Kidney International, Vol. 55 (1999), pp. 2264–2273

l-Arginine supplementation increases mesangial cell injury andsubsequent tissue fibrosis in experimental glomerulonephritis

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

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

six hours but not in normal glomeruli. In the group sacrificedL-Arginine supplementation increases mesangial cell injury andat day 6, day 2 proteinuria was higher in the ATS 1 L-Argsubsequent tissue fibrosis in experimental glomerulonephritis.group compared with the ATS alone group (P , 0.05). Mea-Background. Mesangial cell lysis in the antithymocyte serumsures of fibrotic disease at day 6 all showed large increases(ATS)-induced model of glomerulonephritis is dependent onover control with ATS alone (P , 0.01), and further small,the generation of cytotoxic nitric oxide (NO) through transient

induction of NO synthase (iNOS). We hypothesized that in- but significant increases when L-Arg was combined with ATScreased availability of l-arginine (L-Arg) during mesangial cell (P , 0.05).lysis might provide iNOS with increased substrate leading to Conclusions. The results indicate that if given during diseaseincreased lysis, and that this increased lysis would be reflected induction, L-Arg supplementation can enhance iNOS-depen-in more severe fibrotic disease at day 6. dent tissue injury by providing increased substrate. Although

Methods. To ensure whole body equilibration with high the increase in injury with L-Arg supplementation was small,L-Arg at the time of injury, rats were pretreated with 1% L-Arg it led to increased fibrosis at day 6. These data predict that inin drinking water for one week prior to the administration of diseases with repeated iNOS-dependent tissue injury, L-ArgATS. Animals were sacrificed six hours after ATS injection supplementation may produce cumulative increases in tissuewhen previous experiments had indicated iNOS induction had fibrosis.occurred and at six days. At six hours, plasma was obtainedfor L-Arg levels and nitrite/nitrate (NOx) content. Renal tis-sues were taken for histological evaluation of glomerular cell

It is now clear that fibrotic renal diseases are character-counts, macrophage infiltration (ED-1), and iNOS expression.Glomeruli were isolated for detection of iNOS mRNA and ized by overexpression of the cytokine transformingplaced in culture to study the dependence of NO production growth factor-b (TGF-b) [1, 2]. Once TGF-b is overex-on L-Arg concentration. In rats sacrificed at six days, L-Arg pressed, cells begin a fibrotic “repair” response involvingsupplementation was stopped 16 hours after ATS injection.

increased production and deposition of pathological ma-Fibrotic disease was evaluated by urinary protein excretion,histological assessment of glomerular cell number, matrix accu- trix components and decreased turnover of matrix pro-mulation, and production of transforming growth factor-b1 teins. Although the number of factors leading to TGF-band matrix components fibronectin and plasminogen activator overexpression is growing rapidly [3], all appear to in-inhibitor type-1 (PAI-1) by isolated glomeruli in culture.

volve stimuli that cells may perceive as injurious. InResults. At six hours, the glomerular cell number was sig-normal tissue injury, such as a skin wound, the “injury”nificantly reduced by ATS injection (P , 0.01) and further

significantly (P , 0.05) reduced by L-Arg feeding [normal and “repair” phases are followed by a resolution phase,control (NC) 5 64.2 6 1, ATS 5 53.4 6 0.7, ATS 1 L-Arg 5 in which excess matrix is gradually removed. Although50.8 6 0.7]. Disease increased macrophage infiltration and

perhaps simplistic, it is our current hypothesis that manyiNOS protein and iNOS mRNA levels markedly (P , 0.01),fibrotic renal diseases result from repeated (or continual)whereas L-Arg feeding did not further increase these variables.

Plasma L-Arg levels (nmol/ml) were reduced by disease (NC 5 cycles of tissue “injury” and “repair” following so closely121 6 9, ATS 5 84 6 13, P , 0.01) and elevated by L-Arg that resolution of pathological matrix does not occur.feeding (ATS 1 L-Arg 5 166 6 12, P , 0.01). Plasma NOx The rate at which matrix accumulates depends on thewas significantly increased by ATS and further increased by

severity or frequency of repeated injury and can be grad-ATS 1 L-Arg (P , 0.05). Production of NOx by culturedual, as in diabetic nephropathy, or rapid, as in rapidlyglomeruli showed striking L-Arg concentration dependence inprogressive glomerulonephritis. In addition, at somepoint the accumulated matrix impairs organ functionKey words: fibrotic disease, kidney, tissue injury, nitric oxide, TGF-b,

diabetic nephropathy, rapidly progressive glomerulonephritis. severely enough that injury becomes constant even ifthe original stimulus to TGF-b overexpression is re-

Received for publication November 19, 1997moved. For example, when accumulated matrix de-and in revised form November 25, 1998

Accepted for publication January 4, 1999 creases the functional nephron number sufficiently, hy-pertension, itself a stimulus to TGF-b overexpression, is 1999 by the International Society of Nephrology

2264

Peters et al: L-Arg supplementation and renal fibrosis 2265

always present in the remaining nephrons, and the pro- cell lysis. We studied both the mechanism underlyingthe L-Arg effect and the consequences on fibrosis at daycess becomes self-perpetuating.

