Kidney International, Vol. 57 (2000), pp. 1882–1894

Effects of endothelin or angiotensin II receptor blockadeon diabetes in the transgenic (mRen-2)27 rat

DARREN J. KELLY, SANDFORD L. SKINNER, RICHARD E. GILBERT, ALISON J. COX,MARK E. COOPER, and JENNIFER L. WILKINSON-BERKA

Department of Medicine, Austin and Repatriation Medical Center, Heidelberg West, and Department of Physiology, Universityof Melbourne, Parkville, Victoria, Australia

Effects of endothelin or angiotensin II receptor blockade on betic nephropathy [1]. Angiotensin-converting enzymediabetes in the transgenic (mRen-2)27 rat. (ACE) inhibitors and angiotensin type 1 receptor antag-

Background. Endothelin (ET) and angiotensin II (Ang II) onists (AT1) appear to be particularly effective in thisare vasoactive/trophic peptides that may contribute to the pro-regard in both human and experimental diabetic nephrop-gression of diabetic nephropathy. The transgenic (mRen-2)27athy [2–4]. Micropuncture studies have indicated thatrat exhibits overexpression of Ang II at sites of normal physio-

logical expression. Unlike other rat strains, the streptozotocin- renoprotection with these agents is associated with re-induced diabetic Ren-2 rat develops progressive renal pathol- ductions in intraglomerular pressure [5, 6]. More recently,ogy associated with a declining glomerular filtration rate (GFR) the induction of proinflammatory and prosclerotic growthand provides a convenient model to evaluate the role of these

factors by hyperglycemia [7–10] and possibly tubularvasoactive peptides in the nephropathic process.protein load [11] have also been recognized as importantMethods and Results. Oral administration of either the endo-

thelin A (ETA) and ETB receptor antagonist bosentan or the contributors to the microvascular and tubulointerstitialangiotensin type 1 (AT1) receptor antagonist valsartan for 12 pathology of diabetic renal disease [12]. In various exper-weeks reduced systolic blood pressure (SBP) of nondiabetic

imental models of renal disease, including diabetes,and diabetic Ren-2 rats to normotensive levels. Diabetic renalmRNA for transforming growth factor-b1 (TGF-b1) andpathology was associated with intense renin mRNA and pro-

tein in the proximal tubules and juxtaglomerular cells along collagen IV increases in the kidney with evidence thatwith overexpression of transforming growth factor-b1 (TGF-b1) blockade of the renin-angiotensin system (RAS) de-and collagen IV mRNA in glomeruli and tubules. With valsar- creases expression of these growth factors and matrixtan but not bosentan, renin mRNA and protein in the proximal

proteins as well as renal pathology [12, 13]. This func-tubules were not detected. Valsartan but not bosentan reducedtional interrelationship between angiotensin II (Ang II)TGF-b1 and collagen IV mRNA and the severity of diabetic

renal pathology. A declining GFR with diabetes was attenuated and TGF-b may represent an important trophic/profi-by both treatments. Albuminuria in diabetic rats rose further brotic pathway involved in the progression of experimen-with bosentan but was reduced with valsartan. tal diabetic nephropathy [14].Conclusions. Despite producing normotension, severe dia-

While the experimental streptozotocin (STZ)-inducedbetic renal pathology was not prevented by bosentan, sug-diabetic rat develops proteinuria and modest expressiongesting dissociation of ET, albuminuria, and hypertension from

the structural injury in this diabetic model. The beneficial ef- of extracellular matrix (ECM), it is not associated with afects afforded by valsartan therapy strengthen the importance loss of renal function. In contrast, the diabetic transgenicof the local renin-angiotensin system in mediating progressive (mRen-2)27 rats develop severe renal impairment anddiabetic renal injury.

histologic injury, which resembles human diabetic ne-phropathy. In the transgenic (mRen-2)27 rat, the mouseRen-2 gene is inserted into the genome of a Sprague-The control of systemic and possibly intraglomerularDawley rat such that overexpression of renin occurs athypertension is a key factor in the management of dia-sites of normal physiological expression [15]. This modeldisplays hypertension and develops progressive ne-phropathy with the induction of STZ diabetes, possiblyKey words: glomerulosclerosis, renin-angiotensin system, endothelin,

diabetic nephropathy, hypertension. because of the activation of the intrarenal RAS [16], andprovides an opportunity to explore the relative impor-Received for publication June 8, 1999tance of systolic blood pressure (SBP) versus the localand in revised form October 7, 1999

Accepted for publication November 17, 1999 RAS in the progression of diabetic nephropathy [16].After an initial phase of hyperfiltration in the Ren-2 2000 by the International Society of Nephrology

1882

Kelly et al: Nephropathy in the diabetic Ren-2 rat 1883

rat, STZ diabetes results in the rapid onset of renal units intraperitoneally; Ultratard, Novo Nordisk, Den-impairment and severe renal pathology, with features mark) to promote weight gain and to reduce mortality.similar to those observed in advanced human diabetic Experimental procedures adhered to the guidelines ofnephropathy. This progressive renal disease is prevented the National Health and Medical Research Council ofby ACE inhibitor treatment [16]. Australia’s Code for the Care and Use of Animals for

Endothelin (ET), a potent proinflammatory mediator, Scientific Purposes, and were approved by the Bioethicshas also been implicated in the progression of various Committee of the University of Melbourne.renal pathologies [17]. ET has been shown to be elevated

Renal functional and biochemical studiesin the diabetic kidney [18], and ET antagonists havebeen reported to reduce growth factor expression and Rats were individually housed in metabolic cages ev-ECM accumulation in glomeruli from diabetic rats ery four weeks and habituated for two to three hours,[18, 19]. Furthermore, ET antagonists ameliorate renal and measurements of fluid intake and urine excretionpathology in several experimental models of kidney dis- rate were recorded over 24 hours. Animals continued toease [20, 21]. In the kidney, ET-1 mRNA is reduced by have free access to tap water and standard laboratoryACE inhibition [19], and in vascular tissue, the effects chow during this period. After 24 hours in metabolicof Ang II infusion are ameliorated by ET antagonism cages, an aliquot of urine (5 mL) was collected from the[22]. Together, these studies suggest a functional interre- 24-hour urine sample and stored at 2708C for subse-lationship between ET, Ang II, and structural injury. quent analysis of urinary electrolytes and albumin con-

