MMS19 and IFIH1 Host Genetic Variants Associate with SARS-CoV-2 Infection in Elderly Residents of Long-Term Care Facilities

Abstract

The coronavirus disease 2019 (COVID-19) pandemic has significantly affected older adults. Identifying host COVID-19 susceptibility genes in elderly populations remains a challenge. Here, we aimed to identify host genetic factors influencing the susceptibility to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. We genotyped 12 single-nucleotide polymorphisms (SNPs) previously associated with the innate immune response in a total of 97 elderly (age > 65 years) residents of three long-term care facilities located in Barcelona, Spain. Individuals were PCR-tested during the SARS-CoV-2 outbreaks between September and November 2020. SARS-CoV-2 PCR tests revealed infections in 81 residents. Importantly, the 16 uninfected residents remained SARS-CoV-2 seronegative until vaccination (January and February 2021). After adjusting for sex and age, we found that two SNPs were significantly associated with SARS-CoV-2 infection susceptibility—MMS19 nucleotide excision repair protein homolog (MMS19)/rs2236575 (p = 0.029) and interferon-induced helicase C domain-containing 1 (IFIH1)/rs1990760 (p = 0.034). No association with SARS-CoV-2 infection was found for 10 additional genotyped SNPs, which included 4 SNPs on chromosome 12 in the gene encoding oligoadenylate synthetase (OAS). Our results indicate that MMS19/rs2236575_A and IFIH1/rs1990760_TC genetic variants were associated with a resistance to SARS-CoV-2 infection in a cohort of institutionalized seniors.

1. Introduction

The elderly population is particularly vulnerable to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection and severe coronavirus disease 2019 (COVID-19) [1]. Specifically, residents of long-term care facilities (LTCFs) are at a higher risk of SARS-CoV-2 transmission and infection [2,3]. LTCF population density may account for this increased risk of transmission. Described risk factors for infection and disease severity are sex, age, ethnicity, obesity, and cardiovascular and respiratory diseases. Current vaccines have successfully reduced both SARS-CoV-2 transmission and COVID-19 severity [4,5]; however, the development of new tools to predict virus transmission and disease burden may improve current and future treatments as well as disease prevention strategies. Several large-scale genetic association studies have consistently identified human host single-nucleotide polymorphisms (SNPs) that are implicated in the biology and epidemiology of COVID-19 [6,7,8,9,10,11]. These studies suggest different mechanisms to explain why some individuals are more prone to infection by SARS-CoV-2 or being more severely affected by the disease upon infection [12,13]. Host genetics can inform us about disease etiology and provide new means for patient management. Population-specific risk variants have also been reported [8,14,15].

Chronic inflammatory diseases are more prevalent in older people. Ageing is associated with an immunosenescent phenotype characterized by a progressively proinflammatory state and a reduced immune response against pathogens [16]. Accumulation of damaged DNA as a consequence of ageing is the principal cause of age-related diseases. MMS19 is a component of the iron–sulfur protein assembly machinery involved in DNA metabolism and DNA repair, and its de-regulation has been associated with some particular cancers [17]. Adenosine deaminase acting on RNA 1 (ADAR1) has RNA editing functions and is responsible for activating the intracellular sensors, RIG-I and IFIH1 (MDA5), that sense viral replication, leading to the activation of several genes involved in the innate immune response (IFNAR2, IFNL, OAS) and inflammation [18]. SNP polymorphisms in ADAR-1, IFIH1, and PNPLA3 are involved in liver inflammation [19]. Furthermore, several host proteins involved in immune innate signaling pathways, such as MDA-5/IFIH1, TRIF, IFNAR2, IFITM3, IFN L3/4, are associated with severe COVID-19 [20]. Recently, a haplotype inherited from Neanderthals comprising SNPs in three isoforms of the OAS gene (OAS-1, OAS-2, OAS-3) were found to protect against SARS-CoV-2 infection [9]. Overall, these studies highlight the relevance of SNPs associated with innate signaling with susceptibility to viral infections, but do not address the effect of age. In this study, we aimed to identify host genetic factors influencing susceptibility to SARS-CoV-2 infection in LTCF residents. Due to the extreme importance of age as a risk factor for severe COVID-19, we considered it in our genetic analyses [12]. Our study included uninfected and infected individuals with SARS-CoV-2 that were recruited early in the pandemic before the initiation of vaccination.

