Kidney International, Vol. 55 (1999), pp. 945–955

Mycophenolate mofetil prevents the progressive renal failureinduced by 5/6 renal ablation in rats

FREDDY ROMERO, BERNARDO RODRIGUEZ-ITURBE, GUSTAVO PARRA, LUISANDRA GONZALEZ,JAIME HERRERA-ACOSTA, and EDILIA TAPIA

Renal Service and Laboratory, Hospital Universitario and Instituto de Investigaciones Biomedicas, Maracaibo, Venezuela, andthe Department of Nephrology and Renal Laboratory, Instituto Nacional de Cardiologıa Ignacio Chavez, Mexico City, Mexico

tration with CD43- and ED1-positive cells in glomeruli andMycophenolate mofetil prevents the progressive renal failureinterstitium was two to five times lower in MMF-treated ratsinduced by 5/6 renal ablation in rats.(P , 0.01). Expression of adhesion molecules CD18 and CD11bBackground. Extensive renal ablation is associated with pro-was similarly reduced.gressive sclerosis of the remnant kidney. Because lymphocytes

Conclusion. MMF ameliorates the progressive renal damageand monocytes accumulate in the remnant kidney, it is likelythat they play a role in the renal scarring. Therefore, we treated in the remnant kidney after 5/6Nx. This effect is associated withrats with 5/6 nephrectomy (5/6Nx) with mycophenolate mofetil a reduction in the infiltration of lymphocytes and monocytes,(MMF), a drug that has an antiproliferative effect and that whereas glomerular hypertrophy and systemic hypertensionsuppresses the expression of intercellular adhesion molecules. are unchanged.

Methods. Sprague-Dawley rats with 5/6Nx received MMF(30 mg · kg21 · day21 by daily gastric gavage, N 5 15) or vehicle(N 5 16). Ten additional rats were sham operated. All rats

The progression of renal damage resulting from re-were fed a 30% protein diet. Body weight, serum creatinine,and urinary protein excretion were determined weekly. Lipid duced nephron mass has been extensively studied in theperoxidation, as a measure of oxidative stress observed by 5/6Nx model in the rat, as described originally by Shima-urinary malondialdehyde determinations, was performed every mura and Morrison [1]. Classic studies by Hostetter ettwo weeks. Histologic studies were done in the remnant kidney

al demonstrated that this experimental model was char-four weeks (9 rats from the vehicle-treated group, 7 rats fromacterized by progressive glomerulosclerosis and chronicthe MMF group, and 5 sham-operated rats) and eight weeks

after surgery (the remaining rats). Glomerular volume, sclero- renal failure [2]. The remnant nephron units developedsis in glomeruli (segmental and global) and interstitium (semi- glomerular hypertension, increased single nephron glo-quantitative scale), infiltrating lymphocytes and macrophages merular filtration rate, and glomerular hypertrophy, and(CD43- and ED1-positive cells), and expression of adhesion

over the previous 15 years, all of these factors have beenmolecules (CD54, CD18, and CD11b) were analyzed.postulated to play a pathogenetic role in the progressiveResults. MMF treatment prevented the progressive incre-

ment in serum creatinine and the proteinuria observed in the scarring of the kidney [3]. Surviving nephrons have in-5/6 nephrectomized rats during the eight weeks of observation creased oxygen consumption and are subjected to oxi-(P , 0.01). Weight gain was comparable in the MMF-treated

dant stress [4].and sham-operated rats, whereas weight gain was decreasedBecause the initial event in the subtotal renal ablationin untreated 5/6 nephrectomized rats. Excretion of malondial-

dehyde increased after surgery but returned sooner to control model is an increase in glomerular pressure, treatmentlevels in the MMF-treated rats. Increments in glomerular size strategies have usually been directed to correct this he-and mean arterial blood pressure induced by renal ablation modynamic disturbance, specifically, to reduce intra-were not modified by MMF treatment. Eight weeks after sur-

glomerular pressure. Correction of this pathophysiologicgery, segmental sclerosis was present in 48.4 6 8.35% (6 sd)event and normalization of permselectivity properties ofglomeruli in the vehicle-treated group versus 25 6 10.5% in

the MMF-treated group (P , 0.001). Interstitial fibrosis was the glomerular capillary barrier [5] are major mecha-reduced significantly with MMF treatment (P , 0.001). Infil- nisms underlying the protective effect on renal function

observed with angiotensin-converting enzyme inhibitionand low-protein diet in several models of renal diseaseKey words: sclerosis, remnant kidney, scarring, adhesion molecules,

proteinuria, interstitial fibrosis. in the experimental animal and in humans [6].Less commonly, treatments have been directed to

Received for publication July 20, 1998modify late events in the development of glomeruloscle-and in revised form October 19, 1998

Accepted for publication October 20, 1998 rosis, that is, events deriving from the hemodynamic andhypertrophic glomerular changes. For example, heparin 1999 by the International Society of Nephrology

945

Romero et al: Mycophenolate mofetil therapy in renal ablation946

treatment, which prevents microthrombosis [7], as well rat chow containing 30% protein (Purina, St. Louis, MO,as Ticlopidine, a drug with a wide spectrum of inhibitory USA) and tap water.actions on platelets and platelet–cell interactions, retards After obtaining blood samples for serum creatininethe progression of renal failure in the remnant kidney and 24-hour urine for determination of proteinuria andmodel [8]. More recently, Shimizu et al have reported urinary malondialdehyde excretion, 40 rats had 5/6Nxthat administration of Pirfenidone, a compound that pre- (mentioned later here), and 10 rats had a sham operation.vents and reverses extracellular matrix accumulation, Nine rats in the group that had 5/6Nx died during surgeryameliorates the histologic damage in the remnant kidney or in the following 48 hours. Five days after surgery,of the rat with 5/6Nx [9]. the surviving rats in the 5/6Nx group were randomly

Largely undefined is the role played by infiltrating subdivided in two groups, one of which received MMFlymphocytes and macrophages in the development of (5/6NxMMF group, N 5 15) the other one received vehi-glomerulosclerosis triggered by a reduction of nephron cle (5/6Nx, N 5 16). Four of these rats died in the subse-mass. There is limited information on lymphocyte accu- quent four weeks (two from the 5/6NxMMF group andmulation in this model. However, monocyte/macrophage two from the 5/6Nx group). Histologic studies were doneinfiltration occurs early [10, 11], and their numbers in- four and eight weeks after surgery. At the four-weekcrease fourfold to fivefold in the glomeruli and tubuloin- endpoint, nine rats from the 5/6Nx group, seven ratsterstitial areas of the remnant kidney [12, 13]. Therefore, from the 5/6NxMMF group, and five sham-operated ratsit is likely that these cells play a significant role in the were sacrificed. The remaining rats (5 rats from the sham-subsequent development of fibrosis. In fact, cytokines operated group, 5 rats from the 5/6Nx group, and 6 ratsproduced by infiltrating cells appear to mediate renal from the 5/6NxMMF group) were sacrificed eight weeksscarring in this experimental model [10]. after surgery.

