Transforming growth factor-β1 induces extracellular matrix formation in glomerulonephritis

Cell Differentiation and Development, 32 (1990) 425-432 © 1990 Elsevier Scientific Publishers Ireland, Ltd. 0922-3371/90/$03.50

425

CELDIF 99928

Transforming growth factor-ill induces extracellular matrix formation in glomerulonephritis W a y n e A. Border 1 and Erkki Ruoslahti 2 I Division of Nephrology, University of Utah School of Medicine, Salt Lake City, Utah, U.S.A. and 2 Cancer Research Center, La Jolla Cancer Research Foundation, La Jolla, California, U.S.A.

Extracellular matrices can be important in disease. Glomerulonephritis is an inflammation of the kidney that is characterized by the accumulation of extracellular matrix within the damaged glomeruli. We have shown that transforming growth factor-131 (TGF-I31) is unique in regulating the production of proteoglycans and matrix glycoproteins by glomerular cells in vitro. In an experimental model of glomerulonephritis in rats, we found increased proteoglycan and fibronectin synthesis by cultured nephritic glomeruli, which was greatly reduced by the addition of antiserum to TGF-I31. Conditioned media from glomerular cultures, when added to normal cultured mesangial cells, induced elevated proteoglycan synthesis. The stimulatory activity of the conditioned media was blocked by addition of TGF-~I1 antiserum. Glomerular histology showed mesangial matrix expansion in a time course that roughly paralleled the elevated proteoglycan synthesis by the nephritic glomeruli. At the same time there was an increased expression of TGF-I31 mRNA and TGF-131 protein in the giomeruli. Administration of anti-TGF-131 at the time of induction of glomerulonephritis suppressed the elevated extracellular matrix production and dramatically attenuated histological manifestations of the disease. A small proteoglycan, decorin, also inhibits the activity of TGF-13, potentially providing an alternative format for the prevention of TGF-13 activity. Our results provide direct evidence for a causal role of TGF-I31 in the pathogenesis of the experimental disease and suggest a new approach to the therapy of glomerulonephritis. Growth factors; Transforming growth factor-13; Extracellular matrix; Kidney disease; Glomerulonephritis

Introduction Extracellular matrices perform many important functions in controlling the attachment, migration, proliferation and differentiation of cells [1]. In some situations, however, pathological accumula-

Correspondence address: W.A. Border, Division of Nephrology, University of Utah School of Medicine, Salt Lake City, UT 84132, U.S.A.

tion of extracellular matrix can destroy the function of an organ. One such situation is glomerulonephritis [2]. In the normal glomerulus, the glomerular extracellular matrix exists as a complex superstructure of interacting macromolecules that surrounds the mesangial cells and, in a specialized form as the glomerular basement membrane, separates the capillary endothelium from the visceral epithelium. The glomerular matrices contribute importantly to the structural organization of the glomerulus and to the biophysical properties that

426

govern glomerular filtration and permeability [3]. Glomerulonephritis is an inflammation of the kidney that is characterized by the accumulation of extracellular matrix within the damaged glomeruli. We have studied the matrix accumulation in an experimental model of glomerulonephritis in rats that mimics the most common form of glomerulonephritis in humans, mesangial proliferative glomerulonephritis [4]. The earliest histological finding in this form of glomerulonephritis is an increase in the quantity of the mesangial matrix, with or without mesangial hypercellularity. In its progressive form, there is a continued expansion of the mesangial matrix associated with mesangial cell proliferation. Human mesangial proliferative giomerulonephritis occurs in several forms but most frequently in a primary form as IgA nephropathy (Berger's disease) and secondarily as a complication of systemic lupus erythematosus. In a review of human and experimental mesangial disorders [2] we found that (a) mesangial cell injury can be induced by a variety of immunologic, toxic and hemodynamic stimuli and (b) that despite termination of the initial stimulus, there was often continued accumulation of mesangial matrix. These facts suggested the presence of an intervening process - a missing link between acute injury and the progressive build-up of glomerular matrix. We postulated that the missing link might be the activation and release of TGF-fl [5,6]. TGF-fl has been shown to have widespread effects on extracellular matrix [7,8]. There is growing evidence that some of the multiplicity of TGFfl actions are mediated through regulatory effects on: (a) extracellular matrix synthesis [9-12], (b) enzymes that degrade extracellular matrix [13,14] and (c) the expression of extracellular matrix receptors on cells [15]. Moreover, TGF-fl has been isolated from bovine kidney [16] and rat kidney fibroblasts [17] and murine glomerular cells [18] have been shown, as has nearly every cell line, to possess TGF-fl receptors. We have obtained evidence for a causal role of TGF-fll in the pathogenesis of the experimental disease [19] and shown that antibodies against TGF-fl can prevent the matrix accumulation [20]. Such antibodies, or the TGF-fl inhibitor [21], decorin, may provide new approaches to the therapy of glomerulonephritis.

