Currently several clinical studies addressing various aspects of FXS are underway (clinicaltrials.gov). Most of these studies are based on the 'mGluR' theory formulated in 2004 by Bear, Huber and Warren [
20]. However, FMRP is implicated in various aspects of activity dependent mRNA metabolism, and it has been found to interact with 4% of mRNAs in human fetal brain [
21,
22]. A therapy directed to downstream targets of FMRP is therefore likely to correct only some of several aspects of the FXS phenotype. Reactivation of the gene, on the other hand, would offer a possibility to correct all aspects. Studies in unaffected males with a full mutation who were found to have reduced FMRP levels indicate that the reactivation does not need to be complete [
10,
11]. Furthermore, the description of variable FXS phenotypes within one family due to variable degrees of methylation might indicate that the non-physiological methylation of the
FMR1 gene promoter is less stable than in physiologically methylated promoters, possibly facilitating reactivation without disturbing the methylation of other promoters [
23]. Using 5-azadC it has been shown that it is possible to reactivate the
FMR1 gene by demethylation of the promoter [
13]. We aimed to find a drug that reactivates the
FMR1 gene but with less toxicity than 5-azadC. DNA methylation is mediated by CpG methyltransferases (DNMT1) using S-adenosylmethionine (SAM) as a methyl group donor. Folate in the form of 5-methyltetrahydrofolate is essential for the recycling of SAM after the methylation reaction. Folate deficient diet has been found to cause reduced DNA methylation in humans [
24,
25]. Methotrexate (MTX) is a folate antagonist that acts by inhibiting dihydrofolate reductase (DHFR), an enzyme that catalyzes the conversion of dihydrofolate to tetrahydrofolate. MTX is used in the long-term treatment of rheumatoid arthritis in children and adults. It has been described that treatment with MTX decreases cellular methylation [
16‐
18]. We therefore decided to study MTX as a potential drug for the reactivation of the
FMR1 gene. Firstly we analyzed
FMR1 mRNA expression in fibroblasts from a male FXS patient and from a healthy control. As expected, we did not detect
FMR1 mRNA in the FXS patients’ fibroblasts. Next, we treated the fibroblasts with MTX and used 5-azadC treatment as a positive control. We found that not only treatment with 5-azadC but also treatment with MTX resulted in detectable
FMR1 mRNA levels. The effect of the MTX treatment was positively correlated with the time of treatment and the dosage of MTX. The results that we obtained from the 5-azadC treatment on
FMR1 mRNA levels were comparable to those reported by Chiurazzi
et al. [
13]. However in contrast to them, when we used western blotting we were not able to detect any FMRP, either in the 5-azadC-treated cells or in the MTX-treated cells. Most likely, the protein levels were too low for detection. It has been described in a lymphoblastoid cell line as well as in a mouse model with an unmethylated full mutation that the CGG repeat itself causes mRNA translation to be 40% less efficient [
14,
26]. The FMRP levels might therefore be very low. On the other hand, our experiments might understate the effects of MTX
in vivo because cells grown in culture have been shown to have an excess of
de novo methylation of CpG islands [
27,
28]. In fact, we found that our reactivation experiments were less efficient after the fibroblasts had undergone more cell cycles (data not shown).
MTX affects cell metabolism in many ways. It reduces the synthesis of thymidylates and purines as well as the conversion of homocysteine to methionine. MTX treatment thereby leads to impaired DNA and protein synthesis as well as reduced methylation of DNA and proteins. Consequently, at this point of time, we cannot prove that the expression of FMR1 found after MTX treatment is caused by reduced DNA methylation. In fact DNA methylation after treatment with MTX was not altered. Additionally, the lack of FMR1 protein expression might be caused by the overall reduced protein synthesis. Further experiments directed at understanding the mode of action of MTX in the FXS fibroblasts might also help us to improve the outcome of MTX treatment.
Unfortunately, no mouse model is available for
in vivo testing because even in mice carrying very long CGG repeats no abnormal methylation occurs [
26]. Another opportunity to study the effect of MTX in FXS would be an individual with the comorbidity of FXS and rheumatoid arthritis receiving MTX. Studying
FMR1 mRNA and FMRP expression in lymphocytes of such an individual would help to clarify the potential of MTX in FXS. An inquiry was sent to the German FXS support organization (
http://www.frax.de), but no individual in Germany with such comorbidity is currently known. However, due to the high prevalence of FXS and rheumatoid arthritis, it is likely that such an individual exists in another country.