Deficiency of cannabinoid receptor of type 2 worsens renal functional and structural abnormalities...

Deficiency of cannabinoid receptor of type 2 worsensrenal functional and structural abnormalities instreptozotocin-induced diabetic miceFederica Barutta1, Serena Grimaldi1, Irene Franco2, Stefania Bellini1, Roberto Gambino1, Silvia Pinach1,Alessandro Corbelli3, Graziella Bruno1, Maria Pia Rastaldi3, Teresa Aveta4, Emilio Hirsch2,Vincenzo Di Marzo4 and Gabriella Gruden1

1Diabetic Nephropathy Laboratory, Department of Medical Sciences, University of Turin, Turin, Italy; 2Departments of Genetics, Biologyand Biochemistry, University of Turin, Turin, Italy; 3Renal Research Laboratory, Fondazione IRCCS, Ospedale Maggiore Policlinico andFondazione D’Amico per la Ricerca sulle Malattie Renali, Milan, Italy and 4Endocannabinoid Research Group, Institute of BiomolecularChemistry – CNR, Pozzuoli, Italy

A functionally active endocannabinoid system is present

within the kidney. The cannabinoid receptor type 2 (CB2) is

expressed by both inflammatory cells and podocytes, and its

activation has beneficial effects in experimental diabetic

nephropathy. To further explore the role of CB2 in diabetic

nephropathy, we studied renal functional and structural

abnormalities in streptozotocin-induced diabetic CB2

knockout mice. In diabetic mice, deletion of the CB2 receptor

albuminuria, the downregulation of podocin and nephrin,

mesangial expansion, overexpression of extracellular matrix

components, monocyte infiltration, and reduced renal

function were all exacerbated. To investigate the relative

contributions of podocytes and monocytes to the phenotype

of diabetic knockout mice, bone marrow transplantation

experiments were performed. The lack of CB2 on bone

marrow–derived cells was shown to be important in driving

the enhanced glomerular monocyte accrual found in diabetic

knockout mice. Absence of CB2 on resident glomerular cells

had a major role in worsening diabetic nephropathy, both

functional and structural abnormalities, likely by enhanced

MCP-1 and CB1 signaling. Studies in cultured podocytes

demonstrated that CB2 expression is not altered by a high

glucose milieu but is downregulated by mechanical stretch,

mimicking glomerular capillary hypertension. Thus, CB2

deletion worsens diabetic nephropathy, independent of

bone marrow–derived cells.

Kidney International (2014) 86, 979–990; doi:10.1038/ki.2014.165;

published online 14 May 2014

KEYWORDS: albuminuria; chemokine receptor; chronic kidney disease;

diabetic nephropathy; fibrosis; inflammation

Diabetic nephropathy (DN) is characterized by increasedglomerular permeability to proteins and excessive extracel-lular matrix (ECM) deposition in the mesangium.1 Bothhyperglycemia and hypertension are key determinants in thedevelopment of DN.2 In addition, increasing evidencesuggests that a low-grade inflammatory response also has arole in disease etiology. In particular, the potent chemokinemonocyte chemoattractant protein 1 (MCP-1) is an impor-tant mediator of both functional and structural abnormalitiesof the diabetic kidney.3 The slit diaphragm, a junctionconnecting foot processes (FP) of neighboring podocytes,represents the major restriction site to protein filtration.4 Inhuman DN, there is a downregulation of the slit diaphragmproteins, nephrin, and podocin, and diabetes-related insults,such as advanced glycated products, MCP-1, and mechanicalstretch, which mimics glomerular capillary hypertension,downregulate nephrin in cultured podocytes.5–7

Endocannabinoids (EC), anandamide (AEA) and 2-arachidonoylglycerol (2-AG), bind to two G-protein-coupledseven transmembrane cannabinoid receptors named CB1 andCB2. In experimental diabetes, the CB1 is overexpressed bypodocytes, and CB1 blockade ameliorates albuminuria andnephrin loss.8 The CB2 is mainly expressed by cells of hema-topoietic origin including monocytes–macrophages,9 andCB2 activation is protective in animal models of chronicinflammatory diseases, such as atherosclerosis10 and liverfibrosis,11 whereas CB2 deletion exacerbates tissue damage byenhancing inflammatory, oxidative, and fibrotic processes.12–14

We have previously demonstrated that CB2 receptors are alsoexpressed both in vitro and in vivo by podocytes and undergodownregulation in human DN. Furthermore, in experimentaldiabetes, AM1241, a specific CB2 agonist, reduces albumi-nuria by preventing loss of podocyte proteins important inpreserving glomerular permselectivity.15

To further explore the role of the CB2 receptor in DN, we haveassessed both functional and structural abnormalities of DN inCB2 knockout (KO) mice made diabetic with streptozotocin.

http://www.kidney-international.org b a s i c r e s e a r c h

& 2014 International Society of Nephrology

Correspondence: Federica Barutta, Diabetic Nephropathy Laboratory,

Department of Medical Sciences, University of Turin, C/so Dogliotti 14, Turin

10126, Italy. E-mail: [email protected]

Received 18 July 2013; revised 25 March 2014; accepted 27 March 2014;

published online 14 May 2014

Kidney International (2014) 86, 979–990 979

http://dx.doi.org/10.1038/ki.2014.165

http://www.kidney-international.org

mailto:[email protected]

RESULTSMetabolic and physiological parameters in wild-type (WT)and CB2-KO mice

As shown in Table 1, after 16 weeks of diabetes both bloodglucose and glycated hemoglobin levels were significantlyhigher in diabetic (DM) than in nondiabetic (ND) mice.Furthermore, compared with controls, DM mice showed asignificant decrease in body weight and a significant increasein kidney weight–to–body weight (KW/BW) ratio. Lack ofCB2 receptors did not affect the degree of glycemic control inDM mice and did not alter body weight, KW/BW ratio, andsystolic blood pressure in either ND or DM animals.

Albumin excretion rate (AER), urinary N-acetylglucosamine(NAG), and serum creatinine

There was a significant increase in urinary AER in DM ascompared with ND animals. The absence of CB2 did not alteralbuminuria in the controls but significantly enhanced AERin DM mice. NAG activity/creatinine ratio, a marker oftubular injury, was also significantly augmented in DM micebut did not differ in WT and KO DM animals. Diabetes didnot enhance serum creatinine values in WT mice, and asignificant reduction in renal function was exclusivelyobserved in DM-KO mice (Table 1).

Podocyte abnormalities

As shown in Figure 1, there was a significant diminution inthe glomerular staining for both nephrin and podocin in DMas compared with ND mice, and this effect was significantlyworsened in DM-KO mice. Immunoblotting analyses con-firmed a greater downregulation of both nephrin andpodocin in DM-KO mice. The number of TUNEL (terminaldeoxinucleotidyl transferase-mediated dUTP-fluorescein nickend labeling)- and cleaved caspase 3–positive cells perglomerular cross-sectional area displayed in a podocytedistribution was very modest and did not differ betweengroups (TUNEL: ND-WT: 1.9.0±0.6; DM-WT: 2.1±0.5;ND-KO: 2.0±0.2; DM-KO: 2.1±0.5. Cleaved caspase-3: ND-WT: 1.2±0.2; DM-WT: 1.7±0.3; ND-KO-: 1.3±0.7; DM-KO: 1.5±0.5, positive cells/100 glomeruli, P¼NS). Electronmicroscopy analysis showed areas of podocyte–FP fusion in

DM mice, but the extent of FP effacement was similar inDM-WT and DM-KO mice (Figure 2a).

