Role of superoxide in apoptosis induced by growth factor withdrawal

Wilfred Lieberthal, Veronica Triaca, Jason S. Koh, Patrick J. Pagano, Jerrold S. Levine

Abstract

We have examined the role of reactive oxygen species (ROS) in apoptosis induced by growth factor deprivation in primary cultures of mouse proximal tubular (MPT) cells. When confluent monolayers of MPT cells are deprived of all growth factors, the cells die by apoptosis over a 10- and 14-day period. Both epidermal growth factor (EGF) and high-dose insulin directly inhibit apoptosis of MPT cells deprived of growth factors. Growth factor deprivation results in an increase in the cellular levels of superoxide anion while apoptosis of MPT cells induced by growth factor withdrawal is inhibited by a number of antioxidants and scavengers of ROS. Growth factor deprivation also results in activation of caspase activity, which is inhibited by EGF and high-dose insulin as well as by the ROS scavengers and antioxidants that inhibit apoptosis. The cell-permeant caspase inhibitor, z-Val-Ala-Asp-CH2F (zVAD-fmk), prevents the increase in caspase activity and markedly inhibits apoptosis induced by growth factor deprivation. However, zVAD-fmk had no effect on the increased levels of superoxide associated with growth factor deprivation. Thus we provide novel evidence that ROS play an important role in mediating apoptosis associated with growth factor deprivation. ROS appear to act upstream of caspases in the apoptotic pathway. We hypothesize that oxidant stress, induced by growth factor withdrawal, represents a signaling mechanism for the default pathway of apoptosis.

  • cell viability
  • survival factors
  • programmed cell death
  • reactive oxygen species
  • caspases

many normal mammalian cells have been shown to be dependent upon the presence of specific growth factors for their survival. Examples of this phenomenon include the dependence of some neuronal cells on nerve growth factor (2, 8) and of prostatic epithelial cells on androgens for maintenance of viability (39). In the absence of these “survival factors,” these cells die by apoptosis. A deficiency of essential survival factors has become well recognized as one of many triggers of apoptosis (24). Raff (39) has suggested that the genetically mediated program responsible for apoptosis represents a “default pathway” that is always ready to be activated in many if not all mammalian cells and must constantly be suppressed by growth hormones and other survival factors for cell to remain viable (39).

This “default pathway” for apoptosis may explain, at least in part, the well-demonstrated efficacy of pharmacological doses of renal growth factors such as epidermal growth factor (EGF), insulin-like growth factor-I (IGF-I), and hepatocyte growth factor in facilitating the recovery of renal function in experimental models of acute renal failure (14, 16, 28, 47, 48). However, an inhibitory effect of these renal trophic factors on apoptosis of renal tubular cells has never been demonstrated.

Furthermore, the signaling pathways that mediate the default pathway of apoptosis remain ill defined. In some cells, the removal of growth factors appears to signal apoptosis via increased production of reactive oxygen species (ROS) (2, 46). However, the role of ROS in apoptosis remains controversial (15, 17, 18).

The purpose of this study was twofold:1) to examine the extent to which EGF and IGF-I inhibit apoptosis in renal tubular cells and2) to examine the hypothesis that withdrawal of growth factors induces apoptosis in renal tubular cells by increasing oxidative stress in these cells. We have used primary cultures of mouse proximal tubular (MPT) cells to study this issue, because the default pathway of apoptosis appears to be aberrant in many transformed renal cell lines. Our data suggest that superoxide levels are elevated in MPT cells following removal of growth factors and that the resultant oxidant stress acts as a signal to trigger the default pathway of apoptosis. We also provide novel evidence that ROS act upstream of caspases in inducing apoptosis following growth factor withdrawal.

METHODS

Reagents

Hoechst dye (bisbenzamide or H-33342) was obtained from Calbiochem (San Diego, CA). Culture reagents (culture dishes, culture medium, and trypsin-EDTA) were obtained from GIBCO BRL (Gaithersburg, MD). The caspase inhibitor benzyloxycarbonyl-valinyl-alanyl-aspartyl-fluoromethyl ketone (zVAD-fmk) was obtained from Kamiya Biomedical (Tukwila, WA). All other reagents, including trypan blue, bovine superoxide dismutase (SOD), catalase, Tiron (4,5-dihydroxy-1,3-benzene disulfonic acid), deferoxamine (DFO), and probucol were obtained from Sigma (St. Louis, MO).

Culture of Renal Tubular Cells

Primary culture of MPT cells. MPT cells were cultured from collagenase-digested fragments of proximal tubules obtained from the cortices of kidneys of C57Bl/6 mice by a modification of previously described methods (41). Briefly, cortical tubules were plated in serum-free, defined culture medium (1:1 mixture of DMEM and Ham’s F-12, containing 2 mM glutamine, 15 mM HEPES, 5 μg/ml transferrin, 5 μg/ml insulin, 50 nM hydrocortisone, 500 U/ml penicillin, and 50 μg/ml streptomycin). MPT cells grew to confluence from tubules over 4–5 days and were studied between the 5th and 14th days of culture. Using a combination of morphological, biochemical, and transport characteristics, we have demonstrated that MPT cells are of proximal tubular origin (41).

