Identification of sepsis-related genes by integrating eQTL data with Mendelian randomization analysis

doi:10.21203/rs.3.rs-4964121/v1

Download PDF

Research Article

Identification of sepsis-related genes by integrating eQTL data with Mendelian randomization analysis

https://doi.org/10.21203/rs.3.rs-4964121/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

Sepsis is defined as a life-threatening organ dysfunction caused by a dysfunctional host response to infection and is associated with a high mortality. However, there is currently no effective treatment strategy for sepsis.

Methods

We obtained GSE263789, GSE54514 and GSE66099 from the Gene Expression Omnibus (GEO) database and selected differentially expressed genes (DEGs). We extracted expression quantitative trait loci (eQTL) as exposure and sepsis GWAS as outcome from the IEU Open GWAS database. MR analysis was used to assess causality between eQTL and sepsis. The overlapping genes of DEGs with significant eQTL were identified as key genes. Enrichment analysis and immune cell infiltration analysis were performed and the expression of key genes was verified in a validation cohort.

Results

The 18 genes were identified as sepsis-related key genes, including 11 up-regulated genes (SEMA4A, LRPAP1, FAM89B, TOMM40L, SLC22A15, MACF1, MCTP2, NTSR1, PNKD, ACTR10, CPNE3) and 7 down-regulated genes (IKZF3, TNFRSF25, HDC, HCP5, LYRM4, TFAM, RPS15A). Enrichment analyses showed that these key genes are mainly involved in biological processes related to immune and inflammatory response. Compared with healthy controls, the abundance of neutrophils and activated mast cells increased in the sepsis group. Most of the key genes are correlated with immune cells, including neutrophils, CD8 T cells, resting NK cells, plasma cells, memory B cells, and macrophage subtypes.

Conclusion

By combining bioinformatics and MR analysis, we identified key genes associated with sepsis, enhancing our understanding of the genetic pathogenesis of sepsis and providing new insights into therapeutic targets for sepsis.

sepsis

transcriptome

eQTL

Mendelian randomization

immune cell infiltration

Sepsis, a condition characterized by a life-threatening organ dysfunction due to a dysregulated host response to infection, remains a major cause of death in critical illness and poses a significant threat to human health worldwide[1, 2]. Treatment of sepsis is currently focused on continuous fluid resuscitation, organ support, and anti-infection therapy[3]. There is currently no specific treatment for patients with sepsis, and most clinical trials of potential therapies have failed to improve outcomes[4, 5]. Therefore, it is imperative to deeply understand the pathogenesis of sepsis and explore novel therapeutic strategies.

Increasing evidence supports the central role of immuno-inflammatory response in the progression of sepsis[6]. Sustained immunosuppression is thought to be a predisposing factor for secondary infection and increased mortality in patients with sepsis[7]. The interaction of pro-inflammatory and anti-inflammatory response can lead to excessive inflammation, immunosuppression, and potential secondary complications[8]. The hyper-inflammatory response is characterized by the release of pro-inflammatory mediators and activation of the complement and coagulation systems, leading to intravascular thrombosis, disseminated intravascular coagulation, and multiple organ failure[9]. The anti-inflammatory response is characterized by impaired immune cell function due to effector cell apoptosis, T cell exhaustion, reduced monocyte HLA-DR expression, increased expression of suppressor cells, and suppressed transcription of pro-inflammatory gene[10]. Exploring key genes and immuno-inflammatory profiles may contribute to a better understanding of the pathogenesis of sepsis.

However, the underlying mechanisms involved in the immuno-inflammatory response of sepsis have not been fully elucidated. There is a lack of large-scale, multicentre clinical studies to validate the relationship between the immune system and sepsis.

Genome-wide association studies (GWAS) have identified potential genetic variants associated with the risk of sepsis[11], but the direct causal relationship between these genetic variants and sepsis remains largely unclear. Expression quantitative trait loci (eQTL) are genetic variants that affect gene expression and play a key role in elucidating the biological functions of genetic variations[12].

Mendelian randomization (MR) is a method that explores potential causality between an exposure and an outcome by using genetic variants as the instrumental variables (IVs)[13]. Compared with traditional statistical methods, MR reduces confounding and reverse causation[14], and is increasingly used in the exploration of pathological mechanisms[15].

In this study, we integrated bioinformatics with MR analysis to identify key genes that play an important role in the development of sepsis. Through enrichment analysis and immune cell infiltration analysis, the study will provide important insights into the immune-mediated pathogenesis of sepsis, exploring novel therapeutic strategies for sepsis.

Study design

In this study, we extracted DEGs from sepsis-related datasets GSE263789 and GSE54514. We applied a two-sample MR analysis using data from the GWAS to explore the causal relationship between eQTL and sepsis. By overlapping DEGs and significant eQTLs, we identified key genes for sepsis. Subsequently, Gene ontology (GO) enrichment, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways, gene set enrichment analysis (GSEA) and immune cell infiltration analysis were performed to reveal the role of these key genes in the pathogenesis of sepsis. Finally, GSE66099 was used as the validation set to verify whether there were differences in the expression of key genes between sepsis patients and healthy controls. The flowchart of this study is presented in Fig. 1.

Three fundamental criteria must be satisfied when using MR analysis: 1) the selected IVs must be strongly associated with the exposures; 2) the selected IVs should be independent of the confounders; 3) the selected IVs could only influence the outcomes via the exposure of interest, not via other pathways[16].

Data source

For the current study, three sepsis-related datasets, namely GSE263789 (Platforms: GPL570[HG-U133_Plus_2] Affymetrix Human Genome U133 Plus 2.0 Array), GSE54514 (Platforms: GPL6947 Illumina HumanHT-12 V3.0 expression beadchip) and GSE66099 (Platforms: GPL570[HG-U133_Plus_2] Affymetrix Human Genome U133 Plus 2.0 Array), have been downloaded from Gene Expression Omnibus (GEO) (https://www.ncbi.nlm.nih.gov/geo/). Supplementary Table 1 contains information regarding the three gene expression datasets. GSE263789 and GSE54514 were merged as the training set, and GSE66099 was used as the validation set.

In the training set, GSE263789 consists of 103 patients, including 82 patients with sepsis and 21 normal controls; GSE54514 includes 163 patients, including 127 patients with sepsis and 36 normal controls; while a validation group consisting of 199 patients with sepsis and 47 normal controls was obtained from GSE66099.

