br Results br Discussion We undertook this study

Results
Discussion We undertook this study in an attempt to unravel the epigenetic mechanisms involved in the silencing of the FMR1 gene in fragile X syndrome by the use of a fibroblast line carrying an unmethylated full mutation. There have been several attempts to study epigenetic silencing in fragile X syndrome. Eiges et al. (2007) have shown that FXS human embryonic stem estrogen receptors (hESCs) still express FMRP at a level similar to that in unaffected hESCs, whereas the FMRP level decreases as the hESCs were differentiated. Based on these results, it was expected that by reprogramming FXS fibroblasts into pluripotent stem cells, the hypermethylated state of the FMR1 promoter region would be reversed. However, by now several research groups have shown that iPSCs derived from FXS patients show epigenetic marks characteristic for heterochromatin similar to the full mutation fibroblasts they originated from (Urbach et al., 2010; Sheridan et al., 2011; Bar-Nur et al., 2012). These observations could be explained by the fact that the FXS iPSCs may not have all the characteristics of early pluripotency, but that they represent a later stage of human development (Urbach et al., 2010; Sheridan et al., 2011; Bar-Nur et al., 2012; Gafni et al., 2013). Another approach was used in studies with human fragile X lymphoblastic cells; here, a fully mutated and hypermethylated FMR1 gene was reactivated by treatment with 5-azadeoxycytidine, a hypomethylating agent. Although such treatment significantly reduced DNA methylation in some cells, it could not restore all remaining epigenetic marks to control levels (Chiurazzi et al., 1998, 1999; Coffee et al., 1999, 2002). Drugs such as 4-phenylbutyrate, sodium butyrate or trichostatin A, which block the activity of histone deacetylases, did not restore FMR1 expression to normal levels (Chiurazzi et al., 1999; Coffee et al., 1999, 2002; Tabolacci et al., 2005). In addition, treatment with a compound that reduces the in vitro expression of the FRAXA fragile site, acetyl-l-carnitine, did not restore the FMR1 expression either (Tabolacci et al., 2005). Recently, 5-azadeoxycytidine treatment was also tested on fragile X iPSCs, and it appeared to restore FMR1 expression in both iPSCs and differentiated neurons, which offers possibilities to use these cells as an epigenetic model (Bar-Nur et al., 2012). The availability of a fibroblast cell line carrying an unmethylated full mutation (uFM) provided a new opportunity to study the epigenetic silencing mechanisms in time. We first characterized the uFM fibroblast cell line together with a normal male fibroblast control line and a FXS fibroblast cell line carrying a fully methylated FMR1 promoter. Although increased FMR1 mRNA levels (up to five times) were reported in lymphoblastoid cells of premutation carriers (55∼200 unmethylated CGGs), our findings of normal to slightly increased FMR1 mRNA levels in the uFM fibroblasts are similar to the findings of Pietrobono et al. (2005), who examined a lymphoblastic cell line from the same individual. The lack of DNA methylation ensures that the chromatin is less densely packed and more accessible for transcription, which explains the FMR1 expression in this cell line. Our ChIP results differ from the original ChIP analysis of the uFM lymphoblastoid cell line (Tabolacci et al., 2005). We found a similar increase in H3K4 methylation; however, we did not find decreased H3 acetylation levels or intermediate H3K9 levels in the uFM fibroblasts. These differences could be explained by the fact that we have analyzed a distinct cell type (fibroblasts versus lymphoblastoid cells), and by differences in the ChIP protocol (e.g., quantification methods and reference genes used). Because the uFM fibroblast line lacked methylation of the FMR1 promoter site despite the high number of CGG repeats, we expected to find an unmethylated FMR1 promoter and normal levels of FMR1 mRNA after reprogramming into iPSCs. Surprisingly, we found the promoter region of FMR1 to be hypermethylated in all iPSC clones. Other epigenetic chromatin marks also indicated a repressed FMR1 promoter similar to the marks observed in the fragile X iPSC line. After differentiation of these iPSCs into neural progenitor cells, the FMR1 promoter remained methylated and thus silenced.