Mechanical stretch induces vascular permeability factor in human mesangial cells: Mechanisms of signal transduction

Abstract

Hemodynamic abnormalities have been implicated in the pathogenesis of the increased glomerular permeability to protein of diabetic and other glomerulopathies. Vascular permeability factor (VPF) is one of the most powerful promoters of vascular permeability. We studied the effect of stretch on VPF production by human mesangial cells and the intracellular signaling pathways involved. The application of mechanical stretch (elongation 10%) for 6 h induced a 2.4-fold increase over control in the VPF mRNA level (P < 0.05). There was a corresponding 3-fold increase in VPF protein level by 12 h (P < 0.001), returning to the baseline by 24 h. Stretch-induced VPF secretion was partially prevented both by the protein kinase C (PKC) inhibitor H7 (50 μM: 72% inhibition, P < 0.05) and by pretreatment with phorbol ester (phorbol-12-myristate-13 acetate 10⁻⁷ M: 77% inhibition, P < 0.05). A variety of protein tyrosine kinase (PTK) inhibitors, genistein (20 μg/ml), herbimycin A (3.4 μM), and a specific pp60^src peptide inhibitor (21 μM) also significantly reduced, but did not entirely prevent, stretch-induced VPF protein secretion (respectively 63%, 80%, and 75% inhibition; P < 0.05 for all). The combination of both PKC and PTK inhibition completely abolished the VPF response to mechanical stretch (100% inhibition, P < 0.05). Stretch induces VPF gene expression and protein secretion in human mesangial cells via PKC- and PTK-dependent mechanisms.

Hemodynamic factors have been implicated pathogenically in the increased glomerular permeability to macromolecules, which characterizes diabetic and other glomerulopathies (1, 2), and treatment strategies that reduce glomerular capillary pressure have an antiproteinuric and renoprotective effect in both human and animal studies (3–5).

Mesangial cells are in apposition to, and in continuum with, the glomerular capillaries and thus constitute a primary target for the mechanical insult induced by increased glomerular capillary pressure. The mesangium plays a crucial role in the glomerular trafficking of plasma proteins, and their deposition and accumulation within the mesangium may eventually lead to the development of glomerular sclerotic lesions (6). The vast majority of in vitro studies on mesangial cells have been performed under static conditions, and little is known about the response of mesangial cells to a mechanical insult.

Recently, application of mechanical stretch to mimic a hemodynamic insult has been reported to induce mesangial cell matrix and transforming growth factor (TGF)-β₁ production in human and rat mesangial cells (7–9), suggesting a potential mechanism whereby a hemodynamic insult may be translated into a glomerular sclerotic process. Whether mechanical stretch could also induce the expression of factor(s) that may influence glomerular permeability is unknown.

Vascular permeability factor (VPF), also named vascular endothelial growth factor, is known in four isoforms (10, 11), binds to two high affinity receptors predominantly located on vascular endothelium, and induces endothelial cell proliferation and increased vascular permeability to macromolecules (12–14). VPF is produced by several glomerular cell types (15–18), and VPF receptors are present on glomerular cells, including mesangial cells, which are known to express the mRNA for the VPF receptor (flt-1) (19). As a permeability factor, VPF is 50,000 times more active than histamine (20) and might affect glomerular permeability via autocrine or paracrine mechanisms. Indeed, the infusion of VPF induces proteinuria in experimental animals (21), and elevated VPF plasma levels have been observed in minimal change nephrotic syndrome in humans (22). Little is known as yet about the modulation of kidney VPF production in response to hemodynamic perturbations.

The present study was designed to test whether the application of mechanical stretch induces VPF gene expression and protein secretion in human mesangial cells and to investigate the signal transduction pathways involved in this process. We describe here a molecular effect of mechanical stretch and provide a mechanism whereby a hemodynamic insult might influence glomerular permeability.

METHODS

Materials.

