Colchicine induces multiple axis formation and stalk cell differentiation in Dictyostelium discoideum

During the development of Dictyostelium discoideum, the aggregation of individual amoebae leads to construction of a multicellular pseudoplasmodium. This cell mass ultimately develops into a fruiting body containing a lemonshaped mass of spores, a slender supporting stalk and a basal disc. The pseudoplasmodium contains a polarized axial pattern: future stalk cells occupy its anterior portion, prespore cells are in the middle and a small population of pre-basal disc cells completes the rear end. This pattern is at least roughly proportionate (Takeuchi, Hayashi & Tasaka, 1977) and regulative (Sakai, 1973). At the front of the slug is a nipple-shaped tip, which has been shown to act as an organizer (Raper, 1940; Durston, 1976). The slug tip is known to contain a high concentration of cyclic AMP (Rubin & Robertson, 1976) and this cyclic nucleotide has been shown to induce Dictyostelium discoideum amoebae to differentiate into stalk cells (Bonner, 1970).

Essentially, there are two schools of thought about the establishment of polarity and pattern in the Dictyostelium discoideum pseudoplasmodium, and in similar systems. The first supposes that the pseudoplasmodium contains an axially graded variable, or variables, which provide positional information to cells and thus define their future differentiated state (see Wolpert, 1969; Mc-Mahon, 1973). In most such models the tip would feature as a reference point, which holds a defined value of the variable. The second supposes the existence of a non-positional proportioning mechanism, which partitions appropriate populations of cell types, together with a sorting out process to supply the spatial pattern (Bonner, 1957; Takeuchi, 1969; Garrod & Forman, 1977; Durston, York & Weinberger, 1978); the tip may be supposed a focus for the aggregation of prestalk cells (Durston et al. 1978).

In the course of our investigations into the factors that regulate pattern and polarity in cellular slime moulds we have begun a series of studies with agents that are known to affect these developmental aspects in other systems. In this report, we show that colchicine, which affects polarity and differentiation in Hydra (Corff & Burnett, 1969), dramatically affects pseudoplasmodial organization and induces differentiation of stalk cells in D. discoideum.

Spores of Dictyostelium discoideum strain NC4 were spread on nutrient agar with Escherichia coliB/r as a food source (Durston, 1974). After 3 days, at 21 °C in the dark, the cultures consist of mainly standing and migrating pseudoplasmodia. Pseudoplasmodia were picked from plates using a microscalpel and were replated on 1 · 5 % water agar plates with or without colchicine. Colchicine from two different sources was employed (lot 53C-1680, Sigma Chemical Co.; control no. 7474101/1, Boehringer, Mannheim). The agar plates were prepared by mixing 0 · 5 ml of warm sterile 3 % water agar with 0 · 5 ml of a millipore-filtered (Millex, Millipore Inc.) colchicine solution or sterile distilled water in small plastic Petri dishes (35 × 10 mm; Corning no. 25000). In one series of experiments D. discoideum spores and bacteria were spread on colchicine-containing nutrient agar which had been prepared in a similar manner. The plates were stored in the dark at 21 °C. At selected intervals the gross morphology of the culture was assessed using a dissecting microscope and cell masses were removed from each culture, squashed in distilled water, and examined with phase contrast microscopy. All photographs were taken with a Zeiss Photomicroscope using Panatomic X film.

Neutral red stained slugs were prepared by mixing cells of D. discoideum, of all developmental stages, with bacteria on water agar and adding a drop of neutral red solution (6 mg/ml) so that a gradient of staining existed through the cell mixture. The next day darkly stained slugs with clearly defined prestalk zones were picked up with a microscalpel and plated on control or colchicine plates. The pattern of neutral red staining in these slugs was examined with a dissecting microscope using reflected and transmitted light.

Cell masses were placed in a 20 μl drop of 0 · 1 % Calcofluor white ST (American Cyanamid) on a glass slide and covered with a coverslip. The fluorescence of stained cellulose was detected using a Zeiss standard microscope equipped with a Zeiss IVfl H650 epiluminator and appropriate interference filters (Zeiss combination 4B7703). Photographs were made using Kodak 2475 recording film.

Cell masses from control and colchicine plates were placed in a drop of distilled water on slides that had been previously coated with gelatine (Rogers, 1967). A coverslip was added and the squash preparations were left at room temperature for about 15 – 30 min to allow compression and adhesion of the cell masses. The squash preparations were then placed in a freezer (− 20 °C) until used. The coverslip was pried from the frozen slide and the squashes were fixed in cold methanol for 3–5 min. The fixed preparations were then stained with rabbit antispore serum and goat anti-rabbit FITC-IgG (Nordic Pharmaceuticals). The antispore serum had been prepared against freeze dried spores of D. mucoroides by Dr S. Brahma. The fluorescence of prespore cells was examined using the microscopes described above equipped with appropriate interference filters (Zeiss combination: 487709), and photographed using Kodak recording film.

