A report of two homozygous TERB1 protein-truncating variants in two unrelated women with primary infertility

Abstract

Purpose

To investigate the genetic etiology of patients with female infertility.

Methods

Whole Exome Sequencing was performed on genomic DNA extracted from the patient’s blood. Exome data were filtered for damaging rare biallelic variants in genes with possible roles in reproduction. Sanger sequencing was used to validate the selected variants and segregate them in family members.

Results

A novel homozygous likely pathogenic variant, c.626G>A, p.Trp209*, was identified in the TERB1 gene of the patient. Additionally, we report a second homozygous pathogenic TERB1 variant, c.1703C>G, p.Ser568*, in an infertile woman whose azoospermic brother was previously described to be homozygous for her variant.

Conclusions

Here, we report for the first time two homozygous likely pathogenic and pathogenic TERB1 variants, c.626G>A, p.Trp209* and c.1703C>G, p.Ser568*, respectively, in two unrelated women with primary infertility. TERB1 is known to play an essential role in homologous chromosome movement, synapsis, and recombination during the meiotic prophase I and has an established role in male infertility in humans. Our data add TERB1 to the shortlist of Meiosis I genes associated with human infertility in both sexes.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10815-024-03031-x.

Introduction

Infertility is defined as the inability to conceive after a year of unprotected intercourse [1]. This disorder, affecting globally 15% of the general population [2], can be categorized into two groups: primary and secondary. Unlike primary infertility, secondary infertility patients have had previous successful pregnancies [3]. Due to the complex nature of human reproductive physiology, there is notable variety in how infertility is manifested. In women, ovulatory problems make up the majority of infertility cases. Among these are polycystic ovarian syndrome, premature ovarian insufficiency (POI), and hormonal irregularities caused by hypothalamic dysfunction [4]. In men, abnormalities in sperm count, morphology, motility, and function can lead to infertility [5], with up to 15% of these cases being attributable to genetic defects [6].

Over the years, the growth in access to whole-genome and exome sequencing technologies has facilitated associating many genes of the reproductive axis with human infertility [2]. Although genetic defects in various biological processes, including gonad formation, hormonal regulation, and meiosis, have been described to cause infertility, pathogenic variants in genes contributing to gamete quality and production have been highlighted to have the most severe impact on reproductive success [2]. However, despite the high frequency of infertility, improved insight into its biological mechanisms, and the availability of sequencing technologies, the genetic and pathophysiological heterogeneity of infertility render numerous causative genes and variants undiscovered.

Here, we report two homozygous stop-gain variants in the telomere repeat binding bouquet formation protein 1 (TERB1) gene in two unrelated infertile women from familial cases. So far, TERB1 variants have only been reported in infertile men. Our findings show, for the first time, that they are also associated with primary female infertility.

Materials and methods

Clinical case reports

Case 1

The first family analyzed in this study is of Egyptian origin and includes a female (patient ID 2105) with five years of primary infertility due to diminished ovarian reserve, and her brother (patient ID 2104) with ten years of infertility with three partners due to non-obstructive azoospermia (Fig. 1a). The parents of the infertile siblings are first cousins. Clinical findings of the infertile female are summarized in Table 1.

Fig. 1.

Identification of the likely pathogenic and pathogenic TERB1 variants in two families with primary female infertility. a The family pedigree and chromatograms for Case 1. b The family pedigree [9] and chromatograms for Case 2. ‘ + / + ’ and ‘ +/- ’ indicate family members who are homozygous and heterozygous for the variant, respectively. ‘-/-’ indicate family members with wildtype alleles. The black filled circle and square represent the infertile female and azoospermic male, respectively. Grey circles indicate genetically not tested individuals. Symbols with the black-dot represent the heterozygous carriers. Black arrow indicates the proband. Triangle represents miscarriage. The fertility status of M2073’s homozygous brother is unknown. c Schematic representation of the human TERB1 protein and its functional domains. Interdomains are indicated in grey, and the domains are colored. ARM, armadillo repeats; CC, coiled-coil domain; TBD, TRF-1 binding domain; MYB, Myb-like domain. Mutated amino acids above the protein structure are the TERB1 variants found in our patients (red: novel, blue: reported in [9]), while those below the protein structure in black are the TERB1 variants reported in previous literature

Table 1.