2. Materials and Methods

2.1. Study Participants

A total of 97 residents of three LTCFs located in Barcelona, Spain, were randomly included in this prospective observational study (CoronAVI@S study). Plasma samples were collected before vaccination from September to November 2020 [12]. SARS-CoV-2 serology was performed to sub-classify participants into infected and uninfected individuals depending on their PCR and serology status.

2.2. Ethics

The CoronAVI@S study was approved by the Ethics Board of the Institut Universitari d’Investigació en Atenció Primària Jordi Gol (20-116P). All participants provided written informed consent before inclusion.

2.3. Single-Nucleotide Polymorphism Genotyping

Genomic DNA was extracted from peripheral blood mononuclear cells using Quick Extract DNA Extraction 1.0 solution (Epicentre/Lucigen) as previously described [19]. Two microliters of extracted DNA were used for SNP genotyping (Applied Biosystems, Waltham, MA, USA). The reference numbers of the TaqMan assays used for genotyping on a Quantstudio 5.0 from Applied Biosystems were ADAR1/rs1127313, (C_8724398_10); IFIH1/rs1990760, (C_2780299_30); MMS19/rs2236575 (C_1797584_30); OAS-1/rs4767027 (C_28015278_10); OAS-1/rs10774671 (C_2567433_10); OAS-2/rs1293767 (C_8920276_10); OAS-3/rs10735079 (C_31831768_10); IFNL3/rs12979860 (C_7820464_10); IFNAR2/rs2236757 (C_11354003_30); and PNPLA3/rs738409 (C_7241_10). The SNPs, IFNL3/rs4803217 and TNFAIP8L1/rs1060555, were determined using IDT DNA assay Hs.GT.rs4803217.A.1 and Hs.GT.rs1060555.G.1, respectively, and also assayed by Quantstudio 5.0 (Applied Biosystems).

2.4. Statistical Analysis

Haplotype estimation and association with patient susceptibility to SARS-CoV-2 infection were achieved using Haploview 4.2 (Broad Institute of Harvard and MIT, Cambridge, MA, USA) and SNPSTATS [21]. Dominant, overdominant, recessive, and log-additive inheritance models were fitted using the SNPSTATS online program [21], reporting p values for the SNP association of the inheritance model with the smallest AIC. To enhance the robustness of our findings, we performed a bootstrap resampling with 100 interactions and calculated empirical p-values as the proportion of times the observed p-value was greater than the p-value derived from the bootstrap resampling. Risk factors (sex and age) were included in the genetic analysis to adjust the association of each SNP to infection susceptibility. Hardy–Weinberg equilibrium, minor allele frequency (MAF), and SNP linkage disequilibrium (LD) were also determined by Haploview 4.2. Correlation between the variables was assessed using nonparametric Spearman’s correlation. Mann– Whitney U and chi-squared tests to compare the infected and uninfected individuals were performed using GraphPad Prism 9.0 (GraphPad Software, San Diego, CA, USA).

3. Results

Study participants were tested by PCR during the SARS-CoV-2 outbreaks at their respective LTCFs between September and November 2020. During this period, 81 of the participants tested positive for SARS-CoV-2 (80% female, mean age 87 years) and 16 tested negative (

Table 1. Clinical and biochemical characteristics of participants.

