Genetic and functional evidence identify Carboxypeptidase Q as a regulator of inflammation in respiratory infection

Genetic and functional evidence identify Carboxypeptidase Q as a regulator of inflammation in respiratory infection | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Genetic and functional evidence identify Carboxypeptidase Q as a regulator of inflammation in respiratory infection Can Hou, Yu Zeng, Yudong Yin, Yanan Zhang, Lei Yang, Yan Jiao, and 12 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8104260/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 10 You are reading this latest preprint version Abstract Background Host genetic factors may play a critical role in modulating severity of respiratory infections. However, previous studies have often been limited by pathogen heterogeneity and exposure misclassification. Results Utilizing a relatively homogenous Chinese population consists of 5,151 individuals with first-time infection by syndrome coronavirus 2 (SARS-CoV-2) Omicron variant, we conducted a genome-wide association study and identified a novel locus at 8q22.1 (rs7817424, P = 4.60×10⁻⁸) associated with infection severity. Integrating results from gene mapping and the similarity-based gene prioritization suggested carboxypeptidase Q ( CPQ) gene as the likely causal gene. Single-cell RNA sequencing and transcription factor motif analyses revealed differential CPQ expression in lung immune cells, particularly tissue-resident macrophages and monocyte-derived macrophages, implicating innate immune pathways in severe disease. Functional experiments demonstrated that CPQ overexpression in THP-1 cells suppresses LPS-induced pro-inflammatory cytokines TNF-α and IL-6, while systemic inflammation mouse model showed reduced CPQ expression in lung tissues during severe pneumonia. Conclusions This study identifies establish CPQ as a novel genetic determinant for respiratory infection severity and uncover its previously unrecognized anti-inflammatory role, highlighting its potential as a therapeutic target for controlling hyperinflammatory responses in COVID-19 and beyond. respiratory infection SARS-CoV-2 Omicron genome-wide association study carboxypeptidase Q CPQ gene functional validation Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 INTRODUCTION Infections, in particular respiratory infections, are one of the major causes of deaths worldwide 1 , 2 . While environmental exposure and pathogen characteristics play important roles, accumulating evidence suggests that host genetic factors substantially influence both susceptibility to infection and disease severity 3 , 4 . For example, in the context of infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causative agent of coronavirus disease 2019 (COVID-19), where some individuals experience only mild symptoms while others develop life-threatening complications. Dysregulated immune responses, including hyperinflammatory states, have been implicated in severe infection consequences 5 , 6 . Genome-wide association studies (GWAS) have since identified multiple common genetic variants associated with critical illness in COVID-19, shedding light on host biological pathways involved in antiviral defense and immune regulation 7 , 8 . However, prior to the COVID-19 pandemic, GWAS of infectious diseases were relatively were relatively few and generally limited in scale and resolution 9 . This was due in part to challenges such as pathogen heterogeneity, variable exposure levels, and the lack of large, well-phenotyped cohorts. For example, polymorphisms at the IFNL4 / IFNL3 locus have been shown to influence treatment response differentially across hepatitis C virus genotypes, underscoring how pathogen variation can confound host genetic associations 10 . Another major challenge is pathogen exposure misclassification, which violates the assumption that all individuals in a study are equally at risk. In reality, individuals who have never been exposed to a pathogen are not susceptible to infection, and exposure risk itself is shaped by a complex interplay of environmental and genetic factors, including reinfection 9 , 11 . Moreover, most existing studies have been conducted in populations of predominantly European ancestry, limiting the generalizability of their findings to other ethnic groups. As such, there remains a critical need to conduct well-powered genetic studies in diverse populations using carefully defined clinical phenotypes and uniform pathogen exposure. To address these gaps, we conducted a GWAS in a relatively homogeneous Chinese population consisting of 5,151 individuals who were newly infected with the SARS-CoV-2 Omicron variant since January 2022. During the massive Omicron outbreak in China following the relaxation of COVID-19 restrictions in late 2022, it was estimated that approximately 97% of the population became infected 12 . This context offers a natural, near-universal exposure model, thereby minimizing confounding from differential pathogen exposure, prior immunity, or viral strain heterogeneity. RESULTS Genome-wide association We conducted a GWAS by combining samples from two ongoing multicenter longitudinal cohorts in China: The China Surgery and Anesthesia Cohort (CSAC) and the China Severe Trauma Cohort (CSTC) 13,14 . COVID-19 related information was collected for both cohorts since March 2023 with standardized questionnaire (Supplementary Methods) measuring infection status, medical treatment received, and symptoms experienced during and after infection. By June 2023, 5,151 participants from the CSAC and CSTC cohorts (3,757 and 1,394, respectively) had both phenotype data on COVID-19 severity and genotyping data that passed quality control (QC) procedures (Figure 1 and Methods). Of these, 269 reported first-time infection and severe manifestations of COVID-19 (i.e., underwent inpatient medical care or received medical interventions at primary care centers, Supplementary Table 1). We identified two single-nucleotide polymorphisms (SNPs) that reached genome-wide significance level of P -value<5×10 −8 in the 8q22.1 genomic region (Figure 2a and Supplementary Table 2). The quantile-quantile (QQ) plot (Figure 2b) and the genomic inflation factor (λ=0.998) indicated minimal systematic biases. A high level of concordance between genotyping and whole genome sequencing (WGS) was observed for sub-threshold SNPs (i.e., P -value<5×10 −6 ) in a subset of randomly selected 498 patients (concordance rate ranging from 95.18% to 99.60%; Supplementary Table 3), underscoring the substantial accuracy of both genotyping and imputation techniques employed within the current GWAS. Clumping of these identified significant SNPs led to the identification of a single linkage‑disequilibrium (LD) independent risk locus located within the protein-coding gene Carboxypeptidase Q ( CPQ ). The lead SNP, rs7817424 ( P -value=4.60×10 −8 ), is located within the intronic region of this gene (Figure 2c). We next examined genome-wide significant signals within ±500 kb of this risk locus in prior GWAS. A total of 25 distinct traits were identified as significantly associated with variants within this genomic region (Supplementary Table 4). Among these traits, six were related to the respiratory or hematological systems, including the lung function, chronic obstructive pulmonary disease, mean corpuscular hemoglobin, mean corpuscular volume, red cell distribution width and acute myeloid leukemia. Association between severity by Omicron variant and polygenic risk scores (PRS) using previous COVID-19 GWAS We performed polygenic risk score (PRS) analyses to evaluate the validity of our definition of COVID-19 severity and to compare our results with prior COVID-19 GWAS. GWAS summary statistics of relevant COVID-19 outcomes conducted in East Asian population from the COVID-19 Host Genetics Initiative study (release 7) were used as the discovery set 15 . We found marginal significant associations between PRSs and the COVID-19 severity. Specifically, for the PRS of critical COVID-19 (versus controls without COVID-19), the odd ratio (OR) was 1.12 (95% confidence interval [CI]: 0.99-1.27, P -value=0.07). Similarly, for the PRS of hospitalized COVID-19 (versus controls without COVID-19), the OR was 1.11 (95% CI: 0.98-1.25, P -value=0.09). These findings, while not achieving statistical significance, do provide supporting evidence for the comparability between our GWAS on COVID-19 severity and previous research efforts. Enrichment of the genome-wide significant SNPs in chromatin states As all the significant SNPs within the identified risk locus are positioned in non-coding regions, we further assessed their capacity for regulating gene transcription by examining their enrichments across 15 distinct chromatin states from the Roadmap Epigenomics project (Methods) 16 , each representing diverse regulatory elements. Using a hypergeometric distribution test, we observed strong enrichment signals in chromatin states associated with enhancers, weak transcription, weak polycomb repressor complex, and heterochromatin across various blood and lung tissues/cells (Supplementary Figure 1). Putative risk gene mapping We employed a consensus-based gene mapping approach to identify potential risk genes associated with the identified risk locus 17,18 . This approach combined five analyses: expression quantitative trait loci (eQTL) colocalization, protein quantitative trait loci (pQTL) colocalization, splicing quantitative trait loci (sQTL) colocalization, protein-altering variant (PAV) linkage and similarity-based gene prioritization (Methods). Except for PAV linkage, the other four analyses prioritized CPQ gene as the putative causal candidate for the risk locus (Figure 3a). Specifically, eQTL colocalization analyses using coloc 19 uncovered significant colocalization (posterior probability [PP] for colocalization ≥0.8) between GWAS signals and eQTL signals of CPQ in 17 out of 49 Genotype-Tissue Expression (GTEx) tissues 20 (Supplementary Table 5), including lung and whole blood (PP=0.89 and 0.89; Figure 3b). pQTL colocalization analyses underscored CPQ as the only gene where GWAS signals and serum pQTL signals from previous proteomic GWAS 21 exhibited significant colocalization (PP=0.90; Figure 3b and Supplementary Table 6). Importantly, the direction of effect for the risk allele was consistent at both transcription and protein levels, suggesting that the risk locus might increase susceptibility to severe COVID-19 by up-regulating CPQ expression. Additionally, sQTL colocalization analysis yielded 32 variant-splice pairs with a PP≥0.8 across 12 GTEx tissues, all implicating CPQ gene (Supplementary Table 7). PAV linkage analyses revealed that none of the variants in strong or moderate LD with the lead variant were predicted to be loss of function or missense variants (Supplementary Table 8) 22 . Finally, by integrating GWAS signals with gene features from diverse sources, similarity-based gene prioritization using Polygenic Priority Score (PoPS) method robustly identified CPQ as the most plausible causal gene proximal to the risk locus (Supplementary Table 9) 23 . Characterizing the function of CPQ through single-cell RNA sequencing (scRNA-seq) data To further investigate the potential role of CPQ in SARS‑CoV‑2 infection, we re-analyzed scRNA-seq data from lung tissues and peripheral blood mononuclear cells (PBMCs) in severe COVID-19 patients and healthy controls from two published studies 24,25 . Through differential gene expression (DGE) analyses, we observed a significant up-regulation of CPQ across various cell types in lung (Figure 4a-b), including alveolar macrophages, monocyte-derived macrophages (MDM), and the precursor cells of MDM (i.e., monocytes and transitioning MDM), all of which directly participate in the innate immune response to SARS‑CoV‑2 infection. DGE analyses by disease status further revealed notably higher CPQ expression levels in healthy controls compared to severe COVID-19 patients across all the aforementioned cell types (Supplementary Figure 2). In PBMCs, CPQ also demonstrated significant up-regulation in three monocyte subsets as well as monocyte-derived Dendritic cells (Figure 4c-d), but no significant difference was observed in their expression levels between severe COVID-19 patients and healthy controls (Supplementary Figure 3). Collectively, these findings underscore both tissue-resident macrophages and MDM along with its immature precursor cells in lung as the putative causal cell types for CPQ gene. We observed variations in CPQ expression levels among MDM and its precursor cells, monocytes and transitioning MDM, in severe COVID-19 patients (Supplementary Figure 2), suggesting a potential modulation of CPQ expression during the transition from monocytes to MDM in the context of SARS‑CoV‑2 infection. We then investigated the dynamic changes in CPQ expression throughout this transition process. Employing the pseudotime trajectory analysis approach implemented in Monocle3 26 , we assigned each cell a pseudotime value, representing its relative progress in the transition process (Supplementary Figure 4). Regression of CPQ expression on pseudotime revealed a distinct U-shaped trajectory (Figure 4e), with initial down-regulation during the transition and subsequent recovery post-transition. These results suggest the involvement of CPQ in both monocyte and MDM-specific functions. Focusing on the plausible causal cell types in lung, we conducted co-expression network analysis using hdWGCNA 27 among COVID-19 patients to identify genes co-expressed with CPQ . Nine co-expression modules were identified, comprising over 2,900 genes (Figure 4f and Supplementary Table 10). Enrichment analyses of the top 100 genes in each module revealed distinct biological functions (Supplementary Table 11). Co-expression module 1, for instance, showed enrichment in protein synthesis and translation pathways, while module 8 (containing CPQ ) exhibited enrichment mainly in pathways associated with GTPase activity regulation and signal transduction. We extracted the 20 closest neighboring genes of CPQ in the co-expression network (representing genes with the highest co-expression correlation with CPQ ; Figure 4f) and performed enrichment analyses. Ten enriched pathways emerged, with five related to protein catabolism and localization, involving CPQ and six other genes (Figure 4g and Supplementary Table 12). Apart from CPQ , all six other genes related to protein catabolism and localization were also linked to ubiquitin-dependent protein catabolic processes (Supplementary Table 13). Protein-protein interaction network analysis of these six ubiquitination-related genes using STRING 28 revealed functional connections among USP48 (formerly known as USP31 ), GSK3B , and FBXW11 (Supplementary Figure 5), with GSK3B and FBXW11 identified as the top two closest genes to CPQ in the co-expression network. Identification of transcription factor (TF) binding affecting variants and their target TFs As evidence derived from consensus-based gene mapping and SNP enrichment analyses collectively suggest the variants within the risk locus may alter gene transcription by disrupting or enhancing TF binding, we then conducted a series of analyses to identify candidate variants affecting TF binding and their target TFs (Figure 5a). Among the 109 variants in strong or moderate LD (r 2 ≥0.6) with the lead variant, 64 exhibited high regulatory potential (RegulomeDB score of 1f or higher, Supplementary Table 14) 29 . Motif scanning of ±20 bp sequences flanking these eligible variants against the HOCOMOCD database 30 unveiled 28 variant-TF pairs with significantly increased binding affinity for risk alleles (Supplementary Table 15). Furthermore, we also found 25 variant-TF pairs exhibited a significant decrease in TF binding affinity for the risk alleles (Supplementary Table 16). To focus on TFs active in the putative causal cell types (i.e., monocytes and macrophages in lung), we compared their expression levels with that of background genes in these cell types. We identified a total of six TFs that were notably up-regulated in these cell types, including four TFs with likely enhanced binding affinity ( STAT6 , REL , RUNX1 , and ELF1 ; Figure 5b) and two TFs with likely disrupted binding affinity ( FOXP1 and NFKB1 ; Figure 5c). Among these TFs, three (i.e., STAT6 , ELF1 , and NFKB1 ) showed TF chromatin immunoprecipitation sequencing (ChIP-seq) peak signals near risk variants (rs3802194, rs7844514, and rs73698822, respectively) in K562 cells (Figure 5b-c). Both STAT6 and ELF1 are widely recognized as important transcriptional activators in the innate immune response to virus infection 31,32 , and their increased binding affinity when risk alleles are present aligns with findings of associations between risk alleles and heightened CPQ expression. The third TF, NFKB1 , is the DNA binding subunit of NF-κB and act as both a transcriptional activator (when forming a heterodimer with RELA , RELB , or REL protein) 33 and a repressor (when forming a homodimer with NFKB2 protein) 34 under different conditions. Validation of CPQ function in human cells and mouse model To validate the function of CPQ , a THP-1 cell line stably overexpressing CPQ was generated. Western blotting (WB) analysis confirmed significantly higher CPQ expression in these cells compared to wild-type (WT) controls (Supplementary Figure 6a). An inflammatory model was established in THP-1 cells using lipopolysaccharide (LPS) stimulation and optimizing conditions through time course (Supplementary Figure 6b) and concentration gradient (Supplementary Figure 6c) experiments. Based on these results, stimulation with 100 ng/mL LPS for 24 hours was selected as the standard inflammatory stimulus, while 1600 ng/mL was chosen for high-concentration comparison. ELISA results (Figure 6a-b) demonstrated significantly lower expression of inflammatory cytokines TNF-α and IL-6 in CPQ overexpression cells versus WT cells under high-concentration LPS stimulation (1600 ng/mL). Consistent with this, WB analysis (Figure 6c-f) further confirmed reduced cytokine expression in CPQ overexpression cells. To investigate CPQ 's role in vivo, a systemic inflammation model was established in mice through intraperitoneal LPS injection (10 mg/kg, 6 hours). Successful model induction was confirmed by elevated inflammatory factor expression in lung tissue (Supplementary Figure 6d). Simultaneously assessed CPQ expression levels in lung tissue were significantly lower in LPS-treated mice compared to normal controls (Figure 6g-h), suggesting that inflammatory responses may negatively regulate CPQ expression. Collectively, results from both human cellular and murine models indicate that CPQ exerts potential anti-inflammatory effects during inflammatory responses. DISCUSSION Our study, leveraging genetic data from a relatively homogeneous Chinese population, reveals a previously unrecognized immunomodulatory function of CPQ and highlight it as a potential therapeutic target for mitigating severe outcomes of COVID-19 and other respiratory infections. Through GWAS, we discovered two linked SNPs located at 8q22.1 that reached genome-wide significance. Our integrated analyses, including consensus-based gene mapping, similarity-based prioritization, and regulatory annotation, implicate the CPQ gene as the likely causal gene and suggest that risk variants may influence CPQ expression through regulation of gene transcription. Using scRNA-seq data, we observed differential expression of CPQ in lung tissues between COVID-19 patients and healthy controls. Notably, tissue-resident macrophages and MDMs along with their immature precursors, emerged as putative causal cell types. In these cells, motif-based predictions indicate that risk variants alter binding affinities of transcriptional activators and repressors that are known to play key roles in anti-infection responses. Supporting these observations, functional experiments in human THP-1 cells and systemic inflammation mouse models confirmed the anti-inflammatory role of CPQ , suggesting its involvement in modulating hyperinflammatory responses during severe respiratory infections. The non-coding GWAS variants were found to have significant enrichments across several chromatin states including enhancers, across various blood and lung tissues/cells, which is consistent with GWAS results from other infectious and inflammatory diseases including COVID-19 35–37 . Furthermore, the risk variants were suggested to alter the binding affinity of specific TFs, including those that were likely to be expressed in the