Summary eQTL data utilized in this study were retrieved from the GWAS Catalog website (https://gwas.mrcieu.ac.uk/). Summary-level statistical data for sepsis were derived from the genetic association database of the GWAS summary dataset (IEU) (https://gwas.mrcieu.ac.uk/). The sepsis-related dataset (ieu-b-4980) consist of 11,643 cases and 474,841 controls, including 12243,539 SNPs.

Because all the data used in this study is publicly available, ethical approval for this study was granted based on the original analyses conducted.

Identification of DEGs

Data normalization and standardization were performed using gene expression matrices and annotation files downloaded from the GEO database. The "limma" package along with the combat function from the “sva” package were used to adjust for batch effects and merge GSE263789 and GSE54514 datasets. The "prcomp" package was used for principal component analysis (PCA) to visualize the batch correction. We then used the "limma" package to select the DEGs between sepsis and control cohort in the combined dataset, with the following criterion: adjust p value < 0.05 and |log fold change (FC)| >1. Visualization of DEGs was achieved using the “pheatmap” and “ggplot2” R packages to generate heatmaps and volcano plots, respectively.

Selection of IVs

To improve the accuracy and validity of the causality between eQTL and sepsis risk, SNPs selected as IVs for MR analyses were based on the following criteria: (1) SNPs strongly associated with eQTL were extracted as candidate IVs using a P-value threshold of < 5.0 × 10^− 8; (2) SNPs were eliminated if they have a linkage disequilibrium with other SNPs (linkage disequilibrium r² < 0.001, within 10-Mb distance); (3) PhenoScanner database (http://www.phenoscanner.medschl.cam.ac.uk/phenoscanner) was employed to exclude SNPs associated with confounding genes; (4) The F-statistic for each SNP was calculated using F = R²(n-1-k)/(1-R²)k, where R², K, n refers to the interpreted variance of the IVs, the number of IVs for analysis, and the number of samples, respectively. Only the SNPs with an F statistic > 10 were considered reliable IVs; (5)palindromic or incompatible SNPs and SNPs directly associated with sepsis were excluded. Detailed information on those IVs is shown in Supplementary Table 1.

MR analysis and identification of key genes

MR analysis was performed using R software with the “TwoSampleMR” package. We employed inverse variance weighting (IVW), MR-Egger, weighted median, simple mode and weighted model methods for MR analysis to assess causality between eQTL and sepsis. IVW method was considered as the primary approach, supplemented by other methods. The IVW method calculates a weighted average of Wald ratio estimates and provides the most convincing estimates in the absence of directional pleiotropy[17]. The MR-Egger method can counter the possible horizontal pleiotropy and provide a robust estimate[18]. Weighted median provides reliable causal estimates especially when at least 50% IVs are valid[19]. Simple mode guarantees the robustness for pleiotropy despite being less powerful than IVW[20]. For genetic variables that do not conform to the hypothesis of pleiotropy, the weighted mode method is applicable[21].

A series of sensitivity analyses were conducted to verify the robustness of our MR analysis. Scatter plots, forest plots, and funnel plots were generated to visualize the results. Heterogeneity was assessed using Cochran's Q test, with funnel plots visualizing significant heterogeneity among the selected SNPs. Heterogeneity among IVs needs to be considered when the I² statistic > 50% or the p value of Cochran's Q test < 0.05. The MR Egger test was conducted to evaluate pleiotropic effects, where an MR-Egger intercept P < 0.05 indicated substantial horizontal pleiotropy. The leave-one-out analysis was performed to assess the potential influence of individual SNP on the causality estimated by MR analysis.

Sepsis-related key genes, including both up-regulated and down-regulated genes, were identified by the overlap of DEGs with significant eQTL genes from MR analysis, and visualized by the “Veen” diagram.

Functional and pathway enrichment analyses

Enrichment analysis was applied to verify the possible functions of potential targets. Gene ontology, which is divided into biological process (BP), cellular component (CC), and molecular function (MF), was used to investigate the biological process of DEGs. Potential signaling pathways were explored using KEGG analysis. The “clusterProfiler” R package was used for GO annotation and KEGG pathway enrichment analysis of key genes.

Correlation Analysis Between key genes and Infiltrated Immune Cells

CIBERSORT algorithm was used to evaluate the immune cells signature of 22 types of immune cells in sepsis, and to explore the correlation between key genes and immune cell infiltration. The correlation between sepsis-related key genes and immune cell signatures was calculated by Spearman correlation analysis, and the results were visualized by ggplot2.

GSEA enrichment analysis

GSEA enrichment analysis was employed to identify the association between key genes and signaling pathways. This analysis is implemented in the ‘clusterProfiler’ package in R. Normalized enrichment score (NES) and false discovery rate (FDR) were used to quantify enrichment magnitude and statistical significance, respectively[22]. Adjusted p-value < 0.05 was considered statistically significant.

Validation group differential analysis

The dataset GSE66099 was read using R software (version 4.3.2) to verify whether there were differences in the expression of key genes between sepsis and healthy controls, and to compare the results with our MR analysis.

Identification of DEGs

We combined the expression profiles of sepsis samples from the GSE26378, GSE54514, and GSE 66099 datasets to form a dataset containing 408 sepsis samples and 104 controls. After normalizing to reduce batch effects, we identified 4130 DEGs from the combined GSE26378 and GSE54514 datasets, with 2027 up-regulated DEGs and 2103 down-regulated DEGs in sepsis samples compared to controls. Supplementary Table S1 provides detailed information on DEGs, which are visually represented in the heatmap and the volcano plot (Fig. 2). The heatmap displays the top 50 up-regulated DEGs and the top 50 down-regulated DEGs (Fig. 2A).

MR analysis

By implementing a rigorous set of filtering criteria, we ultimately obtained 24766 SNPs as instrumental variables (Supplementary Table S2 provides detailed information of the IVs). The MR analysis identified 198 sepsis-related genes, of which 98 genes were associated with an increased risk of sepsis and 100 genes with a reduced risk of sepsis (Supplementary Table S3 provides detailed information).

By intersecting these eQTLs with DEGs, we obtained key genes associated with sepsis. There were 11 up-regulated genes (SEMA4A, LRPAP1, FAM89B, TOMM40L, SLC22A15, MACF1, MCTP2, NTSR1, PNKD, ACTR10, CPNE3) and 7 down-regulated genes (IKZF3, TNFRSF25, HDC, HCP5, LYRM4, TFAM, RPS15A), as shown in Fig. 3.

