**3. Oxidative Stress in Diabetic Kidney Disease: Contribution of Sodium Glucose Co-Transporter Type 2 (SGLT2)**

Albeit glucose consumption is low in the proximal tubule (except for segment S3 [62]), PTCs are primarily responsible for its reabsorption via the sodium/glucose co-transporter type 2 (SGLT2) [44]. SGLT2 is located at the brush border of PTCs in the S1 and S2 segments of the proximal tubule, where 90% of the total filtered glucose is reabsorbed [63]. Its transport is coupled with sodium (Na+/Glucose, 1:1) and it accounts for 5.7% of total renal Na+ reabsorption in healthy individuals [64]. Although SGLT2 activity has a role in maintaining the intrarenal osmolarity, the main sodium exchanger at the tubular level is the Na+/H+ Exchanger-3 (NHE3). Interestingly, NHE3 and SGLT2 have been shown to be interdependent [65]. Therefore, SGLT2 activity has a huge impact on both glucose

and sodium balance. The Na+ gradient created by the sodium-potassium pump (Na+/K+- ATPase), located in the basolateral membrane of PTC, is harnessed by SGLT2 to transport glucose and sodium across the plasma membrane down-gradient [63]. At the same time, the basolateral glucose transporter, GLUT2, maintains the glucose gradient by passively easing its backflow into the bloodstream [66].

SGLT2 plays a crucial role in glucose transport and intrarenal osmolarity. For that reason, the dysfunction of this co-transporter, which happens in diabetes, has a direct effect on intrarenal glucose metabolism and also on the redox environment. In diabetic patients, SGLT2 is overactivated, leading to increased glucose and sodium reabsorption [67]. The deleterious effects in kidney function and fluid homeostasis include the alteration of tubular-glomerular feedback (a self-regulated system of the kidney) and the increase of the glomerular filtration rate and the tubular reabsorption capacity. Systematically, the plasma volume and blood pressure increase, leading to hypertension, a common feature of these patients [68]. At the molecular level, this has drastic consequences for the activity of other transporters located in the tubule, such as NHE3, which is sensitive to small changes in sodium concentration. Packer et al. reviewed that Na+/H+ Exchanger-1 (NHE1) in the heart and vasculature and NHE3—the kidney isoform—are upregulated in heart failure and type 2 diabetes mellitus (T2DM) [69]. Additionally, when sodium levels drop in the luminal filtrate, Na+/K+-ATPase activity increases to restore the sodium gradient. As this transporter is ATP-dependent, its overactivation implies an increase in energy demand and oxygen consumption. It is worth mentioning that under normal conditions, 60% of kidney ATP consumption is intended for sodium uptake which is mostly attributed to basal Na+/K+-ATPase activity [70]. Therefore, the energy consumption of this transporter cannot be underestimated, and its impact on mitochondrial function should be further explored. The enhanced mitochondrial phosphorylation produces substantial accumulation of ROS products. In addition, it has been described that interleukin 6 (IL-6) activates SGLT2 through ROS in cultured tubular cells [71], which suggests that an inflammatory context would further promote ROS generation and SGLT2 overactivation. Thus, SGLT2 activity, oxidative stress, and inflammation actively contribute to the onset of DKD. By definition, DKD is a microvascular complication of diabetes mellitus that affects the kidney. These patients are diagnosed by the presence of persistent albuminuria, starting at first stage by hyperfiltration and ending with a progressive decrease in GFR at later stages. The Renal Pathology Society divide DKD pathology into glomerular, tubulointerstitial, and vascular lesions [72]. Focusing on the tubule, tubular membrane thickening, interstitial fibrosis, and tubular atrophy are usually observed. Coughlan et al. mapped the changes in mitochondrial adaptation in experimental rat's kidney mitochondria at 4, 8, 16, and 32 weeks after induction of diabetes by streptozotocin intravenous injection [73]. The authors demonstrated that DKD begins with decreased ATP production in the renal cortex, mitochondrial fragmentation, which is accompanied by increased ROS, and early renal hyperfiltration. Their results suggest that mitochondrial dynamics and bioenergetic function worsen over time, preceding the development of albuminuria and histological lesions. An in vitro experiment with HK-2 cells found out that under high glucose conditions, HK-2 cells were susceptible to mitochondrial fragmentation. These effects were ameliorated with a SGLT2 inhibitor (empagliflozin), by regulating proteins of mitochondrial fission and fusion [74]. Mitochondrial fusion and fission are essential for mitochondrial maintenance but also for mitochondrial DNA (mtDNA) integrity, which can participate in the regulation of cell survival, metabolic processes, and redox-sensitive signals [75]. In addition, high glucose conditions increase superoxide formation in mitochondria, which combined with NO (released by endothelial cells under stress) generates peroxynitrite [76]. Peroxynitrite is a cytotoxic oxidant that induces eNOS uncoupling, restricting vascular relaxation, and promoting diabetic vascular complications. As mentioned before, NOX enzymes have been shown to mediate ROS production. NOX-derived ROS regulate many physiological mechanisms of the kidney, including tubular-glomerular feedback, glucose metabolism and transport, kidney hemodynamics, and electrolyte transport [77]. NOX4 is the major source

