COVID-19 pandemic has proved to be an unrelenting health threat for more than a year now. More contagious and more deadly SARS-CoV-2 variants are a major concern, especially because effective, causal therapy is still unavailable, and vaccination coverage rates are lower than anticipated. Moreover, protection from vaccination is likely to last only for a season or two. One of the strategies to save lives is to identify groups at risk of severe COVID-19 disease and to implement measures aimed at those groups. The elderly, male, obese, chronic disease patients, patients with malignancies, immunocompromised, dark-skinned, socioeconomically disadvantaged and tobacco users suffered the most death toll from COVID-19 (1, 2). Also, interindividual genetic variations might be implicated in more severe COVID-19 (3, 4).

Another strategy directed toward reducing severe COVID-19 symptoms relies on high dose vitamins and trace elements (micronutrients) supplementation. Among micronutrients important for adequate immune function, vitamin D, zinc and selenium are of particular importance for coping with viral, respiratory infections, such as COVID-19 (5). Adequate status of these micronutrients is not only important for immune function and viral clearance, but also might mitigate life-threatening complications of SARS-CoV-2 infection, such as thrombosis and uncontrolled inflammation which leads to cytokine storm. Namely, vitamin D via vitamin D receptor (VDR) mediates immune function and regulation, strengthening of epithelial barriers and antioxidant defense, and also regulates expression of SARS-CoV-2 receptor, ACE2 (6–8). Zinc exerts direct anti-viral effects and also serves as a cofactor of dozens of proteins important for immune function and regulation, and antioxidative defense (9). Selenium via selenoproteins regulates immune function and provides antioxidant defense and vaso-protection (10).

Although important for various processes in our body, a lot of people worldwide have inadequate status of vitamin D, zinc or selenium. There is a pandemic of vitamin D deficiency, especially in higher latitudes due to inadequate sun exposure necessary for the synthesis of this vitamin (11). Selenium deficiency is prevalent in Europe and other parts of the world where soil is poor of this element, while zinc status is compromised predominantly in underdeveloped regions (12). Vitamin D, zinc and selenium deficiency is associated with factors of COVID-19 severity, such as old age, obesity, diabetes, dyslipidemia and chronic and acute inflammation. COVID-19 patients have lower status of vitamin D, zinc and selenium than healthy individuals (13). Also, patients with severe COVID-19 disease have lower status of vitamin D, zinc and selenium than mild and moderate disease patients (14–16). Preliminary evidence from intervention studies suggests that supplementation with these micronutrients protects from severe COVID-19 disease (17, 18).

Apart from chronic and acute health conditions, lifestyle and nutrition, genetic factors also play a role in bioavailability and status of vitamin D, zinc and selenium. Several large genome-wide association studies (GWAS) identified variants located near genes involved in synthesis, transport and metabolism of vitamin D, namely DHCR7/NADSYN1, GC and CYP2R1, respectively (19–21). Variants in these genes are associated with serum level of 25OHD, which is a measure of vitamin D status. Also, genetic variants in VDR gene, which codes for vitamin D receptor that regulates expression of more than 200 human genes (22), have been shown to increase susceptibility to acute lower respiratory infections in children, as well as several inflammatory and autoimmune diseases (23–26). GWAS studies comprising three large independent cohorts associated selenium status measured in whole blood and erythrocytes (27), as well as nails (reflects longer duration of selenium exposure) (28) with a locus near DMGDH gene. The same locus was implicated in a response to selenium supplementation in another GWAS (29). Zinc nutrigenetics is less studied, but a few candidate variants associated with status of this trace elements are identified in a large GWAS, and the most significant association was noted for rs2120019 variant near PPCDC gene (27).

Building on this premise, we hypothesize that genetic variants linked to suboptimal level or decreased bioavailability of vitamin D, zinc and selenium, may also serve as markers of disease course in COVID-19 patients, influenced by the level of these micronutrients. Therefore, we evaluated genetic determinants of vitamin D (DHCR7/NADSYN1 rs12785878, GC rs2282679, CYP2R1 rs10741657, VDR rs2228570), zinc (PPCDC rs2120019) and selenium (DMGDH rs17823744) as risk factors for severe forms of COVID-19. Furthermore, we attempted to investigate the inter-population diversity of the selected genetic variants. Inter-population nutrigenetic variability could suggest different relevance of nutrigenetic factors in populations exposed to different nutritional habits and specific environmental factors. These data should be taken into account when evaluating the possible relationship of nutrigenetic variants with severe COVID-19 forms in other populations struck by the pandemic.

In this study, we focused our attention to genetic variants related to altered level, bioavailability or mechanism of action of vitamin D, zinc and selenium, aiming to investigate their association with different COVID-19 presentation in children and adults. Serbia, like much of Europe and the world, faces micronutrient deficiency/insufficiency which is in part influenced by genetic variability. The prevalence of suboptimal levels of vitamin D and selenium in Serbian population is estimated to be around 50% (32, 33), while zinc insufficiency is around 7% (12). Bearing in mind that even a mild insufficiency of these micronutrients can cause suboptimal immune function, implications for COVID-19 pandemic can be formidable.

