Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Apoptosis or programmed cell death is a physiologic mechanism responsible for the elimination of unwanted cells at various stages of development, injury, and repair (1, 2). Surface epithelial cells frequently become apoptotic when detached from their underlying basement membrane, a process termed anoikis (3). Mechanical tension generated by the link between the cytoskeleton and cell-surface integrins facilitates construction of focal adhesions (4). Periapical rings of actin link to tight and adherens junctions that bind to neighboring cells (5, 6). This combination of mechanical tension and multiple links to cell-cell and cell- matrix junctions contribute to the maintenance of cell shape. How detachment from extracellular matrix or disruption of cell-cell junctions leads to internal rearrangement of the cell cytoskeleton and whether such detachment precedes or follows the initiation of apoptosis are not clear.

Major cytoskeletal filaments, including microtubules, cytokeratin, and actin, are degraded during the execution phase of apoptosis (4). Actin is a substrate for caspases (7, 8) and for calpains (9, 10) in some cell models, and a direct link between actin depolymerization and DNA degradation has been suggested (7). Although actin may be resistant to cleavage in some in vivo models (8), targeting of actin and other intermediate filaments is responsible in part for the collapse of cell shape during the execution phase of apoptosis. Degradation of actin filaments can disrupt required mechanical tension and lead to signals that may facilitate cell detachment.

There is increasing evidence that disruption of cytoskeletal proteins may in itself induce cell death. Disruption of microtubule turnover by taxol and vincristine, which phosphorylate Raf1 and Bcl-2, leads to apoptosis and detachment (11, 12). Cleavage and activation of the actin-associated protein Gas2 by interleukin-1 converting enzyme-like proteases lead to disruption of the actin cytoskeleton (13, 14). Caspase-3 cleaves the actin filament, severing protein gelsolin; the resulting fragment then cleaves actin and in turn leads to morphologic changes in apoptotic cells (15). However, a direct link between cytoskeletal disruption and subsequent apoptosis has not been demonstrated.

Our laboratory recently has demonstrated that airway epithelial cells express both CD95 (Fas) and its ligand (FasL) (16). Situations that permit the apposition of both receptor and ligand may initiate apoptosis, and this may be one explanation for the epithelial damage and denudation noted in asthma (17, 18). Cell deformation and shape changes are seen in the airway epithelium in states of inflammation (17), but the causes are obscure. The aim of this study was to investigate whether disruption of the actin cytoskeleton itself could initiate apoptosis in airway epithelial cells. We studied primary airway epithelial cells, an airway epithelial cell line, and a non-airway epithelial cell line known to withstand actin filament disruption (19). Our data demonstrate that preventing actin filament elongation with cytochalasin D, or inducing actin filament aggregation with jasplakinolide, initiated apoptosis in both primary airway epithelial cells and an airway epithelial cell line. This event did not require ligation of Fas for its initiation or propagation. These data suggest that the actin cytoskeleton, a target of apoptosis, may also be an early modulator of apoptotic commitment.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Materials

Streptomycin, penicillin, L-glutamine, cytochalasin D, d-mannitol, insulin, hydrocortisone, transferrin, triiodothyrinine, epinephrine, human epidermal growth factor (hEGF), bovine pituitary extract (BPE), anti-glucose-6-phosphate dehydrogenase (G6PD) antibody, isoproterenol, and anti-actin monoclonal antibody (mAb) were purchased from Sigma, Inc. (St. Louis, MO) or GIBCO BRL, Inc. (Grand Island, NY). Fetal calf serum (FCS) was purchased from Hyclone (Logan, UT) and was heat-inactivated before use. The peptides glycine-arginine-glycine-glutamic acid-serine-proline (GRGESP) and glycine-arginine-glycine-aspartic acid-serine-proline (GRGDSP) were obtained from Peninsula, Inc. (San Diego, CA). The antihuman CD95 blocking mAb ZB4 and the antihuman FasL mAb NOK1 were purchased from PharMingen, Inc. (San Diego, CA). The anti-CD95 ligating mAb CH11 was purchased from PanVera Corp. (Madison, WI). The anti-vinculin monoclonal immunoglobulin (Ig) G hVin-1 was purchased from Upstate Biotechnology (Lake Placid, NY). Anti-paxillin monoclonal IgG clone 349 and anti-Fas-associated death domain (FADD) monoclonal IgG were purchased from Transduction Laboratories (Lexington, KY). Antihuman caspase-8 monoclonal IgG C15 was a generous gift of Marcus Peter, Ph.D. (University of Chicago, Chicago, IL). The C20 mAb directed against the cytoplasmic tail of Fas was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Rhodamine-labeled phalloidin, and fluorescein isothiocyanate-conjugated and Cy3-conjugated goat antimouse IgG were purchased from Molecular Probes (Eugene, OR). Terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate biotin nick end-labeling (TUNEL) TACS II fluorescent assay kits were purchased from Trevigen, Inc. (Rockville, MD). The caspase inhibitor peptides Z-Val-Ala-Asp-fluoromethylketone (z-VAD-fmk) and N-acetyl-(Asp-Glu-Val)- 3 amino-4-oxobutaonic acid (Ac-DEVD-cho) were purchased from Calbiochem-Novabiochem Corp. (La Jolla, CA). Acetyl-Ile-Glu-Thr-Asp--7-amino-4-trifluoromethylcoumarin (Ac-IETD-AFC) and acetyl-Ile-Glu-Thr-Asp-CHO aldehyde (Ac-IETD-CHO) were purchased from PharMingen, Inc.

Culture of the 1HAEo Airway Epithelial Cell Line

1HAEo cells, a gift of Dieter Gruenert (University of California San Francisco, San Francisco, CA), are human airway epithelial cells transformed with SV-40 (20, 21). Madin-Darby canine kidney (MDCK) epithelial cells were a gift of Eugene Chang (University of Chicago). Both cell lines were grown on collagen IV (10 µg/ml)- coated chamber slides in Dulbecco's modified essential medium containing 10% FCS, 2 mM L-glutamine, 100 µg/ml streptomycin, and 100 U/ml penicillin G. Cells were incubated at 37°C in 5% CO₂ and used when ~ 90% confluent. Slides were washed twice in fresh culture medium after which medium was replaced. Cells were kept in 10% FCS during all experiments to prevent confounding of apoptosis results by withdrawal of any needed growth factors. Agents were dissolved in either dimethyl sulfoxide (DMSO) or phosphate-buffered saline (PBS). Agents or their appropriate vehicle control were added, and cells were incubated for 1 to 24 h at 37°C. In all experiments, the final concentration of DMSO never exceeded 0.1%. In experiments using mannitol, the osmolarity of the medium was measured using a model 3D2 osmometer (Advanced Instruments, Inc., Newberry, MA). At the conclusion of experiments, chamber slides were washed once in fresh medium and fixed in 4% neutral buffered formalin for further processing. Protein from plates was collected as described for Western blot.

