Article

• The EMBO Journal (1998) 17, 2838 - 2845
• doi:10.1093/emboj/17.10.2838

An intersection of the cAMP/PKA and two-component signal transduction systems in Dictyostelium

Peter A. Thomason^1,5, David Traynor^1,5, Guy Cavet², Wen-Tsan Chang³, Adrian J. Harwood⁴ and Robert R. Kay¹

1. MRC Laboratory of Molecular Biology, Hills Road, Cambridge CB2 2QH, UK
2. Present address: Biochemistry Department, B407, Beckman Center, Stanford University Medical Center, Stanford, CA 94305, USA
3. Department of Biochemistry, University of Oxford, South Parks Road, Oxford OX1 3QU, UK
4. Present address: MRC Laboratory for Molecular Cell Biology, University College London, Gower St, London WC1E 6BT, UK
5. P.A.Thomason and D.Traynor contributed equally to this work

Correspondence to:

Robert R. Kay, E-mail: rrk@mrc-lmb.cam.ac.uk

Received 26 January 1998; Accepted 11 March 1998; Revised 6 March 1998

• Keywords:

  • Dictyostelium,
  • phosphorelay,
  • protein kinase A,
  • RegA,
  • response regulator

Introduction

The work described here brings together two well-analysed signal transduction systems in a novel configuration: the cAMP/protein kinase A system and the histidine kinase/two-component system. cAMP is a ubiquitous second messenger which controls the activity of the cAMP-dependent protein kinase (PKA); the system is widespread in eukaryotes, especially in metabolic and neuronal control, and in development (Harwood et al., 1992a,b; Jiang and Struhl, 1995; Li et al., 1995). The level of intracellular cAMP depends on the relative activities of adenylyl cyclase and cAMP phosphodiesterase, both of which can be regulated through extracellular signals. Two-component systems are the basis of most bacterial sensory pathways, and consist, at their simplest, of a sensory histidine kinase that transfers phosphate from a histidine to an aspartate on a second protein, the response regulator (RR), which controls the effector function (Bourret et al., 1991; Parkinson and Kofoid, 1992; Swanson et al., 1994). Recently, two-component systems have been discovered in eukaryotes, including a more elaborate phosphorelay system in yeast which controls osmotic responses (Appleby et al., 1996; Posas et al., 1996). Dictyostelium similarly has an osmosensing histidine kinase (Schuster et al., 1996), and a further histidine kinase (Wang et al., 1996) and an RR (Shaulsky et al., 1996) which have important roles in development.

The fruiting body of Dictyostelium consists of a cellular stalk, stabilized by a basal disc of stalk cells, supporting a mass of spores. It is formed during culmination by the movement of pre-stalk cells up and into the stalk tube, which extends as more pre-stalk cells are added to the top, so lifting the developing spore mass upwards. These morphogenetic events are coupled to the terminal differentiation of each cell type in its appropriate position: stalk cells at the growing tip of the stalk and round its base, spore cells at the top of the stalk, where they mature in a gradient from the top to the bottom of the spore head (Bonner, 1952; Richardson et al., 1994).

Several types of experiment indicate that maturation of both spore and stalk cells is triggered by the activation of PKA. For instance, expression of a dominant inhibitory form of the regulatory subunit of PKA in pre-stalk or pre-spore cells prevents their maturation during normal development (Harwood et al., 1992b; Hopper et al., 1993b). Conversely, overactivity of the catalytic subunit of PKA causes spore and stalk cells to mature prematurely during development and, in the case of the rdeC mutants, which lack a functional PKA regulatory subunit, the final fruiting structure is a mound of differentiated cells (Anjard et al., 1992; Simon et al., 1992; Hopper et al., 1993b; Mann et al., 1994). These same genetic activating manipulations also stimulate the maturation of spore cells when amoebae are allowed to develop as submerged monolayers. Finally, wild-type cells can be induced to form both spore and stalk cells by pharmacological treatment with the cAMP analogue, 8-Br-cAMP, which penetrates the cell and activates PKA directly (Kay, 1989; Riley et al., 1989; Maeda, 1992; Inouye and Gross, 1993; Kubohara et al., 1993; see also Results).

PKA is activated by intracellular cAMP, which is produced largely during development by the adenylyl cyclase, ACA. In early development, the production of intracellular cAMP is stimulated by extracellular cAMP, through the cell surface family of G-protein-linked, cAMP receptors (cARs). Stimulation of these receptors, in turn, initiates a complex signal transduction pathway resulting in the activation of ACA (Firtel, 1996; Parent and Devreotes, 1996).

Thus, it might seem plausible that extracellular cAMP, acting alone through this known pathway, would be sufficient to trigger terminal differentiation during normal development. However, there are two major problems with such a scheme, both deriving from work on cells differentiating in monolayer culture. First, extracellular cAMP barely induces terminal differentiation of spore and stalk cells in culture, even though it adequately promotes pre-spore and pre-stalk cell differentiation (Kay, 1982; Berks and Kay, 1988). Second, extracellular cAMP actually inhibits the final stages of stalk cell formation and must be removed from the monolayer in order for wild-type NC4 stalk cells to mature (Berks and Kay, 1988). The inhibitory pathway seems to involve the protein kinase GSK3, because null mutants in gskA make stalk cells freely, even in the presence of extracellular cAMP (Harwood et al., 1995). These problems of PKA activation would be resolved if there was another pathway for regulating intracellular cAMP levels in Dictyostelium, apart from that acting through adenylyl cyclase. We present evidence for such a pathway, which appears to function by controlling cAMP breakdown through the cAMP phosphodiesterase RegA.

