Cannabinoid Receptor 1 Blockade Ameliorates Albuminuria in Experimental Diabetic Nephropathy

All materials were purchased from Sigma-Aldrich (St. Louis, MO) unless otherwise stated.

N-(piperidin-1-yl)-5-(4-iodophenyl)-1-(2,3-dichlorophenyl)-4-methyl-1H-pyrazole-3-carboxamide (AM251), a CB1 receptor antagonist, was purchased from Cayman Chemical (Ann Arbor, MI), dissolved in ethanol to a stock concentration of 3 mg/ml, and stored at −80°C. Stock solutions were diluted in 18:1:1 ratio of saline/emulphor-620/absolute ethanol immediately prior to use. AM251 conjugated with 5-carboxytetramethylrhodamine was purchased from Tocris (Bristol, U.K.).

Both housing and care of laboratory animals were in accordance with Italian law (D.L.116/1992). Male C57BL6/J mice from The Jackson Laboratory (Bar Harbor, ME) were maintained on a normal diet under standard animal house conditions. Diabetes was induced in mice, aged 8 weeks and weighing ∼22 g, by intraperitoneal injection of STZ-citrate buffer (55 mg · kg body wt⁻¹ · day⁻¹) delivered in five consecutive daily doses. Mice sham-injected with sodium citrate buffer were used as controls. Diabetes onset was confirmed by blood glucose levels >250 mg/dl 4 weeks after the first dose of STZ; only animals with glucose levels >250 mg/dl (95%) were included in the study.

Animals were divided into the following groups: nondiabetic mice given vehicle (n = 7), nondiabetic mice given AM251 (n = 8), diabetic mice given vehicle (n = 11), and diabetic mice given AM251 (n = 10). AM251 was administered daily at the dosage of 1 mg/kg i.p. Mice sham-injected with a mixture 18:1:1 of saline/emulphor-620/absolute ethanol were used as controls. After 14 weeks of experimental diabetes, mice were killed by decapitation. The kidneys were rapidly dissected and weighed. The right kidney was frozen in N₂ and stored at −80°C for mRNA and protein analysis. Half left kidney was fixed in 10% PBS-formalin, then paraffin-embedded for light microscopy; the remaining tissue was embedded in optimal cutting temperature compound and snap-frozen in N₂. For electron microscopy, 1-mm³ pieces of renal cortex were fixed in 2% glutaraldehyde, 4% paraformaldehyde in phosphate buffer 0.12 mol/l for 4 h at room temperature, postfixed in 1% osmium tetroxide for 2 h, dehydrated in graded ethanol, and embedded in Epon 812.

Before killing, blood samples were taken via saphenous vein puncture on alert, 4 h–fasted animals, and glucose levels measured using a glucometer (Accuchek; Roche, Milan, Italy). Systolic blood pressure was assessed by tail-cuff plethysmography in prewarmed unanesthetized animals. Urine was collected over 18 h, with each mouse individually housed in a metabolic cage and provided with food and water ad libitum. Urinary albumin concentration was measured by a mouse albumin ELISA kit (Bethyl Laboratories). Creatinine clearance was estimated from serum and urine creatinine concentrations, as determined by high-performance liquid chromatography (19). Glycated hemoglobin was measured in whole blood samples obtained at the time of killing by quantitative immunoturbidimetric latex determination (Sentinel Diagnostic, Milan, Italy). Urinary N-acetylglucosamine (NAG) activity was assessed by colorimetric assay (Roche), and results were expressed as urinary NAG activity–to–creatinine ratio.

After antigen retrieval and blocking, 4-μm kidney paraffin sections were incubated overnight at 4°C with either anti-CB1 (Cayman Chemical) or WT-1 (Santa Cruz Biotechnology, Glostrup, Denmark) primary antibodies; then the specific staining was detected using the labeled streptavidin biotin + system horseradish peroxidase (Dako, Glostrup, Denmark). Sections were examined using an Olympus microscope (Olympus Bx4I), digitized with a high-resolution camera (Carl Zeiss, Oberkochen, Germany). A negative control was included in which the primary antibody was preincubated with a control peptide. Hippocampus sections were used as positive control for CB1. The percentage area of staining and the podocyte number/glomerular area were quantified by a computer-aided image analysis system (Axiovision 4.7; Carl Zeiss), where 20 glomeruli and 11 renal tubulointerstitial fields for each section were analyzed. Evaluations were performed by two independent investigators in a blinded fashion.