Culture of Primary Airway Epithelial Cells

Primary normal human bronchial epithelial (NHBE) cells were purchased from Clonetics, Inc. (Walkersville, MD). These cells are derived from a single donor and are supplied as first passage cells. Cells were placed into defined medium (Clonetics) containing 5 µg/ml insulin, 0.5 µg/ml hEGF, 10 mg/ml transferrin, 6.5 µg/ml triiodothyrinine, 0.5 mg/ml hydrocortisone, 0.5 mg/ml epinephrine, and 2 ml/liter BPE. Cells were subcultured and used between passages 5 and 7 when ~ 50% confluent. Experiments were done as for the 1HAEo cell line, except that cells were kept in defined medium and not 10% FCS.

In an additional experiment, primary epithelial cells were collected from a human airway using a method we have described previously (22). Briefly, central airway segments were obtained from a patient who underwent surgery for lung cancer. Epithelial cells were grown in serum-free keratinocyte medium (K-SFM; GIBCO BRL, Inc.) supplemented with 0.2 ng/ml epidermal growth factor (EGF), 25 µg/ml BPE, 1 µM isoproterenol, 200 U/ml penicillin, and 200 µg/ml streptomycin. These cells (denoted SHBE) were incubated at 37°C in 5% CO₂, subcultured, and used when ~ 80% confluent. At that time, medium was replaced with K-SFM containing EGF, BPE, penicillin, streptomycin, and 1 mM CaCl₂ but without isoproterenol, and then treated with either cytochalasin D or jasplakinolide as noted.

Assay for Cell DNA Nicking

Apoptotic cells in fixed monolayers were demonstrated by labeling free 3'-hydroxyl groups of DNA using a Trevigen TUNEL fluorescent assay kit. Slides were counterstained with 5 U/ml rhodamine-phalloidin in PBS for 30 min, washed three times in PBS, and then stained with 1 mM Hoechst 33258 in water for 45 s. Representative images were collected using a Sensys 12-bit cooled CCD camera (Photometrics, Inc., Tucson, AZ) and a Nikon (Rolling Meadows, IL) fluorescence microscope. TUNEL-positive nuclei and Hoechst-stained nuclei were counted in each image as the area of the nuclei in pixels after visual thresholding and exclusion of extraneous positive pixels using Spectrum IP software (IP Labs, Vienna, VA) on a Macintosh computer. TUNEL-positive cells were expressed as the percentage of the thresholded area of the TUNEL-stained image divided by the thresholded area of the Hoechst-stained image. Preliminary experiments demonstrated a high correlation with manual counting methods using this technique.

Extraction of F-Actin and G-Actin

After interventions, plates were washed twice with PBS and then once with CSK buffer (0.01 M [1,4-piperazinebis (ethane sulfonic acid)], 0.3 M sucrose, 0.025 NaCl, 1 mM ethyleneglycol-bis-(-aminoethyl ether)-N,N'-tetraacetic acid [EGTA], and 5 mM MgCl₂). Slides then were incubated for 5 min at room temperature in CSK buffer containing 1% Triton X-100. Supernatants were collected and centrifuged at 108,000 × g for 50 min at 25°C. Pellets containing noncytoskeletal F-actin were suspended in 500 µl of 8 M urea each and stored at 70°C. Supernatants were precipitated in 5 vol methanol overnight at 70°C, and then centrifuged at 10,000 × g for 1 h at 4°C. Pellets containing G-actin were suspended in 500 µl 8 M urea each and stored at 70°C. The remaining cell material on the slides was washed once with CSK, collected in 8 M urea, vortexed briefly, and centrifuged for 5 min at 14,000 rpm. Supernatants containing cytoskeletal F-actin were stored at 70°C. Samples were separated on a sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) minigel and transferred to nitrocellulose membranes. Beginning cell numbers in each experiment were equal, and equal volumes of extracted protein were loaded in each lane. Immunodetection was performed using an enhanced chemiluminescence (ECL) protocol (Amersham, Arlington Heights, IL).

Western Blot Assay

After interventions, cells were lysed in buffer containing 1% Nonidet P-40 (NP-40), 0.25% Na-deoxycholate, 150 mM NaCl, 1 mM EGTA, 1 mM phenylmethysulfonyl fluoride, 1 µg/ml aprotinin, 1 µg/ml pepstatin, 1 µg/ml leupeptin, 1 mM Na₃VO₄, and 1 mM NaF for 15 min at 4°C. Samples were centrifuged at 14,000 rpm for 10 min at 4°C, after which supernatants were frozen at 70°C. Proteins were separated on an SDS-PAGE minigel and transferred onto nitrocellulose membranes. Immunodetection was performed using an ECL protocol. In some experiments, membranes were stripped and reprobed with an antibody for G6PD to control for protein loading.

Immunoprecipitation of the Death-Inducing Signaling Complex

This method is similar to that used by Medema and coworkers (23). After treatment, cells were washed once with PBS and collected in lysis buffer for 15 min at 4°C. Immunoprecipitation of the complex was done using the APO-1 mAb directed against the cytoplasmic tail of Fas (20 µg/ml) for 1 h at 4°C. The complex was collected by incubation with 20 µl of protein A/G agarose beads overnight at 4°C. Beads were washed three times with PBS, after which the complex was released by boiling for 2 min and brought to equal volume in Laemmli buffer. Equal volumes of precipitated proteins, reflecting the original volume of lysates and number of cells, were separated and probed using the Western blot protocol and an anti-FADD antibody.