Culture of Madin-Darby canine kidney and opossum kidney (OK) cells. Madin-Darby canine kidney (MDCK) and opossum kidney (OK) cell lines were obtained from the American Type Culture Collection. Both cell lines were grown using the same defined, serum-free medium used to culture MPT cells

Cell Viability Assay

Cell viability was determined by counting the number of cells that remained adherent to the cell monolayer and excluded trypan blue. Although trypan blue positivity and loss of cell adherence cannot be used to distinguish apoptosis from necrosis (24), the combined use of these criteria is a well-established technique for determining whether any intervention alters overall “cell viability.” The exact mechanism of death, whether necrosis (41) or apoptosis (24, 25), must be established by other methods (discussed below). Nonadherent cells were removed by two washes with ice-cold PBS. Adherent cells were harvested by trypsinization, centrifuged at 500 rpm for 5 min, and resuspended in 1 ml DMEM. Trypan blue (0.04 g/dl) was added for 10 min. The number of cells excluding trypan blue was counted in a hemocytometer. Cell viability was expressed as a percentage of the viable cells in experimental wells compared with that in wells containing “freshly confluent” monolayers which develop 5 days after plating tubules (25).

Agarose Gel Electrophoresis of DNA

MPT, MDCK, and OK cells were grown to confluence in P-100 culture dishes over 5 days and then subjected to growth factor withdrawal. Supernatant and detached cells were pooled before extracting a single fraction of DNA per sample for electrophoresis. Cells were lysed in 0.5% Triton X-100, 5 mM Tris ⋅ HCl, pH 7.4, and 20 mM EDTA for 30 min at 4°C. After centrifugation at 15,000g for 20 min, the supernatants were extracted with phenol-chloroform. Sodium acetate, 3.0 M, pH 5.2 (1/10th volume), and MgCl2, 1.0 M (1/30th volume), were added, and the DNA was precipitated in ethanol. Samples were separated by electrophoresis on a 1.2% agarose gel containing ethidium bromide.

Fluorescent Staining of MPT Cells with H-33342

As originally described by Kerr et al. (20) using electron microscopy (EM), apoptotic cells are characterized by unique morphological changes within their nuclei, consisting of chromatin condensation and nuclear fragmentation. These changes can also be identified using the fluorescent DNA-binding dye H-33342 (bisbenzamide), which stains the nuclei of normal cells as well as those that have died by apoptosis or necrosis. As opposed to viable and necrotic cells, the nuclei of apoptotic cells undergo condensation, which results in an increase in the intensity of nuclear fluorescence when a limiting concentration of Hoechst dye is used. The increased intensity of Hoechst staining of apoptotic cells under these conditions is believed to be both the result of structural changes in the DNA (chromatin condensation) as well as of an increased rate of influx of H-33342 into apoptotic compared with normal cells (9, 32). We have previously reported that staining cells with 1 μg/ml H-33342 for 10 min at 37°C resulted in optimal differentiation of apoptotic cells from normal and necrotic cells (25). Our method is a modification of a well-established and widely used technique (18, 42, 49).

Adherent MPT cells were harvested (as described above) and combined in a single Eppendorf tube with cells that had detached spontaneously from the same monolayer. The cells were washed once with PBS before being stained with H-33342. Wet preparations of these cells were made on glass slides, and the same fields were photographed under both phase-contrast and epifluorescence microscopy (excitation wavelength of 348 nm; emission 479 nm) at ×400 magnification.

Flow Cytometric Analysis

MPT cells were stained with H-33342 as described above and placed immediately on ice. Flow cytometry was performed on an Epic ESP flow cytometer (Coulter Electronics, Hialeah, FL) with an ultraviolet (UV)-enhanced argon laser. Hoechst fluorescence was accomplished by excitation with <5 mW of UV laser light (351–364 nm multiline) and detected with a 525-nm bandpass optical filter. Cell debris was gated out electronically. Data were analyzed by Epic Elite software (Coulter Electronics). A constant number of “events” (cells) were analyzed for each sample (10,000 cells/sample). We used two methods of flow cytometry to distinguish normal cells and apoptotic cells.

Light scatter measurements. Normal cells are characterized by relatively high forward light scatter (FLS) (a measure of cell volume or size) and relatively low side light scatter (SLS) (a measure of cell granularity). Apoptotic cells, in contrast, are smaller and more granular than normal cells, and thus are characterized by relatively low FLS and high SLS (29, 44). Thus, the first method we used to quantitate apoptosis by flow cytometry was to compare FLS (x-axis) with SLS (y-axis) as previously described (26, 29). This method continues to be used by some investigators as the sole flow cytometric method for differentiating normal and apoptotic cells (26).