of ROS production in podocytes, and it is upregulated under high-glucose conditions [78]. For this reason, NOX4 has been largely linked to DKD [53,79]. Induction of NOX4 in the Akita model of DKD results in the emergence of the typical structural changes seen in the diabetic kidney: glomerular hyperfiltration, mesangial matrix accumulation, glomerular membrane thickening, albuminuria, and podocyte loss [53]. In addition, metabolomic analysis performed on these mice revealed alterations in tubular energy metabolism with the accumulation Krebs cycle-related urinary metabolites [53]. The Krebs cycle is the central axis of energy production in mitochondria, thus, disproportion of the cycle substrates leads to mitochondrial dysfunction in advanced DKD [80]. Accordingly, a cross-sectional study in patients with and without DKD found that injured tubular cells showed several signs of mitochondrial damage, including (1) changes in mitochondrial dynamics, (2) decreased mtDNA copy number, and (3) increased release of mtDNA into the extracellular space [81]. Paradoxically, Cao et al. reported that plasma mtDNA content was decreased in patients with T2DM with clinically significant proteinuria (24 h urinary protein level >0.5 g) but not mild proteinuria (24 h urinary protein level ≤0.5 g), however, mtDNA abundance was increased in the urine of the former [82]. They propose that hyperglycemia reduces intracellular mtDNA and facilitates its extracellular release, so that circulating mtDNA may be filtered by the kidneys and end up in the urine [82]. In concordance, Kafaji et al. reported that a decreased mtDNA copy number in blood was associated with the severity and the presence of DKD [83]. Their group found out that mtDNA was lower in diabetic patients with macroalbuminuria than in patients with microalbuminuria or normalbuminuria. Therefore, altered mtDNA content can predict the occurrence of DKD and the oxidative stress environment [84]. Interestingly, release of mtDNA under cellular stress has been previously considered as a damage-associated molecular pattern (DAMP) [85]. mtDNA can be recognized by Toll-like receptor 9 (TLR9) and cystosolic Cgas-stimulator of interferon genes (STING) and activate the inflammasome, leading to kidney inflammation and fibrosis [86]. In this line, recent evidence showed an increase in macrophage infiltration in a mouse model of type 2 diabetes [87] and in human progressive DKD [88], which correlates with the disease state and renal injury. Other than the direct effect of mtDNA as DAMP, oxidative stress can induce inflammation in the kidney by stimulating cytokine production [89]. ROS derivates, which participate in cellular signaling, can activate the transcription factors nuclear factor kappa B (NFκB) and activator protein-1 (AP-1), promoting the transcription of cytokines, growth factors, and extracellular membrane proteins [89] (Figure 2).

Adding the immune response to the theme above reaffirms that mitochondrial dysfunction and redox imbalance lead to the emergence of the main precipitating factors: altered metabolism, oxidative stress, and inflammation. These three elements combine and reciprocally feed back into each other. As in the chicken-and-egg dilemma, determining which of these events triggers kidney disease in the first place entails a great difficulty. However, it is clear that pharmacological approaches focused on blocking/stimulating the involved pathways mentioned in this section and targeting the SGLT2 transporter are of particular interest.

**Figure 2.** Mitochondrial dysfunction in DKD. The sodium/glucose co-transporter type 2 (SGLT2) is located in the apical membrane of the proximal tubular cells (PTC) of the kidney and is responsible for glucose and sodium uptake. Diabetic patients have overactivation of SGLT2 that alters glucose and sodium homeostasis, which directly affects mitochondrial function at different levels; imbalance of fission and fusion together with mitochondrial fragmentation (mitochondrial dynamics). Alteration and accumulation of Krebs cycle substrates, uncoupling of the oxidative phosphorylation chain (OXPHOS) with generation of ROS (e.g., formation of H2O2 by the enzyme superoxide dismutase (SOD)), leading to ATP depletion, oxidative stress, and a metabolic shift to oxygen-independent energy sources (mitochondrial bioenergetics). Mitochondrial DNA (mtDNA) is also damaged, leading to a reduction in mtDNA copy number and increased release of mtDNA into the cytosol, which activates the NLR family pyrin domain-containing 3 (NLRP3) inflammasome, and into the extracellular space, triggering the recruitment of immune cells and the onset of the inflammatory response (mitochondrial genetics). DKD: diabetic kidney disease; ROS: reactive oxygen species; ADP: adenosine diphosphate; ATP: adenosine triphosphate; NLR: NOD-like receptor.