Screen for stalk and spore cell maturation mutants

We screened for mutants which are able to form mature spore and stalk cells when developed in tissue culture dishes under a simple salts medium. A non-hydrolysable cAMP analogue served to promote differentiation, and the stalk-specific inducer DIF-1 was provided, as appropriate, to enable stalk cell differentiation to occur (Table I). Wild-type cells do not aggregate in these conditions (due to the presence of the cAMP analogue) but remain amoeboid, and neither spore nor stalk cells form efficiently (NC4 cells form <1 in 10 000 viable spores; <2% stalk cells).

Table 1: Stalk and spore cell formation in culture

View full table

From a collection of >300 mutants created by restriction enzyme-mediated integration (REMI) insertional mutagenesis (Kuspa and Loomis, 1992), we identified strains myc1002 and HM332 which form both stalk and spore cells in the monolayer test conditions (Table I, section A). Subsequently, chemical (Abe and Yanagisawa, 1983) and disruption (Chang et al., accompanying paper) mutants of rdeA were tested, due to their phenotypic similarity to HM332, and were also found to form stalk and spore cells in monolayer culture.

The myc1002 insertion disrupted rdeC, the gene encoding the regulatory subunit of PKA (Simon et al., 1992), whereas rdeA mutants are known to have elevated levels of intracellular cAMP (Abe and Yanagisawa, 1983). Since mutations of both of these genes should result in an activated PKA, it seemed likely that the gene disrupted in strain HM332 might also encode a negative regulator of PKA activity. The HM332 insertion was identified as disrupting regA, a gene previously obtained as a suppressor of a sporulation-defective mutant (Shaulsky et al., 1996). To confirm the results with HM332, regA was disrupted by homologous recombination in Ax2 cells. The resulting strain, HM1015, formed stalk and spore cells even more efficiently than did HM332, perhaps due the greater health of its parental strain (Table I, section A). The regA null phenotype of HM1015 was corrected to wild-type when regA was expressed from its own promoter in this strain (not shown).

The three mutant classes, rdeC, rdeA and regA (see below), are thus all implicated in the regulation of PKA activity. In contrast to these three disruptants, mutations in other relevant genes did not promote stalk and spore maturation: gskA null cells yielded only stalk cells, as expected (Harwood et al., 1995), whereas mutants of the histidine kinases dokA (Schuster et al., 1996) and dhkA (Wang et al., 1996) behaved as wild-type cells (Table I, section B). These results suggest that neither of these kinases are negative regulators of the maturation pathway.

Additional evidence that PKA activation is sufficient to induce stalk cell maturation

Not all previous work has shown that activation of PKA is sufficient to induce stalk cell maturation. Thus, rdeC mutants did not form stalk cells in monolayers (Kay, 1989); however, in these experiments, exogenous DIF-1 was not supplied and, when it is, stalk cells do form (Table I, section C, strain HTY217). Also, strains overexpressing the PKA catalytic subunit in pre-stalk cells form only a very low number of mature stalk cells in normal development (Hopper et al., 1993a; Mann and Firtel, 1993). In contrast, we found that strain A7P, where the catalytic subunit is driven by an actin15 promoter (Anjard et al, 1992), did form stalk cells in monolayer culture with DIF-1, though not with great efficiency (we have observed that strains expressing the neomycin resistance gene, such as A7P, are often more fragile in monolayer culture than other strains). Furthermore, wild-type NC4 cells were induced to form stalk cells if Br-cAMP was added to activate PKA, consistent with earlier work (Inouye and Gross, 1993; Kubohara et al., 1993). Thus, sufficient activation of PKA can trigger stalk, as well as spore, cell differentiation. We next focused our attention on the regA gene product.

RegA protein is present throughout development

The regA transcript, which is present in pre-stalk and pre-spore cells, is expressed at a low level in vegetative cells, and is induced rapidly during aggregation, remaining present throughout development (Shaulsky et al., 1996). Western blots (Figure 1) show that expression of RegA protein is developmentally regulated. It is not observable before the end of aggregation, peaks at the mound stage (9–11 h) and remains at a lower level thereafter. Although RegA is not evident on Western blots before 9 h, the activity can be detected by biochemical assay. Using phosphodiesterase assays of immunoprecipitated material (see below), RegA activity is detectable at early stages of development and even in immunoprecipitates from growing Ax2 cells (data not shown). Western blots (Figure 1) and phosphodiesterase assays (see below) show that RegA protein is absent from strain HM1015; the band appearing at a late time in development is not RegA, but could be a related phosphodiesterase or a GST which reacts with the antiserum.

Figure 1.

Developmental expression of RegA protein. Cells of the wild-type Ax2 and the regA null, HM1015, were developed on agar, lysates prepared and RegA protein detected by Western blotting, as described in Materials and methods. The position of molecular weight standards (kDa) is shown.

