A study of cytokinesis in the heliozoan Actinophrys sol has been made using low-temperature treatments, antimitotic drugs, cytochalasin B, light microscopy and electron microscopy. It reveals that microtubular axopodia remain extended during cell division and play a major role in it.
Data indicate that when organisms are attached to the bases of culture dishes the normal locomotive mechanism of the presumptive daughter cells pulls them apart. However, when Actinophrys are floating freely in their culture medium, they are still able to undertake division. In this situation interactions between axopodia from opposite daughter cell bodies appear to facilitate the movement apart of the prospective daughter cells.
The present study and other published accounts indicate that a type of cytokinesis exists which is not explicable in terms of the ‘contractile ring’ or ‘fusing vesicle’ theories. To account for these observations a ‘barge pole’ model of division is suggested.
Recently the theory that cleavage furrows can be formed by the contraction of restricted ring- or band-shaped regions of the cortex of some dividing cells has received substantial experimental support. Several workers have demonstrated the presence of microfilaments with diameters varying between 3 and 10 nm which are oriented circumferentially in a narrow region that lies beneath the advancing cleavage furrow (for example Schroeder (1968, 1969), Arnold (1969, 1971), Selman & Perry (1970)). Microfilaments in the same regions of the cortex of glycerinated dividing cells bind heavy meromyosin and so are considered to be actin-like (Perry, John & Thomas, 1971; Schroeder, 1973). However ‘contractile rings’ account only for constriction of cleavage furrows. They apparently do not account for final separation of daughter cells (Mitchison, 1952; Tucker, 1971; Mullins & Biesele, 1973).
‘Contractile ring’ mechanisms of cleavage are also not found in all cells (Rappaport, 1971). In some plants fusion of a number of vesicles in the division plane appears to account for the separation of daughter cells (Pickett-Heaps, 1972). Binary fission in some Rhizopoda might involve an unusual type of cytokinesis. The process in Amoeba proteus was studied by Chalkley (1935). He suggested that amoeboid movement is employed and that daughter organisms divide by moving apart.
Actinophrys sol (Heliozoa, Rhizopoda) is approximately spherical and has many long, extremely thin, radially arranged pseudopodia (axopodia) which contain bundles of longitudinally oriented microtubules (Kitching, 1964; Ockleford & Tucker, 1973). The coordinated lengthening and shortening of axopodia apparently accounts for the usual rolling movement of these organisms (Watters, 1968). Binary fission in Actinophrys has been mentioned briefly by Kitching (1964), who saw no obvious activity of axopodia, and by Watters (1968), who concluded that putative daughter organisms simply roll apart. Thus it appears that Heliozoa employ their normal means of locomotion as division mechanisms.
The present observations and experiments on dividing Actinophrys show that in this organism cytokinesis occurs in situations where the normal rolling movement cannot. Evidence is presented which suggests that in these cases microtubular axopodia which contact each other in the region between the dividing cells are responsible for the transmission of forces which separate daughter organisms.
MATERIALS AND METHODS
Culture technique and light microscopy
The procedure used for culturing Actinophrys and the methods used for light microscopy have already been described (Ockleford & Tucker, 1973). Many organisms in 4-day-old cultures are about to divide. These binucleate Actinophrys are recognizable using a dissecting microscope because of their oblong shape. Unless specifically stated to the contrary, only organisms of this type floating freely in the culture medium were used for observations and experiments. Actinophrys were isolated from cultures individually using finely drawn out glass pipettes. Length measurements were made either from photographs of known magnification or with an eyepiece micrometer. Only length measurements of axopodia which remained oriented precisely perpendicular to the optical axis were used.
Actinophrys from 4-day-old cultures were fixed for electron microscopy and flat embedding was carried out as previously described (Ockleford & Tucker, 1973). Embedded material was searched for organisms in the required state of division using a Zeiss inverted phase-contrast microscope. Dividing organisms were cut out with a fret-saw, reoriented and mounted on to Araldite pegs ready for sectioning. Transverse and longitudinal sections were stained in uranyl acetate for 90 min (Gibbons & Grimstone, 1960) and lead citrate for 5 min (Reynolds, 1963) prior to examination in a Siemens Elmiskop I operated at 60 or 80 kV.