Our current study compares an AT1 receptor antago- centration. Urinary Na1 and K1 concentrations andnist with an ETA and ETB receptor antagonist at an equi- plasma electrolytes were measured in duplicate using anpotent antihypertensive dose to explore the relative automated IL 943 flame photometer (Instrumentationinvolvement of Ang II and ET in mediating the renal Laboratories, Milan, Italy).functional and structural abnormalities observed in the Albuminuria was determined by a double antibodySTZ diabetic Ren-2 rat. radioimmunoassay, as previously performed in our labo-

ratories [23]. The glomerular filtration rate (GFR) wasdetermined by injecting a single shot of 99Tc-DTPA intoMETHODSthe tail vein and sampling the blood after 43 minutes,Animalsas previously described [24].

Six-week-old female, heterozygous Ren-2 rats, weigh-ing 125 6 5 g, were randomized to receive either 55 Plasma preparationmg/kg of STZ (Sigma, St. Louis, MO, USA) diluted in

Prior to sacrifice, plasma samples (100 mL) were col-0.1 mol/L citrate buffer, pH 4.5, or citrate buffer alonelected in heparinized vials (10 U/mL) from the tail vein(nondiabetic) by tail vein injection following an over-of conscious rats. The blood samples were centrifugednight fast. Female rats were chosen because they do notat 2000 3 g for five minutes at 48C, and the plasmarequire maintenance antihypertensive therapy. Diabeticsupernatant was removed and frozen for later estima-and nondiabetic Ren-2 (N 5 6 per group) were treatedtions of Na1 and K1.with an AT1 receptor antagonist, valsartan (average dose

of 20 mg/kg/day, in drinking water; Novartis, Basel, Swit- Anatomical studieszerland), or the ET (ETA and ETB) receptor antagonist

Rats were anesthetized (Nembutal 60 mg/kg body wtbosentan (gavaged dose of 100 mg/kg/day; Roche, Basel,intraperitoneally; Rhone Merieux, Pinkenba, QLD, Aus-Switzerland) for 12 weeks post-STZ or vehicle. Treat-tralia), and the abdominal aorta was cannulated with anment commenced within 24 hours of STZ injection. All18-gauge needle. Perfusion-exsanguination commencedrats were housed in a stable environment (maintainedat SBP (180 to 220 mm Hg) via the abdominal aortaat 22 6 18C with a 12-h light-day cycle) and allowed freewith 0.1 mol/L phosphate-buffered saline (PBS), pH 7.4access to tap water and standard rat chow containing(20 to 50 mL), to remove circulating blood, and the0.25% Na1 and 0.76% K1 (GR2; Clark-King & Co.,inferior vena cava adjacent to the renal vein was simulta-Gladesville, NSW, Australia). Each week, rats wereneously severed, allowing free flow of the perfusate.weighed, and blood glucose (nondiabetic, 4 to 8 mmol/L;After clearance of circulating blood, 4% paraformalde-diabetic, 18 to 22 mmol/L) was determined using anhyde in 0.1 mol/L phosphate buffer, pH 7.4, was perfusedAMES glucometer (Bayer Diagnostics, Melbourne, Aus-for a further five minutes (100 to 200 mL of fixative) to fixtralia). Every four weeks, SBP was recorded in preheatedthe tissues [25]. Kidneys were removed from the animal,conscious rats by tail cuff plethysmography. An averagedecapsulated, and sliced transversely. Kidneys wereSBP reading was taken from three consecutive measure-postfixed in the same fixative for two hours, routinelyments to reduce variability (Bunag 1973 ID: 572). Dia-

betic rats received a daily injection of insulin (2 to 4 processed, embedded in paraffin, and sectioned.

Kelly et al: Nephropathy in the diabetic Ren-2 rat1884

Histopathology mouse and rat renin protein have been previously pub-lished by our laboratory [25, 29]. The polyclonal mouseChanges in kidney structure were assessed in a blindedrenin protein antibody was raised against pure mouseprotocol in at least 25 randomly selected tissue sectionssubmandibular gland renin [25], and the polyclonal ratfrom each group studied. Sections were stained withrenin protein was raised in rabbits against a fusion pro-either Mayer’s hematoxylin and eosin (HE) to examinetein that was incorporated into Escherichia coli [29].cell structure, periodic acid-Schiff stain (PAS) to identifyThese antisera specifically label juxtaglomerular cells ofchanges in basement membrane architecture and glyco-the kidney, zona glomerulosa cells of the adrenal, andgen deposition, or Masson’s modified trichrome to dem-macroglial Muller cells of the retina. Renin protein anti-onstrate collagen matrix [26].sera reveal the presence of both prorenin and partiallyor fully processed active renin.Glomerulosclerotic index

From 3 mm kidney sections stained with PAS, 150 to In situ hybridization200 glomeruli from rats were examined in a masked

The 1.4 kb cDNA probe for rat renin (gift of Dr. D.J.protocol. The degree of sclerosis in each glomerulus wasCampbell, Melbourne, Australia) was cloned into pGEMsubjectively graded on a scale of 0 to 4 by the method4 (Stratagene, La Jolla, CA, USA) and linearized withof Saito et al: grade 0, normal; grade 1, sclerotic area upBam HI to produce an antisense riboprobe with T7 RNAto 25% (minimal); grade 2, sclerotic area 25 to 50%polymerase. The 945 bp cDNA probe for TGF-b1 (gift(moderate); grade 3, sclerotic area 50 to 75% (moderateof Dr. Qian, Bethesda, MD, USA) and the 600 bp cDNAto severe); and grade 4, sclerotic area 75 to 100% (severe)probe for a1(IV) collagen (gift of Dr. R. Timpl, Mar-[27]. Glomerulosclerosis was defined as glomerular base-tinsried, Germany) were also cloned into pBluescriptment membrane (GBM) thickening, mesangial hypertro-KS1 (Stratagene) and linearized with XbaI and HindIII

phy, and capillary occlusion. A glomerulosclerotic indexto produce antisense riboprobes with T7 RNA polymer-

