Targeting TGF-b overexpression in renal disease: Maximizing theantifibrotic action of angiotensin II blockade

HARM PETERS, WAYNE A. BORDER, and NANCY A. NOBLE

Division of Nephrology, University of Utah School of Medicine, Salt Lake City, Utah, USA

Targeting TGF-b overexpression in renal disease: Maximizingthe antifibrotic action of angiotensin II blockade.

Background. Overproduction of transforming growth factor-b(TGF-b) is a key mediator of extracellular matrix accumulation infibrotic diseases. We hypothesized that the degree of reduction ofpathological TGF-b expression can be used as a novel index of theantifibrotic potential of angiotensin II (Ang II) blockade in renaldisease.

Methods. One day after induction of Thy 1.1 glomerulonephri-tis, rats were treated with increasing doses of the Ang I convertingenzyme (ACE) inhibitor enalapril and/or the Ang II receptorblocker losartan in the drinking water. Six days after diseaseinduction the therapeutic effect on glomerular TGF-b overexpres-sion was evaluated.

Results. Both enalapril and losartan reduced TGF-b overpro-duction in a dose-dependent manner, showing a moderate reduc-tion at doses known to control blood pressure in renal forms ofhypertension. A maximal reduction in TGF-b expression ofapproximately 45% was seen for both drugs starting at 100mg/liter enalapril and 500 mg/liter losartan, with no furtherreduction at doses of enalapril up to 1000 mg/liter or losartan upto 2500 mg/liter. Co-treatment with both drugs was not superior tosingle therapy. Consistent with our hypothesis that reduction inTGF-b expression is a valid target, other disease measures,including glomerular matrix accumulation, glomerular productionand mRNA expression of the matrix protein fibronectin and theprotease inhibitor plasminogen-activator-inhibitor type 1 (PAI-1)closely followed TGF-b expression.

Conclusions. The data suggest that these therapies act throughvery similar pathways and that, in order to more effectively treatrenal fibrosis, these drugs must be combined with other drugs thatact by different mechanisms.

Progressive accumulation of extracellular matrix is amain cause of chronic renal disease in both animals andhumans [1]. Fibrosis of the glomerulus and the tubuloint-erstitium impairs the kidney function and finally leads to

organ failure. Over the last years considerable evidence hasaccumulated showing that overproduction of the cytokinetransforming growth factor-b (TGF-b) plays a key role inrenal fibrosis [1]. In particular, a number of very differentfactors known to be injurious to the kidney have beenshown to significantly increase TGF-b expression. Thesefactors include angiotensin II (Ang II), hypoxia, highglucose levels, autoantibodies, immune complexes, ad-vanced glycosylation end products, mesangial cell stretchand platelet derived growth factor (PDGF), TGF-b andbasic fibroblast growth factor (bFGF) [reviewed in 2, 3].Overproduction of TGF-b in turn strongly induces andorchestrates an array of events leading to matrix accumu-lation and tissue fibrosis. These events include: (1) in-creased synthesis of matrix proteins such as collagens,fibronectin and proteoglycans; (2) decreased degradationof matrix proteins by suppression of protease expressionand increased production of protease inhibitors such asplasminogen-activator-inhibitor type 1 (PAI-1) or TIMP;and (3) increased synthesis and expression of a group ofcell-matrix receptors called integrins which enhance localmatrix deposition [1].

In rats with anti-thymocyte-serum (ATS)-induced glo-merulonephritis, we have previously shown that treatmentwith TGF-b neutralizing antibodies or the natural TGF-bantagonist decorin greatly reduces pathological glomerularmatrix expansion, suggesting a causative role for TGF-b inrenal fibrosis [4–6]. This notion has recently been con-firmed using gene transfer techniques. Isaka et al havedemonstrated that introduction of the functional TGF-b1gene into normal glomeruli markedly increases productionof extracellular matrix proteins and rapidly leads to glomer-ulosclerosis [7]. Furthermore, Kopp et al have reportedthat transgenic mice with high circulating TGF-b1 developsevere progressive kidney fibrosis and die from renal failure[8].

Drugs that block the actions of Ang II have shownbeneficial effects in a number experimental kidney diseases[2]. ACE inhibitors and the newly available class of Ang IIreceptor type 1 antagonists effectively lessen proteinuria,reduce systemic and glomerular hypertension and prevent

Key words: transforming growth factor-b, enalapril, losartan, fibrosis,glomerulonephritis.

Received for publication December 23, 1997and in revised form March 24, 1998Accepted for publication June 24, 1998

© 1998 by the International Society of Nephrology

Kidney International, Vol. 54 (1998), pp. 1570–1580

1570

development of kidney fibrosis. In humans with diabetic orIgA nephropathy, ACE inhibition slows the progressive lossof renal function and delays the requirement for dialysis ortransplantation [reviewed in 2, 9]. Several recent experi-mental studies have demonstrated that Ang II blockaderesults in a decrease in TGF-b expression and matrixaccumulation, suggesting that the antifibrotic effect is me-diated through a reduction of this fibrogenic cytokine[reviewed in 2, 3]. None of these studies specifically askedhow much reduction in TGF-b is possible with these drugs.

Although systemic or glomerular hypertension is animportant factor in renal injury, several lines of evidencesuggest that the therapeutic effect of Ang II blockade ispartially independent of the hemodynamic effects. Forinstance, in rats with deoxycorticosterone acetate salt hy-pertension, ACE inhibitors and Ang II receptor blockerseffectively limit renal fibrosis with no effect on systemicblood pressure [10]. In addition, ACE inhibitors signifi-cantly slow the progression of renal failure in patients evenin the absence of increased blood pressure [11]. Neverthe-less, the doses of ACE inhibitors or Ang II receptorblockers used are based on their effects on systemic bloodpressure and preservation of filtration rather than by theirantifibrotic effect.

