Multiple roles for cAMP-dependent protein kinase during Dictyostelium development

DEVELOPMENTAL

BIOLOGY

149,

go-99

(19%)

Multiple Roles for CAMP-Dependent Protein Kinase during Dictyostelium Development A. J. HARWOOD, N. A. HOPPER, M-N. SIMON,* S. BOUZID,* M. VERON,* AND J. G. WILLIAMS Imperial

Cancer

Research

Fund,

Clare

and *Unite de Biochimie

Hall Laboratories, Cellulaire, Institut Accepted

South

Mimes,

Herts

Pasteur, 75724 Park

September

EN6 Cedex

3LD, United Kingdom; 15, France

4, 1991

The CAMP-dependent protein kinase (PKA) holoenzyme of Dictyostelium comprises a single regulatory (R) and eatalytic (C) subunit, and both proteins increase in concentration during cellular aggregation. In order to determine the role of the kinase, we have constructed mutants of the R subunit that are defective in CAMP binding, in inhibition of the C subunit, or in both functions. Analysis of these mutants suggests that overexpression of the unmutated R subunit, which is known to block development, occurs by direct inactivation of the C subunit rather than by an effect on intracellular CAMP levels. Cells with an inactive C subunit (PKA- cells) are defective in CAMP relay, the production of CAMP in response to extracellular CAMP stimulation. This presumably accounts for their inability to undertake aggregation. When mixed with wild-type cells, PKA- cells migrate toward the signalling centre but remain confined to the periphery of the tight aggregate and are lost from the back of the migratory slug. This suggests that PKA may be required during the late, multicellular stages of development. Consistent with this, we find that a number of postaggregative genes are not expressed in PKA- cells, even when they are allowed to synergise with normal cells. o 1992 Academic PRSS, IN.

a peak at the end of cellular aggregation (Leichtling et aZ., 1984; de Gunzburg et uL, 1986). Development is blocked prior to aggregation by overexpression of either the Dictyostelium R subunit (Simon et al., 1989) or a mutant form of the mouse type I R subunit cDNA that is unable to bind CAMP (Firtel and Chapman, 1990). We have now constructed a series of mutants of the Dictyostelium R subunit and used them to investigate the role of PKA and the mechanism whereby R subunit overexpression blocks development.

INTRODUCTION

During aggregation, individual Dictyostelium amoebae are attracted to the centre of the developing field by pulses of extracellular adenosine 3’:5’-monophosphate (CAMP) (Konijn et aZ., 1967). Extracellular CAMP binds to cell surface receptors and these couple to G-proteins that are believed to activate a number of intracellular enzymes, including adenylate cyclase, guanlylate cyclase, and phospholipase C (Janssens and Van Haastert, 1987; Snaar-Jagalska et aL, 1988; Firtel et uL, 1989). In addition to acting as the chemotractant, CAMP induces the expression of specific genes, but the precise mechanism of their activation is unknown (reviewed by Kessin, 1988; Mann et aL, 1988; Williams, 1988). As in other eukaryotic cells, elevation of intracellular CAMP increases the activity of the CAMP-dependent protein kinase (PKA) by dissociation of the regulatory (R) subunit from the catalytic (C) subunit (de Gunzburg and Veron, 1982; Leichtling et aZ., 1982; Rutherford et ak, 1982; Majerfeld et al, 1984; Mutzel et aL, 1987). The Dietyostelium and vertebrate R subunits share a high degree of sequence homology. The Dictyostelium holoenzyme differs from PKA in higher species, where the enzyme has a tetrameric structure (Flockhart and Corbin, 1982), by forming a dimer containing one R subunit and one C subunit (de Gunzburg et al, 1984; Mutzel et aZ., 1987). The levels of C and R subunits increase during development, from a very low level in vegetative cells to reach 0012-1606/92 $3.00 Copyright All rights

0 1992 by Academic Press, Inc. of reproduction in any form reserved.

METHODS

Site-Directed

Mutagenesis

and Plasmid

Constructs

Oligonucleotide-mediated site-directed mutagenesis was used to introduce mutations into the Dictyostelium R subunit. Single-stranded templates were prepared after subcloning the R subunit cDNA from pIMS501 as either an EcoRI fragment into Bluescript II-SK’-‘(Stratagene) or an EcoRI-EcoRV fragment into M13mp18. The following mutagenic oligonucleotides were used: 5’-GTAGTTTGAAGAATTAGCT-3’ (mutant AE) 5’-AAGGTGGTACCTTTGAAGAATTAGCTT-8 (mutant Am) 5’-ATTACTTTGAAGAAATTGCA-3’ (mutant BE) 5’-TCTGATTACTCTGCAGAAATTGCATT-3’ (mutant BsA).

