| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Pediatric Nephrology Division, University of Michigan, Ann Arbor, Michigan 48109, USA
1Correspondence: Pediatric Nephrology, University of Michigan Medical Center, 8220D MSRB III, Box 0646, 1150 W. Medical Center Dr., Ann Arbor, MI 48109, USA. E-mail: rransom{at}umich.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: actin • stress fibers • nephrotic syndrome • puromycin aminonucleoside
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2
3
4
5)
.
The pathological mechanisms leading to NS are diverse, including immunological processes, biochemical or genetic defects, and hemodynamically induced glomerular injury (6)
. However, common histological features are observed in each setting, suggesting that a shared molecular mechanism mediates the podocyte structural changes characteristic of NS. One of the primary structural alterations shown to be common to both human disease and animal models of NS is the disruption of the podocyte cytoskeleton, with marked disaggregation and redistribution of actin filaments in the podocyte foot processes (7
8
9)
. Actin filaments have been reported to be the major or sole cytoskeletal protein in podocyte foot processes (10
, 11)
, which suggests that the podocyte actin cytoskeleton may play a critical role in the regulation of foot process structure and function in both health and disease.
We previously reported the presence of high concentrations of the small heat shock protein hsp27 in the glomeruli of normal rats, as well as an increase in glomerular hsp27 and phosphorylation after induction of experimental NS with puromycin aminonucleoside (PAN) (12)
. Although the precise mechanism(s) of action of hsp27 is not yet known, it is a small stress protein reported to have important roles in resistance to thermal and metabolic stress, normal growth and differentiation, signal transduction, molecular chaperoning, and protection from apoptosis (13
14
15
16
17
18
19
20)
. Perhaps its best-described function, however, is that of a regulator of actin polymerization. Hsp27 has been reported to be an actin-associated protein (21)
that was shown to inhibit actin polymerization in vitro (22)
and in vivo (23)
, and its ability to inhibit actin polymerization has been correlated with its state of phosphorylation (15
, 24)
.
Based on these findings we hypothesized that hsp27, via regulation of the actin cytoskeleton, has an important role in regulating both normal podocyte structure and the dramatic structural changes in podocytes that occur during NS. To begin to test this hypothesis, we developed a model of low-concentration (<10 µg/ml) and long-term (7 days) treatment of cultured ‘differentiated’ podocytes with the well-described podocyte toxin PAN, followed by quantitative analyses of changes in podocyte cell survival, morphology, cell area, actin filament structure, and total filamentous actin (F-actin) content. This model was then used to quantitatively determine the specific effects of altered podocyte hsp27 levels on the podocyte response to PAN.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
was a kind gift from Dr. Peter Mundel. Proliferating podocyte cells were grown in RPMI 1640 medium containing 10% fetal bovine serum, 100 U/ml penicillin, 100 µg/ml streptomycin (Gibco-BRL, Rockville, MD), and 10 U/ml mouse 
-interferon (Sigma, St. Louis MO) at 33°C with 100% relative humidity and 5% CO2 atmosphere. Cells were induced to differentiate by transfer to 37°C and removal of -interferon from the above medium to deactivate the transgene for a thermosensitive variant of the SV40 T-antigen. All cultured podocytes used in this study were allowed to differentiate without subcultivation for 10–14 days before treatment. Twice as many antisense transfectant cells were seeded in culture dishes (72,000 cells/well in 6-well dish) for differentiation than vector control or sense transfectants (36,000 cells/well) in order to yield differentiated cultures with similar cell confluence at 10–14 days. NIH 3T3 cells (ATCC #CRL-1658, ATCC, Manassas, VA) and murine mesangial cells (MMC; a kind gift from Dr. Hanna Abboud) were grown in Dulbecco’s modified Eagle’s medium (Gibco-BRL) containing either 10% bovine calf serum (NIH-3T3) or fetal bovine serum (MMC) and 100 U/ml penicillin and 100 µg/ml streptomycin at 37°C with 100% relative humidity and 5% CO2 atmosphere. Cells were seeded at 20,000 cells/well of a 6-well culture dish and treated with PAN after 24 h in culture.
Development of stably transfected podocyte clones
Proliferating podocytes were transfected with human hsp27 sense (26)
, rat hsp27 antisense (27)
, and vector control cDNA (pcDNA3.1, Stratagene, La Jolla, CA) constructs using a transfection reagent (Superfect, Quiagen, Valencia, CA) according to the manufacturer’s instructions. Stably transfected cells were selected in serial passages using G418 antibiotic at 400 µg/ml (100% lethal to untransfected podocytes). The expression vectors contained the entire protein coding region of human hsp27 placed in the sense orientation or the entire protein coding region of rat hsp27 in the antisense orientation, both inserted in the mammalian expression vector pcDNA3.1 under the control of the constitutive cytomegalovirus promoter. Clonal populations of stably transfected hsp27 sense or antisense podocyte clones were isolated by limiting dilution and selected for altered hsp27 levels. Stably transfected clones were then induced to differentiate as described above in the presence of a reduced concentration (200 µg/ml) of G418. Before treatment with PAN, cells were transferred to medium without G418.