(GSI) was then calculated using the formula: ase. In situ hybridization was performed on 4 mm paraffinsections, which were hybridized with antisense probesGSI 5 o

4

i50

Fi (i)to renin, TGF-b1, and collagen IV [12]. Tissue sectionswere dewaxed in histosol, hydrated through graded etha-

where Fi is the percentage of glomeruli in the rat with nol, and immersed in distilled water. Sections were thena given score (i). washed in 0.1 mol/L PBS, pH 7.4, and hybridized with

33P-labeled antisense and sense-specific probes (5 3 105Immunohistochemistry

cpm/25 mL hybridization buffer), which were added toThree micron sections were placed into histosol to hybridization buffer (300 mmol/L NaCl, 10 mmol/L Tris-

remove the paraffin wax, hydrated in graded ethanol, HCl, pH 7.5, 10 mmol/L Na2HPO4, pH 6.8, 5 mmol/Land immersed into tap water before being incubated for ethylenediaminetetraacetic acid (EDTA), pH 8.0, 1 320 minutes with normal goat serum (NGS) diluted 1:10 Denhardt’s solution, 0.8 mg yeast RNA/mL, 50% deion-with 0.1 mol/L PBS, pH 7.4. Sections were then incubated ized formamide, and 10% dextran sulfate), heated tofor 18 hours at 48C with specific primary antisera to renin 858C and 25 mL added to the sections. Cover slips wereprotein. Sections incubated with 1:10 NGS instead of the placed on the sections, and the slides were incubated inprimary antiserum served as the negative control. After a humidified chamber (50% formamide) at 608C for 14thorough washing with PBS (3 3 5 min changes), the to 16 hours. Slides were then washed in 2 3 standardsections were flooded with a solution of 5% hydrogen saline citrate (SSC; 0.3 mol/L NaCl, 0.33 mol/L Na3-peroxide, rinsed with PBS (2 3 5 min), and incubated C6H5O712H2O) containing 50% formamide at 508C towith biotinylated goat antirabbit IgG (Dakopatts, Glo- remove the cover slips. The slides were again washedstrup, Denmark) diluted 1:200 with PBS. Sections were with 2 3 SSC, 50% formamide for a further one hourrinsed with PBS (2 3 5 min) and incubated with an at 558C. Sections were then rinsed three times in RNaseavidin-biotin peroxidase complex (Vector, Burlingame, buffer (10 mmol/L Tris-HCl, pH 7.5, 1 mmol/L EDTA,CA, USA) diluted 1:200 with PBS. Following rinsing pH 8.0, 0.5 mol/L NaCl) at 378C and treated with 150 gwith PBS (2 3 5 min), sections were incubated with RNase A/mL in RNnase buffer for a further one hour0.05% diaminobenzidine and 0.05% hydrogen peroxide at 378C and then washed with 2 3 SSC at 558C for(Pierce, Rockford, IL, USA) in PBS, pH 7.6, for one 45 minutes. Finally, sections were dehydrated throughto three minutes, rinsed in tap water for five minutes, graded ethanol, air dried, and exposed to Kodak X-Omatcounterstained in Mayer’s hematoxylin, differentiated in Autoradiography film for four days at room temperature.Scott’s tap water, dehydrated, cleared, and mounted in Slides were coated with Ilford K5 emulsion (Ilford, 1:1Depex [28]. with distilled water), stored with desiccant at room tem-

perature for 21 days, developed in Kodak D19, fixed inThe nature and specificities of the antisera raised to

Kelly et al: Nephropathy in the diabetic Ren-2 rat 1885

Table 1. Body weight, systolic blood pressure, and plasma glucose of nondiabetic and diabetic transgenic (mRen-2)27 treated for 12 weekswith either valsartan or bosentan

Plasma mmol/L Urine lmol/minBody weight Systolic BP

g mm Hg Glucose Na1 K1 Na1 K1

Nondiabetic 29169 21969 5.260.3 14063 4.460.2 1.3160.3 2.3760.4Nondiabetic 1 V 30167 13366c 6.360.5 14163 4.060.2 1.0460.2 1.8260.2Nondiabetic 1 B 33267 13964c 5.960.6 13864 4.060.2 1.1860.1 2.0360.5Diabetic 21868b 23166 21.760.9b 14362 6.860.3a 2.6460.4a 4.4360.4a

Diabetic 1 V 243614b 13467c 21.460.7b 14064 4.760.2a 3.3060.5a 7.2360.7a,c

Diabetic 1 B 249614b 12967c 19.660.6b 14164 5.460.2a 2.6160.57a 3.3760.71a

Values are expressed as means 6 SEM; N 5 6 rats per group. Abbreviations are: V, valsartan; B, bosentan.aP , 0.05 when compared with respective nondiabetic ratsbP , 0.01 when compared with respective nondiabetic ratscP , 0.05 when compared with respective untreated rats

Table 2. Albuminuria and glomerular filtration rate in nondiabeticIlford Hypam, and stained with HE. For quantitation ofand diabetic transgenic (mRen-2)27 rats treated for 12 weeks with

in situ hybridization, images of kidney sections were either valsartan or bosentancaptured and digitized using a Fujix HC-2000 digital cam-

Albuminuria GFRera (Fuji, Tokyo, Japan). Regional gene expression was mg/24 h mL/minquantitatively measured using an image analysis system

Nondiabetic 0.3060.10 3.3560.1(AIS Imaging, Ontario, Canada), whereby the propor- Nondiabetic 1 V 0.4260.08b 3.2160.1

Nondiabetic 1 B 2.9060.88b 3.6460.2tional area occupied by autoradiographic grains wasDiabetic 2.4260.20a 1.5660.1a

measured against the total area of section from a mini- Diabetic 1 V 1.2060.10a,b 2.4560.1a,b

mum of three sections per animal (N 5 6 animals per Diabetic 1 B 6.1760.4a,b 2.3260.1a,b

group), as previously described [30]. Values are expressed as means 6 SEM; N 5 6 rats per group. V, valsartan;B, bosentan; GFR, glomerular filtration rate.

