Diabetic nephropathy (DN), a distinct manifestation of diabetic kidney disease, affects approximately 30% of patients with diabetes. While most attention has been focused on glomerular changes related to DN, there is growing evidence that tubulopathy is a key feature in the pathogenesis of this disease. The renal proximal tubule cells (RPTCs) are particularly sensitive to the deleterious effect of chronic hyperglycemia. However, the cellular changes that control the dysfunction of the RPTCs are not fully understood. Controlling glucose reabsorption in the proximal tubules via inhibition of glucose transporters (GLUT) has emerged as a promising therapeutic in ameliorating DN. Overactivation of the renal endocannabinoid (eCB) system via the cannabinoid-1 receptor (CB1R) contributes to the development of DN, and its blockade by globally acting or peripherally restricted CB1R antagonists has been shown to ameliorate renal dysfunction in different murine models for diabetes. Recently, we have utilized various pharmacological and genetic tools to show that the eCB/CB1R system contributes to the development of DN via regulating the expression, translocation, and activity of the facilitative GLUT2 located in the RPTCs. These findings have the potential to be translated into therapy, and support the rationale for the preclinical development of novel renal-specific CB1R and/or GLUT2 inhibitors for the treatment of DN.
Glucose reabsorption by renal proximal tubule cells (RPTCs) is usually mediated by 2 central glucose transporters (GLUT): under normoglycemic conditions, the low-affinity high-capacity sodium-dependent glucose cotransporter 2 (SGLT2) is responsible for reabsorbing up to 90% of filtered glucose across the apical/brush border membrane (BBM) of the RPTCs, and the facilitative GLUT2 is responsible for diffusing the reabsorbed or newly synthesized glucose from the tubular cell back to the bloodstream across the basolateral membrane . While the protein expression levels of SGLT2 in the RPTCs are reported to be elevated in patients with type-2 diabetes (T2DM) [2, 3], others have documented a slight opposite trend in its mRNA expression levels in different sets of patients with T2DM and preserved kidney function [4, 5]. Reports have been also quite controversial when exploring SGLT2 expression pattern in different murine models for T1DM and T2DM. For example, SGLT2 has been shown to be upregulated [6, 7], downregulated [8-10], or unchanged [11, 12] under different diabetic paradigms. On the other hand, the protein expression of GLUT2 in RPTCs is dramatically increased in diabetic humans  as well as in murine models for both diabetes and obesity [12-14]. Additionally, a shift in its localization from the RPTC’s basolateral membrane to the BBM, contributing to increased glucose reabsorption during hyperglycemia, was also reported [14-16]. In fact, plasma or luminal glucose concentrations have been shown to regulate GLUT2 expression and/or translocation [12, 17]. This, in turn, may lead to increased glucose concentration within the proximal tubules, which is suggested to enhance tubular damage.
The recent clinical approval of SGLT2 inhibitors for treating hyperglycemia in diabetic patients, and the ongoing clinical trials to test their efficacy in treating diabetic nephropathy (DN) emphasize the therapeutic potential of inhibiting GLUTs in the kidney. SGLT2 inhibitors block the SGLT2 protein, resulting in glucose excretion and lower blood glucose levels. Although SGLT2 inhibitors were found to be particularly helpful in obese and diabetic patients, a few adverse effects, such as diabetic ketoacidosis  as well as genital and urinary tract infection  have been reported with the chronic use of SGLT2 inhibitors. In addition, it has been implicated in increasing the risk for acute kidney injury (AKI) owing to the high, unreabsorbed sodium and glucose levels , and decreased estimated glomerular filtration rate . Although the EMPA-REG and CANVAS trials did not report an increase in AKI [22, 23] in T2DM patients, volume depletion was more common with canagliflozin versus placebo . In fact, post-marketing reports of AKI with SGLT2 inhibitors urged the FDA to revise the labels of the SGLT2 inhibitors in order to minimize the risk of AKI [21, 24]. Careful consideration should also be noted when prescribing SGLT2 inhibitors for patients with low blood pressure or for individuals with reduced kidney function, since they may acutely reduce their glomerular hyperfiltration , and lead to hyperosmolarity and dehydration owing to their diuretic properties . Therefore, renal GLUT2 blockade may be considered a promising strategy for protecting the kidney from the development of DN in these patients, and as a complementary treatment to SGLT2 inhibitors.