Introduction of the mutations was confirmed by sequencing. The mutant R subunit genes were subcloned as EcoRI fragments into the plasmid pIMS5 for expression in E. coli (Simon et uL, 1988). For expression in 90

HARWOOD

ET AL.

Dictyostelium, the EcoRI fragments were end-filled and subcloned into the end-filled &$I1 site of pB1OActBKH. A comparable, unmutated R subunit construct was also made. The other plasmids used in this study have been described previously (Harwood and Drury, 1990; Jermyn and Williams, 1991). Biochemical

Analysis

Cultures of E. co& (strain DH5) cells containing the pIMS/R subunit constructs were disrupted by sonication and centrifuged to give l-2 mg/ml protein extracts. Samples were tested for CAMP binding by incubation with 500 nMrH]cAMP for 30 min at O”C, then diluted to 3 ml, and filtered onto nitrocellulose (de Gunzburg et a& 1984). Filters were washed three times and dried and the bound counts measured. To test for inhibition of PKA activity, partially purified Dictyostelium C subunit was incubated in the presence of [T-~~P]ATP and the peptide substrate Kemptide. The C subunit was preincubated in the presence of R subunit extracts with or without 1 mM CAMP. Strains and Culture Comiitiolzs

of D. discoideum

Roles

for

91

CAMP

firmed the overexpression of the R subunit protein the transformant clones (Simon et al, 1989). Chemotaxis

in

Assay

Vegetative cells were washed in KK, and harvested by centrifugation to form a thick slurry of approximately 10’ cells/ml, and 1 ~1 of each transformant was placed onto a 2% nonnutrient agar containing 10 mM CAMP (Bonner et al., 1966). Assays were incubated for 8 hr at 22°C before scoring. CAMP Assay Extracellular CAMP was measured by the method described by Van Haastert (1984). Cells were starved for 4.5 hr, washed twice in KK,, and resuspended at a density of 6.25 X 10’ cells/ml. The cells were stimulated with 10 mM 2’deoxyadenosine 3’:5’-monophosphate (2’deoxy-CAMP) and loo-ml aliquots were removed every 2 min. The cells were pelleted and the CAMP concentration in the supernatant was determined using the isotope dilution method (Amersham PLC). Histochemical

Dictyostelium discoideum, strain AX2 and derived transformants, were grown in HL-5 medium at 22°C (Watts and Ashworth, 1970). For development, harvested cells were washed three times in KK2 (16.5 mM KH2POI, 3.8 mM K2HPOI, pH 6.2) and spread at a density of 10’ cells/ml onto Millipore nitrocellulose filters supported on a pad soaked in KK, or onto 2% nonnutrient agar to obtain slugs. Transformation

Multiple

Staining

for ,&Galactosidase

Aggregates and slugs were fixed for 15 min in 1% gluteraldehyde in Z buffer (60 mM N%HPOI, 40 mM NaH,PO,, 10 mMKC1, 1 mMMgSO,, 2 mMMgC12). Samples were washed twice in Z buffer and incubated in staining solution (0.1% 5-bromo-4-chloro-3-indolyl-@D-galactoside (X-gal), 5 mM potassium ferricyanide, and 5 mM potassium ferrocyanide in Z buffer) at 37°C (Dingermann et al, 1989).

Cells

Transformation was carried out using the calciumphosphate precipitation method as previously described (Nellen et ah, 1984, Early and Williams, 198’7). Cells (10’) were plated into a petri dish containing MES-buffered HL-5 medium (pH 7.05) 4 hr prior to addition of the precipitate. The precipitate was prepared from a total of 12 rg of plasmid DNA in HBS (137 mMNaC1,5 mMKC1, 0.7 mM N+HPO,, 6 mM dextrose, 21 mM Hepes, pH 7.05) and 125 mM CaCl, and added to the cells for 5 hr. The cells were exposed to a 15% glycerol shock for 10 min allowed to recover overnight in nonselective medium, and then selected in HL-5 medium containing 20 pg/ml G418. Clones were picked after 7 to 10 days. Clones containing higher plasmid copy numbers and hence elevated R subunit expression were selected by growth in medium containing 100 pg/ml G418 (Nellen and Firtel, 1985). Southern and Western analysis con-

Luciferwe

Assay

Dictyostelium cells were developed on nitrocellulose filters or in suspension. Suspension cells were starved for 90 min and then induced by addition of 0.5 mMcAMP after every hour for a total of 4 hr. Cells were washed in KK, and lysed in 100 ~1 of lysis buffer A (100 mM phosphate buffer, pH 7.8, 8 mM MgCl,, 1 mM EDTA, 1 mM DTT, 1% Triton X-100, 15% glycerol, 0.5 mM PMSF). Protein content was determined using dye-binding assay (Bradford, 1976) and BSA was added to 1%. Typically, 5 pg of protein was assayed in 230 ~1 of LBA containing 0.5 mM ATP and 0.1 mM luciferin (Sigma). Luminescence was determined using an LKB 1251 luminometer. Assays were carried out in duplicate. Each experiment was carried out using two independent clones to calculate the mean and standard error of the mean (SEM).