Since overexpression of TGF-b has a central role in renalfibrosis, we asked whether the efficacy of Ang II blockadeon matrix accumulation could be maximized if reduction inTGF-b overexpression is the target. For this study we usedthe anti-thymocyte model of acute mesangioproliferativeglomerulonephritis. There are several reasons for using thismodel. First, we have spent much time evaluating differentmarkers of disease and different sacrifice times with thegoal of using those giving a large difference betweennormal and disease control groups, thereby maximizing thewindow available for detecting small differences in thera-peutic efficacy. Second, although glomerular hypertensionis present for a short period early [12], this model is notconsidered a hypertensive model. Efficacy of angiotensin IIblockade would therefore be expected to be due to otherpathways by which Ang II acts on disease. Finally, we haveundertaken a series of studies aimed at systematicallycomparing currently available and promising treatments forrenal fibrosis with emphasis on the question of what is themaximum effect possible with one drug and whether com-bined therapy enhances disease reduction.

In the present study, we addressed the following ques-tions: (1) At what doses are the antifibrotic effects of theACE inhibitor enalapril or of the Ang II receptor antago-nist losartan maximal? (2) Are enalapril and losartanequipotent at these doses? (3) Will inhibition of Ang IIgeneration and blockade of Ang II type 1 receptors to-gether be more effective than single treatment at maximallyeffective doses? Compared to other models of renal dis-ease, the experimental model used in this study is charac-terized by a very rapid and marked glomerular TGF-b

overexpression and accumulation of extracellular matrix.Using new methods to detect production of TGF-b anddisease markers fibronectin and PAI-1 in the supernatantof cultured nephritic glomeruli, this model is highly sensi-tive to small differences in therapeutic efficacy.

METHODS

Materials

If not otherwise indicated, materials, chemicals or cul-ture media were purchased from Sigma Chemical Co. (St.Louis, MO, USA).

Animals

The studies were performed in male Sprague Dawley(SD) rats (200 to 270 g) obtained from Sasco (Omaha, NE,USA). Animal housing and care were in accordance withthe NIH Guide for the Care and Use of Laboratory Animals,NIH Publication No. 85-23, 1985. Animals were fed anormal protein diet (22% protein, Teklad No. 86 550;Teklad Premier Laboratory Diets, Madison, WI, USA).Glomerulonephritis was induced by tail vein injection ofthe monoclonal anti-Thy 1.1 antibody OX-7 (1.5 mg/1 kg).OX-7 binds to a Thy 1.1-like epitope on the surface ofmesangial cells and causes complement-dependent cell lysis[13]. Control animals were injected with the similar vol-umes of phosphate buffered saline (PBS). Body weight andfood intake was monitored at the beginning and end ofeach experiment.

OX-7 production

The monoclonal antibody was produced in the asciticfluid of adult female BALB/c mice (Charles River, Wil-mington, MA, USA). The hybridoma was obtained fromAmerican Type Culture Collection (Rockville, MD, USA)and grown in RPMI 1640 (Gibco BRL, Long Island, NY,USA) supplemented with 10% FCS, 100 U/ml penicillin,100 mg/ml streptomycin and 25 mM HEPES buffer at37°C/5% CO2. Cells were passaged at confluence. The micewere primed with two intraperitoneal injections of 0.5 mlpristane (2,6,10,14-tetramethyldecanoic acid) one weekapart. Three days after the second pristane administration,the hybridoma cells were spun at 3000 g for five minutesand resuspended in PBS (pH 7.4). Approximately 1 to 2million cells in 0.5 ml PBS were injected intraperitoneallyinto each mouse. When mice had developed noticeableascites, they were anesthetized with ether and the asciticfluid was collected.

To purify the IgG fraction, ascitic fluid was combinedwith 2/10 volumes 1.0 M Tris buffer (pH 8.0) and applied toa pre-equilibrated Protein G column (Pierce, Rockford, IL,USA). The antibody fraction was eluted with 100 mM

glycine (pH 3.0) and promptly neutralized with Tris buffer(pH 9.0). After extensive dialysis against PBS (pH 7.4), the

Peter et al: Targeting TGF-b overexpression 1571

antibody concentration was adjusted to 1 mg/ml by readingthe adsorbance at 280 nm and stored at 270°C until use.

Angiotensin II blockade

The ACE inhibitor enalapril maleate and the Ang IIreceptor type 1 antagonist losartan were used. Losartan waskindly provided by DuPont Merck Pharmaceuticals Co.(Wilmington, DE, USA). Enalapril maleate is a prodrugand is converted in the liver to form the active metaboliteenalaprilate [14]. In rats and humans, losartan is itselfactive but is also converted to the several-fold more activemetabolite, EXP3174 [15]. Both enalapril and losartanshow a relatively long lasting effect on blood pressureallowing single dosing per day [14, 15]. Doses comparableto a drug intake achieved with 20 to 50 mg/liter enalapril or50 to 100 mg/liter losartan given in the drinking water havebeen shown to sufficiently control blood pressure in modelsof renal hypertension, including the remnant kidney andGoldblatt models [14–17].

Experimental design

Experiment 1. Effect of increasing doses of enalapril in thedrinking water on glomerular TGF-b1 production six daysafter disease induction. Sprague Dawley rats were assignedto the following groups: (1) PBS-injected controls (N 5 4),(2) OX-7-injected rats drinking tap water only (N 5 4), and(3) OX-7-injected rats drinking water with 10, 20, 50, 100,200, 500 and 1000 mg/liter enalapril (N 5 3 each).

Experiment 2. Effect of increasing doses of losartan in thedrinking water on glomerular TGF-b1 production six daysafter disease induction. In this experiment, rats were as-signed to the following groups: (1) PBS-injected controls(N 5 4), (2) OX-7-injected rats drinking tap water only(N 5 4), and (3) OX-7-injected rats drinking water with 50,100, 250, 500, 1000 and 2500 mg/liter losartan (N 5 3 each).