putative causal cell types identified by scRNA-seq data analyses. These findings underscore the regulatory potential of these significant SNPs within tissues relevant to the studied outcome. The CPQ gene, where the associated variants locate, is identified as the potential risk gene from converging evidence. CPQ is a protein coding gene belongs to the peptidase M28 family, primarily responsible for catalyzing unsubstituted dipeptide cleavage into amino acids. Although it has not been reported in any prior genetic studies for infectious diseases, we observed its significant up-regulation across various cell types that directly participate in the immune response to SARS‑CoV‑2 infection (i.e., monocyte clusters and alveolar macrophages) in lung tissues and PBMCs using scRNA-seq data. The expression level of CPQ differed in lung tissues between COVID-19 patients and healthy controls, further implying its role in host immune response. Subsequent analyses focused on these putative causal cell types using gene co-expression network, followed by gene set enrichment and protein-protein interaction network, have identified several co-expression modules and distinct biological functions, including pathways associated with protein synthesis and translation, GTPase activity regulation, as well as protein catabolism and localization. Finally, converging evidence from cellular and animal models supporting a potential anti-inflammatory function of CPQ that is not restricted to Omicron infection. In human THP-1 cells, CPQ overexpression resulted in lower levels of TNF-α and IL-6 under high-dose LPS stimulation, demonstrated by ELISA and Western blot analyses, suggesting that CPQ may attenuate the production of inflammatory cytokines under acute inflammatory challenge. Furthermore, systemic inflammation induced by LPS led to a significant downregulation of CPQ expression in lung tissue in mice, probably due to the negative feedback suppression of CPQ during heightened inflammatory states. These results provided mechanistic insights that align with the genetic association between CPQ and severe COVID-19 observed in our GWAS, highlighting its previously unrecognized role in modulating inflammatory pathways. How CPQ participate in the immune response regulation remains unknown. Intriguingly, our results from the protein-protein interaction network and the TF binding analyses suggest that CPQ may participate in the NF-κB pathway activation in the regulatory process of host immune response against infection, likely through involvement in the ubiquitin-proteasome system to degrade the NF-κB inhibitor. NF-κB is a family of transcription factors that play critical roles in inflammation and immunity. The two genes GSK3B and FBXW11 prioritized as the top two nearest genes to CPQ have been demonstrated to play crucial roles in activating NF-κB signal pathway, by marking the inhibitor of NF-κB (i.e., IκB) for proteasome degradation through mediating its ubiquitination 38 – 40 . In addition, four other genes with the highest co-expression with CPQ are also involved in ubiquitin-dependent protein catabolism. In addition, among the TFs that were targets of the variants in LD with the lead variant, NFKB1 is a DNA binding subunit of NF-κB. It could associate with other NF-κB proteins to form distinct dimeric complexes that alter gene regulation in a variety of ways to control the overall levels of transcripts 41 . However, as we found no prior studies on CPQ and NF-κB pathway, the underlying mechanism remains unknown, and our hypothesis needs to be tested in future studies. The major strength of our study is its focus on uncovering the genetic determinants and underlying mechanisms that modulate the host response to respiratory infections, using Omicron infection as an optimal and timely model. Unlike prior GWAS conducted during early phase of the COVID-19 pandemic, when participants were infected with a mixture of SARS-CoV‐2 variants, resulting in heterogenous immune responses and clinical outcomes 42 , our study leverages a relatively homogeneous Chinese population with first-time Omicron infection. This uniform exposure minimizes confounding from prior immunity or variant-related differences, providing a uniquely controlled setting to investigate host genetic contributions to infection severity. Importantly, most individuals in our two cohorts had not been previously exposed to SARS-CoV-2, enabling a clearer dissection of primary immune responses. Another major strength lies in the convergence of evidence from multiple complementary analytic strategies, including GWAS, risk gene prioritization, regulatory annotation, single-cell expression analysis, and functional experimental validation, which together provide robust and mutually reinforcing support for the role of CPQ in modulating infection severity. These analytic steps also underscore how a well-defined, variant-specific infection context can serve as a powerful model for dissecting the host genetic architecture of respiratory infection outcomes more broadly. We also acknowledge several limitations in this study. First, no lab detection has been performed to identify the specific type of SARS-CoV-2 variants in the COVID-19 cases in our cohorts. Despite this, genomic surveillance study has shown that from September 2022 to January 2023, 99.5% of COVID-19 cases in China were infected with Omicron variant 43 . This support the validity of our assumption that vast majority of participants in our cohorts were infected by the Omicron. Second, this study was based on two clinic cohorts comprising patients undergoing surgical or trauma-related care, which may not be representative of the general population. Therefore, caution is needed when generalizing our findings to broader demographic groups. However, our downstream analyses, including those integrating publicly available single-cell sequencing datasets and results from functional experiments, provide additional biological validation and partially alleviate concerns regarding variant misclassification and cohort representativeness. In conclusion, this study provides a comprehensive investigation into the host genetic determinants and underlying mechanisms that modulate the severity of respiratory infections, using first-time Omicron infection as a timely and well-controlled model. We identified CPQ as a novel gene influencing COVID-19 severity in a Chinese population, supported by single‑cell expression and experimental evidence of its anti-inflammatory role. These findings suggest CPQ ’s broader relevance in immune regulation and highlight its potential as a therapeutic target. Further research is needed to validate its association with severity across diverse respiratory infections and populations and to elucidate the underlying molecular pathways linking CPQ with immune regulation. METHODS Study design The current GWAS is based on data obtained from two multicenter cohort studies conducted in China: The China Surgery and Anesthesia Cohort (CSAC) and The China Severe Trauma Cohort (CSTC), both launched in June-July 2020. Detailed study designs for the two cohorts have been previously described 13 , 14 . In brief, the CSAC was conducted in four medical centers in China, with the primary objective of recruiting adults aged 40 to 65 years who underwent elective surgery and general anesthesia. The CSTC aimed to recruit patients with traumatic injuries within three months of admission to the Trauma Center of West China Hospital in Chengdu, China. Both cohorts implemented similar survey protocols and quality control processes, managed through a Cohort Data Collection and Management System (Build 2021SR0484324. ©West China Hospital, Sichuan, China). Upon obtaining informed consent and screening for eligibility, trained data collectors utilized a standardized questionnaire to gather comprehensive information on sociodemographic and lifestyle factors at baseline for each patient. Peripheral blood and fecal samples were also collected from each patient by research nurses following a standard protocol at baseline. Subsequently, patients were regularly followed up at 1, 3, 6, and 12 months after discharge from the hospital, with long-term follow-up conducted through periodic linkage to multiple national or regional databases. In response to the massive Omicron outbreak in China that occurred in late 2022, during which an estimated 97% of the population was infected 12 , both cohorts began collecting COVID-19 related information from March 2023 onwards. A standardized questionnaire (Supplementary Methods) was used to collect data on SARS‑CoV‑2 infection status, medical treatment received, and symptoms experienced during and after the infection, with response rate of 95.0%. As of June 2023, the CSAC and CSTC cohorts recruited a total of 14,538 patients, among which 28 patients had withdrawn later and 6,773 patients have not been genotyped yet. A total of 7737 patients from both cohorts were sent for first wave genotyping and eligible for the present study (Fig. 1 ). 5,151 participants from the CSAC and CSTC cohorts (3,757 and 1,394, respectively) had both measures of COVID-19 severity and genotyping data that passed quality control (QC) procedures, (Fig. 1 and Methods), among whom 269 reported that they experienced the first-time infection and exhibited severe manifestations of COVID-19 (i.e., underwent inpatient medical care or received medical interventions at primary care centers, including transfusion, injection, oxygen therapy, nebulization; Supplementary Table 1). The CSAC, CSTC, and the current study, received ethical approval from the ethics committee of West China Hospital, Sichuan University, with approval numbers 2020 − 243, 2020 − 469, and 2020.661, respectively. Sex as a biological variable Both male and female participants were included in the genome-wide association analyses (2,256 males and 2,895 females). Sex was recorded at baseline and included as a covariate in all analyses to control for potential confounding. No sex-stratified analyses were conducted, and the findings are expected to be relevant to both sexes. For the animal experiments, both male and female C57BL/6 mice were used (3 females and 4 males, 8 weeks old). Sex was not used as an experimental variable, and no sex-specific comparisons were performed. The findings are therefore considered generally applicable to both sexes. Genome-wide genotype data, quality control, imputation and annotation Blood samples were sent for genotyping at the WeGene Clinical Laboratory in Shenzhen, China, using the Illumina Infinium Chinese Genotyping Array (WeGene V3) which covers approximately 700k variants. We performed rigorous quality control and imputation using the Rapid Imputation and Computational Pipeline for Genome-Wide Association Studies (RICOPILI) 44 . A detailed flowchart outlining the steps can be found in Supplementary Methods. In summary, the pre-imputation quality control involved filtering 136,592 variants based on criteria of Hardy-Weinberg Equilibrium P -value (< 1×10 − 6 ), call rate (< 98%), non-biallelic variants, or indel variants, as well as 11 samples with issues of low call rate, sex discrepancy, or duplications. This process resulted in 590,757 single nucleotide polymorphisms (SNPs) on autosomes from 7,726 samples, which were deemed suitable for further analyses. Relatedness assessment and principal component analysis (PCA) were then conducted using a subset of SNPs that passed a more stringent quality control process. Subsequently, we performed imputation based on a reference panel comprising East Asian samples from phase 3 of the 1000 Genomes Project 45 . Following imputation, we performed post-imputation quality control, which involved removing 24,806,907 variants with low imputation quality (information score < 0.8), low allele frequencies (minor allele frequency [MAF] < 0.01), non-biallelic variants and indel variants. Additionally, we excluded 26 samples that showed first or second-degree relationships, Finally, we kept individuals with first-time infection of COVID-19 after 31st December 2021, resulting in a final set of 6,126,560 SNPs from 5,151 samples for the GWAS. We annotated each SNP with its Reference SNP cluster ID (rsID) according to dbSNP Build 151 using ANNOVAR 46 . Whole-genome sequencing data To evaluate the accuracy of the genotyping data, we conducted a comparison between the genotyping results and the results obtained from whole-genome sequencing (WGS) for the SNPs located in the identified risk loci. This analysis was carried out on a subset of patients who had both genotyping and WGS data available. From the CSAC and CSTC cohorts, a total of 733 patients were randomly selected for WGS, and among them, 498 patients were included in the current study. The WGS was conducted by the Institute of Rare Diseases at West China Hospital of Sichuan University. The DNBSEQ-T7 Next-Generation Sequencing (NGS) platform, developed by Complete Genomics and MGI, was utilized for the WGS process. A comprehensive description of the quality control for raw reads, alignment, variant calling, as well as variant and sample filtering can be found in the Supplementary Methods. Definition of severe COVID-19 The primary outcome of interest in this study is severe COVID-19, which is defined as patients requiring inpatient medical care during the infection period. In light of the rapidly spreading Omicron outbreak in China, which has placed a substantial burden on the majority of hospitals, we have broadened the definition of severe COVID-19 to encompass patients who received medical treatments such as transfusion, injection, oxygen therapy, or nebulization therapy in primary care centers during the infection period. This extension is essential as these patients may be unable to be admitted to the hospital due to the overwhelming circumstances. Genome-wide association analysis Using imputed dosage data from the two cohorts, we performed GWAS under an additive model for the severe COVID-19 outcome. To effectively account for population structure and residual kinship among patients, we employed a generalized linear mixed model implemented in GCTA 47 . The model incorporated several covariates, including age, sex, birth year, genotyping batch, cohort, and the first ten principal components. Based on the GWAS results, we generated quantile-quantile (QQ) plots and estimated the genomic inflation factor (λ) to assess whether there were any systematic deviations or inflation in the observed GWAS P-values. A variant was considered genome-wide significant if its P -value ≤ 5×10 − 8 . Identification of linkage disequilibrium independent risk loci To identify linkage disequilibrium (LD) independent risk loci, we employed the “clumping” command in PLINK v1.9 48 . Specifically, for each variant with P -value ≤ 5×10 − 8 (i.e., the index variant), other variants with P -value ≤ 5×10 − 6 , located within 500 kilobase (kb) distance, and LD correlated with the index variant (r 2 ≥ 0.1) were grouped into the index variant’s clump to form LD-independent genomic region. Subsequently, we merged physically overlapping genomic regions using BEDTools 49 to form LD-independent risk loci. The variant with the lowest P-value within each locus was considered the lead variant. To comprehensively characterize the identified risk loci and determine their novelty for COVID-19 severity or related outcomes, we conducted a search in the GWAS Catalog database (accessed on February 2025) 50 . Specifically, we retrieved all genome-wide significant associations from previous GWAS for variants within a 500 kb region surrounding the identified risk locus. PRS analyses We used the GWAS summary statistics (i.e., effect sizes and standard errors for the variants) for very severe respiratory confirmed COVID-19 (i.e., critically ill cases of COVID-19 defined as those individuals who required respiratory support in hospital or who died due to the disease; 784 cases vs. 4,862 population) and summary statistics for hospitalized COVID-19 (2,882 cases vs. 31,200 population) conducted in East Asian populations from the COVID-19 Host Genetics Initiative study (release 7, https://www.covid19hg.org/results/r7/ ) as base data sets 15 . We removed ambiguous SNPs and kept the SNPs available in both base data sets and target data set (the post-QC genotyping data of current study cohorts). PRSs for the two relevant COVID-19 outcomes were generated as weighted sum of the risk alleles through a penalized regression framework known as LASSO method, which allows for heavy shrinkage in the effect estimates of SNPs via regularization 51 . We calculated ORs with 95% CI for the associations between PRSs and COVID-19 severity using logistic regression models, adjusting for birth year, sex, genotyping batch, study cohort and the first ten principal components. Enrichment analyses for genome-wide significant variants We employed the hypergeometric distribution test to assess the enrichment of the identified genome-wide significant variants in specific chromatin states of the genome. The genomic regions representing 15 chromatin states in blood and lung tissues/cells were obtained from the Roadmap Epigenomics project 16 . For each genome-wide significant variant, we randomly selected 100 non-significant variants possessing comparable MAF, LD score, and distance to the nearest gene, to form the background set for the enrichment analyses. Chromatin states showing enrichment with a false discovery rate adjusted P -value (i.e., q-value) of less than 0.05 were considered statistically significant. Consensus-based mapping for putative risk genes We adopted a previously proposed consensus-based gene mapping approach to systematically map the identified risk loci to their putative risk genes 17 , 18 . This approach integrated evidence from the following five distinct analyses that covered various aspects, ranging from mRNA and protein expression to alternative splicing: 1) expression quantitative trait loci (eQTL) colocalization analyses, to identify a gene located near a locus that exhibited significant colocalization between its GWAS signals and eQTL signals in 49 Genotype-Tissue Expression (GTEx) tissues 20 ; 2) splicing quantitative trait loci (sQTL) colocalization analyses, to identify a gene located near a locus that exhibited significant colocalization between its GWAS signals and sQTL signals in 49 GTEx tissues 20 ; 3) protein quantitative trait loci (pQTL) colocalization analyses, to identify a gene located near a locus that exhibited significant colocalization between its GWAS signals and serum pQTL signals from AGES cohort 21 ; 4) protein-altering variant (PAV) linkage, to identify a gene located near a locus that harbored a PAV that is in strong or moderate LD with the identified lead variant, by Variant Effect Predictor 22 ; 5) similarity-based gene prioritization, to identify a gene located near a locus that received top three highest scores from the Polygenic Priority Score (PoPS) 52 . Comprehensive descriptions of each analysis are provided in the Supplementary Methods. For each locus, the putative risk gene was determined as the gene with the highest number of supporting evidence from the above five analyses. Single-cell RNA sequencing (scRNA-seq) data To gain a comprehensive understanding of the identified putative risk gene, we undertook an extensive characterization by re-analyzing previously published single-cell RNA sequencing (scRNA-seq) data of lung tissues and peripheral blood mononuclear cells (PBMCs) collected from severe COVID-19 patients and healthy controls 24 , 25 . Briefly, for the scRNA-seq data of lung tissues, we directly obtained the processed count matrix and cluster annotation data from the original study 25 . Our analysis concentrated on six major cell clusters of lung tissues, including epithelial cells, myeloid cells, fibroblasts, T cells, B cells, and endothelial cells. Within the subset of epithelial cells, myeloid cells, fibroblasts, and T cells, we additionally identified their respective major sub-clusters. This yielded a total of 21 distinct cell types, collectively accounting for approximately 92% of the entire cell population (see Fig. 4 for details). Regarding the scRNA-seq data from PBMCs, we obtained the raw count matrix data from the Gene Expression Omnibus (GEO) repository and processed them using the Seurat v5 53 , following the procedural framework outlined in the original study 24 . Following data normalizing and scaling procedures, we employed the Uniform Manifold Approximation and Projection (UMAP) 54 for dimensionality reduction and the shared nearest neighbor (SNN) modularity optimization algorithm 55 for clustering. The resulting cell clusters were assigned to one of the following 11 major cell types based on a curated list of cell markers outlined in the original study 24 : T cells (CD4 + and CD8+), B cells, natural killing (NK) cells, monocytes (classical, non-classical, and intermediate), platelets, red blood cells (RBC), Dendritic cells (DC), and other undefined cells. Differential gene expression analyses in scRNA-seq data To prioritize the putative causal cell types wherein the risk gene may exert its relevant functions, we conducted differential gene expression (DGE) analyses using the Mann-Whitney U test on the scRNA-seq data of both lung tissues and PBMCs, as previously mentioned. Initially, we conducted DGE analysis for the expression of the risk gene in each cell type in comparison to the rest cells, with the aim of identifying cell types where the risk gene displayed significant overexpression. Furthermore, we conducted DGE analysis for the expression of the risk gene between severe COVID-19 patients and healthy controls for each cell type. The significance level for the DGE analyses was set at a Bonferroni-corrected p-value of 0.002 (0.05/21 cell types) for lung scRNA-seq data and 0.004 (0.05/11 cell types) for PBMCs scRNA-seq data. Pseudotime trajectory analysis in scRNA-seq data To gain insights into the dynamics of the risk gene expression throughout the transition process from monocytes to monocyte-derived macrophages (MDM) in lung tissues during SARS‑CoV‑2 infection, we conducted a pseudotime trajectory analysis for these cell types in severe COVID-19 patients using the R package Monocle 3 26 . Each cell was assigned a pseudotime value, where 0 indicated no progress in the transition from monocyte to MDM, and the maximum value represented 100% progress. We then employed a negative binomial regression model with cubic splines to regress the risk gene expression against the pseudotime. Weighted gene co-expression network analysis in scRNA-seq data We performed weighted gene co-expression network analysis using the R package hdWGCNA 27 , enabling the construction of gene co-expression networks and the clustering of genes into distinct co-expression modules. We specifically focused the analysis on the putative causal cell types in severe COVID-19 patients, as indicated by the DGE analyses. To visualize the gene co-expression network, we employed the dimension reduction algorithm UMAP 54 on the co-expression matrix, to get two-dimensional UMAP features for each gene. Gene set enrichment analysis Based on the module assignment obtained from the weight gene co-expression network analysis, we conducted gene set enrichment analysis using GSEApy 56 . We assessed the enrichment of the top 100 genes within each module in pathways curated from various sources, including BioCarta 57 , Pathway Interaction Database (PID) 58 , WikiPathways 59 , Reactome 60 , and Gene Ontology (GO) 61 . Furthermore, we carried out enrichment analysis for the putative risk gene itself along with the 20 genes nearest to it, which were identified through the application of the K-Nearest Neighbors (KNN) algorithm 62 to the UMAP features of genes within the co-expression network. Statistical significance level for the gene set enrichment analysis was defined as q-value < 0.05. Protein-protein interaction network analysis We performed protein-protein interaction network analysis within a subset of ubiquitination-related genes using STRING ( https://string-db.org/ ) 28 . In particular, we focused on protein-protein interactions that were backed by high-throughput experiments and possessed association scores exceeding the median score of 0.04. Transcription factor binding motifs analyses We conducted transcription factor (TF) binding motif scanning and comparison analyses to identify potential TF binding affecting variants in the risk locus and their target TFs. Initially, we annotated lead variant and variants in strong or moderate LD with it (r 2 ≥ 0.6) within the locus, using the RegulomeDB ranking score 29 . Subsequent analyses were limited to variants with high regulatory potential (i.e., those with a RegulomeDB ranking score of “1f” or higher). For each eligible variant, we extracted both the reference and alternative DNA sequences (i.e., sequences containing the risk allele of the variant) from a ± 20 base pair (bp) region surrounding it, according to the GRCh38 reference sequence. Next, using the motif scanning software FIMO 63 , we calculated the binding affinity for the set of extracted sequences with each human TF binding motif from the HOCOMOCD database (version 11) 30 . We considered a variant to have the potential to increase the binding affinity of a specific TF if the P -value for motif occurrence in the reference sequence was ≤ 1×10 − 4 , while the P-value for motif occurrence in the alternative sequence was > 1×10 − 4 , and the difference in log-odds scores between the two sequences was ≥ 2. Using similar thresholds, we also identified a set of variants with the potential to disrupt TF binding (i.e., variants with P -value > 1×10 − 4 and ≤ 1×10 − 4 for reference and alternative sequences, and difference in log-odds scores≤-2). Filtering of the identified transcription factors Following the identification of the TFs with potential alterations in binding affinity due to the presence of the risk alleles, we performed DGE analyses to assess the expression of the chosen TFs in comparison to background genes for each putative causal cell type, to validate the functional activity of the selected TFs. Then, for TFs with confirmed functional relevance, we visualized their TF chromatin immunoprecipitation-sequencing (ChIP-seq) signals in the regions surrounding the variants that were predicted to affect their binding affinity. The ChIP-seq data used for visualization were obtained from the ENCODE project 64 . Experimental validation in human cells and mouse models Cell culture : THP-1 cells were obtained from National Collection of Authenticated Cell Cultures (NCACC). THP-1 CPQ overexpression cell line was generated using lentiviral transduction in our lab. THP-1 cells and CPQ overexpression cells were cultured in RPMI 1640 medium (Gibco) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin (Thermo Fisher Scientific) at 37°C in a humidified incubator with 5% CO₂. Generation of CPQ overexpression cell line : THP-1 cells were transduced with lentiviruses encoding the CPQ sequence. Viral particles were produced by co-transfecting 293T cells with the lentiviral transfer vector and packaging plasmids (pMD2.G and psPAX2) using Lipofectamine 3000 transfection reagent, according to the manufacturer's protocol. The viral supernatant was harvested 48 hours post-transfection, filtered through a 0.45-µm membrane, and used to transduce target cells cultured in 6-well plates (400 µL per well) in the presence of polybrene (final concentration: 10 µg/mL). After 48 hours of transduction, the medium was replaced with fresh medium containing puromycin (2 µg/mL) for selection. Construction of the Cellular Inflammation Model : THP-1 cells were collected during the logarithmic growth phase and treated with Phorbol 12-myristate 13-acetate (PMA) for 48 hours to induce differentiation. Morphological changes of the cells were monitored under an optical microscope. Following adherence, the cells were differentiated into mature, inactivated macrophages (M0 phenotype). The supernatant was removed, and the cells were washed with serum-free RPMI 1640 medium and incubated in this medium for 24 hours. The serum-free medium was then aspirated. Complete medium containing varying concentrations of LPS was added to the cells to induce inflammation. Mice : C57BL/6 mice at the age of 8 weeks were purchased from Chengdu Dossy Experimental Animals Co., Ltd. All animal experiments in this study were conducted with the approval of the Animal Ethics Committee of Sichuan University (Approval No. 20240308057). Mice were group-housed under a standard light cycle (12-h light–dark; lights on from 7:00 to 19:00) at 20–23°C and about 50% humidity with ad libitum access to water and food. The systemic inflammation model in mice is established through intraperitoneal injection of LPS (10 mg/kg) for 6 hours. Lung tissue is obtained through thoracic dissection, washed with pre-chilled PBS, and stored in liquid nitrogen. Tissue samples are added to lysis buffer (mPER, Thermofisher) at a 10% mass/volume ratio and homogenized. After 10 minutes of lysis on ice, the sample is centrifuged at 4°C for 15 minutes at 16000 xg to remove impurities. The supernatant is transferred to a new pre-chilled tube. Protein concentration is measured using the Bradford method and quantified to 3 µg/µL. ELISA : Enzyme-linked immunosorbent assay (ELISA) for human IL-6 (Boster) and TNF-α (Eeo bioscience) was performed according to the manufacturer’s protocols. The standard curves were calculated by Excel. Cell supernatant samples for ELISA assay were collected by centrifugation at 1200 rpm for 3 min. Western blot analysis : Total protein was extracted from cells or tissue and then analyzed using SDS–PAGE. Following electrophoresis, proteins were transferred to PVDF membranes (Millipore). Membranes were then incubated with the following primary antibodies: rabbit anti-PGCP (1:1000 dilution), rabbit anti-IL-6 (1:500 dilution), or mouse anti-TNF-α (1:500 dilution). Immunoblotting with an anti- β -actin antibody (1:5000 dilution) served as the loading control. Protein bands were visualized using enhanced chemiluminescence (ECL) detection reagents. Band intensities were quantified using Image J Software. All the reagents and chemicals used in this study are listed in Supplementary Methods Table 4. Statistical Data and Reproducibility : Statistical analysis was performed using GraphPad. Sample source, p -values, and the statistical methods employed are described in the corresponding legends or source data. A difference is considered significant when P -value < 0.05. Declarations ACKNOWLEDGEMENTS This work was supported by the 1.3.5 project for disciplines of excellence, West China Hospital, Sichuan University (No. ZYYC21005 to HS, ZYGD20005 to QL), the National Natural Science Foundation of China (82471535 to HS, 82425054 and 82273784 to BK, 82404350 to JS, and 82404391 to CH), and the Science & Technology Department of Sichuan Province (2024NSFSC1637 to JS and 2024NSFSC1568 to CH). We thank the team members involved in West China Biomedical Big Data Center for their support. The computations in this study were supported by the High Performance Computing platform at West China Biomedical Big Data Center, West China Hospital, Sichuan University. Data availability statement The GWAS summary statistics are available upon request. Conflict-of-interest statement The authors declare no competing interests. Contributors CH, JS, BK and HS were responsible for the study’s concept and design. LY and YQ did the acquisition of data. HY contributed to the phenotype definition. CH, WC and YZ did the data and project management. YY, FY did the experiment analyses in human cells and animal models. CH, YY and YZ did the data cleaning and analysis. YZ, CH, YY and JS drafted the manuscript. YZ, CH, YY, YJ, FY, QL, FB, JS, UAV, FF, BK and HS contributed to the interpretation of the results and revised the manuscript. All the authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work. Ethics approval and consent to participate The conduction of CSAC and CSTC were approved by the ethics committee of West China Hospital, Sichuan University (approval number: 2020.469 and 2020.243). The additional collection of COVID-19 data in those Chinese cohorts and this specific study were approved by the ethics committee and animal ethics committee of West China Hospital, Sichuan University (2020.469, 2020.243, 2020.66 and 20240308057). References Jin, X. et al. Global burden of upper respiratory infections in 204 countries and territories, from 1990 to 2019. eClinicalMedicine 37 (2021). Naghavi, M. et al. Global burden of 288 causes of death and life expectancy decomposition in 204 countries and territories and 811 subnational locations, 1990–2021: a systematic analysis for the Global Burden of Disease Study 2021. The Lancet 403 , 2100-2132 (2024). Kousathanas, A. et al. Whole-genome sequencing reveals host factors underlying critical COVID-19. Nature 607 , 97-103 (2022). Chen, H.-H. et al. Host genetic effects in pneumonia. The American Journal of Human Genetics 108 , 194-201 (2021). Chousterman, B.G., Swirski, F.K. & Weber, G.F. Cytokine storm and sepsis disease pathogenesis. Semin Immunopathol 39 , 517-528 (2017). Wong, L.R. & Perlman, S. Immune dysregulation and immunopathology induced by SARS-CoV-2 and related coronaviruses - are we our own worst enemy? Nat Rev Immunol 22 , 47-56 (2022). Pairo-Castineira, E. et al. GWAS and meta-analysis identifies 49 genetic variants underlying critical COVID-19. Nature 617 , 764-768 (2023). Kanai, M. et al. A second update on mapping the human genetic architecture of COVID-19. Nature 621 , E7-E26 (2023). Mozzi, A., Pontremoli, C. & Sironi, M. Genetic susceptibility to infectious diseases: Current status and future perspectives from genome-wide approaches. Infection, Genetics and Evolution 66 , 286-307 (2018). Ansari, M.A. et al. Genome-to-genome analysis highlights the effect of the human innate and adaptive immune systems on the hepatitis C virus. Nature genetics 49 , 666-673 (2017). Duchen, D. et al. Pathogen exposure misclassification can bias association signals in GWAS of infectious diseases when using population-based common control subjects. Am J Hum Genet 110 , 336-348 (2023). Goldberg, E.E., Lin, Q., Romero-Severson, E.O. & Ke, R. Swift and extensive Omicron outbreak in China after sudden exit from 'zero-COVID' policy. Nat Commun 14 , 3888 (2023). Yang, Y. et al. Cohort Profile: The China Severe Trauma Cohort (CSTC). J Epidemiol (2022). Yang, L. et al. Cohort profile: the China surgery and anesthesia cohort (CSAC). Eur J Epidemiol 39 , 207-218 (2024). Initiative, C.-H.G. The COVID-19 Host Genetics Initiative, a global initiative to elucidate the role of host genetic factors in susceptibility and severity of the SARS-CoV-2 virus pandemic. Eur J Hum Genet 28 , 715-718 (2020). Roadmap Epigenomics, C. et al. Integrative analysis of 111 reference human epigenomes. Nature 518 , 317-30 (2015). Aragam, K.G. et al. Discovery and systematic characterization of risk variants and genes for coronary artery disease in over a million participants. Nat Genet 54 , 1803-1815 (2022). Vujkovic, M. et al. A multiancestry genome-wide association study of unexplained chronic ALT elevation as a proxy for nonalcoholic fatty liver disease with histological and radiological validation. Nat Genet 54 , 761-771 (2022). Giambartolomei, C. et al. Bayesian test for colocalisation between pairs of genetic association studies using summary statistics. PLoS Genet 10 , e1004383 (2014). Consortium, G.T. The Genotype-Tissue Expression (GTEx) project. Nat Genet 45 , 580-5 (2013). Gudjonsson, A. et al. A genome-wide association study of serum proteins reveals shared loci with common diseases. Nat Commun 13 , 480 (2022). McLaren, W. et al. The Ensembl Variant Effect Predictor. Genome Biol 17 , 122 (2016). Weeks, E.M. et al. Leveraging polygenic enrichments of gene features to predict genes underlying complex traits and diseases. Nature Genetics 55 , 1267-1276 (2023). Lee, J.S. et al. Immunophenotyping of COVID-19 and influenza highlights the role of type I interferons in development of severe COVID-19. Sci Immunol 5 (2020). Melms, J.C. et al. A molecular single-cell lung atlas of lethal COVID-19. Nature 595 , 114-119 (2021). Cao, J. et al. The single-cell transcriptional landscape of mammalian organogenesis. Nature 566 , 496-502 (2019). Morabito, S., Reese, F., Rahimzadeh, N., Miyoshi, E. & Swarup, V. hdWGCNA identifies co-expression networks in high-dimensional transcriptomics data. Cell Rep Methods 3 , 100498 (2023). Szklarczyk, D. et al. The STRING database in 2023: protein-protein association networks and functional enrichment analyses for any sequenced genome of interest. Nucleic Acids Res 51 , D638-D646 (2023). Dong, S. et al. Annotating and prioritizing human non-coding variants with RegulomeDB v.2. Nat Genet 55 , 724-726 (2023). Kulakovskiy, I.V. et al. HOCOMOCO: a comprehensive collection of human transcription factor binding sites models. Nucleic Acids Res 41 , D195-202 (2013). Chen, H. et al. Activation of STAT6 by STING is critical for antiviral innate immunity. Cell 147 , 436-46 (2011). Seifert, L.L. et al. The ETS transcription factor ELF1 regulates a broadly antiviral program distinct from the type I interferon response. PLoS Pathog 15 , e1007634 (2019). Liu, T., Zhang, L., Joo, D. & Sun, S.C. NF-kappaB signaling in inflammation. Signal Transduct Target Ther 2 , 17023- (2017). Guan, H., Hou, S. & Ricciardi, R.P. DNA binding of repressor nuclear factor-kappaB p50/p50 depends on phosphorylation of Ser337 by the protein kinase A catalytic subunit. J Biol Chem 280 , 9957-62 (2005). Villar, D. et al. Enhancer evolution across 20 mammalian species. Cell 160 , 554-66 (2015). Pairo-Castineira, E. et al. Genetic mechanisms of critical illness in COVID-19. Nature 591 , 92-98 (2021). Farh, K.K. et al. Genetic and epigenetic fine mapping of causal autoimmune disease variants. Nature 518 , 337-43 (2015). Vuillard, L., Nicholson, J. & Hay, R.T. A complex containing betaTrCP recruits Cdc34 to catalyse ubiquitination of IkappaBalpha. FEBS Lett 455 , 311-4 (1999). Suzuki, H. et al. Homodimer of two F-box proteins betaTrCP1 or betaTrCP2 binds to IkappaBalpha for signal-dependent ubiquitination. J Biol Chem 275 , 2877-84 (2000). Fuchs, S.Y., Chen, A., Xiong, Y., Pan, Z.Q. & Ronai, Z. HOS, a human homolog of Slimb, forms an SCF complex with Skp1 and Cullin1 and targets the phosphorylation-dependent degradation of IkappaB and beta-catenin. Oncogene 18 , 2039-46 (1999). Oeckinghaus, A. & Ghosh, S. The NF-kappaB family of transcription factors and its regulation. Cold Spring Harb Perspect Biol 1 , a000034 (2009). Fericean, R.M. et al. Outcomes of Elderly Patients Hospitalized with the SARS-CoV-2 Omicron B.1.1.529 Variant: A Systematic Review. Int J Environ Res Public Health 20 (2023). Wang, S. et al. Genomic Surveillance for SARS-CoV-2 - China, September 26, 2022 to January 29, 2023. China CDC Wkly 5 , 143-151 (2023). Lam, M. et al. RICOPILI: Rapid Imputation for COnsortias PIpeLIne. Bioinformatics 36 , 930-933 (2020). Genomes Project, C. et al. A global reference for human genetic variation. Nature 526 , 68-74 (2015). Wang, K., Li, M. & Hakonarson, H. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res 38 , e164 (2010). Yang, J., Lee, S.H., Goddard, M.E. & Visscher, P.M. GCTA: a tool for genome-wide complex trait analysis. Am J Hum Genet 88 , 76-82 (2011). Purcell, S. et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 81 , 559-75 (2007). Quinlan, A.R. & Hall, I.M. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics 26 , 841-2 (2010). Buniello, A. et al. The NHGRI-EBI GWAS Catalog of published genome-wide association studies, targeted arrays and summary statistics 2019. Nucleic Acids Res 47 , D1005-D1012 (2019). Mak, T.S.H., Porsch, R.M., Choi, S.W., Zhou, X. & Sham, P.C. Polygenic scores via penalized regression on summary statistics. Genet Epidemiol 41 , 469-480 (2017). Weeks, E.M. et al. Leveraging polygenic enrichments of gene features to predict genes underlying complex traits and diseases. 2020.09. 08.20190561 (2020). Hao, Y. et al. Dictionary learning for integrative, multimodal and scalable single-cell analysis. 1-12 (2023). McInnes, L., Healy, J. & Melville, J.J.a.p.a. Umap: Uniform manifold approximation and projection for dimension reduction. (2018). Waltman, L. & Van Eck, N.J.J.T.E.p.j.B. A smart local moving algorithm for large-scale modularity-based community detection. 86 , 1-14 (2013). Fang, Z., Liu, X. & Peltz, G. GSEApy: a comprehensive package for performing gene set enrichment analysis in Python. Bioinformatics 39 (2023). Nishimura, D.J.B.S. & Scient, I.R.T.C.S.J.f. BioCarta. 2 , 117-120 (2001). Schaefer, C.F. et al. PID: the Pathway Interaction Database. Nucleic Acids Res 37 , D674-9 (2009). Martens, M. et al. WikiPathways: connecting communities. Nucleic Acids Res 49 , D613-D621 (2021). Jassal, B. et al. The reactome pathway knowledgebase. Nucleic Acids Res 48 , D498-D503 (2020). Gene Ontology, C. Gene Ontology Consortium: going forward. Nucleic Acids Res 43 , D1049-56 (2015). Fix, E. & Hodges, J.L.J.I.S.R.R.I.d.S. Discriminatory analysis. Nonparametric discrimination: Consistency properties. 