Immunoglobulin G N-Glycome as a biomarker of mortality risk in Escherichia coli induced sepsis

Huachen Wang, Houqiang Li, Zheng Guo, Hongda Hou, Haifeng Hou, Bing ChenFrontiers in Immunology 2025 · DOI

Abstract

Immunoglobulin G N-glycome concentrations at ICU admission are valuable for predicting the in-hospital mortality risk of patients with sepsis, suggesting that immunoglobulin G N-glycome may be a novel biomarker.

Introduction

Sepsis is a life-threatening medical emergency characterized by a dysregulated host response to infection, often leading to multi-organ failure. Escherichia coli is among the most frequent pathogens responsible for these cases, which carry mortality rates between 10% and 50%. While clinicians currently rely on the Sequential Organ Failure Assessment (SOFA) score to monitor patient status, its ability to predict mortality at the time of ICU admission is limited. This study investigates whether the IgG N-glycome—the complex carbohydrate structures attached to antibodies—can serve as a dynamic "molecular barcode" for more accurate risk stratification.

Methodology

The researchers conducted a case-control study involving 100 adult patients with E. coli sepsis and 100 healthy controls. IgG N-glycan profiles were analyzed using Hydrophilic Interaction Chromatography-Ultra Performance Liquid Chromatography (HILIC-UPLC), identifying 24 initial glycan peaks (GPs) and 51 derived traits. To identify the most predictive features and prevent model overfitting, the team employed two machine learning techniques: Least Absolute Shrinkage and Selection Operator (LASSO) and Recursive Feature Elimination (RFE). Logistic regression models were then developed to evaluate the diagnostic and prognostic performance of these glycans.

Results

Alterations in the IgG N-Glycome

Significant differences were observed in the relative intensities of various IgG N-glycans between septic patients and healthy controls.

Relative intensities of 24 IgG N-glycan peaks (GP1-GP24) in healthy controls vs. sepsis patients.

The study identified a notable pro-inflammatory shift, specifically a significant drop in fucosylation. In immunology, the absence of fucose on an IgG molecule enhances its ability to trigger Antibody-Dependent Cellular Cytotoxicity (ADCC), suggesting an aggressive immune response. Additionally, alterations in galactosylation and sialylation levels were found to reflect the systemic "cytokine storm" characteristic of sepsis.

Diagnostic and Prognostic Performance

The combination of specific glycans demonstrated exceptional diagnostic accuracy. Using a subset of six glycans identified through LASSO and RFE, the researchers achieved an Area Under the Curve (AUC) of 0.976 in the testing set for distinguishing sepsis from healthy controls.

Feature selection via LASSO/RFE and ROC curves showing high diagnostic accuracy for sepsis.

Regarding mortality prediction, the SOFA score alone yielded a mediocre AUC of 0.58. However, integrating the SOFA score with glycans GP4 and GP7 significantly improved the predictive power to an AUC of 0.85. Notably, high levels of glycan GP4 were identified as an independent risk factor; patients with elevated GP4 at admission faced a 13-fold increase in the risk of in-hospital mortality.

Discussion

The IgG N-glycome acts as a biological sensor, capturing the current state of immune dysregulation more dynamically than static genetic markers. The synergy between clinical scores (SOFA) and molecular markers (glycans) suggests they provide complementary information: while SOFA measures existing organ damage, glycans likely reflect the underlying immune drivers of that damage.

Conclusion

IgG N-glycans represent a promising new class of biomarkers for sepsis. While further validation in larger, multi-ethnic cohorts and across different pathogens is required, the ability of these molecular signatures to refine mortality risk at ICU admission could lead to more personalized and timely interventions for high-risk patients.

Summary updated

Abstract

Immunoglobulin G <i>N</i>-glycome concentrations at ICU admission are valuable for predicting the in-hospital mortality risk of patients with sepsis, suggesting that immunoglobulin G <i>N</i>-glycome may be a novel biomarker.
BibTeX
@inproceedings{wang2025immunoglobulin,
  author = {Huachen Wang and Houqiang Li and Zheng Guo and Hongda Hou and Haifeng Hou and Bing Chen},
  title = {{Immunoglobulin G N-Glycome as a biomarker of mortality risk in Escherichia coli induced sepsis}},
  booktitle = {Frontiers in Immunology},
  year = {2025},
  doi = {10.3389/fimmu.2025.1532145},
}
Figure