Polymorphisms of IFN signaling genes and FOXP4 influence the severity of COVID-19

The pandemic of coronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has infected billions of people worldwide and posed an enormous threat to global public health and economies. The clinical manifestations of SARS-CoV-2 infection appear widely, ranging from asymptomatic infection, and mild upper respiratory tract illness to severe viral pneumonia with respiratory failure and even death [1]. Reported risk factors for severe COVID-19 include male sex [2], older age, and some comorbidities, such as chronic lung disease, cardiovascular disease, hypertension, diabetes, and obesity [3].

Additionally, host genetic variants have also been shown to modulate the risk of infection and disease severity. Large-scale genome-wide association studies (GWASs) in populations of European ancestry identified some genomic loci associated with COVID-19 severity and susceptibility [4‐6], including rs11385942 (LZTFL1), rs1886814 (near FOXP4), rs657152 (ABO), rs10735079 and rs10774671 (OAS gene cluster), rs74956615 (near TYK2), rs2109069 (DPP9), and rs2236757 (IFNAR2). Several genetic loci related to critical illness in COVID-19 belonged to interferon (IFN) signaling, which was further supported by a recent study that several loss-of-function variants in the IFN pathway were enriched in severe COVID-19 patients [7].

The interferon response functions as the major first line of defense against viruses, including SARS-CoV-2. IFNs interact with their receptors to activate downstream signaling cascades that eventually induce numerous IFN-stimulated genes with various antiviral activities, which contributes to effectively establishing an antiviral state in infected and surrounding cells [8]. Genetic variants in IFN-related genes can influence host antiviral response by affecting IFN production and serum levels. In addition, genetic variants in FOXP4 were associated with severe COVID-19 based on European. Of note, the allele and genotype frequencies of these single nucleotide polymorphisms (SNPs) vary among different populations due to the human genetic background. Recently, most of the associated variants biased towards European-ancestry samples and thus might not be generalizable to other non-European populations, especially for LZTFL1 rs11385942 [4] and TYK2 rs74956615 [6], which were monomorphic in the Asian population. To better understand the host mechanisms that lead to severe COVID-19 among the Guangdong population, we investigated some of the major candidate variants that have been identified as potential genetic factors based on European populations or meta-analyses with multiple populations, as well as common variants of IFN signaling genes that have been reported to be related to the severity of other viral infectious diseases.

Therefore, we selected twenty-two SNPs of IFN signaling genes and FOXP4. Our study aimed to investigate the association of IFN signaling genes and FOXP4 polymorphisms with the severity of COVID-19 in Guangdong population, and thus provide information for effective prevention and individual treatment strategies in the future.

The clinical manifestations of COVID-19 showed apparent heterogeneity, indicating that genetic factors may affect disease severity. In our study, we found that the MX1 rs17000900 CA + AA genotype tended to be correlated with a reduced risk of severe COVID-19 than the CC genotype in males. The AT haplotype consisting of MX1 rs17000900 and rs2071430 was likely to protect against COVID-19 adverse outcomes. Moreover, we observed that rs1886814 and rs2894439 variants of FOXP4 were associated with increased disease severity following SARS-CoV-2 infection. The CA haplotype comprised of risk alleles rs1886814 and rs2894439 was found to have a more significant association with worse clinical outcomes.

The IFN-stimulated gene MX1 encodes myxovirus resistance protein A (MxA), which belongs to the dynamin-like GTPase family and shows broad antiviral activity [11]. Previous studies have reported that MX1 polymorphisms could influence the antiviral and enzymatic activities, which can be expected to insight into the molecular mechanisms of inter-individual variabilities in susceptibility and severity of viral diseases [12]. In addition, a whole-genome sequencing study investigated the association between H7N9 infection and single-nucleotide variants in MX1, and observed that multiple MX1 rare variants increased susceptibility to the H7N9 influenza virus [13]. Among the validated variants in MX1, two promoter single-nucleotide polymorphisms (-123C > A, rs17000900; -88G > T, rs2071430) near the IFN-stimulated response element (ISRE) have been frequently reported the association with various viral diseases, including hepatitis C virus(HCV) [14], enterovirus 71(EV71) [15] and SARS-CoV [16‐18]. Furthermore, higher MX1 expression was associated with a better response to the influenza A H1N1 pandemic in 2009 [19]. MX1 was also a critical responder in SARS-CoV-2 infection. The MX1 expression level was higher in COVID-19 patients, and its expression reduced significantly with age [20], which supported increasing severity in older patients.

MX1 rs17000900 (-123C > A) is a promoter variant near the IFN-stimulated response element. A luciferase reporter assay demonstrated that the rs17000900 A allele contributed to increased promoter activity [14]. Zhang et al. [15]reported that the rs17000900 CA + AA genotype increased expression levels than the CC genotype. The -123A allele also provided stronger binding affinity to nuclear proteins than the wild-type allele [18]. Altogether, the rs17000900 A allele plays a more critical role in the regulation of MX1 antiviral response and then reducing the risk of disease severity. As previous studies reported [21], another five SNPs at the TMPRSS2/MX1 locus were correlated with a reduced risk of developing severe COVID-19 with the high level of MX1 expression in blood.

