Disorders of sex development: a genetic study of patients in a multidisciplinary clinic

Sex development is a process under genetic control directing both the bi-potential gonads to become either a testis or an ovary, and the consequent differentiation of internal ducts and external genitalia. This complex series of events can be altered by a large number of genetic and non-genetic factors. Disorders of sex development (DSD) are all the medical conditions characterized by an atypical chromosomal, gonadal, or phenotypical sex. Incomplete knowledge of the genetic mechanisms involved in sex development results in a low probability of determining the molecular definition of the genetic defect in many of the patients. In this study, we describe the clinical, cytogenetic, and molecular study of 88 cases with DSD, including 29 patients with 46,XY and disorders in androgen synthesis or action, 18 with 46,XX and disorders in androgen excess, 17 with 46,XY and disorders of gonadal (testicular) development, 11 classified as 46,XX other, eight with 46,XX and disorders of gonadal (ovarian) development, and five with sex chromosome anomalies. In total, we found a genetic variant in 56 out of 88 of them, leading to the clinical classification of every patient, and we outline the different steps required for a coherent genetic testing approach. In conclusion, our results highlight the fact that each category of DSD is related to a large number of different DNA alterations, thus requiring multiple genetic studies to achieve a precise etiological diagnosis for each patient.


Introduction
Sex development is a multistep process under genetic control, implying a delicate network of molecular events that direct both the bi-potential gonads to become either a testis or an ovary (sex determination), and the consequent divergent differentiation of internal ducts and external genitalia (sex differentiation). Correct dimorphic sex determination and differentiation achievement can be disrupted by a large number of genetic and non-genetic factors altering any of the molecular signals that specify sex-specific development of sex organs or endocrine function. The term disorders of sex development (DSD) embraces all the medical conditions characterized by an atypical chromosomal, gonadal, or phenotypical sex (1). Thus, a wide number of pathologies are included under the same DSD definition; they show different frequencies and their severity ranges from genital anomalies that do not impair sexual definition or functionality, such as hypospadias, to conditions characterized by sexual ambiguity or discordance between chromosomal and internal or external sex anatomy. In 2006, Hughes et al. (1) proposed the latest recommended classification of DSD, based on the sex chromosomal findings and on the step of gonadal development or phenotypic differentiation in which the alteration had occurred. Current understanding of the genetic control of sex development is still incomplete, resulting a low probability of determining the molecular definition of the causal defect in many of the patients with DSD. Anyhow, proper and thorough clinical evaluation and laboratory investigations are the necessary procedures for obtaining the most accurate diagnostic definition. This can more efficiently be achieved through a multidisciplinary assessment of patients performed by different dedicated specialists with long-standing experience and both pediatric and adult practice. In this study, we describe the results of studies carried out on 88 patients with DSD evaluated and followed in the outpatient 'Centre for diagnosis, care and treatment of DSD' at San Camillo Forlanini Hospital, Sapienza University of Rome. In total, we found a genetic alteration in 56 out of 88 cases, leading to the correct clinical classification of every patient. Each category of DSD was found to be related to a large number of different DNA alterations, thus requiring multiple genetic studies to possibly achieve a precise etiological diagnosis in every patient.

Subjects and methods
Patients A cohort of 88 individuals, aged from 1 day to 41 years affected by non-syndromic DSD, were fully evaluated at the DSD Centre of San Camillo-Forlanini Hospital, Rome (Italy), by an experienced multidisciplinary team including a pediatric surgeon, a pediatric endocrinologist, a clinical psychologist, and a clinical geneticist. Patients were identified on the basis of ambiguous genitalia or discordance among chromosomal, gonadal and/or phenotypic sex, or apparently minor genital abnormalities (Table 1, modified from Hughes et al. (1)). Patients in whom the presence of additional anomalies, such as dysmorphic features and skeletal or visceral abnormalities, was detected were excluded from the cohort with the exception of the three syndromic patients (cases 6, 27, and 32) included in the study. For each patient, hormonal, imaging, and

Karyotyping
Metaphase spreads were obtained from blood lymphocytes using standard procedures. Chromosome analysis was performed using standard G-bands by trypsin using Giemsa (GTG)-banding techniques on cultured lymphocytes.
Testing the presence or absence of SRY gene Recognition of the SRY sequence was carried out on genomic DNA through polymerase chain reaction (PCR) amplification with specific SRY and control (ZP3) gene primers as described by Cui et al. (6).