Results

Effects of growth factors on matrix production in glomerular cell culture Our first experiments were carried out on cultured rat mesangial [5] and glomerular epithelial cells [6] and were designed to characterize the effects of TGF-fl and other growth factors on matrix production by these cells. The other growth factors studied were platelet derived growth factor (PDGF), interleukin-1 (IL-1) and tumor necrosis factor (TNF) because each has been implicated in the pathogenesis of experimental glomerulonephritis [22-24]. The third resident cell in the glomerulus, the endothelial cell, cannot easily be cultured and is not thought to contribute to the composition of the matrices in the mature glomerulus [25]. Mesangial cells and glomerular epithelial cells were isolated from adult rat glomeruli and grown to subconfluence in 6-well multiplates as described [5,6]. The cells were made serum free for 24 h and then placed for 48 h in methionine or sulfate-free medium containing TGF-fll (25 n g / ml), PDGF (10 U/ml), It-1 (5 U / m l ) or TNF (500 U/ml). Growth factor concentrations were selected from a review of publications in which clear biological effects were demonstrated. Five hours before the end of the experiment, [35S]sulfate (to label proteoglycans) or [35S]methionine (to label proteins) were added. The conditioned media were removed and analyzed under reducing conditions using SDS-PAGE and fluorography. Under basal conditions we found that mesangial cells and glomerular epithelial cells secreted two chondroitin/dermatan sulfate proteoglycans (biglycan and decorin), a large proteoglycan of unspecified composition, fibronectin, laminin and type IV collagen. In addition, mesangial cells produced types I, III and VI collagen. Addition of TGF-fll to mesangial cell cultures dramatically increased the density of two broad proteoglycan bands centered at 200 and 120 kDa; while the other growth factors had little effect [5]. TGF-fll did not demonstratively affect the pattern of synthesis of other matrix proteins. SDS-PAGE analysis of the cell layer showed a qualitatively similar effect of TGF-fll on matrix components;

427 however, immunofluorescence microscopy showed a subtle increase in fibronectin deposition. In order to characterize the proteoglycans regulated by TGF-fll, labeled conditioned medium was incubated with a mixture of protease inhibitors and enzymes to degrade the glycosaminoglycan side chains of proteoglycans. The bands were thus identified as being chondroitin/dermatan sulfate proteoglycans and to contain core proteins of approximately 40 kDa. We immunoprecipitated labeled conditioned medium using antibodies raised to synthetic peptides of portions of the proteoglycan core proteins. The immunoprecipitation results showed that the T G F - f l l effect is confined to two proteoglycan bands that are the mesangial cell equivalent of the chondroitin/ dermatan sulfate proteoglycans isolated from bone-biglycan (PG I) and decorin (PG II). Biglycan is named for its two glycosaminoglycan side chains [26]; decorin was so named because of its binding to (or "decorating") collagen fibrils [27]. Although little is known about the function of biglycan, decorin has recently been shown to strongly influence cell proliferation and morphology [28] and to bind TGF-fll [21]. In separate experiments, we found that exposure of glomerular epithelial cells to TGF-fll

TABLE I TGF-fl induction of extracellularmatrix by glomerular cells in culture Matrix Components Proteoglycans Biglycan Decorin Heparan sulfate Fibronectin Laminin Collagen Type I Type III Type IV Type VI

Mesangial c e l l Epithelialcell Medium Mono- Medium Monolayer layer t ttt t t ?t

t t

tttt T

0 0

0

0

0

0

0 0

t 0

ttt t t

t ~t t t

0 0 0 0

0 0 0 0

NP NP 1"t t NP

NP NP ttt NP

t = induction, 0 = no effect, NP = component not produced.

strikingly increased the secretion of biglycan and both secretion and cell layer accumulation of fibronectin and type IV collagen [6]. TGF-fll did not increase the secretion of heparan sulfate proteoglycans and the other growth factors did not demonstrably affect extracellular matrix production by the glomerular epithelial cells. The regulation by T G F - f l l of glomerular cell matrix production is summarized in Table 1.