The results of MR analysis by inverse-variance weighted (IVW) method revealed that 11 up-regulated key genes were significantly positively correlated with sepsis (Fig. 4). Specifically, SEMA4A(OR = 1.062; 95% CI:1.009 to 1.118; P = 0.022), LRPAP1 (OR = 1.104; 95% CI:1.023 to 1.192; P = 0.011), FAM89B (OR = 1.104; 95% CI:1.028 to 1.185 ; P = 0.007), ACTR10 (OR = 1.088; 95% CI:1.012 to 1.170 ; P = 0.023), CPNE3 (OR = 1.080; 95% CI:1.007 to 1.158; P = 0.031), MACF1 (OR = 1.077; 95% CI:1.012 to 1.147; P = 0.020), MCTP2 (OR = 1.076; 95% CI:1.012 to1.145; P = 0.020), NTSR1 (OR = 1.075; 95% CI: 1.009 to 1.145; P = 0.025), PNKD(OR = 1.088; 95% CI: 1.020 to 1.161; P = 0.011), SLC22A15(OR = 1.051; 95% CI: 1.005 to 1.099; P = 0.029) and TOMM40L(OR = 1.131; 95% CI: 1.064 to 1.201; P < 0.001); Conversely, 7 down-regulated key genes showed a significant negative causal relationship with sepsis, namely IKZF3 (OR = 0.935; 95% CI:0.879 to 0.994; P = 0.032), HDC (OR = 0.913; 95% CI:0.838 to 0.994; P = 0.036), HCP5 (OR = 0.823; 95% CI:0.692 to 0.978; P = 0.027), LYRM4 (OR = 0.959; 95% CI:0.922 to 0.997 ; P = 0.033), RPS15A (OR = 0.954; 95% CI:0.914 to 0.996 ; P = 0.033), TFAM (OR = 0.924; 95% CI:0.871 to 0.979 ; P = 0.008) and TNFRSF25(OR = 0.886; 95% CI:0.788 to 0.997; P = 0.044) (Fig. 4).

The results of the WM, MR-Egger, Simple model and Weighted model analysis were consistent with the IVW method in estimating the risk of sepsis (Supplementary Table S3). Scatter plots for each gene were visualized in Fig. 5 and Supplementary Fig. 1.

Sensitivity analysis

Sensitivity analysis revealed no evidence of underlying heterogeneity or horizontal pleiotropy in this study. We performed a Venn analysis to explore sepsis-related key genes. Cochran's Q-test was used to assess the heterogeneity among IVs. No evidence of heterogeneity was detected in our results (Supplementary Table S4). Visualized funnel plots are presented in Fig. 6 and Supplementary Fig. 2. In addition, the MR-Egger intercept test demonstrated the absence of horizontal pleiotropy (Supplementary Table S5). The results of the leave-one-out sensitivity analyses demonstrated that no single SNP could significantly influence the causal estimates (Fig. 7 and Supplementary Fig. 3). Forest plots for each gene are shown in Fig. 8 and supplementary Fig. 4. We visualized the chromosomal distribution of key genes in Fig. 9.

GO and KEGG enrichment analysis

We performed a Venn analysis to explore sepsis-related key genes and further functionally annotated these key genes.

GO and KEGG enrichment analyses were performed to uncover the biological processes and pathways associated with these key genes (Fig. 10). GO enrichment analysis includes biological processes (BP), molecular functions (MF) and cellular components (CC). In terms of biological processes, the genes are primarily involved in small molecule catabolic process, catecholamine metabolic process, catechol-containing compound metabolic process, regulation of neurotransmitter secretion, regulation of axon extension, biogenic amine metabolic process, regulation of neurotransmitter transport, regulation of extent of cell growth, phenol-containing compound metabolic process and amine metabolic process. Cellular component analysis showed that transcription coregulator binding plays an important role in sepsis. In molecular function analysis, these genes were primarily enriched in primary lysosome, azurophil granule and mitochondrial protein-containing complex. KEGG analysis highlighted significant involvement of key genes in the histidine metabolism and huntington disease.

Assessment of immune cell infiltration in sepsis

The functional and pathway analysis of sepsis-related genes shows a close relationship with inflammatory and immune processes. The CIBERSORT algorithm was used to infer immune cell characteristics and to explore the correlation between sepsis-related genes and immune cell infiltration. Figure 11A shows the percentage of 22 types of immune cells in each sepsis sample and control sample. The results showed that the proportion of activated mast cell and neutrophils was significantly different between sepsis and control group (Fig. 11B). Additionally, spearman correlation analysis showed that most of the key genes were associated with immune cells such as neutrophils, CD8 T cells, resting NK cells, plasma cells, memory B cells and macrophage subtypes (Fig. 11C).

GSEA enrichment analysis

We used GSEA analysis to analyse the biological functions of key genes. The top 5 enrichment pathways of key genes are shown in Fig. 12.

The top five active pathways in the high SEMA4A expression group are Fc gamma R-mediated phagocytosis, insulin signaling pathway, lysosome, neurotrophin signaling pathway and regulation of actin cytoskeleton (Fig. 12A); The top five active pathways in the low SEMA4A expression group are DNA replication, intestinal immune network for IgA product, parkinsons disease, porphyrin and chlorophyll metabolism and ribosome (Fig. 12B).

High expression of IKZF3 was closely related to antigen processing and presentation, cell adhesion molecules cams, graft versus host disease, primary immunodeficiency and ribosome (Fig. 12C). The low-expression group of IKZF3 was predominantly enriched in adipocytokine signaling pathway, lysosome, pantothenate and CoA biosynthesis, starch and sucrose metabolism and Toll like receptor signaling pathway (Fig. 12D).

The highly expressed TNFRSF25 plays a central role in allograft rejection, antigen processing and presentation, autoimmune thyroid disease, graft versus host disease and ribosome (Fig. 12E). The low expression of TNFRSF25 is significantly correlated with Fc gamma R-mediated phagocytosis, neurotrophin signaling pathway, regulation of actin cytoskeleton, starch and sucrose metabolism, and Toll like receptor signaling pathway(Fig. 12F).

Notably, high expression of TFAM is mainly enriched with alzheimers disease, graft versus host disease, oxidative phosphorylation, parkinsons disease and ribosome (Fig. 12G). However, at low TFAM expression, it is closely related to Fc gamma R-mediated phagocytosis, focal adhesion, leukocyte transendothelial migration, regulation of actin cytoskeleton and tight junction(Fig. 12H).

