Cyclosporine A (CsA) is a major immunosuppressant widely used in organ transplantation, and its use has improved the survival of kidney, heart, and bone marrow grafts. However, the use of CsA is associated with a number of toxic effects. A major complication is nephrotoxicity, which affects both native and transplanted kidneys. Chronic CsA nephrotoxicity can progress and cause irreversible renal lesions characterized by dilation and atrophy of tubules, hyalinosis of afferent arterioles, and striped interstitial fibrosis (1). These changes can be duplicated in rats and mice by injection of CsA in combination with a low-sodium diet, which accelerates the development of renal lesions (2–5).
A number of studies have implicated transforming growth factor (TGF)β as one of the major pathogenic factors in fibrotic changes seen in renal diseases. Thus, increased expression of TGFβ mRNA and protein has been observed in association with mesangial expansion, glomerular sclerosis, and interstitial fibrosis (6). In both humans and experimental animals with CsA nephropathy, TGFβ mRNA and protein are increased in the kidney (7, 8). Wolf et al. (9) found that in cultured proximal tubular cells, CsA induces expression of TGFβ1 mRNA and protein in a dose-dependent manner. In CsA nephropathy in rats, antagonism of angiotensin II was found to suppress the increase in the renal TGFβ1 level and the histologic and functional alterations of the kidney (10, 11). In light of the well-described profibrotic actions of TGFβ1 (6), these findings indicate a close link between TGFβ1 and CsA-associated interstitial fibrosis, indicating that TGFβ is a potential target for therapeutic interventions in CsA nephrotoxicity.
The tissue level of the extracellular matrix (ECM) is tightly regulated through the balance between production and degradation of ECM components. TGFβ1 induces accumulation of ECM by shifting this balance toward increased production and decreased degradation of ECM (6, 12). It has been suggested that the fibrogenic activity of TGFβ1 is in part mediated by connective tissue growth factor (CTGF), a newly described member of cysteine-rich growth factors (13, 14). Thus, in cultured renal fibroblasts, TGFβ1 induces expression of CTGF mRNA, and TGFβ-stimulated ECM production is reduced by antisense-mediated inactivation of CTGF mRNA (14). Because increased expression of CTGF was described in glomerular and interstitial inflammatory lesions with cellular proliferation and ECM accumulation (13–16), and induction of CTGF was also found in association with CsA-induced myocardial changes (17), a part of the fibrotic changes occurring in CsA nephrotoxicity can be attributed to CTGF.
We have generated the chimeric protein TGFβ receptor (R)II/immunoglobulin (Ig)G Fc, as a novel anti-TGFβ, in which the extracellular domain of human TGFβRII was fused with human IgG1 Fc (18). Our previous study demonstrated that on transfection in vitro of an expression vector carrying cDNA for TGFβRII/IgG Fc, the chimeric protein was produced in a soluble form and neutralized the antiproliferative and matrix-inducing activities of TGFβ1 (18). We also showed that introduction of TGFβRII/IgG Fc by gene transfer in vivo suppressed accumulation of ECM in Thy-1 glomerulonephritis (18) and experimental unilateral ureteral obstruction (19).
To determine whether TGFβ can be a target of therapeutic intervention for CsA nephropathy and to examine the efficacy of TGFβRII/IgG Fc, we introduced TGFβRII/IgG Fc by electroporation in vivo and evaluated its effects on CsA nephrotoxicity in mice, with a specific focus on interstitial fibrosis.
MATERIALS AND METHODS
Animals and Experimental Protocol
Adult female ICR mice weighing 26 to 28 g were used (Japan Clea, Inc., Tokyo, Japan). The experimental protocol was in accordance with the National Institutes of Health “Guide for Care and Use of Laboratory Animals” and was approved by the Committee for Animal Care of Tokai University School of Medicine. Mice were fed a low-sodium diet with free access to water as previously described (20). After 1 week on a low-sodium diet, mice were assigned to the following study groups. All mice were kept on a low-sodium diet throughout the experiment.
