Exploring the Causal Relationship Between Circulating miRNAs and Sepsis Through Mendelian Randomization Analysis

Abstract

Background

MicroRNAs (miRNAs) play a crucial role in the pathophysiology of sepsis, with aberrant expression documented in peripheral blood samples from sepsis patients. This study aims to investigate the causal relationship between circulating miRNAs and sepsis using Mendelian Randomization (MR) approaches.

Methods

We utilized miRNA expression quantitative trait loci (eQTL) data and genome-wide association study (GWAS) data on sepsis. MR analysis was conducted using Inverse Variance Weighted (IVW) and MR-Egger regression methods to identify miRNAs significantly associated with sepsis risk. Potential targets and related pathways of these miRNAs were also analyzed.

Results

IVW and MR-Egger regression analyses revealed that three circulating miRNAs (hsa-miR-125a-5p, hsa-miR-494-3p, and hsa-miR-885-5p) exhibited significant associations with reduced sepsis risk, indicating protective effects. Conversely, hsa-miR-196b-5p, hsa-miR-27b-3p, and hsa-miR-598-3p were linked to increased sepsis risk. The association between hsa-miR-27b-3p and increased sepsis risk was further validated in an independent cohort. Target identification and enrichment analysis indicated that the targets of these miRNAs are closely related to biological processes such as bacterial infection, apoptosis, and the FoxO signaling pathway.

Conclusion

Our study provides novel insights into the causal relationships between circulating miRNAs and sepsis through MR analysis, offering new perspectives for early prevention and treatment strategies for sepsis. These findings provide a foundation for further investigation into the potential application of miRNAs in the diagnosis and treatment of sepsis.

Keywords

miRNAs sepsis Mendelian randomization eQTL GWAS

Introduction

Sepsis is a life-threatening organ dysfunction syndrome triggered by infection, characterized by a dysregulated host response leading to systemic inflammation. In China alone, there are over 9 million cases of severe sepsis annually, with an annual growth rate ranging from 1.5% to 8.0%, resulting in more than 800 000 deaths.¹ Despite significant advances in critical care and antibiotic therapy, sepsis remains one of the major public health concerns globally, associated with high morbidity and mortality rates. According to the World Health Organization, sepsis affects approximately 55 million people worldwide each year, causing 11 million deaths, imposing a substantial burden on global healthcare systems, particularly impacting children and low-income countries.^2,3 The challenge of sepsis prevention and control is further exacerbated by the aging population and an increasing number of immunocompromised patients.⁴ Therefore, investigating the pathogenesis of sepsis, exploring early diagnostic biomarkers, and developing novel therapeutic approaches are critically important.

MicroRNAs (miRNAs), endogenous non-coding RNA molecules approximately 21 nucleotides in length, have been recognized as key regulators of gene expression since their discovery in Caenorhabditis elegans in 1993. By binding to the 3′-untranslated regions of mRNAs, miRNAs can either degrade or inhibit protein translation. Recently, the pivotal roles of miRNAs in the pathophysiology of sepsis have garnered considerable attention. Studies have identified abnormal expressions of various miRNAs in the peripheral blood of sepsis patients, indicating their potential as biomarkers and their involvement in sepsis-related signaling pathways and cellular processes.⁵ For instance, a composite marker comprising miR-150 and miR-4772-5p-iso can distinguish sepsis from systemic inflammatory response syndrome (SIRS) with high specificity and sensitivity.⁶ Additionally, changes in specific miRNAs such as miR-146a⁷ and miR-223⁸ levels correlate with different stages of sepsis, suggesting their significance in predicting disease progression and patient outcomes. Understanding the roles of miRNAs in sepsis provides novel and viable research directions for developing new diagnostic, prognostic, and therapeutic strategies.

Recent studies have demonstrated causal relationships between plasma circulating miRNAs and various diseases, including COVID-19,⁹ amyotrophic lateral sclerosis,¹⁰ and Parkinson's Disease.¹¹ However, no studies have explored the relationship between plasma circulating miRNAs and sepsis. This study aims to investigate the causal relationship between miRNAs and sepsis using Mendelian Randomization (MR) methods. We will utilize existing miRNA expression quantitative trait loci (eQTL) data and genome-wide association study (GWAS) data on sepsis to identify miRNAs significantly associated with sepsis risk and analyze their potential targets and related pathways. Through this research, we aim to provide new insights and evidence for the early prevention and treatment of sepsis.