Binucleate organisms in ring preparations were maintained at 4 °C in a constant temperature room. Together with control organisms which had been maintained at room temperature (20 °C) they were examined hourly and the percentage of each group which had divided was determined.
Actinophrys with 2 nuclei, contained in a small Teflon chamber (about 0·5 cm diameter) with a dialysis membrane covering its base, were treated with a 7·5 mg ml−1 solution of colchicine (Calbiochem) dissolved in culture medium. Colchicine was washed out by pipetting away the majority of the solution and replenishing it with fresh culture medium and then by placing the dialysis membrane of the Teflon chamber on to the surface of a reservoir of fresh culture medium which was agitated with a magnetic stirrer. After dialysis for 1 h organisms were isolated individually on to microscope slides using finely drawn-out glass pipettes. The dialysis membrane was almost transparent, so organisms could be seen clearly using transmitted illumination under the dissecting microscope. Actinophrys washed free of external colchicine and untreated Actinophrys in ring preparations were maintained at room temperature (20 °C). The number of organisms which had divided was counted every hour.
Cytochalasin B treatments
Cytochalasin B (Imperial Chemical Industries) was dispersed in a solution of 0·1% dimethyl sulphoxide (DMSO) dissolved in culture medium. Ring preparations were made which contained oblong Actinophrys suspended in culture medium alone or in 0·1% DMSO or 50 μg ml−1 cytochalasin B dispersed in 0 · 1% DMSO. At 50 fig ml−1 solutions of the drug appeared to be saturated, for they contained microscopically visible crystals of cytochalasin B. No higher concentration was applied. The preparations were maintained at room temperature (23 °C) and the number of cells divided was counted every hour.
Inhibition of cytoplasmic streaming in the staminal hairs of Zebrina pendula was used to demonstrate that cytochalasin B solutions were active despite the negative results obtained here with Actinophrys. Inhibition of cytoplasmic streaming was noted when more than 10μ g ml−1 of cytochalasin B was dsisolved in water containing 0·1% DMSO which was used to mount the stamen hairs (Figs. 21, 22).
Binary fission: major structural changes
As cytokinesis proceeds putative daughter Actinophrys gradually move apart (Figs. 9–16). The nuclei lie approximately at the centre of each prospective daughter cell during the period when the bridge of cytoplasm which connects the two forming daughter cells lengthens and attenuates (Fig. 19). Frequently food vacuoles or other vacuolcs occur in bridge regions (Figs. 11–13). Such vacuoles often have greater diameters than the parts of the bridge to either side of them and thus appear to occlude the bridge between the 2 cells. Finally the connecting bridge breaks. Bridges do not remain rigidly extended. The 2 broken ends move out of the plane of focus when the bridge parts separate and are gradually resorbed into their respective cell bodies. The length and diameter of the bridge between daughter cells at the time of final separation varies from individual to individual. It is often as much as 100 μm in length and as little as 2 μm in diameter. Daughter organisms continue to move apart after breakage of the bridge between the cells (Table 1).
Rate of division
The rate at which dividing cells move apart can be calculated from measurements made on photographs taken at intervals. Table 1 indicates that the rate at which daughter organisms move apart varies between 5 and 22 μm min.−1. During the 5-min period before bridge-breaking some cells move faster, and other cells move more slowly than in the 5 min after bridge-breaking.
