Testicular histological and immunohistochemical aspects in a post-pubertal patient with 5 alpha-reductase type 2 deficiency: case report and review of the literature in a perspective of evaluation of potential fertility of these patients

The 46, XY disorders of sex development (DSD) are presently classified in three main categories [1]: disorders of gonad development such as gonadal dysgenesis, disorders of androgen biosynthesis and metabolism and disorders related to androgen sensitivity (the androgen insensitivity syndrome, AIS). Within the second category, genetic causes have been identified, such as loss of function mutations of the LH receptor gene (LHCGR), 17β hydroxysteroid dehydrogenase deficiency type 3, related to mutations of the HSD17B3 gene and loss of function mutations of the SRD5A2 gene, responsible for 5α-reductase type 2deficiencywhich is one of the three 5α-reductase isoforms expressed in humans [2]. In the third category, androgen receptor (AR) mutations have been identified in patients with mild, partial or complete androgen insensitivity syndromes (AIS) [1].

AIS have variable phenotypic presentations, mainly related to the severity of the deleterious effects of AR mutations. Indeed, subjects with complete androgen insensitivity (CAIS) present as girls and women with feminine aspect of external genitalia but with absent pubic hair, blind ending vagina and uterine agenesis. In partial androgen insensitivity syndrome (PAIS) phenotypes, there are varying degrees of masculinization, ranging from perineoscrotal hypospadias to minor forms of male infertility, more or less associated with gynecomastia, undervirilization, hypotrophic gonads and micropenis [1].

5α-R2 deficiency phenotype is also variable, ranging from a complete female phenotype at birth to a more or less complete virilization of genitalia [3]. In girls with 5α-R2 deficiency, spontaneous virilization occurring at the onset of puberty usually reveals the condition [3].

In both pubertal and adult subjects with CAIS and 5α-R2 deficiency, very few studies of testicular histology showed severely altered spermatogenesis [4, 5]. The severe spermatogenesis impairement in CAIS and 5α-R2 deficiency could be related in part to the bilateral cryptorchidism [6] found in both conditions, but also by either altered levels of testicular testosterone action or the lack of metabolic activation to a more active androgen due to mutated 5α-reductase type 2.

An interesting cellular marker of Sertoli cells, used for the evaluation of the androgen signaling within seminiferous tubules is the anti-Müllerian hormone (AMH). AMH expression is usually recognized as a marker of differentiation of Sertoli cells and considered to be negatively regulated by androgens at puberty and adulthood [7, 8].

The first objective of this work was to evaluate the contribution of testicular testosterone versus its 5α-reduced metabolite, dihydrotestosterone (DHT), on AMH repression after puberty and its relation with spermatogenesis. We, therefore, compared the testicular morphology, as well as the AR and AMH expression in a postpubertal case of genetically demonstrated 5α-R2 deficiency with that of a postpubertal case of CAIS. We also compared our findings to published data describing the histological testicular features of post pubertal individuals considered to be 5α-R2 deficient, but in which the diagnosis was not genetically confirmed. Finally, we discuss the testicular histological findings in the more general context of the fertility potential of these patients in adult life, if they were assigned to a male gender.

The testicular samples of the 5α-R2 deficiency patient were obtained immediately after the bilateral gonadectomy performed when she was 17-years-old. The patient was a post-pubertal XY female with primary amenorrhea, failure of pubertal breast development and virilization, clitoromegaly and bilateral cryptorchidism, with testes located in the inguinal canals. This phenotype was related to a recurrent [3], homozygous, missense deleterious mutation in the exon 2 of the SRD5A2 gene (c.344G > A; Gly115Asp). The residual 5α-reductase type 2 activity in cells transfected with this mutant was of less than 0.2% [9] The detailed clinical and hormonal characteristics of this patient (testosterone-T: 7.2 ng/mL (normal range in post-bertal males: 3.5-8.5), DHT: 0.16 ng/mL (0.25-1.1), FSH:14.5 IU/L (2.7-7.4), LH: 4.2 IU/L (2.6-6.5), AMH: 65 pmol/L (15–89) as well as the results of the pelvic magnetic resonance imaging (MRI) have been previously described [10].

