Blood Gene Signature Tied to STAT3 and ARG1 Reveals Immune Shutdown in Sepsis

Sepsis remains one of the most feared conditions in modern medicine, a runaway response to infection that can send the immune system into chaos and claim millions of lives worldwide each year. Yet for all its deadliness, sepsis has long frustrated clinicians with a paradox: many patients do not die from an immune system raging out of control, but from one that has quietly collapsed. Now, a team of researchers in China has distilled that collapse into a compact molecular fingerprint detectable in a simple blood sample, offering a reproducible computational framework that could sharpen diagnosis and, eventually, guide immunotherapy for the sickest patients.

The study, published in the journal 3 Biotech, was led by Dongmei Liu of Lianyungang TCM Hospital at Nanjing University of Chinese Medicine, together with colleagues including corresponding author Xiaoyong Xu of the Fourth Affiliated Hospital of Nanjing Medical University. Rather than recruiting new patients, the team mined publicly available peripheral-blood transcriptomic datasets, turning archived gene-expression measurements into a fresh window on the immune biology of sepsis. Their analysis converged on a signature centered on two molecules with well-established roles in immune suppression: the transcription factor STAT3 and the enzyme arginase 1, known as ARG1.

The researchers built their discovery around a dataset known as GSE28750, which contained whole-blood expression profiles from ten sepsis patients and twenty healthy controls. Differential expression analysis identified 482 genes whose activity distinguished septic blood from healthy blood, with 301 genes ramped up and 181 dialed down. The direction of these changes told a coherent story. Genes associated with myeloid cells, the family of innate immune cells that includes monocytes and neutrophils, were prominently upregulated, as were genes tied to immunoregulatory function. Meanwhile, two of the most important flags of immune competence, the HLA-DRA and HLA-DRB1 genes that help immune cells display fragments of pathogens to T cells, were significantly downregulated, a molecular echo of the impaired antigen presentation that characterizes septic immunoparalysis.

Enrichment analysis sharpened the picture. The upregulated genes clustered in pathways governing myeloid cell activation, IL6-JAK-STAT3 signaling, and arginine metabolism, while pathways responsible for antigen processing and presentation were depleted. Each of these threads converges on a cell type that has become a central character in sepsis research: the myeloid-derived suppressor cell, or MDSC. These immature myeloid cells, which expand dramatically during sepsis, dampen T-cell responses through a repertoire of mechanisms, including the depletion of arginine by arginase enzymes, starving lymphocytes of a nutrient they need to proliferate and function.

To quantify this shift, the team applied single-sample gene set enrichment analysis, or ssGSEA, a computational technique that estimates the activity of predefined gene signatures in bulk expression data. Sepsis samples scored markedly higher on an MDSC-like signature than healthy controls, with a mean score of 0.72 plus or minus 0.21 compared with minus 0.36 plus or minus 0.18 in controls, a difference that reached a P value of 1.1 times ten to the minus seven. In plain terms, the blood of septic patients did not merely show scattered signs of immune disturbance; it carried a coordinated, statistically robust imprint of suppressive myeloid remodeling.

The critical question for any biomarker study is whether the finding holds beyond the data in which it was discovered. The researchers therefore turned to an independent validation cohort, GSE65682, comprising 120 sepsis-related samples drawn at baseline or ICU admission alongside 40 healthy controls. There, they checked that the candidate genes moved in the same direction as in the discovery set and tested the performance of a composite signature built from the key genes. The composite score achieved an area under the receiver operating characteristic curve, or AUC, of 0.94 in the discovery cohort and 0.88 in the validation cohort, values that indicate strong discriminatory power for separating septic patients from healthy individuals, though the authors frame this as exploratory diagnostic performance rather than a clinically validated test.

The choice of STAT3 and ARG1 as the signature’s anchors is biologically telling. STAT3 is a signaling hub downstream of interleukin-6 and other inflammatory cytokines, and its persistent activation in myeloid cells is a recognized driver of suppressor-cell differentiation, promoting the expansion of cells that blunt adaptive immunity. ARG1, the gene encoding arginase 1, is one of the classic effector molecules of these cells, catabolizing arginine in the microenvironment and thereby constraining T-cell responses. Recent work cited by the authors has shown that blocking arginase activity in a specific neutrophil subset can combat sepsis-induced lymphopenia, the loss of lymphocytes that leaves survivors vulnerable to secondary infections. The new signature thus captures, in transcriptomic terms, a pathway whose perturbation is already being explored therapeutically.

The study arrives amid a broader effort to bring genomic precision to critical care. A consensus blood transcriptomic framework for sepsis published in Nature Medicine in 2025 and a companion consensus immune dysregulation framework reflect a growing recognition that sepsis is not a single disease but a spectrum of host responses that may demand different treatments. Monocyte HLA-DR expression, the cell-surface readout of the antigen-presentation machinery whose genes the new study found downregulated, has been monitored in septic shock cohorts for two decades and remains a focal point for trials of immunostimulatory agents such as GM-CSF. A compact gene signature measurable from bulk blood, requiring no flow cytometry or specialized cell sorting, could complement these approaches and fit more easily into routine laboratory workflows.

The authors are careful about the limits of their work. The analysis relied on bulk transcriptomics, which measures average gene activity across mixed cell populations and cannot resolve which specific cells express which genes; single-cell studies have, however, independently identified immunosuppressive HLA-DR-low monocyte populations in sepsis, lending plausibility to the bulk-level signal. The validation cohort confirmed directionality and composite performance but did not test whether the signature predicts mortality, tracks disease trajectory over time, or distinguishes sepsis from other causes of critical illness. And because the data come from public repositories, the findings await prospective evaluation in newly recruited, diverse patient populations before any diagnostic use could be contemplated.

Even so, the study offers something the sepsis field badly needs: a transparent, reproducible pipeline. The authors report that processed expression matrices, candidate-gene lists, the complete immune-signature gene sets, and the R analysis scripts are available from the corresponding author on reasonable request, and the underlying data are freely accessible through the Gene Expression Omnibus. In a discipline where irreproducible biomarker claims have slowed progress, that openness matters. If future studies confirm and extend the STAT3-ARG1-associated myeloid signature, clinicians may one day add a molecular gauge of immune suppression to the vital signs and lab panels that currently guide sepsis care, identifying which patients need not just antibiotics and fluids, but a deliberate effort to restart a silenced immune system.

Subject of Research: A bulk peripheral-blood transcriptomic signature associated with STAT3 and ARG1 reflecting myeloid immunosuppression in sepsis

Article Title: Bulk peripheral-blood transcriptomic analysis identifies a STAT3/ARG1-associated myeloid immunosuppressive signature in sepsis

Article References: Bulk peripheral-blood transcriptomic analysis identifies a STAT3/ARG1-associated myeloid immunosuppressive signature in sepsis. (n.d.). https://doi.org/10.1007/s13205-026-05087-1

Image Credits: AI Generated

DOI: 10.1007/s13205-026-05087-1

Keywords: sepsis, transcriptomics, STAT3, ARG1, myeloid-derived suppressor cells, immunosuppression, biomarker, gene expression, ssGSEA, HLA-DR, IL6-JAK-STAT3 signaling, diagnostics