Experiment 3. Combined effect of maximally effective dosesof enalapril and losartan on glomerular TGF-b1 productionsix days after disease induction. Based on results fromexperiments 1 and 2, the minimum doses of enalapril (100mg/liter) and losartan (500 mg/liter), which maximallyreduced TGF-b overexpression, were used. Rats wereassigned to the following groups: (1) PBS-injected controls(N 5 4), (2) OX-7-injected rats drinking tap water only(N 5 4), (3) OX-7-injected rats with 100 mg/liter enalapril(N 5 4), (4) OX-7-injected rats with 500 mg/liter losartan(N 5 3), and (5) OX-7 injected rats drinking water with 100mg/liter enalapril and 500 mg/liter losartan (N 5 4).

In all three experiments, drug treatment was started 24hours after disease induction. At that time, based on anaverage water intake of 40 ml per day, 60% of the dailydose was administered by gavage. Thereafter, the animalswere housed in metabolic cages to monitor the water anddrug intake individually. The amount of drug in the drink-ing water was adjusted daily to that amount each animalwould receive by drinking 40 ml water. Systolic blood

pressure was measured in the conscious state by tail cuffmethod as previously described the day before the end ofthe experiment [18].

Six days after induction of glomerulonephritis, the ani-mals were anesthetized with ether. Following a midlineabdominal incision, 5 to 10 ml blood was drawn from thelower abdominal aorta and the kidneys were subsequentlyperfused with 30 ml ice-cold PBS. For histological exami-nation cortical tissue was snap frozen and fixed in 10%neutral buffered formalin. Glomeruli from individual ratswere isolated by a graded sieving technique (150, 125, 106and 75 mm mesh metal sieves) as described previously [19]and resuspended at 2000 glomeruli/ml/well in DMEMsupplemented with 0.1 U/ml insulin, 100 U/ml penicillin,100 mg/ml streptomycin and 25 mM HEPES buffer. After a72 hours incubation at 37°C/5% CO2, the supernatant washarvested and stored at 270°C until analysis of glomerularproduction of TGF-b1, fibronectin and PAI-1.

Measurement of TGF-b1, fibronectin and PAI-1

TGF-b1 production of cultured glomeruli was measuredafter acid activation using a commercially available inhibi-tory enzyme-linked immunosorbant assay (ELISA) kit(Quantikiney, R&D Systems, Minneapolis, MN, USA)according to the manufacturer’s instructions. Therefore,this method only measured total TGF-b1. Fibronectin andPAI-1 synthesis were measured with modified ELISA as-says according to published methods [20]. Three samplesfrom each rat were analyzed.

Light microscopy

All microscopic examinations were performed in ablinded fashion. Three micrometer sections of paraffin-embedded tissue were stained with periodic acid-Schiff(PAS) and glomerular matrix expansion was evaluated aspreviously described [19]. Briefly, in 30 glomeruli from eachrat the percentage of mesangial matrix occupying eachglomerulus was rated as: 0 5 0%, 1 5 25%, 2 5 50%, 3 575% and 4 5 100%.

RNA preparation and Northern hybridization

Total RNA was extracted by a guanidinium isothiocyantemethod using Trizol® reagent according to the manufac-turers instructions. Isolated glomeruli from three to fourrats from the experimental group 3 were pooled for subse-quent RNA extraction. For Northern analysis, RNA wasdenatured and fractionated by electrophoresis through a1.8% agarose gel (30 mg/lane) and transferred to a Hybondnylon membrane (Amersham Corp., UK). Nuclei acidswere immobilized by UV irradiation. Membranes wereprehybridized with 50% formamide, 10% Denhardt’s solu-tion, 0.1% SDS, 5 3 SSC and 200 mg/ml denatured salmonsperm DNA and hybridized with DNA probes labeled with32P-dCTP by random oligonucleotide priming (BoehringerMannheim Biochemicals, Indianapolis, IN, USA). The

Peter et al: Targeting TGF-b overexpression1572

blots were washed in 2 3 SSC, 0.1% SDS at roomtemperature for 15 minutes and in 0.1 3 SSC, 0.1% SDS at55°C for 30 minutes. DNA probes used were: (1) ratglyceraldehyde-3-phosphate dehydrogenase (GAPDH)cDNA (a gift form Drs. P. Kondaiah and M.B. Sporn) [21],(2) TGF-b1 cDNA (kindly provided from Dr. D.L. Moses)[22], (3) fibronectin-EDA cDNA (a generous gift from Dr.R.O. Hynes) [23], and (4) PAI-1 cDNA (kindly provided byDr. T.D. Gelehrter) [24]. Three blots per probe wereperformed.

Statistical analysis

Data are expressed as mean 6 SEM. Statistical analysis ofdifferences between the groups was performed by ANOVAand subsequent t-testing with Bonferroni correction for

multiple comparison. Comparisons with a P value , 0.05were considered significantly different.

RESULTS

Body weight, food consumption, drinking volume

There were no significant differences in body weight,food intake and drinking volume between the groupsinvestigated. All animals drank more than 20 ml per day.

In all three experiments, measures of disease severitywere very similar (Figs. 1 to 3). Compared to the normalcontrol group, the mean TGF-b1 production of culturednephritic glomeruli from untreated animals was increased8.7-fold, and glomerular production of fibronectin andPAI-1 synthesis were increased by 7.4-fold and 16.6-fold,

Fig. 1. Effect of increasing doses the ACE inhibitor enalapril in the drinking water on: (A) glomerular matrix accumulation, (B) glomerular TGF-b1production, (C) glomerular fibronectin production, and (D) glomerular PAI-1 production. In male Sprague Dawley rats treatment was started one dayafter injection of OX-7 (GN). The kidneys were harvested six days after disease induction.

Peter et al: Targeting TGF-b overexpression 1573

respectively. The mean glomerular matrix score increasedfrom 1.39 in the normal animals to 2.45 in the diseasecontrols.

Systolic blood pressure in the PBS-injected controls was114 6 11 mm Hg in experiment 1, 109 6 9 mm Hg inexperiment 2 and 108 6 12 mm Hg in experiment 3. In thedisease controls drinking tap water only, blood pressurewas 111 6 11, 118 6 14 and 117 6 15 mm Hg, respectively.Treatment with enalapril, losartan or both together did notsignificantly change systolic blood pressure.