View full figure (56 KB)

One domain of RegA is a cAMP phosphodiesterase

RegA is homologous to mammalian cyclic nucleotide phosphodiesterases (Shaulsky et al., 1996), but not to the Dictyostelium PdsA phosphodiesterase (Lacombe et al., 1986). We used a biochemical assay to confirm that RegA is a cAMP phosphodiesterase (PDE). Phosphodiesterase activity was evident in immunoprecipitates from developing cells (Figure 2), made using a polyclonal antiserum (R1/2F) against RegA. The R1/2F antiserum selectively precipitated RegA: negligible activity was precipitated from the regA null strain HM1015 (<1% that of Ax2 cells, the same level of activity isolated from Ax2 cells using pre-immune serum), whereas the level of activity precipitated from UK7, a strain disrupted in the pdsA gene, was very similar to that obtained from Ax2 cells (data not shown). Similar phosphodiesterase activity was seen by expressing intact RegA or its PDE domain in bacteria. The K_m of RegA for cAMP is 5 M (5.0 M for the endogenous enzyme, 6.7 M for the GST–PDE fusion). The enzyme is specific for cAMP, because cGMP does not compete with cAMP (Figure 2). Like the mammalian PDEs, RegA is sensitive to the general PDE inhibitor 3-isobutyl-1-methylxanthine (50% inhibitory concentration 250 M at 1 M cAMP). The PDE4 inhibitors rolipram or RO20-1724 did not inhibit. RegA is not regulated by Ca²⁺/calmodulin in vitro, and cGMP, at concentrations up to 20 mM, shows no allosteric effects on RegA activity.

Figure 2.

RegA is a cAMP phosphodiesterase. RegA was immunoprecipitated from wild-type (Ax2) cells at 8 h of development. Phosphodiesterase assays were performed using 0.4 M [³H]cAMP as substrate, plus the indicated concentrations of unlabelled cAMP or cGMP. The data shown are representative of three experiments, and are the mean of duplicate assays range (where no error bars are shown, these are contained within the symbols). Similar results were obtained with bacterially expressed GST–PDE and GST–RegA (data not shown).

View full figure (51 KB)

Terminal differentiation in regA mutants is still dependent on PKA

RegA phosphodiesterase activity could control terminal differentiation by regulating PKA, or through some other route involving cAMP. To test whether terminal differentiation in regA mutants still requires PKA, differentiation was followed in regA null cells expressing the Rm dominant inhibitor of PKA, a form of the PKA R-subunit which cannot bind cAMP or dissociate from the catalytic subunit (Harwood et al., 1992b). Expression of Rm in pre-stalk (strain HM2013) or pre-spore (HM2011) cells reduced stalk and spore differentiation compared with control cells expressing the inactive inhibitor, Rc (Table II). Therefore, in regA mutants, stalk and spore maturation requires PKA activity.

Table 2: Cell maturation in regA nulls is dependent on PKA

View full table

The second domain of RegA is a response regulator

The second domain of RegA has homology to RRs of two-component systems. RRs receive the phosphate controlling their activity from a histidine of an upstream histidine kinase. This phosphorylation occurs on a conserved aspartate residue (Volz, 1993) corresponding to Asp212 of RegA. Low molecular weight phospho-compounds are also effective phosphate donors, since the RR domain itself catalyses the phosphotransfer (Lukat et al., 1992). Accordingly, the RegA RR domain has phosphotransferase activity: a purified GST fusion protein containing the RegA RR domain became phosphorylated when incubated in vitro with acetyl-[³²P]phosphate (Figure 3). Mutant proteins lacking Asp212 were not labelled by acetyl phosphate, nor was bovine serum albumin (BSA) (Figure 3). Like other RRs, phosphotransferase activity of the wild-type RegA RR domain is Mg²⁺ dependent (Lukat et al., 1992; McCleary and Stock, 1994), and the label can be completely removed from the protein by brief heat treatment (100°C, 2 min).

Figure 3.

Phosphorylation of the RegA RR domain by acetyl phosphate. WT, wild-type RR domain; D212N, Asp212Asn mutant RR domain; D212E, Asp212Glu mutant RR domain; +EDTA, 20 mM EDTA present in the reaction (with the wild-type RR domain). Numbers at the side indicate the size of molecular weight standards (kDa).

View full figure (75 KB)

A second property of RRs, phosphdonor phosphatase activity (Lukat et al., 1992), can be detected in the RegA RR domain using phosphoramidate as substrate, but the rate of this RR-dependent phosphoramidate hydrolysis was very low. Using ³¹P NMR analysis, spontaneous hydrolysis of 18 mM phosphoramidate occurred at a rate of 0.5 mol/ml/h over 16 h at 25°C (see also Lukat et al., 1992). The RegA RR domain (at a concentration of 10 M) increased this rate by only 10% (not shown). No enzymatic breakdown of acetyl phosphate could be detected over its high rate of spontaneous hydrolysis. These results suggest that the RegA RR domain has low intrinsic phosphatase activity, resembling Spo0F, whose phosphorylated form is quite stable (Zapf et al., 1996), rather than CheY, whose phosphorylated form is very unstable due to its high intrinsic phosphatase activity (Lukat et al., 1992).

Phosphorylation of the RR domain activates the phosphodiesterase domain

Phosphorylation of RRs modulates their activity, e.g. the activities of CheB, NtrC and BvgA are enhanced by phosphorylation (Feng et al., 1992; Lukat et al., 1992; Sanders et al., 1992; Boucher et al., 1994). Addition of phosphoramidate, as phosphodonor, to RegA stimulated its PDE activity by up to 8-fold in vitro. Activation was dose dependent and saturable (not shown). Half-maximal activation occurred at 2 mM phosphoramidate and, by extrapolation to saturating phosporamidate concentrations, a maximal stimulation of 10-fold is expected. Kinetic analysis (Figure 4) showed that the activation of RegA by phosphoramidate is due to an increase in the V_max of the phosphodiesterase, with no appreciable change in the K_m for cAMP. Activation was also observed with immunoprecipitated RegA from Dictyostelium lystates, though to a lesser extent than with bacterially expressed protein.

