Mechanical Stretch-Induced Fibronectin and Transforming Growth ...

Mechanical Stretch-Induced Fibronectin and Transforming Growth Factor- 1 Production in Human Mesangial Cells Is p38 Mitogen-Activated Protein Kinase–Dependent Gabriella Gruden, Silvia Zonca, Anthea Hayward, Stephen Thomas, Sabrina Maestrini, Luigi Gnudi, and GianCarlo Viberti

Hemodynamic abnormalities are important in the pathogenesis of the excess mesangial matrix deposition of diabetic and other glomerulopathies. p38-Mitogen-activated protein (MAP) kinase, an important intracellular signaling molecule, is activated in the glomeruli of diabetic rats. We studied, in human mesangial cells, the effect of stretch on p38 MAP kinase activation and the role of p38 MAP kinase in stretchinduced fibronectin and transforming growth factor- 1 (TGF- 1) accumulation. p38 MAP kinase was activated by stretch in a rapid (11-fold increase at 30 min, P < 0.001) and sustained manner (3-fold increase at 33 h, P < 0.001); this activation was mediated by protein kinase C (PKC). Stretch-induced fibronectin and TGF- 1 protein levels were completely abolished (100% inhibition, P < 0.001; and 92% inhibition, P < 0.01, respectively) by SB203580, a specific p38 MAP kinase inhibitor. At 33 h, TGF- 1 blockade did not affect stretch-induced fibronectin production, but partially prevented stretch-induced p38 MAP kinase activation (59% inhibition, P < 0.05). TGF- 1 induced fibronectin accumulation after 72 h of exposure via a p38 MAP kinase–dependent mechanism (30% increase over control, P < 0.01). In human mesangial cells, stretch activates, via a PKC-dependent mechanism, p38 MAP kinase, which independently induces TGF- 1 and fibronectin. In turn, TGF- 1 contributes to maintaining late p38 MAP kinase activation, which perpetuates fibronectin accumulation. Diabetes 49:655–661, 2000

From the Unit for Metabolic Medicine, Department of Endocrinology, Diabetes and Internal Medicine, Division of Medicine, Guy’s, King’s, and St Thomas’ School of Medicine, King’s College London, London, U.K. Address correspondence and reprint requests to GianCarlo Viberti, MD, FRCP, Department of Endocrinology, Diabetes and Internal Medicine, Division of Medicine, Fifth Floor, Thomas Guy House, Guy’s Hospital, London SE1 9RT, U.K. E-mail: [email protected]. Received for publication 11 June 1999 and accepted in revised form 22 December 1999. C V, coefficient of variation; ELISA, enzyme-linked immunosorbent assay; FCS, fetal calf serum; HRP, horseradish peroxidase; JNK, c-Jun NH2terminal kinase; MAP, mitogen-activated protein; PKC, protein kinase C; P M S F, phenylmethylsulfonyl fluoride; rhTGF- 1, human recombinant transforming growth factor- 1; TGF- 1, transforming growth factor- 1; TRE, 12-O-tetradecanoylphorbol 13-acetate (TPA)-response element. DIABETES, VOL. 49, APRIL 2000

A

lteration in intraglomerular hemodynamics is a common feature of several forms of renal disease, including diabetic nephropathy, and the mesangium is a primary target of injury in these conditions (1,2). Treatments that lower glomerular capillary pressure have a potent renoprotective effect (3–5). Mesangial cells, via their cellular projections, are directly connected with the glomerular capillary basement membrane and are thus affected by the mechanical insult of an increased glomerular capillary pressure. The application of mechanical stretch to rat mesangial cells results in an increased production of extracellular matrix and transforming growth factor- 1 (TGF- 1), a potent prosclerotic cytokine (6,7). p38 Mitogen-activated protein (MAP) kinase, a member of the MAP kinase family, is activated in response to various extracellular stresses, such as hyperosmolar and oxidative stress, and by several cytokines (8–10), including TGF- 1 (11). p38 MAP kinase is activated in the glomeruli of diabetic rats (10) and induces in various cell types transcription factors (12–14), which can enhance extracellular matrix gene expression (15–18). MAP kinases are induced by stretch in rat mesangial cells (19), and Erk1 and Erk2 have been implicated in matrix production (20), but there is no information to date on the role of p38 MAP kinase in stretch-induced TGF- 1 and matrix production in human mesangial cells. The present study was designed to test in human mesangial cells whether p38 MAP kinase is implicated in stretchinduced fibronectin and TGF- 1 production and to examine the interaction between p38 MAP kinase and TGF- 1 in relation to extracellular matrix production. RESEARCH DESIGN AND METHODS Materials. All materials were purchased from Sigma (St. Louis, MO), unless otherwise stated. Fetal calf serum (FCS) was obtained from Gibco-BRL (Paisley, U.K.) and Flex I and Flex II plates from Flexcell International (McKeesport, PA). Pan–antiTGF- , anti–TGF- 1 neutralizing antibodies, and human recombinant TGF- 1 (rhTGF- 1) were obtained from R & D Systems (Minneapolis, MN). TGF- 1 enzymelinked immunosorbent assay (ELISA) and anti–phospho-p38 MAP kinase antibody were obtained from Promega (Southampton, U.K.). The p38 MAP kinase inhibitor SB203580 [4-(4-fluorophenyl)-2-(4-methylsulfinynphenyl)-5-(4-pyridyl)] imidazole and calphostin C were from Calbiochem (Nottingham, U.K.), and the anti–p38 MAP kinase antibody from Insight Biotechnology (Wembley, U.K.). The rabbit anti-fibronectin antibody was from Dako (Cambridge, U.K.) and the in vitro p38 MAP kinase activity assay from New England Biolabs (Hitchin, U.K.). 655