Experiment 1

As shown in Figure 1B, treatment with enalapril in thedrinking water reduced glomerular TGF-b1 production in adose-dependent manner. A moderate decrease in patho-logical TGF-b1 production of approximately 20 to 25% wasseen in rats treated with 10 to 20 mg enalapril/liter drinkingwater (Fig. 1). With increasing doses of enalapril, a maxi-mal reduction of approximately 45% in TGF-b1 productionwas seen starting in animals treated with 100 mg enalapril.A further increase in the enalapril dose did not furtherdecrease glomerular TGF-b1 production (200 mg: 41%;500 mg: 45%; 1000 mg: 45%). The accumulation of extra-cellular matrix in the glomeruli following OX-7 injectionshowed a similar pattern (Fig. 1A). Compared to thedisease control group (2.56 6 0.04) lower doses of enalapril(10, 20 and 50 mg/liter) reduced the matrix score slightly to2.28 6 0.04, 2.12 6 0.12 and 2.16 6 0.20, respectively.Higher doses of enalapril (100, 200, 500 and 1000 mg/liter)prevented glomerular matrix accumulation more effectivelygiving matrix scores of 1.92 6 0.20, 2.00 6 0.20, 2.04 6 0.08and 2.00 6 0.04, respectively.

Since TGF-b1 is a potent stimulator of fibronectin andPAI-1 synthesis, we used their production by isolatedglomeruli as an additional indicator of therapeutic efficacyof Ang II blockade on TGF-b1 expression in our system.Similar to the glomerular TGF-b production and matrixaccumulation, treatment of nephritic animals with increas-ing doses of enalapril reduced glomerular fibronectin andPAI-1 production in a dose-dependent manner. Glomeruliisolated from the disease control animals produced 4953 6729 ng/ml fibronectin and 185 6 19 ng/ml PAI-1 (Fig. 1, Band C, respectively). Treatment with 10, 20, 50, 100, 200,500 and 1000 mg enalapril in the drinking water decreasedthe glomerular fibronectin by 13%, 24%, 40%, 46%, 36%,40% and 42%, respectively. The glomerular PAI-1 produc-tion was decreased by 36%, 35%, 64%, 59%, 66%, 58% and64%, respectively.

Experiment 2

Similar to enalapril in experiment 1, treatment of ne-phritic rats with increasing doses of losartan decreasedglomerular TGF-b1 overproduction in a dose-dependentmanner as well. Addition of 50 mg or 100 mg losartan to thedrinking water reduced pathological glomerular TGF-b1

production only by 20 to 25% (Fig. 2B). Higher doses oflosartan decreased the TGF-b1 production more effec-tively. A maximal decrease of approximately 45% was seenat 500 mg/liter losartan, however, a further increase in thelosartan dose did not further decrease TGF-b1 production(1000 mg, 38%; 2500 mg, 40%). Compared to the matrixaccumulation seen in the disease controls (2.40 6 0.04),treatment of nephritic rats with 50, 100, 250, 500, 1000 and2500 mg losartan decreased the matrix score to 2.24 6 0.08,2.20 6 0.12, 2.00 6 0.08, 1.84 6 0.04, 1.88 6 0.16 and1.84 6 0.04, respectively (Fig. 2A). Glomerular fibronectinand PAI-1 production were decreased dose-dependently aswell (Fig. 2 B, C). Treatment of nephritic rats with increas-ing doses of losartan reduced glomerular fibronectin pro-duction by 21%, 14%, 28%, 41%, 49% and 36%, respec-tively. The glomerular PAI-1 production decreased by 29%,37%, 58%, 61%, 54% and 62%, respectively.

Experiment 3

The minimum doses of enalapril and losartan showingmaximal TGF-b1 reduction were combined to look foradditive effects of these drugs on disease severity. Fromexperiments 1 and 2, the doses of 100 mg/liter enalapril and500 mg/liter losartan (Figs. 1 and 2), were chosen. Theresults clearly show that, at these doses, a combination ofenalapril and losartan was not superior to one form of AngII blockade alone (Fig. 3). Treatment of rats with 100mg/liter enalapril in the drinking water reduced glomerularTGF-b1 synthesis by 41%, therapy with 500 mg/liter losar-tan by 45% and treatment with both 100 mg/liter enalapriland 500 mg/liter losartan by 45% reduction (Fig. 3B).Compared to the disease control animals, glomerular ma-trix accumulation was also equally reduced in the threetreatment groups (Fig. 3A; matrix score of 2.40 6 0.04 vs.1.84 6 0.04, 1.84 6 0.12 and 1.80 6 0.04, respectively). Theglomerular production of fibronectin and PAI-1 showed asimilar picture (Fig. 3 C, D). Treatment of nephritic ratswith enalapril or losartan or both decreased the glomerularfibronectin by 41%, 37% and 41%, respectively. The glo-merular synthesis of PAI-1 decreased by 52%, 61% and56%, respectively.

In experiment 3, the glomerular mRNA expression ofTGF-b1, fibronectin and PAI-1 in addition to their glomer-ular protein production was analyzed (Fig. 4). Reduction ofthese mRNAs were not as great as the decrease in theirprotein production. Again, as shown in Figure 4, treatmentof nephritic rats with maximally effective doses of enalaprilor losartan or both decreased the glomerular TGF-b1,fibronectin, and PAI-1 mRNA levels significantly (P ,0.05), and to a similar extent (P . 0.05). Compared to thedisease control animals, treatment with 100 mg/liter enala-pril or 500 mg/liter losartan or both reduced glomerularTGF-b1 mRNA expression by 16%, 16% and 21%, respec-tively. Glomerular fibronectin mRNA levels were de-creased by 26%, 30% and 22% and glomerular PAI-1

Peter et al: Targeting TGF-b overexpression1574

mRNA expression were decreased by 25%, 25% and 25%,respectively.