Intensity of H-33342 fluorescence. A number of investigators have used Hoechst staining to distinguish normal and necrotic cells (low-intensity nuclear fluorescence) from apoptotic cells (high-intensity nuclear fluorescence) (9, 32, 43). We have adapted this method to analyze our cells, using a combination of cell size (FLS, x-axis) and H-33342 intensity (y-axis) (37).

Electron Microscopy of FACS-Sorted MPT Cells

We used flow cytometry to sort MPT cells into two populations with characteristics of normal and apoptotic cells and then examined these populations by EM (23). In this way, we were able to confirm that comparison of the intensity of Hoechst fluorescence vs. FLS is a reliable way to distinguish normal from apoptotic MPT cells. Two 6-well plates of freshly confluent monolayers and two 6-well plates of MPT cells subjected to 5 days of incubation in growth factor-free medium were stained with H-33342 as described above and combined into a single sample (of ∼24 × 106 cells in 1.0 ml PBS). Analysis of these cells by flow cytometry, and the subsequent plotting of Hoechst fluorescence vs. cell size (FLS), demonstrated two major cell populations. One population was characterized by low-intensity Hoechst staining and relatively high FLS. Cells from these two populations were sorted into separate test tubes at a flow rate of 800–1000 cells/s, and each population was examined by EM.

The two populations of sorted MPT cells were centrifuged at 500g for 15 min, fixed with 2.5% glutaraldehyde in PBS for 1 h at 4°C, and washed three times in Sabatini’s solution (PBS with 6.8% sucrose). Samples were postfixed with 1% osmium tetroxide (1 h), washed three times in Sabatini’s solution, passed through a graded series of alcohols (30, 50, 70, 90, and 100% for 15 min each), and treated with propylene oxide (15 min), a 1:1 Epon-propylene oxide mix (1 h), and three changes in pure Epon (3 h, 3 h, and overnight). Polymerization occurred overnight at 64°C. Ultrathin sections (∼50 nm) were cut with a MT2 Sorvall ultramicrotome, stained with lead citrate and uranyl citrate, and examined with a JEOL 100CX transmission electron microscope at 60 kV using a 20-μm objective aperture.

Measurement of Superoxide Anion in MPT Cells

Superoxide levels were measured in MPT cells using well-described techniques adapted by Pagano and his associates (34, 35). Freshly confluent MPT cells or cells exposed to growth factor withdrawal for 5 days were trypsinized and placed in suspension in 1 ml of PBS. A 20-μl aliquot of these cells was to determine the number of cells in each sample. Lucigenin (250 μM) was added to each of the rest of the samples, which were placed in 1.6-ml polypropylene 8 × 50-mm tubes (Evergreen Scientific, Los Angeles, CA). After allowing the samples to equilibrate with lucigenin at 37°C for 15 min, the tubes were placed in a luminometer (model 20e; Turner Designs, Mountain View, CA), the light chamber of which was maintained at 37°C. The luminometer was programmed to report units of light (analog voltage of the photomultiplier) over a 5-min period at 10 consecutive 30-s intervals. The 10 readings obtained for each sample were averaged to obtain a single readout per specimen. Periodic tests with an external standard (tritiated toluene) were done to verify that the sensitivity of the luminometer is constant. Furthermore, dark current readings (the photomultiplier background signal when the shutter is closed) are subtracted automatically by the luminometer. A blank containing PBS alone gave low and constant readings comparable to readings obtained from a blank containing lucigenin. The chemiluminescence of the sample containing freshly confluent cells was 7- to 10-fold above the background signal. Because the background signal was low and constant, they were not subtracted from the sample measurement. The specificity of the light reaction induced by lucigenin in the presence of superoxide has been established (13).

After the chemiluminescence of each sample was determined, Tiron (10 mM) was added and measurements of chemiluminescence were repeated. Tiron, a nonenzymatic scavenger of superoxide, was used to determine the superoxide-specific chemiluminescent signal of each specimen. Tiron quenched the chemiluminescent signal of each specimen to a value that was within 10% of background. The difference in luminescence in the absence and presence of Tiron was compared with a standard calibration curve as previously described (34).

