Proteomics of human duodenum in pre-diabetes and type 2 diabetes reveals potential novel therapeutic targets for aetiology and therapeutics

Remission from pre-diabetes and type 2 diabetes (T2D) is frequently observed immediately after a duodenal/jejunal bypass. This demonstrates the reversibility of T2D and involvement of the proximal small intestine in T2D pathology. This study investigates the role of the duodenum in the pathophysiology of T2D, in line with the hypothesis, that exclusion of duodenum/jejunum in T2D (for example, in RYGB surgery) prevents release of unidentified factors that impair insulin sensitivity, leading to T2D remission. The study aimed at identifying novel protein therapeutic targets in human duodenum in T2D and clarifying aetiology. Proteome analysis using nano LC-MS/MS was performed on human duodenum surgical samples: T2D (n = 9), pre-diabetics (n = 6) and non-diabetics (n = 11). Proteins were quantified and analysed using MaxQuant and Perseus software. Linear discriminant (LDA), repeated k-fold cross-validation and receiver operating characteristic (ROC) analyses were used to segregate disease groups for the highest-ranking proteins. Pilot mRNA expression of five selected protein targets was analysed using real-time RT-qPCR on duodenum from human and the pre-clinical high-fat diet C57BL/6J mouse model of pre-diabetes. Proteomics identified and quantified significant changes in 23 proteins (one-way-ANOVA and Holm adjusted pairwise comparisons), differentiating pre-diabetes and T2D. LDA analyses accurately distinguished disease states and identified highest ranking for 10 proteins distinguishing type 2 diabetics, pre-diabetics and non-diabetics with first linear discriminant LD1 explaining 92% of differences; of which three proteins increased in pre-diabetes (DYNC1LI1, KIF5B and MAPRE1; ROC AUC = 0.89,0.82 and 0.86) are involved in vesicle and organelle transport along microtubules to and from the periphery; and seven proteins decrease in T2D (FUBP3, TPPP, NDUFAB1, ATP2B4, NPLOC4, DHRS4, CHMP3; ROC AUC = 0.81, 0.81, 0.77, 0.83, 0.82, 0.83, 0.76) revealing disruption in gene transcription, microtubule stability, mitochondrial function, Ca2+ signalling, unfolded protein response, redox status and multivesicular body formation. Two secreted proteins (APOA4 and RBP4; ROC AUC = 0.79 and 0.80) were also increased in T2D. Modelling using k-fold cross-validation, identified predictive power for 17 of the 23 proteins. At the mRNA level only borderline significant changes in Chmp3 and Fubp3 were detected using the duodenum of the pre-clinical high-fat diet C57BL/6J mouse model of pre-diabetes. Overall, this study addresses a critical knowledge gap, supports our hypothesis, and contributes to a deeper understanding of intestinal involvement in T2D pathogenesis, by examining how changes in the duodenum proteome landscape might reveal mechanisms that contribute to metabolic imbalance in T2D. While these results are currently purely cross-sectional and associative, and their contribution to causality and pathophysiological interpretation for T2D remain to be determined, they expand the frontiers of research on gut-targeted therapies in T2D, and lay groundwork for future translational research into gut-centred drug development to treat T2D or achieve controlled prevention or medical remission of T2D.