ABSTRACT
Cells disaggregated from slugs of Dictyostelium discoideum were cultured in Bonner’s salt solution in roller tubes. Cells rapidly stuck together to form an amorphous loose agglutinate which was later transformed into a spheroidal tight agglutinate surrounded by slime sheath material. Prespore cells in the loose agglutinate underwent partial dedifferentiation by starting to decompose their specific antigen until formation of the tight agglutinate, in which the antigen was resynthesized. During the process, there was some decrease in the proportion of prespore cells.
Changes in the distribution of prespore and prestalk cells in the agglutinates were examined by using immunocytochemical staining. They were randomly distributed in the early agglutinates, but became well separated in 4 h agglutinates in such a way that prestalk cells were completely enveloped by prespore cells. Prestalk cells later came outside to be partially enveloped and finally occupied a hemisphere side by side with prespore cells. During the process, cells in one or two outer layers differentiated into prestalk or stalk cells. Similar changes in the distribution pattern were observed, when labelled prestalk cells were cultured with unlabelled prespore cells and their distribution in co-agglutinates was followed by autoradiography. It was concluded from these results that the majorities of prestalk and prespore cells isolated from slugs are sorted out in agglutinates without changing their original cell types, and that the sorting-out occurs in the complete absence of polar structures and movement of the cell mass. The distribution patterns in agglutinates of prestalk and prespore cells were discussed with reference to intercellular adhesion among/between them.
INTRODUCTION
Upon, depletion of the food supply, amoebae of the slime mould Dictyostelium discoideum aggregate chemotactically to form a slug-shaped cell mass. The slug finally constructs a fruiting body consisting of a terminal spore mass on a supporting cellular stalk. During formation of the fruiting body, the anterior cells of the slug differentiate into stalk cells, while the posterior cells differentiate into spores. It has been revealed by various means that the anterior prestalk cells of a slug have characteristics different from the posterior prespore cells. When sections of D. discoideum slugs are stained with fluorescein-conjugated antiserum produced against spores of D. mucoroides, the prestalk cells have no staining while the prespore cells are strongly stained in cytoplasmic granules (Takeuchi, 1963, 1972).
It has been shown that when prestalk and prespore cells in the slug are displaced or mixed they are sorted out to occupy their original positions. This was first shown by Bonner (1952) who grafted vitally stained cells from the anterior end onto the posterior end of a colourless migrating slug and found that the coloured cells move up to the anterior positions as the slug migrates. The finding was later refuted by Farnsworth & Wolpert (1971), but recently confirmed by Yamamoto (1977). On the other hand, Takeuchi (1969) who mixed cells disaggregated from the anterior and the posterior portions of slugs and allowed them to reaggregate on agar found that they are sorted out during slug formation to occupy their original positions in a reformed slug.
The present work was attempted to examine whether or not such sorting out of cells is brought about by morphogenetic movements of a cell mass, i.e. formation and migration of a polarized slug. For this purpose, prestalk and prespore cells disaggregated from slugs were cultured in roller tubes to produce round spherical co-agglutinates, in which their distributions as well as states of differentiation were followed immunocytochemically and autoradiographically. As a result, it was found that the two cell types are sorted out in an agglutinate completely devoid of polar structure and movement with most cells maintaining the original cell types.
MATERIALS AND METHODS
Organism and culture
D. discoideum strain NC 4 was used in this study. Spores of the strain were inoculated with Escherichia coli B/r on a standard nutrient agar medium (Bonner, 1947). Amoebae were harvested at the stationary growth phase and washed by repeated centrifugations (1500 rev./min, 1·5 min). To obtain slugs, washed amoebae were streaked on non-nutrient agar (2%) and incubated at 21 °C. To obtain preculminating cell masses, amoebae were grown, at 21 °C, with E. coli in a nutrient liquid medium on a rotary shaker (150 rev./min). After being washed, 107 amoebae were resuspended in 0-5 ml KCl-containing phosphate buffer (K-P buffer: 20mM-KCl, 10 mM phosphate buffer, pH 6-5) and deposited on a 5·5 cm Whatman No. 50 filter paper supported by three pieces of 5-5 cm Toyo No. 2 filter paper saturated with K-P buffer. Development was allowed to proceed at 21 °C, and preculminating cell masses were collected after 17 h of incubation.