Materials and Methods

Acquisition of miRNA eQTL Data

In this study, miRNA eQTLs data were obtained from two distinct cohorts: 1) The dataset from Huan et al,¹² which included eQTL data for 280 high-quality miRNAs across 5329 individuals, was used as the training set. To minimize potential horizontal pleiotropy, we focused on cis-miRNA-eQTLs and excluded SNPs located within coding regions with synonymous or non-synonymous mutation consequences. Instrumental variables (IVs) were selected using a false discovery rate (FDR) adjusted by the Benjamini-Hochberg method, set at 0.1 (corresponding to an original P-value of 6.6 × 10⁻⁵). 2) The second cohort provided by Nikpay et al¹³ included eQTL data for 2083 miRNAs in 710 individuals, serving as the validation set, with a P-value threshold of 1 × 10⁻⁵ for selecting relevant IVs. To harmonize miRNA identifiers across both datasets, we utilized the miRCarta v1.1 database. This approach not only effectively identified disease-related miRNAs but also ensured the accuracy and reliability of the results.

Sepsis GWAS Data Retrieval

Summary statistics from the sepsis GWAS cohort were obtained from the MRC-IEU OpenGWAS database (dataset ID: ieu-b-5088), based on resources from the UK Biobank. This dataset comprised 11 568 cases and 451 301 controls, totaling 462 869 European individuals. Additionally, the GWAS covered 12 243 540 SNPs.

Mendelian Randomization Analysis

TwoSampleMR v0.6.8 R package was employed for conducting two-sample Mendelian Randomization (MR) analysis¹⁴ to investigate the association between miRNA expression levels and sepsis risk, following a previously described methodology.^9–11 Initially, linkage disequilibrium analysis of IVs was performed within the training set using a clustering window of r² < 0.5 and 10 kb to select independent IVs. Corresponding SNPs were then extracted from the GWAS dataset, and normalization was applied to exposure and outcome data to ensure consistent effect allele usage. Palindromic SNPs were excluded. Two primary analytical strategies were adopted: Inverse Variance Weighted (IVW) and MR-Egger regression analyses. Criteria for identifying causally associated miRNAs included: (1) P-values < .05 in IVW tests and FDR < 0.1; (2) consistent direction of estimated effects in both IVW and MR-Egger models; (3) MR analysis based on at least three SNPs; and (4) significant results in validation studies to confirm high-confidence relationships. Horizontal pleiotropy was assessed using MR-Egger regression, and Leave-One-Out analysis was conducted to examine the influence of individual SNPs on overall causal inference. Heterogeneity among the causal estimates for each miRNA was assessed using Cochran's Q statistic. A P-value > .05 for Cochran's Q was considered indicative of a lack of significant heterogeneity, suggesting that the instrumental variables for a given miRNA provide consistent causal estimates. This methodology enhances the reliability of the findings and provides new perspectives on the role of miRNAs in disease.

Target Identification and Enrichment Analysis

To explore the functional impact of identified potential causal miRNAs, we utilized miRNet 2.0 and its online platform to identify their target genes. This tool is accessible via https://www.mirnet.ca/miRNet/home.xhtml (accessed on January 12, 2025). Only experimentally validated target genes sourced from the miRTarBase v9.0 database were considered. These miRNAs and their corresponding target genes were imported into Cytoscape software v3.10.2 to construct detailed miRNA-gene interaction network maps. For further analysis of biological processes involving these target genes, ClusterProfiler package was used for Gene Ontology (GO) annotation enrichment analysis, focusing particularly on Biological Process (BP) enrichment. An FDR threshold of less than 0.05 was set to adjust for multiple comparisons.

Results

Circulating miRNA eQTLs

From the cohort provided by Huan et al, we initially selected 9612 single nucleotide polymorphisms (SNPs) as instrumental variables. After filtering, 9517 SNPs not directly associated with coding gene regions were identified for further analysis. Additionally, 17 SNPs linked to mir-213 were excluded due to ineffective mapping to known miRNA identifiers. Ultimately, this study focused on 9500 SNPs significantly associated with 75 circulating mature human miRNAs for in-depth investigation (Table 1, Supplemental Materials: Table S1). Figure 1 illustrates the workflow diagram of this MR analysis.

Figure 1.

The Workflow Diagram of this MR Analysis.

Table 1.

The List of Top SNPs Per miRNA in the Huan Et al Cohort.