DISCUSSION

Fibrotic diseases are characterized by the accumulationof extracellular matrix. The current state of our knowledgesuggests that TGF-b overexpression is responsible for thisaccumulation through its actions to induce production ofextracellular matrix, inhibit its degradation and increaseintegrin expression on cells, thereby enhancing matrixdeposition. TGF-b overexpression may be viewed as thefinal common element in fibrotic diseases [1]. A largenumber of stimuli induce its expression. Of relevance torenal disease are Ang II, cell damage or lysis, immunecomplex deposition, high glucose, advanced glycosylatedend products, shear stress and mesangial stretch as well as

exogenously delivered cytokines TGF-b, PDGF and bFGF[reviewed in 3]. It is noteworthy that most models of renaldisease, as well as most human renal diseases, involveincreases in one or more of these stimuli.

Evidence is mounting rapidly that reduction of TGF-boverexpression in models of fibrosis decreases pathologicalmatrix accumulation, whether the therapy is neutralizingantibodies to TGF-b, decorin, low protein intake, angio-tensin II blockade or L-arginine supplementation [2, 4–6,18, 25, 26]. Very recently, two human studies reported inabstract form that angiotensin blockade is associated withreduction in TGF-b overexpression. The first showed areduction in glomerular TGF-b mRNA expression in IgApatients taking ACE inhibitors [27], and the second re-ported reductions in plasma TGF-b levels in diabeticpatients after six months treatment with captopril [28].

Fig. 2. Effect of increasing doses of the Ang II receptor type 1 antagonist losartan in the drinking water on: (A) glomerular matrix accumulation, (B)glomerular TGF-b1 production, (C) glomerular fibronectin production, and (D) glomerular PAI-1 production. Treatment was started one day afterinjection of OX-7 (GN). The kidneys were harvested six days after disease induction.

Peter et al: Targeting TGF-b overexpression 1575

Since fibrotic diseases are characterized by TGF-b overex-pression, we took a further step to quantitate TGF-boverexpression as a marker of disease severity and to testthe hypothesis that the of degree of reduction of TGF-b isa useful molecular marker of the therapeutic efficacy ofcommonly used and new antifibrotic therapies.

The model of mesangioproliferative glomerulonephritisused in this study is one in which administration of anantibody to the Thy 1.1 epitope on glomerular mesangialcells leads to immune-mediated, complement-dependentmesangial cell lysis. Evidence supports the notion that thismodel has great relevance because it contains all of thedisease elements demonstrated in human fibrotic diseases,whether acute or chronic. Another advantage of this modelis that it is rapid, reproducible and, as clearly demonstratedin this study, can be used to very sensitively discriminate

between therapeutic regimens. It is clear that a singleinjury, whether mesangial cell lysis, or chemical damagedue to administration of aminonucleoside or adriamycin,rarely leads to chronic disease. However, if injury in thesemodels is repeated, disease does often progress. Similarlyin humans, one bout of post-streptococcal glomerulone-phritis rarely produces progressive disease. Other modelsmay also be progressive because the injury is repeated orcontinual. For example, removal of 5/6 of the renal massproduces persistent hypertensive injury. Similarly, ligating aureter to produce ureteral obstruction results in continualhydrostatic injury. In diabetes, periodic episodes of hyper-glycemia or repeated deposition of advanced glycosylatedendproducts may be perceived as repeated injury. It is alsopossible, if injury subsided in these models, that renal tissuemight have some capacity to recover.

Fig. 3. Effect of the maximally effective doses of enalapril (Ena, 100 mg/liter) or losartan (Los, 500 mg/liter) or both (100 mg/liter enalapril and 500mg/liter losartan) on: (A) glomerular matrix accumulation, (B) glomerular TGF-b1 production, (C) glomerular fibronectin production, and (D)glomerular PAI-1 production. Treatment was started one day after injection of OX-7 (GN). The kidneys were harvested six days after disease induction,*P , 0.01 versus GN.

Peter et al: Targeting TGF-b overexpression1576

The two drugs used in this study are both importantantihypertensives. Studies comparing ACE inhibitors withother drugs reducing blood pressure have long suggestedthat Ang II has pressure-independent fibrogenic actions[29]. However, only recently has Ang II been shown bedirectly fibrogenic by up-regulating TGF-b, matrix proteinsynthesis and PAI-1 in culture, in the absence of increasedpressure [reviewed in 3]. Angiotensin blockade with ACEinhibitors and/or Ang II type 1 receptor antagonists hasbeen shown to reduce fibrosis in a large number of diseasemodels and in an ever-growing number of human trials[reviewed in 30]. In a few of these studies TGF-b overex-

pression was investigated. We have previously noted thatthe observed reductions in TGF-b overexpression withangiotensin blockade were usually on the order of 40 to60% [2]. A primary goal of the present study was todetermine whether higher does of these drugs can reduceTGF-b overproduction and fibrosis more completely. Asecondary goal was to determine whether, when thesedrugs are combined, at doses showing maximal effectsalone, any additive effects are seen.

The results demonstrate that increasing doses of enala-pril or losartan results in a dose-dependent antifibroticeffect. The maximal reduction in pathological glomerular

Fig. 4. Effect of maximally effective doses of enalapril (Ena, 100 mg/liter) or losartan (Los, 500 mg/liter) or both (100 mg/liter enalapril and 500mg/liter losartan) on mRNA expression. (A) Representative Northern blot. The relative glomerular mRNA expression of TGF-b1 is depicted in (B),fibronectin in (C), and PAI-1 in (D). Treatment was started one day after injection of OX-7 (GN). The glomerular mRNA was isolated six days afterdisease induction and is expressed in relation to GAPDH expression, *P , 0.05 versus GN. Glomerular size markers are shown on the right.

