Diabetic nephropathy is the leading cause of end-stage renal disease in humans in the Western world. The recent development of Na+-glucose cotransporter 2 (SGLT2) inhibitors offers a new antidiabetic therapy via enhanced glucose excretion. Whether this strategy exerts beneficial effects on the development of type 2 diabetic nephropathy is still largely unclear. We investigated the effects of the specific SGLT2 inhibitor empagliflozin in BTBR.Cg-Lep<ob>/WiscJ (BTBR ob/ob) mice, which spontaneously develop type 2 diabetic nephropathy. In the first experiment, BTBR ob/ob mice received either a diet containing 300 ppm empagliflozin or equicaloric placebo chow for 12 wk. In the second experiment, BTBR ob/ob mice received 1 μg·kg body wt−1·day−1 ANG II to induce arterial hypertension and were separated into the same two diet groups for 6 wk. In both experiments, empagliflozin treatment enhanced glucosuria, thereby lowering blood glucose. Independently of hypertension, empagliflozin reduced albuminuria in diabetic mice. However, empagliflozin treatment affected diabetes-related glomerular hypertrophy, markers of renal inflammation, and mesangial matrix expansion only in BTBR ob/ob mice without hypertension. In summary, empagliflozin demonstrated significant antihyperglycemic effects, differentially ameliorating early features of diabetic nephropathy in BTBR ob/ob mice with and without hypertension.
- diabetes mellitus type 2
- diabetic nephropathy
- renal growth
- Na+-glucose cotransporters
diabetic nephropathy is the leading cause of end-stage renal disease in humans in the Western world (11). The mechanisms leading to the pathological changes seen in patients with diabetic nephropathy are still not fully understood. However, well-controlled blood glucose levels are essential for the prevention of diabetes-induced end-organ damage.
The transport of glucose across cell membranes is accomplished by two gene families: facilitative glucose transporters (GLUTs) and Na+-glucose cotransporters (SGLTs). In the kidney, high-capacity SGLT2 (SLC5A2) is responsible for the majority of glucose reabsorption in the early proximal tubule, while low-capacity SGLT1 (SLC5A1) reabsorbs the remaining glucose further distal in the proximal tubule (26, 27). Investigations of transgenic mice have shown that the lack of SGLT1 only has minor effects on glucose reabsorption (15), whereas SGLT2 deficiency results in pronounced glucosuria (33). These experimental findings correspond with clinical data, in which patients with mutations in the Sglt2 gene have sustained renal glucosuria, whereas mutations in the Sglt1 gene have little or no impact on glucosuria (28). Recently, the development of specific and potent SGLT2 inhibitors offers a new antidiabetic therapy via enhanced glucose excretion (23).
Mice carrying the ob/ob mutation, an inactivating mutation in the leptin gene, develop hyperglycemia, hypercholesterolemia, elevated triglycerides, insulin resistance, and subsequently diabetes mellitus type 2, depending on the genetic background (9, 10). In contrast to other background strains, BTBR.Cg-Lep<ob>/WiscJ (BTBR ob/ob) mice have been recently described as an excellent animal model for diabetic nephropathy (18, 24). BTBR ob/ob mice develop progressive proteinuria and a renal histomorphological picture that is quite similar to that seen in human patients with advanced diabetic nephropathy (18). In contrast to the majority of patients with diabetic nephropathy, BTBR ob/ob mice do not develop hypertension but are rather slightly hypotensive (18). Therefore, besides investigation of usual diabetic BTBR ob/ob mice, we tried to further aggravate this diabetic nephropathy model by the induction of hypertension via chronic ANG II infusion in these mice.
To elucidate whether SGLT2 inhibition with empagliflozin (BI-10773, Boehringer Ingelheim) has beneficial effects on the development and progression of diabetic nephropathy in type 2 diabetes, we investigated the effects of empagliflozin compared with placebo in diabetic BTBR ob/ob mice without or with ANG II-mediated hypertension.
