Afferent arteriolopathy is the most characteristic lesion of chronic cyclosporine (CsA) nephrotoxicity. We investigated the effect of therapeutic doses of mycophenolate mofetil (MMF) in a model of chronic CsA nephrotoxicity where transforming growth factor-β (TGF-β) was shown to play a central role.
Rats treated with vehicle, MMF 10 mg/kg/day, CsA 10 mg/kg/day or CsA + MMF were sacrificed at 7 or 28 days. Physiologic and histologic changes were studied in addition to TGF-β1 mRNA and protein expressions, and mRNA expression of plasminogen activator inhibitor-1 (PAI-1) and the extracellular matrix (ECM) proteins biglycan and types I and IV collagen.
While MMF markedly ameliorated afferent arteriolopathy, it had no significant effect on interstitial fibrosis and tubular atrophy. In addition, MMF treatment reduced both TGF-β1 mRNA and protein levels by 39% and 32%, respectively (p < 0.05 vs. CsA only). The expression of the ECM proteins followed that of TGF-β1 and was significantly decreased with MMF; a similar effect was observed with PAI-1, suggesting an increase in ECM degradation.
These results suggest that MMF exerts a beneficial effect on CsA arteriolopathy and that it decreases TGF-β1. While this drug combination may be useful clinically, long-term studies are needed to determine if MMF has a lasting benefit.
Cyclosporine (CsA) remains an important immunosuppressive drug used in transplantation (1). However, the major dose-limiting adverse effect of long-term CsA administration is chronic nephrotoxicity, a histologic lesion characterized by striped interstitial fibrosis, tubular atrophy and an arteriolar lesion (2). Renal arteriolopathy is highly specific for CsA-induced injury and is characterized by eosinophilic granular transformation of vascular smooth muscle cells of afferent glomerular arterioles that may eventually progress to muscle cell necrosis and hyalinization of the vessel wall (1–4). This arteriolopathy may be reversed in a manner that can be dissociated from progressive tubulointerstitial fibrosis (3,4).
Our previous studies using an experimental model of chronic CsA nephrotoxicity have shown that the fibrogenic cytokine transforming growth factor-β1 (TGF-β1) is involved in the fibrosis of this model by increasing extracellular matrix (ECM) synthesis and by decreasing ECM degradation through increasing plasminogen activator inhibitor-1 (PAI-1) (5). Apoptosis plays also a role and is directly correlated with fibrosis where cell loss may prevent the kidney's ability to remodel effectively (6). In this model, we have shown that the renin-angiotensin system (RAS) was up-regulated and that blocking its activity reduced fibrosis through a mechanism that involves modulation of TGF-β1 expression (7). In addition, l-arginine protected CsA-treated animals from impaired renal function and from the development of interstitial fibrosis, at least partly by decreasing TGF-β1 and ECM proteins (8). All this evidence point to a central role for TGF-β in development of fibrosis in chronic CsA nephrotoxicity (9). Thus, attempts at antagonizing TGF-β, for example with pirfenidone, were beneficial in ameliorating CsA-induced renal lesions (10).
Mycophenolic acid, the active metabolite of mycophenolate mofetil (MMF), is a specific inhibitor of inosine monophosphate dehydrogenase involved in de novo purine synthesis and a suppressor of T- and B-lymphocyte proliferation (11). Thus, MMF has been successfully used as an effective immunosuppressant in transplantation. In addition to its effect on lymphocytes, MMF also inhibits collagen deposition and the proliferation of fibroblasts, vascular smooth muscle and mesangial cells (12–16). Furthermore, MMF inhibits the production of lymphocyte and macrophage-derived cytokines and growth factors (17–20). Mycophenolate mofetil has been shown to have a protective effect in immune glomerulonephritis and in ischemia-reperfusion injury (21–26). In the remnant kidney model, MMF reduces interstitial myofibroblast infiltration and preserves renal function (13,27–29). In most studies, MMF reduced the infiltration of macrophages and lymphocytes and the expression of adhesion molecules. Thus, MMF may be able to act on inflammatory cell activation and on myofibroblast differentiation and proliferation, which are early events in the development of fibrosis.