Glomerular hypertrophy, mesangial expansion, andexpression of ECM components

Glomerular volume was significantly increased in DM ascompared with ND animals, and this effect was not altered byCB2 deletion (ND-WT: 140±9.05; DM-WT: 345±12.03;ND-KO: 134±4.05; DM-KO: 357±5.23 mm3, Po0.05 DM-WT and DM-KO vs. ND). Histological assessment by PAS(periodic acid–Schiff) staining revealed a mild mesangialexpansion and scattered tubulointerstitial damage in DMmice. The degree of glomerular injury was greater in DM-KOmice. Indeed, a more prominent mesangial expansion wasobserved by both light and electron microscopy (Figure 2band Figure 3a and b). As shown in Figure 3e and h, the rise inboth fibronectin and type IV (a4) collagen mRNA levelsinduced by diabetes was significantly greater in KO mice andan increase in type I collagen was exclusively observed inDM-KO animals (Figure 3d). Although changes in total renalcortex expression may also reflect tubulointerstitial abnorm-alities, immunohistochemical analysis and picrosirius redstaining confirmed that glomerular expression of bothcollagen and fibronectin was greater in DM-KO comparedwith DM-WT mice (Figure 3c, f, and g). Finally, there was asignificant increase in both transforming growth factor b1and connective tissue growth factor mRNA expression in DMas compared with ND mice, and the expression of theseprosclerotic cytokines was significantly greater in DM-KOthan in DM-WT mice (Figure 3i and j).

Glomerular monocyte infiltration and MCP-1/C-C motifchemokine receptor 2 (CCR2) system

The number of glomerular cells positive for the monocytemarker F4/80 was significantly augmented in DM-WT ascompared with ND mice. The absence of CB2 furtherincreased diabetes-induced monocyte accrual without affect-ing monocyte infiltration in ND mice (Figure 4a and b).Expression of GR1, a marker of the M1 subpopulation ofproinflammatory monocytes,16 was greater in DM mice, andthis effect was exacerbated by CB2 deficiency (Figure 4c).

Table 1 | Metabolic and physiological parameters in wild-type and CB2� /� knockout mice

ND-WT ND-KO DM-WT DM-KO

Body weight (g) 28.37±0.57 28.40±0.77 24.20±0.82a 23.86±0.51a

BG (mg/dl) 101±8.56 104±14.61 309±11.76a 338±15.68a

Glycated Hb (%) 5.01±0.10 5.02±0.09 10.94±0.22a 11.15±0.55a

SBP (mm Hg) 109±7.27 115±7.08 117±6.46 115±5.69KW/BW ratio 5.96±0.23 5.32±0.14 7.53±0.24a 7.57±0.22a

SCr (mmol/l) 6.33±0.31 6.27±0.35 6.45±0.31 8.46±1.02b

NAG (U/g) 86.27±2.42 81.29±6 142.74±10.05a 130.98±5.33a

AER (mg/18 h) 87.4 (70.5–106.2) 87.3 (74.9–104.9) 216.2 (192–274.4)a 345.8 (329.3–352.8)a,b

Abbreviations: AER, albumin excretion rate; BG, blood glucose; DM, diabetic mice; KO, CB2� /� knockout; KW/BW, kidney weight/body weight; NAG, N-acetylglucosamineactivity/creatinine ratio; ND, nondiabetic mice; SBP, systolic blood pressure; SCr, serum creatinine; WT, wild type.Data are shown as means±s.e.m. or geometric means (25th–75th percentile).aPo0.001 DM-WT and DM-KO vs. ND-WT and ND-KO.bPo0.05 DM-KO vs. others.

b a s i c r e s e a r c h F Barutta et al.: CB2 deletion and experimental diabetes

980 Kidney International (2014) 86, 979–990

Diabetes increased MCP-1 mRNA levels withoutaffecting the expression of the MCP-1 receptor CCR2(Figure 4d). On the contrary, CB2 deletion did not alterMCP-1 expression but enhanced CCR2 mRNA levels in bothND and DM mice (Figure 4e). Assessment of glomerularCCR2 expression by immunohistochemistry confirmedCCR2 upregulation in mice lacking the CB2 receptor(Figure 4f and g).

Bone marrow (BM) transplantation study

To investigate the relative contributions of glomerularresident cells and BM-derived cells to the phenotype ofDM-KO mice, BM transplantation experiments were per-formed. As shown in Table 2, BM transplantation did notaffect metabolic/physiological parameters in either WT or KOanimals. After 14 weeks of diabetes, there was a significantincrease in AER, serum creatinine levels, and fibronectin

NephrinN

DD

MPodocin

WT KO

ND

DM

WT KO

*

ND ND DM DM0

10

20

30

Glo

mer

ular

pod

ocin

posi

tive

area

(%

)

40

50KO mice

*

WT mice

ND

Nephrin Podocin

Tubulin Tubulin

ND-WT

ND-KO

DM-W

T

DM-K

O

ND-WT

ND-KO

DM-W

T

DM-K

O

0

1

0.5

Nep

hrin

/tubu

lin r

atio

1.5

*

ND DM DM

ND ND DM DM0

0.2

0.4

0.6

0.8

1P

odoc

in/tu

bulin

rat

io1.2

*

#

ND ND DM DM

0

5

10

15

Glo

mer

ular

nep

hrin

posi

tive

area

(%

)

20

25

KO mice

WT mice

KO mice

WT mice

KO mice

WT mice

Figure 1 | Effect of the CB2 deficiency on diabetes-induced nephrin and podocin downregulation. Both nephrin and podocin expressionswere assessed in ND-WT (n¼ 12), ND-KO (n¼ 6), DM-WT (n¼ 9), and DM-KO (n¼ 7) mice by immunofluorescence and immunoblotting.Representative immunofluorescence images of (a) nephrin and (b) podocin are shown (�400), and (c, d) quantification of glomerular staining isreported in the graphs (*Po0.001 DM-WT vs. ND-WT, ND-KO, and DM-KO). (e, g) Representative immunoblots of nephrin and podocin proteinexpression in total protein extracts from the renal cortex are shown. Tubulin was used as internal control. Results of densitometry analyses arereported in panel (f) for nephrin (*Po0.01 DM-WT vs. ND-WT, ND-KO, and DM-KO) and in panel (h) for podocin (*Po0.001 DM-WT vs. ND-WTand ND-KO; #Po0.05 DM-KO vs. DM-WT). DM, diabetic; KO, CB2� /� knockout; ND, nondiabetic; WT, wild type.