FOXP4 encodes a transcription factor associated with neurodevelopmental disorders and lung cancer [22]. Tian et al. [23] reported that FOXP4 was an important regulator of non-small cell lung cancer (NSCLC) and was significantly highly expressed in NSCLC cell lines and NSCLC patients. FOXP4 played a crucial role in regulating lung secretory epithelial cell fate and regeneration during lung development, and FOXP4 downregulation can impair epithelium regeneration in lung tissue [24, 25]. Thus, FOXP4 could protect the lung against pathogens by affecting the production of mucus. Furthermore, FOXP4 was also expressed in CD4 + and CD8 + T cells and necessary for memory T-cell cytokines recall responses to viral infection [26]. For COVID-19, SARS-CoV-2 cross-reactive T-cell immunity due to an exposure history to common cold coronaviruses [27] may affect disease severity.

The variant rs1886814 is located within FOXP4-AS1, a lncRNA gene that upregulates FOXP4 [28]. In the meantime, a recent colocalization analysis reported that this lung-specific expression quantitative trait loci (eQTL) signal for FOXP4 identified rs1886814 as the variant with the highest likelihood of causality, which is located ∼10 kb upstream of FOXP4 and is likely a regulatory variant associated with increased FOXP4 expression when A allele mutates to C [29]. The GWAS meta-analysis of COVID-19 Host Genetics Initiative (HGI) demonstrated that rs1886814 was associated with the severity of COVID-19 manifestation, which was increased odds of more severe COVID-19 phenotypes [5]. Ani Manichaikul et al. [30] conducted a GWAS of subclinical interstitial lung disease (ILD) in the population-based Multi-Ethnic Study of Atherosclerosis Study, which reported a novel FOXP4 region variant rs2894439 associated with emphysema and in LD with the lead variants rs1886814 (r² = 0.7) [29], suggesting a potential role of this gene in disease severity. These data suggested the minor, expression-increasing allele is associated with an increased risk of COVID-19 disease severity and interstitial lung disease. Interestingly, the risk alleles of these two variants are much more common in East Asians than in other 1 KG populations (rs1886814 MAF = 0.381 in East Asians and 0.043 in 1KGP Europeans, rs2894439 MAF = 0.430 in East Asians and 0.031 in 1KGP Europeans). Therefore, we hypothesized that rs1886814 A allele to C change and rs2894439 G to A change may regulate FOXP4 gene and protein expression, and then affect the production of mucus or memory T-cell recall responses to SARS-CoV-2. Wu et al. [31] identified a significant intronic variant in FOXP4 locus associated with severe COVID-19 (rs1853837, OR = 1.28, p = 2.51E-10, LD r² = 0.64 with rs1886814), supporting the validity of our result. The CA haplotype composed of rs1886814 and rs2894439 was found to have a more significant association with worse clinical outcomes, which was consistent with the single SNP analysis. This result further suggested that rs1886814 and rs2894439 played an important role in the severity of COVID-19.

Previous studies reported that virus-neutralizing antibodies were associated with COVID-19 severity [32, 33]. One possibility is that severe disease caused by hyperinflammation or uncontrolled viral replication induces overproduction of antibodies that serve as a “biomarker” of severity [33]. Thus, consistent with previous studies, our result suggests severity of SARS-CoV-2 infection significantly correlates with higher antibody levels. Interestingly, we observed a significant association between genotypes of rs1886814 and rs2894439 at the FOXP4 locus and the serum level of neutralizing antibodies. Further investigations are needed to understand how the FOXP4 gene affects antibody response and disease severity.

No haplotype of the OAS gene cluster was found to have an association with severe disease in our study. Related studies from Zhou [34] found that the protective alleles at both rs4767027-T (the OAS1 pQTL) and rs10774671-G (the OAS1 sQTL) are found on a Neanderthal haplotype. Banday [35] found that rs10774671-A and rs1131454-A formed a common haplotype, which decreased OAS1 expression and contributed to COVID-19 severity through allele-specific regulation of splicing and nonsense-mediated decay in European and African populations. The human genetic background influences the susceptibility to and the severity of infectious diseases. Further studies of larger samples in the Asian population should be carried out.

Sex-specific differences in clinical outcomes and immune response to SARS-CoV-2 were observed in our study. This is, at least in part, due to sex-based differences in innate and adaptive immune responses that are influenced by sex-related genes and sex hormones [36]. Sex hormones regulate immune-related gene expression by binding to the receptors expressed on cells of the immune system. Notably, estrogens can promote beneficial immune system activation, which may protect against severe COVID-19, but the effects of androgens on immune function are largely suppressive [37]. The X chromosome contains many immune-related genes, and incomplete X-inactivation can provide another immune advantage to females [38].

There are still certain limitations to our study. The sample size of subjects was relatively small, thus providing only limited power. However, both the case group and the control group in this study met the minimum sample size required for the significant sites at the level of statistical efficacy of 0.80. Meanwhile, the incidence rate of the outcome event was 20.38% in our study, which was observed much the same as in previous studies [39], a report of 72,314 cases in China with 14% severe cases and 5% critical cases. We believe the association could be revealed even with a modest sample size by the candidate gene approach. The Bonferroni correction was made considering the number of SNPs analyzed. That a more conservative analysis considering the number of tests carried out could lead to a type II error, considering the sample size. In addition, these observations were consistent with large genomic studies in diverse populations by HGI [5], as this is a replication study among the Guangdong population. Finally, the specific mechanism of how the variants regulate our immune system and thus contribute to diverse clinical presentations of COVID-19 remains unknown. Therefore, further studies with large participants are required to investigate such associations and explore potential mechanisms of how SNPs affect this.

In conclusion, our study found that the MX1 gene promoter variant tended to be associated with a reduced risk of developing severe COVID-19 in males, and the polymorphisms near FOXP4 were significantly associated with increased COVID-19 severity in the Guangdong population. Distinguishing high-risk patients who develop severe COVID-19 will provide information for early intervention and individual treatment strategies.