Induction by acid load of the maturation of prestalk cells in Dictyostelium discoideum

The fruiting body of the cellular slime mould Dictyostelium discoideum consists of a spherical mass of spores and a supporting stalk, the latter being composed of vacuolated cells with cell walls. The fruiting body is formed from a mass of amoeboid cells (the pseudoplasmodium, or ‘slug’) in which the precursor cell types of spores and stalk cells are discernible, namely prespore cells and prestalk cells (Bonner, 1967).

During the process of fruiting body construction, prestalk cells first turn into stalk cells at the apical region of the slug, and the stalk elongates by continuously adding new stalk cells to its tip (Bonner, 1952). In other words, prestalk cells become mature stalk cells in an orderly manner, suggesting the presence of a mechanism that controls stalk cell maturation.

Since the discovery that starved D. discoideum cells can be induced by cAMP to become mature stalk cells (Bonner, 1970), this nucleotide has been suggested by some workers to be the natural inducer of stalk cell differentiation. Ammonia, on the other hand, inhibits stalk cell differentiation (Gross et al. 1981, 1983) and also prevents fruiting body formation (Schindler & Sussman, 1977). In a model proposed by Sussman & Schindler (1978), antagonism between cAMP and NH₃ was supposed to play the key role in the control of stalk cell differentiation. The finding that ammonia inhibits the intracellular accumulation of cAMP (Schindler & Sussman, 1979) was taken as evidence for this hypothesis. However, the recent finding by Berks & Kay (1988) that cAMP inhibits some step in stalk cell differentiation threw doubt on its originally claimed role in stalk cell maturation.

Other than inhibiting the accumulation of cAMP, ammonia has the effect of elevating cytoplasmic pH (pH,, Inouye, 1988) and the pH of intracellular acidic vesicles (pH_v, Yamamoto & Takeuchi, 1983), suggesting the possibility that pH, and pH_v may play, in addition to their proposed role in the choice between spore and stalk pathways (Gross et al. 1983, 1988; Inouye, 1985), an important role in the control of stalk cell maturation. To examine this possibility, I investigated in this study the effects of acid load on migrating slugs of D. discoideum. Carbon dioxide and ammonia proved to be particularly useful for investigating the effects of weak acid and weak base, respectively, on intact migrating slugs, because they permeate efficiently into the tissue without the need of physically disturbing the slug structure. The effects of other weak acids and bases, as well as of protonpump inhibitors and DIF, were also investigated. The results obtained clearly indicate that cytoplasmic acidification induces the maturation of prestalk cells to become stalk cells. Based on these and other results, a possible mechanism for the control of stalk cell maturation during normal development will be discussed.

Dictyostelium discoideum, V12M2 and NC4 (wild-type strains), were used. They were grown on Klebsiella aerogenes using a modified SM medium (4·4 g KH₂PO₄, 2·0 g Na₂HPO₄, 1·0 g MgSO₄ · 7H₂O, 7·5 g glucose, 10-0g Bacto peptone (Difco), 1·0 g yeast extract (Difco), 15 g agar, 11 H₂O). Cells about to clear the bacterial lawn were collected and washed in 20 mm-potassium phosphate buffer, pH 6-0, (PB) by repeated centrifugation.

A thick suspension of washed vegetative cells (approx. 1-6×10⁸ cells ml⁻¹ in PB supplemented with 2mm-MgSO₄) was deposited on agar substrata (3 ml of 1·5 % bacto agar, Difco, USA, in H₂O) made in 50 ml tissue culture flasks (Becton Dickinson, USA). Each flask received a spot of 10 pl cell suspension in the centre. After the cells had settled on the agar substrata, excess fluid was removed. The flasks were then capped and incubated at 21°C in the dark. After approx. 24 h, when slugs had moved away from the original spot, the cap of each flask was replaced with a silicon double stopper, and the CO₂ concentration in the flask was adjusted in such a way that the O₂ concentration remained unchanged at 21 %. Usually, x % CO₂ (initial concentration) was prepared by withdrawing 18·5 ml air from the flask, followed by an addition of ( 15− 0·75x) ml air, 0-16x ml O₂ and 0-6x ml CO₂. The gasses had been saturated with H₂O. The air space of the flask with 3 ml agar and a double stopper was 60·0 ml. Plastic syringes were used for withdrawing and adding the gases. The withdrawal of air caused decreases in the pressure ( −0·25 atm) and in the air-space volume (− 4 · 5 ml) of the flasks, which were taken into consideration in calculating the volumes of the gasses to be added. Using the solubility of CO₂ in water at 21 °C (0 · 8541CO₂ in 11 H₂O at 1 atm CO₂), the actual CO₂ concentration was calculated to be 0-957 × [initial concen tration]. In the text, CO₂ concentrations are given as the initial concentrations. In some experiments, CO₂ was removed by perfusing approx. 5 litres air using an aspirator after the indicated periods of exposure.

