Ammonium Sulfate Modifies Adenylate Cyclase and the Chemotactic Receptor of Dictyosteliurn discoideurn EVIDENCE FOR A G PROTEIN EFFECT*

(NH4),S04 was found to activate adenylate cyclase in Dictyostelium discoideum membranes. The effect of (NH4),S04 on the enzyme was observed after pretreat- ment of membranes but could not be observed if the salt was added to the assay mixture. Activation was seen when membranes were pretreated with 0.16 M (NH4),S04 and was maximal at 0.6-1.0 M. The maximal activation of the enzyme was observed within 3 min of pretreatment and was not readily reversible. The effect was specific for the N&+ ion since pretreat- ment of membranes with other NHt salts could activate the enzyme, whereas pretreatment with NaCl or KC1 could not. Pretreatment of plasma membranes with (NH4),S04 eliminated the sensitivity of the enzyme to the inhibitory effect of guanine nucleotides. (NH4),S04 pretreatment also significantly attenuated the inhibi- tion by guanine nucleotides of cAMP binding to its plasma membrane receptor. The effect of (NH4),S04 on GTP inhibition of cAMP binding to its receptor was even more dramatic when the salt was present in the binding assay. (NH4),S04 also increased the ADP-ri-bosylation by cholera toxin of a 39,000-Da membrane protein. The data support the hypothesis that (NH,),SO,-induced changes in adenylate

(NH4),S04 was found to activate adenylate cyclase in Dictyostelium discoideum membranes. The effect of (NH4),S04 on the enzyme was observed after pretreatment of membranes but could not be observed if the salt was added to the assay mixture. Activation was seen when membranes were pretreated with 0.16 M (NH4),S04 and was maximal at 0.6-1.0 M. The maximal activation of the enzyme was observed within 3 min of pretreatment and was not readily reversible.
The effect was specific for the N&+ ion since pretreatment of membranes with other NHt salts could activate the enzyme, whereas pretreatment with NaCl or KC1 could not. Pretreatment of plasma membranes with (NH4),S04 eliminated the sensitivity of the enzyme to the inhibitory effect of guanine nucleotides. (NH4),S04 pretreatment also significantly attenuated the inhibition by guanine nucleotides of cAMP binding to its plasma membrane receptor. The effect of (NH4),S04 on GTP inhibition of cAMP binding to its receptor was even more dramatic when the salt was present in the binding assay. (NH4),S04 also increased the ADP-ribosylation by cholera toxin of a 39,000-Da membrane protein. The data support the hypothesis that (NH,),SO,-induced changes in adenylate cyclase and the cAMP receptor are due to an alteration of a putative G protein.
The mechanism(s) of regulation of Dictyostelium discoideum adenylate cyclase, the enzyme responsible for synthesis of the chemoattractant for this organism, remains an intriguing subject of investigation. Rhythmic changes in adenylate cyclase activity underlie the production of cAMP in the form of 5-6-min pulses (1, 2). The newly synthesized cAMP is rapidly released into the medium. When cells are stimulated with extracellular CAMP, they respond with a brief activation of adenylate cyclase and are then transiently desensitized to further stimulation (3). It is not clear how the chemotactic receptor is coupled to adenylate cyclase. In higher eukaryotes, regulatory G proteins ( Gi and G8)' are involved in transducing * This work was supported by National Science Foundation Grant DCB 86-03554 (to C. K.) and National Institutes of Health Grant NS 16513 (to A. H.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 11 To whom correspondence should be addressed E. A. Doisy Dept. of Biochemistry, St. Louis University School of Medicine, 1402 S. Grant Blvd., St. Louis, MO 63104.
Our recent studies (5) have shown that guanine nucleotides inhibit D. discoideum adenylate cyclase. This was the first indication that guanine nucleotides may play a role in the regulation of this enzyme. This inhibition of guanine nucleotides was noncompetitive with respect to substrate, supporting the possibility that a regulatory G protein may constitute part of the adenylate cyclase complex. However, some of the characteristics of nucleotide-mediated inhibition of adenylate cyclase were unusual compared with those reported for mammalian systems. For example, a greater inhibition of the D. discoideum enzyme by guanine nucleotides was observed in the presence of Mn2+. If a protein comparable to a regulatory G protein does exist as part of the D. discoideum adenylate cyclase complex, it possesses some unique features.
In the course of experiments designed to examine the regulation of adenylate cyclase and its coupling to the chemotactic receptor, we observed that pretreatment of membranes with (NHJ2SO4 activates the enzyme. The experiments reported here provide strong evidence that this activation reflects a n altered interaction of both the adenylate cyclase and the chemotactic receptor with a regulatory G protein(s).