Mice (n=4) were given subcutaneous injection of the vehicle daily for an additional 3 weeks.
Mice received a daily subcutaneous injection of CsA (Sigma, St. Louis, MO) at a dose of 25 mg/kg/d. On days 1 and 7 of CsA treatment, mice were transfected with a control vector containing cDNA for β-galactosidase. CsA injection was continued until sacrifice at 2 (n=4) or 3 weeks (n=10).
Mice received a daily subcutaneous injection of CsA at a dose of 25 mg/kg/d. On days 1 and 7 of CsA treatment, a vector containing cDNA for TGFβRII/IgG Fc was transfected. CsA injection was continued until sacrifice at 2 (n=4) or 3 weeks (n=10).
On the day of sacrifice, plasma samples were obtained with EDTA, and kidneys were harvested and used for RNA isolation or processed for histologic analyses.
DNA Transfection into the Skeletal Muscle
DNA was introduced into the muscle by in vivo electroporation as described previously (21). Briefly, while the mice were under light anesthesia with diethylether, 60 μL of 0.5% bupivacaine was injected into the bilateral tibialis anterior muscles. Three days later, the bupivacaine-treated portions were injected with 50 μg of an expression vector, either pCAGGS-TGFβRII/IgG Fc carrying cDNA for the chimeric protein consisting of the extracellular domain of human TGFβRII and the Fc portion of human IgG1 (18) or, as a control, pCAGGS-lacZ containing cDNA for β-galactosidase at a concentration of 1.5 μg/μL in saline (21). A pair of electrode needles were inserted into the muscle to a depth of 5 mm to encompass the DNA injection sites, and electric pulses were delivered using the CUY21 electroporator (Nepa Gene Co., Chiba, Japan). The voltage was maintained constant at 100 V during the pulse duration. Electrodes consisted of a pair of stainless steel needles of 5 mm in length and 0.4 mm in diameter, fixed with a distance of 5 mm. Three pulses (50 msec each) and an additional three pulses of the opposite polarity were administered to each injection site at a rate of 1 pulse/sec.
Determination of Plasma TGFβ1 Concentration
Plasma samples were kept frozen at −80°C until the assay. Samples were acid-activated to convert latent TGFβ1 to active form. The concentration of total TGFβ1 was determined by a sandwich enzyme-linked immunosorbent assay (R&D Systems, Minneapolis, MN) with a standard curve constructed using recombinant human TGFβ1. Values were expressed as nanograms/milliliter.
Two micrometer-thick sections were stained with either periodic acid-Schiff (PAS) or Masson’s trichrome method. The degree of tubulointerstitial injury was semiquantitatively evaluated on PAS-stained sections as previously described (5). Tubulointerstitial injury was defined by lesions including tubular casts, dilation and atrophy, thickening of the tubular basement membrane, and interstitial matrix expansion. A minimum of 30 fields were examined in each section with 20× objectives, and tubulointerstitial injury was graded using the following scores: 0=less than 5% of the area involved; 1=5% to 25% of the area involved; 2=25% to 50% of the area involved; and 3=more than 50% of the area involved.
The extent of interstitial fibrosis was determined as the collagen fractional volume of renal cortical interstitium by a point-counting method on Masson’s trichrome-stained sections as we have described (22). A minimum of 50 fields were examined with 40x objectives using a grid of 0.0625 mm2 with 100 points. Values were expressed as the number of positive points per 100 points.
For staining of TGFβ1, 4 μm-thick frozen sections were fixed in acetone for 10 min at 4°C and incubated in 0.3% H2O2 in methanol for 30 min. After blocking with normal goat serum, sections were incubated with 1:25 dilution of rabbit anti-mouse TGFβ1 antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA), followed by peroxidase-conjugated anti-rabbit Ig (DAKO Corp., Carpinteria, CA). Color was developed with diaminobenzidine, and sections were counterstained with PAS. The number of TGFβ1-expressing tubulointerstitial cells was determined in the cortex excluding glomeruli using a grid of 0.0625 mm2. A minimum of 50 grids were examined in each section, and the number of TGFβ1-expressing cells was expressed as cells/mm2.