Roller tube culture of slug cells
Migrating slugs were washed by two centrifugations (500 rev./min, 30 sec) in K-P buffer to be freed of unaggregated amoebae and disaggregated at 4 °C by forced pipetting in EDTA-P buffer (1 mM EDTA, 20mM-NaCl, 40 mM phosphate buffer, pH 7·0). The buffer was then filtered through nylon mesh (23 μm openings) to remove cell clumps. Disaggregated cells were washed, with Bonner’s salt solution (BSS: 10 mM-NaCl, 10mM-KCl, 3 m.M-CaCl2) and resuspended in BSS at 1·7 × 106 cells/ml. Roller tube culture of disaggregated cells was made as described previously (Takeuchi, Hayashi & Tasaka, 1977). Each tube containing 3 ml of the cell suspension was rolled at 21 °C, at 28 rev./ min. Agglutinates formed were collected at intervals by centrifugations at 500 rev./min for 30 sec.
Fractionation of presumptive cells
The method employed was a modification of that previously used for D. mucoroides cells (Oohata & Takeuchi, 1977). Stepwise density gradients consisted of Urografin (Schering) solutions of specific gravities, ρ 1·21 (1 ml), ρ 1·17 (2·5 ml), ρ 1·61 (2 ml), and ρ 1·15 (2 ml), which were adjusted to fractionate D. discoideum cells disaggregated at the preculmination stage. Disaggregated cells were suspended in 1-5 ml EDTA-P buffer at 3x108 cells/ml and were centrifuged as described by Oohata & Takeuchi (1977). Cells were collected from bands by syringes and washed with BSS by three centrifugations at 2000 rev./min for 5, 3, and 2 min.
Immunocytochemistry and microfluorometry
Immunocytochemical staining of prespore cells was made with fluorescein isothiocyanate (FITC)-conjugated antiserum produced against spores of D. mucoroides, as described by Takeuchi (1963). Agglutinates were disaggregated in a 50 mM Tris-buffered (pH 7·2) solution of 0·1 % pronase containing 25 mM dimercaptopropanol (Takeuchi & Yabuno, 1970). After being washed, disaggregated cells were placed on a coverglass, fixed in methanol and stained. The preparations were observed in a dark field with a Reichert fluorescence microscope (Zetopan). Fluorescence intensities of stained cells were determined using a Reichert microspectrophotometer attached to the microscope, as described by Hayashi & Takeuchi (1976). For histological staining, agglutinates were fixed, embedded in Paraplast (Sherwood) and sectioned 3-5 μm thick. The sections were stained with the antiserum and observed.
Autoradiography
To label amoebae, they were grown with a thymidine-requiring mutant strain (B3) of E. coli in a nutrient liquid medium containing [3H]thymidine (5μCi/ml, Radiochemical Centre, Amersham) plus 5 μg/ml thymidine. After being washed, cells were allowed to develop on filter papers. Labelled prestalk and unlabelled prespore cells were fractionated from preculminating cell masses as described, mixed at 1:3 ratio and incubated in roller tubes. Agglutinates formed were fixed in Famer’s fixative and sectioned 5 μm thick. The preparations were dipped in liquid emulsion, Sakura NR-M2, in a dark room. They were stored in a black box at room temperature for 2 weeks and developed in Sakura Conidol-X for 6 min at 21 °C, fixed in Coni fixer and washed with distilled water.