EXPERIMENTAL PROCEDURES
Plasma Membrane Preparations-D. discoideum Ax-2 cells were grown in HL5 medium, harvested, and starved as spinner suspensions in 20 mM MES, pH 8.0, containing 2 mM MgClZ at a density 1 X lo7 cells/ml (6,7). Aggregation-competent cells were washed twice with 10 mM Tris-HC1, pH 8.0, and resuspended at a density of 1 X 10s cells/ml in buffer A (30% sucrose, 50 mM Tris-HC1, pH 8.0, 1 mM EDTA, 1 mM dithiothreitol, and 15 mM MgCl,). The cell resuspension was passed through a polycarbonate filter (25-mm diameter, 5-pm pore size) (&lo). The lysate was centrifuged at 30,000 X g for 15 min, and the supernatant was discarded. Only the top fluffy layer of the pellet (Pz) was collected. This fraction, which consisted primarily of plasma membranes, was washed in buffer A and resuspended to a protein concentration of 2.5-3 mg/ml in buffer B (10 mM Tris-HC1, pH 8.0,4 mM dithiothreitol, 30% sucrose). The characteristics of this membrane preparation are described in more detail elsewhere (8)(9)(10).
Pretreatment of Membranes with (NHJzS04-Unless indicated otherwise, the P, plasma membrane fraction was incubated on ice with the desired concentration of (NH&S04 for 20 min. The fraction was diluted six times with buffer A and centrifuged at 15,000 X g. The pellet was resuspended in 7.5 volumes of buffer B and immediately assayed for adenylate cyclase activity. Control Pz was incubated for the same period of time without additional salt and then treated the same way as the experimental sample. Adenylate Cyclase Assay-Unless indicated otherwise, enzyme activity was determined in a reaction mixture containing 25 mM Tris-HC1, pH 8 contained five protease inhibitors: 2 pM leupeptin, 7 p M antipain, 30 p~ aprotinin, 10 p~ w-p-tosyl-L-lysine chloromethyl ketone, and 300 p~ phenylmethylsulfonyl fluoride. Membranes were incubated for 20 min at 0 "C with or without (NH4)~SO4 and then washed with buffer B containing protease inhibitors. Plasma membranes were incubated for 30 min at 27 "C in the presence or absence of 10 pg/ml cholera toxin which had been activated with 33 mM dithiothreitol for 10 min at 37 "C. The incubation mixture consisted of 25 mM Tris-HCl, pH 7.2, 1 mM ATP, 10 mM thymidine, 1 mM EDTA, 150 p~ GTP, 2 mM MgC12, 3 mM phosphoenolpyruvate, 10 pg/ml pyruvate kinase, 8 p~ [32P]NAD (10 Ci/mmol), and 200 units/ml DNase. The reaction was stopped by the addition of 1 ml of ice-cold 50 mM Tris-HCl, pH 8.0, containing the protease inhibitors described previously. Membranes were pelleted by a 10-min centrifugation at 15,000 X g. Pellets were solubilized in 100 pl of sample buffer (1% SDS, 20 mM NazC03, 28 mM dithiothreitol, 80 mg/ml bromphenol blue, 10% sucrose), heated at 100 "C for 1 min, and analyzed by SDS-polyacrylamide gel electrophoresis according to Laemmli (13). After staining (Coomassie Blue) and destaining, gels were dried, and autoradiography was performed at -70 "C using XR5 film and Du Pont-New England Nuclear Lightning Plus intensifying screens. Densitometer tracings of autoradiograms were obtained using an LKB laser densitometer.
CAMP Binding Actiuity-Plasma membranes were incubated with 5 nM I3H]cAMP in 20 mM phosphate buffer, (pH 6.4), containing 10 mM dithiothreitol in a final volume of 400 pl. Additions are indicated in the table legends. After 30 s (the time of maximum binding), 200pl aliquots were vacuum-filtered onto 0.2-pm polycarbonate filters. Radioactivity bound to the filters was determined by liquid scintillation counting. Duplicate filters agreed within values of 5%. Nonspecific binding was determined by performing the assay in the presence of 10 ptd nonradioactive CAMP. Such values were less than 0.5% of the total radioactivity. Binding activity was linear with protein concentrations ranging from 0.5 to 20 pg. The experiments are representative of at least two experiments.
Materiaki-[c~~~P]ATP, [3H]cAMP, and [32P]NAD were purchased from ICN Pharmaceuticals or Du Pont-New England Nuclear. ATP, creatine phosphate, and creatine phosphokinase were purchased from Sigma. Cholera toxin was purchased from List Biochemicals. Other chemicals were from Behring Diagnostics. Polycarbonate filters were purchased from Bio-Rad.