Validation group differential analysis

The expression of eighteen key genes identified in the MR analysis was validated in GSE66099. The results showed that SEMA4A, LRPAP1, FAM89B, TOMM40L, SLC22A15, MACF1, MCTP2, NTSR1, PNKD, ACTR10 and CPNE3 were significantly up-regulated in the sepsis group, while IKZF3, TNFRSF25, HDC, HCP5, LYRM4, TFAM and RPS15A were significantly down-regulated (Fig. 13). The expression trend of these eighteen genes between the two groups was consistent with the results proposed in the MR analysis, providing greater credibility to the MR results.

Sepsis is defined as a fatal organ dysfunction resulting from a dysregulated immune response to infection[23]. It has a high rate of morbidity and mortality and is a serious public health problem worldwide[23]. However, the genetic mechanism of sepsis is not fully understood, and there are currently no effective strategies against sepsis. Therefore, exploring novel biomarkers and therapeutic targets is critical for the integrated management of sepsis. In this study, we combined bioinformatics analysis with MR analysis to identify key genes involved in the pathogenesis of sepsis. Our findings contribute to a better understanding of the immunomodulatory mechanisms underlying sepsis and reveal potential therapeutic targets for sepsis.

This study identified 11 up-regulated and 7 down-regulated key genes associated with sepsis. Up-regulated genes include SEMA4A, LRPAP1, FAM89B, TOMM40L, SLC22A15, MACF1, MCTP2, NTSR1, PNKD, ACTR10 and CPNE3, and down-regulated genes include IKZF3, TNFRSF25, HDC, HCP5, LYRM4, TFAM and RPS15A. Through functional enrichment analysis and immune cell infiltration analysis, we discovered that these genes predominantly engage in inflammatory and immune processes. Activated mast cells and neutrophils were abundant in patients with sepsis compared to healthy controls. Notably, most of the key genes exhibited correlations with diverse immune cells, including neutrophils, CD8 T cells, resting NK cells, plasma cells, memory B cells and macrophage subtypes.

SEMA4D is a homodimeric protein belonging to the fourth class of semaphorin protein family, has immunoregulatory function and plays an important role in T cell activation, antibody production, and intercellular adhesion[24, 25]. A recent study showed that asiaticoside could alleviate lipopolysaccharide-induced acute lung injury by blocking Sema4D/CD72 pathway[26]. Cui Y, et al.[27] reported that SEMA4D/VEGF surface enhances endothelialization by diminished-glycolysis-mediated M2-like macrophage polarization.

Interferon (IFN) signaling plays a key role in n the restriction or eradication of pathogen invasion[28]. It has been reported that the N-terminus of secreted LRPAP1 effectively binds and causes IFNAR1 degradation that enhances both DNA and RNA viral infections, including herpesvirus HSV-1, hepatitis B virus (HBV), EV71, and beta-coronavirus HCoV-OC43[29].

CD4 T helper cells are able to differentiate into a number of effector subsets that perform diverse functions in adaptive immune responses. Cytokine signaling pathway plays an important role in regulating the differentiation of CD4 T helper cells. Ikaros zinc finger (IkZF) transcription factors are known regulators of immune cell development,especially that of effector CD4 T cell populations[30]. It has been suggested that aiolos may negatively regulate T_H1 differentiation by repressing autocrine IL-2 signaling[31]. Various studies have shown that Th17 is inextricably linked to the pathogenesis of sepsis[32, 33]. During the course of sepsis, Th17 regulates the inflammatory response by secreting pro-inflammatory cytokines, recruiting neutrophils, activating innate immune cells, and enhancing B lymphocyte function[34]. Using Aiolos-deficient mice, Quintana FJ,et.al demonstrated that Aiolos promotes Th17 differentiation by directly silencing Il2 expression in vitro and in vivo[35].

T follicular helper (Tfh) cells is important to promote the development of germinal centers and maturation of high affinity antigen-specific B cells[36]. Impaired B-cell maturation contributes to reduced B cell numbers and poor prognosis in sepsis, the numbers of circulating Tfh cell positively correlated with the numbers of mature B cell and immunoglobulin concentrations[37]. IkZF transcription factors Aiolos regulated Tfh cell differentiation by interacting with STAT3 to form a transcriptional complex capable of inducing Bcl-6 expression in CD4 T cell populations[38]. As an antagonist of IL-2 signaling, Aiolos has been shown to be a positive regulator of T_FH cell differentiation[39].

TNFRSF25 is a member of the TNF receptor superfamily (TNFRSF) and binds to the TNF-like protein TL1A. This receptor is preferentially expressed in lymphocyte-rich tissues and may play a role in regulating lymphocyte homeostasis[40]. TNFRSF25 signaling has been shown to stimulate NF-kappaB, which in turn regulate cell apoptosis[41]. Activation of TNFRSF25 in primary T cells has been found to stimulate proliferation, cell activation, and effector function[42].

LYRM4 is necessary to maintain the stability and activity of the human cysteine desulfurase complex NFS1-LYRM4-ACP. Disruption of this gene can negatively affect mitochondrial and cytosolic iron homeostasis[43].

TFAM(transcription factor A, mitochondria) is a major regulator of mitochondrial function, and its expression is responsible for mtDNA transcription initiation[44]. TFAM is a ~ 24 kDa protein with non-specific DNA-binding properties. After synthesis as a precursor protein (~ 29 kDa) in the cytoplasm, TFAM is shuttled to the mitochondria, where mature TFAM is generated by cleavage of a targeting sequence (~ 5 kDa) by a processing peptidase in the mitochondrial matrix[45].Insufficient TFAM is associated with the failure of mitochondrial biological energy supply and apoptosis, leading to mitochondrial diseases[46]. Mitochondrial dysfunction in sepsis has been reported to be associated with diminished intramitochondrial TFAM[47]. Studies have shown that TFAM plays a central role in restoring mitochondrial function in sepsis-induced organ failure[48]. Deng Z,et.al showed that melatonin attenuated sepsis-induced acute kidney injury by promoting mitophagy through SIRT3-mediated TFAM deacetylation[49]. Zhang F,et.al reported that TFAM-Mediated mitochondrial transfer of mesenchymal stem cells (MSCs) improved the permeability barrier in sepsis-associated acute lung injury[50].

