Does the presence of AGG interruptions within the CGG repeat tract have a protective effect on the fertility phenotype of female FMR1 premutation carriers?

Abstract

Purpose

While FMR1 premutation carriers (CGG 55–200) were shown to have reduced success with IVF treatment (lower oocyte yield), studies reporting on the association between the number of CGG repeats and patients’ response to controlled ovarian hyperstimulation (COH) are inconsistent. In the present study, we aim to explore whether the number of CGG repeats in women with premutation in FMR1 gene, undergoing COH for IVF, correlates with COH variables and whether the number of AGG interruptions may function as a “protective factor” by improving the ovarian response to COH.

Methods

Retrospective study, in an academic IVF-PGD unit. Fifty-seven consecutive FMR1 premutation carriers who underwent 285 IVF treatment cycles were included. The numbers of CGG repeats and AGG interruptions were retrieved and correlated to the demographics and COH variables.

Results

There were no significant association between the numbers of CGG or the AGG interruptions and the number of oocyte retrieved or the peak estradiol levels. The lack of association was also observed when including all the IVF treatment cycles or only the first or last IVF treatment cycle. Moreover, no associations were found between the number of CGG repeats or AGG interruptions and other COH variables, i.e., duration of stimulation, the total dose of gonadotropin used, or the number of top-quality embryos.

Conclusions

No associations were observed between the number of CGG repeats or AGG interruptions and any of the COH variables. Further studies are required to identify early biomarkers of POI to empower FMR1 premutation carriers with risk assessment tools to consider procedures such as fertility preservation.

Background

Trinucleotide expansion in the 5′ untranslated region of the fragile X mental retardation 1 (FMR1) gene located at Xq27.3 is associated with a spectrum of clinical phenotypes depending on the number of CGG repeats. Patients with fragile X–related mental retardation carry the full mutation CGG repeat expansions (> 200 repeats) [1], while expansion of the CGG triplet number above the normal range (n = 5–44) towards the so-called premutation status is associated with increased risk for fragile X–associated premature ovarian insufficiency (FXPOI) in females [2, 3] and fragile X–associated tremor/ataxia syndrome (FXTAS) predominantly in males [4].

Premature ovarian insufficiency (POI) is defined by the presence of postmenopausal levels of FSH (> 40 IU/L) in women under 40 years of age, with four or more months of secondary amenorrhea [5, 6]. POI affects ~ 1% of women before the age of 40 and 0.1% before the age of 30 [7, 8]. Previous studies suggest that premutation repeat length (n = 55–200) is associated with a spectrum of ovarian dysfunction with overt POI being at the extreme end of the scale [9, 10]. Carriers of the premutation on one allele have 16–35% risk of POI [11], compared with only 1% of females in the general population [12]. Moreover, fragile X premutation carriers with regular menstrual cycles were found to have an impaired ovarian function, as evident by abnormal ovarian reserve biomarkers [13–18].

A variety of models have been proposed as the culprits of FXTAS and FXPOI (summarized in [19]): (1) the toxic RNA gain-of-function model: This model suggests that DNA containing CGG-expanded repeats leads to the formation of dynamic intra-nuclear long rCGG RNA aggregates that sequesters specific RNA binding proteins. Proteins such as Sam68 and DGCR8 and its partner DROSHA directly bind to the double-stranded RNA hairpin structure of long rCGG RNA aggregates resulting in the loss of normal cell function and cell death. (2) Repeat-associated non-AUG (RAN) initiated translation: Due to translation of the expansion mutation, as part of a larger open-reading frame, mutant protein named polyglycine-containing protein is expressed in neuronal cells and granulosa cells, with the consequent disruption of cellular function, leading to cell toxicity.

Since ovarian reserve greatly affects IVF success, Pastore et al. [20] have recently studied the effect of FMR1 CGG repeat lengths on IVF outcomes after controlled ovarian hyperstimulation (COH). In their meta-analysis, they could demonstrate that FMR1 premutation carriers (CGG 55–200) were shown to have reduced success with IVF treatment (lower oocyte yield) than women with a normal CGG repeat length or a full mutation, although findings have yielded conflicting results. Moreover, the association between the number of CGG repeats and patients’ response to COH was also inconsistent [21–25].

