Size regulation in Dictyostelium

In Dictyostelium discoideum the size of the slug is regulated at two points in the life cycle, one of which occurs at aggregation and the other at the time of slug formation. Aggregation is the period when the cell mass initially forms by the collection of cells to the centre of the aggregation territory. When there is a sparse population of cells the size is determined solely by the territory (Bonner & Dodd, 1962). In contrast, crowded populations do not show a one to one correspondence between aggregate number and the number of fruiting bodies because large aggregate cell masses divide to form separate slugs. In studying the control of size in large aggregates Hohl & Raper (1964) found there was a ‘critical mass’ or volume of cells above which complete intercellular integration could not be maintained.

In this report we are concerned with mechanisms of size control of large aggregates. Ultimately we would like to have a molecular understanding of the control process, and as a first step toward this goal the experiments to be described elucidate some of the properties of the process. A quantitative method for studying this kind of size control is described in this paper.

What is novel in the studies presented here is evidence that this size control is achieved by two interlocking factors; the tip, which is the presumptive stalk region of the fruiting body, suppresses the formation of secondary tips in the surrounding cell mass, and the surrounding cells can differ in their ability to be affected by the dominance of the tip. There is an upper limit to the volume of the cell mass that can be inhibited ; aggregates smaller than this volume will generally maintain integrity whereas aggregates larger than this volume will divide. An assay using a microchamber with precisely known dimensions was developed to estimate the volume and distance over which inhibition occurs, the time over which it acts and the effect of the tip size and cell mass shape on size regulation. Theoretical calculations using these results were found to support the hypothesis that an inhibitor spreads out from the tip by simple diffusion and are consistent with the finding that the inhibition can act across a thin agar membrane. In addition, the tip inhibition mechanism will be compared to the aggregation process.

The wild-type strain of D. discoideum, DdH and a mutant strain, P-4, (Chia, 1975) were used. Although P-4 has normal growth and aggregation, each aggregate forms a number of small slugs. Amoebae were grown and harvested on agar plates as described using Escherichia coli as the bacterial associate (Bonner, 1947). In some experiments cells were stained with vital dyes in order to distinguish the darkly staining prestalk region of the slug; neutral red or nile blue sulphate stock solutions contained 7 · 5 mg/100 ml water (Allied Chemical and Dye, Corp.). To stain cells, vegetative amoebae were washed off the plates in cold distilled water and spun down at 150 g for 2 min. The cell pellet was resuspended in 15 ml of water containing 1 ml of the stock solution for each 0 · 1 ml of packed cells in the pellet and spun down again. After harvesting, the cells were used immediately, or they were resuspended in 100% Bonner’s saline solution (Bonner, 1947), dispensed on 1% nonnutrient water agar (Bacto-agar) and allowed to develop in the dark at 22 °C. When slugs had formed prestalk and prespore cell masses were obtained by transecting the slugs with a microknife as described by Gregg (1967).

In order to determine precisely the sizes of slugs small wells were made by embedding wire of appropriate diameter in electron microscopy plastic (Spurr embedding kit, Polysciences, Inc.). The plastic was polymerized in gelatin capsules at 60 °C for two days and the wires removed after softening the plastic for 10 min in boiling water. The plastic blocks were sectioned on an AO 820 microtome to the appropriate thickness for the volume required and the sections placed over Millipore filters (HABP 04700). Slit wells were made in a similar manner by embedding shim stock (3 · 45 mm × 125 μm). Cells were transferred to the wells using a hair loop and the wells were placed in Petri dishes over Millipore prefilters (AP10 04700) saturated with 2 ml of 17 mM-Sorensen phosphate buffer, pH 6 · 0. The cell masses were then incubated in the light at 22 °C. The transformation of a cell mass to a slug is directed by a single tip and therefore a measurement of the numbers of tips formed in each well provides a means to determine the average slug volumes by dividing the well volume by the number of tips per well.