We comparatively analyzed the testicular histological aspect of this case with those of a CAIS and in a man with obstructive azoospermia but normal testicular spermatogenesis (see below). The CAIS patient was an 18-year-old, XY female, with primary amenorrhea (FSH: 3 IU/L, LH:18 IU/L) and absent axillary and pubic hair and high testosterone (19 ng/mL) and AMH (170 pmol/L) levels. On clinical and ultrasound examinations were observed a blind ending vagina, absent uterus and bilateral cryptorchidism (the left gonad was situated in the pelvis while the right one was located in the inguinal canal). In this patient, the diagnosis of CAIS was confirmed by the identification of a missense mutation in the exon 5 of the AR gene. This recurrent, loss of function mutation (c.2194G > A, p.Asp732Asn or D732N upon a previous classification), has already been reported in several CAIS patients [11] and its deleterious character has been previously demonstrated, as functional analysis revealed that this mutant had lost 95% of the wild-type transcriptional activity [11].

The testicular sample, used as a normal control, was obtained from a 29-year-old patient, with normal testosterone and gonadotropin levels (T: 6.8 ng/mL, FSH: 4.5 IU/L, LH: 4.2 IU/L, AMH: 33 pmol/L) undergoing testicular biopsy for obstructive azoospermia. Indeed, for ethical reasons, it was not possible to perform testicular biopsies in healthy men. However, as already reported, more than 86% of patients with obstructive azoospermia present normal spermatogenesis on testicular biopsies [12]. For this patient, the histological analysis of the testicular biopsy showed, as expected, normal structures and ongoing spermatogenesis within seminiferous tubules.

Herein, we report the histological and immunohistochemical testicular analysis of a posptubertal patient with 5α-R2 deficiency; the diagnostic was considered in front of the clinical presentation, associated with a high T/DHT ratio, and confirmed by the genetic analysis, which revealed a homozyguous missense mutation in the SRD5A2 gene (c.344G > A; Gly115Asp). The deleterious character of this recurrent mutation was clearly demonstrated by Wigley et al. [9] who showed a 99% loss of the enzymatic activity of the mutated protein. To our knowledge, we present the first testicular histological description in a postpubertal patient, with a genetic confirmation of the 5α-R2 deficiency. Indeed, the eight previous publications (Table 1) describing with more or less details the testicular histology in postpubertal patients, putatively affected by this pathology, are relatively old and lack any information related to patients SRD5A2 gene analysis [5, 19‐25].

Overall, histological features in those postpubertal patients with 5α-R2 deficiency seem to be rather heterogeneous, varying from complete lack of spermatogenesis to apparently normal spermatogenesis (Table 1).

The histological analysis of the testicular samples obtained from our patient revealed a heterogeneous aspect with a wide majority of seminiferous tubules presenting a mature aspect but a severely altered spermatogenesis, and containing only Sertoli cells. In a minority of seminiferous tubules of prepubertal appearance, where Sertoli cells exhibited a pseudostratified distribution, few germ cells were identified. Overall, the histological aspect of our patient’s testicular samples was similar to that described by Steger et al. [5], in three postpubertal patients (Table 1, patients 2, 9 and 10), presenting with clinical features resembling a 5α-R2 deficiency. The combined histological study of our case, and the analysis of the available literature (Table 1) also showed a strong percentage of impaired spermatogenesis (7/8, 87%) in cryptorchid 5 alpha-reductase type 2 deficiency patients [5, 21, 22]. These findings suggest that cryptorchidism by itself could be deleterious for spermatogenesis. Similarly, we compared serum dihydrotestosterone (DHT) levels in subjects with histologically normal spermatogenesis with those measured in patients presenting with impaired spermatogenesis (Table 1). Within the first group of patients (n = 4), mean DHT levels were of 0.26 ± 0.05 ng/mL whereas in the second group mean DHT levels were 0.12 ± 0.07 ng/mL. We therefore noticed a trend toward reduction in serum DHT levels in patients with impaired spermatogenesis suggesting that a more severe impairment in 5 alpha-reductase type 2 activity could also contribute to the spermatogenesis alteration.

Histological analysis of the testicular samples of the 5α-R2 deficiency patient reported here also revealed an oncocytic transformation of Sertoli cell cytoplasm in very few seminiferous tubules. This minoritary peculiar histological feature is similar to that described in several infertile subjects [26].The pathophysiological significance of this feature is still unknown, but it remains to be determined whether this oncocytic transformation may be associated to testis tumor development. Important areas of Leydig cell hyperplasia were also observed in our case; such morphological observation has been previously found in 5α-R2 deficient testis [22, 24, 25].

Similarly, Leydig cell hyperplasia has been already reported in CAIS postpubertal patients [4], or in men with Klinefelter syndrome [27]. In all these different pathological conditions, Leydig cell hyperplasia could be related to chronic Leydig cell stimulation induced by the long-lasting LH excessive secretion [14, 24, 27, 28]. The risk of progression of these lesions to a Leydig cell tumor is unknown in 5α-R2 deficiency but deserves to be monitored if the testes are preserved.

