Changes in the cytoskeletal structure of cultured smooth muscle cells induced by calyculin-A

Calyculin-A (CL-A) is a toxin isolated from the marine sponge Discodermia calyx (Kato et al., 1986). Initial studies indicated that CL-A inhibits the embryonic development of the starfish Asterina pectinifera and the sea urchin Hemi centrotus pulccherumis (Kato et al., 1986), and exhibits a strong cytotoxicity against L1210 and P388 leukemia cells, Ehrlich ascites tumor cells, and 3Y1 fibroblast cells (Kato et al., 1986, 1988). Recently, it has also been found that CL-A induces maturation of the oocyte of the starfish Aste rina pectinifera (Tosuji et al., 1991). In other experiments, it was shown that CL-A is a potent inhibitor of types 1 and 2A protein phosphatases (Ishihara et al., 1989; Hartshorne et al., 1989). The IC₅₀ values for inhibition by CL-A for both classes of phosphatases are approximately 1 nM. This is distinct from the other well-known phosphatase inhibitor, okadaic acid, whose IC₅₀ value for type 1 phosphatase is several hundred-fold higher (Ishihara et al., 1989a). CL-A has been widely used to examine the role of phosphatases in various biological systems. In isolated smooth muscle tissues, CL-A increased the extent of myosin phosphorylation via inhibition of phosphatase activity and caused contraction (Hartshorne et al., 1989; Ishihara et al., 1989b). These data are consistent with the hypothesis that phosphorylation mechanism(s) control the contractile activity of smooth muscle (Hartshorne, 1987). Another system in which the effects of CL-A have been examined extensively is 3T3 fibroblasts. In these cells, it was shown that CL-A induced a rapid morphological change in which the cells rounded and detached from the substratum (Chartier et al., 1991). During this process stress fibers disappeared and Factin distribution became irregular, then formed a localized ball-like structure (Hirano et al., 1992). This structure was attached to the nucleus via cables of vimentin filaments. Accompanying the changes in cytoskeletal structure were changes in protein phosphorylation. The three major phosphorylated components were the myosin light chain, vimentin and an unidentified high molecular weight protein (Chartier et al., 1991). To explain the formation of the unusual ball-like structure it was suggested that increased phosphorylation of myosin caused a contractile event in which the actin and myosin were colocalized in an aggregated form. This contractile process, presumably, dragged the attached intermediate filaments with it so that these were concentrated between the ball-like structure and the nucleus (Hirano et al., 1992).

In the present study we have focused on the earlier stages of shape change induced by CL-A in smooth muscle-derived cultured A10 cells. Changes in cytoskeletal structure were observed and these results suggest that alterations in the focal contacts and in the microtubular network precede the rounding-up of cells.

Cultured A10 smooth muscle cells, originating from rat aorta (American Type Culture Collection) were purchased from Dainihon Seiyaku Co. (Osaka, Japan) and maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal calf serum (FCS).

Cells were grown on coverslips in plastic tissue culture dishes (21 cm², Corning, USA). The monolayers were incubated for 30 min with phosphate-buffered saline (PBS; pH 7.4) containing 2 µM fura-2/AM at 37°C. After incubation, cells were excited alternately at 340 nm and 380 nm, and the 500 nm emissions induced by 340 nm excitation (F340) and those induced by 380 nm excitation (F380) were measured with a fluorescence microscope (CAM-200, JASCO, Japan). The high K⁺ solution was made by substituting NaCl with equimolar KCl. The ratio of F340 to F380 was used as an indicator of the cytosolic Ca²⁺ level and the quantitative comparison of the cytosolic Ca²⁺ level was made by taking the resting and high K⁺-stimulated cytosolic Ca²⁺ levels as 0% and 100%, respectively.

Cells were washed three times with PBS and treated with various concentrations of CL-A in PBS at room temperature (22°C). The percentage of cells rounded up were determined at intervals and plotted against the incubation time (min) with CL-A.