Mycophenolate mofetil (MMF) is a new immunosup- Body weight, serum creatinine, and urinary proteinpressant agent in which the active component, mycophe- excretion were determined weekly. Urinary malondialde-nolic acid, is an inhibitor of inosine 59-monophosphate hyde (MDA) determinations were done every two weeks.dehydrogenase, which controls the synthesis of guano- Histologic and immunohistologic studies were done insine triphosphate [14, 15]. MMF depletes guanosine tri- remnant kidneys and in sham-operated animals at thephosphate pools in lymphocytes and monocytes and sup- time of sacrifice.presses the de novo synthesis of purines, thereby exertinga selective and reversible antiproliferative activity on Dose of mycophenolate mofetilthese cells. In addition, the expression of intercellular

Preliminary studies were done to determine theadhesion molecules, which is critical for the migration

maximal tolerated dose of MMF. Doses higher than 30of leukocytes from circulation and their subsequent accu-mg · kg21 · day21 were usually associated with diarrhea;mulation in the tissues, is suppressed by this drug [16].consequently, this dose was chosen. Daily gastric gavageTherefore, we decided to investigate whether treat-was used to administer 30 mg · kg21 · day21 of MMFment with MMF would reduce the renal cellular infiltra-diluted in 200 ml of water to the rats of the 5/6NxMMFtion and the expression of adhesion molecules, as wellgroup. Because mycophenolate (CellCept; Roche Phar-as the oxidant-induced lipid peroxidation observed aftermaceutical Co., Caracas, Venezuela) is insoluble in wa-5/6Nx, and whether these modifications would be associ-ter, we prepared each dose individually and agitated theated with improvement in the progression of renal dis-mixture vigorously to obtain a suspension, and then theease. We found that deterioration of renal function andsuspension was immediately given to the rat before theproteinuria, which is characteristic of this model, wasMMF precipitated. Rats in the 5/6Nx group and shamprevented and that the hallmark histologic changes weregroups received the same volume of water by the samemarkedly improved. Because the hematologic, liver, androute of administration daily.renal toxicity of MMF are relatively minor [17, 18] and

this drug is used with relative safety in the long-termSurgical procedurestreatment of renal transplantation, MMF could be poten-

Subtotal renal ablation was done in a single surgicaltially useful to retard the progressive damage associatedprocedure by subcapsular removal of the right kidneywith reduced nephron mass in humans.and selective infarction of two thirds of the left kidneyby ligation of the posterior and one or two of the anterior

METHODS extrarenal branches of the renal artery, to leave approxi-Design of the study and experimental groups mately one sixth of the total kidney tissue mass. Sham

operations consisted in laparotomy and manipulation ofStudies were done in 50 male Sprague-Dawley ratsthe kidneys and renal pedicle without destruction of re-weighing 200 to 280 g at the start of the experiments.

Throughout the study, all rats were fed ad libitum with nal tissue.

Romero et al: Mycophenolate mofetil therapy in renal ablation 947

Determination of lipid peroxidation sclerosis in superficial and juxtamedullary sections. Glo-merulosclerosis was evaluated in PAS-stained tissue andWe used the method of Ohkawa et al to determinewas defined as the increment of PAS-positive material,thiobarbituric acid reactive substances, an index of lipidwith loss of cellular elements, collapse of capillary lumina,peroxidation [19]. MDA was measured in 24-hour urineand entrapment of amorphous hyaline material associ-centrifuged at 2000 r.p.m. and processed the same day.ated or not with adhesions to Bowman’s capsule. Sclero-Four hundred microliters of the sample were added tosis involving more than 80% of the glomerular tuft wasa reaction mixture consisting of 200 ml of 8.1% sodiumconsidered global, and sclerosis involving less than 80%dodecyl sulfate, 1.5 ml of 20% acetic acid (pH 3.5), 1.5was considered segmental. No attempt was made to sub-ml of 0.5% thiobarbituric acid, and 400 ml of bidistilledclassify the segmental lesions in accordance with thewater. This mixture was heated at 958C in water bathpercentage involvement of the glomerulus. At least 100for 60 minutes. After removal from the water bath, 1 mlglomeruli were evaluated per biopsy. Results are ex-of water and 5 ml N-butanol-pyridine were added, andpressed as the percentage glomeruli examined (more thanthe mixture was agitated and centrifuged at 2000 g for100 per biopsy) showing segmental or global sclerosis.15 minutes. The upper organic layer was pipetted, and

Tubulointerstitial lesions. Severity of tubular dilationthe absorbance of this fraction was read at 532 nm in aand atrophy, as well as tubulointerstitial infiltration (he-Shimatzu model UV21100 S spectrophotometer (Kyoto,matoxylin and eosin staining) and fibrosis (trichromeJapan). Malondialdehyde bis-dimethyl acetal was usedstaining), were evaluated individually, determining theas external standard. Results were expressed as nmol ofpercentage in the biopsy section affected by the corre-MDA excreted in 24 hours.sponding finding with the help of a grid-fitted eyepiece.

