Bi-allelic Missense Pathogenic Variants in TRIP13 Cause Female Infertility Characterized by Oocyte Maturation Arrest

Abstract

Normal oocyte meiosis is a prerequisite for successful human reproduction, and abnormalities in the process will result in infertility. In 2016, we identified mutations in TUBB8 as responsible for human oocyte meiotic arrest. However, the underlying genetic factors for most affected individuals remain unknown. TRIP13, encoding an AAA-ATPase, is a key component of the spindle assembly checkpoint, and recurrent homozygous nonsense variants and a splicing variant in TRIP13 are reported to cause Wilms tumors in children. In this study, we identified homozygous and compound heterozygous missense pathogenic variants in TRIP13 responsible for female infertility mainly characterized by oocyte meiotic arrest in five individuals from four independent families. Individuals from three families suffered from oocyte maturation arrest, whereas the individual from the fourth family had abnormal zygote cleavage. All displayed only the infertility phenotype without Wilms tumors or any other abnormalities. In vitro and in vivo studies showed that the identified variants reduced the protein abundance of TRIP13 and caused its downstream molecule, HORMAD2, to accumulate in HeLa cells and in proband-derived lymphoblastoid cells. The chromosome mis-segregation assay showed that variants did not have any effects on mitosis. Injecting TRIP13 cRNA into oocytes from one affected individual was able to rescue the phenotype, which has implications for future therapeutic treatments. This study reports pathogenic variants in TRIP13 responsible for oocyte meiotic arrest, and it highlights the pivotal but different roles of TRIP13 in meiosis and mitosis. These findings also indicate that different dosage effects of mutant TRIP13 might result in two distinct human diseases.

Introduction

The meiotic maturation of oocytes is a fundamental prerequisite for successful human reproduction. In prophase of meiosis I, genetic information is exchanged between homologous maternal and paternal chromosomes through the process of homologous recombination,^1,2 which ensures that chromosomes segregate accurately when the meiotic cell divides.¹ After the diplotene stage of prophase I, oocytes are arrested as germinal vesicle (GV) oocytes.³ In puberty, a surge of luteinizing hormone triggers meiotic resumption, which consists of chromatin condensation, breakdown of the nuclear envelope, spindle formation, and chromosome alignment,^3,4 and meiosis I is completed upon extrusion of the first polar body (PB1).⁵ Therefore, defects in the process of meiosis may cause oocyte maturation arrest. Knockouts of a number of genes in mice, including Pde3a (MIM: 123805), Cdc25b (MIM: 116949), and Marf1 (MIM: 614593), have been reported to cause oocyte maturation arrest and infertility.^{6, 7, 8} We recently discovered human oocyte maturation arrest as a new Mendelian disease (MIM: 616780 and MIM: 617743) and identified pathogenic variants in TUBB8 (MIM: 616768) and PATL2 (MIM: 614661) as being responsible for the corresponding phenotypes.^9,10 Pathogenic variants in TUBB8 cause oocyte metaphase I (MI) stage arrest (MIM: 616780), whereas pathogenic variants in PATL2 mainly result in oocyte GV stage arrest (MIM: 617743). Until now, these were the only two mutant genes known to result in human oocyte maturation arrest. However, pathogenic variants in TUBB8 and PATL2 account for around 30% of the individuals with oocyte maturation arrest,^{9, 10, 11, 12, 13, 14, 15, 16} and the underlying genetic causes for the majority of affected individuals remain unknown.

