Limpet haemocytes: III. effects of cytochalasin b and colchicine on cell spreading and aggregation

Colchicine and cytochalasin B have been used widely in attempts to interfere in a fairly specific way with a number of processes involving motility of cells (Spooner, Yamada & Wessells, 1971). Among other effects, cytochalasin B (CCB) has been shown to inhibit the aggregation of blood platelets (White, 1971; Haslam, 1972; Kay & Fudenberg, 1973), to inhibit the spreading (Weiss, 1972) and locomotion (Wessells, Spooner & Luduena, 1973) of cells on a solid substrate and to alter cell shape, particularly by preventing the formation and maintenance of microspikes (Wessells et al. 1971). Colchicine too affects the shape, spreading and solid substrate locomotion of some cells (Vasiliev et al. 1970; Spooner et al. 1971). In view of the suggested associations between cell shape, especially the presence of microspikes, cell adhesion (Pethica, 1961 ; Weiss, 1972), and locomotion (Taylor & Robbins, 1963 ; Partridge & Davies, 1974), it would be interesting to correlate the effects of CCB and colchicine on each of these cellular activities. The systems previously used to study the effects of these agents on such aspects of cell behaviour are so disparate that comparisons between them are not very useful. For this reason we have investigated the effects of CCB and colchicine on the development and maintenance of microspikes, on aggregation and on the cell-substrate relationships of limpet blood cells. This is a particularly convenient system, for the haemocytes display rapid aggregation and substrate spreading. The great majority of the cells, the amoebocytes, develop prominent spikes which appear to be large microspikes (Davies & Partridge, 1972) and which participate in aggregation, spreading and locomotion (Partridge & Davies, 1974). Conversely, a small proportion of cells, macrophages, spread on a glass substrate without forming spikes. Thus we have 2 different models of cell spreading in this system.

Blood was withdrawn from the palliai veins of limpets (Patella vtdgata) as described previously (Davies & Partridge, 1972) and preparations plated on to glass slides in humid chambers (Partridge & Davies, 1974) were used to study the spreading and locomotion in various media.

The media used were constituted as follows:

Artificial seawater: NaCl 23·44 g, KC1 0·725 g, Na₂SO₄ 3·90 g, CaCl₂.6H₂O 2·20 g, MgC1₂.6H₂O 10·64 g.

EDTA-seawater: NaCl 17·20 g, KC1 0·725 g, Na₂SO₄ 3·90 g, EDTA 37·20 g.

CMF-seawater: NaCl 28·90 g, KC1 0·725 g, Na₂SO₄ 3·90 g.

Each of the above solutions was made up to 1 dm³ with glass-distilled water and the pH adjusted to 8·20 with 0·5 mol dm⁻³ NaOH.

Cytochalasin B was obtained from ICI Ltd., and made up in a stock solution in DMSO at a concentration of 0·5 mg/cm³. A Hamilton microlitre syringe 702-N was used to dispense aliquots from the stock solution as required. The DMSO used in the making of the stock solution was obtained from the same source as the DMSO used in the controls (Koch-Light Laboratories Ltd.). Colchicine was obtained from BDH and made up to a concentration of 50 μg/cm³ in artificial seawater.

The formation of spikes, which normally occurs during the first 3 min after withdrawal of amoebocytes from the limpet (Fig. 2) (Davies & Partridge, 1972), is completely inhibited by addition to the withdrawn blood of an equal quantity of seawater containing CCB (Fig. 3). The cells become lobulated into a number of smooth-edged blebs and lamellae, which results in a somewhat mulberry-like shape. This effect is equally evident at all CCB concentrations tested (0·0625–1·0 μg/cm³). Similarly, fully formed spikes on suspended cells collapse rapidly when exposed to CCB. Addition of seawater containing DMSO at a concentration of 1 mm³/cm³ is without effect on the form of suspended cells.

