Biallelic mutations in PSMC3IP are associated with secondary amenorrhea: expanding the spectrum of premature ovarian insufficiency

Abstract

Premature ovarian insufficiency (POI) has a strong genetic component, but, in most cases, the etiology remains unidentified. PSMC3IP is an autosomal recessive gene for POI and ovarian dysgenesis, and so far, biallelic mutations in this gene have been described in only four independent families, with all affected members showing primary amenorrhea. Here, we report on the first family with recessive variants in the PSMC3IP gene and POI in a patient with secondary amenorrhea. Whole-exome sequencing (WES) was performed on a 29-year-old woman with secondary amenorrhea and POI; she was found to carry compound heterozygous variants in the PSMC3IP gene: c.206_208delAGA and c.189 G > T. Her younger sister, who also presented with a suspect of POI due to infertility and very low levels of anti-müllerian hormone (AMH), was found to carry the same PSMC3IP variants. Our case report shows the importance to include PSMC3IP in designed POI NGS panels or in WES/WGS studies in patients with either primary or secondary amenorrhea.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10815-022-02471-7.

Introduction

Premature ovarian insufficiency (POI) is the term defining amenorrhea before 40 years of age. It has a strong genetic component and can present with either primary or secondary amenorrhea, because of early ovarian exhaustion [1, 2]. Its incidence in the general population is 1% [3] and a recent global prevalence study concluded that the pooled prevalence of POI is as high as 3.7% (95% confidence interval [CI] 3.1–4.3) [4].

POI is characterized by infertility and hypergonadotropic hypogonadism (syndromic or non-syndromic), resulting in low estradiol and elevated follicle-stimulating hormone (FSH), within the menopausal range. It can be the consequence of either follicle depletion or follicle dysfunction. Reduced number of initial germ cells, increased apoptosis, or destruction of primordial follicles can cause follicle depletion; conversely, failure of follicles to respond to hormonal stimuli underlines follicle dysfunction [5, 6]. POI is a highly distressing condition, which deserves great attention due to the multiple implications for women’s health and well-being [7].

The most common genetic cause of POI is Turner syndrome due to monosomy X in complete or mosaic forms. Chromosomal aberrations such as X chromosome deletions and translocations account for 5–10% of all cases with POI [8]. Approximately 2–5% of sporadic cases with POI and secondary amenorrhea can be explained by the FMR1 premutation. Indeed, the prevalence of this condition is around 1/250 and up to 20% of FMR1 premutation carriers present with POI [6, 9, 10].

In the past few years, next-generation sequencing techniques such as whole-exome sequencing (WES) or whole genome sequencing (WGS) have allowed the discovery of a high number of genes responsible for non-syndromic POI. Mutations in identified genes have been demonstrated to carry deleterious effect on the underlying germ cell pool and/or on steroidogenesis [2, 11, 12]. To date, over 50 genes encoding transcription factors, oocyte secreted factors, hormone receptors, and proteins involved in meiotic recombination and or DNA repair have been found to cause ovarian dysfunction in POI [12–15]. Despite all available genetic causes of POI, its etiology remains unknown in most clinical cases [5, 11, 13].

PSMC3IP is an autosomal recessive gene for POI and ovarian dysgenesis, and so far, biallelic mutations in this gene have been described in only four independent families with primary amenorrhea [13, 16–18]. PSMC3IP has nine exons and encodes a nuclear protein that functions in meiotic recombination and works as a coactivator of ligand-dependent transcription mediated by nuclear hormone receptors [13, 16]. The N-terminal amino acids constitute the double-strand DNA-binding domain, the C-terminal region harbor the single-strand DNA-binding domain, and the RAD51/DMC1 interaction site for ligand-dependent transcriptional coactivation [16]. The PSMC3IP protein binds to the alpha- and beta-estrogen receptors as well as the glucocorticoid, thyroid, androgen, and progesterone receptors, and acts as a coactivator of hormonally dependent transcriptional activation [19, 20].

Zangen et al. [16] described a consanguineous family with several female members affected by POI showing primary amenorrhea and streak or undetectable gonads. All patients shared a homozygous 3-bp deletion (c.600_602del, p.Glu201del) in the PSMC3IP gene detected by linkage analysis and WES [16].

Yang et al. [13] reported on a patient carrying loss of function compound heterozygous variants in PSMC3IP (c.496_497delCT, p.Arg166Alafs; c.430_431insGA, p.Leu144*) and presenting with primary amenorrhea and non-visualized ovaries on ultrasound examination.

