Li et al. 2025 Immunometabolic Alterations in Type 2 Diabetes

Li et al. 2025 analyzed public PBMC single-cell RNA-seq data to characterize immune-cell composition, T-cell metabolic heterogeneity, T-cell-monocyte communication, transcription-factor activity, and drug-enrichment signals in type 2 diabetes.

Citation

• Li H, Zou L, Long Z, Zhan J. 2025. “Immunometabolic alterations in type 2 diabetes mellitus revealed by single-cell RNA sequencing: insights into subtypes and therapeutic targets.” Frontiers in Immunology 15:1537909. DOI: 10.3389/fimmu.2024.1537909.

Study Design From Li et al. 2025

• The study used public GEO single-cell RNA-seq data from 37 T2D patients in GSE268210 and 11 healthy-control individuals in GSE244515.
• Seurat v5.0.1 was used for filtering, normalization, variable-gene detection, dimensionality reduction, clustering, and differential expression.
• Harmony integrated the datasets to reduce batch effects and technical variation.
• Cells were retained if they had 500-3,500 detected genes, genes expressed in at least five cells, and mitochondrial expression not exceeding 5%.
• Major PBMC cell types included CD4+ T cells, CD8+ T cells, NK cells, B cells, monocytes, dendritic cells, and plasma cells.

Main Findings From Li et al. 2025

• T2D samples showed higher monocyte proportions and lower CD4+ T-cell proportions than healthy controls.
• T2D samples showed higher cytotoxic CD8+ T-cell and naive CD8+ T-cell proportions and lower regulatory CD4+ T-cell proportions.
• Monocyte reclustering found higher intermediate monocyte proportions and lower classical monocyte proportions in T2D.
• GSVA over 42 KEGG metabolic pathways separated samples into four groups: T2D subtypes A-C and a healthy-control group D.
• CellChat inferred stronger T-cell-monocyte communication in T2D, with MHC-I signaling contributing broadly across T2D subtypes.
• DoRothEA/VIPER analysis identified 126 active transcription factors, including immune and metabolic regulators discussed by the authors as relevant to NF-kB, STAT3, FOXO1, and subtype-specific T-cell states.
• Machine-learning models classified T2D subtypes from T-cell metabolic signatures with validation AUC values above 0.8 and average AUC above 0.84.
• DGIdb drug-enrichment analysis nominated suloctidil for subtypes A and B and chlorpropamide for subtype C.

Manuscript Use

• This paper strengthens the claim that T2D PBMC immune remodeling includes immunometabolic T-cell states, not just broad cell-composition shifts.
• It is most useful for mechanistic framing around T2D T-cell immunometabolic subtypes, T-cell-monocyte communication, and T2D transcription-factor activity.
• It does not provide direct evidence for ancestry-associated immune differences, so it should support background and hypothesis framing rather than ancestry-specific claims.

Limitations From Li et al. 2025

• The study used public datasets, had a limited sample size, and was cross-sectional, limiting generalizability and causal inference.
• The T2D and healthy-control cells came from different GEO accessions, so residual cohort or batch differences may remain despite Harmony integration.
• Drug-enrichment hits nominate hypotheses and should not be treated as evidence of clinical efficacy for T2D subtypes.
• The paper notes demographic variation as a limitation, but it does not perform ancestry-stratified analysis.

Links

• Gu et al. 2024 × Li et al. 2025 — synthesis

Sources

• _raw/Li2025immunometabolic.pdf