The endocannabinoid system, composed of lipid-based endogenous ligands (anandamide and 2-arachidonoylglycerol) that bind to 2 G-protein-coupled receptors (GPCRs; known as cannabinoid-1 receptor [CB1R] and CB2R), is an important biological system that has been implicated in various physiological and pathological processes both in the central nervous system and periphery. CB1R is the most abundant GPCR in the mammalian brain; however, it is also present in multiple peripheral tissues at much lower, but functionally relevant densities, such as the kidney (reviewed in ). Using murine models for both T1DM and T2DM, several groups have tremendously contributed to our understanding of the role of the endocannabinoid/CB1R system in the development of DN. Under both conditions, the protein and mRNA expression levels of CB1R were found to be upregulated in podocytes, mesangial cells, and RPTCs, and its pharmacological inhibition by globally acting or peripherally restricted antagonists ameliorates diabetes-induced renal dysfunction, inflammation, and fibrosis [28-35].
While SGLT2 inhibitors are still under clinical evaluation for the treatment of DN, the current and most effective therapeutic interventions for diabetic kidney disease are blockers of the renin-angiotensin system (RAS) . Interestingly, it was suggested that CB1R has the ability to form a heterodimer with type 1 angiotensin II receptor (AT1R) and amplify its activity . Moreover, administering Losartan, an AT1R antagonist, was found to attenuate DN by downregulating the expression of CB1R in podocytes . In addition, CB1R blockade inhibits the effect of angiotensin II in cultured human podocytes similarly to losartan, emphasizing the obligatory role of CB1R on AT1R signaling in podocytes ; however, the influence of CB1R antagonism on RAS-mediated hypertension is yet to be investigated. As for SGLT2, we have recently shown that either peripheral blockade or genetic deletion of CB1R specifically in the RPTCs did not affect the downregulated renal expression of SGLT2 in 2 type 1 diabetic mouse models . On the other hand, we found that CB1R activation or blockade modulates the expression and dynamic membrane translocation of GLUT2 in RPTCs . This, in turn, affects the susceptibility of the diabetic mice to develop DN. First, we reconfirmed the observation that GLUT2 expression is upregulated and recruited to the BBM of the RPTCs during hyperglycemia [14, 15], and then we defined, for the first time, that CB1R is involved in these processes. Next, using a novel mouse strain that lacks CB1R from RPTCs , we showed that the renal downregulation of GLUT2 in these mice results in preserved renal function (with little or no signs of renal fibrosis or inflammation) during hyperglycemia. Moreover, knocking down GLUT2 from RPTCs led to the cells’ reduced ability to reabsorb glucose, and it alleviates the deleterious effects of hyperglycemia on these cells. Lastly, we found that CB1R regulates GLUT2 expression, membrane translocation, and activity by a signaling mechanism that involves elevating cytosolic Ca2+ levels and activating the upstream modulator of GLUT2, protein kinase C beta-1  (Fig. 1). The renal protective effect of tubular GLUT2 downregulation, mediated by CB1R, does not affect blood glucose levels in these mouse models, and therefore, GLUT2 does not possess the risk for the development of ketoacidosis or osmotic diuresis, which in turn, may lead to AKI, as it has been suggested with the use of SGLT2 inhibitors and discussed above. However, the effect of downregulating GLUT2 in RPTCs in models for T2DM has not yet been tested, and therefore, we cannot postulate what would be the effect under this condition.
In conclusion, CB1R seems to play a fundamental role in the development of DN via regulating the RAS system, activated during hypertension, and by modulating glucose reabsorption during hyperglycemia. Thus, its inhibition in the kidney, particularly in podocytes and/or RPTCs, has a great potential in treating DN. We also suggest that based on our recent findings, modulating/blocking GLUT2 in the apical membrane of the RPTCs should be further tested preclinically for ameliorating the deleterious effects of hyperglycemia on the kidney.
This work was supported by an ERC-2015-StG grant (#676841) as well as by the Israel Science Foundation (ISF) grant (#158/18) to Joseph Tam.
The authors declare that they have no conflicts of interest to disclose.
Contribution from the CME Course of the DIABESITY Working Group of the ERA-EDTA, Lisbon, November 24–25, 2017.