92

VOLUME

DEVELOPMENTALBIOLOGY

149,19%?

TABLE 1 Mutant substitutions Mutant name ATE BSA AEBE ATEBSA

Position

Wild-type residue + mutant residue

133; 135 260; 261 135; 261 133; 135; 260; 261

Serine; glycine -, threonine; glutamic acid Phenylalanine; glycine + serine; alanine Glycine; glycine + glutamic acid; glutamic acid Serine; glycine; phenylalanine; glycine + threonine; glutamic acid; serine; alanine

Site A Wild-type Mutant

Site B

E G G S F G E L A L I Y G S P R A A T V EGGTFEELALIYGSPRAATV I I I I 130 133 135 140

Enzymatic Assay for /3-Galactosidase Dictyostelium cells were harvested by centrifugation and washed twice in KK, buffer. Assays were performed by a slight modification of the method described by Miller (1972). Cells (5 x 107) were lysed in lysis buffer (100 mM phosphate buffer, pH 7.8, 8 mM MgCl,, 1 mM EDTA, 1 mMDTT, 1% Triton X-100,15% glycerol) and cleared by centrifugation for 5 min. Two microliters of extract was used to determine protein concentrations and fi-galactosidase activity was assayed from 100 ~1 of extract by addition of 500 ~1 of o-nitrophenyl-P-D-galactopyranoside solution (1.6 mg/ml in Z buffer) for 30 min at 22°C. Reactions were stopped by adding 400 ~1 of 1 M Na,CO,. The samples were centrifuged at 120~7 for 5 min and diluted with 500 ~1 of Z buffer, and their optical density at 420 nm was determined. Assays were performed in triplicate. Each experiment was performed using three independent clones.

PSDYFGEIALLTDRPRAATV PSDYSAEIALLTDRPRAATV I 260

I 265

grees in the mutant proteins. Proteins mutated in both sites A and B (AEBE and ATEBSA) display an approximate loo-fold reduction in CAMP binding in comparison to the wild-type protein, whereas proteins mutated only at site A or B show intermediate reductions of 26-fold (ATE) and 2.3-fold (BSA), respectively. The wild-type, ATE, and AEBE mutant proteins all inhibit the activity of the C subunit (Table 2). Cyclic AMP prevents inhibition by the wild-type and the ATE proteins (confirming that the ATE protein is indeed able to bind CAMP, albeit at a greatly reduced level relative to the unmutated protein when assayed in vitro), but has no effect on the repression of catalytic activity by the AEBE protein. This is expected, because the AEBE protein is unable to bind CAMP in vitro. Unexpectedly, however, the two substitutions within BSA render mutant TABLE 2

RESULTS

Mutants of the Dictyostelium R Subunit Based on sequence homology with mammalian PKA proteins, two potential CAMP binding sites can be identified in the Dictyostelium R subunit protein (Mutzel et al., 1987). We refer to them as the A and B sites. Using site-directed mutagenesis, individual amino acid residues within each putative binding site were altered to create four mutant R subunit cDNAs. The names of the mutants indicate the site of alteration and the nature of the substitution. Thus the ATE mutant has a Ser-Thr and a Gly-Glu substitution within the A binding site and AEBE has Gly-Glu substitutions within both the A and the B binding sites (Table 1). The mutant R subunit proteins were synthesized in E. coli (Simon et al., 1988) and total cell extracts were assayed to establish the ability of each mutant protein to bind CAMP and to inhibit the C subunit (Table 2). These two biochemical properties are altered to varying de-

Activity of C subunit” Extract” pIMS5“ Wild-type R subunit ATE BSA AEB E ATEBSA

CAMP binding” 0.1 23 0.9 10 0.27 0.13

f f f f f f

0.00 2.0 0.09 1.0 0.05 0.09

CAMP-

CAMP+

100 18 27 81 31 100

100 100 86 100 25 100

a Crude protein extract from transformed E. co&. * PHICAMP binding (pmol/mg protein). ’ Partially purified Dictyostelium C subunit is used as a source of kinase activity. Extracts from R subunit expressing E. coli transformants are added in either the presence or the absence of 1 mMcAMP. The resulting values are normalised to the figure obtained from the extract alone; consequently 100% indicates that the extract does not inhibit the C subunit. ‘Extract from E. coli transformed with plasmid pIMS5, not expressing R subunit protein.