Snap-frozen sections (3 μm), fixed in cold acetone and blocked in 3% BSA, were incubated with guinea pig anti-nephrin (Progen Biotechnik, Maaßstraße, Germany), rabbit anti-podocin, anti-synaptopodin (Synaptic System, Gottingen, Germany), or anti–zonula occludens-1 (ZO-1; Zymed Laboratories, San Francisco, CA) primary antibodies. After washing, fluorescein isothiocyanate–conjugated anti–guinea pig (Santa Cruz Biotechnology) and anti-rabbit (Dako) secondary antibodies were added. Sections were examined using an Olympus epifluorescence microscope (Olympus Bx4I), digitized with a high-resolution camera (Carl Zeiss). Results were calculated as percentage positively stained tissue within the glomerular tuft. On average, 20 randomly selected hilar glomerular tuft cross-sections were assessed per mouse.

Double immunofluorescent staining was performed for CB1 and the specific podocyte marker nephrin. After blocking, sections were incubated with an anti-nephrin antibody and washed; then a fluorescein isothiocyanate–conjugated donkey anti–guinea pig antibody was added. After washing, sections were incubated with a rabbit anti-CB1 antibody for 18 h at 4°C, washed, and overlaid with a biotinylated swine anti-rabbit IgG (Dako) and then with Alexa-555–conjugated streptavidin (Invitrogen, Milan, Italy).

Binding of AM251 to the CB1 receptor was assessed by double fluorescence. After immunostaining for the CB1 receptor, as described above, sections were incubated with the AM251 fluorescent analog at 1 μmol/l concentration for 30 min, then washed, mounted, and visualized by epifluorescence microscopy.

Total kidney was homogenized in radioimmunoprecipitation assay buffer containing 0.5% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 10 mmol/l EDTA, and protease inhibitors. Proteins were separated by SDS-PAGE and electrotransferred to nitrocellulose membrane. After blocking in 5% nonfat milk, membranes were incubated with a rabbit anti-CB1 antibody (Calbiochem, Darstadt, Germany) overnight at 4°C. After washing, secondary anti-rabbit horseradish peroxidase–linked (Amersham) antibody was added. Detection was performed using SuperSignal West Femto chemiluminescence substrate (Pierce, Rockford, IL) and visualized on a Gel-Doc system (Bio-Rad). Band intensities were quantified by densitometry. Tubulin was used as internal control.

Total RNA was extracted from the renal cortex using TRIZOL reagent (Invitrogen). Total RNA (1 μg) was reverse transcribed into cDNA using the high-capacity reverse transcription kit (Applied Biosystems, Monza, Italy). CB1, nephrin, podocin, ZO-1, transforming growth factor-β1 (TGF-β1), connective tissue growth factor (CTGF), and fibronectin mRNA expression was analyzed by TaqMan real-time PCR using predeveloped TaqMan reagents (Applied Biosystems; CB1: Mm00432621; nephrin: Mm00497828; podocin: Mm00499929; ZO-1: Mm01320537; TGF-β1: Mm00441724; CTGF: Mm00515790; fibronectin: Mm01256744). Fluorescence for each cycle was analyzed quantitatively and gene expression normalized relative to the expression of HPRT. Because housekeeping genes ubiquitously expressed in the renal cortex do not control for variations in the glomerular number per specimen or changes in podocyte number (20), WT-1, a podocyte-specific gene, was used as endogenous reference for the evaluation of nephrin, podocin, and ZO-1 mRNA expression.

Apoptosis was assessed by the transferase-mediated dUTP nick-end labeling (TUNEL) method using the ApopTag In Situ Apoptosis Detection Kit (Millipore, Billerica, MA). Results were expressed as the number of positive cells per glomerulus in at least 20 random glomeruli. Slides pretreated with 20,000 units/ml DNase were used as positive control.