Activation of Plasma Membrane Adenylate
Cyclase by (NH4)2S04-(NH4)2S0, could activate adenylate cyclase when plasma membranes (P2) were pretreated with the salt (Fig. 1). Activation of adenylate cyclase activity was observable with 0.16 M (NH4),S04 and was maximal at 0.6-1.0 M. In this experiment, a 4.5-fold activation was observed with the higher concentration of salt. The maximum degree of FIG. 1. Dose-dependent activation of adenylate cyclase by (NH4)&304. Plasma membranes were preincubated for 20 min with the indicated concentrations of (NHd2S04, and adenylate cyclase was assayed after membranes were washed. enzyme activation did vary in different experiments between 3-and 10-fold. The variability probably reflects the fact that new preparations of plasma membranes were used in each experiment. The activation of the enzyme was not transitory in that adenylate cyclase activity remained linear with increasing times of incubation in the assay mixture (Fig. 2). Thus, this activation is different from the transitory activation of the enzyme observed when cells are stimulated with CAMP (1,2).
The activation of adenylate cyclase by (NH4),S04 could not be attributed to the presence of residual (NH,),SO, in the assay. A 6-fold dilution of the membrane fraction prior to its assay did not change the degree of activation observed (data not shown). Also, the maximal concentration of (NH4)&304 which could remain was calculated to be 5 mM. This concentration was ineffective when added to the assay mixture. In contrast to the results observed upon pretreatment, no activation of adenylate cyclase was observed when the salt was present in the incubation mixture. Higher concentrations of (NH4)2S04 (100-200 mM) inhibited enzyme activity by 20-65% (data not shown). Comparable inhibition of membrane and solubilized brain adenylate cyclase activity has been observed by Neer and Salter (14) when such concentrations of (NH4),S04 were present in their assay mixture. It is probable that (NH4),S04 has an inhibitory effect on the catalytic protein since both Mn2+-activated and basal activities are affected.
The length of time plasma membranes needed to be pretreated with (NH4),S04 in order to observe an activation of adenylate cyclase was examined (Fig. 3). Activation of adenylate cyclase occurred rapidly. Within 3 min of pretreatment, plasma membranes showed increased enzyme activity. The degree of activation did not change with longer times of pretreatment.
The ion involved in activating adenylate cyclase was examined by pretreating plasma membranes with a variety of salts. Plasma membranes were incubated on ice for 20 min with 0.64 M (NH4)2S04, NH4HC03, NH4CH3C02, NH,Cl, Na2S04, NaHC03, NaCl, or KC1 (Fig. 4). All ammonium salts activated the adenylate cyclase. (NH4),S04 and NH4HCOB  were equipotent in their action and activated adenylate cyclase in this experiment approximately 3-4-fold. Ammonium acetate was less potent, resulting in a 2-fold activation. Slight or no effect was observed with NaCl or KC1, respectively.
Effect of Divalent Cations on Activation of Adenylate Cyclase by (NH4)2S04-Previously, we have shown that Mn2+ stimulates D. discoideum adenylate cyclase 3-8-fold when present in the assay mixture either as the sole divalent cation or in addition to M$+. This activation appears to reflect a divalent cation regulatory site on the catalytic subunit (15). The effect of Mn'+ was still observed in plasma membranes that had been pretreated with (NH4),S04 ( Table I). The activity measured in untreated membranes was increased in this experiment approximately 3-fold when assayed in the presence of Mn2+. Pretreatment of plasma membranes with (NH,),SO, activated the enzyme approximately 3-fold when assayed in  Effect of Pretreatment on Guanine Nucleotide Sensitivity of Adenylate Cyclase-D. discoideum adenylate cyclase is inhibited by GTP and its analogs when they are present in the assay medium (5). We therefore examined the possibility that (NH4),S04 alters the sensitivity of the enzyme to guanine nucleotides (Table 11). Plasma membranes which had not been preactivated with (NH4),S04 showed a 33% inhibition of adenylate cyclase activity by 100 p~ Gpp(NH)p. As reported earlier, no effect of guanine nucleotides on enzyme activity occurred at 1 p~. The enzyme which had been activated by (NH4),S04 pretreatment of plasma membranes remained unresponsive to low concentrations of guanine nucleotides, showing neither stimulation nor inhibition with 1 p~ Gpp(NH)p. However, activation of the enzyme by (NH4),S04 rendered the activity insensitive to the inhibitory effects of 100 ~L M Gpp(NH)p.
ADP-ribosylation of (NH4),S04-treated Membranes-Bacterial toxin-catalyzed ADP-ribosylation of D. discoideum plasma membranes and the effects of guanine nucleotides on that process have allowed us to identify a putative G protein of 39,000 Da in D. discoideum.' It is possible that the activation of adenylate cyclase and the loss of inhibition by guanine nucleotides induced by (NH4)'S04 reflect an alteration of this protein. Such an alteration was evidenced by monitoring its sensitivity to ADP-ribosylation (Fig. 5). ADP-ribosylation of plasma membranes by cholera toxin identified a band of 39,000 Da. An increased labeling of this protein was evident in membranes that had been preincubated with (NH4)$304 under conditions which lead to an activation of adenylate cyclase. In addition, another band of approximately 52,000 Da was now identifiable by ADP-ribosylation.