For staining of collagen type IV and macrophages, rabbit anti-mouse collagen type IV antibody (Chemicon, Temecula, CA) and monoclonal rat anti-mouse F4/80 (Serotec, Oxford, UK) were used on 0.1% trypsin-treated, 4 μm-thick, paraffin-embedded sections. Counterstaining was performed with hematoxylin or methyl green. F4/80-positive macrophages were enumerated in the cortex excluding glomeruli using a grid of 0.0625 mm2. A minimum of 50 grids were examined in each section, and the average number of macrophages/ mm2 was obtained. The extent of collagen type IV deposition in the cortical tubulointerstitium was determined as previously described (23). In each section, 30 non-overlapping fields were randomly selected and graded using a score of 0 to 3, where 0 was equivalent to the staining of a normal kidney, 1 was mild change, 2 was moderate change, and 3 was severe change (24). Scores from these 30 fields were averaged.
Northern Blot Analysis
Total RNA was isolated from the kidney using RNA-Solv Reagent (Omega Biotek, Inc., Doraville, GA), and a 20-μg aliquot was separated on 1% agarose gel containing formaldehyde, transferred onto Hybond N+ membrane (Amersham Biosciences, Piscataway, NJ), and hybridized with [α-32P]-labeled cDNA probe for TGFβ1, CTGF, or β-actin. Membranes were exposed to the imaging plate, and the intensity of hybridization bands was measured as a ratio to that of β-actin. cDNA probes for TGFβ1 and β-actin were from Clontech (Palo Alto, CA), and CTGF cDNA probe was generated by polymerase chain reaction using primers 5′-GAGTGGGTGTGTGACGAGC-CCAAGG-3′ and 5′-ATGTCTCCGTACTTCCTGTAGT-3′ (17).
The results are expressed as means±standard error. The unpaired Student t test was used to analyze the difference between two groups, and values were regarded significant at P less than 0.05.
Cyclosporine A-Induced Increase in Plasma and Renal TGFβ1 and Its Suppression by TGFβRII/IgG Fc
By 2 weeks of CsA treatment, the plasma TGFβ1 level in the CsA group was significantly elevated (23.9±1.7 vs. 12.6±1.7 ng/mL in the control group; P <0.05) (Fig. 1). This increase was markedly suppressed in the CsA+TGFβRII/IgG Fc group compared with the control group (9.7±0.3 ng/mL; P <0.001 vs. control group). At 3 weeks of CsA treatment, plasma TGFβ1 levels in the CsA and CsA+TGFβRII/IgG Fc groups were comparable to that in the control mice (14.5±2.1 and 16.3±0.6 ng/mL; not significant [NS] vs. control group).
Renal TGFβ1 expression was assessed by immunohistochemistry as the number of TGFβ1-expressing tubulointerstitial cells (Fig. 2A). In the control group, a small number of TGFβ1-expressing cells were found throughout the tubulointerstitial area (21.9±0.7 cells/mm2). On CsA treatment, an increasing number of TGFβ1-expressing cells were found in the tubulointerstitial area and focally around Bowman’s capsule (Fig. 2A). After 2 weeks, the number of TGFβ1-expressing cells increased in the CsA group to 10 times the value seen in the control group (205.4±40.0 cells/mm2; P <0.01 vs. control group). By 3 weeks of CsA treatment, the level of TGFβ1-expressing cells in the CsA group declined to 50% of that seen at 2 weeks, but the value was significantly higher than that in the control group (114.6±24.4 cells/mm2; P <0.05 vs. CsA group at 2 weeks, P <0.01 vs. control group). The increase in TGFβ1-expressing cells occurring at 2 weeks of CsA treatment was largely reduced in the TGFβRII/IgG Fc group by more than 60% (90.8±25.1 cells/mm2; P <0.05 vs. CsA group). By 3 weeks, the number of TGFβ1-expressing cells was also suppressed by TGFβRII/IgG Fc, with the value similar to that found at 2 weeks (Fig. 2). Thus, at 3 weeks, the number of TGFβ1-expressing cells in the TGFβRII/IgG Fc group was comparable to that in the CsA group and significantly higher than that in the control group (81.8±9.5 cells/ mm2; NS vs. CsA group at 3 weeks and P <0.01 vs. control group).