Slugs (1-day-old) migrating on plain agar plates (4 ml of 1 · 5% water agar in 5 cm plastic dishes) were exposed to NH₃ by placing the plates (after removing the lids) in plastic boxes (3 · 061) containing 0 · 51 ammonia source solution. The boxes were then tightly sealed. The ammonia source was made by adding 0 – 7 ml of 1 M-NH₄C1 to 0·51 of 0·5 M-NaOH. The partial pressures of ammonia in the air space were calculated from the solubility constant of NH₃ at 21 °C (Denmead et al. 1982) and the concentrations of NH₃ in the ammonia source solutions. The latter were measured using an ammonia electrode (Horiba, Japan). After an exposure to NH₃ for the indicated times, the plates were transferred to a box containing 1·51 solution of neutralized bromothymol blue. Decrease in the NH₃ concentration of the ammonia source solution during this procedure was negligible. The same method was used for trimethylamine by replacing NH₄Cl with trimethylamine chloride.

Slugs treated with CO₂ or NH₃ were further incubated for 24–40h at 21 °C in the dark. Resultant structures formed were then observed using a dissecting microscope (Olympus, Japan). They were classified into several groups according to their morphology, as indicated in the figure legends, and the number of individuals of each group was scored. Whole-mount specimens and squashed specimens of representative samples of each group have been examined at higher magnifications using phase-contrast and Nomarski microscopy. Sections of fixed samples of some groups have also been examined. For visualizing the cell wall, specimens were immersed in 0·1% Calcofluor white ST (American Cyanamide, USA) solution in PB and observed using a fluorescence microscope (excitation 405nm, emission >470nm, Nikon, Japan; Harrington & Raper, 1968).

1-day-old slugs were transferred onto agar plates containing 20 mm-potassium phosphate buffer, at the indicated pH, and the substances to be tested. About half of the slugs were dissected lengthwise at a position about one fifth from the tip, and each fragment was dissociated separately with a hair loop. Then the slugs, both intact and dissected, were covered with coverslips. The plates were incubated at 21 °C in the dark. After approx. 24 h incubation, the plates were examined using a phase-contrast microscope. The coverslips were then carefully lifted off the agar surface (almost all the cells adhered to the coverslips), placed on a glass slide carrying a drop of 0·1% Calcofluor and observed as described above.

1-day-old slugs were transferred onto agar plates containing the indicated amounts of the test substances. The plates were then incubated in unidirectional light at 21 °C. Induc-tion of fruiting by each added substance was estimated by counting the numbers of fruiting bodies and migrating slugs 20 h later.

For neutral red staining, washed vegetative cells were incubated in 0·005% neutral red solution in PB for 10 min at 0°C. Tetramethylrhodamine staining and grafting of slugs were carried out as described previously (Akiyama & Inouye, 1987). The effects of CO₂ and NH₃ on the pH of the agar substratum were assessed by measuring the pH of a 20mm-KCl solution using a glass pH electrode under exactly the same conditions as in the experimentals.

DIF-1 (natural and synthetic, giving the same results) was kindly provided by Dr R. Kay (MRC, Cambridge, England). DES and miconazole were purchased from Sigma (St Louis, USA), and zearalenone from Mikor Chemicals (Jerusalem, Israel). CO₂ (over 99 9% purity) and other chemicals were obtained from local suppliers).

