TGF-β1 as an Endogenous Defender Against Macrophage-Triggered Stromelysin Gene Expression in the Glomerulus¹

Transforming growth factor-β is a pleiotropic cytokine involved in a wide range of cell functions (1). The expression of TGF-β is observed in various embryonic tissues and adult organs, and physiologic levels of TGF-β are supposedly necessary for tissue repair and the maintenance of organ functions (2, 3). On the other hand, overexpression of TGF-β is closely linked to certain diseases, including tissue fibrosis (4). In the kidney, a body of literature has reported on the “dark side” of TGF-β, i.e., its contribution to renal fibrogenesis (5). In glomerulonephritis, for example, up-regulated TGF-β may induce the accumulation of extracellular matrices via stimulation of matrix production, inhibition of matrix-degrading proteinases, and up-regulation of proteinase inhibitors. However, little attention is currently being paid to the “bright side” of this molecule in glomerular pathophysiology.

TGF-β is known to be a potent regulator of immune systems and inflammatory processes, and generally functions as an immunosuppressor (6). This molecule represses B cell proliferation and Ig secretion, mitogenesis and cytokine production by thymocytes/T lymphocytes, and function of NK cells. TGF-β inhibits neutrophil and T cell adhesion to the endothelium (7) and strongly deactivates macrophages (Mφ)³ at picomolar concentrations (8). Based on the fact that Mφ play a crucial role in the generation of various glomerular diseases, TGF-β may function as a potential inhibitor of Mφ-mediated glomerular injury (9). Recent investigations have provided evidence regarding the beneficial action of TGF-β1 in the glomerulus. For example, mesangial cells produce biologically active TGF-β1 that inhibits the adhesion of Mφ on various substrata and leads to their subsequent deactivation (10). Mesangial cell-derived TGF-β1 depresses Mφ production of injurious inflammatory mediators, including IL-1β, IL-6, TNF-α, and monocyte chemoattractant protein-1 (11, 12). These data indicate that TGF-β may inhibit the activation of glomerular cells via suppression of local Mφ function.

Stromelysin/transin is a matrix-degrading metalloproteinase that plays an important role in the metabolism of the extracellular matrix. Tightly controlled expression of this enzyme supposedly required for organogenesis and wound healing, and its uncontrolled expression is associated with certain pathologies, including tumor invasion/metastasis, inflammation, and atherosclerosis (13, 14, 15). Stromelysin is a multipotent matrix degrader; i.e., it degrades collagens, glycoproteins, and proteoglycans (16), which are all components of the glomerular basement membrane and the mesangial matrix. Overexpression of stromelysin induces an aberrant breakdown of the glomerular matrix and thereby contributes to its structural and functional alteration (17, 18). Mesangial cells have the ability to produce stromelysin abundantly in response to the Mφ-derived proinflammatory cytokine, IL-1β (19). When stromelysin was overexpressed via genetic manipulation, mesangial cells exhibited an altered phenotype: enhanced mitogenesis and migration (20). Furthermore, when activated Mφ were transferred into normal rat glomeruli, stromelysin was substantially induced in resident glomerular cells (21). These data suggest a pathologic implication of stromelysin in Mφ-associated glomerular injury.

Using stromelysin as an indicator molecule, the present investigation examines whether and how endogenous TGF-β1 modulates the glomerular cell activation triggered by Mφ. This study provides evidence that mesangial cell-derived TGF-β1 is a glomerular “defender” that attenuates Mφ-mediated activation of resident cells.

Rat mesangial cells were cultured from the isolated glomeruli of an adult male Sprague-Dawley rat (22) and identified as being of mesangial cell phenotype (23). The normal alveolar Mφ NR8383 (24), derived from a Sprague-Dawley rat, and the mink lung epithelial cell line CCL64 (25), were purchased from the American Type Culture Collection (Rockville, MD). These cells were maintained in DMEM/Ham’s F-12 (DMEM-F12) (Life Technologies, Gaithersburg, MD) supplemented with 10 to 15% FCS.