F Barutta et al.: CB2 deletion and experimental diabetes b a s i c r e s e a r c h

expression in DM-KO as compared with DM-WT animals,whereas podocin expression was significantly reduced. Theseabnormalities were not rescued by transplantation with BMfrom WT mice. Moreover, in DM-WT animals, transplanta-tion of CB2� /� BM cells did not magnify albuminuria,fibronectin overexpression, renal function loss and podocindownregulation, thus failing to mimic the phenotype of

DM-KO mice (Table 2 and Figure 5a–i). Collectively, thesedata suggest that BM-derived cells do not have a major rolein mediating the deleterious effects of CB2 deprivation.

To confirm the efficacy of BM transplantation in affectingCB2 expression on infiltrating monocytes, double stainingfor CB2 and F4/80 was performed in chimeric animals. Asexpected, in DM-WT animals transplanted with BM from KO

ND

DM

ND

DM

WT

**

*

KO

#

#

Figure 2 | Transmission electron microscopy analysis. Representative electron microscopy images showing glomeruli from both nondiabetic(ND) and diabetic (DM), wild-type (WT) and CB2� /� knockout (KO) mice. (a) The extent of foot processes fusion (#) was similar in WTand KO diabetic mice. (b) Glomeruli from DM-KO mice showed a more prominent mesangial expansion (*) than DM-WT mice.



mice (DM-WT-cKO) glomerular immunostaining for CB2showed a podocyte distribution and did not localize toinfiltrating F4/80-positive cells (Figure 5j–l). On the contrary,in DM-KO animals transplanted with BM from WT mice(DM-KO-cWT), the staining exclusively localized to cellsexpressing F4/80 (Figure 5m–o). The number of infiltratingmonocytes was similar in DM-KO-cWT and DM-WT mice,whereas it was enhanced in DM-WT-cKO and of a compar-able magnitude to that observed in DM-KO mice (ND:0.22±0.07; DM-WT: 1.09±0.13; DM-KO: 2.34±0.11; DM-KO-cWT: 1.25±0.08; DM-WT-cKO: 2.53±0.17; Po0.01 DM-KO and DM-WT-cKO vs. DM-WT and DM-KO-cWT). Thissuggests that lack of CB2 on BM-derived cells drives theenhanced monocyte recruitment observed in DM-KO mice.Differences were not due to local changes in MCP-1 levels asthe 2.3-fold increase in MCP-1 mRNA expression that wasobserved in DM as compared with ND mice was not

significantly modified by either CB2 deletion or BM trans-plantation (DM-KO: 2.17±0.37; DM-KO-cWT: 2.16±0.15,DM-WT-cKO: 2.12±0.02; fold increase over control).

Effect of CB2 deletion on the EC system

We also investigated whether compensatory changes in thelocal EC system could account for the phenotype of DM-KOmice. In DM mice, expression of CB1 was enhanced(Figure 6a and b) and localized predominantly to podocytesas confirmed by co-staining with WT-1 (Figure 6c), whereasno overlap was found with the mesangial cell marker a-smooth muscle actin (Figure 6d). Similarly, CB2 had apredominant podocyte distribution and did not colocalizewith mesangial cells (Supplementary Figure S1 online).Importantly, no changes in CB1 expression/distribution wereobserved in DM-KO mice with respect to DM-WT mice.

DM

ND

WT KO

DM

ND

WT KO

ND DM DM0

0.5

1

1.5

2

2.5

Rel

ativ

e ex

pres

sion

of ty

pe IV

col

lage

nα4

mR

NA

#

§

ND

*

05

101520

% M

esan

gial

are

a

253035

ND DM DM

DM

ND

WT KO

ND

§§

ND

*

05

101520

Glo

mer

ular

fibr

onec

tinpo

sitiv

e ar

ea (

%)

253035

ND DM DM

§

ND0

0.51

1.5

0

0.5

1

1.5

† †

‡2

Rel

ativ

e ex

pres

sion

of T

GF

-β1

mR

NA

Rel

ativ

e ex

pres

sion

of C

TG

F m

RN

A

2.5

2

Rel

ativ

e ex

pres

sion

of fi

bron

ectin

mR

NA

2.53

3.5

DM DM ND DM DM ND0

1

#

2

3

4

5

DM DM

†

#

0

1

Rel

ativ

e ex

pres

sion

of ty

pe I

colla

gen

mR

NA

2

3

DM DM

KO mice

WT miceKO mice

WT mice

KO mice

WT mice

KO mice

WT miceKO mice

WT miceKO mice

WT mice

KO mice

WT mice

Figure 3 | CB2 absence worsens diabetes-induced glomerular structural abnormalities and enhances expression of fibrosis markers.Renal cortex samples from ND-WT (n¼ 12), ND-KO (n¼ 6), DM-WT (n¼ 9), and DM-KO (n¼ 7) mice were studied 16 weeks after diabetesinduction. (a) Representative PAS staining images (�400) and (b) quantitation of percentage of mesangial areas are shown. (c) Collagenstaining and both (d) type I and (e) type IV collagen mRNA expression were studied by picrosirius red staining and real-time PCR, respectively.(f, g) Glomerular fibronectin expression was assessed by immunofluorescence and (h) renal cortex fibronectin gene expression by real-timePCR. (i) Transforming growth factor (TGF)-b1 and (j) connective tissue growth factor (CTGF) mRNA levels were measured by real-time PCR intotal renal cortex. (#Po0.05, zPo0.01, and *Po0.001 DM-WT vs. ND-WT; wPo0.05, and yPo0.01 DM-KO vs. DM-WT.) DM, diabetic; KO, CB2� /�

knockout; ND, nondiabetic; WT, wild type.



Figure 7 shows the levels of EC, EC-related molecules,and EC-related enzymes in the renal cortex of studied mice.2-AG, the main CB2 ligand, was diminished in the renalcortex of DM mice in parallel with a reduced expression ofthe 2-AG synthetic enzyme diacylglycerol lipase-a, withno change in the levels of the degrading enzyme, mono-acylglycerol lipase (Figure 7a–c). CB2 deletion did notmodify these effects, but increased the levels of both AEA,

the CB1 ligand, and oleoylethanolamine (OEA), an EC-related molecule (Figure 7a and b). This is was likely due toreduced degradation as levels of fatty acid amide hydrolase,which hydrolyzes both AEA and OEA, were significantlylower in DM-KO mice (Figure 7f), whereas the levelsof the AEA and OEA-biosynthesizing enzyme N-acylphosphatidylethanolamine-specific phospholipase D wereunaltered.