GO and KEGG enrichment analysis revealed key biological processes and pathways associated with sepsis, which are closely related to inflammatory and immune processes. The abundance of activated mast cells and neutrophils were increased in patients with sepsis compared with healthy controls. Consistent with the result of previous studies, sepsis-related genes may play an important role in regulating immune cell infiltration[51]. Spearman’s correlation analysis was performed to evaluate the correlation between key DEGs and infiltrating immune cell types.

During sepsis, neutrophils play a critical role in the host's inflammatory response against invading pathogens[52]. Activated neutrophils exert effector functions primarily through phagocytosis, degranulation and releasing neutrophil extracellular traps (NETs)[53–55]. However, excessive neutrophil activation and NET release can further induce inflammation and organ injury, leading to the progression of sepsis[56]. Neutrophils exhibit increased lifespan and impaired migration, resulting in overwhelming vascular inflammation through the release of cytokines, reactive oxygen species (ROS) and NETs[57]. Neutrophils and NETs induce pro-inflammatory and pro-angiogenic responses in endothelial cells via NF-κB activation[58]. Neutrophils and NETs degrade glycocalyx present on the surface of endothelial cells and increase endothelial permeability through junction cleavage, high expression of adhesion molecules, and apoptosis[59]. Additionally, neutrophils and NETs induce a pro‐coagulant endothelial cell phenotype via degradation of the anti‐coagulation system and up‐regulation of tissue factor[60]. Mast cells stimulated by TNF-α can release cytokines, proteases, histamine and heparinase, contributing to further glycocalyx degradation[61]. Endothelial dysfunction leads to impaired microcirculatory blood flow, tissue hypoperfusion, and life-threatening organ failure in the late phase of sepsis.

Neutrophils are often found to be elevated in sepsis, which is thought to be associated with the inhibition of apoptosis in neutrophils and the release of immature neutrophils[62]. In this study, key genes were significantly associated with infiltration of multiple immune cells, especially neutrophils. These key genes may be involved in the development of sepsis as immunomodulatory molecules. More detailed and in-depth mechanisms of how key genes regulate neutrophils are worth investigating. In a mouse model of CLP-induced septic peritonitis, mast cells were systematically and locally activated and released pre-stored inflammatory mediators [63]. In addition, MCs have shown immunological implications in regulating cell death in sepsis[64]. Yue J, et al.[65] showed that MCs activation could mediate blood-brain barrier impairment and cognitive dysfunction in septic mice in a histamine-dependent pathway.

Depletion of B cells, CD4 and CD8 T cells due to increased apoptosis accounts for "lymphocyte exhaustion" and immunosuppression in sepsis[66, 67]. In sepsis, lymphopaenia has been observed to be associated with increased mortality[68, 69]. Increased expression of inhibitory receptors on lymphocytes in patients with sepsis directly affects their ability to respond to infection[70]. In addition, the function of Th1, Th2, and Th17 cells has been shown to be suppressed in patients suffering from sepsis[71]. Several experimental and clinical trials have shown that sepsis enhances Treg function, which suppresses monocytes, neutrophils and effector T cells, leading to immune paralysis and ultimately septic death[72, 73]. In a murine sepsis model, suppression of T cell autophagy lead to decreased viability and function of T cells through accelerated apoptosis[74]. Reversing lymphocyte apoptosis has become a challenging aspect of sepsis treatment.

Sepsis is characterized by M1-like macrophage activation. Enhanced autophagy has been reported to inhibit M1-like macrophage polarization and reduce pro-inflammatory cytokines, thereby alleviating CLP-induced sepsis[75]. Patients with sepsis experience immune disorders that manifest as pro-inflammatory response and immunosuppression, which occur sequentially or simultaneously[76]. Sepsis-related deaths primarily occur during the period of immunosuppression[77]. Apoptosis of immune cells is an important factor in the development of immunosuppression in sepsis[78]. Abnormalities in the counts and functions of B cells lead to impaired B cell-mediated immune response, exacerbating the development of sepsis[79]. It has been shown that depletion of memory B cells contributes to sepsis-induced immunosuppression and increases the risk of secondary infection. Reduced circulating B-cell and IgM levels are associated with reduced survival in patients with sepsis[80].

Correlation analysis showed that neutrophils, CD8 T cell, resting NK cells, memory B cells and plasma cells signatures were associated with most key genes. Most up-regulated genes, including SEMA4A, LRPAP1, FAM89B, TOMM40L, SLC22A15, MACF1, MCTP2, NTSR1, ACTR10 and CPNE3, were positively correlated with neutrophils. Down-regulated genes including IKZF3, TNFRSF25, HDC, HCP5 and LYRM4 were negatively correlated with neutrophils. These results highlighted the complex interactions between sepsis-related genes and immune cells, emphasizing the importance of further research into immune-related pathogenesis of sepsis.

This study has several strengths. To our knowledge, this study is the first to combine bioinformatics with MR analysis to explore the genetic pathogenesis of sepsis. The MR method avoided the bias caused by confounders and reverse causality in conventional observational studies. In addition, our study reveals the potential role of key genes and immune cell signatures in sepsis, which may provide novel therapeutic targets for the clinical management of patients with sepsis.

However, there are still some limitations. Firstly, the datasets we analyzed were downloaded from the GEO, so detailed clinical data was not available. Secondly, although we combined bioinformatics with MR analysis to identify key genes, the exact role of these genes in the pathogenesis of sepsis needs to be further elucidated through in vitro and in vivo experiments. Finally, because of the diversity of infectious sources, ethnicity, severity and course of sepsis patients, our findings may not be generalizable to all sepsis patients. Therefore, collecting more clinical specimens and conducting more in-depth analysis will become one of our future research work.

In summary, our study combined bioinformatics and MR analysis to identify key genes associated with sepsis. Neutrophils and actived mast cells were found to correlate with most of the key genes. These findings may provide novel biomarkers and potential therapeutic targets for sepsis, deepening our understanding of the immune-mediated pathogenesis of sepsis.

Acknowledgments

The authors thank the IEU OpenGWAS Project for providing summary-level GWAS data.

Author contributions

Yuling Li and Jian Kang contributed to design the study. Chao Wen wrote the original manuscript. Dongliang Yang and Hongyan Guo performed the statistical analysis. Chuankun Dong, Qingyun Peng and Jiangwei Zhao contributed to literature search. Runan Wang, Yingqi Li and Yuanhao Li contributed to methodology and visualization. Suosuo Yang and Yanbo Ren participated in the revision of the manuscript. All authors have reviewed the manuscript and approved the submitted version.