Recent studies revealed that AGG interruptions within the CGG repeat area might function as a “protective factor,” by decreasing the risk for intergenerational expansion, with a proven substantial influence upon the risk of a full mutation expansion in a given number of repeats [26, 27]. Moreover, Domniz et al. [28] have recently demonstrated that the risk of transmitting full mutation expansions decreased as the number of the AGG interruptions increased.

Prompted by the aforementioned observations, we aimed to explore whether the number of CGG repeats in women with premutation in FMR1 gene, undergoing COH for IVF, correlates with COH variables and whether the number of AGG interruptions may function as a “protective factor” by improving the ovarian response to COH.

Methods

Patients and IVF treatment

The study population consisted of all consecutive FMR1 premutation carriers referred to our IVF unit for IVF-pre-implantation genetic diagnosis (PGD) or fertility preservation treatment, between September 2011 and April 2019, who reached the ovum pickup (OPU) stage. The study was approved by the Institutional Ethical Review Board of Sheba Medical Center, Israel.

Data on patients’ age, infertility treatment-related variables and ovarian stimulation characteristics, number of oocytes retrieved, and number of embryos transferred per cycle were collected and correlated to the patients’ number of CGG repeats and AGG interruptions.

The study required no modification of our routine IVF protocols. The selection of type of COH protocol used was the decision of the treating physician. In all protocols, gonadotropins were administered in variable doses, depending on patients’ age and/or ovarian responsiveness in previous cycles, and further adjusted according to serum estradiol (E2) levels and vaginal ultrasound measurements of follicular diameter obtained every 2 or 3 days. Thirty-four to 36 h after HCG injection, oocytes were aspirated by the ultrasound-guided transvaginal route. Top-quality embryos on day 3 were defined as embryos with ≥ 7 blastomeres and ≤ 15% fragmentations.

Molecular characterization

The number of FMR1 CGG repeats in the polymorphic zone and the number of AGG interruptions were determined using a triplet-primed PCR method as reported in previous studies [28]. The FMR1 CGG Primer is specific for CGG repeats and will not hybridize to AGG sequences commonly found in FMR1 alleles. The accuracy and validity of our systems were assured using AmplideX© PCR process control and mPCR and sensitivity.

Statistical analysis

Analysis of variance (ANOVA) and regression analysis were used to identify statistically significant effects between the number of CGG repeats, number of AGG interruptions, duration of treatment, patients’ age, and the amount of gonadotropin used in each cycle, as factors that might potentially affect COH outcome (peak E2 levels and the number of oocytes retrieved). Furthermore, charts and descriptive statistics were also used to evaluate any possible relationship between the number of CGG repeats and AGG interruptions and the fertility and COH variables. All statistical analyses were performed using ®JMP Statistical Discovery software, version 14.3.0, from ®SAS Institute Inc., Cary NC. A p value of 0.05 was considered significant.

Results

Fifty-seven consecutive FMR1 premutation carriers who underwent 285 IVF treatment cycles were evaluated. Patients’ characteristics are presented in Table 1.

Table 1.

FMR premutation carriers’ characteristics and COH variables (n = 57)

Patients characteristics	Mean + SD
Age	31.7 ± 3.9
Day 3 FSH-IU/L	9.6 ± 6.6
Day 3 LH-IU/L	5.5 ± 3.5
Day 3 FSH/LH ratio	1.8
Day 3 Estradiol (pmol/L)	224.7 ± 191
Total gonadotrophins used for stimulation-IU	2747 ± 2268
Peak estradiol (pmol/L)	5401 ± 3187
Number of oocytes retrieved	7.6 ± 5
Number of day 3 top-quality embryos	2.5 ± 2.5
CGG repeats	87.8 ± 024
Range of CGG repeats	55–200
AGG interruptions	0.7 ± 0.7
Range of AGG interruptions	0–2

There were no significant association between the numbers of CGG or the AGG interruptions and the number of oocyte retrieved or the peak estradiol levels (Figs. 1 and 2). Moreover, when stratifying patients according to their response to COH: poor responder, those yielding 1–3 oocytes; suboptimal response, 4–9 oocytes; normal response, ≥ 10 oocytes [28]; again, no associations were observed between the patients’ response to COH and the number of CGG repeats or AGG interruptions (Figs. 1 and 2).