The mesangial cell clone MTG6 that stably expresses the active form of porcine TGF-β1 was established using the calcium phosphate coprecipitation method (26). This clone produces and secretes high levels of biologically active TGF-β1 (26). Mock transfectants were established by transfecting mesangial cells with an empty plasmid, pHβAPr-1-neo (27). To abrogate function of the TGF-βR type II (RII), mesangial cells were cotransfected with pRIIDN (a gift from Dr. R. Derynck, University of California, San Francisco; 28 , which codes for a dominant-negative mutant of TGF-β RII and pRc/CMV (Invitrogen, Carlsbad, CA) that introduces a neomycin phosphotransferase gene. Overexpression of the mutant receptor gene leads to impaired TGF-β signaling in a dominant-negative fashion, as demonstrated previously (28). Stable transfectants were selected in the presence of neomycin analogue G418 (750 μg/ml), and clones MRIIDN4 and MRIIDN8 were established. Untransfected cells and mock-transfected cells were used as controls.

Reporter Mφ BAGMAC^NR were generated as described previously (21). These cells express a bacterial β-galactosidase gene that is under the control of the long terminal repeat of the Moloney murine leukemia virus. BAGMAC^NR cells also express a neomycin phosphotransferase gene that renders the cells resistant to G418. In contrast to BAGMAC^NR cells, resident glomerular cells are susceptible to G418. This difference allows for the selective subtraction of resident cell mRNAs from the total mRNAs expressed in BAGMAC^NR-containing, chimeric glomeruli (neomycin subtraction method) (21).

Mφ-conditioned media (MφCM) were prepared as follows. NR8383 cells cultured in 1% FCS/DMEM-F12 were stimulated with or without 1 μg/ml of LPS (Escherichia coli 0111:B4; Sigma, St. Louis, MO) for 6 h. Cells were washed three times and reseeded in fresh DMEM-F12 containing 1% FCS. After 24 h, the media were collected and filtered. MφCM diluted at 1:1 (50%) was used for experiments.

MφCM were prepared from isolated, normal and nephritic glomeruli. To create the latter, anti-Thy-1 glomerulonephritis was induced in adult male Sprague-Dawley rats (four rats with a body weight of 250–350 g) (23) by a mAb, 1-22-3 (29). Media conditioned by normal or nephritic (day 7) glomeruli were prepared by incubating isolated glomeruli (1 × 10⁴) in 1 ml of 0.5% FCS/DMEM-F12 for 24 h. MφCM diluted at 1:4 (20%) was used for the assessment of TGF-β bioactivity.

A growth inhibition assay of the mink lung epithelial cell line CCL64 was used to measure the bioactivity of TGF-β (26). In brief, CCL64 cells were seeded in 24-well plates at a density of 5 × 10³ cells/well. After incubation for 24 h, cells were further incubated in 1% FCS/DMEM-F12 for 24 h and then exposed to diluted test media. After 48 h, [³H]thymidine (1 μCi/well) was added and incubated for 3 h, and incorporation of radioactivity was subsequently measured by liquid scintillation counting. Human TGF-β1 (Genzyme, Cambridge, MA) was used as a standard. Assays were performed in quadruplicate.

The expression of stromelysin was examined by Northern blot analysis. Mesangial cells or isolated glomeruli (∼1 × 10⁴) were exposed to 5 to 10 ng/ml of human rIL-1β (a gift from Otsuka Pharmaceutical, Tokushima, Japan) or 1:1-diluted MφCM for 18 h and subjected to analyses. To examine the effect of TGF-β, isolated, normal glomeruli were pretreated with TGF-β1 (20 ng/ml) for 6 h and then stimulated with MφCM. After 18 h, glomeruli were harvested and used for RNA extraction.