#### **4. Antioxidant Properties of SGLT2 Inhibitors**

SGLT2 inhibitors (SGLT2is) are hypoglycemic drugs that target SGLT2 producing glycosuria. Consequently, SGLT2is decrease blood glucose in an insulin-independent manner, improving insulin resistance in diabetes [90,91]. To date, different SGLT2is, such as canagliflozin, dapagliflozin, ertugluflozin, and empagliflozin, have been approved to treat T2DM and DKD. Large clinical trials have demonstrated that these drugs (mainly canagliflozin, dapagliflozin, and empagliflozin) delay the progression of DKD on top of the standard of care with renin-angiotensin system (RAS) blockers [92–95]. The protective effect of the SGLT2is has been widely attributed to normalization of glycemia, natriuresis, body weight reduction, and decrease of intraglomerular pressure [63,96,97]. However, the use of SGLT2is has been associated with reduction of oxidative stress and inflammation as well (Table 2). In type 2 diabetic patients, twelve-week treatment with dapagliflozin significantly reduced the urinary excretion of 8-oxo-7,8-dihydro-2 -deoxyguanosine (8 oxo-dG), a marker of DNA oxidation [98]. In addition, empagliflozin treatment increased 2,2¢-azino-bis-(3-ethylbenzthiazoline-6-sulphonic acid) (ABTS) radical scavenging capacity (a measure of antioxidant capacity) in diabetic patients, suggesting a beneficial equilibrium between oxidant and antioxidant mechanisms despite empagliflozin also increasing serum levels of the thiobarbituric acid reactive substances (TBARS) and malondialdehyde (MDA). The authors suggest that the increased levels of TBARS and MDA (both indicators of lipid peroxidation) can be explained by the ketogenesis induced by SGLT2is that may contribute

to lipid peroxidation [99]. Contrarily, Nabrdalik-Le´sniak et al. have described that therapy of at least one month with either canagliflozin or empagliflozin decreased total antioxidant capacity (estimated by ABTS urine levels) in diabetic patients as compared to that of diabetic patients not treated with SGLT2is and healthy controls. However, the urinary activity of the antioxidant enzymes SOD and MnSOD was increased in diabetic patients treated with SGLT2is, suggesting that these drugs can also promote the activation of antioxidant mechanisms [100]. In this line, Iannantuoni et al. have described that 24-week long treatment with empagliflozin reduces superoxide production, increases glutathione content (an antioxidant), and enhances the expression of glutathione s-reductase and catalase in leukocytes of diabetic patients. In this study, the authors found that empagliflozin increased the serum levels of the anti-inflammatory interleukin 10 (IL-10) and decreased C reactive protein and myeloperoxidase serum levels suggesting, an antioxidant and antiinflammatory role of SGLT2is [101]. Similarly, CD34+ endothelial progenitors (markers of endothelial function) isolated from peripheral blood of diabetic patients treated with canagliflozin showed increased SOD2, catalase, and glutathione peroxidase gene expression, which suggests that SGLT2is induce a beneficial antioxidant profile on these cells [102].


**Table 2.** Summary of the antioxidant effects of SGLT2is observed in clinical and preclinical studies.


**Table 2.** *Cont*.

In diabetic experimental models, similar results have been found (Table 2). The administration of SGLT2is in rat and mice models decreases glycemia [106], and in experimental models of diabetic nephropathy, these drugs have demonstrated cardiorenal protection [103,107–110]. In a high-fat diet (HFD)-induced obesity mouse model, 8 weeks of

empagliflozin administration improved cardiac dysfunction via Sestrin2-mediated AMPKmammalian target of rapamycin (mTOR) pathways by maintaining the redox stability [104]. Similarly, canagliflozin administration for 12 weeks improved atherosclerosis lesions and endothelial dysfunction in streptozotocin-induced diabetic apolipoprotein E-deficient (ApoE−/−) mice. These beneficial effects could be in part ascribed to the reduction of oxidative stress, as canagliflozin decreased the expression of NOX2 and p22phox (both NADPH oxidase subunits) and the urinary excretion of 8-OHdG [105]. Regarding the kidney, Kamezaki et al. have demonstrated that ipragliflozin reduced renal NOX4 expression and oxidative stress in both the tubular epithelium and glomerular podocytes using a model of early diabetic nephropathy (db/db mice) [103]. In obese prediabetic rats, dapagliflozin suppressed renal gluconeogenesis and oxidative stress [111]. Thus, in a diabetic milieu, SGLT2is seem to promote antioxidant effects that could participate in the cardiorenal protective mechanisms of these drugs. Interestingly, blockade of SGLT2 (with SGLT2is or by silencing with siRNAs) ameliorates oxidative stress in endothelial [112,113], tubular [114–116], and mesangial cells [117] in culture. This suggests that these drugs have direct antioxidant effects on kidney cells and possibly on other organs (even off-target [113]).