STRETCH AND p38 MAPKINASE IN MESANGIALCELLS

Cell culture. Human mesangial cells were isolated as previously described (21). Briefly, normal renal cortex was taken from the opposite tumor-free pole of nephrectomy specimens. Nephrectomy was performed for localized, capsulated grade 1 hypernephromas. Tissue was analyzed by light microscopy and immunofluorescence to exclude the presence of tumor cells and glomerular abnormalities (22). Intact glomeruli were collected by serial sieving of cortical homogenates. Cells obtained from three separate kidneys were used in this study. After digestion with collagenase (type IV, 750 U/ml), the isolated glomeruli were seeded in culture flasks. After the outgrowth of mesangial cells, the glomeruli were removed by washing with phosphate-buffered saline (PBS) and the cells were cultured in RPMI 1640 medium, supplemented with insulin-transferrin-selenium and L-glutamine and containing 20% FCS, 7 mmol/l glucose, 100 IU/ml penicillin, and 100 mg/ml streptomycin in a humidified 5% CO2 incubator at 37°C. Mesangial cells were harvested using 0.25% trypsin and 0.5% EDTA. The cells were stellate or fusiform in appearance, grew in multilayers, formed hillocks in long-term culture, and stained for -smooth muscle actin by direct immunofluorescence. Cells did not stain for cytokeratin, factor VIII, common leukocyte antigen (Dako, High Wycombe, U.K.), and Thy-1 (Serotec, Oxford, U.K.), respectively excluding contamination of epithelial cells, endothelial cells, lymphomonocytes, and human fibroblasts (23). Studies were performed between passages 4 and 7, while the cells retain their peculiar morphologic and immunofluorescent phenotypic features. Application of mechanical stretch to cultured cells. Mesangial cells were seeded in equal numbers (12,000/cm2) into six-well type I collagen–coated silicon elastomer-base culture plates (Flex I plates) and control plates (Flex II plates). After 3–5 days, they were serum deprived and incubated in insulin-free medium for 48 h, and then subjected to repeated stretch/relaxation cycles by mechanical deformation using a stress unit. The stress unit is a modification of the unit initially described by Banes et al. (24) and consists of a vacuum unit and a baseplate. A vacuum was cyclically applied (60 cycles/min) to the rubber-base plates via the baseplate, which was placed in a humidified incubator with 5% CO2 at 37°C. Cells were exposed to an average 10% uniaxial elongation, which mimics that present in vivo in glomeruli exposed to supernormal pressure levels and is known in vitro to induce mesangial cell cytokines and matrix production (6,7,25). In preliminary experiments, we found that an average cell elongation of 4%, believed to correspond to the physiological stretch induced by a normal intracapillary pressure (26), had no effect on p38 MAP kinase activity or TGF- 1 and fibronectin production (data not shown). Stretch and control experiments were carried out simultaneously with cells derived from a single pool. Control cells were grown in nondeformable but otherwise identical plates (Flex II plates) in parallel. Mesangial cell exposure to hyperosmolar stress. Serum- and insulindeprived human mesangial cells were exposed for 20 min to either iso-osmolar (290 mosm/l) or hyperosmolar (440 mosml/l) media. Hyperosmolar medium was obtained by supplementation with 150 mmol/l NaCl. Osmolarity was confirmed by a cryoscopic osmometer (Osmomat 30; Gonotec, Berlin). Cell number determination. Cells were harvested with 0.25% trypsin and 0.5% EDTA and the cell number determined by a Coulter Cell Counter (Coulter Electronics, Luton Beds, U.K.). Fibronectin and TGF- 1 protein measurement. Culture supernatants from all experimental conditions were collected, centrifuged to remove cell debris, and stored at –70°C for analysis. For each experiment, both fibronectin and TGF- 1 protein levels were determined within a single-assay run. After activation of latent TGF- 1 by acidification, total TGF- 1 protein concentration was measured by ELISA (range 16–1,000 pg/ml; intra-assay coefficient of variation [CV], 1.6%) using a mouse monoclonal and a rabbit polyclonal anti–human TGF- 1. Fibronectin protein levels were measured with a modification of the competitive inhibition ELISA described by Rennard et al. (27) (range 0.05–0.5 µg/ml; intra-assay CV, 5%). Briefly, 96-well cluster plates were coated overnight at 4°C with 100 ng/well of human fibronectin. Samples were incubated overnight at 4oC with rabbit–anti-human fibronectin as primary antibody. After washing, samples and standards were added to the plates and incubated for 1 h at 37°C. Immunocomplexes were detected by horseradish peroxidase (HRP)-conjugated goat–anti-rabbit IgG and revealed by 3,3 ,5,5 -tetramethylbenzidine dihydrochloride substrate. The reaction was stopped with H2SO4, and the absorbance measured at 450 nm. Measurements of pericellular polymeric fibronectin. Levels of pericellular polymeric fibronectin were determined in deoxycolate-insoluble and SDS-soluble protein extracts as previously described (28). Cells were lysed in 2% deoxycolate, 0.02 mol/l Tris-HCl (pH 8.8), 2 mmol/l phenylmethylsulfonyl fluoride (PMSF), 2 mmol/l EDTA, 2 mmol/l iodoacetic acid, and 2 mmol/l N-ethylmaleimide. The cell lysates were centrifuged at 10,000 rpm for 20 min to spin out deoxycolate-insoluble material, which was then dissolved in 1% SDS, 0.02 mol/l Tris-HCl (pH 8.8), 2 mmol/l PMSF, 2 mmol/l EDTA, 2 mmol/l iodoacetic acid, and 2 mmol/l N-ethyl-maleimide. Deoxycolate-insoluble and SDS-soluble extracts were separated by 5% SDS-PAGE and electrotransferred to nitrocellulose membranes. The membranes were blocked in 5% skimmed milk in Tris bufferTween-20 (pH 7.6), then incubated with a rabbit anti-fibronectin antibody. Membranes were then washed in Tris buffer-Tween-20 and incubated with HRP-linked 656