Funding

None.

Data availability

This study used publicly available datasets. The summary data from the Genome-Wide Association Study used in this research can be obtained from the respective organizations that provided them. The detailed information about these data sets are provided in the methods section of this article.

Conflict of interest The authors declare no competing interests.

Ethics approval and consent to participate Summary-level data for Mendelian randomization analysis in this study were obtained from published studies. All studies were conducted in accordance with the Declaration of Helsinki and were conducted with the approval of institutional ethics committees, and therefore did not require additional ethical approval.

D.C. Angus, T. van der Poll (2013) Severe sepsis and septic shock. The New England journal of medicine 369(9):840-51
S. Esposito, G. De Simone, G. Boccia, et al. (2017) Sepsis and septic shock: New definitions, new diagnostic and therapeutic approaches. Journal of global antimicrobial resistance 10:204-212
A. Mushtaq, F. Kazi (2022) Updates in sepsis management. The Lancet. Infectious diseases 22(1):24
M.S. Rizvi, A. Gallo De Moraes (2021) New Decade, Old Debate: Blocking the Cytokine Pathways in Infection-Induced Cytokine Cascade. Critical care explorations 3(3):e0364
H. Liang, H. Song, R. Zhai, et al. (2021) Corticosteroids for Treating Sepsis in Adult Patients: A Systematic Review and Meta-Analysis. Frontiers in immunology 12:709155
X. Zhang, Y. Zhang, S. Yuan, et al. (2024) The potential immunological mechanisms of sepsis. Frontiers in immunology 15:1434688
L.K. Torres, P. Pickkers, T. van der Poll (2022) Sepsis-Induced Immunosuppression. Annu Rev Physiol 84:157-181
W.J. Wiersinga, T. van der Poll (2022) Immunopathophysiology of human sepsis. EBioMedicine 86:104363
M. Cao, G. Wang, J. Xie (2023) Immune dysregulation in sepsis: experiences, lessons and perspectives. Cell death discovery 9(1):465
D. Liu, S.Y. Huang, J.H. Sun, et al. (2022) Sepsis-induced immunosuppression: mechanisms, diagnosis and current treatment options. Mil Med Res 9(1):56
F. Degenhardt, D. Ellinghaus, S. Juzenas, et al. (2022) Detailed stratified GWAS analysis for severe COVID-19 in four European populations. Human molecular genetics 31(23):3945-3966
R.C. Jansen, J.P. Nap (2001) Genetical genomics: the added value from segregation. Trends Genet 17(7):388-91
S. Burgess, A. Butterworth, S.G. Thompson (2013) Mendelian randomization analysis with multiple genetic variants using summarized data. Genetic epidemiology 37(7):658-65
D.A. Lawlor, R.M. Harbord, J.A. Sterne, et al. (2008) Mendelian randomization: using genes as instruments for making causal inferences in epidemiology. Statistics in medicine 27(8):1133-63
P. Sekula, M.F. Del Greco, C. Pattaro, et al. (2016) Mendelian Randomization as an Approach to Assess Causality Using Observational Data. Journal of the American Society of Nephrology : JASN 27(11):3253-3265
A.G. Boef, O.M. Dekkers, S. le Cessie (2015) Mendelian randomization studies: a review of the approaches used and the quality of reporting. International journal of epidemiology 44(2):496-511
G. Hemani, J. Zheng, B. Elsworth, et al. (2018) The MR-Base platform supports systematic causal inference across the human phenome. eLife 7
S. Burgess, S.G. Thompson (2017) Interpreting findings from Mendelian randomization using the MR-Egger method. European journal of epidemiology 32(5):377-389
J. Bowden, G. Davey Smith, P.C. Haycock, et al. (2016) Consistent Estimation in Mendelian Randomization with Some Invalid Instruments Using a Weighted Median Estimator. Genetic epidemiology 40(4):304-14
R.L. Milne, K.B. Kuchenbaecker, K. Michailidou, et al. (2017) Identification of ten variants associated with risk of estrogen-receptor-negative breast cancer. 49(12):1767-1778
F.P. Hartwig, G. Davey Smith, J. Bowden (2017) Robust inference in summary data Mendelian randomization via the zero modal pleiotropy assumption. International journal of epidemiology 46(6):1985-1998
A. Subramanian, P. Tamayo, V.K. Mootha, et al. (2005) Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proceedings of the National Academy of Sciences of the United States of America 102(43):15545-50
C. Fleischmann-Struzek, L. Mellhammar, N. Rose, et al. (2020) Incidence and mortality of hospital- and ICU-treated sepsis: results from an updated and expanded systematic review and meta-analysis. 46(8):1552-1562
K.T. Maleki, M. Cornillet, N.K. Björkström (2016) Soluble SEMA4D/CD100: A novel immunoregulator in infectious and inflammatory diseases. Clinical immunology (Orlando, Fla.) 163:52-9
H. Kikutani, A. Kumanogoh (2003) Semaphorins in interactions between T cells and antigen-presenting cells. Nature reviews. Immunology 3(2):159-67
J. Zhang, H. Zhao, F. Wang, et al. (2024) Asiaticoside alleviates lipopolysaccharide-induced acute lung injury by blocking Sema4D/CD72 and inhibiting mitochondrial dysfunction in RAW264.7 cell and mice. Naunyn-Schmiedeberg's archives of pharmacology
Y. Cui, X. Jiang, M. Yang, et al. (2023) SEMA4D/VEGF surface enhances endothelialization by diminished-glycolysis-mediated M2-like macrophage polarization. Materials today. Bio 23:100832
A. Hosseini, V. Hashemi, N. Shomali, et al. (2020) Innate and adaptive immune responses against coronavirus. Biomedicine & pharmacotherapy = Biomedecine & pharmacotherapie 132:110859
H. Li, X. Wang, Y. Wang, et al. (2023) Secreted LRPAP1 binds and triggers IFNAR1 degradation to facilitate virus evasion from cellular innate immunity. Signal transduction and targeted therapy 8(1):374
Y. Fan, D. Lu (2016) The Ikaros family of zinc-finger proteins. Acta pharmaceutica Sinica. B 6(6):513-521
M.D. Powell, K.A. Read, B.K. Sreekumar, et al. (2019) Ikaros Zinc Finger Transcription Factors: Regulators of Cytokine Signaling Pathways and CD4(+) T Helper Cell Differentiation. Frontiers in immunology 10:1299
X. Liu, L. Chen, W. Peng, et al. (2024) Th17/Treg balance: the bloom and wane in the pathophysiology of sepsis. Frontiers in immunology 15:1356869
X. Chen, X. Chen, Y. Yang, et al. (2023) Protective role of the novel cytokine Metrnl/ interleukin-41 in host immunity defense during sepsis by promoting macrophage recruitment and modulating Treg/Th17 immune cell balance. Clinical immunology (Orlando, Fla.) 254:109690
V. Kumar (2018) T cells and their immunometabolism: A novel way to understanding sepsis immunopathogenesis and future therapeutics. European journal of cell biology 97(6):379-392
F.J. Quintana, H. Jin, E.J. Burns, et al. (2012) Aiolos promotes TH17 differentiation by directly silencing Il2 expression. Nature immunology 13(8):770-7
F.V. Sjaastad, S.A. Condotta, J.A. Kotov, et al. (2018) Polymicrobial Sepsis Chronic Immunoparalysis Is Defined by Diminished Ag-Specific T Cell-Dependent B Cell Responses. Frontiers in immunology 9:2532
S. Duan, Y. Jiao, J. Wang, et al. (2020) Impaired B-Cell Maturation Contributes to Reduced B Cell Numbers and Poor Prognosis in Sepsis. Shock (Augusta, Ga.) 54(1):70-77
K.A. Read, M.D. Powell, C.E. Baker, et al. (2017) Integrated STAT3 and Ikaros Zinc Finger Transcription Factor Activities Regulate Bcl-6 Expression in CD4(+) Th Cells. 199(7):2377-2387
K.A. Read, D.M. Jones, S. Pokhrel, et al. (2023) Aiolos represses CD4(+) T cell cytotoxic programming via reciprocal regulation of T(FH) transcription factors and IL-2 sensitivity. 14(1):1652