Northern blot analysis was performed as previously described (20). In brief, total RNA was extracted by a single-step method (30), and RNA samples were electrophoresed on 1.2% agarose gels and transferred onto nitrocellulose membranes. For hybridization, a truncated human TGF-β RII cDNA (28) and a rat stromelysin-1 cDNA (31) were labeled with [³²P]deoxy-CTP using the random priming method. As loading controls, a chicken β-actin cDNA and a rat glyceraldehyde-3-phosphate dehydrogenase cDNA were used. The membranes were hybridized with probes at 65°C overnight in a solution containing 4× SSC (600 mM sodium chloride, 60 mM sodium citrate), 5× Denhardt’s solution, 10% dextran sulfate, 50 μg/ml herring sperm DNA, and 50 μg/ml poly(A). These membranes were subsequently washed at 50°C and exposed to Kodak XAR films at −80°C. Densitometric analysis was performed using a computerized program, the National Institutes of Health Image program. The intensity of each message normalized by the density level of β-actin mRNA was used to compare data in different groups.

To investigate the role of Mφ in the expression of stromelysin in normal and nephritic glomeruli, a technique for in vivo Mφ transfer was used (21). Via the renal circulation, BAGMAC^NR cells that had been stimulated by 1 μg/ml LPS for 16 h (1 × 10⁶ cells) were transferred into normal rat glomeruli or into glomeruli that were in the regeneration phase of acute anti-Thy-1 glomerulonephritis (day 7), in which TGF-β1 is up-regulated (a total of 16 Sprague-Dawley rats with a body weight of 250–350 g). Immediately after the cell injection, both kidneys were removed and processed for glomerular isolation. One-third (0.5–1 × 10⁴) of the glomeruli isolated from each kidney were immediately frozen at −80°C. The remaining glomeruli were incubated for 24 h in 1% FCS/DMEM-F12 with or without G418 (200 μg/ml) and then stored at −80°C. The induction of stromelysin was examined by Northern blot analysis. In this experimental setting, mRNA expression in resident glomerular cells, but not in BAGMAC^NR cells, is selectively abolished by G418 (21).

To evaluate cell transfer efficiency, isolated glomeruli containing BAG MAC^NR cells were subjected to 5-bromo-4-chloro-3-indolyl β-d-galactopyranoside (X-gal) assay (23, 32). In brief, isolated glomeruli were fixed in 0.5% glutaraldehyde, 2 mM MgCl₂, and 1.25 mM EGTA in PBS at 4°C overnight and then incubated at 37°C for 1 to 2 h in a reaction buffer containing 1 mg/ml X-gal (Sigma), 20 mM K₃Fe(CN)₆, 20 mM K₄Fe(CN)₆ · 3H₂O, 2 mM MgCl₂, 0.01% sodium desoxycholate, and 0.02% Nonidet P-40 in PBS. More than 100 glomeruli were randomly selected, and percentages of X-gal-positive glomeruli were determined by light microscopy.

All experiments were repeated two to four times, and representative results are shown. Data are expressed as means ± SE. Statistical analyses were performed using the nonparametric Mann-Whitney U test. A p value of <0.05 was used to indicate a statistically significant difference.

Cultured mesangial cells have the ability to produce biologically active TGF-β1 (10, 11, 33). To investigate whether the overexpression of TGF-β modulates the responsiveness of mesangial cells to Mφ-derived cytokines, rat mesangial cells were stably transfected with a cDNA encoding the active form of TGF-β1. The established MTG6 cells expressed a transgene transcript, produced higher levels of active TGF-β, and exhibited lower mitogenic activity than mock-transfected cells (26). Using these transfectants, the expression of stromelysin in response to IL-1β was examined. Northern blot analysis revealed that IL-1β markedly induced stromelysin gene expression in mock-transfected cells. In contrast, this induction was dramatically depressed in MTG6 cells overexpressing active TGF-β1 (Fig. 1).

To examine whether endogenous TGF-β acts as an autocrine inhibitor of the cytokine response, mesangial cells were stably transfected with a cDNA coding for a dominant-interfering form of the TGF-β RII. When overexpressed, this mutant receptor effectively blocks TGF-β signaling (28). As shown in Figure 2,A, the established MRIIDN4 and MRIIDN8 cells, but not untransfected or mock-transfected cells, expressed abundant transgene mRNAs. Using these clones, the induction of stromelysin by IL-1β was examined. Compared with untransfected cells and mock transfectants, the expression of stromelysin was substantially enhanced in both mutant cells (Fig. 2 B).