In vitro model of podocyte process retraction
Differentiated podocyte clones were treated with a single application of PAN (Sigma) to a final concentration of 1.25 or 5.0 µg/ml. Controls received an equivalent volume of sterile water. Cells were cultured for an additional 7 days without subcultivation. Subconfluent MMC and NIH-3T3 cells were similarly treated with PAN at concentrations identical to or markedly higher than those for podocytes. Because, in contrast to differentiated podocyte cultures, both MMC and NIH-3T3 cells proliferate continuously in culture, cells were seeded at low density and cultured for 7 days in the presence of PAN.
Phase contrast microscopy
Phase contrast microscopy was performed as described previously (28)
on an inverted Zeiss Axiovert 135TV using a 20x objective lens, an 0.55 NA long-working distance condenser lens, and phase contrast optics. Images were acquired using a monochrome CCD camera (Dage model RT3000) driven by NIH Image 1.61 software and captured using a Scion LG3 image capture board installed in an Apple Macintosh computer (PowerPC 7500). Images were adjusted for contrast and gain in Adobe Photoshop v. 5.0.2.
Immunohistochemistry
After PAN treatment for 7 days, cells grown on glass coverslips were washed briefly in warm PBS and fixed in 3.7% paraformaldehyde in PBS for 30 min. Cells were lysed by addition of an equal volume of 0.5% Triton X-100 in PBS to the fixation solution and incubation for another 5 min. Cells were washed with 0.05% Tween-20 in PBS (T-PBS) and incubated with Texas red-conjugated phalloidin (Molecular Probes, Eugene, OR) diluted 1:250 in 5% heat-inactivated horse serum in T-PBS for 30 min. Coverslips were washed and mounted in 0.05% phenylenediamine in 90% glycerol in PBS and sealed with fingernail polish. Fluorescence microscopy was performed as described previously (28)
. Prepared cells were imaged on an inverted Zeiss Axiovert 135TV with an AtoArc 100W epifluorescent illuminator and 20x objective lens. Images were collected using an integrating monochrome CCD camera (Dage model RT3000) driven by NIH Image 1.61 software (with capture hardware described above) using macros permitting collection and signal averaging over several video frames. An average of 16 frames were integrated for fluorescent images. Images were adjusted for contrast and gain, and levels were adjusted in Adobe Photoshop v. 5.0.2. All immunofluorescence micrographs shown were taken of one of the two clones of hsp27 sense (clone 23) and hsp27 antisense (clone 9) cells, but were representative of results gathered for both clonal lines of each type.
Phalloidin binding assay
Phalloidin binding was determined using a modification of the protocol of Singhal et al. (29)
. After treatment, cells were washed briefly in warm PBS and fixed in 3.7% paraformaldehyde in PBS for 30 min. Fixed cells were washed once with PBS and incubated in 0.1% saponin in PBS containing a saturating amount (0.4 µM) of Texas red phalloidin for 1 h at room temperature in darkness with gentle agitation. Cells were washed twice with 0.1% saponin in PBS and the labeled phalloidin was extracted from cells into 1 ml methanol by agitation for 30 min in darkness. The methanol was collected, centrifuged at 14,000 g for 5 min and supernatant fluorescence was measured in an LS-50B luminescence spectrometer (Perkin Elmer Analytical Instruments, Norwalk, CT) with excitation at 591 nm and emission measured at 608 nm. Linear regression analyses of Texas red phalloidin standards was used to derive absolute amounts of bound phalloidin from each sample’s relative fluorescence values. Protein was extracted from duplicate wells and assayed by the bicinchoninic acid (BCA) microassay (see below); these values were used to normalize the amount of bound phalloidin to protein quantity.
Cell survival assay
After PAN or sham treatment, culture medium was collected and the adherent cells were collected by trypsinization. The cells suspended in trypsin solution were combined with the original culture medium and cells were pelleted by centrifugation at 1500 g for 3 min. The cells were resuspended in PBS and an equal volume of 0.4% trypan blue in PBS was added. After 1 min, two aliquots of each cell suspension were counted using a hemocytometer. Cells that freely took up dye (deep blue) were scored as dead, and pale blue or clear cells were scored as live.