Thyroid hormone receptor interactor 13 (TRIP13, MIM: 604507) functions in both mitosis and meiosis,^{17, 18, 19, 20, 21} and it is widely expressed in all kinds of tissues, including germ cells.²² During mitosis, TRIP13 catalyzes a structural transition in MAD2 and subsequent spindle checkpoint silencing.¹⁷ Furthermore, TRIP13 is highly expressed in many cancer types, suggesting an oncogenic property. In 2017, a recurrent homozygous nonsense pathogenic variant, c.1060C>T (p.Arg354^∗), and a homozygous splicing pathogenic variant in TRIP13 were shown to cause Wilms tumors (MIM: 617598) in children by impairing chromosome segregation and causing spindle assembly checkpoint deficiency.²³ It has been reported that TRIP13 is a negative regulator of the HORMA proteins, including HORMAD1 and HORMAD2.²⁴ TRIP13 is required for completing meiotic recombination in mice, and it plays a critical role in chromosome synapsis by removing HORMAD2 from synapsed chromosomal axes.^20,22,25,26 Both male and female Trip13-knockout mice are infertile.^22,26,27

In this study, we identified different homozygous and compound heterozygous missense pathogenic variants in TRIP13 (GRCh37, GenBank: NM_004237.4) that are responsible for female infertility mainly characterized by oocyte maturation arrest instead of Wilms tumors in five affected individuals from four independent families. We investigated the molecular mechanism of the corresponding pathogenic variants in HeLa cells and in proband-derived lymphoblastoid cell lines (LCLs). We also explored a potential therapeutic treatment in one of the affected individuals by injecting TRIP13 cRNA into her oocytes. Our study established a causal relationship between TRIP13 and female infertility and revealed the interesting phenomenon that different types of TRIP13 recessive pathogenic variants cause distinct diseases through different effects on mitosis and meiosis.

Subjects and Methods

Clinical Samples

In total, 1,253 women with oocyte maturation arrest, fertilization failure, and embryonic arrest were evaluated in this study. The individuals in this cohort were recruited from 62 collaborating hospitals and reproductive centers in China from 2015. The phenotypes and related numbers are shown in Table S1. In this study, the four affected families with individuals with pathogenic variants in TRIP13 were from the Shanghai Ji Ai Genetics and IVF Institute affiliated with the Obstetrics and Gynecology Hospital of Fudan University, the Reproductive Medicine Center of the Shanghai Ninth Hospital affiliated with Shanghai Jiao Tong University, the First Hospital affiliated with Nanjing Medical University, and the Assisted Reproductive Technology Laboratory of Shenyang Jinghua Hospital. Whole blood of control individuals and family 1 (II-1) was sampled, and their lymphoblastoid cells were transformed into immortal cell lines by Epstein–Barr virus via standard protocols.²⁸

All of the studies were approved by the ethics committees of the Medical College of Fudan University and the Reproductive Medicine Center of the Shanghai Ninth Hospital affiliated with Shanghai Jiao Tong University, and written informed consent was obtained from the affected individuals and controls. All oocytes (from control and affected individuals) were donated following informed consent and approval from these committees.

Genetic Analysis

We extracted genomic DNA from peripheral blood by following the instructions of the QIAGEN DNA extraction kit. We used Sanger sequencing to exclude pathogenic variants in TUBB8. We performed whole-exome sequencing (WES) to identify pathogenic gene variants in affected individuals from family 1 and 2. Whole-exome capture and sequencing was performed according to a previously reported protocol, included mapping the raw FASTQ files to the human reference sequence (NCBI Genome build GRCh37) and variant calling and annotation with GRCh37 (dbSNP version 138).^10,29 Because family1 was a consanguineous family, candidate variants were filtered with the following criteria: (1) variants with a minor allele frequency < 0.1% in the ExAC Browser and not existing in our control database, (2) homozygous variants, (3) variants located within homozygous regions greater than 2.0 Mb, and (4) variants functionally predicted by at least one prediction software to be damaging. We performed homozygosity mapping with the online software HomozygosityMapper to identify homozygous regions, and we used Mutation Taster and PolyPhen-2 to predict loss-of-function alleles and damaging missense variants. We used Sanger sequencing to confirm the likely pathogenic variants in all members of the families. For family 2, both homozygous and compound heterozygous variants shared in two sisters were considered. Finally, a recurrent missense homozygous variant in TRIP13 was most likely pathogenic. Mutational screening of the pathogenic gene was then performed in the remaining 1,250 samples with oocyte maturation arrest, fertilization failure, and embryonic arrest.