Glomerular cross-sectional area (A_G) was measured in 20 glomerular profiles per mouse using a computerized image analysis system (Axiovision 4.7; Carl Zeiss). The glomerular volume (V_G) was then calculated as V_G = β/K[A_G I^3/2, where β = 1.38 is the size distributor coefficient and K = 1.01 is the shape coefficient for glomeruli idealized as a sphere (21).

Ultrathin sections for ultrastructural examination were cut with the Ultracut E Reichert-Jung ultramicrotome, stained with uranylacetate and lead citrate, and examined with the transmission electron microscope Philips CM 10. Microphotographs were taken using a SIS Megaview II digital camera. Mean foot process width (FPW) was assessed as previously described (22,23) (supplementary Materials, available in an online appendix at http://diabetes.diabetesjournals.org/cgi/content/full/dc09-1336/DC1).

Data, presented as means ± SEM, geometric mean (25th–75th percentile), or fold change over control, were analyzed by Student t test or ANOVA, as appropriate. Least significant difference test was used for post hoc comparisons. Values for P < 0.05 were considered significant.

As shown in Table 1, after 14 weeks of diabetes both blood glucose and glycated hemoglobin levels were significantly higher in diabetic than in nondiabetic mice. Furthermore, compared with sham-injected control animals, diabetic mice showed a significant decrease in body weight. Systolic blood pressure was slightly higher in diabetic than in control mice, but this did not reach statistical significance.

To establish whether the CB1 receptor is expressed within the glomeruli and its expression is modulated by diabetes, we studied CB1 expression in kidney sections from both diabetic and control mice by immunohistochemistry. In control animals, only few glomerular cells stained positively for CB1 (Fig. 1,A). CB1 protein expression was enhanced in diabetic mice (Fig. 1,B) and semiquantitative analysis showed that the percentage positive area was 1.5-fold greater than in the controls (Fig. 1,E). Specificity of the antibody binding was confirmed by disappearance of the signal when the antibody was preabsorbed with a 10-fold excess of control peptide. In the tubuli, staining for the CB1 receptor was faint and no differences were observed between control and diabetic mice (nondiabetic: 2.07 ± 0.73; diabetic: 2.54 ± 0.32; P = not significant) (Fig. 1 C and D).

To clarify which glomerular cell type overexpressed the CB1 receptor, double-labeling immunofluorescence was performed in diabetic mice for the detection of both CB1 and nephrin. The CB1 receptor was expressed primarily by glomerular podocytes as the positive staining for nephrin (Fig. 1,L) colocalized with CB1 staining (Fig. 1,I), resulting in a partial yellow overlap (Fig. 1 M and N).

To confirm immunohistochemistry findings with more quantitative techniques, we measured CB1 mRNA and protein expression in total renal cortex by real-time PCR and immunoblotting. There was a significant two-fold increase in CB1 mRNA levels in diabetic mice (Fig. 1,F) compared with nondiabetic control animals. Immunoblotting showed a band migrating at ∼60 kDa, corresponding to the reported molecular weight of CB1 (Fig. 1,G), and densitometric analysis demonstrated that CB1 protein expression was significantly increased in the diabetic mice (Fig. 1 G and H).

As shown in Fig. 1 O–Q, glomerular staining using a fluorescent AM251 analog overlapped with the immunofluorescent staining for the CB1 receptor, confirming that AM251 binds to the CB1 receptor.

As shown in Table 1, treatment with AM251 did not affect the degree of glycemic control because both blood glucose and glycated hemoglobin levels were similar in treated and untreated diabetic animals. In addition, in both diabetic and control animals administration of AM251 did not alter body weight. Finally, no differences were observed in systolic blood pressure.