Caspase-8-Like Activity Assay

After interventions, cells were lysed in buffer containing 1% NP-40, 0.25% Na-DOC, 150 mM NaCl, and 1 mM EGTA, after which samples were frozen quickly in dry ice-ethanol. Samples (200 µl) were resuspended in 25 mM N-2-hydroxyethylpiperazine-N'-ethane sulfonic acid, pH 7.5, 0.1% 3-[(3-cholamidopropyl) dimethyamino]- 1-propanesulfonate, 5 mM ethylenediaminetetraacetic acid, and 10 mM dithiothreitol to which 5 µg/ml Ac-IETD-AFC was added. Ac-IETD-CHO (2 µg/ml) was added to duplicate aliquots of samples to subtract caspase-8-like activity. Samples were incubated for 1 h at 37°C. Fluorescence then was measured at an excitation wavelength of 400 nm and emission wavelength of 505 nm, and was expressed finally as fluorescence per milligram cell lysate protein.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Cytochalasin D Causes Apoptosis in 1HAEo Cells

Cytochalasin D caps the barbed or rapidly growing end of actin filaments and prevents the addition of G-actin monomers (24, 25). This causes complete disruption of actin filament integrity by creating short filaments and aggregates that function at a reduced capacity (26). Treatment of confluent 1HAEo monolayers with 0.04 to 0.5 µg/ml cytochalasin D for 0.5 to 24 h caused a concentration- and time-dependent derangement in actin filament structure. Disordered actin filaments, loss of stress fibers, and cell shrinkage were seen within 1 h at concentrations of cytochalasin D  0.1 µg/ml (Figure 1) and were maximal with the highest concentration of cytochalasin D (0.5 µg/ml) used. Fewer than 10% of cells detached from the matrix during this period as determined by cell nuclei counting from fields selected at random in the monolayers. The changes seen correlated to an increase in both G-actin and noncytoskeletal F-actin over 5 h on Western blot (Figure 1).

View larger version (109K): [in this window] [in a new window]   View larger version (24K): [in this window] [in a new window]   Figure 1.   Morphology of actin filaments in 1HAEo cells after treatment with cytochalasin D. (A) Confluent 1HAEo cells were treated with control medium (a) or medium containing 0.5 µg/ml cytochalasin D for 0.5 (b), 1 (c), or 5 h (d). Cells were fixed and actin filaments were visualized by staining with rhodamine-phalloidin. Bar represents 50 µm. (B). Confluent 1HAEo cells were treated with control medium (a) or medium containing 0.04 (b), 0.1 (c) or 0.5 µg/ml cytochalasin D (d) for 5 h. Cells were fixed and actin filaments were visualized by staining with rhodamine-phalloidin. In both A and B, there is a time- and concentration-dependent disruption of actin filaments after treatment with cytochalasin D. Bar represents 50 µm. (C) Confluent 1HAEo cells or MDCK cells were treated with control medium or medium containing 0.5 µg/ml cytochalasin D for 0 to 5 h. Cell lysate fractions containing cytoskeletal and noncytoskeletal F-actin and G-actin were resolved on SDS-PAGE and probed for the presence of actin. There is a decrease in cytosolic F-actin and increase in monomeric G-actin over time. Numbers represent time of treatment in hours.

Cytochalasin D treatment also elicited apoptosis in 1HAEo cells. Cells treated with 0.04 to 0.5 µg/ml cytochalasin D for 5 h demonstrated small nuclei with chromatin condensation as demonstrated by Hoechst 33258 stain (Figure 2). TUNEL labeling was increased in a time- and concentration-dependent manner after cytochalasin D treatment (Figure 2). The number of TUNEL-positive cells after 5 h of treatment with 0.5 µg/ml cytochalasin D was 37.0 ± 4.0, versus 5.5 ± 0.8% for control (P < 0.000001 by t test; n = 15). The number of TUNEL-positive cells after 24 h was 29.1 ± 2.8 versus 4.9 ± 0.8% for control (P = 0.00004 by t test; n = 10).

View larger version (69K): [in this window] [in a new window]   View larger version (15K): [in this window] [in a new window]   View larger version (129K): [in this window] [in a new window]   Figure 2.   Apoptosis of 1HAEo cells grown on collagen IV after treatment with cytochalasin D. (A) Confluent 1HAEo cells were treated with control medium (a) or medium containing 0.5 µg/ml cytochalasin D for 0.5 (b), 1 (c), or 5 h (d). Cells were fixed and stained with Hoechst 33258 to demonstrate nuclear morphology. Images represent matched fields to corresponding images in Figure 1A. Bar represents 50 µm. (B) Confluent 1HAEo cells were fixed 1 to 5 h after treatment with 0.04 to 0.5 µg/ml cytochalasin D and stained to demonstrate DNA nicking by fluorescent TUNEL assay. In both A and B, there is a time- and concentration-dependent increase in apoptosis after treatment with cytochalasin D. Data are presented as mean ± standard error of the mean (n = 5 to 10). (C) Confluent MDCK cells were treated with control medium (a and c) or medium containing 0.5 µg/ml cytochalasin D (b and d) for 5 h. Cells were fixed 5 h later and stained with rhodamine-phalloidin (a and b) to demonstrate actin filaments and Hoechst 33258 (c and d) to demonstrate nuclear morphology. In contrast to the 1HAEo cells, there is little change in nuclear morphology in MDCK cells, even though significant cell shrinkage has occurred and actin filament integrity has been substantially compromised. Bar represents 50 µm.

In contrast, confluent MDCK cells treated with 0.5 µg/ ml cytochalasin D for 5 or 24 h did not undergo apoptosis. Whereas cytochalasin D induced similar changes in actin filament morphology and cell shape as seen in the 1HAEo cell line, there was little change in nuclear morphology, even after 24 h (Figure 2). The number of TUNEL-positive cells after 5 h of treatment with 0.5 µg/ml cytochalasin D was 3.9 ± 1.1, versus 1.6 ± 0.7 for control (n = 7), and after 24 h was 5.3 ± 1.1 versus 3.0 ± 1.0% for control (n = 7). These data suggested a clear difference between an airway and nonairway epithelial cells, and further suggested that cell shape change or shrinkage alone was insufficient to elicit apoptosis.

Withdrawal of cytochalasin D treatment did not reverse apoptosis. In these experiments, confluent 1HAEo monolayers were treated with 0.5 µg/ml cytochalasin D for 5 h, after which medium was replaced with medium containing 10% FCS for an additional 19 h. The number of TUNEL-positive nuclei in cells treated in this manner was 25.1 ± 1.8 versus 28.2 ± 1.4% in cells treated with 0.5 µg/ml cytochalasin D for 24 h without withdrawal (n = 4). Cytoskeletal morphology and cell shape were normal in nonapoptotic cells after withdrawal of cytochalasin D, whereas apoptotic cells were smaller and had a collapsed actin cytoskeleton (Figure 3).