Western Blotting for Poly-(ADP-Ribose)Polymerase

MPT cells were washed three times in cold PBS and lysed in buffer containing 50 mM Tris ⋅ HCl, pH 7.5, 50 mM glucose, 10 mM EDTA, aprotinin (0.5 μg/μl),N-tosyl-l-phenylalanine chloromethyl ketone (2 μg/μl), phenylmethylsulfonyl fluoride (1 mM), DNase (5 μg/ml), RNase (5 μg/ml), and SDS (1 g/100 ml). The lysate was centrifuged for 10 min at 1,000g at 4°C to remove nuclei and remaining intact cells and heated at 65°C for 15 min. After the addition of sample buffer, lysates were run on an 8% polyacrylamide-SDS gel. Proteins were then electrophoretically transferred to nitrocellulose filters. After transfer, the filters were washed in 150 mM NaCl, 100 mM Tris ⋅ HCl, pH 7.5, and 0.05% Tween 20 (TBST buffer), and blocked for 1 h in TBST containing 5% (wt/vol) nonfat powdered milk (TBST + M) before incubation with the anti-poly-(ADP-ribose)polymerase (anti-PARP) antibody PY20 (Boehringer Mannheim, Indianapolis, IN) at a dilution of 1:2,000 in TBST + M at room temperature for 90 min. The nitrocellulose filters were then washed three times with TBST and incubated in secondary horseradish peroxidase-labeled goat anti-mouse antibody (1:2,000 in TBST + M) for 45 min at room temperature. After three further washes with TBST, bound secondary antibody was detected using the enhanced chemiluminescence system (12).

Statistics

All data are expressed as the means ± SE. Comparisons between two groups were made using a two-tailed Student’st-test. When more than two groups were compared, the Bonferroni correction was used.P < 0.05 was considered significant.

RESULTS

Effect of Growth Factor Withdrawal on Viability of Cultured Renal Tubular Epithelial Cells

The response of growth factor deprivation on the viability of three different renal tubular cell lines in culture was compared:1) primary MPT cells (21, 40, 41);2) OK cells, an immortalized line of cells derived from the proximal tubules of the opossum kidney; and3) MDCK cells, an immortalized line of canine distal tubular cells.

Freshly confluent monolayers of MPT cells incubated in growth factor-free medium fail to maintain confluence. The loss of cell viability occurs asynchronously over many days, resulting in gradual loss of the monolayer. As individual MPT cells die, they become small and rounded, separate from surrounding cells and the substratum, and then detach from the monolayer. After 10 days of growth factor deprivation, only 18 ± 2% of the cells remain viable, whereas in the presence of EGF (10 nM) or high-dose insulin (5 μg/ml) (which stimulates IGF-I receptors), 86 ± 4% and 71 ± 10% of cells remain viable, respectively (Fig. 1). Two weeks after growth factor deprivation all MPT cells are dead (Fig.2). Subconfluent MPT cells respond to growth factor withdrawal in a manner comparable to confluent cells (data not shown).

Fig. 1.

Response of MPT cell viability to growth factor withdrawal. Confluent MPT monolayers were incubated in culture medium lacking growth factors or in presence of epidermal growth factor (EGF, 10 nM) or high-dose insulin (5 μg/ml) for 10 days. Numbers of viable cells remaining were counted and expressed as a percentage of the number of cells in freshly confluent monolayers. * P < 0.01 compared with freshly confluent monolayers. † P < 0.05 compared with MPT cells grown in absence of growth factors.

Fig. 2.

Comparison of response of MPT, OK, and MDCK cells to growth factor withdrawal. Confluent MPT, MDCK, and OK cells were subjected to growth factor deprivation, and the numbers of viable cells were determined after 14 days. Numbers of viable cells remaining were counted and expressed as a percentage of the number of cells in freshly confluent monolayers. * P < 0.01 compared with freshly confluent monolayers of the same type. † P < 0.01 compared with MPT cells grown in absence of growth factors.

In contrast to the effect of growth factor deprivation on MPT cells, OK and MDCK cells do not lose confluence over a period of 14 days in response to growth factor withdrawal. Instead, the number of viable OK and MDCK cells markedly increases during this period despite the absence of growth factors (Fig. 2).

DNA Fragmentation in Cultured Renal Tubular Epithelial Cells Exposed to Growth Factor Withdrawal

We compared the extent to which growth factor deprivation induces DNA fragmentation in MPT, MDCK and OK cells. Agarose gel electrophoresis of DNA obtained from MPT cells subjected to 5 days of growth factor deprivation demonstrates a typical “ladder” pattern of oligonucleosomal fragmentation of DNA consistent with apoptotic cell death (Fig. 3). A small amount of DNA fragmentation is discernible in freshly confluent MDCK as well as OK cells (Fig. 4; lanes 1 and 3, respectively). In OK cells, there is no increase in the amount of DNA fragmentation after 14 days of growth factor deprivation (Fig. 4;lane 4). In contrast to OK cells, there is a substantial increase in the amount of DNA laddering in MDCK cells subjected to 14 days of growth factor withdrawal (Fig. 4;lane 2).

Fig. 3.

DNA electrophoresis of MPT cells subjected to growth factor deprivation. Confluent monolayers of mouse proximal tubular (MPT) cells were incubated in growth factor-free medium for 2, 3, 4, and 5 days (lanes 2,3, 4, and 5, respectively). DNA markers are shown in lane 1. DNA fragmentation first becomes evident after 4 days of growth factor withdrawal and is marked on the 5th day when a “ladder” pattern is evident.

