Home LiteratureArticle Details
PMID: 42723823 Published · epublish English

A high-risk sepsis subtype identified by regulated cell death signatures and MAD2L2 validation.

Frontiers in immunology ·Vol. 17 ·2026-00-00

Luo L, Cao D, Liao Z, Lin C, Wang A, Zeng L, Yang J, Song W, Deng Y

Abstract

Regulated cell death (RCD) is closely involved in immune dysregulation and organ injury during sepsis, yet the coordinated transcriptional features of multiple RCD modalities and their relevance to sepsis heterogeneity remain insufficiently characterized. Using the whole-blood transcriptomic cohort GSE65682, we performed consensus clustering of RCD-associated genes and identified four molecular subtypes among 479 patients with sepsis. Cluster 3 showed the poorest 28-day survival (log-rank p = 0.034) and was characterized by a distinct mixed immune transcriptional profile. A sensitivity analysis restricted to apoptosis-, ferroptosis-, necroptosis-, and pyroptosis-associated genes reproduced the primary four-cluster structure and retained the high-risk features of Cluster 3. Integrated differential-expression analysis, weighted gene co-expression network analysis, and complementary machine-learning approaches identified a six-gene signature comprising SMC4, SLC5A6, POLR3GL, MAD2L2, EXOC7, and COG1. The resulting RCD.score closely captured the transcriptional phenotype of Cluster 3. In the external GSE54514 cohort, patients with higher RCD.score values had higher APACHE II scores, and exploratory joint stratification identified a subgroup with both high APACHE II and high RCD.score that had the highest observed mortality. Single-cell analysis localized much of the RCD.score-associated signal to monocyte and dendritic-cell populations. Sample-level analyses showed the clearest sepsis-associated increase in monocytes, with MAD2L2 making the largest positive contribution among the six signature genes. In THP-1 cells, LPS stimulation increased MAD2L2, TNF-α, and IL-1β protein expression, whereas MAD2L2 knockdown reduced these inflammatory proteins and decreased GSDMD-N abundance and caspase-1 processing under both LPS and LPS + ATP conditions. These findings define an RCD-associated high-risk sepsis subtype, provide a cell-type-resolved view of its molecular features, and identify MAD2L2 as a candidate involved in monocyte inflammatory and pyroptosis-associated responses.

Keywords
MAD2L2 RCD.score machine-learning molecular endotyping regulated cell death (RCD) sepsis
Article Info
Journal
Frontiers in immunology
Abbr.
Front Immunol
ISSN
1664-3224
Published
2026-00-00
Language
English
Country/Region
Switzerland
NLM ID
101560960
Analysis Services
Analysis Services

Contact

No. 2 Wenbo Road, Zhangqiu District, Jinan, Shandong

Qilu Normal University · Genelibs Bioinformatics Lab

750 Shunhua Rd, Jinan

2F, Bldg F, University Science Park

Tel: 0531-88819269

WeChat Official Account

Follow our WeChat subscription account for real-time updates and the latest in medical and biological research.


Business Email

E-mail: product@genelibs.com