DM

ND

WT KO

DM

ND

WT KO

ND ND DM DM

ND ND DM

OO

DMND01

34567

2

ND DM DM

† †

ND ND DM DM

*

§

*

0

1

2

3

46

4

2

0

Num

ber

of F

4/80

+

posi

tive

cells

Rel

ativ

e ex

pres

sion

of G

R1+

mR

NA

Rel

ativ

e ex

pres

sion

of C

CR

2 m

RN

A

ND ND DM DM

# #

0

2

4

6

8

Rel

ativ

e ex

pres

sion

of M

CP

-1 m

RN

A

02468

101214

CC

R2

prot

ein

expr

essi

on(%

pos

itive

are

a)

‡KO mice

WT mice

KO mice

WT miceKO mice

WT mice

KO mice

WT mice

KO mice

WT mice

Figure 4 | Effect of CB2 deletion on markers of inflammation in diabetic (DM) and nondiabetic (ND) mice. ND-WT (n¼ 12), ND-KO (n¼ 6),DM-WT (n¼ 9), and DM-KO (n¼ 6) mice were studied 16 weeks after diabetes onset. Glomerular monocyte accrual was assessed bycounting the number of F4/80-positive cells within the glomeruli. (a) Representative images are shown (�400) and (b) the number ofF4/80-positive cells are reported in the graph. (c) GR1þ , (d) monocyte chemoattractant protein 1 (MCP-1), and (e) C-C motif chemokinereceptor 2 (CCR2) mRNA levels were measured by real-time PCR on total RNA extracts from the renal cortex, and results corrected for theexpression of the housekeeping gene HRPT (hypoxantine phosphoribosyltransferase). (f) Representative images of immunohistochemistrystaining for CCR2 are shown (�400) and (g) quantification of glomerular staining are reported in the graphs. (*Po0.05 DM-WT vs. ND;zPo0.05 and yPo0.01 DM-KO vs. DM-WT; #Po0.001 DM-WT and DM-KO vs. others; wPo0.01 and 1Po0.001 DM-KO and ND-KO vs. others.)KO, CB2� /� knockout; WT, wild type.

Table 2 | Metabolic and physiological parameters in the bone marrow transplantation study

ND-WT ND-WT-cko ND-KO-cwt DM-WT DM-KO DM-WT-cko DM-KO-cwt

Body weight (g) 31.1±0.8 30.3±1.3 29.9±0.4 24.9±0.7a 24.4±0.9a 23.1±1.9a 23.4±0.7a

BG (mg/dl) 98.7±6.5 100.5±2.3 102.5±3.2 360.2±23.5a 301.4±23.5a 365.7±18.5a 358.3±19.3a

Glycated Hb (%) 4.1±0.1 3.9±0.2 3.9±0.2 10.1±0.4a 9.4±0.6a 9.7± 0.8a 9.1±0.2a

SBP (mm Hg) 122±2.0 118±3.5 120±1.8 120±1.0 118±2.6 117±36.3 113±2.5KW/BW ratio 5.3±0.3 5.1±0.4 4.4±0.4 7.4±0.4b 7.3±0.94b 8.1±0.8b 7.6±0.3b

SCr (mmol/l) 5.27± 0.83 5.04±0.82 5.13±0.79 5.49±2.08 7.33±0.41d 5.17±0.70 6.85±0.56d

AER (mg/18 h) 68.1 (64.6–75.3) 59.6 (56.7–65.7) 61.6 (57.4–68.9) 234.0c (200.0–285.9) 359.9d (227.9–519.4) 236.2c (211.5–257.4) 376.8d (353.2–368.5)

Abbreviatons: AER, albumin excretion rate; BG, blood glucose; cko, mice transplanted with bone marrow from CB2� /� mice; cwt, mice transplanted with bone marrow fromCB2þ /þ mice; DM, diabetic mice; KO, CB2� /� knockout; KW/BW, kidney weight/body weight; NAG, N-acetylglucosamine activity/creatinine ratio; ND, nondiabetic mice;SBP, systolic blood pressure; SCr, serum creatinine; WT, wild type.Data are shown as means±s.e.m. or geometric means (25th–75th percentile).aPo0.001 all DM groups vs. all ND groups.bPo0.05 all DM groups vs. all ND groups.cPo0.05 DM-WT and DM-WT-cko vs. ND.dPo0.05 DM-KO and DM-KO-cwt vs. DM-WT and DM-WT-cko.



In vitro study

Glomerular CB2 deficiency is a feature of DN in humans, andwe explored in cultured podocytes the potential underlyingmechanism. Exposure of podocytes to a high glucosemilieu for 1, 2, and 3 days did not alter CB2 expression as

assessed by immunoblotting. On the contrary, exposure tomechanical stretch induced a significant reduction in CB2expression at 12 h, suggesting that a rise in intraglomerularpressure may contribute to CB2 downregulation in diabetes(Figure 8).

ND-WT ND-WT-cko

DM-WT-cko

CB2 F4/80+ Merge

DM-KO

ND-KO-cwt

DM-KO-cwt

DM-WT

WT-cko

* *

#

DMND

#

KO-cwt

KO mice50

4045

353025

Glo

mer

ular

pod

ocin

posi

tive

area

(%

)

20151050

WT mice

O

† †

§

DMND

5

44.5

3.53

2.5

Rel

ativ

e ex

pres

sion

of fi

bron

ectin

mR

NA

21.5

10.5

0

WT-ckoKO-cwt

KO miceWT mice

Figure 5 | Bone marrow (BM) transplantation study. Wild type, CB2� /� knockout (KO), and lethally irradiated recipient mice reconstitutedwith BM from either CB2� /� (chimera-KO cko) or WT (chimera-WT cwt) mice were studied 14 weeks after diabetes onset (4–6 mice per group).Podocin expression was measured in renal cortex sections by immunofluorescence. (a–g) Representative immunofluorescence images areshown (�400) and (h) quantification of glomerular staining is reported in the graph (*Po0.01 DM and DM-WT-cko vs. ND groups; #Po0.05DM-KO and DM-KO-cwt vs. DM and DM-cko). Fibronectin mRNA levels were measured by real-time PCR on total RNA extracts from therenal cortex, and (i) results were corrected for the expression of the housekeeping gene HRPT (hypoxantine phosphoribosyltransferase)(wPo0.05 DM and DM-WT-cko vs. ND groups, yPo0.05 DM-KO vs. DM-WT and DM-WT-cko; 1Po0.01 DM-KO-cwt vs. DM-WT and DM-WT-cko).In (j–l) DM-WT-cko and (m–o) DM-KO-cwt mice, double immunofluorescence for (j, m) CB2 and (k, n) the monocyte marker F4/80þ wasperformed and (l, o) merged images are shown. The insert in panel (m) shows nuclei counterstained with DAPI (4,6-diamidino-2-phenylindole).



DISCUSSION

In this study, we have provided evidence that, in experimentaldiabetes, deletion of the CB2 receptor worsens albuminuria,renal function, podocyte protein loss, overexpression of ECMcomponents, and glomerular monocyte accrual.

Body weight, blood glucose levels, glycated hemoglobin,and systolic blood pressure were similar in WT and KOdiabetic mice, indicating that the effects observed wereindependent of both metabolic and hemodynamic factors. Astudy has reported enhanced body weight and reducedinsulin resistance in CB2� /� mice with aging. However, CB2deprivation affects these parameters exclusively in mice much

older that those used in our study.17 Importantly, nodifferences in albuminuria were observed between WT andCB2� /� mice in the absence of diabetes.

