In high glucose (HG), mesangial cells (MCs) lose their contractile response to endothelin-1 (ET-1) coincidently with filamentous (F)-actin disassembly. We postulated that these MC phenotypic changes are mediated by altered protein kinase C (PKC) isozyme activity, myosin light chain (MLC20) phosphorylation, or Ca2+ signaling. MCs were growth arrested for 24 h in 0.5% fetal bovine serum (FBS)-DMEM in 5.6 (normal glucose; NG) or 30 mM glucose (high glucose; HG). In HG, the planar area was reduced [2,608 ± 135 vs. 3,952 ± 225 (SE) μm2 in NG, P < 0.01, n = 31] with no contractile response to 0.1 μM ET-1. Mannitol did not affect cell size or ET-1 response. Confocal imaging of fluo 3- loaded cells revealed that the peak intensity of ET-1-induced Ca2+signaling was not altered in HG vs. NG. Immunoblotting of phosphorylated MLC20 showed that HG increased mono- and decreased unphosphorylated MLC20 (42 ± 16 and 49 ± 15 vs. 13 ± 3 and 80 ± 4% of total in NG,P < 0.05, n = 3), but the peak phosphorylation responses to ET-1 were identical in NG and HG. ET-1 stimulated translocation of PKC-δ and -ε from cytosolic to membrane and particulate fractions identically in NG and HG but did not cause PKC-ζ translocation. In HG, membrane accumulation of PKC-ζ was observed. Membrane PKC-ζ activity measured by immunoprecipitation and32P phosphorylation of PKC-ε pseudosubstrate peptide was 190 ± 18% of NG (P < 0.01,n = 4), which was completely inhibited by pretreatment with a myristoylated peptide inhibitor (ZI). In HG, pretreatment with ZI for 24 h restored normal MC size and contractile and F-actin disassembly responses to ET-1. In conclusion, in HG, decreased MC size is due to decreased F-actin assembly, and loss of contractile response to ET-1 occurs in the presence of normal Ca2+ signaling and normal MLC20 phosphorylation. In HG, altered F-actin and contractile functions in MCs are mediated by PKC-ζ.
- calcium signaling
- myosin light chain phosphorylation
- protein kinase C-ζ
endothelin-1(ET-1) is a potent vasoconstrictor that is synthesized by glomerular endothelial cells (33), epithelial cells (45), and mesangial cells (40). ET-1-stimulated mesangial cell contraction can regulate glomerular capillary surface area and filtration rate (3). However, mesangial cells cultured in high glucose (8, 17) and glomeruli isolated from streptozotocin (STZ)-diabetic rats fail to contract in response to ANG II, (24), ET-1, or raised intracellular Ca2+ (17). In diabetes, high glucose-induced loss of preglomerular afferent arteriolar contractile response to vasoconstrictors leads to elevated intraglomerular pressure, accelerating the progression of glomerulosclerosis (58).
The precise mechanisms by which high glucose alters the phenotype of mesangial cell and vascular smooth muscle cell (VSMC) responsiveness are not known. Reduced Ca2+ signaling mediated by either high glucose-induced protein kinase C (PKC) inhibition of inositol 1,4,5-trisphosphate (IP3) release (34) or via high glucose-induced inhibition of receptor-operated channels (35) have been proposed. Work from our laboratory has demonstrated that isolated glomeruli and mesangial cells cultured in 25 mM (high) glucose display normal Ca2+ signaling to ET-1, ANG II, and arginine vasopressin (AVP) as measured by45Ca2+ efflux and fura 2 spectrofluorimetry, respectively (53). The normal IP3 response observed in high glucose is attributed to upregulation ofmyo-inositol transport, which prevents mesangial cellmyo-inositol depletion (4).
The initial Ca2+ transient induced by vasoconstrictor peptides in VSMCs and mesangial cells is accompanied by Ca2+/calmodulin (CaM)-dependent activation of myosin light chain kinase (MLCK) and the phosphorylation of the regulatory 20-kDa myosin light chain (MLC20) at serine-19 and threonine-18. This leads to the activation of the actin-activated myosin ATPase, interaction of filamentous (F)-actin and myosin, and cellular contraction (42, 46). PKC, a ubiquitous serine-threonine protein kinase, is also implicated in the regulation of contraction in both smooth muscle and mesangial cells (8, 9, 50, 52). Activation of mesangial cell PKC isozymes by hyperglycemia in diabetes represents an important pathway potentially contributing to altered cytoskeletal responsiveness (2, 8, 30). PKC can directly phosphorylate MLC20 at serine-1, serine-2, and threonine-9 without activating contraction (52) and indirectly stimulates MLC20 phosphorylation at serine-19 and threonine-18 by inhibiting myosin light chain phosphatase (MLC-PP). PKC can also stimulate contraction, independent of MLC20phosphorylation, by phosphorylating the actin regulatory proteins calponin and caldesmon (15, 16).
Although we have previously shown that ET-1-stimulated mesangial cell contraction involves the activation of PKC-δ, -ε, and, to a lesser extent, -α in normal glucose (9), little is known about PKC isozyme regulation of mesangial cell MLC20phosphorylation. In high glucose, mesangial cell PKC isozyme expression patterns, subcellular distribution, and activity are altered (1,2, 20, 21, 23, 55). In mesangial cells, Kikkawa et al. (23) and, recently, Amiri and Garcia (1) demonstrated membrane translocation of PKC-ζ and -α after 72 and 120 h of high glucose. Therefore, high glucose-induced activation of selective PKC isozymes that mediate cytoskeletal restructuring may be implicated in the mechanism of reduced contractility. We have observed that mesangial cells in high glucose for 48 h demonstrate F-actin disassembly, which is reversed by inhibition of the polyol pathway and appears to be mediated through a PKC-dependent mechanism (61). Loss of mesangial cell contractility to ET-1 is also restored with aldose reductase inhibition (8).