secondary antibodies. Detection was performed by enhanced chemiluminescence and the band intensity quantitated by laser densitometry. Measurement of active p38 MAP kinase levels and p38 MAP kinase activity. Cells were harvested in a Tris-HCl (50 mmol/l, pH 7.6) lysis buffer containing 0.01 mmol/l sodium pyrophosphate, 100 mmol/l NaF, 20 mmol/l EDTA, 20 mmol/l EGTA, 1 mmol/l Na3VO4, 1 mmol/l PMSF, 1 mg/ml leupeptin, 1 mg/ml aprotinin, and 1% Triton X-100. To determine total and phospho-p38 MAP kinase levels, total protein extracts were separated by 10% SDS-PAGE and electrotransferred to nitrocellulose membranes. The membranes were blocked in 5% skimmed milk in Tris buffer-Tween-20 (pH 7.6), then incubated with either a goat anti–p38-MAP kinase antibody, detecting total p38 MAP kinase, or with a rabbit anti–phospho-p38-MAP kinase antibody, binding only to the active form of p38 MAP kinase. Membranes were then washed in Tris buffer-Tween-20 and incubated with HRP-linked secondary antibodies. Detection was performed by enhanced chemiluminescence and the band intensity quantitated by laser densitometry. To determine p38 MAP kinase activity, total p38 MAP kinase was immunoprecipitated from total cell lysates using a rabbit anti–p38 MAP antibody and protein A sepharose beads. Immunoprecipitates were then incubated with ATF-2 fusion protein (2 µg) in the presence of ATP (200 µg) and kinase buffer for 30 min at 30°C. Phosphorylation of ATF-2 at Thr71 (14) was determined by Western blotting using a specific rabbit anti–phospho-ATF-2 (Thr71) antibody as described above. Inhibition of protein kinase C and p38 MAP kinase. In the experiments to inhibit protein kinase C (PKC) activity, serum- and insulin-deprived human mesangial cells were exposed to cyclical stretch for 30 min in the presence or absence of calphostin C, a specific PKC inhibitor. Calphostin C was used at a 1-µmol/l concentration added to the culture media 4 h before the experiment. In the experiments to inhibit p38 MAP kinase, cells were serum- and insulindeprived for 48 h and then exposed to stretch for various time periods up to 33 h, because preliminary experiments showed that longer periods of serum deprivation affected cell viability. p38 MAP kinase was inhibited by the addition of the imidazole derivative SB203580. This compound inhibits p38 MAP kinase activity by binding to the p38 MAP kinase ATP pocket (29), and it does not inhibit other relevant kinases such as extracellular signal-regulated kinase (ERK) and PKC (30–33). Because SB203580 has an inhibitory effect on the c-Jun NH2-terminal kinase (JNK)-2 isoforms at high concentrations (>10 µmol/l) (33), we used lower concentrations of 1, 5, and 10 µmol/l. Inhibition experiments on basal protein production were carried out simultaneously. Data presentation and statistical analysis. The number of experiments, carried out in triplicate, is reported in each figure legend. All data are presented as means ± SE. Student’s t test was used for the comparison between two groups. When more than two groups were studied, data were analyzed by analysis of variance and, if significant, the Newman-Keuls procedure was used for post hoc comparisons. Values for P < 0.05 were considered significant.

RESULTS

p38 MAP kinase is expressed by human mesangial cells and is activated in response to hyperosmolar stress. Hyperosmolar stress is an established p38 MAP kinase activator in various cell types (9). To determine whether p38 MAP kinase can be activated in human mesangial cells, we exposed human mesangial cells for 20 min to either isoosmolar (290 mosm/l) or hyperosmolar (440 mosm/l) media and analyzed total protein extracts by immunoblotting. Using a specific anti–p38 MAP kinase antibody, a single band (~40 kDa), consistent with the reported molecular weight of p38 MAP kinase (9), was detected, indicating that p38 MAP kinase is expressed as a single isoform in human mesangial cells. Activated phospho-p38 MAP kinase levels were very low in cells exposed to iso-osmolar media, but were induced by exposure to hyperosmolar media, which, however, had no effect on total p38 MAP kinase (Fig. 1). Stretch induces p38 MAP kinase activation in human mesangial cells. To test the effect of mechanical stretch on p38 MAP kinase, serum- and insulin-deprived human mesangial cells were exposed to cyclical stretch (elongation 10%) for short (1, 5, 10, 30, and 60 min) and long (6, 12, and 33 h) time periods. We found that stretch induced p38 MAP kinase activation by 5 min, with a maximum 11-fold increase at 30 min. High levels were sustained for up to 33 h. Total p38 MAP DIABETES, VOL. 49, APRIL 2000