To determine quantitative aspects of tip inhibition, such as the cell mass volume and the distance over which inhibition is effective, tips taken from average-sized slugs were grafted to prespore cell masses in standardized sizes and shapes (20 × 10⁶μm³ disc wells and 21 · 5 × 10⁶μm³ slit wells). The number of tips forming in wells with and without tip grafts were compared and, if the intact grafted tip inhibits secondary tips from forming in prespore cell masses, the average slug volume will be large. The distance over which inhibition acts can most easily be determined using slit wells.

To test for an effect of tip size the tips to be grafted were of a large size range and the recipient prespore cell mass volume was 30 × 10⁶μm³. The volumes of the tips to be grafted were estimated by measurements of the tip base and height, assuming that the tip is a cylinder.

In order to test whether or not the inhibitor can diffuse across an extracellular space tip cells and prespore cells were separated by a thin agar membrane (ca. 100 – 125 μm) made of 3% Bacto-agar in 17 mM-Sorensen phosphate buffer. Dialysis membranes (65 μm with MW cutoffs of 6 – 8000 and 12000 No. 3787-D20 from A. H. Thomas Co.), Millipore filters (0 · 45 μm) and Nucleopore membranes (0 · 6, 0 · 8 and 1 · 0 μm) were also used in some experiments in place of the agar membrane.

The volumes of the wells used in these experiments ranged from 2 · 5 to 50 × 10⁶μ m³, each with nearly equal surface to volume ratios with the exception of the smallest wells. The disc-shaped cell masses obtained by artificially clumping the cells mimic the size and shape of the cell masses obtained normally during aggregation, and provide a convenient way to obtain cell masses of any volume desired.

The average number of slugs formed 11 h after placing the freshly harvested cells in the wells is plotted for each volume in Fig. 1. The results with P-4 show clearly that the number of slugs/well increases with cell volume and that P-4 cell masses form approximately three times as many slugs than the wildtype DdH strain. The increase in number of slugs with increased cell mass is taken to mean that there is an upper limit to the size of the cell mass.

Both the P-4 and DdH lines were determined by linear regression analysis of the data: for P-4 N = 93 wells, r = 0 · 99 and the mean ± S.D. is 3 · 7 ± 1 · 1 × 10⁶μm³; for DdH only the N = 73 wells above 20 × 10⁶μ m³ were used in the analysis since as can be seen DdH cell masses less than 20 × 10⁶μm³ almost always form only one slug, r = 0 · 28 and the mean ± S.D. is 15 · 3 ± 3 · 0 × 10⁶μm³. The correlation coefficients for the P-4 and DdH data are significantly different from 0 at P < 0 · 001 and 0 · 02 respectively.

The DdH slug volume determined here is almost identical to the value of 15 × 10⁶μm³ reported by Hohl & Raper (1964) using their totally different method.

The following experiments examine whether or not the anterior tip of the slug inhibits the formation of independent rival tips. The tip inhibition assay is shown approximately to scale in Fig. 2. The results allow us to determine whether the reduced size in P-4 slugs is a result of the P-4 tip producing a less effective inhibitor or the prespore cells being more resistant to the inhibition or both.

The first experiment was a homotypic graft (one strain was used for both the tip graft and the prespore recipient). With DdH it can be seen that when a tip is placed on the prespore cell mass as compared to a cell mass without a tip the average slug volume significantly increases (Table 1 expt 1). Since the well volumes were 20 × 10⁶μ m³, it is obvious from the average slug volumes in the inhibited wells (17 · 3 × 10⁶μm³) that the grafted tip almost completely inhibited new tips from arising among the prespore cells. In addition, it should be noted that the volume of cell mass dominated by the transplanted tip is nearly the same as the volume of an average slug formed from cells placed directly into the wells. This implies that tip grafts are as effective in inhibition as tips which arise autonomously from cell masses.

