Spinal muscular atrophy (SMA) is a neurodegenerative disease characterized by progressive motor neuron loss and caused by mutations in SMN1 (Survival Motor Neuron 1). The disease severity inversely correlates with the copy number of SMN2, a duplicated gene that is nearly identical to SMN1. We have delineated a mechanism of transcriptional regulation in the SMN2 locus. A previously uncharacterized long noncoding RNA (lncRNA), SMN-antisense 1 (SMN-AS1), represses SMN2 expression by recruiting the Polycomb Repressive Complex 2 (PRC2) to its locus. Chemically modified oligonucleotides that disrupt the interaction between SMN-AS1 and PRC2 inhibit the recruitment of PRC2 and increase SMN2 expression in primary neuronal cultures. Our approach comprises a gene-up-regulation technology that leverages interactions between lncRNA and PRC2. Our data provide proof-of-concept that this technology can be used to treat disease caused by epigenetic silencing of specific loci.spinal muscular atrophy | lncRNA | PRC2 | SMN S pinal muscular atrophy is the leading genetic cause of infant mortality and is caused by deletions or mutation of Survival Motor Neuron 1 (SMN1) (1). Unique to humans, SMN1 is duplicated in the genome as SMN2, which is nearly identical in sequence. However, a C-to-T point mutation in exon 7 of SMN2 results in preferential skipping of this exon during pre-mRNA splicing and production of a truncated and unstable protein. A small fraction (10-20%) of pre-mRNA transcribed from SMN2 is spliced correctly to include exon 7 and produces a full-length SMN (SMN-FL, inclusive of exon 7) that is identical to the SMN1 gene product (2-4).Spinal motor neurons are highly sensitive to SMN1 deficiency, and their premature death causes motor function deficit in SMA patients (5, 6). The SMN2-derived SMN protein can extend spinal motor neuron survival, yet insufficient levels of SMN eventually lead to cell death. Overall, SMA patients with higher SMN2 genomic copy number have a less severe disease phenotype (7, 8). Type 0 or I patients, carrying one or two copies of SMN2, show onset of SMA within a few months of life with a life expectancy of less than 2. In contrast, type III and IV patients, carrying three or more copies, respectively, show juvenile or adult onset and slower disease progression (9). As further genetic evidence, SMA mouse models have been produced in which smn1 −/− mice, which would otherwise be embryonic lethal (10), can be rescued in the presence of high copy numbers of the human SMN transgene (11-13). Similar to the human disease spectrum, increased copy number of a human SMN transgene is inversely associated with decreased disease severity and mortality. We reasoned that increasing SMN2 transcription could phenocopy the beneficiary effect of SMN2 gene amplification and compensate for SMN1 deficiency. In addition, SMN1 heterozygotes are asymptomatic, whereas affected homozygotes have 10-20% of normal SMN levels. Therefore, we predict that modest SMN2 up-regulation will provide significant therapeutic benefit. H...

Mutations in the retinitis pigmentosa GTPase regulator (RPGR) gene cause severe X-linked retinitis pigmentosa (XLRP). More than 80% of the mutations are located in the terminal exon ORF15 of the RPGR gene. Genome editing, which represents a novel approach to treat monogenic disorders, is based on highly specific nucleases that cleave or nick at a chosen position within the complex genome followed by the repair of the double or single strand break (DSB or SSB) by endogenous repair mechanisms. The major pathways include error prone non-homologous end joining (NHEJ), microhomology-mediated end-joining (MMEJ), and homologous recombination (HR), the latter two with the help of a donor template. Currently, endonucleases for inducing the DSB are based on the CRISPR-Cas9 system or TALE proteins fused to the non-specific FokI nuclease (TALEN). However, specificity and toxicity of both endonuclease types raise concerns about their use for therapeutic in vivo applications. In order to study in vivo genome editing for the treatment of XLRP, our lab has generated a mouse model containing a point mutation in the ORF15 exon. In the present study, we characterize advanced variants of both endonuclease types (Cas9-FokI and TALE-MutH) for their activity and toxicity at the murine Rpgr-ORF15 locus for later usage in the mouse model. In total, ten sequences within or near the ORF15 exon have been targeted for the induction of DSB or SSB. Nine target sites for CRISPR/Cas9-FokI were chosen: three before, within, and behind the exon, respectively, and one target site for TALE-MutH within the exon. These sequences have been cloned into the traffic light reporter (TLR) gene expression system at the homing endonuclease I-SceI site. The TLR system has been modified to express either GFP in case of successful HR or BFP in case of NHEJ. Plasmids containing substrate, nucleases and template DNA were transfected into HEK293T cells. Efficiency of DNA modification was measured by FACS analysis and T7 surveyor assay, and toxicity was assessed by cell survival assay. In addition to the episomal TLR system within a human cell line (HEK293T), the genome of murine C2C12 cells was targeted by all endonuclease variants and toxicity was analysed via the T7 surveyor assay. Toxicity of Cas9-FokI and TALE-MutH are comparable to the golden standard ISceI while standard Cas9 nucleases showed slightly increased toxicity in HEK293T cells. Two different concentrations of the nucleases were used in a toxicity assay and were equally tolerated. Cas9-FokI showed preferences in its activity at the nine target sites with activities well above ISce-I level, while the one target site of TALE-MutH was as efficient as ISceI. Activity results were confirmed in the murine cell line C2C12. Off target toxicity in C2C12 cells was non-detectable. The characterization of the activity and toxicity of the tested endonucleases helped us to identify the most promising tailored nuclease and its target sequence in our gene targeting approach to treat XLRP. With the help of mouse reti...