Embedded Files
intronic splicing silencers (ISSs), which facilitate exon 7 exclusion in SMN2 (reviewed in 160). www.annualreviews.org • 25 Years of Spinal Muscular Atrophy Research 237 Annu. Rev. Genom. Hum. Genet. 2020.21:231-261. Downloaded from www.annualreviews.org Access provided by 2600:1700:3c30:a3d0:89dc:b3d1:ac73:5085 on 01/23/22. For personal use only. Exon 7 of both SMN genes contains a centrally placed ESE with a GA-rich region, and mutations in this ESE abolish exon 7 inclusion in both genes. The main splicing factor that recognizes this ESE is SFRS10 (hTRA2-β1) (69, 70). In addition, the splicing factors SRSF9 (SRp30c), heterogeneous nuclear ribonucleoprotein (hnRNP) G, hnRNP M, TDP43, and PSF promote exon 7 inclusion via direct or indirect binding to this ESE (160 and references therein). This network of splicing factors binding to this particular ESE is likely responsible for the ∼10% full-length mRNA generated by SMN2. Overexpression of these splicing factors, either separately or in combination, restores the splicing capacity of SMN2 minigenes up to 80% and increases endogenous SMN protein levels (69, 70). Complete depletion of murine Sfrs10 had almost no effect on exon 7 inclusion, highlighting the complexity of the in vivo splicing (112). Histone deacetylase inhibitors, such as valproic acid, increase SFRS10 (hTRA2-β1) (21) and enhance SMN2 exon 7 inclusion and thus the amount of SMN protein. These findings sparked the groundbreaking idea to modulate exon 7 splicing in vivo as a strategy to treat SMA (69). Similarly, tailed 5 GGA antisense oligonucleotides (ASOs) facilitate recruitment of SFRS10 to the tail, markedly increasing exon 7 inclusion and SMN levels (161). The most important difference between the two SMN genes is the C6U substitution (c.840C>T) in exon 7. Twenty years ago, in a time when silent mutations were mostly underestimated, Wirth’s and Androphy’s groups showed that SMA is caused by a silent mutation that disrupts an ESE, causing exon 7 skipping in SMN2 (106). This seminal finding opened the door to therapy based on restoring SMN2 splicing. Next, Krainer’s group found that this C6U substitution disrupts a heptamer motif of ESE, CAGACAA, recognized by the SR-rich splicing factor SRSF1 (ASF/SF2) (25). By contrast, Manley’s group showed that the C6U substitution creates a new exonic splicing silencer, which is bound by the splicing repressor hnRNA A1 (89). The latter, together with an ISS localized in intron 6 from −112 to −68 base pairs (bp) (element 1) and intron 7 from +10 to +24 bp (ISS-N1), facilitates exon 7 skipping (124, 158). Among all cisregulatory domains, the discovery of ISS-N1 (158) and its targeting by specific ASOs (79) opened a completely new therapeutic era for the treatment of SMA individuals. SMN Protein and Cellular Function SMN protein is essential for every cell and species. The depletion of SMN is early embryonic lethal (155). Since only humans have two SMN copies, SMA occurs naturally only in humans; all other species developing SMA are genetically engineered by introducing human SMN2 copies, SMN2 cDNA, or mutations that promote exon 7 skipping or decrease SMN functionality (19, 75, 125, 126). SMN participates in different protein complexes involved in small nuclear ribonucleoprotein (snRNP) biogenesis, translation, transcription, microRNA metabolism, stress granule formation, cell survival, ubiquitin homeostasis, DNA damage response, actin cytoskeleton dynamics, endocytosis, vesicular transport, vesicle trafficking along neurons, and energy homeostasis (extensively reviewed in 72, 159, 168, and references therein). Ribonucleoprotein Function and Splicing The canonical function of SMN is snRNP biogenesis and splicing (103, 104, 135). SMN depletion mainly affects the U12 minor spliceosome (57). Consequently, an accumulation of U12-dependent intron retention transcripts has been observed, but only a few were reproducibly confirmed, such as TMEM41B (Stasimon) (42, 107). Nonetheless, the restoration of TMEM41B expression level failed to rescue the SMA phenotype in mice (171). Other significantly misspliced transcripts impair calcium homeostasis and voltage-gated calcium channel clustering, which
Report abuse