Ammonia promotes accumulation of intracellular cAMP in differentiating amoebae of Dictyostelium discoideum

Mechanisms that regulate differentiation of Dictyo stelium discoideum amoebae during multicellular devel opment can be conveniently studied under in vitro culture conditions that permit differentiation of single cells. Monolayers of wild-type V12M2 cells differen tiate as stalk cells if supplied with cAMP, and mono layers of sporogenous derivatives of V12M2 form both stalk cells and spores (Town et al. 1976; Kay et al. 1978; Kay, 1982). The dependence of stalk cell formation on cell density led to the identification of another essential morphogen, differentiation inducing factor (DIF) (Town et al. 1976; Morris et al. 1987). It is now widely held that the choice between stalk cell and spore differentiation is regulated in part by levels of cAMP and DIF. Extracellular cAMP is required to initiate development but inhibits terminal stalk cell differentiation (Sobolewski et al. 1983; Berks and Kay, 1988). Developing cells degrade exogenous cAMP and secrete DIF, which antagonizes extracellular cAMP late in development (Wang et al. 1986). DIF is essential for the completion of stalk cell differentiation and induces several genes that are expressed only in prestalk and stalk cells (Kopachik et al. 1983; Jermyn et al. 1987). In contrast, maintaining high levels of extracellular cAMP and inhibiting DIF accumulation favors spore differentiation (Ishida, 1980; Riley and Barclay, 1986; Berks and Kay, 1988).

Extracellular cAMP functions by binding to cell surface receptors that are analogous and homologous to mammalian G-protein regulated hormone receptors (Gomer et al. 1986; Oyama and Blumberg, 1986; Haribabu and Dottin, 1986; Klein et al. 1988). Binding causes rapid accumulation of several intracellular second messengers including Ca²⁺ ions, inositol tri phosphate, cAMP and cGMP (Newell et al. 1987), any of which could regulate the choice between stalk cell and spore differentiation.

In previous studies, we used caffeine and progester one, which inhibit accumulation of intracellular cAMP by different mechanisms (Brenner and Thoms, 1984; Klein and Bracket, 1975), to investigate the role of intracellular cAMP in differentiation. Both drugs prevent spore and increase stalk cell formation during in vitro differentiation of sporogenous mutants (Riley and Barclay, 1986). These results are not due to pleiotropic effects on other second messenger systems because simultaneous exposure to 8-Br-cAMP, a membrane-permeable cAMP analog with little affinity for the cell surface cAMP receptor (Van Haastert and Kein, 1983), completely restores spore formation to caffeine- and progesterone-treated cultures (Riley et al. 1989). In the absence of caffeine and progesterone, conditions that increase rates of endogenous cAMP synthesis and accumulation increase spore and decrease stalk cell formation during standard development on agar as well as during in vitro differentiation (Riley et al. 1989). From these data, we proposed that intracellular cAMP levels regulate cell fate, with high levels promoting spore and/or inhibiting stalk cell differentiation. This conclusion was supported in a separate series of exper iments (Kay, 1989).

Previous studies showed that high ambient pH and ammonia promote spore formation during in vitro differentiation of sporogenous mutants (Gross et al. 1981; Gross et al. 1983). Do these environmental factors affect cell fate by promoting intracellular cAMP ac cumulation or is another mechanism involved? To address this, we examined the effects of high pH and ammonia on cell fate and accumulation of intracellular cAMP in cultures of wild-type strain V12M2 and a sporogenous derivative, HB200. Our results support the hypothesis that high intracellular cAMP levels promote spore and/or inhibit stalk cell formation in cultures exposed to ammonia or high pH.

In all experiments, we used wild-type D. discoideum strain V12M2 or a spontaneous sporogenous derivative, HB200. Amoebae were grown as previously described (Riley et al. 1989). For in vitro differentiation, washed amoebae were distributed on 6 cm tissues culture dishes at 5×10⁴ cells cm⁻² and submerged in 2.5ml of KM (10mM KC1, 5mM MgCh, 200qgml⁻¹ streptomycin sulfate, and either 10mM MES, pH 6.2 or 10 mM Hepes, pH 7.5) containing ImM cAMP. Under these conditions, cells terminally differentiate within 12–24 h as highly vacuolated stalk cells or phase-bright spores. This occurs without formation of large cell aggregates, although loose clumps of 10–20 cells often form by 6h. For submerged aggregation, cells were distributed at 2·5×10⁵ cellscm⁻² and submerged in KM without exogenous cAMP to promote aggregation. In this case, cells form large cohesive aggregates (up to 1000 cells), but roughly half of these cells fail to complete differentiation. The remainder differentiate asynchronously as stalk cells or spores after 2–4 days. Despite these differences, early development (through 8h) follows the same time course during in vitro differentiation and submerged aggregation as judged by accumulation of cellular phosphodiesterase activity: Levels peak at 6h and decline by 8h (Riley, unpublished data).