Al-Agha et al. [17] identified a consanguineous Yemeni family in which four daughters presented with POI and a brother had azoospermia. They all shared a homozygous nonsense mutation in exon 6 of the PSMC3IP gene (c.489 C.G, p.Tyr163*).

Finally, Mei et al. [18] reported on a Chinese patient with primary amenorrhea and bilateral small dysgenetic ovaries, carrying two compound heterozygous variants in PSMC3IP c.597 + 1G > T and c.268G > C.

Here, we report on the first family with recessive variants in the PSMC3IP gene and POI in which the proband was a patient with secondary amenorrhea.

Clinical case report

The proband is a 29-year-old woman of Italian origin, born from non-consanguineous parents. She presented at our Unit at the age of 27 years for evaluation of 6 months amenorrhea ensued after combined hormonal contraception (CHC) was suspended. Her past medical history was uneventful with no positive history of POI, early menopause, or long-term infertility within the family. She reported regular pubertal development with spontaneous menarche at 12 years of age and regular menstrual cycles thereafter. She experienced sporadic episodes of polymenorrhea (menstrual cycle shorter than 24 days) since the age of 19 years; at 22 years of age, she initiated CHC (Ethinyl estradiol 30 mcg/Dienogest 2 mg) for contraceptive needs. Following 4 years, CHC was suspended for wish to become pregnant. At first evaluation, she reported mild vasomotor symptoms as sole complaint. Hormonal evaluation showed increased levels of FSH (64.3 mIU/ml), which remained elevated 6 months apart (37.9 mIU/ml), thus confirming the diagnosis of POI according to current recommendations [7, 21]. Antimullerian hormone measurement (AMH 0.1 ng/ml) and antral follicular count (AFC 1) at pelvic ultrasound revealed diminished ovarian reserve [22], meaning early loss of ovarian follicles. Further laboratory investigations confirmed normal values of prolactin, TSH, and thyroid hormones, as well as negative thyroid autoimmunity. As a collateral finding, a mild elevation of antinuclear antibody (ANA 1:80) was observed, without clinical symptoms suggestive for connective tissue diseases. Karyotype was normal (46,XX) and FMR1 repeat lengths were in the normal range. Array-CGH did not detect any pathogenic or likely pathogenic variant. Magnetic resonance imaging (MRI) excluded diseases of the hypothalamic-hypophyseal region.

Soon after diagnosis, the patient was started on sequential hormone replacement therapy (HRT, estradiol 1 mg/dihydrogesterone 10 mg) for treating the POI condition [7].

The proband’s two brothers were referred in good health, one of them with healthy children. Her younger sister, 22 years old, was also referred in good health, but she reported progressive shortening of menstrual cycles, with overt polymenorrhea over the last 2 months, and failure to conceive naturally in the past 12 months. At the time of the study, no in-depth examinations or endocrinological evaluations had been carried out on her in this regard.

Genetic analysis

WES was performed using Twist Human Core Exome Kit (Twist Bioscience) on a NovasSeq6000 platform (Illumina, San Diego, CA). Variant filtering and prioritization were performed exploiting the eVAI tool (EnGenome) and an in-house developed gene panel for POI (Supplementary Table 1). The proband (II-3; Fig. 1a) was found to carry compound heterozygous variants in the PSMC3IP gene (NM_016556.4): c.206_208delAGA (p.Lys69del) and c.189 G > T (p.Lys63Asn) (Fig. 1b). Both variants are absent in homozygous state from the Genome Aggregation Database (gnomAD). The variant c.206_208delAGA is classified as likely pathogenic (PM2, PM4, and PP3) and the variant c.189 G > T is classified as variant of unknown significance (PM2, PP3) based on the ACMG criteria (Supplementary Table 2) [23]. No other potentially relevant variants were identified (the coverage of the known genes associated with POI is reported in Supplementary Table 1) and coverage-based analysis of CNVs excluded genomic imbalances. Segregation analyses on the mother, the two brothers and the sister (respectively I-2, II-1, II-2, and II-4 in Fig. 1a) confirmed the variants are in trans.

Fig. 1.

Pedigree and genetic analysis. a Family pedigree: filled black circle—affected proband; filled gray symbol—mildly affected sister; empty symbols—unaffected subjects; the variants identified are reported below each tested subject. b Sequence electropherograms showing the c.206_208delAGA and the c.189 G > T heterozygous mutations

Interestingly, the younger sister (II-4, Fig. 1a) was found to be compound heterozygous for the PSMC3IP variants (Fig. 1b), as the proband. Hormonal testing showed FSH 3.88 mUI/ml and LH 5.1 mUI/ml, both in the normal range. However, AMH levels at 22 years of age were 0.11 ng/ml, far below the expected AMH values in a young woman and, then, predictive of early ovarian insufficiency [24]. Pelvic ultrasound evaluation revealed normal uterus and ovaries, with signs of follicular activation. Hormonal status was reassessed 6 months apart, showing gonadotropins still in the normal range (FSH 10.43 mUI/ml and LH 4,2 mUI/ml) and confirming low AMH (0.22 ng/ml). Based on hormonal features and clinical history, we referred the proband’s sister to an infertility clinic.