aP , 0.05 when compared with respective nondiabetic ratsStatistics bP , 0.05 when compared to respective untreated rats

Data are expressed as means 6 SEM unless otherwisestated. Statistical significance was determined by a two-way analysis of variance (ANOVA) with a Fishers post

played the largest increase in urinary electrolyte outputhoc comparison. Albuminuria was analyzed using log(P , 0.05; Table 1).transformed data and a represented as geometric means

Renal parameters. Diabetes was associated with an3/4 tolerance factors. Analyses were performed using

increase in urinary albumin excretion. Bosentan treat-Statview II 1 Graphics package (Abacus Concepts,ment increased albumin excretion further in both nondi-

Berkeley, CA, USA) on an Apple Macintosh G3 com-abetic and diabetic rats, while valsartan treatment was

puter (Apple Computer, Inc., Cupertino, CA, USA). A associated with reduced albumin excretion in diabeticP value , 0.05 was regarded as statistically significant. rats (Table 2). The impairment in GFR observed in dia-

betic Ren-2 rats was attenuated by all treatment regi-mens, although the GFR was still significantly lower thanRESULTSthe respective drug-treated nondiabetic rats (Table 2).Renal functional and biochemical studies

Body weight, systolic blood pressure and plasma glu- Anatomical studiescose. Diabetic rats had reduced body weight indepen- In the kidney cortex of nondiabetic rats, there wasdent of treatment (P , 0.01). SBP was decreased to a no significant renal pathology (Fig. 1A). However, insimilar level within one week of commencement of either diabetic rats, most glomeruli displayed thickened GBM,valsartan or bosentan in diabetic rats (P , 0.01; Table mesangial expansion, and capillary occlusion, as well as1). Plasma glucose was elevated to a similar extent in all efferent arteriole hyalinization and tubular vacuolationdiabetic rat groups, independent of treatment (P , 0.01; (Fig. 1B). In both nondiabetic and diabetic rats treatedTable 1). with valsartan (Fig. 1 C, D) and nondiabetic rats treated

Plasma and urinary electrolytes. In drug-treated dia- with bosentan (Fig. 1E), only minimal evidence of GBMbetic rats, plasma K1 was increased compared with drug- thickening was apparent. Diabetic rats treated with bo-treated nondiabetic controls (P , 0.05). Urinary electro- sentan (Fig. 1F) displayed severe glomerulosclerosis andlyte excretion was increased in diabetic rats, irrespective tubulointerstitial disease.

In the medulla of nondiabetic and diabetic rats treatedof drug treatment. Valsartan-treated diabetic rats dis-

Kelly et al: Nephropathy in the diabetic Ren-2 rat1886

Fig. 1. Histopathology of the kidney cortex in non-diabetic and diabetic transgenic (mRen-2)27 treated for 12 weeks with either valsartan or bosentan.Sections are stained with PAS. G, glomerulus. (A) In nondiabetic rats, there is minimal evidence of glomerulosclerosis. (B) Diabetic rats havesevere glomerulosclerosis, arteriolar hyalinization (*). Valsartan-treated nondiabetic (C ) and diabetic rats (D) and bosentan-treated nondiabeticrats (E ) have minimal evidence of glomerulosclerosis. Bosentan-treated diabetic rats (F ) have severe glomerulosclerosis and cortical interstitialfibrosis (arrow, magnification 3300). Reproduction of this figure in color was made possible by Novartis Pharmaceuticals, Bern, Switzerland.

with valsartan (Fig. 2 C, D) and nondiabetic rats treated rats treated with either valsartan or bosentan (data notshown) and diabetic rats treated with valsartan displayedwith bosentan (Fig. 2E), there was no evident pathology.

However, in diabetic rats (Fig. 2B) and diabetic rats labeling for renin mRNA (Fig. 3C) and protein (Fig.4C) in JGA only. Diabetic rats treated with bosentantreated with bosentan (Fig. 2F), there was florid deposi-

tion of collagen matrix, tubular dilation and degenera- displayed intense labeling for renin mRNA (Fig. 3D)and protein (Fig. 4D) in both JGA and proximal tubules.tion, and the presence of inflammatory cells.

Renin mRNA and immunohistochemistry Transforming growth factor-b1 and collagen IV mRNA

In nondiabetic rats, TGF-b1 (Fig. 5 A, B) and collagenIn nondiabetic rats, some juxtaglomerular apparatuses(JGAs) were positive for renin mRNA (Fig. 3A) or pro- IV (Fig. 6 A, B) mRNA were barely detected in the

kidney, while in diabetic rats, TGF-b1 (Fig. 5 C, D) andtein (Fig. 4A), whereas with diabetes, most JGAs andproximal tubules displayed intense labeling for renin collagen IV (Fig. 6 C, D) mRNA were increased (Fig.

7) and associated with sclerotic glomeruli and regions ofmRNA (Fig. 3B) and protein (Fig. 4B). Nondiabetic

Kelly et al: Nephropathy in the diabetic Ren-2 rat 1887

Fig. 2. Histopathology of the kidney medulla in nondiabetic and diabetic transgenic (mRen-2)27 treated for 12 weeks with either valsartan or bosentan.Sections are stained with Masson’s Trichrome. VR, vasa recta. (A) Nondiabetic rats have no evidence of medullary pathology, whereas (B) diabeticrats have dilated tubules, degenerated vasa recta bundles, and the presence of inflammatory cells (arrow), and there appeared to be more collagen(stained blue). In valsartan-treated nondiabetic (C ) and diabetic rats (D) and bosentan-treated nondiabetic rats (E ), there is no evidence ofmedullary pathology. Bosentan-treated diabetic rats (F ) have dilated tubules, degenerated vasa recta bundles and the presence of inflammatorycells (arrow), and there appears to be more collagen (stained blue) throughout the interstitium (magnification 3300). Reproduction of this figurein color was made possible by Novartis Pharmaceuticals, Bern, Switzerland.

tubulointerstitial disease. In valsartan-treated nondia- nificantly lower than in diabetic rats (P , 0.05; Tablebetic (not shown) and diabetic rats, the levels of TGF-b1 3). The GSI in diabetic rats treated with bosentan was(Fig. 5 E, F) and collagen IV (Fig. 6 E, F) mRNA were not significantly reduced when compared with diabeticsimilar to nondiabetic rats (Fig. 7). In bosentan-treated rats (Table 3), but there were fewer severely scleroticnondiabetic rats, both TGF-b1 and collagen IV mRNA glomeruli.(data not shown) were similar to nondiabetic rats (Fig.7). In diabetic rats treated with bosentan, TGF-b1 (Fig.