The method used to measure the shape changes of platelets quantitatively (Karaki et al., 1989) was also applied to analysis of the shape changes in the A10 cells. Briefly: cells were grown in tissue culture flasks (75 cm², Corning, USA); monolayers were incubated in PBS containing 0.05% trypsin and 0.02% EDTA (with no added CaCl₂ or MgCl₂) at 37°C until the cells began to detach (approximately 1 min). The detached cells (2.3×10⁶ cells/ml) were resuspended in PBS, and the absorbance at 700 nm was measured (UVDEC-460 spectrophotometer, JASCO, Japan).

Cells were washed with PBS and treated with 1×10^-7 M CL-A in PBS. The CL-A-treated cells were immersed in 2.5% glutaraldehyde and 2% paraformaldehyde in PBS for 2 h at 4°C. The fixed cells were washed briefly in the same buffer, post-fixed with buffer containing 1% OsO₄ for 2 h at 4°C and dehydrated through a graded ethanol series. For transmission electron microscopy, cells were embedded in Epon 812. Thin sections were cut using a LKB ultramicrotome (4800), doubly stained with uranyl acetate and lead citrate, and examined in a JEM 1200 EX electron microscope (JEOL, Japan) operated at 80 kV. For scanning electron microscopy, cells were dried with liquid CO₂ at the critical point, coated with gold by sputtering in a vacuum evaporator and observed with an S-430 (Hitachi, Japan).

For the observation of stress fibers and focal contacts, cells grown on coverslips were washed with PBS, and treated with 1×10^-7 M CL-A in PBS. At intervals, the cells were briefly washed with PBS and immediately fixed with 3.7% formaldehyde in PBS followed by permeabilization with cold acetone. Fixed cells were then rehydrated with PBS and incubated for 60 min with a mouse monoclonal antibody to vinculin (ICN ImmunoBiologicals, IL, USA; diluted 50 times with PBS). After washing with PBS, the cells were incubated for 60 min in a mixture of fluorescein-conjugated goat anti-mouse IgG (Tago, Inc., Burlingame, CA, USA; diluted 10 times with PBS) and rhodamine-phalloidin (Molecular Probes, Inc., Eugene, OR, USA; diluted 20 times with PBS). Cells were washed with PBS and mounted in PBS containing 50% (v/v) glycerol and 100 mg/ml 1,4-diazabicyclo[2,2,2]-octane (DABCO; Sigma, St. Louis, MO, USA) to prevent photo-bleaching.

For the observation of microtubules, cells were fixed with cold methanol. After rehydration in PBS, the cells were incubated for 60 min with anti-α-tubulin monoclonal antibody (Amersham Int. plc., Buckinghamshire, UK; diluted 100 times with PBS), followed by incubation with fluorescein-conjugated goat anti-mouse IgG (diluted 10 times with PBS). Cells were then mounted as described above. Localization of F-actin, vinculin and microtubules was observed with a Nikon fluorescence microscope (Nikon, Japan). Images were recorded on Kodak Tri-X film (ASA 400).

CL-A caused rapid and marked changes in the morphology of A10 cells. Within a few minutes after the addition of CL-A (1×10^-7 M), the cells started to become rounded (Fig. 1a-c) and finally detached from the substratum. The rounding-up of the cells was dependent on the concentration of CL-A as shown in Fig. 2a. With higher CL-A concentrations, rounding-up was induced more rapidly. In contrast, okadaic acid, a cytotoxic compound isolated from marine sponges and known as an effective inhibitor of type 2A phosphatase (Ishihara et al., 1989; Hartshorne et al., 1989), did not induce similar morphological changes at concentrations as high as 1×10^-5 M (data not shown).

The surface changes of CL-A-treated A10 cells were observed by scanning electron microscopy. The surface of untreated cells was flat, showing only relatively thin spikelike protuberances (Fig. 1d). Following the addition of CL-A, numerous blebs were produced on the surface of the treated cells (Fig. 1e and f). Bleb formation was observed during the early stages of shape change, while the cells were still attached to the substratum, and was found also with detached cells.