HARWOOD ET AL. actln 15 promoter 8. inltlatlon COdOIl

R subunit

Multiple actin 15 terminator

cDNA

93

Roles for CAMP actin

6-neo

R

FIG. 1. The structure of the plasmid pA15Rsub. The Dictyostelium R subunit cDNA is expressed from the actin 15 promoter and utilises the actin 15 initiation codon. As a consequence the protein contains 12 N-terminal o&in 15 residues in place of the 4 N-terminal R subunit residues. Translation termination occurs at the R subunit termination codon, while the mRNA processing sequences are derived from the actin 15 gene. The presence of the neo marker expressed from the actin 6 promoter allows selection for introduction of the plasmid into Lktyostelium cells. The Dictyostelium R subunit contains a region of interaction with the catalytic site of the C subunit (I) and two putative CAMP binding sites (sites A and B). When the glycine residues in other species equivalent to those at positions 135 and 261 are mutated CAMP binding is abolished (Clegg et al, 1987; Kuno et al, 1988).

proteins containing this altered CAMP binding site (BSA and ATEBSA) incapable of inhibiting the C subunit. We have used the BSA mutant, which retains the ability to bind CAMP but has lost the ability to interact with the C subunit, to dissect the mechanism whereby overexpression of the R subunit blocks development. @m-expression of the R Subunit Inhibition of the C Subunit

Blocks Development

by

The wild-type R subunit and the various mutants were fused in frame to the actin 15 gene (Fig. 1). The actin 15 promoter directs gene expression during growth and early development (Cohen et al, 1986). In agreement with previous results (Simon et ak, 1989), overexpression of the unmutated R subunit arrests development prior to aggregation (Table 3). Only those mutant R subunits that retain the ability to interact with the C subunit act to block development. The number of clones that exhibited the blocked phenotype could be increased by further selection to 100 fig/ml G418 (Table 3). This is most likely due to an increase in plasmid copy number resulting in increased expression of R subunit protein (Simon et ab, 1989). All clones expressing the BSA or the

ATEBSA cDNA exhibited normal development, even after selection on 100 pg/ml of G418 cells. We refer to the cells where PKA is inhibited and development is blocked as PKA- cells. In the experiments that follow we compare PKA- cells generated using the mutants ATE, which binds to the C subunit in a CAMPreversible manner, and AEBE, which cannot bind CAMP. The mutant BSA, which does not bind the subunit and hence does not block development, is used as a control. Inhibitim of PKA Prevents Chernotaxis

CAMP Relay but Not

Extracellular CAMP release was measured after stimulation with X-deoxy-CAMP, a CAMP analogue that stimulates the receptor but does not interfere with the subsequent assay of CAMP. Control BSA cells released a pulse of CAMP, but no CAMP pulse was generated in PKA- ceils of either mutant type (Fig. 2). The ability to sense and move up a CAMP concentration gradient was determined by a “spreading assay,” in which cells are

TABLE 3 Percentage of clones’ blocked in development” Mutant Unmutated Mutant ATE Mutant BSA Mutant AEBE Mutant ATEBSA

After initial selection 33 47 0 87.5 0

(n = (72 = (n = (n = (72 =

9) 15) 21) 8) 21)

After growth in 100 wg/ml G 41Sb 100 0 84 0

(n (n (n (n

= = = =

5) 4) 6) 5)

a Cells (107) were developed on nitrocellulose filters. Clones were generated by two independent transformations and each clone was tested three times (n = total number tested). b Clones were grown in 100 pg/ml G418 and then tested as in footnote a (see text).

00 0

2 4 6 Time (mins) after addition of 2’ deoxy-CAMP

8

FIG. 2. The rise in extracellular CAMP after stimulation with 2’deoxy-CAMP. Each curve represents the results of three independent clones normalised to the mean CAMP level at t = 0 (error bars indicate the SEM). A CAMP pulse is generated by the control mutant Bs* (a), but not by the PKA- cells ATE (0) or AeBa (0).

94

DEVELOPMENTALBIOLOGY

placed on an agar plate containing CAMP (Bonner et al., 1966). In this assay chemokinesis (increased random movement of amoebae) generates a diffuse disc of cells, but chemotaxis results in a ring of amoebae expanding away from the outermost edge of the disc. This is due to directed movement away from the centre, where CAMP becomes depleted by enzymatic activity. Clear outward migration was observed for both the control and the two classes of PKA- cells (Fig. 3). Synergy of PKA-