After 14 weeks of diabetes, there was a more than 10-fold increase in urinary albumin excretion rate in diabetic compared with nondiabetic animals. Treatment with AM251 did not alter albuminuria in the controls, but induced a significant 50% reduction in albumin excretion rate levels in the diabetic mice (Table 1). Results were similar when they were expressed as urinary albumin/creatinine ratio (nondiabetic: 146 ± 6.88; nondiabetic + AM251: 147.9 ± 33.87; diabetic: 2447.5 ± 607.7; diabetic + AM251: 806.3 ± 190.5; P < 0.001 diabetic vs. others). Urinary NAG activity/creatinine ratio, a marker of tubular damage, was slightly increased in diabetic mice compared with controls, although not significantly. In addition, no differences were observed between treated and untreated diabetic mice (nondiabetic: 88.11 ± 8.37; nondiabetic + AM251: 93.16 ± 16.52; diabetic: 178.13 ± 21.43; diabetic + AM251: 180.64 ± 40.86 units/g; P = not significant). Renal function was similar among groups as assessed by measurement of creatinine clearance (nondiabetic: 0.38 ± 0.07; diabetic: 0.41 ± 0.1; diabetic + AM251: 0.5 ± 0.1 ml/min; P = not significant).

To clarify the underlying mechanism of the beneficial effect of CB1 blockade on albuminuria, we assessed the effect of treatment with AM251 on the expression of nephrin, podocin, ZO-1, and synaptopodin by immunofluorescence. After 14 weeks of diabetes, there was a significant reduction in nephrin, podocin, and ZO-1 protein expression, and this effect was completely prevented in diabetic mice treated with AM251 (Fig. 2,A–F). No significant changes in synaptopodin protein levels were observed among groups (Fig. 2 G and H).

We also assessed nephrin, podocin, and ZO-1 mRNA in total renal cortex from all groups by real-time PCR and found a significant reduction in nephrin, podocin, and ZO-1 mRNA levels in diabetic compared with control mice. Treatment with AM251 prevented diabetes-induced nephrin and podocin mRNA downregulation (Fig. 3,A and B). A similar trend was also observed for ZO-1, although it did not reach statistical significance (Fig. 3 C).

To establish whether in diabetic mice treated with AM251, podocyte protein expression was preserved because CB1 blockade prevented podocyte damage, podocyte number, apoptosis, and ultrastructure were assessed by WT-1 staining, TUNEL assay, and electron microscopy, respectively. The number of both WT-1– and TUNEL-positive cells per glomerular cross-sectional area displayed in a podocyte distribution did not differ among groups (Fig. 4) (nondiabetic: 2.0 ± 0.5; diabetic: 2.2 ± 0.4; nondiabetic + AM251: 2.5 ± 0.3; diabetic + AM251: 2.1 ± 0.5, TUNEL-positive cells per 100 glomeruli; P = not significant). Electron microscopy analysis did not show any evidence of ultrastructural glomerular (Fig. 4,A–F) and tubular (Fig. 5,G–L) damage in the studied animals. In particular, the normal arrangement of interdigitating foot processes was maintained in all groups, and podocyte foot processes appeared tall and narrow in both treated and untreated diabetic mice (Fig. 5 A–F). Furthermore, mean FPW was comparable among groups (nondiabetic: 350 ± 1.33; nondiabetic + AM251: 352 ± 1.27; diabetic: 375 ± 1.11; diabetic + AM251: 364 ± 0.99 nm; P = not significant).

Glomerular volume was significantly increased in diabetic mice compared with control animals and this effect was not altered by treatment with AM251 (nondiabetic: 136 ± 10.33; diabetic: 358 ± 19.06; nondiabetic + AM251: 127 ± 4.98; diabetic + AM251: 369 ± 5.72 μm³; P < 0.05 diabetic and diabetic + AM251 vs. nondiabetic).

At electron microscopy, the degree of mesangial expansion and glomerular basement membrane thickening was very mild in diabetic mice, and no major differences were observed among groups (Fig. 4,A–D). However, in the renal cortex levels of mRNA encoding for fibronectin, TGF-β1, and CTGF were significantly greater in diabetic than in control animals. Treatment of diabetic mice with AM251 did not affect the overexpression of these markers of fibrosis, suggesting that CB1 blockade does not have antifibrotic properties in this model (Figure 6 A–C).