Laboratory determinations Gradation of damage (0 to 51) was done by modifyingthe scale used by Zoja et al as follows: 0 5 changes inProteinuria was determined in 24-hour urine collec-less than 10% of the histologic section; 11 5 changestions by the sulfosalicylic acid method and the serumin up 20% of the section; 21 5 changes in up to 40%creatinine by modification of the Jaffe reaction in auto-of the section; 31 5 changes in up to 60% of the section;analyzer methodology.41 5 changes in up to 80% of the section; 51 5 changes

Tissue preparation in more than 80% of the section [8]. The average scoreobtained in the sections of each biopsy (10 to 15 sectionsFrom every animal, coronal sections of remnant renalper biopsy) was used as the score of the biopsy.tissue were obtained at the time of sacrifice. One frag-

ment was fixed in 15% formalin, and 10 were embeddedImmunofluorescence studiesin Paraplast Plus (Monoject, Sherwood Medical Scien-

Identification of intercellular adhesion molecules wastific Division, St. Louis, MO, USA); the rest were in-done in frozen 3 mm sections by indirect immunofluo-cluded in Tris buffered saline (TBS) freezing mediumrescence technique, as described in previous publications(Triangle Biomedical Sciences, Durham, NC, USA) and[20, 21] using the monoclonal antibodies listed later. Aswere immediately frozen in dry ice and acetone and weresecondary antibody we used fluorescein-conjugated af-stored at 2708C.finity-pure F(ab9)2 rat antimouse IgG with minimal cross

Light microscopy studies reactivity with rat proteins (Accurate Chemical & Scien-tific Corp., Westbury, CT, USA) at a concentration ofParaffin-embedded sections were stained with hema-60 mg/ml (1:50). Ethidium bromide was used for nucleartoxylin eosin, Masson’s Gomori’s stain, and periodicstaining.acid-Schiff reagent (PAS). Coronal sections cut at 4 mm

The following monoclonal antibodies were used: anti-thickness, including superficial and juxtamedullary glo-CD43 (Accurate Chemical & Scientific Corp., Westbury,meruli, were evaluated.NY, USA; mouse IgG1 clone W3/13, distribution in TEvaluation of histologic damage. Sections were exam-lymphocytes, concentration 5 5 mg/ml); anti-ED1 (Accu-ined by two investigators without previous knowledgerate Chemical & Scientific; mouse IgG1, distribution inof the experimental group of the animal from which themonocytes, concentration 5 30 mg/ml); anti-ICAM1tissue was taken. At least 10 sections were analyzed(CD54; Seikegaku Corp. Tokyo, Japan; IgG1, distribu-for each kidney, and the evaluation that was assignedtion in endothelial cells, monocytes, T and B lymphocytes,represented the mean values of these sections. Agree-epithelial cells, dendritic cells, mesangial cells; concentra-ment between the two observers was within 10%. Evalu-tion 5 10 mg/ml); anti-CD11b (Biosource International,ation was done as follows.Camarillo, CA, USA; mouse IgG2 A, distribution inGlomerular sclerosis. Glomerular lesions were evalu-monocytes/macrophages, neutrophils; concentration 5 5ated by determining the number of normal glomeruli and

the number with either segmental or total glomerular mg/ml) and anti-LFA1 (CD18; Seikegaku Corp.; IgG1,

Romero et al: Mycophenolate mofetil therapy in renal ablation948

distribution in lymphocytes, monocytes, neutrophils; con-centration 5 10 mg/ml).

Double-staining procedures were done as follows: Themonoclonal antibody (mAb) used to identify the inte-grins (mAb anti-CD18 or anti-CD11b) was applied tothe tissue (concentrations listed earlier here) and wasincubated in a humid chamber at room temperature for45 minutes. Afterward, tetramethylrhodamine isothiocy-anate (TRIC)-conjugated, affinity-pure F(ab9)2 rat anti-mouse IgG, with minimal cross-reaction with rat proteins(Accurate Chemical & Scientific) was added. After wash-ing for 10 minutes with PBS 0.01 m, 0.15 m NaCl (onechange of PBS) and 100 ml of blocking solution (PBS0.01 m, 0.15 m NaCl containing 10% fetal calf serum and

Fig. 1. Serum creatinine concentration in sham-operated animals (h)10% goat serum) were added, incubated for 10 minutes,and rats with reduction in renal mass, untreated (5/6Nx, j) and treatedand washed as described earlier here. Afterward, the with mycophenolate mofetil (5/6NxMMF, d). Data from five to eight

tissues were incubated with the mAb used to identify weeks correspond to five animals in each group; otherwise, the valuesrepresent 10 to 16 animals in each group (Methods section, experimentalthe leukocyte populations (mAb anti-CD43 or mAb anti-design). Abbreviations are: B, baseline; weeks, weeks after surgery.ED1 at the dilutions described earlier) and were then*P , 0.05, **P , 0.01, ***P , 0.001 vs. sham. Bars are standard

reacted with 50 ml of fluorescein-conjugated, affinity- deviation. The differences between the 5/6Nx group and the MMFpure F(ab9)2 rat antimouse IgG (20 mg/ml) with minimal treated group are also significant (P , 0.001) after the fourth week.cross-reaction with rat proteins (Accurate Chemical &Scientific). As a control for the double-staining procedure,five tissues were double stained with CD18 and ED1, as test or Kruskal Wallis tests with Dunn’s multiple compar-described earlier here. ison post-test were used. Differences were considered

Tissues were examined with a Zeiss microscope with significant when P was less than 0.05.epifluorescence system, phase contrast, and appropriatefilters. Cells with red nuclei surrounded by a green cyto-

RESULTSplasmatic membrane were counted as positive. Countingwas done with the help of a grid-fitted eyepiece. Within Figure 1 shows the serial changes in serum creatininethe glomerular tuft, staining results were expressed as concentration observed in the sham-operated and in thepositive cells per glomerular cross-section, with the ex- rats with 5/6 reduction in renal mass with (5/6NxMMF)ception of CD54 staining that was expressed as a 11 to and without (5/6Nx) treatment with MMF. The un-41 score, described previously [17], which takes into treated 5/6Nx group of rats experienced a progressiveaccount the number of positive cells and the intensity of rise in serum creatinine (P , 0.01 vs. baseline) from thethe fluorescent staining. In the interstitial areas, results second week up to the seventh to eighth week afterwere expressed as positive cells per mm2. nephrectomy when it appeared to reach a plateau, at

levels of 156 6 6.61 mmol/liter, approximately 2.4 timesGlomerular volumeabove control levels. The levels of serum creatinine the

At least 100 glomeruli in each kidney tissue were ex-5/6NxMMF group and in the sham-operated group wereamined, and the glomerular diameter was measured us-lower than in the untreated 5/6Nx rats (P , 0.001 in Fig.ing a microscopic scale attached to the ocular lens. Then,1). In contrast, the 5/6NxMMF group had essentiallythe mean random cross-sectional glomerular area (AG)similar serum creatinine levels than the sham-operatedwas calculated, and the glomerular volume (VG) wasgroup the first four weeks after surgery. Thereafter, thedetermined using the following formula:level of the creatinine in 5/6NxMMF group was slightly

VG 5 (b/k 3 AG)3/2above the levels in the sham-operated group but wassteady at each time interval (Fig. 1).where b 5 1.38 is the shape coefficient for spheres, the

Urinary protein excretion is shown in Figure 2. Theidealized shape of the glomerular tuft, and k 5 1.1 isanimals receiving MMF (5/6NxMMF) and the sham-the size distribution coefficient [22].operated group showed essentially similar levels of pro-

Statistical analysis teinuria throughout the eight-week observation period.Urinary protein excretion in these groups (5/6NxMMFResults are expressed as mean 6 sd. Comparisonsgroup and sham-operated group) is unchanged from thebetween groups were done by one-way analysis of vari-baseline levels. In contrast, the untreated animals withance followed by Tukey-Kramer post-tests. When differ-reduced renal mass (5/6Nx group) showed a progressiveences in the standard deviation of the groups under anal-

ysis were significant, nonparametric Mann–Whitney rank increment in proteinuria.