Northern analysis showed that the renal TGFβ1 mRNA level increased persistently on CsA treatment (Fig. 2B). A significant increase was detected by 2 and 3 weeks of CsA treatment (0.21±0.04 and 0.18±0.04 vs. 0.10±0.02 in control group; P <0.05). The observed CsA-induced increases in the TGFβ1 mRNA expression were little affected by the TGFβRII/IgG Fc intervention, with the values at 2 and 3 weeks comparable to those in the CsA group (0.19±0.04 and 0.16±0.06; NS vs. CsA group).
We also examined the expression of CTGF mRNA, a recently described growth factor that has been implicated in renal interstitial fibrosis (13–16). After 2 weeks of CsA treatment, there was an increase in the renal CTGF mRNA level to 2.18±1.04, which did not attain statistical significance (vs. 1.15±0.64 in control group; P ±0.07). By 3 weeks, CTGF mRNA expression returned to the baseline level (1.02±0.18; NS vs. control group). The TGFβRII/IgG Fc intervention had no appreciable effect on the renal CTGF mRNA level at 2 or 3 weeks of CsA treatment (2.29±1.20 and 0.96±0.26; NS vs. CsA group).
Persistent Attenuation of Tubulointerstitial Injury by TGFβRII/IgG Fc
PAS staining showed a normal kidney architecture in the control group, except for hypertrophy of juxtaglomerular cells, which is known to occur in the setting of salt depletion (Fig. 3). CsA treatment induced characteristic histologic changes by 2 weeks, including tubular casts, atrophy and dilation, early striped cortical fibrosis, and inflammatory cell infiltration (Fig. 3). By 3 weeks, these lesions were more advanced and extensive in the CsA group, with severe interstitial fibrosis (Fig. 3). The tubulointerstitial lesions were characterized by collapse of tubular segments, with remaining tubules showing dilation and vacuolization, and increased interstitial matrix deposition (Fig. 3). Semiquantitative scoring showed progressive aggravation of the tubulointerstitial injury in the CsA group (0.87±0.16 and 2.01±0.24 at 2 and 3 weeks, respectively, vs. 0.04±0.01 in control group; P <0.001 and P <0.0001). In the CsA+TGFβRII/IgG Fc group, there was mild tubulointerstitial injury with a score similar to that found in the CsA group (Fig. 3; 0.78±0.12; NS vs. CsA group). However, the subsequent progression of these lesions was largely suppressed by the TGFβRII/IgG Fc intervention (Fig. 3). Thus, the intervention induced marked attenuation of the tubulointerstitial injury, with the score significantly less than that in the CsA group at 3 weeks (0.86±0.16; P <0.01 vs. CsA group).
The degree of interstitial fibrosis was quantitatively assessed as the collagen fractional volume of renal cortical interstitium by a point-counting method on Masson’s trichrome-stained sections (Fig. 4). The CsA group showed progression of interstitial fibrosis with 9.98±1.50 and 18.01±1.05 seen at 2 and 3 weeks of CsA treatment (vs. 2.08±0.10 in control group; P <0.01 and P <0.001). Parallel with the findings on tubulointerstitial injury, the TGFβRII/IgG Fc intervention effectively suppressed the development of interstitial fibrosis at both 2 and 3 weeks (6.60±1.01 and 6.71±0.49 at 2 and 3 weeks, respectively; P <0.05 and P <0.001 vs. CsA group).
