Abstract Study question Can profiling of variation in follicle-stimulating hormone receptor (FSHR) gene expression forecast response to Controlled Ovarian Stimulation (COS) in Assisted Reproductive Techniques? Summary answer Full-length FSHR transcript was the most common mRNA species. Nine distinct transcripts were characterised, but only one (deletion Exon 6) was associated with altered response. What is known already AMH levels and age are considered useful predictors of ovarian response to COS (high AMH and young age typically associated with high yields of oocytes; older ages and low AMH associated with fewer oocytes). However, it is not uncommon to see patients who do not fit this pattern, highlighting the limitations of these parameters in predicting response. Age-independent biomarkers capable of revealing patients at elevated risk of hyperstimulation, or those who might produce fewer than the optimal number of oocytes, are desirable. Alternative transcripts of the FSHR gene have been proposed as an explanation for variation in response to COS. Study design, size, duration FSHR splice variants in mRNA in granulosa cells from 126 patients were examined utilising long-read nanopore sequencing. This strategy permitted the sequencing of mRNA variants without fragmentation, allowing the detection of all splicing events in each mRNA strand. In addition to being categorised based on age and AMH levels, patients were also classified by the expression of FSHR transcript variants and their response to stimulation (high, low, or normal response). Participants/materials, setting, methods Patterns of FSHR mRNA splice variants were analysed in each of 126 patients. This involved extraction of RNA from the granulosa cells, followed by long-read nanopore sequencing, permitting analysis of whole transcripts in a single, contiguous sequence ‘read’. Patients were classified based upon their level of response to COS (low, normal, high), taking into account their age, AMH level and the number of oocytes retrieved. Patterns of FSHR transcript variation were compared to COS response. Main results and the role of chance Full-length FSHR transcript was detected in 97.6% of patients and skipping of Exon 6 was observed in 68%. Five novel transcripts were discovered, including partial insertion of ∼100bp segments of introns 1 or 2, seen in 6.55% of samples. In total, nine distinct FSHR mRNA variants were detected. As expected, increasing age was associated with higher FSH requirement per oocyte collected (p = 0.018), and AMH levels declined with age (2.8± 1.42 in women <38 versus 1.58 ±1.29 in women ≥38; p = 0.011). Consistent with previous analyses, AMH levels and amount of FSH per oocyte retrieved were negatively correlated (p < 0.001). One-third of the patients (33%) in this cohort displayed responses to COS that were not considered to be consistent with expectations based upon age and AMH levels. Expression of the novel FSHR transcript variants was not associated with FSH needed per oocyte retrieved (p = 0.78), nor the proportion of mature oocytes (p = 0.95).The only transcript potentially linked to an altered response to COS was the previously reported mRNA species, lacking exon 6, associated with significantly greater numbers of oocytes produced (p = 0.0087). Interestingly, women ≥38 expressed novel variants more frequently than their younger counterparts (p = 0.018), potentially indicative of an age-related decline in transcriptional fidelity. Limitations, reasons for caution It is possible that the use of controlled ovarian stimulation might alter the dynamics of transcription in cumulus cells. Therefore, it is possible that the spectrum of transcripts seen in this study might differ from those present in unstimulated cycles. Wider implications of the findings Expected associations between age, AMH levels and response to COS were observed. Skipping exon 6 may be associated with a superior response to COS, suggesting FSHR mRNA profiling might have a role in tailoring controlled ovarian stimulation protocols. Novel transcript variants were detected and seen most often in older women. Trial registration number not applicable
Read full abstract