Clinical data of patient 2105

Parameter	Measurement
Age at menarche	13
Menstrual cycle	30 days, lasting 3–5 days
Luteinizing Hormone levels	Elevated (18.5 IU/L)
Follicle Stimulating Hormone levels	Elevated (22.3 IU/L)
Anti-Mullerian Hormone levels	Severely decreased (0.3 ng/mL)
Ovary size	Small ovaries Right → 1,96 × 1,70 × 1,52 Left → 2,26 × 1,07 × 1,51
Partner sperm parameters	Normal andrological parameters Ejaculate volume = 2 mL Sperm count = 60 million/ejaculate Sperm motility = 85% Progressive motility = 54% Morphology index = 40%

The clinical data was acquired on day 2 of menses when the patient was 26 years and 8 months old. The reference level of the hormones at follicular phase are the following: Luteinizing Hormone = 1–12 IU/L [7], Follicle Stimulating Hormone = 1–9 IU/L [7], Anti-Mullerian Hormone = 2–6.8 ng/mL [8].

Case 2

The second infertile female is from a previously reported family (Fig. 1b) consisting of two unaffected siblings and an infertile non-obstructive azoospermic male who is homozygous for a pathogenic protein-truncating TERB1 variant, c.1703C > G, p.Ser568* [9]. The fertility status of the second homozygous brother is unknown. The affected sister has an unfulfilled wish to have a child for 4–5 years and had one spontaneous pregnancy that resulted in a very early miscarriage. Afterwards, medically assisted reproduction treatment did not result in pregnancy. No DNA was available from the infertile sister prior to publishing the original report.

Subjects and DNA extraction

All participants of this study have provided their written informed consents. Blood samples were collected from the patients as well as family members when available. Flexigene DNA Kit (Qiagen, Hilden, Germany) was used to isolate genomic DNA from whole blood following manufacturer instructions.

Library preparation and whole exome sequence filtering

Roche Nimbelgen SeqCap EZ Human Exomes or MedExomes capture kits were used to capture five hundred nanograms of peripheral blood leukocyte DNA from the patients. The DNA was sequenced with paired-end 100 base-pair reads on Illumina HiSeq 6000. Burrows-Wheeler Aligner (V.0.7.17) [10] was utilized to map the sequence reads to the human reference genome (hg19). Picard (V2.27.4) [11] was used as described previously [12] to flag duplicate reads, which were excluded from our analysis. Variant calling was completed with GATK HaplotypeCaller (V.4.2.4.0). ANNOVAR and custom scripts were used to complete variant calling as described previously [12]. The resulting annotated variants were filtered against frequent germline polymorphisms found in The 1000 Genomes Project, Genome Aggregation Database (gnomAD) (V2.1.1) [13] and dbSNP135. Next, only biallelic variants with a maximum population minor-allele frequency (MAF) of less than or equal to 0.01 were kept. Among these, variants seen in ≤ 5 individuals out of approximately 4400 in-house controls, that lead to insertion/deletion, protein-truncating, or affect conserved missense, and have a Combined Annotation Dependent Depletion (CADD) [14] score ≥ 10 were selected for. Remaining variants in genes with a potential role in reproduction were then classified according to the American College of Medical Genetics and Genomics (ACMG) guidelines using Varsome [15], which utilizes a point system as previously described [16] to evaluate and score the pathogenic potential of the variant of interest. Only variants that are classified as pathogenic (score ≥ 10), likely pathogenic (6 ≤  score ≤ 10), or variant of uncertain significance (VUS) (0 ≤  score ≤ 5) were selected.