Fluorescent in situ hybridization
SRY translocations in 46,XX male patients and DMRT1 deletions were investigated by fluorescent in situ hybridization (FISH) using two specific probes selected from a public database (http://genome.ucsc.edu), respectively, for SRY and DMRT1 genes.

Direct sequencing
The search for DNA point mutations in SRY, DHH, NR5A1, SRD5A2, AR, AMH, CYP21A2, CYP11B1, RSPO1, and WNT4 genes was carried out by PCR followed by direct sequencing. SRD5A2, CYP21A2, AMH, and CYP11B1 were analyzed as described previously (7,8,9). Primer sequences and annealing temperatures employed for amplification The reference gene telomerase reverse transcriptase (TERT) was simultaneously quantified in a separate tube for each specimen. For each primer pair, the reaction efficiency parameter (R2) was assessed by a standard-curve analysis as reported in Supplementary Table 2. Results for each sample were expressed as N-fold changes in copies of each test exon, and normalized to TERT relative to the copy number of the test exon in the calibrator DNA, according to the following equation: amount of targetZ2 KDDCt (11).

Classification of patients with DSD
Classical cytogenetic techniques were employed in order to categorize each patient into the correct DSD class according to the karyotype. The five 46,XX testicular (SRY-negative) patients and the single 46,XX ovotesticular patient with DSD ( Fig. 1) were then investigated for the presence of RSPO1 point mutations. We identified the homozygous c.286C1GOA (p.I32_l95del) splicing mutation in case 6 ( Table 2). This patient was affected by 46,XX ovotesticular DSD associated with palmoplantar hyperkeratosis, congenital bilateral corneal opacity, and strabismus and this has already been reported (12). In total, we did not find any genetic alteration in four out of seven 46,XX testicular patients with DSD.
One of these patients, in whom the hormonal profile was indicative of a rare form of the disease, was secondly analyzed for CYP11B1 point mutations. In total, we found CYP21A2 DNA alterations in 14 out of 18 (78%) patients (  . 1), demonstrated to be negative in a previously performed array comparative genomic   Table 2) showing complete gonadal dysgenesis carried a heterozygous whole NR0B1 gene duplication. We identified a heterozygous deletion encompassing the DMRT1 gene in patient 32 (Table 2) manifesting partial gonadal dysgenesis. Finally, we did not detect a WNT4 imbalance in any of the investigated patients.

DSD 46,XY disorders in androgen synthesis or action
Sequence analysis of AR and SRD5A2 genes was performed in 29 patients with suspected DSD 46,XY defects in the synthesis or action of androgens (Fig. 1). The AMH gene was sequenced in a single case with the clinical diagnosis of persistent Mü llerian duct syndrome. We found a genetic alteration in 24 out of 29 (83%) patients (Fig. 2). Among them, 15 individuals (    and c.165_788del (p.L56_L263del)). Eight patients carried a total of 12 different alterations in the SRD5A2 gene in compound heterozygosity (  367COT (p.R123W)) and a novel nonsense (c.564COA (p.C188X)) mutation in the AMH gene. In total, we found a genetic alteration in 56 out of 88 (64%) patients (Table 2). Figure 3