In this study, we have provided evidence that in experimental diabetes the CB1 receptor is overexpressed within the glomeruli, predominantly by glomerular podocytes. Second, we have shown that blockade of the CB1 receptor with AM251 ameliorates albuminuria and prevents downregulation of nephrin, podocin, and ZO-1. Taken together, these data suggest a role of the CB1 receptor in the pathogenesis of proteinuria in diabetes.

In nondiabetic mice, only a few glomerular cells stained positively for the CB1 receptor, but after 14 weeks of experimental diabetes there was a significant 1.5-fold increase in glomerular CB1 expression. A previous study has shown that the CB1 receptor is expressed at low level in rat kidneys, but no differences in renal CB1 expression were observed between Zucker fatty rats and lean animals (18). Therefore, this is the first evidence of glomerular CB1 overexpression in experimental diabetes and, to our knowledge, in any renal disease.

In agreement with previous data in rat kidneys (18), immunoreactivity for CB1 was predominantly localized to the glomeruli. A faint signal was also detectable in the tubuli, but it was not enhanced in the diabetic mice. Both pattern of staining and colocalization with the podocyte marker nephrin strongly indicate that in diabetic mice the CB1 receptor was primarily overexpressed by glomerular podocytes. This is not surprising because similarities have been reported between podocytes and neurons (24,25), the predominant CB1 receptor–expressing cell type in physiological conditions (13). The underlying mechanism/s of CB1 receptor induction in diabetic podocytes is entirely unknown. However, oxidative stress, which is enhanced in the diabetic glomeruli, is known to induce CB1 receptor expression in other cell types (26).

Consistent with our immunohistochemical data, we found a significant 1.8- to 2-fold increase in CB1 both protein and mRNA expression in renal cortex from diabetic mice as assessed by immunoblotting and real-time PCR. Diabetes-induced enhanced transcription of the CB1 receptor is a potential underlying mechanism as both activator protein 1 (AP-1) and nuclear factor-κB (27,28), the transcription factors predominantly implicated in the pathogenesis of diabetic nephropathy, have binding sites on the promoter region of the CB1 receptor gene (29).

To establish whether upregulation of the CB1 receptor in podocytes plays a role in the pathogenesis of proteinuria in diabetes, we studied the effect of treatment with AM251. AM251, a potent and specific CB1 receptor inverse agonist (30,,–33), is structurally very close to rimonabant, but exhibits better binding affinity and selectivity for the CB1 receptor (33). In our model, selectivity of AM251 binding to the CB1 receptor was indirectly confirmed by colocalization of AM251 and CB1 within the diabetic glomeruli. AM251 was administered daily at the dosage of 1 mg/kg on the basis of previous studies that have proven both efficacy and safety of this dose in mice (34,35). Delivery was via intraperitoneal injection to ensure that all animals were given an equal amount of drug. After 14 weeks of diabetes, there was a 10-fold increase in albuminuria in diabetic mice compared with controls. Treatment with AM251 induced a significant 50% reduction in albuminuria in diabetic mice, supporting the hypothesis that signaling through the CB1 receptor contributes to enhanced glomerular permeability to albumin.

Administration of AM251 did not affect body weight in either diabetic or control mice. This is in agreement with previous reports showing that CB1 blockade prevents weight gain in animals with diet-induced obesity, but is much less efficacious at exerting this effect in lean animals fed a standard diet (36,–38). Furthermore, blood glucose levels, glycated hemoglobin, and systolic blood pressure were similar in treated and untreated diabetic mice, consistent with the antiproteinuric effect of CB1 blockade observed in these mice being independent of both metabolic and hemodynamic factors. A significant reduction in proteinuria has been recently reported in obese Zucker fatty rats treated with rimonabant (18); however, in this model CB1 receptor was not overexpressed in the glomeruli, and the antiproteinuric effect was most likely due to amelioration of the metabolic profile because it was associated with body weight loss and diminution of both blood glucose and lipid levels. In our study, we have purposely chosen an animal model of type 1 diabetes without obesity, insulin resistance, and lipid abnormalities to ensure that the beneficial effects of CB1 blockade were independent of changes in metabolism.