Plasmid Construction and Mutagenesis

The full-length coding sequence of TRIP13 was amplified and cloned into the pCMV6-Entry vector with or without FLAG-tag (Origene). The KOD-Plus Mutagenesis Kit was used to introduce six variants (c.77A>G, c.518G>A, c.907G>A, c.592A>G, c.739G>A, and c.1060C>T) (GRCh37, GenBank: NM_004237.4). We linearized the TRIP13 plasmid with AgeI enzyme, and we used the HiScribe T7 ARCA mRNA Kit (E2060S, New England Biolabs) to transcribe TRIP13 cRNA according to the manufacturer’s instructions.

Immunoblotting and TRIP13 ATPase Activity Assay

Wild-type (WT) and mutant TRIP13 were transfected into HeLa cells. At 36 h, cells were harvested for immunoblotting. For proband-derived LCLs, the cells were harvested directly for immunoblotting. In brief, total protein was extracted in RIPA lysis buffer (Shanghai Wei AO Biological Technology). After denaturing with 5× SDS loading buffer, the cell lysate was loaded for SDS-PAGE, electrophoretically separated, and then transferred to a nitrocellulose membrane. After blocking with 5% fat-free milk in phosphate-buffered saline with Tween-20 (PBST), the nitrocellulose membrane was incubated with TRIP13 (1:1,000 dilution, ab128153, Abcam), HORMAD2 (1:1,000 dilution, ab89961, Abcam), Vinculin (1:5,000 dilution, 13901, CST), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (1:3,000 dilution, 2118, CST) antibodies overnight at 4°C. We used goat anti-rabbit IgG (1:5,000 dilution, M21002, Abmart) and goat anti-mouse IgG (1:5,000 dilution, M21001, Abmart) conjugated to horseradish peroxidase to detect the primary antibodies. Enhanced chemiluminescence (ECL) imaging was performed on a chemiluminescent imaging system (Tanon, 5200). To test the TRIP13 ATPase activity, we transfected WT or mutant TRIP13 into HeLa cells. The cells were lysed with passive cell-lysis buffer, and the activity was measured with the ADP-Glo Kinase Assay (Promega).

Oocyte Collection and Microinjection

The MI oocytes from the affected individuals were donated for investigation after the individuals provided written and informed consent. The control PB1 oocytes used in this study were matured in vitro from oocytes at the GV or MI stage. Oocytes were retrieved from the ovaries of each individual after human menopausal gonadotropin (hMG, Fengyuan Pharmaceutical) stimulation. Then oocytes were degranulated and randomly divided into the injected and non-injected groups. Approximately 500 ng/μL TRIP13 cRNA solution was used for injection. After injection, oocytes were maintained in G-MOPS medium (Vitrolife) for in vitro maturation and considered to be matured when they extruded PB1. After 10 h, intracytoplasmic sperm injection (ICSI) was performed. Fertilization was evaluated at 16–18 h after ICSI for the formation of two pronuclei. The zygotes and embryos were monitored for 6 days after ICSI.

Flow-Cytometry Cell-Cycle Analysis and Live-Cell Imaging Analysis of Chromosome Segregation Errors

Proband-derived and control LCLs were harvested and centrifuged at 1000 × g for 5 min, washed with cold PBS, and fixed in cold 70% ethanol for at least 2 h. Cells were incubated with propidium iodide at 37°C for 30 min. Samples of 10,000 cells were analyzed by FACSCalibur flow cytometry (Becton-Dickinson). For live-cell imaging, cells were synchronized in early S phase for 24 h with 2 mM thymidine. Cells were then released from thymidine and co-incubated with SiR-DNA dye (Cytoskeleton) for 4 h and then plated on a 96-well plate with 0.4% low-melting agarose. Images were obtained on an EVOS M7000 live-cell workstation.

