4. Discussion
Pharmacological intervention targeting SGLT2 has recently attached great concern as a novel therapeutic option for diabetes. Increasing evidences have confirmed the extra renal protective effect of SGLT2 inhibitors beyond the controlling of hyperglycemia.(Perkovic et al., 2019) However, the mechanism through which SGLT2 inhibitors exert the renal protective effect has been rarely reported. Here we find that the Empa can relieve the tubular injury by alleviating the mitochondrial fission though an AMPK-dependent pathway in diabetic mice. By activating AMPK, Empa can remarkably rescues the dephosphorylation of DRP1S637 and inhibits the mitochondria-translocation of DRP1, which eventually result in less mitochondrial fission. The present study revealed the existence of a glycemia independent mechanism of Empa that protects against mitochondrial dysfunction and alleviates the tubular injury in diabetic mice and HK2 cells.
DKD was traditionally considered as a glomerular disease characterized by mesangial expansion, reduction in capillary surface, and podocyte loss which differ it from other glomerulonephritis.(Thomas et al., 2015) However, increasing evidences have revealed the primary contribution of tubular injury in the pathogenesis of DKD.(Bonventre, 2012; Tang & Lai, 2012) In the present study, we found the urinal biomarkers NAG and NGAL which reflect the early tubular injury increased obviously at the 10th week without change in albuminuria in KK-Ay mice. Since the albuminuria is an early clinical manifestation of DKD and reflects glomerulopathy, our finding verifies the notion that diabetic tubulopathy occurs earlier than glomerular disease.
Recent clinical trials have confirmed the renal protective effect of SGLT2-inhibitors(Mahaffey et al., 2018; Wanner, Inzucchi & Zinman, 2016). In present study, both the TUNEL and Western Blotting assay consolidated that the Empa could alleviate apoptosis in the kidney of KK-Ay mice. Recent animal and clinical studies have confirmed the critical role of mitochondrial fission in the pathogenesis of DKD.(Galvan et al., 2019; Jiang et al., 2019) However, the signaling pathways by which hyperglycemia leads to mitochondrial fission are not fully understood.(Wang et al., 2012) Here we found that the DRP1S637 decreased significantly and the mitochondria translocation of DRP1 increased remarkably in the kidney of KK-Ay mice. The DRP1 phosphorylation on serine-637 was previously reported to be involved in Parkinson’s disease,(Dagda et al., 2011) myocardial ischemia-reperfusion injury(Sharp et al., 2014) and regulation of ER morphology in stressed pancreatic β-cells through its anti-fission effect.(Wikstrom et al., 2013) Zhou et al. reported that the Empa could rescue the dephosphorylation of DRP1S637 in cardiac microvascular endothelial cells and alleviated the mitochondrial fission, but the intrinsic mechanism remained unclear.(Zhou, Wang, Zhu, Hu, Chen & Ren, 2018)
To detect the exact mechanism through which the Empa impacts the DRP1 phosphorylation, we conducted a series of investigations. In present study, we found an obviously decreased phosphorylation of AMPK and DRP1S637 in KK-Ay mice kidney, and this change was reversed in the Empa group. We presumed that the Empa may rescue the dephosphorylation of DRP1S637 and alleviate the mitochondrial fission via an AMPK dependent pathway. We also found an obvious increase of SP1 and PGAM5 expression in KK-Ay mice kidney which were reversed in Empa group. These findings remind us that some causative relationship may exist among the AMPK, SP1 and PGAM5.
To verify our deduction, we performed in vitro experiments using HK2 cells. We found that the apoptosis increased significantly in HG treated cells. Meanwhile, the increased mitochondrial fission and disrupted mitochondrial membrane potential were detected in HK2 cells subjected to HG. All these changes were alleviated in Empa treated cells. These findings were consistent with the in vivo experiments.
Then we tested the phosphorylation of AMPK, the expression of SP1 and PGAM5 in HK2 cells to testify our presumption. Here we found that the AMPK activity decreased obviously in HG treated cells and the expressions of SP1 and PGAM5 increased significantly. These changes were rescued in Empa treated cells. Those above findings confirmed the involvement of AMPK, SP1 and PGAM5 in the development of mitochondrial fission and the Empa’s anti-fission effect. But if there is a causality among these changes was unclear. So, we introduced the AMPK activator and inhibitor to our study. We found that not only the AMPK activator but also the Empa could increase the AMPK phosphorylation. Both the two chemicals could decrease the SP1 and PGAM5 expressions and alleviate the mitochondrial fission. However, the Compound C, an AMPK inhibitor reversed the declining of SP1 and PGAM5 and blocked the anti-fission effect in Empa treated cells. These findings suggested that the AMPK located at the upstream of the AMPK/SP1/PGAM5 axis and played a central role in the process of HG induced mitochondrial fission in HK2 cells.
To further illustrate how the SP1 and PGAM5 participate in the development of the mitochondrial fission, si-SP1 and si-PGAM5 were introduced to our later investigations. We found that both the si-SP1 and si-PGAM5 could alleviated the HG induced mitochondrial fission, but neither of them could activate the AMPK. Meanwhile, the Empa could not only activate the AMPK but also decrease the SP1 and PGAM5 expression and relive the HG induced mitochondrial fission in HK2 cells. These phenomena consolidated our deduction that the AMPK located at the upstream of the AMPK/SP1/PGAM5 axis and the Empa could alleviated the mitochondrial fission via an AMPK dependent pathway. Meanwhile, we found that the si-SP1 could decrease the PGAM5 expression, but the si-PGAM5 had no effect on the SP1 protein. This result suggests that the SP1 could mediate the PGAM5 expression but it is not contrarily. Finally, we verified the direct interaction between the SP1 and the PGAM5 by luciferase reporter assay and chromatin immunoprecipitation (ChIP), and our findings clearly showed that the SP1 could bind to the promotor of PGAM5 at -539 nt and increased its expression.
In conclusion, the present study depicted an AMPK regulated and PGAM5 involved mitochondrial fission pathway in the development of diabetic tubular disease (Figure 6), and we also revealed a glycemia independent mechanism of Empa that protects against the diabetic tubular injury. Our finding could not only provide the experimental and theoretical basis for the expansion of new indications of Empa, but also revealed a new prospective therapeutic target for DKD.