Our study demonstrates that i) in human AC16 cells, high glucose treatment induces significant demethylation in the promoter regions of NF-kB and SOD2; ii) the observed DNA demethylation is mediated by an increase of TET2 binding to the CpGs island in NF-kB and SOD2 promoters; iii) empagliflozin prevents HG-induced demethylation changes by reducing TET2 binding to the investigated promoter region and counteracts the altered genes expression; iv) transient SGLT2 gene silencing prevents the DNA demethylation observed in promoter regions thus suggesting a role of SGLT2 as a potential target of the anti-inflammatory and antioxidant effect of empagliflozin in cardiomyocytes.
Hyperglycemia, the mainly pathogenetic mechanism of diabetes, induces changes in redox status, inflammation, metabolic profiles, intracellular signaling pathways, and energy production, predisposing to cardiovascular diseases [23].
Preclinical models and human studies have addressed the link between epigenetic factors, type 2 diabetes, and cardiovascular diseases. Hyperglycemia induces epigenetic changes that lead to the altered expression of gene implicated in oxidative stress and inflammation [24]. Actually, it was demonstrated that high glucose correlates with a modified DNA methylation pattern [25]. In our previous studies, we demonstrated that in human aortic endothelial cells, high glucose induced an increase in TET2 binding on NF-kB and SIRT6 promoter region, leading to significant demethylation and, consequently, an increase of gene expression. In agreement, also diabetic patients showed statistically significant lower levels of NF-κB and SIRT6 DNA methylation compared to nondiabetic patients [5, 6].
In addition, SET7, a lysin methyltransferase, in response to a change in glucose concentration, translocates in the nucleus regulating the NF-kB pathway [26, 27]. Moreover, hyperglycemia reduced H3K4me1 and -me2 and increased the binding of LSD1 and Sp1 at the Sod2 gene [28].
Interestingly enough, our results demonstrate that, also in human cardiomyocytes, high glucose exposure induces an increase in NF-κB and SOD2 expression through an increment in the demethylation levels of specific CpGs islands located in their promoter regions, which might affect cardiac function and be associated with the development and progression of cardiovascular disease.
The concomitant activation of the epigenetic machinery and the increased binding of TET2 in the promoter region of the investigated gene demonstrates that glucose exposure, DNA methylation and gene expression changes are causally linked. More intriguing, our results firstly showed that treatment with empagliphozin reduced the expression levels and TET2 binding to the promoter region of NF-kB and SOD2, preventing the HG-induced demethylation and restoring the normal levels of gene expression.
SGLT2i antidiabetic class demonstrates large cardiovascular benefits in both diabetic and non-diabetic patients mainly due to systemic effects derived from glycemic control that improved metabolic, hormonal, and hemodynamic whole-body homeostatic [29, 30]. However, additional mechanisms due to direct effects on cardiac cells, as effects on inflammation, oxidative stress, and intracellular ion homeostasis were also identified [18].
In this regard, recent studies demonstrated that SGLT2i reduce cardiac inflammation through the inhibition of cardiac NLRP3 inflammasome [19, 31], and a reduction of the levels of myocardial pro-inflammatory cytokines, including ASC, caspase-1, IL-1β, IL-6 and TNFα [31, 32]. Furthermore, SGLT2i has been shown to also play an important role in the reduction of oxidative stress that is a main contributor to the pathogenesis of cardiovascular disease.
Nishitani S et al. showed that dapagliflozin-treated mice, had higher circulating and tissue levels of β-hydroxybutyrate, a molecule involved in histone modification [33] and speculated that the beneficial health effects of SGLT2 I could be associated with epigenetic mechanism. Indeed, any convincing data supporting their hypothesis were provided.