The antithymocyte serum (ATS) model of glomerulo- 6 of disease.nephritis involves a sequence of distinct “injury” and“repair” phases of disease, making it amenable to dissec-

METHODStion of effects on both processes. An injection of an

Experimental designantibody to a thy-1.1–like antigen present in glomerularmesangial cells initiates a complement-mediated lysis of Two experiments with three groups of eight rats each

were conducted. Groups were normal control, ATSa portion of these cells. This “injury” is followed by up-regulation of TGF-b, and a fibrotic “repair” response alone, and ATS 1 L-Arg. L-Arg supplementation in-

volved 1% L-Arg added to drinking water from sevenfollows.We have recently demonstrated that mesangial cell days before ATS injection to ensure total body equilibra-

tion until just after the injury phase of disease. The goallysis in this model is mediated through cytotoxic nitricoxide (NO) production from substrate l-arginine was to isolate the effects of L-Arg supplementation on

the injury phase of disease. In experiment 1, the effect of(L-Arg) [4]. Both pretreatment with the NO synthase(NOS) blocker l-monomethylarginine and restriction of L-Arg supplementation on the severity of ATS-induced

mesangial cell lysis and on variables known to be in-L-Arg intake during the period of lysis significantly re-duced cell lysis and subsequent disease severity [4]. In volved in mesangial cell lysis was examined six hours

after disease induction. Previous studies had revealedaddition, a time course study of the activities of L-Argmetabolic pathways revealed that mRNA for inducible that at this time, induction of iNOS had occurred, and

mesangial cell lysis was almost complete [5]. ExperimentNOS (iNOS) was seen only transiently during the mesan-gial cell lysis phase of disease at six hours after ATS 2 was similar to experiment 1 except that L-Arg supple-

mentation was stopped at 16 hours after disease induc-injection [5]. Because iNOS utilization of L-Arg to pro-duce NO has been reported to be very substrate depen- tion, and sacrifice occurred six days after ATS injection,

at a time when previous experiments had shown that thedent [6, 7], we hypothesized that L-Arg supplementation,given during the period of mesangial cell lysis, might fibrotic response is maximal. The goal was to determine

whether changes in mesangial cell injury caused byprovide the transiently induced iNOS with additionalsubstrate, leading to greater NO generation and greater L-Arg supplementation given only until 16 hours after

disease induction were reflected in the severity of fibrosiscell lysis. We further hypothesized that this increase incell lysis would be reflected in the severity of the fibrotic at day 6.response at day 6.

Methods used in both experimentsThere is now considerable evidence that manipulationof L-Arg intake alters the course of disease in complex Materials. Unless otherwise indicated, materials,

chemicals, and culture media were obtained from Sigmaways. In addition to our published data showing thatrestriction of L-Arg at the time of mesangial cell lysis Chemical Co. (St. Louis, MO, USA).

Animals. Male Sprague-Dawley (SD) rats (200 to 270 g)reduces lysis [4], other work has shown increased severityof nephrotoxic nephritis if L-Arg is systemically removed were obtained from Sasco (Omaha, NE, USA) and fed

a normal protein diet (22% protein, Teklad diet No.by injection of arginase [8]. Further complexity comesfrom studies showing that dietary L-Arg restriction dur- 86550; Teklad Premier Laboratory Diets, Madison, WI,

USA) with an L-Arg content of 0.85%.ing “repair” in ATS decreases the extent of fibrosis [9],whereas supplementation improves functional and mor- Disease induction. Glomerulonephritis was induced

by tail vein injection of a submaximal dose of ATS underphological parameters in a number of other models ofexperimental kidney disease [10–19]. Unlike ATS, in light methoxyflurane anesthesia. ATS was produced in

sheep as previously described [20]. ATS binds to amost of these models, the injurious stimulus is continualand not immune mediated: glomerular hypertension in thy-1.1–like epitope on the surface of mesangial cells and

causes complement-dependent cell lysis. Control animals5/6 nephrectomy, hydronephrosis in ureteral obstruc-tion, high glucose in diabetes, etc. Thus, there is no dis- were injected with equal volumes of phosphate-buffered

saline (PBS). Body weight, food, and water intake weretinct “injury” phase of disease, which can be studiedseparately from the “repair” phase, as there is in ATS. monitored individually at the beginning and end of each

experiment.In an attempt to further unravel the complexities ofL-Arg manipulations so that decisions concerning use in Sacrifice. At sacrifice, animals were anesthetized with

ether. Following a midline abdominal incision, blood washumans may be grounded in a more complete under-standing of mechanisms, we used the ATS model of drawn through a heparinized syringe from the lower

abdominal aorta. The blood was immediately centri-glomerulonephritis and studied L-Arg supplementationgiven only before and during the period of mesangial fuged, and plasma was frozen at 2708C until analysis.

Peters et al: L-Arg supplementation and renal fibrosis2266

The kidneys were subsequently perfused with 30 ml of DNA and hybridized with DNA probes labeled with 32P-dCTP by random oligonucleotide priming (Boehringerice-cold PBS and harvested. For histological examina-

tion, cortical tissue was snap frozen and fixed in 10% Mannheim Biochemicals, Indianapolis, IN, USA). Theblots were washed in 2 3 SSC, 0.1% SDS at room tem-neutral-buffered formalin. Glomeruli from individual

rats were isolated by a graded sieving technique (150, perature for 15 minutes and in 0.1 3 SSC, 0.1% SDS at558C for 30 minutes. DNA probes used were iNOS125, 106, and 75 mm mesh metal sieves) as described

previously [20]. (kindly provided by J. Pollock, Medical College of Geor-gia) [21] and rat glyceraldehyde-3-phosphate dehydroge-Light microscopy. All microscopic examinations were

performed in a blinded fashion. Three micrometer sec- nase (GAPDH) cDNA [22].Plasma L-arginine levels. These levels were deter-tions of paraffin-imbedded tissue were stained with peri-

odic acid-Schiff (PAS). Cells per glomerulus were mea- mined by a standardized high-performance liquid chro-matography technique [23].sured by counting nuclei in 30 glomeruli of 80 to 100

mm diameter from each animal. Nitrite assay. Plasma NO2/NO3 levels were measuredby the Griess reaction [24] after nitrate reductase treat-Statistical analysis. Data are expressed as mean 6 sem.