In this study we postulated that high glucose may alter mesangial cell Ca2+ signaling, MLC20 phosphorylation, or PKC isozyme activity, mediating cytoskeletal dysfunction and hypocontractility. We examined the effects of high glucose on ET-1-stimulated mesangial cell planar area reduction (contraction), Ca2+ signaling, MLC20phosphorylation, PKC isozyme distribution, and F-actin disassembly. The use of a myristoylated PKC-ζ peptide inhibitor (myr-RRGARRWRK; ZI) was used to determine the role of the PKC-ζ isozyme in the regulation of contraction and F-actin assembly in high glucose.
MATERIALS AND METHODS
DMEM, penicillin, streptomycin, and trypsin were purchased from GIBCO BRL Life Technologies (Burlington, ON). Fetal bovine serum (FBS) was purchased from Wisent (St. Bruno, PQ). 12 Phorbol-13 myristate (PMA), ET-1, leupeptin, pepstatin A, aprotinin, benzamidine, Tween 20, sodium orthovanadate, dithiothreitol (DTT), and polyclonal anti-PKC-α, -δ, -ε, and -γ antibodies were purchased from Sigma (St. Louis, MO). Polyclonal anti-PKC-βI and -ζ antibodies and protein G-Sepharose were from Santa Cruz Biotechnology (Santa Cruz, CA). Rhodamine-phalloidin and FITC-DNase 1 were from Molecular Probes (Eugene, OR). Horseradish peroxidase (HRP)-labeled goat anti-rabbit and goat anti-mouse secondary antibodies were from Bio-Rad (Hercules, CA) and Jackson Immunoresearch Laboratories (West Grove, PA), respectively. Polyclonal anti-MLC20 antibody and a monoclonal anti-phospho-Ser19-MLC20antibody were kindly supplied by Drs. Subah Packer (Indiana University, Indianapolis, IN) and Yasuharu Sasaki, (Asahi Chemical, Shizuoka, Japan), respectively. Dr. Michael P. Walsh (University of Calgary, Calgary, AB) generously provided purified chicken gizzard MLC20. The myristoylated PKC-ζ inhibitor ZI was synthesized by the Hospital for Sick Children Peptide Synthesis Laboratory (Toronto, ON). The purity of the peptide was >95% as determined by high-pressure liquid chromatography and mass spectroscopy. All other chemicals were of analytic or electrophoresis grade and purchased from BDH (Toronto, ON).
Mesangial Cell Planar Area Measurements
Mesangial cells (passages T5–T10) were cultured as previously described (18) in 20% FBS-DMEM in 5.6 mM (normal) glucose, pH 7.4, supplemented with 100 U/ml penicillin, 100 μg/ml streptomycin, and 10 mM HEPES.
Mesangial cells (100,000 cells/dish) were directly growth arrested for 24 h on 35-mm plastic dishes (Falcon, Becton-Dickinson Laboratories, Lincoln Park, NJ) in either 2 ml 0.5% FBS-DMEM with 5.6 or 30 mM (high) glucose. The dishes were transferred to the heated stage of an inverted phase-contrast light microscope (Bausch and Lomb, New York, NY) and maintained in 5% CO2 in air at 30°C. The images were captured by a Hitachi KP-113 solid-state television camera (Hitachi Denshi, Tokyo, Japan). After visualization, the image was digitized to represent time 0. To downregulate PKC in normal glucose, cells were pretreated for 24 h with 0.1 μM PMA. To inhibit PKC-ζ, cells were pretreated for 24 h with 10 μM of ZI. This peptide has been shown to inhibit PKC-ζ activity inXenopus laevis oocytes (10) and in pancreatic islet β-cells (47). ET-1 was added directly to the medium to achieve a final concentration of 0.1 μM for cells in normal or high glucose in the absence or presence of ZI. To control for osmolarity, cells were growth arrested in 0.5% FBS in normal glucose supplemented with 24.4 mM mannitol. The same group of cells (n = 10–20/group) was digitized serially for the next 60 min. Images were captured on a 486 DX PC and digitized with Sigma Scan morphometric software (Jandel Scientific, San Rafael, CA). Only those cells with a clearly defined border perimeter were used for planar surface area measurement. Changes in individual cell surface planar area were calculated for all of the cells at each time point.
Confocal Imaging of Intracellular Ca2+
Cells were sparsely plated on glass coverslips to minimize overlap of individual cells. With the use of our previously published methods (54), growth-arrested cells were loaded with 2.5 μM fluo 3 in DMEM (containing CaCl2) with 0.02% Pluronic F-127 and 1 mg/ml BSA for 30 min at 37°C. The cells were washed once in DMEM without fluo 3 and then incubated in the dark in DMEM for 30 min at room temperature (RT). The coverslip was mounted in a chamber on the stage of a Zeiss confocal microscope (LSM 410, Düsseldorf, Germany), and the cells were imaged at 30°C in response to 0.1 μM ET-1. An argon laser was focused with an inverted objective lens (Axiovert 100, ×10), and the pixel resolution was set at 512 × 512 with a gray-scale level of 0 (minimum) to 255 (maximum). A section through the maximum diameter of each cell was digitized at consecutive time points after the introduction of ET-1. To standardize the fluorescence intensity measurement between experiments, the image contrast and brightness levels were adjusted optimally and then kept constant for the experiments in the study. The pinhole (size = 25), scanning time (0.546 s), zoom (1), and magnification of the confocal scanning system were identical for each experiment. The digitized confocal images were analyzed using National Institutes of Health (NIH) Image software (version 1.62) for the Macintosh (NIH, Bethesda, MD), without enhancement.
Analysis of MLC20 Phosphorylation
Phosphorylation of the 20-kDa regulatory MLC20 was measured using a modified method from Persechini et al. (36). The separation of unphosphorylated, monophosphorylated, and diphosphorylated forms of MLC20 was accomplished with glycerol PAGE followed by electrophoretic transfer of protein to nitrocellulose membranes and immunoblot analysis.