View larger version (68K): [in this window] [in a new window]   Figure 3.   Apoptosis of 1HAEo cells after treatment with, and then removal of, cytochalasin D. Confluent 1HAEo cells were treated with 0.5 µg/ml cytochalasin D for 5 h, after which cells were recovered in control medium plus 10% FCS for 19 h. Cells were then fixed and stained with rhodamine-phalloidin (a) to demonstrate actin filaments and Hoechst 33258 (b) to demonstrate nuclear morphology. Only those cells that fail to respond to the withdrawal of cytochalasin D are apoptotic. Bar represents 50 µm.

Jasplakinolide Causes Apoptosis in 1HAEo Cells

Jasplakinolide is a cell-permeable monocyclic peptide (27, 28) that binds to the same sites on F-actin as phalloidin (29, 30) and induces actin polymerization and aggregation (28, 29, 31). Jasplakinolide can inhibit the growth of prostate carcinoma epithelial cell lines in vitro (30) and can enhance apoptosis induced by cytokine deprivation in CTLL-20 lymphoma cells (32). In these studies, the effects of jasplakinolide were associated with its ability to polymerize F-actin. Neither report, however, demonstrated an independent effect of jasplakinolide on cell apoptosis. To examine whether jasplakinolide would induce apoptosis in airway epithelial cells, 1HAEo cells were treated with 1.0 µM jasplakinolide for 5 to 24 h. This treatment caused aggregation of actin filaments as noted on rhodamine-phalloidin staining (Figure 4). This correlated to a substantial loss of G-actin and an increase in noncytoskeletal F-actin over 5 h on Western blot (Figure 4). As in the experiments with cytochalasin D, fewer than 10% of cells detached from the matrix over 24 h after treatment with jasplakinolide as determined by cell nuclei counting from fields selected at random in the monolayer.

View larger version (133K): [in this window] [in a new window]   View larger version (84K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   View larger version (162K): [in this window] [in a new window]   Figure 4.   Apoptosis of 1HAEo cells grown on collagen IV after treatment with jasplakinolide. (A) Confluent 1HAEo cells were treated with control medium (a) or with medium containing 0.03 (b), 0.1 (c), 0.3 (d), 1.0 (e), or 3.0 µM ( f ) jasplakinolide for 5 h. Cells were fixed and actin filaments were visualized by staining with rhodamine-phalloidin. Bar represents 50 µm. There is a concentration-dependent disruption of actin filaments after treatment with jasplakinolide. (B) Confluent 1HAEo cells or MDCK cells were treated with control medium or medium containing 1.0 µM jasplakinolide for 0 to 5 h. Cell lysate fractions containing cytoskeletal and noncytoskeletal F-actin and G-actin were resolved on SDS-PAGE and probed for the presence of actin. There is an increase in cytosolic F-actin and decrease in monomeric G-actin over time. Numbers represent time of treatment in hours. (C) Confluent 1HAEo cells were treated with control medium (a) or with medium containing 0.03 (b), 0.1 (c), 0.3 (d), 1.0 (e), or 3.0 µM ( f ) jasplakinolide for 5 h. Cells were fixed and stained with Hoechst 33258 to demonstrate nuclear morphology. Bar represents 50 µm. (D) Cells were treated with 0.03 to 3.0 µM jasplakinolide for 5 h, then fixed and stained to demonstrate DNA nicking by fluorescent TUNEL assay. In both C and D, there is a concentration-dependent increase in apoptosis after treatment with jasplakinolide. Data are presented as mean ± standard error of the mean (n = 4 experiments). (E) Confluent MDCK cells were treated with control medium (a and d) or with medium containing 0.1 (b and e) or 1.0 µM (c and f ) jasplakinolide for 5 (a-c) or 24 h (d-f ). Cells were fixed and actin filaments were visualized by staining with rhodamine-phalloidin. Bar represents 50 µm. There is a concentration-dependent disruption of actin filaments after treatment with jasplakinolide at both time points.

Treatment with jasplakinolide also elicited apoptosis as demonstrated by changes in nuclear morphology and positive TUNEL staining (Figure 4). The number of TUNEL-positive cells after 5 h of treatment with 1.0 µM jasplakinolide was 23.2 ± 5.3 versus 4.4 ± 0.7% for control (P = 0.006 by t test; n = 11). The number of TUNEL-positive cells after 24 h was 25.7 ± 5.0 versus 8.3 ± 1.8% for control (P = 0.04 by t test; n = 4).

As in the cytochalasin experiments, treatment of confluent MDCK cells with 1.0 µM jasplakinolide for 5 or 24 h elicited actin filament aggregation (Figure 4). However, these cells did not undergo apoptosis as noted by either nuclear morphology or by TUNEL labeling. The number of TUNEL-positive cells after 5 h of treatment with 1.0 µM jasplakinolide was 0.9 ± 0.3 versus 0.7 ± 0.1% for control (n = 4), and after 24 h was 2.1 ± 0.9 versus 0.3 ± 0.1% for control (n = 4). These data demonstrated that cellular aggregation of actin induced by jasplakinolide could initiate apoptosis in 1HAEo cells independent of another required stimulus. As with cytochalasin D, whereas the changes in both actin filament integrity and cell shape after treatment with jasplakinolide were similar in the two cell lines, only the airway epithelial cells became apoptotic.

Changes in Cell Shape Associated with Hyperosmolarity Do Not Elicit Apoptosis in 1HAEo Cells

Disruption of actin filament integrity with either cytochalasin D or jasplakinolide is associated with changes in cell shape as well as in nuclear morphology. It is possible that apoptosis is associated with membrane changes associated with cell shrinking and not changes in actin filament integrity. To test this, 1HAEo cells were treated with increasing concentrations of mannitol to the culture medium. Incubation of cells in medium (325 mOsm alone) plus 100 to 1,000 mM mannitol (420 to 1,330 mOsm) elicited concentration-dependent changes in cell and nuclear morphology (Figure 5). Cells pulled away from neighboring cells and had a smaller volume as noted by phase-contrast microscopy. Cell nuclei became vacuolated and misshapen, but there were no increases in cell chromatin intensity on Hoechst staining. Despite the changes in cell shape and nuclear morphology, rhodamine phalloidin staining demonstrated the continued presence of stress fibers and actin filaments (Figure 5). These data suggested that the changes in cell shape occurred without any disruption in actin filament morphology.