Furthermore, in a heterotypic graft (a different strain was used for the tip and the prespore cells) in which P-4 tips are used (expt. 2) it was found that P-4 tips are nearly as effective as DdH tips in inhibition strength. However, when P-4 prespore cell masses are used (expts 3 and 4) the prespore cells are not significantly inhibited, regardless of whether a DdH or P-4 tip is grafted. This evidence suggests that P-4 slugs are small as a result of some altered prespore function which makes these cells much less sensitive to the normal inhibitor levels.

The distance over which inhibition can extend is also an important parameter to consider because pre-tip aggregates are irregular in shape and the long and narrow aggregates can be mimicked by prespore cell masses in slit wells. The inhibition range in such cell masses can easily be measured by tip grafts to the approximate middle of the slit wells. Figure 3 shows the distribution of the inhibition distances for the four graft types.

The mean distance between the fruiting bodies formed under the grafted tip and the nearest neighbouring secondary tips was determined to be 852 ± 332 μm (mean ± S.D.) for DdH. The distribution of the distances (Fig. 3 a) is significantly greater than the distribution of distances found between the tips arising from the prespore cell masses (Fig. 3e) without the tip grafts (P < 0 · 005; one sided Smirnov test). Each tip forming in the regulating prespore cell mass inhibits approximately half of the distance separating it from its neighbours, that is 282 μm. The average DdH tip graft distance of inhibition was estimated by assuming that the secondary tips adjacent to the tip graft inhibit over a distance of 282 μm so that the difference of 570 μm is the tip graft distance of inhibition in slit wells. Clearly then the tip graft prevents newly arising tips from forming in an adjacent region.

The histograms show that there is total inhibition at distances less than 200 μm and almost 95 % inhibition up to 400 μm (Fig. 3 c). At greater distances inhibition is drastically reduced and quite variable.

The results also support the conclusions reached regarding the cell type responsible for the small P-4 phenotype. The P 4 tip inhibition distance (Fig. 3 c) is not significantly different from the DdH distance (Fig. 3 a) in a two-sided Smirnov test, P > 95 %, whereas the comparison tip grafts to DdH and P-4 prespore cell masses (Figs. 3 a v. d and c v. b) shows that the inhibition distance is significantly shorter when P-4 prespore cell masses are used (onesided Smirnov test, P < 0 · 005).

The results just mentioned allow us to examine in more detail the interesting possibility that the geometry of the cell mass surrounding a tip is an important factor affecting inhibition and the size of the cell mass. In Table 2 the Slug volumes resulting from tip grafts onto prespore cells in disc wells or slit wells are listed. Here it can be easily seen that tips grafted onto disc cell masses inhibit more cell volume than do tips grafted onto slit cell masses. The difference in the number of slugs/well in disc and slit wells serves to emphasize this effect of the cell mass shape on tip inhibition; this evidence suggests that inhibition is more effective in circular cell masses than in long and narrow cell masses.

The important question to consider now is whether any of the size variation is determined by the tip. It is possible that all tips regardless of the slug size have equal inhibition strength and that only differences in the cell’s response to inhibition determine the slug size. An alternative hypothesis to consider is that tip strength varies with tip size. A test was made to determine whether or not tip strength varies with tip size, and if it does, whether large tips can inhibit larger cell masses (Fig. 4).

There is a definite effect of tip size on the ability to inhibit secondary tip formation; the larger the tip the greater its inhibition (Table 3). This result will be discussed later because the evidence suggests that the attraction power of aggregate centres is not affected by size.

The following experiment involves separation of tip cells and prespore cells to examine whether or not the tip inhibitor requires cell contact for transmission of the inhibition. Dissociated prestalk cells in a 50 ± 10⁶μm³ disc well were placed under a layer of agar, dialysis membrane, Millipore or Nucleopore filters (Fig. 5). It is essential to put the prestalk cells under the barrier in order to prevent the prestalk cells from rapidly reforming slugs and breaking contact with the agar, dialysis membrane or filters.

