RNA-mediated pathogenesis is emerging as an important disease

mechanism in unstable microsatellite disorders (Poulos et al., 2011). The most recent example is a potential role for rGGGGCC repeats in autosomal dominant frontotemporal dementia/amyotrophic lateral sclerosis (FTD/ALS) (DeJesus-Hernandez et al., 2011; Renton et al., 2011). While the toxic RNA model for DM is supported by considerable experimental evidence, recent studies have suggested that other factors might contribute to disease phenotypes (Sicot et al., 2011; Zu et al., 2011). Thus, it is important to discriminate between the relative effects of toxic RNAs and proteins in unstable microsatellite diseases. The MBNL loss-of-function MG-132 concentration model for DM allows this distinction because specific disease

manifestations, such as myotonia, are replicated in mouse models in the absence of microsatellite Bosutinib expansions (Poulos et al., 2011). Myotonic dystrophy is classified as a muscular dystrophy but the development and maintenance of normal brain function is also profoundly affected in this disease. Although DM symptoms are proposed to result from dysregulation of alternative splicing, the extent of missplicing induced by C(C)UGexp RNAs has been unclear particularly since previous studies have reported only a few missplicing events in the DM1 CNS (Jiang et al., 2004; Sergeant et al., 2001). Here, we tested the hypothesis that MBNL2 is an important splicing regulator during brain development and this function is compromised in DM. We generated Mbnl2 knockout mice and discovered that in contrast to Mbnl1, Mbnl2 is not an essential alternative splicing factor during skeletal muscle development. until However, Mbnl2 may play a compensatory role when Mbnl1 expression is compromised. The discordance between our results on the effect of Mbnl2 loss on skeletal muscle and a previous report using Mbnl2 gene traps may be attributable to differences in knockout strategy and the fact that prior studies did not evaluate alternative splicing in the CNS ( Hao

et al., 2008; Lin et al., 2006). While Mbnl2 knockout mice did not display pronounced muscle pathology, loss of Mbnl2 resulted in widespread splicing abnormalities in the brain. During this study, we uncovered a remarkable similarity between the control of alternative splicing during postnatal development by Mbnl1 in skeletal muscle and Mbnl2 in the brain. Both factors promote adult isoform expression and, similar to Mbnl1, Mbnl2 regulates the developmental splicing of hundreds of alternative cassette exons via the recognition of a YGCY motif in a manner reminiscent of the Nova, Rbfox, and PTBP splicing factor families, although the binding motifs for these factors are quite different ( Du et al., 2010; Li et al., 2007; Licatalosi and Darnell, 2010; Licatalosi et al., 2008; Witten and Ule, 2011; Zhang et al., 2008).