Haemophilus-Informed Genetic and Transcriptomic Analyses Identify IL-7 and Vegfa as Drivers of Immunosuppressive Sepsis Endotypes
Jinwei Dai, Nianzhe Sun, Lina Zhang, Zhaoxin Qian, Zhihong ZuoBackground:
Sepsis exhibits marked biological heterogeneity, yet the causal pathways linking host immune mediators to clinically relevant endotypes remain poorly defined, particularly in pathogen-specific contexts.
Methods:
We applied Mendelian randomization (MR) to investigate the causal effects of circulating immune mediators on the risk of Haemophilus septicemia, identifying interleukin-7 (IL-7) and vascular endothelial growth factor A (VEGFA) as putative causal factors. To elucidate the biological context underlying these genetic associations, we integrated pathogen-informed gene sets with bulk blood transcriptomic data from a large sepsis cohort (GSE185263). Single-sample gene set enrichment analysis and Pathway responsive genes (PROGENy) were used to infer pathway activities, which were subsequently aligned with previously reported sepsis endotypes.
Results:
A Haemophilus-informed transcriptional score captured a distinct signaling hierarchy dominated by TGFβ, TNFα/NFκB, hypoxia, and cell death pathways, accompanied by suppression of regenerative signaling. These pathway activities aligned most closely with the neutrophil-dominated, immunosuppressive (NeutroSupp) endotype described in the original cohort. While VEGFA expression showed strong coupling with VEGF and hypoxia signaling, IL-7-related immune maintenance pathways, inferred through Janus Kinase–signal transducer and activator of transcription activity, were only modestly engaged, consistent with impaired adaptive immune recovery.
Conclusions:
By integrating genetic causality with transcriptomic and endotype-level analyses, our findings suggest that IL-7 and VEGFA contribute to sepsis risk through distinct immune–endothelial axes that converge on an immunosuppressive sepsis endotype. This work provides a pathogen-informed framework for interpreting genetic associations within clinically relevant sepsis phenotypes.