Romero et al: Mycophenolate mofetil therapy in renal ablation 949

Fig. 2. Urinary protein excretion in sham-operated animals (h) and Fig. 3. Urinary malondialdehyde (MDA) excretion in sham-operatedrats with reduction of renal mass untreated (5/6Nx, j) and treated with animals (h) and rats with reduction of renal mass untreated (5/6Nx, j)mycophenolate mofetil (5/6NxMMF, d). Data from five to eight weeks and treated with mycophenolate mofetil (5/6NxMMF, d). From five tocorrespond to five animals in each group; otherwise, the values represent eight weeks, all groups consisted of five animals each; otherwise, all10 to 16 animals in each group (Methods section, experimental design). values represent at least 10 animals in each group (Methods section,Bars are standard deviation. Abbreviations are: B, baseline; weeks, weeks experimental design). Bars are standard deviation. Abbreviations are:after surgery. *P , 0.01 and **P , 0.001 vs. sham and MMF group. B, baseline; weeks, weeks after surgery. *P , 0.001 vs. sham; **P , 0.01

vs. sham. The difference between the 5/6Nx group and the 5/6NxMMFgroup at three to four weeks and at five to six weeks is also significant(P , 0.001).

All animal groups had similar weights (g) at the begin-ning of the experiment (5/6Nx group 5 255 6 11.42,5/6NxMMF group 5 236.9 6 10.09, sham-operated of the glomeruli in sham-operated animals and increasedgroup 5 255 6 5.00) and gained weight during the period 13 times at four weeks and 17 times at eight weeks afterof observation. However, the 5/6NxMMF group and the 5/6Nx in the untreated group (5/6Nx group). Eight weekssham-operated groups gained weight similarly at four after renal ablation, segmental glomerulosclerosis in ratsweeks (5/6NxMMF group 5 320 6 6.70, sham-operated treated with MMF (5/6NxMMF group) was 25 6 10.5%,group 5 315 6 4.30) and at eight weeks (5/6Nx,MMF half of the value in untreated rats at the same period ofgroup 5 409 6 2.45, sham-operated group 5 411.3 6 time (P , 0.001).2.40); in contrast, the 5/6Nx group gained less weight at Tubular dilation and atrophy, as well as tubulointersti-four weeks and eight weeks (286.1 6 9.44 and 338 6 tial infiltration and fibrosis, were substantially reduced8.00, respectively, P , 0.01). in the 5/6NxMMF group. In contrast, the untreated ne-

Mean arterial blood pressure at the time of sacrifice phrectomized animals (5/6Nx group) had increasing se-(4 weeks after surgery) was 121.1 6 10.33 in the sham- verity of tubular and interstitial damage, involving 40%operated rats, 165.5 6 16.5 in the 5/6Nx group (P , to 80% of the examined remnant renal tissue after eight0.001 vs. sham), and 158.3 6 15.4 in the 5/6NxMMF weeks (Table 1).group (P , 0.01 vs. sham). Figure 4A shows typical sclerotic changes in glomeruli

Figure 3 shows the urinary excretion of MDA. Both and interstitium in untreated animals, in contrast withgroups of nephrectomized rats (5/6Nx group and animals treated with MMF (Fig. 4B).5/6NxMMF group) had their urinary excretion of MDA Cells expressing adhesion molecules (CD54, CD18,doubled after surgery; however, the group treated with and CD11b) and CD43- and ED1-positive cells in theMMF showed a progressive decrease in urinary MDA glomeruli and in tubulointerstitial areas are shown instarting three to four weeks after surgery, reaching con- Tables 2 and 3, respectively, as well as in Figure 4 C–F,trol levels during the fifth to sixth week. In contrast, the and Figure 5.untreated animals (5/6Nx group) increased the MDA In biopsies examined four and eight weeks after sur-excretion nine times, peaking at the fourth week and gery, the number of lymphocytes (CD43-positive cells)decreasing thereafter (Fig. 3). and macrophages (ED1-positive cells) in the glomeruli

The histologic observations are shown in Table 1. The (Table 2) and interstitium (Table 3) was not statisticallyglomerular volume is higher in the two groups of rats different in sham-operated rats and in rats treated withwith reduced renal mass. There were no significant differ- MMF (5/6NxMMF group). In contrast, the untreatedences in the glomerular hypertrophy in the 5/6Nx group 5/6Nx group showed values three to four times higherand the 5/6NxMMF group. Glomerular sclerosis, both (Table 2).

Four weeks after surgery, the untreated 5/6Nx groupsegmental and global, was present in a small percentage

Romero et al: Mycophenolate mofetil therapy in renal ablation950

Table 1. Histological observations

4 weeks after surgery 8 weeks after surgery

Sham 5/6 Nx 5/6 Nx 1 MMF Sham 5/6 Nx 5/6 Nx 1 MFF(N 5 5) (N 5 9) (N 5 7) (N 5 5) (N 5 5) (N 5 5)

VG lm3 3 10 6 1.6160.10 2.2360.43 2.27 60.34 1.8160.42 2.6160.21 2.5860.90Glomerular sclerosis

Segmental % 3.060.71 38.764.33 21.7 69.41a 2.860.83 48.468.35 25.0610.51a

Global % 0.060.00 7.5763.65 4.29 66.45 0.060.00 20.6615.27 6.20 67.95b

Tubular dilation & atrophy (0–5) 0.060.00 2.8960.78 0.71 60.76a 0.060.00 4.260.45 0.860.84a

Interstitial infiltration (0–5) 0.060.00 1.5660.73 1.0 60.89 0.060.00 3.861.48 1.262.16Interstitial fibrosis (0–5) 0.060.00 1.6760.50 0.42 60.54a 0.060.00 2.060.71 0.4 60.89b

Segmental and global glomerular sclerosis expressed as the percent of the glomeruli examined presenting the corresponding lesion. Tubulointerstitial damage wasgraded according to the percent of the biopsy affected by the corresponding finding (see Methods section). N corresponds to the number of renal biopsies in eachgroup; the value in each biopsy is the mean of 10-15 sections.