View larger version (113K): [in this window] [in a new window]   View larger version (96K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   Figure 5.   Changes in cell shape, cell nuclei, and actin filaments in 1HAEo cells after treatment with mannitol. (A) Confluent 1HAEo cells were treated with control medium (a) or medium containing 1,000 mM mannitol (b). Phase-contrast microscopy was done 10 min after adding mannitol. Concentration-dependent changes in cell morphology are seen, with cell shrinkage and pulling away from neighboring cells. Bar represents 250 µm. (B) Confluent 1HAEo cells were treated with control medium (a) or medium containing 100 (b), 200 (c), 300 (d), 500 (e), or 1,000 mM mannitol ( f ) for 5 h. Cells were fixed and stained with rhodamine-phalloidin to demonstrate actin filament morphology. Even with high concentrations of mannitol, stress fibers are clearly seen. Bar represents 50 µm. (C) Confluent 1HAEo cells were treated with control medium (a) or medium containing 100 (b), 200 (c), 300 (d), 500 (e), or 1,000 mM mannitol ( f ) for 5 h. Cells were fixed and stained with Hoechst 33258 to demonstrate nuclear morphology. Concentration-dependent changes in nuclear morphology are seen, with increasing vacuolization but no changes in chromatin density. Bar represents 100 µm. Images in B and D represent the same fields. (D) Confluent 1HAEo cells were fixed 5 h after treatment with 100 to 1,000 mM mannitol and stained to demonstrate DNA nicking by fluorescent TUNEL assay. There is no significant increase in apoptosis after treatment with mannitol (n = 4).

Cell shape changes induced by mannitol did not elicit apoptosis. The number of TUNEL-positive cells after 5 h of incubation in medium plus 1,000 mM mannitol (1,330 mOsm) was 4.6 ± 1.7 versus 2.2 ± 1.0% in medium (325 mOsm) alone (n = 4; P = not significant) (Figure 5). Incubation of cells in medium plus 500 mM mannitol followed by treatment with either cytochalasin D or jasplakinolide did not elicit greater apoptosis than treatment with either actin-disrupting agent alone (Table 1). Taken together, these data suggest that airway epithelial cells did not undergo apoptosis as a function of cell shrinking alone and that the apoptotic changes seen after addition of either cytochalasin D or jasplakinolide was due to actin filament disruption.

Cytochalasin, but Not Jasplakinolide, Treatment Decreases Focal Adhesion Protein Abundance in 1HAEo Cells

It is possible that disruption of actin filaments could cause dissolution of focal adhesions that bind cells to matrix, which then could initiate apoptosis (33). To test this possibility, 1HAEo cells were treated with either 0.5 µg/ml cytochalasin D or 1.0 µM jasplakinolide for 1 to 24 h. Whole cell protein lysates were then resolved on Western blot for the presence of paxillin, which localizes to the ends of actin stress fibers in focal adhesions (34), and vinculin, a structural component of focal adhesions (35). Treatment with jasplakinolide did not cause significant changes in the abundance of either protein (Figure 6). However, the abundance of both proteins was decreased after exposure to cytochalasin D for  1 h.

View larger version (55K): [in this window] [in a new window]   Figure 6.   Abundance of focal adhesion proteins in 1HAEo cells after treatment with either cytochalasin D or jasplakinolide. Confluent 1HAEo cells were treated with 0.5 µg/ ml cytochalasin D (CYD) or 1.0 µM jasplakinolide (JAS) for 1 to 24 h, after which whole cell protein lysates were resolved on Western blot and probed for the abundance of vinculin (mAb clone hVin-1) and paxillin (mAb clone 3A9). Equal quantities of protein were loaded in each lane. Membranes were stripped and reprobed with an antibody to G6PD to demonstrate lane loading. Cytochalasin D, but not jasplakinolide, decreases the abundance of both vinculin and paxillin. The numbers represent time in hours of treatment. Representative of three experiments.

Cytochalasin D-Induced Apoptosis Occurs Regardless of Binding of Available Fibronectin Receptors

Airway epithelial cells have fibronectin (FN) receptors consisting of ₅₁ integrin subunits (36), which can be induced during repair (37). Activation of these receptors either by binding FN or by soluble Arg-Gly-Asp-Ser peptide, which mimics the RGD sequence located within the III-10 module of FN (38, 39), prevents cell death (40). To test whether FN receptor binding would prevent cytochalasin D-induced apoptosis, 1HAEo cells were grown to confluence on collagen IV and then treated for 60 min with either 10⁴ M GRGDSP peptide or 10⁴ M of the control peptide GRGESP. Cells were then treated with 0.5 µg/ml cytochalasin D or control vehicle for 5 h. Addition of GRGDSP did not prevent either the disruption of actin filaments or the induction of apoptosis (Table 2). These data demonstrated that 1HAEo cells could not be protected from cytochalasin D-induced apoptosis by prior FN binding and activation.

Cytochalasin-Induced and Jasplakinolide-Induced Apoptosis Does Not Require Ligation of CD95

Both primary airway epithelial cells and 1HAEo cells have Fas and FasL (16). It is possible that cytochalasin D treatment could cause either the release of FasL, which then could bind the receptor, or that membrane deformations after cytochalasin D treatment could bring FasL into approximation with the receptor. To test this, 1HAEo cells were pretreated with both the Fas-blocking mAb ZB4 (10 µg/ml) and the FasL-binding monoclonal antibody NOK1 (10 µg/ ml) for 60 min. Cells were then treated with 0.5 µg/ml cytochalasin D for 5 h. Cells treated in this manner demonstrated similar TUNEL-positive staining compared with cells treated with cytochalasin D alone (Table 3). Similar results were obtained when cells were treated with jasplakinolide and both mAbs (Table 3). The same combination of NOK1 and ZB4 blocked completely apoptosis induced by the CD95-ligating mAb CH11 (Table 3). These observations demonstrated that both cytochalasin D- and jasplakinolide-induced apoptosis did not require binding of CD95.

Cytochalasin D-Induced and Jasplakinolide-Induced Apoptosis Requires Activation of Caspases

To determine whether the observed apoptosis after cytochalasin D treatment required the activation of caspases, 1HAEo cells were treated with 0.5 µg/ml cytochalasin D plus either 100 nM of z-VAD-fmk or 100 nM Ac-DEVD-cho for 5 h. Both caspase inhibitors attenuated TUNEL-positive staining induced by cytochalasin D (Table 4). In similar experiments, 1HAEo cells were grown to confluence and then treated with 1.0 µM jasplakinolide plus either 100 nM of z-VAD-fmk or 100 nM Ac-DEVD-cho for 5 h. Both caspase inhibitors attenuated TUNEL-positive staining induced by jasplakinolide (Table 4).