Real-Time Quantitative PCR

Total RNA was extracted from proband-derived and control LCLs by an miRNeasy Micro Kit (QIAGEN), and first-strand cDNA was generated with the PrimeScript RT Reagent Kit with gDNA Eraser according to the manufacturer’s protocol. Then qPCR was performed with a SYBR Green Premix Ex Taq (Takara) via an Applied Biosystems Prism7300 machine (Applied Biosystems). Expression of HORMAD2 (MIM: 618842) was analyzed by qPCR, and the results were normalized by comparison to the expression of an internal GAPDH (MIM: 138400) control. The qPCR primers for HORMAD2 and GAPDH are shown in Table S2.

Statistical Analysis

All experiments were performed independently at least three times. We analyzed the data by using a one-way ANOVA, Student’s t test, and Pearson’s chi-square test, and p < 0.05 was considered statistically significant. ImageJ software was used for quantitation of the immunoblotting.

Results

Clinical Characteristics of the Affected Individuals

All five affected individuals had been diagnosed with primary infertility of unknown cause for several years despite their having had normal menstrual cycles. Their male partners had normal sperm counts and normal sperm morphologic features and motility. The proband in family 1 was from a consanguineous family (Figure 1A and Table 1). She was 33 years old and had been diagnosed with primary infertility for 9 years. She underwent two failed in vitro fertilization (IVF) attempts. Nineteen MI oocytes were retrieved, but none could be matured in vitro (Figure 1B and Table 1). The proband in family 2 was 39 years old. During her 10 years of infertility, she underwent three IVF cycles. Altogether, 44 oocytes were retrieved, of which 43 arrested at the MI stage and only one PB1 oocyte was obtained (Figure 1B and Table 1). Her sister was also diagnosed with infertility for many years, but her detailed clinical information was unavailable. The proband in family 3 was 27 years old and had been infertile for 9 years. In her three IVF cycles, 42 oocytes were retrieved, of which 39 oocytes were immature, including 37 MI oocytes and two GV oocytes (Table 1). The proband in family 4 was 29 years old and had suffered from infertility for 7 years. She underwent only one IVF cycle, and 14 PB1 oocytes were retrieved. Slightly different from the other probands, 12 of her PB1 oocytes could be fertilized successfully. Nine zygotes remained at the one-cell stage, whereas the other three underwent cleavage in the following culture but arrested before the eight-cell stage (Table 1).

Figure 1.

Identification of Pathogenic Variants in TRIP13 and the Morphology of Normal and Affected Individuals’ Oocytes

(A) Pedigrees of four affected families. Family 1 was a consanguineous family with a recessive inheritance pattern. All five affected individuals had bi-allelic missense pathogenic variants with a recessive inheritance pattern. Sanger sequencing confirmation is shown below the pedigrees. The equal sign indicates infertility, and the black circles represent the affected individuals.

(B) The morphology of normal and affected individuals’ GV, MI, and PB1 oocytes by light and polarization microscopy. The normal oocytes extruded PB1 (red arrow) and entered metaphase-II. Oocytes from family 1 (II-1) and family 2 (II-1) were arrested at the MI stage. The polarization microscope images indicate the normal spindles (white arrows). The scale bar represents 80 μm.

(C) The mutation pattern and conservation of mutated amino acids in TRIP13. The green balls indicate the mutations identified in this study, and all mutations were missense mutations. The red balls indicate the mutations in the study of Shawn et al.²³

Table 1.