DISCUSSION5 G, H), and collagen IV (Fig. 6 G, H), mRNA wasBlockade of the RAS with AT1 antagonists or ACElocalized to glomeruli and areas of tubulointerstitial dis-

inhibitors has been shown previously to improve glomer-ease, similar to that seen in diabetic rats (Fig. 7).ular hemodynamics and structure in human and rodent

Glomerulosclerotic index diabetic nephropathy [2, 4, 5, 31, 32]. The present findingsprovide further evidence for a direct causal relationshipDiabetes was associated with an increase in the GSI.

The GSI in diabetic rats treated with valsartan was sig- between the activity of the intrarenal RAS and the pro-

Kelly et al: Nephropathy in the diabetic Ren-2 rat1888

Fig. 3. Renin mRNA of the kidney in nondiabetic and diabetic transgenic (mRen-2)27 treated for 12 weeks with either valsartan or bosentan.Sections are stained with HE. (A) Nondiabetic rats displayed renin mRNA (arrow) in the JGA, whereas in (B) rats with diabetes there was intenserenin mRNA (arrow) in both the JGA and the proximal tubules. In diabetic rats treated with valsartan (C ), renin mRNA (arrow) was detectedin the JGA only. In diabetic rats treated with bosentan (D), gene expression for renin mRNA (arrow) was present in both JGA and the cytoplasmof proximal tubules (magnification 3300). Reproduction of this figure in color was made possible by Novartis Pharmaceuticals, Bern, Switzerland.

gression of diabetic nephropathy. As with ACE inhibi- present study, hypertension in the Ren-2 rat appeared tobe partly ET dependent, yet blockade of the ET receptortion in our previous study [16], treatment with the AT1

antagonist valsartan reduced SBP, ameliorated the se- failed to attenuate the diabetic renal pathology. There-fore, it appears that the vasoconstrictor effects of hor-vere pathology, and improved the renal function of dia-

betic Ren-2 rats (Table 4). By contrast, bosentan, al- mones such as ET and Ang II can be functionally sepa-rated from their actions in the microvascular tissues.though normalizing blood pressure and improving GFR,

had little or no effect on the florid renal pathology and Indeed, agents such as AT1 antagonists or ACE inhibi-tors can confer renoprotective benefits without majorwas associated with increased albuminuria (Table 4).

Bosentan reduced SBP to normotensive levels in both effects of systemic blood pressure, as has been observedin studies performed in normotensive diabetic rats [3].nondiabetic and diabetic Ren-2 rats, suggesting that the

hypertension of this Ren-2 strain is ET dependent. These Despite only a modest reduction in SBP, agents thatinterrupt the RAS confer renoprotection in diabetes [5].findings are in accordance with other investigators who

have shown that the hypertension in the Ren-2 rat is The possibility of renoprotection independent of effectson SBP with agents that interrupt the RAS has alsoprobably mediated by locally generated Ang II and ET

[33]. This is in accordance with experimental data, which been suggested in a range of clinical studies of diabeticnephropathy [4].have shown that the chronic pressor properties of infused

Ang II in normal rats are blunted by bosentan [34]. In the The finding that bosentan did not reduce renal pathol-

Kelly et al: Nephropathy in the diabetic Ren-2 rat 1889

Fig. 4. Renin immunohistochemistry of the kidney in nondiabetic and diabetic transgenic (mRen-2)27 treated for 12 weeks with either valsartanor bosentan. Sections are counterstained with Mayer’s hematoxylin. G, glomerulus. (A) Nondiabetic rats displayed renin protein (arrow) in theJGA, whereas with (B) diabetic animals there was intense renin protein (arrow) in both the JGA and the proximal tubules. (C ) In diabetic ratstreated with valsartan, renin protein (arrow) was detected in the JGA only. (D) In diabetic rats treated with bosentan, renin protein (arrow) wasdetected in both the JGA and the cytoplasm of proximal tubules (magnification 3300).

ogy in the diabetic Ren-2 contrasts with other reports This may explain the improvement in GFR in diabeticRen-2 rats treated with bosentan. A discrepancy betweenof both hemodynamic and pathological benefits with the

ET blockade in various forms of renal disease [20, 21]. the effects of ET receptor antagonism on glomerularfunction and structural injury has also been reportedIn other models of experimental diabetes, treatment with

another nonselective ET antagonist has reduced both in a model of cyclosporine nephrotoxicity [36]. Otherinvestigators have shown a correlation between GFRalbuminuria and SBP [21], whereas in the present study,

ET antagonism decreased SBP but increased albumin- and occluded glomeruli in diabetes [37, 38]. This im-provement in renal function could relate to the knownuria. One must be cautious when comparing the two

studies since the present studies were performed in trans- beneficial effects of bosentan on glomerular and systemichemodynamics [39]. ET-1 has been reported to increasegenic rats that develop renal failure. In the diabetic

Ren-2 rat, it is possible that the disease is predominantly intraglomerular pressure with ET blockade lowering glo-merular pressure [40]; however, this is not a universalAng II dependent, although kidney and plasma Ang II

levels actually decrease with diabetes [35]. The results finding, as a study found that ET can reduce intraglomer-ular pressure, albeit at high doses [41]. ET-1 inducesof the present study suggest that local cell-specific auto-

crine/paracrine mechanisms may be more relevant than mesangial cell contraction [42], and therefore, ET block-ade would be expected to increase the glomerular filtra-whole tissue peptide levels. Indeed, in the present study,

there was evidence of a local increase in renin in proximal tion surface area through mesangial cell relaxation. Thiseffect on the glomerular barrier may explain the in-tubules of diabetic rat kidneys consistent with specific

activation of the RAS within the kidney. creased albuminuria associated with bosentan in theRen-2 rats, irrespective of the presence of diabetes. SinceAlthough bosentan did not improve the overall GSI,

there were significantly fewer severely sclerotic (grades albumin occurs as a result not only of the glomerulusbut also tubular factors, one cannot exclude the effects3 and 4) glomeruli when compared with untreated rats.