Mesangial cells were growth arrested for 24 h on 100-mm plastic dishes in 0.5% FBS-DMEM containing 5.6 or 30 mM glucose. Cells were treated with either 0.1 μM ET-1, PMA, or ionomycin for 0, 2, 5, 10, 20, and 40 min at 30°C. To inhibit PKC-dependent MLC20phosphorylation, cells were exposed to 0.1 μM PMA for 24 h to downregulate PKC. Trypsin-detached cells served as a negative control. After treatment, cells were rinsed with PBS and cellular proteins were precipitated in 600 μl of 10% trichloroacetic acid containing 2 mM DTT at 4°C. Precipitates were scraped from the plates and centrifuged at 15,000 g for 15 min. The supernatant fraction was decanted, and the pellet was extracted three times in 6 ml of diethyl ether containing 2 mM DTT at 4°C for 2 min, followed by air drying at room temperature (RT) for 30 min to remove residual ether. The pellet was then dissolved in 100 μl of sample buffer containing 8 M urea, 20 mM Tris base, 23 mM glycine, and 2 mM DTT, pH 8.6. After centrifugation and protein determination, 100 μg of the urea-solubilized samples were electrophoresed at 400 V for 1.5 h at 10°C in 1.0-mm minigels containing 10% polyacrylamide, 0.5% bisacrylamide, 40% glycerol, 20 mM Tris base, and 23 mM glycine. The reservoir buffer contained (in mM) 20 Tris, 23 glycine, 1 sodium thioglycolate, and 1 DTT. Separated proteins were then electroblotted to 0.22-μm nitrocellulose (Bio-Rad) membranes at 24 V for 3 h, employing a 20 mM Tris-23 mM glycine-20% methanol transfer buffer, pH 7.6.
After transfer, membranes were blocked in 5% skim milk in Tris-buffered saline plus Tween 20 (TTBS) and then incubated for 1 h at RT in a rabbit polyclonal anti-MLC20 antibody (1:1,000) in 5% skim milk in TTBS. For chemiluminescent detection, the membranes were incubated with an HRP-conjugated anti-rabbit secondary antibody, dipped into luminol substrate solution for 1 min, and developed on film, as described above. Densitometry was performed using NIH Image 1.62 analysis software, and MLC20phosphorylation was expressed as the percentage of MLC20 in un-, mono-, or diphosphorylated form.
Immunoblotting of Ser19-MLC20 and the ETA Receptor
In a separate series of experiments, phosphorylation of Ser19-MLC20 and ETA receptor expression were analyzed in total mesangial cell lysates. Cells were growth arrested in six-well plates for 24 h in 0.5% FBS-DMEM in normal or high glucose. To determine if high glucose altered ETA receptor expression, cells were growth arrested in 0.5% FBS-DMEM in normal or high glucose for 1, 2, 3, 5, and 7 days and then lysed in 150 μl of 2× SDS sample buffer at 100°C. After protein determination, 15 μg of cellular protein were subjected to SDS-PAGE using 15 and 10% minigels for Ser19-MLC20 and the ETA receptor, respectively. The proteins were transferred to polyvinylidene difluoride (PVDF; Millipore, Bedford, MA) membranes, blocked as described above, and probed for 1 h at RT (1:1,000 in 5% skim milk in TTBS) with either a monoclonal anti-Ser19-MLC20 antibody or a sheep polyclonal anti-ETA receptor antibody (US Biologicals, Swampscott, MA). Chemiluminescent detection and densitometry were performed as described above.
Cell Fractionation and PKC Immunoblots
PKC isozymes were probed in cytosolic, membrane, and particulate fractions of 24-h-growth-arrested mesangial cells in normal and high glucose in the absence and presence of 0.1 μM ET for 10 min at 30°C. As described previously (9, 21), cellular fractions were obtained using sequential ultracentrifugation in the absence (cytosolic) or presence of the detergents 1% Triton X-100 (membrane) and 10% SDS (particulate). Total PKC isozyme expression was determined in cells lysed in 2× SDS sample buffer at 100°C. Protein content was measured using a Bio-Rad Dc Lowry protein assay (Bio-Rad). Fifteen microliters of protein were subjected to SDS-PAGE on 10% gels followed by overnight transfer onto PVDF membranes. Membranes were then exposed to polyclonal anti PKC-α, -βI, -δ, -ε, -γ, or -ζ antibodies for 1 h at 1:40,000 dilution for PKC-α; 1:15,000 for PKC-ζ; 1:10,000 for PKC-δ and -ε; 1:8,000 for PKC-γ; and 1:1,000 for PKC-βI. This was followed by a 20-min incubation with a HRP-conjugated affinity-purified goat anti-rabbit IgG antibody (Jackson Immunochemicals) diluted 1:5,000. The secondary antibody was detected by chemiluminescence (Kirkegaard & Perry Labs, Gaithersburg, MD), and the membranes were developed on Kodak X-Omat Blue film (Eastman Kodak, Rochester, NY). Densitometry was performed using NIH Image 1.62 analysis software. The specificity of primary antibody interaction with the sample was eliminated in the presence of the peptide that was used to generate the antibody.
Confocal Fluorescence Imaging
Mesangial cells were cultured on glass coverslips under conditions identical to those described above. The cells were fixed with 3.7% formaldehyde for 15 min at RT followed by plasma membrane and nuclear membrane permeabilization with 100% methanol at −20°C for 10 min, as previously published (61). After being washed three times with PBS, the cell proteins were blocked with 1% goat serum plus 0.1% BSA in PBS for 60 min at RT. For Ser19-MLC20, cells were permeabilized with 0.1% Triton X-100 for 30 min. Rabbit polyclonal anti-PKC-α, -δ, -ε, -ζ, and -γ and monoclonal anti-Ser19-MLC20 were diluted to 1:100 (1:10 PKC-βI) in blocking solution and added to each coverslip for 60 min at 37°C. After being washed three times with PBS, FITC-conjugated anti-rabbit or anti-mouse secondary antibodies were diluted 1:160 in blocking solution and added for 60 min at 37°C. The coverslips were mounted on glass slides with Aqua-Poly-mount (Polysciences, Warrington, PA) and SlowFade antifade reagent. The cells were visualized with an oil-immersion, inverted objective lens (Axiovert 100, ×63), and the image pixel resolution was 512 × 512 with a gray-scale level of 0 (minimum) to 255 (maximum) intensity. To standardize the fluorescence intensity for all the experimental preparations, the confocal image contrast and brightness levels were adjusted optimally for each PKC isozyme and then kept constant for all image analysis.