The activation of caspase-8 is an early step in the caspase protease cascade induced by CD95 (41) or by exposing the caspase-8 precursor to an active death-inducing signal complex (DISC) (42, 43). A recent report suggests that membrane oligomerization and cleavage also may activate caspase-8 and initiate apoptosis (44). To examine whether pro-caspase-8 was cleaved, 1HAEo cells were treated with either 0.5 µg/ml cytochalasin D or 1.0 µM jasplakinolide for 5 or 24 h. Pro-caspase-8 abundance was then determined by Western blot using the C15 mAb, which detects the caspase-8/a and 8/b isoforms (45). Depletion of the 55-kD pro-caspase-8 and appearance of cleaved fragments were seen 5 h after treatment with cytochalasin D (Figure 7). Increased abundance of cleavage fragments was not seen at any time point after treatment with jasplakinolide (Figure 7). Similarly, significant caspase-8-like activity was seen in whole cell protein lysates after treatment with cytochalasin D but not after treatment with jasplakinolide (Figure 7).

View larger version (38K): [in this window] [in a new window]   View larger version (15K): [in this window] [in a new window]   Figure 7.   Effect of cytochalasin D and jasplakinolide on cleavage of pro-caspase-8. (A) Confluent 1HAEo cells were treated for 1, 3, 5, or 24 h with 0.5 µg/ml cytochalasin D, 1.0 µM jasplakinolide, or sham. Whole cell lysates were resolved by Western blot and probed for abundance of pro-caspase-8. Equal quantities of protein were loaded in each lane. Membranes were stripped and re-probed with an antibody to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) to demonstrate lane loading. At both 5 and 24 h, both cytochalasin D and jasplakinolide induce an increase in caspase-8 fragments (arrows). Numbers at left denote molecular weights in kD; numbers at top denote time in hours. (B) Confluent 1HAEo cells were treated for 1 or 5 h with either 0.5 µg/ml cytochalasin D or 1.0 µM jasplakinolide (JAS). Caspase-8-like activity in whole cell lysate was determined by measurement of amino-4-trifluoromethylcoumarin fluorescence after incubation with Ac-IETD-AFC for 1 h. Treatment with cytochalasin D, but not jasplakinolide, leads to activation of caspase-8, (n = 3 or 4 experiments at each time point).

As noted previously, cytoskeletal rearrangement elicited by actin disruption could induce apoptosis by causing aggregation of Fas at the cell surface. Conformational changes then could lead to the generation of a DISC associated with the cytoplasmic tail of Fas, which in turn would elicit caspase-8 cleavage and activation (23). To test this possibility, we treated confluent 1HAEo cells with 0.5 µg/ml cytochalasin D or with 1.0 µM jasplakinolide for 1 to 5 h. The signaling complex was precipitated from cell protein lysates using an antibody for the cytoplasmic tail of Fas. The presence of the DISC-associated peptide FADD was assessed by Western blot. Treatment with either cytochalasin D or jasplakinolide increased the formation of the DISC, relative to control, as demonstrated by the relative abundance of FADD (Figure 8). Treatment with jasplakinolide also elicited formation of a second band, running slightly faster than FADD, most prominent at 3 h (Figure 8).

View larger version (28K): [in this window] [in a new window]   Figure 8.   Effect of cytochalasin D and jasplakinolide on the formation of the Fas-associated DISC. Confluent 1HAEo cells were treated for 1, 3, or 5 h with 0.5 µg/ml cytochalasin D (CYD), 1.0 µM jasplakinolide (JAS), or control vehicle (DMSO alone). The DISC was precipitated from whole cell lysates using the C20 (APO-1) antibody for the cytoplasmic tail of Fas. Proteins were then resolved by Western blot and probed for abundance of FADD. Equal volumes of immunoprecipitated protein, reflecting results from equal numbers of cells, were loaded in each lane. JAS elicited formation of a DISC as demonstrated by increased abundance of FADD within 3 h. CYD elicited some increased abundance of FADD at 5 h. The presence of FADD in control cells was unchanged; the presence represents the baseline rate of apoptosis seen in this cell line.

Cytochalasin D and Jasplakinolide Cause Apoptosis in Primary Airway Epithelial Cells

To confirm our findings in the 1HAEo cell line, we examined whether disruption of actin filaments could elicit apoptosis in two sets of primary airway epithelial cells. Primary NHBE cells were treated with either agent for 5 h and then examined for filament disruption and TUNEL labeling. Compared with nontreated control cells, cells treated with either 0.5 µg/ml cytochalasin D or 1.0 µM jasplakinolide were small with condensed actin filaments and had small nuclei with chromatin condensation (Figure 9). The number of TUNEL-positive cells in the NHBE cells after 5 h of treatment with 0.5 µg/ml cytochalasin D was 19.3 ± 2.4 versus 1.1 ± 0.6% for control cells (P < 0.0002; n = 4) (Figure 9). The number of TUNEL-positive cells after 5 h of treatment with 1.0 µM jasplakinolide was 15.4 ± 3.5 versus 1.1 ± 0.6% for control cells (P = 0.002; n = 4) (Figure 9). Apoptosis elicited by either agent was blocked almost completely by addition of 100 nM Ac-DEVD-cho (Figure 9), although disruption of actin filament morphology was not changed (data not shown).