Mutation analyses

Primers designed with Primer3Plus [17] were used to amplify target regions with the variants using PCR conditions previously described [18]. Sanger sequencing was used to validate candidate variants and segregate the validated variants among available family members. All variants detected have been submitted to the Leiden Open Variation Database (https://databases.lovd.nl/shared/individuals/00440109) (patient IDs 00446588 for case 1 and 00446590 for case 2).

Results

Case 1

To potentially identify the genetic etiology of infertility in this family, whole exome sequencing was performed on 2105 using blood DNA. Given the parents are first cousins and did not have problem conceiving, the recessive mode of inheritance of the defect was prioritized. Exome data were filtered according to the criteria described in the Materials and Methods and resulted in the identification of two genes with candidate variants: a novel homozygous stop-gain variant, NM_001136505.2:c.626G>A, p.Trp209*, in TERB1, and two novel multiple heterozygous nonsynonymous variants, NM_053006:c.2T>A, p.M1K and NM_053006: c.728G>T, p.C243F, in TSSK2 (Supplementary Figs. 1 and 2). The identified variants were classified as likely pathogenic, likely pathogenic, and likely benign, respectively [15]. Sanger sequencing was used to validate and segregate the candidate variants, which led to the confirmation of the TERB1 variant (Fig. 1a) and the exclusion of the TSSK2 variants since c.2T>A, p.M1K was not validated in the patient (Supplementary Figs. 1 and 2).

The TERB1 variant is located in a run of homozygosity of 11.1-Mb on chromosome 16 and was predicted by the R package masonmd (Make Sense of nonsense-mediated decay (NMD)) [19] and Mutation Taster [20] to trigger NMD. Segregation analysis showed that this variant is also homozygous in the affected azoospermic brother, while the three unaffected siblings and parents are all heterozygous carriers (Fig. 1a). To exclude the presence of any other recessive causative variants responsible for the infertility of the azoospermic brother, we next performed exome sequencing on his blood DNA and filtered the exome data under the same criteria described above. Our analysis did not reveal any other candidate gene with plausible recessive variants that may explain the phenotype of the infertile brother (Supplementary Fig. 1).

TERB1 is involved in the pairing of homologous chromosomes during meiotic prophase I [21] and has an established role in the causation of male infertility due to meiotic arrest and consequently azoospermia (Table 2). Based on its function, the phenotype of the affected brother, and the pathogenicity classification of the variant, we conclude that the novel homozygous likely pathogenic variant, c.626G>A, p.Trp209*, in TERB1 is the most plausible candidate to explain the infertility of the two siblings.

Table 2.

Previously reported recessive TERB1 variants

Reference	Patient ID	Variant*	Zygosity	ACMG Classification by Varsome	Phenotype
Kherraf et al., 2022 [22]	P0145	c.733G>A, p.Gly245Arg	Hom	VUS (4P: 0B)	NOA
Salas-Huetos et al., 2021 [9]	Individual 2	c.977A>G, p.Glu326Gly	Hom	VUS (1P: 1B)	NOA
	M2073**	c.1703C>G, p.Ser568*	Hom	Pathogenic	NOA
Krausz et al., 2020 [23]	M468	c.236C>T, p. Ala79Val	Hom	VUS (2P: 0B)	NOA
	10–200 & brother	c.289_290del, p. Leu97Valfs*7	Multiple Het	Pathogenic	NOA
		c.1813C>T, p. Arg605*		Pathogenic	NOA
Alhathal et al., 2020 [24]	19DG1792	c.733G>A, p.Gly245Arg	Hom	VUS (4P: 0B)	NOA
	19DG1816		Hom		NOA

*All variants are provided in NM_001136505.2. **Previously reported male proband whose infertile sister is reported in this study. ‘Hom’ and ‘Multiple Het’ are short for homozygous and multiple heterozygous, respectively. ‘VUS’ is for variant of uncertain significance, ‘P’ for pathogenic points, and ‘B’ for benign points.’NOA’ stands for non-obstructive azoospermia.

