Compound heterozygous ZP1 mutations cause empty follicle syndrome in infertile sisters

Empty follicle syndrome (EFS) is a condition in which no oocyte is retrieved from mature follicles after proper ovarian stimulation in an in vitro fertilization procedure. Genetic evidence accumulates for the etiology of recurrent EFS without pharmacological or iatrogenic problems. In this study, we present two infertile sisters in a family with EFS after three cycles of standard ovarian stimulation with human chorionic gonadotrophin and/or gonadotropin‐releasing hormone agonist therapy. Via whole‐exome sequencing and cosegregation test, we identified compound heterozygous mutations in the gene of ZP1 in both of the infertile sisters. Coimmunoprecipitation tests and homology modeling analysis confirmed that both mutated ZP1 disrupt the formation of oocyte zona pellucida by interrupting the interaction among ZP1, ZP2, and ZP3. We thus propose that the specific mutations in ZP1 gene render a causality for the intractable EFS.

family with EFS after three cycles of standard ovarian stimulation with human chorionic gonadotrophin and/or gonadotropin-releasing hormone agonist therapy.
Via whole-exome sequencing and cosegregation test, we identified compound heterozygous mutations in the gene of ZP1 in both of the infertile sisters.
In this study, we identified compound heterozygous mutations of zona pellucida glycoprotein 1 gene (ZP1; MIM# 195000) from a family with two sisters of abnormal oogenesis with EFS (Empty ovarian follicle, HP:0031067). The allelic mutants were transmitted from either parent. We hypothesized that these two ZP1 mutations caused the morphological defect of the oocytes, thus incurring degeneration of oocytes and empty follicles.  Table S1). Her Chromosomal karyotyping was 46,XX. The proband received two cycles of IVF treatment. In the first attempt, a gonadotropin-releasing hormone (GnRH) agonist protocol was performed. The second IVF attempt was carried out with GnRH antagonist Protocol 3 months later (Table 1, Supporting Information Methods). Several cumulus-corona complexes were recovered but none of a recognizable oocyte was obtained in both of the two attempts.
The sister of the proband (Figure 1a, II-2) was 28-years-old and suffered from primary infertility (Female infertility, HP:0008222) for 5 years. She had regular menses (26-28 days in length) since menarche at the age of 18, her anti-Müllerian hormone (AMH) level was also normal (6.1 ng/ml), and the ultrasound revealed a total of 10 antral follicles, 2-5 mm in diameter (Table S1). She underwent one IVF attempt with GnRH antagonist protocol (Table 1)  Then followed by filtering, annotation, and prioritization for the WES data on the (Hu et al., 2014) proband, the cosegregation test was performed for the selected candidate variants on all six family members (Figure 1a, I-1, I-2, II-1, II-2, II-3, II-4) via Sanger-sequencing (Supporting Information Methods and Table S3). This strategy revealed a compound heterozygous variants, c.170_174delGCCAG  To investigate the molecular mechanism of how p.I390Tfs*16 or p.G57Dfs*9 mutant of ZP1 led to the absence of oocyte zona pellucida, oocyte degeneration, and empty follicles, we performed the coimmunoprecipitation (co-IP) analysis to explore the interaction of ZP1 with other zona pellucida glycoproteins. It has been shown that the ZP4 protein does not coprecipitate with other ZP proteins (Chen et al., 2017), so here we analyzed the interaction between ZP1, ZP2, and ZP3.
The reconstructed vectors with tag protein of full-length ZP2 and ZP3 with full-length ZP1 or two types of truncated ZP1 were cotransfected into 293T cells with an equal amount. And the proteins were extracted from the transfected cells with targeting the anti-FLAG antibody or anti-HA antibody, respectively. Then the immunoblotting was performed to detect the ZP proteins (Supporting Information Methods).
Three-dimensional structural models of the wild-type ZP1 (ZP1_WT) and ZP1-mutant (ZP1_mut1) were predicted by the I-TASSER web tool (Roy, Kucukural, & Zhang, 2010; http://zhang. bioinformatics.ku.edu/I-TASSER/). The new cartoon representation of models and structural superposition were generated by software package visual molecular dynamics (Humphrey, Dalke, & Schulten, 1996). Three-dimensional structural models of the wild-type ZP1 (ZP1_WT) and p.I390Tfs*16-truncated ZP1 (ZP1_mut1) show that more than half of the ZP1 protein was absent in p.I390Tfs*16truncated ZP1 (ZP1_mut1) model, and this structural defect may well cause dysfunction of ZP1 (Supporting Information Methods and Figure S1).  (Huang et al., 2014). Although in our case, the more severe condition, namely, without observable oocyte, could be explained by the remaining half amount of the previously reported truncated ZP1 protein, which was insufficient to maintain the oocyte normal development, so it caused the morphological defect of the oocytes, leading to degeneration and empty follicles, whereas another allelic mutation (p.G57Dfs*9) caused the complete loss of ZP1 protein in our patients.
In conclusion, we proposed a scenario in which allelic mutations in the ZP1 gene may well induce refractory "genuine" EFS, so besides the recessive EFS caused by LHCGR gene mutations and the dominant EFS caused by ZP3 gene mutations, we added another candidate into the EFS genetic repertoire, providing a novel glimpse into the heterogeneity of EFS etiology.