Intracellular and extracellular cAMP levels were measured by radio immune assay as previously described (Riley et al. 1989).

Phosphodiesterase assays were performed according to Boudreau and Drummond (1975), with minor modifications noted in the legend for Fig. 5.

We repeated and extended earlier studies (Gross et al. 1981; Gross et al. 1983) showing that increasing extra cellular pH or ammonia levels during in vitro differentiation increases spore formation and decreases stalk cell formation in sporogenous mutants. Fig. 1 shows that, in the absence of added NH₄C1, nearly two times more HB200 cells formed spores at pH 7.5 than at pH 6.2. The effects of exogenous NH4CI on cell fate varied with ambient pH. The ratio of NH₃:NH4⁺ is 20 times higher at pH 7.5 than at pH 6.2 and 15–20 times more NH₄CI had to be added at pH 6.2 to optimize spore formation (Fig. 1). These data indicate that NH₃ rather than NH₄⁺ is the active species in promoting spore formation. The increase in HB200 spore pro duction in response to ammonia or high pH was highly reproducible and similar in magnitude to that observed previously with other sporogenous strains (Gross et al. 1981; Gross et al. 1983). Spore formation increased to as much as 80 % when cAMP hydrolysis and accumulation of DIF (a prestalk morphogen) were inhibited by plating cells at lower cell densities (not shown).

We were puzzled that continuous exposure to NH₄Cl concentrations exceeding ImM at pH7.5 or 15mM at pH6.2 reduced both stalk cell and spore formation. However, high ammonia levels delay early develop-ment (see later) which could allow cells to prematurely degrade exogenous cAMP. This could inhibit spore formation because terminal differentiation of spores is sensitive to extracellular cAMP concentrations (Ishida, 1980; Riley and Barclay, 1986; Berks and Kay, 1988). Fig. 2A shows that adding NH₄CI (up to 20mM) to cultures at pH 7.5 after 10 h of differentiation bypassed the early delay and promoted spore formation. Further more, continuous exposure to 20mM NH₄CI at pH 7.5 promoted spore differentiation if extracellular cAMP concentrations were increased (Fig. 2B). Thus, high ammonia levels promote spore formation if the devel opmental delay associated with ammonia is bypassed or if high exogenous cAMP levels are maintained.

Raising the pH of the media could affect cell fate directly by a pH sensitive mechanism. Alternatively, high pH might act indirectly by increasing the ratio of NH₃:NH₄⁺ secreted into the medium. From published rates of ammonia secretion (Schindler and Sussman, 1977; Aeckerie et al. 1985), average concentrations of NH₃+NH₄⁺ could approach 100pM during the course of in vitro differentiation, and concentrations within loose cell clumps could be much higher. Such ammonia levels are probably too low to affect cell fate at pH 6.2 but might be adequate at pH 7.5 because the NH₃:NH₄⁺ ratio is 20 times higher. If pH acts only to raise the NH₃:NH₄⁺ ratio, then raising pH should not affect cell fate if secreted ammonia is completely removed.

We removed ammonia enzymatically by adding a cocktail containing glutamate dehydrogenase, alpha ketoglutarate and NADPH (Schindler and Sussman, 1977). To prevent exhaustion of the cocktail, and because the spore-promoting activity of ammonia is strongest after 10 h of in vitro differentiation (see below), the effect of enzymatic ammonia depletion was tested at this time (Fig. 3). Increasing the pH from 6.2 to 7.5 nearly doubled spore formation in control cul tures that received no cocktail. Cultures at pH7.5 that received incomplete cocktails (missing one or more reagents) formed the same number of spores as cultures at pH 7.5 that received no cocktail. In contrast, cultures at pH 7.5 that received complete cocktail formed the same number of spores as cultures at pH 6.2 that received either complete cocktail or no cocktail at all. These results were not due to greater accumulation of enzymatic byproduct (glutamate) at pH 7.5 since 10 mM glutamate by itself had no effect on cell fate (not shown). Thus, high pH alone cannot enhance spore formation because enzymatic removal of ammonia abolished the spore-promoting activity of high pH. This supports the hypothesis that high pH increases spore formation by increasing the NH₃:NH₄⁺ ratio.