MATERIALS AND METHODS
Eight-week-old female BTBR ob/ob (18, 24) mice and sex- and age-matched wild-type (WT) control mice were used in the experiments. Animals of both genotypes were littermates from heterozygous parents. Genotypes were verified with PCR using the following primers: LEPWT, forward 5′-AATGACCTGGAGAATCTCC-3′; LEPOB, reverse 5′-GCAGATGGAGGAGGTCTCA-3′; and RFLP, forward 5′-TGAGTTTGTCCAAGATGGACC-3′ and reverse 5′-GCCATCCAGGCTCTCTGG-3′ (13). Animals were housed at constant humidity (60 ± 5%) and temperature (24 ± 1°C) and with a 12:12-h light-dark cycle (6 AM to 6 PM light). Mice had access to water ad libitum.
All animal experiments were performed in accordance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals (NIH Pub. No. 85-23, Revised 1996) and the Federal Law on the Use of Experimental Animals in Germany and were approved by the local authorities.
A group BTBR ob/ob mice (n = 12) received a diet containing 300 ppm empagliflozin (kindly provided by Boehringer Ingelheim) for 12 wk. A second group of BTBR ob/ob mice (n = 12) received equicaloric placebo chow for 12 wk. WT mice (n = 11) received a placebo diet for 12 wk and served as baseline controls. Blood glucose levels were measured before and on weeks 3, 6, and 12 after the start of treatment with an Accu Check Aviva glucose meter (Roche Diagnostics, Mannheim, Germany).
Induction of hypertension.
To determine the effects of empagliflozin treatment in hypertensive type 2 diabetes, BTBR ob/ob mice (n = 16) received Alzet osmotic minipumps subcutaneously, which released 1 μg·kg body wt−1·day−1 ANG II. Three days after implantation, mice were divided equally into the same diet groups for 6 wk. Blood glucose levels were measured before and on weeks 3 and 6 after the start of treatment.
Mice were placed for 24 h in metabolic cages and urine was collected and stored at −80°C for further analysis. Urinary creatinine, glucose, and Na+ concentrations were measured in the Clinical Chemistry Department of the University Hospital Dresden using standard procedures. Urinary albumin concentrations were determined using a commercially available ELISA kit (Bethyl Laboratories, Montgomery, TX) according to the manufacturer's protocol.
Blood pressure measurement.
Under light isoflurane (0.5%) anesthesia, blood pressure was recorded by the tail-cuff method using BP Recorder 58500 (Ugo Basile Srl, Comerio, Italy).
Renal survival biopsies for mRNA quantification were taken from mice in the normotensive groups 6 wk after the start of treatment as previously described (19). At the end of the experiment, mice were terminally anesthetized with isoflurane. Blood was collected from the vena cava and split into tubes containing either Li-heparin or Na-EDTA. Plasma was generated by centrifugation (10,000 rpm at 4°C for 15 min). Plasma was snap frozen and stored at −80°C. Mice were perfused with 0.9% NaCl solution, and kidneys were harvested, decapsulated, cut in half, and either fixed for paraffin embedding and cryosectioning or were snap frozen and stored at −80°C for protein and mRNA expression experiments.
Plasma (Li-heparin) Na+ concentrations were measured in the Clinical Chemistry Department using standard procedures. Plasma (Na-EDTA) creatinine levels were determined using an enzymatic creatinine assay (Crystal Chem, Downers Grove, IL) according to the manufacturer's instructions.
Immunohistochemistry and kidney morphology.
Harvested kidneys were fixed in Zn fixative overnight and were further processed as previously described (12). Paraffin-embedded kidneys were cut into 2-μm-thick sections. Sections were stained with periodic acid-Schiff (PAS) reagent, and the mesangial index fraction and glomerular hypertrophy were assessed in a minimum of 30 glomeruli/mouse similarly to previously described methods (29). In brief, the glomerular tuft was highlighted using the polygon selection tool in ImageJ software (NIH; http://rsb.info.nih.gov/ij/), and the glomerular area was calculated. The PAS-positive mesangial matrix area was assessed accordingly and quantified as the percent fraction of the glomerular tuft. For immunohistochemical detection of SGLT1 and SGLT2, renal biopsies were processed as previously described (5, 26) and stained using antibodies against SGLT1 (5) or SGLT2 (27) (both antibodies were kind gifts of Dr. H. Koepsell).