All materials were purchased from Sigma unless otherwise stated. RPMI 1640 culture medium, fetal calf serum, and Triazol were obtained from GIBCO/BRL. Flex I and Flex II plates were from Flexcell International (McKeesport, PA). H7 1-(5-isoquinolinesulfonyl)-2-methylpipirazine, genistein (4′, 5, 7-trihydroxyisoflavone), herbimycin A, tyrosine kinase peptide inhibitor (p60src residues 137–157), and phorbol-12-myristate-13 acetate (PMA) were purchased from Calbiochem. The reverse transcription system was purchased from Promega, oligonucleotide primers were purchased from Oswel (Southampton, United Kingdom), and AmpliTaq was purchased from Perkin–Elmer. Monoclonal and rabbit polyclonal anti-human VPF antibodies were obtained, respectively, from R & D System and Serotec.

Cell Culture.

Human mesangial cells were isolated as described (23). In brief, normal renal cortex was taken from the opposite tumor-free pole of nephrectomy specimens and removed for localized, capsulated grade 1 hypernephromas. Intact glomeruli were collected by serial sieving of cortical homogenates. Tissue was analyzed by light microscopy and by immunofluorescence to confirm the absence of tumor cells and to exclude the presence of glomerular abnormalities (24). Cells obtained from three separate kidneys were used in this study. After digestion with collagenase (type IV, 750 units/ml), the isolated glomeruli were seeded in culture flasks. After the outgrowth of mesangial cells, the glomeruli were removed by washing with PBS, and the cells were cultured in RPMI 1640 medium supplemented with insulin–transferrin–selenium and L-glutamine and containing 20% fetal calf serum, 7 mM glucose, 100 units/ml penicillin/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 α-smooth muscle actin by direct immunofluorescence. Cells did not stain for cytokeratin, factor VIII, common leukocyte antigen (DAKO), or Thy-1 (Serotec), excluding contamination of epithelial and endothelial cells, lymphomonocytes, and human fibroblasts (25). Studies were performed between passages 4 and 7, and the cells retained all of the morphologic and immunofluorescent features described above. A subset of experiments was performed on human skin fibroblasts, a cell type producing VPF in basal conditions (26), to evaluate the cell specificity of the stretch-induced effects and to control for potential nonspecific protein release from plasma membrane stretch-induced damage. Skin fibroblasts were obtained from a forearm skin biopsy performed on normal subjects, and cells were cultured in standard conditions and used after five passages. The experimental protocol paralleled that described for mesangial cells.

Application of Mechanical Stretch to Cultured Cells.

Mesangial cells were seeded in equal number (12,000/cm²) into 6-well, collagen-coated, silicone 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 Barnes and coworkers (27) and consists of a vacuum unit and a baseplate. Vacuum was applied cyclically (60 cycles/min) to the rubber-base dishes via the baseplate, which was placed in a humidified incubator with 5% CO₂ at 37°C. Vacuum pressures were applied to induce an average of either a 4 or 10% uniaxial elongation in the culture surface. 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.

Cell Number Determination.

Cells were harvested with 0.25% trypsin and 0.5% EDTA, and the cell number was determined by a Coulter cell counter (Coulter).

mRNA Analysis.

Total RNA was isolated using a commercial preparation based on a guanidinium and phenol extraction (Triazol) and reverse transcribed (1 μg) according to standard protocols using avian myeloblastosis virus reverse transcriptase and poly d(T). The PCR was performed with oligonucleotide primers complementary with sequences located in exons 2 and 5–7 (28, 29). The primer for exon 2 was based on that described by Iijima et al. (18), and the primer for exon 5–7 was designed to amplify specifically the 165 isoform of human VPF, which lacks exon 6 (28, 29). A single PCR product of 317 bp was obtained, the identity of which was confirmed by digestion with the restriction enzyme HintIII (Promega), yielding two fragments of 181 and 136 bp as predicted from the known cDNA sequence for VPF¹⁶⁵ (29). Expression of the housekeeping gene GAPDH was determined in parallel to control for amount of RNA input and reverse transcription efficiency using a primer sequence reported (30). VPF and GAPDH mRNA levels were quantitated by competitive reverse transcriptase-PCR using deletion-mutated cDNA to control for PCR amplification efficiency and for use in quantitative analysis as described (31). PCR products were resolved in a 3% Nu-Sieve/1% agarose gel containing ethidium bromide, analyzed by an image system (Eagle Eye System, Stratagene), and quantitated using densitometry analysis software (qgel, Stratagene).