Cultured mesangial cells express stromelysin in response to Mφ-derived cytokines (19). However, it is undetermined whether glomeruli that are stimulated by cytokines express this metalloproteinase. To examine this possibility, isolated, normal rat glomeruli were stimulated with either IL-1β or medium conditioned by activated Mφ, and the expression of stromelysin was examined by Northern blot analysis. A modest expression of stromelysin was observed in unstimulated glomeruli. This expression was substantially up-regulated when the glomeruli were stimulated by either IL-1β or MφCM (Fig. 3 A).

The effect of TGF-β on the glomerular expression of stromelysin was examined. Isolated, normal glomeruli were pretreated with or without TGF-β1 and then stimulated by MφCM. Northern blot analysis revealed that TGF-β1 inhibited the induction of stromelysin in the glomerulus (Fig. 3 B).

To examine whether endogenous TGF-β1 is able to inhibit the expression of stromelysin in the glomerulus, an acute model of anti-Thy-1 glomerulonephritis was used. In this model, TGF-β1 is up-regulated in activated mesangial cells during the regeneration of the glomerulus (34). Glomeruli were isolated from normal rat kidneys or kidneys that were subjected to the anti-Thy-1 glomerulonephritis (day 7). Compared with normal glomeruli, the regenerating glomeruli produced higher levels of biologically active TGF-β, as evidenced by the inhibition of [³H]thymidine incorporation by CCL64 cells (7.5 ± 12.3% inhibition (not significant) in normal glomeruli and 34.5 ± 6.7% inhibition (p < 0.05) in nephritic glomeruli). Isolated glomeruli were subsequently stimulated by IL-1β or MφCM, and the expression of stromelysin was examined by Northern blot analysis. Consistent with the effect of externally added TGF-β1, the induction of stromelysin was dramatically suppressed in the glomeruli producing active TGF-β1 (Fig. 3 C).

Using a technique for in vivo Mφ transfer, we previously reported that activated Mφ induce stromelysin expression in the normal glomerulus (21). To examine whether the Mφ-triggered stromelysin expression is inhibited by endogenous TGF-β1, BAGMAC^NR Mφ stimulated with LPS were transferred into normal rat glomeruli or into glomeruli that had been subjected to acute anti-Thy-1 glomerulonephritis (day 7). After the cell transfer, glomeruli were isolated and used for X-gal assay and Northern blot analysis. X-gal assay on isolated glomeruli showed that there was no significant difference in the percentages of X-gal-positive glomeruli between normal (73.2 ± 3.4%, n = 10) and nephritic conditions (71.0 ± 2.9%, n = 6). The expression of stromelysin in isolated chimeric glomeruli was subsequently examined by Northern blot analysis (Fig. 4). Immediately after the cell transfer, the stromelysin transcript was not detectable in both normal and inflamed glomeruli. After ex vivo incubation of these chimeric glomeruli, stromelysin mRNA was induced in Mφ-transferred, normal glomeruli. This induction was abolished by the treatment of glomeruli with G418, indicating that resident cells, but not transferred Mφ, are responsible for the expression of stromelysin. In the regenerating glomeruli that express TGF-β1, the stromelysin expression triggered by Mφ was markedly suppressed (19.1 ± 2.4%, n = 3) compared with its induction in normal glomeruli (100%).

IL-1 is one of the pivotal pathogenic mediators elaborated by activated Mφ. IL-1 stimulates resident glomerular cells to cell proliferation, aberrant matrix metabolism, and the release of inflammatory mediators (35). The injection of IL-1 exacerbates, but the blocking of IL-1 activity attenuates, experimental glomerulonephritis, suggesting its crucial pathologic contribution (36). IL-1 induces metalloproteinase stromelysin in glomerular cells (19). Using this molecule as an indicator, the present study investigated how endogenous TGF-β modulates the stimulatory effect of Mφ-derived cytokines on the glomerulus. In both mesangial cells and isolated, normal glomeruli, externally supplied or internally produced TGF-β inhibited the expression of stromelysin in response to Mφ-derived factors, including IL-1β. Consistent with these results, the glomeruli producing the active form of TGF-β1 exhibited resistance against the Mφ-triggered stromelysin induction. This effect may be due to the suppressive effect of TGF-β1 on the production of cytokines by activated Mφ (10, 11). Another explanation could be that TGF-β inhibited the cytokine response of glomerular cells via intervention in certain signaling pathways (26). In support of this possibility, transfection experiments showed that the dominant-negative inhibition of TGF-β signaling enhanced the IL-1 response of mesangial cells.