G. GRUDEN AND ASSOCIATES

FIG. 1. p38 MAP kinase is expressed by human mesangial cells and is activated in response to hyperosmolar stress. Serum- and insulindeprived human mesangial cells were exposed for 20 min to iso-osmolar (290 mosml/l) and hyperosmolar (440 mosml/l) media. Total protein extracts were analyzed by immunoblotting using rabbit anti– phospho-p38 MAP kinase and goat anti–p38 MAP kinase antibodies. MAPK, mitogen-activated protein kinase.

kinase protein levels were unaltered by stretch. A stretchinduced increase in p38 MAP kinase activity peaking at 30 min was also observed in an in vitro kinase assay. The addition of SB203580 (1 µmol/l) resulted in a 94% reduction of stretchinduced p38 MAP kinase activity, indicating the inhibitory efficacy of this compound in our experimental setting (Fig. 2).

Stretch-induced p38 MAP kinase activation is PKC dependent. To investigate the molecular mechanism of p38 MAP kinase activation by stretch, we examined whether the PKC pathway was involved. Mesangial cells were exposed to stretch for 30 min in the presence or absence of calphostin C (1 µmol/l). Stretch-induced p38 MAP kinase activation was completely abolished by calphostin C (n = 3, P < 0.05 stretch + calphostin vs. stretch) (Fig. 3) and was thus considered PKC dependent. Stretch-induced fibronectin in human mesangial cells is p38 MAP kinase dependent. The application of stretch to rat mesangial cells induces fibronectin and other matrix components (34). p38 MAP kinase induces transcription factors, which bind to the promoter region of extracellular matrix genes (15–18). Therefore, we investigated the role of p38 MAP kinase in stretch-induced fibronectin. No changes were seen in mesangial cell number up to 33 h of stretch (nonstretch [26,390 ± 1,343] vs. stretch [25,657 ± 380]), indicating that stretch did not stimulate cell proliferation over this time period. Cyclical stretch induced a significant 67% increase in fibronectin levels in the medium at 33 h, with no increase at 12 or 24 h (Fig. 4A). There was also a significant increase in pericellular polymeric fibronectin (1.3 ± 0.03–fold increase over control, P < 0.05), indicating that the increase in fibronectin in the supernatant was paralleled by an enhanced fibronectin incorporation in the matrix.

FIG. 2. Stretch induces a rapid and sustained p38 MAP kinase activation in human mesangial cells. Serum- and insulin-deprived human mesangial cells were exposed to cyclical stretch (elongation 10%) for short (1, 5, 10, 30, and 60 min) and long (6, 12, and 33 h) time periods. Active and total p38 MAP kinase levels were determined by immunoblotting using rabbit anti–phospho-p38 MAP kinase and goat anti–p38 MAP kinase antibodies (A and B). p38 MAP kinase activity was determined in the presence and absence of SB203580 1 µmol/l (SB) by in vitro kinase assay using ATF-2 fusion protein (35 kDa) as substrate. Inhibition experiments with SB203580 were carried out after 30 min exposure to stretch (C). Results were quantitated by densitometric analysis (D and E) and expressed as fold increase over control values represented by the dotted line (number of experiments = 3–5). *P < 0.001 stretched vs. nonstretched cells. DIABETES, VOL. 49, APRIL 2000

657


FIG. 3. Stretch-induced p38 MAP kinase activation is PKC dependent. Serum- and insulin-deprived mesangial cells were exposed to stretch for 30 min in the presence or absence of the specific PKC inhibitor, calphostin C. Total protein extracts were analyzed by immunoblotting using rabbit anti–phospho-p38 MAP kinase.

At the 33-h time point, SB203580 (1, 5, and 10 µmol/l) completely abolished at all concentrations the increase in fibronectin levels in the supernatant seen in response to stretch, with no effect on basal fibronectin levels (Fig. 4B). Stretch induces TGF- 1 via p38 MAP kinase. Given the importance of TGF- 1 in matrix production in renal disease, we tested the role of p38 MAP kinase in stretch-induced TGF- 1. Mesangial cells were exposed to cyclical stretch for 6, 12, 24, and 33 h in the presence or absence of SB203580 (1 µmol/l). A rise in TGF- 1 protein levels was seen after 24 h of stretch and reached a peak 1.5-fold increase by 33 h. The addition of SB203580 abolished stretch-induced TGF- 1 protein levels at 33 h, but did not affect basal TGF- 1 levels (Fig. 4C), indicating that stretch-induced TGF- 1 is p38 MAP kinase dependent.

Late p38 MAP kinase activation by stretch is TGF- 1 dependent. TGF- 1 is a known activator of p38 MAP kinase in other cell types (11). Having established that stretch activates p38 MAP kinase, which induces TGF- 1, we investigated the effect of TGF- 1 on p38 MAP kinase in human mesangial cells. After preliminary dose-response experiments, mesangial cells were exposed to rhTGF- 1 (1.1 ng/ml) for 1, 10, 15, and 30 min. TGF- 1 induced a twofold increase in activated phospho-p38 MAP kinase levels by 1 min, which rose to a fourfold increase at 15 and 30 min (data not shown). To test the role of TGF- 1 in stretch-induced p38 MAP kinase activation, mesangial cells were exposed to cyclical stretch for 6, 12, and 33 h in the presence or absence of a specific neutralizing anti–TGF- 1 antibody (1 µg/ml). Stretchinduced p38 MAP kinase activation was not affected at the early time points of 6 and 12 h, but was significantly reduced, though not abolished, at 33 h by TGF- 1 neutralization (Fig. 5). This suggests that late stretch-induced p38 MAP kinase activation is, at least in part, TGF- 1 dependent. TGF- 1 requires p38 MAP kinase for the induction of fibronectin. To test the role of TGF- 1 in stretch-induced fibronectin, human mesangial cells were exposed to cyclical stretch for 33 h in the presence of a pan-neutralizing anti–TGF- antibody (20 µg/ml) or control nonimmune IgG. A pan-neutralizing anti–TGF- antibody previously used by others (35,36) was chosen to allow comparison with previous studies. Stretch produced a significant increase in fibronectin, which was not prevented by TGF- blockade (stretch = 1.47 ± 0.06–fold increase vs. control, P < 0.01; stretch + anti–TGF- = 1.43 ± 0.09–fold increase vs. control,