RESULTS
Formation of agglutinates from disaggregated slug cells
When disaggregated slug cells were cultured in BSS in roller tubes, cells quickly stuck together to form amorphous agglutinates such as shown in Fig. 1A. When placed on a slide glass, cells of these agglutinates adhered to the glass and tended to disperse. They were thus similar to the ‘loose’ agglutinates which form in an early roller tube culture of washed vegetative amoebae (Takeuchi et al. 1977). When the culture was continued, cells in the loose agglutinate secreted on its surface viscous material like the slime sheath surrounding the slug. The agglutinate now assumed a smooth spheroidal shape and no longer showed a tendency to be dispersed on glass (Fig. 1B). Such changes in the agglutinate appear comparable to those from the ‘loose’ to the ‘tight’ agglutinates which have previously been described in a roller tube culture of washed vegetative amoebae (Takeuchi et al. 1977). Formation of tight agglutinates was completed after 3-4 h of culture. Although no apparent changes of the agglutinates were discernible during further culture, an increasing number of cells adjacent to the outer sheath were observed to form heavy cellulose walls and large vacuoles in a fashion characteristic of stalk cells (Fig. 1C). On rare occasions, a few spore-like cells appeared in the agglutinates after 12 h of culture.
Agglutinates formed in roller tube culture of disaggregated slug cells. The culture was conducted according to the Method section. (A) 2 h agglutinate: amorphous loose agglutinates without slime sheath material. (B) 6h agglutinate: a spheroidal tight agglutinate with surrounding sheath. (C) Crushed 12 h agglutinate: mature stalk cells are differentiated in the outer layer.
Agglutinates formed in roller tube culture of disaggregated slug cells. The culture was conducted according to the Method section. (A) 2 h agglutinate: amorphous loose agglutinates without slime sheath material. (B) 6h agglutinate: a spheroidal tight agglutinate with surrounding sheath. (C) Crushed 12 h agglutinate: mature stalk cells are differentiated in the outer layer.
Changes of prespore cells in agglutinates
Taking advantage of the fact that the prespore cells of the migrating slug were specifically stained in cytoplasmic granules with heteroplastic antispore serum (Takeuchi, 1963), changes in the proportion of prespore cells and their antigenic contents in agglutinates were examined during the roller tube culture. Agglutinates were collected at various times of culture. Cells disaggregated therefrom were stained with the FITC conjugated anti-Z). mucoroides spore serum, and the numbers of cells with and without stained granules were counted as described above. The ratio of prespore to total cells was initially about 75 % (in slugs), but gradually decreased in agglutinates to about 63 % within 5-6 h of culture (Fig. 2A). Fluorescence intensities of individual cells were microspectrophotometrically determined, and their frequency distributions at indicated times were shown in Fig. 3. The frequency distribution pattern of 0 h cells indicates that cells comprising slugs were grouped into two classes : those of high fluorescence intensities corresponded to prespore cells and those of low to prestalk cells (Hayashi & Takeuchi, 1976). When the culture was continued, the former gradually decreased both in number and fluorescence intensity, with the concomitant increases of the latter. Thus, the distribution became too spread out to be clearly bimodal, as observed with cells of 4 and 6 h agglutinates (Fig. 3). Formation of tight agglutinates, however, was soon followed by increases in number and intensity of prespore cells, resulting in restoration of a bimodal distribution. On the basis of these histograms, the average fluorescence intensity of prespore cells, indicative of their antigen content, was calculated at each culture time, regarding cells exceeding 15 (arbitrary unit) in intensity as the prespore cells. As shown in Fig. 2B, the average fluorescence intensity began to decrease after 2 h of culture, but showed a considerable increase after formation of tight agglutinates to the level equal to that of prespore cells contained in the slug.