Figure 4.

Phosphoramidate activates RegA by increasing the V_max. cAMP phosphodiesterase assays were performed in the absence (Control) or presence (10 mM PA) of 10 mM phosphoramidate, using cAMP concentrations in the range 2–50 M. The kinetics of the reaction were determined using a Lineweaver–Burk double-reciprocal plot. In the example shown (representative of three experiments), the kinetic parameters are as follows: V_max = 2.2 (control) and 17.5 arbitrary units (with phosphoramidate; 8-fold increase); K_m = 5.4 M (control) and 4.2 M (with phosphoramidate).

View full figure (38 KB)

Other potential phosphodonors (acetyl phosphate, ATP and phosphoenol pyruvate) had no effect at a concentration of 10 mM, whilst carbamyl phosphate gave a small stimulation (2-fold; not shown). Acetyl phosphate is probably ineffective due to the low stoichiometry of phosphorylation, as labelling of the RegA RR domain with 10 mM acetyl-[³²P]phosphate yielded no more than 4% phosphorylated protein.

Phosphoramidate did not activate the PDE domain expressed alone, nor mutant forms of RegA in which Asp212 of the RR domain was mutated to either asparagine or glutamic acid (Figure 5). This shows that activation is not due to a non-specific effect of phosphoramidate on the PDE domain, but depends on the RR domain and, within it, on the putative phosphoacceptor Asp212.

Figure 5.

Activation of RegA cAMP phosphodiesterase activity by phosphoramidate is dependent on the RR domain. (A) Outline of proteins used in in vitro activation experiments. Hatched squares, GST; white rectangles, RR domain (the wild-type has aspartate, D, at residue 212, mutants have either asparagine, N, or glutamate, E, at this position); black ovals, PDE domain. GST fusions were produced in bacteria; the native RegA protein was immunoprecipitated from Dictyostelium cell lysates. (B) In vitro phosphodiesterase assays were performed without (Control; white bars) or with (10 mM PA; black bars) 10 mM phosphoramidate. Results are expressed as activity measured in the presence of PA relative to that in its absence (basal activity = 100%). Results show the mean values, error bars are the standard error (n = 6).

View full figure (50 KB)

The flow of phosphate onto RegA determines the activity of the maturation pathway

Preventing the flow of phosphate onto Asp212 of RegA should prevent its activation and, therefore, decrease RegA activity in the cell. To attempt to divert the phosphate flow away from RegA, the RegA RR domain was expressed in wild-type Dictyostelium cells using the strong, non-specific actin15 promoter (giving strain HM2045). This caused a similar phenotype to the regA null mutant. Aggregation was completed by 6 h in HM2045, compared with 10 h for wild-type Ax2 cells. The terminal structures made by HM2045 were club-shaped (Figure 6), rather similar to regA nulls, and caused distinct yellowing of the agar, a common feature of strains which have precocious spore maturation. Indeed, spore production during development of HM2045 was accelerated compared with wild-type cells: HM2045 yielded 50% of its total spores by 14 h of development, whereas Ax2 cells required 19 h to make 50% of its final spore yield, both strains giving very similar spore numbers. Finally, HM2045 formed spores efficiently in monolayers (>10%), unlike the wild-type but similarly to regA, rdeC and rdeA mutant strains. In contrast, expression of the full-length regA gene, driven by the actin15 promoter, resulted in slower development than for the wild-type (although fruiting bodies eventually formed).

Figure 6.

Developmental morphology of wild-type and mutant strains. Wild-type (Ax2), regA null (HM1015), regA^-/actin15::RegA (regA null cells transformed with regA cDNA driven by the actin15 promoter; HM2042) and Ax2 cells overexpressing the RR domain (actin15::RR; HM2045) were developed on KK2 agar for 26 h. The final morphology of all strains is shown, except for regA^-/actin15::RegA (see text). The regA null strain has a greatly thickened lower portion of the stalk, often more severe than in the image shown, such that the structures have a pyramid of mature stalk/spore cells at their base, from which a tapering stalk extends. Likewise, the HM2045 strain can also form these structures.

View full figure (86 KB)

Rapid development and de-regulated cell maturation in HM2045 indicate that the phosphorelay pathway governing RegA activity is operative during early as well as late development, and thus that RegA is likely to have roles throughout Dictyostelium development.

The RegA protein functions at an intersection between the cAMP and two-component signal transduction systems, combining in itself domains characteristic of both. The PDE domain has good homology to mammalian phosphodiesterases (Shaulsky et al., 1996), and we have shown biochemically that it is a cAMP-specific phosphodiesterase (see also Shaulsky et al., 1998). The RR domain has clear homology to other response regulators, and it can accept phosphates from acetyl phosphate, in a reaction typical of RRs, with transfer dependent on the predicted phosphoaccepting residue, Asp212. Our results using phosphoramidate indicate that phosphorylation of the RR domain activates the PDE domain.

In the simplest configuration of a two-component system, phosphates are transferred directly from a histidine kinase to the Asp of the RR (Parkinson and Kofoid, 1992; Swanson et al., 1994). However, in more complex systems such as those controlling initiation of sporulation in Bacillus subtilis, and in the yeast osmo-regulatory pathway, phosphates flow through a four-step phosphorelay (Burbulys et al., 1991; Posas et al., 1996). In the yeast system, phosphates are relayed from His to Asp on the kinase (Sln1p) and then to a His on an intermediate phosphotransfer protein, Ypd1p, before reaching the RR (Ssk1p).