and

Wild-Type

Cells

The above results suggest that PKA- cells can respond to, but not generate, CAMP signals. In confirmation of this, we find that PKA- cells synergise with nontransformed cells to form a multicellular aggregate. Mixtures containing more than 50% of wild-type cells readily form mature culminants (data not shown). By tracing the fate of the mutant cells we show that they differ radically from wild-type cells in their behaviour after the completion of aggregation. The plasmid pActl5Gal contains the E. coli LczcZgene expressed from the Dictyostelium actin 15 promoter. Cells that express the LacZ gene can be detected by staining with X-Gal, and the fi-galactosidase enzyme persists throughout development (Dingermann et al, 1989). Cotransformants, containing the pActl5Gal plasmid, and mutant R subunits were generated. These were used to examine the fate of PKA- cells in mixtures with untransformed AX2 cells. During initial development, individual amoebae migrate separately towards the CAMP signalling centre but, later in aggregation, cells come into contact to form multicellular streams. PKA- cells enter multicellular streams apparently normally (Fig. 4d) but, at the later tipped aggregate and first finger stages, they become localised to a ring around the base (Fig. 4e). They remain in this position throughout subsequent development, hence PKA- cells are never found in the mature spore and stalk cells but can be found in the basal disc. If a synergised aggregate forms a migratory slug, then PKA- cells display a more extreme form of this same behaviour. In control slugs, containing BSA cells cotransformed with pActl5Ga1, stained cells were dispersed throughout the entire structure. In contrast, in slugs containing marked PKA- cells, staining cells were almost exclusively localised to the posterior region and many stained cells were present in the slime trail that is deposited behind the slug (Figs. 4c and 4f).

VOLUME

l&,1992

aggregate when provided with an exogenous CAMP signalling source, they are in some way defective in later, multicellular development. We have analyzed the expression of late gene products using the same approach of allowing PKA- cells to synergise with wild-type cells. We first analyzed the expression of cysteine proteinase 2 (CP2), a CAMP-inducible, non-cell-type-specific gene that is expressed at the end of cellular aggregation (Pears et ab, 1985). Cotransformants were made that contain various mutant forms of the R subunit and a luciferase reporter gene expressed from the CP2 promoter. Levels of luciferase expression in control cells after 8 hr of development were 170-fold higher than those in vegetative cells, while PKA- cells showed only a 1.6-fold increase during this time period (Table 4a). The CP2 gene is induced in cell suspension by exogenous CAMP and this response was tested for PKA- cells. Cyclic AMP induced a 250-fold elevation of luciferase activity in control cells but, again, PKA- cells exhibited only a tiny induction (Table 4b). The low-level accumulation observed in PKA- cells for both the CP2 gene and the coordinately expressed cysteine proteinase 1 gene, as seen in a previous study (Simon et al., 1989), may reflect a low level of expression in all cells. Alternatively, it could reflect the presence of a small number of cells that have a low copy number of the R subunit and so escape repression. Finally, we investigated whether PKA- cells that had entered an aggregate were competent to express celltype-specific genes. The psA, ecmA, and ecmB genes are expressed in the precursors of spore and stalk cells and are induced by different extracellular signals. The psA (formerly called the D19) gene is expressed in prespore cells and is induced by CAMP (Barklis and Lodish, 1983); Early et ah, 1988). The ecmA (formerly called the pDd63) gene is expressed in prestalk cells and is inducible by DIF (Jermyn et al, 1987). The ecmB (formerly called the pDd56) gene is expressed in a subpopulation of the prestalk cells (Jermyn et ak, 1989). Its expression is also induced by DIF but is repressed by CAMP (Jermyn et al, 1987; Berks and Kay, 1988,199O). Constructs containing the promoter of each gene fused to LacZ were cotransformed into cells with the mutant R subunit genes. The resulting marked strains were synergised with AX2 cells for 15 hr and @galactoside expression was measured enzymatically. The PKAcells did not express any of the three genes at a detectable level (Fig. 5). DISCUSSION

Late Developmental Cells

Genes Are Not Expressed in PKA-

Taken in conjunction, the above observations suggest that, although PKA- cells are competent to enter the

While the protein sequence of the Dictyostelium R subunit indicates the presence of two potential CAMP binding sites, kinetic analysis suggests there is only one (Arents and van Driel, 1982; Majerfeld et al., 1984; de Gunz-

HARWOOD ET AL,

CAMP -

Multiple Roles for CAMP

95

CAMP +

A

on t-t

‘K A-

FIG. 3. Nontransformed AX2 cells, mutant control cells, mutant BSA, and PKA- cells, in this case ATE, all undergo chemotaxis in response to CAMP by forming an expanding ring of cells. The uneven appearance of the disc of cells seen with nontransformed and control cells is due to cellular aggregation, a process which is blocked in PKA- cells.

96

DEVELOPMENTAL BIOLOGY

VOLUME 142,1992

FIG. 4. Development of Dict~ostelium cotransformants synergised with an equal number of nontransformed AX2 cells. Each clone contains pActl5gal in addition to the control (mutant BSA;a, c, e) or the PKA- construct (in this case, mutant A “; b, d, f). Photographs show representative whole mounts stained with X-gal (see text).

burg et ah, 1984). Mutation in either the A or the B site reduces, but does not completely eliminate, CAMP binding, showing there to be two sites. The failure to detect them kinetically may indicate that the two sites differ greatly in their affinities for CAMP and this possibility is currently being investigated.