(A) Changes in the ratio of prespore to total cells in agglutinates during roller tube culture of disaggregated slug cells. The culture was conducted as described in the Method section. Agglutinates formed were collected at intervals, and cells disaggregated from them were stained with the fluorescein-conjugated anti-D. mucoroides spore serum. Prespore cells were identified as those having stained granules. The value at 0 h represents the ratio in slugs. The horizontal bar indicates the time of appearance of tight agglutinates, and the vertical ones standard deviations. (B) Changes in the fluorescence intensities (arbitary unit) of prespore cells in agglutinates during the culture. The average fluorescence intensities were calculated from each histogram shown in Fig. 3, regarding cells exceeding 15 in intensity as the prespore cells.
(A) Changes in the ratio of prespore to total cells in agglutinates during roller tube culture of disaggregated slug cells. The culture was conducted as described in the Method section. Agglutinates formed were collected at intervals, and cells disaggregated from them were stained with the fluorescein-conjugated anti-D. mucoroides spore serum. Prespore cells were identified as those having stained granules. The value at 0 h represents the ratio in slugs. The horizontal bar indicates the time of appearance of tight agglutinates, and the vertical ones standard deviations. (B) Changes in the fluorescence intensities (arbitary unit) of prespore cells in agglutinates during the culture. The average fluorescence intensities were calculated from each histogram shown in Fig. 3, regarding cells exceeding 15 in intensity as the prespore cells.
Frequency distributions of the fluorescence intensities of cells comprising agglutinates at various times of roller tube culture. Agglutinates were collected at times indicated, and cells disaggregated from them were stained with the fluorescein conjugated antiserum. The fluorescence intensity of a cell was measured with a microphotometer and is expressed as a relative value to a standard.
Frequency distributions of the fluorescence intensities of cells comprising agglutinates at various times of roller tube culture. Agglutinates were collected at times indicated, and cells disaggregated from them were stained with the fluorescein conjugated antiserum. The fluorescence intensity of a cell was measured with a microphotometer and is expressed as a relative value to a standard.
Distribution of presumptive cells as revealed by immunocytochemistry
Changes in the distribution of prespore and prestalk cells in agglutinates were examined by means of immunocytochemical staining of sections of agglutinates fixed at various times of culture. In 2 h (loose) agglutinates, prespore cells were either clustered or intermingled with unstained, probably prestalk cells (Fig. 4 A). In 4 h (tight) agglutinates covered by stained sheath, prespore cells were well separated from and completely enveloped unstained (prestalk) cells (Fig. 4B). The latter was confirmed by serial sections not to come in contact with any part of the surface of an agglutinate. Later, however, the unstained region came outside to the surface to be partially enveloped by prespore cells after about 6 h of culture (Fig. 4C). Simultaneously, many cells in one or two outer layers became unstained and some of them differentiated into stalk cells. Final distribution of both types of cells was attained in 8 h agglutinates, in which prespore and prestalk cells occupied each hemisphere (Fig. 4D). All the cells in outer layers were now either prestalk or stalk cells.
Sections of agglutinates stained with the fluorescein-conjugated antiserum. Agglutinates were fixed and sectioned after various times of roller tube culture. (A) 2 h agglutinate: prespore cells (stained) and prestalk cells (unstained) are distributed randomly. (B) 4 h agglutinate: prestalk cells are completely enveloped by prespore cells, and the agglutinate is covered by slime sheath (stained). (C) 6 h agglutinate: the prestalk cell mass is partially enveloped by prespore cells. (D) 8 h agglutinate: the prestalk and prespore cell masses occupy each hemisphere. Cells in outer one or two layers of the agglutinate are prestalk or stalk cells.
Sections of agglutinates stained with the fluorescein-conjugated antiserum. Agglutinates were fixed and sectioned after various times of roller tube culture. (A) 2 h agglutinate: prespore cells (stained) and prestalk cells (unstained) are distributed randomly. (B) 4 h agglutinate: prestalk cells are completely enveloped by prespore cells, and the agglutinate is covered by slime sheath (stained). (C) 6 h agglutinate: the prestalk cell mass is partially enveloped by prespore cells. (D) 8 h agglutinate: the prestalk and prespore cell masses occupy each hemisphere. Cells in outer one or two layers of the agglutinate are prestalk or stalk cells.