In the case of Dictyostelium, mutants of a second gene, rdeA, strongly resemble regA mutants phenotypically: fruiting body formation is accelerated and final morphogenesis abherrant (Abe and Yanagisawa, 1983), and both strains are able to form mature stalk and spore cells in our monolayer test. Recent work demonstrates that the Dictyostelium rdeA gene in fact encodes a functional homologue of the yeast Ypd1 gene, as shown most strikingly by the complementation of an rdeA mutant by Ypd1 (Chang et al., accompanying paper). rdeA mutants are also known to have high cAMP levels (Abe and Yanagisawa, 1983), consistent with reduced RegA activity. It therefore seems likely that RegA is controlled by a phosphorelay system configured in a similar way to the osmo-regulatory pathway of yeast, and that RdeA is the immediate upstream phosphodonor (Figure 7). Thus we propose that stalk and spore cell maturation are both controlled by a common pathway, involving RdeA and RegA, that controls cAMP levels and hence PKA activity.

Figure 7.

Proposed phosphorelay pathway controlling RegA activity. The upper scheme is the current model of the yeast osmo-regulatory phosphorelay system (Posas et al., 1996), the lower scheme is our proposed phosphorelay pathway for RegA regulation. The flow of phosphate is shown by curved arrows (the first of which is autophosphorylation of the histidine kinase), and occurs by a four-step process. The first component of the pathway in Dictyostelium is not known, but is shown as being a hybrid histidine kinase (labelled His Kin); phosphates flow from this to RdeA and then to RegA, where phosphorylation at Asp212 of the RR domain activates the phosphodiesterase domain (this is shown by the crooked arrow). In both schemes, when the phosphorelay pathway is active, the downstream signalling pathways are less active.

View full figure (47 KB)

The cognate upstream histidine kinase(s) and ligand(s) that control RegA activity remain unknown. Since phosphorylation on Asp212 appears to activate the RegA phosphodiesterase, we predict that the flow of phosphates from the upstream kinase should also be activating and, therefore, that kinase-null mutants should resemble regA mutants phenotypically. Consistent with this prediction, genetic manipulations expected to interrupt the flow of phosphates to RegA (rdeA knock out or RR domain overexpression) cause similar phenotypes to a regA null. On these grounds, the suggestion by Loomis and co-workers (Loomis et al., 1997; Shaulsky et al., 1998) that phosphorylation is inhibitory to RegA and that the upstream kinase is DhkA seem unlikely, especially since dhkA null mutants behave as wild-type in monolayers and differ significantly from regA in their developmental phenotype: they develop long, fragile stalks and are defective in spore production (Wang et al., 1996).

The discovery of the regA pathway for controlling cAMP levels also suggests a solution to the problem of why extracellular cAMP alone is insufficient to induce stalk and spore cell maturation in cultured cells, even though it can activate adenylyl cyclase. We propose that only the highest levels of intracellular cAMP (and hence PKA activity) are sufficient to drive spore and stalk maturation and that they are attained only when adenylyl cyclase is active and the RegA phosphodiesterase is inactive (Houslay and Milligan, 1997). Since the activity of these proteins is probably controlled by two different extracellular signals, this condition will only be met where both ligands are present at the appropriate concentrations. In this way, the very precise spatial and temporal control over terminal differentiation can be obtained, which is presumably necessary for morphogenesis of the fruiting body.

Cell methods

Cells were grown and developed at 22°C (Watts and Ashworth, 1970; Kay, 1987) and REMI mutants isolated in the DH1 background as before (Harwood et al., 1995). Monolayer differentiation was as before (Kay, 1987) but used 10 M cAMP-S (Sigma), and (for stalk assays) 100 nM DIF-1 or 15 mM 8-Br-cAMP, as indicated.

Rc and Rm cell lines were made by introducing the pre-stalk- or pre-spore-specific constructs (Harwood et al., 1992a; Hopper et al., 1993a) into HM1015 by the standard CaPO₄ method (Harwood et al., 1992a). Initial transformants were selected at 40–80 g/ml and maintained at 50 g/ml G418; stable transformants were maintained at 20 g/ml G418 (10 g/ml for 'rescued' strains). HM1015 was made by electroporation of 50 g of HindIII–SacII-linearized pRegAKO into Ax2 (1.610⁷ cells). Transformants were selected with 20 g/ml blasticidin S. Null mutants were confirmed by Western and Southern blotting.

Molecular biology

The REMI insert and flanking regions in myc1002 were isolated by plasmid rescue; this failed for HM332 and so inverse PCR was used instead. The PCR product (3' end of regA) was used to probe a gt11 cDNA library, prepared from cAMP pulse-induced cells (a gift of P.Devreotes). This yielded an 80% full-length cDNA which was used to construct a full-length cDNA in conjunction with a genomic DNA clone covering the missing 5' end of the gene, by PCR (introducing a BssSI site at nucleotides 487–492; this is a silent mutation).

The regA KO vector was made from a 3.5 kb genomic fragment of the regA promoter and the 5' end of the gene (isolated from a genomic minilibrary of EcoRI–BclI-digested Ax2 DNA in pBluescript KS II). 5' distal sequences were removed using HindIII, and the blasticidin S deaminase cassette (from pRHI119, obtained from R.H.Insall) inserted, giving pRegAKOD3'. A 1.2 kb genomic NotI–SacII PCR product from regA, covering the 3' end of the gene to the stop codon, was inserted 3' to the bsr cassette in pRegAKOD3', giving rise to pRegAKO, with a 263 amino acid deletion (229–491).