snpID	miRNA_FHS	sample.Size	MAF	Estimate	Std.Error	Pval
rs2370747	miR_100_5p	3502	0.43	−3.48	0.13	1.28 × 10⁻¹³⁷
rs12980971	miR_125a_5p	5205	0.39	0.38	0.04	7.84 × 10⁻¹⁸
rs2370747	miR_125b_5p	5225	0.37	−1.7	0.07	1.45 × 10⁻¹¹²
rs28576121	miR_1270	4812	0.13	−1.2	0.08	2.00 × 10⁻⁵⁵
rs4905998	miR_127_3p	5062	0.33	0.82	0.05	2.04 × 10⁻⁶⁴
rs965235	miR_1303	3812	0.33	−0.88	0.04	3.90 × 10⁻¹¹⁶
rs685149	miR_130a_3p	5070	0.36	−0.33	0.06	6.93 × 10⁻⁸
rs373001	miR_130b_3p	4884	0.42	0.29	0.05	3.28 × 10⁻⁹
rs861814	miR_130b_5p	5108	0.47	0.19	0.03	5.37 × 10⁻⁹
rs3803809	miR_132_3p	5035	0.15	−0.39	0.06	7.48 × 10⁻¹¹
rs4429540	miR_133a	5156	0.37	−3.05	0.14	5.55 × 10⁻¹⁰⁶
rs12588718	miR_134	3316	0.32	0.38	0.06	4.93 × 10⁻¹⁰
rs6538758	miR_135a_5p	1644	0.33	0.26	0.06	1.01 × 10⁻⁵
rs808721	miR_138_5p	4175	0.42	−0.19	0.04	2.38 × 10⁻⁵
rs11235550	miR_139_3p	3196	0.08	−1.01	0.19	1.70 × 10⁻⁷
rs904469	miR_139_5p	5242	0.17	−0.52	0.05	6.18 × 10⁻²⁴
rs7789194	miR_148a_3p	2026	0.33	0.44	0.09	2.79 × 10⁻⁶
rs11167021	miR_151a_3p	5131	0.47	−0.47	0.06	2.51 × 10⁻¹⁴
rs6578139	miR_151a_5p	4727	0.44	−0.25	0.03	1.79 × 10⁻¹⁶
rs2737	miR_152	5049	0.22	0.33	0.05	6.88 × 10⁻¹²
rs1434282	miR_181a_5p	4906	0.29	−0.37	0.05	1.39 × 10⁻¹⁵
rs6467265	miR_182_5p	4660	0.42	−0.43	0.07	5.25 × 10⁻⁹
rs6467265	miR_183_3p	5161	0.42	−0.47	0.05	1.20 × 10⁻¹⁷
rs12451384	miR_193a_5p	5115	0.23	−0.4	0.05	1.23 × 10⁻¹⁴
rs30227	miR_193b_3p	5158	0.39	0.2	0.05	1.36 × 10⁻⁵
rs2071243	miR_196b_5p	4148	0.4	−0.4	0.05	1.05 × 10⁻¹⁴
rs28640110	miR_204_5p	711	0.07	−1.69	0.17	1.69 × 10⁻²²
rs11246184	miR_210	5100	0.44	0.24	0.06	2.71 × 10⁻⁵
rs13165104	miR_218_2_3p	325	0.35	−0.97	0.18	1.52 × 10⁻⁷
rs13165104	miR_218_5p	5155	0.04	−3.32	0.13	7.10 × 10⁻¹²⁷
rs4603591	miR_22_3p	4823	0.28	0.65	0.09	6.30 × 10⁻¹³
rs12939051	miR_22_5p	5055	0.51	−0.15	0.03	7.08 × 10⁻⁸
rs10990614	miR_24_3p	5275	0.12	0.26	0.06	1.79 × 10⁻⁵
rs7607369	miR_26b_5p	4794	0.43	0.42	0.07	3.60 × 10⁻⁹
rs6479588	miR_27b_3p	5143	0.11	−0.43	0.06	3.01 × 10⁻¹³
rs6793253	miR_28_3p	4884	0.2	−0.38	0.09	4.50 × 10⁻⁵
rs259958	miR_296_5p	4985	0.48	0.33	0.05	5.95 × 10⁻¹³
rs9342836	miR_30a_3p	1015	0.16	−1.54	0.13	2.28 × 10⁻³²
rs7033598	miR_31_5p	4946	0.35	−0.42	0.04	7.84 × 10⁻²¹
rs17099623	miR_323a_3p	1696	0.16	0.29	0.07	2.42 × 10⁻⁵
rs11160594	miR_329	3193	0.23	0.34	0.06	9.34 × 10⁻⁸
rs7155649	miR_337_3p	285	0.08	−2.36	0.53	1.32 × 10⁻⁵
rs11763793	miR_339_3p	5193	0.16	−0.6	0.05	5.76 × 10⁻³⁵
rs10266519	miR_339_5p	5216	0.17	−0.58	0.05	2.58 × 10⁻²⁸
rs11627407	miR_345_5p	5148	0.48	0.17	0.04	2.20 × 10⁻⁵
rs4905998	miR_370	4573	0.31	0.83	0.06	9.54 × 10⁻⁴⁸
rs12147008	miR_376a_3p	4927	0.27	0.65	0.07	4.37 × 10⁻²⁰
rs4905998	miR_382_5p	4342	0.31	0.67	0.06	2.57 × 10⁻²⁸
rs4905998	miR_409_3p	5252	0.33	0.57	0.04	9.89 × 10⁻⁴¹
rs4905998	miR_411_3p	3908	0.3	0.48	0.05	7.88 × 10⁻²⁶
rs4905998	miR_431_5p	1939	0.26	0.39	0.08	8.30 × 10⁻⁷
rs12881545	miR_433	2727	0.24	0.44	0.06	6.67 × 10⁻¹⁵
rs34817776	miR_454_5p	5102	0.41	0.16	0.04	3.16 × 10⁻⁵
rs11042699	miR_483_3p	1250	0.17	0.55	0.11	3.91 × 10⁻⁷
rs12881545	miR_494	3031	0.26	0.54	0.06	1.29 × 10⁻¹⁷
rs4905998	miR_543	4662	0.31	0.68	0.05	3.57 × 10⁻⁴²
rs1839612	miR_550a_3p	1769	0.49	−0.27	0.04	6.39 × 10⁻¹⁰
rs3087896	miR_570_3p	283	0.1	−2.61	0.49	1.70 × 10⁻⁷
rs7808842	miR_589_5p	3602	0.03	−0.86	0.21	3.67 × 10⁻⁵
rs17747335	miR_598	4699	0.53	−0.2	0.03	2.35 × 10⁻⁸
rs2127869	miR_625_3p	5165	0.34	−0.34	0.06	2.14 × 10⁻⁸
rs2127868	miR_625_5p	5019	0.26	−0.25	0.04	1.01 × 10⁻⁹
rs28483325	miR_628_3p	5133	0.1	−0.32	0.05	2.08 × 10⁻¹⁰
rs2291985	miR_629_5p	4777	0.02	1.06	0.11	3.27 × 10⁻²⁰
rs12453629	miR_635	1157	0.29	−0.51	0.12	1.50 × 10⁻⁵
rs12881545	miR_654_5p	4232	0.27	0.69	0.06	9.14 × 10⁻³²
rs4905998	miR_668	4686	0.31	0.84	0.06	7.72 × 10⁻⁴⁴
rs6967282	miR_671_3p	5156	0.2	0.14	0.03	3.94 × 10⁻⁵
rs11658975	miR_744_3p	5126	0.25	−0.14	0.03	2.92 × 10⁻⁵
rs2969226	miR_744_5p	5107	0.34	−0.15	0.04	1.29 × 10⁻⁵
rs12710081	miR_7_5p	5016	0.15	0.61	0.13	5.07 × 10⁻⁶
rs6792305	miR_885_5p	4223	0.03	−1.22	0.16	1.49 × 10⁻¹⁴
rs6062577	miR_941	3628	0.14	0.79	0.09	1.07 × 10⁻¹⁸
rs12022333	miR_942	5257	0.05	−0.31	0.07	9.47 × 10⁻⁶
rs11670644	miR_99b_5p	5169	0.26	0.36	0.05	7.63 × 10⁻¹⁵