More recent studies have shown that the late addition of MMF does not prevent the development of chronic CsA nephrotoxicity in the rat (19). On the other hand, when relatively large doses of MMF (40 mg/kg/day) where used either alone or in combination with losartan, a beneficial effect was observed (20). In this study, we used clinically relevant doses of MMF (10 mg/kg/day) in an experimental model of chronic CsA nephrotoxicity. We tried to dissociate the effect of MMF on the specific lesion of arteriolopathy from the other CsA-induced renal lesions. As TGF-β plays a central role in this model, we also studied the effect of MMF administration on the expression of TGF-β1 and a number of ECM proteins. Our findings suggest that MMF has a beneficial effect on CsA-induced afferent arteriolopathy and is associated with a decrease in the expression of TGF-β1, PAI-1 and ECM proteins.
Materials and Methods
Adult male Sprague-Dawley rats (Charles River, Wilmington, MA) weighing 325–350 g were housed in a temperature- and light-controlled environment. They received a low-salt diet (0.05% sodium, Teklad Premier, WI) and were allowed free access to tap water. Animals were pair-fed and weighed daily. After 1 week on a low-salt diet, weight-matched animals were assigned to one of four groups of 12 animals. Six animals in each group were sacrificed at 7 days (Groups A to D) and 6 were sacrificed at 28 days (Groups E to H). The experimental groups (n = 6/group) were: Groups A and E, vehicle (VH) olive-oil control; Groups B and F, MMF; Groups C and G, CsA; and Group D and H, CsA + MMF. Systolic blood pressure was measured by tail plethysmography (Natsume Seisakusho Co. Ltd, Tokyo, Japan) and 24-h urine samples were collected in metabolic cages (Nalge Company, Rochester, NY). The following day, rats were anesthetized with intraperitoneal ketamine, the abdomen was opened through a midline incision and the aorta was cannulated retrogradely below the renal arteries with an 18-gauge needle. With the aorta occluded by ligation above the renal arteries and the renal veins opened by a small incision for outflow, the kidneys were perfused with 20 mL of cold heparinized saline. The left kidney was removed and processed for light microscopy. After removing the right kidney, the cortex was dissected from the medulla, and the cortex was processed for RNA and protein analysis. Following the experiment, the animals were euthanized by deep anesthesia with ketamine followed by exsanguination.
Cyclosporine (Novartis Research Institute, East Hanover, NJ) was diluted in olive oil and administered subcutaneously (sc) at a dose of 10 mg/kg/day. The vehicle group received olive oil at 1 mL/kg/day sc. Mycophenolate mofetil (Hoffman-La Roche Laboratories, Base, Switzerland) was suspended in sterile water to a final concentration of 10 mg/mL and was administered by gavage once daily at a dose of 10 mg/kg.
Blood was collected from the jugular vein in plastic syringes transferred to metal-free tubes and chilled on ice. Plasma was harvested immediately by centrifugation at 4 °C and stored at – 70 °C until determined. Creatinine and blood urea nitrogen levels were measured by a Cobas autoanalyzer (Roche Diagnostics, Division Hoffman-La Roche Inc., Nutley, NJ). Cyclosporine blood level was measured by a monoclonal radioimmunoassay (Incstar Co., Stillwater, MN).
Tissue was fixed in 10% buffered formalin and in Methyl Carnoy's solution, and then embedded in paraffin. Sections 2–4μ thick were stained with periodic acid-Schiff's reagent and trichrome stain and evaluated by light microscopy by a blinded pathologist for tubular injury, interstitial fibrosis and afferent arteriolopathy. Arteriolopathy of the afferent arteriole was characterized by expansion of the cell cytoplasm of terminal arteriolar smooth muscle cells by eosinophilic, granular material. Findings ascribed to tubular injury included cellular and intercellular vacuolization, tubular collapse and tubular distention. Interstitial fibrosis consisted of matrix expansion with tubular distortion and collapse and basement membrane thickening.