Statistical analysis between the groups was performed ment. Briefly, samples were diluted 1:4 with PBS andincubated at room temperature in the presence of 0.05by analysis of variance and subsequent t-testing with

Bonferroni correction for multiple comparison. A P mm NADPH and 0.05 U nitrate reductase (BoehringerMannheim Biochemicals). After 45 minutes of incuba-value of less than 0.05 was considered as significant.tion at room temperature, proteins were precipitated

Methods used in experiment #1: Severity of mesangial with 70% zinc sulfate solution, and samples were centri-cell injury fuged at 14,000 g for five minutes. One hundred microli-

ters of sample were then mixed with 100 ml Griess reagentImmunofluorescence microscopy. One to two mi-crometer sections were made with a cryostat microtome [0.05% N-(1-naphthyl) ethylene diamine dihydrochlo-

ride, 0.5% sulfanilamide in 45% glacial acetic acid] in(Leica CM 1800; Leica Instruments, Nussloch, Ger-many). Glomerular macrophage infiltration was ana- 96-well plates. After 10 minutes of incubation in the

dark, adsorbance was read at 546 nm in an automatedlyzed using a mouse monoclonal antirat monocyte andmacrophage IgG1 (ED-1; Serotec, Oxford, UK) as pri- plate reader (Thermomax; Molecular Devices, Menlo

Park, CA, USA). Standard samples were prepared withmary and a fluorescein isothiocyanate-coupled rat anti-mouse antibody (Jackson ImmunoResearch Labora- sodium nitrate and were also nitrate reductase treated.

Glomerular nitric oxide production. The L-Arg con-tories, West Grove, PA, USA) as secondary antibody.ED-1–positive cells in 20 glomeruli from each animal centration dependency of glomerular NO synthesis was

measured in glomeruli from four nephritic and threewere counted. Glomerular iNOS expression was ana-lyzed using a rabbit antirat iNOS antibody (Alexis Cor- normal rats. Six hours after ATS or PBS injection, glo-

meruli were isolated from each individual animal andporation, San Diego, CA, USA) as primary and a rhoda-mine-coupled antirabbit IgG (Jackson ImmunoResearch resuspended at a density of 2000 glomeruli/ml in phenol-

red-free RPMI 1640 with increasing L-Arg concentra-Laboratories) as secondary antibody. The intensity ofthe glomerular staining was rated as follows: 0 5 no tions (0 to 1000 nmol/ml). Culture medium was prepared

using a RPMI 1640 Select-Aminet Kit (GIBCO BRL,staining; 1 5 weak and spotty staining; 2 5 moderateand segmental staining; 3 5 moderately strong and seg- Grand Island, NY, USA). The medium was supple-

mented with 0.1 U/ml insulin, 100 U/ml penicillin, 100mental or moderate but diffuse staining; and 4 5 strongand diffuse staining. Twenty glomeruli from each animal mg/ml streptomycin, and 25 mm HEPES buffer. After a

48-hour incubation at 378C/5% CO2, the samples werewere scored.RNA preparation and Northern hybridization. Total centrifuged at 14,000 g for five minutes, and the cell-free

supernatant was stored at 2708C until nitrite analysisRNA was extracted by a guanidinium isothiocyantemethod using Trizolw Reagent according to the manu- when 100 ml samples were mixed with 100 ml of Griess

reagent. Standards were prepared using sodium nitrite.facturer’s instructions. Glomeruli from four rats werepooled for subsequent RNA extraction. For Northern

Methods used in experiment #2: Evaluation of theanalysis, RNA was denatured and fractionated by elec-severity of fibrotic disease at day 6trophoresis through a 1.8% agarose gel (30 mg/lane) and

transferred to a Hybond nylon membrane (Amersham Matrix scoring. Glomerular matrix expansion wasevaluated by rating the mesangial matrix occupying eachCorp., Little Chalfont, Buckinghamshire, UK). Nucleic

acids were immobilized by ultraviolet irradiation. Mem- of 30 glomeruli per rat using the following rating system:0 5 0%; 1 5 25%; 3 5 75%; and 4 5 100%.branes were prehybridized with 50% formamide, 10%

Denhardt’s solution, 0.1% sodium dodecyl sulfate Urinary protein excretion. Twenty-four hour urinesamples were taken eight and one days before, on the(SDS), 5 3 SSC, and 200 mg/ml denatured salmon sperm

Peters et al: L-Arg supplementation and renal fibrosis 2267

same day, and on the second and fifth days after ATSor PBS injection. Proteinuria was measured using theBradford method and is expressed as mg per 24 hours.

Production of transforming growth factor-b1, fibro-nectin, and plasminogen activator inhibitor type-1 by day6 glomeruli in culture. Glomeruli were isolated from in-dividual rats by graded sieving and resuspended inDMEM supplemented with 0.1 U/ml insulin, 100 U/mlpenicillin, 100 mg/ml streptomycin, and 25 mm HEPESbuffer. After 72 hours of incubation at 378C/5% CO2,the samples were harvested and stored at 2708C untilanalysis of glomerular TGF-b1, fibronectin, and plasmin-ogen activator inhibitor type-1 (PAI-1) production.TGF-b1 production of cultured glomeruli was measuredafter acid activation using a commercially available kit(Quantikinew; R&D Systems, Minneapolis, MN, USA)according to the manufacturer’s instructions. Fibronectinand PAI-1 synthesis were measured with modified inhibi-tory enzyme-linked immunosorbent assays (ELISA) ac-cording to published methods [25]. Three samples fromeach rat were analyzed and averaged, and then statistics

Fig. 1. Glomerular cell number as an indicator of mesangial cell lysiswere preformed. Each group thus contained data from six hours after injection of anti-thymocyte serum (ATS). Control ani-eight rats. mals were injected with similar volumes of phosphate-buffered saline.