In five out of six trials with agar between the prestalk and prespore cells, tip formation occurred at an average of 4 · 8 h after plating; control cell masses formed tips in ca. 2 h, therefore the mean delay in tip formation was 2 · 8 h. A view of one such experimental cell mass and its respective control is shown in Fig. 6. The photographs were taken 3 h after plating and it can be seen that a well-formed tip is present on the control prespore cell mass but not où the cell mass over the prestalk cells. Generally the prestalk cells under the agar migrate out of the well along the underside of the agar and the prespore cells above spread out on the upper surface following the wave of the prestalk cells. It is at this time some of the prespore cells form a tip. In addition, prespore cells were placed above and below the agar; no inhibition or delay of tip formation was observed.

Further experiments with dialysis membranes, Millipore and Nucleopore filters showed no delay even when, to give more time for diffusion, the prestalk cells were put into the wells 7 h before the prespore cells were placed on top of the barrier. It appears then that the inhibitor can act without direct cell contact.

The objective of this study was to uncover general rules governing size regulation and the control of prestalk cell differentiation in the cellular slime mould, D. discoideum. The tip was first shown to be the dominant organizer region in slime moulds by Raper (1940). By grafting tips onto the sides of migrating slugs he found that the slug divided approximately equally between the tips. In extending these experiments, the experiments reported here, and those of Hohl & Raper (1964), have shown that the average size of the cell mass dominated by a single tip of strain DdH is around 15 × 10⁶μ m³. Thus, there is an upper limit to the volume of cell mass that can be inhibited ; aggregates smaller than this volume will generally maintain integrity whereas aggregates larger than this volume will subdivide. Therefore the size of the slugs and fruiting bodies formed, even in large aggregates from crowded populations, are kept under a certain size.

It is interesting to contrast what is now known about the regulatory mechanism of tip inhibition with the well-studied aggregation process. First the average distance over which inhibition acts (570 μm) is somewhat less than the average attraction distance of centres (1270 μm) reported by Bonner & Dodd (1962). Nevertheless both processes involve distances which could be traversed by a diffusing small molecule within roughly several hours (Crick, 1970).

A second observation is that tips form centrally in the cell mass, but this is in contrast to the placement of centres during their initiation in small drops of aggregating cells (Konijn, 1961). In drops at low density, Konijn found that initiation of centres was random. The reasons why tip formation is invariably position dependent are obscure, but one possibility is that some diffusible activator builds up in the region with the highest cell density. Consequently, it will be especially interesting to see if the recently discovered stalk-cell-inducing factor (DIF) might play such a role in pretip cell masses (Kay, Town & Gross, 1979).

Lastly, although tip inhibition varies with the size of the tip, there is no effect of size on center attraction power. Konijn (1968) clearly showed that in D. discoideum the chemotactic response of responding populations of amoebae is independent of the size of the attracting cell mass over the range of 400 – 5000 cells. The important consequence is that the aggregate territory dominated becomes independent of the cell density. He later speculated that cyclic AMP release by each cell was reduced at high cell density and therefore there is an upper limit to the centre attraction power. In contrast, the tip graft experiments reported here show that inhibition strength steadily increases with tip size. This would suggest that each cell in a tip gives off a constant amount of the postulated inhibitor regardless of the density. It is possible that the apparent proportional increase in inhibition of tips, which are the presumptive stalk cell zones of the fruiting body, may be related to the intriguing problem of size-regulated stalk to spore ratios (Bonner & Slifkin, 1949; MacWilliams & Bonner, 1979).

In summary then, although both centre dominance over an aggregation territory and tip inhibition over the surrounding cell mass seem to involve diffusible factors, their regulatory mechanisms differ in at least two fundamental ways.