Fig. 4.

DNA electrophoresis of MDCK and OK and cells subjected to growth factor deprivation. Lanes 1 and3 represent DNA from freshly confluent MDCK cells (lane 1) and OK cells (lane 3), whereaslanes 2 and4 represent DNA from MDCK cells (lane 2) and OK cells (lane 4) that have been maintained in the absence of growth factors for 14 days. A small amount of DNA fragmentation is present in a ladder pattern in both lines of freshly confluent cells (MDCK; lane 1 and OK cells lane 3). There is a substantial increase in the amount of DNA laddering in growth factor-deprived MDCK cells (lane 2). However, there is no increase in DNA fragmentation in growth factor-deprived OK cells (lane 3) compared with freshly confluent OK cells (lane 1).

Morphology of MPT Cells Dying Following Growth Factor Withdrawal

Phase-contrast and fluorescent microscopy. The majority of freshly confluent MPT cells stained with H-33342 appear normal morphologically when examined under phase-contrast microscopy (Fig.5 A) or fluorescence microscopy (Fig. 5 B). In contrast, many of the cells subjected to growth factor withdrawal show the characteristic morphological features of apoptosis. Apoptotic cells, viewed under phase contrast, are smaller and more granular-appearing than viable cells (Fig.5 C). The nuclei of apoptotic cells fluoresce more intensely when stained with H-33342 and also demonstrate two classic features of apoptosis, i.e., chromatin condensation and fragmentation (Fig. 5 D). The process of nuclear fragmentation, which is characteristic of apoptosis, results in the disintegration of nuclei into multiple, small but brightly fluorescent fragments of DNA (Fig.5 D).

Fig. 5.

Phase-contrast and fluorescence microscopy of MPT cells obtained from freshly confluent monolayers (A andB) and after growth factor withdrawal for 5 days (C andD). In all experiments, adherent MPT cells together with those that have detached from the monolayer were combined and stained with H-33342 and mounted onto glass slides as a wet preparation. The same fields were photographed under phase-contrast (A andC) and fluorescence (B andD) microscopy at ×280 magnification. Cells from freshly confluent monolayers show normal morphology on phase-contrast (A) and fluorescence microscopy (B). Normal nuclei stained with H-33342 fluoresce faintly and demonstrate a delicate chromatin substructure (B). In contrast, many of the cells subjected to growth hormone withdrawal (C andD) demonstrate the characteristic features of apoptosis. Apoptotic nuclei stain intensely with H-33342 and become condensed (C, arrows) so that the normal chromatin pattern is lost. In later stages of apoptosis, the condensed nuclei undergo fragmentation (C, arrowheads). Cells containing these nuclei, when viewed under phase-contrast microscopy (B), are much smaller and more granular than cells with normal nuclear staining.

Transmission EM of cells of MPT cells subjected to growth factor deprivation and sorted by flow cytometry.Freshly confluent MPT cells as well as cells subjected to 5 days of growth factor withdrawal were each combined into a single sample, stained with H-33342, and analyzed by flow cytometry (see methods). MPT cells were sorted into two separate fractions for examination by transmission EM. Cells demonstrating faint H-33342 staining and high FLS were all viable on EM (Fig. 6,AC). In contrast, the majority (∼90%) of cells demonstrating bright H-33342 staining and low FLS demonstrated apoptotic morphology (Fig. 7,AC). A minority of cells from this population (∼10%) were morphologically normal.

Fig. 6.

Electron microscopy (EM) of MPT cells subjected to growth factor withdrawal and sorted by fluorescence-activated cell sorting (FACS). Cells sorted by flow cytometry and characterized by faint H-33342 staining and high forward light scatter (FLS) were all normal on EM. Low-power view (A, ×3,500 magnification) and high-power views (B, ×4,200 magnification; andC, ×4,200 magnification) of two different MPT cells show cells with viable nuclear morphology and an intact surface containing microvilli.

Fig. 7.

EM of MPT cells subjected to growth factor withdrawal and sorted by FACS. Cells characterized by bright H-33342 staining and reduced cell size (FLS) demonstrated apoptotic morphology on EM. Low-power view (A, ×3,700 magnification) demonstrates many apoptotic cells with condensed nuclei or containing apoptotic bodies. B: high-power view (×6,700 magnification) of an apoptotic cell in which the nucleus has undergone condensation and fragmentation into two pieces. Nuclear fragments appear to be in the process of being extruded from the cell.C: high-power view (×5,200 magnification) of another apoptotic cell demonstrating the classic features of loss of cytoplasmic volume and of crescentic condensation of chromatin against the nuclear membrane. All views (AC) also demonstrate that the apoptotic cells are rounded and have lost almost all their plasma membrane microvilli. This is another typical morphological feature of apoptosis when cells are examined under EM.