Shape change was also assessed spectrophotometrically (see Materials and Methods) as shown in Fig. 2b. At 1×10^-8 M CL-A there was no detectable change within a few minutes. Shape change was clearly detected at higher concentrations of CL-A and the rate of change was faster at 1 ×10^-6 M compared to 1×10^-7 M. At the latter two concentrations of CL-A the cytosolic Ca²⁺ concentrations were estimated to determine if there was a relationship between shape change and internal Ca²⁺ levels. As shown in Fig. 3a, both concentrations of CL-A induced an increase in cytosolic [Ca²⁺]. A significant increase in [Ca²⁺] was observed after 3.5 to 4 min with 1×10^-7 M CL-A and after about 1.5 min with 1×10^-6 M CL-A. In order to examine if the increase in cytosolic [Ca²⁺] triggers the shape change, effects of ethyleneglycol-bis(β-aminoethyl ether)-N,Ntetraacetic acid (EGTA) and verapamil, a calcium antagonist, were examined. Shape change was monitored in the presence of 4 mM external EGTA (to eliminate Ca²⁺ influx) and was identical to that in the control experiment (Fig. 4). However, the rate of shape change was prolonged (at 1×10^-7 M CL-A) by the inclusion of a nonselective kinase inhibitor, K-252a (Kase et al., 1986; Yasuzawa et al., 1986), at 1 ×10^-5 M. Verapamil (1×10^-5 M) blocked both the K⁺-induced increase in cytosolic [Ca²⁺] and that induced by CL-A (Fig. 3b), but did not prevent shape change (data not shown). Finally, it should be pointed out that shape change was not induced by high [K⁺] (data not shown) although this did increase the cytosolic [Ca²⁺] as shown in Fig. 3a. These results suggest that the observed shape changes reflect alterations in protein phosphorylation that are not directly dependent on Ca²⁺ transients.

Fig. 5 shows the effect of CL-A on the cytoskeletal structures of A10 cells. The distribution of F-actin and focal contacts was examined by fluorescence microscopy. In control cells, stress fibers, stained with rhodamine-phalloidin, were observed along the long axis of the cell (Fig. 5a). Addition of CL-A (1 ×10^-7 M) caused relatively rapid alteration in F-actin distribution. After 2 min exposure to CL-A the intensity of stress fiber staining decreased (Fig. 5c) and after 4 min it appeared irregular and in patches (Fig. 5e).

Vinculin distribution in the same cells was also examined. For the control cells the vinculin staining showed a characteristic localization at the termini of stress fibers at the periphery of the cell (Fig. 5b, arrowheads). After the addition of CL-A (2 min), while the cells remained flat and attached to the substratum, the vinculin distribution was altered. The staining became less intensive and appeared to move along the stress fibers toward the center of the cell (Fig. 5d). In the more rounded cells, vinculin staining was often diffuse (Fig. 5f) and did not show the discrete punctate pattern characteristic of control cells.

This sequence of changes in both morphology and focal contacts of A10 cells, induced by CL-A, was further examined by transmission electron microscopy. In control cells, the electron density of the focal contacts were evident at the points where the cells attached to the substratum (Fig. 6a, shown by arrows). The focal contacts became less obvious after CL-A treatment (Fig. 6b) and finally were not observed when the cells became spherical (Fig. 6c). Instead, numerous blebs were detected on the cell surface opposite to the substratum (Fig. 6c).

The time course of changes in microtubule structure was also investigated. In control cells (Fig. 7a and b) the network of microtubules was well defined throughout the cytoplasm. After treatment with CL-A microtubule staining altered rapidly. The staining pattern became less extensive and many ‘nicks’ or ‘gaps’ were observed along the microtubules (Fig. 7c), suggesting that the microtubules were fractured following exposure to CL-A. Microtubular structure progressively became less defined (Fig. 7d and e) until only local areas of tubulin concentration were evident (Fig. 7f).

Mitotic spindles in dividing cells were also detected both before (Fig. 8a and b) and after (Fig. 8c and d) the CL-A treatment. Mitotic cells seemed to be less sensitive to CL-A. Microtubules in the spindles were clearly detectable after CL-A treatment even when the interphase microtubules had already disappeared (Fig. 8c and d); thus suggesting that the microtubules of the spindles are more resistant to CL-A than the cytoplasmic microtubules observed in interphase cells.