Collagen type IV was detected by immunohistochemistry. After 2 weeks of CsA treatment, a significant increase was noted in collagen type IV, followed by a further increase by 3 weeks (Fig. 5). Accumulation of collagen type IV occurred in the areas of tubulointerstitial fibrosis, with most pronounced increases seen around the thickened basement membrane of Bowman’s capsule and degenerated tubules (Fig. 5). At 2 weeks of CsA treatment, collagen type IV deposition increased in the CsA+TGFβRII/IgG Fc group to a level similar to that seen in the CsA group (0.96±0.18 vs. 0.98±0.18 in the CsA group; NS;Fig. 5). However, progressive accumulation of collagen type IV occurring by 3 weeks was largely suppressed by the TGFβRII/IgG Fc intervention (Fig. 5), with the score significantly less than that in the CsA group (0.86±0.13 vs. 1.50±0.24 in the CsA group; P <0.05).
CsA treatment induced marked and continuous infiltration of F4/80-positive macrophages throughout the experiment (Fig. 6). Accumulation of macrophages was observed around the tubules and within the interstitium. Interstitial macrophages assumed a periglomerular distribution, and most intense accumulation was seen in areas of interstitial fibrosis (Fig. 6). Thus, compared with the control group, the number of macrophages increased persistently in the CsA group at 2 and 3 weeks (361.0±67.3 and 433.4±35.7 cells/mm2 vs. 95.0±6.2 cells/mm2 in the control group; P <0.05 and P <0.001 vs. control group). In the CsA+TGFβRII/IgG Fc group, macrophage accumulation was somewhat attenuated, but was comparable to that in the CsA group by 2 weeks (251.2±60.2 cells/mm2; NS vs. CsA group). Thereafter, accumulation of macrophages was markedly suppressed by the TGFβRII/IgG Fc intervention by 3 weeks (177.3±44.1 cells/mm2; NS and P <0.01 vs. CsA group).
In the present study, CsA administration to mice maintained on a low-sodium diet caused significant increases in both the plasma and renal TGFβ1 levels in association with induction of renal TGFβ mRNA. In parallel, CsA-treated mice developed histologic alterations characteristic of CsA nephrotoxicity, including tubular dilation, vacuolization and atrophy, tubulointerstitial fibrosis, and accumulation of macrophages and collagen type IV. These morphologic changes were more pronounced by 3 weeks of CsA administration, with severe striped interstitial fibrosis. We demonstrated that the intervention with the chimeric protein TGFβRII/IgG Fc suppressed the increases in the plasma and renal TGFβ1 levels and resulted in attenuation of tubulointerstitial changes throughout the course of the experiment.
TGFβ is a multifunctional cytokine that plays an important role in the regulation of ECM. It was demonstrated in rats and mice that CsA administration induces histologic and functional alterations of the kidney parallel with an increase in the renal and plasma TGFβ1 levels (2–5, 8), and that these CsA-induced alterations are attenuated concomitantly by antagonism of angiotensin II (10, 11). These observations provide circumstantial evidence that TGFβ1 is an effector involved in CsA nephrotoxicity. In recent studies, anti-TGFβ1 antibodies were used to seek more direct evidence for the potential link between TGFβ1 and CsA-induced histologic alterations, and variable effects were observed (24, 25). In mice, a single administration of neutralizing anti-TGFβ1 antibody resulted in attenuation of early CsA-induced changes in histology and collagen mRNA level (24). Multiple administration of the antibodies in rats ameliorated the decline of renal function and arteriolar hyalinosis, with limited effects on histologic changes (25). The present study with the TGFβ-capturing chimeric protein TGFβRII/IgG Fc demonstrated that this anti-TGFβintervention abrogated CsA-induced chronic interstitial alterations and, thus, provides another piece of evidence that the development of these lesions are attributable to TGFβ1.
