The SMN1 common variant c.22 dupA in Chinese patients causes spinal muscular atrophy by nonsense-mediated mRNA decay in humans

Spinal muscular atrophy (SMA) is a common autosomal recessive neuromuscular disorder that is mostly caused by homozygous deletion of the SMN1 gene. Approximately 5%-10% of SMA patients are believed to have SMN1 variants. c.22 dupA (p.Ser8lysfs*23) has been identified as the most frequent variant in the Chinese SMA population and to be associated with a severe phenotype. However, the exact molecular mechanism of the variant on the pathogenesis of SMA is unclear. We observed that SMN1 mRNA and the SMN protein in the peripheral blood cells of a patient with c.22 dupA were lower than those of controls. The aim of this study is to investigate whether nonsense-mediated mRNA decay (NMD) plays a role in the mechanism of the c.22 dupA variant of the SMN1 gene as it causes SMA. Two lymphoblasts cell lines from two patients (patient 1 and 2) with the c.22 dupA, and one dermal fibroblasts cell line from patient 2 were included in our study. Two-stage validation of the NMD mechanism was supplied. We first measured the changes in the transcript levels of the SMN1 gene by real-time quantitative PCR after immortalized B-lymphoblasts and dermal fibroblasts cells of the SMA patients were treated with inhibitors of the NMD pathway, including puromycin and cyclohemide. Next, lentivirus-mediated knockdown of the key NMD factor-Up-frameshift protein 1 (UPF1)-was performed in the fibroblasts cell line to further clarify whether the variant led to NMD, as UPF1 recognizes abnormally terminated transcripts as NMD substrates during translation. SC35 1.7-kb transcripts, a physiological NMD substrate was determined to be a NMD positive gene in our experiments. The two inhibitors resulted in a dramatic escalation of the levels of the full-length SMN1 (fl-SMN1) transcripts. Additionally, the SC35 1.7-kb mRNA levels were also increased, suggesting that NMD pathway is suppressed by the two inhibitors. For the 3 cell lines, the fold increase of the SMN1 transcript levels of cycloheximide ranged from 2.5 ± 0.4 to 8.3 ± 0.1, 1.9 ± 0.2 to 5.0 ± 0.7 and 2.2 ± 0.1 to 4.9 ± 0.2 for two lymphoblastoid cell lines and one fibroblasts cell line, respectively. For these cell lines, the fold increases of the SMN1 transcript levels of puromycin were as follows: 5.5 ± 0.2 to 19.5 ± 4.0, 3.1 ± 0.3 to 9.9 ± 1.8 and 1.5 ± 0.2 to 6.5 ± 0.5. Meanwhile, the SC35 1.7-kb transcript levels were markedly increased in all 3 cell lines. In addition, lentivirus-mediated UPF1 knockdown lead to a reduction of the UPF1 protein level to 22.5% compared to the negative control lentivirus. Additionally, knockdown of the UPF1 gene also promoted mRNA expression of the SC35 1.7kb and fl-SMN1 genes. The increases of the SMN1 and SC35 1.7-kb mRNA levels reached about 4- and 6.5-fold in fibroblasts derived from the patient 2, respectively. Altogether, our study provides the first evidence that the c.22 dupA variant in the SMN1 gene triggers NMD. SMA pathogenesis in the patient is associated with mRNA degradation of SMN1, but not the truncated SMN protein.