Peter et al: Targeting TGF-b overexpression 1577

TGF-b1 overexpression and matrix accumulation was seenat doses higher than those known to control blood pressurein experimental forms of hypertension [15–17]. At thelower doses of Ang II blockade, only a moderate antifi-brotic effect was seen. This finding is consistent with aprevious report showing that, following 5/6 nephrectomy inrats, a higher dose of ACE inhibitor reduced histologicalevidence of renal fibrosis more effectively than a lower onedespite a similar effect on blood pressure [31]. The presentstudy expands this observation to a broader range of doses,adds a receptor antagonist, and again shows that dosing isvery important. Two other studies have used Ang IIblockade in the ATS model. One showed a therapeuticeffect in hypertensive animals [32] and the other, veryrecently published, showed data similar to that presentedhere except without multiple doses or combination therapy[33].

After a point, increasing doses did not lead to furtherreduction in TGF-b expression or any other measure ofdisease. It was impossible to normalize TGF-b expression.This result is consistent with human data where Ang IIblockade slows, but does not halt the progressive loss ofrenal function [reviewed in 2, 9] and strongly implies thatAng II blockade alone cannot prevent end-stage renalfibrosis. The data also suggest that, at least in this model ofdisease, the two drugs used are almost identical in theantifibrotic effect seen at maximal doses. To our knowledgethis is the first such side-by-side comparison of ACEinhibition and angiotensin receptor antagonism at manydoses.

New data are beginning to appear suggesting that thesedrugs have actions other than blocking Ang II that lead toincreases in beneficial, vasodilatory nitric oxide. ACE in-hibitors block proteases other than ACE, producing anaccumulation of bradykinin and increased nitric oxidesynthesis [9]. Ang II receptor antagonists induce accumu-lation of large amounts of Ang II and Ang II breakdownproducts [34], and these breakdown products have beenreported to induce vasodilation via a kinin/nitric oxidepathway [35, 36]. High levels of circulating Ang II causeincreased signaling through the Ang II type 2 receptor. Thismay also lead to increased nitric oxide production [36, 37].That these pathways can be important has been recentlysuggested in a model of cardiac fibrosis where selectiveblockade of the Ang II type 2 receptor or the bradykininreceptor largely reversed the beneficial effect seen inanimals treated only with an Ang II receptor type 1antagonist [38]. While the present study does not addressthese questions directly, the striking equipotency of maxi-mally effective doses of enalapril and losartan, and theabsence of any additive therapeutic effect, suggest thatthere are similarities in the mechanisms of action. If indeedenhanced nitric oxide production is an important action ofboth drugs, albeit via different pathways, the data suggestthat at maximally effective doses, both drugs enhance nitric

oxide to levels that are maximally therapeutic. Otherwisethere should be some drug additivity. It remains most likelythat both of these drugs act primarily to reduce the actionsof Ang II that occur through the type 1 receptor.

It has been suggested that the simultaneous use of ACEinhibition and Ang II receptor antagonism is required toachieve maximal inhibition of Ang II [39, 40]. Our resultssuggest that this may be true if both drugs are at sub-maximal doses, but not if given at maximal antifibroticdoses. However, the use of both drugs together at sub-optimal doses might reduce the frequency of side effectswhile providing maximal therapeutic efficacy.

The mechanisms of action of these drugs are clearlycomplex, and some of their actions may be as yet unknown.For example, because Ang II may stimulate mesangial cellproliferation [41, 42], blocking its action may reduce pro-liferation so that the number of activated mesangial cellsoverproducing TGF-b, fibronectin and PAI-1 is reduced.Thus, one could argue that the therapeutic effects shownhere result only from less mesangial cell proliferation.However, since Ang II induces increases in TGF-b [43, 44]and TGF-b stimulates PDGF receptor expression [45–47],potentially making mesangial cells more sensitive to theproliferative actions of PDGF, one could also argue thatreduced cell proliferation in response to Ang II blockadeare TGF-b mediated. It appears to us that the in vivosituation is simply too complex to be certain which path-ways are operating. However, the paradigm “where there isoverexpression of TGF-b there is fibrosis” is well estab-lished and is the basis for our therapeutic targeting ofTGF-b overproduction, regardless of the complexity of themechanisms underlying the observed TGF-b reduction.

In summary, the present study demonstrates that quan-titation of TGF-b overexpression can be a powerful param-eter with which to compare and contrast different thera-peutic approaches to fibrotic renal diseases and that theThy 1.1 model of glomerulonephritis is an excellent modelin which to do this. TGF-b overexpression and fibroticdisease were reduced more effectively at doses of enalapriland losartan that are higher than those known to reduceblood pressure. Enalapril and losartan are equipotent atmaximally effective doses and combined therapy yields noadditional effect. Perhaps most importantly, in our effortsto prevent renal fibrosis in humans, is this study’s strongsuggestion that neither ACE inhibition nor Ang II typereceptor antagonism will be sufficient to normalize over-production of TGF-b. While these therapies clearly haveefficacy, it is likely that they must be combined with otheragents, which act through different mechanisms, in order toultimately stop disease.

ACKNOWLEDGMENTS

This work was supported by National Institute of Diabetes and Diges-tive and Kidney Diseases (NIDDK) with grants DK 49342 to N.A.N. andDK 43609 and DK 49374 to W.A.B., and a matching fellowship from the

Peter et al: Targeting TGF-b overexpression1578

National Kidney Foundation of Utah to H.P. We thank Ms. Linda Hogeand Ms. Diane Miller for excellent technical assistance.

Reprint requests to Nancy A. Noble, Ph.D., Division of Nephrology,University of Utah School of Medicine, Salt Lake City, Utah 84132, USA.E-mail: nnoble@hsc.utah.edu

APPENDIX

Abbreviations used in this article are: ACE, angiotensin convertingenzyme; Ang II, angiotensin II; ATS, anti-thymocyte serum; bFGF, basicfibroblast growth factor; ELISA, enzyme-linked immunosorbant assay;FCS, fetal calf serum; GAPDH, glyceraldehyde-3-phosphate dehydroge-nase; IgA, immunoglobulin A; PAI-1, plasminogen activitor inhibitor type1; PBS, phosphate buffered saline; PDGF, platelet-derived growth factor;TGF-B, transforming growth factor-B.