When migrating slugs were exposed to 10% CO₂, they immediately stopped migrating and all the prestalk cells, and the majority of prestalk-like cells in the posterior region (see below), became mature stalk cells. (In this paper, no distinction will be made between stalk cell and basal disk cell.) Meanwhile, the prespore cells of each slug underwent regulation to make one or more normal fruiting bodies, consisting of mature spores and stalk cells, within 24 h (Fig. 1A–D). The fates of both precursor cell types after a CO₂ exposure were revealed by prestalkspecific staining with neutral red (Bonner, 1952; Sternfeld & David, 1981); while many of the mature stalk cells formed within the original slug had large red granules, most cells in the stalk of the fruiting body structures did not (Fig. 2). Experiments using chimeric slugs with a rhodamine-labelled prestalk region also indicated that only a small fraction of the original prestalk cells contributed to the stalk of the fruiting body structures (data not shown). Time-lapse films of the entire process revealed that the cells in the posterior region resumed mobility within about 12 h and formed a fruiting body, while the anterior cells remained immobile throughout.

When CO₂ (10%) was removed after a 10-min exposure, no stalk cell mass was formed. However, a 20-min exposure to 10 % CO₂ was sufficient for stalk cell induction. In such a case, the anterior region of the slugs became a mass of mature stalk cells, just as in the case of a continuous exposure to CO₂. On the other hand, prespore cells, instead of making fruiting body structures, reorganized themselves to form a new slug which resumed migration. A similar situation occurred with some of the slugs that had been continuously exposed to 5−10% CO₂. At lower concentrations of CO₂ (1−5%), only the normal fruiting body formation was induced. When the CO₂ concentration exceeded 10%, some slugs failed to form the fruiting body structures, giving rise to slugshaped masses of stalk cells with some amoeboid cells hidden inside (Fig. 1E-F). When the CO₂ concentration was very high, some cells, especially those in the tip region, lysed. The effects of various concentrations of CO₂ on migrating slugs are summarized in Fig. 3.

An exposure to CO₂ causes a transient decrease in the pH_i of slug cells (Inouye, 1988), which raises the possibility that cytoplasmic acidification induces the maturation of prestalk cells to become true stalk cells. Supporting this is the observation that simultaneously added ammonia nullified the effects of CO₂ (data not shown). To examine further this possibility, migrating slugs were first exposed to NH₃ and then transferred into an atmosphere without NH₃. (pH₁ can also be decreased by removing weak base, Inouye, 1988). Meanwhile, the relative humidity was kept at 100%. Upon removal of NH₃, the slugs stopped migration. Subsequently, if the NH₃ concentration and the length of the exposure were appropriate, prestalk cells became mature stalk cells while prespore cells formed fruiting bodies, giving rise to exactly the same structures as those formed in 10% CO₂. Fig. 4 shows an example of the structure formed after a 10 min exposure to 0·012% NH₃. At higher concentrations of NH₃, many of the cells in the slug lysed, whereas at lower concentrations (0·002−0·005%), normal fruiting body formation was induced (Fig. 5).

If cytoplasmic acidification is the cause of stalk cell maturation, the effect of NH₃ removal is expected to vary depending on the length of the exposure to NH₃, because the degree of acid load by NH₃ removal should become larger if the length of the exposure is prolonged (Roos & Boron, 1981). Indeed, increasing the period of exposure to the same concentration of NH₃ resulted in correspondingly more pronounced effects on slugs (Fig. 6). If NH₃ was not removed but present continuously, all the slug cells lysed even at a concentration of 0·002 %.

Addition followed by removal of trimethylamine, another weak base, also induced the formation of fruiting bodies and the maturation of prestalk cells in situ. As in the case of NH₃ removal, the magnitude of its effects was dependent on the length of the exposure period (data not shown).