A color-image analyzer (Nikon E400, Nikon Inc., Tokyo, Japan, Pixera Professional digital camera, Macintosh Powebook G3, NIH Image vs. 1.5) was used for semiquantitative scoring. Afferent arteriolopathy was estimated by counting the percentage of juxtaglomerular afferent arterioles with arteriolopathy per total afferent arterioles available for examination (magnification, ×200), with a minimum of 100 glomeruli per biopsy assessed. The following 0–3+ score was used: 0 = no arterioles injured; 0.5 =≤15% of arterioles injured; 1 = 15–30%; 1.5 = 31–45%; 2 = 46–60%; 2.5 = 61–75%; 3 =≥75%. Interstitial fibrosis was estimated by counting the percentage of injured areas per fields of cortex and medulla with a minimum of 30 fields (magnification, ×200) reviewed using the following score: 0 = normal interstitium; 0.5 =≤5% of areas injured; 1 = 5–20%; 1.5 = 21–35%; 2 = 36–50%; 2.5 = 51–65%; 3 =≥65%. The extent of tubular injury in cortical tubules was graded using the following criteria: 0 = no tubular injury; 0.5 = <5% of tubules injured; 1 = 5–20%; 1.5 = 21–35%; 2 = 36–50%; 2.5 = 51–65% and 3 =≥65%.
Northern blot analysis
Renal tissue was finely minced with a razor blade on ice and then homogenized in TRIzol reagent (GibcoBrl, Grand Island, NY). RNA extraction was performed according to the manufacturer's protocol. After resuspension in Tris-EDTA buffer, RNA concentrations were determined using spectrophotometric readings at Absorbance260. Thirty micrograms of RNA were electrophoresed in each lane in 0.9% agarose gels containing 2.2 m formaldehyde and 0.2 m Mops (pH 7.0) and transferred to a nylon membrane (ICN Biomedicals, Costa Mesa, CA) overnight by capillary blotting. Nucleic acids were crosslinked by ultraviolet irradiation (Stratagene, La Jolla, CA). The membranes were prehybridized for 2 h at 42 °C with 50% formamide, 10% Denhardt's solution, 0.1% SDS, 5 × standard saline citrate (SSC), and 200 μg/mL of denatured salmon sperm DNA. They were then hybridized at 42 °C for 18 h with cDNA probes labeled with 32P-dCTP by random oligonucleotide priming (Boehringer Mannheim). The blots were washed in 2 ×€SSC, 0.1% SDS at room temperature for 15 min and in 0.1 ×€SSC, 0.1% SDS at 50 °C for 15 min. Films were exposed at −70 °C for different time periods to ensure linearity of densitometric values and exposure time. Autoradiographs were scanned on an imaging densitometer (GS-700, Bio-Rad Laboratories, Hercules, CA). The density of bands for the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA was used to control for differences in the total amount of RNA loaded onto each gel line. For quantitative purposes, the values were divided by the density of bands for GAPDH in the same lane. The cDNA probes used for Northern blotting were: a mouse TGF-β1 cDNA probe (plasmid MUI5); a rat PAI-1 cDNA probe [plasmid pBluescript SK (–)]; a human biglycan cDNA probe (plasmid P16); a rat procollagen α1 cDNA (plasmid pα1R1); a rat collagen IV cDNA probe (plasmid pCIV-1-PE16) (American Type Culture Collection, Rockville, MD); and a rat GAPDH cDNA probe (plasmid pBluescript KS II).
Frozen (–70 °C) kidney tissue sections embedded in Tissue-Tek O.C.T. compound (Miles Inc. Diagnostics Division, Elkhart, IN) were processed for protein extraction. Tissue was thawed at 4 °C in 5 mL of lysis buffer (GITC/BME) then rinsed three times in 5 mL of cold 4 °C PBS pH 7.4. Protein extracts were prepared by homogenizing kidney tissue in 1 mL of cold PBS containing 0.05% Tween 20 (PBS-T) with a glass tissue homogenizer (Kontes Glass, Vineland, NJ). The homogenate was centrifuged at 4 °C for 15 min at 15 000 g and the supernatant was collected and then centrifuged to remove cellular debris. Protein concentration in the supernatant was determined using the Micro BCA (bicinchoninic acid) assay (Pierce, Rockford, IL). Latent TGF-β1 was activated to immunoreactive TGF-β1 and was detected by a commercially available TGF-β1-specific sandwich ELISA (Quantikine, R & D Systems, Minneapolis, MN). The manufacturer's recommendations were followed for sample activation, reagent preparation, assay procedure and calculation of results.
Results were expressed as mean ± standard error. Comparisons between groups were performed using analysis of variance (Tukey-Kramer analysis). A p-value less than 0.05 was considered statistically significant.