Sepsis-Associated Circulating miRNAs

Using IVW and MR-Egger regression, we identified all miRNAs significantly associated with sepsis risk (Figure 2, Supplemental Materials: Figure S1), without observing significant horizontal pleiotropy (Table 2). After Bonferroni correction for multiple testing, three blood miRNA markers were significantly associated with reduced IDD risk: hsa-miR-125a-5p (IVW, OR: 0.971, 95% CI: 0.952-0.991, P-value: .004), hsa-miR-494-3p (IVW, OR: 0.976, 95% CI: 0.963-0.989, P-value: .001), and hsa-miR-885-5p (IVW, OR: 0.954, 95% CI: 0.924-0.984, P-value: .003). Conversely, three blood miRNA markers were significantly associated with increased sepsis risk: hsa-miR-196b-5p (IVW, OR: 1.069, 95% CI: 1.028-1.111, P-value: .001), hsa-miR-27b-3p (IVW, OR: 1.097, 95% CI: 1.035-1.162, P-value: .002), and hsa-miR-598-3p (IVW, OR: 1.066, 95% CI: 1.031-1.103, P-value < .001). Leave-One-Out results for miRNAs are shown in Figures S2-S7, indicating good stability and reliability of the findings. Cochran's Q test indicated no significant heterogeneity for the instrumental variables of these miRNAs (all P > .05, Table S2), supporting the consistency of the effect estimates.

Figure 2.

Associations of Serum-Circulating miRNAs with Protective and Harmful Effects on Sepsis in the Discovery Dataset. nSNP: Number of Single Nucleotide Polymorphism; CI: Confidence Interval; MR: Mendelian Randomization.

Table 2.