Cell area determination
Phase contrast micrographs (n=6) were collected randomly from fixed cells after PAN treatment. The 10 cells in each digital micrograph that were entirely within the image field and closest to the arbitrary upper right-hand corner of the image were hand-traced using a graphics tablet (WACOM, Vancouver, WA) and NIH Image 1.61 software. Measurements were made of the total number of pixels within each outline and converted to unit area using a conversion factor derived from pixel measurements of a standard measure using the same microscopic optics arrangements.
Protein extraction
Proteins were extracted from cell cultures by washing adherent cells with warm PBS, followed by addition of 1 ml per well of a 6-well tissue culture plate of 1:1 water-saturated phenol:extraction buffer. The extraction buffer consisted of 50 mM tris-Cl pH 7.0, 700 mM sucrose, 50 mM KCl, 100 mM NaF, 2 mM each of MgCl2, EGTA, and 2-mercaptoethanol, 1 mM sodium orthovanadate, and 0.1 mM PMSF. Cells were extracted for 5 min at room temperature and the extracts centrifuged at 1500 g for 15 min at 10°C to separate phases. The upper, phenol phase was removed and mixed with 5 vol of -20°C methanol acidified with 100 mM ammonium acetate. After precipitation of protein overnight at -20°C, precipitates were collected by centrifugation at 1500 g for 25 min at 4°C. Pellets were washed twice with ice-cold 80% methanol and dried under vacuum. Protein was solubilized from dried pellets in a minimal volume of 2% SDS, 62.5 mM tris-Cl pH 6.8, 10% glycerol for SDS-PAGE analysis or into isoelectric focusing (IEF) sample buffer (9 M urea, 2% (w/v) Nonidet P-40, 50 mM NaF, 5 mM sodium orthophosphate, 67 mM [ß-D] glycerophosphate pH 7.2, 1 mM sodium orthovanadate, 1 mM EDTA, 2% 2-mercaptoethanol, 0.1 mM PMSF, and 0.5 µg/ml each of leupeptin, pepstatin, and aprotinin) by prolonged, gentle agitation. The protein content in SDS-PAGE samples was determined using the BCA microassay.
SDS-PAGE, slab IEF separations, and Western blotting
SDS-PAGE sample buffer extracts containing equal quantities of protein from each sample were brought to 5% 2-mercaptoethanol, heated to 100°C for 4 min, and separated on 13% polyacrylamide gels by SDS-PAGE, followed by transfer to PVDF membranes by immersion blotting. Urea buffer protein extracts containing similar amounts of hsp27 (as determined by prior analysis of SDS-PAGE sample buffer extracts by quantitative Western blotting) were separated by slab-gel IEF using the Bio-Rad Model 111 Mini-IEF Cell essentially according to the manufacturer’s instructions, except that the support film was omitted to permit subsequent transfer of proteins to PVDF membranes. Proteins were focused at 100 V for 15 min, 200 V for 15 min, and finally at 450 V for 1 to 2 h. After focusing, proteins were transferred to PVDF membranes by semi-dry transfer for 30 min at 75 mA.
Protein binding sites on PVDF membranes were blocked with 5% nonfat dry milk in T-PBS overnight at 4°C. Membranes were incubated with primary antibodies [1:5000 rabbit anti-murine hsp27 polyclonal or mouse anti-human hsp27 monoclonal (StressGen, Victoria, BC, Canada)] in 5% BSA in T-PBS for 1 h at room temperature. After three washes, blots were incubated in secondary antibody solution [1:10,000 goat anti-rabbit IgG horseradish-peroxidase conjugate or 1:5000 goat anti-mouse IgG horseradish-peroxidase conjugate (Jackson ImmunoResearch, West Grove, PA)] in 5% nonfat dry milk in PBS-T for 1 h at room temperature. After three additional washes, antibody binding was visualized using the ECL chemiluminescence system (Amersham, Arlington Heights, IL) as detected using the Bio-Rad Chemidoc system (Bio-Rad, Hercules, CA). Densitometric analyses of captured images were performed using version 4.1.0.026 of the Bio-Rad Quantity One software. Quantitation of the absolute amounts of specific protein in each band was performed by linear regression analysis using six different amounts of standard protein (murine or human hsp27; StressGen) on each blot to generate standard curves of protein content vs. densitometric units (mean density value x area).
Protein assay
The BCA protein microassay (30)
was used to quantify total protein in samples. Equal volumes of sample solubilization buffer were added to albumin protein standards to normalize for the effect of buffer components on the assay results.
Statistics
Results were analyzed for statistical significance using Statview v. 4.57 software (Abacus Concepts, Berkeley, CA) by the unpaired, two-tailed t test. Comparisons with P values < 0.05 were considered significant and P values < 0.01 were separately noted. Separate statistical analyses were performed to compare PAN-treated vs. sham-treated individual transfectant clones (*P<0.05 and **P<0.01 vs. sham-treated cells) and to compare PAN-treated and sham-treated vector control cells with the combined results from both clones of similarly treated hsp27 sense or antisense transfectants (+P<0.05, ++P<0.01 vs. vector cells).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
.