Full-length regA cDNA, the RR domain and the PDE domain were expressed as GST fusion proteins in Escherichia coli, using pGEX-2T (Pharmacia). The RR and PDE domain constructs encompassed amino acids 127–335 and 385–793, respectively. Soluble GST fusion proteins were purified using glutathione–agarose resin (Fluka). The RR domain construct expressed from the actin15 promoter in Dictyostelium cells encompassed RegA amino acids 1–430. All constructs were confirmed by DNA sequencing (ABI377).

Mutagenesis and rescue plasmids

Asp212 of RegA was mutagenized to Asn or Glu using the QuikChange site-directed mutagenesis kit (Stratagene). Actin15::RegA was made by replacing the gfp cDNA from actin15::gfp with the regA cDNA from pKSIISalI:regA (BamHI–XhoI fragment). The regA promoter rescue construct was made by digesting the regA 3.5 kb genomic fragment with BglII–SalI and inserting the 1.8 kb promoter fragment into pKSIISalI: regA digested with BamHI–SalI, to give an expression cassette of the regA cDNA driven in-frame by its own promoter. This cassette was removed as a 5.0 kb XbaI–XhoI fragment and used to replace the actin15::gfp cassette. Both rescue vectors conferred resistance to G418.

Immunochemistry

Polyclonal antiserum (R1/2F) was raised in rabbits to the GST–PDE fusion. RegA was detected by Western blotting of cell lysates in 50 mM Tris pH 7.4, 1 mM EDTA, 0.05% Triton X-100, containing protease inhibitors [1 mM benzamidine, 10 M L64, 10 g/ml leupeptin, 1 M pepstatin, 1 mM phenylmethylsulfonyl fluoride (PMSF) and 0.1 mM TLCK]. Proteins (40 g/lane) were resolved by 10% SDS–PAGE, electroblotted onto Immobilon P (Millipore), incubated with R1/2F and then with horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG (Sigma), and bands were visualized by ECL (Amersham). R1/2F detects a band of 110 kDa on a Western blot.

RegA was immunoprecipitated from cell lysates of 1.110⁸ cells/ml in IP buffer (50 mM K₁K₂PO₄ pH 7.3, 1 mM MgCl₂ and 10% glycerol plus protease inhibitors as above), pre-cleared by centrifugation. The equivalent of 16 mg (dry weight) of protein A–Sepharose CL-4B (Pharmacia) per 10⁸ cells and 30 l of R1/2F serum were added. After 1 h at 4°C, beads were pelleted and washed six times in eight bed volumes of IP buffer, then assayed for PDE activity. Protein concentrations were determined using the Bio-Rad dye-binding assay.

Phosphodiesterase assays

Assays were in 50 mM Tris–HCl, 50 mM KCl, 5 mM MgCl₂, 10% glycerol, 1 mM dithiothreitol (DTT), pH 8.0 at 25°C, usually containing 0.33 M (410⁵ d.p.m.) [³H]cAMP (Amersham), 1–100 M total cAMP, final volume 20 l, at 25°C. Reactions were stopped with trichloroacetic acid (TCA) (to 5%), and nucleotides were resolved by TLC on PEI-cellulose plates with a fluorescent indicator (Sigma), using 1 M ammonium acetate pH 7.5/ethanol (30:75 v/v). Nucleotide spots were excised and radioactivity measured by scintillation counting. Using potassium phosphate buffers, the pH optimum of RegA was 8.0, and 5 mM MgCl₂ was optimal. Tris buffer was used instead of phosphate buffer for assays to avoid any potential problems when analysing the effects of phosphodonor compounds. Phosphoramidate was pre-incubated with enzyme for 5 min. Phosphoramidate was the kind gift of Dr Ann Stock.

In vitro ³²P labelling

Acetyl-[³²P]phosphate was synthesized as described in Kornberg et al. (1956), but using only 50 mol of K₂HPO₄ as substrate. Five g of fusion protein was used per labelling reaction (10 l), containing 10 mM MgCl₂ (Lukat et al., 1992) for 10 min at 25°C. SDS–PAGE and electroblotting were done at 4°C to minimize loss of label from proteins.

Accession number

The DDBJ/EMBL/GenBank accession number for the regA locus, including 1.8 kb of promoter sequence, is AJ005398.

We are grateful to Drs Bill Loomis and Gad Shaulsky for sharing data prior to publication, and for the dhkA strain, to Drs Richard Sucgang and Rich Kessin for providing strain UK7, to Drs M.-Y.Chen and P.Devreotes for strain myc1002, and to Dr Stephan Schuster for the dokA strain. We thank Dr Ann Stock for help with phosphoramidate experiments, Dr Stephan Schuster for advice on acetyl phosphate labelling, and Dr JiChun Yang for performing NMR experiments. We thank Drs Julian Gross and Peter Newell for helpful discussions, and Drs Chris Thompson, Hugh Pelham and particularly Jacqueline Milne for useful comments on the manuscript. This work was supported in part by an International Research Scholars award from the Howard Hughes Medical Institute to R.R.K.