To clarify the mechanism/s of the beneficial effect of AM251, we tested whether CB1 receptor blockade prevents downregulation of podocyte proteins implicated in the maintenance of glomerular permselectivity to proteins. After 14 weeks of diabetes, there was a significant reduction in nephrin, podocin, and ZO-1 both mRNA and protein expression. Treatment with AM251 prevented diabetes-induced downregulation of nephrin, podocin, and ZO-1 protein expression. Results were confirmed by mRNA analysis for both nephrin and podocin, and a similar trend was also observed for ZO-1. Loss of slit diaphragm and slit diaphragm–associated proteins has been implicated in the pathogenesis of proteinuria (2); therefore, downregulation of nephrin, podocin, and ZO-1 is a possible mechanism whereby CB1 overexpression may lead to increased glomerular permeability to albumin.

The number of both total and apoptotic podocytes was not increased in diabetic mice. Furthermore, there was no evidence of podocyte foot process effacement at the ultrastructural level or changes in mean FPW. It is, thus, unlikely that levels of nephrin, podocin, and ZO-1 were normal in diabetic mice treated with AM251 because of prevention of podocyte damage. These data also suggest that in early experimental diabetes, downregulation of slit diaphragm proteins precedes the development of podocyte foot process effacement/loss and that podocyte structural damage is not strictly required for the development of proteinuria. Consistent with this view, in nephrin knockout animals proteinuria occurs even in the absence of any defects in the podocyte foot processes (39). The degree of nephrin reduction required for the development of proteinuria is unknown; however, the parallel downregulation of other slit diaphragm proteins is likely to cause a rise in this threshold level (40).

In the diabetic mice, tubular ultrastructure was normal and urinary activity of NAG, a marker of tubular damage, was comparable in treated and untreated mice. Therefore, prevention of tubular injury is not a likely explanation of AM251 antiproteinuric activity. However, we cannot exclude the possibility that at a later stage of experimental diabetes, when severe tubular damage occurs, CB1 blockade may have further beneficial effects due to prevention of tubulointerstitial injury, as previously shown in the Zucker fatty rat model (18).

Glomerular hypertrophy and accumulation of extracellular matrix components, which is mediated predominantly by the prosclerotic cytokines TGF-β1 and CTGF, are other characteristic features of diabetic nephropathy (1). In our study, after 14 weeks of mild diabetes, C57BL6/J mice, which are relatively resistant to the development of glomerulosclerosis, did not show major ultrastructural abnormalities in the mesangium and the glomerular basement membrane. However, there was a significant increase in both glomerular volume and renal mRNA expression of fibronectin, TGF-β1, and CTGF in diabetic mice. These diabetes-induced effects were left unchanged by treatment with AM251, indicating failure of CB1 blockade in interfering with glomerular hypertrophy and renal fibrogenesis. In animal models of chronic liver injury, blockade of the CB1 receptor decreases fibrosis by lowering hepatic TGF-β1 expression and reducing accumulation of fibrogenic cells (14). The different effect of CB1 blockade on renal and liver fibrosis is not entirely surprising because the cell types overexpressing the CB1 receptor in liver cirrhosis, namely myofibroblasts and hepatic stellate cells, have much greater profibrotic potential than podocytes. In addition, a tissue-specific effect of CB1 receptor activation cannot be excluded.

In conclusion, our findings may have important implications for diabetic nephropathy in humans. Proteinuria is a characteristic feature of diabetic nephropathy and a key determinant of progression (1). Nephrin is downregulated in early diabetic nephropathy, and this has been implicated in the pathogenesis of the diabetic proteinuria (10). Our data, showing upregulation of CB1 receptors in podocytes and a beneficial effect of AM251 on both albuminuria and nephrin loss, suggest that an elevated CB1 receptor tone (41) is also involved in the pathogenesis of the diabetic proteinuria and identify a new target for therapeutic intervention.

This work was supported by the Compagnia di San Paolo, the Italian Society of Diabetes, the University of Turin (ex-60% grant), and the Piedmont Region Research Grant.

No potential conflicts of interest relevant to this article were reported.