Population genetics of IFITM3 in Portugal and Central Africa reveals a potential modifier of influenza severity

Abstract

Influenza epidemics are a serious global public health and economic problem. The IFITM3 allele (rs12252-C) was suggested as a population-based genetic risk factor for severe influenza virus infection by A(H1N1)pdm09. We analyzed the population genetics of IFITM3 variants in the Portuguese general population (n = 200) and Central Africans (largely Angolan) (n = 148) as well as its association to influenza severity in Portuguese patients (n = 41). Seven SNPs, within the 352 bp IFITM3 amplicon around rs12252, were identified. SNP distributions in the Portuguese appeared at an intermediate level between the Africans and other Europeans. According to HapMap, rs34481144 belongs to the same linkage disequilibrium (LD) block as rs12252 and is in strong LD with rs6421983. A negative association with severe relative to mild disease was observed for allele rs34481144-A, indicating a protective effect under the dominant model. Moreover, haplotype Hap4 with rs34481144-A, not including rs12252-C, was significantly associated to mild influenza. Conversely, although with borderline significance, haplotype Hap1 with rs34481144-G, not including rs12252-C, was associated to severe disease. Moreover, in comparison to the general Portuguese population, statistical significant differences in the frequencies of the protective allele rs34481144-A in the severe disease group, the deleterious Hap1 in the mild disease group, and the protective Hap4 in the severe disease group were observed. The population attributable risk (PAR) for the targeted rs34481144 allele or genotype was of 55.91 and 64.44% in the general population and the mildly infected individuals, respectively. Implication of these variants in disease phenotype needs further validation, namely through functional analysis as is discussed.

Antiviral Protection by IFITM3 In Vivo

“…Several follow-up studies examining Europeans and other populations in which rs12252-C is rare have failed to find an association between the SNP and severe influenza virus infection [ 57 – 62 ], although some studies support a link between the SNP and mild influenza [ 56 , 57 ]. In contrast, a strong overrepresentation of homozygous rs12252-C genotypes in severe influenza patients has been repeatedly observed in studies examining Chinese individuals in which the allele frequency for this SNP is greater than 50% in the general population [ 52 – 55 ].…”

Association between rs12252 and influenza susceptibility and severity: an updated meta-analysis

“…We also made a comparison between severe and mild influenza patients, and for this analysis, we used the mild group as the control group. A total of five studies [ 18 , 19 , 22 , 25 ] were included in this analysis, which included 312 patients with severe influenza and 468 patients with mild influenza. Significant difference was shown in CC vs. TT (OR = 3.11, 95% CI (1.58–6.13), P = 0.001) and CC vs. CT + TT models (OR = 3.90, 95% CI (2.44–6.21), P < 0.00001).…”

“…[ 18 ] and David et al . [ 25 ], while Randolph et al . [ 20 ] found rs12252 was not associated with susceptibility to paediatric influenza infection.…”

“…(1) Only nine English studies were included in our meta-analysis. For instance, although we found some eligible studies based on the title and abstract, we had to exclude a study because it was published in Russian; (2) most of the included studies were on Asian and Caucasian populations, while only one study included Portuguese [ 25 ] and one studied African [ 20 ] populations. More studies in different ethnic groups are needed; (3) the definitions of severe influenza in the eligible studies are not exactly the same, which may increase the heterogeneity of our meta-analysis; (4) different kinds of influenza are analysed together which may contribute to the heterogeneity; (5) some studies were multi-arm studies, in subgroup meta-analysis was classified according to degree of patients symptoms; however, the data of control group were repeatedly used, which would enlarge the sample size by mistake and increase the risk of type I error.…”

Human IFITM3 restricts Chikungunya virus and Mayaro virus infection and is susceptible to virus-mediated counteraction

“…The potential importance of human IFITM3 protein as antiviral factor has been addressed in clinical observation studies of Influenza A. Specifically, the SNP rs12252-C allele might associate with increased IAV mortality and morbidity (Everitt, Clare et al, 2012, Pan, Yang et al, 2017, Williams et al, 2014, although other cohorts failed to display this association (Carter, Hebbring et al, 2018, David, Correia et al, 2018, Lopez-Rodriguez, Herrera-Ramos et al, 2016, Mills, Rautanen et al, 2014, Randolph, Yip et al, 2017. A debated mechanistic working model based on the idea that the genetic variation induces alteration of a splice acceptor site, resulting in expression of a truncated IFITM3 protein which lost its antiviral activity (Everitt et al, 2012).…”