Abe K and Yanagisawa K (1983) A new class of rapid developing mutants in Dictyostelium discoideum: implications for cyclic AMP metabolism and cell differentiation. Dev Biol, 95, 200–210. | Article | PubMed | ISI | ChemPort |

Anjard C, Pinaud S, Kay RR and Reymond CD (1992) Overexpression of DdPK2 protein kinase causes rapid development and affects the intracellular cAMP pathway of Dictyostelium discoideum. Development, 115, 785–790. | PubMed | ISI | ChemPort |

Appleby JL, Parkinson JS and Bourret RB (1996) Signal transduction via the multistep phosphorelay: not necessarily a road less traveled. Cell, 86, 845–848. | Article | PubMed | ISI | ChemPort |

Berks M and Kay RR (1988) Cyclic AMP is an inhibitor of stalk cell differentiation in Dictyostelium discoideum. Dev Biol, 126, 108–114. | Article | PubMed | ISI | ChemPort |

Bonner JT (1952) The pattern of differentiation in amoeboid slime molds. Am Nat, 86, 79–89. | Article | ISI

Boucher PE, Menozzi FD and Locht C (1994) The modular architecture of bacterial response regulators. J Mol Biol, 241, 363–377. | Article | PubMed | ISI | ChemPort |

Bourret RB, Borkovich KA and Simon MI (1991) Signal transduction pathways involving protein phosphorylation in prokaryotes. Annu Rev Biochem, 60, 401–441. | Article | PubMed | ISI | ChemPort |

Burbulys D, Trach KA and Hoch JA (1991) Initiation of sporulation in B.subtilis is controlled by a multicomponent phosphorelay. Cell, 64, 545–552. | Article | PubMed | ISI | ChemPort |

Feng J, Atkinson MR, McCleary W, Stock JB, Wanner BL and Ninfa AJ (1992) Role of phosphorylated metabolic intermediates in the regulation of glutamine synthase synthesis in Escherichia coli. J Bacteriol, 174, 6061–6070. | PubMed | ISI | ChemPort |

Firtel RA (1996) Interacting signaling pathways controlling multicellular development in Dictyostelium. Curr Opin Genet Dev, 6, 545–554. | Article | PubMed | ISI | ChemPort |

Harwood AJ, Hopper NA, Simon MN, Bouzid S, Veron M and Williams JG (1992a) Multiple roles for cAMP-dependent protein kinase during Dictyostelium development. Dev Biol, 149, 90–99. | Article | PubMed | ISI | ChemPort |

Harwood AJ, Hopper NA, Simon MN, Driscoll DM, Veron M and Williams JG (1992b) Culmination in Dictyostelium is regulated by the cAMP-dependent protein kinase. Cell, 69, 615–624. | Article | PubMed | ISI | ChemPort |

Harwood AJ, Plyte SE, Woodgett J, Strutt H and Kay RR (1995) Glycogen synthase kinase 3 regulates cell fate in Dictyostelium. Cell, 80, 139–148. | Article | PubMed | ISI | ChemPort |

Hopper NA, Anjard C, Reymond CD and Williams JG (1993a) Induction of terminal differentiation of Dictyostelium by cAMP-dependent protein kinase and opposing effects of intracellular and extracellular cAMP on stalk cell differentiation. Development, 119, 147–154. | ISI | ChemPort |

Hopper NA, Harwood AJ, Bouzid S, Veron M and Williams JG (1993b) Activation of the prespore and spore cell pathway of Dictyostelium differentiation by cAMP-dependent protein kinase and evidence for its upstream regulation by ammonia. EMBO J, 12, 2459–2466. | PubMed | ISI | ChemPort |

Houslay MD and Milligan G (1997) Tailoring cAMP signalling responses through isoform multiplicity. Trends Biochem Sci, 22, 217–224. | Article | PubMed | ISI | ChemPort |

Inouye K and Gross J (1993) In vitro stalk cell differentiation in wild-type and slugger mutants of Dictyostelium discoideum. Development, 118, 523–526. | ISI | ChemPort |

Jiang J and Struhl G (1995) Protein kinase A and hedgehog signaling in Drosophila limb development. Cell, 80, 563–572. | Article | PubMed | ISI | ChemPort |

Kay RR (1982) cAMP and spore differentiation in Dictyostelium discoideum. Proc Natl Acad Sci USA, 79, 3228–3231. | Article | PubMed | ChemPort |

Kay RR (1987) Cell differentiation in monolayers and the investigation of slime mold morphogens. Methods Cell Biol, 28, 433–448. | PubMed | ISI | ChemPort |

Kay RR (1989) Evidence that elevated intracellular cyclic AMP triggers spore maturation in Dictyostelium. Development, 105, 753–759. | ISI | ChemPort |

Kornberg A, Kornberg SR and Simms ES (1956) Metaphosphate synthesis by an enzyme from Escherichia coli. Biochim Biophys Acta, 20, 215–227. | Article | PubMed | ISI | ChemPort |

Kubohara Y, Maeda M and Okamoto K (1993) Analysis of the maturation process of prestalk cells in Dictyostelium discoideum. Exp Cell Res, 207, 107–114. | Article | PubMed | ISI | ChemPort |

Kuspa A and Loomis WF (1992) Tagging developmental genes in Dictyostelium by restriction enzyme-mediated integration of plasmid DNA. Proc Natl Acad Sci USA, 89, 8803–8807. | Article | PubMed | ChemPort |

Lacombe ML, Podgorski GJ, Franke J and Kessin RH (1986) Molecular cloning and developmental expression of the cyclic nucleotide phosphodiesterase gene of Dictyostelium discoideum. J Biol Chem, 261, 16811–16817. | PubMed | ChemPort |

Li W, Ohlmeyer JT, Lane ME and Kalderon D (1995) Function of protein kinase A in hedgehog signal transduction and Drosophila imaginal disc development. Cell, 80, 553–562. | Article | PubMed | ISI | ChemPort |