Mesangial cell F- and G-actin were simultaneously labeled with the fluorescent probes rhodamine-phalloidin (F-actin) and FITC-DNase 1 (G-actin) according to our previously published methods (60). Cells were cultured on glass coverslips under experimental conditions described above. After being washed in PBS (4°C), the cells were fixed in 3.7% formaldehyde in PBS for 15 min at RT and permeabilized with 0.1% Triton X-100 for 10 min. After additional washes in PBS, the cells were incubated simultaneously in the dark with 200 μl of 0.165 μM rhodamine-phalloidin and 0.3 μM FITC-DNase 1 for 20 min. Coverslips were mounted on glass slides and imaged with a dual-channel Zeiss LSM 410 confocal laser-scanning microscope, as described above (Confocal Imaging of Intracellular Ca 2+).
Measurement of PKC-ζ and PKC-δ Activity
PKC-ζ activity was determined in membrane fractions using immunoprecipitation and 32P phosphosphorylation of Ser159-PKC-ε pseudosubstrate peptide (44). To inhibit PKC-ζ, cells were pretreated with 10 μM ZI for 24 h. Cells cultured in normal glucose supplemented with 24.4 mM mannitol for 24 h served as an osmotic control. Membrane fractions were obtained as previously described (9, 21). The supernatant was precleared by incubation with protein G-Sepharose (100 μl/ml lysate) at 4°C for 20 min on a rocker, followed by centrifugation at 4°C at 15,000 g for 10 min. Samples of the lysate were removed for protein assay (Bio-Rad, Dcprotein assay), and 2 μg of rabbit polyclonal anti-PKC-ζ (Santa Cruz Biotechnology) were added to 300 μg of total cellular protein. The samples were allowed to rock at 4°C for 3 h, followed by immunoprecipitation with 40 μl of protein G-Sepharose for 1 h. The immunocomplexes were centrifuged at 15,000 gfor 10 min at 4°C and then washed three times with ice-cold lysis buffer, followed by three washes in kinase buffer. The kinase activity of immunoprecipitated PKC-ζ was assayed in 40 μl of kinase buffer that contained (in mM) 50 Tris, pH 7.5, 5 MgCl2, 2 NaF, 0.1 Na3VO4, and 0.1 Na2P2O7 · 10 H2O, as well as 5 μg of Ser159-PKC-149–164-ε pseudosubstrate peptide (Biomol Research Laboratories, Plymouth Meeting, PA), 40 μM ATP, 5 μCi of [γ-32P]ATP, and 10 nM microcystin. After 15 min at 30°C, the reaction was terminated by the addition of 25 μl of ice-cold quenching buffer containing 0.1 mM ATP and 100 mM EDTA, pH 8.0. Fifty microliters of the reaction mixture were spotted on P81 filter disks and washed four times in 75 mM phosphoric acid followed by a final wash in 80% ethanol. The disks were air dried and placed in 10 ml of Ready Protein scintillant (Beckman Industries); then,32P radioactivity was counted in a Beckman LSC5 scintillation counter.
PKC-δ activity was determined in total cell lysates using immunoprecipitation and 32P phosphorylation of Ser25-PKC-α pseudosubstrate peptide. To determine the specificity of the PKC-ζ inhibitor, cells were pretreated with 10 μM ZI for 24 h. Cells were lysed in 500 μl of lysis buffer containing (in mM) 150 NaCl, 25 HEPES, 1 EGTA, 2 EDTA, 10 NaF, 50 β-glycerophosphate, 1 Na3VO4, 2.5 bensamidine, and 1 phenylmethylsulfonyl fluoride, as well as 10 μg/ml leupeptin, pepstatin, and aprotinin, 10 nM microcystin, and 1% Triton-100, pH 7.5. Lysates were triturated five times through a 25-gauge needle, placed on ice for 30 min, and centrifuged at 15,000g for 5 min to remove cellular debris. Three hundred micrograms of total cellular protein were immunoprecipitated with 2 μg of rabbit polyclonal anti-PKC-δ (Santa Cruz Biotechnology) for 3 h, followed by the addition of 40 μl protein G-Sepharose for 1 h. The immunocomplexes were centrifuged, washed in lysis buffer, followed by three washes in kinase buffer. Immunoprecipitated PKC-δ was assayed in 40 μl of kinase buffer that contained 5 μg of Ser25-PKC-19–31-α pseudosubstrate peptide (GIBCO BRL Life Technologies) but was otherwise identical to that used for determining PKC-ζ activity. The reaction was terminated by the addition of quenching buffer, followed by spotting on P81 filter disks, subsequent washing in 75 mM phosphoric acid, and scintillation counting, as described above. Data for PKC-ζ and PKC-δ activities are expressed as the percentage of those observed in normal glucose.
All results are expressed as means± SE. Statistical analysis was performed using InStat 2.01 statistics software (Graph Pad, Sacramento, CA). The means of two groups were compared using an unpaired Student's t-test. The means of three or more groups were compared by one-way ANOVA with Bonferroni's multiple comparison. To compare multiple groups with basal normal glucose, Dunnett's multiple comparison was utilized. Differences described as significant are P < 0.05 unless stated otherwise.
Effects of High Glucose and Chronic Phorbol Ester on Mesangial Cell Planar Area
Figure 1 demonstrates the change in planar surface area of mesangial cells growth arrested for 24 h in 5.6 (normal) and 30 mM (high) glucose and stimulated with 0.1 μM ET-1 at 30°C. Basal planar surface area in cells cultured in 5.6 mM glucose was 3,952 ± 225 μm2 (n = 53 cells). Cells incubated in 30 mM glucose displayed a smaller basal planar area of 2,608 ± 135 μm2 (n = 31 cells, P < 0.01 vs. normal glucose at time 0). This was not due to osmolarity as cells cultured in normal glucose supplemented with 24.4 mM mannitol were no different from cells in normal glucose alone (3,424 ± 238 μm2,n = 28 cells).