All groups were fed a normal protein diet (22% casein). Drinking waterin group “ATS 1 L-Arg” was supplemented with 1% L-Arg. *P , 0.01vs. control; #P , 0.05 vs. ATS.RESULTS

L-Arginine intake

Supplementation of the drinking water with 1% L-Argtion does not alter macrophage infiltration in responseresulted in an approximate fourfold increase in dailyto ATS-induced disease.L-Arg intake (614 6 46 mg/day) compared with the

Inducible nitric oxide synthase expression. A ninefoldcontrol rats (158 6 13 mg/day) and those injected withincrease in glomerular iNOS staining intensity was seenATS and drinking tap water only (153 6 18 mg/day).with ATS (Fig. 2B). Again, this increase was unchangedFood consumption and weight gain during the experi-by L-Arg supplementation. The lack of an effect ofment were not different between the groups of animalsL-Arg supplementation on iNOS expression at the pro-(data not shown).tein level was confirmed at the RNA level by Northern

Experiment #1 analysis of glomerular iNOS mRNA expression (Fig. 2C, D). Although a 104-fold increase in diseased animalsMesangial cell injury. As shown in Figure 1, ATS in-was seen, no further change was seen with L-Arg supple-jection resulted in a rapid drop in glomerular cell numbermentation. These results indicate that L-Arg supplemen-in six hours. Normal cell counts averaged 64.2 6 1.0 cellstation does not affect iNOS expression at either theper glomerulus, whereas ATS animals’ cell counts weremRNA or protein level, suggesting that any increase inreduced to 53.4 6 0.7 cells per glomerulus (P , 0.01).iNOS activity and NO-mediated tissue injury is due toSupplementation of drinking water with 1% L-Arg re-increased metabolism of L-Arg to NO because of in-sulted in a significant increase in mesangial cell lysis fromcreased substrate availability.53.4 6 0.7 cells per glomerulus with ATS to 50.8 6 0.7

Plasma L-arginine and NOx content. An injection ofcells remaining after six hours (P , 0.05). This resultindicates that L-Arg supplementation leads to a small ATS and dietary L-Arg supplementation led to signifi-

cant changes in plasma L-Arg six hours after diseasebut significant increase in mesangial cell lysis comparedwith rats given ATS alone. induction (Fig. 3A). An injection of ATS was associated

with an average decrease in plasma L-Arg from 121 6Glomerular macrophage infiltration. Macrophages mea-sured six hours after ATS injection were increased nine- 9 nmol/ml to 84 6 13 nmol/ml, (P , 0.01), suggesting

increased utilization of L-Arg during the period of mes-fold with ATS (P , 0.01; Fig. 2A). A small but nonsig-nificant increase in macrophages was seen with ATS 1 angial cell lysis. Supplementation with L-Arg more than

overcame this decrease, resulting in plasma L-Arg levelsL-Arg. These results suggest that L-Arg supplementa-

Peters et al: L-Arg supplementation and renal fibrosis2268

Fig. 2. Glomerular monocyte/macrophage in-filtration and inducible nitric oxide synthase(iNOS) expression six hours after disease in-duction. (A) Number of ED-1–positive cellsper glomerulus. (B) Glomerular staining foriNOS. (C) Representative Northern blots ofiNOS and GAPDH mRNA. (D) Quantitationof glomerular iNOS mRNA levels normalizedto GAPDH expression (N 5 4 blots). *P ,0.01 vs. control.

Fig. 3. L-Arginine (L-Arg) concentrations(A) and arterial nitrite/nitrate (NOx; B) levelssix hours after disease induction. All groupswere fed a normal protein diet. Drinking waterin group “ATS 1 L-Arg” was supplementedwith 1% L-Arg. *P , 0.01 vs. control; *P ,0.01 vs. ATS, #P , 0.05 vs. ATS.

that were 137% of normal (166 6 12 vs. 121 6 9 nmol/ evidence for increased utilization of plasma L-Arg dur-ing mesangial cell lysis came from measurement ofml) and 198% higher than in the ATS alone group (166 6

12 vs. 84 6 13 nmol/ml, P , 0.001; Fig. 3A). Further plasma NOx content, as shown in Figure 3B. Despite

Peters et al: L-Arg supplementation and renal fibrosis 2269

Fig. 4. In vitro nitrite/nitrate production by glomeruli harvested sixhours after disease induction in relation to the extracellular L-Argconcentration. Symbols are: (j) ATS; (r) control.

their lower plasma L-Arg level, the plasma of ATS alone Fig. 5. Time course of proteinuria in rats injected with anti-thymocyteserum (ATS) or phosphate-buffered saline (PBS; control). Symbolsrats contained significantly (P , 0.01) more nitrite andare: (d) control; (r) ATS; (j) ATS 1 L-Arg. Drinking water in thenitrate, indicative of greater in vivo NO generation, thangroup “ATS 1 L-Arg” was supplemented with 1% L-Arg for seven

normal animals. L-Arg supplementation resulted in a days before until 16 hours after ATS injection. *P , 0.05 vs. ATS.further significant increase in NOx compared with ATSalone, consistent with additional L-Arg utilization byiNOS in glomeruli during mesangial cell lysis (ATS 5

after ATS injection in vivo should be very much higher31.0 6 1.3 nmol/ml; ATS 1 L-Arg 5 35.7 6 1.1 nmol/in ATS 1 L-Arg compared with ATS alone animals.ml, P , 0.05).

