Renoprotective effects of sodium-glucose co-transport 2 inhibitors (SGLT2i) represent a hallmark success after decades searching for novel diabetic kidney disease (DKD) treatments. In 2019, results from the Canagliflozin and Renal Events in Diabetes with Established Nephropathy Clinical Evaluation (CREDENCE) showed that SGLT2i treatment significantly reduced progression to end-stage kidney disease (ESKD), doubling of creatinine, as well as the incidence of several cardiovascular outcomes [1]. CREDENCE cemented SGLT2i as pillars, along with angiotensin-converting enzyme inhibitors (ACEi) and angiotensin II receptor blockers (ARBs), of pharmacological DKD therapy. Despite ever-growing use, controversy persists regarding how blocking one transporter produces such remarkable clinical improvements.

Dominant DKD pathogenesis theories revolve around podocyte injury, stemming from observations that early DKD manifestations include albuminuria. Most benefits of ACEi or ARB therapy are attributed to the correction of intraglomerular pressures and single nephron GFR (SNGFR). Glomerular filtration rate (GFR) is tightly controlled by two mechanisms: glomerular vascular tone (afferent and efferent arterioles) and urinary ion concentration (i.e., Cl−). The macula densa, via Na–K–Cl co-transporter 2 (NKCC2), responds to increased Cl− delivery with adenosine-mediated afferent arteriole vasoconstriction and decreased SNGFR. This phenomenon, known as tubuloglomerular feedback (TG), helps maintain renal physiological homeostasis, and its dysregulation prominently mediates DKD. Hyperglycemia increases proximal Na+/Cl− reabsorption, decreasing concentrations sensed by macula densa cells and preventing appropriate SNGFR regulation. SGLT2i prevents this phenomenon, increases distal tubule Cl− delivery, and, ostensibly, corrects TG/intraglomerular hemodynamics similarly to ACEi/ARB therapy. Many believe this mechanism predominantly explains clinical trial findings. However, by using measured GFR (mGFR) in type 2 DM patients on dapagliflozin, van Bommel et al. [2] demonstrated that dapagliflozin does not cause afferent arteriole constriction but, rather, post-glomerular vasodilatation. Whether these conflicting results represent a “class-effect” or differences among individual SGLT2i would require further studies to elucidate.

Aside from correcting TG feedback issues, how do SGLT2i protect from DKD progression? Growing evidence points to improved tubular homeostasis behind SGLT2i effects on DKD pathogenesis. SGLT2 is primarily expressed on brush borders of S1 proximal tubule segments, and its expression is increased by 40–80% with persistent glycosuria. Induced apical membrane SGLT2 expression maximizes glucose reabsorption, but also increases Na+–K+-ATPase activity and oxygen consumption—causing hypoxia and tubular injury. Blocking SGLT2 may thereby improve proximal tubule oxygenation. Secondly, hyperglycemia itself also triggers tubular growth in response to various stimuli during the early hyperplasic to late hypertrophic phase, reflected by the enlarged size of kidneys commonly seen in diabetic patients (comprehensive review in [3]). After empagliflozin treatment in Akita mice, however, kidney growth was only partially attenuated, suggesting that tubular growth is regulated by not only luminal glucose delivery but also by other mechanisms. For instance, forms of familial renal hyperglycosuria where SGLT2 is missing do not develop progressive renal disease [4]. Lastly, a notion that increasing tubular senescence in both human and murine models of DKD might lead to maladaptive repair in chronic hyperglycemia milieu, but whether SGLT2i would modify the aging process requires further studies.

In a recent review [5], Packer suggests that restoration of autophagy represents an additional mechanism contributing to the renoprotective effects of SGLT2i. Autophagy, an intracellular catabolic process, maintains energy resources during nutritional deprivation by selectively sequestering and degrading organelles. This intricate process is predominantly regulated by two ubiquitously expressed proteins, mammalian target of rapamycin (mTOR) and AMP-activated protein kinase (AMPK). The mTOR complex 1 (mTOR1), a major sensor of intracellular amino acid, promotes protein synthesis (anabolism) and inhibits autophagy. Conversely, energy stress and an increased AMP/ATP ratio enhance AMPK activity as well as autophagy. AMPK-triggered autophagy regenerates the necessary nutrients from damaged/excess organelles. Increasing evidence indicates intact autophagy is necessary to slow DKD. For instance, podocyte-specific deletion of Atg5, a crucial autophagy protein, accelerated podocyte loss in murine models of both type 1 and 2 DM. Given its glycosuric effects, SGLT2i remarkably alters cellular metabolism in proximal tubules and, possibly, podocytes [6], toward a nutritional deprivation state, and thereby activates the AMPK pathway. Moreover, recent data support that SGLT2i augment ketolysis and inhibit mTORC1—an effect that ostensibly increases autophagy [7].

Another important modulator of autophagy and glucose homeostasis is sirtuin-1 (SIRT1). SIRT1 downregulation occurs during DKD in both proximal tubules and podocytes, likely due to increased insulin resistance, and has been implicated in diminished autophagy and mitochondrial injuries. An indirect effect of tubular SIRT1 expression on podocytes was revealed by Hasegawa et al. [8], who used genetically modified murine models and demonstrated that overexpression of SIRT1 mitigated albuminuria in DKD (db/db) mice. By contrast, deleting SIRT1 in proximal tubules exacerbated glomerular dysfunction from increased expression of claudin-1, a protein resulting in aberrant tight junction by interacting with podocin and nephrin. Similar to AMPK, SGLT2i may upregulate SIRT1 expression in proximal tubules due to glucose wasting, and therefore stabilize slit diaphragm between podocytes in addition to enhanced autophagy influx in proximal tubules.

Proximal tubules are the main focus of putative SGLT2i effects, but evidence suggests that other areas are targeted. For instance, Cassis et al. [9] reported an upregulation of podocyte SGLT2 expression in a murine model of protein-overloaded proteinuria induced by bovine serum albumin (BSA), and treatment with SGLT2i dapagliflozin significantly improved albuminuria via restoration of cytoskeletal remodeling in the podocytes. More recently, abnormal tubular cell glucose metabolism has been implicated in acute kidney injury (AKI)-associated mortality, further highlighting the pivotal role of glucose homeostasis particularly in tubules. In conclusion, the extraordinary clinical success of SGLT2i, while improving outcomes of DKD, paved the way for a better understanding of DKD pathophysiology, possibly leading to the identification of new therapeutic targets for this and other progressive kidney diseases.