In this paper, we found a novel FMN2 heterozygous variant c.1949C>T (p.Ser650Leu) carried by all three patients in a Chinese family, and another FMN2 heterozygous variant c.1967G>A(p.Arg656His) in a sporadic case, through the whole exome sequencing. Since no FMN2 missense mutation is reported for causing human POI, we preliminarily assessed p.Ser650Leu variant via cross-species alignment and 3D modeling and found it possibly deleterious. A series of functional evidence was also consistent with our hypothesis. We proved the expression of FMN2 in different stage of human oocytes, and observed a statistical difference of chromosomal breakages between the POI patient carrying p.Arg656His variant and the health control (p = 0.0013). Western Blot also suggested a decrease in FMN2 and P21 in the mutant type, and an associated increase in H2AX. The p.Arg656His variant filtered from 20 sporadic cases with an extremely low variant frequency also indicated that the gene FMN2 might play an essential role in the genetic etiology of POI. To the best of the authors’ knowledge, this is the first report on missense variants of FMN2 that cause POI.
POI and DNA damage repair
Premature Ovarian Insufficiency has been a severe problem for female reproductive health. As a heterogenous and irreversible disease, many new techniques have been used to achieve the purpose of early diagnosis. About 5% of POI patients have a clear family history of early menopause, suggesting that familial aggregation exists. However, most POI patients still have no clear etiology.
Major causes of POI are follicular dysfunction and premature depletion of functional primordial follicles [17]. The eukaryotic cell has a certain ability of self-repair. If the damage is severe enough beyond repair, the apoptosis mechanism may be activated to remove permanently damaged cells, leading to a decrease in the number of normal cells [18]. For the follicle reserve is usually limited, the function of DNA repair seems far more critical. Before mammals were born, oocytes have been long-term arrested at the first meiotic prophase [19]. A large number of DNA replicates during meiosis will be completed during the same period. Therefore, to ensure the correct replication of genetic information in meiosis Ⅰ and the proper separation of homologous chromosomes, oocytes have a mechanism of checkpoints at the pachytene stage of meiosis prophase [20]. When DNA damage occurs, checkpoints are activated to arrest the cell cycle and initiate the downstream DNA repair pathway. DNA can be quickly repaired during this time if the damage is minor or there is a perfect DNA repair system. Otherwise, abnormal repair may be formed, leading to cell apoptosis or even tumorigenesis [21-24]. In our cases, in addition to a three-generation POI family, the proband’s grandmother was diagnosed with mammary cancer as well, giving her first and second-degree relatives a much higher rate of morbidity than normal [25, 26].
Mammalian oocytes are particularly vulnerable to DNA damage. Physiologically, they may lie dormant in the ovary for many years (> 40 in humans) until they receive the stimulus to grow and acquire the competence to become fertilized. The implication of this is that in some organisms, such as human, oocytes face the danger of cumulative genetic damage for decades. Given the critical role of meiosis in germ-cell survival, endocrine and fertility problems often occur in persons with a defect in DNA repair [18]. In our lymphocyte models, we found a slightly higher number of chromosomal breaks in the mutant type in each concentration of MMC. And under the condition of 150nM, the number of breaks between the two groups showed a statistically significant difference, consistent with other MMC induced models [27-30]. Besides, the higher concentration of MMC we used, the more necrotic cells exist. And at 300nM condition, it was unprecedentedly challenging to find karyotypes and count the breaks because of the full view of necrotizing cells under a microscope. However, we did not find a suitable statistical method to measure the apoptotic cells in our slides, which may cause inaccuracy in the comparison results between the mutant type and wild type. Further functional studies and larger sample size are needed in the future work.
WES and POI
The theory goes hand in hand with reports of POI cases caused by DNA repair-related genetic abnormalities [31]. Through whole-exome sequencing, Qin et al. [10] discovered three novel mutations in CSB-PGBD3, which caused a dysfunctional DNA repair mechanism. Novel mutations in the gene BRCA2 [9] and FANCA [32] related to DNA repair were also found in the same way.
We have been collecting basic information and peripheral blood samples of POI patients in the Second Xiangya Hospital of Central South University since 2018. In our past work, familial and sporadic POI samples were counted.
In this study, WES and Sanger sequencing are performed in a non-syndromic Chinese POI family, identified the c.1949C>T variant in FMN2. Cross-species alignment and 3D modeling predicted it as possibly deleterious. Functional studies further implied the hypersensitivity it caused towards mitomycin C. Generally, these results suggest that the novel mutation leads to lower expression in protein FMN2, caused a deficiency in DNA repair. P21 also shows a lower expression, referring to a shorter cell cycle arrest and time for those damage sites to be fixed. Among 20 sporadic POI cases, one additional missense mutation ENST00000319653.9:c.1967G>A (p.Arg656His) in FMN2 was found simultaneously, which can be a lateral confirmation of the essentiality of FMN2 on DNA repairing. As a candidate gene causing POI, it needs to be intact for normal ovarian development and maintenance.