57 , 238-247 (1989). Grant, C.E., Bailey, T.L. & Noble, W.S. FIMO: scanning for occurrences of a given motif. Bioinformatics 27 , 1017-8 (2011). Consortium, E.P. The ENCODE (ENCyclopedia Of DNA Elements) Project. Science 306 , 636-40 (2004). Additional Declarations No competing interests reported. Supplementary Files STables20250717.xlsx Supplementaryfiguresandmethods20250717.docx Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 25 Mar, 2026 Reviews received at journal 12 Mar, 2026 Reviews received at journal 11 Mar, 2026 Reviewers agreed at journal 15 Feb, 2026 Reviewers agreed at journal 13 Feb, 2026 Reviewers agreed at journal 09 Feb, 2026 Reviewers invited by journal 09 Feb, 2026 Editor assigned by journal 17 Nov, 2025 Submission checks completed at journal 13 Nov, 2025 First submitted to journal 13 Nov, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8104260","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":589720159,"identity":"2a7c6b2c-7ead-4d33-996c-5f6999550430","order_by":0,"name":"Can Hou","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Can","middleName":"","lastName":"Hou","suffix":""},{"id":589720162,"identity":"a3bc452b-3c95-40c4-989c-8873e5d28874","order_by":1,"name":"Yu Zeng","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Yu","middleName":"","lastName":"Zeng","suffix":""},{"id":589720164,"identity":"d6c7ab40-7374-4c54-8e64-2d1938424f02","order_by":2,"name":"Yudong Yin","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Yudong","middleName":"","lastName":"Yin","suffix":""},{"id":589720165,"identity":"a87ea258-df95-479e-884c-f9171c518e1b","order_by":3,"name":"Yanan Zhang","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Yanan","middleName":"","lastName":"Zhang","suffix":""},{"id":589720166,"identity":"eb5dd3e0-78e8-4110-99ff-3f7f747ba939","order_by":4,"name":"Lei Yang","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Lei","middleName":"","lastName":"Yang","suffix":""},{"id":589720168,"identity":"a20f5f13-24da-48a4-887e-1f309fbe20af","order_by":5,"name":"Yan Jiao","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Yan","middleName":"","lastName":"Jiao","suffix":""},{"id":589720170,"identity":"9058283d-fabd-44ee-bdb4-404a2c805ae9","order_by":6,"name":"Fei Yan","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Fei","middleName":"","lastName":"Yan","suffix":""},{"id":589720172,"identity":"19f09d09-747b-450d-9a2e-de87096bc51c","order_by":7,"name":"Huazhen Yang","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Huazhen","middleName":"","lastName":"Yang","suffix":""},{"id":589720173,"identity":"867e8ab1-c72f-4f41-a145-79dd96e26660","order_by":8,"name":"Wenwen Chen","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Wenwen","middleName":"","lastName":"Chen","suffix":""},{"id":589720176,"identity":"e0734ef9-a369-435e-9d4f-de2493ee84bb","order_by":9,"name":"Yuanyuan Qu","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Yuanyuan","middleName":"","lastName":"Qu","suffix":""},{"id":589720177,"identity":"ab8be016-814a-4d48-bc4b-2849db3359a4","order_by":10,"name":"Qian Xu","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Qian","middleName":"","lastName":"Xu","suffix":""},{"id":589720179,"identity":"f0690cf4-6e49-49c7-b75f-6792dbebb91c","order_by":11,"name":"Fengxiao Bu","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Fengxiao","middleName":"","lastName":"Bu","suffix":""},{"id":589720181,"identity":"617d3c5d-bd48-4850-89d6-bc8ae2a233c6","order_by":12,"name":"Unnur Valdimarsdóttir","email":"","orcid":"","institution":"University of Iceland","correspondingAuthor":false,"prefix":"","firstName":"Unnur","middleName":"","lastName":"Valdimarsdóttir","suffix":""},{"id":589720183,"identity":"3e0c58cd-165d-4919-9b82-dd0965aa79f6","order_by":13,"name":"Fang Fang","email":"","orcid":"","institution":"Karolinska Institutet","correspondingAuthor":false,"prefix":"","firstName":"Fang","middleName":"","lastName":"Fang","suffix":""},{"id":589720184,"identity":"74c776e2-8cac-4ceb-8c80-7ed3534810cc","order_by":14,"name":"Qian Li","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Qian","middleName":"","lastName":"Li","suffix":""},{"id":589720185,"identity":"a56b8d9d-ad5f-4bff-aa88-f9139e644821","order_by":15,"name":"Bowen Ke","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Bowen","middleName":"","lastName":"Ke","suffix":""},{"id":589720187,"identity":"30409fd4-52cb-4dd9-a118-9db74a0e8ecb","order_by":16,"name":"Jie Song","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Jie","middleName":"","lastName":"Song","suffix":""},{"id":589720189,"identity":"1966d2eb-4725-41e5-bd6f-37d2ccc073ed","order_by":17,"name":"Huan Song","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAzElEQVRIiWNgGAWjYJACZhDBz8DARqIWyQaStRgcIFaLfHvv4deFbXZ5xjeSnz38UXOYgX92A34tBmfOpVnPbEsuNruRZm7Mc+wwg8SdAwS0SOSYGfO2MSduu53DJs3YcBgokkDAYTPAWuoTN8/OYZP8SYwWhhs5xo952w4nbpDOYZPgJUaLwZkzZsw8544nzrj/zEya51g6j8QNQg5r7zH+zFNWndjfc/iZ5I8aazn+GYQcBox0CWQeD0H1QMD8gRhVo2AUjIJRMIIBALNmP1c6UqoSAAAAAElFTkSuQmCC","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":true,"prefix":"","firstName":"Huan","middleName":"","lastName":"Song","suffix":""}],"badges":[],"createdAt":"2025-11-13 09:53:22","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8104260/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8104260/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":102534028,"identity":"edaa4ab7-8791-4e52-845c-194d8ba957f2","added_by":"auto","created_at":"2026-02-12 16:58:19","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":481237,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eOverview of the study design, analyses and findings. a. \u003c/strong\u003eStudy population of genome-wide association analysis. The analysis was performed on 269 severe COVID-19 patients (cases) and non-severe COVID-19 patients (controls) from two Chinese cohorts, the China Surgery and Anesthesia Cohort (CSAC) and the China Severe Trauma Cohort (CSTC). A linkage disequilibrium (LD) independent risk locus at 8q22.1 (leadSNP: rs7817424) in the intronic region of protein-coding gene Carboxypeptidase Q (\u003cem\u003eCPQ\u003c/em\u003e) was identified to be significantly associated with COVID-19 severity (P \u0026lt; 5 × 10\u003csup\u003e−8\u003c/sup\u003e). \u003cstrong\u003eb.\u003c/strong\u003e Prioritization of \u003cem\u003eCPQ\u003c/em\u003e as the putative causal gene through a consensus-based gene mapping approach, including expression quantitative trait loci (eQTL) colocalization, protein quantitative trait loci (pQTL) colocalization, splicing quantitative trait loci (sQTL) colocalization, protein-altering variant (PAV) linkage and Polygenic Priority Score (PoPS) method. Symbol ‘√’ indicated the supporting evidence. \u003cstrong\u003ec. \u003c/strong\u003eCharacterizing the function of \u003cem\u003eCPQ\u003c/em\u003e in single-cell RNA sequencing (scRNA-seq) data. Tissue-resident macrophages and monocyte-derived macrophages were identified as the putative causal cell types for the CPQ gene by re-analyzing scRNA-seq data from lung tissues in severe COVID-19 patients and healthy controls. Results of multiple analyses suggested \u003cem\u003eCPQ\u003c/em\u003e pCPQ potentially play a role in severe Omicron symptoms via innate immune pathways.\u003cstrong\u003e d. \u003c/strong\u003eIdentification of transcription factor (TF) binding affecting variants and their target TFs. A series of analyses and rigorous selection procedures identified three variants in strong or moderate LD with the lead SNP that may affect the binding affinity of \u003cem\u003eSTAT6, ELF1 \u003c/em\u003eand \u003cem\u003eNFKB1. \u003c/em\u003e\u003cstrong\u003ee. \u003c/strong\u003eValidation of \u003cem\u003eCPQ\u003c/em\u003e function in human cells and mouse model. THP-1 cells stably overexpressing \u003cem\u003eCPQ \u003c/em\u003e(\u003cem\u003eCPQ \u003c/em\u003eOE) were generated by lentiviral transduction. Upon LPS stimulation, \u003cem\u003eCPQ\u003c/em\u003e OE cells showed reduced expression of inflammatory cytokines TNF-α and IL-6 compared to controls. In a mouse model of systemic inflammation, \u003cem\u003eCPQ\u003c/em\u003e expression was significantly downregulated in lung tissue. These results suggest that \u003cem\u003eCPQ \u003c/em\u003emay play an anti-inflammatory role in both in vitro and in vivo settings.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-8104260/v1/13ca81e6d14a0a38c775acda.png"},{"id":102534036,"identity":"073d3ccf-eb47-4678-add5-487185a0acf7","added_by":"auto","created_at":"2026-02-12 16:58:19","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":223066,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSummary of the Genome-Wide Association Study (GWAS) findings of Omicron COVID-19 severity within the Chinese population.\u003c/strong\u003e \u003cstrong\u003ea.\u003c/strong\u003e Manhattan plot displaying the result of the severe COVID-19 GWAS. The analysis utilized a generalized linear mixed model implemented in GCTA. The dashed horizontal line in the plot indicates the threshold for genome-wide significance (\u003cem\u003eP-\u003c/em\u003evalue = 5 × 10\u003csup\u003e−8\u003c/sup\u003e). \u003cstrong\u003eb.\u003c/strong\u003e Q-Q plot showing the distributions of the observed to expected \u003cem\u003eP\u003c/em\u003e-values. \u003cstrong\u003ec. \u003c/strong\u003eRegional plot of the COVID-19 GWAS with \u003cem\u003eP\u003c/em\u003e-values for variants surrounding the risk locus and their linkage disequilibrium (LD) with lead SNP rs7817424 based on the East Asian reference panel from1000 Genome project phrase 3. The region has been annotated using 15 core chromatin states from lung tissue data derived from the NIH Roadmap Epigenomics project. These states include enhancers (Enh), quiescent/low (Quies), heterochromatin (Het), active transcription start site (TssA), flanking active transcription start site (TssAFlnk), weak transcription (TxWk), ZNF genes \u0026amp; repeats (ZNF/Rpts), strong transcription (Tx), and genic enhancers (EnhG). Genomic positions are based on the human genome assembly hg19.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-8104260/v1/91d10541cd484fb7d4b55ed1.png"},{"id":102746409,"identity":"c60f90cc-9fe5-4c73-a20b-b991bb980468","added_by":"auto","created_at":"2026-02-16 08:57:32","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":251729,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eConsensus-based gene mapping\u003c/strong\u003e. \u003cstrong\u003ea.\u003c/strong\u003e Convergence of evidence from five distinct analyses in consensus-based gene mapping, all pointing to \u003cem\u003eCPQ\u003c/em\u003e as the putative causal gene for the risk locus. \u003cstrong\u003eb.\u003c/strong\u003e Regional plot of the COVID-19 GWAS, lung expression quantitative trait loci (eQTL), blood eQTL, and protein quantitative trait loci (pQTL) p-values for variants surrounding the risk locus and their linkage disequilibrium (LD) with the lead variants based on the East Asian (for COVID-19 GWAS) or European (for eQTL and pQTL) reference panel. The lung and blood eQTL GWAS summary data were derived from the GTEx V8, while the pQTL GWAS summary data were derived from the AGES cohort. The posterior probabilities for the presence of a single variant responsible for both COVID-19 GWAS and eQTL/pQTL signals (i.e., PP\u003csub\u003eshared\u003c/sub\u003e from colocalization analysis) are provided in the upper right corner of each regional plot.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-8104260/v1/4687b41100c14684180bf041.png"},{"id":102534031,"identity":"ab3d2cb0-26fa-48f1-949b-5c9c18faf134","added_by":"auto","created_at":"2026-02-12 16:58:19","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":582503,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSingle-cell sequencing data analyses for characterizing the function of \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eCPQ.\u003c/strong\u003e\u003c/em\u003e \u003cstrong\u003ea. \u003c/strong\u003eUniform Manifold Approximation and Projection (UMAP) plot illustrating the cell type clusters in lung single-cell RNA sequencing (scRNA-seq) data obtained from severe COVID-19 patients and healthy controls.\u003cstrong\u003e b.\u003c/strong\u003e Dot plot displaying the expression levels of the \u003cem\u003eCPQ\u003c/em\u003e gene across 21 different cell types within the lung scRNA-seq data. Seven cell types exhibiting a significant up-regulation of \u003cem\u003eCPQ\u003c/em\u003e gene expression after Bonferroni correction are highlighted in red. \u003cstrong\u003ec. \u003c/strong\u003eUMAP plot illustrating the cell type clusters in peripheral blood mononuclear cells (PBMCs) scRNA-seq data obtained from severe COVID-19 patients and healthy controls.\u003cstrong\u003e d. \u003c/strong\u003eDot plot displaying the expression levels of the \u003cem\u003eCPQ\u003c/em\u003e gene across 11 cell types within the PBMCs scRNA-seq data. Five cell types exhibiting a significant up-regulation of \u003cem\u003eCPQ\u003c/em\u003e gene expression after Bonferroni correction are highlighted in red.\u003cstrong\u003e e.\u003c/strong\u003e Expression dynamics of the \u003cem\u003eCPQ\u003c/em\u003e gene during the transition process from monocytes to monocyte-derived macrophages (MDM) within lung tissues from severe COVID-19 patients. Each dot within the plot represents either a monocyte, MDM, or transitioning MDM cell. The x-axis displays pseudotime values, indicating the relative progression in the transition process, while the y-axis reflects the normalized \u003cem\u003eCPQ\u003c/em\u003e gene expression levels. The curve in the plot represents the non-linear relationship between pseudotime and \u003cem\u003eCPQ\u003c/em\u003e expression levels, fitted using a negative binomial regression model with cubic splines. \u003cstrong\u003ef. \u003c/strong\u003eUMAP lower-dimensional manifold of co-expression network of genes within the putative causal cell types found in lung tissues of severe COVID-19 patients. Each dot in the plot represents a gene, with the distance between genes indicating their degree of co-expression. The color of each dot corresponds to the specific module to which the gene belongs. \u003cstrong\u003eg.\u003c/strong\u003eEnriched terms for the 20 nearest genes to \u003cem\u003eCPQ\u003c/em\u003e within the co-expression network.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-8104260/v1/612b8caf34ada30fd56398d1.png"},{"id":102534034,"identity":"c7155db9-55ca-46e5-8def-b60107c3eb80","added_by":"auto","created_at":"2026-02-12 16:58:19","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":538533,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAnalyses for identifying putative risk variants influencing transcription factor (TF) binding affinity and the target TFs\u003c/strong\u003e. \u003cstrong\u003ea. \u003c/strong\u003eOverview of the analysis procedures, including the number of eligible risk variants and TFs remained after each analysis. \u003cstrong\u003eb and c.\u003c/strong\u003e The left plot presents the expression levels of the selected TFs across 21 distinct cell types in the lung single-cell RNA sequencing (scRNA-seq) data. These TFs were predicted to exhibit increased or decreased binding affinity at genomic regions surrounding the risk variants when risk alleles are present, as indicated by TF binding motif analyses. TFs demonstrating significant up-regulation compared to background genes within the putative causal cell types (i.e., monocytes and macrophages) are considered to have confirmed functional relevance and are highlighted in red. For the selected TFs with confirmed functional relevance, the right plot illustrates their potentially altered TF-binding motifs from HOCOMOCD database (version 11). Below these motifs, genome sequences according to GRCh38 are presented, with the locations of the risk variants and their alternative alleles (which are also risk alleles) highlighted. Additionally, we provide the TF chromatin immunoprecipitation-sequencing (ChIP-seq) signals within the regions surrounding the risk variants (positioned in the middle) for K562 cells, presented below the genome sequence for each TF. The ChIP-seq data used for visualization were obtained from the ENCODE project.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-8104260/v1/8af0926acdeea5b6055e6055.png"},{"id":102534032,"identity":"dd680624-11dc-47fd-acca-280a3574e42f","added_by":"auto","created_at":"2026-02-12 16:58:19","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":207589,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eOverexpression of carboxypeptidase Q inhibits inflammation.\u003c/strong\u003e \u003cstrong\u003ea\u003c/strong\u003e. TNF-α and \u003cstrong\u003eb. \u003c/strong\u003eIL-6\u003cstrong\u003e \u003c/strong\u003eexpression levels measured by ELISA. THP-1 wild-type (WT) and carboxypeptidase Q (CPQ) overexpression (CPQ OE) cells (1 × 10⁶ cells per ml), seeded in six-well plates, were treated with LPS (100 ng ml⁻¹ or 1600 ng ml⁻¹) for 24 h (n = 3).\u003cstrong\u003ec and d.\u003c/strong\u003e Western blot analysis of TNF-α expression under conditions identical to (a) (20 μg per lane; n = 3). \u003cstrong\u003ee and\u003c/strong\u003e \u003cstrong\u003ef\u003c/strong\u003e Western blot analysis of IL-6 expression under conditions identical to \u003cstrong\u003ea\u003c/strong\u003e (20 μg per lane; n = 3). \u003cstrong\u003ed\u003c/strong\u003e. TNF-α and \u003cstrong\u003ef\u003c/strong\u003e. IL-6 expression levels normalized to \u003cem\u003eβ\u003c/em\u003e-actin. Grayscale analysis was performed using Image J software. \u003cstrong\u003eg and h\u003c/strong\u003e. Systemic inflammation was induced in mice via intraperitoneal injection of LPS (10 mg per kg) for 6 h (model group). Lung tissues from wild-type and model mice were harvested for analysis. CPQ expression in lung tissue was determined by western blot using anti-PGCP antibody (20 μg per lane), normalized to \u003cem\u003eβ\u003c/em\u003e-actin (n = 7). Grayscale analysis was performed using Image J software. Data are shown as the mean of three independent experiments ± s.e.m. P values were determined by unpaired two-tailed t-tests. NS, not significant.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-8104260/v1/823df2644bfb0b7511ed4898.png"},{"id":102962142,"identity":"fe172c90-cb4b-4571-94c7-5a6b4a243203","added_by":"auto","created_at":"2026-02-19 04:03:31","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3629968,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8104260/v1/487c4200-f342-423b-aad9-8ae89c40518d.pdf"},{"id":102534030,"identity":"0a5616d8-a023-4a03-8439-46381eff023c","added_by":"auto","created_at":"2026-02-12 16:58:19","extension":"xlsx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":285916,"visible":true,"origin":"","legend":"","description":"","filename":"STables20250717.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8104260/v1/8251a25dcd587a17caed2a85.xlsx"},{"id":102534033,"identity":"1c34c83d-c6bf-45cd-98d3-8f37cf2a7ad4","added_by":"auto","created_at":"2026-02-12 16:58:19","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":2061702,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementaryfiguresandmethods20250717.docx","url":"https://assets-eu.researchsquare.com/files/rs-8104260/v1/492265158a8c8524ffc59c2e.