This study is the first to report significantly raised levels of mRNA in a sample of males carrying intermediate size, or grey zone, FMR1 alleles, compared with the carriers of normal alleles. This finding deals with the controversy concerning the functioning of the slightly expanded FMR1 alleles by showing that transcription is indeed disturbed in a significant proportion of these alleles. Earlier studies have shown that the raised mRNA levels reflect increased transcriptional activity rather than an increased mRNA stability.8,19
Using the transcription inhibitor actinomycin D, it was shown that the elevation of the mRNA transcript cannot be explained by its increased stability.8
Moreover, direct evidence for an increased transcription rate has recently been obtained using a nuclear run‐on experiment.25
We also show that there is a relationship of these levels with the size of CGG repeat in the grey zone class, and that the sharp increase in mRNA levels corresponding to the number of CGG repeats seen in the grey zone range is significantly reduced in the premutation range. Our data show a statistically significant trend for mRNA level to be a linear function of the repeat size in the grey zone sample, but there is also a considerable variation in this sample, and an overlap with the normal range. The onset of the transcription increase indicated by our data is at approximately 39 repeats; however, more observations, especially at the upper end of the “normal” alleles, are required to estimate this threshold more precisely, and we do not recommend, at this stage, its use as the sole criterion of classification of small expansion alleles.
Our results confirm the findings from two earlier studies,6,19
but are discrepant with those from the study of Allen et al
in which no significant relationship was detected between mRNA and CGG repeat number in grey zone alleles of identical range. The possible reasons for this discrepancy include a smaller grey zone sample of 21 males of unspecified composition (compared with 33 individuals of European descent in this study) and/or different procedures used to assess mRNA levels. The procedures involved in running each individual test in the present study ensured high reliability and minimised the inherent variation. It is less likely, however, that the discrepancy reflects difference in an ascertainment between the two studies. Although more than half of the individuals in our grey zone sample were identified in a sample pre‐selected for SEN, mRNA level, and not the behavioural phenotype, was the primary trait in this study. Furthermore, we found no correlation between any cognitive/behavioural scores and either mRNA or CGG repeat size in this sample (own unpublished data).
Based on the discovery of late‐onset tremor/ataxia in some males carrying a premutation gene and showing elevation of mRNA, a toxic “gain‐of‐function” effect of excessive FMR1 mRNA on the brain was considered.12,26,27
. It seems that the POF occurring in a significant proportion of female premutation carriers may also be attributed to a similar effect on ovarian cells.10
If indeed raised levels of FMR1 mRNA cause late‐onset abnormal conditions in premutation carriers, we can speculate that the raised levels observed in carriers of grey zone alleles, even if more modest, could have a similar effect. Such an assumption is supported by a significant association recently reported, in population‐based studies, between POF and grey zone alleles.14,15
Taken together with the present results, it seems that screening studies of both POF and late‐onset neurodegenerative conditions with tremor/ataxia should consider the intermediate, as well as premutation, FMR1 allele sizes. The effect of grey zone on neurodevelopment is, however, still controversial, with some studies showing a noticeable increase of grey zone carriers among SEN children,18,28
but others failing to do so.13,16,29,30
Our results also deal with the ambiguity currently associated with the definition of grey zone alleles. The threshold for the onset of a progressive increase in FMR1 transcription was determined to be approximately 39 CGGs. Although more data are required to narrow the error margins, it seems that the lower bound for the grey zone currently recommended at 453
may have to be revisited. As for the upper bound, our data suggest a continuous scale of involvement, rather than the clear‐cut separation of grey zone from premutation that has been recommended.3
The present data also showed that the rate of increase in mRNA levels relative to the number of CGGs tends to slow in the premutation range, an observation that appears at variance with the hypothesis that slight FMRP deficit in premutation5,6
leads to the transcriptional enhancement.7,8
However, the results from a more recent study indicate that this enhancement may result from a direct (cis
) effect of the CGG element.19
In conclusion, our results, in combination with some phenotypic effects of grey zone alleles reported earlier, on the one hand, and the transmission instability of these alleles,31
on the other, suggest that premutation–grey zone differences are quantitative rather than qualitative.
It is recommended, however, that our findings be verified on a larger, population‐based, sample of grey zone carriers, and that future studies cover the whole range of both grey zone and premutation alleles, to allow clarification of the demarcation lines within a broad category of small expansions.