To determine when ammonia or high pH are most effective in promoting spore formation, HB200 cells were allowed to differentiate in vitro for 10 h under one set of conditions (defined by pH and NH₄CI levels) after which the media were changed and cells were allowed to complete differentiation under another set of con ditions. 10 h corresponds to a relatively late stage of development because HB200 cells differentiate quite rapidly. Transcript levels of the prespore-specific gene D19, which encodes a cell surface glycoprotein of unknown function (Early et al. 1988), are maximal by 8 h and are almost undetectable by 12 h, the time when the first mature spores appear (not shown). Adding 15 DIM NH4CI (pH 6.2) or increasing pH from 6.2 to 7.5 after 10 h increased spore formation to the same extent as continuous exposure to high pH or ammonia (Table 1, compare line 1 with lines 2, 3, 5, and 6). In contrast, spore formation did not increase over the control if cells were exposed for the first 10 h of differentiation to either pH 7.5 or to 15 mM NH4CI at pH 6.2 and then shifted back to the control conditions of pH 6.2 with no exogenous NH4CI (Table 1, compare line 1 with lines 4 and 7). Thus, exposure to high pH or ammonia after 10 h of differentiation is completely effective in promoting spore formation, but early ex posure is neither necessary nor sufficient for optimal spore formation.

Because previous work showed that spore formation correlates with elevated intracellular cAMP levels (Kay, 1989; Riley et al. 1989), we performed similar experiments with drugs that raise or lower intracellular cAMP levels (Table 1). 8-Br-cAMP is a membrane-permeable cAMP analogue that has high affinity for intracellular targets of cAMP, such as the regulatory subunit of cAMP-dependent protein kinase, but has very low affinity for the cell surface cAMP receptor (DeWit et al. 1984; Van Haastert and Kein, 1983). This allows 8-Br-cAMP to enter cells and mimic endogenous cAMP without stimulating other second messenger systems via the cell surface receptor. Adding 1 mM 8-Br-cAMP after 10 h was as effective in promoting spore formation as continuous exposure to 8-Br-cAMP (Table 1, lines 8 and 9). However, spore formation did not increase over the control if 8-Br-cAMP was added for the first 10 h and then removed (compare lines 1 and 10). Thus, the effects of adding or removing 8-Br-cAMP after 10 h were the same as adding or removing NH₃, suggesting that these agents might operate by a common mechanism.

In contrast, treating cells with caffeine or progester one, drugs that lower intracellular cAMP levels (Bren ner and Thoms, 1984; Klein and Brachet, 1975; Riley et al. 1989) gave quite different results. Adding 5 HIM caffeine or 20μM progesterone after 10 h completely inhibited spore differentiation and promoted stalk cell differentiation (Table 1, lines 11 and 14). Moreover, the effects of caffeine and progesterone on cell fate per sisted even after the drugs were removed. Spore forma tion was inhibited and nearly all cells differentiated as stalk cells when caffeine or progesterone were added for the first 10 h and then removed (Table 1, lines 13 and 16). These results are not due to retention of caffeine or progesterone inside cells after washing because the effects of these drugs on cAMP synthesis and aggre gation are rapidly reversible (Brenner and Thoms, 1984; Klein and Brachet, 1975). Instead, our results may reflect earlier observations that prestalk differentiation is only slowly reversible (Raper, 1940; Bonner, 1949).

In summary, these data show that (1) drugs that increase (8-Br-cAMP) or decrease (caffeine and pro gesterone) intracellular cAMP levels effectively alter cell fate when added late in development, (2) adding caffeine or progesterone early in development induces stable changes in cell fate while raising cAMP levels early has no lasting effects, and (3) high pH and ammonia promote spore formation by a mechanism that is consistent with elevation, but not reduction, of intracellular cAMP levels.

To establish whether a link exists between extracellular NH₃ and intracellular cAMP, we determined the effects of high pH and ammonia on cAMP accumulation during ‘submerged aggregation’, culture conditions that permit reliable measurements of both intracellular and secreted cAMP levels (see Methods to distinguish from ‘in vitro differentiation’). Under these conditions, cells in control cultures began streaming by 4.5 h and com pleted aggregation by 6h. Adding 15mM NH₄C1 to cultures at pH 6.2 or raising the pH from 6.2 to 7.5 delayed the onset of aggregation by 1 h. Adding 15 mM NH4CI to cultures at pH7.5 delayed aggregation by 3–4h. However, once initiated, aggregation proceeded normally in cultures exposed to ammonia and/or high pH (not shown).

