Skip to main content

Conference Program

Subpage Hero

Loading

Splice rescue of protein-truncating variants is associated with reduced selection pressure and attenuated disease phenotype

Clinical Genetics and Therapeutics
  • Primary Categories:
    • Genomic Medicine
  • Secondary Categories:
    • Genomic Medicine
Introduction:
Human genetic variants that alter the canonical splicing sequence could result in various splicing patterns, which is challenging for variant interpretation. It will be more complex when a coding variant occurs near the canonical splice site, and thereby the protein product is possibly influenced by both the mutated codon and the altered splicing pattern. Particularly, if the altered splicing results in an in-frame splice-out of the coding variant, the deleterious effect of the variant on the protein product could presumably be ‘rescued’. However, the prevalence of this event in the human genome and the extent to which it influences the population genetic structure and the manifestation of hereditary diseases are little-known.

Methods:
We selected 8,747,281 variants (including 145,791 nonsense variants, 200,642 frameshift variants, 5,624,824 missense variants, and 2,776,024 synonymous variants) in the gnomAD database and analyzed the splicing consequence using SpliceAI (cutoff=0.5). If the predicted spliced-out base number in a variant was a multiple of 3, the splicing consequence was “inframe splicing” and thus the possible deleterious effect caused by the variant was supposed to be “rescued”. Vice versa, if the spliced-out base number was not a multiple of 3, the splicing consequence was “not-inframe splicing” and the possible deleterious effect caused by the variant was supposed to be “not-rescued”. We also analyzed all LoF variants (n=116,859) in the ClinVar dataset and chose a nonsense variant in APC (NM_000038.6:c.288T>A) as an example to validate the rescue effect with minigene assay.

Results:
In loss-of-function (LoF) variants, compared with the variants that were predicted to be not-rescued, those which were predicted to be rescued had higher population maximum allele frequency (p=0.043) and lower proportion of singleton (p<0.001), corresponding to reduced selection pressure. In ClinVar dataset, the nonsense variant in APC (NM_000038.6:c.288T>A) was previously reported to cause attenuated familial adenomatous polyposis (AFAP). Minigene assay for this variant showed the existence of in-frame (69bp) splicing products, where the original deleterious nonsense variant was spliced out as predicted. This may explain the “attenuated” disease phenotype. 

Conclusion:
In summary, we found that the splice-out of putative null variants is possibly associated with reduced selection pressure and attenuated disease phenotype, providing new insights into the human genetic architecture and clinical variant interpretation.

Agenda

Sponsors