The significance of FMR1 CGG repeats in Chinese women with premature ovarian insufficiency and diminished ovarian reserve

The fragile X mental retardation 1 (FMR1) gene, located at Xq27.3, is an X-linked gene, carrying CGG repeats in the 5′-untranslated region. Based on the criteria stated by the American College of Medical Genetics and Genomics (ACMG), FMR1 CGG repeats can be classified as normal (< 45 repeats), intermediate (45–54 repeats), premutation (55–199 repeats), and full mutation (> 200 repeats) [1].

The ovarian reserve indicates a woman’s reproductive potential, which is a function of the number and quality of oocytes [2]. Premature ovarian insufficiency (POI) is a condition that accentuates the extreme spectrum of impaired ovarian function. It is characterized by failure of ovarian function in women less than 40 years of age [3]. The decline of ovarian reserve is a continuous, gradual process and the concept of decreased or diminished ovarian reserve (DOR) has been explored in few reports earlier [4]. DOR is not an overt phenotype and was once described as ‘occult POI’ [5]. Compared with women of similar age, women with DOR commonly have regular menses but they also have reduced number of ovarian follicles and reduced fecundity [2]. It is thought that DOR may or may not progress into POI eventually (depending on whether amenorrhea occurs before the age of 40 years).

Numerous studies have examined the association between FMR1 CGG repeats and ovarian dysfunction [6‐9]. In western countries, premutation of FMR1 are reportedly correlated with POI in women [7‐9], which is now referred to as fragile-X-associated primary ovarian insufficiency (FXPOI). It has been estimated that around 11–14% of familial and 2–6% of sporadic POI cases are associated with FMR1 premutations [3]. Association between normal or intermediate range FMR1 alleles and a reproductive risk has also been explored [10, 11]. Multiple association studies of FMR1 alleles, i.e. CGG repeats > 36 [12], 41–58 [13], 45–54 [9], and 35–54 [14] have been reported to be associated with POI. However, the distribution of FMR1 CGG repeats also varies with ethnicity [15]. Fewer patients with POI from Asia carried the FMR1 premutations [16].

The pathogenic role of FMR1 premutations in Chinese women is controversial. Three earlier studies have shown very low prevalence of premutation carriers in Chinese women with POI (< 1%) [17‐19], which is lower than studies from western countries [3]. Therefore, FMR1 premutation may not be a common explanation for POI in Chinese women.

With regard to the relationship between DOR and FMR1, the results are inconsistent: Women with DOR might be at a risk of carrying alleles in the premutation range [5, 20]. Some studies found an association between DOR and normal/intermediate CGG repeats, including < 26 [21], < 28 [22], 35–54 [23], 45–54 [9, 24], or > 40 [5] and have reported a negative effect on ovarian ageing. By contrast, some studies reported that ovarian reserve was not affected by CGG repeats [23‐27]. The role of FMR1 in Chinese women with DOR has not been extensively investigated.

This study protocol was approved by the Institutional Review Board (IRB) of Peking Union Medical College Hospital (PUMCH) (No. JS-1604).

In the present study, we compared the distribution of FMR1 CGG repeats among 124 patients with POI, 57 patients with DOR and 111 controls in Chinese population. No statistically significant differences were identified between the DOR group and the control group, whereas the distribution of allele 1 CGG repeat in patients with POI was difference from that in the control group. Frequency of premutation was relatively low in both healthy women and patients with ovarian dysfunction in China: two cases with premutation carriers were identified in the POI group, 1 premutation carrier in the control group, and no carriers were found in the DOR group.

POI occurs in 1% of women and has severe consequences, including infertility and chronic hypoestrogenism that may result in increased cardiovascular risk, impaired bone health, and considerable psychosocial sequelae [32]. A higher prevalence of spontaneous POI has been reported from a very recent meta-analysis, i.e. 3.7% among all women worldwide [33]. Hormone replacement therapy, the principal therapeutic approach for POI, helps alleviate the related symptoms although this does not effectively solve the issues related to fertility.

The main cause of POI is unknown, but genetic factors, autoimmune ovarian damage, iatrogenic and environmental factors are the known causes. Among all the genetic factors implicated for POI, the FMR1 premutation is regarded as the leading single-gene cause of POI [34]. If FMR1 screening for the target population could select women with high risk of POI, it has long-term benefits to family and enables them in planning the families with an opportunity to adopt alternate methods. The American College of Obstetricians and Gynecologists (ACOG) and the European Society of Human Reproduction and Embryology (ESHRE) recommend population-based FMR1 screening for women younger than 40 years of age and presenting with ovarian insufficiency [35, 36]. FMR1 testing has a dual role in patients with ovarian insufficiency: determining the probable cause of ovarian failure and identifying women at risk of transmitting mutations to their offspring. However, studies have shown wide heterogeneity and inconsistency in the association between FMR1 CGG repeats and POI susceptibility across different ethnicities [15, 37].