As dividing organisms move apart contacts are occasionally seen between 2 axopodia based on different cell bodies (Figs. 17, 19). These are seen most frequently in the region between the dividing organisms. Axopodial contacts are of 3 types: terminal for both axopodia (double terminal); terminal for just one axopodium (single terminal); not terminal for either (double non-terminal). These contacts are V, T and X-shaped, respectively. Frequently a webbed region is visible at the point of contact between axopodia (Fig. 17). Sometimes one or both of the contacting axopodia are bent (Fig. 17). The degree of curvature and the direction of the bend sometimes vary as division proceeds. Axopodia involved in double non-terminal contacts appear to move with respect to each other in a slow, incomplete ‘knitting motion’. Kinetocysts are small organelles which lie between the microtubular axoneme and the axopodial membrane. They move up and down axopodia and are thought to be involved in food capture (Bardele, 1970). Contact points between axopodia do not block the passage of kinetocysts. Kinetocysts sometimes move into webbed regions between contacting axopodia. They then move out again into the main axoneme-containing portion of the axopodium. However, kinetocysts do not cross over from one axopodium to another via a contact point. This may indicate that the 2 contacting axopodia are physically separated in some way, for instance by a membrane.
Formation of new axopodia
Small axopodia occur frequently in the area to either side of the cytoplasmic bridge which connects dividing organisms (Fig. 18). Their lengths change continually. They alternate between periods of growth and resorption. Table 2 provides examples of the types of axopodial length changes which have been observed.
The maximum average elongation rate recorded for these axopodia is 4·5 μm min−1; the maximum rate of shortening observed has about the same value. Longer-growing axopodia have greater diameters at their bases than shorter-growing axopodia. Although axopodia near to the bridge of cytoplasm connecting the 2 cells are usually shorter than those further from the bridge, a precisely graded arrangement of axopodia in decreasing order of length is not apparent (Fig. 18). The appearance of these axopodia which grow and are resorbed during this part of the life cycle is restricted to a localized area of the surface of the organism. Examination of the surface of the 2 cell bodies further from the connecting bridge of cytoplasm over protracted periods (2 h) fails to reveal any small growing axopodia at a time when growing axopodia are continually present near the connecting bridge of cytoplasm.
Rolling movements and division
Examination of living oblong Actinophrys., which lie on the bottom of Petri dishes, shows that as division proceeds the connecting bridge of cytoplasm between the daughter organisms rises slightly. This is consistent with the suggestion (Watters, 1968) that daughter organisms divide by rolling apart. Using finely pointed needles to make scratches on the bases of Petri dishes adjacent to organisms at early division stages, an assessment can be made of the relative movement of daughter organisms as the separation between them increases. These studies show that sometimes one daughter moves and at other times both move away from each other.
Single free-floating oblong Actinophrys which are tumbled and spun in the culture medium with a fine glass needle can be observed continuously with a dissecting microscope. Binucleate organisms which are never allowed to settle on to the base or side of the Petri dish will usually complete an apparently normal division within 30 min of the start of manipulation. This demonstrates that these organisms are able to undertake cytokinesis when they are not attached to a solid substrate.
The connecting bridge of cytoplasm
Transverse and median longitudinal sections of dumbbell-shaped dividing organisms reveal no concentrations of microtubules or microfilaments. It is pertinent that in heliozoans preserved with the fixative used here microtubules are adequately demonstrated both within the cell bodies and in the axopodia which extend out from them (Roth, Pihlaja & Shigenaka, 1970; Ockleford & Tucker, 1973). Cytoplasm in the bridges is highly vacuolated. The vacuoles are often arranged in such a way that no path of cytoplasm is available in which a straight rigid skeletal structure could run from one cell body to the other (Figs. 11–13).
Influence of low temperature, colchicine and cytochalasin B on division
Half of a group of binucleate organisms maintained at room temperature (20 °C) divides within 6 h. In the same length of time none of a similar group maintained at 4 °C divides. This total inhibition of cytokinesis by low-temperature treatment lasts for at least up to 24 h (Fig. 2). Organisms returned to room temperature after up to 24 h of low-temperature treatment proceed to divide in apparently the same way as organisms maintained at room temperature. The percentage of cells divided after each hour in both control (20 °C) and experimental (4 °C followed by 20 °C) groups of organisms is recorded in Figs. 1 and 2.