Fig. 5. Transforming growth factor-b1 mRNA in the kidney in nondiabetic and diabetic transgenic (mRen-2)27 treated for 12 weeks with eithervalsartan or bosentan. Sections were stained with HE. (A, C, E, and G ) Bright-field micrographs. (B, D, F, and H ) Corresponding dark-fieldmicrographs. In both nondiabetic (A and B) and diabetic (E and F) rats treated with valsartan, there was minimal gene expression for TGF-b1detected. However, in both diabetic (C and D) and diabetic rats treated with bosentan (G and H), there was intense gene expression for TGF-b1in the glomerulus (G) and regions of tubulointerstitial injury (magnification 3300).

Kelly et al: Nephropathy in the diabetic Ren-2 rat 1891

Fig. 6. Collagen IV mRNA in the kidney in nondiabetic and diabetic transgenic (mRen-2)27 treated for 12 weeks with either valsartan or bosentan.Sections were stained with HE. (A, C, E, and G) Bright-field micrographs. (B, D, F, and H) Corresponding dark-field micrographs. In both nondiabetic(A and B) and diabetic rats treated with valsartan (E and F), there was minimal gene expression for collagen IV detected. However, in bothdiabetic (C and D) and diabetic rats treated with bosentan (G and H), there was intense gene expression for collagen IV in the glomerulus (G)and regions of tubulointerstitial injury (magnification 3600).

Kelly et al: Nephropathy in the diabetic Ren-2 rat1892

Fig. 7. Quantitation of gene expression for TGF-b1 (A) and collagen IV (B). Values are expressed as means 6 SEM (N 5 6). *P , 0.05 whencompared with respective nondiabetic rats; #P , 0.05 when compared with untreated diabetic rats.

Table 3. Glomerulosclerotic index and the percentage of glomerulosclerosis in kidneys from nondiabetic and diabetic transgenic(mRen-2)27 rats treated for 12 weeks with either valsartan or bosentan

%

GSI Grade 0 Grade 1 Grade 2 Grade 3 Grade 4

Nondiabetic 0.8060.10 4465 2964 2463 361 0Nondiabetic 1 V 0.2460.06b 7666 2466 0 0 0Nondiabetic 1 B 0.4260.18 67612 2569 663 260.6 0Diabetic 1.7560.10a 1967 2462 2862 21 63a 862a

Diabetic 1 V 1.0760.21a,b 21610 52611 1565 10 66a,b 261a,b

Diabetic 1 B 1.6060.3a 3066 3965 2062 8.763a,b 2.360.7a,b

Values are expressed as means 6 SEM; N 5 6 rats per group.aP , 0.05 when compared with respective nondiabetic ratsbP , 0.05 when compared with respective untreated rats

Table 4. Summary of results

Control Diabetic Diabetic (V) Diabetic (B)

SBP → → ↓↓ ↓↓Albuminuria → ↑ ↓ ↑↑Serum K1 → ↑ → ↑GFR → ↓↓ ↓ ↓Histopathology

Cortex → ↑↑ ↑ ↑↑Medulla → ↑↑ → ↑↑

Renin mRNAJGA → ↑↑ ↑↑ ↑PCT N/D ↑↑ N/D ↑↑

Renin proteinJGA → ↑↑ ↑↑ ↑PCT N/D ↑↑ N/D ↑↑TGF-b1 mRNA → ↑↑ → ↑↑Collagen IV mRNA → ↑↑ → ↑↑Abbreviations are: V, valsartan; B, bosentan; GFR, glomerular filtration rate; SBP, systolic blood pressure; JGA, juxtaglomerular cells; PCT, proximal tubules;

ND, not detected.

Kelly et al: Nephropathy in the diabetic Ren-2 rat 1893

donation of the transgenic rats required to establish our existing colony,of ET antagonists on tubular reabsorption of albumin,and Ms. Bronwyn Rees and Mr. Matthew Waldron for expert technical

an area not previously investigated in detail. assistance. Reproduction of Figures 1 to 3 in color was made possibleby Novartis Pharmaceuticals, Bern, Switzerland.We have previously hypothesized from findings with

the ACE inhibitor perindopril in this model that theReprint requests to Darren J. Kelly, Ph.D., Department of Medicine,

appearance of renin in proximal tubules represents a Austin and Repatriation Medical Center (Repatriation Campus), Heidel-berg West, Victoria, Australia, 3081.potential pathogenic mechanism for tubulointerstitialE-mail: dkelly@austin.unimelb.edu.audisease [16]. As with perindopril, valsartan treatment of

diabetic Ren-2 prevented the appearance of renin inREFERENCESproximal tubular cells and the progression of surround-

ing fibrosis and inflammation. On the other hand, bosen- 1. Cooper ME: Pathogenesis, prevention, and treatment of diabeticnephropathy. Lancet 352:213–219, 1998tan neither prevented the appearance of tubular renin

2. Anderson S, Rennke HG, Garcia DL, Brenner BM: Short andnor the development of interstitial disease. The appear- long term effects of antihypertensive therapy in the diabetic rat.