In normal glucose, administration of ET-1 for 60 min decreased planar area to 77 ± 3% of the original surface area (P< 0.05 vs. normal glucose at time 0) (Fig. 1 B). Mannitol in normal glucose did not alter mesangial cell responsiveness to ET-1 at 60 min (81 ± 2% of basal area).
Effects of High Glucose on ETA Receptor Expression
To determine whether the loss of ET-1 responsiveness in high glucose was due to reduced ETA receptor expression, mesangial cells were cultured in 0.5% FBS-DMEM in 5.6 or 30 mM glucose for up to 7 days, and total cell lysates were examined by immunoblot for the presence of the ETA receptor. In the representative immunoblot shown in Fig. 2, the antibody generated against a 39-kDa cytosolic fragment of the ETAreceptor reacted with cellular proteins of 64-, 50-, and 39-kDa molecular mass. The intensities of the three dominant bands were not altered by high glucose. Equal protein loading was confirmed by staining the PVDF membrane with Ponceau S.
Confocal Fluorescence Imaging of ET-1-Stimulated Intracellular Ca2+ Signaling
In normal glucose, fluo 3-loaded cells displayed perinuclear and cytosolic Ca2+ fluorescence staining at time 0(Fig. 3 A). The administration of 0.1 μM ET-1 stimulated an increase in cytosolic and nuclear Ca2+ beginning at 20 s and peaking at 40–60 s. Fluorescence intensity returned to basal levels at 80–100 s. This rapid Ca2+ response to ET-1 was seen in >90% of cells examined (n = 44 cells in 3 independent experiments).
In Fig. 3 B at time 0, cells in high glucose displayed a basal perinuclear and cytosolic staining pattern that was no different from that seen in normal glucose. However, an increase in cytosolic and nuclear Ca2+ did not occur until 40 s after the addition of 0.1 μM ET-1 and reached a peak at 70–80 s. Fluorescence intensity returned to basal levels at 120–200 s. Although the Ca2+ response was delayed compared with that for normal glucose, peak fluorescence pixel intensity (130 ± 13 in normal glucose vs. 117 ± 12 in high glucose) and the number of cells responding to ET-1 (>90%, n = 44 cells) were not different.
Phosphorylation of MLC20
Effects of ET-1, phorbol ester, and ionomycin on MLC20phosphorylation.
The change in ET-1-induced mesangial cell planar surface area was used as an indirect measurement of contraction. Glycerol-urea PAGE was used to measure MLC20 phosphorylation, the key regulatory step preceding smooth muscle cell contraction. As shown in Fig.4, at time 0 the majority of MLC20 was unphosphorylated [81 ± 5 (SE)%,n = 3], with 13 ± 4 and 6 ± 2% mono- and diphosphorylated, respectively. In comparison, trypsinized suspended cells contained 97 ± 5 un- and 3 ± 1% monophosphorylated MLC20. ET-1 at 2 min significantly stimulated MLC20 phosphorylation. Mono- and diphosphorylated MLC20 increased to 60 ± 9 and 24 ± 11% of total, respectively (P < 0.05 vs. time 0,n = 3), whereas unphosphorylated MLC20decreased to 16 ± 7% of total (P < 0.05 vs.time 0, n = 3). The effect of ET-1 was sustained at 40 min, with 29 ± 7, 40 ± 4, and 31 ± 5% un-, mono-, and diphosphorylated, respectively (P< 0.05 vs. time 0). Downregulation of PKC with 24-h PMA did not alter time 0 MLC20 phosphorylation (72 ± 4, 22 ± 10, 6 ± 2% un-, mono-, and diphosphorylated, respectively) and did not affect ET-1-stimulated phosphorylation at 10 min (22 ± 6, 40 ± 10, and 38 ± 13% un-, mono-, diphosphorylated, respectively, vs. 23 ± 8, 45 ± 2, and 32 ± 8% with no PMA).
As shown in the representative immunoblots, 0.1 μM acute PMA did not stimulate phosphorylation of MLC20, whereas 0.1 μM ionomycin stimulated phosphorylation to lower levels than seen with ET-1.
Effects of high glucose on ET-1-stimulated MLC20phosphorylation.
Densitometric analysis of ET-1-stimulated MLC20phosphorylation in high glucose is shown in Fig.5. At time 0, high glucose significantly increased monophosphorylated MLC20, whereas it decreased unphosphorylated MLC20 (42 ± 16 and 49 ± 15% of total, respectively, vs. 13 ± 3 and 80 ± 4% of total in normal glucose at time 0, P< 0.05, n = 3). ET-1 at 2 min significantly stimulated diphosphorylation of MLC20. Mono- and diphosphorylated MLC20 increased to 62 ± 20 and 28 ± 11% of total, respectively (P < 0.05 vs. time 0,n = 3), whereas unphosphorylated MLC20decreased to 10 ± 1% of total (P < 0.05 vs.time 0, n = 3). The effect of ET-1 on diphosphorylated MLC20 was sustained at 40 min, with 34 ± 20% of total MLC20 diphosphorylated (P < 0.05 vs. time 0). Downregulation of PKC with 24-h PMA did not alter basal MLC20 phosphorylation in high glucose (54 ± 20, 37 ± 19, and 9 ± 4% un-, mono-, di-phosphorylated, respectively) and did not affect ET-1-stimulated phosphorylation at 10 min (20 ± 10, 39 ± 13, and 41 ± 21% un-, mono-, diphosphorylated, respectively, vs. 24 ± 11, 40 ± 6, and 36 ± 15% with no PMA).
Effects of high glucose on phosphorylated Ser19-MLC20.
The use of glycerol-urea PAGE effectively separated the un-, mono-, and diphosphorylated MLC20. However, this method cannot distinguish which serine or threonine is phosphorylated. To determine whether Ser19-MLC20, the dominant site phosphorylated by MLCK and Rho kinase, was the monophosphorylated MLC20, we immunoblotted total cell lysates with a monoclonal phospho-specific Ser19-MLC20antibody. As shown in Fig. 6 A, basal levels of phospho-Ser19-MLC20 were not altered by high glucose (92 ± 8% of normal glucose,n = 8).
Confocal imaging of ET-1-stimulated phosphorylated Ser19-MLC20.