Dependence of glomerular nitric oxide production onExperiment #2l-arginine concentration. Because the previously men-

Proteinuria. Rats given L-Arg supplemented drinkingtioned data suggested that L-Arg–supplemented rats hadwater before and during ATS-induced mesangial cellhigher L-Arg utilization but similar glomerular iNOSlysis developed significantly higher proteinuria at day 2expression compared with ATS-injected animals with aof disease (P , 0.05; Fig. 5). This difference in protein-normal L-Arg intake, we determined the NOx produc-uria was not seen at day 5, suggesting that it results fromtion of nephritic glomeruli isolated six hours after diseasedifferences in injury that are not carried through to laterinduction in the presence of increasing L-Arg concentra-phases of disease.tions. As shown in Figure 4, the NOx synthesis of ne-

Fibrotic disease at day 6. From past experience withphritic glomeruli was highly dependent on the extracellu-the ATS model, we know that extracellular matrix accu-lar L-Arg concentration. A dramatic increase in NOxmulation at day 6 is greatly increased, although the peaksynthesis was seen starting at 50 nmol/ml, which peakedis several days later in Sprague-Dawley rats. However,at 500 nmol/ml L-Arg. In contrast, NOx production byother measures of disease, including production ofglomeruli from control rats did not change with differentTGF-b1, fibronectin, and PAI-1, which are very reliableL-Arg concentrations in the culture medium. Taken withvariables of disease severity, peak at day 6. As seen inthe data on plasma L-Arg levels shown in Figure 3A,Figure 6A, a large increase in the quantity of pathologicalthese data indicate that at six hours after ATS injection,matrix is seen with ATS alone at day 6 (P , 0.01) com-plasma L-Arg levels are in the range in which smallpared with control. L-Arg supplementation before andincreases in L-Arg concentration lead to large differ-during the period of mesangial cell lysis results in a smallences in NO generation. Thus, glomerular iNOS utiliza-

tion of substrate L-Arg and NO generation six hours but significant (P , 0.05) increase in matrix accumula-

Peters et al: L-Arg supplementation and renal fibrosis2270

Fig. 6. Effect of L-Arg supplementation onthe anti-thymocyte serum (ATS)-induced tis-sue injury as measured by the disease severitysix days after injection with ATS or phos-phate-buffered saline (PBS; control). Drink-ing water in the group “ATS 1 L-Arg” wassupplemented with 1% L-Arg for seven daysbefore until 16 hours after ATS injection.Shown are the following: glomerular matrixaccumulation (A) and glomerular productionof TGF-b1 (B), fibronectin (C), and PAI-1(D) in culture supernatant after 72 hours.*P , 0.01 vs. control, #P , 0.05 vs. ATS.

tion. Similar patterns were seen for glomerular produc- in this study, when dietary intake or endogenous utiliza-tion of TGF-b1, fibronectin, and PAI-1 (Fig. 6 B, C, and tion is increased as during inflammation, plasma L-ArgD). Thus, the small increase in mesangial cell lysis seen levels change.at six hours with L-Arg supplementation (Fig. 1) results With L-Arg as its substrate, iNOS generation of largein small but significant increases in fibrotic disease at amounts of NO has been identified as a key mediatorday 6. of the mesangial cell injury that follows the injection of

Glomerular cell number. The day 6 glomerular cell ATS [4]. A similar critical role of NO in renal tissuenumbers for the three groups of rats were 64.6 6 3.0 injury has been shown in models of acute ischemic renalcells per glomerulus for normal control, which increased failure and lupus nephritis [28, 29] and is suggested into 75.4 6 1.8 for the ATS alone group (P , 0.01). A experimental kidney transplant rejection [30]. This studyfurther significant (P , 0.05) increase to 82.5 6 2.8 for demonstrates that in ATS-induced glomerulonephritis,the ATS 1 L-Arg group was seen. Taken with the data dietary L-Arg supplementation given during the time ofpresented in Figure 1, this indicates that compared with cell lysis results in a significant augmentation of mesan-ATS alone, L-Arg feeding during the period of mesangial gial cell damage (Fig. 1). L-Arg administration did notcell lysis results in fewer cells per glomerulus at six hours appear to reduce the inflammatory response followingbut a greater number of cells per glomerulus at day 6. ATS injection because macrophage infiltration and glo-

merular iNOS and protein levels did not differ in L-Arg–supplemented and –nonsupplemented rats (Fig. 2). TheDISCUSSIONED-1 staining here contrasts with a report that L-Argl-Arginine has been classified as a semi-essentialsupplementation reduces renal macrophage infiltrationamino acid [26]. Under normal physiological conditions,in obstructive nephropathy and puromycin-induced ne-endogenous L-Arg synthesis is sufficient to maintainphrosis [12].whole body L-Arg metabolic homeostasis. Endogenous

Two measures suggest that at six hours after diseaseL-Arg synthesis occurs primarily in the proximal tubuleinduction, utilization of L-Arg to produce NO is in-of the kidney using the nonprotein amino acid l-citrul-creased by ATS and further increased when ATS is givenline as precursor [27]. The rate of L-Arg production in

renal tubules is constant, however, and, as demonstrated to L-Arg–supplemented rats. First, plasma L-Arg levels

Peters et al: L-Arg supplementation and renal fibrosis 2271

are lower and plasma NOx content is higher in ATS results from the reaction of the cytotoxic radical peroxi-nitrite and the amino acid tyrosine in proteins. Peroxini-alone rats (Fig. 3), and second, plasma NOx levels are

significantly higher in ATS rats with L-Arg supplementa- trite is an important mediator of NO cytotoxicity andis a product of the reaction of NO with superanions.tion than in ATS rats drinking tap water only (Fig. 3).

