GENETIC VARIANT INTERPRETATION OVER PROTEIN CHANGES MASKS SPLICING DEFECTS IN THE SODIUM IODIDE SYMPORTER-CODING PRE-MESSENGER RNA

Curitiba

Congreso; XIX Latin American Thyroid Congress; 2023

Sociedad Latinoamericana de Tiroides

Introduction: The interpretation of disease-causing genetic variants frequently ignores that either exonic or intronic variants canpotentially lead to mRNA splicing defect. Exonic variants ? including synonyms, missense or nonsense variants based on protein-codingsequence ? can affect or create cryptic motifs recognized by the splicing machinery. The prediction of splicing defects is particularlyrelevant because, if present, a single nucleotide substitution can lead to a very significant effect on the protein, and they can be maskedas synonymous or small effect variants if only interpreted at the protein level. Objective: To investigate the effect of congenitalhypothyroidism-causing SLC5A5 gene variant on sodium iodide symporter (NIS) pre-mRNA splicing. Methods: Bibliographiccompilation of congenital hypothyroidism-causing SLC5A5 gene variants. Bioinformatic prediction of the effect of variants oncentral and auxiliary elements involved in NIS pre-mRNA splicing. Results: Central elements involved in NIS pre-mRNA splicing,including donor and acceptor splicing sites, branch points, and poly-pyrimidine tracts) were identified. The variants c.970-3C>Aand c.1315_1329del were predicted to affect donor and acceptor splicing sites, respectively. Complementary, the variants c.572A>Gand c.749G>T, and the variants c.809T>A, c.859_864del, c.1157T>G and c.1593C>G were predicted to generate cryptic donorand acceptor splicing sites, respectively. In line, the variant c.1593C>G was reported to cause a deleterious NIS pre-mRNA splicingdefect due to a cryptic acceptor site. Moreover, the variants c.816C>A, c.1106A>T, c.1628G>A, and c.1679C>T were predicted toaffect auxiliary elements involved in NIS-pre-mRNA splicing. In line with high performance functional splicing assays, bioinformaticpredictions indicated that the variants c.371G>A and c.1183G>A do not affect NIS pre-mRNA splicing. Conclusions: Although mostdisease-causing exonic variants causes changes in the protein-coding sequence, approximately 30% of disease-causing variants lead topre-mRNA splicing defects. The present study predicted that 36% of congenital hypothyroidism-causing SLC5A5 gene variants, manyof them experimentally tested at the protein level, are predicted to be deleterious for NIS pre-mRNA splicing. The results highlightthe importance to assess the pathogenicity of novel variants on pre-mRNA splicing, before consider a potential effect on the protein.