Reverse transcription and quantitative real-time PCR.
Whole kidney RNA was prepared using a RNeasy Protect Mini Kit (Qiagen, Hilden, Germany). cDNA was synthesized, and gene-specific primers (Table 1) were used with SYBR Green PCR Master Mix (Bioline, Luckenwalde, Germany) for quantification. Gene expression was normalized against the housekeeping gene L32.
Western blot analysis.
Protein samples were prepared from the whole kidney using a Precelys homogenizator (Peqlab, Erlangen, Germany), and membrane fractions were prepared as previously described (26, 27). Membrane samples were separated by SDS-PAGE and transferred to a polyvinylidene difluoride membrane. SGLT1 and SGLT2 antibody incubation was performed as previously described (26, 27). Band intensities were quantified using ImageJ software and normalized to β-actin.
Data are expressed as means ± SE and were analyzed by an unpaired Student's t-test. Significance was considered from a value of P < 0.05.
Expression of glucose transporters in the kidney.
Immunofluorescent staining of renal sections from WT mice using SGLT1-specific antibody showed strong positive staining of the tubular brush-border membrane, especially in S3 segments of proximal tubules (Fig. 1A). SGLT2 visualization demonstrated specific and selective staining of the apical brush-border membrane of early proximal convoluted tubules (Fig. 1B).
To analyze the effects of diabetes and empagliflozin treatment on protein expression of SGLT1 and SGLT2, we quantified their expression in Western blot experiments. Renal SGLT1 protein expression was increased by 65% in diabetic BTBR ob/ob mice compared with WT mice (P < 0.01 vs. WT mice; Fig. 1, C and D). This increase was not influenced by empagliflozin treatment (+89%, P = 0.44 vs. placebo) in diabetic mice. SGLT2 protein expression was equally increased in diabetic mice with and without empagliflozin treatment but due to higher variability did not reach a significant level (placebo: +64% and empagliflozin: +56%, Fig. 1, C and E).
In contrast, renal SGLT1 mRNA expression was reduced by ∼40% in diabetic BTBR ob/ob mice compared with WT mice (Fig. 1F). Treatment with empagliflozin did not affect SGLT1 mRNA expression in diabetic mice. In contrast to the previously described downregulation of SGLT2 in streptozotocin (STZ)-induced diabetic mice (8, 34), renal mRNA expression of SGLT2 was unchanged in BTBR ob/ob mice (Fig. 1G). Empagliflozin treatment did not regulate renal mRNA expression. To investigate the expression of other glucose transporters in kidneys of BTBR ob/ob mice, we quantified GLUT1 (Fig. 1H) and GLUT2 mRNA (Fig. 1I), but neither diabetic state in BTBR ob/ob mice compared with WT mice nor treatment with empagliflozin influenced the expression levels of the two transport proteins.
Basal characteristics of empagliflozin-treated BTBR ob/ob mice.
WT mice at the age of 8 wk weighed 26 ± 0.5 g and gained 6 ± 0.6 g over the 12 wk of the experiment (Table 2). Obese BTBR ob/ob mice on both diets had a significantly higher baseline body weight (placebo: 47 ± 1.1 g and empagliflozin: 47 ± 1.3 g). Independently of the drug therapy, BTBR ob/ob mice also gained more weight in the 12 wk than WT mice (placebo: 19 ± 2.3 g and empagliflozin: 19 ± 4.4 g) in the same experimental period.
