Genomics of dendritic cell pyruvate metabolism reveals stage-associated circulating immune signatures in diabetic retinopathy.

BACKGROUND

The role of circulating pyruvate in diabetic retinopathy (DR) progression is poorly defined. Unravelling its cell-specific genomic regulation is vital.

METHODS

Integrative genomics analysis combined two-sample Mendelian randomization (MR) to evaluate the association between circulating pyruvate and DR risk with single-cell RNA sequencing (scRNA-seq) metabolic profiling in peripheral blood mononuclear cells (PBMCs). Machine learning was used to derive exploratory blood-based candidate biomarkers from metabolism-associated genes, followed by qPCR assessment in an independent peripheral blood cohort.

RESULTS

MR indicated an inverse association between elevated pyruvate and DR risk (IVW OR = 0.617, 95% CI = 0.399-0.953, p = 0.029). scRNA-seq identified dendritic cells (DCs) as the immune subset with the highest pyruvate-pathway activity score. Trajectory-based analysis suggested a late-state increase in pyruvate-pathway activity in DR DCs that was not observed in controls. A 7-gene metabolic signature showed moderate discrimination between groups (validation AUC = 0.79), and qPCR showed higher expression of SMIM5, PFKFB2, and PLXNA4 in DR peripheral blood. Database-based toxicogenomic analysis highlighted benzo(a)pyrene and sodium arsenite as candidate exposures associated with regulation of these target genes.

CONCLUSION

Our findings support an association between DC-related pyruvate metabolism and stage-associated immune states in DR. We identify blood-based candidate biomarkers and database-derived environmental hypotheses for future validation. These results should be interpreted as observational and hypothesis-generating pending functional experiments.