Furthermore, Solini et al. demonstrated that dapagliflozin modulates miRNA expression. Indeed, upregulating miRNA-30e-5p, dapa inhibits myocardiocyte autophagy and heart failure, and reducing miRNA-199a-3p, dapa causes a reduction in cardiac PPAR levels, ameliorating mitochondrial fatty acid oxidation and improving cardiac function in patients with heart failure. Moreover, dapagliflozin exerts nephroprotection by preserving renal vasodilating capacity by reducing miRNA-27b expression [34]. These results first suggested epigenetic mechanisms of SGLT2i in improving cardiac functions through miRNAs modulations.
Our data provide, for the first time, evidence of the SGLT2i ability to exert their anti-inflammatory and antioxidant effect by modulating NF-kB and SOD2 DNA methylation, directly targeting cardiomycyte SGLT2.
Previous studies clearly demonstrated that SGLT2i effects at cardiac level are mediated through the modulations of SGLT1, Na+/H + exchanger 1 (NHE1), Ca2+/calmodulin-dependent protein kinase II (CaMKII), and late Na + current (late INa) [35]. Indeed, we recently provided evidences that SGLT2 protein is expressed in human hearts of diabetic and non-diabetic patients and in human cardiomyocyte and that hyperglycemia condition induces its overexpression. In addition, the observed high glucose induced cardiomyocytes SGLT2 overexpression is associated with increased oxidative stress, inflammation, and apoptosis which in turn leads to heart dysfunction. More intriguing, the silencing of SGLT2 blunted mitochondrial oxidative proteins COX-IV, Cytochrome c and increased the expression levels of the guardian SIRT3 in cardiomyocytes exposed to high glucose [36].
Therefore, in light of such recent evidences, it cannot be ruled out that SGLT2i might prevents the HG-induced demethylation and expression changes observed in NF-kB and SOD genes, by acting through a direct inhibition of SGLT2. To verify such hypothesis, the effect of EMPA, an SGLT2i with the greatest selectivity for SGLT2 [37], on epigenetic machinery was tested in cardiomyocyte exposed to high glucose condition and treated with small SGLT2 RNA interfering. Interestingly, our results showed that the SGLT2 silencing prevented the HG-induced hypo-methylation in the promoter region of NF-kb and SOD2, without any significant differences with cells treated only with HG + EMPA. These results confirm the hypothesis that the observed epigenetic effects of SGLT2 inhibitors might be mainly explained by the interaction with SGLT2. In addition, the stronger effect observed with a double inhibition treatment (EMPA + SGLT2 silencing) also suggest that other targets, ie. SGLT1, Na channel might be probably also involved.
Potential In Vivo Implications
Our study showing an “in vitro” causal link between SGLT 2 inhibition, DNA methylation and gene expression changes demonstrates that SGLT2 inhibitors are potential therapeutic epigenetic regulators thus suggesting a potential clinical implication of our results.
An “in vivo” detrimental cardiac effect of hyperglycemia on DNA methylation has been previously demonstrated in diabetic patients [5, 6, 38]. More specifically, plasma glucose levels were found negatively correlated with DNA methylation in peripheral leukocytes of the promoter region of NF-kB genes. Interestingly enough, in diabetic patients oral hypoglycemic agent therapy resulted a significant predictor of NF-κB DNA methylation, independently of age, sex, body mass index (BMI), glucose and plasma lipid levels [5, 6]. Furthermore, a significant correlation analysis of DNA methylation profiles with intima-media thickness (IMT), a surrogate marker for early atherosclerosis, left ventricular mass (LVM), left ventricular ejection fraction (LVEF), and cardiac performance index (MPI) was also found in 365 healthy subjects independently of the other risk factors [38].
We acknowledge that results obtained only in vitro is a potential limitation of our study and that further in vivo studies are necessary for validating our data. Notwithstanding, these previous evidences strongly suggest the potential clinical implication of our results in terms of cardiovascular outcome. Furthermore, results showing that both EMPA and SGLT2 silencing, reduced the expression levels and TET2 binding to the promoter region of NF-kB and SOD2, preventing the HG-induced demethylation and restoring the normal levels of gene expression, clearly demonstrates that SGLT 2 inhibition, DNA methylation and gene expression changes are causally linked.