As with the R subunit proteins of other species (Clegg et aZ., 1987; Kuno et aZ., 1988), a single Gly-Glu substitution within each site eliminates CAMP binding without affecting inhibition of the C subunit. Unexpectedly, the two amino acid substitutions (Phe,-Ser and Gly,,Ala) in the B site prevent inhibition of the C subunit.

HARWOOD ET AL.. TABLE

4

Percentage increase in CP2-luciferase expression over the levels in vegetative cells (mean + SEM) (4 Smw PKAControl (BsA mutant) (b) CAMP induction PKACAMPCAMP+ Control (BsA mutant) CAMPCAMP+

1.6 + 0.2 170. + 46 1.16 + 0.11 2.00 * 0.14 2.18 + 0.35 252 + 50

This is not due to overall degradation of the BSA R subunit protein in E. coli, because extracts retain the ability to bind CAMP at the A site and, in both E. coli and Dictyostelium cells, full-length protein is detectable by Western analysis (data not shown). A more likely explanation is that the mutant protein is unable to interact with the C subunit. This may be due to an overall destabilisation of the holoenzyme. The hinge region at the N-terminus of the protein is generally considered to be responsible for the R/C interaction; however, it has been shown recently that residues situated outside this region are also required for the formation of the holoenzyme in the absence of CAMP (Wang et ah, 1991). Calculation indicates that, using the actin 15 promoter, sufficient R subunit protein is produced to bind most or all of the cytosolic CAMP (Simon et ah, 1989). This suggests two potential explanations for the block to development. The excess R subunit (i.e., the protein that remains uncomplexed with CAMP) might bind to and inactivate the C subunit. Alternatively, the block could be the direct result of CAMP depletion by some, as yet uncharacterised mechanism. The fact that cells expressing the ATE and AEBE mutants, which inhibit the C subunit but bind little CAMP, are blocked in development while those expressing the BSA mutant, which does bind CAMP but cannot inhibit the C subunit, develop normally strongly supports the former hypothesis. The failure of PKA- cells to aggregate is most likely due to their inability to relay CAMP signals. We do not know the reason for the CAMP signalling defect. It may be due to a reduction in CAMP production, to feedback control by the kinase of components of the CAMP signalling system, or to inhibition of CAMP secretion. PKAcells can undergo chemotaxis towards CAMP and may aggregate when mixed with wild-type cells by restoration of the CAMP signal. In apparent contradiction, a previous study led to the opposite conclusions (Firtel and Chapman, 1990). This difference may be due either

Multiple

Roles for

97

CAMP

to the use of a heterologous protein, the previous study utilised a mutant form of the murine R subunit, or to differences in the way the cellular properties were assayed. While PKA- cells are able to aggregate and enter streams apparently normally, they remain confined to the periphery of the tipped aggregate and are selectively lost from the migratory slug. Group 2 Agip mutants can form an aggregate only when mixed with wild-type cells or exposed to CAMP pulses, but cannot complete development (Darmon et aL, 1977). When synergised with wildtype cells, these mutants do not form mature spores and undifferentiated cells are selectively localized in the base of the aggregate. Because of these parallels with PKA- cells, it would be interesting to examine these mutants for lesions in the CAMP second messenger pathway. At present we can only speculate as to the cause of this aberrant developmental behaviour of PKA- cells. Even when synergised with wild-type cells, they show only a very low level of expression of the CP2 gene and the cell-type-specific pspA, ecmA, and ecmB genes are totally inactive. This radical defect in gene expression may be the cause of their eventual exclusion from the multicellular aggregate, perhaps because they fail to accumulate a cellular component required for chemotaxis within, or a cell adhesion molecule required to remain within, the tipped aggregate. We cannot conclude that PKA is directly involved in induction of any of the four

pspA II

ecmA

Control

ecmB

pspA

ecmA

ecmB

PKA-

FIG. 5. Cotransformants, which in addition to the mutant BsA (control) or ATE mutant R subunit gene (PKA-) contain the h’. coli La& gene fused to the pspA, the ecmA, or the ecmB gene promoter, are synergised with an equal number of nontransformed AX2 cells. Samples were assayed for fl-galactosidase activity after development for 15 hr. Each assay was carried out in duplicate for three independent clones calculated. The mean enzymatic activity is plotted as fold increase, i.e., the multiple with which it is increased over the levels in vegetative cells (error bars indicate the SEM).

98

DEVELOPMENTAL BIOLOGY

genes. because inactivation of it mav Dreclude subse&en; gene activation simply by bloc&g development. The AEBE mutant gene provides a potent inhibitor of the C subunit of PKA and should greatly facilitate further molecular genetic approaches to investigating the role of PKA. We thank P. Skehel for his numerous and interesting discussions concerning this work. This work was supported by grants from CNRS (UA1129) and the Association de Recherche contre le Cancer (ARC 6438).