FIG. 4. Stretch-induced fibronectin and TGF- 1 in human mesangial cells are p38 MAP kinase dependent. A: Time course of fibronectin accumulation in human mesangial cells exposed to stretch (10% elongation) for 12, 24, and 33 h. B: Effect of SB203580 (SB) on stretch-induced fibronectin at 33 h. C: Effect of SB203580 (SB) on stretch-induced TGF- 1 at 33 h. Results are expressed as fold increase over control values represented by the dotted line (number of experiments = 5). *P < 0.001, #P < 0.01 stretch vs. others. 658

DIABETES, VOL. 49, APRIL 2000


FIG. 5. TGF- 1 contributes to late stretch-induced p38 MAP kinase activation. Serum- and insulin-deprived mesangial cells were exposed to stretch for 6, 12, and 33 h in the presence ( ) or absence ( ) of a specific neutralizing anti–TGF- 1 antibody (1 µg/ml). Active p38 MAP kinase levels were determined by immunoblotting using an anti–phospho-p38 MAP kinase antibody and quatitated by densitometric analysis. Results are expressed as fold increase over control and representative immunoblots are shown in the inserts. Number of experiments = 4. *P < 0.05 stretch and stretch + anti–TGF- 1 vs. nonstretch at 6 and 12 h; #P < 0.05 stretch vs. stretch + anti–TGF- 1; and P < 0.05 stretch vs. nonstretch at 33 h.

P < 0.01; stretch vs. stretch + anti–TGF- = NS; n = 4), confirming that at this 33 h time point, stretch-induced fibronectin was TGF- independent. To further investigate this relationship, we performed time-course experiments to examine the effect of TGF- 1 on fibronectin production. The addition of exogenous TGF- 1 (1.1 ng/ml) to mesangial cells induced a significant 1.3-fold increase in fibronectin protein levels after 72 h with no increase at earlier time points (6, 12, 24, and 48 h). In the presence of the p38 MAP kinase inhibitor SB203580 (1 µmol/l), TGF- 1–induced fibronectin production was completely abolished (Fig. 6), indicating that this late TGF- 1–induced fibronectin production was p38 MAP kinase–dependent. DISCUSSION

This study demonstrates that stretch activates p38 MAP kinase in human mesangial cells and suggests that p38 MAP kinase activation has a central role in the process of extracellular matrix accumulation initiated by a mechanical insult. The application of cyclical stretch to human mesangial cells induced a significant 11-fold increase in activated p38 MAP kinase levels. This provides the first evidence of stretchinduced MAP kinase activation in any human cell. The complete abrogation of p38 MAP kinase activation in the presence of the PKC inhibitor calphostin C strongly supports a role of PKC as an intracellular mediator of stretchinduced p38 MAP kinase activation. PKC is rapidly activated within minutes in stretched rat mesangial cells (37) and mediates glucose-induced p38 MAP kinase activation in vascular smooth muscle cells (38). p38 MAP kinase activation occurred within 5 min, reached a peak at 30 min, and was sustained for up to 33 h. This sustained p38 MAP kinase activation is of interest, because in rat mesangial cells, stretch-induced activation of other MAP kinases, such as Erk1 and Erk2, occurs within 10 min, returnDIABETES, VOL. 49, APRIL 2000

FIG. 6. TGF- 1 induces fibronectin via a p38 MAP kinase–dependent mechanism. Serum- and insulin-deprived human mesangial cells were exposed to rhTGF- 1 (1.1 ng/ml) for 12, 24, 48, and 72 h in the presence or absence of SB203580 (SB 1 µmol/l). Fibronectin protein levels were measured as described in the RESEARCHDESIGNANDMETHODS and expressed as fold increase versus control (· · ·). Number of experiments = 5. *P < 0.01 rhTGF- 1 vs. rhTGF- 1 + SB.

