Supervisors info:
Δημήτριος Λουτράδης, Καθηγητής, Ιατρική, ΕΚΠΑ
Πέτρος Δρακάκης , Αναπληρωτής Καθηγητής, Ιατρική, ΕΚΠΑ
Κωνσταντίνος Καλλιανίδης , Αναπληρωτής Καθηγητής, Ιατρική, ΕΚΠΑ
Summary:
Variability in the subfertile patient population excludes the possibility of a single approach to controlled ovarian stimulation (COS) covering all the requirements of a patient. In the future, genetic screening may allow an individual patient's response to COS as well as IVF outcomes to be predicted based on genotype.
In the last 5-10 years a lot of attention has been drawn to the FMR1 gene and its implication in reproductive medicine. Our study focuses on possible effect of FMR1 gene on IVF outcomes. The FMR1 gene is located on the long arm of the X chromosome at position Xq27.3. A normal female karyotype is assigned as 46,XX meaning that every woman carries 2 FMR1 alleles which contribute to her genotype.
The most interesting feature of this gene lies in its 5’end untranslated region and it is a (CGG)n numerically polymorphic repeat.This (CGG)n repeat element can exert both positive and negative effect on translation where optimal translation appears near the modal repeat number within the general population. Fu et al. (1991), well characterized the distribution pattern of the different alleles among normal individuals and found that 29 CGG repeats is the most frequent allele while in normal Greek population 28 CGG repeats appear most frequently. In unaffected individuals the CGG ranges between 5-45 leading to normal mRNA transcription, translation and expression. The premutation allele (55-200 repeats) results in elevated mRNA transcription but reduced FMRP expression. Male carriers of a premutation allele have an increased risk of being affected from FXTAS while female carriers may be affected from FXPOI. A full mutation allele (>200 repeats) will lead to FXS. Our study focuses on the normal range.
FMRP is a major cellular translation repressor protein. It controls the level of translation of multiple transcripts (including its own). In humans,FMRP is expressed after birth in female germline predominantly in granulosa cells and in a rat experimental model an increasing FMRP expression has been observed with advancing follicle development. Women with (CGG)n outside the range associated with normal folliculogenesis (2634) are expected to have relaxed FMR1 transcription control, altered FMRP levels which will affect their folliculogenesis.
In the recent bibliography 29-30 CGG repeats have been proposed to be reflective of normal ovarian reserve while higher and lower counts denote risk for Premature Ovarian Senescence. Additionally, new genotypes, within the typical range, have been proposed for the ovarian function of the FMR1 gene: normal, if (CGG)n in both alleles are within 26-34, heterozygous if one allele is outside the normal range and homozygous if both alleles are outside the normal range. FMR1 genotypes have been found to be predictive of pregnancy, het-norm/low most significantly and with decreasing chance in comparison to norm genotypes. A low FMR1 allele (CGG<26 ) has been associated with significant poorer morphologic embryo quality and pregnancy chance.
We have studied 62 infertile women undergoing IVF, 26 poor responders (according to Bologna criteria) and 36 good responders, if having ≥8 oocytes, a basal FSH measurement of ≤10 mIU/mL and a peak E2 level ≥1500pg/mL.Their infertility reason was tubal, male or unexplained. Available data were age, weight, BMI, years of infertility, previous IVF, their hormonal profile (FSH, LH, PRL, TSH), parameters of ovarian stimulation/induction (days of stimulation, total FSH dose, peak E2) and their IVF outcome (number of follicles, number of oocytes, maturation rate, fertilization rate, embryo quality, pregnancy). Our workflow consisted of sample and data collection, DNA extraction, PCR amplification, capillary electrophoresis, software analysis and statistical analysis.
The distribution curve of FMR1 allele frequencies based on the CGG repeat number of the studied infertile Greek population was found in accordance with Fu’s Distribution curve. We found no shift towards higher ends neither in the total sample, nor the good or the poor responders separately. No intermediate or premutation alleles were found. (CGG)28 was the most frequent allele observed in the total sample, good or poor responders (43,5%, 42,3%, 44,4% respectively) and this was no different from Greek general population. The most prevalent genotype in both poor and good responders was norm (52,8% , 46,2% respectively).The distribution of genotypes between good and poor responders did not statistically differ (p-value 0,256). With regards to their FMR1 genetic background: infertile women carrying a norm genotype had less mean days of stimulation, higher mean of oocyte maturation rates and presented the best pregnancy odds. Infertile women carrying a hom genotype needed the most mean days of stimulation and had the lowest mean of oocyte maturation rates. The comparative pregnancy odds in this underrepresented subgroup remains for us inconclusive as a sample effect cannot be ruled out. Infertile women carrying a het genotype had 83% less pregnancy odds compared to those carrying a norm genotype. In the whole study population women not carrying a Low Allele in their Genotype had less mean number of days of stimulation compared to those carrying a Low Allele (p-value = 0,033). In the whole study population the presence of a low allele (CGG<26) was not associated with differences in pregnancy odds (p-value=0,198).
Conclusively, FMR1 genotypes seem to have both a quantitative (days of stimulation) and a qualitative (maturation rates) effect on IVF outcomes, as well as on pregnancy odds. The FMR1 could be considered as a candidate gene implicated in IVF success. A rationale of building up a multi-genetic, individualized profile with other genes involved in the IVF process is also an option, where FMR1 could prove informative.
The relative small size of the study population allows for only preliminary results, which require confirmation in a larger study population. The random X-inactivation could act as a possible modifier of the impact of FMR1 genotypes in the ovary.