Stable podocyte transfectants contain dramatically different amounts of Hsp27
Protein extracts of independently cloned podocyte cell lines stably transfected with either human hsp27 sense (sense) or rat hsp27 antisense (antisense) constructs and differentiated for 10 days were analyzed by quantitative Western blotting using anti-human hsp27 or anti-murine hsp27 antibodies. These reagents allowed independent analysis of the human or murine hsp27 isoforms without interspecies cross-reaction. Vector control podocytes contained 2.34 ng of murine hsp27/µg protein (Fig. 1
) and, as expected, no detectable human hsp27 (data not shown). Both sense clones had significantly greater amounts of human hsp27 (17.1 and 14.9 ng hsp27/µg protein in clones 23 and 31, respectively) than the amount of murine hsp27 in vector controls whereas both antisense clones contained significantly smaller amounts of murine hsp27 (0.41 and 0.21 ng hsp27/µg protein in clones 9 and 20, respectively) than vector-transfected cells (Fig. 1)
. No endogenous murine hsp27 protein was detected (limit of detection 
0.02 ng/µg protein) in sense clones (data not shown). The mean amount of hsp27 present in all sense clones combined was > 680% of vector control values whereas the mean quantity of hsp27 in antisense clones was only 13% of vector control values, equivalent to a 51-fold difference in hsp27 expression between sense and antisense clones.
|
Stable podocyte transfectants have less Hsp27 phosphorylation than vector transfectants
Protein extracts from differentiated vector control cells or sense or antisense clones were separated by IEF, followed by Western blotting to determine the relative amount of hsp27 phosphorylation. Only monophosphorylated and unphosphorylated hsp27 isoforms were detected in extracts of transfectants, despite two possible serine phosphorylation sites on murine hsp27 and three possible sites on human hsp27 (31)
. The percentage of hsp27 phosphorylated was greater in vector control cells (30%) than in antisense clone 9 (21%) or in either of the sense clones (13% and 17% in clones 23 and 31, respectively), as shown in Fig. 2
. When the results from individual clones were combined and analyzed in aggregate, there was significantly more phosphorylated hsp27 in vector cells than in either antisense or sense clones. Aggregate significance values were determined by comparing vector values (n=4, mean=30%) with combined antisense (n=8, four each of clones 9 and 20, mean=23%) or sense (n=8, four each of clones 23 and 31, mean=15%) values. These results demonstrated that in murine podocyte transfectants, both the endogenous murine hsp27 in vector and antisense transfectants and the exogenous human hsp27 in sense transfectants are phosphorylated.
|
Because the absolute amounts of phosphorylated and unphosphorylated hsp27 present in cells may be more biologically relevant than percentage of phosphorylated hsp27, absolute quantities of phosphorylated and unphosphorylated hsp27 (Table 1
) in each podocyte clone were calculated from the data shown in Figs. 1
and 2
. This analysis revealed that although a smaller percentage of the hsp27 in sense clones was phosphorylated than in vector controls (Fig. 2)
, the absolute amount of phosphorylated hsp27 in sense cells was > threefold greater than the amount in vector controls (>2.2 ng/µg vs. 0.7 ng/µg; Table 1
). In contrast, both the percentage and absolute amount of phosphorylated hsp27 in antisense clones were less than the corresponding values in vector controls (Fig. 2
and Table 1
). Similarly, the amount of unphosphorylated hsp27 in sense transfectants was > sixfold greater than the amount in vector controls and sixfold more unphosphorylated hsp27 was present in vector controls than in antisense clones. The finding that a smaller percentage but larger absolute amount of hsp27 is phosphorylated in sense transfectant clones vs. vector controls and that there is a larger absolute amount of unphosphorylated hsp27 in sense (and lesser amount in antisense) than vector controls raises the possibility that the observed phenotypic differences between transfectants and controls (see below) may be due to differences in either the amount of total hsp27 (Fig. 1)
or the amount of a particular phosphorylated isoform of hsp27 (Table 1)
.
|
Hsp27 regulates normal podocyte morphology and PAN-induced morphological changes
Differentiated sense and antisense clones displayed dramatically different cellular morphologies compared with vector controls as examined by phase contrast microscopy (Fig. 3
). Untreated, differentiated vector control cells displayed a characteristic flattened morphology with numerous cell processes (Fig. 3A
) whereas untreated sense clones were often elongated, with smooth cell borders and many contiguous areas of cell–cell contact (Fig. 3D
); antisense clones were flattened but exhibited less frequent and smaller processes than controls (Fig. 3G
).