	SARS-CoV-2 Infected	SARS-CoV-2 Uninfected	p-Value
N (%)	81 (83.5)	16 (16.5)
Age, years	87 (81–90)	80 (74–91)	0.1668
Female	65 (80.2)	8 (50)	0.0104 2
AGM level 1			0.1460 2
1	0	0
2	4	3
3	38	6
4	39	7
Albumin (g/L)	37.55 (35.9–39.75)	38.2 (36.93–39.5)	0.2725
Alanine aminotransferase (U/L)	11 (8.5–14)	12.5 (9.25–17.75)	0.3694
Aspartate aminotransferase (U/L)	17 (14–20)	18 (15.25–21.25)	0.3431
Creatine kinase (U/L)	42.5 (33.25–71)	47 (34.5–73.25)	0.7907
High-density lipoprotein (mg/dL)	46.2 (40.73–56.53)	42.4 (36–47.8)	0.1135
Total cholesterol (mg/dL)	193 (162–228)	198.5 (131.8–230.8)	0.6147
Creatinine mg/dL	0.81 (0.675–1.035)	0.82 (0.6925–1.25)	0.8186
Alkaline phosphatase (U/L)	78 (66.25–104)	89 (72.25–112.8)	0.2704
Ferritin (ng/mL)	66.5 (34–144.3)	108 (50–195.8)	0.1604
Fibrinogen (mg/L)	484 (415–539.5)	471.5 (386.3–535.5)	0.8820
Phosphate (mmol/L)	3.4 (3.1–3.8)	3.4 (3.225–3.6)	0.9087
Gamma-glutamyltransferase (U/L)	17 (13–26)	25.5 (15–32)	0.1615
Glucose (mg/dL)	92 (83–106)	111 (90–136)	0.0329
Hematocrit (%)	38.2 (35.55–40.35)	37.25 (32.88–40.3)	0.5044
Hemoglobin (g/dL)	33.5 (33–34.1)	33.95 (33.33–34.45)	0.0891
Lactate dehydrogenase (U/L)	171 (143.8–187.8)	172.5 (156.5–192.5)	0.7683
Leucocyte count (×109/L)	6.2 (5–7)	6.3 (4.97–7.8)	0.8605
Lymphocyte count (×109/L)	1.7 (1.35–2.05)	1.75 (1.225–2.35)	0.8263
Magnesium (mg/dL)	2.06 (1.898–2.188)	2.03 (1.893–2.115)	0.2356
Platelet (×109/L)	205 (168–249.5)	180 (145.8–221.5)	0.1236
Potassium (mmol/L)	4.32 (4.153–4.53)	4.39 (4.073–4.588)	0.6842
Serum protein (g/L)	65.95 (63.15–68.78)	65.35 (61.18–68.98)	0.7870
Sodium (mmol/L)	140.9 (139.1–142)	140 (138.6–141.6)	0.2790
Prothrombin time (s)	11.75 (11.2–12.4)	11.8 (11.1–13.2)	0.6207
Triglycerides (mg/dL)	115.5 (81.5–148.8)	111 (88.25–165.5)	0.8935
Partial thromboplastin Time (s)	30.05 (28.53–33.15)	30.55 (28.8–35.68)	0.3548
Urea (mg/dL)	40 (34–51)	43.5 (32–53)	0.7094

Values are given as n (%) or median (interquartile range, 25–75% percentiles). ¹ AGM level: Stratum of Adjusted Morbidity ranging from 1 to 4 according to the number of comorbidities and their need for health care. ² Chi-squared test, whereas all other p-values were determined by the Mann–Whitney U test, both tests were performed using GraphPad Prism 9.0 (GraphPad Software).

). The number of females was significantly higher in the infected group than the uninfected group (p = 0.0104). Only three infected individuals required hospitalization after an acute infection. Interestingly, 16 uninfected individuals remained seronegative for antibodies against the virus nucleoprotein until vaccination (January and February 2021) [18]. Table 1 shows the clinical and biochemical characteristics of the study patients. Compared with SARS-CoV-2-infected individuals, uninfected individuals had significantly higher glucose levels (Table 1). The higher glucose level in the uninfected group may be due to the higher number of diabetic individuals in this group (n = 8/16) compared with the infected group (n = 17/81; p = 0.0153, chi-squared). No other differences were found between the infected and uninfected individuals.

For the entire cohort of 97 individuals, we analyzed 12 SNPs from seven host genes previously associated with the innate immune response [9,10,11,18,19] or essential host cofactors involved in SARS-CoV-2 replication [22,23]. The observed allele SNP frequencies were characteristic of Southern European populations. All the variants had a minor allele frequency >0.01 and were in Hardy–Weinberg equilibrium (

Table 2. Association of 12 single nucleotide polymorphisms from 7 host genes with susceptibility to SARS-CoV-2 infection.