Within 2 h of plating on 0·05 M colchicine agar each slug reorganized up into two to five irregular shaped cell masses (Fig. 1B). This subdivision of slugs continued for at least 24 h (Fig. 1 C, D) with the number of mounds produced generally being related to the original slug length. After 24 h the cell mounds became very discrete in their morphology, each one possessing a sharply defined tip. At the base and between each cell mound was a mass of material. The cell mounds move about continuing to leave this material behind (Fig. 1F). By 7 days very little remained of the cell masses (Fig. 1G). When the material that lay at the bases and between the cell masses was examined by phase microscopy it was seen to contain stalk cells and extracellular material. When a group of cell masses were stained in situ with Calcofluor, the cell mounds exhibited no fluorescence while the basal and inter-mass material fluoresced brightly (Fig. 2A). Staining of individual mounds revealed that the fluorescence originated at their bases (Fig. 2B). Stalk cells were present in the fluorescent material, becoming evident after 20 h (Fig. 1E). By 5–7 days all the cells had transformed into stalk cells (Fig. 1H) but spores were never observed. Of 276 slugs plated on 0·05 M colchicine, none deviated from the above description.

Slugs plated on 0·01 M colchicine retained their normal morphology except that a small posterior protrusion was usually in evidence (Fig. 3 A). The migrating slugs left behind small mounds of stalk cells in their slime trails. The projection from the rear of migrating slugs was also made up of stalk cells (Fig. 3B, C). Calcofluor staining of these slugs revealed the same pattern of fluorescence as for the cell mounds produced on 0·05 M colchicine (Fig. 2B). Slugs plated on 0·01 M colchicine agar rarely fruit, even in the presence of bright overhead light. Of 187 slugs only 27 culminated, forming fruiting bodies which lacked a basal disc. Slugs left on the treated agar are mostly transformed into stalk cells but even by 7 days some undifferentiated cells were still evident.

Slugs plated on 0·025 M colchicine showed one or other of the patterns described above. Usually both induced phenotypes were expressed on the same plate by different slugs. Of 187 slugs plated on 0·025 M colchicine only 10 culminated, with each fruit lacking in a basal disc. At 0·005 and 0·0025 M colchicine did not detectably affect slug morphology but fruiting occurred in only 41 % (46 of 112) or 70 % (28 of 40) of the cases, respectively. In these cases fruiting was normal.

Slugs or cell mounds, from all colchicine concentrations, that were replated onto colchicine-free water agar after 3 days underwent culmination. Fruits formed from slugs treated with 0·01 M colchicine or above formed fruiting bodies that lacked a basal disc and that contained normal stalk cells but small round spores.

When neutral red-stained slugs are plated on water agar and left in the dark, their staining patterns become increasingly sharper so that a clearly defined anterior red region and smaller posterior stained region are evident (Fig. 4).

By 24 h on 0·05 M colchicine, each cell mass had a clearly defined bright red tip (Fig. 4). No staining was evident in other regions. By 48 h, the tip remained dark red but red-stained cells also became evident in the main body of the cell masses. By 72 h the staining was reminiscent of the control staining patterns except that basal staining was enhanced and red stained stalk cells projected from the rear of the cell masses. The neutral red staining pattern of slugs plated on 0·01 M colchicine was markedly different from that of control slugs (Fig. 4). Within the first 24 h the zone of neutral red staining had begun to extend two-thirds to three quarters of the way backward and appeared to increase in intensity. At 48 h most slugs are uniformly stained. By 72 h the staining was very bright and was limited to the rear end of the slugs. Slugs plated on 0 025 M showed one or other of the above patterns.

Squashes of control slugs, of all ages, stained with rabbit antispore serum and goat anti-rabbit FITC IgG revealed an anterior zone containing cells with no evident fluorescence and a posterior zone containing cells with bright particles of fluorescence (Fig. 5 A). In some slugs a small region at the rear end was also devoid of fluorescence. After treatment of slugs with colchicine, at all concentrations employed in this study, these regions remained definable although the fluorescent cells were less bright in their fluorescence. For example, even after 48 h treatment on 0 · 05 M colchicine all mounds revealed a tip region deficient in fluorescence and a rear region with bright fluorescent cells (Fig. 5B).

When colchicine (up to 0·05 M) was included in the nutrient agar plates growth was slowed but was not stopped at any concentration. At 0·025–0·05 M development proceeded to late aggregation, producing many mounds as in Fig. 1D. At 0 · 005 – 0 · 01 M development progressed to pseudoplasmodium formation as described earlier (Fig. 2), but the slugs did not culminate. At lower drug concentrations development was apparently normal. When slugs or aggregates from untreated cultures were transferred to and disaggregated on colchicine agar plates, the developmental results were exactly as described above.

In a series of experiments, a total of 185 slugs were cut into prestalk and prespore regions and plated on colchicine agar. The results were the same as described above with both prespore and prestalk pieces developing in the same manner as whole slugs.