Effect of Growth Factor Deprivation on Levels of Superoxide Anion in MPT Cells

The level of superoxide anion was 0.367 ± 0.007 pmol/1,000 cells in freshly confluent monolayers and was increased to 0.512 ± 0.028 pmol/1,000 cells after 5 days of growth factor deprivation (P < 0.05) (Fig.8). Superoxide anion levels in MPT monolayers maintained for 5 days in the presence of EGF (0.347 ± 0.004 pmol/1,000 cells) or high-dose insulin (0.368 ± 0.021 pmol/1,000 cells) were comparable to those in freshly confluent monolayers (Fig. 8). In contrast to the effect of growth factors, the caspase inhibitor zVAD-fmk (5 μM) did not prevent the increase in superoxide levels in MPT maintained for 5 days in the absence of growth factors cells (0.494 ± 0.55 pmol/1,000 cells) (Fig. 8)

Fig. 8.

Measurement of superoxide levels in MPT cells. Superoxide levels are elevated in MPT cells subjected to growth factor withdrawal compared with freshly confluent monolayers. EGF and insulin inhibit the increase in superoxide levels induced by growth factor deprivation. However, the caspase inhibitor benzyloxycarbonyl-valinyl-alanyl-aspartyl-fluoromethyl ketone (zVAD-fmk) has no effect on the increase in superoxide levels associated with growth factor withdrawal. * P < 0.05 compared with freshly confluent monolayers.

Effect of Antioxidants on Viability of MPT Cells Subjected to Growth Factor Withdrawal

Cell viability of MPT cells maintained in the absence of growth factors for 10 days was substantially increased by a number of different antioxidants. SOD enzymatically converts superoxide to hydrogen peroxide, whereas Tiron is a nonenzymatic scavenger of superoxide (13). Catalase converts hydrogen peroxide to water, whereas the iron chelator DFO inhibits the iron-dependent conversion of hydrogen peroxide to the highly reactive hydroxyl radical (3). SOD (100 U/ml), Tiron (25 μM), catalase (300 U/ml), and DFO (10 μM) (Table1) were all as effective as EGF and high-dose insulin in maintaining viability of MPT cell monolayers (Fig.1).

View this table:
Table 1.

Effect of scavengers of ROS and antioxidants on viability of MPT cells following growth factor deprivation

Effect of Caspase Inhibition on the Viability of MPT Cells Subjected to Growth Factor Withdrawal

The effect of zVAD-fmk, a cell-permeant inhibitor of caspase activity, on the viability of MPT cells subjected to growth factor deprivation was determined. zVAD-fmk resulted in a dose-dependent increase in survival in MPT cells in the absence of growth factors (Fig.9).

Fig. 9.

Dose-dependent effect of the caspase inhibitor, zVAD-fmk, on confluence of MPT cells subjected to growth factor-free conditions for 10 days. Number of viable MPT cells (adherent and excluding trypan blue) present after 5 days of incubation in growth factor-free medium or supplemented with zVAD are expressed as a percent of the absolute number of cells in freshly confluent control monolayers (223,000 ± 3,000/well;n = 10). * P < 0.01 compared with growth factor-free conditions.

Flow Cytometric Analysis of the Effect of Growth Factors, Antioxidants, and zVAD-fmk on Apoptosis of MPT Cells Subjected to Growth Factor Deprivation

Because EGF and high-dose insulin are both potent growth factors for MPT cells, the preservation of monolayer confluence by these hormones (Fig. 1) could theoretically be due entirely to an increased rate of proliferation of MPT cells. To confirm that EGF and insulin inhibit apoptosis of MPT cells, we used fluorescence-activated cell sorting (FACS) analysis to quantify the proportion of cells remaining viable after incubation in the presence and absence of these factors for 5 days (Fig. 10). The proportion of viable cells was increased from 16% (in the absence of growth factors) (Fig. 10 A) to 59% in the presence of EGF (Fig. 10 B) and to 58% in the presence of insulin (Fig.10 C). We have also confirmed by FACS analysis that apoptosis of MPT cells in the absence of growth factors is inhibited by the antioxidant DFO (Fig.10 D) and the caspase inhibitor zVAD-fmk (Fig. 10 E). The proportions of viable cells in the absence of growth factors (16%) (Fig.10 A) was increased to 60% (Fig.10 D) and 56% (Fig.10 E) by DFO (10 μM) and zVAD-fmk (5 μM) respectively.

Fig. 10.