Pleiotropy Tests of Instrument Effects in the Discovery Dataset.

miRNAs	Egger Intercept	SE	P-Value
hsa-miR-127-3p	−0.0038	0.0054	.4877
hsa-miR-370-3p	−0.0053	0.0061	.3933
hsa-miR-668-3p	0.0006	0.0053	.9117
hsa-miR-409-3p	−0.0025	0.0059	.6680
hsa-miR-654-5p	−0.0035	0.0059	.5542
hsa-miR-382-5p	−0.0043	0.0061	.4816
hsa-miR-494-3p	0.0017	0.0088	.8478
hsa-miR-134-5p	−0.0011	0.0298	.9709
hsa-miR-598-3p	0.0083	0.0189	.6636

Mechanisms of miRNA Regulation in Sepsis Risk

To further explore the potential mechanisms by which these causal miRNAs might influence sepsis, we predicted 1047 experimentally validated targets using miRNet. Among these, hsa-miR-27b-3p had the most targets (439), followed by hsa-miR-125a-5p (280 targets), hsa-miR-196b-5p (162 targets), hsa-miR-494-3p (142 targets), hsa-miR-885-5p (73 targets), and hsa-miR-598-3p (26 targets) (Figure 3). KEGG pathway enrichment analysis revealed that these targets are involved in pathways related to bacterial infections, proteoglycans in cancer, cellular senescence, FoxO signaling, relaxin signaling, and various cancer pathways (Figure 4). GO annotation enrichment indicated significant enrichment in biological processes such as myeloid cell differentiation, response to oxygen levels, cell growth, and apoptosis (Figure 5). These analyses provide plausible biological hypotheses for future functional studies to validate.

Figure 3.

The miRNA-Gene Network of the Causal miRNAs Associated with Sepsis.

Figure 4.

Significantly Enriched KEGG Pathways for the Targets of Causal miRNAs.

Figure 5.

Significantly Enriched Biological Processes for the Targets of Causal miRNAs.

Validation of Circulating miRNAs in an Independent Cohort

Next, we used the cohort from Nikpay et al to validate potential miRNAs. We collected genome-wide miRNA eQTL data for all six candidate causal miRNAs. A P-value threshold of 1 × 10⁻⁶ was set to select relevant IVs. The same quality control procedures applied in the original study were also performed here. After data preprocessing, all six miRNAs were validated in this new study, without observing significant horizontal pleiotropy (Figure 6, Table S3). IVW analysis showed that hsa-miR-27b-3p exhibited a detrimental effect on sepsis, with an OR of 1.22 (95% CI: 1.061-1.403), consistent with observations from the discovery dataset. Other miRNAs did not show significant associations with sepsis risk in this cohort.

Figure 6.

Associations of serum-Circulating miRNAs with Protective and Harmful Effects on Sepsis in the Validation Dataset. nSNP: Number of Single Nucleotide Polymorphism; CI: Confidence Interval; MR: Mendelian Randomization.

Discussion

This study employed MR to systematically investigate, for the first time, the causal relationship between circulating miRNAs and sepsis, providing new insights for early diagnosis and therapeutic strategies. Sepsis is a severe systemic inflammatory response syndrome with a complex pathogenesis involving alterations in various biomarkers. We identified a panel of circulating miRNAs that may exert causal effects on sepsis risk. Although their precise mechanisms require further experimental validation, these findings offer potential directions for improving sepsis management.

The most immediate application lies in the development of biomarkers. A multi-miRNA signature, constructed based on the protective miRNAs (eg, hsa-miR-125a-5p, hsa-miR-494-3p) and risk miRNAs (eg, hsa-miR-27b-3p) identified here, could be utilized for early diagnosis and risk stratification in patients with suspected infections.¹⁵ This would assist clinicians in identifying high-risk individuals and initiating timely, targeted interventions. Furthermore, protective miRNAs represent promising candidates for novel therapeutic strategies, such as miRNA mimics.¹⁶ Conversely, risk miRNAs may indicate the need for closer monitoring or interventions targeting their downstream pathways (eg, apoptosis regulation).

We found that hsa-miR-125a-5p, hsa-miR-494-3p, and hsa-miR-885-5p significantly reduce the risk of sepsis. Studies have shown that in sepsis patients, serum extracellular vesicles contain elevated levels of hsa-miR-125a-5p, which are transported to lung macrophages to inhibit Tnfaip3, thereby promoting reactive oxygen species (ROS) production and pro-inflammatory cytokine release, leading to acute lung injury.¹⁷ Additionally, sepsis is a major cause of acute kidney injury, and hsa-miR-125a-5p may mitigate renal inflammation through the TRAF6/NF-κB axis.¹⁸ Compared to healthy controls, serum levels of hsa-miR-494-3p are significantly lower in sepsis patients; this miRNA can regulate inflammation by targeting PTEN mRNA's 3′-untranslated region¹⁹ and modulate TLR6 expression to alleviate sepsis-related inflammation.²⁰ Moreover, miR-494-3p targets Lp-PLA2 and inhibits its secretion, thereby reducing sepsis-associated inflammation.²¹ MiR-885-5p has been shown to regulate cell growth, senescence, and/or apoptosis. Research indicates that exosomes from sepsis patients promote cardiomyocyte apoptosis via HMBOX1-mediated miR-885-5p, suggesting new therapeutic directions for sepsis-induced cardiac suppression.²²

Furthermore, we found that hsa-miR-196b-5p, hsa-miR-27b-3p, and hsa-miR-598-3p are associated with increased sepsis risk. Chrysophanol has been shown to inhibit inflammation and apoptosis through the miR-27b-3p-PPARG axis, offering protective effects against sepsis-induced acute myocardial injury.²³ In sepsis patients, miR-27b-3p is upregulated in both cellular and extracellular vesicle compartments and exhibits different expression patterns at various disease stages, suggesting its potential role in diagnosing and assessing sepsis severity.²⁴ However, there is no direct experimental evidence linking hsa-miR-196b-5p and hsa-miR-598-3p to sepsis, although they are widely implicated in regulating inflammation and apoptosis.^25,26 Future research should further explore the roles and mechanisms of these miRNAs in sepsis.