Further confirmation of these findings were in vitro data Immunohistochemical studies have shown evidence forincreased staining of nitrotyrosine in a number of humanin which nephritic glomeruli showed markedly higher

NOx production at L-Arg concentrations comparable to diseases, including atherosclerosis, myocardial ischemia,inflammatory bowel disease, and septic lung disease [43].those seen in the plasma of animals with L-Arg supple-

mentation. These results strongly support the concept In human kidney disease, Kashem et al have detectedincreased expression of iNOS mRNA and protein in thethat the detrimental effect of L-Arg supplementation on

ATS-induced glomerulonephritis is directly mediated by majority of patients with mesangioproliferative glomeru-lonephritis, primarily IgA nephropathy [44]. The iNOSgreater substrate availability. When L-Arg is stopped

16 hours after ATS injection and the “repair” phase enzyme was colocalized with areas of macrophage infil-tration, but was also found in resident renal cells. Inter-proceeds with normal L-Arg intake, the increased injury

that occurred by six hours continues to be reflected by estingly, the iNOS expression correlated with the sever-ity of histological signs of renal damage, as well as thesmall but significant increases in glomerular cell number,

matrix accumulation, and production of TGF-b1, fibro- loss of renal function [44]. We have also reported apreliminary study showing significantly increased iNOSnectin, and PAI-1 at six days of disease (Fig. 6). Other

disease models in which L-Arg administration worsens protein expression in glomeruli and the tubulointerstit-ium of patients with acute and chronic renal transplanttissue injury appear to share the feature that NO syn-

thesis is greatly increased through induction of NO rejection (abstract; Ikegaya et al, J Am Soc Nephrol7:1563, 1996). Finally, nitrotyrosines have recently beensynthase. These include immune-mediated dermal vas-

culitis, hypoxic lung injury, toxic colitis, and brain reper- detected in failing human kidney transplants [45]. There-fore, it appears likely that induction of iNOS plays a rolefusion [31–34].

In terms of the relevance of rat models of tissue injury, in a number of human disorders.In addition to being a substrate for iNOS, L-Arg is athere has been some debate about whether humans re-

spond with similar induction of iNOS. In rodents, the substrate for several other NO synthases, including aconstitutive enzyme produced by endothelial cells [35].central role of iNOS expression in host defense, tumor

killing, and autoimmune tissue destruction in vivo is un- Renal endothelial cells produce small quantities of NOby this enzyme. This endothelial NO serves as a criticaldisputed [35]. However, a high output iNOS pathway has

been difficult to show in humans. Rodent macrophages in vasodilator and may play a protective role in both normalglomeruli and those exposed to increased pressure. It isvitro produce large amounts of NO in response to stimuli

such as lipopolysaccharides or cytokines such as inter- generally believed that the mechanism underlying thetherapeutic effects seen with dietary L-Arg supplemen-feron-g or tumor necrosis factor-a, whereas human mac-

rophages or monocytes in vitro show little or no response tation in a number of models of renal disease is increasedendothelial NO generation [46].[36]. On the other hand, there is considerable recent in

vivo evidence that this pathway is operating in humans. l-Arginine is also a substrate for the generation ofagmatine, polyamines, and l-proline [27]. Both cell pro-The expression of iNOS and increased NO synthesis

has been shown in cells from several human diseases, liferation, dependent on polyamine generation, and col-lagen synthesis, dependent on l-proline production, areincluding pulmonary alveolar macrophages from pa-

tients with tuberculosis, urinary neutrophils from pa- major characteristics of many renal diseases. A timecourse study of the activity of these pathways in ATStients with bacterial urinary tract infection and synovio-

cytes, chondrocytes and endothelial cells from patients shows that, in addition to the iNOS pathway, both thepolyamine and the l-proline pathways are highly acti-with inflammatory arthritis [37–39]. In HIV type 1 en-

cephalitis, the presence of iNOS has been reported in vated [5], suggesting that cell proliferation and collagensynthesis could be affected by L-Arg intake. Supportingbrain astrocytes [40], and its induction has been recently

linked to the HIV protein gP41 [41]. Human natural this hypothesis are data in which dietary restriction ofL-Arg, started one day after ATS disease induction, atkiller cell cytotoxicity appears to be mediated through

a NO pathway [42]. Similar to rodents, the killing of the the start of the “repair” phase of disease, resulted in lessfibrosis at day 7 [9].pathogen Leishmania major by human macrophages is

NO dependent but requires the ligation of a specific In apparent contrast to these data, there are now manystudies showing a therapeutic effect of high L-Arg intakesurface antigen [36].

Another indication for an active tissue damaging NO on renal disease [10–19]. That this therapeutic approachmay have merit is undisputed. However, in most of thesepathway in humans has been provided by immunohisto-

chemical detection of nitrotyrosine [43]. Nitrotyrosine studies, supplementation began after induction of dis-

Peters et al: L-Arg supplementation and renal fibrosis2272

Gath I, Kleinert H: Nitric oxide synthase isozymes. Hypertensionease, suggesting that this positive effect may act primarily23:1121–1131, 1994

on the “repair” phase. Indeed, to complicate things fur- 8. Waddington S, Cook HT, Reaveley D, Jansen A, Cattell V:L-arginine depletion inhibits glomerular nitric oxide synthesis andther, we have recently found a significant therapeuticexacerbates rat nephrotoxic nephritis. Kidney Int 49:1090–1096,effect of L-Arg supplementation on ATS day 6 fibrotic1996

disease if treatment is started 24 hours after ATS injec- 9. Narita I, Border WA, Ketteler M, Ruoslahti E, Noble NA:L-arginine may mediate the therapeutic effects of low protein diets.tion (unpublished observation). It appears obvious thatProc Natl Acad Sci USA 92:4552–4556, 1995if both L-Arg restriction and L-Arg supplementation

10. Reyes AA, Pukerson ML, Karl I, Klahr S: Dietary supplementa-have positive therapeutic effects on the “repair” phase tion with L-arginine ameliorates the progression of renal disease

in rats with subtotal nephrectomy. Am J Kidney Dis 20:168–176,of ATS disease, then these effects act through different1992L-Arg pathways.