Basal phospho-Ser19-MLC20 cellular distribution in mesangial cells in normal glucose is shown in Fig. 6 B. Fluorescence was localized primarily to the cell membrane along with a fibrillar, cytosolic staining pattern. ET-1 administration failed to significantly alter the staining pattern. In high glucose, the intensity and staining pattern were not different from that in normal glucose. In high glucose, ET-1 also failed to change the distribution and fluorescence intensity of the membrane staining.
Mesangial Cell PKC Isozymes
Effects of high glucose on basal and ET-1-stimulated PKC isozyme distribution.
In normal glucose, ET-1 stimulates the translocation of PKC-α, -δ, and -ε isozymes, but not PKC-ζ, from cytosolic to membrane and particulate fractions (9). To determine whether high glucose alters ET-1-stimulated PKC isozyme translocation, we immunoblotted cytosolic, membrane, and particulate cellular fractions with PKC isozyme-specific antibodies. As shown in Fig.7, PKC-α, -δ, and -ε were present in all three fractions in both normal and high glucose and translocated in response to ET-1 from cytosolic to membrane and particulate fractions to a similar extent. In response to ET-1 in normal vs. high glucose, PKC-α distribution in membrane and particulate fractions, respectively, was 87 ± 17 and 133 ± 10% of that in basal normal glucose (P < 0.05) vs. in high glucose 92 ± 20 and 142 ± 6% of basal normal glucose levels (P < 0.05, n = 4). PKC-δ distribution in membrane and particulate fractions, respectively, was 300 ± 76 and 134 ± 3% of that in basal normal glucose (P < 0.01 and P < 0.05, respectively) vs. in high glucose 287 ± 74 and 137 ± 24% of that in basal normal glucose (P < 0.01 and P< 0.05, respectively). PKC-ε distribution in membrane and particulate fractions, respectively, was 205 ± 42 and 133 ± 16% of that in basal normal glucose (P < 0.01 andP < 0.05, respectively) vs. in high glucose 184 ± 52 and 190 ± 56% of that in basal normal glucose (P < 0.01 and P < 0.05, respectively,n = 4). Cells treated with 0.1 μM PMA for 10 min served as a positive control and demonstrated the disappearance of PKC-α, -δ, and -ε from the cytosolic fraction (12 ± 9, 2 ± 1, and 4 ± 1% of distribution in basal normal glucose, respectively, P < 0.01, n = 4). This was accompanied by enhanced recovery of isozymes in membrane (289 ± 97, 513 ± 90, and 267 ± 43% of basal normal glucose, respectively, P < 0.01, n = 4) and particulate fractions (376 ± 46, 205 ± 33, and 244 ± 70% of basal normal glucose, respectively, P < 0.01,n = 4).
Subcellular distribution of PKC-ζ and -βI are also shown in Fig. 7. Both isozymes were present in all three fractions but did not respond to ET-1 in either normal or high glucose. PKC-ζ was PMA unresponsive, whereas PKC-βI demonstrated translocation from cytosolic to membrane and particulate fractions (2.6 ± 1, 295 ± 117, and 322 ± 111% of that in cytosolic and membrane and particulate fractions with basal normal glucose, respectively, P < 0.01, n = 4). High glucose alone enhanced recovery of PKC-ζ in the membrane fraction (294 ± 106% of basal normal glucose, P< 0.01, n = 4) with no change in either the cytosolic or particulate fractions (96 ± 7 and 91 ± 7%, respectively, of basal normal glucose, n = 4).
Effects of high glucose on total PKC isozyme content.
In high glucose, it was postulated that the enhanced recovery of PKC-ζ in the membrane might be attributed to enhanced total PKC expression. Therefore, we determined the effects of 24-h high glucose on total PKC expression by immunoblotting total cell lysates (Fig.8). PKC-α, -δ, -ε, -ζ, and -βI were expressed as a single band in mesangial cells, but PKC-γ was not expressed. High glucose stimulated a significant 2.7-fold increase in PKC-δ (272 ± 75%, P < 0.05, vs. normal glucose, n = 6), whereas PKC-α (125 ± 12%, n = 6), -ε (90 ± 10%,n = 6), -ζ (102 ± 2%, n = 6), and -βI (90 ± 12%, n = 6) remained unchanged.
PKC-ζ and High Glucose-Induced Loss of Mesangial Cell Contraction to ET-1
Effects of high glucose on immunoprecipitated PKC-ζ and PKC-δ activity.
Figure 9 shows the effects of high glucose on membrane PKC-ζ activity in the absence and presence of the ZI. High glucose increased PKC-ζ activity to 190 ± 18% of normal glucose (P < 0.01, n = 4). Twenty-four-hour pretreatment with 10 μM ZI in either normal or high glucose significantly inhibited activity to 45 ± 10 and 73 ± 4% of normal glucose, respectively (P < 0.01,n = 4). Mannitol did not significantly affect PKC-ζ activity (99 ± 11% of normal glucose, n = 4).
In total cell lysates, high glucose increased PKC-δ activity to 241 ± 41% of normal glucose (P < 0.01,n = 3). Twenty-four-hour pretreatment with 10 μM ZI in either normal or high glucose did not alter total PKC-δ activity (110 ± 22 and 256 ± 37% of normal glucose, respectively,n = 3). Mannitol did not significantly affect PKC-δ activity (129 ± 12% of normal glucose, n = 4).
Effects of ZI on mesangial cell planar area and ET-1-stimulated contraction.
Basal planar surface area in cells cultured in normal glucose was 3,952 ± 225 μm2 (n = 53 cells) (Fig. 10 A). In comparison, cells incubated in 30 mM glucose displayed a smaller basal planar area of 2,608 ± 135 μm2 (n = 31 cells,P < 0.01 vs. normal glucose at time 0). This was not due to osmolarity as cells cultured in normal glucose supplemented with 24.4 mM mannitol were no different from cells in normal glucose (3,624 ± 238 μm2, n= 28 cells). Pretreatment of cells in high glucose with 10 μM ZI for 24 h restored the area to values seen in normal glucose (3,875 ± 213 μm2, n = 40 cells). In normal glucose, pretreatment with ZI did not alter basal planar area (3,871 ± 277 μm2, n = 23 cells). ZI pretreatment in normal glucose did not alter MC responsiveness to ET-1 at 60 min (69 ± 5 vs. 77 ± 3% of basal area in normal glucose) as shown in Fig. 10 B. Cells in high glucose did not respond to ET-1. In high glucose, pretreatment with ZI restored the ET-1 response, decreasing planar area to 80 ± 2% of basal area, similar to the change in normal glucose.