Kidney Int 36:526–536, 1989ance of tubular renin appears to indicate activation of a3. Allen TJ, Cao Z, Youssef S, Hulthen UL, Cooper ME: Thenormally quiescent cellular RAS, which may then precip-

role of angiotensin II and bradykinin in experimental diabeticitate a cascade of Ang II-dependent prosclerotic and nephropathy: Functional and structural studies. Diabetes 46:1612–

1618, 1997proinflammatory growth factors such as TGF-b [13]. Evi-4. Lewis ES, Hunsicker LG, Bain RP, Rohde RD: The effect ofdence of synthesis of components of the local RAS in

angiotensin converting enzyme inhibition on diabetic nephropathy.the tubule is confirmed by the presence of increased N Engl J Med 329:1456–1462, 1993

5. Zatz R, Dunn R, Meyer TW, Anderson S, Rennke HG, Brennerrenin mRNA within the tubules of untreated diabeticBM: Prevention of diabetic glomerulopathy by pharmacologicaland bosentan-treated diabetic Ren-2 rats.amelioration of glomerular capillary hypertension. J Clin Invest

Increased glomerular and tubulointerstitial expression 77:1925–1930, 19866. Anderson S, Jung FF, Ingelfinger JR: Renal renin-angiotensinof TGF-b1 and collagen IV mRNA in the diabetic and

system in diabetes: Functional, immunohistochemical, and molecu-bosentan-treated diabetic Ren-2 rats may represent alar biological correlations. Am J Physiol 265:F477–F486, 1993

pathogenic mechanism that involves amplification of the 7. Han DC, Isono M, Hoffman BB, Ziyadeh FN: High glucosestimulates proliferation and collagen type I synthesis in renal corti-glomerular and tubular RAS with subsequent growthcal fibroblasts: Mediation by autocrine activation of TGF-beta? Jfactor induction, resulting in the development of renalAm Soc Nephrol 10:91891–91899, 1999

pathology. Elevated glucose also enhances the sensitivity 8. Rocco M, Chen Y, Goldfarb S, Ziyadeh FN: Elevated glucosestimulates TGF-beta gene expression and bioactivity in proximalof Ang II in the kidney [9], as well as increasing TGF-btubule. Kidney Int 41:107–114, 1992and ECM [43]. Studies in a proximal tubule cell line

9. Arima S, Ito S, Omata K, Takeuchi K, Abe K: High glucosehave suggested that glucose per se may, in certain con- augments angiotensin II action by inhibiting NO synthesis in in

vitro microperfused rabbit afferent arterioles. Kidney Int 48:683–texts, itself activate the RAS [44], providing a further689, 1995mechanism for a diabetic induced amplification of Ang

10. Sharma K, Ziyadeh FN: Hyperglycaemia and diabetic kidney dis-II-mediated tissue injury. Elevated glucose and local Ang ease: The case for transforming growth factor-beta as a key media-

tor. Diabetes 44:1139–1146, 1998II may then act synergistically to increase TGF-b1 and11. Benigni A, Remuzzi G: Mechanism of progression of renal disease-collagen IV, thereby accelerating renal damage further.

growth factors and related mechanisms. J Hypertens 16(Suppl):S9–In summary, this study has reported a difference in S12, 1998

12. Gilbert RE, Cox A, Wu LL, Allen TJ, Hulthen L, Jerums G,the renal effects between AT1 and ETA&B receptor block-Cooper ME: Expression of transforming growth factor-beta andade in this model of progressive diabetic nephropathy,type IV collagen in the renal tubulointerstitium in experimental

despite equivalent reduction in blood pressure. These diabetes: Effects of angiotensin converting enzyme inhibition. Dia-betes 47:414–422, 1998findings highlight a dissociation of hypertension from

13. Wu L, Roe C, Cox A, Dziadek M, Cooper ME, Gilbert RE:diabetic renal disease and provide strong evidence forTransforming growth factor beta 1 and renal injury following subto-

intrarenal RAS, particularly in the proximal tubules, to tal nephrectomy in the rat: Role of the renin-angiotensin system.Kidney Int 51:1553–1567, 1997mediate tissue injury in the diabetic Ren-2 rat. This

14. Kagami S, Border WA, Miller DE, Noble NA: Angiotensin IIwould ultimately result in the glomerular and tubuloin-stimulates extracellular matrix protein synthesis through induction

terstitial disease observed in this model, possibly via the of transforming growth factor-beta expression in rat glomerularmesangial cells. J Clin Invest 93:2431–2437, 1994prosclerotic cytokine TGF-b1. This leads to ECM accu-

15. Mullins JJ, Peters J, Ganten D: Fulminant hypertension in trans-mulation, glomerulosclerosis, tubulointerstitial disease,genic rats harbouring the mouse Ren-2 gene. Nature 344:541–544,

and renal failure. 199016. Kelly DJ, Wilkinson-Berka JL, Cooper ATJ, ME, Skinner SL:

A new model of progressive diabetic renal impairment in the trans-ACKNOWLEDGMENTSgenic (mRen-2)27 rat. Kidney Int 54:343–352, 1998

17. Ruiz Ortega M, Gomez Garre D, Alcazar R, Palacios I, BustosThis project was supported by grants from the National Healthand Medical Research Council of Australia, the Juvenile Diabetes C, Gonzalez S, Plaza J, Gonzalez E, Egido J: Involvement of

angiotensin II and endothelin in matrix protein production andFoundation International, and the Baxter Extramural Grant Program.Dr. Richard Gilbert is a recipient of a Career Development Award renal sclerosis. J Hypertens S51–S58, 1994

18. Fukui M, Nakamura T, Ebihara I, Osada S, Tomino Y, Masakifrom the Juvenile Diabetes Foundation International. The authorsthank Professor Detlev Ganten and Ms. Ursula Ganten for their kind T, Goto K, Furuichi Y, Koide H: Gene expression for endothelins

Kelly et al: Nephropathy in the diabetic Ren-2 rat1894

and their receptors in glomeruli of diabetic rats. J Lab Clin Med Doyle AE: Enalapril retards albuminuria and glomerular base-ment membrane thickening in experimental diabetic nephropathy.122:149–156, 1993Diabetologia 32:326–328, 198919. Fukui M, Nakamura T, Ebihara I, Makita Y, Osada S, Tomino