Distribution of presumptive cells as revealed by autoradiography
To examine whether the above-described changes in distribution of the two cell types are mainly attributable to rearrangement of cells whose original differentiated states are maintained or to dedifferentiation of cells followed by their redifferentiation after formation of agglutinates, labelled prestalk cells were mixed with unlabelled prespore cells and their locations in co-agglutinates were traced by autoradiography. Amoebae were labelled with [3H]thymidine as described in the Method section. Autoradiography of these cells revealed grains concentrated on the nucleus, except for some scattered in the cytoplasm, probably on mitocondria.
Separation of prestalk from prespore cells
To fractionate the two cell types, cells disaggregated from preculminating cell masses were centrifuged through a discontinuous Urografin gradient. Preculminating cell masses were used, because they gave better separation than slugs. After centrifugation, three bands formed at interfaces, as shown in Fig. 5. Cells in each band were stained with the fluorescent antispore serum to identify prespore cells. As shown in Table 1, most cells in the band I were unstained, while those in the band III were stained. In contrast, the band II was a mixture of stained and unstained cells. It was thus concluded that about 95 % of cells in the bands I and III were prestalk and prespore cells respectively. It was noticed during the experiments that the specific gravities of disaggregated cells are considerably affected by many factors, such as the conditions of slug formation, the method of disaggregation, the number of cells layered on the gradient and so on. If different conditions are used, the densities of Urografin solutions should be adjusted to obtain clear separation of the two cell types.
The numbers of stained and unstained cells in the bands I and III produced by fractionation of disaggregated cells on a Urografin gradient

A centrifugal pattern on a stepwise Urografin gradient of cells disaggregated at the preculmination stage. Three bands (I, II and III) formed at interfaces.
Distribution of labelled cells in agglutinates
Labelled prestalk cells and unlabelled prespore cells were mixed and cultured in roller tubes as described above. Agglutinates formed were fixed and sectioned at subsequent times. Distributions of the two types of cells in agglutinates were examined using the method of autoradiography. Labelled prestalk cells were first scattered randomly in 2 h (loose) agglutinates (Fig. 6A), indicating that both prestalk and prespore cells adhered without any selection. In 4 h (tight) agglutinates, however, the majority of labelled prestalk cells were concentrated in the centre (Fig. 6B). Later in 8 h agglutinates, labelled cells came outside to be partially enveloped by unlabelled cells (Fig. 6C). Some labelled cells were also observed scattered on the surface of the agglutinates, but most of them were those adsorbed on the slime sheath after its formation. It is thus concluded that the majority of labelled cells which had previously been prestalk cells in the slug followed the same pattern of distribution as that of cells unstained by immunocytochemical staining (cf. Fig. 4). This indicates that most prestalk cells are sorted out in the newly formed agglutinates without changing their differentiated state. When labelled prespore and unlabelled prestalk cells were mixed in a roller tube culture, their distribution pattern underwent the same type of changes as described above.
Autoradiographs of sections of agglutinates. Labelled prestalk cells and unlabelled prespore cells were mixed and cultured in roller tubes. Co-agglutinates formed were fixed and sectioned at intervals, and their autoradiographs were prepared. (A) 2 h agglutinate: original prestalk (labelled) and prespore (unlabelled) cells are randomly dispersed in the agglutinate. (B) 4 h agglutinate: original prestalk cells are accumulated in the inside of the agglutinate. (C) 8 h agglutinate: most original prestalk cells are partially enveloped by original prespore cells.
Autoradiographs of sections of agglutinates. Labelled prestalk cells and unlabelled prespore cells were mixed and cultured in roller tubes. Co-agglutinates formed were fixed and sectioned at intervals, and their autoradiographs were prepared. (A) 2 h agglutinate: original prestalk (labelled) and prespore (unlabelled) cells are randomly dispersed in the agglutinate. (B) 4 h agglutinate: original prestalk cells are accumulated in the inside of the agglutinate. (C) 8 h agglutinate: most original prestalk cells are partially enveloped by original prespore cells.