|
Vector and sense cell morphology was largely unchanged after treatment with 1.25 µg/ml PAN (Fig. 3B, E
) whereas many antisense cells had partially retracted and some cells had detached (Fig. 3H
). After treatment with 5.0 µg/ml PAN, the majority of attached vector and antisense cells retracted (Fig. 3C, I
) and the vast majority of antisense cells detached and died (data not shown). In marked contrast, sense cells (Fig. 3F
) displayed relative morphological preservation after PAN treatment vs. either vector control (Fig. 3C
) or antisense (Fig. 3I
) cells. These results indicated that hsp27 was able to both alter normal podocyte morphology and modulate PAN-induced podocyte morphological changes.
Hsp27 regulates podocyte cell survival after PAN treatment
Trypan blue assays after treatment revealed there were no significant differences in the number of nonvital cells (cells that took up dye) either between vector control, sense, and antisense transfectants or between untreated and PAN-treated cells. In all cell suspensions regardless of hsp27 protein expression or treatment, 10% of cells were dead (data not shown). Cell counts of living cells, however, showed that stable transfection and PAN treatment both influenced the number of vital cells remaining in podocyte cultures after differentiation and treatment. Approximately half of the number of vector control cells initially seeded into plates (3.6x104) were present after differentiation and sham treatment (1.8x104, Fig. 4
A, left bar). There were significantly more live vector cells present in cultures treated with 1.25 µg/ml PAN than in untreated cultures, but significantly fewer after treatment with 5.0 µg/ml PAN (Fig. 4A
). In contrast, the number of live sense clone 31 cells increased significantly vs. untreated sense cells at both PAN concentrations (Fig. 4B
, right bars) whereas the number of live sense clone 23 cells increased significantly only after treatment with 5.0 µg/ml PAN (Fig. 4B
, left bars). In marked contrast to sense clones and vector control cells, a significant decline in the number of live antisense clone 20 cells was observed after 1.25 µg/ml PAN treatment (Fig. 4C
, right bars), and both antisense clone populations dramatically declined compared with untreated controls after 5.0 µg/ml PAN treatment.
|
When compared with identically treated vector control cells, the mean number of sense cells (results from both clones in aggregate) was significantly greater in sham-treated controls (mean of 2.8x104 sense cells vs. 1.8x104 vector control cells) and after PAN treatment at either concentration (Fig. 4B
vs. A). In contrast, even though cultures were initially seeded with twice as many antisense cells (7.2x104), when considered in aggregate there were no significant differences between vector controls and antisense clones after sham treatment or treatment with PAN at 1.25 µg/ml (Fig. 4C
vs. A). Notably, however, the number of antisense cells remaining after treatment with 5.0 µg/ml PAN was significantly less than identically treated vector controls (Fig. 4C
vs. A). These results confirmed the ability of hsp27 to influence podocyte cell survival during differentiation as well as modulating the response to PAN treatment.
Hsp27 regulates podocyte polymerized actin content after PAN treatment
Since disruption of podocyte actin filaments is a characteristic feature of NS and hsp27 is a known regulator of actin polymerization, the amount of phalloidin bound by podocytes was assayed to provide a quantitative measure of the amount of polymerized actin present in cells. We found a significant increase in the polymerized actin content of vector cells after treatment with 1.25 µg/ml PAN vs. untreated controls and a significant decrease after 5.0 µg/ml PAN (Fig. 5
A). In contrast, there was no significant change in phalloidin binding in either sense clone after PAN treatment except for a slight decrease in sense clone 23 after treatment at 5.0 µg/ml (Fig. 5B
, left bars), whereas both antisense clones displayed a significant decrease in phalloidin binding compared with untreated cells after PAN treatment at both concentrations (Fig. 5C
).
|
Combined phalloidin binding assay results from both sense or both antisense clones in aggregate were also compared with identically treated vector controls. After treatment with 5.0 µg/ml PAN, significantly more phalloidin was bound by sense cells than vector controls (Fig. 5B
vs. A) whereas antisense cells bound less phalloidin than vector controls after 1.25 (Fig. 5C
vs. A) and similar amounts after 5.0 µg/ml PAN treatment. Since phalloidin binding represents a measure of cellular polymerized actin, these results indicated clearly that sense clones were significantly more resistant to PAN-induced disruption of podocyte actin filaments than either vector control or antisense cells.