Determination of the plasma and renal levels of TGFβ1 showed that the intervention with TGFβRII/IgG Fc suppressed the increases occurring in TGFβ1 levels at 2 weeks of CsA administration. Because the chimeric protein functions as a soluble receptor and captures TGFβ1 molecules, the efficiency of these assays using anti-TGFβ1 antibodies can be altered as a consequence of the capture of TGFβ1 by TGFβRII/IgG Fc. In this regard, we found that the CsA-induced increase in renal TGFβ1 mRNA was not affected by the intervention. These contrasting effects seen on TGFβ1 mRNA and protein indicate that the capture of TGFβ1 molecule by the chimeric protein did not affect CsA-induced TGFβ1 production per se, but masked TGFβ1 molecules from the recognition by anti-TGFβ1 antibodies used in the enzyme-linked immunosorbent assay and immunohistochemistry. Taken together, the present results indicate that the intervention with the chimeric protein can effectively capture and neutralize the TGFβ1 molecules in the plasma and the kidney.
CTGF is a cysteine-rich growth factor previously suggested to mediate a part of TGFβ1-induced increase in ECM production (13, 14). CsA treatment caused a slight increase in the renal CTGF mRNA expression at 2 weeks, followed by a decline to the baseline level. This transient increase, however, did not reach statistical significance (P =0.07 vs. control group) and was not affected by the TGFβRII/IgG Fc intervention (Fig. 2B). Thus, our findings indicate that in all likelihood, CTGF has no appreciable role in the fibrotic changes within 3 weeks of CsA treatment.
A prevailing view of chronic CsA nephrotoxicity dictates that CsA exerts acute effects on the vasculature and hemo-dynamics and causes early hypoperfusion in the medulla, with such clinical manifestations as an acute decrease in glomerular filtration rate and an increase in serum creatinine (4, 26). This acute phase is followed by activation of reninangiotensin system, which potentiates to chronic, low-grade ischemic injury to poorly perfused areas in the cortex and corticomedullary junctions (10, 11). Ischemic tissue insult and CsA induce production of macrophage chemoattractants that recruit macrophages into the lesions. Infiltrating macrophages secrete growth-promoting factors and profibrotic cytokines including TGFβ1 (4, 6, 12). Production of TGFβ1 is also induced in tubular epithelial cells through direct action of CsA (9). Because TGFβ1 is a potent chemoattractant for monocytes (6) and stimulates its own production through a positive feedback loop (6), the sequence of events further enhances accumulation of macrophages and local production of TGFβ1. Finally, striped interstitial fibrosis, one of the hallmarks of CsA nephrotoxicity, develops as a consequence of chronic tissue injury and TGFβ1-induced accumulation of ECM. In this scenario, the role of TGFβ1 is limited to the induction of fibrotic changes. Recently, we demonstrated in a model of ureteral obstruction that reduction of macrophages infiltrating the obstructed kidneys is associated with more severe interstitial fibrosis, shedding light on the heretofore unappreciated antifibrotic function of infiltrating macrophages (27). On neutralization of profibrotic and macrophage chemoattractive activities of TGFβ by the chimeric protein, the antifibrotic function of infiltrating macrophages contributes to the salutary effects of the chimeric protein maintained after reversion of the TGFβ levels.
The present study showed that in addition to the use of neutralizing anti-TGFβ1 antibody, the intervention with TGFβ-capturing TGFβRII/IgG Fc is an effective anti-TGFβ intervention. These anti-TGFβ interventions have advantages and disadvantages. Because of the half-life of exogenous antibody of approximately 24 hr, neutralizing anti-TGFβ1 antibody is limited in the duration of effects and may require multiple administrations. Also, the specificity of the antibody may be confined to a single isoform of TGFβ. In contrast, an expression vector introduced in vivo by gene transfer can remain active for up to 14 days (18, 21). The chimeric protein with IgG-Fc has an extended plasma half-life (18). In addition, TGFβRII recognizes all three isoforms of TGFβ (28). An inherent concern with anti-TGFβ interventions is the theoretic potential that nullification of anti-inflammatory and antiproliferative activities of TGFβ can augment inflammation and tumor formation. In this regard, the present and previous studies with anti-TGFβ interventions showed no discernible signs of these adverse effects during the course of the experiment (24, 25). Inhibition of TGFβ action might also reduce the efficacy of CsA, because TGFβ is believed to mediate a part of the immunosuppressive activity of CsA.