Generation of Competitor cDNA.

Competitor cDNAs with a 50-bp deletion were generated by PCR according to Celi et al. (32), and the product obtained was isolated by gel and column purification and quantitated by densitometry. Native and competitor cDNAs had similar amplification kinetics.

Protein Analysis.

Culture supernatants from all experimental conditions were collected, centrifuged to remove cell debris, and stored at −70°C for analysis. VPF protein concentration was measured by an in-house, two-site immunoenzymometric assay using a mouse monoclonal and a rabbit polyclonal anti-human VPF¹⁶⁵ (range 1–40 pM, intra-assay coefficient of variation: 5.3%). For each experiment, VPF protein levels were determined within a single assay; 96-well cluster plates were coated overnight at 4°C with a mouse monoclonal anti-VPF antibody as the capture antibody. The plates were blocked with BSA, after which the samples were added and incubated for 5 h. After washing, a rabbit polyclonal anti-human VPF¹⁶⁵ as the detection antibody was added and incubated overnight. Immunocomplexes were detected by horseradish peroxidase-conjugated goat-anti-rabbit IgG and revealed by 3,3′,5,5′-tetramethylbenzidine dihydrochloride substrate. The reaction was stopped with H₂S0₄, and the absorbance was measured at 450/690 nm. The assay also detects the VPF¹²¹ isoform, but no cross-reactivity was detected with human platelet-derived growth factor, human TGF-β_1–5, or bovine VPF. All protein results were adjusted for cell number.

Inhibition Experiments.

Serum- and insulin-deprived mesangial cells were exposed to protein kinase C (PKC) inhibition by preincubation for 1 h with H7 (50 μM) or down-regulation by preincubation for 24 h with PMA (10⁻⁷M). PTK inhibition was obtained by preincubation for 1 h with genistein (20 μg/ml), herbimycin A (3.4 μM), or pp60^src tyrosine kinase peptide inhibitor (peptide A, 21 μM), a 21-residue peptide corresponding to a part (residues 137–157) of the noncatalytic domain of pp60^src (33). Cells were then subjected to mechanical stretch for 12 h. Appropriate control experiments were conducted in parallel. Inhibition experiments on basal VPF protein secretion were carried out simultaneously.

Data Presentation and Statistic Analysis.

Number of experiments for each experimental condition is reported in the legend to figures. Comparisons among experimental conditions were made by ANOVA, and comparisons between experiments were performed using the Student–Newman–Keuls test. Values for P < 0.05 were considered significant. All data are presented as mean ± SEM.

RESULTS

Effect of Stretch on Mesangial Cell VPF mRNA Levels.

Cells were made quiescent by serum and insulin-deprivation for 48 h before the experiment because VPF mRNA expression is stimulated by low concentrations of fetal calf serum (18). Both stretched and control cells expressed the VPF165 isoform. At 10% elongation, the stimulus VPF mRNA level rose to a peak by 6 h (2.4-fold increase over control, P < 0.05) and diminished toward the basal level by 12 h (1.3-fold increase over control) (Fig. 1A).

Figure 1.

Effect of mechanical stretch on VPF mRNA and protein levels. Human mesangial cells were serum- and insulin-deprived for 48 h, then exposed to mechanical stretch for the indicated time periods. (A) VPF mRNA level was quantitated as described in Materials and Methods and expressed as fold increase vs. control nonstretched cells (n = 4). ∗, P < 0.05 stretched vs. control nonstretched cells at 6 h. (B) VPF protein level was measured as described in Materials and Methods and expressed as fold increase vs. control nonstretched cells (n = 15). ∗∗, P < 0.001 stretched vs. control nonstretched cells at 12 h.

Effect of Cell Stretch on VPF Protein Levels.