BTBR ob/ob mice already have increased urinary glucose excretion compared with WT mice (2,182 ± 972 vs. 8 ± 3 μmol/mg creatinine, P < 0.001; Table 2). Already 3 wk after the start of empagliflozin treatment, glucosuria was increased more than threefold in BTBR ob/ob mice, and this increase lasted for the whole experimental time (6,987 ± 987 vs. 2,182 ± 972 μmol/mg creatinine, P < 0.001; Table 2). The empagliflozin-induced augmented glucosuria led to a pronounced drop of blood glucose levels in diabetic mice (381 ± 29 vs. 230 ± 13 mg/dl, P < 0.001) 3 wk after the start of empagliflozin treatment (Table 2). The blood glucose-lowering effect was still present at the end of the experiment.
At the age of 8 wk, BTBR ob/ob mice already have a lower blood pressure than WT mice (84 ± 1.1 vs. 116 ± 2.1 mmHg, P < 0.001; Table 2). Empagliflozin treatment for 12 wk had no significant impact on blood pressure in BTBR ob/ob mice (81 ± 1.4 mmHg).
Renal effects of empagliflozin treatment.
To analyze renal function in diabetic BTBR ob/ob mice, creatinine clearance was determined as an indirect measure of the glomerular filtration rate. Diabetic placebo-treated BTBR ob/ob mice at the age of 20 wk have an increased creatinine clearance compared with WT mice (565 ± 123.7 vs. 254 ± 25.7 μl/min, P < 0.01; Table 2), whereas creatinine clearance of diabetic empagliflozin-treated BTBR ob/ob mice was not different from WT mice. Nevertheless, empagliflozin treatment only tended to decrease creatinine clearance compared with placebo-treated diabetic mice (412 ± 64.1 μl/min, P = 0.11). Increased creatinine clearance in diabetic BTBR ob/ob mice was associated with significant glomerular hypertrophy (as measured using PAS-stained kidney sections) compared with WT mice (3,646 ± 59.8 vs. 2,486 ± 61.7 μm2, P < 0.001; Fig. 2, A–C). Empagliflozin treatment of BTBR ob/ob mice led to a significant reduction of the glomerular hypertrophy (3,392 ± 100.9 μm2, P < 0.05; Fig. 2, A and D). Hyperfiltration and glomerular hypertrophy are associated with pronounced albuminuria in human diabetic nephropathy, as demonstrated in BTBR ob/ob mice (2,291 ± 524.4 vs. 213 ± 33.1 μg/mg creatinine, P < 0.001; Fig. 2E). Albuminuria was markedly decreased in empagliflozin-treated BTBR ob/ob mice (834 ± 134.9 μg/mg creatinine, P < 0.05).
Glomerular matrix expansion was quantified by the mesangial index fraction of the glomerular tuft. Diabetic BTBR ob/ob showed a significantly increased mesangial index fraction compared with WT mice (Fig. 3A). Similar to the effects on albuminuria and glomerular hypertrophy, empagliflozin treatment was able to ameliorate the increased glomerular matrix expansion. Furthermore, we analyzed the mRNA expression of the inflammatory markers monocyte chemoattractant protein (MCP)-1 [chemokine (C-C motif) ligand 2 (Ccl2)], regulated on activation normal T cell expressed and secreted (RANTES) (Ccl5), and IL-6 in kidney samples. In 14-wk-old diabetic BTBR ob/ob mice (6 wk of placebo diet), MCP-1 (∼1.4-fold, P < 0.05 vs. WT mice; Fig. 3B), RANTES (∼2.6-fold, P < 0.05 vs. WT mice; Fig. 3C), and IL-6 (∼2.4-fold, P < 0.05 vs. WT mice; Fig. 3D) mRNA expression were significantly increased. None of these inflammatory markers was significantly affected after 6 wk of empagliflozin treatment. However, 12 wk of empagliflozin treatment reduced renal mRNA expression of MCP-1 by ∼75% (P < 0.05 vs. placebo; Fig. 3B, inset), RANTES by ∼55% (P < 0.05 vs. placebo; Fig. 3C, inset), and IL-6 by 52% (P < 0.05 vs. placebo; Fig. 3D, inset).
Characterization of hypertensive diabetic mice.