Another factor which I have found to affect the outcome of tip inhibition is the geometry of the distribution of the cells surrounding the tip. The main finding in this regard was that long, thin cell masses in the slit wells formed slugs which were 2 · 5 – 3 times smaller than the slugs resulting from disc-shaped cell masses. It is possible that the cell mass shape affects size in this way because tip dominance is limited by the inhibitor’s range of diffusion. In these same experiments I found that inhibition is very strong close to the tip and that it weakens drastically at distances greater than 400 μm. Using an entirely different grafting method Durston (1976) and Durston & Vork (1977) made a significant observation which supports and complements the tip inhibition experiments reported here. In their experiments segments of slugs were grafted end to end and by this they showed that tip inhibition was strongest near the tip and progressively weaker in segments farther from it.

It is possible with the available information to examine whether or not simple diffusion is a mechanism by which inhibition is transmitted from the tip. By simple diffusion I mean that the paths of the inhibitor molecules are random walks from the tip and the motion of the individual molecules is not facilitated by active transport or relay on the part of the surrounding cells. It should be noted that pulsatile release and relay of the cAMP signal is a well established part of the aggregation process and it is important to see if what is now known about tip inhibition is also consistent with such a mechanism.

First of all, the gradient of inhibition in the slug, as pointed out by Durston (1976), and the distance dependence of inhibition seen in slit wells is consistent with the simple diffusion model. A relay mechanism however should not show such a pattern of decreasing intensity with distance. Then there is the tip grafting result which does not fit well with inhibition mediated by a pacemaker signal. In a pacemaker model inhibition strength should either not be affected by the size of the tip, or perhaps increase with decreasing size of the tip (Clark & Steck, 1979). Either pacemaker model is difficult to reconcile with the fact that large tips have greater inhibition power.

It has already been mentioned that the distance of inhibition is consistent with a simple diffusion mechanism. The time for setting up a tip inhibition gradient also supports this hypothesis. I found that a papilla of cells appeared on the prespore cell mass in about an hour from the time of isolation of the cells. Farnsworth (1973), however, made a more precise estimate of the time needed for a cell mass to become completely resistant to inhibition. In his experiments tips were removed from cell masses and the remaining cell mass bisected for various periods of time by insertion of an impermeable plastic barrier. He found that a barrier left in place for 34 min resulted in 50 % of the cell masses forming two tips and thus, on the average, it takes 34 min for a group of cells to become resistant to inhibition by a nearby rival tip. It might be argued that this is too short for diffusion mechanisms to be operating.

However, according to calculations by Crick (1970) a diffusion gradient of a small molecule could be set up over a distance of about 0 · 1cm in several hours assuming a reasonable diffusion coefficient. But the average distance of inhibition in slit wells is about half of that distance and since the time required varies with the length squared, the time necessary should be roughly one quarter, or less than one hour. On this basis, and on the results of the experiment which showed that the tip inhibition can act across an agar barrier, simple diffusion is supported as a mechanism by which the tip dominates surrounding cells.

Let me reiterate that the size of a slug is also significantly affected by the cell strain and the position of the cells in the cell mass. There are two points to be made. First of all, although tip strength in DdH varies with tip size, no differences were found in tip strength between strain DdH and the smaller strain P-4. This evidence is consistent with the hypothesis that the size of the slug cell mass is determined by two factors; one is the ability of a tip to dominate or inhibit cells from forming an independent, rival tip and the other is the ability of the surrounding cells to resist being subjected to the inhibition of the newly arisen tip. If the surrounding cells are easily inhibited, then the slugs produced will be large; if they are resistant to inhibition the slugs resulting will be correspondingly small.

At the present time the nature of the changes in prestalk and prespore cells and in the mutant strain causing these differences in resistance to inhibition are unclear. It is hoped that the experiments here will open the way to a gençtic approach to the problem of inhibition and resistance to inhibition.

This investigation was supported by National Institutes of Health Research Service Award T32GM07312, a Sigma Xi Grant-in-Aid of Research, the Whitehall Foundation and grants from the National Science Foundation and National Institutes of Health to Dr J. T. Bonner.

I would like to thank John T. Bonner and Edward C. Cox for comments which improved the manuscript and David Trevan for his help with the figures.