Fig. 1.

Oocyte number and peak E2 vs. number of CGG repeats

Fig. 2.

Oocyte number and peak E2 vs. number of AGG interruptions

The lack of association was also observed when including all the IVF treatment cycles, only the first or only the last IVF treatment cycle. Moreover, no associations were found between the number of CGG repeats or AGG interruptions and other COH variable, i.e., duration of stimulation, the total dose of gonadotropin used, or the number of top-quality embryos. Re-analysis results on different categories of CGG repeat numbers (< 70, 70–90, > 90, and < 100, > 100) using regression models revealed no effects of either CGG repeats or AGG interruptions, on COH and embryological variables.

Discussion

In Israel, as part of a national prenatal screening program, all women who wish to conceive are advised to determine their FMR1 CGG carrier status. All FMR1 premutation carriers (55–200 CGG repeats) are further referred to a genetic consultation to consider IVF and PGD in order to avoid the risk of CGG expansion in the offspring. Moreover, those who do not intend to conceive, between the ages of 30 and 41 years old, are offered IVF treatment for fertility preservation of their own oocytes or embryos. Therefore, in all women undergoing IVF in our center, the CGG repeat status is known (either FMR1 premutation carriers (55–200 CGG repeats) or normal (< 55 CGG repeats).

In the present study of FMR1 premutation carriers undergoing IVF treatment, no associations were observed between the number of CGG and the number of oocyte retrieved or peak estradiol levels. Moreover, when stratifying patients according to their response to COH (poor responder, suboptimal response, and normal response [29]), there was no in-between groups difference in CGG repeats. Same lack of associations was observed when correlating the numbers of AGG interruptions and the aforementioned COH variable.

Ennis et al. [30] found a significant nonlinear relationship between the premutation size and menopause age with a minimal value at 80 repeats. While FMR1 premutation carriers with less than 80 repeats demonstrate increased severity of the phenotype with increasing repeat number. Thereafter, with increased repeat number (over 80 repeats), there was a reduction in the phenotype severity with increased menopausal age. Sullivan et al. [12] examined the association between the CGG repeat size and various measures of ovarian function. They found that the repeat size significantly influenced the risk for ovarian dysfunction. While the risk for POI increased significantly when the repeat sizes were between 80 and 99 compared with those with ≤ 40 repeats, the risk did not increase for those with ≥ 100 repeats compared with those with 80–99 repeats. Tejada et al. [31] have retrospectively studied the number of CGG repeats in FMR1 premutation carriers with and without POI. They observed a nonlinear association between CGG repeat number and the manifestation of POF, showing that the likelihood of having POF was significantly higher in carriers with fewer than 100 CGG repeats compared with carriers with 100 or more CGG repeats.

Pastore et al. [20] have recently studied the effect of CGG repeat lengths on IVF outcomes after COH. In their meta-analysis, they could demonstrate that women with CGG repeats in the normal (< 45 CGG), intermediate range (45–54 CGG), or both showed a minimal effect on IVF outcomes. Moreover, premutation carriers (CGG 55–200) were shown to have reduced success with IVF treatment (lower oocyte yield) than women with a normal CGG repeat length or a full mutation, although findings were inconsistent.

Several studies examined the association between the number of CGG repeats and patients’ response to COH. Rohr et al. [21] examined AMH levels (a marker of ovarian reserve) among FMR1 premutation carriers (n = 145) to characterize their ovarian function. AMH levels were significantly lower among longer repeat allele (≥70) carriers compared with shorter repeat allele (< 70) carriers. On the other hand, Bibi et al. [22] found that FMR1 premutation carriers (n = 18) with < 100 CGG repeats suffer from impaired ovarian response and decreased fertilization rate during IVF treatment compared with patients with ≥ 100 CGG repeats. Elizur et al. [23] studied 21 premutation carriers and demonstrated a significant nonlinear association between the number of CGG repeats and ovarian function with the mid-size range (80–120) having the worst prognosis [23], with significantly less oocyte retrieved compared with FMR1 premutation carriers with higher or lower CGG repeat tracts.