Clinical Characteristics of Affected Individuals and Their Retrieved Oocytes

Age (Years)	Duration of Infertility (Years)	IVF/ICSICycles	Total Number of Oocytes Retrieved	GV Oocytes	MI Oocytes	PB1 Oocytes	Fertilized Oocytes	Cleaved Embryos
Family 1 (II-1)
33	9	1(IVF)	9	0	9	0	0	0
		2(IVF)	10	0	10	0	0	0
Family 2 (II-1)
39	10	1(IVF)	25	0	25	0	0	0
		2(IVF)	5	0	5	0	0	0
		3(IVF)	19	0	18	1	0	0
Family 3 (II-1)
27	9	1(IVF)	13	0	10	3	0	0
		2(IVF)	14	0	14	0	0	0
		3(IVF)	15	2	13	0	0	0
Family 4 (II-1)
29	7	1(IVF)	14	0	0	14	12	3

GV, germinal vesicle; MI, metaphase I; PB1, first polar body. Clinical information for Family 2 (II-2) was not available.

Identification of Bi-allelic Missense Pathogenic Variants in TRIP13

Because family 1 was consanguineous, homozygous variants were prioritized in the analysis. On the basis of the WES data and our filtering criteria (see Subjects and Methods), we first identified a likely pathogenic homozygous missense variant in TRIP13 (c.77A>G (p.His26Arg), GRCh37, GenBank: NM_004237.4) in family 1 (Figure 1A). TRIP13 localized within the homozygous region (6.49 M) of the proband (Figures 1A and S1). We further identified the recurrent homozygous variant in TRIP13 in two affected sisters in family 2, implying the genetic contribution of TRIP13 to oocyte maturation arrest. We then performed mutational screening of TRIP13 in a cohort of 1,250 women with oocyte maturation arrest, fertilization failure, and embryonic arrest. As a result, we found two other families (families 3 and 4) carrying TRIP13 compound heterozygous pathogenic variants. The affected individual in family 3 had the compound heterozygous variants c.518G>A (p.Arg173Gln) and c.907G>A (p.Glu303Lys) (GRCh37, GenBank: NM_004237.4) (Figure 1A), whereas the proband in family 4 carried the compound heterozygous variants c.592A>G (p.Ile198Val) and c.739G>A (p.Val247Met) (GRCh37, GenBank: NM_004237.4) (Figure 1A). The variants and recessive inheritance pattern were confirmed by Sanger sequencing (Figure 1A). The allelic frequencies of the five variants in the ExAC Browser and gnomAD are shown in Table 2. Most of the corresponding residues affected by the missense variants are conserved among different species, whereas the variant c.77A>G (p.His26Arg) is conserved among other species, but not in mice (Figure 1C).

Table 2.

TRIP13 Pathogenic Variants Observed in the Four Families

Families	Genomic Position	cDNA Change	Protein Change	Mutation Type	Sifta	PPH2a	ExAC Eb	gnomADc
1 and 2	chr5: 893190	c.77A>G	p.His26Arg	missense	D	B	0.0002	9.2 × 10−6
3	chr5: 901529	c.518G>A	p.Arg173Gln	missense	D	D	0	8.0 × 10−6
	chr5: 911998	c.907G>A	p.Glu303Lys	missense	T	P	0	1.8 × 10−5
4	chr5: 904319	c.592A>G	p.Ile198Val	missense	T	P	NA	NA
	chr5: 908169	c.739G>A	p.Val247Met	missense	D	D	NA	NA

Abbreviations are as follows: B, benign; T, tolerance; D, damaging; P, probably damaging; and NA, not available.

^aMutation assessment by Sift and PolyPhen-2 (PPH2).

^bFrequency of corresponding mutations in the East Asian population of the ExAC Browser.

^cFrequency of corresponding mutations in gnomAD.