In the above experiments, unbuffered agar was used as substratum. If buffered agar had been used, far more CO₂, NH₃ or trimethylamine would have dissolved in the agar, resulting in a considerable increase in their concentrations (in ionized form) in the agar and a significant decrease in their concentration in the air. The choice of unbuffered agar to avoid such problems, however, unavoidably led to considerable changes in the substratum pH. For example, 10% CO₂ lowered the substratum pH from the original value of 5·7 to 4·8 within 10min, approaching the equilibrium value of 4·46. Upon exposure to 0·012 % NH₃, the substratum pH exceeded 10 within 6min. However, such changes in environmental pH cannot be the cause of stalk cell maturation. First of all, slugs transferred onto agar substrata of pH 4·5 or 4·0 (20 mm-citrate buffer) were only induced to fruit and no stalk cell maturation like that seen in 10 % CO₂ was observed. Second, after removal of NH₃, it took a very long time for the pH to return to its original value, and anyway there is no reason that the agar substratum should become acidic in this case. Third, CO₂-induced maturation of stalk cells occurred with nearly the same efficiency when buffered agar (20mm-phosphate, pH7·0) was used. In this case, even when 20% CO₂ (initial concentration) was applied, the substratum pH only decreased to pH6-5, still higher than the pH of unbuffered agar without CO₂ addition (pH5·7).

The results described above clearly indicate that cytoplasmic acidification induces the maturation of prestalk cells in intact slugs. To see whether other inducers of stalk cell differentiation under in vitro conditions, such as proton-pump inhibitors and DIF (Kay & Jermyn, 1983; Gross et al. 1983, 1988), can also induce the maturation of prestalk cells, their effects on slug cells were investigated using an agarsandwich method (see Materials and methods). These substances differ from each other, and from weak acids, in their effects on pH_i (Inouye, 1988). For the sake of comparison, cAMP and adenosine, which influence the cell differentiation of slug cells (Weijer & Durston, 1985), as well as some weak acids, were examined under the same experimental condition. Because the method described above cannot be used for these nonvolatile substances, they were included in the agar substrata.

To see the effects of weak acids, migrating slugs were transferred onto agar plates containing various concentrations of weak acids, and then covered with coverslips. Fig. 7 shows the effects of 4mm-propionate at pH 6·0. Apparently, only prestalk cells, and perhaps prestalk-like cells in the posterior region, were induced by propionate to become stalk cells. The first sign of vacuolation characteristic of maturing stalk cells was evident within a few hours. Under these conditions, cells in the anterior region showed no movement at all after exposure to propionate whereas the majority of the posterior cells resumed amoeboid movement and either piled up or moved away from their original positions by forming streams. Propionate at concentrations below 3mm was not effective, while at concentrations over 6mm most cells rounded up and no stalk cells were formed.

To verify further the origin of the stalk cells formed, the anterior region and the posterior region of the slug were separated by dissection, gently spread with a hair loop, and exposed to propionate in the same way as described above. Many of the anterior cells became stalk cells, whereas the majority of the posterior cells remained amoeboid without further differentiation (Fig. 8A-D). It was noticed that stalk cells are sometimes formed in contact with what appeared to be the sheath material, even under the control condition (data not shown). Yet the effects of weak acids, as well as the difference in inducibility between anterior and posterior cells, were clear.

Other weak acids, such as acetate (15− 20 mm), benzoate (1 mm), and p-phenylbenzoate (20− 50 μM) also induced stalk cell maturation at pH 6·0. The ability of weak acids to induce stalk cell differentiation was pH dependent; when the external pH was 7·0, no stalk cells were formed with the same concentrations of weak acids that effectively induced stalk cell differentiation at pH 6·0.

The effects of weak acids are summarized in Table 1 together with the effects of other inducers, which will be described in the following sections.

Some inhibitors of the plasma membrane H⁺-ATPase were examined for their effects on slugs. Diethylstil-boestrol (DES) not only induced stalk cell differentiation in the anterior region of the slug but, unlike weak acids, also converted many prespore cells into stalk cells (Fig. 8E-H). Miconazole, another inhibitor of the H⁺-ATPase, had similar effects, but zeara-lenone was much less effective.

Externally applied DIF (100– 3000 units ml⁻¹) turned most of the posterior cells, as well as the anterior cells, into mature stalk cells (Table 1). Although some amoeboid cells were often found among large masses of vacuolated cells, virtually all isolated cells became stalk cells.

For the sake of comparison, cAMP and adenosine were examined for their influence on stalk differentiation under the same experimental condition. At 10 μm, cAMP converted almost all anterior and posterior cells into stalk cells, though some anterior cells remained amoeboid for at least 2 days (Table 1, see also Town & Stanford, 1977). On the other band, adenosine (1-5 ITIM) did not induce fruiting or stalk cell differentiation.