To determine whether altered VPF mRNA levels were accompanied by an increased protein secretion, the VPF concentration was measured in the culture supernatant. VPF protein was detected in control conditions, and the application of a low level of mechanical stretch (4% elongation) did not alter VPF protein secretion (1.03 ± 0.08-fold increase over control). In contrast, higher degrees of mechanical stretch (10% elongation) resulted in a marked increase in VPF protein levels by 12 h (3.1 ± 0.3-fold increase over control, P < 0.001) with a return to the baseline by 24 h (Fig. 1B), in keeping with the changes in mRNA. The application of mechanical stretch (elongation 10%) for 12 h did not alter VPF protein level in cultured human fibroblasts (1.05 ± 0.37-fold increase over control), confirming the cell specificity of the stimulus and the absence of nonspecific protein release due to stretch-induced cell membrane damage.

Effect of PKC and PTK Inhibition on Stretch-Induced VPF Secretion.

To investigate the intracellular mechanisms by which stretch induces VPF production, we examined the effect of PKC and protein tyrosine-kinase (PTK) inhibition. Pretreatment with H7 significantly inhibited the VPF secretion in response to stretch (72% inhibition; P < 0.05), and PKC depletion by pretreatment with PMA also significantly reduced the VPF response by 74% (P < 0.05), confirming the involvement of PKC (Fig. 2A). The addition of genistein significantly reduced stretch-induced VPF response by 63% (P < 0.05; Fig. 2B). Because genistein, a PTK inhibitor that competes with ATP, also is reported to have a PKC inhibitory activity at high concentrations, we tested the effect of herbimycin A, which blocks PTK activity through benzoquinone interactions with protein sulphydryl groups, and of peptide A, which has a direct and specific effect on PTK pp60^src (33). Neither of these compounds has an inhibitory activity on PKC (33, 34). Herbimycin A reduced stretch-induced VPF secretion by 80%, and peptide A induced a 75% inhibition (P < 0.05 for both; Fig. 2B). Pretreatment with a combination of H7 and genistein totally abolished stretch-induced VPF protein secretion (100% inhibition, P < 0.05; Fig. 2C). Basal secretion of VPF was unaffected by pretreatment with H7, PMA, genistein, herbimycin A, and peptide A (Fig. 2D). Cell viability, confirmed by trypan blue exclusion, was similar in all experimental conditions.

Figure 2.

Effect of PKC and PTK inhibition on basal and stretch-induced VPF protein secretion. Serum- and insulin-deprived mesangial cells were exposed for 12 h to mechanical stretch in the presence or absence of (A) PKC inhibition by 1-h preincubation with H7 (50 μM) or PKC depletion by 24-h PMA preincubation (10⁻⁷ M), (B) PTK inhibition by 1-h preincubation with genistein (20 μg/ml) or herbimycin A (3.4 μM) or peptide A (21 μM), (C) PKC and PTK inhibition by 1-h preincubation with genistein (20 μg/ml) plus H7 (50 μM). The VPF protein level was measured as described in Materials and Methods and expressed as fold increase vs. control nonstretched cells (n = 4). Stretched cells vs. stretched cells plus inhibitors: ∗, P < 0.05 for all. (D) The effect of PKC and PTK inhibition on basal VPF secretion in control nonstretched cells. Absolute values in nanogram adjusted for cell number are reported (n = 4).

DISCUSSION

In the present study, we have demonstrated that mechanical stretch of cultured human mesangial cells induces VPF, one of the most powerful promoters of vascular permeability. Mesangial cells subjected to cyclic stretching, to simulate an in vivo hemodynamic insult producing an ≈10% elongation, showed a significant increase in VPF mRNA level, which was maximal after 6 h and diminished by 12 h. These changes in VPF mRNA level were related temporally to a significant 3-fold increase in the secretion of VPF peptide, which peaked after 12 h, returning to the baseline values by 24 h. Using precise quantitative determinations by immunoenzymometric assay, we confirmed here that secretion of VPF protein occurs in human mesangial cells under static basal conditions, as suggested by other studies in experimental animal mesangial cells (18). The time course of the VPF response to stretch was similar to that described for other VPF-inducers, such as serum, platelet-derived growth factor and 12-o-tetradecanoyl-phorbol-13-acetate (18, 35). Moreover, in rat mesangial cells, stretch induces the gene expression of other molecules, such as matrix components laminin and fibronectin, with a similar pattern (36). The progressive reduction in response with time is a common feature of response to in vitro stimuli. The explanation for this is unclear; however, down-regulation or depletion of cellular components necessary for the signal transduction may be important.