Discussion
DSDs are complex conditions related to a vast number of different causes. Establishing the specific etiology may be crucial for choosing the more adequate sex of assignment, for the clinical management of patients, and to permit the family to plan informed further pregnancies. However, molecular characterization cannot be reached in a consistent number of cases, due to the still limited knowledge of etiological determinants. In this report, we describe the clinical assessment and the cytogenetic and molecular findings in a large cohort of patients with DSD. In our hands, it was possible to identify a genetic defect in 64% of them and to assign each of the examined patients to a specific category in accordance with the current DSD classification. Accordingly, a specific survey could be planned for each patient. This study excluded 45,X as well as 47,XXX and 47,XXY patients and individuals with sex chromosome mosaicism identified during prenatal diagnosis, but that did not display abnormalities of the genital tract after birth. Karyotype analysis performed during this study showed that five out of 88 (6%) patients harbored mosaic sex chromosome anomalies, confirming that standard cytogenetic analyses can detect frequent   genetic causes of DSD. Moreover, the initial classification based on clinical and cytogenetic findings was revealed to be an important starting point to carry out the further appropriate molecular testing, specific for each DSD subgroup. Out of 88 patients, 7 (8%) were classified as 46,XX testicular DSD and, between them, two out of seven carried a SRY translocation onto the pseudoautosomal region PAR1 of one X chromosome. This aberration is considered the major cause of testicular development in individuals with 46,XX testicular DSD (20,21,22). Conversely SRY translocation appeared to be involved only in less than a third of our cases. Thus, we assumed that other molecular determinants were responsible for the other 46,XX testicular DSD cases not carrying the SRY translocation. The involvement of the SOX9 gene has already been demonstrated in a number of 46,XX testicular patients with DSD (5,23,24). Particularly, Cox et al. (24) and Vetro et al. (5) reported a 178-kb duplication and a 96 kb triplication, respectively, 600kb and 500 kb upstream of SOX9, in 46,XX, SRY-negative male patients. These alterations were assumed to enhance the promoter activity leading to SOX9 overexpression. In this study, we demonstrated the presence of a SOX9 exon 1 duplication in a 46,XX testicular patient with DSD ( Table 2, case 9) not harboring an SRY translocation. Although it is not possible to affirm with certainty that the duplication identified in our patient is causative of his DSD, data from the literature permit speculation about its role in the determination of the abnormal gonadal development (5,23,24). Owing to the lack of the patient's DNA, it was not possible to investigate whether the rearrangement identified in case 9 extended upstream of SOX9, but we cannot exclude the involvement of its promoter. In addition, as patient 9 belongs to north African ethnic group and lives in Africa, DNA neither from other family members nor from healthy controls of his population was available for testing the origin and the possible recurrence of the rearrangement. Our series of SRY-negative 46,XX testicular patients with DSD were also investigated for RSPO1 gene alterations as this gene has already been described as recessively mutated in two familial cases with 46,XX testicular DSD (25). Those patients showed genital anomalies accompanied by additional features, in particular palmoplantar hyperkeratosis. We did not find any RSPO1 point mutations in our 46,XX testicular DSD cases, implying that the RSPO1 gene may not be involved in 46,XX testicular DSD without palmoplantar hyperkeratosis. Our series of patients included a single 46,XX ovotesticular DSD case showing palmoplantar hyperkeratosis. This patient was born from consanguineous parents and harbored the c.286C1GOA (p.I32_I95del) RSPO1 mutation in homozygosity (12). The 21-hydroxylase deficiency is considered the most frequent cause of DSD with genital ambiguity. Genetic analysis of CYP21A2 performed in patients with a definitive or presumptive clinical diagnosis of adrenogenital syndrome allowed the identification of the molecular defect in 14 out of 18 (78%) cases. Among the four negative patients, case 24 was afterward recognized to be affected by a very rare form of congenital adrenal hyperplasia related to 11-bhydroxylase deficiency. This patient was born as the result of a consanguineous mating and presented ambiguous genitalia at birth. Her clinical and hormonal profile could be defined only after the first weeks of life, when the CYP21A2 study had already been started. She was found to carry the homozygous p.G379V alteration in the CYP11B1 gene. The remaining three out of 18 CYP21A2-negative patients, in whom adrenogenital syndrome was suspected, showed regression of clitoral hypertrophy throughout the late neonatal period. The evolution of their clinical presentation together with the molecular and hormonal findings led to definitive exclusion of the initial diagnostic hypothesis in these infants.