Effects of ZI on mesangial cell F-/G-actin in high glucose.
In Table 1, the average pixel intensity/cell values for F-actin and G-actin are presented. F-/G-actin ratio was calculated for each cell and the mean ± SE is reported for each group. In high glucose, the F-/G-actin ratio was significantly reduced to 1.5 ± 0.2 (P < 0.05 vs. normal glucose, n = 44 cells) from a value of 1.9 ± 0.1 in normal glucose. The addition of 0.1 μM ET-1 in normal glucose resulted in an F-/G-actin ratio that was not different from that seen in the basal high glucose state (1.5 ± 0.1, n = 44). In high glucose, ET-1 did not alter the mesangial cell F-/G-actin ratio (1.4 ± 0.1, n = 36). Mannitol (normal glucose+24.4 mM mannitol for 24 h) had no effect on the F- or G-actin content or F-/G-actin ratio (2.0 ± 0.2, n= 26 cells).
The lack of a contractile response to ET-1 in mesangial cells in high glucose may involve reorganization of the cellular cytoskeleton. In normal glucose, F-actin staining filled the cytoplasm and was organized in a bundled stress fiber pattern (Fig.11). G-actin was homogeneously distributed in the cytosol and followed a granular staining pattern. In high glucose, F-actin intensity was reduced, whereas cytosolic and perinuclear G-actin fluorescence intensity was increased. In normal glucose, ZI pretreatment had no affect on F-/G-actin staining pattern or intensity. However, pretreatment of cells in high glucose with ZI resulted in a staining pattern that was seen in normal glucose.
In this study, we examined the signaling events related to mesangial cell contractile and cytoskeletal dysfunction in high glucose. PKC has been implicated as an underlying mechanism of high glucose-induced VSMC dysfunction (25). In high glucose, lack of mesangial cell contractility in response to ET-1 and F-actin disassembly is not due to altered Ca2+ signaling or phosphorylation of MLC20 but is PKC-ζ mediated.
The observations that after exposure to 24-h high glucose mesangial cells are smaller in size, unresponsive to ET-1 (8), and display PKC-mediated F-actin disassembly (61) resemble our earlier studies in higher passage cells. We reported previously that mesangial cell reduced planar area, responsiveness to ET-1, and F-actin disassembly were reversed by tolrestat, an aldose reductase inhibitor (21). Aldose reductase inhibition of the polyol pathway prevented the de novo synthesis of diacylglycerol (DAG) (8, 29, 49) and thus activation of DAG-sensitive PKC isozymes at 48 h (21). At 48 h, high glucose stimulated the membrane translocation of PKC-δ and -ε (21). In the present study, total PKC-δ but not PKC-ε was increased by 24-h high glucose, whereas PKC-ζ was increased in the membrane fraction only. Although the exact mechanisms of PKC-ζ activation are not known, phosphatidic acid (PA), a precursor of DAG, is reported to activate PKC-ζ (32). In our earlier studies (21), tolrestat, through inhibition of de novo PA synthesis, may have reduced PKC-ζ activity concomitantly with either PKC-δ and/or -ε.
It is possible that the diminished responsiveness of mesangial cells to ET-1 in high glucose may be mediated by reduced endothelin-receptor A (ETA) expression (19). Hargrove et al. (14) demonstrated in cultured rat mesangial cells that high glucose for 24 h upregulates the production of ET-1 and ET-1 mRNA twofold. Increased ET-1 production may cause downregulation and decreased expression of mesangial cell ETA receptors (19). By contrast, glomerular ETA receptor expression was upregulated in glomeruli from alloxan-diabetic rabbits (22), whereas hyperglycemia did not alter kidney ETA receptor expression in humans or in STZ-diabetic rats (41). In our study, 1–7 days of high glucose did not alter mesangial cell ETA receptor expression. This suggests that reduced mesangial cell ETA receptor expression does not likely account for the high glucose-induced contractile dysfunction in response to ET-1.
As observed in Fig. 3, A and B, in high glucose ET-1 stimulated intracellular Ca2+ to similar levels observed in normal glucose. Although peak intensity of fluo 3-labeled release was not different, a time delay in the measured response in high glucose was consistently observed. We have previously shown that under the identical culture conditions used in this study (9), acute PMA does not elicit a Ca2+ signal. This suggests that PKC activation does not directly stimulate Ca2+ signaling in mesangial cells. However, others have shown that prior PKC activation may inhibit signaling. Interaction may be at the level of G protein coupling (57) or the cytoskeleton (37). Disruption of F-actin with cytochalasin D in platelets inhibits Ca2+ entry (39). Treatment with wortmannin, an inhibitor of phosphatidylinositol 3-kinase (PI 3-kinase), also inhibits Ca2+ entry (39). PI 3-kinase is a potent activator of atypical PKC-ζ (59). In our hands, cytochalasin D pretreatment of mesangial cells did not affect either timing or intensity of ET-1-stimulated Ca2+ release (data not shown). This suggests that the time delay observed in Ca2+ signaling in response to ET-1 in high glucose is not attributable to mesangial cell F-actin disassembly. The potential role of PI 3-kinase in high glucose-induced activation of mesangial cell PKC-ζ and stress fiber disassembly was not explored.