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Genetic and functional evidence identify Carboxypeptidase Q as a regulator of inflammation in respiratory infection","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eInfections, in particular respiratory infections, are one of the major causes of deaths worldwide\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. While environmental exposure and pathogen characteristics play important roles, accumulating evidence suggests that host genetic factors substantially influence both susceptibility to infection and disease severity\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. For example, in the context of infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causative agent of coronavirus disease 2019 (COVID-19), where some individuals experience only mild symptoms while others develop life-threatening complications. Dysregulated immune responses, including hyperinflammatory states, have been implicated in severe infection consequences\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Genome-wide association studies (GWAS) have since identified multiple common genetic variants associated with critical illness in COVID-19, shedding light on host biological pathways involved in antiviral defense and immune regulation\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eHowever, prior to the COVID-19 pandemic, GWAS of infectious diseases were relatively were relatively few and generally limited in scale and resolution\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. This was due in part to challenges such as pathogen heterogeneity, variable exposure levels, and the lack of large, well-phenotyped cohorts. For example, polymorphisms at the \u003cem\u003eIFNL4\u003c/em\u003e/\u003cem\u003eIFNL3\u003c/em\u003e locus have been shown to influence treatment response differentially across hepatitis C virus genotypes, underscoring how pathogen variation can confound host genetic associations\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. Another major challenge is pathogen exposure misclassification, which violates the assumption that all individuals in a study are equally at risk. In reality, individuals who have never been exposed to a pathogen are not susceptible to infection, and exposure risk itself is shaped by a complex interplay of environmental and genetic factors, including reinfection\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. Moreover, most existing studies have been conducted in populations of predominantly European ancestry, limiting the generalizability of their findings to other ethnic groups. As such, there remains a critical need to conduct well-powered genetic studies in diverse populations using carefully defined clinical phenotypes and uniform pathogen exposure.\u003c/p\u003e \u003cp\u003eTo address these gaps, we conducted a GWAS in a relatively homogeneous Chinese population consisting of 5,151 individuals who were newly infected with the SARS-CoV-2 Omicron variant since January 2022. During the massive Omicron outbreak in China following the relaxation of COVID-19 restrictions in late 2022, it was estimated that approximately 97% of the population became infected\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. This context offers a natural, near-universal exposure model, thereby minimizing confounding from differential pathogen exposure, prior immunity, or viral strain heterogeneity.\u003c/p\u003e"},{"header":"RESULTS","content":"\u003cp\u003e\u003cstrong\u003eGenome-wide association\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe conducted a GWAS by combining samples from two ongoing multicenter longitudinal cohorts in China: The China Surgery and Anesthesia Cohort (CSAC) and the China Severe Trauma Cohort (CSTC)\u003csup\u003e13,14\u003c/sup\u003e. COVID-19 related information was collected for both cohorts since March 2023 with standardized questionnaire (Supplementary Methods) measuring infection status, medical treatment received, and symptoms experienced during and after infection. By June 2023, 5,151 participants from the CSAC and CSTC cohorts (3,757 and 1,394, respectively) had both phenotype data on COVID-19 severity and genotyping data that passed quality control (QC) procedures (Figure 1 and Methods). Of these, 269 reported first-time infection and severe manifestations of COVID-19 (i.e., underwent inpatient medical care or received medical interventions at primary care centers, Supplementary Table 1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe identified two single-nucleotide polymorphisms (SNPs) that reached genome-wide significance level of \u003cem\u003eP\u003c/em\u003e-value\u0026lt;5\u0026times;10\u003csup\u003e\u0026minus;8\u003c/sup\u003ein the 8q22.1 genomic region (Figure 2a and Supplementary Table 2). The quantile-quantile (QQ) plot (Figure 2b) and the genomic inflation factor (\u0026lambda;=0.998) indicated minimal systematic biases. A high level of concordance between genotyping and whole genome sequencing (WGS) was observed for sub-threshold SNPs (i.e., \u003cem\u003eP\u003c/em\u003e-value\u0026lt;5\u0026times;10\u003csup\u003e\u0026minus;6\u003c/sup\u003e) in a subset of randomly selected 498 patients (concordance rate ranging from 95.18% to 99.60%; Supplementary Table 3), underscoring the substantial accuracy of both genotyping and imputation techniques employed within the current GWAS.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eClumping of these identified significant SNPs led to the identification of a single linkage‑disequilibrium (LD) independent risk locus located within the protein-coding gene \u003cem\u003eCarboxypeptidase Q\u003c/em\u003e (\u003cem\u003eCPQ\u003c/em\u003e). The lead SNP, rs7817424 (\u003cem\u003eP\u003c/em\u003e-value=4.60\u0026times;10\u003csup\u003e\u0026minus;8\u003c/sup\u003e), is located within the intronic region of this gene (Figure 2c). We next examined genome-wide significant signals within \u0026plusmn;500\u0026thinsp;kb of this risk locus in prior GWAS. A total of 25 distinct traits were identified as significantly associated with variants within this genomic region (Supplementary Table 4). Among these traits, six were related to the respiratory or hematological systems, including the lung function, chronic obstructive pulmonary disease, mean corpuscular hemoglobin, mean corpuscular volume, red cell distribution width and acute myeloid leukemia.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAssociation between severity by Omicron variant and polygenic risk scores (PRS) using previous COVID-19 GWAS\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe performed polygenic risk score (PRS) analyses to evaluate the validity of our definition of COVID-19 severity and to compare our results with prior COVID-19 GWAS. GWAS summary statistics of relevant COVID-19 outcomes conducted in East Asian population from the COVID-19 Host Genetics Initiative study (release 7) were used as the discovery set \u003csup\u003e15\u003c/sup\u003e. We found marginal significant associations between PRSs and the COVID-19 severity. Specifically, for the PRS of critical COVID-19 (versus controls without COVID-19), the odd ratio (OR) was 1.12 (95% confidence interval [CI]: 0.99-1.27, \u003cem\u003eP\u003c/em\u003e-value=0.07). Similarly, for the PRS of hospitalized COVID-19 (versus controls without COVID-19), the OR was 1.11 (95% CI: 0.98-1.25, \u003cem\u003eP\u003c/em\u003e-value=0.09). These findings, while not achieving statistical significance, do provide supporting evidence for the comparability between our GWAS on COVID-19 severity and previous research efforts.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEnrichment of the genome-wide significant SNPs in chromatin states\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAs all the significant SNPs within the identified risk locus are positioned in non-coding regions, we further assessed their capacity for regulating gene transcription by examining their enrichments across 15 distinct chromatin states\u0026nbsp;from the Roadmap Epigenomics project (Methods)\u003csup\u003e16\u003c/sup\u003e, each representing diverse regulatory elements. Using a hypergeometric distribution test, we observed strong enrichment signals in chromatin states associated with enhancers, weak transcription, weak polycomb repressor complex, and heterochromatin across various blood and lung tissues/cells (Supplementary Figure 1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePutative risk gene mapping\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe employed a consensus-based gene mapping approach to identify potential risk genes associated with the identified risk locus\u003csup\u003e17,18\u003c/sup\u003e. This approach combined five analyses: expression quantitative trait loci (eQTL) colocalization, protein quantitative trait loci (pQTL) colocalization, splicing quantitative trait loci (sQTL) colocalization, protein-altering variant (PAV) linkage and similarity-based gene prioritization (Methods). Except for PAV linkage, the other four analyses prioritized \u003cem\u003eCPQ\u003c/em\u003e gene as the putative causal candidate for the risk locus (Figure 3a). Specifically, eQTL colocalization analyses using coloc\u003csup\u003e19\u003c/sup\u003e uncovered significant colocalization (posterior probability [PP] for colocalization \u0026ge;0.8) between GWAS signals and eQTL signals of \u003cem\u003eCPQ\u003c/em\u003e in 17 out of 49 Genotype-Tissue Expression (GTEx) tissues\u003csup\u003e20\u003c/sup\u003e (Supplementary Table 5), including lung and whole blood (PP=0.89 and 0.89; Figure 3b). pQTL colocalization analyses underscored \u003cem\u003eCPQ\u003c/em\u003e as the only gene where GWAS signals and serum pQTL signals from previous proteomic GWAS\u003csup\u003e21\u003c/sup\u003e exhibited significant colocalization (PP=0.90; Figure 3b and Supplementary Table 6). Importantly, the direction of effect for the risk allele was consistent at both transcription and protein levels, suggesting that the risk locus might increase susceptibility to severe COVID-19 by up-regulating \u003cem\u003eCPQ\u003c/em\u003e expression. Additionally, sQTL colocalization analysis yielded 32 variant-splice pairs with a PP\u0026ge;0.8 across 12 GTEx tissues, all implicating \u003cem\u003eCPQ\u003c/em\u003e gene (Supplementary Table 7). PAV linkage analyses revealed that none of the variants in strong or moderate LD with the lead variant were predicted to be loss of function or missense variants (Supplementary Table 8)\u003csup\u003e22\u003c/sup\u003e. Finally, by integrating GWAS signals with gene features from diverse sources, similarity-based gene prioritization using Polygenic Priority Score (PoPS) method robustly identified \u003cem\u003eCPQ\u003c/em\u003e as the most plausible causal gene proximal to the risk locus (Supplementary Table 9)\u003csup\u003e23\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCharacterizing the function of \u003cem\u003eCPQ\u003c/em\u003e through single-cell RNA sequencing (scRNA-seq) data\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo further investigate the potential role of \u003cem\u003eCPQ\u003c/em\u003e in SARS‑CoV‑2 infection, we re-analyzed scRNA-seq data from lung tissues and peripheral blood mononuclear cells (PBMCs) in severe COVID-19 patients and healthy controls from two published studies\u003csup\u003e24,25\u003c/sup\u003e. Through differential gene expression (DGE) analyses, we observed a significant up-regulation of \u003cem\u003eCPQ\u003c/em\u003e across various cell types in lung (Figure 4a-b), including alveolar macrophages, monocyte-derived macrophages (MDM), and the precursor cells of MDM (i.e., monocytes and transitioning MDM), all of which directly participate in the innate immune response to SARS‑CoV‑2 infection. DGE analyses by disease status further revealed notably higher \u003cem\u003eCPQ\u003c/em\u003e expression levels in healthy controls compared to severe COVID-19 patients across all the aforementioned cell types (Supplementary Figure 2). In PBMCs, \u003cem\u003eCPQ\u003c/em\u003e also demonstrated significant up-regulation in three monocyte subsets as well as monocyte-derived Dendritic cells (Figure 4c-d), but no significant difference was observed in their expression levels between severe COVID-19 patients and healthy controls (Supplementary Figure 3). Collectively, these findings underscore both tissue-resident macrophages and MDM along with its immature precursor cells in lung as the putative causal cell types for \u003cem\u003eCPQ\u003c/em\u003e gene.\u003c/p\u003e\n\u003cp\u003eWe observed variations in \u003cem\u003eCPQ\u003c/em\u003e expression levels among MDM and its precursor cells, monocytes and transitioning MDM, in severe COVID-19 patients (Supplementary Figure 2), suggesting a potential modulation of \u003cem\u003eCPQ\u003c/em\u003e expression during the transition from monocytes to MDM in the context of SARS‑CoV‑2 infection. We then investigated the dynamic changes in \u003cem\u003eCPQ\u003c/em\u003e expression throughout this transition process. Employing the pseudotime trajectory analysis approach implemented in Monocle3\u003csup\u003e26\u003c/sup\u003e, we assigned each cell a pseudotime value, representing its relative progress in the transition process (Supplementary Figure 4). Regression of \u003cem\u003eCPQ\u003c/em\u003e expression on pseudotime revealed a distinct U-shaped trajectory (Figure 4e), with initial down-regulation during the transition and subsequent recovery post-transition. These results suggest the involvement of \u003cem\u003eCPQ\u003c/em\u003e in both monocyte and MDM-specific functions.\u003c/p\u003e\n\u003cp\u003eFocusing on the plausible causal cell types in lung, we conducted co-expression network analysis using hdWGCNA\u003csup\u003e27\u003c/sup\u003e among COVID-19 patients to identify genes co-expressed with \u003cem\u003eCPQ\u003c/em\u003e. Nine co-expression modules were identified, comprising over 2,900 genes (Figure 4f and Supplementary Table 10). Enrichment analyses of the top 100 genes in each module revealed distinct biological functions (Supplementary Table 11). Co-expression module 1, for instance, showed enrichment in protein synthesis and translation pathways, while module 8 (containing \u003cem\u003eCPQ\u003c/em\u003e) exhibited enrichment mainly in pathways associated with GTPase activity regulation and signal transduction. We extracted the 20 closest neighboring genes of \u003cem\u003eCPQ\u003c/em\u003e in the co-expression network (representing genes with the highest co-expression correlation with \u003cem\u003eCPQ\u003c/em\u003e; Figure 4f) and performed enrichment analyses. Ten enriched pathways emerged, with five related to protein catabolism and localization, involving \u003cem\u003eCPQ\u003c/em\u003e and six other genes (Figure 4g and Supplementary Table 12). Apart from \u003cem\u003eCPQ\u003c/em\u003e, all six other genes related to protein catabolism and localization were also linked to ubiquitin-dependent protein catabolic processes (Supplementary Table 13). Protein-protein interaction network analysis of these six ubiquitination-related genes using STRING\u003csup\u003e28\u003c/sup\u003e revealed functional connections among \u003cem\u003eUSP48\u003c/em\u003e (formerly known as \u003cem\u003eUSP31\u003c/em\u003e), \u003cem\u003eGSK3B\u003c/em\u003e, and \u003cem\u003eFBXW11\u003c/em\u003e (Supplementary Figure 5), with \u003cem\u003eGSK3B\u003c/em\u003e and \u003cem\u003eFBXW11\u003c/em\u003e identified as the top two closest genes to \u003cem\u003eCPQ\u003c/em\u003e in the co-expression network.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIdentification of transcription factor (TF) binding affecting variants and their target TFs\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAs evidence derived from consensus-based gene mapping and SNP enrichment analyses collectively suggest the variants within the risk locus may alter gene transcription by disrupting or enhancing TF binding, we then conducted a series of analyses to identify candidate variants affecting TF binding and their target TFs (Figure 5a). Among the 109 variants in strong or moderate LD (r\u003csup\u003e2\u003c/sup\u003e\u0026ge;0.6) with the lead variant, 64 exhibited high regulatory potential (RegulomeDB score of 1f or higher, Supplementary Table 14)\u003csup\u003e29\u003c/sup\u003e. Motif scanning of\u0026nbsp;\u0026plusmn;20 bp sequences flanking these eligible variants against the HOCOMOCD database\u003csup\u003e30\u003c/sup\u003e unveiled 28 variant-TF pairs with significantly increased binding affinity for risk alleles (Supplementary Table 15). Furthermore, we also found 25 variant-TF pairs exhibited a significant decrease in TF binding affinity for the risk alleles (Supplementary Table 16). To focus on TFs active in the putative causal cell types (i.e., monocytes and macrophages in lung), we compared their expression levels with that of background genes in these cell types. We identified a total of six TFs that were notably up-regulated in these cell types, including four TFs with likely enhanced binding affinity (\u003cem\u003eSTAT6\u003c/em\u003e, \u003cem\u003eREL\u003c/em\u003e, \u003cem\u003eRUNX1\u003c/em\u003e, and \u003cem\u003eELF1\u003c/em\u003e; Figure 5b) and two TFs with likely disrupted binding affinity (\u003cem\u003eFOXP1\u003c/em\u003e and \u003cem\u003eNFKB1\u003c/em\u003e; Figure 5c). Among these TFs, three (i.e., \u003cem\u003eSTAT6\u003c/em\u003e, \u003cem\u003eELF1\u003c/em\u003e, and \u003cem\u003eNFKB1\u003c/em\u003e) showed TF chromatin immunoprecipitation sequencing (ChIP-seq) peak signals near risk variants (rs3802194, rs7844514, and rs73698822, respectively) in K562 cells (Figure 5b-c). Both \u003cem\u003eSTAT6\u003c/em\u003e and \u003cem\u003eELF1\u003c/em\u003e are widely recognized as important transcriptional activators in the innate immune response to virus infection\u003csup\u003e31,32\u003c/sup\u003e, and their increased binding affinity when risk alleles are present aligns with findings of associations between risk alleles and heightened \u003cem\u003eCPQ\u003c/em\u003e expression. The third TF, \u003cem\u003eNFKB1\u003c/em\u003e, is the DNA binding subunit of NF-\u0026kappa;B and act as both a transcriptional activator (when forming a heterodimer with \u003cem\u003eRELA\u003c/em\u003e, \u003cem\u003eRELB\u003c/em\u003e, or \u003cem\u003eREL\u003c/em\u003e protein)\u003csup\u003e33\u003c/sup\u003e and a repressor (when forming a homodimer with \u003cem\u003eNFKB2\u003c/em\u003e protein)\u003csup\u003e34\u003c/sup\u003e under different conditions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eValidation of CPQ function in human cells and mouse model\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo validate the function of \u003cem\u003eCPQ\u003c/em\u003e, a THP-1 cell line stably overexpressing \u003cem\u003eCPQ\u003c/em\u003e was generated. Western blotting (WB) analysis confirmed significantly higher CPQ expression in these cells compared to wild-type (WT) controls (Supplementary Figure 6a). An inflammatory model was established in THP-1 cells using lipopolysaccharide (LPS) stimulation and optimizing conditions through time course (Supplementary Figure 6b) and concentration gradient (Supplementary Figure 6c) experiments. Based on these results, stimulation with 100 ng/mL LPS for 24 hours was selected as the standard inflammatory stimulus, while 1600 ng/mL was chosen for high-concentration comparison. ELISA results (Figure 6a-b) demonstrated significantly lower expression of inflammatory cytokines TNF-\u0026alpha; and IL-6 in \u003cem\u003eCPQ\u003c/em\u003e overexpression cells versus WT cells under high-concentration LPS stimulation (1600 ng/mL). Consistent with this, WB analysis (Figure 6c-f) further confirmed reduced cytokine expression in \u003cem\u003eCPQ\u003c/em\u003e overexpression cells.\u003c/p\u003e\n\u003cp\u003eTo investigate \u003cem\u003eCPQ\u003c/em\u003e\u0026apos;s role in vivo, a systemic inflammation model was established in mice through intraperitoneal LPS injection (10 mg/kg, 6 hours). Successful model induction was confirmed by elevated inflammatory factor expression in lung tissue (Supplementary Figure 6d). Simultaneously assessed \u003cem\u003eCPQ\u003c/em\u003e expression levels in lung tissue were significantly lower in LPS-treated mice compared to normal controls (Figure 6g-h), suggesting that inflammatory responses may negatively regulate \u003cem\u003eCPQ\u003c/em\u003e expression.\u003c/p\u003e\n\u003cp\u003eCollectively, results from both human cellular and murine models indicate that \u003cem\u003eCPQ\u003c/em\u003e exerts potential anti-inflammatory effects during inflammatory responses.\u003c/p\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eOur study, leveraging genetic data from a relatively homogeneous Chinese population, reveals a previously unrecognized immunomodulatory function of \u003cem\u003eCPQ\u003c/em\u003e and highlight it as a potential therapeutic target for mitigating severe outcomes of COVID-19 and other respiratory infections. Through GWAS, we discovered two linked SNPs located at 8q22.1 that reached genome-wide significance. Our integrated analyses, including consensus-based gene mapping, similarity-based prioritization, and regulatory annotation, implicate the \u003cem\u003eCPQ\u003c/em\u003e gene as the likely causal gene and suggest that risk variants may influence \u003cem\u003eCPQ\u003c/em\u003e expression through regulation of gene transcription. Using scRNA-seq data, we observed differential expression of \u003cem\u003eCPQ\u003c/em\u003e in lung tissues between COVID-19 patients and healthy controls. Notably, tissue-resident macrophages and MDMs along with their immature precursors, emerged as putative causal cell types. In these cells, motif-based predictions indicate that risk variants alter binding affinities of transcriptional activators and repressors that are known to play key roles in anti-infection responses. Supporting these observations, functional experiments in human THP-1 cells and systemic inflammation mouse models confirmed the anti-inflammatory role of \u003cem\u003eCPQ\u003c/em\u003e, suggesting its involvement in modulating hyperinflammatory responses during severe respiratory infections.