11. Katoh T, Takahashi K, Klahr S, Reyes AA, Badr KF: DietaryIn summary, using the ATS model of glomerulone- supplementation with L-arginine ameliorates glomerular hyperten-

sion in rats with subtotal nephrectomy. J Am Soc Nephrol 4:1690–phritis, we have shown that L-Arg supplementation,1694, 1994given before and during the period of mesangial cell

12. Reyes AA, Porras BH, Chasalow FI, Klahr S: L-arginine de-injury, leads to small but significant increases in cell lysis creases the infiltration of the kidney by macrophages in obstructive

nephropathy and puromycin-induced nephrosis. Kidney Int 45:that are reflected in more severe fibrotic disease at day1346–1354, 19946. The data strongly support the hypothesis that supple-

13. Reyes AA, Karl IE, Yates J, Klahr S: Low plasma and renalmental L-Arg provides sufficiently high levels of sub- tissue levels of L-arginine in rats with obstructive nephropathy.

Kidney Int 45:782–787, 1994strate to iNOS, which is transiently but dramatically in-14. Morrissey JJ, Ishidoya S, McCracken R, Klahr S: Nitric oxidecreased, to generate additional cytotoxic NO leading to

generation ameliorates the tubulointerstitial fibrosis of obstructivemeasurable increases in mesangial cell lysis. Because nephropathy. J Am Soc Nephrol 7:1–11, 1996

15. Reyes AA, Karl IE, Kissane J, Klahr S: L-arginine administra-iNOS expression is also increased in many human dis-tion prevents glomerular hyperfiltration and decreases proteinuriaeases, because L-Arg is used in many metabolic path-in diabetic rats. J Am Soc Nephrol 4:1039–1045, 1993

ways, and because current data in animals do not provide 16. Chen PY, St. John PL, Kirk KA, Abrahamson DR, Sanders PW:Hypertensive nephrosclerosis in the Dahl/Rapp rat: Initial sites ofus with definite guidelines for when L-Arg supplementa-injury and effect of dietary L-arginine supplementation. Lab Investtion may be beneficial, and when it may be harmful, we68:174–184, 1993

conclude that L-Arg supplementation should be used in 17. De Nicola L, Thomson SC, Wead LM, Brown MR, Gabbai FB:Arginine feeding modifies cyclosporine nephrotoxicity in rats.humans only with great caution.J Clin Invest 92:1859–1865, 1993

18. Ingram A, Parbtani A, Thai K, Ly H, Shankland SJ, MorrisseyACKNOWLEDGMENTS G, Scholey JW: Dietary supplementation with L-arginine limits

cell proliferation in the remnant glomerulus. Kidney Int 48:1857–This work was supported by National Institute of Diabetes and1865, 1995Digestive and Kidney Diseases (NIDDK) grant DK 49342 to N.A.N.

19. Ashab I, Peer G, Blum M, Wollman Y, Chernihovsky T, Hass-and DK 43609 and DK 49374 to W.A.B. H.P. was supported by aner A, Schwartz D, Cabili S, Silverberg D, Iaina A: Oral admin-grant from the Deutscher Akademischer Austauschdienst (DAAD)istration of L-arginine and captopril in rats prevents chronic renaland a matching fellowship from the National Kidney Foundation offailure by nitric oxide production. Kidney Int 47:1515–1521, 1995Utah.

20. Okuda S, Languino LR, Ruoslahti E, Border WA: Elevatedexpression of transforming growth factor-beta and proteoglycanReprint requests to Nancy A. Noble, Ph.D., Division of Nephrology,production in experimental glomerulonephritis: Possible role inUniversity of Utah School of Medicine, Salt Lake City, Utah 84132,expansion of the mesangial extracellular matrix. J Clin InvestUSA.86:453–462, 1990E-mail: Nancy.Noble@hsc.utah.edu

21. Nakane M, Pollock JS, Klinghofer V, Basha F, Marsden PA,Hokan A, Ogura T, Esumi H, Carter GW: Functional expressionof three isoforms of human nitric oxide synthase in baculovirus-REFERENCESinfected insect cells. Biochem Biophys Res Commun 206:511–517,

1. Border WA, Noble NA: Transforming growth factor-b in tissue 1995fibrosis. N Engl J Med 331:1286–1292, 1994 22. Fort P, Marty L, Piechaczyk M, Sabrouty SE, Dani C, Jeanteur

2. Border WA, Noble NA: Evidence that TGF-b should be a thera- PH, Blanchard JM: Various rat adult tissues express only onepeutic target in diabetic nephropathy. (editorial) Kidney Int major mRNA species from the glyceraldehyde-3-phosphate-dehy-54:1390–1391, 1998 drogenase multigenic family. Nucleic Acids Res 13:1431–1442, 1985

3. Noble NA, Border WA: Angiotensin II in renal fibrosis: Should 23. Schuster R: Determination of amino acids in biological, pharma-TGF-b rather than blood pressure be the therapeutic target? Semin cological, plant and food samples by automated precolumn derivat-Nephrol 17:455–466, 1997 ization and high-performance liquid chromatography. J Chro-

4. Narita I, Border WA, Ketteler M, Noble NA: Nitric oxide matogr 431:271–284, 1988mediates immunologic injury to kidney mesangium in experimental 24. Green LC, Wagner DA, Glogowski J, Skipper PL, Wishnok JS,glomerulonephritis. Lab Invest 72:17–24, 1995 Tannenbaum SR: Analysis of nitrate, nitrite and [15N]nitrate in