Colchicine has been shown to have dramatic effects on the development of D. discoideum. At 0 · 01 M (4 mg/ml) colchicine prevents pseudoplasmodia from culminating. This effect has previously been reported by Cappuccinelli & Ashworth (1976). In the present study, stalk cell production at the rear of pseudoplasmodia and cell mounds was also observed at colchicine concentrations of 0 · 01 M or more. That these were true stalk cells is verified by their morphology and their fluorescence in the presence of Calcofluor (Bonner, 1970). During normal development, stalk cell production occurs only at culmination with the cells for the stalk differentiating at the tip and the basal disc cells at the base (rear) of the squat culminating cell mass (Raper, 1940; Bonner, 1967). The induction of stalk cell production at the rear of colchicine-treated slugs suggests that the future basal disc cells are being prematurely transformed into their differentiated state. This idea is supported by the evidence that treated slugs replated on agar lacking colchicine produce a fruiting body lacking a basal disc.

The second major effect on D. discoideum development was the induction of multiple axes. At concentrations above 0 · 025 M (10 mg/ml) colchicine induces pseudoplasmodia to divide into many cell mounds, each of which develops a discrete tip. Immunofluorescent staining of prespore cells revealed the absence of staining in the mound tips and bright fluorescence in the main body of the mounds. A distinct prespore-prestalk pattern resembling that seen in normal pseudoplasmodia (Tacheuchi, 1963) was seen in all treated pseudoplasmodia and cell mounds, though prespore cells stained less brightly than controls in the colchicine-treated cultures. Thus colchicine induces the formation of multiple axes from single pseudoplasmodia.

The third major effect of colchicine was on the neutral red staining patterns in the slug. Pseudoplasmodia that had been stained with neutral red underwent unusual changes in stain distribution after plating on 0 · 01 M colchicine. The staining of the anterior region of these slugs increased in intensity and progressed backwards so that they exhibited a very bright uniform staining after 48 h. The staining continued backwards until only a bright rear staining was detectable. In contrast, stained control slugs developed a clearly defined red anterior region of staining and a smaller but definite posterior region which continued to sharpen as development progressed. This control pattern has been observed by other workers (Bonner, 1952; Francis & O’Day, 1971). At 0 · 05 M, the staining pattern resembled that of controls, although there was an excessive accumulation of red stained cells at the slug’s rear end. We cannot yet account for this unusual concentration dependence.

Bonner (1952) first suggested that the neutral red staining pattern in the Dictyostelium discoideum slug occurs because neutral red is a specific stain for prestalk cells. Neutral red stains large autophagic vacuoles which are a specific organelle for the prestalk cell in slugs that have migrated for more than 24 h (Bluemink, Durston, v. Maurik and Vork, in preparation; Francis & O’Day, 1971). It is, therefore, possible that the neutral red staining change seen in colchicine-treated slugs reflect an effect of colchicine on the differentiated state of cells in the slug (so that posterior cells become prestalk-like). This conclusion is not indicated by our results using immunofluorescent staining since the immunofluorescent staining pattern is unchanged in colchicine-treated slugs. The apparent dramatic effects of colchicine on one pattern (neutral red) but not the other (fluorescence) remains to be resolved.

Corff & Burnett (1969) have previously shown that colchicine at 15 – 25 mg/ml has a marked effect on regenerating Hydra, tending to inhibit production of distal structures (tentacles), while supporting formation of proximal parts (peduncle). The basis of this effect is unknown. Our findings here indicate that colchicine can suppress the apical dominance of the Dictyostelium discoideum slug tip and can also suppress differentiation of one of the two cell types found in final fruiting body (spores), while causing premature differentiation of the others (stalk cells). Either or both of these effects may be analogous to the effect on Hydra, but the analogy remains to be clarified.

Cappuccinelli, Hames & Cuccureddu (1977) have found that colchicine binds specifically to purified D. discoideum tubulin, thus showing that the drug has its typical effect in cellular slime moulds (Ohmsted & Borisy, 1973). By this token, the drug can be used as a diagnostic tool for the analysis of microtubular importance in cellular functions in D. discoideum. From this it can be interpreted that microtubules play a critical role in morphogenesis and cellular differentiation in this organism. Of the many proposed functions for microtubules (Ohmsted and Borisy, 1973; Snider & McIntosh, 1976) two may be relevant: cellular secretion and cell division. The immediate effect of colchicine (0 · 25 – 0 · 05 M) on slug architecture suggests that the drug is interfering with the normal signals that maintain slug morphology. Possibly, this effect is due to an inhibition of the secretion of factors essential for slug structure maintenance. The immediate effect is unlikely to be due to a prevention of cell division since only a small number of slug cells undergo division and the cell cycle of D. discoideum takes over 8 h (Zada-Hames & Ashworth, 1977). The latter effects resulting in the induction of stalk cells could be due to an effect on cell division. Durston & Vork (1978) have recently shown a correlation between DNA synthetic patterns and the prespore-prestalk boundary in D. discoideum. It is also possible that it is due to an effect on cellular secretion. Rudolph, Greengard & Malawista (1977) have shown that drug enhancement of cAMP secretion from human leucocytes is increased in the presence of colchicine. It is conceivable that colchicine mimicry of cAMP action (i.e. stalk cell induction and multiple tip formation) may represent an analogous situation.

We would like to thank Peter Poot for his assistance in some of the early work on the immunofluorescent staining and Carmen Kroon for her photographic and artistic work. Dr S. Brahma kindly supplied the antispore serum.