View larger version (97K): [in this window] [in a new window]   View larger version (12K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   View larger version (15K): [in this window] [in a new window]   Figure 9.   Effect of cytochalasin D and jasplakinolide on actin filaments morphology and apoptosis in primary airway epithelial cells. (A) Subconfluent NHBE cells were treated with control medium (a), medium containing 0.5 µg/ml cytochalasin D (b), or medium containing 1.0 µM jasplakinolide (c) for 5 h. Cells were fixed and actin filaments were visualized by staining with rhodamine-phalloidin. Bar represents 50 µm. There is disruption of actin filaments after treatment with either cytochalasin D or jasplakinolide. (B) Subconfluent NHBE cells were treated with control medium (a), medium containing 0.5 µg/ ml cytochalasin D (b), or medium containing 1.0 µM jasplakinolide (c) for 5 h. Cells were fixed and stained with Hoechst 33258 to demonstrate nuclear morphology. Bar represents 50 µm. (C) Confluent NHBE cells were fixed 5 h after treatment with 0.1 or 0.5 µg/ ml cytochalasin D, or 0.5 µg/ml cytochalasin D plus 100 nM Ac-DEVD-cho, and stained to demonstrate DNA nicking by fluorescent TUNEL assay. Data are presented as mean ± standard error of the mean. Cytochalasin D elicits apoptosis that is blocked almost completely by Ac-DEVD-cho (n = 4 experiments). (D) Confluent NHBE cells were fixed 5 h after treatment with 0.1 or 1.0 µM jasplakinolide, or 1.0 µM jasplakinolide plus 100 nM Ac-DEVD-cho, and stained to demonstrate DNA nicking by fluorescent TUNEL assay. Data are presented as mean ± standard error of the mean. Jasplakinolide elicits apoptosis that is blocked almost completely by Ac-DEVD-cho. For C and D: *P = 0.001 and §P < 0.0001 versus control (DMSO diluent only) calculated by Fisher's protected least significant difference test after analysis of variance. P = 0.002 and ¶P = 0.0002 versus agonist alone calculated by Fisher's protected least significant difference test after analysis of variance (n = 4 experiments). (E) Confluent SHBE cells were fixed 5 or 24 h after treatment with either 0.5 µg/ml cytochalasin D or 1.0 µM jasplakinolide, and stained to demonstrate DNA nicking by fluorescent TUNEL assay. Data are presented as mean ± standard error of the mean. Both agents elicited apoptosis compared with matched controls treated with diluent (DMSO) only (n = 2 experiments).

In an additional experiment, SHBE cells demonstrated both characteristic morphologic changes and TUNEL staining after treatment with either cytochalasin D or jasplakinolide for 5 or 24 h (Figure 9). Both sets of results were similar to that seen with 1HAEo and NHBE cells.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Cytochalasin D and jasplakinolide disrupt actin filament integrity in cells: the former by preventing the addition of G-actin monomers to the barbed end of the filament (24, 25) and the latter by causing aggregation of filaments (26, 29). Both agents effectively modify the actin cytoskeleton and are useful to demonstrate a potential role for actin in the induction of apoptosis. In this report, we show that apoptosis is initiated rapidly with either cytochalasin D or jasplakinolide in both primary airway epithelial cells and in an airway epithelial cell line.

The activity of cytochalasin D on cell apoptosis has been reported previously. Primary rat embryo fibroblasts, and a cell line derived from these cells, treated with cytochalasin D undergo growth arrest (46). This process required at least 24 h of treatment. Insect ovarian follicle and nurse cells treated with cytochalasin D undergo concurrent apoptosis and cell detachment (47). Treatment of T lymphoma cells with cytochalasin B to disrupt actin filaments augmented the growth-arresting and apoptotic effects of anhydroretinol (48). In contrast to these reports, we find that cytochalasin D-induced apoptosis in airway epithelial cells occurred early ( 5 h), occurred before detachment of cells from underlying matrix, required no other apoptotic stimulus, and required activation of caspase proteases.

There are cell-type differences in the response to cytochalasin D. Treatment with cytochalasin D elicited apoptosis in both primary airway epithelial cells and the 1HAEo cell line. Similar treatment of MDCK cells with concentrations of cytochalasin D did not. Actin filament disruption in MDCK cells is fully reversible once cytochalasin D is withdrawn. In 1HAEo cells, filament disruption and nuclear condensation was reversible only in a subset of cells. MDCK cells are capable of undergoing apoptosis after detachment from extracellular matrix (49, 50) or after exposure to DNA-damaging agents (51). Our data demonstrate that airway epithelial cells and the MDCK cell line have different responses and different survival after actin filament disruption. The reasons for this difference were not elucidated in the present study. Such differences may relate to differences in the regulation of actin filament integrity, to relative differences in resistance to detachment or dissolution of focal adhesion complexes, or to differences in how apoptosis regulating proteins associate with the actin scaffolding.

Apoptosis in the 1HAEo cell line induced by either cytochalasin D or jasplakinolide was not prevented by pretreatment with blocking antibodies to Fas and FasL. These data suggest that changes in cell shape do not elicit apoptosis by inducing the apposition and ligation of FasL to Fas.

Apoptosis induced by cytochalasin D was blocked by inhibition of caspase proteases. Neither caspase inhibitor prevented the disruption of actin filaments, suggesting that the inhibition of apoptosis occurred distal to any signal activated by filament derangement. Furthermore, cytochalasin D-induced apoptosis in the 1HAEo cell line was accompanied by cleavage of pro-caspase-8. It has been reported previously that caspase-8 is cleaved and activated after its recruitment to the DISC that forms on the cytoplasmic portion of CD95 and other death receptors (42, 43). Membrane oligomerization and cleavage also may activate caspase-8 and initiate apoptosis (44), suggesting that the activation of caspase-8 is not restricted to CD95. Treatment with cytochalasin D induced elevated levels of the DISC, relative to controls, in 1HAEo cells. Our data further support the idea that activation of caspase-8 may occur with DISC assembly but independently of activation of CD95.

Jasplakinolide, a cyclodepsipeptide isolated from the marine sponge Jaspis johnstoni, induces F-actin aggregation by binding to the same site as phalloidin (28, 29), and has antiproliferative activity in acute myeloid leukemia and prostate carcinoma cells (30, 52). Jasplakinolide enhanced apoptosis elicited by cytokine withdrawal in the CTLL-20 T lymphocyte cell line and in the Ba/F3 cell line after growth factor withdrawal, but did not itself elicit apoptosis in these cells (32). In that report, the ability of jasplakinolide to enhance apoptosis correlated to its ability to stabilize F-actin. Actin polymerization elicited by jasplakinolide elicits apoptosis directly in HL-60 cells over 4 to 24 h (53). This was associated with a decrease in G-actin and an increase in F-actin, similar to that seen in our experiments, and was associated with membrane blebbing and DNA fragmentation. In our experiments, jasplakinolide elicited apoptosis in both primary airway epithelial cells and in the 1HAEo cell line independent of any additional stressor signal. Like cytochalasin D, treatment with jasplakinolide initiated apoptosis within 5 h and did so via activation of caspases. This occurred without the participation of caspase-8 in the cascade.