Chronic ANG II infusion using Alzet minipumps over 6 wk induced profound and lasting hypertension in diabetic BTBR ob/ob mice (Table 3), reaching blood pressure values of almost 150 mmHg. Similar to nonhypertensive BTBR ob/ob mice, ANG II-infused mice on both diets were obese (placebo: 52 ± 1.3 g and empagliflozin: 52 ± 1.6 g; Table 3) and, independently of the drug therapy, gained weight fast (placebo: 9 ± 1.4 g and empagliflozin: 12 ± 3.5 g). Chronic ANG II infusion led to more pronounced glucosuria in BTBR ob/ob mice before the treatment started (hypertensive 4,814 ± 630 μmol/mg creatinine and normotensive: 2,182 ± 972 μmol/mg creatinine, P < 0.001; Tables 2 and 3). In BTBR ob/ob mice, ANG II infusion itself led to a further increase of glomerular hypertrophy (3,877 ± 88.6 vs. 3,646 ± 59.8 μm2, P < 0.05; Figs. 2A and 4A) but decreased creatinine clearance by >50% (171 ± 27.2 vs. 565 ± 123.7 μl/min, P < 0.01) and aggravated albuminuria by twofold compared with normotensive BTBR ob/ob mice (4,869 ± 1,776 vs. 2,291 ± 524.4 μg/mg creatinine; Figs. 2E and 4B).
Effects of empagliflozin treatment in hypertensive BTBR ob/ob mice.
Similar to normotensive mice, empagliflozin treatment in these hypertensive mice reduced blood pressure not significantly by 6 mmHg (Table 3) but was still able to further increase glucose excretion already 3 wk after the start of treatment (empagliflozin: 9,716 ± 658 μg/mg creatinine and placebo: 5,303 ± 947 μg/mg creatinine, P < 0.05), which was sustained for the full experimental time. Empagliflozin successfully reduced blood glucose levels also in ANG II-infused mice after 3 wk by 140 and 240 mg/dl after 6 wk, respectively (Table 3). In contrast to normotensive BTBR ob/ob mice, empagliflozin had no significant impact on creatinine clearance (Table 3) and subsequently no effect on glomerular hypertrophy in hypertensive BTBR ob/ob mice (Fig. 4A). Nevertheless, empagliflozin still ameliorated albuminuria by 70% in hypertensive BTBR ob/ob mice (1,470 ± 226 vs. 4,869 ± 1,776 μg/mg creatinine, P < 0.05; Fig. 4B). However, empagliflozin treatment had no effect on glomerular matrix expansion (Fig. 4C) in hypertensive BTBR ob/ob mice.
In hypertensive BTBR ob/ob mice, empagliflozin treatment did not influence renal mRNA expression of SGLT1 (data not shown) or SGLT2 (data not shown). Similar to normotensive diabetic mice, 6 wk of empagliflozin treatment was not able to reduce renal mRNA expression of MCP-1 (data not shown), RANTES (data not shown), or IL-6 (data not shown) in ANG II-induced hypertensive mice.
To our knowledge, this is the first study investigating the impact of long-term SGLT2 inhibitor treatment on the development and progression of diabetic nephropathy in a murine model of type 2 diabetes without and with hypertension. Besides sustained hyperglycemia, hypertension strongly contributes to the development and progression of diabetic nephropathy in humans (1). Although the BTBR ob/ob model is considered to be a good murine type 2 diabetic nephropathy model, to further improve this model system and to better resemble the human situation, we induced hypertension in diabetic BTBR ob/ob mice by chronic ANG II infusion. The key findings of the present study are that the selective SGLT2 inhibitor empagliflozin, via increasing urinary glucose excretion, lowers blood glucose levels in type 2 diabetic BTBR ob/ob mice independently of accompanying hypertension. Empagliflozin markedly decreased albuminuria in both experimental settings, whereas empagliflozin ameliorated early diabetes-related features in the kidney, such as glomerular hypertrophy/hyperfiltration, markers of inflammation, and mesangial matrix expansion only in normotensive but not hypertensive BTBR ob/ob mice.