Statistical significances between 5/6NxMMF vs. 5/6Nx are: a P , 0.001, b P , 0.05. The percentage glomeruli showing global sclerosis in the 5/6Nx group is higherat 8 weeks than at 4 weeks (P , 0.001). Data in untreated 5/6Nx are different (P ranging from ,0.05 to ,0.001) from data in sham operated rats.

had increased expression of adhesion molecules CD45, 5/6Nx and the 5/6NxMMF groups were similarly elevatedCD18, and CD11b (Tables 2 and 3). Eight weeks after with respect to the sham-operated group. The doses ofsurgery, the expression of CD54 in the untreated 5/6Nx MMF used in this work are not expected to have a sig-group had decreased, whereas in contrast, CD11b had nificant antiplatelet effect or to decrease the numbers ofmore than doubled in tubulointerstitial areas (Table 3). circulating leukocytes, but this possibility was not ruledSham-operated animals and 5/6Nx rats treated with out in this study.MMF had essentially similar expression of CD54, CD18, Other recent studies addressing the final commonand CD11b at four weeks (Table 2). At eight weeks, the pathway by which hemodynamic alterations induce renalonly significant difference between the sham-operated scarring are those of Shimizu et al, who used Pirfenidone,and 5/6Nx group was a reduction in the interstitial infil- a drug that prevents and reverses collagen accumulationtration of cells expressing CD54 in the MMF-treated [9]. They found attenuation of the chronic renal failuregroup (Table 3). in the rats with 5/6Nx [9].

Suppression of lymphocyte activity and monocyte/macrophage infiltration is yet another strategy of poten-DISCUSSIONtial use for amelioration of the progressive renal damage

The 5/6Nx model in the rat has been widely used toresulting from renal ablation. Macrophages are knownevaluate the effect of treatment strategies on the progres-to participate in the inflammatory reaction, producing asive scarring of the kidney that is associated with exten-number of soluble mediators, including cytokines, prote-sive renal ablation. Because the initial effects of theases, and free oxygen radicals [27], and several studiesreduction of nephron mass involves hypertension andhave shown that macrophage infiltration and prolifera-hyperfiltration of the remnant nephron units, a vast ma-tion are prominent findings in the remnant kidney [11–jority of research efforts have been directed to determine13, 28]. Furthermore, Schiller et al have recently demon-if the correction of the altered hemodynamics improvesstrated that mechanical injury resulting from glomerularthe renal damage. Understandably, drugs that lower sys-capillary hypertension induces up-regulation of mono-temic and glomerular pressure are among the mostcyte chemoattractant protein-1 (MCP-1) [10]. Interest-prominently studied [23–26].ingly, a low-protein diet modulates the expression ofOther treatment strategies have been investigated lessMCP-1 [10], which suggests that protein restriction, infrequently. Pukerson et al showed that thrombotic mech-addition to improving glomerular hemodynamics in theanisms are important in the progression of the glomeru-remnant nephrons, could have beneficial effects resultinglar lesion and that heparin had a beneficial effect [7].from inhibition of macrophage infiltration.Suppression of platelet adhesion and platelet-derived

Mycophenolate mofetil is the morpholinoethytl esterfactors was a treatment modality chosen by Zoja et al,of mycophenolic acid, and it inhibits inosine 59-mono-who found that Ticlopidine, a drug with a wide spectrumphosphate dehydrogenase, which controls the de novoof antiplatelet effects, improved the progressive renalsynthesis of guanosine triphosphate [29]. This drug hasdamage of the remnant kidney [8]. Interestingly, thea relatively specific effect on lymphocytes because otherbeneficial effects of this treatment occur without modifi-cells have alternative purine synthesis pathways notcation of the systemic blood pressure [8], which alsoblocked by MMF. In the experimental setting of experi-appears to be the case in the improvement induced with

MMF reported here, as the blood pressure levels in the mental renal ablation, MMF is a drug well suited for the

Romero et al: Mycophenolate mofetil therapy in renal ablation 951

Fig. 4. Changes observed by light microscopy and lymphocyte and macrophage infiltration of remnant kidneys four weeks after 5/6Nx. Severesegmental sclerosis (approximately 60% of the glomerular tuft) observed in untreated rat (receiving vehicle; A) is contrasted with essentiallynormal glomerular appearance in rat treated with MMF (B). In addition, interstitial fibrosis is evident in untreated rat and absent in the biopsyfrom the rat treated with MMF (Mason’s Trichromic stain, original magnification &time;400). Immonuhistology showing remnant kidney tissuestained with mAb anti-CD43 (C, untreated rat; D, rat treated with MMF) and with mAb anti-ED1 (E, untreated rat; F, rat treated with MMF).Original magnification was 3630.

purpose of testing the effects of suppressing macrophage brane glycoproteins, which are ligands of the selectinadhesion molecules in the vascular endothelium [29]; ininfiltration and proliferation for several reasons. First,

MMF depletes the pools of guanosine triphosphate not addition, MMF inhibits up-regulation of CD18 in lym-phocytes [30]. Consequently, the infiltration of mono-only in lymphocytes but also in monocytes and thereby

decreases the synthesis of fucose and mannose of mem- cytes, as well as lymphocytes, is inhibited by this drug

Romero et al: Mycophenolate mofetil therapy in renal ablation952

Table 2. Cellular infiltration in the glomeruli

4 weeks 8 weeks

Positive cells Sham 5/6Nx 5/6NxMFF Sham 5/6Nx 5/6NxMFFper gcs (N 5 5) (N 5 9) (N 5 7) (N 5 5) (N 5 5) (N 5 5)

CD43 1.1460.27 3.7261.30 1.4660.94b 0.4660.31 3.4760.54 1.1560.56a

ED1 0.7460.11 3.1460.48 0.9760.27b 0.8660.27 2.2860.37 1.2660.44b

CD54 1.2760.13 2.2160.79 0.8660.18a 0.8560.12 0.5460.11 1.0860.08a

CD18 0.9860.27 4.4461.45 1.9461.02a 1.2160.14 1.3060.25 0.8060.24b

CD11b 1.0460.60 4.7562.18 1.3060.22a 1.0760.70 3.5060.94 0.8360.16a

N corresponds to the number of biopsies studied; the value in each biopsy represents the mean of 10-15 sections.Statistical significances between 5/6Nx MMF and 5/6 Nx are: a P , 0.001, b P , 0.01. In the 5/6Nx group, the differences between the 4th and 8th week in CD54

and CD18 are significant (P , 0.01). The data of untreated 5/6Nx are significantly different (P ranging from ,0.05 to ,0.001) from the data in sham operated rats.