REFERENCES

1. BORDER WA, NOBLE NA: Transforming growth factor-b in tissuefibrosis. N Engl J Med 331:1286–1292, 1994

2. NOBLE NA, BORDER WA: Angiotensin II in renal fibrosis: ShouldTGF-b rather than blood pressure be the therapeutic target? SeminNephrol 17:455–466, 1997

3. BORDER WA, NOBLE NA: Interactions of TGF-b and angiotensin II inrenal fibrosis. Hypertension S31:181–188, 1998

4. BORDER WA, OKUDA S, LANGUINO LR, SPORN MB, RUOSLAHTI E:Suppression of experimental glomerulonephritis by antiserum againsttransforming growth factor beta-1. Nature 346:371–374, 1990

5. BORDER WA, NOBLE NA, YAMAMOTO T, HARPER JR, YAMAGUCHI Y,PIERSCHBACHER MD, RUOSLAHTI E: Natural inhibitor of transforminggrowth factor-b protects against scarring in experimental kidneydisease. Nature 360:361–364, 1992

6. ISAKA Y, BREES DK, IKEGAYA K, KANEDA Y, IMAI E, NOBLE NA,BORDER WA: Gene therapy by skeletal muscle expression of decorinprevents fibrotic disease in rat kidney. Nature Med 2:418–423, 1996

7. ISAKA Y, FUJIWARA Y, UEDA N, KANEDA Y, KAMADA T, IMAI E:Glomerulosclerosis induced by in vivo transfection of transforminggrowth factor-b or platelet-derived growth factor into the rat kidney.J Clin Invest 92:2597–2601, 1993

8. KOPP JB, FACTOR VM, MOZES M, NAGY P, SANDERSON N, BOTTINGEREP, KLOTMAN PE, THORGEIRSSON SS: Transgenic mice with increasedplasma levels of TGF-b1 develop progressive renal disease. Lab Invest74:991–1003, 1996

9. ICHIKAWA I: Will angiotensin II receptor antagonists be renoprotec-tive in humans? Kidney Int 50:684–492, 1996

10. KIM S, OHTA K, HAMAGUCHI A, OMURA T, YUKIMURA T, MIURA K,INADA Y, WADA T, ISHIMURA Y, CHATANI F, IWAO H: Role ofangiotensin II in renal injury of deoxycorticosterone acetate-salthypertensive rats. Hypertension 24:195–204, 1994

11. PARVING HH, HOMMEL E, DAMKJAER NIELSEN M, GIESE J: Effect ofcaptopril on blood pressure and kidney function in normotensiveinsulin dependent diabetics with nephropathy. BMJ 299:533–536, 1989

12. YAMAMOTO T, MUNDY CA, WILSON CB, BLANTZ RC: Effect ofmesangial cell lysis and proliferation on glomerular hemodynamics inthe rat. Kidney Int 40:705–713, 1991

13. BAGCHUS WM, HOEDEMAEKER PJ, ROZING J, BAKKER WW: Glomer-ulonephritis induced by monoclonal anti-thy 1.1 antibodies. Lab Invest55:680–687, 1986

14. TODD PA, HEEL RC: Enalaprin. A review of its pharmacodynamic andpharmacokinetic properties, and therapeutic use in hypertension andcongestive heart failure. Drugs 31:198–248, 1986

15. SMITH RD, CHIU AT, WONG PC, HERBLIN WF, TIMMERMANS PBWM:Phamacology of nonpeptide angiotenisin II receptor antagonists.Annu Rev Pharmacol Toxicol 32:135–165, 1992

16. RYAN MJ, BOUCHER DM, COHEN DM, OLSZEWSKI BJ, SINGER RM,SMITH RD, KAPLAN HR: Antihypertensive profile of the angiotensin-converting enzyme inhibitors CI-906 and CI-907. Fed Proc 43:1330–1332, 1984

17. LAFAYETTE RA, MAYER G, PARK SK, MEYER TW: Angiotensin IIreceptor blockade limits glomerular injury in rats with reduced renalmass. J Clin Invest 90:766–771, 1992

18. OKUDA S, NAKAMURA T, YAMAMOTO T, RUOSLAHTI E, BORDER WA:Dietary protein restriction rapidly reduces transforming growth factorb1 expression in experimental glomerulonephritis. Proc Natl Acad SciUSA 88:9765–9769, 1991

19. OKUDA S, LANGUINO LR, RUOSLAHTI E, BORDER WA: Elevatedexpression of transforming growth factor-beta and proteoglycan pro-duction in experimental glomerulonephritis: Possible role in expan-sion of the mesangial extracellular matrix. J Clin Invest 86:453–462,1990

20. RENNARD SI, BERG R, MARTIN GR, FOIDART JM, ROBEY PG:Enzyme-linked immunoassay (ELISA) for connective tissue compo-nents. Anal Biochem 104:205–214, 1980

21. FORT P, MARTY L, PIECHACZYK M, SABROUTY SE, DANI C, JEANTEURPH, BLANCHARD JM: Various rat adult tissues express only one majormRNA species from the glyceraldehyde-3-phosphate-dehydrogenasemultigenic family. Nucl Acids Res 13:1431–1442, 1985

22. DERYNCK R, JARRETT A, CHEN EY, GOEDDEL DV: The murinetransforming growth factor-B precursor. J Biol Chem 261:4377–4379,1986

23. SCHWARZBAUER JE, TAMKUN JW, LEMISCHKA IR, HYNES RO: Threedifferent fibronectin mRNAs arise by alternative splicing within thecoding region. Cell 35:421–431, 1983

24. ZEHEB R, GELEHRTER TD: Cloning and sequencing of cDNA for therat plasminogen activator inhibitor-1. Gene 73:459–468, 1988

25. REYES AA, PUKERSON ML, KARL I, KLAHR S: Dietary supplementa-tion with L-arginine ameliorates the progression of renal disease inrats with subtotal nephrectomy. Am J Kidney Dis 20:168–176, 1992

