Evolutionary and in silico analysis of the antiviral TRIM22 gene

• Main Author: Kelly, Jenna
• Format: Text
• Language: English
• Published: Scholarship@Western 2014
• Subjects: Tripartite motif protein 22; host restriction factors; evolution; interferon; innate immune system; single nucleotide polymorphism; in silico analysis; Immunity; greenlandic; inuit
• Online Access: https://ir.lib.uwo.ca/etd/2327; https://ir.lib.uwo.ca/cgi/viewcontent.cgi?article=3790&context=etd

Tripartite motif protein 22 (TRIM22) is an evolutionarily ancient interferon-induced protein that been shown to potently inhibit human immunodeficiency virus (HIV), hepatitis B virus (HBV), and influenza A virus (IAV) replication. Altered TRIM22 expression levels have also been linked to autoimmune disease, cancer, and cellular proliferation. Despite its important role in a number of biological processes, the factors that influence TRIM22 expression and/or antiviral activity remain largely unknown. To identify key functional sites in TRIM22, we performed extensive evolutionary and in silico analyses on the TRIM22 coding region. These tools allowed us to pinpoint multiple sites in TRIM22 that have evolved under positive selection during mammalian evolution, including one site that coincides with the location of a common non-synonymous SNP (nsSNP) in the human TRIM22 gene (TRIM22 rs1063303:G>C). Remarkably, we found that the frequency of TRIM22 rs1063303:G>C varied considerably among different ethnic populations and African (AFR), American (AMR), and European (EUR) populations contained an excess of intermediate frequency TRIM22 rs1063303:G>C alleles when compared to a neutral model of evolution. The latter is typically indicative of balancing selection, a non-neutral selective process that maintains polymorphism in a population. Interestingly, we also found that the TRIM22 nsSNP rs1063303:G>C had an inverse impact on TRIM22 function. TRIM22 rs1063303:G>C increased TRIM22 expression levels, but decreased its anti-HIV activity and altered its subcellular localization pattern. In addition to these studies, we used a variety of in silico methods to prioritize and delineate other functional sites in TRIM22. We showed that the majority of positively selected sites in the C-terminal B30.2 domain of TRIM22 are located in one of four surface-exposed variable loops that are critical for the anti-HIV effects of the closely-related TRIM5α protein. Moreover, we used six different in silico nsSNP prediction programs to screen all of the nsSNPs in the TRIM22 gene and identified 14 high-risk nsSNPs that are predicted to be highly deleterious to TRIM22 function. Finally, to examine the TRIM22 nsSNP rs1063303:G>C in a more isolated population, we genotyped this nsSNP in two Inuit populations (Canadian and Greenlandic Inuit). We found that the TRIM22 rs1063303:C allele is inordinately prevalent in the Inuit compared to non-Inuit populations and that these two populations do not contain an excess of intermediate frequency TRIM22 rs1063303:G>C alleles compared to a neutral model of evolution, indicating that site TRIM22 rs1063303:G>C has not evolved under balancing selection in the Inuit. Lastly, we found an interesting association between the TRIM22 rs1063303:C allele and serum levels of triglycerides (TG) and high-density lipoprotein (HDL). Taken together, the results presented here identify a number of pertinent sites in the TRIM22 protein that likely influence its biological and/or antiviral functions.