5. Ketteler M, Ikegaya N, Brees DK, Border WA, Noble NA: biological fluids. Anal Biochem 126:131–138, 1982L-arginine metabolism in immune-mediated glomerulonephritis in 25. Rennard SI, Berg R, Martin GR, Foidart JM, Robey PG: En-the rat. Am J Kidney Dis 28:878–887, 1996 zyme-linked immunoassay (ELISA) for connective tissue compo-

6. Bogle RG, Baydoun AR, Pearson JD, Moncada S, Mann GE: nents. Anal Biochem 104:205–214, 1980L-Arginine transport is increased in macrophages generating nitric 26. Castillo L, Ajami A, Branch S, Chapman TE, Yu YM, Burkeoxide. Biochem J 284:15–18, 1992 JF, Young YR: Plasma arginine kinetics in adult man: Response

to an arginine-free diet. Metabolism 43:114–122, 19947. Forstermann U, Closs EI, Pollock JS, Nakane M, Schwarz P,

Peters et al: L-Arg supplementation and renal fibrosis 2273

27. Peters H, Noble NA: Dietary L-arginine in renal disease. Semin Mumford R, Weidner JR, Calaycay J, Boechat N: Induciblenitric oxide synthase in pulmonary alveolar macrophages fromNephrol 6:567–575, 1996

28. Noiri E, Peresleni T, Miller F, Goligorsky MS: In vivo targeting patients with tuberculosis. J Exp Med 183:2293–2302, 199638. Sakurai H, Kohsaka H, Liu MF, Higashiyama H, Hirata Y,of inducible NO synthase with oligodeoxynucleotides protects rat

kidney against ischemia. J Clin Invest 97:2377–2383, 1996 Kanno K, Saito I, Miyasaka N: Nitric oxide production and induc-ible nitric oxide synthase expression in inflammatory arthritidis.29. Weinberg JB, Granger DL, Pisetsky DS, Seldin MF, Misukonis

MA, Mason SN, Pippen AM, Ruiz P, Wood ER, Gilkeson GS: J Clin Invest 96:2357–2363, 199539. Wheeler MA, Smith SD, Garcia Cardena G, Nathan CF, WeissThe role of nitric oxide in the pathogenesis of spontaneous murine

autoimmune disease: Increased nitric oxide production and nitric RM, Sessa WC: Bacterial infection induces nitric oxide synthasein human neutrophils. J Clin Invest 99:110–116, 1997oxide synthase expression in MRL-lpr/lpr mice, and reduction of

spontaneous glomerulonephritis and arthritis by orally adminis- 40. Bukrinsky MI, Nottet HS, Schmidtmayerova H, Dubrovsky L,Flanagan CR, Mullins ME, Lipton SA, Gendelman HE: Regu-tered NG-monomethyl-L-arginine. J Exp Med 179:651–660, 1994

30. Cattell V, Smith J, Jansen A, Riveros-Moreno V, Moncada S: lation of nitric oxide synthase activity in human immunodeficiencyvirus type I (HIV-1)-infected monocytes: Implications for HIV-Localization of inducible nitric oxide synthase in acute renal allo-

graft rejection in the rat. Transplantation 58:1399–1402, 1994 associated neurological disease. J Exp Med 181:732–745, 199541. Adamson DC, Wildemann B, Sasaki M, Glass JD, McArthur31. Mulligan MS, Moncada S, Ward PA: Protective effects of inhibi-

tors of nitric oxide synthase in immune complex-induced vasculitis. JC, Christov VI, Dawson TM, Dawson VL: Immunologic NOsynthase: Elevation in severe AIDS dementia and induction byBr J Pharmacol 107:1159–1162, 1992

32. Zhang F, Casey RM, Ross ME, Iadecola C: Aminoguanidine HIV-1 gp41. Science 274:1917–1921, 199642. Xiao L, Eneroth PH, Qureshi GA: Nitric oxide synthase pathwayameliorates and L-arginine worsens brain damage from intralumi-

nal middle cerebral artery occlusion. Stroke 27:317–323, 1996 may mediate human natural killer cell cytotoxicity. Scand J Immu-nol 42:505–511, 199533. Nozik ES, Huang Y-CT, Piantadosi CA: L-arginine enhances

injury in the isolated rabbit lung during hyperoxia. Respir Physiol 43. Beckman JS, Koppenol WH: Nitric oxide, superoxide, and peroxy-nitrite: The good, the bad, and ugly. Am J Physiol 271:C1424–100:63–74, 1995

34. Neilly PJD, Gardiner KR, Rowlands BJ: Experimental colitis C1437, 199644. Kashem A, Endoh M, Yano N, Yamauchi F, Nomoto Y, Sakaiis ameliorated by inhibition of nitric oxide synthase activity. (letter)

Gut 38:475, 1996 H: Expression of inducible-NOS in human glomerulonephritis: Thepossible source is infiltrating monocytes/macrophages. Kidney Int35. Moncada S, Higgs A: The L-arginine-nitric oxide pathway. N Engl

J Med 329:2002–2012, 1993 50:392–399, 199645. MacMillan-Crow LA, Crow JP, Kerby JD, Beckman JS, Thomp-36. Vouldoukis I, Riveros Moreno V, Dugas B, Quaaz F, Becherel

P, Debre P, Moncada S, Mossalayi MD: The killing of Leishmania son JA: Nitration and inactivation of manganese superoxide dismu-tase in chronic rejection of human renal allografts. Proc Natl Acadmajor by human macrophages is mediated by nitric oxide induced

after ligation of the Fc epsilon RII/CD23 surface antigen. Proc Sci USA 93:11853–11858, 199646. Klahr S, Morrissey J: Renal disease: The two faces of nitric oxide.Natl Acad Sci USA 92:7804–7808, 1995

37. Nicholson S, Bonecini Almeida M, Lapa S, Nathan C, Xie QW, Lab Invest 72:1–3, 1995