ing to basal levels by 60 min (20). This late p38 MAP kinase activation is temporally related to an increase in TGF- 1 protein levels and is reduced by TGF- 1 inhibition, indicating that while the rapid stretch-induced p38 MAP kinase activation is TGF- 1 independent, TGF- 1 becomes important in the sustained activation of p38 MAP kinase. Accordingly, the addition of TGF- 1 to human mesangial cells resulted in p38 MAP kinase activation, confirming findings previously reported in nonrenal cell types (11). The intracellular mediator(s) responsible for TGF- 1–induced p38 MAP kinase activation remain(s) to be elucidated, but the TGF- –activated kinase 1, known to stimulate p38 MAP kinase activity, is a likely candidate (39). The magnitude of stretch-induced p38 MAP kinase activation was comparable to that observed in human mesangial cells exposed to hyperosmolar media. Furthermore, it was greater than that reported in cardiac and vascular smooth muscle cells exposed to angiotensin II, endothelin 1, and high glucose (39–41). A physiological degree of stretch (4% elongation) does not induce cytokine or matrix production in human mesangial cells (25,26), nor does it induce p38 MAP kinase activation in cardiac cells (42) or human mesangial cells. Thus, degrees of stretch-simulating pathological states (10% elongation) seem to be necessary for these effects. Our data indicate that the stretch-induced increase in mesangial cell TGF- 1 and fibronectin protein production is p38 MAP kinase dependent since it was abolished by SB203580, a specific p38 MAP kinase inhibitor. This inhibitor was highly effective in blocking p38 MAP kinase activity in response to stretch in our experimental setting. Although the highest dose of SB203580 used (10 µmol/l) may not be entirely selective for p38 MAP kinase because of its inhibitory effect on JNK-2 (33), an equal effect was seen at lower doses with no loss of efficacy. Moreover, stretch inducing a 10% elongation does not activate JNK in rat mesangial cells (19,20,44). 659


p38 MAP kinase phosphorylates transcription factors inducing c-fos and c-jun transcription (12–14). c-fos and c-jun combine to form the AP1 transcription complex that binds to and activates genes with a 12-O-tetradecanoylphorbol 13acetate (TPA)-response element (TRE) in their promoter region. TGF- 1 and fibronectin both contain TRE regions (15,16) and p38 MAP kinase is thus likely to induce fibronectin and TGF- 1 production via this mechanism. The expression of c-fos has been shown to be induced by stretch in rat mesangial cells (6). Under our experimental conditions, the effect of stretch and TGF- 1 on fibronectin was more modest compared with that of other studies (7). Previous studies have used rat mesangial cells, which may have a different quantitative response. In addition, we specifically used serum- and insulin-free experimental conditions to test purely the effect of a mechanical insult. Serum induces a 50–60% augmentation in stretchinduced matrix production in mesangial cells (34,38) and insulin enhances TGF- 1 and fibronectin production (45), thus confounding and amplifying the effect of stretch. The addition of TGF- 1 to mesangial cells resulted in fibronectin accumulation after 72 h via a p38 MAP kinase–dependent mechanism, suggesting that p38 MAP kinase may be an important pathway for a number of matrix-promoting stimuli. At 33 h, however, TGF- 1 blockade did not affect stretchinduced fibronectin, in accord with the observation that stretch-induced (I) collagen expression precedes that of TGF- 1 (34) and that in the rat remnant kidney model, early alterations in matrix accumulation precede any increase in local TGF- 1 expression (46). Furthermore, in immunohistochemistry studies using different degrees of cell-stretching in a single culture well, the magnitude of matrix production correlated directly with the degree of cell elongation (34,47), a finding against a role of a secreted cytokine, such as TGF- 1, and in favor of a direct effect of stretch. Hirakata et al. (35) reported that stretch-induced fibronectin gene expression in rat mesangial cells was abolished by an anti–TGF- 1 antibody, whereas Yasuda et al. (34) and Riser et al. (36) found no effect of TGF- 1 inhibition on stretchinduced collagen production. Our data suggest that stretch induces fibronectin via a biphasic signaling pathway. Stretch, via PKC, causes early activation of p38 MAP kinase, which

FIG. 7. Proposed sequence of molecular events from mechanical stretch to fibronectin production in human mesangial cells. Stretch via PKC activates p38 MAP kinase (indicated by *), which induces TGF- 1 and fibronectin (indicated by ↑). TGF- 1, in turn, maintains activation of p38 MAP kinase, perpetuating fibronectin production. MAPK, mitogen-activated protein kinase. 660

induces TGF- 1 and fibronectin production. In turn, TGF- 1 maintains late p38 MAP kinase activation, which perpetuates fibronectin production (Fig. 7). These findings could have important clinical pathophysiological implications for renal disease. p38 MAP kinase is important in mediating inflammatory responses in neutrophils and cell hypertrophy in the heart (8,48), and this is the first evidence that p38 MAP kinase participates in the intracellular signaling cascade leading to mesangial cell extracellular matrix production and, presumably, glomerulosclerosis. The interplay between p38 MAP kinase and TGF- 1, described here, provides a potential mechanism for a progressive and sustained sclerotic process, initially triggered by a hemodynamic insult. Our observations may have particular relevance for diabetes because p38 MAP kinase is also activated by high glucose via a PKC-dependent mechanism (38), and active p38 MAP kinase levels are augmented in vivo in glomeruli from diabetic rats (10). ACKNOWLEDGMENTS