Amoebocytes suspended in serum or seawater containing CCB are greatly inhibited with regard to their ability to spread on to a glass surface. Individual amoebocytes in serum, artificial seawater or a mixture of the two, containing 1 mm³/cm³ DMSO spread completely in 15-30 min. In contrast, suspensions in media containing 0·0625–0·5 μg/cm³ CCB have barely begun to spread within the first hour and require 2–3 h to spread to their full extent. In addition, the spreading in the presence of CCB (Fig. 4) is unlike that seen in untreated cells (Fig. 5) in that the extending lamellae contain no obvious spikes. Within the range of concentrations used, there is little or no dose effect of CCB on spreading.

Cells which have been allowed to spread in serum show no obvious changes when the medium is replaced by artificial seawater containing 1 mm³/cm³ DMSO for periods of several hours and such cells move normally on the glass substrate. Within 1 min of addition of seawater containing CCB to a preparation of spread amoebocytes, effects on the spikes and cell margins are observed. These changes are usually so rapid that they are difficult to photograph, but in some cases, where the coverslip is closely applied to the slide, the penetration of replacement media into the preparation occurs slowly and the sequence of effects on the spread cells is retarded. In such cases it is possible to obtain a better series of micrographs (Figs. 8–12) which illustrate the nature of the changes but do not reflect the speed with which these changes can occur. Where they project beyond the margin of the cell, the spikes become beaded or irregular in shape and reduced in phase density. Simultaneously, the phase-dense proximal regions of the spike which lie within the lamellar portions of the cells become indistinct, break up into irregularly shaped structures, or become invisible (Fig. 9). Subsequently the extended margins of the cells retract. Sometimes this retraction is quite pronounced and the cells contract into tight clumps, leaving long tenuous retraction fibrils trailing back to attachment points of the original spike tips (Fig. 10, cell a). In other cases the retraction is less marked, and consists of the withdrawal of the lamellar web between the spikes. The spikes themselves remain as retraction fibrils which are rather flaccid as judged by their movement in currents in the medium. In less well spread cells the lamellar web merges with the bases of the spikes to give a polygonal outline to the cell. No net translocation of amoebocytes on the substrate occurs in CCB-containing media, apart from those movements interpreted as retraction.

Replacement of media containing CCB with fresh seawater containing 1 mm³/cm³ DMSO results in a rapid reversal of the effects in either suspended or spread amoebocytes. Within a few minutes new spikes begin to form (Fig. 11) and normal spreading occurs within 30 min (Fig. 12).

Suspension of haemocytes in seawater containing 50 μg/cm³ colchicine causes no obvious change in their morphology within the few minutes that they remain in suspension. Spreading of such cells also proceeds quite normally. Similarly, in a preparation of fully spread cells, replacement of the serum or seawater by seawater containing 50 μg/cm³ colchicine has no observable effect in the first 30 min but by 60 min many cells have lost their bipolar form (Fig. 6). By 4 h they have lost virtually all trace of polarity and are very extensively spread between spike lamella structures extending from the entire cell margin, assuming a stellate appearance (Fig. 7).

Slide preparations made from sparse suspensions of haemocytes in (a) artificial seawater, (ft) 0·5 μg/cm³ CCB in artificial seawater, or (c) 50 μg/cm³ colchicine in artificial seawater, were kept at room temperature for 4 h then fixed in 4% glutaraldehyde in seawater and stained. The average spread area per cell was estimated on 100 isolated amoebocytes in each experimental group, by means of a random spot Chalkley array as described by Curtis (1960). The results are shown in Table 1. It can be seen that CCB greatly reduces the extent to which cells spread on to the substrate compared with that shown by cells in seawater alone. In contrast, colchicine considerably enhances the amount of spreading exhibited by the cells, perhaps reflecting the larger proportion of the cell margin occupied by spike-supported lamella, which appears to be the organelle concerned with spreading.

Macrophages are observed to spread normally and to remain spread in the presence of up to 1·0 μg/cm³ CCB (the highest concentration used). Likewise 50 μg/cm³ colchicine produces no noticeable effect on these cells.

Considerable variation was found between individual aggregation experiments but the main effects were quite consistent between experiments and the following examples are representative of the results obtained.