Table 3. Cellular infiltration in the interstitium

4 weeks 8 weeks

Positive cells Sham 5/6Nx 5/6NxMFF Sham 5/6Nx 5/6NxMFFper mm2 (N 5 5) (N 5 9) (N 5 7) (N 5 5) (N 5 5) (N 5 5)

CD43 68.4656.5 135.6664.2 59.2639.3c 36.8617.3 431.2659.2 91.7659.9a

ED1 35.267.6 194.4661.2 34.069.6a 56.0611.6 140.0622.8 75.664.91a

CD54 252.0627.5 531.66216 314.46124c 261.6644.5 128.8615.3 73.6642.6c

CD18 25.4610.8 229.66132 72.6631.2b 41.6612.5 74.4619.9 75.0627.9CD11b 32.0628.4 121.2643.1 25.2611.9 21.6617.8 352.86105 36.1621.8a

N corresponds to the number of biopsies studied; the value in each biopsy represents the mean of 10-15 sections.Statistical differences between 5/6Nx MMF and 5/6Nx are: a P , 0.001, b P , 0.01, c P , 0.05. In the 5/6Nx group, the differences between the 4th and 8th week

in CD54 and CD18 are significant (P , 0.01). The data in the untreated 5/6Nx group are significantly different (p ranging from ,0.05 to ,0.001) from the data inthe sham operating group, except for CD18 positive cells at 8 weeks.

[30–32]. Second, MMF exerts a long-term inhibition of after four weeks in the 5/6NxMMF group, whereasin contrast, it increases in the untreated 5/6Nx groupthe production of lymphocyte and macrophage-derived

cytokines [33], a valuable characteristic in the prevention (Table 1). Tubulointerstitial changes are ameliorated im-portantly in the MMF-treated group (P . 0.05 vs. sham;of chronic scarring in the remnant kidney [34]. Third,

MMF, in contrast to other antirejection drugs, is not Table 1), in contrast with the changes seen in the 5/6Nxgroup (Fig. 5 A, B).associated with renal toxicity [15] and is reasonably well

tolerated as a long-term treatment [17, 18]. The observed beneficial effect on renal scarring islikely the result of direct and indirect actions of MMFTreatment with MMF completely prevented the pro-

teinuria induced by renal ablation. This protective effect because cytokine “cross-talk” between macrophages andintrinsic renal cells is central to the cell proliferation andis maintained for at least eight weeks (Fig. 2). To our

knowledge, there is no direct effect of MMF on glomeru- matrix expansion that leads to glomerulosclerosis andinterstitial fibrosis [36]. Careful studies by Floege et allar permselectivity. Reduction of proteinuria, noted in

anecdotal reports of several kidney diseases treated with and Kliem et al have shown that in the remnant kidneymodel, proliferation of intrinsic glomerular and tubuloin-this drug [35], is presumed to be a consequence of its

immunosuppressant properties, reduction of cellular in- terstitial cells is an early event in the development ofrenal scarring; macrophage infiltration appears later andfiltration, and decrease in cytokine production. Renal

function is also markedly improved by MMF treatment. contributes to amplify the inflammatory response [12,13]. To our knowledge, there is no evidence of a directAs reported by others, the 5/6Nx group presents a pro-

gressive rise in serum creatinine, whereas in contrast, suppressive effect of MMF on cytokine production byintrinsic renal cells. However, this drug reduces cell ad-the rats with 5/6 reduction in renal mass treated with

MMF showed only a modest increment in serum creati- hesion molecules in endothelial cells [16], and there is awell-defined inhibition of the production of all lympho-nine that remained at steady levels from the fourth to

the eighth week (Fig. 1). cyte and macrophage-derived cytokines and growth fac-tors. Specifically, platelet-derived growth factor, inter-The histologic findings are in accordance with the renal

function data. The number of glomeruli with segmental leukin 2, tumor necrosis factor a, and transforminggrowth factor-b are profoundly reduced [33]. Suppres-and global sclerosis is substantially reduced with MMF

treatment. Moreover, the percentage of glomeruli with sion of macrophage-derived platelet-derived growth fac-tor should prevent further activation of the mesangialthese changes (Fig. 5 A, B) remains essentially unchanged

Romero et al: Mycophenolate mofetil therapy in renal ablation 953

Fig. 5. Intercellular adhesion molecules in remnant kidneys four weeks after 5/6Nx. Intense interstitial infiltration of cells expressing CD18 inkidney from untreated rat (A) is contrasted with the negative appearance of the kidney from a rat treated with MMF (B). Intense glomerularexpression of CD54 in untreated 5/6Nx rat (C) contrasts with moderate expression of CD54 in 5/6Nx rats treated with MMF (D).

cells, in addition to limiting amplification of the prolifera- It is interesting that a similar increment in glomerularvolume was found in the 5/6Nx and the 5/6NxMMFtive reaction of infiltrating cells. Suppression of macro-

phage-derived transforming growth factor-b should have groups. This finding firstly suggests that the hemody-namic disturbances that trigger glomerular hypertrophya beneficial effect on glomerulosclerosis by limiting the

synthesis and favoring the degradation of extracellular are similar in both groups of rats with renal ablation, andsecondly suggests that renal hypertrophy is unaffected bymatrix by renal metalloproteinases [36, 37].

An additional influence that may induce synthesis of MMF treatment. This is in contrast with the reductionin the renal hypertrophy and hyperplasia associatedchemokynes is the generation of reactive oxygen species.