32. Cooper ME, Allen TJ, O’Brien RC, Papazoglou D, Clarke B,Y, Koide H: Effects of enalapril on endothelin-1 and growth factorJerums G, Doyle AE: Nephropathy in a model combining geneticgene expression in diabetic rat glomeruli. J Lab Clin Med 123:763–hypertension with experimental diabetes: Enalapril versus hydra-768, 1994lazine and metoprolol therapy. Diabetes 39:1575–1579, 199020. Benigni A, Zoja C, Corna D, Orisio S, Longaretti L, Bertani

33. Gardiner SM, March PA, Kemp JJ, Mullins JJ, Bennett T:T, Remuzzi G: A specific endothelin subtype A receptor antagonistHaemodynamic effects of losartan and the endothelin antagonist,protects against injury in renal disease progression. Kidney IntSB 209670, in conscious, transgenic ((mRen-2)27) hypertensive44:440–444, 1993rats. Br J Pharmacol 2244:2237, 199521. Benigni A, Colosio W, Brena C, Bruzzi I, Bertani T, Remuzzi

34. Balakrishnan SM, Wang HD, Gopalakrishnan V, Wilson TW,G: Unselective inhibition of endothelin receptors reduces renalMcNeil JR: Effect of an endothelin antagonist on hemodynamicdysfunction in experimental diabetes. Diabetes 47:450–456, 1998responses to angiotensin II. Hypertension 28:806–809, 199622. Moreau P, d’Uscio LV, Shaw S, Takase H, Barton M, Luscher

35. Campbell DJ, Kelly DJ, Wilkinson-Berka JL, Cooper ME, Skin-TF: Angiotensin II increases tissue endothelin and induces vascularner SL: Increased bradykinin and normal angiotensin peptide levelshypertrophy: Reversal by ETA-receptor antagonist. Circulationin streptozotocin-diabetic Sprague-Dawley and TGR (mRen-2)2796:1593–1597, 1997rats. Kidney Int 56:211–221, 199923. Soulis T, Cooper ME, Vranes D, Bucala R, Jerums G: The

36. Hunley TE, Fogo A, Iwasaki S, Kon V: Endothelin A receptoreffects of aminoguanidine in preventing experimental diabetic ne-mediates functional but not structural damage in chronic cyclospo-phropathy are related to duration of treatment. Kidney Int 50:627–rine nephrotoxicity. J Am Soc Nephrol 5:1718–1723, 1995634, 1996

37. Harris RD, Steffes MW, Bilous RW, Sutherland DE, Mauer24. Allen TJ, Cooper ME, O’Brien RC, Bach LA, Jackson B, JerumsSM: Global glomerular sclerosis and glomerular arteriolar hyalino-G: Glomerular filtration rate in the streptozotocin diabetic rat:sis in insulin dependent diabetes. Kidney Int 40:107–114, 1991The role of exchangeable sodium, vasoactive hormones and insulin 38. Osterby R, Parving HH, Nyberg G, Hommel E, Jorgensen HE,therapy. Diabetes 38:1182–1190, 1990 Lokkegaard H, Svalander C: A strong correlation between glo-25. Berka JLA, Alcorn D, Ryan G, Skinner SL: Renin processing merular filtration rate and filtration surface in diabetic nephropa-

studied by immunogold localisation of prorenin and renin in granu- thy. Diabetologia 31:265–270, 1988lar juxtaglomerular cells in mice treated with enalapril. Cell Tissue 39. Calo G, Gratton JP, Telemaque S, Orleans-Juste PD, RegoliRes 268:141–148, 1992 D: Pharmacology of endothelins: Vascular preparations for study-

26. Bradbury P, Gordon KC, Cook HC: Connective tissues and stains, ing ETA and ETB receptors. Mol Cell Biochem 154:31–37, 1996in Theory and Practice of Histological Techniques (vol 1, 2nd ed), 40. King AJ, Brenner BM, Anderson S: Endothelin: A potent renaledited by Bancroft JD, Stevens A, London, Churchill Living- and systemic vasoconstrictor peptide. Am J Physiol 256:F1051–stone, 1982, pp 122–216 F1058, 1989

27. Saito T, Sumithran E, Glasgow EF, Atkins RC: The enhance- 41. Lin H, Sangmal M, Smith MJ, Young DB: Effect of endothelin-1ment of aminonucleoside nephrosis by the co-administration of on glomerular hydraulic pressure and renin release in dogs. Hyper-protamine. Kidney Int 32:691–699, 1987 tension 21:845–851, 1993

28. Rong P, Berka JL, Kelly DJ, Alcorn D, Skinner SL: Renin 42. Dlugosz JA, Munk S, Zhou XP, Whiteside CI: Endothelin-1-processing and secretion in adrenal and retina of transgenic induced mesangial cell contraction involves activation of protein(mRen-2)27 rats. Kidney Int 46:1583–1587, 1994 kinase c-alpha-delta, and -epsilon. Am J Physiol 44:423–432, 1998

29. Campbell DJ, Valentijn AJ, Berka JL: Production of rat renin 43. Ayo SH, Radnik RA, Glass WFD, Garoni JA, Rampt ER, Ap-fusion protein in Escherichia coli and the preparation of renin- pling DR, Kreisberg JI: Increased extracellular matrix synthesisspecific antisera. Mol Cell Endocrinol 73:83–91, 1990 and mRNA in mesangial cells grown in high-glucose medium. Am

30. Gilbert RE, Vranes D, Berka JL, Kelly DJ, Cox A, Wu LL, J Physiol 260:F185–F191, 1991Stacker SA, Cooper ME: Vascular endothelial growth factor and 44. Zhang SL, Filep JG, Hohman TC, Tang SS, Ingelfinger JR,its receptors in control and diabetic rat eyes. Lab Invest 78:1–11, Chan JSD: Molecular mechanisms of glucose action on angioten-1998 sinogen gene expression in rat proximal tubular cells. Kidney Int

55:454–464, 199931. Cooper ME, Allen TJ, Macmillan PA, Clarke BA, Jerums G,