\u003c/p\u003e \u003cp\u003eThe non-coding GWAS variants were found to have significant enrichments across several chromatin states including enhancers, across various blood and lung tissues/cells, which is consistent with GWAS results from other infectious and inflammatory diseases including COVID-19\u003csup\u003e35\u0026ndash;37\u003c/sup\u003e. Furthermore, the risk variants were suggested to alter the binding affinity of specific TFs, including those that were likely to be expressed in the putative causal cell types identified by scRNA-seq data analyses. These findings underscore the regulatory potential of these significant SNPs within tissues relevant to the studied outcome.\u003c/p\u003e \u003cp\u003eThe \u003cem\u003eCPQ\u003c/em\u003e gene, where the associated variants locate, is identified as the potential risk gene from converging evidence. \u003cem\u003eCPQ\u003c/em\u003e is a protein coding gene belongs to the peptidase M28 family, primarily responsible for catalyzing unsubstituted dipeptide cleavage into amino acids. Although it has not been reported in any prior genetic studies for infectious diseases, we observed its significant up-regulation across various cell types that directly participate in the immune response to SARS‑CoV‑2 infection (i.e., monocyte clusters and alveolar macrophages) in lung tissues and PBMCs using scRNA-seq data. The expression level of \u003cem\u003eCPQ\u003c/em\u003e differed in lung tissues between COVID-19 patients and healthy controls, further implying its role in host immune response. Subsequent analyses focused on these putative causal cell types using gene co-expression network, followed by gene set enrichment and protein-protein interaction network, have identified several co-expression modules and distinct biological functions, including pathways associated with protein synthesis and translation, GTPase activity regulation, as well as protein catabolism and localization.\u003c/p\u003e \u003cp\u003eFinally, converging evidence from cellular and animal models supporting a potential anti-inflammatory function of CPQ that is not restricted to Omicron infection. In human THP-1 cells, \u003cem\u003eCPQ\u003c/em\u003e overexpression resulted in lower levels of TNF-α and IL-6 under high-dose LPS stimulation, demonstrated by ELISA and Western blot analyses, suggesting that \u003cem\u003eCPQ\u003c/em\u003e may attenuate the production of inflammatory cytokines under acute inflammatory challenge. Furthermore, systemic inflammation induced by LPS led to a significant downregulation of \u003cem\u003eCPQ\u003c/em\u003e expression in lung tissue in mice, probably due to the negative feedback suppression of \u003cem\u003eCPQ\u003c/em\u003e during heightened inflammatory states. These results provided mechanistic insights that align with the genetic association between \u003cem\u003eCPQ\u003c/em\u003e and severe COVID-19 observed in our GWAS, highlighting its previously unrecognized role in modulating inflammatory pathways.\u003c/p\u003e \u003cp\u003eHow \u003cem\u003eCPQ\u003c/em\u003e participate in the immune response regulation remains unknown. Intriguingly, our results from the protein-protein interaction network and the TF binding analyses suggest that \u003cem\u003eCPQ\u003c/em\u003e may participate in the NF-κB pathway activation in the regulatory process of host immune response against infection, likely through involvement in the ubiquitin-proteasome system to degrade the NF-κB inhibitor. NF-κB is a family of transcription factors that play critical roles in inflammation and immunity. The two genes \u003cem\u003eGSK3B\u003c/em\u003e and \u003cem\u003eFBXW11\u003c/em\u003e prioritized as the top two nearest genes to \u003cem\u003eCPQ\u003c/em\u003e have been demonstrated to play crucial roles in activating NF-κB signal pathway, by marking the inhibitor of NF-κB (i.e., IκB) for proteasome degradation through mediating its ubiquitination\u003csup\u003e\u003cspan additionalcitationids=\"CR39\" citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. In addition, four other genes with the highest co-expression with \u003cem\u003eCPQ\u003c/em\u003e are also involved in ubiquitin-dependent protein catabolism. In addition, among the TFs that were targets of the variants in LD with the lead variant, \u003cem\u003eNFKB1\u003c/em\u003e is a DNA binding subunit of NF-κB. It could associate with other NF-κB proteins to form distinct dimeric complexes that alter gene regulation in a variety of ways to control the overall levels of transcripts\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. However, as we found no prior studies on \u003cem\u003eCPQ\u003c/em\u003e and NF-κB pathway, the underlying mechanism remains unknown, and our hypothesis needs to be tested in future studies.\u003c/p\u003e \u003cp\u003eThe major strength of our study is its focus on uncovering the genetic determinants and underlying mechanisms that modulate the host response to respiratory infections, using Omicron infection as an optimal and timely model. Unlike prior GWAS conducted during early phase of the COVID-19 pandemic, when participants were infected with a mixture of SARS-CoV‐2 variants, resulting in heterogenous immune responses and clinical outcomes\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e, our study leverages a relatively homogeneous Chinese population with first-time Omicron infection. This uniform exposure minimizes confounding from prior immunity or variant-related differences, providing a uniquely controlled setting to investigate host genetic contributions to infection severity. Importantly, most individuals in our two cohorts had not been previously exposed to SARS-CoV-2, enabling a clearer dissection of primary immune responses. Another major strength lies in the convergence of evidence from multiple complementary analytic strategies, including GWAS, risk gene prioritization, regulatory annotation, single-cell expression analysis, and functional experimental validation, which together provide robust and mutually reinforcing support for the role of \u003cem\u003eCPQ\u003c/em\u003e in modulating infection severity. These analytic steps also underscore how a well-defined, variant-specific infection context can serve as a powerful model for dissecting the host genetic architecture of respiratory infection outcomes more broadly.\u003c/p\u003e \u003cp\u003eWe also acknowledge several limitations in this study. First, no lab detection has been performed to identify the specific type of SARS-CoV-2 variants in the COVID-19 cases in our cohorts. Despite this, genomic surveillance study has shown that from September 2022 to January 2023, 99.5% of COVID-19 cases in China were infected with Omicron variant\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. This support the validity of our assumption that vast majority of participants in our cohorts were infected by the Omicron. Second, this study was based on two clinic cohorts comprising patients undergoing surgical or trauma-related care, which may not be representative of the general population. Therefore, caution is needed when generalizing our findings to broader demographic groups. However, our downstream analyses, including those integrating publicly available single-cell sequencing datasets and results from functional experiments, provide additional biological validation and partially alleviate concerns regarding variant misclassification and cohort representativeness.\u003c/p\u003e \u003cp\u003eIn conclusion, this study provides a comprehensive investigation into the host genetic determinants and underlying mechanisms that modulate the severity of respiratory infections, using first-time Omicron infection as a timely and well-controlled model. We identified \u003cem\u003eCPQ\u003c/em\u003e as a novel gene influencing COVID-19 severity in a Chinese population, supported by single‑cell expression and experimental evidence of its anti-inflammatory role. These findings suggest \u003cem\u003eCPQ\u003c/em\u003e\u0026rsquo;s broader relevance in immune regulation and highlight its potential as a therapeutic target. Further research is needed to validate its association with severity across diverse respiratory infections and populations and to elucidate the underlying molecular pathways linking CPQ with immune regulation.\u003c/p\u003e"},{"header":"METHODS","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eStudy design\u003c/h2\u003e \u003cp\u003e The current GWAS is based on data obtained from two multicenter cohort studies conducted in China: The China Surgery and Anesthesia Cohort (CSAC) and The China Severe Trauma Cohort (CSTC), both launched in June-July 2020. Detailed study designs for the two cohorts have been previously described\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. In brief, the CSAC was conducted in four medical centers in China, with the primary objective of recruiting adults aged 40 to 65 years who underwent elective surgery and general anesthesia. The CSTC aimed to recruit patients with traumatic injuries within three months of admission to the Trauma Center of West China Hospital in Chengdu, China. Both cohorts implemented similar survey protocols and quality control processes, managed through a Cohort Data Collection and Management System (Build 2021SR0484324. \u0026copy;West China Hospital, Sichuan, China). Upon obtaining informed consent and screening for eligibility, trained data collectors utilized a standardized questionnaire to gather comprehensive information on sociodemographic and lifestyle factors at baseline for each patient. Peripheral blood and fecal samples were also collected from each patient by research nurses following a standard protocol at baseline. Subsequently, patients were regularly followed up at 1, 3, 6, and 12 months after discharge from the hospital, with long-term follow-up conducted through periodic linkage to multiple national or regional databases. In response to the massive Omicron outbreak in China that occurred in late 2022, during which an estimated 97% of the population was infected\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e, both cohorts began collecting COVID-19 related information from March 2023 onwards. A standardized questionnaire (Supplementary Methods) was used to collect data on SARS‑CoV‑2 infection status, medical treatment received, and symptoms experienced during and after the infection, with response rate of 95.0%.\u003c/p\u003e \u003cp\u003eAs of June 2023, the CSAC and CSTC cohorts recruited a total of 14,538 patients, among which 28 patients had withdrawn later and 6,773 patients have not been genotyped yet. A total of 7737 patients from both cohorts were sent for first wave genotyping and eligible for the present study (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). 5,151 participants from the CSAC and CSTC cohorts (3,757 and 1,394, respectively) had both measures of COVID-19 severity and genotyping data that passed quality control (QC) procedures, (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e and Methods), among whom 269 reported that they experienced the first-time infection and exhibited severe manifestations of COVID-19 (i.e., underwent inpatient medical care or received medical interventions at primary care centers, including transfusion, injection, oxygen therapy, nebulization; Supplementary Table\u0026nbsp;1).\u003c/p\u003e \u003cp\u003e The CSAC, CSTC, and the current study, received ethical approval from the ethics committee of West China Hospital, Sichuan University, with approval numbers 2020\u0026thinsp;\u0026minus;\u0026thinsp;243, 2020\u0026thinsp;\u0026minus;\u0026thinsp;469, and 2020.661, respectively.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eSex as a biological variable\u003c/h2\u003e \u003cp\u003eBoth male and female participants were included in the genome-wide association analyses (2,256 males and 2,895 females). Sex was recorded at baseline and included as a covariate in all analyses to control for potential confounding. No sex-stratified analyses were conducted, and the findings are expected to be relevant to both sexes.\u003c/p\u003e \u003cp\u003eFor the animal experiments, both male and female C57BL/6 mice were used (3 females and 4 males, 8 weeks old). Sex was not used as an experimental variable, and no sex-specific comparisons were performed. The findings are therefore considered generally applicable to both sexes.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eGenome-wide genotype data, quality control, imputation and annotation\u003c/h2\u003e \u003cp\u003eBlood samples were sent for genotyping at the WeGene Clinical Laboratory in Shenzhen, China, using the Illumina Infinium Chinese Genotyping Array (WeGene V3) which covers approximately 700k variants. We performed rigorous quality control and imputation using the Rapid Imputation and Computational Pipeline for Genome-Wide Association Studies (RICOPILI)\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. A detailed flowchart outlining the steps can be found in Supplementary Methods. In summary, the pre-imputation quality control involved filtering 136,592 variants based on criteria of Hardy-Weinberg Equilibrium \u003cem\u003eP\u003c/em\u003e-value (\u0026lt;\u0026thinsp;1\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;6\u003c/sup\u003e), call rate (\u0026lt;\u0026thinsp;98%), non-biallelic variants, or indel variants, as well as 11 samples with issues of low call rate, sex discrepancy, or duplications. This process resulted in 590,757 single nucleotide polymorphisms (SNPs) on autosomes from 7,726 samples, which were deemed suitable for further analyses. Relatedness assessment and principal component analysis (PCA) were then conducted using a subset of SNPs that passed a more stringent quality control process. Subsequently, we performed imputation based on a reference panel comprising East Asian samples from phase 3 of the 1000 Genomes Project\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. Following imputation, we performed post-imputation quality control, which involved removing 24,806,907 variants with low imputation quality (information score\u0026thinsp;\u0026lt;\u0026thinsp;0.8), low allele frequencies (minor allele frequency [MAF]\u0026thinsp;\u0026lt;\u0026thinsp;0.01), non-biallelic variants and indel variants. Additionally, we excluded 26 samples that showed first or second-degree relationships, Finally, we kept individuals with first-time infection of COVID-19 after 31st December 2021, resulting in a final set of 6,126,560 SNPs from 5,151 samples for the GWAS. We annotated each SNP with its Reference SNP cluster ID (rsID) according to dbSNP Build 151 using ANNOVAR\u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eWhole-genome sequencing data\u003c/h2\u003e \u003cp\u003eTo evaluate the accuracy of the genotyping data, we conducted a comparison between the genotyping results and the results obtained from whole-genome sequencing (WGS) for the SNPs located in the identified risk loci. This analysis was carried out on a subset of patients who had both genotyping and WGS data available. From the CSAC and CSTC cohorts, a total of 733 patients were randomly selected for WGS, and among them, 498 patients were included in the current study. The WGS was conducted by the Institute of Rare Diseases at West China Hospital of Sichuan University. The DNBSEQ-T7 Next-Generation Sequencing (NGS) platform, developed by Complete Genomics and MGI, was utilized for the WGS process. A comprehensive description of the quality control for raw reads, alignment, variant calling, as well as variant and sample filtering can be found in the Supplementary Methods.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eDefinition of severe COVID-19\u003c/h2\u003e \u003cp\u003eThe primary outcome of interest in this study is severe COVID-19, which is defined as patients requiring inpatient medical care during the infection period. In light of the rapidly spreading Omicron outbreak in China, which has placed a substantial burden on the majority of hospitals, we have broadened the definition of severe COVID-19 to encompass patients who received medical treatments such as transfusion, injection, oxygen therapy, or nebulization therapy in primary care centers during the infection period. This extension is essential as these patients may be unable to be admitted to the hospital due to the overwhelming circumstances.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eGenome-wide association analysis\u003c/h2\u003e \u003cp\u003eUsing imputed dosage data from the two cohorts, we performed GWAS under an additive model for the severe COVID-19 outcome. To effectively account for population structure and residual kinship among patients, we employed a generalized linear mixed model implemented in GCTA\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e. The model incorporated several covariates, including age, sex, birth year, genotyping batch, cohort, and the first ten principal components. Based on the GWAS results, we generated quantile-quantile (QQ) plots and estimated the genomic inflation factor (λ) to assess whether there were any systematic deviations or inflation in the observed GWAS P-values. A variant was considered genome-wide significant if its \u003cem\u003eP\u003c/em\u003e-value\u0026thinsp;\u0026le;\u0026thinsp;5\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;8\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eIdentification of linkage disequilibrium independent risk loci\u003c/h2\u003e \u003cp\u003eTo identify linkage disequilibrium (LD) independent risk loci, we employed the \u0026ldquo;clumping\u0026rdquo; command in PLINK v1.9\u003csup\u003e48\u003c/sup\u003e. Specifically, for each variant with \u003cem\u003eP\u003c/em\u003e-value\u0026thinsp;\u0026le;\u0026thinsp;5\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;8\u003c/sup\u003e (i.e., the index variant), other variants with \u003cem\u003eP\u003c/em\u003e-value\u0026thinsp;\u0026le;\u0026thinsp;5\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;6\u003c/sup\u003e, located within 500 kilobase (kb) distance, and LD correlated with the index variant (r\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;\u0026ge;\u0026thinsp;0.1) were grouped into the index variant\u0026rsquo;s clump to form LD-independent genomic region. Subsequently, we merged physically overlapping genomic regions using BEDTools\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e to form LD-independent risk loci. The variant with the lowest P-value within each locus was considered the lead variant.\u003c/p\u003e \u003cp\u003eTo comprehensively characterize the identified risk loci and determine their novelty for COVID-19 severity or related outcomes, we conducted a search in the GWAS Catalog database (accessed on February 2025)\u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e. Specifically, we retrieved all genome-wide significant associations from previous GWAS for variants within a 500 kb region surrounding the identified risk locus.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003ePRS analyses\u003c/h2\u003e \u003cp\u003eWe used the GWAS summary statistics (i.e., effect sizes and standard errors for the variants) for very severe respiratory confirmed COVID-19 (i.e., critically ill cases of COVID-19 defined as those individuals who required respiratory support in hospital or who died due to the disease; 784 cases vs. 4,862 population) and summary statistics for hospitalized COVID-19 (2,882 cases vs. 31,200 population) conducted in East Asian populations from the COVID-19 Host Genetics Initiative study (release 7, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.covid19hg.org/results/r7/\u003c/span\u003e\u003cspan address=\"https://www.covid19hg.org/results/r7/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) as base data sets \u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. We removed ambiguous SNPs and kept the SNPs available in both base data sets and target data set (the post-QC genotyping data of current study cohorts). PRSs for the two relevant COVID-19 outcomes were generated as weighted sum of the risk alleles through a penalized regression framework known as LASSO method, which allows for heavy shrinkage in the effect estimates of SNPs via regularization \u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e. We calculated ORs with 95% CI for the associations between PRSs and COVID-19 severity using logistic regression models, adjusting for birth year, sex, genotyping batch, study cohort and the first ten principal components.