Effects of Pathogenic Variants on TRIP13 Abundance, ATPase Activity, and Downstream HORMAD2 Abundance

To evaluate the functional effects of the pathogenic variants in vitro, we performed immunoblot analysis in cells transfected with WT or mutant TRIP13 constructs. Compared with the WT, the TRIP13 abundance was significantly decreased in the c.77A>G (p.His26Arg), c.907G>A (p.Glu303Lys), and c.739G>A (p.Val247Met) mutants (Figure 2A).We also observed that the variant p.Arg354^∗, which causes Wilms tumors, resulted in almost no detectable TRIP13, whereas the missense variants we identified in this study resulted in different degrees of TRIP13 reduction (Figure S2). We further measured the TRIP13 abundance in the proband-derived LCLs from family 1 (II-1). Similar to the results in vitro, the TRIP13 abundance was dramatically decreased compared to in three normal controls (Figure 2B). It has been shown previously that TRIP13 functions in mice as a checkpoint in completing meiotic recombination by removing HORMAD2 from synapsed chromosome axes.²⁷ In addition, Hormad2 and Trip13 double-knockout mice rescue the infertility phenotype of Trip13-knockouts, suggesting that HORMAD2 might be a downstream target of TRIP13.²⁰ We therefore supposed that TRIP13 pathogenic variants might cause the infertility phenotype by affecting the abundance of HORMAD2, and we measured the abundance of HORMAD2 both in vitro and in vivo. In HeLa cells, WT TRIP13 had an inhibitory effect on HORMAD2 (Figure 2C). In contrast, the inhibitory effects on HORMAD2 were reduced by the mutants c.77A>G (p.His26Arg), c.907G>A (p.Glu303Lys), and c.739G>A (p.Val247Met) (Figure 2C). Similar results were obtained in the proband-derived LCLs. As shown in Figures 2D and 2E, HORMAD2 accumulated significantly more than did the normal control at both the protein and mRNA level, suggesting that the accumulation of HORMAD2 might play a key role in the oocyte MI arrest phenotype.

Figure 2.

Protein Abundance of Mutant Proteins and Their Effects on HORMAD2 and ATPase Activity

(A) The effects of the pathogenic variants on TRIP13 level by immunoblotting in HeLa cells transfected with WT or mutant constructs. (B) TRIP13 level in proband-derived and control LCLs. (C) The effects of the pathogenic variants on HORMAD2 level by immunoblotting in HeLa cells transfected with WT or mutant constructs. (D) TRIP13 pathogenic variants led to increased HORMAD2 level in proband-derived LCLs. The black arrow indicates the band for HORMAD2. (E) The mRNA level of HORMAD2 in proband-derived and control LCLs. (F) The effects of the pathogenic variants on TRIP13 AAA-ATPase activity. Bars indicate the means ± SEM. We used one-way ANOVA analysis or Student’s t test analysis. ^∗∗∗p < 0.001, ^∗p < 0.05.

TRIP13 is an AAA-ATPase, and its ATPase activity is essential for its checkpoint function.^17,30 To determine if the pathogenic variants affect its ATPase activity, we transfected the WT and mutated TRIP13 plasmids into HeLa cells. Using the ADP-Glo Kinase Assay, we found that the mutant c.739G>A (p.Val247Met) led to a significant decrease in ATPase activity, whereas the other mutants had no obvious effects (Figure 2F), demonstrating that the c.739G>A pathogenic variant in family 4 has a distinct influence on TRIP13 function compared with the other pathogenic variants.

Effects of TRIP13 Pathogenic Variants on Cell-Cycle Progression and Chromosome Mis-segregation in Mitosis

Homozygous nonsense and splicing pathogenic variants have been reported to affect mitosis in individuals with Wilms tumors.²³ To test whether the homozygous missense and compound missense pathogenic variants we identified might also affect mitosis, we examined cell cycle progression and the frequency of chromosome mis-segregation in proband-derived and control LCLs. However, there was no significant difference in the cell cycle progression and the chromosome mis-segregation rate between proband-derived and control LCLs (Figures 3 and S3), suggesting that homozygous missense or compound missense pathogenic variants in TRIP13 did not have effects on mitosis.

Figure 3.