Results of our study showed association of DHCR7/NADSYN1 rs12785878 and CYP2R1 rs10741657 variants with severe form of COVID-19 in adult patients. In the pediatric COVID-19 group, we did not detect any associations between analyzed genetic factors and disease severity.

Variant rs12785878 is located 8 kilobases upstream of the DHCR7 gene on chromosome 11q12. DHCR7 codes for 7-dehydrocholesterol reductase, a key enzyme in cholesterol biosynthesis and vitamin D cutaneous production. Particularly, the activity of DHCR7 enzyme decreases the level of vitamin D precursor availability, 7-dehydorcholesterol, shunting it in the direction of cholesterol biosynthesis (34). DHCR7 rs12785878 variant was firstly noted in a large GWAS of 25OHD concentrations in 33 996 individuals of European descent by (21), and subsequently confirmed in other numerous studies, where the G allele was shown to be associated with lower 25OHD serum levels. In our comparative genetic population analysis, we showed that frequency of G allele markedly varied across different populations; it ranged from 0.75 in African and Asian populations to 0.22 in European populations. Considerable differences in allele frequencies between populations may indicate a locus that has undergone positive selection in a specific geographical area. It has been suggested that positive selection influenced the DHCR7 gene, increasing the frequency of reduced activity variant (allele T) in European populations (35). Reduced activity of DHCR7 leads to increased availability of 7-dehydrocholesterol for vitamin D synthesis allowing Europeans to avoid deficiency in northern latitudes. Our results suggested an association of the rs12785878 variant with the severe form of COVID-19. However, in our study, carriers of the TT genotype which has been linked to higher circulating level of 25OHD, were more likely to develop a severe form of COVID-19. Given the immune-enhancing aspects of vitamin D, one might expect the opposite. However, it should be noted that the roll of DHCR7 variants to vitamin D status is constrained to vitamin D synthesis in the skin exposed to UV light and probably does not influence contribution of dietary sources and supplementation to the level of this vitamin. COVID-19 patients enrolled in our study were hospitalized with this disease in the spring of 2020 during or right after lockdown measures were in place. That means that vitamin D was scarcely synthetized in skin during that time and, therefore, DHCR7 variants had limited influence on vitamin D level. Regardless of vitamin D production, DHCR7 variants could influence immune function mediated through cholesterol metabolites. Namely, reduced activity of DHCR7 leads to increased 7-dehydrocholesterol levels which influences interferon production and viral clearance (36).

After initial UV-light mediated production of vitamin D3 in the skin, the second step in the synthesis of active vitamin D is catalyzed in the liver by cytochrome P450 (CYP) enzyme which hydroxylates carbon 25, producing the intermediate 25-hydroxyvitamin D3, or 25OHD, the major circulatory form of the vitamin D (37). This enzyme is encoded by the CYP2R1 (cytochrome P450, family 2, subfamily R, polypeptide 1) gene. Variant rs10741657 located near the CYP2R1 gene was linked by several studies to 25OHD serum concentrations (21, 38), with the allele G shown to be associated with lower 25OHD serum levels, and with the GG homozygotes having the lowest levels. Our results indicated an association between severe COVID-19 and CYP2R1 GG genotype. In the logistic regression model, CYP2R1 rs10741657 GG carriers had 5.9 times higher odds to develop severe disease. We noticed low level of variability of allele G frequency for CYP2R1 rs10741657 variant across analyzed populations (dAF 0.15); G allele frequencies ranged from 0.58 in Serbian, which was the lowest among European populations, to 0.73 in African population. The highest frequency of the G allele among Europeans was observed in Italian and Spanish populations (0.66 and 0.69, respectively). Recent study that reported negative correlations between mean levels of vitamin D in various European countries with COVID-19 cases per million people, suggested that Spain, Italy and Switzerland are the most vulnerable countries since they have very low levels of vitamin D in the aging population (39). Our results could be of particular interest for these populations. Relatively high occurrence of the CYP2R1 rs10741657 G variant in different populations, reproducible association with decreased 25OHD levels and observed relationship with severe COVID-19 in our group of patients, gives a solid ground for future studies to examine relationship of this CYP2R1 variant with the clinical course of COVID-19 in other populations.

A role of vitamin D in the pathogenesis of COVID-19 has been extensively studied since the beginning of the pandemic. Calcitriol (1, 25-dihydroxyvitamin D3) has an important role in regulating renin angiotensin system by enhancing the expression of ACE2, which is the main target of SARS-CoV2 cell entry (7, 8). Also, vitamin D has an immunomodulatory effect and it can prevent macrophages to release excessive proinflammatory cytokines and chemokines (6, 40). A recent study showed that the intake levels of relevant micronutrients such as vitamin D are inversely associated with higher COVID-19 incidence and/or mortality (41).