CL-A induces a dramatic change in the morphology of cultured smooth muscle A10 cells, as reported for 3T3 fibroblasts (Chartier et al., 1991; Hirano et al., 1992). We also observed that CL-A causes similar changes in other cell lines, including HeLa cells, and normal and transformed rat embryo fibroblast cells (REF and 4A cells, respectively). Even at 3×10^-9 M CL-A, these cells became rounded and began to detach from the substratum after several hours (data not shown).

Vandre and Wills (1992) observed the rounding up of the pig kidney cell line LLC-PK at 0.5-1.0 × 10^-6 M okadaic acid. However, we found that CL-A is considerably more potent than okadaic acid and this probably reflects the finding that CL-A is a more effective inhibitor of type 1 phosphatases than okadaic acid (Ishihara et al., 1989). It would follow, therefore, that type 1 phosphatases have an important role in regulating the structure of the cytoskeleton and in controlling shape change. A contributory factor may be that CL-A is more permeable to cells than okadaic acid.

In A10 cells, CL-A treatment caused notable changes in the localization of focal contacts preceding the rounding up of the cells; (i) the intensity of vinculin staining decreased, and (ii) vinculin staining moved towards the inside of the cell. Furthermore, electron microscopy revealed that the structure of the focal contacts actually disappeared during CL-A treatment. Vinculin is known to be a substrate protein for pp60src in the focal contacts (Sefton et al., 1981; Kellie, 1988; Kellie et al., 1991). Analysis of cells transformed by sarcoma viruses, however, indicated that tyrosine-specific phosphorylation of vinculin seems to be insufficient to induce the rounded morphology characteristic of transformed cells (Kellie, 1988). Since CL-A is a potent inhibitor of type 1 and 2A phosphatases, which are serine and threonine phosphatases (Ishihara et al., 1989; Hartshorne et al., 1989), phosphorylation of serine and/or threonine residue(s) of the protein(s) in focal contacts following CL-A treatment, i.e. due to a shift in the kinasephosphatase equilibrium, may be responsible for the changes in cell shape. Further investigation is necessary to elucidate whether vinculin is actually phosphorylated following CL-A treatment. It will be particularly interesting if vinculin is phosphorylated on both serine/threonine residues and tyrosine. Since it is known that the placental phosphotyrosine phosphatase is not inhibited by CL-A (Ishihara et al., 1989), the presence of phosphotyrosine would imply regulation of the phosphatase via serine or threonine phosphorylation.

We found that microtubule staining became punctate after CL-A treatment. This may indicate that several breaks were induced along the microtubule filament by CL-A treatment. Vale (1991) found that extracts of Xenopus eggs could rapidly sever stable microtubules along their length. This severing activity was low in interphase and increased during mitosis via a post-translational mechanism. Addition of purified cyclin to an interphase extract increased the microtubule-severing activity as well as histone H1 kinase activity. Involvement of types 1 and 2A phosphatases in the increase in mitotic promoting factor activity has also been seen in starfish oocytes and in mouse oocytes (Picard et al., 1989, 1991; Rime and Ozon, 1990). Therefore, the punctate staining of microtubules observed here may indicate the induction of microtubule-severing activity as a result of phosphorylation.

Spindle microtubules were more resistant to CL-A than interphase microtubules. Since Vale (1991) reported that severing activity found in the mitotic extract was inhibited by brain microtubule-associated proteins (MAPs) bound to the microtubule wall, it is possible that similar protein(s) that were bound to the spindle microtubules protected the polymers against severing. It is also possible that tubulin monomers in spindle microtubules are post-translationally modified, e.g. phosphorylated at a unique site, and become resistant to severing. Recently, Vandre and Wills (1992) reported that okadaic acid at 8-40 nM blocked LLC-PK cells at a metaphase-like mitotic state, and the spindle microtubules were present throughout the period of this mitotic block. It is possible that CL-A treatment of the A10 cells induced a similar effect and stabilized the metaphase spindles. The protein(s) that are phosphorylated as a result of phosphatase inhibition and that regulate microtubule-severing and/or mitotic block remain to be determined.