Regarding 46,XX patients with DSD with abnormal development of the Müllerian structures, our results demonstrated the presence of a duplication involving SHOX exon 6 in two out of ten unrelated cases with MRKH syndrome, a condition of still mostly unclear etiology. Nevertheless, our targeted investigations permitted replication of the results obtained from the study by Gervasini et al. (4), which reported a SHOX intragenic duplication in five patients with abnormal development of the Müllerian ducts. Although the mechanism that may relate SHOX duplications and the development of Müllerian ducts has not been clarified, our data and those described by Gervasini (27,28), it is possible that the involvement of this gene is restricted to cases with an atypical form of the syndrome.
Investigations of patients affected by 46,XY DSD with a defect in testicular development led to the molecular characterization of six out of 16 (37%) cases.
These outcomes are consistent with the still incomplete understanding of the molecular events that underlie testicular development, indicating the need to search for novel genes associated with gonadal dysgenesis in 46,XY patients. The NR5A1 gene, studied in 46,XY patients with a diagnosis of partial gonadal dysgenesis, was found to be mutated in heterozygosity in 3 out of 8 (37%) cases. These results consistent with those described in recent reports that identify mutations in NR5A1 as a major cause of 46,XY DSD with a defect in the testicular development. Among the three mutations identified in NR5A1, the genetic location of c.86COT (p.T29M) seems to affect the binding of the protein to DNA, while p.L231_233dup and p.V291 lay in the domain regulating transcription after hormone binding. Interestingly, in the three NR5A1-mutated patients, no Müllerian structures seemed to be present. The analysis of the DHH gene yielded negative results in all cases of 46,XY DSD with partial gonadal dysgenesis, even if DHH alterations have already been described as the possible cause of a consistent number of 46,XY DSD cases with a defect in the testicular development (29).
Concerning patients with 46,XY DSD with a defect in the synthesis or action of androgens, 24 out of 29 (83%) cases were characterized at a molecular level. Among the 12 different identified SRD5A2 genetic alterations, one maps in the transmembrane region, possibly affecting the protein localization, and 11 in the protein catalytic domain. Sequence analysis of the AR gene identified 13 different mutations, including nine alterations lying in the functional protein domains: three out of nine in the zinc finger domain responsible for the DNA binding, and six out of nine in the domain that regulates the transcription after hormone binding. The incomplete diagnostic sensitivity of the applied molecular studies in 46,XY patients with DSD with a defect in the synthesis or action of androgens might be the cause of the failure of diagnosis in those partial androgen insensitivity syndrome (PAIS) patients for whom negative results were obtained according to AR analysis. These patients may indeed harbor DNA alterations in non-canonically investigated AR regions (introns or regulatory sequences), or in different known or as yet unidentified genes.
Our results highlight that each category of DSD is related to a large number of different DNA alterations, thus requiring multiple genetic studies to possibly achieve a precise etiological diagnosis in every patient. Currently, as a consequence of the incomplete knowledge concerning the genetic factors involved in the differentiation of testes and ovaries, DSD associated with anomalies in gonadal development still often lacks a molecular diagnosis.
A multidisciplinary and specialized DSD center is the key for the correct clinical management of neonates in cases of ambiguous genitalia. Moreover, the introduction of new technologies for massive parallel sequencing is becoming helpful for the molecular characterization of patients with DSD by analyzing previously known genes as well as candidate genes.
Supplementary data This is linked to the online version of the paper at http://dx.doi.org/10.1530/ EC-14-0085.

Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

Funding
This research did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector.
Author contribution statement L Laino participated in the design and coordination of the study, carried out the cytogenetic studies, and drafted the manuscript; S Majore provided the genetic counseling to the patients, established the clinical diagnosis, participated in the study design, and helped to draft the manuscript; N Preziosi carried out the molecular genetic studies; B Grammatico participated in the design and helped to draft the manuscript; C De Bernardo participated in the design and helped to draft the manuscript; S Scommegna established the clinical diagnosis, participated in the study design, and helped to draft the manuscript; A M Rapone provided the psychological support to the families of the patients; G Marrocco conceived the study and gave the final approval of the version to be published; I Bottillo conceived the study, drafted the manuscript, and gave the final approval of the version to be published; and P Grammatico conceived the study and gave the final approval of the version to be published. All authors read and approved the final manuscript.