Stress fibers are composed not only of actin but also contain myosin, tropomyosin, and α-actinin and require the small GTP-binding protein Rho for their formation (38). Stress fiber assembly is usually preceded by MLC20 phosphorylation, whereas stress fiber disassembly is a consequence of MLC20dephosphorylation (5). In mesangial cells, Kreisberg et al. (28) demonstrated that cAMP-mediated stress fiber disassembly preceded MLC20 dephosphorylation. In our study, basal MLC20 phosphorylation was markedly increased in high glucose, whereas F-actin was disassembled, resembling the effects of agonist stimulation. Fukuda et al. (11) reported increased basal MLC20 phosphorylation in platelets from type 2 diabetic patients, but the organization of actin was not studied. Because MLC20 phosphorylation is a balance between kinase and phosphatase activities (42), it is possible that high glucose may enhance MLCK activity or increase PKC- or Rho-mediated inhibition of phosphatase activity (31, 43). In this study, neither acute nor chronic PMA, in either normal or high glucose, affected basal MLC20 phosphorylation. The lack of MLC20 phosphorylation in response to acute PMA suggests that the PMA-induced mesangial cell contraction observed in our earlier studies (8, 9) must be MLC20 phosphorylation independent. In our earlier study, chronic PMA inhibited contraction by downregulating PKC-α, -δ, and -ε in the cytosolic, membrane, and particulate fractions (9), thus reducing DAG-sensitive PKC isozyme content. PKC isozymes may stimulate contraction by initiating a kinase cascade involving the phosphorylation of the actin regulatory proteins calponin, caldesmon (16), and/or myristoylated alanine-rich C-kinase substrate (48).
In high glucose, ET-1 stimulated phosphorylation of MLC20to levels seen in normal glucose, which was not accompanied by cell contraction. This might be explained by the preferential phosphorylation of PKC-specific sites serine-1, serine-2, or threonine-9 of the MLC20 by ET-1 in high glucose. Phosphorylation of MLC20 at these sites does not result in myosin-ATPase activation and contraction (52). If high glucose enhanced PKC-mediated inhibition of MLC-PP, phosphorylation of Ser19-MLC20 would increase. This was not detected by either immunoblotting or confocal imaging with a Ser19-MLC20 antibody. The effects of high glucose on either Rho expression or activity are not known. Therefore, in high glucose, MLC20 phosphorylation does not appear to correlate with either F-actin disassembly or lack of contraction in response to ET-1. High glucose must be exerting its actions at another cytoskeletal target.
In this study, high glucose for 24-h enhanced membrane accumulation of PKC-ζ. This was much earlier than the 5-day time point reported by others (1, 24). This was not accompanied by a concomitant decrease in cytosolic and/or particulate fraction and may be explained by the fact that the relative abundance of PKC-ζ in these fractions may hinder the detection of small compartmental changes in distribution. Immunocomplexed PKC-ζ membrane activity was significantly increased by high glucose and was normalized by pretreatment with the myristoylated ZI. ZI was specific for the PKC-ζ isozyme, as pretreatment with ZI did not affect immunoprecipitated PKC-δ activity in either normal or high glucose. Although total protein expression and activity of PKC-δ were increased by high glucose, this was not accompanied by changes in fractional distribution. PKC-δ may be activated by PKC-ζ-mediated Ser660 phosphorylation (62). Furthermore, overexpression of constitutively active PKC-ζ results in phosphorylation of PKC-δ coexpressed in HEK-293 cells (62). In our study, high-glucose-induced PKC-ζ-mediated activation of PKC-δ is unlikely, as pretreatment of cells in high glucose with the ZI, which inhibited PKC-ζ activity, did not inhibit PKC-δ activity. In COS-7 cells, H2O2 causes phosphorylation of Tyr502 and Tyr523, which results in prolonged activation of PKC-δ (26). Recent studies by Konishi et al. (27) using phosphorylation-site specific antibodies and mass spectrometric analyses have shown that H2O2-induced PKC-δ activation requires phosphorylation of Tyr311 and is not accompanied by membrane translocation. Mesangial cells cultured in 30 mM glucose for as little as 1 h generate H2O2, as detected by dichlorofluorescein (13), but it is not known whether this is associated with PKC isozyme tyrosine phosphorylation and activation.
The observations that 24-h high glucose-stimulated, PKC-ζ-dependent F-actin disassembly was reversed by ZI, which also reversed cell size and response to ET-1, suggest that the smaller mesangial cell planar area is the result of F-actin disassembly. The mechanism underlying PKC-ζ-mediated disassembly is not known. PKC-ζ can directly associate with the actin cytoskeleton in fibroblasts (12). In NIH-3T3 fibroblasts, atypical PKC-ζ and -λ are reported to mediate Cdc42-mediated stress fiber disassembly (6) and have also been implicated in regulating Ras-mediated stress fiber disassembly (51). In mesangial cells, high glucose for 48 h does not affect the expression of Ras but increases membrane associated Ras activity (56). It is not known whether high glucose-induced Ras activation results in increased PKC-ζ activity and F-actin disassembly. The recent findings of Cortes et al. (7) that perfused glomeruli from 9-mo-old STZ-diabetic rats display F-actin disassembly and loss of stress fibers strongly suggest that the F-actin disassembly we have observed in cultured mesangial cells in high glucose mimic the in vivo findings.
In summary, this study identifies the importance of PKC-ζ in the mediation of high glucose-induced loss of mesangial cell responsiveness to ET-1. Loss of the contractile response to ET-1 is not due to reduced Ca2+ signaling, phosphorylation of MLC20, or altered DAG-sensitive PKC isozyme activation but may be related to PKC-ζ-dependent F-actin disassembly. This cytoskeletal disorganization alters the mesangial cell phenotype and may account for the early effects of high glucose-induced cellular dysfunction.
The authors thank Hong Hua for technical assistance with confocal microscopy. We also thank Dr. C. S. Packer (Indiana University, Indianapolis, IN) and Dr. Yasuharu Sasaki (Asahi Chemical, Ltd., Shizuoka, Japan) for MLC20 and anti-phospho-Ser19-MLC20 antibodies, respectively.
This work was jointly funded by the Juvenile Diabetes Foundation International and Canadian Institutes of Health Research.
Address for reprint requests and other correspondence: C. I. Whiteside, Medical Sciences Bldg., Rm. 7302, 1 King's College Cir., University of Toronto, Toronto, Ontario, Canada M5S 1A8 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
First published August 15, 2001; 10.1152/ajprenal.00055.2001
- Copyright © 2002 the American Physiological Society