Although both cytochalasin D and jasplakinolide elicited apoptosis in airway epithelial cells, the mechanism by which this occurs was different for each agonist. Both agonists elicited apoptosis that could be blocked by the caspase inhibitors z-VAD-fmk and Ac-DEVD-cho in the 1HAEo cell line, and by Ac-DEVD-cho in primary airway epithelial cells. However, treatment with cytochalasin D induced cleavage and activation of pro-caspase-8, whereas treatment with jasplakinolide elicited cleavage but not activation. These data suggest different mechanisms by which actin disruption and aggregation may induce apoptosis.

Anoikis has been demonstrated to occur after disruption of focal adhesion contacts by blocking integrin receptors (54) or by treatment of cells with an antisense oligonucleotide or injected mAb for focal adhesion kinase (55, 56). Cytochalasin D prevents sphingosylphosphorylcholine-mediated tyrosine phosphorylation of this kinase and stress fiber formation in Swiss 3T3 fibroblasts (57). In our experiments using the 1HAEo cell line, cytochalasin D treatment disrupted actin filaments and initiated apoptosis concurrently with decreased abundance of both vinculin and paxillin, whereas jasplakinolide initiated apoptosis without changes in the abundance of either focal adhesion protein. This may suggest differences in signaling to apoptotic pathways. In support of this, one recent report (58) suggests that interrupting the function of focal adhesion kinase can elicit apoptosis, even in anchored cells, that requires both FADD and caspase-8. Whether changes in focal adhesion complex proteins are a prerequisite for cytochalasin-induced apoptosis or whether these changes are parallel but unrelated remains to be determined.

FN, an extracellular matrix protein found in airway basement membranes (59), can provide a survival signal to prevent apoptosis after cell detachment in several cell types (38, 60). This process is mediated by the binding of FN to ₅₁ integrin (61) and may depend on suppression of p53 (57). In our present experiments, treatment of 1HAEo cells with a saturating concentration of a soluble RGDS peptide that binds ₅₁ integrin did not prevent cytochalasin D-induced apoptosis. These data demonstrate that an integrin-mediated survival signal does not overcome apoptosis induced by actin filament disruption.

Actin is a substrate for caspases during the later execution phase of apoptosis (7, 62), although actin may be resistant to cleavage in some cell lines (8). This may be mediated by caspases, either directly (63) or indirectly through calpains (9), and may be preceded by downregulation of actin-encoding genes (64). Although treatment with cytochalasin D shifted actin from cytoskeletal, polymerized F-actin to monomeric G-actin and jasplakinolide shifted actin to a noncytoskeletal compartment, neither elicited actin cleavage at the early time points (data not shown). Our data suggest that change in actin morphology other than cleavage is responsible for initiating apoptosis. The mechanism by which this occurs in airway epithelial cells is unclear. One possibility is that cytochalasin D and jasplakinolide interact with an actin-binding protein; this in turn invokes a signal leading to apoptosis. In support of this theory, the concentrations of both agents required to elicit apoptosis are far lower than the content of actin in the cell. Displacement or inactivation of an actin regulatory protein may signal for apoptosis to proceed. One such regulator, gelsolin, is a substrate of caspase-3 (15) and can itself disrupt actin filaments (65). The disparate effects of cytochalasin D and jasplakinolide on actin filament morphology suggest that an action via a single regulatory protein is unlikely.

A second possibility is that perturbation of the actin cytoskeletal architecture, either by dissolution or by aggregation of actin, disrupts signaling intermediates and regulators that have been localized to its scaffold. Either anti-apoptotic regulators could be disrupted or pro-apoptotic regulators could be approximated, leading to activation. One example of the latter is the assembly of a DISC such as that normally associated with Fas (23). In our experiments, both cytochalasin D and jasplakinolide induced assembly of a DISC. As noted by the presence of FADD in the APO-1 precipitated complex, both cytochalasin D and jasplakinolide in association with actin cytoskeleton derangement induce approximation of at least some members of the DISC. This in turn leads to its activation. Whereas treatment with either agent induced increased abundance of the FADD-containing complex, jasplakinolide treatment led to the accumulation of a second band that also bound the anti-FADD antibody. This band of slightly lower molecular weight could represent a post-translational modification of FADD (e.g., dephosphorylation) or represent another separate component of the complex. This difference may translate into caspase-8 cleavage but not its activation, and suggests an alternate entry point into the caspase cascade.

A final possibility is that the apoptosis-inducing effects of both cytochalasin D and jasplakinolide are the result not of disrupting actin, but rather of changing cell shape and inducing cell shrinkage. Alterations in cell volume are a known trigger to apoptosis in several cell systems (66, 67). However, there are several reasons to suggest that this was not the reason in our present experiments. First, the disruptive effects of both cytochalasin D and jasplakinolide on actin and resulting shape change preceded the development of nuclear fragmentation. Second, the MDCK cell line demonstrated equal disruption of cell shape and morphology after treatment with both agents and within the same time-frame, as seen in the 1HAEo airway epithelial cell line. However, apoptosis in these cells was < 5% at all time points and concentrations of actin-disrupting agents used. Finally, similar cell shape changes elicited by hyperosmolarity, in which actin filament integrity remained intact, did not elicit apoptosis. However, addition of either cytochalasin D or jasplakinolide in cells pretreated with mannitol promptly induced apoptosis to a similar degree as cells treated with either disrupting agent in a normal osmolar environment. These data suggest that cell shape change, in and of itself, is not a sufficient stimulus to initiate cell death in airway epithelial cells.

There are potential physiologic implications to apoptosis elicited by disruption of actin filament architecture in airway epithelial cells. One recent report raises the intriguing possibility that Rho-family guanidine triphosphatases may regulate apoptosis via their control of actin filament integrity (68). Changes in actin filament integrity, whether initiated by external environmental factors or by internal signals, may lead to failed repair of the epithelium after injury, both by preventing migration and by eliciting cell death. This may exacerbate chronic epithelial damage in inflammatory airways diseases such as asthma. How such factors initiate apoptosis in airway epithelium, while not to epithelia of other cell types, is not clear but could relate to differences in actin morphology or in the regulation of actin scaffolding and branching. Such differences need to be elucidated.

The data reported here provide evidence that disruption of actin filaments, either by cytochalasin D or by jasplakinolide, initiates apoptosis in airway epithelial cells. Our data, combined with recent reports on lymphocytes (32), suggest that the actin cytoskeletal network may be both a target of apoptosis and an early signaling component toward apoptotic commitment.