The minimum concentration of colchicine which causes visible shortening of microtubular axopodia in Actinophrys is 2·5 mg ml−1 of culture medium. The application of 7·5 mg ml−1of colchicine to dividing binucleate Actinophrys causes withdrawal of axopodia, regression of the dumb-bell outline to approximately that of an oblate spheroid and a general flattening of the organism. Fig. 4 shows that the colchicine treatments (7 · 5 mg ml−1) totally inhibit division of binucleate Actinophrys under conditions where 50% of a similar group of organisms, not treated with colchicine (Fig. 3), divide within 2 h. After washing, dialysis and resuspension in fresh culture medium, binucleate colchicine-treated organisms proceed to divide at the rate shown in Fig. 4. The lower rate of division noted after recovery from colchicine treatment compared with that observed after low-temperature treatment is probably not caused by inadequate washing. Comparison of the absorption of ultraviolet light by the culture medium in which the organisms are resuspended with that of a series of standard dilutions of colchicine shows that it is present at a concentration of less than 100 /μg ml−1. Figs. 3 and 4 compare the percentages of organisms which have divided every hour in a group of binucleate cells recovering from colchicine treatment with the percentages in a group of similar cells maintained in culture medium alone. When cells divide after recovery from colchicine treatment they bear apparently normal axopodia.
Division of binucleate Actinophrys is not inhibited in the presence of cytochalasin B (50 μg ml−1) dispersed in a solution of 0·1% DMSO dissolved in culture medium. Likewise, a solution of 0·1% DMSO alone in culture medium has no inhibitory effect on division. The percentages of organisms which have divided after each hour in groups of organisms maintained in their normal culture medium alone, treated with DMSO, or cytochalasin B and DMSO is shown in Figs. 5, 6 and 7, respectively. Movement of kinetocysts in axopodia, the rolling motion of whole organisms and the operation of contractile vacuoles in Actinophrys are not inhibited even after 24 h in the presence of 50 μg ml−1 cytochalasin B. Differences between rates of division in control organisms (Figs. 1, 3 and 5) are a result of differences in physiological condition between cultures. These are inevitable when non-axenic culture methods are used. Such differences are of little importance with regard to this study because in each case control and experimental populations were derived from the same cultures. Under the conditions of the experiments inhibition with both cold and colchicine treatments was total. This effect was reproducible.
One item of information which is relevant to the observations above is based on observations of cells which are aggregated into groups. This colonial behaviour is thought to allow heliozoans to capture prey organisms that are too powerful to be caught by a single individual (Jepps, 1956). When a group of these organisms has digested its common meal, it splits up. The process by which individuals separate from groups is similar to certain events occurring during binary fission. Cytoplasmic bridges are formed which are similar in shape to the bridges between daughter cells undertaking binary fission. The cytoplasmic bridges eventually snap, separating individual organisms from clumps. There are bent axopodia between separating cells and clumps (Fig. 20). These occur at similar places and appear similar to the bent axopodia between Actinophrys undertaking binary fission.
Watters’ (1968) suggestion that when Actinophrys divides the 2 daughter organisms move apart by rolling over the substrate accounts for many of the features observed during fission. It cannot account for the simple demonstration that freely floating Actinophrys arc capable of dividing. This observation indicates that floating organ isms must push against each other when they are dividing, for there is no fixed object sufficiently close to the cells against which a pushing or pulling force could be exerted. Hence if heliozoan locomotion does provide the force for division it can only be in those instances where the organism is attached to some fixed object.
Cold and colchicine destroy axonemal microtubules in Heliozoa (Tilney & Porter, 1967; Ockleford & Tucker, 1973) and completely inhibit the separation of daughter Actinophrys. It is likely therefore that division depends upon the presence of micro-tubules. Microtubules are strong elastic structures (Ockleford & Tucker, 1973), so their essential role probably lies in the transmission of some force required to separate the daughter organisms. In free-floating dividers this must be a ‘pushing force’. To function effectively, such a force must operate in the region between the 2 dividing cells. The requirement then is probably for a microtubular force-transmitting component which lies in the region between 2 dividing daughter Actinophrys.