On the contrary, Tsafrir et al. [24] have studied 22 FMR1 premutation carriers underwent 63 IVF treatment cycles. While FMR1 premutation carriers required higher doses of gonadotrophins and had lower peak estradiol levels and fewer oocytes retrieved, no association was found between the number of CGG repeats and the number of oocytes retrieved. Another study [25] has studied 51 FMR1 premutation carriers (an extension of their previous study [22] underwent IVF-PGD). A multivariate mixed model analysis showed no association between the number of CGG repeats and the square root of the number of oocytes retrieved. Similarly, no association was found between the number of CGG repeats and the fertilization rate or cleavage rate. Adding quadratic transformation on the number of CGG repeats or a dummy variable that categorized the number of repeats into three groups (< 80, 80–99, and > 99 repeats) or two groups (< 100, ≥ 100 repeats) did not improve the model either.

At the DNA level, AGG interruptions within the CGG repeats play an important stabilizing and protective role [32], as demonstrated by the reduced risk of instability during inheritance with increased number of AGG interruptions [26, 28]. Prompted by the theoretically ameliorating effect of AGG interruptions on inheritance, another objective of our study was to assess the number of AGG interruptions on COH outcome. The number of AGG interruptions within the CGG repeat area did not protect the patients, or improved ovarian response to COH. Lekovich et al. [33] studied the effect of AGG on ovarian reserve in permutation carriers and found higher ovarian reserve among women with two AGGs compared with women with none or 1 AGG interruption. On the other hand, Allen et al. [34] investigated the influence of AGG interruptions on the age of secondary amenorrhea in premutation carriers and could not demonstrate any association between those parameters.

While the smaller studies (n < 50) have demonstrated an association between the CGG repeat size and ovarian response to COH [22, 23], larger studies consisting of > 50 FMR1 premutation carriers could not demonstrate any association between the number of CGG repeats and the number of oocytes retrieved ([25], and the present study). Moreover, similar to the study of Allen et al. [33], we demonstrated a lack of protection against poor ovarian response with increased number of AGG interruptions. The differences between ours and Lekovich et al.’s [33] study might be explain by the differences in the sample size (57 vs. 32 FMR1 premutation carriers), and the different patient characteristics (55–200 vs. < 100 CGG repeats).

While the limitation of this study stems from its retrospective nature which limits the ability to control for potential unknown confounding factors, its strengths come from the relatively large sample size, consisting of 57 FMR1 premutation carriers who underwent 285 IVF treatment cycles.

Conclusion

In conclusion, our study could not demonstrate any association between the number of CGG repeats or AGG interruptions and the number of oocytes retrieved or any COH variables. Further studies are required to better understand the molecular mechanisms responsible for POI in permutation patients, in order to develop novel remedies or to identify early biomarkers of imminent POI that will allow patients to attend for fertility preservation procedures, in advance.

Abbreviations

COH

Controlled ovarian hyperstimulation

E2

Estradiol

FMR1

Fragile X mental retardation 1

FXPOI

Fragile X–associated premature ovarian insufficiency

FXS

Fragile X syndrome

FXTAS

Fragile X–associated tremor/ataxia syndrome

IVF

In vitro fertilization-embryo transfer

OPU

Ovum pick-up

PGD

Pre-implantation genetic testing

POI

Premature ovarian insufficiency

Authors’ contributions

MFG designed the study and analyzed the data. MK reviewed and analyzed the data and drafted the manuscript. AM and ND contributed and retrieved patient information. SE and HR reviewed the data. RO designed the study, reviewed and analyzed the data, and drafted the manuscript. YC designed the study and reviewed and analyzed the data. All authors read and approved the final manuscript.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Compliance with ethical standards

Ethics approval and consent to participate

This study was approved by the Research and Ethical Committee of the Sheba Medical Center (Tel Hashomer), Ramat-Gan, Israel.

Competing interests

The authors declare that they have no conflicts of interest.