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eEnrichment analyses for genome-wide significant variants\u003c/h2\u003e \u003cp\u003eWe employed the hypergeometric distribution test to assess the enrichment of the identified genome-wide significant variants in specific chromatin states of the genome. The genomic regions representing 15 chromatin states in blood and lung tissues/cells were obtained from the Roadmap Epigenomics project\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. For each genome-wide significant variant, we randomly selected 100 non-significant variants possessing comparable MAF, LD score, and distance to the nearest gene, to form the background set for the enrichment analyses. Chromatin states showing enrichment with a false discovery rate adjusted \u003cem\u003eP\u003c/em\u003e-value (i.e., q-value) of less than 0.05 were considered statistically significant.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eConsensus-based mapping for putative risk genes\u003c/h2\u003e \u003cp\u003eWe adopted a previously proposed consensus-based gene mapping approach to systematically map the identified risk loci to their putative risk genes \u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e,\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. This approach integrated evidence from the following five distinct analyses that covered various aspects, ranging from mRNA and protein expression to alternative splicing: 1) expression quantitative trait loci (eQTL) colocalization analyses, to identify a gene located near a locus that exhibited significant colocalization between its GWAS signals and eQTL signals in 49 Genotype-Tissue Expression (GTEx) tissues\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e; 2) splicing quantitative trait loci (sQTL) colocalization analyses, to identify a gene located near a locus that exhibited significant colocalization between its GWAS signals and sQTL signals in 49 GTEx tissues\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e; 3) protein quantitative trait loci (pQTL) colocalization analyses, to identify a gene located near a locus that exhibited significant colocalization between its GWAS signals and serum pQTL signals from AGES cohort\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e; 4) protein-altering variant (PAV) linkage, to identify a gene located near a locus that harbored a PAV that is in strong or moderate LD with the identified lead variant, by Variant Effect Predictor \u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e; 5) similarity-based gene prioritization, to identify a gene located near a locus that received top three highest scores from the Polygenic Priority Score (PoPS)\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eComprehensive descriptions of each analysis are provided in the Supplementary Methods. For each locus, the putative risk gene was determined as the gene with the highest number of supporting evidence from the above five analyses.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003eSingle-cell RNA sequencing (scRNA-seq) data\u003c/h2\u003e \u003cp\u003eTo gain a comprehensive understanding of the identified putative risk gene, we undertook an extensive characterization by re-analyzing previously published single-cell RNA sequencing (scRNA-seq) data of lung tissues and peripheral blood mononuclear cells (PBMCs) collected from severe COVID-19 patients and healthy controls\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. Briefly, for the scRNA-seq data of lung tissues, we directly obtained the processed count matrix and cluster annotation data from the original study\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. Our analysis concentrated on six major cell clusters of lung tissues, including epithelial cells, myeloid cells, fibroblasts, T cells, B cells, and endothelial cells. Within the subset of epithelial cells, myeloid cells, fibroblasts, and T cells, we additionally identified their respective major sub-clusters. This yielded a total of 21 distinct cell types, collectively accounting for approximately 92% of the entire cell population (see Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e for details). Regarding the scRNA-seq data from PBMCs, we obtained the raw count matrix data from the Gene Expression Omnibus (GEO) repository and processed them using the Seurat v5\u003csup\u003e53\u003c/sup\u003e, following the procedural framework outlined in the original study\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. Following data normalizing and scaling procedures, we employed the Uniform Manifold Approximation and Projection (UMAP)\u003csup\u003e\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e for dimensionality reduction and the shared nearest neighbor (SNN) modularity optimization algorithm\u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e for clustering. The resulting cell clusters were assigned to one of the following 11 major cell types based on a curated list of cell markers outlined in the original study\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e: T cells (CD4\u0026thinsp;+\u0026thinsp;and CD8+), B cells, natural killing (NK) cells, monocytes (classical, non-classical, and intermediate), platelets, red blood cells (RBC), Dendritic cells (DC), and other undefined cells.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003eDifferential gene expression analyses in scRNA-seq data\u003c/h2\u003e \u003cp\u003eTo prioritize the putative causal cell types wherein the risk gene may exert its relevant functions, we conducted differential gene expression (DGE) analyses using the Mann-Whitney U test on the scRNA-seq data of both lung tissues and PBMCs, as previously mentioned. Initially, we conducted DGE analysis for the expression of the risk gene in each cell type in comparison to the rest cells, with the aim of identifying cell types where the risk gene displayed significant overexpression. Furthermore, we conducted DGE analysis for the expression of the risk gene between severe COVID-19 patients and healthy controls for each cell type. The significance level for the DGE analyses was set at a Bonferroni-corrected p-value of 0.002 (0.05/21 cell types) for lung scRNA-seq data and 0.004 (0.05/11 cell types) for PBMCs scRNA-seq data.\u003c/p\u003e \u003cdiv id=\"Sec23\" class=\"Section3\"\u003e \u003ch2\u003ePseudotime trajectory analysis in scRNA-seq data\u003c/h2\u003e \u003cp\u003eTo gain insights into the dynamics of the risk gene expression throughout the transition process from monocytes to monocyte-derived macrophages (MDM) in lung tissues during SARS‑CoV‑2 infection, we conducted a pseudotime trajectory analysis for these cell types in severe COVID-19 patients using the R package Monocle 3\u003csup\u003e26\u003c/sup\u003e. Each cell was assigned a pseudotime value, where 0 indicated no progress in the transition from monocyte to MDM, and the maximum value represented 100% progress. We then employed a negative binomial regression model with cubic splines to regress the risk gene expression against the pseudotime.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003eWeighted gene co-expression network analysis in scRNA-seq data\u003c/h2\u003e \u003cp\u003eWe performed weighted gene co-expression network analysis using the R package hdWGCNA\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e, enabling the construction of gene co-expression networks and the clustering of genes into distinct co-expression modules. We specifically focused the analysis on the putative causal cell types in severe COVID-19 patients, as indicated by the DGE analyses. To visualize the gene co-expression network, we employed the dimension reduction algorithm UMAP\u003csup\u003e\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e on the co-expression matrix, to get two-dimensional UMAP features for each gene.\u003c/p\u003e \u003cdiv id=\"Sec25\" class=\"Section3\"\u003e \u003ch2\u003eGene set enrichment analysis\u003c/h2\u003e \u003cp\u003eBased on the module assignment obtained from the weight gene co-expression network analysis, we conducted gene set enrichment analysis using GSEApy\u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e. We assessed the enrichment of the top 100 genes within each module in pathways curated from various sources, including BioCarta\u003csup\u003e\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e, Pathway Interaction Database (PID)\u003csup\u003e\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e, WikiPathways\u003csup\u003e\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e, Reactome\u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e, and Gene Ontology (GO)\u003csup\u003e\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e\u003c/sup\u003e. Furthermore, we carried out enrichment analysis for the putative risk gene itself along with the 20 genes nearest to it, which were identified through the application of the K-Nearest Neighbors (KNN) algorithm\u003csup\u003e\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e\u003c/sup\u003e to the UMAP features of genes within the co-expression network. Statistical significance level for the gene set enrichment analysis was defined as q-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec26\" class=\"Section3\"\u003e \u003ch2\u003eProtein-protein interaction network analysis\u003c/h2\u003e \u003cp\u003eWe performed protein-protein interaction network analysis within a subset of ubiquitination-related genes using STRING (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://string-db.org/\u003c/span\u003e\u003cspan address=\"https://string-db.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e)\u003csup\u003e28\u003c/sup\u003e. In particular, we focused on protein-protein interactions that were backed by high-throughput experiments and possessed association scores exceeding the median score of 0.04.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec27\" class=\"Section3\"\u003e \u003ch2\u003eTranscription factor binding motifs analyses\u003c/h2\u003e \u003cp\u003eWe conducted transcription factor (TF) binding motif scanning and comparison analyses to identify potential TF binding affecting variants in the risk locus and their target TFs. Initially, we annotated lead variant and variants in strong or moderate LD with it (r\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;\u0026ge;\u0026thinsp;0.6) within the locus, using the RegulomeDB ranking score\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. Subsequent analyses were limited to variants with high regulatory potential (i.e., those with a RegulomeDB ranking score of \u0026ldquo;1f\u0026rdquo; or higher). For each eligible variant, we extracted both the reference and alternative DNA sequences (i.e., sequences containing the risk allele of the variant) from a\u0026thinsp;\u0026plusmn;\u0026thinsp;20 base pair (bp) region surrounding it, according to the GRCh38 reference sequence. Next, using the motif scanning software FIMO\u003csup\u003e\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e\u003c/sup\u003e, we calculated the binding affinity for the set of extracted sequences with each human TF binding motif from the HOCOMOCD database (version 11)\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. We considered a variant to have the potential to increase the binding affinity of a specific TF if the \u003cem\u003eP\u003c/em\u003e-value for motif occurrence in the reference sequence was \u0026le;\u0026thinsp;1\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;4\u003c/sup\u003e, while the P-value for motif occurrence in the alternative sequence was \u0026gt;\u0026thinsp;1\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;4\u003c/sup\u003e, and the difference in log-odds scores between the two sequences was \u0026ge;\u0026thinsp;2. Using similar thresholds, we also identified a set of variants with the potential to disrupt TF binding (i.e., variants with \u003cem\u003eP\u003c/em\u003e-value\u0026thinsp;\u0026gt;\u0026thinsp;1\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;4\u003c/sup\u003e and \u0026le;\u0026thinsp;1\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;4\u003c/sup\u003e for reference and alternative sequences, and difference in log-odds scores\u0026le;-2).\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec28\" class=\"Section2\"\u003e \u003ch2\u003eFiltering of the identified transcription factors\u003c/h2\u003e \u003cp\u003eFollowing the identification of the TFs with potential alterations in binding affinity due to the presence of the risk alleles, we performed DGE analyses to assess the expression of the chosen TFs in comparison to background genes for each putative causal cell type, to validate the functional activity of the selected TFs. Then, for TFs with confirmed functional relevance, we visualized their TF chromatin immunoprecipitation-sequencing (ChIP-seq) signals in the regions surrounding the variants that were predicted to affect their binding affinity. The ChIP-seq data used for visualization were obtained from the ENCODE project\u003csup\u003e\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec29\" class=\"Section2\"\u003e \u003ch2\u003eExperimental validation in human cells and mouse models\u003c/h2\u003e \u003cp\u003e \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eCell culture\u003c/span\u003e: THP-1 cells were obtained from National Collection of Authenticated Cell Cultures (NCACC). THP-1 \u003cem\u003eCPQ overexpression\u003c/em\u003e cell line was generated using lentiviral transduction in our lab. THP-1 cells and \u003cem\u003eCPQ overexpression\u003c/em\u003e cells were cultured in RPMI 1640 medium (Gibco) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin (Thermo Fisher Scientific) at 37\u0026deg;C in a humidified incubator with 5% CO₂.\u003c/p\u003e \u003cp\u003e \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eGeneration of\u003c/span\u003e \u003cspan type=\"ItalicUnderline\" class=\"ItalicUnderline\" name=\"Emphasis\"\u003eCPQ\u003c/span\u003e \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eoverexpression cell line\u003c/span\u003e: THP-1 cells were transduced with lentiviruses encoding the \u003cem\u003eCPQ\u003c/em\u003e sequence. Viral particles were produced by co-transfecting 293T cells with the lentiviral transfer vector and packaging plasmids (pMD2.G and psPAX2) using Lipofectamine 3000 transfection reagent, according to the manufacturer's protocol. The viral supernatant was harvested 48 hours post-transfection, filtered through a 0.45-\u0026micro;m membrane, and used to transduce target cells cultured in 6-well plates (400 \u0026micro;L per well) in the presence of polybrene (final concentration: 10 \u0026micro;g/mL). After 48 hours of transduction, the medium was replaced with fresh medium containing puromycin (2 \u0026micro;g/mL) for selection.\u003c/p\u003e \u003cp\u003e \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eConstruction of the Cellular Inflammation Model\u003c/span\u003e: THP-1 cells were collected during the logarithmic growth phase and treated with Phorbol 12-myristate 13-acetate (PMA) for 48 hours to induce differentiation. Morphological changes of the cells were monitored under an optical microscope. Following adherence, the cells were differentiated into mature, inactivated macrophages (M0 phenotype). The supernatant was removed, and the cells were washed with serum-free RPMI 1640 medium and incubated in this medium for 24 hours. The serum-free medium was then aspirated. Complete medium containing varying concentrations of LPS was added to the cells to induce inflammation.\u003c/p\u003e \u003cp\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eMice\u003c/span\u003e: C57BL/6 mice at the age of 8 weeks were purchased from Chengdu Dossy Experimental Animals Co., Ltd. All animal experiments in this study were conducted with the approval of the Animal Ethics Committee of Sichuan University (Approval No. 20240308057). Mice were group-housed under a standard light cycle (12-h light\u0026ndash;dark; lights on from 7:00 to 19:00) at 20\u0026ndash;23\u0026deg;C and about 50% humidity with ad libitum access to water and food.\u003c/p\u003e \u003cp\u003eThe systemic inflammation model in mice is established through intraperitoneal injection of LPS (10 mg/kg) for 6 hours. Lung tissue is obtained through thoracic dissection, washed with pre-chilled PBS, and stored in liquid nitrogen. Tissue samples are added to lysis buffer (mPER, Thermofisher) at a 10% mass/volume ratio and homogenized. After 10 minutes of lysis on ice, the sample is centrifuged at 4\u0026deg;C for 15 minutes at 16000 xg to remove impurities. The supernatant is transferred to a new pre-chilled tube. Protein concentration is measured using the Bradford method and quantified to 3 \u0026micro;g/\u0026micro;L.\u003c/p\u003e \u003cp\u003e \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eELISA\u003c/span\u003e: Enzyme-linked immunosorbent assay (ELISA) for human IL-6 (Boster) and TNF-α (Eeo bioscience) was performed according to the manufacturer\u0026rsquo;s protocols. The standard curves were calculated by Excel. Cell supernatant samples for ELISA assay were collected by centrifugation at 1200 rpm for 3 min.\u003c/p\u003e \u003cp\u003e \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eWestern blot analysis\u003c/span\u003e: Total protein was extracted from cells or tissue and then analyzed using SDS\u0026ndash;PAGE. Following electrophoresis, proteins were transferred to PVDF membranes (Millipore). Membranes were then incubated with the following primary antibodies: rabbit anti-PGCP (1:1000 dilution), rabbit anti-IL-6 (1:500 dilution), or mouse anti-TNF-α (1:500 dilution). Immunoblotting with an anti-\u003cem\u003eβ\u003c/em\u003e-actin antibody (1:5000 dilution) served as the loading control. Protein bands were visualized using enhanced chemiluminescence (ECL) detection reagents. Band intensities were quantified using Image J Software. All the reagents and chemicals used in this study are listed in Supplementary Methods Table\u0026nbsp;4.\u003c/p\u003e \u003cp\u003e \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eStatistical Data and Reproducibility\u003c/span\u003e: Statistical analysis was performed using GraphPad. Sample source, \u003cem\u003ep\u003c/em\u003e-values, and the statistical methods employed are described in the corresponding legends or source data. A difference is considered significant when \u003cem\u003eP\u003c/em\u003e-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eACKNOWLEDGEMENTS\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the 1.3.5 project for disciplines of excellence, West China Hospital, Sichuan University (No. ZYYC21005 to HS, ZYGD20005 to QL),\u0026nbsp;the National Natural Science Foundation of China (82471535 to HS, 82425054 and 82273784 to BK, 82404350 to JS, and 82404391 to CH), and the Science \u0026amp; Technology Department of Sichuan Province (2024NSFSC1637 to JS and 2024NSFSC1568 to CH). We thank the team members involved in West China Biomedical Big Data Center for their support.\u0026nbsp;The computations in this study were supported by the High Performance Computing platform at West China Biomedical Big Data Center, West China Hospital, Sichuan University.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe GWAS summary statistics are available upon request.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict-of-interest statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eContributors\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCH, JS, BK and HS were responsible for the study\u0026rsquo;s concept and design. LY and YQ did the acquisition of data. HY contributed to the phenotype definition. CH, WC and YZ did the data and project management. YY, FY did the experiment analyses in human cells and animal models. CH, YY and YZ did the data cleaning and analysis. YZ, CH, YY and JS drafted the manuscript. YZ, CH, YY, YJ, FY, QL, FB, JS, UAV, FF, BK and HS contributed to the interpretation of the results and revised the manuscript. All the authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe conduction of CSAC and CSTC were approved by the ethics committee of West China Hospital, Sichuan University (approval number: 2020.469 and 2020.243). The additional collection of COVID-19 data in those Chinese cohorts and this specific study were approved by the ethics committee and animal ethics committee of West China Hospital, Sichuan University (2020.469, 2020.243, 2020.66 and 20240308057).