The finding that vacuoles occlude the cytoplasmic bridges rules out the possibility that microtubules extend the length of the bridge of cytoplasm between the cells. The fine-structural evidence suggests the same conclusion. The only microtubule containing components which span the space between the cell bodies of floating daughter organisms are the axonemes of those axopodia based on opposite cell bodies that contact each other. Ockleford & Tucker (1973) have presented evidence that reveals that axopodia elastically resist bending; a bent axopodium is therefore normally an indication of a deforming force. Some of the axopodia which do make contact between dividing cells are bent. They are probably bent because they are transmitting the force that is separating the daughter organisms.
It appears reasonable that in all cases the force separating daughter Actinophrys is transmitted by axopodia. In the case of dividing organisms attached to the bases of Petri dishes there is probably a pulling force exerted against the base of the dish. According to Watters (1968), leading axopodia shorten when their tips are firmly attached to the substrate and thus cause putative daughter organisms to ‘pull’ apart. In contrast to this suggestion it has been argued above that the axopodia of free floating dividers must exert a pushing force against other axopodia. These division processes may not be distinctly separate. They could represent extremes of a continuous range in which organisms undertake intermediate forms of division when some axopodia pull on the substrate and some axopodia push against the substrate and/or axopodia on the other daughter Actinophrys. Such an arrangement could explain the variation in rates of division shown in Table 1. The chances of an axopodium from a floating organism making contact with another axopodium would seem to be far lower than the chances of an axopodium from a substrate-attached organism making contact with the bottom of the Petri dish. The free-floating division could be thought of as a more ‘difficult’ type of division which the organisms take longer to accomplish.
How are the contacts made between pushing axopodia on free-floating dividers? Newly forming axopodia are situated in the area where such contacts occur. The continual growth and resorption of these axopodia may represent a trial and error system for forming contacts. When a correct contact is made by an axopodium it might ‘trigger’ continued growth where otherwise the axopodium would be resorbed.
The rates of axopodial outgrowth (about 4 μm min−1) in this naturally occurring situation are much slower than average rates (10μm min−1) reported for early out-growth of retracted axopodia (Ockleford & Tucker, 1973). This implies that in these naturally growing axopodia there is some extra limitation on the speed of outgrowth. The limitation may be for example the availability of some substance required for the outgrowth of the axopodium.
The fact that cytochalasin B does not prevent the division of Actinophrys is not compelling proof of the absence of a contractile ring. Although this drug is thought to have an inhibitory effect on microfilamentous systems (Wessells et al. 1971), there is no certainty that it actually enters the organism. Many protozoa have low permeability to some drugs; for instance, the lowest concentration of colchicine which induces shortening of axopodia in Actinophrys is many times greater than the highest concentrations of colchicine normally used to arrest mitosis in tissue-cultured cells (Priest, 1969). Also, the apparent absence of microfilaments in the connecting bridge of cytoplasm could be the result either of their genuine absence in vivo or the failure of the fixation technique used to preserve them in a recognizable form.
However, because colchicine is a drug which causes microtubule breakdown but which is thought to have no effect on microfilamentous systems (Wessells et al. 1971), the fact that it causes reversion from a dumbbell- to an oblong-shaped cell is inconsistent with a microfilamentous contractile ring division mechanism. Further, the changing shape of Actinophrys during division is different from the changing shape of cells where contractile rings are thought to be operative (for example, see Szollosi, 1970). It seems unlikely, therefore, that a contractile ring plays a major role in cytokinesis in Actinophrys. It might be argued that colchicine exerts its anti-cytokinetic effect on Actinophrys as a result of an interaction with membranes (Stadler & Franke, 1972) and a resulting change in membrane properties. However, the blocking action described here does not merely arrest cleavage. It causes a change in cell shape which includes regression of the cleavage furrow. There is apparently no evidence that colchicine causes regression of furrows in other situations. As a general rule major changes of cell shape as a result of colchicine action can be satisfactorily explained on the basis of the drug’s destructive effect on cytoplasmic microtubules (for example, see Tilney, 1969).