Values for weight gain, systolic blood pressure, serum creatinine level, blood urea nitrogen level and CsA whole blood trough level are summarized in Table 1. Weight gain was progressive in all treatment groups. There were no significant differences in body weight or in the rate of weight gain, suggesting that total food intake was comparable between the groups for the entire study period. Systolic blood pressure remained similar in all the experimental groups. Similar to VH, treatment with MMF did not affect kidney function. As expected, CsA treatment significantly increased serum creatinine at 28 days (p < 0.05) and blood urea nitrogen at 7 and 28 days (p < 0.05) compared with the VH groups. While concomitant therapy with MMF in the CsA rats improved serum creatinine and blood urea nitrogen at 28 days compared with the CsA-treated rats, the changes did not reach statistical significance except for serum creatinine at 28 days. In addition, the values remained significantly higher than in the VH-or MMF-treated rats.
Table 1. Physiologic changes in the experimental groups
Scores for the histologic changes observed are summarized in Table 2. Both the VH- and MMF-treated rats had normal kidney histology (Figures 1A,B and 2A,B). By contrast, kidneys of the CsA-treated rats developed tubular atrophy and striped interstitial fibrosis at 28 days (Figure 1C). Cyclosporine-treated rats also showed prominent renal vascular lesions characterized by hypertrophied smooth muscle cells of afferent arterioles with characteristic granular eosinophilic transformation (Figure 2C). Although treatment with MMF improved CsA-induced interstitial fibrosis and tubular atrophy, these changes did not reach statistically significance. On the other hand, the specific lesion of afferent arteriolopathy was significantly improved with MMF treatment (Figure 2D). As shown in Table 2, the score for afferent arteriolopathy decreased by 47% in the CsA + MMF group at 28 days (p < 0.05) compared with the CsA group.
Transforming growth factor-β1 mRNA was significantly increased in the CsA groups (p < 0.05) when compared with the VH-treated groups (Figure 3). Mycophenolate mofetil-treated rats had a TGF-β1 mRNA expression that was similar to the VH-treated rats. On the other hand, the TGF-β1 mRNA expression in the CsA + MMF groups was consistently and significantly lower at both 7 and 28 days (p < 0.05 vs. CsA group). Similar results were observed for the TGF-β1 protein expression by ELISA technique (Figure 4); as expected, the TGF-β1 protein expression lagged behind that of TGF-β1 mRNA. While CsA significantly increased TGF-β1 expression (p < 0.05) when compared with the VH or MMF groups, MMF treatment of the CsA rats resulted in a reduction in TGF-β1 protein expression (p < 0.05 vs. CsA only group). Overall, by 28 days, treatment of the CsA rats with MMF resulted in a reduction in TGF-β1 mRNA expression by 39% and in TGF-β1 protein expression by 32%.
Expression of PAI-1
The expression of PAI-1, a plasmin protease inhibitor that blocks ECM degradation and that is directly stimulated by TGF-β, is shown in Figure 5. The expression of PAI-1 mRNA paralleled that of TGF-β1 in this study with a significant increase in PAI-1 expression with CsA treatment (p < 0.05 vs. VH groups). Mycophenolate mofetil-treated rats had a PAI-1 mRNA expression similar to that of the VH-treated rats. Concomitant treatment of the CsA rats with MMF in the CsA + MMF group was associated with a significant decrease in PAI-1 mRNA expression by approximately 40–50% at both time points (p < 0.05 vs. VH or MMF groups).
Expression of ECM proteins
The expression of biglycan and of types I and IV collagen by Northern blotting is shown in Figure 6. These ECM proteins are directly stimulated by TGF-β. Their expression was similar to that of TGF-β1 and progressively increased in the CsA groups compared with the VH or MMF groups (p < 0.05), suggesting active ECM synthesis. On the other hand, MMF treatment in the CsA rats significantly lowered the expression of these ECM proteins as early as 7 days. By 28 days, MMF treatment in the CsA rats had significantly decreased the expression of biglycan, type I collagen and type IV collagen by 45%, 54% and 38%, respectively, compared with the CsA only treated rats and down to levels not statistically different from those of the VH- or MMF-treated rats.