Hsp27 regulates both normal podocyte actin cytoskeleton and the changes induced by PAN
Figure 6
shows representative micrographs of the structure of actin filaments as revealed by fluorescent phalloidin labeling in cells after differentiation and PAN treatment. Untreated vector control cells (Fig. 6A
) contained central, regular arrays of large actin stress fibers surrounded by a peripheral region containing mostly smaller actin filaments. Cellular processes contained actin filaments, but not exclusively in the form of large stress fibers. In contrast, untreated sense clones (Fig. 6D
) contained abundant, parallel arrays of comparatively thin actin stress fibers that extended across the entire cell body. Untreated antisense clones (Fig. 6G
) contained a mixture of large and smaller stress fibers, often with a central bundle of large stress fibers and smaller peripheral stress fibers.
|
After treatment with PAN at 1.25 µg/ml, actin filaments in vector control cells (Fig. 6B
) were partially disrupted, with large stress fibers spread more diffusely than in untreated cells; strongly staining, punctate concentrations of polymerized actin appeared in cells. At the higher PAN concentration (Fig. 6C
), the large stress fibers had almost completely disappeared, though finer actin filaments were still visible in the cell body, often arranged in a circular pattern. The punctate actin concentrations were more prominent than at 1.25 µg/ml PAN. In contrast to these changes, actin filaments in PAN-treated sense clones were largely unaffected by PAN treatment at either concentration (Fig. 6E, F
), although some subcortical and punctate concentrations of polymerized actin were visible in sense cells treated with 5.0 µg/ml PAN (Fig. 6F
). Compared with either the vector control or sense clones, the actin cytoskeleton in antisense clones was more severely disrupted by PAN. Treatment with the lower PAN dosage (1.25 µg/ml) caused loss of most large stress fibers (Fig. 6H
). After treatment with 5.0 µg/ml PAN (Fig. 6I
), large stress fibers essentially disappeared from the few remaining antisense cells, leaving only a thin subcortical ring of polymerized actin and fine actin filaments in the cell body. Similar to the phalloidin binding results above, these results clearly demonstrated that hsp27 was able to modulate PAN-induced disruption of podocyte actin filaments.
Hsp27 regulates podocyte cell area changes induced by PAN treatment
Digital phase contrast micrographs of transfectants were analyzed to determine cell area by manually outlining cell borders and integrating the enclosed area. The mean cell area of vector control cells treated with 1.25 µg/ml PAN was significantly greater than sham-treated cells whereas the mean cell area was significantly smaller in vector cells treated with 5.0 µg/ml PAN (Fig. 7
A). There were no significant changes in mean cell area in the sense clones after treatment with 1.25 µg/ml PAN whereas treatment with 5.0 µg/ml PAN caused a significant increase in cell area in both clones (Fig. 7B
). The changes in mean cell area of antisense clone 9 in response to PAN treatment were similar to the responses of vector control cells whereas the area of clone 20 cells showed no significant change after 1.25 µg/ml PAN treatment and a significant decrease after 5.0 µg/ml PAN treatment.
|
When analyzed in aggregate, only sense clones displayed significant differences in mean cell area vs. identically treated vector control cells. Sense clones were significantly smaller than vector controls after 1.25 µg/ml PAN treatment and significantly larger than controls after treatment with 5.0 µg/ml PAN (Fig. 7B
vs. A). These results together clearly indicated hsp27 was able to modulate PAN-induced changes in cultured podocyte cell area.
PAN effects on cell morphology and actin cytoskeleton are relatively podocyte specific
To determine the cell specificity of the effects of PAN observed in murine podocytes, we treated both NIH-3T3 fibroblasts and kidney MMC with various concentrations of PAN. We found that the morphology of both MMC and NIH-3T3 cells was unaffected by culture in the presence of 1.25 µg/ml PAN (compare Fig. 8
A, B with C and D, respectively). Treatment of either cell type with 5.0 µg/ml PAN resulted in a more flattened cell morphology (Fig. 8E, F
) as well as partial disruption of actin filaments and increased cell area (data not shown), similar to the results observed in differentiated podocytes after treatment with 1.25 µg/ml PAN. The retraction, detachment, and severe actin filament disruption observed in differentiated podocytes after treatment with 5.0 µg/ml PAN were observed in MMC cells only after treatment with 25 µg/ml PAN (Fig. 8G
) and in NIH-3T3 cells at a PAN concentration >25 µg/ml (Fig. 8H
). These results suggest that differentiated, cultured podocytes are specifically affected by concentrations of PAN that do not affect other cell types, in agreement with the earlier observation that podocytes derived from primary glomerular outgrowths are highly sensitive to PAN treatment (32)
.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. In the present study, we tested the hypothesis that hsp27 has an important role in regulating normal podocyte structure and the dramatic structural changes in podocytes that occur during NS. To do this, we developed cultured podocyte clones stably transfected with sense and antisense hsp27 constructs and analyzed the effects of altered hsp27 levels using a newly developed in vitro model of PAN-induced podocyte injury.