In DM mice, genetic deletion of the CB2 receptor induceda significant increase in albuminuria and magnified diabetes-induced downregulation of both nephrin and podocin. Thesedata corroborate previous results, showing that treatmentwith the CB2 agonist AM1241 reduces diabetes-inducedalbuminuria and prevents nephrin loss.15 Taken together, ourcurrent and previous findings indicate that the CB2 is animportant regulator of both slit diaphragm protein expres-sion and glomerular permselectivity. Ultrastructural analysis

ND-WT

ND

ND

-KO

DM

DM

-KO

CB1 WT-1 Merge CB1 SMA Merge

ND-KO DM-WT DM-KO ND0

5

10

15

20

25 WT* *

KO

Glo

mer

ular

CB

1po

sitiv

e ar

ea (

%)

ND DM DM

Figure 6 | Glomerular CB1 expression/distribution is not altered in CB2� /� knockout mice. CB1 protein expression was assessed inrenal sections from ND-WT (n¼ 12), ND-KO (n¼ 6), DM-WT (n¼ 9), and DM-KO (n¼ 7) mice by (a) immunohistochemistry and (b) quantificationof glomerular staining reported in the graph (*Po0.01 DM-WT and DM-KO vs. ND). Double immunofluorescence for CB1 (red) and (c) either thepodocyte marker WT-1 (green) or (d) the mesangial cells’ marker a-smooth muscle actin (SMA) (green) was performed in all studiedgroups and overlapped pictures (Merge) shown (c, d). DM, diabetic; KO, CB2� /� knockout; ND, nondiabetic; WT, wild type.



showed areas of podocyte–FP fusion in DM mice; however,the extent of FP effacement and podocyte apoptosis wassimilar in WT and KO DM mice, suggesting that slitdiaphragm protein loss has a predominant role in exacer-bating proteinuria. Consistently, development of albuminuriain the presence of nephrin loss and before FP effacement andpodocyte apoptosis has been reported.18,19

CB2 deletion also worsened diabetes-induced mesangialexpansion as proven by both histological and ultrastructuralanalyses. Moreover, expression of fibronectin, type I collagen,and type IV a4 chain collagen, a specific component of the

glomerular basement membrane,20 was further enhanced inCB2� /� DM mice, possibly because of a greater rise intransforming growth factor-b1 and connective tissue growthfactor expression. An increase in serum creatinine levels wasexclusively observed in DM CB2� /� mice, indicating thatexacerbation of diabetes-induced glomerular damage had asignificant impact on renal function. It is unlikely thattubular damage was implicated as NAG activity wascomparable in diabetic KO and WT mice.

The observation that CB2� /� mice display enhancedsusceptibility to diabetes-induced renal fibrosis is in agreement

WTKO

#35

30

25R

enal

cor

tex

endo

cann

abin

oid

leve

ls20

15

10

5

0

25

20

15N

AP

E-P

LD/tu

bulin

rat

io

10

5

0

ND

ND

DM

DM

ND

ND

DM

DM

2-AG(pmol/mg)

AEA(pmol/g)

* *

ND

ND

DM

DM

ND

ND

DM

DM

ND ND DM DM ND ND DM DM

NAPE-PLD

Tubulin

ND0

0.5

FA

AH

/tubu

lin r

atio

1

1.5

2

2.5

ND DM DM

#

Tubulin

FAAH

ND0

0.51

1.5 * *

Tubulin

DAGL-α

2

DA

GL-

α/tu

bulin

rat

io

2.53

3.5

ND DM DM0

0.05

0.10

0.15

0.20

0.30WTKO

#

WTKO

WTKO

WTKO

WT

MAGL

Tubulin1.81.61.41.2

MA

GL/

tubu

lin r

atio

10.80.60.40.2

0

KOR

enal

cor

tex

PE

A a

nd O

EA

leve

ls

0.25

PEA(pmol/mg)

OEA(pmol/mg)

Figure 7 | Effect of CB2 deletion on the endocannabinoid system. Concentrations of (a) endocannabinoid (EC) (anandamide (AEA),2-arachidonoylglycerol (2-AG)) and (b) endocannabinoid-related molecules (palmitoylethanolamide (PEA) and oleoylethanolamide (OEA))were measured in the renal cortex from ND-WT (n¼ 4), ND-KO (n¼ 3), DM-WT (n¼ 5), and DM-KO (n¼ 4) mice by isotope dilution liquidchromatography–mass spectrometry. Protein expression of (c) diacylglycerol lipase-a (DAGL-a), (d) monoacylglycerol lipase (MAGL), (e) fattyacid amide hydrolase (FAAH), and (f) N-acyl phosphatidylethanolamine-specific phospholipase D (NAPE-PLD) was assessed in total renal cortexprotein extracts by immunoblotting. Tubulin was used as an internal control. Representative immunoblots and results of densitometry analysesare shown. *Po0.01 DM-WT and DM-KO vs. ND; #Po0.05 DM-KO vs. others. DM, diabetic; KO, CB2� /� knockout; ND, nondiabetic;WT, wild type.

1.41.2

CB

2/tu

bulin

rat

io(f

old

chan

ge o

ver

cont

rol)

10.80.60.40.2

0

1.41.2

CB

2/tu

bulin

rat

io(f

old

chan

ge o

ver

cont

rol)

10.80.60.40.2

0Control Stretch

*

CB2

Tubulin

24 48 72

(Hours)

TubulinCB2

Figure 8 | Effect of high glucose and mechanical stretch on CB2 receptor expression in cultured podocytes. Serum-depriveddifferentiated podocytes were exposed to either (a) high glucose concentration (25 mmol/l) for 24, 48, and 72 h or (b) mechanical stretch for12 h (black columns). Control cells in normal glucose concentration (7 mmol/l made iso-osmolar with mannitol) and in a mechanically stableenvironment were studied in parallel (white columns). Total proteins were extracted and CB2 expression assessed by immunoblotting.Densitometric analyses and representative immunoblots are shown. Results are expressed as fold change over control (n¼ 3, *Po0.05 stretchvs. control).



with previous studies demonstrating that CB2� /� miceexhibit augmented cirrhosis when exposed to CCl4,13,21

exacerbated cardiac fibrosis after ischemia/reperfusion,22

and enhanced dermal fibrosis in response to bleomycin.12

Our results were, however, relatively unexpected inexperimental diabetes as in this model treatment with theCB2 agonist AM1241 fails to ameliorate renal fibrogenesis.15

The reason of this discrepancy is unclear; however, it ispossible that the extent to which the CB2 receptor wasmodulated pharmacologically by AM1241 was not ofsufficient magnitude to elicit a protective response on renalfibrosis. In DM animals, 2-AG levels, the major CB2 ligand,are reduced, thus a greater dose of exogenous agonist may berequired to achieved specific therapeutic effects.