Fig. 4A shows intracellular cAMP levels during sub merged aggregation of HB200 cells. Cells cultured at pH 6.2 and pH7.5 had identical levels of intracellular cAMP at 6h of development. However, cAMP ac cumulated more quickly and was maintained at maxi mal levels through 8h of development at pH7.5, while cAMP levels declined sharply by 8 h in control cultures.

Still more striking was the finding that, in cells cultured with 15 mM NH4CI at pH 6.2 or pH 7.5, intracellular cAMP were nearly maximal at 2h. These levels equaled or exceeded the 6h peak in control cells and were maintained through 8h of development. Identical re sults were obtained with V12M2 cells (Fig. 4B). Thus, increasing ammonia levels, either by increasing the pH or by direct addition of NH4CI, stimulated precocious accumulation and prolonged maintenance of high intra cellular cAMP levels in both wild-type and sporogenous amoebae. This could explain the role of NH₃ as a prespore morphogen because elevating intracellular cAMP levels is sufficient to promote spore formation (Kay, 1989; Riley et al. 1989).

Ammonia might promote accumulation of intracellu lar cAMP by altering rates of cAMP synthesis, degra dation or secretion. However, we were unable to detect any effects of ammonia on specific activities of cellular phosphodiesterase (PDE) or adenylate cyclase activi ties in cells developed for 2h, the time when ammonia-treated cells first achieved maximal intracellular cAMP levels. In fact, enzyme activities in cells developed for 2h (with or without ammonia) were not significantly higher than in vegetative cells (not shown). In contrast, ammonia did affect the amount of cAMP secreted into the medium. Cultures exposed to 15 mM NH4CI at pH7.5 accumulated only 1/10 as much cAMP in the extracellular medium as control cultures at pH6.2 (Fig. 4C). Intermediate levels of extracellular cAMP accumulated in cultures at pH 7.5 with no exogenous NH4CI and at pH6.2 with 15mM NH4CI (Fig. 4C). Decreased levels of extracellular cAMP accumulation were not due to increased rates of cAMP hydrolysis because secreted PDE activities were the same in ammonia-treated and control cultures (Fig. 5). These data indicate that ammonia raises intracellular cAMP levels by inhibiting cAMP secretion. This mechanism explains the apparent discrepancy between our results and those of previous studies (see Discussion) and could also explain why ammonia delays aggregation by several hours.

HB200 amoebae terminally differentiate as stalk cells when caffeine or progesterone are used to lower intra cellular cAMP levels (Table 1). However, simultaneous exposure to 8-Br-cAMP completely reverses inhibition of spore formation in cultures exposed to caffeine or progesterone (Riley et al. 1989). This change in cell fate is thought to result from diffusion of 8-Br-cAMP into cells where it binds to intracellular targets of cAMP, such as cAMP-dependent protein kinase, for which it has high affinity (DeWit et al. 1984). Table 2 shows that ammonia also restores spore formation to caffeine- and progesterone-treated cultures. Spore formation was inhibited and nearly all cells differentiated as stalk cells when cultured with 2.5 mM caffeine or 10 μM progester one, but spore inhibition was completely reversed if 15 mM NH4CI was also added. It is possible that ammonia bypasses the effects of caffeine and progester-one by inhibiting cAMP secretion and thereby raising intracellular cAMP levels. (Basal rates of cAMP syn thesis persist in the presence of caffeine and progester one.) Alternatively, ammonia might directly activate targets of intracellular cAMP, such as cAMP-depen dent protein kinase, or another second messenger system that is dominant over changes in intracellular cAMP.

We cannot test the effects of ammonia on intracellu lar cAMP accumulation during in vitro differentiation because high levels of exogenous cAMP (1 mM) pre clude reliable measurement of intracellular cAMP levels. However, ammonia reversed the effects of caffeine and progesterone on intracellular cAMP ac cumulation during submerged aggregation. Fig. 6 shows that 2.5 mM caffeine and 10;IM progesterone reduced internal cAMP levels unless 15 mM NH4CI was also present, in which case normal cAMP levels ac cumulated. These data support the hypothesis that ammonia reversed the effects of caffeine and progester one on cell fate by restoring normal intracellular cAMP levels.