Furthermore, other studies sought for more detailed nutrigenetic markers as factors that might contribute to the bioavailability of vitamin D and therefore influence COVID-19 susceptibility and/or clinical course. A study which aimed to assess the association between variants in the GC gene that encodes vitamin D binding/transport protein (DBP), and the prevalence and mortality rates of COVID-19, pointed out to rs7041 variant (42). Another recent study (not peer-reviewed yet) performed in Portuguese population of COVID-19 patients found an association of GC rs2282679 variant with COVID-19 disease severity (43). Our study indicated an association between DHCR7/NADSYN1 rs12785878 and CYP2R1 rs10741657 variants and the severity of COVID-19, but the variants GC rs2282679 and VDR rs2228570 were not linked to a higher risk of severe COVID-19 in adults. In a recent Mendelian randomization study (not peer-reviewed yet) on vitamin D and COVID-19 in individuals of European ancestry, genetically increased 25OHD concentrations did not protect against COVID-19 susceptibility, but increased the odds for hospitalization and severe disease (44). Another Mendelian randomization study failed to find evidence of a linkage between vitamin D deficiency and COVID-19 infection rates or severe disease (45). However, described studies did not consider true vitamin D deficiency in participants, and it can be possible that vitamin D supplementation could benefit insufficient/deficient patients (44, 45).

Nutrigenetics of zinc and selenium were discussed in the context of COVID-19 as a promising strategy to implement personalized approach to strengthening antioxidant and antiviral defense (41). However, the present study is the first study on genetic determinants of zinc and selenium that included COVID-19 patients. Our results did not show association of these variants with severe COVID-19, however, our univariate analysis pointed out to DMGDH rs17823744 variant, which warrants further investigation. DMGDH is implicated in homocysteine metabolism and it is postulated that there is a connection between selenium exposure and the homocysteine metabolic pathways (28). Apart from locus near DMGDH gene, which was noted in several GWAS cohorts focused on selenium status, other studies analyzed genes coding for selenoproteins. Candidate gene approach studies or studies that analyzed protein activity, associated glutathione peroxidase 1 (GPX1), GPX4 and selenoprotein P (SELENOP) with response to selenium supplementation (46–48). Although important for selenium action, these genes were not associated with selenium status in GWAS studies.

Apart from nutrigenetics, other studies also dealt with genetic factors associated with COVID-19 severity, but the results were not consistent. A GWAS study of almost 4,000 healthy volunteers and severely ill Italian and Spanish COVID-19 patients associated loci on chromosomes 3 and chromosome 9 (near ABO locus) (3), while another GWAS study of 2,244 critically ill UK patients associated variants located on chromosomes 12, 19 and 21 with more severe disease or poor outcome (4). None of these loci are located near nutrigenetic variants analyzed in the present study.

Results of this study should be evaluated in larger cohorts of COVID-19 patients. Small number of patients is the main limitation of this study, although, both mild and moderate, as well as severe groups are well-represented among adult patients. Among pediatric patients, the majority of patients had mild symptoms so we could not investigate risk factors of severe disease in children. Instead, genetic variants were analyzed as risk factors of symptomatic disease.

Another limitation of this study is lack of micronutrient measurement needed to assess actual status of vitamin D, zinc and selenium. We focused solely on the genetic variants associated with altered levels of vitamin D, zinc and selenium. Status of the micronutrient is highly confounded variable that depends on numerous factors. Beside genetics, nutrition habits, individual lifestyle and presence of comorbidities are also important. Therefore, comprehensive nutrigenetic studies complemented with measures of nutritional intake, micronutrient's serum level and detailed lifestyle evaluations are needed in the future.

This study highlights the importance of personalized as well as population based strategies directed toward reduction of COVID-19 burden. These strategies can be implemented through employment of established nutrigenetic markers, such as genetic variants involved in vitamin D disposition. Our results pointed out to DHCR7/NADSYN1 rs12785878 and CYP2R1 rs10741657 variants, both involved in vitamin D synthesis, as potential risk factors of severe COVID-19. Holistic approach comprising nutrigenetics, micronutrient, status, lifestyle and clinical parameters could effectively combat micronutrient deficiency, and potentially improve anti-viral defense. This approach is especially important in populations at risk of micronutrient deficiency.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

This study was approved by the Ethics Committee of the Institute of Molecular Genetics and Genetic Engineering University of Belgrade (approval for sample collection and biobank formation O-EO-016/2020, 06.05.2020.; approval for the genetic study O-EO-016/2020/1, 03.09.2020).

NK: conceptualization, investigation, statistical analysis, writing - draft preparation, and editing. AS: methodology, investigation, and writing – draft preparation. KK: investigation, methodology, and writing – draft preparation. VG: data analysis and interpretation, investigation, and writing – draft preparation. BZ: methodology, investigation, writing — review, and editing, VS-T, MS, ZZ, OO, GS, and LL: methodology, sample collection, and clinical data analysis. SP: concept and design of the study, writing –review, and editing, BS: concept and design of the study, statistical analysis, results interpretation, drafting, and review of final manuscript. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.