Haemocytes from a single limpet were divided into 4 aliquots, pelleted by centrifugation at 300 g and suspended in the 4 aggregation media shown in Table 2. In comparison with the cells in artificial seawater, inhibition of the aggregation of haemocytes is evident even at 0·125μg/cm³ CCB. More or less complete inhibition of aggregation was found at 0·5 μg/cm³ in all experiments, but the 2 lower concentrations of CCB sometimes caused more complete inhibition than is seen in this example. We suspect that the aggregation which occurs in 0·25and 0·125Ag/cm³ CCB represents collisions occurring between cells before they have been affected by these low concentrations of the reagent, but a more discriminating measure of aggregation is required to confirm this.

To test whether the effect of CCB on aggregation is reversible, the experimental schedule shown in Fig. 1 was followed. The results of the total particle counts and the percentage aggregation between counts 1–2 and 3–4 are given in Table 3. They show that haemocytes which fail to aggregate significantly when shaken for 5 min in 0·5 μg/cm³ CCB in seawater will aggregate when subsequently shaken in seawater containing 1 mm³/cm³ DMSO. The reduction in particle number during the centrifugation and resuspension between counts 2 and 3 may be due to adhesion of cells to the centrifuge tube or to one another during centrifugation.

To test the effect of CCB on established cell contacts, aggregates previously formed by shaking a suspension of haemocytes in seawater for 5 min, were suspended in 0·5 μg/cm³ CCB in seawater and agitated again on the shaker. Counts made on samples taken at zero time and after 5 and 10 min shaking did not reveal any significant change in total particle number, indicating that no disaggregation caused by breakdown of established contacts between cells had occurred.

The effects of colchicine on spread cells were evident only after about 1 h. Therefore haemocytes were maintained in suspension in CMF-seawater or in 50 μg/cm³ colchicine in CMF-seawater for 1 h. They were then pelleted by centrifugation and resuspended in seawater and in 50 μg/cm³ colchicine in seawater respectively, for aggregation by shaking in the normal way. The results of a sample experiment given in Table 4 show that colchicine has no significant effect on the aggregation of haemocytes.

In the moderately short-term experiments conducted, colchicine was found to have a limited effect on the spreading of amoebocytes, consisting of a gradual loss of polarity over the first few hours. This is in accord with the observations of Vasiliev et al. (1970) on fibroblasts, which gradually lost oriented movement and morphological polarity when cultured in colchicine. It appears that the microtubular system in amoebocytes is probably concerned with regulation of the orientation of the spikes rather than with their structure or their function in spreading and aggregation.

CCB at very low concentrations was found to have rapid and profound effects on the formation and maintenance of spikes, and on the aggregation, spreading and solid substrate locomotion of the amoebocytes. The action on all of these functions was very rapid and almost completely reversible on removal of CCB, observations which suggest that major disruption of cellular mechanisms is not involved, and that the phenomena resulting from CCB treatment of amoebocytes may be the result of interference with a single system.

Several observations point to the spikes as the CCB-sensitive organelles whose disruption is responsible for most if not all of the changes found. The earliest observed effect of CCB on spread cells is on the optical integrity of these structures. Although the change in form of the spikes where they project beyond the margin of the lamella could be attributed to the collapse of deformable structures following a reduction of cell-substrate adhesiveness, the simultaneous disruption of the portion of the spike lying within the lamella is not amenable to such an explanation and seems to indicate that the structure of the spike itself is affected. Furthermore, the retraction of cells which follows the disruption of the spikes provides no evidence of loss of adhesiveness, for the tips of the spikes retain their attachment to the substrate and the adhesions between the cells remain intact. Rather, the retraction appears to result from a loss of rigidity of the lamellipodial portions of the cells due perhaps to the loss of supportive spikes (Partridge & Davies, 1974). The fact that the spike-supported lamellae of amoebocytes in suspension collapse in the presence of CCB is also in agreement with this view.