FMR1 premutation carriers showed an increased risk of POI, especially in a population of European descent [37]. However, this factor may not contribute to the POI susceptibility in the Asian population [37, 38]. Studies of Iranian [39] and Indian [37] populations found no significant relationship between the FMR1 CGG repeat and POI. Studies from China also showed FMR1 premutation to be an uncommon explanation for POI [18, 19]. The frequency of premutation carriers among Chinese women with sporadic POI ranges between 0.5 and 0.9% in three previous studies [17‐19]; these studies analyzed FMR1 CGG repeats using PCR and capillary electrophoresis. Our study shows the premutation frequency in the POI group was 1.6% using PCR and microfluidic capillary electrophoresis. A previous study showed that the repeat sizes determined from both methods were largely concordant and, on average, were within one repeat size difference [30]. Then, we speculated that the error rate within the two methods is acceptable. Our results combined with three previous reports in China, shows the premutation frequency in Chinese women with sporadic POI may be 0.87% (6/693). Typically, in healthy women, the premutation carrier frequency has been reported as 1/579–1/1955 in China [40‐42], 1/781 in Korea [43]. The permutation frequency in Asian women [41] is much lower than the statistics in white (1/169), African American (1/124), and Hispanic (1/287) healthy women [44]. Although limited by sample size, our study found no statistical difference in the premutation carrier frequency among POI and healthy women. It is possible that the premutation carrier frequency in Chinese POI patients is higher than that in healthy women, but the overall incidence is relatively low in both groups. Further study with a larger sample size is needed to get a definite conclusion.

Studies have reported controversial outcomes, wherein intermediate and normal alleles are associated with POI frequency [13, 14, 19]. Our study, in agreement with others, found the distribution of allele 1 in the POI group was different from healthy women. Another study from China suggested that the CGG repeats in allele 1, and not allele 2, were significantly associated with POI occurrence [17]. The risk of POI occurrence for < 26 and ≥ 29 CGG repeats in allele 1 was higher than that for 26–28 CGG repeats. Gleicher et al. [45] demonstrated that < 26 repeats of both alleles have negative effects on reproduction.

Various studies also explored the association of FMR1 CGG repeats with DOR. Most of these studies have shown association of DOR with the number of CGG repeats of the FMR1 gene [5, 9, 21‐24, 46], while few other studies including ours suggest that the number of CGG repeats of the FMR1 gene seems to be independent of DOR [26, 27]. Our study found no difference in the CGG repeats of allele 1 between the DOR and control groups in Chinese women. Pastore et al. [25] found a significant difference in the CGG repeats of allele 1 between the DOR cases and women with a normal reproductive history among the Whites, but not the Asians, since Asian women seem less likely to have an allele with ≤25 CGG repeats than other races [25, 47]. Race variation may be associated with the different results.

The previously reported frequency of 35–44 repeats is 14.5–17% in the DOR group and 3.9% in the controls [5, 23], suggesting that the CGG repeats of 35–44 may be markedly overrepresented in women with DOR. We compared the proportion of 35–44 CGG repeats among groups in this study, and found no difference of the prevalence between the DOR group and control. In the control group, which included women with normal menopausal age, the frequency of 35–44 CGG repeats is 28.8%, which is much higher than in previous reports. The difference might be attributable to the presence of the secondary modal peak seen in the Asian population. Women in the present study had a primary modal peak at 29 to 31 repeats and a secondary modal peak at 36 to 38 repeats of allele 2. Reports showed other Asians such as Japanese [12, 48] and Indonesians [41] also have a secondary peak, which was not identified in studies of western populations [12]. The presence of a secondary modal peak may be related to varied outcomes across different races.

The underlying mechanism behind CGG repeats regulating FMR1 gene expression in the ovary, and thereby affecting ovarian function remains unknown. It is unclear whether the reduced ovarian reserve represents a pathological condition resulting from abnormally accelerated atresia in a normal antral follicular pool or an abnormally small initial pool of oocytes [49]. A recent study showed that various CGG expansions of an FMR1 allele may lead to changes in RNA level and ratios of distinct RNA isoforms, which could regulate the translation and/or cellular localization of fragile X mental retardation protein (FMRP), affecting the expression of steroidogenic enzymes and hormonal receptors, that result in ovarian dysfunction [50]. In addition, Dioguardi et al. found that the permutation transcript contributes to the mitochondrial and ovarian abnormalities in permutation mouse models [51]. Studies showed that ovaries from Fmr1 knockout mice show increased mTOR protein [52, 53], and the YAC mice with premutation CGG repeat show reduced phosphorylated mTOR levels [54]. Then, both underexpression and overexpression of mTOR can result in ovarian dysfunction [55]. A recent study showed a potential relationship between the regulation of FMR1/FMRP expression and the AKT/mTOR signaling pathway in a human proliferating granulosa cell model system [56]. The above experiments suggest the mTOR pathway as a potential therapeutic target. However, the result of CGG repeats affecting the ovarian ageing process in various ethnicities is inconsistent, and the mechanism of heterogeneity in varied ethnicity is unclear. Further functional studies are needed to explain the inconsistent results across different ethnicities and susceptibility to ovarian insufficiency.

The primary strength of this study was the identification of women with a well-defined phenotype, independent of any potential risk factors for analysis of FMR1 CGG repeats. Second, as the association FMR1 gene and DOR has not been evaluated, by enrolling both POI and DOR patients in this study, we could have a comparison between these two groups in one Center. The primary limitation of the study however is the relatively small sample size.

No difference of FMR1 alleles in the premutation ranges were found between POI or DOR and healthy women and the frequency of premutation was relatively low in Chinese women. The distribution of allele 1 CGG repeat in patients with POI showed some different from that in healthy women. Although FMR1 testing has been recommended in the evaluation of the etiology of ovarian insufficiency in western countries, screening for the same among Chinese women is not warranted as the low frequency of occurrence. Further investigations with larger sample sizes is necessary to study the incidence of FMR1 expansions in all forms of ovarian insufficiency to confirm the results of this pilot study.