26. PETERS H, BORDER WA, NOBLE NA: Low protein diet and L-argininesupplementation produce additive therapeutic effects on mesangipro-liferative glomerunephrtitis in the rat. (abstract) J Invest Med 45:277A,1997

27. NISHIMURA M, OKAMURA M, KONISHI Y, INOUE K, YOUNG CK,YOSHIOKA T, NEGORO N, INOUE T, KANAYAMA Y: Effect of treatmentwith ACE inhibitor (ACEI) on TGF-beta gene expressions in renalbiopsy from patients with IgA nephropathy. (abstract) J Am SocNephrol 6:397, 1995

28. ELTAYEB B, ALZAHABI B, ZIYADEH FN, ROHDE R, LEWIS EJ, SHARMAK: Effects of captopril on serum levels of TGF-b1 in insulin-dependent diabetic patients. (abstract) J Am Soc Nephrol 8:110A, 1997

29. ANDERSON S, RENNKE HG, BRENNER BM: Therapeutic advantage ofconverting enzyme inhibitors in arresting progressive renal diseaseassociated with systemic hypertension in the rat. J Clin Invest 77:1993–2000, 1986

30. MATSUSAKA T, HYMES J, ICHIKAWA I: Angiotensin in progressive renaldiseases: Theory and practice. J Am Soc Nephrol 7:2025–2043, 1996

31. IKOMA M, KAWAMURA T, KAKINUMA Y, FOGO A, ICHIKAWA I: Causeof variable therapeutic efficiency of angiotensin converting enzymeinhibitor on glomerular lesions. Kidney Int 40:195–202, 1991

32. KIYAMA S, NANISHI F, TOMOOKA S, OKUDA S, ONOYAMA K, FU-JISHIMA M: Inhibitory effects of antihypertensive drugs on mesangialcell proliferation after anti-thymocyte serum (ATS)-induced mesan-giolysis in spontaneously hypertensive rats. Life Sci 54:1891–1900,1994

33. ZOJA C, ABBATE M, CORNA D, CAPITANIO M, DONADELLI R, BRUZZII, OLDROYD S, BENIGNI A, REMUZZI G: Pharmacologic control ofangiotensin II ameliorates renal disease while reducing renal TGF-bin experimental mesangioproliferative glomerulonephritis. Am J Kid-ney Dis 31:453–463, 1998

34. CAMPBELL DJ, KLADIS A, VALENTIJN AJ: Effects of losartan onangiotensin and bradykinin peptides and angiotensin-converting en-zyme. J Cardiovasc Pharmacol 26:233–240, 1995

35. BROSNIHAN KB, LI P, FERRARIO CM: Angiotensin-(1–7) dilatescanine coronary arteries through kinins and nitric oxide. Hypertension27:523–528, 1996

36. SEYEDI N, XU X, NASJLETTI A, HINTZE TH: Coronary kinin genera-tion mediates nitric oxide release after angiotensin receptor stimula-tion. Hypertension 26:164–170, 1995

37. SIRAGY HM, CAREY RM: The subtype-2 (AT2) angiotensin receptorregulates renal cyclic guanosine 39, 59-monophosphate and AT1receptor-mediated prostaglandin E2 production in conscious rats.J Clin Invest 97:1978–1982, 1996

38. LUI YH, YANG XP, SHAROV VG, NASS O, SABBAH HN, PETERSON E,CARRETERO OA: Effects of angiotensin-converting enzyme inhibitors

Peter et al: Targeting TGF-b overexpression 1579

and angiotensin II type 1 receptor antagonists in rats with heartfailure. Role of kinins and angiotensin II type 2 receptor. J Clin Invest99:1926–1935, 1997

39. GREENE EL, KREN S, HOSTETTER TH: Role of aldosterone in theremnant kidney model in the rat. J Clin Invest 98:1063–1068, 1996

40. AZIZI M, CHATELLIER G, GUYENE T-T, MURIETA-GEOFFROY D,MENARD J: Additive effects of combined angiotensin-convertingenzyme inhibition and angiotensin II antagonism on blood pressureand renin release in sodium-depleted normotensives. Circulation92:825–834, 1995

41. WOLF G, HABERSTROH U, NEILSON EG: Angiotensin II stimulates theproliferation and biosynthesis of type I collagen in cultured murinemesangial cells. Am J Pathol 140:95–107, 1992

42. RAY PE, AQUILERA G, KOPP JB, HORIKOSHI S, KLOTMAN PE:Angiotensin II receptor-mediated proliferation of cultured humanfetal mesangial cells. Kidney Int 40:764–771, 1991

43. ITOH H, MUKOYAMA M, PRATT RE, GIBBONS GH, DZAU VJ: Multiple

autocrine growth factors modulate vascular smooth muscle cell growthresponse to angiotensin II. J Clin Invest 91:2268–2274, 1993

44. KAGAMI S, BORDER WA, MILLER DE, NOBLE NA: Angiotensin IIstimulates extracellular matrix protein synthesis through induction oftransforming growth factor-b expression in rat glomerular mesangialcells. J Clin Invest 93:2431–2437, 1994

45. BATTEGAY EJ, RAINES EW, SEIFERT RA, BOWEN-POPE DF, ROSS R:TGF-b induces bimodal proliferation of connective tissue cells viacomplex control of an autocrine PDGF loop. Cell 63:515–524, 1990

46. HABERSTROH U, ZAHNER G, DISSER M, THAISS F, WOLF G, STAHL

RAK: TGF-b stimulates rat mesangial cell proliferation in culture:Role of PDGF b-receptor expression. Am J Physiol 264:F199–F205,1993

47. STOUFFER GA, OWENS GK: TGF-b promotes proliferation of culturedSMC via both PDG-AA-dependent and PDGF-AA-independentmechanisms. J Clin Invest 93:2048–2055, 1994

Peter et al: Targeting TGF-b overexpression1580