S. Bittner, G. Knoll, M. Ehrenschwender (2017) Death receptor 3 signaling enhances proliferation of human regulatory T cells. 591(8):1187-1195
T.S. Migone, J. Zhang, X. Luo, et al. (2002) TL1A is a TNF-like ligand for DR3 and TR6/DcR3 and functions as a T cell costimulator. Immunity 16(3):479-92
V. Valatas, G. Kolios, G. Bamias (2019) TL1A (TNFSF15) and DR3 (TNFRSF25): A Co-stimulatory System of Cytokines With Diverse Functions in Gut Mucosal Immunity. Frontiers in immunology 10:583
M.T. Boniecki, S.A. Freibert (2017) Structure and functional dynamics of the mitochondrial Fe/S cluster synthesis complex. 8(1):1287
N.A. Bonekamp, M. Jiang (2021) High levels of TFAM repress mammalian mitochondrial DNA transcription in vivo. 4(11)
A.R. Stiles, M.T. Simon, A. Stover, et al. (2016) Mutations in TFAM, encoding mitochondrial transcription factor A, cause neonatal liver failure associated with mtDNA depletion. Molecular genetics and metabolism 119(1-2):91-9
F. Ullah, W. Rauf, K. Khan, et al. (2021) A recessive variant in TFAM causes mtDNA depletion associated with primary ovarian insufficiency, seizures, intellectual disability and hearing loss. 140(12):1733-1751
T. Rahmel, B. Marko, H. Nowak, et al. (2020) Mitochondrial dysfunction in sepsis is associated with diminished intramitochondrial TFAM despite its increased cellular expression. Scientific reports 10(1):21029
E.C. van der Slikke, B.S. Star, M. van Meurs, et al. (2021) Sepsis is associated with mitochondrial DNA damage and a reduced mitochondrial mass in the kidney of patients with sepsis-AKI. Critical care 25(1):36
Z. Deng, M. He, H. Hu, et al. (2024) Melatonin attenuates sepsis-induced acute kidney injury by promoting mitophagy through SIRT3-mediated TFAM deacetylation. Autophagy 20(1):151-165
F. Zhang, X. Zheng, F. Zhao, et al. (2023) TFAM-Mediated mitochondrial transfer of MSCs improved the permeability barrier in sepsis-associated acute lung injury. Apoptosis : an international journal on programmed cell death 28(7-8):1048-1059
P. Conti, A. Caraffa, C.E. Gallenga, et al. (2020) Coronavirus-19 (SARS-CoV-2) induces acute severe lung inflammation via IL-1 causing cytokine storm in COVID-19: a promising inhibitory strategy. Journal of biological regulators and homeostatic agents 34(6):1971-1975
K. Kienle, K.M. Glaser (2021) Neutrophils self-limit swarming to contain bacterial growth in vivo. 372(6548)
M. Sekheri, D. El Kebir, N. Edner, et al. (2020) 15-Epi-LXA(4) and 17-epi-RvD1 restore TLR9-mediated impaired neutrophil phagocytosis and accelerate resolution of lung inflammation. 117(14):7971-7980
B.A. Rosa, M. Ahmed, D.K. Singh (2021) IFN signaling and neutrophil degranulation transcriptional signatures are induced during SARS-CoV-2 infection. 4(1):290
J.M. Adrover, A. Aroca-Crevillén, G. Crainiciuc, et al. (2020) Programmed 'disarming' of the neutrophil proteome reduces the magnitude of inflammation. 21(2):135-144
H. Zhang, D. Wu, Y. Wang, et al. (2024) Ferritin-mediated neutrophil extracellular traps formation and cytokine storm via macrophage scavenger receptor in sepsis-associated lung injury. Cell communication and signaling : CCS 22(1):97
C.L. Zhu, Y. Wang, Q. Liu, et al. (2022) Dysregulation of neutrophil death in sepsis. Frontiers in immunology 13:963955
A. Barkaway, L. Rolas, R. Joulia, et al. (2021) Age-related changes in the local milieu of inflamed tissues cause aberrant neutrophil trafficking and subsequent remote organ damage. Immunity 54(7):1494-1510.e7
Y. Ma, X. Yang, V. Chatterjee, et al. (2019) Role of Neutrophil Extracellular Traps and Vesicles in Regulating Vascular Endothelial Permeability. Frontiers in immunology 10:1037
H. Zhang, Y. Wang, M. Qu, et al. (2023) Neutrophil, neutrophil extracellular traps and endothelial cell dysfunction in sepsis. 13(1):e1170
B.F. Becker, M. Jacob, S. Leipert, et al. (2015) Degradation of the endothelial glycocalyx in clinical settings: searching for the sheddases. British journal of clinical pharmacology 80(3):389-402
M.J. Delano, P.A. Ward (2016) The immune system's role in sepsis progression, resolution, and long-term outcome. Immunological reviews 274(1):330-353
E.J. Seeley, R.E. Sutherland, S.S. Kim, et al. (2011) Systemic mast cell degranulation increases mortality during polymicrobial septic peritonitis in mice. Journal of leukocyte biology 90(3):591-7
L. Ramos, G. Peña, B. Cai, et al. (2010) Mast cell stabilization improves survival by preventing apoptosis in sepsis. Journal of immunology (Baltimore, Md. : 1950) 185(1):709-16
J. Yue, Y. Tan, R. Huan, et al. (2023) Mast cell activation mediates blood-brain barrier impairment and cognitive dysfunction in septic mice in a histamine-dependent pathway. Frontiers in immunology 14:1090288
M. Kox, G.R. Manjeri, J.A. Wagenaars, et al. Nature immunology
C. Cao, M. Yu, Y. Chai (2019) Pathological alteration and therapeutic implications of sepsis-induced immune cell apoptosis. 10(10):782
C. Krautz, S.L. Maier, M. Brunner, et al. (2018) Reduced circulating B cells and plasma IgM levels are associated with decreased survival in sepsis - A meta-analysis. Journal of critical care 45:71-75
R.S. Hotchkiss, K.W. Tinsley, P.E. Swanson, et al. (2001) Sepsis-induced apoptosis causes progressive profound depletion of B and CD4+ T lymphocytes in humans. Journal of immunology (Baltimore, Md. : 1950) 166(11):6952-63
R. Shao, C.S. Li, Y. Fang, et al. (2015) Low B and T lymphocyte attenuator expression on CD4+ T cells in the early stage of sepsis is associated with the severity and mortality of septic patients: a prospective cohort study. Critical care 19(1):308
J. Cabrera-Perez, S.A. Condotta, V.P. Badovinac, et al. (2014) Impact of sepsis on CD4 T cell immunity. Journal of leukocyte biology 96(5):767-77
L.F. Huang, Y.M. Yao, N. Dong, et al. (2010) Association between regulatory T cell activity and sepsis and outcome of severely burned patients: a prospective, observational study. Critical care 14(1):R3
Q. Liu, L. An, Z. Qi, et al. (2017) Increased Expression of Programmed Cell Death-1 in Regulatory T Cells of Patients with Severe Sepsis and Septic Shock: An Observational Clinical Study. 86(5):408-417
T. Oami, E. Watanabe, M. Hatano, et al. (2017) Suppression of T Cell Autophagy Results in Decreased Viability and Function of T Cells Through Accelerated Apoptosis in a Murine Sepsis Model. Critical care medicine 45(1):e77-e85
F. Xu, Y. Ma, W. Huang, et al. (2020) Typically inhibiting USP14 promotes autophagy in M1-like macrophages and alleviates CLP-induced sepsis. 11(8):666
M. Cecconi, L. Evans, M. Levy, et al. (2018) Sepsis and septic shock. Lancet 392(10141):75-87
G.P. Otto, M. Sossdorf, R.A. Claus, et al. (2011) The late phase of sepsis is characterized by an increased microbiological burden and death rate. Critical care 15(4):R183
R.Q. Yao, C. Ren, L.Y. Zheng, et al. (2022) Advances in Immune Monitoring Approaches for Sepsis-Induced Immunosuppression. Frontiers in immunology 13:891024
X. Dong, Q. Liu, Q. Zheng, et al. (2020) Alterations of B Cells in Immunosuppressive Phase of Septic Shock Patients. Critical care medicine 48(6):815-821
X. Dong, H. Tu, X. Bai, et al. (2023) INTRINSIC/EXTRINSIC APOPTOSIS AND PYROPTOSIS CONTRIBUTE TO THE SELECTIVE DEPLETION OF B CELL SUBSETS IN SEPTIC SHOCK PATIENTS. Shock (Augusta, Ga.) 60(3):345-353