Effects of TRIP13 Missense Pathogenic Variants on Cell-Cycle Progression and Chromosome Mis-segregation in Mitosis

(A) Cell cycle stages of proband-derived and control LCLs. We used Pearson’s chi-square test. Samples of 10,000 cells were analyzed. The control data in the graph represents the average of three different controls. (B) Representative images and frequency of chromosome mis-segregation and lagging. The left panel shows representative images of chromosome mis-segregation and lagging chromosomes. The white arrow indicates chromosome mis-segregation, and the green arrow indicates the lagging chromosome. The scale bar represents 5 μm. The right panel shows the statistic result of chromosome mis-segregation and lagging. Pearson’s chi-square test, control (n = 57), family 1 (II-1) (n = 67).

Phenotypic Rescue by TRIP13 cRNA Injection Into Oocytes from One Affected Individual

In order to explore potential therapeutic treatment for these affected individuals, we injected TRIP13 cRNA into oocytes of proband II-1 from family 1 in her two most recent ICSI cycles. A total of 22 MI oocytes were retrieved, of which 13 oocytes were injected with TRIP13 cRNA. All injected MI oocytes extruded PB1, and 11 of the injected oocytes were successfully fertilized as indicated by the formation of two pronuclei on day 1, and seven of the oocytes developed into blastocysts on day 6 (Figure 4). In contrast, non-injected oocytes remained in the MI stage even after long-term culture. These results further confirm that TRIP13 pathogenic variants are responsible for infertility and suggest a possible treatment for affected individuals with pathogenic variants in TRIP13.

Figure 4.

Phenotypic Rescue by TRIP13 cRNA Injection into Oocytes of the Proband from Family 1

(A) In the control group, the two retrieved MI oocytes remained arrested at MI on day 1. (B) Three representative oocytes were injected with TRIP13 cRNA, and all three oocytes successfully extruded PB1 and entered MII. MII oocytes underwent ICSI, and the oocytes were monitored for 6 days after ICSI. The red arrows indicate PB1. The scale bar represents 80 μm.

Discussion

In this study, we identified a second mutant gene—TRIP13—responsible for human oocyte arrest at the MI stage (MIM: 616780). A recurrent homozygous missense pathogenic variant was identified in two families, whereas different compound heterozygous missense pathogenic variants were found in the other two families. These pathogenic variants caused a decrease in the TRIP13 abundance and led to HORMAD2 accumulation in HeLa cells and in proband-derived LCLs, which might be the reason for oocyte MI arrest. Injecting normal TRIP13 cRNA into oocytes of the proband II-1 from family 1 could rescue the phenotype.

TRIP13 is a widely expressed gene that functions in both mitosis and meiosis. In mitosis, TRIP13 acts as a mitotic checkpoint-silencing protein that catalyzes the conversion of C-MAD2 to O-MAD2 with the help of the cofactor p31comet.^18,30,31 In meiosis, TRIP13 is required at an early step in meiotic recombination and is involved in recombination pathways.^22,25 TRIP13 deficiency in mice results in the accumulation of HORMAD2 at synapsed chromosome axes,²⁷ and thus TRIP13 might function through different pathways in mitosis and meiosis. In addition, Trip13-knockout female mice are infertile and are totally devoid of follicles,²⁷ whereas female mice with Trip13 hypomorphic mutations have reduced TRIP13 abundance and are infertile but still show a few growing follicles in their ovaries.²⁷ Both mouse models showed no obviously abnormal phenotypes in somatic tissues or organs other than the reproductive system.²⁷ It is likely that in mice a certain dosage of Trip13 is required for normal meiosis, whereas there is a backup mechanism for mitosis under conditions of low or absent Trip13 expression. Although infertile Trip13-knockout mice show no phenotype in other normal tissues and organs, a previous study showed that individuals with homozygous nonsense or splicing TRIP13 pathogenic variants had Wilms tumors and chromosomal mis-segregation.²³ This suggests different roles for TRIP13 in mitosis in mice and humans.