These studies used cultured podocytes derived from the ‘Immortomouse’, which conditionally expresses the SV40 T-antigen to permit growth as ‘differentiated’ podocytes expressing differentiation-specific protein markers and displaying a flattened morphology, distinctive cellular processes, and a low cell proliferation rate (25)
. Although these differentiated podocytes do not form the extensive interdigitating networks of secondary cell processes spanned by slit membranes characteristic of podocytes in vivo, they do have several fundamental characteristics of in vivo podocytes that were lacking in earlier podocyte culture systems; at this time they represent the best available in vitro cell culture system for analysis of podocyte responses to relevant effectors such as PAN. We analyzed changes in podocyte morphology and actin cytoskeleton after 7 days of PAN treatment, a time selected to closely model the time of the most dramatic podocyte structural changes in PAN-induced NS in vivo.
We found that this in vitro podocyte injury model could reproducibly mimic many of the podocyte morphological and actin cytoskeletal changes seen during NS in vivo. These changes included alterations in cell morphology and size, process retraction, and actin filament disruption. In addition, the low-dose PAN treatment induced significant hsp27 accumulation in podocytes, similar to our earlier findings of glomerular hsp27 accumulation in experimental NS in vivo, providing further validation of the in vitro model. Because of these similarities, we believe this model has significant potential to permit more mechanistic studies of podocyte pathobiology during PAN-induced injury in the future.
We tested several concentrations of PAN in this treatment regime and discovered two distinct responses of cultured murine podocytes at either a low (1.25 µg/ml) or high (5 µg/ml) concentration. Previously described PAN treatments of cultured podocytes (32
, 33)
used higher PAN concentrations (50–100 µg/ml) over shorter periods (20–72 h) with results similar to the response of vector control cells treated with 5.0 µg/ml PAN for 7 days in this study. This response was characterized by cell retraction, rounding, and detachment, as well as disruption of the actin cytoskeleton. Although treatment with 1.25 µg/ml PAN for 7 days also resulted in discernible rearrangements of actin stress fibers, the number of live cells, the mean cell area, and the F-actin content of these cells all actually increased significantly, reflecting a fundamentally distinct biological response from that previously reported.
This reciprocal response of podocytes to low vs. higher concentrations of PAN demonstrates that podocytes can respond to a stressor with relevance to NS in at least two distinct fashions: 1) by active processes such as expansion of cell area, an increase in F-actin content, and increased cell survival or proliferation, or 2) by responses more characteristic of a toxic effect such as cell process retraction, a decrease in F-actin content, and decreased cell survival. We thus speculate that the low-concentration PAN treatment regime may be more relevant to podocyte changes that occur in typical idiopathic NS in vivo, since little or no podocyte death has been associated with this disease (34)
and the effacement observed in vivo is characterized by a loss of distinctive foot process morphology and a decrease in cell perimeter (35)
, but not by a radical reduction in the podocyte area in contact with the GBM. In contrast, the higher concentration PAN treatment regime may be more relevant to podocyte changes that occur in more resistant forms of NS, such as focal segmental glomerulosclerosis, where more severe podocyte injury and retraction may result in 1) detachment of foot processes from the GBM and subsequent formation of synechiae, 2) decreased podocyte survival that results in podocyte loss in the urine (36)
, and ultimately 3) development of chronic renal insufficiency.
We also found that alterations in hsp27 levels were able to regulate a variety of podocyte characteristics in the absence of any treatment. Hsp27 has been reported by numerous investigators to have both positive and negative roles in the regulation of actin filament dynamics. It has been identified as an barbed-end filament capping protein in vitro (37)
and has been reported to inhibit recovery of actin filaments and focal adhesions induced by disruption via heat shock and to delay reattachment and limit recovery of actin filaments (23)
. In contrast to these negative effects, hsp27 has also been reported to confer protection against actin filament disruption caused by oxidants (38)
, to provide partial protection against cell death caused by hydrogen peroxide treatment (38)
, and to suppress actin filament disruption induced by cytochalasin D as well as accelerating the reassembly of actin stress fibers (15
, 39)
. Our results are entirely consistent with a protective role for hsp27 in the regulation of actin filaments. We found that modulation of hsp27 expression alone (in the absence of PAN treatment) in podocytes consistently resulted in morphological and actin cytoskeletal alterations.