Of note, although the renal TGFβ1 mRNA and protein were only partially suppressed by the intervention with TGFβRII/IgG Fc at 3 weeks of CsA treatment, tubulointerstitial injury and accumulation of collagen type IV were abrogated not only at 2 weeks but also at 3 weeks. As described previously, the results with neutralizing anti-TGFβ1 antibody all demonstrate that anti-TGFβ1 intervention is effective in suppressing CsA nephrotoxicity in vivo (24, 25). Furthermore, the present and previous studies in mice demonstrated that neutralization of the early increase in the TGFβ1 level was associated with attenuation of the later aggravation of renal lesions, suggesting that when given early, anti-TGFβ intervention can attenuate the later progression of CsA nephropathy. In this regard, a recent study demonstrated that even in established chronic CsA nephropathy, pan-specific neutralizing anti-TGFβ antibody ameliorated histologic alterations and renal function (29). Thus, a possibility is raised that anti-TGFβ intervention can improve CsA-induced morphologic and functional changes even in the progression phase. It should be noted that because of the limited lifespan of the introduced cDNA and the chimeric protein, long-term suppression of fibrosis would require repeated introduction of the expression vector.
The present study demonstrated that a novel anti-TGFβ1 intervention with TGFβ-capturing TGFβRII/IgG Fc is an effective alternative to the use of neutralizing anti-TGFβ1 antibody to suppress the development of tubulointerstitial alterations in chronic CsA nephrotoxicity in vivo. Together with the recently described magnesium supplementation for prevention of CsA nephrotoxicity (30), anti-TGFβ1 interventions can substantially decrease the adverse renal effects of CsA.
The authors thank Drs. Fumio Niimura and Takako Asano for discussion and technical advice; Suguri Niwa, Atsuko Satoh, and Chie Sakurai for their excellent technical assistance; and Mari Inoue for superb editorial assistance.
1. Mihatsch MJ, Thiel G, Spichtin HP, et al. Morphological findings in kidney transplants after treatment with cyclosporine. Transplant Proc 1983; 15(Suppl 1): 2821.
2. Rosen S, Greenfeld Z, Brezis M. Chronic cyclosporine-induced nephropathy in rats. Transplantation 1990; 49: 445.
3. Young BA, Burdmann EA, Johnson RJ, et al. Cyclosporine A induced arteriolopathy in a rat model of chronic cyclosporine nephropathy. Kidney Int 1995; 48: 431.
4. Young BA, Burdmann EA, Johnson RJ, et al. Cellular proliferation and macrophage influx precede interstitial fibrosis in cyclosporine nephrotoxicity. Kidney Int 1995; 48: 439.
5. Shihab FS, Andoh TF, Tanner AM, et al. Sodium depletion enhances fibrosis and the expression of TGF-beta 1 and matrix proteins in experimental chronic cyclosporine nephropathy. Am J Kidney Dis 1997; 30: 71.
6. Border WA, Noble NA. Transforming growth factor β in tissue fibrosis. N Engl J Med 1994; 331: 1286.
7. Pankewycz OG, Miao L, Isaacs R, et al. Increased renal tubular expression of transforming growth factor beta in human allografts correlates with cyclosporine toxicity. Kidney Int 1996; 50: 1634.
8. Khanna A, Kapur S, Sharma V, et al. In vivo hyperexpression of transforming growth factor-β1 in mice: stimulation by cyclosporine. Transplantation 1997; 63: 1037.
9. Wolf G, Thaiss F, Stahl RAK. Cyclosporine stimulates expression of transforming growth factor-β in renal cells. Possible mechanism of cyclosporines antiproliferative effects. Transplantation 1995; 60: 237.
10. Burdmann EL, Andoh TF, Nast CC, et al. Prevention of experimental cyclosporine-induced interstitial fibrosis by losartan and enalapril. Am J Physiol 1995; 269: F491.
11. Shihab FS, Bennett WM, Tanner AM, et al. Angiotensin II blockade decreases TGF-beta 1 and matrix proteins in cyclosporine nephropathy. Kidney Int 1997; 52: 660.