Our experiments, designed to determine the intracellular mechanisms mediating VPF induction by stretch, provided evidence that PKC and PTK are components of the signaling mechanisms involved. The significant blunting of the increase in VPF protein secretion in response to stretch by the addition of the PKC inhibitor, H7, or PMA-induced PKC depletion strongly supports a role of PKC as an intracellular mediator of stretch. PKC-responsive loci have been described on the VPF gene promoter, and other stimuli have been shown to induce VPF via a PKC-dependent mechanism (28, 35). PKC translocation to the cell membrane with phosphorylation of an 80-kDa endogenous substrate of PKC occurs within minutes of mechanical stretch (7, 37), and several other observations in both cultured myocardial cells and rat mesangial cells underscore the role of PKC activation in response to mechanical overload (37–39).

We further have demonstrated that mechanical stretch-induced VPF protein secretion also was significantly impaired by genistein and herbimycin A, two mechanistically unrelated protein tyrosine kinase inhibitors. These results prove that protein tyrosine kinases are directly implicated in the VPF response to mechanical stretch in human mesangial cells. Hypoxia was so far the only other stimulus known to induce VPF through a protein tyrosine kinase-dependent mechanism (40). Of interest, tyrosine kinases recently have been implicated also in stretch-induced TGF-β expression in rat mesangial cells (41). Which protein tyrosine kinases are responsible for VPF induction remains to be elucidated, but the protein tyrosine kinase pp60^src is a likely candidate because it is activated by stretch in fetal lung cells (42) and has been implicated in VPF induction by hypoxia (40). In keeping with this hypothesis, we observed an inhibition of stretch-induced VPF protein secretion in the presence of a tyrosine-kinase peptide inhibitor, which is highly specific for pp60^src.

The response to stretch was partially prevented by either PKC or PTK inhibition, and the combined inhibition of both PKC and protein tyrosine kinases totally abolished VPF induction by mechanical stretch. These results suggest that separate signaling cascades are likely to be implicated and raise the possibility that they may act independently.

These in vitro findings may have important implications for in vivo pathophysiological conditions in renal disease. Rises in intraglomerular pressure as measured in experimental diabetes and remnant kidney models can be calculated to result in an expansion in glomerular volume, which entails an ≈10% radial elongation (43, 44). Of interest, the increase in VPF protein seen in response to stretch was only evident at these higher degrees of elongation whereas the application of more “physiological” levels of mechanical stretch (4% elongation) did not alter VPF protein secretion. These data suggest that stretch-induced VPF expression may be a specific feature of glomeruli with high intraglomerular pressure and provide a potential molecular mechanism by which a hemodynamic insult might be translated into a greater transcapillary transit of macromolecules resulting in proteinuria, a key factor in the progression of renal disease (45–48).

We applied a cyclical mechanical stretch on the evidence that, in the normal glomerulus capillary, pressure is pulsatile (49) and that, in situations such as diabetes, this pulsatility may be enhanced because of defective autoregulation (44–51).

The observation that stretch-induced VPF is mediated via PKC activation has particular relevance for diabetes because PKC is also activated by high glucose concentration (52). Moreover, an oral inhibitor of the β₂ isoform of PKC significantly prevents the development of proteinuria in the experimental diabetic rat (53).

In conclusion, our data help to explain the molecular basis for the injurious role of the altered hemodynamic forces that lead to glomerular hyperpermeability. The knowledge of the signaling pathways that mediate VPF production in human mesangial cells may provide novel therapeutic targets for molecular pharmacology.