Our MR analysis identifies a causal link, but the underlying mechanisms are derived from computational predictions and thus are speculative. For instance, while bacterial infection is a primary cause of sepsis and the targets of causal miRNAs were significantly enriched in related pathways, this association is suggestive rather than proven, and experimental confirmation is necessary.²⁷ Similarly, the proposed associations with myeloid cell differentiation, apoptosis, and the FoxO signaling pathway represent interesting hypotheses generated from our analysis that merit further investigation in mechanistic studies.²⁸ Multiple apoptosis regulatory mechanisms are affected in sepsis patients. For instance, the expression of pro-apoptotic proteins like Bax increases, while anti-apoptotic proteins such as Bcl-2 decrease, promoting apoptosis.²⁹ Targeting apoptosis regulation pathways could be a potential therapeutic strategy. FoxO transcription factors can regulate inflammation,³⁰ apoptosis and survival,³¹ and immune cell function, including T lymphocytes and macrophages.³² Furthermore, sepsis often involves severe metabolic disturbances, and FoxO transcription factors also participate in cellular metabolism regulation, particularly adjusting energy metabolism under nutrient deficiency or stress conditions to maintain cell survival.³³

Moreover, the independent validation cohort results showed that hsa-miR-27b-3p exhibits detrimental effects on sepsis, consistent with findings from the discovery dataset, enhancing the reliability of our conclusions. Notably, however, other candidate miRNAs failed to exhibit statistically significant associations. This discrepancy may stem from multiple contributing factors: First, the validation cohort's relatively modest sample size (n = 710) compared to the discovery set (n = 5329) may have compromised the statistical power to detect miRNA associations with moderate effect sizes.³⁴ Second, inherent biological variability in miRNA expression profiles across diverse populations, compounded by technical discrepancies in eQTL profiling platforms, could undermine reproducibility.³⁵ Third, certain miRNA effects might be context-dependent, influenced by specific genetic architectures or environmental exposures that differ between cohorts.³⁶

Several limitations of this study should be acknowledged. First, the GWAS and eQTL data utilized were primarily derived from European populations, which may limit the generalizability of our findings to other ethnic groups and necessitate further validation in diverse cohorts. Second, although the MR analysis suggests causal relationships and pathway analyses reveal potential biological implications, these findings remain hypothesis-generating and are not yet directly translatable to clinical practice. Future work should include functional validation in experimental models of sepsis, as well as well-designed prospective cohort studies to evaluate the diagnostic and prognostic utility of these miRNAs before clinical application can be considered.

It is also noteworthy that the effect sizes (odds ratios) of the identified miRNAs, while statistically significant, were modest—a common feature in MR studies of complex traits. This indicates that individual miRNAs likely contribute only limited effects to sepsis risk and are unlikely to serve as stand-alone diagnostic biomarkers. The principal value of these findings lies in highlighting potential causal players in sepsis pathophysiology, supported by the significant enrichment of their target genes in key biological pathways. Future research could explore the construction of a multi-miRNA signature or a polygenic risk score integrating these miRNAs. Combining such a profile with established clinical scores (eg, SOFA) and biomarkers (eg, PCT) may enhance the accuracy of risk stratification, early diagnosis, and prognosis assessment in sepsis.

Conclusion

In summary, this MR study provides genetic evidence supporting a causal role for specific circulating miRNAs in sepsis pathogenesis. Our findings highlight several promising candidates for therapeutic development. Specifically, miRNA mimics for the protective miRNAs (eg, hsa-miR-125a-5p, hsa-miR-494-3p) or antagomirs for the risk miRNAs (eg, hsa-miR-27b-3p) represent novel biological strategies worthy of investigation. Furthermore, the enriched pathways (eg, apoptosis, FoxO signaling) unveil potential downstream druggable targets. While these possibilities are compelling, they remain hypothetical and require rigorous functional validation in experimental models of sepsis to confirm the mechanisms and assess the efficacy and safety of such interventions before any clinical application can be considered.