Diabetes enhanced glomerular monocyte infiltration, andthis feature was further exacerbated by CB2 deletion.Similarly, in bleomycin-induced dermal fibrosis, CB2deletion is associated with enhanced monocyte infiltrationwithin the skin.12 Moreover, CB2 activation attenuates bothdiabetes- and cisplatin-induced renal monocyte accrual,15,23

whereas lack of CB2 magnifies cisplatin-induced inflamma-tion and tissue injury.23 Binding of the chemokine MCP-1 tothe CCR2 receptor controls glomerular monocyte recruit-ment,3 and our data suggest that overexpression of MCP-1and CCR2, which are induced by diabetes and by CB2deficiency, respectively, may enhance MCP-1 signaling andthus monocyte recruitment. Consistently, expression of GR1,a marker highly expressed by the subpopulation of inflam-matory monocytes that are specifically recruited through theMCP-1/CCR2 system,16 was significantly increased in DMCB2� /� mice.

Enhanced glomerular monocyte infiltration is a potentialmechanism whereby the absence of CB2 could indirectlyaffect both AER and mesangial expansion, because locallyrecruited monocytes can release cytotoxic products, reactiveoxygen species, and cytokines that can alter neighboringglomerular cell phenotypes.3 On the other hand, as podocytesalso express both the CB2 and the CCR2,6,15 enhancedMCP-1 signaling can have direct deleterious effects in this celltype.6,24 To clarify which of these two alternative mechanismshas a predominant role and to explore the relative con-tributions of glomerular cells and BM-derived monocytes tothe phenotype of DM CB2� /� mice, we studied the effect ofBM transplantation in our experimental model. The studyduration was shortened, because a significant change inalbuminuria between diabetic CB2þ /þ and CB2� /�

animals was detectable at earlier time points and becausesub-lethal irradiation may affect survival and enhance therisk of concomitant infections in laboratory animals,particularly in the presence of diabetes. In DM chimericmice, the degree of glomerular monocyte infiltration waspredominantly of the donor phenotype, indicating that CB2deficiency on BM cells has a dominant role in enhancingmonocyte accrual in DM-KO mice. However, neither trans-plantation with BM from CB2þ /þ mice in DM CB2� /�

mice nor transplantation with BM from CB2� /� mice into

DM CB2þ /þ animals altered the magnitude of albuminuria,renal function loss, podocin downregulation, and fibronectinoverexpression. Therefore, BM-derived cells affected glome-rular monocyte accrual but did not have a major role inmediating the deleterious effects of CB2 deficiency. Enhancedinfiltration of inflammatory cells was by itself insufficient tomagnify the glomerular damage in this model, and residentglomerular cells appeared to have a major role. The mech-anism by which the absence of CB2 on resident glomerularcells contributes to the phenotype of DM CB2� /� mice isunknown; however, in cultured podocytes, both geneticdeletion and pharmacological blockade of CB2 lead to CCR2overexpression,15 and MCP-1 binding to CCR2 downregulatespodocyte proteins.6 Although podocytes have a relatively lowprofibrotic potential and CB2 was not expressed by mesangialcells, mesangial expansion has been reported in an animalmodel of early podocyte-specific injury,18 and CB2� /�

podocytes may affect deposition of ECM components actingon mesangial cells in a paracrine manner.

At variance with our findings, previous studies on otherexperimental models of chronic inflammatory diseases, suchas dermal fibrosis, atherosclerosis, and nerve injury,10,12,25

have shown that enhanced monocyte infiltration is thepredominant mechanism of the deleterious effects ofCB2� /� deletion. However, in these models, CB2 and CCR2are weakly expressed by resident cells and predominantlyexposed by infiltrating monocytes, whereas the oppositeoccurs within the glomeruli.

There is increasing evidence of opposing effects of the twoEC receptors in various organs, including the kidney,26–29 andwe have previously shown that not only CB2 activation15 butalso CB1 blockade can ameliorate experimental DN.8

Therefore, we explored the possibility that compensatorychanges in the local EC system could account for thephenotype of DM-KO mice. Levels of AEA and OEA weresignificantly increased in DM-KO mice likely because offatty acid amide hydrolase downregulation. Because AEApredominantly binds to CB1 receptors,30 the increase in AEAmay result in enhanced CB1 signaling, particularly in theabsence of CB2 receptors. This raises the possibility that thedeleterious effects of CB2 deprivation in our model were, atleast in part, mediated through the CB1 receptor. On the otherhand, OEA is a PPAR-a (peroxisome proliferator-activatedreceptor-a) agonist, and there is evidence of a renoprotectiveand anti-proteinuric effect of PPAR-a activation in variousanimal models of kidney injury, including DN.31–33 Further-more, in cultured podocytes, PPAR-a agonists reduce apop-tosis and increase nephrin expression.31,34 It is thus temptingto speculate that the increase in OEA levels may represent aprotective mechanism compensating for the deleterious effectsof both CB2 deletion and enhanced CB1 signaling.

Our findings may have important implications for DN inhumans. Genetic deletion of CB2 in experimental animalsmimics the marked downregulation of glomerular CB2receptors reported in patients with DN.15 The underlyingmechanism of CB2 downregulation is unclear; however, in



our study, podocyte exposure to stretch significantly reducedCB2 expression, while high glucose, which is known toincrease CB1 expression in podocytes,35 was ineffective. Thissuggests that the hemodynamic insult of glomerular hyper-tension, a well-established cause of DN progression,36 is animportant modulator of the glomerular response to CB2agonists and that strategies reducing blood pressure in theglomerular capillaries may have the additional benefit ofrestoring CB2 receptor expression. Glomerular CB2 expres-sion is downregulated in patients in an advanced stage of thecomplication, whereas a reduced CB2 signaling due todiminished 2-AG synthesis appears to be the major abnor-mality in early experimental diabetes.15 This raises the possi-bility that treatment with CB2 agonists may be beneficial inan early stage of DN, particularly in combination withrenin–angiotensin system inhibitors. In this regard, it isnoteworthy that b-caryophyllene, which is a natural CB2agonist, has been shown to ameliorate cisplatin-inducednephrotoxicity in mice.37 Further preclinical studies arerequired to assess whether the addition of a CB2 agonist tocurrent therapy with renin–angiotensin system inhibitorsmay provide additional benefits.

