1. IntroductionAs of December 2022, there have been 648 million cases worldwide of coronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and over six million deaths. COVID-19 is notable for its wide range of clinical presentations, with most patients exhibiting mild disease, but with nearly 20% exhibiting symptoms ranging from hypoxia to respiratory failure and death [1]. Highly effective vaccines and boosters now offer reliable (yet imperfect) protection against infection and severe disease [2,3]. While factors such as age and the presence of comorbidities have been clearly linked to the likelihood of developing severe disease [4,5,6,7], less is known about the host molecular factors that underlie disease severity. Identification and characterization of molecular biomarkers of COVID-19 disease states may enable early identification of patients at increased risk for developing severe disease. Additionally, biomarkers may point to specific biological pathways that contribute to the development of symptomatic or severe disease.Many studies have identified serum protein and metabolite biomarkers of COVID-19 in an effort to determine the molecular basis for severe disease [8,9,10,11,12]. These studies identify proteins and metabolites present at different concentrations in the serum of healthy individuals and individuals with varying degrees of COVID-19 severity. Identified biomarkers have pointed to disruptions in several biological pathways associated with COVID-19, including amino acid and lipid metabolism, and heme biosynthesis, among others. Several studies have also sought to identify alterations in erythrocytes as a result of COVID-19 infection, and have found a range of changes, including disruptions in levels of proteins involved in amino acid metabolism, changes in RBC morphology, and evidence of increased oxidative stress in RBCs [13,14,15]. In a previously performed classic twin study of erythrocyte proteins and metabolites, the authors identified over 700 unique molecules present in red blood cells and determined correlations among their levels as well as their heritability [16,17]. In addition to being easily collected, erythrocytes are metabolically active cells that maintain homeostasis in many metabolic pathways, enabling their use to identify relationships among proteins and metabolic pathways. Data from this twin study revealed heritability of entire metabolic pathways, including energy metabolism pathways [16]. Measurements of biomarker heritability, which frequently rely on twin studies but can also use studies of close relatives, offer a way to quantify the role of genetics in determining levels of specific proteins and metabolites. The heritability percentage of a biomarker is, therefore, the percentage of variation in its level that can be attributed to genetics rather than environmental factors [18]. While our previous twin study examined heritability of proteins and metabolites in the erythrocytes of healthy individuals, biomarker heritability has been studied in the context of numerous conditions, including chronic kidney disease, atherosclerosis, and metabolic syndrome [19,20,21,22].We found that many molecules identified as heritable in our previous twin study overlap with those found in early, pre-vaccine COVID-19-omics studies. Here, we cross-reference two pre-vaccine studies of COVID-19 metabolomics and proteomics with the aforementioned twin dataset to identify heritable biomarkers and biological pathways associated with COVID-19 severity [8,9]. Understanding the heritability of known biomarkers of COVID-19 severity may not only explain whether some individuals are predisposed to severe disease, but also provide insights into SARS-CoV-2 pathogenesis. 4. DiscussionIn this study, we interrogated a twin study database of erythrocyte proteins and metabolites to show that many of the plasma biomarkers associated with COVID-19 disease states are highly heritable in erythrocytes. The twin study previously showed inheritance of energy metabolism and redox pathways in erythrocytes, setting a precedent for its use to determine heritability of disease-associated pathways [16,17]. Metabolomics and proteomics studies of COVID-19 patients have exhaustively identified variations in protein and metabolite levels associated with the disease, but none have specifically addressed the heritability of biomarkers. Our results suggest that, in addition to individual biomarkers being heritable, several metabolic pathways downregulated in COVID-19, in particular amino acid biosynthesis and metabolism pathways, are heritable in erythrocytes. Understanding the heritability of disease biomarkers creates an opportunity to understand why certain individuals may be at heightened risk of severe disease and complications. To our knowledge, the finding that many COVID-19 biomarkers are heritable in erythrocytes is novel, as is the observation that many of these heritable biomarkers are part of amino acid metabolism pathways. These observations suggest that there may be familial risk of symptomatic or severe COVID-19 that warrants further exploration. Indeed, many studies have attempted to identify potential genetic risk factors for severe COVID-19. One study identified mutations in genes involved in type I interferon production as being associated with the development