In contrast to STZ-induced (34) and Akita/+ (32) type 1 diabetes models, SGLT2 inhibition induced a profound glucosuria in BTBR ob/ob mice, as expected. Tubular glucose reabsorption is predominantly mediated by the two Na+-glucose transporters SGLT1 and SGLT2 and glucose transporters of the GLUT family (17). We hypothesized that the effectiveness of the SGLT2 inhibitor empagliflozin depends on the expression level of the renal target SGLT2 during the diabetic disease process. To verify this, we analyzed mRNA expression of SGLT1, SGLT2, GLUT1, and GLUT2. In contrast to the STZ-induced diabetes type 1 model, in which SGLT2 expression is downregulated (8), SGLT2 mRNA and protein expression are unaltered in the BTBR ob/ob model of type 2 diabetes. This may explain why the SGLT2 inhibitor empagliflozin more efficiently (by ∼50%) lowered blood glucose levels compared with STZ-induced diabetic mice with a complete lack of SGLT2 (34). In addition, BTBR ob/ob mice compared with BTBR WT mice demonstrated decreased renal SGLT1 mRNA expression but increased renal SGLT1 protein levels, both of which were not further altered under complete SGLT2 blockade. While this result was surprising and cannot be explained, a previous study (4) has also demonstrated differential expression of SGLT1 mRNA and protein. From these results, we interpret that diabetic BTBR ob/ob mice do not show a mechanism of compensatory increased SGLT1-mediated glucose reabsorption under complete SGLT2 blockade, as has been demonstrated by Vallon and coworkers in C57Bl/6 mice (25).
The most common early pathophysiological changes seen in diabetic nephropathy in both human and animal models are glomerular hyperfiltration/hypertrophy and albuminuria. In the “tubulocentric model,” the glomerular hyperfiltration and hypertrophy in diabetes are caused by an enhanced tubuloglomerular feedback mechanism (31, 35). Recently, it has been shown that SGLT2 inhibition activates the tubuloglomerular feedback mechanism (35) and thereby normalizes glomerular hyperfiltration under diabetic conditions (30). These results are consistent with our study, in which empagliflozin treatment attenuated the glomerular hypertrophy seen in normotensive BTBR ob/ob mice, which was associated with a numerical reduction of the increased creatinine clearance.
Empagliflozin treatment reduced blood glucose levels and increased glucosuria also in ANG II-infused hypertensive BTBR ob/ob mice, whereas it had no effect on both creatinine clearance and glomerular hypertrophy. The results in our diabetic mice suggest that ANG II infusion markedly overrides the milder effects that may be achieved by the use of empagliflozin in regard to the tubuloglomerular feedback mechanism. Consistent with this assumption, a study by Moniwa and coworkers (22) implicated that ANG II inhibits the tubuloglomerular feedback, which seems to be necessary for the empagliflozin effect independent of successful reduction of the blood glucose levels.
An increase in urinary albumin excretion is one of the earliest clinical symptoms of microvascular damage (2). In diabetic conditions, albuminuria is the result of an increased amount of filtered albumin in the injured glomeruli and a disturbed tubular function, which leads to reduced albumin reabsorption in the proximal tubules by endocytosis (6). Diabetic BTBR ob/ob mice reveal a profound albuminuria. Treatment of normotensive and hypertensive BTBR ob/ob mice with empagliflozin led to a significant reduction of albuminuria. In accordance with a previous study in Akita/+ mice (32), our data suggest that the amelioration of albuminuria by empagliflozin is a result of a mechanism secondary to the substantial blood glucose reduction and apparently independent of the tubuloglomerular feedback loop. These early changes in diabetic nephropathy are subsequently followed by gradual and progressive accumulation of the mesangial extracellular matrix, which appeared to be reduced via empagliflozin treatment. Although this effect was only mild and restricted to BTBR ob/ob mice without hypertension, it further supports the potential of empagliflozin as a treatment for diabetic nephropathy. In recent years, data from experimental models (38) and clinical trials (16) have established inhibition of the renin-angiotensin system (RAS) as the major preventive strategy for patients with type 2 diabetes at risk to develop diabetic nephropathy. While we did not compare the effects of empagliflozin treatment with effects of RAS inhibition, the lower efficacy of empagliflozin in our hypertensive disease model, taken together with experimental data from diabetic endothelial nitric oxide synthase-deficient db/db mice (38), suggest that angiotensin-converting enzyme inhibition is more potent in hypertensive diabetic nephropathy. However, further studies are needed to dissect the individual effects of RAS and SGLT2 inhibitors in the same disease model and to elucidate the potential of a combined therapeutic approach.