The beneficial antioxidant effects of SGLT2is observed in DKD can in part be attributed to improvement of glycemia and blood pressure control, as other hypoglycemic drugs also reduce oxidative stress [98]. However, other protective mechanisms cannot be ruled out. Recent data obtained from EMPEROR-Reduced and DAPA-CKD studies have confirmed the efficacy of SGLT2is in patients with heart failure or CKD without a history of diabetes [118]. In the EMPEROR-Reduced double-blind trial, patients receiving empagliflozin had a lower risk of cardiovascular death or hospitalization and lesser GFR decline over time, regardless of the presence or absence of diabetes [119]. In addition, among CKD patients participating in the DAPA-CKD clinical trial, those treated with dapagliflozin had better stabilization of GFR and a lower risk of progression to end-stage renal disease or death, again, independent of the presence of diabetes [120]. Although the renal antioxidant effect of the SGLT2is in nondiabetic human CKD needs to be assessed in future research, studies in nondiabetic experimental models of kidney damage have already shown that these drugs protect the kidney by decreasing oxidative stress [121,122]. Furthermore, SGLT2is also have antioxidant properties in organs other than the kidney. Olgar et al. demonstrated that the increase of SGLT2 activity in cardiomyocytes of old Wistar male rats (24-month age) triggers ROS accumulation, alters mitochondrial dynamics, and promotes the loss of mitochondrial membrane potential. Interestingly, these cardiac events were mitigated by dapagliflozin treatment [123].

It has also been demonstrated that targeting SGLT2 with canagliflozin reduces cancer cell proliferation by inhibiting mitochondrial complex I-mediated respiration [124]. In the same line, dapagliflozin and canagliflozin produced cell growth arrest in breast cancer cells both in vitro and in vivo via the AMPK-mTOR pathway. The activation of the AMPK-mTOR pathway basically inhibits energy-consuming processes such as the mitochondrial phosphorylation and increases catabolic processes to restore cellular energy homeostasis [125]. Thus, as SGLT2is have antioxidant effects in nondiabetic CKD and other pathologies not related to the kidney, it is possible that the antioxidant properties of SGLT2is observed in DKD patients happen due to a combination of several factors: (1) glycemia and blood pressure control, (2) reduction of glucose concentration in the target cells, and (3) activation of pathways that improve glucose utilization. Further research is needed to ascertain the exact mechanism by which SGLT2is decrease ROS, but it is a fact that these drugs promote the restoration of the oxidant/antioxidant balance in the diabetic kidney as well as in other scenarios.

#### **5. Conclusions**

In conclusion, oxidative stress has an important role in DKD onset. High levels of glucose together with increased ROS production overactivate the SGLT2 transporter in tubular cells, which, in turn, exacerbates oxidative stress. The use of SGLT2is has demonstrated clear cardiovascular benefits which may be in part ascribed by a beneficial balance between oxidant and antioxidant pathways.

**Author Contributions:** Writing—original draft preparation, C.L.-C., M.M.-V.d.B. and A.V.; writing review and editing, C.J.-C. and M.J.S.; figures, C.L.-C. and M.M.-V.d.B. All authors have read and agreed to the published version of the manuscript.

**Funding:** The authors are current recipients of grants from FONDO DE INVESTIGACIÓN SANITARIA-FEDER, ISCIII (PI17/00257 and RICORS RD21/0005/0016), and Fundació la Marató de TV3 (421/C/2020, 759/U/2020 and 215/C/2021).

**Conflicts of Interest:** A.V. reports personal fees from MUNDIPHARMA, and nonfinancial support from MUNDIPHARMA, SANOFI, and NOVONORDISK outside this. C.J.C. declares travel support and a research grant from TRAVERE THERAPEUTICS outside this work. M.J.S. reports grants from BOEHRINGER, personal fees from NOVONORDISK, JANSSEN, BOHERINGER ASTRAZENECA, FRESENIUS, MUNDIPHARMA, PFIZER, ICU, GE Healthcare BAYER, TRAVERE THERAPEUTICS, and VIFOR, and nonfinancial support from ELI LILLY and ESTEVE, outside this work.