Formation of mature stalk cells in agar sandwiches, irrespective of the inducer, was inhibited by 10 nw-NH4CI if the external pH was 7·5 but not at pH 6·0.

During normal development, the maturation of prestalk cells at the apical region of the slug is the initial event in the process of fruiting body construction. The effects of various agents on the probability of the transition from the migrating phase to the fruiting phase can therefore be a measure of their activity in situ in inducing stalk cell maturation. Slugs were transferred onto agar substrata containing each test substance but left uncovered, and observed 24 h later.

Slugs transferred onto agar substrata containing 4 mm-propionate at pH 6-0 often formed fruiting bodies without further migration. Induction of fruiting by propionate has already been reported (Sussman et al. 1978). Those slugs that resumed migration, on the other hand, dropped numerous cells behind during migration. Such cells, also seen under normal conditions (Bonner, 1967), are trapped in a tube of slime sheath and differentiate into stalk-like cells within several hours (data not shown). At higher concentrations, some slugs formed structures similar to those shown in Figs 1 and 4. Such structures were consistently observed on agar plates containing 2– 3mm-benzoate at pH 6·0.

Freely migrating slugs were not as efficiently induced to fruit by DES as they were by weak acids (Table 1). Slugs migrating on DES agar also dropped a large number of cells behind, but unlike the result with weak acids, they mostly remained amoeboid. Migrating slugs were not induced to fruit by DIF within the concentration range (100– 3000 units ml⁻¹) where most of slug cells became stalk cells under the agarsandwich condition (Table 1). Some induction of fruiting was obvious only at 30 000 units ml⁻¹, and it was only at these concentrations, too, that the number of dropped cells increased significantly. These cells turned into mature stalk cells within several hours (data not shown).

Slugs freely migrating on agar substratum containing 10 μm-CAMP left a large number of cells behind (data not shown). Higher concentrations of cAMP caused severe aberration of slug structure (data not shown; see also Nestle & Sussman, 1972; Chia, 1975; George, 1977; Feit et al. 1978). On the other hand, adenosine did not induce fruiting or the cell-dropping effect.

The data presented thus far are for a wild-type strain V12M2. Most experiments were also done in exactly the same way using another wild-type strain NC4. The results were essentially the same as those described above, but it should be noted that in general NC4 cells were less efficiently induced to differentiate into stalk cells when the agar-sandwich method was used.

It was shown in this study that prestalk cells were induced to mature when they were exposed to weak acids, or when externally applied weak bases were removed. From this, I conclude that cytoplasmic acidification is the primary event during the process of stalk cell maturation. This conclusion is based on the following observations. (1) Exposure of cells to weak acids and removal of weak bases both induce a transient decrease in pH_i, reaching its minimum value within 3 min (Inouye, 1988). (2) The extent of the pH, decrease induced by CO₂ addition and NH₃ removal depends on the concentrations of CO₂ and NH₃, respectively. Also the pH_i decrease induced by NH₃ removal becomes larger as the length of the exposure time is prolonged (Roos & Boron, 1981). The effects of CO₂ addition and NH₃ removal on migrating slugs varied correspondingly and agreed perfectly with the effects having been mediated by the pH, changes induced by them (see, Figs 3, 5, 6). (3) NH₃ itself elevates the pH_i of prestalk cells (Inouye, 1988). Correspondingly, NH₃ inhibits the CO₂-induced maturation of prestalk cells. It also prevents stalk cell maturation during culmination (Schindler & Sussman, 1977). (4) The ability of weak acids to lower pH_i, and their ability to induce stalk cell maturation are both dependent on the external pH, being more effective at lower external pHs.

Williams et al. (1984) have shown that weak acids prevent the secretion of cAMP whereas ammonia suppresses its production, thus having opposite effects on the intracellular level of cAMP. It therefore seems possible that the effects of weak acids and weak bases may be mediated by changes not in pH_i but in the intracellular level of cAMP, which may play the key role in the control of stalk cell differentiation. Recent observations that 8-bromo cAMP, a membrane-permeable derivative of cAMP, induces the maturation of prestalk cells (Kwong et al. 1988; Maeda, 1988) seems to support this possibility. However, it has not been demonstrated that NH₃ removal, which was also shown to induce stalk cell maturation, elevates the intracellular level of cAMP. In any case, the rapid changes in pH, induced by weak acids and weak bases, being consequences of simple physicochemical reactions, are likely to precede changes in cAMP-release and -production rates and, in fact, it is quite possible that the former are the cause of the latter.