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

AnimalsCB2-intact C57BL6/J and CB2-deficient B6.129P2-Cnr2tm1Dgen/Jmice, backcrossed at least five generations to C57BL/6 mice, werefrom Jackson Laboratories (Bar Harbor, ME). Animals were main-tained on a normal diet under standard animal house conditions.Principles of laboratory animal care (NIH publication no. 85–23,revised 1985; http://grants1.nih.gov/grants/olaw/references/phspol.htm) were followed. Both housing and care of laboratory animalswere in accordance with Italian law (D.L.116/1992), and the studywas approved by the local Ethical Committee. Diabetes was inducedin mice aged 8 weeks by streptozotocin intraperitoneal injection aswe have previously described.15

KO studyGroups of DM-WT (n¼ 9) and DM-KO (n¼ 7) mice withequivalent blood glucose levels and ND-WT (n¼ 12) and ND-KO(n¼ 6) mice were studied in parallel. Before euthanasia, bloodsamples were collected via saphenous vein puncture on alert 4-h-fasted animals. Glucose levels were measured using a glucometer(Accu-chek; Roche, Milan, Italy) and the glycated hemoglobin byquantitative immunoturbidimetric latex determination (SentinelDiagnostic, Milan, Italy). Systolic blood pressure was assessed bytail-cuff plethysmography. Urinary albumin was measured byenzyme-linked immunosorbent assay (Bethyl Laboratories, Milan,Italy) in 18-h urine collections. Serum creatinine levels weredetermined by high-performance liquid chromatography andurinary NAG levels using a colorimetric assay (Roche). Sixteenweeks after diabetes onset, mice were killed by decapitation. Thekidneys were rapidly dissected, weighed, and processed for sub-sequent analyses.

BM transplantation studyIn this study, 6–8-week-old female CB2� /� and CB2þ /þ mice wereused as BM donors. BM cells were flushed from tibial and femurcavities under sterile conditions, filtered through 70-mm nylonmeshes (BD Biosciences, Milan, Italy), and then transplantedwithout further purification or in vitro expansion. Before trans-plantation, recipient male mice, aged 8 weeks, underwent whole-body irradiation with 8 Gy. After 24 h, postirradiated mice wereinjected with 2.0�106 BM cells via the tail vein. CB2� /� micereceived a BM transplant from CB2þ /þ mice (KO-cwt; n¼ 15) andCB2þ /þ mice from CB2� /� mice (WT-cko; n¼ 15). Diabetes wasinduced as described above, and 14 weeks thereafter mice werekilled. DNA was isolated (DNeasy Blood and Tissue Kit, Qiagen,Milan, Italy), and chimerism was confirmed by PCR. Only BM-chimeric mice that showed 480% reconstitution efficiency withdonor cells were used.

In vitro experimentsConditionally immortalized mouse podocytes were kindly providedby P. Mundel (Harvard Medical School, Boston, MA) and culturedas previously described.38 After serum deprivation (0.5% FCS) for24 h, differentiated podocytes were exposed to either a high(25 mmol/l) or normal (7 mmol/l made iso-osmolar with mannitol)glucose concentration for 24, 48, and 72 h. Cells, seeded on type Icollagen-coated silicon elastomer-base culture plates (Flex I plates)were subjected for 6 and 12 h to repeated stretch–relaxation cyclesby mechanical deformation using a Stress Unit as describedpreviously.7,39,40 Cells grown in a mechanically stable environment(Flex II plates) were used as controls.

EC measurementEC and EC-related molecules were measured by isotope-dilutionliquid chromatography–mass spectrometric analysis in lipid extractsof frozen kidney tissue samples from ND-WT (n¼ 4), ND-KO(n¼ 3), DM-WT (n¼ 5), and DM-KO (n¼ 4) mice as we describedpreviously15 and detailed in the Supplementary Information online.

Immunohistochemistry and immunofluorescenceGlomerular expression of CB1, CB2, CCR2, F4/80, cleaved caspase-3, nephrin, podocin, and fibronectin was assessed by eitherimmunohistochemistry or immunofluorescence.6,8,15 Co-stainingfor either CB1 or CB2 and specific markers of podocytes (WT-1),mesangial cells (a smooth muscle actin), and monocytes (F4/80)was performed by double immunofluorescence (SupplementaryInformation online).

ImmunoblottingExpression of nephrin, podocin, CB2, diacylglycerol lipase-a,monoacylglycerol lipase, fatty acid amide hydrolase, and N-acylphosphatidylethanolamine-specific phospholipase D was measuredin total protein extracts from either renal cortex or culturedpodocytes by immunoblotting as detailed in the SupplementaryInformation online.

mRNA analysismRNA expression was analyzed by real-time PCR using pre-developed TaqMan reagents (Applied Biosystems, Monza, Italy) asdetailed in the Supplementary Information online.



http://grants1.nih.gov/grants/olaw/references/phspol.htm

http://grants1.nih.gov/grants/olaw/references/phspol.htm

Histological and ultrastructural analysesParaffin-embedded tissue sections were stained with PAS. Mesangialarea was analyzed (percentage of glomerular area) from digitalpictures of 15–20 glomeruli per kidney per animal using theAxiovision 4.7 software (Milan, Italy).41 Glomerular collagen wasassessed by picrosirius red staining. The glomerular cross-sectionalarea was measured in 20 glomerular profiles per mouse using theAxiovision software.42 Electron microscopy was performed asdetailed in the Supplementary Information online.

Statistical analysisData, presented as mean±s.e.m. or geometric mean (25–75%percentile), were analyzed by analysis of variance. Least significantdifference test was used for post hoc comparisons. Values for Po0.05were considered significant.

DISCLOSUREAll the authors declared no competing interests.

ACKNOWLEDGMENTSThis work was supported by the Compagnia di San Paolo, the ItalianSociety of Diabetes, the University of Turin (ex-60% grant), and thePiedmont Region Research Grant.

SUPPLEMENTARY MATERIALFigure S1. The CB2 receptor is not expressed by mesangial cells.Supplementary material is linked to the online version of the paper athttp://www.nature.com/ki

REFERENCES1. Molitch ME, DeFronzo RA, Franz MJ et al. American Diabetes Association.

Nephropathy in diabetes. Diabetes Care 2004; 27: S79–S83.2. Cooper ME. Interaction of metabolic and haemodynamic factors in mediat-

ing experimental diabetic nephropathy. Diabetologia 2001; 44: 1957–1972.3. Giunti S, Barutta F, Perin PC et al. Targeting the MCP-1/CCR2 system in

diabetic kidney disease. Curr Vasc Pharmacol 2010; 8: 849–860.4. Pavenstadt H, Kriz W, Kretzler M. Cell biology of the glomerular podocyte.

Physiol Rev 2003; 83: 253–307.5. Doublier S, Salvidio G, Lupia E et al. Nephrin expression is reduced in

human diabetic nephropathy: evidence for a distinct role for glycatedalbumin and angiotensin II. Diabetes 2003; 52: 1023–1030.

6. Tarabra E, Giunti S, Barutta F et al. Effect of the monocytechemoattractant protein-1/CC chemokine receptor 2 system on nephrinexpression in streptozotocin-treated mice and human culturedpodocytes. Diabetes 2009; 58: 2109–2118.

7. Miceli I, Burt D, Tarabra E et al. Stretch reduces nephrin expression via anangiotensin II-AT(1)-dependent mechanism in human podocytes: effectof rosiglitazone. Am J Physiol Renal Physiol 2010; 298: F381–F390.

8. Barutta F, Corbelli A, Mastrocola R et al. Cannabinoid receptor 1 blockadeameliorates albuminuria in experimental diabetic nephropathy. Diabetes2010; 59: 1046–1054.