Affected individuals with a recurrent TRIP13 homozygous c.1060C>T (p.Arg354^∗) nonsense pathogenic variant or a homozygous c.673−1G>C splicing pathogenic variant were previously shown to develop Wilms tumors as a result of chromosome mis-segregation,²³ whereas the affected individuals with a recurrent homozygous missense pathogenic variant (p.His26Arg) or different compound heterozygous missense pathogenic variants in TRIP13 in the present study suffered from infertility characterized by oocyte maturation arrest (Figure 1C). The two distinct diseases resulting from different types of TRIP13 recessive pathogenic variants might be caused by different effects of pathogenic variants. The homozygous nonsense pathogenic variant c.1060C>T (p.Arg354^∗) and homozygous splicing pathogenic variant (c.673−1G>C) associated with Wilms tumors and abnormal cellular mitosis led to total loss of TRIP13 function,²³ whereas in the present study all of the pathogenic variants we identified were missense pathogenic variants and only affected oocyte meiosis, and thus the function of TRIP13 was partly retained in the women we evaluated (Figures 2A and S2). This suggests that different dosage effects of mutant TRIP13 have different effects on human meiosis and mitosis, and these different effects result in the two distinct diseases.

There was phenotypic variability among affected individuals with bi-allelic missense pathogenic variants in the present study. Families 1–3 showed oocyte MI arrest, whereas family 4 had morphologically normal oocytes but showed zygote cleavage arrest. One possible explanation is that different variants have different effects on the function of TRIP13. Our ATPase activity assay indicated that, compared with other variants, only variant c.739G>A (p.Val247Met) from family 4 reduced the ATPase activity of TRIP13, which might be essential for the first zygote cleavage and cause a different phenotype in the individual. In addition, probands in families 3 and 4 carried compound heterozygous variants. Thus, the functional consequence of compound heterozygous variants on the protein was not the result of either one of the variants but was the result of the combined effects from both alleles. Therefore, even if one of the variants has mild or no influence on the protein abundance in vitro, the overall effect of compound heterozygous variants could affect TRIP13 function in vivo.

Previously, we and others identified several pathogenic genes responsible for human oocyte maturation arrest (TUBB8 and PATL2), human fertilization failure (TLE6, MIM: 612399; WEE2, MIM: 614084), and early embryonic arrest (PADI6, MIM: 610363; NLRP2, MIM: 609364; NLRP5, MIM: 609658).^{32, 33, 34, 35} To the best of our knowledge, there was no convincingly pathogenic gene reported to cause first zygotic cleavage failure in human. The present study suggests that TRIP13 might play a role during the first cleavage of the zygote, but the exact mechanism needs to be clarified in future affected humans or animal models.

In summary, we identified bi-allelic missense pathogenic variants in TRIP13 that cause female infertility and oocyte maturation arrest as a result of the accumulation of HORMAD2. We also explored TRIP13 cRNA treatment for rescuing the TRIP13 mutation phenotype. Our findings indicate that different dosage effects of mutant TRIP13 might result in two distinct human diseases. Although TRIP13 cRNA injection has implications for the future treatment of women with bi-allelic missense pathogenic variants in TRIP13, the effectiveness and safety should be carefully evaluated in mouse or non-human primate models in the future. If these can be fully evaluated and investigated, the strategy will have a rationale for preclinical trials, and this would be a first step toward a precise therapy for clinical infertility.

Declaration of Interests

The authors declare no competing interests.

Published: May 29, 2020

Web Resources

• dbSNP, https://www.ncbi.nlm.nih.gov/projects/SNP/

• ExAC Browser, http://exac.broadinstitute.org/

• GenBank, https://www.ncbi.nlm.nih.gov/genbank/

• gnomAD, https://gnomad.broadinstitute.org/

• HomozygosityMapper, http://www.homozygositymapper.org/

• MutationTaster, http://mutationtaster.org/MutationTaster/

• OMIM, https://www.omim.org/

• PolyPhen-2, http://genetics.bwh.harvard.edu/pph2/

• SIFT, https://sift.bii.a-star.edu.sg/