Elevation of TGF-fl and matrix synthesis in experimental glomerulonephritis In order to show the relevance of our glomerular cell culture results to disease, we utilized a model of mesangial proliferative glomerulonephritis in the rat [19]. Injection of antithymocyte serum into the rat produces a dose and complement dependent, selective injury to mesangial cells causing an acute mesangial proliferative glomerulonephritis [29]. The reason for the injury is that mesangial cells, but not glomerular endothelial or epithelial cells, express a thy-I-like antigen on their surface [30]. The dose of antithymocyte serum we administered produced an acute form of mesangial injury glomerulonephritis [5]. There was a decrease in matrix noted on days 1 and 4, presumably caused by complement-mediated lysis of a portion of the mesangial cells, followed by a definite increase in the mesangial extracellular matrix, beginning on day 7, becoming maximal on day 14, and decreasing thereafter. Functional changes in this model of glomerulonephritis consisted of: (a) transient proteinuria during the first week, (b) no significant change in levels of serum creatinine and, (c) a slight but significant elevation of systolic blood pressure only on day 14. Groups of nephritic animals were sacrificed 1, 4, 7, 14 and 28 days after being injected with antithymocyte serum. Their glomeruli were isolated, placed in culture and biosynthetically labeled to identify newly synthesized proteoglycans, as a marker of TGF-fl activity, and fibronectin, the most prominent glycoprotein found in extracellular matrix. One day after antithymocyte serum injection, proteoglycan synthesis was the same as in normal controls; however, on day 4 there was a striking induction of proteo-

428

glycan production that reached a 4900% increase on day 7, and which then declined on days 14 and 28. Glomerular fibronectin production also increased on day 4 and remained elevated through day 14 before declining toward control levels by day 28. Immunofluorescent examination of nephritic kidney tissue showed that the pathologic matrix, found on days 7 and 14, stained more brightly with antibodies against fibronectin, than did matrix in normal glomeruli. To verify the increase of TGF-fl expression in the nephritic kidneys suggested by the proteoglycan assays, the presence of TGF-fll mRNA and protein in the glomeruli was analyzed. Increased levels of TGF-fll mRNA were found in nephritic glomeruli on days 4, 7 and 14. Scanning of the RNA transfer blots showed that the increase was about 5-fold at its peak on day 7. The amount of RNA applied to the gel was controlled by assaying for the RNA of a "housekeeping" enzyme, and this RNA remained constant during the disease process. Anti-LC is an antibody made against a synthetic peptide from TGF-fll that reacts with cells that are thought to be synthesizing TGF-fll [31]. We used anti-LC to detect TGF/31 production by glomerular cells throughout the 28 day course of glomerulonephritis. Glomeruli from control rats showed an average of about 20 cells per glomerulus that were positive with the anti-LC serum. In glomeruli from nephritic animals, the number of glomerular cells stained by anti-LC was unchanged on day 4, doubled on day 7, and decreased later roughly paralleling the other indicators of TGF-fll expression.

Causal role of TGF-fl in elevated matrix synthesis by nephritic glomeruli An antiserum raised against a synthetic peptide from TGF-fll was used to determine if TGF-fl is responsible for the induction of proteoglycan and fibronectin synthesis by the nephritic glomeruli. The TGF-fll antiserum or control serum were added to cultures of nephritic glomeruli harvested from kidneys on day 7 following antithymocyte serum administration. The TGF-B1 antiserum, compared to control serum, reduced the glomerular synthesis of biglycan, decorin and fibronectin by an average of 70%.