Cyclosporine-associated arteriolopathy was described in up to 30% of kidney biopsies in the first year after kidney transplantation and is also present in the kidney when CsA is used for other organ allografts and autoimmune diseases (1–4,30). Afferent arteriolopathy is considered the hallmark of chronic CsA nephrotoxicity and is usually associated with the other pathological features of nephrotoxicity such as striped tubulointerstitial fibrosis and tubular atrophy. In the present study, we used clinically relevant doses of MMF and were able to show that MMF treatment significantly improved CsA arteriolopathy. However, the other chronic nephrotoxicity lesions of tubulointerstitial fibrosis and tubular atrophy, although improved, were not significantly ameliorated by MMF therapy at the current dose. This is slightly different from a recent report where relatively large doses of MMF (four times the dose used in this study), alone or in combination with losartan, had a beneficial effect on both tubulointerstitial fibrosis and arteriolopathy (20). We believe that these results are complementary and suggest that arteriolopathy is the CsA lesion, which is most sensitive to MMF therapy. Because arteriolopathy is the most specific histological lesion of chronic CsA nephrotoxicity, this early beneficial effect also suggests a specific effect of MMF on CsA nephrotoxicity.
The regression of CsA arteriolopathy was associated with an improvement in renal function in this study. Previous reports in which CsA was withdrawn or in which the dose was reduced have also shown an improvement in renal function without altering the progression of the tubulointerstitial disease (31). The dissociation in the arteriolar response to MMF therapy compared with the tubulointerstitial lesions confirms previous reports where drug discontinuation resulted in improved CsA arteriolopathy independent of tubulointerstitial fibrosis (4,32). These findings suggest that somewhat separate mechanisms are responsible for these two histological lesions. Our findings are also similar to a recently published report in the remnant kidney model where MMF therapy was effective in ameliorating afferent arteriolopathy (33). In this model of progressive renal injury, MMF decreased afferent arteriolar resistance despite persistent arterial hypertension and was accompanied by an improvement in the glomerular hemodynamics. Mycophenolate mofetil has well-documented antiproliferative effects and was shown to inhibit vascular smooth muscle proliferation and collagen deposition (11,13,14,16). This effect may, in turn, result in maintaining the functional capacity of preglomerular vessels. Mycophenolate mofetil can also inhibit cytokine-induced nitric oxide (NO) biosynthesis in vitro (34). In the chronic NO synthase inhibition model, MMF treatment at doses of 10 mg/kg limited the extent of renal injury and partially decreased proteinuria (35). In addition, pretreatment with MMF inhibited the induction of the NO synthase gene during ischemia-reperfusion injury (36).
The net impact of MMF therapy is likely to be a function of the timing of its introduction in the course of the nephropathy. It appears that, in order for MMF to exert an effective renal protection, treatment must be instituted early, and starting it after the disease is established substantially decreases its efficacy. In our study, MMF was coadministered with CsA right from the onset in order to approximate the clinical encountered situation in organ transplantation where MMF is frequently used as maintenance therapy with CsA. In a prior report, late MMF administration in the course of the nephropathy (after 5 weeks of CsA therapy) did not reverse CsA-associated renal injury, thus underscoring the importance of early therapy (19). In addition, two separate reports in the remnant kidney model confirmed the fact that initiation of MMF therapy immediately after mass removal is more effective than when started later when renal injury is established (29,37). Similar results were also observed in the Heyman nephritis model (23). The reason for this discrepancy is unclear. However, we know that lymphocyte and monocyte proliferation and increased expression of adhesion molecules are early events in both immune- and nonimmune-mediated nephropathies. Those early events are influenced by a number of cytokines including, but not limited to, Ang II. In chronic CsA nephrotoxicity, an early macrophage infiltration has been clearly observed and precedes the development of interstitial fibrosis (38). Along those lines, MMF was shown to have a direct effect on those early events. For example, MMF inhibits the proliferation of monocytes and fibroblasts and induces the apoptosis of monocytes (13,39,40). Moreover, MPA inhibits the glycosylation and, consequently, the affinity of adhesion molecules expressed by a variety of cells (41,42). The net result is a decrease by MMF in early lymphocyte and macrophage infiltration. If initiated at later stages, when other cellular events acquire greater pathogenic importance, MMF can still exert a protective effect but requires the association of other agents with complementary mechanisms of action, such as drugs that block the RAS (20,29,37,43).