Axopodia are associated with at least 3 types of movement other than the movement apart of dividing daughter organisms. These are, the movement of kinetocysts within axopodia, the rolling movement of the whole organism, and the movement of material over the external surface of axopodia. Kitching (1962) first drew attention to this latter type of movement when he described the lifting of a ‘skin’ from the surface of the cell body of Actinophrys and its subsequent centrifugal transport out along axopodia. The origin of this skin is probably the result of simultaneous discharge of many kinetocysts, because after it has lifted, the axopodia appear smooth and needle-like. A similar type of movement occurs when material egested from the cell body is transported ccntripetally along axopodia at rates up to 60 μm min−1.
The evidence which implicates axopodia in the causation of these 3 other movements makes it seem likely that the motive force for division is developed there also. The present experiments and observations do not strongly suggest how this force is produced. A complete discussion of possible mechanisms would necessitate a review of the several theoretical models which have been proposed; most of these deal with the mechanics of the spindle. Models of 2 main types can be considered. One type suggests that force production occurs by loss or addition of oriented molecules (for example Inoue & Sato, 1967) and the other attempts to explain movement in terms of a sliding or shearing mechanism (for example Mclntosh, Hepler & Van Wie, 1969). Although growth of new axopodia in the furrow region suggests a mechanism of the former type, neither of these types of theories can be invalidated with present evidence.
There are recent reports of actin-like protein localized in the spindles of locust spermatocytes (Gawadi, 1971) and crane fly spermatocytes (Forer & Behnke, 1972). As Huxley (1973) has shown, there are unexpected similarities at the ultrastructural and biochemical levels between many apparently unrelated types of force-producing mechanisms. All of the systems he has reviewed appear to contain oriented actin filaments. It is possible that actin is present in heliozoan axopodia, but that it is as yet undetected.
The theory of cytokinesis suggested here for Actinophrys sol is similar to the model proposed by Chalkley (1935) to account for his observations of the binary fission of Amoeba proteus. The most remarkable agreement is in the case of the modified free-floating division which also occurs in that organism (Chalkley, 1935; Liesche, 1938). In Actinophrys sol this seems to involve axopodia pushing on axopodia, whereas in Amoeba proteus pseudopodia apparently push against each other. It is of interest in this respect that outgrowing pseudopodia in Amoeba proteus which fail to make contact with pseudopodia on the other cell body become extremely long and thin, imparting a stellate heliozoan-like appearance to the organism. In the closely related Amoeba sphaeronucltolosus the process is generally similar but modified in that the pseudopodia which apparently interact to cause division are unusually short and blunt. Wittmann (1951) describes these pseudopodia as Presswiilsten (pressure-pads). These mechanisms are illustrated in Fig. 8. Although Netzel & Heunert (1971) do not comment on the mechanism of division, it is possible from their micrographs that cytokinesis in the testacean Arcella vulgaris is accomplished by the exertion of forces such that the edges of the new and old shell are pressed against each other or the substrate. Thus it appears that this mechanism of cytokinesis is relatively common at least among one class of Protozoa (Rhizopoda).
In view of this finding it would be of interest to examine other cell types in order to determine whether this division mechanism occurs in other phyla and to define exactly how widespread it is. Possible indicators in the search for cells or organisms which might undertake cytokinesis in a similar way to that reported here are the occurrence of a temporally discrete nuclear and cytoplasmic division and a tendency not to round up totally and become immobile during fission.
I am indebted to Dr f. B. Tucker for his encouragement and guidance and to Professor K. G. Grell for critically reading the manuscript. This work was undertaken during tenure of a Research Studentship awarded by the Science Research Council of Great Britain and was also supported by a Royal Society European Programme Fellowship.