Quantitation of apoptosis of MPT cells using flow cytometric analysis. For each sample, adherent and detached MPT cells were pooled and stained with H-33342 and analyzed by flow cytometry. A constant number of events (10,000) were analyzed per sample. Each sample was analyzed in two ways. 1). Cells were analyzed in all scattergrams on the left on the basis of size (FLS) vs. granularity (side scatter). The normal population (low granularity and normal size) is denoted asregion 1(R1), and the percentage of cells inR1 is shown in each scattergram.2) Cells are also analyzed as the log fluorescent intensity of Hoechst staining vs. size (FLS); these analyses are represented by the rightcolumn of scattergrams. Viable cells were defined in these scattergrams on the basis of normal size plus faint H-33342 fluorescence (right bottom quadrant), whereas apoptotic cells were defined as small cells with increased intensity of H-33342 intensity (left top quadrant). For each variable, the percentage of normal and apoptotic cells determined on the basis of cell size/granularity (leftscattergrams) or cell size/H-33342 intensity (right scattergrams) was comparable. Percentage of viable cells after 10 days in hormone-free medium (A) is 23% (left) to 16% (right). Incubation of cells in presence of either EGF (B) or high-dose insulin (C) increases the proportion of viable cells to ∼60%. Incubation of cells in presence of deferoxamine (DFO, D) or zVAD-fmk (E) also markedly increases the proportion of viable cells to ∼60%. Figure is representative of 6 experiments.

Effect of Growth Factor Deprivation on Caspase Activity in MPT Cells

PARP is a DNA repair protein that has been identified as a substrate for caspase 3 (apopain/CPP32). Caspase 3 inactivates PARP by cleaving the intact 113-kDa PARP molecule into a smaller and inactive 89-kDa fragment (12, 31, 45). The recognition of the smaller 89-kDa degradation product of PARP in cell lysates by Western blotting has become an established method for detecting caspase activation. We demonstrate that subjecting MPT cells to growth factor deprivation for 10 days results in degradation of PARP (Fig.11; lane 1) and, by inference, causes activation of caspase 3. We also demonstrate that activation of caspase 3 by growth factor deprivation is inhibited by the presence in the culture medium of EGF, insulin, Tiron, and zVAD-fmk (Fig. 11; lanes 2, 3,4, and5, respectively).

Fig. 11.

Western blot of poly(ADP-ribose) polymerase (PARP). In lysates of MPT cells subjected to growth factor deprivation for 5 days (lane 1), two bands of PARP are visible; a 113-kDa band, which represents intact PARP, and an 89-kDa fragment, which is the cleavage product of PARP after proteolysis by caspase 3. Presence of the 89-kDa band represents activation of caspase 3 in growth-factor-deprived cells. In contrast, MPT cells incubated in presence of EGF (lane 2), insulin (lane 3), 4,5-dihydroxy-1,3-benzene disulfonic acid (Tiron, lane 4), or zVAD-fmk (lane 5) demonstrate only a single intact band of PARP (113 kDa), indicating that caspase 3 activity was suppressed by all these conditions. Blot is representative of 3 experiments.

DISCUSSION

We initially compared the response of three renal tubular cell lines, primary cultures of MPT cells, OK cells, and MDCK cells, to growth factor deprivation. As we have reported previously (23), MPT cells respond to a lack of growth factors by rapidly losing viability and confluence over a 10- and 14-day period (Figs. 1 and 2). We have demonstrated that these cells die by apoptosis using the biochemical marker of DNA fragmentation (Fig. 3) as well as by morphological criteria including phase-contrast and fluorescent microscopy (Fig. 5) and EM (Figs. 6 and 7). These data demonstrate that survival of MPT cells depends on the presence of these renal growth factors and are consistent with the hypothesis that apoptosis of renal tubular cells represents a default pathway that is always ready to be activated unless constantly inhibited by essential survival factors in adequate concentrations (39).

Interestingly, the response of two immortalized renal cells lines (OK and MDCK cells) to growth factor deprivation is aberrant. Unlike MPT cells, both OK and MDCK cells continue to proliferate in the absence of renal growth factors, so that after 2 wk, the monolayers of both cell lines are still confluent and the numbers of viable cells in each cell line have increased two- to threefold above those in freshly confluent monolayers (Fig. 2). The degree of apoptosis induced by growth factor deprivation, as assessed by DNA fragmentation, differs between OK and MDCK cells. DNA fragmentation after 14 days of growth factor deprivation is substantial in MDCK cells (Fig. 4; lane 4), indicating that, while some MDCK cells are actively proliferating, others are dying. The net effect in MDCK cells is an increase in cell number (Fig. 2). In contrast, in OK cells, there is no clear increase in the degree of DNA fragmentation compared with freshly confluent cells (Fig. 4; lane 2 vs. 3). Thus, while OK cells, like MDCK cells, continue to grow despite the absence of growth factors, they differ from both MPT cells and MDCK cells in that apoptosis is not a prominent response to lack of growth factors in this cell line. These studies indicate that the response to growth factor deprivation is abnormal in both of these immortalized cell lines. For these reasons, we have chosen to study the default pathway of apoptosis in monolayers of primary cultures of MPT cells, which rapidly lose confluence and undergo apoptosis in the absence of growth factors.