To determine if the nephritic glomeruli were releasing increased amounts of TGF-fl, conditioned media from normal and nephritic glomeruh were assayed for their ability to induce proteoglycan synthesis when added to normal cultured mesangial cells. We, and others, have shown that the ability to stimulate proteoglycan production is a relatively specific property of TGF-fl (or a marker of TGF-fl activity); thus, the response of the mesangial cell cultures to the conditioned media can be considered as a bioassay for TGF-fl [51. The temporal pattern of proteoglycan synthesis induced by the conditioned media resembled the proteoglycan production seen in the glomeruli cultures. The increase in proteoglycan synthesis induced by the conditioned media at its peak on day 7 varied between 1200 and 3942%. When the TGF-fll antiserum was added to the conditioned media taken from glomerular cultures on days 4 and 7, it blocked the ability of the conditioned media to stimulate proteoglycan production. Preincubation with the immunizing TGF-fll synthetic peptide abolished the blocking effect of the antiserum on the induction of proteoglycan synthesis by conditioned media from day 7 nephritic glomeruli. Since anti-TGF-fll antibody suppressed proteoglycan and fibronectin production by nephritic glomeruli, we decided to test its effects in vivo [20]. The experimental design was to induce glomerulonephritis in rats with a single injection of antithymocyte serum followed by treatment of the rats with either injection of anti-TGF-fll or control sera. The anti-TGF-fll serum was prepared by immunizing a rabbit with a synthetic peptide containing residues 78-109 from human mature TGF-fll; the cyclized form of this peptide has been shown to elicit an antiserum capable of inhibiting the binding of TGF-fll to cells [32]. The specificity of the antiserum was established by showing it to be capable of neutralizing the activity of purified TGF-fll in a bioassay. The effects of the anti-TGF-fll antiserum and control sera on the glomerulonephritis model were evaluated by using histological evaluation of glomerular extracellular matrix accumulation as a measure of disease activity. Quantitation of the

429

glomerular extracellular matrix demonstrated that the glomeruli contain much less extraceUular matrix material in the anti-TGF-fll-treated groups than in the groups that received control serum. We previously found that the injured glomeruli in culture produced more extraceUular matrix components than did cells from normal glomeruli, and that the synthesis of two proteoglycans, biglycan and decorin, was particularly high [19]. We therefore used the production of proteoglycans as a bioassay that would reflect the activity of TGF-fl and its influence on extracellular matrix synthesis in our disease model. Such an analysis showed that proteoglycan production by glomerular cells was suppressed to a near normal rate by antiTGF-fll [20]. Scanning of the proteoglycan bands indicated that the suppression of this measure of the disease process was about 60% on day 4 and 80% on day 7 following injection, mRNA analysis of the anti-TGF-fll-treated and control glomeruli in the current study revealed equally elevated levels of TGF-fll mRNA in both groups of rats. A similar increase in the number of TGF-fll-positive cells was also seen in the treated and control

animals. These results indicate that anti-TGF-B1 did not interfere with induction of glomerular injury, as reflected by the increased synthesis of TGF-fll mRNA and TGF-fll protein in glomeruli. The cells responsible for the increased TGF-fll expression have not been identified, but they could be proliferating mesangial cells and/or infiltrating monocyte/macrophages. Proteoglycans as natural inhibitors of TGF-fl activity In a separate study it was found that decorin and biglycan can bind TGF-fl and suppress its activity [21]. Since the synthesis of decorin and biglycan is stimulated by TGF-fl, these proteoglycans may constitute a negative regulatory loop in the control of TGF-fl activity.

Conclusion

Our findings establish a central role for TGF-fl in the pathogenesis of acute mesangial proliferative glomerulonephritis (Fig. 1). Whereas the

Glomerular Cell Injury

,1

TGF-g gene expression

Increased expression of matrix receptors on cells

Increased synthesis of proteoglycans and matrix proteins

Decreased expression of proteases and increased expression of protease inhibitors

Accumulation of extracellular matrix Fig. 1.