The beneficial effect of MMF on CsA-induced arteriolopathy seems to be, at least partly, related to a decrease in TGF-β1 expression, a key fibrogenic cytokine implicated in a number of chronic diseases of the kidney and other organs. The central role of TGF-β in the pathogenesis of chronic allograft nephropathy and chronic CsA nephrotoxicity has been clearly demonstrated. Chronic nephrotoxicity is likely to represent a major contributor to the fibrosis of chronic allograft nephropathy and there is evidence to suggest that reducing CsA exposure may improve long-term allograft survival. We have previously shown that TGF-β plays an important role in the pathogenesis of chronic CsA nephrotoxicity by increasing ECM deposition and by decreasing its degradation through PAI-1, a plasmin protease inhibitor (5). Moreover, therapy with MMF was previously shown to be beneficial in decreasing TGF-β expression in chronically rejecting rat renal allografts (43). A similar benefit with CsA nephrotoxicity was also observed when large doses of MMF were used (19). In this study, we were able to demonstrate that clinically relevant doses of MMF were sufficient to decrease TGF-β1 expression in chronic CsA nephrotoxicity. This, in turn, was accompanied by a significant decrease in biglycan and collagen, suggestive of a decrease in ECM deposition in addition to a decrease in PAI-1, suggestive of an increase in ECM degradation. However, it remains unclear whether the effect of MMF on PAI-1 is independent of TGF-β or is related to TGF-β, which can up-regulate PAI-1 expression. Additional evidence that MMF is capable of reducing ECM accumulation comes from the 5/6 renal ablation model where MMF-reduced myofibroblasts and collagen III expression (13). Of interest is the observation that TGF-β is increased, in a dose-dependant fashion, in the granules of the juxtaglomerular apparatus (JGA) cells in CsA-treated rats (44). Although these observations may provide an explanation for decreased TGF-β expression in association with amelioration of the arteriolopathy, it remains unclear whether there is a pathogenic association between local cytokines and the arteriolopathy. In addition, it is to be emphasized that TGF-β1 was measured in the whole kidney, which may not entirely explain the changes seen in the vessels.
Other molecular mediators of fibrosis may also have been influenced by MMF. There is evidence to suggest that Ang II play a role in the development of chronic CsA nephrotoxicity (45). Of interest is that losartan can almost completely block CsA arteriolopathy, although tubulointerstitial fibrosis is significantly but only partially ameliorated (46). As Ang II is known to induce TGF-β in this model, regulation of Ang II by MMF may very well provide one of the missing links. Along those lines, we had previously shown that the use of losartan or enalapril in this model directly reduces fibrosis and also TGF-β1 expression (7). Moreover, the addition of blockers of the RAS to MMF is synergistic in the remnant kidney model and in a rat model of chronic allograft nephropathy (29,37,43). Mycophenolate mofetil therapy also decreases the number of Ang II-positive cells in the glomeruli of CsA-treated rats (20) and in the tubulointerstitium of salt-sensitive hypertensive rats where the expression of AT1 receptors remained unchanged (47). Moreover, MPA-treated cell cultures from the medullary thick ascending limb had a reduced AT1 receptor activity without affecting ACE activity (48). However, whether this represents a direct effect of MMF on the RAS or a suppression of Ang II-driven cytokines remains unclear.
Recent clinical studies have confirmed a beneficial effect associated with the addition of MMF and a reduction in the dose of CsA in the prevention of chronic allograft nephropathy. As chronic CsA nephrotoxicity is a major contributor to this lesion, this may partly explain some of the beneficial effects described with MMF addition on long-term renal allograft survival. We conclude that the use of MMF in maintenance immunosuppressive therapy in a CsA-based regimen may provide protection against CsA-associated arteriolopathy. This finding may offer another rationale for the use of MMF in solid organ transplantation in addition to its antirejection property. However, long-term clinical and histological studies are needed to determine if MMF has a lasting benefit on CsA-induced renal lesions.
This work was supported in part by a grant from the Dialysis Research Foundation (F.S.S.) and by a grant from the Legacy Research Foundation (W.M.B. and T.F.A.). Different parts of this work were presented at the American Society of Nephrology Annual Meeting, Philadelphia, PA, 2002, and at the 2003 American Transplant Congress, Washington, DC.