Supplemental Material

sj-xlsx-1-cat-10.1177_10760296251386028 - Supplemental material for Exploring the Causal Relationship Between Circulating miRNAs and Sepsis Through Mendelian Randomization Analysis

Supplemental material, sj-xlsx-1-cat-10.1177_10760296251386028 for Exploring the Causal Relationship Between Circulating miRNAs and Sepsis Through Mendelian Randomization Analysis by Yan Jin and Yuan Zhang in Clinical and Applied Thrombosis/Hemostasis

Supplemental Material

sj-xlsx-2-cat-10.1177_10760296251386028 - Supplemental material for Exploring the Causal Relationship Between Circulating miRNAs and Sepsis Through Mendelian Randomization Analysis

Supplemental material, sj-xlsx-2-cat-10.1177_10760296251386028 for Exploring the Causal Relationship Between Circulating miRNAs and Sepsis Through Mendelian Randomization Analysis by Yan Jin and Yuan Zhang in Clinical and Applied Thrombosis/Hemostasis

Supplemental Material

sj-xlsx-3-cat-10.1177_10760296251386028 - Supplemental material for Exploring the Causal Relationship Between Circulating miRNAs and Sepsis Through Mendelian Randomization Analysis

Supplemental material, sj-xlsx-3-cat-10.1177_10760296251386028 for Exploring the Causal Relationship Between Circulating miRNAs and Sepsis Through Mendelian Randomization Analysis by Yan Jin and Yuan Zhang in Clinical and Applied Thrombosis/Hemostasis

Supplemental Material

sj-doc-4-cat-10.1177_10760296251386028 - Supplemental material for Exploring the Causal Relationship Between Circulating miRNAs and Sepsis Through Mendelian Randomization Analysis

Supplemental material, sj-doc-4-cat-10.1177_10760296251386028 for Exploring the Causal Relationship Between Circulating miRNAs and Sepsis Through Mendelian Randomization Analysis by Yan Jin and Yuan Zhang in Clinical and Applied Thrombosis/Hemostasis

Footnotes

Abbreviation

Ethics Approval

There is no need for informed consent in our study since the unidentified data were free from medical ethics review.

Consent for Publication

Institutional review board approval and informed consent were not required in the current study because research data are publicly available and all patient data are de-identified.

ORCID iD

Yuan Zhang

Authors’ Contributions

Yan Jin and Yuan Zhang designed, extracted, analyzed, and interpreted the data from databases. Yan Jin and Yuan Zhang wrote the manuscript and substantively revised it. All authors have read and approved the final manuscript.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Availability of Data and Materials

The original contributions presented in this study are included in the article/Supplemental Material. Further inquiries can be directed to the corresponding author(s).

Supplemental Material

Supplemental material for this article is available online.

References

Weng

Yin

, et al. National incidence and mortality of hospitalized sepsis in China. Crit Care. 2023;27(1):84.

Tong

Venkatesh

McCreary

. Acute kidney injury with empirical antibiotics for sepsis. JAMA. 2023;330(16):1531–1533.

Rudd

Johnson

Agesa

, et al. Global, regional, and national sepsis incidence and mortality, 1990-2017: Analysis for the Global Burden of Disease Study. Lancet. 2020;395(10219):200-211.

Guo

Huang

Dou

, et al. Aging and aging-related diseases: From molecular mechanisms to interventions and treatments. Signal Transduct Target Ther. 2022;7(1):391.

Formosa

Turgeon

Dos

. Role of miRNA dysregulation in sepsis. Mol Med. 2022;28(1):99.

Vilanova

Atalar

, et al. Genome-wide sequencing of cellular microRNAs identifies a combinatorial expression signature diagnostic of sepsis. PLoS One. 2013;8(10):e75918.

Wang

Zhu

Gao

Guo

. The effect of hybrid blood purification combined with ulinastatin for the treatment of severe sepsis on APACHE II score and levels of miR-146a and miR-155. Int J Gen Med. 2024;17:5897-5905.

Zhang

Wang

Dong

. Predictive value of miR-7110-5p and miR-223-3p as biomarkers for sepsis secondary to pneumonia. Technol Health Care. 2024;32(5):2931-2939.

Song

, et al. Identifying putative causal links between MicroRNAs and severe COVID-19 using Mendelian randomization. Cells. 2021;10(12):3504.

10.

Zhu

, et al. Evaluating the causal association between microRNAs and amyotrophic lateral sclerosis. Neurol Sci. 2023;44(10):3567-3575.

11.

Shi

, et al. Systematic genome-wide Mendelian randomization reveals the causal links between miRNAs and Parkinson's disease. Front Neurosci. 2024;18:1385675.

12.

Huan

Rong

Liu

, et al. Genome-wide identification of microRNA expression quantitative trait loci. Nat Commun. 2015;6:6601.

13.

Nikpay

Beehler

Valsesia

, et al. Genome-wide identification of circulating-miRNA expression quantitative trait loci reveals the role of several miRNAs in the regulation of cardiometabolic phenotypes. Cardiovasc Res. 2019;115(11):1629-1645.