430

earlier results have provided correlative evidence for a role of TGF-fl in glomerulonephritis [19], the new antibody inhibition data establish that TGF-fl has a direct, causal role, at least in the experimental glomerulonephritis model we have used [20]. The type of TGF-fl responsible for the elevated production of extracellular matrix in our model is likely to be TGF-fll because the antiserum was prepared against a peptide from this TGF-fl. However, the antiserum could also have affected other types of TGF-fl. The mechanism of the TGF-fll action is likely to be the extremely strong stimulation of extracellular matrix synthesis, manifested as a particularly marked stimulation of proteoglycan synthesis by TGF-fll in cultured mesangial cells [5]. On the other hand, the extracellular matrix produced as a result of TGF-fl action, especially the proteoglycan in it, can downregulate the TGF-fl activity [22]. The balance of this regulation appears to be perturbed in glomerulonephritis, as shown by the impressive suppression of matrix accumulation in the experimental disease achieved by anti-TGF-fll. TGF-fl may play a similar role in human glomerular diseases, and perhaps other diseases where fibrosis is a factor, suggesting new approaches to therapy of such diseases.

Acknowledgements Some of this work has been presented as part of a Ciba Foundation Symposium and will be published [Border, W.A. (1990) Role of Transforming Growth Factor-B1 in Experimental Glomerulonephritis. In: Clinical Applications of TGF-fl. John Wiley & Sons Ltd (Ciba Foundation Symposium 157): in press; and was supported by a departmental grant to W.A.B., grants CA 42507 and CA 28896 to E.R. We thank Drs L.W. Fisher, K.C. Flanders and M.B. Sporn for antibodies.

References 1. Ruoslahti, E. and M.O. Pierschbacher: New perspectives in cell adhesion. RGD and integrins. Science 238, 491-497, (1987).

2. Klahr, S., G. Schreiner and I. Ichikawa: The progression of renal disease. N. Engl. J. Med. 318, 1657-1666 (1988). 3. Kanwar, Y.S: Biology of disease. Biophysiology of glomerular filtration and proteinuria. Lab. Invest. 51, 7-21(1984). 4. Border, W.A.: Distinguishing minimal change disease from mesangial disorders. Kidney Int. 34, 419-434 (1988). 5. Border, W.A., S. Okuda, L.R. Languino and E. Ruoslahti: Transforming growth factor-fl regulates production of proteoglycans by mesangial cells. Kidney Int. 37, 689-695 (1988). 6. Nakamura, T., S. Okuda, D. Miller, E. Ruoslahti and W. Border: Transforming growth factor-fl (TGF-fl) regulates production of extracellular matrix (ECM) components by glomerular epithelial cells. Kidney Int. 37, 221, (1990). 7. Massague, J.: The TGF-fl family of growth and differentiation factors. Cell 49, 347-438 (1987). 8. Roberts, A.B., K.C. Flanders, P. Kondaiah Pet al: Transforming growth factor fl: biochemistry and roles in embryogenesis, tissue repair and remodeling, and carcinogenesis. Recent Prog. Horm. Res. 44, 157-197 (1988). 9. Bassols, A. and J. Massague: Transforming growth factor ft regulates the expression and structure of extraceihilar matrix chondroitin/dermatan sulfate proteoglycans. J. Biol. Chem. 263, 3039-3045 (1988). 10. Ignotz, R.A. and J. Massague: Transforming growth factor beta stimulates the expression of fibronectin and collagen and their incorporation into the extracellular matrix. J. Biol. Chem. 261, 4337-4345 (1986). 11. Roberts, A.B., M.B. Spore, R.K. Assoian et al. Transforming growth factor type-ft. Rapid induction of fibrosis and angiogenesis in vivo and stimulation of collagen formation in vitro. Proc. Natl. Acad. Sci. USA 83, 4167-4171 (1986). 12. Morales, T.I. and A.B. Roberts: Transforming growth factor ft regulates the metabolism of proteoglycans in bovine cartilage organ cultures. J. Biol. Chem. 263, 12828-12831 (1988). 13. Edwards, D.R., G. Murphy, J.J. Reynolds et al: Transforming growth factor beta modulates the expression of collagenase and metalloproteinase inhibitor. EMBO J. 6, 1899-1904 (1987). 14. Laiho, M., O. Saksela and J. Keski-Oja: Transforming growth factor-ft induction of type I plasminogen activator inhibitor. J. Biol. Chem. 262, 17467-17474 (1987). 15. Ignotz, R.A. and J. Massague: Cell adhesion protein receptors as targets for transforming growth factor beta action. Cell 51, 189-197 (1987). 16. Roberts, A.B., M.A. Anzano, C.A. Meyers et al: Purification and properties of a Type fl transforming growth factor from bovine kidney. Biochem. 22:5692-5698 (1983). 17. Frolik, C.A., L.M. Wakefield, D.M. Smith and M.B. Sporn: Characterization of a membrane receptor for transforming growth factor-ft in normal rat kidney fibroblasts. J. Biol. Chem. 259, 10995-11000 (1984). 18. MacKay, K., L.J. Striker, J.W. Tauffer, T. Doi, L.Y. Agodoa and G.E. Striker: Transforming growth factor-ft. Murine glomerular receptors and responses of isolated glomerular cells. J. Clin. Invest. 83, 1160-1167 (1989).