DISCUSSION
When cells disaggregated from slugs were cultured in roller tubes, they first formed a loose agglutinate which later became tight by secreting slime sheath material on its surface. During the period of loose agglutinates, both the proportion of prespore cells and their antigenic contents decreased, indicating that cells in such an agglutinate underwent the same process of dedifferentiation as completely isolated cells (Takeuchi & Sakai, 1971). After formation of a tight agglutinate, however, prespore cells resynthesized the antigen, but their proportion remained about 60%. That this ratio is characteristic of such a submerged agglutinate was shown by the following fact. When slug cells disaggregated by the use of pronase-dimercaptopropanol were cultured, the transition from a loose to a tight agglutinate was delayed about 2 h. As a result, the proportion of prespore cells decreased to about 25 % in loose agglutinates, but the final proportion in tight agglutinates was about 60% (unpublished).
Recently, roller tube or rotary shake cultures of washed vegetative amoebae were conducted by Forman & Garrod (1977), Sternfeld & Bonner (1977) and Takeuchi et al. (1977), who observed differentiation of prespore cells, spores and stalk cells in agglutinates kept under submerged conditions. In their roller tube culture of washed vegetative amoebae, Takeuchi et al. (1977) found the same type of transition from a loose to a tight agglutinate as observed in the present study and proposed that formation of tight agglutinates is a prerequisite for differentiation of prespore cells. The proportion of prespore cells in such an agglutinate attained to about 50 %.
The distribution of prestalk and prespore cells in agglutinates was investigated by two different methods, immunocytochemical staining for prespore cells and autoradiographical tracing of labelled prestalk or prespore cells. The fact that the distributions as revealed by both the methods were essentially the same indicates that the majority of presumptive cells are rearranged in an agglutinate without changing their original cell types. However, this does not completely exclude occurrence of the conversion of cell types in agglutinates. In fact, the conversion, though on a small scale, is suggested to have occurred by a certain decrease during the culture in the proportion of prespore cells. The decrease was, at least in part, attributable to redifferentiation of prespore cells located in the outer layers of agglutinates into prestalk or stalk cells. The reason why they underwent such differentiation is unknown, but the phenomenon parallels with the fact that the prespore cells next to the agar surface in a migrating slug lose the prespore antigen during migration (Takeuchi et al. 1977).
The present study showed that prestalk cells are first sorted out inside to be completely enveloped by prespore cells, but that they later move outside to be only partially enveloped. The fact that such sorting out proceeded in a cell mass which, unlike the slug, underwent no morphogenetic movement suggests that it may have been brought about by differential intercellular adhesion.
Steinberg (1963) described in his differential adhesion hypothesis that sorting out behaviour of embryonic cells is attributed to differences in the strength of their intercellular adhesion. According to his hypothesis, the fact that prespore cells completely enveloped prestalk cells in the 4 h agglutinate indicates that the work of adhesion among prestalk cells (Wpst-pst) was greater than that among prespore cells (Wpsp-psp) and that the work between prespore and prestalk cells (Wpsp-pst) was in between: Wpst-pst > Wpsp-pst > Wpsp-psp. Furthermore, the changes from complete to partial envelopment of prestalk cells as observed in later agglutinates are interpreted to indicate that the work of adhesion among the same types of cells was relatively increased during the culture than that between the different types to become Wpst-pst > Wpsp-psp > Wpsp-pst. That prestalk cells are more adhesive than prespore cells agrees well with the previous observations made by Maeda & Takeuchi (1969), Takeuchi & Yabuno (1970) and Yabuno (1971).
ACKNOWLEDGEMENTS
This work was in part supported by grants-in-aid to I. T. from the Ministry of Education of Japan (No. 248010) and Takeda Science Foundation.