In agreement with studies of other sporogenous strains (Gross et al. 1981; Gross et al. 1983), we have shown that high concentrations of NH4CI or high pH enhance spore formation by HB200 cells during in vitro differentiation. While others have speculated that both agents exert this effect by a common mechanism (Sussman, 1982), we provide the first direct evidence that exogen ous NH4CI and high pH function by increasing the concentration of NH₃ in the medium. High pH has no separate role. In addition, we have shown that concen trations of ammonia that promote spore formation lead to precocious accumulation and maintenance of high intracellular cAMP levels in HB200 and V12M2 cells. This is probably the mechanism by which ammonia promotes spore formation because other conditions that raise intracellular cAMP levels also promote spore formation (Kay, 1989; Riley et al. 1989).

Ammonia raises intracellular cAMP levels by inhibit ing cAMP secretion. This explains how ammonia per mits cells with low adenylate cyclase activities (early in development or in the presence of caffeine) to accumu late or maintain high intracellular cAMP levels (Figs 4 and 6). Inhibition of cAMP secretion also explains the ability of ammonia to antagonize the effects of pro gesterone (Table 2 and Fig. 6), a drug that normally stimulates cAMP secretion (Klein and Brachet, 1975). The mechanism of cAMP secretion in Dictyostelium is unknown, but differential regulation of this function might be important during development because pre stalk cells of migrating slugs secrete much more cAMP than do prespore cells (Bonner and Slifkin, 1949). This raises the interesting possibility that ammonia induces differentiation of prespore cells in migrating slugs by inhibiting cAMP secretion and thereby raising intra cellular cAMP levels.

Stimulation of intracellular cAMP accumulation by ammonia would not have been expected based on earlier studies showing that 15 mM NH₄C1 at pH7.2 inhibits cAMP synthesis by aggregation-competent cells of wild-type strain NC4 (Schindler and Sussman, 1979; Williams et al. 1984). However, the previous studies examined the short term effects of NH4CI on cAMP synthesis by aggregating cells, whereas we looked at accumulation, not the rate of synthesis, in cells cultured from the time of starvation to 8 h of development. As we have shown, ammonia can increase accumulation of intracellular cAMP even when adenylate cyclase ac tivity is low. Moreover, it is possible that adenylate cyclase of Dictyostelium is only transiently sensitive to changes in pH or ammonia concentrations and that, after a period of adjustment, high rates of synthesis return. Indeed, Khachatrian et al. (1987) showed that preincubating Dictyostelium membranes with very high levels of NH4SO4 (>100 mM) increases activation of adenylate cyclase by inhibiting an inhibitory G protein.

Low rates of cAMP degradation could also play a part in maintaining high cAMP levels in ammonia-treated cells. We have recently shown that expression of the Dictyostelium cAMP phosphodiesterase gene and accumulation of enzyme activity are negatively regu lated by high intracellular cAMP levels: treatment with ammonia or 8-Br-cAMP reduces cellular PDE activity by nearly half (B. B. Riley and S. L. Barclay, unpub lished data). This suggests that, as ammonia raises intracellular cAMP levels, reduction of cellular PDE activity helps to maintain high cAMP levels.

These results and other recent studies suggest that intracellular cAMP regulates Dictyostelium develop ment, perhaps by more than one mechanism. Over expression of the regulatory subunit of cAMP-depen dent protein kinase disrupts development at a stage prior to aggregation (Simon et al. 1989; Firtel and Chapman, 1990). This block presumably results from low kinase activity. Another cAMP-binding protein, CABP1, may function to transduce the cAMP signal from the cell membrane into the nucleus (Kay et al. 1987) and has been implicated in regulation of the rate of development (Tsang et al. 1987). It is not yet clear what cell functions these proteins regulate or how they affect cell fate. However, this study and others (Riley and Barclay, 1986; Riley et al. 1989; Kay, 1989) clearly show that cell fate correlates with intracellular cAMP levels.

Genes that respond to changes in intracellular cAMP levels should be useful in elucidating the mechanisms by which cAMP affects development. Use of drugs like ammonia, 8-Br-cAMP, caffeine, and progesterone to alter intracellular cAMP levels during in vitro differentiation could be a powerful system for detecting such genes and studying their expression.

This research was supported by NIH grant GM35432 awarded to S. L. Barclay.