of severe pneumonia in the setting of COVID-19 [29]. A study of hospitalized Chinese patients found that those homozygous for a C at the rs12252 SNP in interferon-induced transmembrane protein 3 (IFITM3) gene (IFITM3) had more severe disease [30]. An analysis of UK Biobank data identified the ApoE e4e4 genotype, which is associated with dramatically increased risk of Alzheimer’s disease, as being associated with COVID-19 test positivity and severity [31]. Interestingly, a study using twin data found heritability of symptoms associated with COVID-19 [32]. The identification of many genetic variants that are associated with other respiratory illnesses suggests a potentially important role for genetics in COVID-19 susceptibility [33]. Our study suggests that amino acid metabolism pathways disrupted in COVID-19 [9] are heritable. In addition to the COVID-19-omics datasets our study referenced [8,9], many other studies have also described dysregulation of amino acid metabolism in COVID-19, although the directionality of these alterations in some cases differs among studies [34,35,36,37,38,39]. Amino acid metabolism is known to be required for immunity—beyond simply providing nutrients for protein synthesis, amino acids play a role in instructing cells to proliferate and carry out effector functions [40]. Levels of specific amino acids have been found to correlate with levels of proinflammatory cytokines [38], and amino acid metabolism has been linked to the development of cytokine release syndrome in COVID-19 [39]. A decrease in arginine is frequently seen among studies [9,34,37], and one study found that supplementation of arginine decreased the release of proinflammatory cytokines by peripheral blood mononuclear cells collected from SARS-CoV-2-infected macaques [39]. Taken together, these observations suggest that the dysregulation of amino acid metabolism pathways may be a targetable feature of severe COVID-19 infection. Our finding that these pathways are heritable suggests that genetics may govern, at least in part, the up- or downregulation of amino acid pathways that play a role in COVID-19 immune responses. We also found that aminoacyl-tRNA synthesis was a highly heritable pathway associated with COVID-19. Aminoacyl-tRNA synthetases have been found to be involved in immune cell development, and have also been associated with viral infections, in some cases being upregulated in response to infection and in others being specifically inhibited by viral RNA motifs [41]. The heritability of this pathway again points to a possible heritable impact on immunity that could influence COVID-19 severity. Additionally, other studies have identified changes in energy metabolism pathways in COVID-19 patients [13,38], which a previous study using our twin study dataset also found to be highly heritable in erythrocytes [16]. Unfortunately, many of the biomarkers identified as being predictive for COVID-19 severity were not in the twin erythrocyte data set, therefore we could not comment on heritability; however, it is possible that non-heritable biomarkers may present an opportunity to influence either likelihood of developing severe COVID-19 or alter the course of the illness. Limitations of this study include some aspects of the referenced datasets and the twin erythrocyte dataset. First, the COVID-19 datasets did not control for age, as noted by [9], so age may have acted as a confounder. Second, both COVID-19 datasets used plasma and serum for their metabolomics and proteomics experiments, while the twin multi-omics dataset used erythrocytes. To our knowledge, no systematic studies have been undertaken to demonstrate correlations between erythrocyte and plasma/serum protein and metabolite levels. One study published by Thomas et al. looked directly at protein and metabolite changes in the erythrocytes of COVID-19 patients. They found that while RBC parameters themselves were not significantly altered in COVID-19 patients, there were changes in energy and lipid metabolism pathways that pointed to changes in cell membrane homeostasis [13]. Analysis of the heritability of these specific changes in erythrocyte proteins and metabolites seen in COVID-19 patients using our erythrocyte-omics dataset is an important next step in evaluating the heritability of COVID-19 biomarkers. In spite of these limitations, we believe that our use of erythrocyte-omics data in this study provides valuable insight into the heritability of COVID-19 biomarkers. Erythrocytes enable detailed analyses of metabolic pathways in a viable cell, and our group and others have previously shown whole metabolic pathways to be heritable in erythrocytes [16,17,42]. A study of plasma metabolite heritability found that many of the amino acids whose metabolism pathways we identified as being heritable were themselves heritable in plasma [43]. With the exception of leucine and isoleucine, which were found to not be heritable in the twin erythrocyte data set (though their overall metabolism pathways were heritable) but were ≥30% heritable in plasma, amino acid heritability was higher in the erythrocyte dataset than in the plasma dataset. This may be reflective of the competing and complicating factors, such as kidney function and gut absorption, that impact serum and plasma more directly than they do erythrocytes.