Family tree and the segregation studies are shown in Fig. 1.

Discussion

Our patient is the first ever-reported case of secondary amenorrhea caused by mutations in PSMC3IP. With the development of next-generation sequencing techniques, our results show the importance to include PSMC3IP in designed POI NGS panels or in WES/WGS studies not only in patients with primary amenorrhea but also in patients with secondary amenorrhea, in order to detect new cases and new mutations.

The diagnosis is of considerable importance even for other family members in the monogenic forms, at variance with chromosomal causes of POI that are usually “de novo.” Indeed, family members may be at increased risk of infertility and guidance in taking the best reproductive choices is mandatory.

In the family presented in here, the possibility of carrying out the segregation study in the parents of the proband made it possible to re-classify the c.206_208delAGA and c.189 G > T changes as likely pathogenic variants based on ACMG classification [23]. In addition, the diagnosis in the proband made it possible to predict a possible early ovarian insufficiency (confirmed by very low values of AMH) in the younger sister, thus possibly allowing her to undergo cryopreservation of gametes for future in vitro fertilization techniques. Indeed, the relevance of combining genetics with AMH to facilitate the prediction of POI has been recently reviewed and it seems essential to safeguard reproductive potentials within families [25].

Eskenazi et al. recently described a series of patients with “Idiopathic “ POI that showed a high rate (38%) of gene variants detected by NGS analysis [26]. Unfortunately, the ACMG criteria for the interpretation of the variants were not applied and segregation studies were not performed in these patients. Thus, a high rate of false positive results may be observed. Results by Eskenazi et al. [26] demonstrate how difficult it is to classify the variants identified by NGS analysis and how family segregation plays a fundamental role, helping the classification of the large number of variants of unknown significance.

In our case report, we were able to identify a causative gene that had not initially been included in the virtual panel designed for secondary amenorrhea because we performed the WES analysis.

NGS panels play a well-known clinical role in the investigation of clear diagnosis with well-defined etiology. However, WES in heterogeneous or poorly defined phenotypes such as POI is able to confer a superior diagnostic performance identifying new genes or new phenotypes for known genes. In addition, WES allows a reanalysis of data even after many years, when new causative genes are discovered.

We hypothesize that PSMC3IP mutations might cause POI by means of a combination of reduced prenatal follicular pool and increased postnatal follicular atresia with variable expressivity, from primary amenorrhea to secondary amenorrhea (as presented by our patient), depending on the severity and on the localization of the mutations. This variable expressivity is a well-known mechanism described in other XX-gonadal dysgenesis genes [2, 27].

Pathogenic mutations in PSMC3IP described so far in primary amenorrhea have been demonstrated to alter the single-strand DNA-binding domain and the RAD51/DMC1 interaction site (both in the C-terminus of the protein). These effects impair meiosis and compromise the estrogenic pathway downstream of the FSHR expressed in developing follicles [13, 16, 17, 28–30].

In our family, an explanation for secondary amenorrhea could be the possibility that both the missense variants c.206_208delAGA and c.189 G > T are closer to the N-terminus, in the double-strand DNA-binding domain. Mutations in this region could cause a milder phenotype unlike those in the C-terminus. Further functional studies are required to prove this hypothesis.

While the altered estrogen-dependent transcription could affect only the gonadal development in females, the impaired meiosis in PSMC3IP mutation could impair also spermatogenesis with a block at meiosis I and at the primary spermatocyte stage, as demonstrated in mice [31]. The stop gain mutation described by Al-Agha in a family with POI and azoospermia confirms such hypothesis [17].

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

F.S. and R.E.N. contributed to sample collection, interpretation of data, writing of the manuscript, critical revision, and final approval of the manuscript. E.G. contributed to genetic analysis, writing of the manuscript, critical revision, interpretation of data, and final approval of the manuscript. L.C. contributed to data acquisition, literature review, writing of the manuscript, and final approval of the manuscript. E.M.V. contributed to critical revision, interpretation of data, and final approval of the manuscript. All authors have read and approved the manuscript.

Data availability

The data underlying this article will be shared on request to the corresponding author.

Declarations

Conflict of interest

The authors declare no competing interests.