Electron-microscopic observations have revealed that microspikes are based on a structure of microfilaments (Goldman & Follett, 1969; Wessells et al. 1971; Allison, 1973) and initial investigations suggest that the spikes of limpet amoebocytes possess a similar structure. The suggestion that the activity of CCB is attributable to an effect on spike structure is, therefore, in accordance with the proposal of Wessells et al. (1971) that CCB disrupts the 5-nm microfilament lattice system. This hypothesis is disputed (Carter, 1972; Goldman, 1972) and there is strong evidence that CCB acts on cell surface components involved in the movement of molecules across membranes (Mizel & Wilson, 1972; Plagemann & Estensen, 1972). However, these 2 hypotheses are not incompatible: a hydrophobic molecule like CCB (Rothweiler & Tamm, 1966) which may disturb the structure of lipid membranes could interfere with the insertion of microfilaments into them, and thus disorganize microfilament-based structures and functions.

Spreading of amoebocytes on to a glass substrate is greatly inhibited by CCB with regard both to rate and final extent of spreading. Similarly, the locomotion of these cells, which seems to depend on the same basic mechanism as spreading (Partridge & Davies, 1974), was completely inhibited by CCB. This is probably a reflexion of the important role of spikes in the normal process of advance of the leading edge of amoebocytes over a solid substrate. However, the fact that amoebocytes do spread in the presence of CCB, albeit slowly and to a reduced extent, suggests that some other mechanism plays a part in the process. This may be a surface tension effect as proposed by Carter (1967) or a physicochemical interaction based on the DLVO theory for lyophobic colloids, a theory of cell adhesion favoured by Curtis (1973) and Gingell (1971). It is of interest in this respect that the macrophages spread normally in the presence of CCB ; these cells do not display spikes and it is possible that they spread by means of the mechanism which is responsible for the partial spreading of amoebocytes in CCB.

It appears, from the aggregation experiments, that CCB interferes with the mechanism whereby single cells develop initial adhesions rather than with the stability of existing adhesive contacts between cell surfaces. Aggregates, in which cells have formed stable adhesive contacts, are not dispersed by concentrations of CCB and conditions of shear which completely inhibit the aggregation of single cells. The inhibition could arise from the suggested inability of cells to form close contacts except by means of low-diameter probes (Pethica, 1961; Weiss, 1964), that is, by means of microspikes. On the other hand, it seems possible that cell surfaces do approach one another closely enough to form adhesions, because amoebocytes are able to adhere to a glass surface in the presence of CCB. In this case, the inhibition of aggregation by CCB could be explained by the low rate of spreading of amoebocytes in the absence of spike-supported lamellae, for the initial contact would have to be rapidly stabilized, for example, by an increase in contacting area, to resist the high shear forces in shaken suspensions (Davies & Partridge, 1972). These alternative hypotheses of the mechanism of inhibition of aggregation by CCB are under investigation (Partridge & Jones, in preparation). There is some circumstantial evidence in favour of the view that the inhibition of aggregation induced by CCB operates by its effect on spikes and the rate of stabilization of contact. Thus platelet aggregation, which is a fast aggregating system involving the production of spikes, is inhibited by CCB (White, 1971; Haslam, 1972; Kay & Fudenberg, 1973), whereas the comparatively slow aggregation of dissociated chick tissue cells which do not form prominent spikes, is not inhibited by CCB (Armstrong & Parenti, 1972; Steinberg & Wiseman, 1972).

It appears that the effects of CCB on the spreading and locomotion of amoebocytes on a glass surface and on the aggregation of cells in shaken suspensions can be attributed to the action of CCB on the integrity of the spikes, which seem to play important roles in these aspects of cell behaviour. Thus, the amoebocyte may represent one extreme of spike-dependent cell behaviour while the macrophage falls at the other end of the range.

On this basis an investigation is currently in progress on the ultrastructure of the 2 cell types, with regard especially to the role played by the spikes in the adhesion and locomotion of the amoebocyte. The effects of colchicine and CCB on these processes should provide some indication of the mechanisms by which the spikes operate in these aspects of cell behaviour.