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eJin, X.\u003cem\u003e et al.\u003c/em\u003e Global burden of upper respiratory infections in 204 countries and territories, from 1990 to 2019. \u003cem\u003eeClinicalMedicine\u003c/em\u003e \u003cstrong\u003e37\u003c/strong\u003e(2021).\u003c/li\u003e\n\u003cli\u003eNaghavi, M.\u003cem\u003e et al.\u003c/em\u003e Global burden of 288 causes of death and life expectancy decomposition in 204 countries and territories and 811 subnational locations, 1990\u0026amp;#x2013;2021: a systematic analysis for the Global Burden of Disease Study 2021. \u003cem\u003eThe Lancet\u003c/em\u003e \u003cstrong\u003e403\u003c/strong\u003e, 2100-2132 (2024).\u003c/li\u003e\n\u003cli\u003eKousathanas, A.\u003cem\u003e et al.\u003c/em\u003e Whole-genome sequencing reveals host factors underlying critical COVID-19. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e607\u003c/strong\u003e, 97-103 (2022).\u003c/li\u003e\n\u003cli\u003eChen, H.-H.\u003cem\u003e et al.\u003c/em\u003e Host genetic effects in pneumonia. \u003cem\u003eThe American Journal of Human Genetics\u003c/em\u003e \u003cstrong\u003e108\u003c/strong\u003e, 194-201 (2021).\u003c/li\u003e\n\u003cli\u003eChousterman, B.G., Swirski, F.K. \u0026amp; Weber, G.F. Cytokine storm and sepsis disease pathogenesis. \u003cem\u003eSemin Immunopathol\u003c/em\u003e \u003cstrong\u003e39\u003c/strong\u003e, 517-528 (2017).\u003c/li\u003e\n\u003cli\u003eWong, L.R. \u0026amp; Perlman, S. Immune dysregulation and immunopathology induced by SARS-CoV-2 and related coronaviruses - are we our own worst enemy? \u003cem\u003eNat Rev Immunol\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 47-56 (2022).\u003c/li\u003e\n\u003cli\u003ePairo-Castineira, E.\u003cem\u003e et al.\u003c/em\u003e GWAS and meta-analysis identifies 49 genetic variants underlying critical COVID-19. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e617\u003c/strong\u003e, 764-768 (2023).\u003c/li\u003e\n\u003cli\u003eKanai, M.\u003cem\u003e et al.\u003c/em\u003e A second update on mapping the human genetic architecture of COVID-19. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e621\u003c/strong\u003e, E7-E26 (2023).\u003c/li\u003e\n\u003cli\u003eMozzi, A., Pontremoli, C. \u0026amp; Sironi, M. Genetic susceptibility to infectious diseases: Current status and future perspectives from genome-wide approaches. \u003cem\u003eInfection, Genetics and Evolution\u003c/em\u003e \u003cstrong\u003e66\u003c/strong\u003e, 286-307 (2018).\u003c/li\u003e\n\u003cli\u003eAnsari, M.A.\u003cem\u003e et al.\u003c/em\u003e Genome-to-genome analysis highlights the effect of the human innate and adaptive immune systems on the hepatitis C virus. \u003cem\u003eNature genetics\u003c/em\u003e \u003cstrong\u003e49\u003c/strong\u003e, 666-673 (2017).\u003c/li\u003e\n\u003cli\u003eDuchen, D.\u003cem\u003e et al.\u003c/em\u003e Pathogen exposure misclassification can bias association signals in GWAS of infectious diseases when using population-based common control subjects. \u003cem\u003eAm J Hum Genet\u003c/em\u003e \u003cstrong\u003e110\u003c/strong\u003e, 336-348 (2023).\u003c/li\u003e\n\u003cli\u003eGoldberg, E.E., Lin, Q., Romero-Severson, E.O. \u0026amp; Ke, R. Swift and extensive Omicron outbreak in China after sudden exit from \u0026apos;zero-COVID\u0026apos; policy. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 3888 (2023).\u003c/li\u003e\n\u003cli\u003eYang, Y.\u003cem\u003e et al.\u003c/em\u003e Cohort Profile: The China Severe Trauma Cohort (CSTC). \u003cem\u003eJ Epidemiol\u003c/em\u003e (2022).\u003c/li\u003e\n\u003cli\u003eYang, L.\u003cem\u003e et al.\u003c/em\u003e Cohort profile: the China surgery and anesthesia cohort (CSAC). \u003cem\u003eEur J Epidemiol\u003c/em\u003e \u003cstrong\u003e39\u003c/strong\u003e, 207-218 (2024).\u003c/li\u003e\n\u003cli\u003eInitiative, C.-H.G. The COVID-19 Host Genetics Initiative, a global initiative to elucidate the role of host genetic factors in susceptibility and severity of the SARS-CoV-2 virus pandemic. \u003cem\u003eEur J Hum Genet\u003c/em\u003e \u003cstrong\u003e28\u003c/strong\u003e, 715-718 (2020).\u003c/li\u003e\n\u003cli\u003eRoadmap Epigenomics, C.\u003cem\u003e et al.\u003c/em\u003e Integrative analysis of 111 reference human epigenomes. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e518\u003c/strong\u003e, 317-30 (2015).\u003c/li\u003e\n\u003cli\u003eAragam, K.G.\u003cem\u003e et al.\u003c/em\u003e Discovery and systematic characterization of risk variants and genes for coronary artery disease in over a million participants. \u003cem\u003eNat Genet\u003c/em\u003e \u003cstrong\u003e54\u003c/strong\u003e, 1803-1815 (2022).\u003c/li\u003e\n\u003cli\u003eVujkovic, M.\u003cem\u003e et al.\u003c/em\u003e A multiancestry genome-wide association study of unexplained chronic ALT elevation as a proxy for nonalcoholic fatty liver disease with histological and radiological validation. \u003cem\u003eNat Genet\u003c/em\u003e \u003cstrong\u003e54\u003c/strong\u003e, 761-771 (2022).\u003c/li\u003e\n\u003cli\u003eGiambartolomei, C.\u003cem\u003e et al.\u003c/em\u003e Bayesian test for colocalisation between pairs of genetic association studies using summary statistics. \u003cem\u003ePLoS Genet\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, e1004383 (2014).\u003c/li\u003e\n\u003cli\u003eConsortium, G.T. The Genotype-Tissue Expression (GTEx) project. \u003cem\u003eNat Genet\u003c/em\u003e \u003cstrong\u003e45\u003c/strong\u003e, 580-5 (2013).\u003c/li\u003e\n\u003cli\u003eGudjonsson, A.\u003cem\u003e et al.\u003c/em\u003e A genome-wide association study of serum proteins reveals shared loci with common diseases. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 480 (2022).\u003c/li\u003e\n\u003cli\u003eMcLaren, W.\u003cem\u003e et al.\u003c/em\u003e The Ensembl Variant Effect Predictor. \u003cem\u003eGenome Biol\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 122 (2016).\u003c/li\u003e\n\u003cli\u003eWeeks, E.M.\u003cem\u003e et al.\u003c/em\u003e Leveraging polygenic enrichments of gene features to predict genes underlying complex traits and diseases. \u003cem\u003eNature Genetics\u003c/em\u003e \u003cstrong\u003e55\u003c/strong\u003e, 1267-1276 (2023).\u003c/li\u003e\n\u003cli\u003eLee, J.S.\u003cem\u003e et al.\u003c/em\u003e Immunophenotyping of COVID-19 and influenza highlights the role of type I interferons in development of severe COVID-19. \u003cem\u003eSci Immunol\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e(2020).\u003c/li\u003e\n\u003cli\u003eMelms, J.C.\u003cem\u003e et al.\u003c/em\u003e A molecular single-cell lung atlas of lethal COVID-19. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e595\u003c/strong\u003e, 114-119 (2021).\u003c/li\u003e\n\u003cli\u003eCao, J.\u003cem\u003e et al.\u003c/em\u003e The single-cell transcriptional landscape of mammalian organogenesis. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e566\u003c/strong\u003e, 496-502 (2019).\u003c/li\u003e\n\u003cli\u003eMorabito, S., Reese, F., Rahimzadeh, N., Miyoshi, E. \u0026amp; Swarup, V. hdWGCNA identifies co-expression networks in high-dimensional transcriptomics data. \u003cem\u003eCell Rep Methods\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 100498 (2023).\u003c/li\u003e\n\u003cli\u003eSzklarczyk, D.\u003cem\u003e et al.\u003c/em\u003e The STRING database in 2023: protein-protein association networks and functional enrichment analyses for any sequenced genome of interest. \u003cem\u003eNucleic Acids Res\u003c/em\u003e \u003cstrong\u003e51\u003c/strong\u003e, D638-D646 (2023).\u003c/li\u003e\n\u003cli\u003eDong, S.\u003cem\u003e et al.\u003c/em\u003e Annotating and prioritizing human non-coding variants with RegulomeDB v.2. \u003cem\u003eNat Genet\u003c/em\u003e \u003cstrong\u003e55\u003c/strong\u003e, 724-726 (2023).\u003c/li\u003e\n\u003cli\u003eKulakovskiy, I.V.\u003cem\u003e et al.\u003c/em\u003e HOCOMOCO: a comprehensive collection of human transcription factor binding sites models. \u003cem\u003eNucleic Acids Res\u003c/em\u003e \u003cstrong\u003e41\u003c/strong\u003e, D195-202 (2013).\u003c/li\u003e\n\u003cli\u003eChen, H.\u003cem\u003e et al.\u003c/em\u003e Activation of STAT6 by STING is critical for antiviral innate immunity. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e147\u003c/strong\u003e, 436-46 (2011).\u003c/li\u003e\n\u003cli\u003eSeifert, L.L.\u003cem\u003e et al.\u003c/em\u003e The ETS transcription factor ELF1 regulates a broadly antiviral program distinct from the type I interferon response. \u003cem\u003ePLoS Pathog\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, e1007634 (2019).\u003c/li\u003e\n\u003cli\u003eLiu, T., Zhang, L., Joo, D. \u0026amp; Sun, S.C. NF-kappaB signaling in inflammation. \u003cem\u003eSignal Transduct Target Ther\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 17023- (2017).\u003c/li\u003e\n\u003cli\u003eGuan, H., Hou, S. \u0026amp; Ricciardi, R.P. DNA binding of repressor nuclear factor-kappaB p50/p50 depends on phosphorylation of Ser337 by the protein kinase A catalytic subunit. \u003cem\u003eJ Biol Chem\u003c/em\u003e \u003cstrong\u003e280\u003c/strong\u003e, 9957-62 (2005).\u003c/li\u003e\n\u003cli\u003eVillar, D.\u003cem\u003e et al.\u003c/em\u003e Enhancer evolution across 20 mammalian species. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e160\u003c/strong\u003e, 554-66 (2015).\u003c/li\u003e\n\u003cli\u003ePairo-Castineira, E.\u003cem\u003e et al.\u003c/em\u003e Genetic mechanisms of critical illness in COVID-19. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e591\u003c/strong\u003e, 92-98 (2021).\u003c/li\u003e\n\u003cli\u003eFarh, K.K.\u003cem\u003e et al.\u003c/em\u003e Genetic and epigenetic fine mapping of causal autoimmune disease variants. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e518\u003c/strong\u003e, 337-43 (2015).\u003c/li\u003e\n\u003cli\u003eVuillard, L., Nicholson, J. \u0026amp; Hay, R.T. A complex containing betaTrCP recruits Cdc34 to catalyse ubiquitination of IkappaBalpha. \u003cem\u003eFEBS Lett\u003c/em\u003e \u003cstrong\u003e455\u003c/strong\u003e, 311-4 (1999).\u003c/li\u003e\n\u003cli\u003eSuzuki, H.\u003cem\u003e et al.\u003c/em\u003e Homodimer of two F-box proteins betaTrCP1 or betaTrCP2 binds to IkappaBalpha for signal-dependent ubiquitination. \u003cem\u003eJ Biol Chem\u003c/em\u003e \u003cstrong\u003e275\u003c/strong\u003e, 2877-84 (2000).\u003c/li\u003e\n\u003cli\u003eFuchs, S.Y., Chen, A., Xiong, Y., Pan, Z.Q. \u0026amp; Ronai, Z. HOS, a human homolog of Slimb, forms an SCF complex with Skp1 and Cullin1 and targets the phosphorylation-dependent degradation of IkappaB and beta-catenin. \u003cem\u003eOncogene\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e, 2039-46 (1999).\u003c/li\u003e\n\u003cli\u003eOeckinghaus, A. \u0026amp; Ghosh, S. The NF-kappaB family of transcription factors and its regulation. \u003cem\u003eCold Spring Harb Perspect Biol\u003c/em\u003e \u003cstrong\u003e1\u003c/strong\u003e, a000034 (2009).\u003c/li\u003e\n\u003cli\u003eFericean, R.M.\u003cem\u003e et al.\u003c/em\u003e Outcomes of Elderly Patients Hospitalized with the SARS-CoV-2 Omicron B.1.1.529 Variant: A Systematic Review. \u003cem\u003eInt J Environ Res Public Health\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e(2023).\u003c/li\u003e\n\u003cli\u003eWang, S.\u003cem\u003e et al.\u003c/em\u003e Genomic Surveillance for SARS-CoV-2 - China, September 26, 2022 to January 29, 2023. \u003cem\u003eChina CDC Wkly\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 143-151 (2023).\u003c/li\u003e\n\u003cli\u003eLam, M.\u003cem\u003e et al.\u003c/em\u003e RICOPILI: Rapid Imputation for COnsortias PIpeLIne. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e36\u003c/strong\u003e, 930-933 (2020).\u003c/li\u003e\n\u003cli\u003eGenomes Project, C.\u003cem\u003e et al.\u003c/em\u003e A global reference for human genetic variation. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e526\u003c/strong\u003e, 68-74 (2015).\u003c/li\u003e\n\u003cli\u003eWang, K., Li, M. \u0026amp; Hakonarson, H. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. \u003cem\u003eNucleic Acids Res\u003c/em\u003e \u003cstrong\u003e38\u003c/strong\u003e, e164 (2010).\u003c/li\u003e\n\u003cli\u003eYang, J., Lee, S.H., Goddard, M.E. \u0026amp; Visscher, P.M. GCTA: a tool for genome-wide complex trait analysis. \u003cem\u003eAm J Hum Genet\u003c/em\u003e \u003cstrong\u003e88\u003c/strong\u003e, 76-82 (2011).\u003c/li\u003e\n\u003cli\u003ePurcell, S.\u003cem\u003e et al.\u003c/em\u003e PLINK: a tool set for whole-genome association and population-based linkage analyses. \u003cem\u003eAm J Hum Genet\u003c/em\u003e \u003cstrong\u003e81\u003c/strong\u003e, 559-75 (2007).\u003c/li\u003e\n\u003cli\u003eQuinlan, A.R. \u0026amp; Hall, I.M. BEDTools: a flexible suite of utilities for comparing genomic features. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e26\u003c/strong\u003e, 841-2 (2010).\u003c/li\u003e\n\u003cli\u003eBuniello, A.\u003cem\u003e et al.\u003c/em\u003e The NHGRI-EBI GWAS Catalog of published genome-wide association studies, targeted arrays and summary statistics 2019. \u003cem\u003eNucleic Acids Res\u003c/em\u003e \u003cstrong\u003e47\u003c/strong\u003e, D1005-D1012 (2019).\u003c/li\u003e\n\u003cli\u003eMak, T.S.H., Porsch, R.M., Choi, S.W., Zhou, X. \u0026amp; Sham, P.C. Polygenic scores via penalized regression on summary statistics. \u003cem\u003eGenet Epidemiol\u003c/em\u003e \u003cstrong\u003e41\u003c/strong\u003e, 469-480 (2017).\u003c/li\u003e\n\u003cli\u003eWeeks, E.M.\u003cem\u003e et al.\u003c/em\u003e Leveraging polygenic enrichments of gene features to predict genes underlying complex traits and diseases. 2020.09. 08.20190561 (2020).\u003c/li\u003e\n\u003cli\u003eHao, Y.\u003cem\u003e et al.\u003c/em\u003e Dictionary learning for integrative, multimodal and scalable single-cell analysis. 1-12 (2023).\u003c/li\u003e\n\u003cli\u003eMcInnes, L., Healy, J. \u0026amp; Melville, J.J.a.p.a. Umap: Uniform manifold approximation and projection for dimension reduction. (2018).\u003c/li\u003e\n\u003cli\u003eWaltman, L. \u0026amp; Van Eck, N.J.J.T.E.p.j.B. A smart local moving algorithm for large-scale modularity-based community detection. \u003cstrong\u003e86\u003c/strong\u003e, 1-14 (2013).\u003c/li\u003e\n\u003cli\u003eFang, Z., Liu, X. \u0026amp; Peltz, G. GSEApy: a comprehensive package for performing gene set enrichment analysis in Python. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e39\u003c/strong\u003e(2023).\u003c/li\u003e\n\u003cli\u003eNishimura, D.J.B.S. \u0026amp; Scient, I.R.T.C.S.J.f. BioCarta. \u003cstrong\u003e2\u003c/strong\u003e, 117-120 (2001).\u003c/li\u003e\n\u003cli\u003eSchaefer, C.F.\u003cem\u003e et al.\u003c/em\u003e PID: the Pathway Interaction Database. \u003cem\u003eNucleic Acids Res\u003c/em\u003e \u003cstrong\u003e37\u003c/strong\u003e, D674-9 (2009).\u003c/li\u003e\n\u003cli\u003eMartens, M.\u003cem\u003e et al.\u003c/em\u003e WikiPathways: connecting communities. \u003cem\u003eNucleic Acids Res\u003c/em\u003e \u003cstrong\u003e49\u003c/strong\u003e, D613-D621 (2021).\u003c/li\u003e\n\u003cli\u003eJassal, B.\u003cem\u003e et al.\u003c/em\u003e The reactome pathway knowledgebase. \u003cem\u003eNucleic Acids Res\u003c/em\u003e \u003cstrong\u003e48\u003c/strong\u003e, D498-D503 (2020).\u003c/li\u003e\n\u003cli\u003eGene Ontology, C. Gene Ontology Consortium: going forward. \u003cem\u003eNucleic Acids Res\u003c/em\u003e \u003cstrong\u003e43\u003c/strong\u003e, D1049-56 (2015).\u003c/li\u003e\n\u003cli\u003eFix, E. \u0026amp; Hodges, J.L.J.I.S.R.R.I.d.S. Discriminatory analysis. Nonparametric discrimination: Consistency properties. \u003cstrong\u003e57\u003c/strong\u003e, 238-247 (1989).\u003c/li\u003e\n\u003cli\u003eGrant, C.E., Bailey, T.L. \u0026amp; Noble, W.S. FIMO: scanning for occurrences of a given motif. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 1017-8 (2011).\u003c/li\u003e\n\u003cli\u003eConsortium, E.P. The ENCODE (ENCyclopedia Of DNA Elements) Project. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e306\u003c/strong\u003e, 636-40 (2004). \u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"genome-biology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"gbio","sideBox":"Learn more about [Genome Biology](https://genomebiology.biomedcentral.com/)","snPcode":"13059","submissionUrl":"https://submission.springernature.com/new-submission/13059/3","title":"Genome Biology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"respiratory infection, SARS-CoV-2 Omicron, genome-wide association study, carboxypeptidase Q, CPQ gene, functional validation","lastPublishedDoi":"10.21203/rs.3.rs-8104260/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8104260/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eHost genetic factors may play a critical role in modulating severity of respiratory infections. However, previous studies have often been limited by pathogen heterogeneity and exposure misclassification.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eUtilizing a relatively homogenous Chinese population consists of 5,151 individuals with first-time infection by syndrome coronavirus 2 (SARS-CoV-2) Omicron variant, we conducted a genome-wide association study and identified a novel locus at 8q22.1 (rs7817424, P\u0026thinsp;=\u0026thinsp;4.60\u0026times;10⁻⁸) associated with infection severity. Integrating results from gene mapping and the similarity-based gene prioritization suggested carboxypeptidase Q (\u003cem\u003eCPQ)\u003c/em\u003e gene as the likely causal gene. Single-cell RNA sequencing and transcription factor motif analyses revealed differential CPQ expression in lung immune cells, particularly tissue-resident macrophages and monocyte-derived macrophages, implicating innate immune pathways in severe disease. Functional experiments demonstrated that CPQ overexpression in THP-1 cells suppresses LPS-induced pro-inflammatory cytokines TNF-α and IL-6, while systemic inflammation mouse model showed reduced \u003cem\u003eCPQ\u003c/em\u003e expression in lung tissues during severe pneumonia.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eThis study identifies establish \u003cem\u003eCPQ\u003c/em\u003e as a novel genetic determinant for respiratory infection severity and uncover its previously unrecognized anti-inflammatory role, highlighting its potential as a therapeutic target for controlling hyperinflammatory responses in COVID-19 and beyond.\u003c/p\u003e","manuscriptTitle":"Genetic and functional evidence identify Carboxypeptidase Q as a regulator of inflammation in respiratory infection","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-12 16:58:14","doi":"10.21203/rs.3.rs-8104260/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-03-25T14:55:50+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-12T04:42:01+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-11T09:28:56+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"162164878111920570920360906429050947176","date":"2026-02-15T15:52:57+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"20737282590843139588594266293153720644","date":"2026-02-13T17:08:40+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"130070147512591595283425158628552471864","date":"2026-02-09T21:57:15+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-02-09T18:15:27+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-11-17T11:54:48+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-11-13T13:16:56+00:00","index":"","fulltext":""},{"type":"submitted","content":"Genome Biology","date":"2025-11-13T09:41:48+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"genome-biology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"gbio","sideBox":"Learn more about [Genome Biology](https://genomebiology.biomedcentral.com/)","snPcode":"13059","submissionUrl":"https://submission.springernature.com/new-submission/13059/3","title":"Genome Biology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"9deee648-d0e8-4ec9-9dbb-811fc0ad88eb","owner":[],"postedDate":"February 12th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[],"tags":[],"updatedAt":"2026-03-25T15:09:42+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-12 16:58:14","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8104260","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8104260","identity":"rs-8104260","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}