12. Blobe GC, Schiemann WP, Lodish HF. Role of transforming growth factor β in human diseases. N Engl J Med 2000; 342: 1350.
13. Yokoi H, Sugawara A, Mukoyama M, et al. Role of connective tissue growth factor in profibrotic action of transforming growth factor-beta: a potential target for preventing renal fibrosis. Am J Kidney Dis 2001; 38(Suppl 1): S134.
14. Yokoi H, Mukoyama M, Sugawara A, et al. Role of connective tissue growth factor in fibronectin expression and tubulointerstitial fibrosis. Am J Physiol 2002; 282: F933.
15. Ito Y, Aten J, Bende RJ, et al. Expression of connective tissue growth factor in human renal fibrosis. Kidney Int 1998; 53: 853.
16. Rister BL, Denichilo M, Cortes P, et al. Regulation of connective tissue growth factor activity in cultured rat mesangial cells and its expression in experimental diabetic glomerulosclerosis. J Am Soc Nephrol 2000; 11: 25.
17. Finckenberg P, Lassila M, Inkinen K, et al. Cyclosporine induces myocardial connective tissue growth factor in spontaneously hypertensive rats on high-sodium diet. Transplantation 2001; 71: 951.
18. Isaka Y, Akagi Y, Ando Y, et al. Gene therapy by transforming growth factor-β receptor-IgG Fc chimera suppressed extracellular matrix accumulation in experimental glomerulonephritis. Kidney Int 1999; 55: 465.
19. Imai E, Akagi Y, Isaka Y. Towards gene therapy for renal diseases. Nephrologie 1998; 19: 397.
20. Kon V, Hunley T, Fogo A. Combined antagonism of endothelin A/B receptors links endothelin to vasoconstriction whereas angiotensin II effects fibrosis. Studies in chronic cyclosporine nephrotoxicity in rats. Transplantation 1995; 60: 89.
21. Aihara H, Miyazaki J. Gene transfer into muscle by electroporation in vivo. Nat Biotechnol 1998; 16: 867.
22. Ma J, Nishimura H, Fogo A, et al. Accelerated fibrosis and collagen deposition develop in the renal interstitium of angiotensin type 2 receptor null mutant mice during ureteral obstruction. Kidney Int 1998; 53: 937.
23. Kaneto H, Morrissey J, McCracken R, et al. Enalapril reduces collagen type IV synthesis and expansion of the interstitium in the obstructed kidney. Kidney Int 1994; 45: 1637.
24. Khanna AK, Cairns VR, Becker CG, et al. Transforming growth factor (TGF)- β mimics and anti-TGF-β antibody abrogates the in vivo effects of cyclosporine. Demonstration of a direct role of TGF-β in immunosuppression and nephrotoxicity of cyclosporine. Transplantation 1999; 67: 882.
25. Islam M, Burke JF Jr, McGowan TA, et al. Effect of anti-transforming growth factor-β antibodies in cyclosporine-induced renal dysfunction. Kidney Int 2001; 59: 498.
26. Thomas SE, Andoh TF, Pichler RH, et al. Accelerated apoptosis characterizes cyclosporine-associated interstitial fibrosis. Kidney Int 1998; 53: 897.
27. Nishida M, Fujinaka H, Matsusaka T, et al. Absence of angiotensin II type 1 receptor in bone-arrow-derived cells is detrimental in the evolution of renal fibrosis. J Clin Invest 2002; 110: 1859.
28. Wrana JL, Attisano L, Wieser R, et al. Mechanism of activation of the TGF-β receptor. Nature 1994; 370: 341.
29. Ling H, Li X, Jha S, et al. Therapeutic role of TGF-β-neutralizing antibody in mouse cyclosporin A nephropathy: morphologic improvement associated with functional preservation. J Am Soc Nephrol 2003; 14: 377.
30. Asai T, Nakatani T, Yamanaka S, et al. Magnesium supplementation prevents experimental chronic cyclosporine A nephrotoxicity via reninangiotensin system independent mechanism. Transplantation 2002; 74: 784.