Other weak acids and substances known to induce stalk cell differentiation, such as diethylstilboestrol (DES), differentiation inducing factor (DIF) and cAMP, were also examined for their effects on migrating slugs (Table 1). To analyse the mode of action of these substances, it is convenient to divide the process of stalk cell differentiation into two steps; the differentiation of aggregated cells into prestalk cells (step 1) and the maturation of prestalk cells to become true stalk cells (step 2).

It is evident from the above results that weak acids induce step 2. Induction of fruiting by weak acids or by removal of weak bases can be interpreted as a consequence of relatively weak cytoplasmic acidification; the process of fruiting body construction begins with the maturation of prestalk cells in the tip region followed by successive maturation of the adjacent prestalk cells (Raper & Fennell, 1952). It is conceivable that a relatively mild acid load would trigger maturation only at the tip region (probably the most susceptible region within the slug to environmental changes) and, in this way, initiate the sequence of events. Stronger acid loads, on the other hand, would overcome the intrinsic control of stalk cell maturation (see below) and induce step 2 simultaneously in all the prestalk cells. The stalk cells formed in the posterior region of the slug probably originated from ‘anterior-like cells’ (Stemfeld & David, 1981) and ‘rear-guard cells’ (Bonner, 1957, 1967). These cell types are almost indistinguishable from prestalk cells, except that they are located in the posterior region (Bonner, 1967; Devine & Loomis, 1985; Kakutani & Takeuchi, 1986) and are prevented from maturing by the existence of real prestalk cells (Sternfeld & David, 1982). Conceivably, these cells became mature stalk cells after prestalk cells in the same slug ceased to emit the hypothetical maturationinhibiting signal by becoming true stalk cells (see Stemfeld & David, 1982).

DES, DIF and cAMP were different from weak acids in that they not only caused the maturation of prestalk cells but also induced the transdifferentiation of prespore cells to become mature stalk cells. DIF has been shown to induce step 1 at the transcriptional level (Williams et al. 1987) and also to cause transdifferentiation from prespore to prestalk (Kay & Jermyn, 1983). DES has also been shown to induce step 1 (Kay et al. 1986). However, DIF and DES, by themselves, do not seem sufficiently active to induce step 2 in intact slugs. Although not conclusive, this view seems to be favoured by the following argument which is based on the idea described above that the inducer of step 2 should also induce fruiting. DIF did not efficiently induce fruiting, nor did it cause the maturation of prestalk cells in situ within migrating slugs, even at concentrations almost two orders of magnitude higher than the effective concentration for inducing stalk cell maturation under the agarsandwich condition. In the case of DES and other proton-pump inhibitors, fruiting and stalk cell induction are both induced at about the same concentration. This is in contrast with the case of weak acids where slugs were efficiently induced to fruit by lower concentrations of weak acids than those needed for in situ maturation of prestalk cells. These observations suggest that, whereas weak acids induce step 2 in intact migrating slugs, DIF and DES do not. This may appear to contradict the fact that a large number of mature stalk cells are formed in cell monolayers in agar sandwiches, or in submerged cultures, when DES or DIF are present. However, these results do not necessarily mean that they induce step 2. The agar-sandwich method used in this study, and the submerged conditions used so far to examine the effects of DIF and the other inducers, could require the inducer of step 1 for maintaining prestalk cells but could allow them to mature without the need for the inducer of step 2. Conceivably, the action of DIF or DES is counteracted in intact migrating slugs by some factor(s), which are diluted away under the agarsandwich and submerged conditions. Recently, Sobolewski & Weeks (1988) showed that the transition from prestalk cells to stalk cells is restrained in the slug until the culmination stage but unrestrained in vitro, supporting the above interpretation.

