Protein-protein interaction network construction and analysis to identify synergistic effects for sodium-glucose cotransporter-2, glucagon-like peptide-1 and mineralocorticoid receptor

Abstract Introduction Recent studies have reported benefits of combining sodium-glucose cotransporter-2 (SGLT2) inhibitors and glucagon-like peptide-1 (GLP1R) agonists. There may be additional benefit by further adding mineralocorticoid receptor (MR) antagonists. This bioinformatic study conducted protein-protein interaction analysis for the genes encoding for SGLT2 (SLC5A2), GLP1R (GLP1R) and MR (NR3C2). Methods Network interaction analysis of protein products from the SGLT2, GLP1R and MR genes was conducted using the STRING database and analysis tools. Nodes represented shells of interactors whilst edges represented protein-protein interactions. Known interactions from curated databases and experiments, and predicted interactions from gene neighbourhoods, fusions or co-occurrences were denoted by different colours. Gene co-expression analysis was conducted to identify strengths of gene partners. Gene ontology enrichment analysis was used to identify relevant physiological processes and diseases. Epidemiological analysis was performed on using the Hong Kong diabetes cohort comparing patients receiving their first prescription of GLP1R agonists with or without SGLT2 inhibitors between 2015 and 2019 on cardiovascular outcomes. Results Interaction analysis revealed stronger confidence for interactions of SLC5A2-GLP1R (combined score: 0.909) than of SLC5A2-NR3C2 (combined score: 0.715) or GLP1R-NR3C2 (combined score: 0.436) (Figure 1). Gene ontology enrichment analysis revealed the strongest signals and the lowest false discovery rates for sodium ion import and inorganic cation transport (Figure 2). In the propensity score matched cohort between GLP1R agonist-SGLT2 inhibitor users and GLP1R agonist users, combination therapy was associated with lower risks of myocardial infarction (hazard ratio HR: 0.34 0.15-0.77) and heart failure (HR: 0.40, 0.20-0.82) but not for atrial fibrillation (HR: 0.66 0.28-1.54 or stroke/transient ischaemic attack (HR: 0.58 0.24-1.43). Conclusions Bioinformatic analysis revealed protein-protein interactions between SLC5A2 with GLP1R and NR3C2. Such effects may explain the lower risks of myocardial infarction and heart failure in patients receiving GLP1R agonist-SGLT2 inhibitor combination therapy. Further epidemiological analysis is required to identify whether addition of MR antagonists to this combination is associated with better outcomes.Figure 1For image description, please refer to the figure legend and surrounding text. Figure 2For image description, please refer to the figure legend and surrounding text.