This work was supported by the British Diabetic Association Equipment Grants RG/98/0001608 and RG98/0001862 and by a Novo Nordisk Grant. Dr. Gabriella Gruden is an R.D. Lawrence fellow of the British Diabetic Association. Dr. Silvia Zonca was a visiting fellow from the Department of Pediatrics III at the University of Milan. Dr. Sabrina Maestrini was a visiting scientist from the Istituto Auxologico Italiano, Milan, Italy. We thank Mr. R. Gruden for helpful technical assistance. REFERENCES 1. Osterby R, Parving H-H, Hommel E, Jorgensen HE, Lokkegard H: Glomerular structure and function in diabetic nephropathy. Diabetes 39:1057–1063, 1990 2. Hostetter TH, Rennke HG, Brenner BM: The case for intra-renal hypertension in the initiation and progression of diabetic and other glomerulopathies. Am J Med 72:375–380, 1982 3. Viberti G, Moghensen CE, Groop LC, Pauls JF, for the European Microalbuminuria Captopril Study Group: Effect of captopril on progression to clinical proteinuria in patients with insulin-dependent diabetes mellitus and microalbuminuria. JAMA 271:275–279, 1994 4. Lewis EJ, Hunsicker LG, Bain RP, Rohde RD, for the Collaborative Study Group: The effect of angiotensin-converting-enzyme inhibition on diabetic nephropathy. N Engl J Med 329:1456–1462, 1993 5. Zatz R, Dunn BR, Meyer TW, Anderson S, Rennke HG, Brenner BM: Prevention of diabetic glomerulopathy by pharmacologic amelioration of glomerular capillary pressure. J Clin Invest 77:1925–1930, 1986 6. Harris RC, Akai Y, Yasuda T, Homma T: The role of physical forces in alterations of mesangial cell function. Kidney Int Suppl 45:S17-S21, 1994 7. Riser BL, Cortes P, Heiling C, Grondin J, Ladson-Worford S, Patterson D, Naris RG: Cyclic stretching force selectively up-regulates transforming growth factor- isoforms in cultured mesangial cells. Am J Pathol 148:1915– 1923, 1996 8. Raingeaud J, Gupta S, Rogers J, Dickens M, Han J, Ulevitch RJ, Davis RJ: Proinflammatory cytokines and environmental stress cause p38 MAP kinase activation by dual phosphorylation on tyrosine and threonine. J Biol Chem 270: 7420–7426, 1995 9. Han J, Lee JD, Bibbs L, Ulevitch RJ: A MAP kinase targeted by endotoxin and hyperosmolarity in mammalian cells. Science 265:808–811, 1994 10. Dunlop ME, Johnson EIM, Muggli EE: Glomerular activation of p38 mitogenactivated protein kinase (p38MAPK) and accumulation of small stress protein hsp25 in experimental diabetes (Abstract). Diabetes 47 (Suppl. 1):A51, 1998 11. Hannigan M, Zhan L, Ai Y, Huang C-K: The role of p38 MAP kinase in TGF- 1induced signal transduction in human neutrophils. Biochem Biophys Res Commun 246:55–58, 1998 12. Janknecht R, Hunter T: Convergence of MAP kinase pathways on the ternary complex factor SAP-1a. EMBO J 16:1620–1627, 1997 13. Whitmarsh AJ, Yang S-H, Su MS-S, Sharrocks AD, Davis RJ: Role of p38 and JNK mitogen activated protein kinases in the activation of ternary complex factors. Mol Cell Biol 17:2360–2371, 1997 DIABETES, VOL. 49, APRIL 2000


14. Gupta S, Campbell D, Derijard B, Davis RJ: Transcription factor ATF2 regulation by the JNK signal transduction pathway. Science 267:389–393, 1995 15. Kim SJ, Angel P, Lafyatis R, Hattori K, Kim KY, Sporn MB, Karin M, Roberts AB: Autoinduction of transforming growth factor 1 is mediated by the AP-1 complex. Mol Cell Biol 10:1492–1497, 1990 16. Dean DC, McQuillian JJ, Weintraub S: Serum stimulation of fibronectin gene expression appears to result from rapid serum-binding of nuclear proteins to a cAMP response element. J Biol Chem 265:3522–3527, 1990 17. Muro AF, Bernath VA, Kornblihtt AR: Interaction of the 170 cyclic AMP response element with the adjacent CCAAT box in the human fibronectin gene promoter. J Biol Chem 267:12767–12774, 1992 18. Corsi PS, Ransone LJ, Verma IM: Cross-talk in signal transduction: TPAinducible factor Jun/AP1 activates cAMP-responsive enhancer elements. Oncogene 5:427–431, 1990 19. Ingram AJ, Ly H, Thai K, Kang M, Scholey JW: Activation of mesangial cell signaling cascades in response to mechanical strain. Kidney Int 55:476–485, 1999 20. Ishida T, Haneda M, Koya D, Kikkawa R: Stretch-induced overproduction of fibronectin in mesangial cells is mediated by the activation of mitogen-activated protein kinase. Diabetes 48:595–602, 1999 21. Striker GE, Striker LJ: Biology of disease: glomerular cell culture. Lab Invest 53:122–131, 1985 22. Beaufils H, Patte R, Aubert P, Camey M, Kuss R, Barbagelatta M, Chomette G: Renal immunopathology in renal cell carcinoma. Virchows Arch A Pathol Anat Histopathol 404:87–97, 1984 23. Sacks S, Zhou W, Campbell RD, Martin J: C3 and C4 gene expression and interferon-gamma-mediated regulation in human glomerular mesangial cells. Clin Exp Immunol 93:411–417, 1993 24. Banes AJ, Gilbert J, Taylor D, Monbureau O: A new vacuum-operated stressproviding instrument that applies static or variable duration cycling tension or compression to cells in vitro. J Cell Sci 75:35–42, 1985 25. Gruden G, Thomas S, Burt D, Lane S, Chusney G, Sacks S, Viberti GC: Mechanical stretch induces vascular permeability factor in human mesangial cells: mechanisms of signal transduction. Proc Natl Acad Sci U S A 94:12112– 12116, 1997 26. Cortes P, Riser BL, Zhao X, Narins RG: Glomerular volume expansion and mesangial cell mechanical stain: mediators of glomerular pressure injury. Kidney Int 45:S11–S16, 1994 27. Rennard SI, Berg R, Martin GR, Foidart JM, Robey PG: Enzyme-linked immunoassay (ELISA) for connective tissue components. Anal Biochem 104:205–214, 1980 28. Coin MG, Hynes RO: Biosynthesis and processing of fibronectin in NIL.8 hamster cells. J Biol Chem 254:12050–12055, 1979 29. Tong L, Pav S, White DM, Rogers S, Crane KM, Cywin CL, Brown ML, Pargellis CA: A highly specific inhibitor of human p38 MAP kinase binds in the ATP pocket. Nat Struct Biol 4:311–316, 1997 30. Cuenda A, Rouse J, Doza YN, Meier R, Cohen P, Gallagher TF, Young PR, Lee JC: SB 203580 is a specific inhibitor of a MAP kinase homologue which is stimulated by cellular stresses and interleukin-1. FEBS Lett 364:229–233, 1995 31. Rausch O, Marshall CJ: Cooperation of p38 and extracellular signal-regulated kinase mitogen-activated protein kinase pathways during granulocyte colony-stimulating factor-induced hemopoietic cell proliferation. J Biol Chem 274:4096–4105, 1999