The biologic actions of TGF-β are mediated by specific cell surface receptors, TGF-βR type I (RI) and RII. TGF-β1 initially binds to RII that has a constitutively active kinase. RI is then recruited to the complex and phosphorylated by RII, leading to the propagation of the signal to downstream substrates (37). Based on this current dogma, both RI and RII are required for TGF-β signaling, and the functional inhibition of either RI or RII should abrogate the biologic actions of TGF-β. However, some early reports suggested separate signaling pathways for the antiproliferative and matrix-regulatory effects of TGF-β. For example, the functional inactivation of RII did inhibit the TGF-β regulation of mitogenesis, but it did not affect the stimulatory effects of TGF-β on the synthesis of fibronectin and plasminogen activator inhibitor I (28). It has been suggested that RII is essential for the control of mitogenesis by TGF-β but might be not necessary for its regulation of the genes associated with matrix turnover. The present study showed that the dominant-negative inactivation of RII abrogated the effect of TGF-β on the expression of stromelysin in mesangial cells. Therefore, RII is required in certain cell types for the TGF-β regulation of genes involved in matrix turnover. This finding is consistent with a recent report using primary hepatocytes derived from dominant-negative RII-transgenic mice (38).

The infiltration of monocytes/Mφ is a common pathologic feature of a wide range of human and experimental glomerular diseases (39). Mφ infiltration is correlated with both structural and functional changes of the glomerulus, and depletion of Mφ attenuates glomerular damage and proteinuria in several forms of experimental glomerulonephritis (39). However, in the process of acute inflammation, the Mφ-initiated pathologic change is self-limiting. Endogenous, antiinflammatory machinery may exist to turn off the inflammatory process toward resolution. The TGF-β1 produced by activated mesangial cells is a possible molecular switch (9). Together with the fact that TGF-β1 is an autocrine inhibitor of mesangial cell proliferation (26, 33), the present results further support the possibility that TGF-β1 is a glomerular defender that attenuates Mφ-mediated glomerular cell activation.

Several cytokines are regarded as deactivators of Mφ. Those include IL-4, IL-10, IL-13, and TGF-β (40). In the present study, the role of TGF-β was highlighted, but other cytokines could also function against the inflammatory process initiated by Mφ. Recent investigations have shown that, under pathologic circumstances, glomeruli express IL-4, IL-10, and IL-13 (41, 42, 43, 44). When these cytokines are administered into animals that have been subjected to experimental glomerulonephritis, histopathologic changes are attenuated (45). Although these cytokines are regarded as “Th2 factors” released by Th cells, recent reports have disclosed the ability of resident glomerular cells, especially mesangial cells, to produce IL-4 and IL-13 (44, 46). At inflammatory sites, the cooperation of these Mφ deactivators released from mesangial cells may play a role in the prevention of or recovery from glomerulonephritis.

The role of TGF-β in glomerular pathophysiology is still undefined. Normally, constitutive expression of TGF-β1 is observed in the glomerulus (34, 47, 48, 49). This expression is up-regulated in a wide range of glomerular diseases, especially in glomerulonephritis, where mesangial cells are activated (5). Under these pathologic conditions, TGF-β1 stimulates matrix production by glomerular cells and is thereby postulated to be a molecule responsible for the generation of glomerulosclerosis (5). In addition to this effect, however, TGF-β1 is potentially antiinflammatory through its suppressive effects on Mφ, the “blackguards” in glomerular injury (9). The present results reinforce the idea of a beneficial potential of TGF-β1 in Mφ-mediated, early glomerular inflammation.

I thank Dr. F. Shimizu for generous support and Otsuka Pharmaceutical Co. for the human rIL-1β.