REFERENCES ARENTS, J., and VAN DRIEL, R. (1982). Soluble, cytoplasmic cyclic AMP binding proteins of Dictyostelium discoideum. FEBS L&t. 137,201204.

BAR&IS, E., and LODISH, H. F. (1983). Regulation of Dictyostelium discoideum mRNAs specific for prespore or prestalk cells. Cell 32, 1139-1148. BERKS, M., and KAY, R. R. (1988). Cyclic AMP is an inhibitor of stalk cell differentiation in Dictyostelium discoideum. Dev. BioL 126,108114. BERKS, M., and KAY, R. R. (1990). Combinatorial control of cell differentiation by CAMP and DIF-1 during development of Dictyostelium discddeum. Development 110,977-984. BONNER,J., KELSO, A., and GILLMOR, R. (1966). A new approach to the problem of aggregation in the cellular slime molds. BioL Bull. 130, 28-42.

BRADFORD,M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein using the principle of protein-dye binding. Anal. Biochem. 72,248-254. CLEGG, C., CORRELL, L., CADD, G., and MCKNIGHT, G. (1987). Inhibition of intracellular CAMP-dependent protein kinase using mutant genes of the regulatory type I. J. Biochem. Chem. 262,13,111-13,119. COHEN, S. M., KNECHT, D., LODISH, H. F., and LOOMIS, W. F. (1986). DNA sequences required for expression of a Dictyostelium actin gene. EMBO J. 5,3361-3366. DARMON, M., BARRAND, P., BRACHET, P., KLEIN, C., and PEREIRA, D. S. L. (1977). Phenotypic suppression of morphogenetic mutants of Dietyostelium

discoidearn.

Dev.

BioL 58,174-184.

DE GIJNZBURG,J., and VERON, M. (1982). A CAMP-dependent proteinkinase is present in differentiating Dictyostelium discoideum cells. EMBO J. 1,1063-1068. DE GLJNZBURG,J., PART, D., GUISO, N., and VERON, M. (1984). An unusual adenosine 3’5’-phosphate dependent protein kinase from Dietyostelium

disco&urn.

Biochemistry

23,3805-3812.

DE GIJNZBURG,J., FRANKE, J., KESSIN, R., and VERON, M. (1986). Detection and developmental regulation of the messenger RNA for the regulatory subunit of the CAMP-dependent protein-kinase of Dietyostelium diswideum cells by cell free translation. EMBOJ. 5,363367.

EARLY, A. E., WILLIAMS, J. G., MEYER, H. E., POR, S. B., SMITH, E., WILLIAMS, K. L., and GOOLEY, A. A. (1988). Structural characterization of Dictyostelium disundeum prespore-specific gene D19 and of its product, cell surface glycoprotein PsA. Mol. CeU. BioL 8, 345% 3466.

FIRTEL, R. A., VAN HAASTERT, P. J. M., KIMMEL, R. A., and DEVREOTES, P. N. (1989). G protein linked signal transduction pathways in development: Dictyostelium as an experimental system. CeU 58,253-259. FIRTEL, R. A., and CHAPMAN, A. L. (1990). A role for CAMP-dependent protein kinase A in early Dictyostelium development. Genes Dev. 4, 18-28. FLOCKHART, D., and CORBIN,J. (1982). Regulatory mechanisms in the control of protein kinases. @it. Rev. Biochem. 12,133-186. HARWOOD, A. J., and DRURY, L. (1990). New vectors for expression of the E. eoli 1acZ gene in Dictyostelium. Nucleic Acids Res. l&4292. JANSSENS, P., and VAN HAASTERT, P. (1987). Molecular basis of transmembrane signal transduction in Dictyostelium disc&teum. Micre &oL Rev. 51,396-418. JERMYN, K., BERKS, M., KAY, R., and WILLIAMS, J. (1987). Two distinct classes of prestalk-enriched messenger RNA sequences in Dictyostelium discoideum Development 100,745-755. JERMYN K. A., DIJFFY, K., and WILLIAMS, J. G. (1989). A new anatomy of the prestalk zone of Dictyostelium. Nature 340,144-146. JERMYN, K., and WILLIAMS, J. (1991). An analysis of culmination in Dietyostelium using prestalk and stalk specific cell autonomous markers. Development 111,779-787. KESSIN, R. (1988). Genetics of early Dictyostelium discoideum development. MicrobioL Rev. 52,29-49. KONIJN, T. M., VAN DE MEENE, J. G. C., BONNER, J. T., and BARKLEY, D. S. (1967). The acrasin activity of adenosine-3’,5’-cyclic phosphate. Proc NatL Acad Sci. USA 58,1152-1154. KIJNO, T., SHIJNTOH, H., MOTOYOSHI, S., KIYOFUMI, S., TAKEDA, T., FUKUDA, K., and TANAKA, C. (1988). Site-directed mutagenesis of the CAMP-binding sites of the recombinant type I regulatory subunit of CAMP-dependent protein kinase. B&hem. Biophys. Res. Commun. 153,1244-1250. LEICHTLING, B., MAJERFELQ I., COFFMAN, D., and RICKENE~ERG,H. (1982). Identification of the regulatory subunit of a CAMP dependent protein kinase in Dictyostelium discoideum Biochem. Biophys. Res. Commun.

discwideum

Gene 59,99-106.