DIABETES, VOL. 49, APRIL 2000

32. Lee JC, Laydon JT, McDonnell PC, Gallagher TF, Kumar S, Green D, McNulty D, Blumenthal MJ, Heys JR, Landvatter SW, Strickler JE, McLaughlin MM, Siemens IR, Fisher SM, Livi GP, White JR, Adams JL, Young PR: A protein kinase involved in the regulation of inflammatory cytokine biosynthesis. Nature 372:739–746, 1994 33. Whitmarsh AJ, Yang S-H, Su MS-S, Sharrocks AD, Davis RJ: Role of p38 and JNK mitogen-activated protein kinase in the activation of ternary complex factors. Mol Cell Biol 17:2360–2371, 1997 34. Yasuda T, Kondo S, Homma T, Harris RC: Regulation of extracellular matrix by mechanical stress in rat glomerular mesangial cells. J Clin Invest 98:1991– 2000, 1996 35. Hirakata M, Kaname S, Chung UG, Joki N, Hori Y, Noda M, Takuwa Y, Okazaki T, Fujita T, Katoh T, Kurokawa K: Tyrosine kinase dependent expression of TGF- induced by stretch in mesangial cells. Kidney Int 51:1028–1036, 1997 36. Riser BL, Cortes P, Yee J, Sharba AK, Asano K, Rodriguez-Barbero A, Narins RG: Mechanical strain- and high glucose-induced alterations in mesangial cells collagen metabolism: role of TGF- 1. Am J Soc Nephrol 9:827–836, 1998 37. Homma T, Akai Y, Burns KD, Harris RC: Activation of S6 kinase by repeated cycles of stretching and relaxation in rat glomerular mesangial cells: evidence for involvement of protein kinase C. J Biol Chem 267:23129–23135, 1992 38. Igarashi M, Wakasaki H, Takahara N, Ishii H, Jiang Z-Y, Yamauchi T, Kuboki K, Meier M, Rhodes CJ, King GL: Glucose or diabetes activates p38 mitogen-activated protein kinase via differential pathways J Clin Invest 103:185–195, 1999 39. Yamaguchi K, Shirakabe K, Shibuya H, Irie K, Oishi I, Ueno N, Taniguchi T, Nishida E, Matsumoto K: Identification of a member of the MAPKKK family as a potential mediator of TGF- 1 signal transduction. Science 270:2008–2011, 1995 40. Kusuhara M, Takahashi E, Peterson TE, Abe J-I, Ishida M, Han J, Ulevitch R, Berk BC: p38 kinase is a negative regulator of angiotensin II signal transduction in vascular muscle cells. Circ Res 83:824–831, 1998 41. Choukroun G, Hajjar R, Kyriakis JM, Bonventre JV, Rosenzweig A, Force T: Role of the stress-activated protein kinase in endothelin-induced cardiomyocytes hypertrophy. J Clin Invest 102:1311–1320, 1998 42. MacKenna DA, Dolfi F, Vuori K, Ruoslahti E: Extracellular signal-regulated kinase and c-Jun NH2-terminal kinase activation by mechanical stretch is integrin-dependent and matrix-specific in rat cardiac fibroblasts. J Clin Invest 101:301–310, 1998 44. Kawata Y, Mizukami Y, Fujii Z, Sakumura T, Yoshida K, Matsuzaki M: Applied pressure enhances cell proliferation through mitogen-activated protein kinase activation in mesangial cells. J Biol Chem 273:16905–16912, 1998 45. Anderson PW, Zhang XY, Tina J, Correale JD, Xi XP, Yang D, Graf K, Law RE, Hsueh WA: Insulin and angiotensin II are additive in stimulating TGF- 1 and matrix mRNAs in mesangial cells. Kidney Int 50:745–753, 1996 46. Lee LK, Meyer TW, Polock AS, Lovett DH: Endothelial cell injury initiates glomerular sclerosis in the rat remnant kidney. J Clin Invest 96:953–964, 1995 47. Riser BL, Cortes P, Zhao X, Bernstein J, Dumier F, Narins RG: Intraglomerular pressure and mesangial stretching stimulate extracellular matrix formation in the rat. J Clin Invest 90:1932–1943, 1992 48. Nemoto S, Sheng Z, Lin A: Opposing effects of Jun kinase and p38 mitogenactivated protein kinases on cardiomyocyte hypertrophy. Mol Cell Biol 18: 3518–3526, 1998

661