Uncovering the molecular networks of ferroptosis in the pathogenesis of type 2 diabetes and its complications: a multi-omics investigation.

BACKGROUND

Diabetes is a multi-factorial disorder and related complications constitute one of the principal causes of global mortality and disability. The role of ferroptosis in diabetes and its complications is intricate and significant. This study endeavors to disclose the role of ferroptosis in the aforementioned diseases from multiple perspectives through multi-omics.

METHODS

We performed genetic correlation analyses via the Linkage Disequilibrium Score and High-Definition Likelihood approaches for type 2 diabetes (T2D) and its complications. The data concerning the expression of ferroptosis-related genes (FRGs) were obtained from the meta-analysis of studies on gene expression and protein abundance. Mendelian randomization analyses and cross-validation were implemented using the discovery cohort, replication cohort, and imaging genomics cohort of T2D and its complications. Moreover, we conducted colocalization analyses on T2D and tissue-specific single-cell RNA sequencing investigations on the complications to complement the results.

RESULTS

Genetic association analysis indicated that the selected datasets could be incorporated into a secondary analysis of T2D complications. In the primary analysis, six FRGs (CDKN1A, ENO3, FURIN, RARRES2, TYRO3, and YTHDC2) were found to be positively associated with T2D risk. Conversely, eight FRGs (ARNTL, CAMKK2, CTSB, FADS2, KDM5A, MEG3, SREBF1, and STAT3) were inversely associated with T2D risk. The 14 FRGs were included in the secondary analysis. Within the FRGs, which received full support from both the discovery and replication cohorts, and were further validated by imaging genomics, higher levels of CDKN1A were positively associated with DKD risk. Higher levels of CAMKK2 and KDM5A were associated with a decreased risk of DKD. For DCM, higher levels of CTSB were positively associated with DCM risk. And genetically predicted higher levels of ARNTL and SREBF1 were associated with a decreased risk of NAFLD. Finally, we validated the tissue-specific expression of each complication with scRNA-seq datasets.

CONCLUSIONS

This study identified FRGs in relation to T2D and its complications, which may enhance the understanding of the pathogenic mechanisms of their development. Meanwhile, it offers cross-validation for imaging genomics and further indicates the direction for non-invasive diagnosis.