The immunohistochemical analysis also revealed that in our patient, the androgen receptor (AR) was expressed in all three testicular compartments (Sertoli cells, Leydig cells and peritubular myoid cells), suggesting the possibility of a global cellular response to the intratesticular testosterone in this case of 5α-R2 deficiency. Most notably, AR was expressed in 92% of the seminiferous tubules, namely in those with a mature aspect. However, AR expression was below detectable levels in 8% of the seminiferous tubules, specifically those with an immature aspect. Along this line, it is worth noting that AMH immunodetection was rather heterogeneous, as previously reported in a case of 5α-R2 deficiency with “Sertoli cell only” histology [5]. We showed that AMH was repressed in the majority of the seminiferous tubules, whereas AMH was only detectable in immature tubules lacking AR expression. These results suggest that intratesticular testosterone seems sufficient to repress AMH, when AR is expressed, in spite of the 5α-R2 deficiency, and thus irrespective of the conversion of testosterone into dihydrotestosterone. The absence of AMH repression in few seminiferous tubules, in which AR was lacking, is in accordance with previous data published [7, 8], supporting the need of AR expression in Sertoli cells for AMH repression. Interestingly, in our patient, AMH was repressed in Sertoli cells harboring mature morphology, even in the absence of germ cells. This suggests that testosterone could exert direct inhibitory effects on AMH expression, independently from spermatogenesis initiation [29]. Alternatively, the absence of spermatogonia in most seminiferous tubules of our patient might have led to functional alterations in Sertoli cells, disrupting their ability to synthesize AMH.

In our patient with 5 alpha-Reductase type 2 Deficiency, serum AMH level was within normal range for postpubertal male as reported in few cases with this condition [30]. This contrasts with the very high serum AMH levels we observed in our CAIS patient, which were in line with the increase in AMH levels reported in adult CAIS patients [31]. This difference in serum AMH levels in 5 alpha-Reductase type 2 Deficiency versus CAIS reinforces the hypothesis that testosterone conversion into dihydrotestosterone seems not essential for AMH repression.

Identification of various spermatogenesis stages in patients with 5α-R2 deficiency is of importance in the context of sex assignment and fertility preservation. Indeed, due to recent progress in the assisted reproduction techniques, successful results with testicular sperm extraction (TESE) and intracytoplasmic sperm injection (ICSI) are now reported in azoospermic patients, either with androgen insensitivity [14] or with Klinefelter syndrome [27]. Concerning 5α-R2 deficient patients, normal spermatogenesis with spontaneous fertility is extremely rare (two published cases) [32]. On the same way, spermatozoa detection in the seminal fluid is also a very rare event, found in only nine 5α-R2 deficient cases [33‐35], allowing either intrauterine insemination of the partner in one case [33], or an in vitro fertilization by ICSI, as recently reported for two patients [34, 35]. Except these uncommon cases, fertility remains an important issue for 5α-R2 deficient patients, notably those harboring a severe enzymatic defect that precludes spontaneous parenthood and sperm recovery in the ejaculate (i.e. azoospermic patients). In these latter cases, of unknown prevalence, the only fertility therapeutic possibility could be the testicular sperm extraction (TESE). This therapeutic option seems reasonable since as indicated above, TESE has been used as a successful therapeutic strategy in a number of patients with non obstructive azoospermia of various origins, including those with cryptorchidism [36]. In this context, a detailed testicular histological analysis is crucial since this is currently the only reliable method to know if there is spermatozoa in the testis and if spermatozoon are present, to carry out a cryopreservation for ICSI. Finally, a direct relationship between the SRD5A2 gene mutation severity and the presence of spermatozoa still remains an unsolved question, opening further studies with regards to the prognostic and risk stratification in relation with the severity of the enzymatic defect.

Written informed consent was obtained from the patient with 5α-R2 deficiency, as well as from both the CAIS patient and the individual with obstructive azoospermia, for genetic and histological analyses and for publication of this case report and the accompanying images. Copies of written consents are available upon request for review by the Journal Editor.

LV is PhD student at INSERM-Paris Sud Research Unit U693 and assistant at Department of Biophysics and Nuclear Medecine at Bicêtre Hospital, France. ML is head of the INSERM-Paris Sud Research Unit U693, France. JY is Professor of Endocrinology at Université Paris-Sud and Medical practitioner at the Department of Reproductive Endocrinology, Bicêtre Hospital, France.