SNP	Gene	Chr: pos	MAF	HW p-Value	Alleles	COVID-19 (n = 81)			Uninfected (n = 16)			OR (95% CI)	p-Value	p-Value Bootstrap	Model
					1/2	1/1	1/2	2/2	1/1	1/2	2/2
rs1127313	ADAR-1	Chr1: 154583949	0.48	0.11	G/A	18	47	16	4	10	2	0.91 (0.25–3.34)	0.89	0.98	D
rs1990760	IFIH1	Chr2: 162267541	0.43	0.84	T/C	28	35	18	4	11	1	3.57 (1.04–12.29)	0.034	0.031	OD
rs2236575	MMS19	Chr10: 97465981	0.46	0.55	T/A	25	39	16	1	12	3	6.57 (0.80–53.87)	0.029	0.046	D
rs4767027	OAS-1	Chr 12: 112921352	0.38	0.40	C/T	30	37	14	9	5	2	0.43 (0.14–1.33)	0.14	0.14	D
rs10774671	OAS-1	Chr 12: 112919388	0.40	0.53	A/G	27	38	15	9	5	2	0.33 (0.10–1.07)	0.061	0.050	D
rs1293767	OAS-2	Chr 12: 112987349	0.37	0.51	G/C	33	35	13	7	7	2	0.81 (0.26–2.49)	0.71	0.65	D
rs10735079	OAS-3	Chr 12: 112942203	0.42	0.84	A/G	26	39	16	7	7	2	0.60 (0.19–1.90)	0.39	0.40	D
rs12979860	IFNL3	Chr 19: 39248147	0.29	0.14	C/T	42	31	8	10	3	3	0.40 (0.10–1.58)	0.17	0.19	OD
rs4803217	IFNL3	Chr 19: 39243580	0.27	0.44	C/A	43	31	7	10	4	2	0.51 (0.14–1.81)	0.28	0.31	OD
rs1060555	TNFAIP8L1	Chr 19: 4652810	0.32	0.17	C/G	36	38	7	5	11	0	1.29 (0.38–4.35)	0.68	0.71	D
rs2236757	IFNAR2	Chr 21: 33252612	0.30	0.09	G/A	41	28	11	10	5	1	0.41 (0.12–1.42)	0.14	0.13	D
rs738409	PNPLA3	Chr 22: 43928847	0.24	1.00	C/G	48	28	5	7	9	0	2.16 (0.69–6.80)	0.18	0.16	D

Adjusted by sex and age. Chr, chromosome; Pos, position; MAF, minor allele frequency; HW p-val, p-value from the Hardy-Weinberg equilibrium test (p > 0.05); OR (odds ratio); CI (confidence interval). D (Dominant; heterozygous and homozygous have the same risk, 1/1 vs. 1/2 + 2/2), and OD (Over Dominant; heterozygous are compared to a pool of both homozygous alleles, 1/1 + 2/2 vs. 1/2) models.

). The four SNPs in the OAS gene were in high LD (R2 > 0.81). The SNPs in the IFNL3 gene were also in high LD (R2 = 0.94). No other significant LD was observed among the other study SNPs. For the association analyses for each SNP, we adjusted for sex and age, identifying genetic associations between susceptibility to SARS-CoV-2 infection and two SNPs— MMS19/rs2236575 in the dominant model (odds ratio [OR] 6.57, 95% CI 0.80–53.87, p = 0.029) and overdominant model (OR 3.60, 95% CI 0.97–13.30, p = 0.041), and IFIH1/rs1990760 in the overdominant model (OR 3.57, 95% CI 1.04–12.29, p = 0.034; Table 2). After adjusting for multiple comparisons, by means of bootstrap techniques, MMS19/rs2236575 and IFIH1/rs1990760 remained significantly associated with susceptibility to SARS-CoV-2 infection (Table 2). In the unadjusted analysis, MMS19/rs2236575 was the only variant related to susceptibility to SARS-CoV-2 infection in the dominant (p = 0.022) and overdominant (p = 0.049) models.