The effects of altered hsp27 phosphorylation on podocyte morphology and actin cytoskeletal structure, however, were less clear. Although the hsp27 sense cells contained a larger amount of phosphorylated hsp27 than either the vector or antisense cells, they also had a lower percentage of hsp27 phosphorylation than either of the other cells. The relative biological importance of the absolute amount vs. relative phosphorylation of hsp27 is not yet understood, but hsp27 phosphorylation itself is known to have important effects on hsp27 function. Phosphorylation has been shown to abolish the ability of hsp27 to inhibit actin polymerization in vitro (37)
and to regulate the organization of hsp27 into multimers (40)
. Chinese hamster fibroblasts overexpressing a nonphosphorylatable mutant hsp27 protein have been reported to have a reduced F-actin content and a delay in growth factor-induced actin polymerization (15)
, and actin filaments in these cells were more sensitive to heat shock-induced disruption (24)
. Although these studies made no attempt to quantify the absolute or relative amounts of phosphorylated hsp27 present in transfectants, they do suggest that the amount of phosphorylated (or unphosphorylated) hsp27 present in cells may be a determining factor in regulating actin filament structure by hsp27. Our results can thus be interpreted in one of two ways: that the quantity 1) of total hsp27 or 2) of either the un- or monophosphorylated hsp27 isoform is primarily responsible for the effects of hsp27 on podocyte morphology and the actin cytoskeleton. However, since both sense and antisense transfection resulted in less relative phosphorylation of hsp27 than controls yet displayed divergent and largely opposite effects on podocytes, we conclude that differences in the relative phosphorylation of hsp27 were not responsible for the changes observed. Further studies to determine the effects of stable expression of hsp27 mutant proteins that are nonphosphorylatable or pseudo-phosphorylated are under way in our laboratory to clarify the role of hsp27 phosphorylation on podocyte structure.
Our findings also demonstrated the ability of hsp27 to regulate a variety of cellular responses of podocytes to PAN-induced injury. Podocyte hsp27 levels correlated directly with resistance to PAN-induced cell death. This finding is entirely consistent with hsp27’s known chaperonin function (41
, 42)
. Alternatively, although we did not determine whether cell death was due to apoptosis vs. necrosis, this protection may have resulted from hsp27’s reported role in protection against apoptosis (43)
. Hsp27 was also able to regulate the podocyte morphological and actin cytoskeletal responses to PAN. Compared with the morphological changes and quantitative and qualitative disruption of actin filaments seen in vector control cells, cells containing greater amounts of hsp27 were markedly protected whereas cells expressing lesser amounts of hsp27 were markedly more sensitive to PAN-induced injury. Although hsp27 has been reported to function as either a chaperonin or regulator of the actin cytoskeleton, the pronounced actin cytoskeletal effects seen in the current studies suggest that hsp27’s major role in podocytes is probably in the regulation of the actin cytoskeleton.
In summary, we analyzed murine podocytes stably transfected with hsp27 sense, antisense, and vector control constructs using a newly developed in vitro PAN model system designed to mimic the podocyte structural changes seen in NS. We found that hsp27 is able to regulate many key structural features of podocytes. More importantly, we also found that hsp27 is able to regulate the morphological and actin cytoskeletal response of podocytes to injury induced by PAN, a toxic compound used in a well-established in vivo model of NS. In conjunction with our earlier findings of increased glomerular hsp27 expression and phosphorylation during PAN-induced NS, these findings suggest that hsp27 may have an important role in regulating the podocyte structural changes observed during the development of and/or recovery from nephrotic syndrome.
|
ACKNOWLEDGMENTS |
|---|
Received for publication September 6, 2001. Accepted for publication November 5, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
B-crystallin confers resistance to TNF- and oxidative stress-induced cytotoxicity in stably transfected murine L929 fibroblasts. J. Immunol. 154,363-374[Abstract]
This article has been cited by other articles:
|
|
 |
|
  J.-K. Park, N. Ronkina, A. Hoft, C. Prohl, J. Menne, M. Gaestel, H. Haller, and M. Meier Deletion of MK2 signalling in vivo inhibits small Hsp phosphorylation but not diabetic nephropathy Nephrol. Dial. Transplant., June 1, 2008; 23(6): 1844 - 1853. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. De Petris, K. A. Hruska, S. Chiechio, and H. Liapis Bone morphogenetic protein-7 delays podocyte injury due to high glucose Nephrol. Dial. Transplant., December 1, 2007; 22(12): 3442 - 3450. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Eichler, Q. Ma, C. Kelly, J. Mishra, S. Parikh, R. F. Ransom, P. Devarajan, and W. E. Smoyer Single and Combination Toxic Metal Exposures Induce Apoptosis in Cultured Murine Podocytes Exclusively via the Extrinsic Caspase 8 Pathway Toxicol. Sci., April 1, 2006; 90(2): 392 - 399. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Shibata, M. Nagase, and T. Fujita Fluvastatin Ameliorates Podocyte Injury in Proteinuric Rats via Modulation of Excessive Rho Signaling J. Am. Soc. Nephrol., March 1, 2006; 17(3): 754 - 764. [Abstract] [Full Text] |