14.

Hemani

Zheng

Elsworth

, et al. The MR-Base platform supports systematic causal inference across the human phenome. Elife. 2018;7:e34408.

15.

Bindayna

. MicroRNA as Sepsis biomarkers: A comprehensive review. Int J Mol Sci. 2024;25(12):6476.

16.

Zhang

Chen

, et al. Biomimetic immunosuppressive exosomes that inhibit cytokine storms contribute to the alleviation of sepsis. Adv Mater. 2022;34(19):e2108476.

17.

Wang

Huang

, et al. Macrophage biomimetic nanoparticle-targeted functional extracellular vesicle micro-RNAs revealed via multiomics analysis alleviate sepsis-induced acute lung injury. J Nanobiotechnology. 2024;22(1):362.

18.

Yang

Huang

, et al. Reduced expression of MiR-125a-5p aggravates LPS-induced experimental acute kidney injury pathology by targeting TRAF6. Life Sci. 2022;288:119657.

19.

Kong

. MicroRNA-494-3p protects rat cardiomyocytes against septic shock via PTEN. Exp Ther Med. 2019;17(3):1706–1716.

20.

Wang

Dou

. MicroRNA-494-3p alleviates inflammatory response in sepsis by targeting TLR6. Eur Rev Med Pharmacol Sci. 2019;23(7):2971-2977.

21.

Yan

Luo

, et al. Lipopolysaccharide (LPS)-induced inflammation in RAW264.7 cells is inhibited by microRNA-494-3p via targeting lipoprotein-associated phospholipase A2. Eur J Trauma Emerg Surg. 2024;50(6):3289-3298.

22.

, et al. Exosome-Derived from sepsis Patients’ blood promoted pyroptosis of cardiomyocytes by regulating miR-885-5p/HMBOX1. Front Cardiovasc Med. 2022;9:774193.

23.

Zhao

Wang

Zhu

. Chrysophanol exerts a protective effect against sepsis-induced acute myocardial injury through modulating the microRNA-27b-3p/peroxisomal proliferating-activated receptor gamma axis. Bioengineered. 2022;13(5):12673–12690.

24.

Reithmair

Buschmann

Märte

, et al. Cellular and extracellular miRNAs are blood-compartment-specific diagnostic targets in sepsis. J Cell Mol Med. 2017;21(10):2403-2411.

25.

Xue

Guo

Wang

, et al. miR-196b-5p affects macrophage polarization and inflammation in endometriosis. Iran J Immunol. 2024;21(3):201-211.

26.

Lin

Huang

Wang

Liu

. The long noncoding RNA MALAT1/microRNA-598-3p axis regulates the proliferation and apoptosis of retinoblastoma cells through the PI3K/AKT pathway. Mol Vis. 2022;28:269-279.

27.

Lonsdale

Shah

Lipman

. Infection, sepsis and the inflammatory response: Mechanisms and therapy. Front Med (Lausanne). 2020;7:588863.

28.

Liang

. Macrophages in sepsis-induced acute lung injury: Exosomal modulation and therapeutic potential. Front Immunol. 2024;15:1518008.

29.

Deng

Huang

Wang

, et al. Regulated programmed cell death in sepsis associated acute lung injury: From pathogenesis to therapy. Int Immunopharmacol. 2025;148:114111.

30.

Chen

Wang

Cai

. Euphorbiasteroid induces neurotoxicity through the FOXO/NF-κB/apoptosis signaling pathway. Brain Res Bull. 2024;219:111129.

31.

Oyama

Brashears

Rathore

, et al. PHGDH Inhibition and FOXO3 modulation drives PUMA-dependent apoptosis in osteosarcoma. Cell Death Dis. 2025;16(1):89.

32.

Tao

Jiang

Wang

, et al. Pseudokinase STK40 promotes T(H)1 and T(H)17 cell differentiation by targeting FOXO transcription factors. Sci Adv. 2024;10(47):eadp2919.

33.

Kaur

Khan

Grewal

, et al. Exploring therapeutic strategies: The relationship between metabolic disorders and FOXO signalling in Alzheimer's disease. CNS Neurol Disord Drug Targets. 2025;24(3):196-207.

34.

Mustafa

Mens

van Hilten

, et al. A comprehensive study of genetic regulation and disease associations of plasma circulatory microRNAs using population-level data. Genome Biol. 2024;25(1):276.

35.

Liu

Luo

Shi

, et al. Dual nanozyme bubble-propelled biosensor for multiple miRNAs detection overcomes spatial hurdles of DNA probes. Biosens Bioelectron. 2025;286:117624.

36.

Vattathil

Gerasimov

Canon

, et al. Mapping the microRNA landscape in the older adult brain and its genetic contribution to neuropsychiatric conditions. Nat Aging. 2025;5(2):306-319.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

2.37 MB

1.78 MB

0.02 MB

0.01 MB