We made a Spearman’s correlation between our clinical data (Table 1) and MMS19/rs2236575 and IFIH1/rs1990760 individually in the entire cohort in the infected and uninfected individuals. In the entire cohort, we did not find any significant correlation with any of the clinical parameters described in Table 1. In the infected group, only MMS19_AA genotype was significantly correlated with lower levels of Creatine Kinase (Spearman r = −0.25, p = 0.0231). In the uninfected group, MMS19_AA genotype was significantly correlated with lower levels of phosphate (r = −0.66, p = 0.0043). We also found that the IFIH1_CC genotype was significantly correlated with higher levels of albumin (r = 0.52, p = 0.0398), platelets (r = 0.51, p = 0.0410), total protein (r = 0.618, p = 0.0112) and significant lower levels of prothrombin time (r = −0.5820, p = 0.0185) and phosphate (r = −0.54, p = 0.0297).

4. Discussion

In this study, we tested the association between host genetic polymorphisms and susceptibility to SARS-CoV-2 infection. After adjusting for sex and age, we found that residents of the LTCFs carrying MMS19/rs2236575_A and IFIH1/rs1990760_TC genetic variants were protected against SARS-CoV-2 infection.

The IFIH1/rs1990760 TT variant was previously reported to confer resistance to SARS-CoV-2 infection and explain the epidemiological features of the pandemic in different countries [24]. Importantly, COVID-19 patients with the IFIH1/rs1990760 TT variant have an attenuated inflammatory response and better outcomes [25]. The human gene IFIH1 encodes a helicase, melanoma differentiation-associated gene-5, which acts as a cytoplasmic virus receptor and triggers the transcription of type-1 interferon genes and the systemic inflammatory response. IFIH1/rs1990760 has been previously linked to bowel disease, type 1 diabetes, vitiligo, and liver disease progression in HIV-1 patients [19,26].

In the overdominant model, the heterozygote gives an advantage to the carrier; in our case, individuals with the IFIH1rs1990760_TC genotype have a lower risk of becoming infected with SARS-CoV-2. This observation is consistent with prior ones where the overdominant model of inheritance of rs73064425 in the LZTFL1 gene was most significant, suggesting C/T as risky genotype with susceptibility to COVID-19 [27].

In contrast to the IFIH1/rs1990760 variant, we are the first to describe a possible implication of MMS19/rs2236575 variants in susceptibility to SARS-CoV-2 infection. MMS19 is a component of the cytosolic iron-sulfur protein assembly machinery that transfers Fe/S clusters to various DNA metabolism-associated Fe/S proteins [28]. In the Turkish Genome Project data (https://tgd.tuseb.gov.tr/en/variant/10-97465981-T-A, accessed on 31 July 2024), we found that the SNP MMS19/rs2236575 is an intronic mutation located in a regulatory region of the MMS19 transcript 14. Whether MMS19/rs2236575 associates with the loss-of-function (LOF) of MMS19 remains to be clarified. This LOF would then likely reduce the ability of the virus to replicate because of its dependence on MMS19 for acquisition of its required iron sulfur cofactors. MMS19 is implicated in DNA replication and repair, and its deregulation has been linked to numerous human diseases, including cancer [17]. MMS19 degradation and its downregulation is a common feature in cancer and genomic stability and is evolutionarily controlled [29]. MMS19/rs2236575 has been explored as a possible marker of advanced epithelial ovarian cancer [30]. Finally, the MMS19 protein has been described as a potential cofactor of the SARS-CoV-2 RNA-dependent RNA polymerase [22]. The virus helicase (nonstructural protein 3) has a Fe/S cluster that modulates its RNA-binding and unwinding activities [23].

Isoforms in the Neanderthal haplotype on chromosome 12 in the gene encoding oligoadenylate synthetase (OAS) play a protective role for individuals of European ancestry against COVID-19 susceptibility and severity [9,10,11]. An additional finding of our study is that we did not find support for a protective role of OAS after interaction with SARS-CoV-2.

One limitation of our study is the reduced size of our cohort, especially in the uninfected group and, for this reason, our findings require validation in larger cohorts. We identified two host genetic variants associated with susceptibility to SARS-CoV-2 infection in an elderly population. Current SARS-CoV-2 variants tend to cause mild symptoms, with few hospitalizations and deaths [13,31,32]. Nevertheless, our findings can lead to better understanding of SARS-CoV-2 infections.