Nath, Croatt and Hostetter have shown that oxidant with other immunosuppressive drugs such as cyclospor-ine A [39].stress is increased in the 5/6Nx model [4], and our studies,

in addition to confirming their findings, have shown that Lymphocyte (CD43-positive cells) and macrophage(ED1-positive cells) infiltration is comparable in theurinary MDA is significantly reduced by MMF treatment

from the first to the sixth week after nephrectomy (Fig. MMF-treated group and the sham-operated group atfour weeks after surgery. In contrast, the 5/6Nx group3). This effect of MMF is likely the result of suppression

of the inflammatory infiltrate, but by itself, a reduction shows the cellular infiltration of the remnant kidney re-ported previously by others [10–13, 28]. These findingsin reactive oxygen species would reduce the expression

of MCP-1, RANTES, and ICAM-1 in the mesangial cells indicate that MMF administration was capable of effec-tively counteracting the infiltration and proliferative[reviewed in 34]. In addition, a decrease in oxidative

stress would eliminate a stimulus for apoptosis, which is stimuli triggered by the glomerular hypertension.Expression of adhesion molecules has not been stud-an early event in the remnant kidney model [38].

Romero et al: Mycophenolate mofetil therapy in renal ablation954

ied in the renal ablation model. All adhesion molecules REFERENCEStested increased several times in the remnant kidney 1. Shimamura T, Morrison AB: A progressive glomerulosclerosis

occurring in partial five-sixths nephrectomy. Am J Pathol 79:95–(glomeruli and interstitium) in the 5/6Nx rats examined101, 1975four weeks after nephrectomy. In the following month,

2. Hostetter TH, Olson JL, Venkatachalam MA, Brenner BM:the expression of CD54 and CD18 in glomeruli and inter- Hyperfiltration of remnant nephrons: A potentially adverse re-

sponse to renal ablation. Am J Physiol 241:F85–F93, 1981stitium fell to levels comparable to those present in sham-3. Zatz R: Haemodynamically mediated glomerular injury: The endoperated rats (Tables 2 and 3). The reasons for the reduc-

of a 15-year-old controversy? Curr Opin Nephrol Hypertens 5:468–tion in expression of CD54 and CD18 are not clear, 475, 1996

4. Nath K, Croatt AJ, Hostetter TH: Oxygen consumption andbut it is likely that progressive scarring of the kidney isoxidant stress in surviving nephrons. Am J Physiol 258(Renal Fluidassociated with a decrease in the number of CD54-posi-Electrolyte Physiol 27):F1354–F1362, 1990

tive cells, which includes endothelial cells, lymphocytes, 5. Olson JL, Hostetter TH, Rennke HG, Brenner BM, Venkata-chalam MA: Altered glomerular permselectivity and progressiveand neutrophils. As shown in Table 3, the rats from thesclerosis following extreme ablation of renal mass. Kidney Int5/6Nx group showed a minor reduction (P . 0.05) in22:112–126, 1982

tubulointerstitial ED1-positive cells at four and eight 6. Brenner BM, Lawler EV, Mackenzie HS: The hyperfiltrationtheory: A paradigm shift in nephrology. Kidney Int 49:1774–1777,weeks after nephrectomy (194 6 61.2 and 140 6 22.81996cells per mm2, respectively), whereas the CD11b-positive 7. Pukerson M, Hoffsten PE, Klahr S: Pathogenesis of the glomeru-

cells almost triple in number (Table 3). Because CD11b lopathy associated with renal infarction in rats. Kidney Int 9:407–417, 1976stains granulocytes in addition to monocytes and macro-

8. Zoja C, Perico N, Bergamelli A, Pasini M, Morigi M, Dadanphages, we interpret these findings as indicative of an J, Belloni A, Bertani T, Remuzzi G: Ticlopidine prevents renalincrement in nonmacrophage leukocyte infiltration after disease progression in rats with reduced renal mass. Kidney Int

37:934–942, 1990eight weeks. Interestingly, the numbers of CD11b-posi-9. Shimizu T, Fukagawa M, Kuroda T, Hata S, Iwasaki Y, Nemototive cells in rats treated with MMF are comparable to M, Shirai K, Yamauchi S, Margolin S, Shmizu F, Kurokawa K:

the sham-operated animals (Tables 2 and 3). Pirfenidone prevents collagen accumulation in the remnant kidneyin rats with partial nephrectomy. Kidney Int 52:S239–S243, 1997The relatively stable number of macrophages from the

10. Schiller B, Moran J: Focal glomerulosclerosis in the remnantfourth to the eight weeks in the remnant kidney observed kidney model: An inflammatory disease mediated by cytokines.by us is in contrast with the findings in a recent report Nephrol Dial Transplant 12:430–437, 1997

11. Wu LL, Yang N, Roe CJ, Cooper ME, Gilbert CR, Atkins RC,by Yang et al [28], who found that ED1-positive cellsLan HY: Macrophage and myofibroblast proliferation in remnantincreased progressively up to 16 weeks after nephrec- kidney: Role angiotensin II. Kidney Int 52:S221–S225, 1997

12. Floege J, Burns MW, Alpers CE, Yoshimura A, Pritzl P, Gor-tomy. Their study did not use a high-protein diet [28],don K, Seifert RA, Bowen-Pope DF, Couser WG, Johnson RJ:and consequently, the tubulointerstitial damage and theGlomerular cell proliferation and PDGF expression precede glo-

increment in serum creatinine progressed at a lower rate merulosclerosis in the remnant kidney model. Kidney Int 41:297–309, 1992than in this study. It is possible that this explains, at least

13. Kliem V, Johnson RJ, Alpers CE, Yoshimura A, Couser WG,in part, the discrepancy.Koch KM, Floege J: Mechanisms involved in the pathogenesis of

These studies confirm recent preliminary reports from tubulointerstitial fibrosis in 5/6-nephrectomized rats. Kidney Int49:666–678, 1996our group [40] and Fujihara et al [41], and suggest that

14. Allison AC, Eugui EM: Mycophenolate mofetil a rationally de-MMF may be useful in the treatment of progressive renalsigned immunosuppressive drug. Clin Transplant 7:96–112, 1993

failure associated with reduced renal mass. Yet to be 15. Sollinger HW: Mycophenolate mofetil. Kidney Int 48(Suppl52):S14–S17, 1995determined is whether MMF treatment would give com-

16. Hauser IA, Johnson DR, Thevenod F, Goppelt-Strube M: Effectparable or additional beneficial effects to those obtainedof mycophenolic acid on TNF alpha-induced expression of cell

by other treatment modalites in this experimental model. adhesion molecules in human venous endothelial cells in vitro. BrJ Pharmacol 122:1315–1322, 1997The route of administration, as well as the dose of MMF