WES has helped us discover many candidate genes related to DNA repair in POI pedigrees and sporadic POI cases. However, it is still in the initial phase revealing the genetic etiology of this disease.
FMN2 and POI
Formin homology proteins are actin regulators with scaffold function (implicated in organogenesis), normal tissue homeostasis, and invasion and metastasis of cancer cells through the regulation of actin remodeling. Human FMN2 (1722 aa) showed 74.7% total-amino-acid identity with mouse Fmn2 and 31.9% total-amino-acid identity with human FMN1. Although the N-terminal half was divergent between FMN2 orthologs and FMN1 orthologs, FH1 and FH2 domains were conserved among FMN2 and FMN1 orthologs [13]. Little work has been done on human FMN2, but mouse and drosophila FMN2 homologs are relatively well studied. FMNL1, FMNL2, FMNL3, DIAPH1, DIAPH2, DIAPH3, DAAM1, DAAM2, Fmn2, FHOD1, FHOD3, GRID2IP, and FHDC1 are Formin homology proteins with FH1 and FH2 domains [33-36]. With multiple proline-rich motifs, the FH1 domain is the binding region for Profilin, SRC, EMS1, FNBP1, FNBP2, FNBP3, FNBP4, and WBP4 [37-41], while the FH2 domain is the actin-structure modification region [42, 43]. Recent studies declared that mammalian Spir1 and Spir2 KIND domains were reported to bind directly to the C-terminal tail, distal to the FH2 domains, of Fmn1 and Fmn2 [44]. The binding domain (residues 1023–1059), which formed Spir1/Fmn2 Complex with Spir1, are necessary and sufficient to stabilize actin [15]. A study claimed that the nuclear actin assembled by Formin-2 and Spire-1/Spire-2 plays an important role in the DNA repairing procedure, which implies another role of Fmn2 [14].
However, the function of the rest residues remains unknown. Our work predicted that p.Ser650Leu missense mutation might lead to a tremendous change in the 3D structure of the FH2 domain, which may affect or even lose the function of actin binding. FMN2 is now reported as a critical regulator of the cyclin-dependent kinase inhibitor p21 [11]. Some existing research claimed that increasing FMN2 protein levels promotes cell cycle arrest by inhibiting the degradation of p21, which is closely related to the procedure of DNA damage repair [18], verifying the conjecture above in another hypothesis. This phenomenon is highly consistent with our experimental results and provides evidence for the pathogenicity of FMN2 we identified in this pedigree. Our work proves a lower FMN2 and P21 expression under MMC treatment, suggesting that when under severe environmental stimuli, DNA damage repair increases, FMN2 expression increases, and P21 increases accordingly, consistent with previous research. And according to our results, p.Ser650Leu heterozygous variant may lead to a decrease in FMN2 and p21 expression under the same conditions. H2AX is the first step in recruiting and localizing DNA damage proteins. Former studies in DNA damage models showed a dose-dependent decrease in H2AX when treated with radiation or chemotherapy drugs [46]. In our study, H2AX was decreased in both types under 150nM MMC, which means more H2AX consumption was needed for DNA repair, suggesting a more serious DNA damage and a DNA repair that is difficult to keep up with on time. Besides, there was a comparatively higher H2AX expression in the mutant type in each condition, indicating a higher DNA damage rate. To all appearances, this heterozygous missense mutation in FMN2 caused more damage and a shorter repairing time.
FMN2 has been proved as a significant gene related to mouse reproduction. In a mouse model knocked out 433aa of the FH1 domain, female FMN2-/- mice shown a phenotype with abnormal oocyte morphology, abnormal polar body morphology, abnormal female meiosis, abnormal pregnancy, and reduced female fertility [45]. The pathogenicity of FMN2 in human may be slightly different. According to THE HUMAN PROTEIN ATLAS (https://www.proteinatlas.org/ENSG00000155816-FMN2/tissue), human FMN2 protein is expressed in brain, ovary, testis, and other tissues. Previous research reported mental retardation cases with FMN2 nonsense or truncation mutations on different sites [46, 47]. In 2016, another study found a significantly higher copy number variation (CNV) in the FMN2 gene while statistically comparing the sporadic POF patient population with healthy controls. However, they did not provide a more functional experimental basis [48]. Yet, no FMN2 missense has been found underlying any POI case. Our work proves the existence of FMN2 in different stages of human oocytes on protein levels, consistent with the Human Protein Atlas. WES reports of a family POI case and 20 sporadic POI cases provided the novel missense variant p.Ser650Leu, on which we mainly focused in this work, and another rare variant p.Arg656His of FMN2, which was predicted to be possibly deleterious by three in-silico tools. Although our sample size is too small to clarify the impact of these two mutations, it is suggested to deem FMN2 as a candidate gene responsible for POI, acting through insufficient DNA repairing.