results in a wide range of genital malformations [71].

Wnt4 (chromosome 1p36-p35) presents both an anti-testis function by repressing male differentiation and a pro-ovary function by supporting germ cells [14]. Homozygotic inactivation in mice results in total failure of Müllerian duct development [49, 95]. In human, homozygotic inactivation results in the SERKAL syndrome (female-to-male sex reversal, dysgenesis of kidneys, adrenals and lungs) which is embryonic lethal [66]. Recent studies suggest that there are no genetic alterations involved in the aetiology of MRKH syndrome or other Müllerian duct abnormalities [19, 25]. Interestingly, mutations have been described in two women presenting with absence of Müllerian duct derivatives, unilateral renal agenesis and androgen excess [13, 14]. It is possible that Wnt4 deficiency results in a phenotype that is similar but certainly different to the classic MRKH syndrome, as it is characterized by hyperandrogenism.

Wnt5a (chromosome 3p21-p14) mutated female mice present with a shortened uterus and poorly defined cervix and vagina [102]. Evidence from human studies is quite limited, however no causal Wnt5a mutations were recently observed among 189 Chinese women [100].

Wnt7a (chromosome 3p25) plays an important role in guiding uterine growth and hormonal responses. It is possible also that it mediates the expression of anti- Müllerian Hormone Receptor type II (AMHR2). Wnt7a mutations in mice result in severe changes in the size, the morphology and the cytoarchitecture of the uterus [18, 70, 91, 102]. In human, apart from some sporadic polymorphisms, no mutations have been detected in women with various malformations of the genital tract, suggesting no correlation [31, 91].

The genes for the AMH (chromosome 19p13) and its receptor (AMHR) have been considered as candidate genes for cases of aplasia, as they are responsible for Müllerian duct regression in male fetuses. In a similar way to the genetic males, activating mutations could cause Müllerian duct regression in a genetic female during embryogenesis [61]. Indeed, anomalies similar to those observed in the human MRKH syndrome were evident in female transgenic mice over-expressing AMH [9]. However, no mutations in those genes have been found in association with uterine aplasia or other anomaly of the genital tract in female mice. Possible role of other genes participating in the AMH signaling pathway (ALK2, ALK3) cannot be excluded from the aetiology of such anomalies [80]. Interestingly, Müllerian duct regression was evident in cases of ALK6 knock-out mice [24].

Apart from some rare polymorphisms (present both in patients and controls), no deleterious mutations of AMH/AMHR genes have been detected so far in women with MRKH syndrome [81, 103]. Other potential mechanisms have also been suggested, like high maternal AMH levels during pregnancy. Estradiol (E2) has been shown to induce AMH expression in vitro [20]. It can be assumed that high E2 levels or exposure to other estrogen-like substances in early pregnancy could induce AMH expression, resulting in Müllerian duct regression in the developing female fetus. On the other hand, over-expression of AMH would only justify a complete lack of Müllerian derivatives, which is not a common finding; the vast majority of anomalies correspond to partial rather than total agenesis.

Estrogens seem to be involved in the organogenesis and differentiation of the female genital tract. It is well known the biologic responses to estrogens are mediated through the estrogen receptors. Both types of those receptors, ERa (chromosome 6q25) and ERb (chromosome 14q23), are expressed on the mesenchyme and the epithelium of the paramesonephric ducts and possibly function as ligand modulated transcription factors. As a result of estrogen binding, the ER undergoes a conformational change which allows dimerization, DNA binding and recruitment of co-factors. The final result is either transcriptional activation or regression of target genes, mainly HOX and Wnt [3]. ERa deficient mice commonly present with a hypoplastic uterus and vagina [26].

In human, there have not been ER mutations reported in relation to female genital tract malformations. However, other mechanisms of ER involvement in the pathogenesis of such anomalies could be suggested. Estrogen and progesterone regulate HOX gene expression in both the embryonic reproductive tract and the adult reproductive tract. HOX A10 and HOX A11 expression is up-regulated by 17β-estradiol and progesterone. The regulation is direct and is achieved by the estrogen or the progesterone receptor binding to regulatory areas of the genes [16]. Wnt-7a mediates normal growth in the absence of estrogenic activity but is also required at the time of the initial estrogenic response which induces increased cellularity of the uterine tissues [18]. It is possible that altered expression of ER genes leads to impaired expression of certain HOX and Wnt genes along the developing Müllerian ducts, which in turn results in abnormal phenotypes of the female genital tract.

A number of chemicals released in the environment can bind to the ERs and exhibit estrogenic activity similar to 17β-estradiol. Epidemiological studies have shown that those chemicals, known as environmental endocrine disruptors or xenoestrogens, had an adverse impact on the woman’s health and fertility over the past few decades. Apart from carcinogenesis, they have also been implicated in the pathogenesis of the congenital anomalies of the female genital tract [79]. It seems that the carcinogenic and teratogenic defects of such endocrine disruptors are caused after binding to the ERa [27, 28]. Moreover, it is not unlikely that those adverse effects might have been transmitted to subsequent generations through epigenetic modifications [67].

Diethylstilbestrol (DES) is a non-steroidal estrogen and the most well known example of a chemical compound with an adverse effect on the woman’s reproductive health. Millions of women had been prenatally exposed to DES in the past, until the increased incidence of genital tract malformations and tumors became evident [50].

DES-induced malformations (T-shaped uterus, class U1a) were the result of abnormal morphogenesis of the Müllerian ducts and were similar to anomalies observed in HOX A10 mutant mice, with transformation of the upper part of the uterus into an oviduct-like morphology. According to developmental studies, prenatal DES administration shifts the expression of HOX A9 from the oviducts to the uterus and decreases both HOX A10 and HOX A11 expression on the uterus. Decreased expression of the genes that provide uterine identity and increased expression of a gene providing oviductal identity seems to be the cause of the T-shaped uterus, characterized by branching and narrowing into a tube-like form. The uterus in not fully transformed into an oviduct possibly due to the redundancy provided by other HOX genes [35].

Targeted mutations of ERa in mice prevent the effects of DES on HOX expression. It is possible that DES impairs the conformation of the ER, so as the receptor to interact selectively with atypical coactivators or corepressors, inducing differential HOX gene activation which in turn leads to genital tract malformations [28]. DES shifts the expression pattern of HOX genes also in human uterine cell cultures, suggesting a similar role in human uterine malformations [40].