Case 2

DNA analysis of the infertile sister revealed that she is also homozygous for the TERB1 variant found in her brother, NM_001136505.2:c.1703C>G, p.Ser568* (Fig. 1b). The variant was also predicted by masonmd [19] and Mutation Taster [20] to trigger NMD.

Discussion

In this paper, we report a novel likely pathogenic protein-truncating variant, p.Trp209*, in the TERB1 gene of an infertile woman and her azoospermic brother. We also report a second infertile female, from a previously described family [9], who is homozygous for another pathogenic protein-truncating variant in TERB1.

TERB1 encodes for a 727 amino acid nuclear protein consisting of two ARM, one coiled-coil, one TERF1-interacting, and one Myb domains (Fig. 1c) [21]. TERB1 plays a critical role in the attachment of telomeres to the nuclear envelope and is required for homologous chromosome movement, pairing, synapsis, and recombination. It interacts with the telomeric repeat binding factor 1 (TERF1 in humans and trf1 in mice) and mediates the assembly of the meiotic telomere complex (MTC) [25], which includes another TERB protein, TERB2, and membrane-anchored junction protein (MAJIN) [26], a transmembrane protein of the inner nuclear membrane. The MTC localizes to the nucleus and anchors the chromosomes to SUN domain-containing protein 1 (SUN1), another transmembrane protein of the inner nuclear membrane, and KASH domain-containing protein 5 (KASH5), a transmembrane protein of the outer nuclear membrane [27]. SUN1 and KASH5 form the Linker of Nucleoskeleton and Cytoskeleton (LINC) complex and attach the MTC to the Dynein-Dynactin complex in the cytoplasm [28]. Altogether, these interacting proteins between the chromosomes in the nucleus and the cytoplasmic cytoskeleton ensure the rapid movement of homologous chromosomes for pairing, synapsis, and recombination [29].

In mice, deleterious mutations in any of Terb1 [26], Terb2 [26], Majin [26], Sun1 [30], or Kash5 [31] result in the loss of germ cells, gonadal dysgenesis, and infertility in both sexes. In humans, biallelic variants in TERB1 [9, 22–24], TERB2 [9], MAJIN [9], SUN1 [32], and KASH5 [32] have been shown to cause non-obstructive azoospermia. However, of these five genes, only biallelic variants in KASH5 have so far been reported in infertile women (Supplementary Table 1). Some of these women had primary infertility due to POI [33, 34] while others had diminished ovarian reserve and/or recurrent miscarriage [35].

In addition to the members of the MTC and LINC complexes, recessive defects in approximately 50 Meiosis I genes have been shown to cause infertility in both male and female mice [36]. However, to our knowledge, biallelic variants in about 20 Meiosis I genes have been reported to cause male and female infertility in humans (summarized in Supplementary Table 1). Results from both mice and human studies have demonstrated a consistent difference in the severity of the phenotype between the two sexes. While in men, deleterious variants cause mostly a complete arrest of Meiosis I and lead to non-obstructive azoospermia, in women, the meiotic arrest appears to be partial and results in a spectrum of reproductive outcomes, ranging from infertility to POI, diminished ovarian reserve, early embryonic arrest after medically assisted reproduction, recurrent molar pregnancy and/or miscarriage (Supplementary Table 1). This sexual dimorphism was described a long time ago in mice [36, 37] and humans [38]. However, due to the phenotype-based nature of human studies, this spectrum of diverse reproductive outcomes seen in women adds another layer of complexity to the highly heterogeneous entity of female infertility, consequently, hampering and delaying the identification of its causative genes and associated variants.

Herein, we describe, for the first time, the association of recessive pathogenic TERB1 variants with primary female infertility in two unrelated families. Our report adds TERB1 to the, as of yet, short list of Meiosis I genes associated with human infertility in both sexes.

Supplementary Information

Below is the link to the electronic supplementary material.

Declarations

Conflict of interest

The authors have declared that no conflict of interest exists.