Unlocking the power of empagliflozin: Rescuing inflammation in hyperglycaemia-exposed human cardiomyocytes through comprehensive multi-level analysis

Abstract

Aims: Hyperglycaemic conditions increase cardiac stress, a common phenomenon associated with inflammation, aging, and metabolic imbalance. Sodium-glucose cotransporter 2 inhibitors, a class of anti-diabetic drugs, showed to improve cardiovascular functions although their mechanism of action has not yet been fully established. This study investigated the effects of empagliflozin on cardiomyocytes following high glucose exposure, specifically focusing on inflammatory and metabolic responses.

Methods and results: A three-part strategy was formulated: (i) a meta-analysis of selected randomized clinical trials was carried out to assess the anti-inflammatory effects of empagliflozin in diabetic patients; (ii) the impact of empagliflozin on human cardiomyocyte AC16 cells exposed to normal (5 mM) and high (33 mM) glucose concentrations for 2 and 7 days was explored by evaluating gene expression and protein levels of pivotal markers associated with cardiac inflammation, stress, endoplasmic reticulum damage, and calcium modulation; (iii) in silico data from bioinformatic analyses were exploited to construct an interaction map delineating the potential mechanism of action of empagliflozin on cardiac tissue. Empagliflozin reversed high-glucose mediated alterations at the transcriptional level, decreasing inflammatory, metabolic, and aging signatures. Specifically, in vitro experiments on human cardiomyocytes, meta-analyses of clinical data on inflammatory biomarkers from diabetic peripheral blood samples, and sequencing of pathological human heart tissues, all support that empagliflozin exerts anti-inflammatory effects both systemically and directly in cardiac tissue, on cardiomyocytes.

Conclusion: Our study provides insights based on robust mechanistic data for optimizing heart failure management and highlights the intricate interplay between diabetes, inflammation, aging, and cardiovascular health.

Keywords: Empagliflozin; Glucose; Human cardiomyocytes; Inflammaging.

Figure 1

Forest plot of diabetic patients untreated or treated with sodium–glucose cotransporter 2 inhibitors. Data are shown as mean change in biomarkers to compare differences between groups. Hattori (trial registration: UMIN Clinical Registry, UMIN000021552), Dekkers (Netherlands Trial Register, NTR 4439), Latva‐Rasku (trial registration: ClinicalTrials.gov, NCT02426541), Suhrs (EU Clinical Trials Register, 2017‐000240‐17), and Phrueksotsai (clinicaltrials.in.th, TCTR20170511001). CI, confidence interval; CRP, C‐reactive protein; IL‐6, interleukin‐6; SD, standard deviation; TNF‐α, tumour necrosis factor‐alpha.

Figure 2

(A) Volcano plots displaying differentially expressed genes (DGEs) from normalized RNA‐seq reads. The x and y axes indicate fold‐change resulting from comparisons expressed in log2 and p‐value in −log10, respectively; alluvial plots showing gene ontology (GO) for biological processes of the top 20 DGEs (high glucose [HG] vs. control [CTR] conditions, left; HG + empagliflozin [EMPA] vs. HG conditions, right). (B) Gene intersection across DGEs from HG versus CTR and HG + EMPA versus HG comparisons with DGEs from diabetic and non‐diabetic patients affected by post‐ischaemic heart failure. (C) Venn diagram showing the intersection of GO biological processes from common genes of gene intersection analysis.

Figure 3

(A) Enrichment plots displaying processes altered upon high glucose (HG) stimulation with (left) and without (right) empagliflozin (EMPA). Normalized enrichment score (NES) and p‐value indicate differential expression in pathway enrichment across comparisons and related statistical significance, respectively. (B) Significant processes from enrichment analysis in HG versus control (CTR) and HG + EMPA versus HG. Terms enriched are significant for p < 0.5, reported transformed as −log10 in the label. NES indicates the value of enrichment. The size of each term is reported as gene count in the label. (C) Alluvial plot illustrating the main pathways altered by HG and EMPA. The plot shows that EMPA rebalances gene expression in the direction of normal glucose (NG).

Figure 4

(A) Venn diagram showing the intersection between genes recovered upon empagliflozin (EMPA) treatment in high glucose (HG) conditions. (B) Heatmaps showing gene ontology analysis of biological processes from common differentially expressed genes downregulated upon HG exposure from the HG versus control (CTR) comparison (red) and upregulated with EMPA from the HG + EMPA versus HG comparison (blue). Data are reported as the fold‐change from each comparison in log2. (C) Dot plots showing biological processes with opposite patterns of expression and related gene signature expression. Data are reported as the fold‐change from each comparison in log2. (D) Venn diagrams showing the intersection between genes recovered upon EMPA treatment in HG conditions and the inflammaging signature. Dot plot showing significant genes involved in inflammaging reduced with HG and recovered upon EMPA treatment. Data are reported as the fold‐change from each comparison in log2. (E) Gene ontology chart indicating significant biological processes associated with common genes involved in inflammaging.

Figure 5

(A) Heatmaps illustrating cytokine content in a time course setting (3, 6, and 12 h and then 1, 2, 3, 4, 7 days) upon high glucose (HG) exposure with/without empagliflozin (EMPA). Data are reported as log2 ratio normalized on protein content. (B) Western blot analysis of interleukin‐6 (IL‐6), NLRP3, ASC, and GLUT1 in normal and high glucose conditions for days, untreated or treated with EMPA at 0.5 μM. (C) Western blot analysis of total and phosphorylated form of NF‐κB in normal and high glucose conditions for 2 and 7 days, untreated or treated with EMPA at 0.5 μM. (D) Histogram showing Fura‐2 AM fluorescence values at a wavelength of 335 nm for excitation and 512 nm emission resulting from HG exposure with/without EMPA for 2 and 7 days. Statistical significance was calculated with unpaired t‐test and is reported as follows: *<0.05, **<0.01, ***<0.001. (E) Histogram showing fatty acid accumulation as absorbance values at 570 nm resulting from HG exposure with/without EMPA for 2 and 7 days. Statistical significance was calculated with unpaired t‐test and is reported as follows: *<0.05, **<0.01, ***<0.001.

Figure 6

(A–E) Metabolic parameters based on oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) values upon high glucose (HG) exposure with/without empagliflozin (EMPA). Statistical significance was calculated with unpaired t‐test and is reported as follows: *<0.05, **<0.01, ***<0.001. Metabolic phenotypes based on basal respiration and glycolysis. (F) Western blot analysis of total and phosphorylated form of eIF2α in AC16 cells in normal and high glucose conditions for 2 and 7 days, untreated or treated with EMPA at 0.5 μM, EX‐527 at 10 μM, and combo treatment (EMPA 0.5 μM + EX‐527 10 μM).