9. Di Marzo V. Endocannabinoids: synthesis and degradation. Rev PhysiolBiochem Pharmacol 2008; 160: 1–24.

10. Steffens S, Veillard NR, Arnaud C et al. Low dose oral cannabinoid therapyreduces progression of atherosclerosis in mice. Nature 2005; 434: 782–786.

11. Munoz-Luque J, Ros J, Fernandez-Varo G et al. Regression of fibrosisafter chronic stimulation of cannabinoid CB2 receptor in cirrhotic rats.J Pharmacol Exp Ther 2008; 324: 475–483.

12. Akhmetshina A, Dees C, Busch N et al. The cannabinoid receptor CB2exerts antifibrotic effects in experimental dermal fibrosis. Arthritis Rheum2009; 60: 1129–1136.

13. Julien B, Grenard P, Teixeira-Clerc F et al. Antifibrogenic role of the canna-binoid receptor CB2 in the liver. Gastroenterology 2005; 128: 742–755.

14. Calkin A, Giunti S, Jandeleit-Dahm K et al. PPAR-a and -g agonistsattenuate diabetic kidney disease in the apolipoprotein E knockoutmouse. Nephrol Dial Transplant 2006; 21: 2399–2405.

15. Barutta F, Piscitelli F, Pinach S et al. Protective role of cannabinoidreceptor type 2 in a mouse model of diabetic nephropathy. Diabetes2011; 60: 2386–2396.

16. Tacke F, Randolph GJ. Migratory fate and differentiation of bloodmonocyte subsets. Immunobiol 2006; 211: 609–618.

17. Agudo J, Martin M, Roca C et al. Deficiency of CB2 cannabinoid receptorin mice improves insulin sensitivity but increases food intake and obesitywith age. Diabetologia 2010; 53: 2629–2640.

18. Inoki K, Mori H, Wang J et al. mTORC1 activation in podocytes is a criticalstep in the development of diabetic nephropathy in mice. J Clin Invest2011; 121: 2181–2196.

19. Kalluri R. Proteinuria with and without renal glomerular podocyteeffacement. J Am Soc Nephrol 2006; 17: 2383–2389.

20. Kalluri R, Shield CF, Todd P et al. Isoform switching of type IV collagen isdevelopmentally arrested in X-linked Alport syndrome leading toincreased susceptibility of renal basement membranes toendoproteolysis. J Clin Invest 1997; 99: 2470–2478.

21. Teixeira-Clerc F, Belot MP, Manin S et al. Beneficial paracrine effects ofcannabinoid receptor 2 on liver injury and regeneration. Hepatology2010; 52: 1046–1059.

22. Defer N, Wan J, Souktani R et al. The cannabinoid receptor type 2promotes cardiac myocyte and fibroblast survival and protects againstischemia/reperfusion-induced cardiomyopathy. FASEB J 2009; 23:2120–2130.

23. Mukhopadhyay P, Rajesh M, Pan H et al. Cannabinoid-2 receptor limitsinflammation, oxidative/nitrosative stress, and cell death in nephropathy.Free Radic Biol Med 2010; 48: 457–467.

24. Burt D, Salvidio G, Tarabra E et al. The monocyte chemoattractantprotein-1/cognate CC chemokine receptor 2 system affects cell motility incultured human podocytes. Am J Pathol 2007; 171: 1789–1799.

25. Racz I, Nadal X, Alferink J et al. Crucial role of CB2 cannabinoid receptor inthe regulation of central immune responses during neuropathic pain.J Neurosci 2008; 28: 12125–12135.

26. Tam J, Liu J, Mukhopadhyay B et al. Endocannabinoids in liver disease.Hepatology 2011; 53: 346–355.

27. Pacher P, Mukhopadhyay P, Mohanraj R et al. Modulation of theendocannabinoid system in cardiovascular disease: therapeutic potentialand limitations. Hypertension 2008; 52: 601–607.

28. Pacher P, Kunos G. Modulating the endocannabinoid system in humanhealth and disease-successes and failures. FEBS J 2013; 280: 1918–1943.

29. Horvath B, Mukhopadhyay P, Hasko G et al. The endocannabinoid systemand plant-derived cannabinoids in diabetes and diabetic complications.Am J Pathol 2012; 180: 432–442.

30. Gonsiorek W, Lunn C, Fan X et al. Endocannabinoid 2-arachidonyl glycerolis a full agonist through human type 2 cannabinoid receptor: antagonismby anandamide. Mol Pharmacol 2000; 57: 1045–1050.

31. Zhou Y, Kong X, Zhao P et al. Peroxisome proliferator-activated receptor-a is renoprotective in doxorubicin-induced glomerular injury. Kidney Int2011; 79: 1302–1311.

32. Zuo Y, Yang HC, Potthoff SA et al. Protective effects of PPARg agonist inacute nephrotic syndrome. Nephrol Dial Transplant 2012; 27: 174–181.

33. Park CW, Kim HW, Ko SH et al. Accelerated diabetic nephropathy in micelacking the peroxisome proliferator-activated receptor-a. Diabetes 2006;55: 885–893.

34. Ren S, Xin C, Beck KF et al. PPARa activation upregulates nephrinexpression in human embryonic kidney epithelial cells and podocytes bya dual mechanism. Biochem Biophys Res Commun 2005; 338: 1818–1824.

35. Nam DH, Lee MH, Kim JE et al. Blockade of cannabinoid receptor 1improves insulin resistance, lipid metabolism, and diabetic nephropathyin db/db mice. Endocrinology 2012; 153: 1387–1396.

36. Giunti S, Barit D, Cooper ME. Mechanisms of diabetic nephropathy: role ofhypertension. Hypertension 2006; 48: 519–526.

37. Horvatha B, Mukhopadhyaya P, Kechrida M et al. Caryophylleneameliorates cisplatin-induced nephrotoxicity in a cannabinoid 2 receptor-dependent manner. Free Radic Biol Med 2012; 52: 1325–1333.

38. Shankland SJ, Pippin JW, Reiser J et al. Podocytes in culture: past, present,and future. Kidney Int 2007; 72: 26–36.

39. Gruden G, Thomas S, Burt D et al. Mechanical stretch induces vascularpermeability factor in human mesangial cells: mechanisms of signaltransduction. Proc Natl Acad Sci USA 1997; 94: 12112–12116.

40. Barutta F, Pinach S, Giunti S et al. Heat shock protein expressionin diabetic nephropathy. Am J Physiol Renal Physiol 2008; 295:F1817–F1824.

41. Watson AM, Gray SP, Jiaze L et al. Alagebrium reduces glomerularfibrogenesis and inflammation beyond preventing RAGE activationin diabetic apolipoprotein E knockout mice. Diabetes 2012; 61:2105–2113.

42. Weibel ER. Stereological methods. In Practical Methods for BiologicalMorphometry. Academic: London, UK, 1979 pp 51–57.



http://www.nature.com/ki

Deficiency of cannabinoid receptor of type 2 worsens renal functional and structural abnormalities...

Documents

Transcript of Deficiency of cannabinoid receptor of type 2 worsens renal functional and structural abnormalities...