105,949-955.

LEICHTLING, B., MAJERFELD, I., SPITZ, E., SCHALLER, K., WOFFENDIN, C., KAKINIJMA, S., and RICKENBERG, H. (1984). A cytosolic cyclic AMP dependent protein kinase in Dictyostelium diswideum. 2. Developmental regulation. J. BioL Chem. 259,662-668. MAJERFELD, I., LEICHTLING, B., MELIGENI, J., SPITZ, E., and RICKENBERG, H. (1984). A cytosolic cyclic AMP-dependent protein kinase in Dictyostelium diseoideum 1. Properties. J. BioL Chem. 259,654-661. MANN, S. K., PINKO, C., and FIRTEL, R. A. (1988). Control of early gene expression in Dictyostelium. Dev. Genet. 9,337-350. MILLER, J. (1972). Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. MUTZEL, R., LACOMBE, M., SIMON, M., DEGUNZBIJRG,J., and VERON, M. (1987). Cloning and cDNA sequence of the regulatory subunit of CAMP dependent protein-kinase from Dictyostelium discoideum. Proc.

DINGERMANN, T., REINDL, N., WERNER, H., HILDEBRANDT, M., NELLEN, W., HARWOOD, A., WILLIAMS, J., and NERKE, K. (1989). Optimization and in situ detection of Escherichia coli @-galactosidase gene expression in Dictyostelium disu%!eum. Gene 85,353-362. EARLY, A., and WILLIAMS, J. (1987). Two vectors which facilitate gene manipulation and a simplified transformation procedure for Dietyostelium

VOLUME 149. I992

NatL

Acad.

Sci. USA 84.6-10.

NELLEN, W., SILAN, C., and FIRTEL, R. (1984). DNA-mediated transformation in Dictyostelium disco&urn-regulated expression of an actin gene fusion. MoL CelL BioL 4,2890-2898. NELLEN, W., and FIRTEL, R. A. (1985). High-copy-number transformants and co-transformation in Dictyostelium. Gene 39, 155-163. PEARS, C. J., MAHBUBANI, H., and WILLIAMS, J. G. (1985). Characterization of two highly diverged but developmentally co-regulated cys-

HARWOOD ET AL. teine proteinase genes in Dictyostelium diswideum Nucleic Acids 13,8853-8866. RUTHERFORD, C. L., TAYLOR, R. D., FRAME, L. T., and AUCH, R. L. (1982). A cyclic AMP dependent protein kinase in Dictyostelium discoideum. B&hem. Biophys. Res. Commun. 108,1210-1220. SIMON, M. N., MUTZJZL,R., MUTZEL, H., and VERON, M. (1988). Vectors for expression of truncated coding sequences in Eschwichia coli. Plasmid 19,94-102. SIMON, M. N., DRISCOL, D. M., MUTZEL, R., PART, D., and WILLIAMS, J. G. (1989). Overproduction of the regulatory subunit of the CAMPdependent protein kinase blocks the differentiation of DictyosteRes.

lium

discaideum.

EMBO

J. 8,2039-2043.

SNAAR-JAGALSKA, B. E., KESBEKE, F., PUPILLO, M., and VAN HAASTERT, P. J. (1988). Immunological detection of G-protein a-subunits

Multipk

Roles for

CAMP

99

in Dictyostelium discoideum. Biochem. Biophys. Res. Commun. 156, 757-761. VAN HAASTERT, P. (1984). A method for studying CAMP-relay in Dietyostelium discoideum: The effect of temperature on CAMP-relay. J. Gen. Microbial 130,2559-2564. WANG, Y., Scorr, J. D., MCKNIGHT, S. G., and KREBS, E. G. (1991). A constitutively active holoenzyme form of the CAMP-dependent protein kinase. Proc NatL Acud Ski. USA 88.2446-2450. WATT& D. J., and ASHWORTH, J. M. (1970). Growth of myxameobae of the cellular slime mould Dictyostelium discoideum in axenic culture. Biochem. J. 119,171-174. WILLIAMS, J. G. (1988). The role of diffusible molecules in regulating the cellular differentiation of Dictyostelium discoideum Develop me-at 103,1-16.