No competing interests reported.

SupplementaryFigure1.docx
Supplementary Figure 1 Scatter plots of the causal effect of key genes on the risk of sepsis. (A) FAM89B; (B) TOMM40L; (C) SLC22A15; (D) MACF1; (E) MCTP2; (F) NTSR1; (G) PNKD; (H) ACTR10; (I) CPNE3; (J) HDC; (K) HCP5; (L) RPS15A
SupplementaryFigure2.docx
Supplementary Figure 2 Funnel plots of the causal relationship between key genes and sepsis. (A) FAM89B; (B) TOMM40L; (C) SLC22A15; (D) MACF1; (E) MCTP2; (F) NTSR1; (G) PNKD; (H) ACTR10; (I) CPNE3; (J) HDC; (K) HCP5; (L) RPS15A
SupplementaryFigure3.docx
Supplementary Figure 3 Leave-one-out plots of the causality estimation between key genes and sepsis. (A) FAM89B; (B) TOMM40L; (C) SLC22A15; (D) MACF1; (E) MCTP2; (F) NTSR1; (G) PNKD; (H) ACTR10; (I) CPNE3; (J) HDC; (K) HCP5; (L) RPS15A
SupplementaryFigure4.docx
Supplementary Figure 4 Forest plots for each gene.
SupplementaryTableS1.xlsx
SupplementaryTableS2.csv
SupplementaryTableS3.csv
SupplementaryTableS4.docx
SupplementaryTableS5.docx

Download PDF

Editor assigned by journal
24 Aug, 2024
Submission checks completed at journal
24 Aug, 2024
First submitted to journal
23 Aug, 2024

You are reading this latest preprint version

Identification of sepsis-related genes by integrating eQTL data with Mendelian randomization analysis

Status:

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

Introduction

Materials and methods

Study design

Data source

Identification of DEGs

Selection of IVs

MR analysis and identification of key genes

Functional and pathway enrichment analyses

Correlation Analysis Between key genes and Infiltrated Immune Cells

GSEA enrichment analysis

Validation group differential analysis

Results

Identification of DEGs

MR analysis

Sensitivity analysis

GO and KEGG enrichment analysis

Assessment of immune cell infiltration in sepsis

GSEA enrichment analysis

Validation group differential analysis

Discussion

Conclusion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1