The effects of cAMP, under the agar-sandwich condition, could at least partly be interpreted as a consequence of the stimulation of DIF production by cAMP (Brookman et al. 1982). In this connection, it may be worth noting that prespore cells were more efficiently induced by cAMP to become stalk cells than prestalk cells (see Results). This could be explained if prespore cells are more efficient DIF producers. In fact, the prespore region of the slug has been shown to contain at least twice as much DIF as the prestalk region (Brookman et al. 1987), and this difference might reflect a difference between the two cell types in rate of DIF production. The effects of cAMP on freely migrating slugs will be further complicated by disturbance of the chemotactic signalling system, which is thought to be operating also in the slug.

It will be interesting to compare the differences between weak acids, DES and DIF with the difference in their influences on the pH₁, of slug cells; whereas weak acids, as noted above, elicit an abrupt decrease in pH_i, which is followed by a gradual recovery, DES induces a slow and steady decline of pH_i. DIF, on the other hand, caused only a slight and transient drop of pH_i (Inouye, 1988). It therefore seems possible that these pH₁, changes induced by weak acids, DES and DIF would be reflected, respectively, in the strong, weak and negligible induction of step 2. This, together with the fact that ammonia inhibits both the pH_i decrease and stalk cell maturation, suggests that for prestalk cells to mature their pH_i is required to drop sufficiently. However, it remains to be elucidated whether the pH_i decrease directly induces the maturation process or whether its effect is further mediated by other factors, such as a decrease in pH_v as has been proposed to be the case in prestalk differentiation (Gross et al. 1988) or an increase in the intracellular level of cAMP as mentioned above (Williams et al. 1984). These possibilities are not mutually exclusive. If pH_v proved to be decisive, a small pH_v drop would induce differentiation into prestalk cells (step 1) while a larger pH_v drop would be required for their maturation (step 2).

Based on the above considerations, the following hypothesis can be offered about the construction of the fruiting body (Sobolewski & Weeks, 1988 give a similar discussion). While DIF induces stalk cell differentiation, it is somehow blocked at step 2 during slug migration. Ammonia is one of the candidates for the blocker, because it is continuously produced by cells (Gregg et al. 1954; Schindler & Sussman, 1977; Feit & Sollitto, 1987), and because exogenously applied ammonia blocks step 2. Alternatively, the cAMP signalling system may be responsible for the inhibition of DIF action, as cAMP has recently been shown to interfere with DIF induction of stalk cell differentiation (Berks & Kay, 1988) and also to raise the pH₁, of aggregative cells (Aerts et al. 1987). If the former is the case, loss of NH₃ by evaporation from the slug tip would relieve the block of step 2 and initiate the maturation of prestalk cells. This is similar to the model proposed by Sussman & Schindler (1978), but, in their model, ammonia inhibits the accumulation of cAMP which is postulated to be the inducer of stalk cell maturation.

However, there is some doubt as to whether preferential loss of ammonia by evaporation from the tip would be sufficient for restricting the maturation of prestalk cells within the limited region of the growing end of the stalk. The finding of this study that weak acids induce stalk cell maturation raises another possibility which is worthy of consideration. It has been shown that stalk cells in the process of maturation degrade proteins and RNAs at a high rate releasing a significant amount of ammonia (Gregg et al. 1954; Wilson & Rutherford, 1978). While ammonia (with its very high permeability through the cell membrane) will diffuse quickly through the tissue, other degradation products such as organic acids will diffuse relatively slowly. Therefore, it is reasonable to assume that some organic acid, which are weak acids, accumulate near the apical end of the growing stalk, causing cytoplasmic acidification (and consequently the maturation of nearby prestalk cells), while fast-diffusing ammonia would suppress acidification in the rest of the prestalk region. This type of regulation mechanism based on a short-range activation signal and long-range inhibition has been thought to be important in a variety of patternforming systems (Turing, 1952; for a review, see Meinhardt, 1982). In order to know whether such a mechanism is at work during the construction of the fruiting body in the cellular slime moulds, further biochemical and cytological studies will be necessary.

I wish to thank Dr Julian Gross for discussion and for suggestions for improvement of the manuscript, Dr Michael Hanna for comments on an earlier draft of the manuscript and Dr Rob Kay for suggesting the ammonia-removal experiments and for kindly providing DIF.