We have examined the hypothesis that oxidant stress is involved in the signaling pathway that leads to the default pathway of apoptosis. Two lines of evidence support a role for oxidant stress in this pathway. First, antioxidants were shown to maintain the viability of MPT cells in the absence of growth factors (Table 1). Using flow cytometry (Fig.10), we demonstrate that the maintenance of viability by the antioxidant DFO is mediated by inhibition of apoptosis. These antioxidants are as effective as the renal growth factors EGF and high-dose insulin in maintaining MPT confluence (Fig. 1) and inhibiting apoptosis (Fig. 10) of MPT cells. The second line of evidence supporting a role for oxidant stress in the default pathway of apoptosis is that direct measurement of superoxide demonstrates a substantial (40%) elevation in the level of superoxide anion in MPT cells subjected to growth factor withdrawal compared with freshly confluent cells (Fig. 8). EGF and insulin prevent the increase in superoxide levels in MPT cells induced by growth factor deprivation (Fig. 8). Our data are consistent with other studies demonstrating that antioxidants protect against apoptosis induced by various triggers, such as cisplatin (25) and tumor necrosis factor-α (17). The role played by ROS in the commitment and/or execution phases of apoptosis remains uncertain (5, 17). ROS most likely promote apoptosis by acting as second messengers in the complex signal transduction pathways involved in activation of the death program (17).

We have tried to further elucidate the role of oxidant stress in the signaling pathway of apoptosis by determining whether oxidant stress acts upstream or downstream of activation of the capsases. Substantial evidence now suggests that the execution program is similar in all cell types (1, 11, 19, 22, 27). The Ced-3gene, which is a death effector in the nematodeCaenorhabditis elegans, has structural and functional homology to the mammalian protease interleukin converting enzyme (ICE). Although ICE itself does not appear to be an effector of apoptosis in mammalian cells (6), several ICE/CED-3 homologs have been identified that appear to be critical components of the cell death machinery (27, 33, 36, 50). These “ICE-like” proteases, recently named caspases (1), are members of an emerging family of aspartate-specific cysteine proteases that are constitutively produced and present in an inactive form in the cytosol of most if not all cells (27). Proteolytic activation of these proenzymes to form a heterodimeric catalytic domain (1) precipitates a cascade of ICE-like protease activation (33) that ultimately mediates apoptosis by cleaving multiple substrates within the cells (27, 50). In further support of a generalized role for ICE-like proteases in the execution of apoptosis, a number of inhibitors of these proteases have been shown to prevent apoptosis by diverse stimuli (4, 10, 19, 31, 45)

In this study, we show that caspases are activated by growth factor deprivation as evidenced by cleavage of the DNA repair enzyme PARP, a well-known substrate for caspase 3. Caspase 3 cleaves intact, active PARP (113 kDa) into a smaller and inactive 89-kDa fragment (7, 12, 31,45). We demonstrate that cleavage of PARP, and by inference activation of caspase 3, is inhibited by renal growth factors (EGF and high-dose insulin), by the superoxide scavenger Tiron, and by zVAD-fmk a cell-permeant inhibitor of caspase 3 (19, 30, 38) (Fig. 11).

We also show that inhibition of caspase 3 with zVAD-fmk inhibits apoptosis of MPT cells induced by growth-factor withdrawal in a dose-dependent fashion (with an IC50 of ∼1 μM) (Figs. 9 and10). zVAD-fmk has also been shown to inhibit apoptosis induced by other stimuli such as Fas (30), staurosporine (19), ceramide (38), and dexamethasone (10). However, it is important to point out that zVAD-fmk inhibits activation of caspases and cell death without preventing the increase in superoxide levels associated with growth factor deprivation (Fig. 8). These data suggest that oxidant stress plays a role in the default pathway of apoptosis by acting upstream of caspases.

In summary, we have demonstrated that the renal trophic factors EGF and IGF-I are potent inhibitors of the default pathway of apoptosis in MPT cells. We provide novel evidence that ROS play a role in apoptosis of renal tubular cells induced by growth factor deprivation. Our data are consistent with the hypothesis that oxidant stress participates in the signaling pathways responsible for inducing activation of caspases that mediate the cell death program.

Acknowledgments

We are grateful to John Daley for technical assistance in flow cytometric analysis and to Dr. Yuhui Xu for the electron microscopic analysis of apoptotic and nonapoptotic cells.

Footnotes

  • Address for reprint requests: W. Lieberthal, Renal Section, Evans 428, Boston Medical Center, 88 East Newton St., Boston, MA 02118.

  • This work was supported by National Institutes of Health Grants DK-37105 (W. Lieberthal), DK-52898 (W. Lieberthal), HL-53031 (W. Lieberthal), AR/AI-42732 (J. S. Levine), and HL-55425 (P. J. Pagano), American Cancer Society Grant IN97-N (J. S. Levine), by a Clinical Scientist award from the National Kidney Foundation (J. S. Levine), a National American Heart Association Grant-in-Aid 95011900 (P. J. Pagano).

REFERENCES

View Abstract