17. European Mycophenolate Mofetil Cooperative Study Group:in this study, is comparable to those used routinely inPlacebo-controlled study of mycophenolate mofetil combined with

humans with relatively minor side-effects. Therefore, cyclosporin and corticosteroids for prevention of acute rejection.Lancet 345:1321–1325, 1995consideration may be given to trials with this drug in the

18. Schiff MH, Gloldblum R, Rees MMC: 2-Morpholinoethyl myco-clinical setting of progressive renal failure. phenolic acid (ME-MPA) in the treatment of refractory rheuma-toid arthritis (RA). (abstract) Arthritis Rheum 33:s155, 1990

19. Ohkawa H, Ohishi N, Yagi K: Assay for lipid peroxides in animalACKNOWLEDGMENTStissues by thiobarbituric acid reaction. Anal Biochem 95:351–358,

The authors are grateful for financial support from Asociacion de 1979Amigos del Rinon, Maracaibo, Venezuela. Parts of this article were 20. Parra G, Romero M, Henriquez-La Roche C, Pineda R, Rodri-presented in the 30th Annual Meeting of the American Society of Ne- guez-Iturbe B: Expression of adhesion molecules in poststrepto-phrology (San Antonio, Texas, November 2–5, 1997) and were published coccal glomerulonephritis. Nephrol Dial Transplant 9:1412–1417,in abstract form (Romero et al, J Am Soc Nephrol 8, 628A, 1997). 1994

21. Parra G, Moreno P, Rodriguez-Iturbe B: Glomerular prolifera-Reprint requests to Bernardo Rodrıguez-Iturbe, Apartado Postal tive activity and T lymphocyte infiltration in acute serum sickness.

1430, Maracaibo 4001-A, Venezuela. Clin Immunol Immunopathol 82:299–302, 199722. Weibel ER: Stereological Methods: Practical Methods for Biologi-E-mail: bri@iamnet.com

Romero et al: Mycophenolate mofetil therapy in renal ablation 955

cal Morphometry. London, Academic Press, 1979, pp 40–62, 101– 31. Heeman U, Azuma H, Schmid C, Philipp T, Tilney N: Effect ofmycophenolic acid mofetil on acute rejection of kidney allograft116in rats. Clin Nephrol 5:355–359, 199623. Anderson S, Rennke HG, Brenner BM: Therapeutic advantage

32. Laurent AF, Dumont S, Poindron P, Muller CD: Mycophenolicof converting enzyme inhibitors in arresting progressive renal dis-acid suppresses protein N-linked glycosilation in human monocytesease associated with systemic hypertension in the rat. J Clin Investand their adhesion to endothelial cells and to some substrates.77:1993–2000, 1986(abstract) Exp Hematol 24:59–67, 199624. Remuzzi A, Imberti G, Puntorieri S, Malanchini B, Macconi

33. Nadeau KC, Azuma H, Tilney NL: Sequential cytokine expres-D, Magrini L, Bertani T, Remuzzi G: Dissociation between anti-sion in renal allografts in rats immunosuppressed with maintenanceproteinuric and anti-hypertensive effects of angiotensin-convertingcyclosporine or mycophenolate mofetil. Transplantation 62:1362–enzyme inhibitors in rats. Am J Physiol 267:F1034–F1044, 19941366, 196625. Lewis EJ, Hunsicker LG, Bain RP, Rhode RD: The effect of

34. Schena FP, Gesualdo L, Grandaliano G, Montinaro V: Pro-angiotensin-converting enzyme inhibition on diabetic nephropa-gression of renal damage in human glomerulonephritides: Is therethy. N Engl J Med 329:1456–1462, 1993sleight of hand in winning the game? Kidney Int 52:1439–1457,26. Maschio G, Alberti D, Locatelli F, Mann JF, Motolese M,1997Ponticelli C, Ritz E, Zucchelli P: The angiotensin-converting

35. Briggs WA, Choi MJ, Scheel PJ Jr: Successful mycophenolateenzyme inhibition in the progressive renal insufficiency studymofetil treatment of glomerular disease. Am J Kidney Dis 31:213–group: Effect of angiotensin-converting enzyme inhibitor benaze-217, 1998pril on the progression of chronic renal insufficiency. N Engl J

36. Schlondorff D: The role of chemokines in the initiation andMed 334:939–945, 1996progression of renal disease. Kidney Int 47(Suppl):S44–S47, 199527. Cattell V: Macrophages in acute glomerular inflammation. (edito-

37. Roberts WA, Noble NA: Transforming growth factor-b in tissuerial review) Kidney Int 45:945–952, 1994fibrosis. N Engl J Med 331:1286–1292, 199428. Yang N, Wu LL, Nikolic-Paterson J, Ng Y-Y, Yang W-C, Mu 38. Sugiyama H, Kashihara N, Makino H, Yamasaki I, Ota Z: Apo-

W, Gilbert RE, Cooper ME, Atkins RC, Lan HY: Local macro- ptosis in glomerular sclerosis. Kidney Int 49:103–111, 1996phage and myofibroblast proliferation in progressive renal injury 39. Batlle DC, Gutterman C, Keilani T, Peces R, Lapointe M:in the rat remnant kidney. Nephrol Dial Transplant 13:1967–1974, Effect of cyclosporine A on renal function and kidney growth in1998 the uninephrectomized rat. Kidney Int 37:21–28, 1990

29. Sintchak MD, Fleming MA, Futer O, Raybuck SA, Chambers 40. Romero F, Parra G, Rodriguez-Iturbe B, Gonzalez L: Myco-SP, Caron PR, Murcko MA, Wilson KP: Structure and mecha- phenolate mofetil (MMF) prevents the renal damage induced bynism of inosine monophosphate dehydrogenase in complex with 5/6 nephrectomy in the rat. (abstract) J Am Soc Nephrol 8:628A,the immunosuppressant mycophenolic acid. Cell 85:921–930, 1996 1997

30. Allison AC, Eugui EM: Purine metabolism and immunosuppres- 41. Fujihara CK, Malheiros DMAC, Oliveira SG, Antunes GR,sive effects mycophenolate mofetil (MMF). Clin Transplant 10:77– Zatz R, Noronha I: Mycophenolate mofetil attenuates renal injury

in the remnant kidney. (abstract) J Am Soc Nephrol 8:615A, 199784, 1996