431 19. Okuda, S., L.R. Languino, E. Ruoslahti and W.A. Border: Elevated expression of transforming growth factor-fl and proteoglycan production in experimental glomerulonephritis. Possible role in expansion of the mesangial extracellular matrix. J. Clin. Invest. 86, 453-462 (1990). 20. Border, W.A., S. Okuda, L.R. Languino, M.B. Sporn and E. Ruoslahti. An antiserum against transforming growth factor fll suppresses experimental glomerulonephritis. Nature, 346, 371-374 (1990). 21. Yamaguchi, Y., D.M. Mann and E. Ruoslahti: Transforming growth factor fl is negatively regulated by a proteoglycan. Nature 346, 281-284 (1990). 22. Johnson, R.J., R.L. Garcia, P. Pritzl and C.E. Alpers: Platelets mediate glomerular cell proliferation in immune complex nephritis induced by antimesangial cell antibodies in the rat. Am. J. Pathol. 136, 369-374 (1990). 23. Werber, H.I., S.N. Emancipator, M.L. Tykocinski and J.R. Sedor: The interleukin-I gene is expressed by rat glomerular mesangial cells and is augmented in immune complex glomerulonephritis. J. Immunol. 138, 3207-3212 (1987). 24. Boswell, J.M., M.A. Yui, D.W. Burt and V.E. Kelley: Increased tumor necrosis factor and IL-I beta gene expression in the kidneys of mice with lupus nephritis. J. Immunol. 141, 3050-3054 (1988). 25. Abrahamson, D.R.: Structure and development of the glomerular capillary wall and basement membrane. Amer. J. Physiol. 253 (Ren. Fluid Electrolyte Physiol. 22), F783F794 (1987).

26. Fisher, L.W., J.D. Termine and M.F. Young: Deducedprotein sequence of bone small proteoglycan I (biglycan) shows homology with proteoglycan II (decorin) and several nonconnective tissues in a variety of species. J. Biol. Chem. 264, 4571-4577 (1989). 27. Krusius, T. and E. Ruoslahti: Primary structure of an extracellular matrix proteoglycan core protein deduced from cloned cDNA. Proc. Natl. Acad. Sci. USA 83, 76837687 (1986). 28. Yamaguchi, Y. and E. Ruoslahti: Expression of human proteoglycan in Chinese hamster ovary cells inhibits cell proliferation. Nature 336, 244-246 (1988). 29. Yamamoto, T. and C.B. Wilson: Quantitative and qualitative studies of antibody induced mesangial cell damage in the rat. Kidney Int. 32, 514-525 (1987). 30. Bagchus, W.M., Ph.J. Hoedemaeker, J. Rozing and W.W. Bakker. Glomerulonephritis induced by monoclonal antithy I.I antibodies. Lab. Invest. 55, 680-687 (1986). 31. Flanders, K.C., N.L. Thompson, D.S. Cissel et al: Transforming growth factor-r: histochemicai localization with antibodies to different epitopes. J. Cell. Biol. 108, 653-660 (1989). 32. Flanders, K.C., A.B. Roberts, N. Ling, B.E. Fleurdleye and M.B. Sporn: Antibodies to peptide determinants in transforming growth factor-fl and their applications. Biochemistry 27:739-746 (1988).