Loomis WF, Shaulsky G and Wang N (1997) Histidine kinases in signal transduction pathways of eukaryotes. J Cell Sci, 110, 1141–1145. | PubMed | ISI | ChemPort |

Lukat GS, McCleary WR, Stock AM and Stock JB (1992) Phosphorylation of bacterial response regulator proteins by low molecular weight phospho-donors. Proc Natl Acad Sci USA, 89, 718–722. | Article | PubMed | ChemPort |

Maeda M (1992) Efficient induction of sporulation of Dictyostelium prespore cells by 8-bromocyclic AMP under both submerged- and shaken-culture conditions and involvement of protein kinase(s) in its action. Dev Growth Differ, 34, 263–275. | Article | ISI | ChemPort |

Mann SKO and Firtel RA (1993) cAMP-dependent protein kinase differentially regulates prestalk and prespore differentiation during Dictyostelium development. Development, 119, 135–146. | PubMed | ISI | ChemPort |

Mann SKO, Richardson DL, Lee S, Kimmel AR and Firtel RA (1994) Expression of cAMP-dependent protein kinase in prespore cells is sufficient to induce spore cell differentiation in Dictyostelium. Proc Natl Acad Sci USA, 91, 10561–10565. | Article | PubMed | ChemPort |

McCleary WR and Stock JB (1994) Acetyl phosphate and the activation of two-component response regulators. J Biol Chem, 269, 31567–31572. | PubMed | ISI | ChemPort |

Parent CA and Devreotes PN (1996) Molecular genetics of signal transduction in Dictyostelium. Annu Rev Biochem, 65, 411–440. | Article | PubMed | ISI | ChemPort |

Parkinson JS and Kofoid EC (1992) Communication modules in bacterial signalling proteins. Annu Rev Genet, 26, 71–112. | Article | PubMed | ISI | ChemPort |

Posas F, Wurgler-Murphy SM, Maeda T, Witten EA, Thai TC and Saito H (1996) Yeast HOG1 MAP kinase cascade is regulated by a multistep phosphorelay mechanism in the SLN1–YPD1–SSK1 'two-component' osmosensor. Cell, 86, 865–875. | Article | PubMed | ISI | ChemPort |

Richardson DL, Loomis WF and Kimmel AR (1994) Progression of an inductive signal activates sporulation in Dictyostelium discoideum. Development, 120, 2891–2900. | PubMed | ISI | ChemPort |

Riley BB, Jensen BR and Barclay SL (1989) Conditions that elevate intracellular cAMP levels promote spore formation in Dictyostelium. Differentiation, 41, 5–13. | Article | PubMed | ISI | ChemPort |

Sanders DA, Gillece-Castro BL, Burlingame AL and Koshland DEJ (1992) Phosphorylation site of NtrC, a protein phosphatase whose covalent intermediate activates transcription. J Bacteriol, 174, 5117–5122. | PubMed | ISI | ChemPort |

Schuster SC, Noegel AA, Oehme F, Gerisch G and Simon MI (1996) The hybrid histidine kinase DokA is part of the osmotic response system of Dictyostelium. EMBO J, 15, 3880–3889. | PubMed | ISI | ChemPort |

Shaulsky G, Escalante R and Loomis WF (1996) Developmental signal transduction pathways uncovered by genetic suppressors. Proc Natl Acad Sci USA, 93, 15260–15265. | Article | PubMed | ChemPort |

Shaulsky G, Fuller D and Loomis WF (1998) A cAMP-phosphodiesterase controls PKA-dependent differentiation. Development, 125, 691–699. | PubMed | ISI | ChemPort |

Simon MN, Pelegrini O, Veron M and Kay RR (1992) Mutation of protein kinase-A causes heterochronic development of Dictyostelium. Nature, 356, 171–172. | Article | PubMed | ISI | ChemPort |

Swanson RV, Alex LA and Simon MI (1994) Histidine and aspartate phosphorylation: two-component systems and the limits of homology. Trends Biochem Sci, 19, 485–490. | Article | PubMed | ISI | ChemPort |

Volz K (1993) Structural conservation in the CheY superfamily. Biochemistry, 32, 11741–11753. | Article | PubMed | ISI | ChemPort |

Wang N, Shaulsky G, Escalante R and Loomis WF (1996) A two-component histidine kinase gene that functions in Dictyostelium development. EMBO J, 15, 3890–3898. | PubMed | ISI | ChemPort |

Watts DJ and Ashworth JM (1970) Growth of myxamoebae of the cellular slime mould Dictyostelium discoideum in axenic culture. Biochem J, 119, 171–174. | PubMed | ISI | ChemPort |

Zapf JW, Hoch JA and Whiteley JM (1996) A phosphotransferase activity of the Bacillus subtilis sporulation protein Spo0F that employs phosphoramidate substrates. Biochemistry, 35, 2926–2933. | Article | PubMed | ISI | ChemPort |

Main navigation

• Search EMBO Journal

  Search

• Advanced search

• The EMBO Journal

  • Home
  • Aims and scope
  • Current issue
  • Advance online publication
  • Web focuses
  • Archive:
  • Browse by issue
  • Browse by subject
  • Browse by category
  • Free online sample issue
  • Press releases

• Authors and Referees

  • Editorial process
  • Guide for authors
  • Submit an article
  • Guide for referees
  • Editors and Editorial Board
  • Contact editorial office

• Customer Service

  • Subscribe
  • Order sample copy
  • Purchase articles
  • Reprints and permissions
  • Contact NPG
  • Advertising