Under diabetic conditions, hyperglycemia and the resulting glucotoxicity are the driving force for inflammatory processes in the kidney. Treatment of BTBR ob/ob mice with empagliflozin for 12 wk led to reduced renal inflammation, as demonstrated by the significant reduction in mRNA expression of the proinflammatory cytokines MCP-1, RANTES, and IL-6. Despite the significant reduction of blood glucose in ANG II-infused BTBR ob/ob mice, empagliflozin treatment for 6 wk did not influence the expression of those chemokines in hypertensive BTBR ob/ob mice. Again, ANG II elicits direct effect on the expression of such proinflammatory cytokines (21, 37), which could mask the positive empagliflozin effects seen in the normotensive group of animals. However, treatment of hypertensive mice was performed only for 6 wk. Such a short treatment period was also not sufficient in normotensive mice to reduce the expression of these cytokines.
Beside these “on target” effects, SGLT2 inhibition is also associated with additional favorable effects for type 2 diabetic patients, such as weight loss (7) and decreased blood pressure (3). In animal (20, 32) and clinical studies with type 2 diabetic patients (3, 36), it has been shown that SGLT2 inhibition leads to weight loss. Increased urinary Na+ and glucose excretion under the SGLT2 inhibitor therapy induces osmotic diuresis, which contributes to weight loss. In BTBR ob/ob mice, empagliflozin treatment did not influence body weight, which also could be a consequence of the manifested hyperphagia. Furthermore, polyuria in empagliflozin-treated diabetic mice was not further increased in our experiments, and, thus, osmotic diuresis does not contribute to the reduction of body weight. In type 2 diabetic patients, empagliflozin treatment slightly reduces blood pressure by 3–7 mmHg (14). In Akita/+ mice, empagliflozin treatment prevented the rise in blood pressure seen in these mice on a placebo diet (32). In diabetic BTBR ob/ob mice without or with hypertension, empagliflozin slightly reduced systolic blood pressure to a comparable extent as seen in humans, but the detected blood pressure effect did not reach significance, which was most likely due to the limited number of animals as well as some variability caused by the tail-cuff system used for blood pressure measurements.
Taken together, our present results demonstrate that SGLT2 inhibition by empagliflozin is a good therapeutic option to lower blood glucose levels in type 2 diabetes. Moreover, empagliflozin treatment differentially ameliorates markers of renal injury, such as albuminuria, hyperfiltration/hypertrophy, inflammation, and mesangial matrix expansion in murine diabetic nephropathy without or with hypertension, supporting the concept of SGLT2 inhibition for the prevention of diabetic nephropathy.
C. Hugo has received, within the past 12 mo, research grant support for basic studies from Boehringer Ingelheim Pharma. E. Mayoux is full-time employee at Boehringer Ingelheim Pharma.
Author contributions: F.G., C.B., and N.J. performed experiments; F.G. and C.B. analyzed data; F.G., V.T.T., B.H., and C.H. interpreted results of experiments; F.G. prepared figures; F.G., C.B., E.M., V.T.T., B.H., and C.H. drafted manuscript; F.G., C.B., N.J., E.M., V.T.T., B.H., and C.H. approved final version of manuscript; E.M., V.T.T., B.H., and C.H. conception and design of research.
The technical assistance of Anika Lüdemann, Manuela Brandt, and Annett Heinrich is gratefully acknowledged.
- Copyright © 2014 the American Physiological Society