Membranes were also ADP-ribosylated with cholera toxin in the absence or presence of different concentrations of (NH4LS04. No effect on ADP-ribosylation was observed with concentrations of salt below 100 mM. Concentrations of (NH4),S04 between 100 and 600 mM progressively increased ADP-ribosylation of the 39,000-Da protein. The effects were specific for ammonium in that salts not containing that ion were ineffective in eliciting an increase in ADP-ribosylation (data not shown).

L. Khachatrian, C. Klein, and A. Howlett, manuscript submitted
for publication.  (NH,),SO, Effects on cAMP Binding to Plasma Membranes-To examine if the activation of adenylate cyclase by (NH,),SO, could be explained by an activation of cAMP receptor binding activity in the plasma membrane, plasma membranes were pretreated with (NH,),SO, under conditions which lead to an activation of adenylate cyclase and were then assayed for cAMP binding activity (Table 111). In this experiment, a slight decrease in cAMP binding activity occurred. In other experiments, no difference was observed between plasma membranes preincubated in the absence or presence of (NH,),SO,. The possibility that (NH,),SO, altered the sensitivity of the cAMP receptor to guanine nucleotides was examined by performing the binding assay in the presence or absence of G T P Binding of cAMP to control plasma membranes was inhibited approximately25% by GTP. The degree of inhibition varied between 22 and 33% in different experiments. (NH,),SO,-pretreated membranes showed significantly less inhibition by GTP, generally on the order of 6-10%. The loss of G T P sensitivity by (NH,),SO4 was also evident when (NH4)'S04 was present in the binding assay. The addition of 1.2 M (NH,),SO, completely eliminated inhibition by GTP. (NH4),S04 at 0.6 M decreased the inhibitory effect of GTP by 40%, whereas 0.15 M (NH,),SO, had no effect on the ability of GTP to inhibit cAMP binding activity (Table 111). At the highest concentration of (NH,),SO, tested (1.4 M), a 30% increase in cAMP binding occurred. When cAMP binding to intact cells is measured in the presence of 3.2 M (NH,),SO,, as high as a 30-fold increase in binding has been reported (16). It is not clear if the small increase in binding to plasma membrane preparations seen here reflects the same process seen with intact cells. The results support the hypothesis that activation of adenylate cyclase by (NH,)ZSO, may reflect an alteration of a G protein as demonstrated by the loss of GTP effects on the cAMP receptor.

DISCUSSION
The data presented here provide new insights into the unique regulatory features of the adenylate cyclase in D.
discoideurn. Pretreatment of plasma membranes with ammonium salts resulted in a dramatic, stable, activation of the enzyme. The data presented strongly support the hypothesis that the activation of adenylate cyclase by (NH,),SO, reflects an altered activity of a regulatory G component. We have previously demonstrated inhibition of adenylate cyclase by guanine nucleotides in a manner consistent with the presence of a coupling G protein (5). The ability of guanine nucleotides to inhibit adenylate cyclase is eliminated by (NH,),SO, pretreatment of membranes. In our system, (NH4),S04 activation of adenylate cyclase could only be demonstrated when plasma membranes were pretreated with the salt since the enzyme was inhibited by 0.1-0.2 M (NH,),SO, when it was present in the assay mixture. Such inhibition would not allow the detection of a possible stimulatory effect elicited by a concurrent event, e.g. alteration of G protein when the salt was present in the assay.
Modulation of a G protein is also consistent with the results of experiments designed to examine the effects of (NH4),S04 on cAMP binding to the chemotactic receptor in plasma membranes. Janssens et al. (17) have reported that guanine nucleotides decrease the level of equilibrium binding of cAMP to membranes. The authors suggest that this reflects a complex formation between the receptor and a guanine nucleotide regulatory protein. We have shown here that when cAMP binding is measured in the presence of (NH,),SO, or in membranes which have been pretreated with (NH4)'S04, the cAMP receptor displays a reduced sensitivity to guanine nucleotides. Thus, (NH,),SO, seems to alter a regulatory G protein as assessed by the inability of guanine nucleotides to inhibit both the cAMP receptor binding and adenylate cyclase.
We have recently identified a candidate for a regulatory G protein by its ability to be ADP-ribosylated by both pertussis and cholera toxins.' The properties of this protein are distinct from G, and Gi, and thus, we refer to it as Gd. Evidence that this 39,000-Da protein, Gd, may indeed confer guanine nucleotide sensitivity to the cAMP receptor and adenylate cyclase was obtained by examining the effects of (NH,),SO, on ADP-ribosylation of this putative G protein. Pretreatment of membranes with (NH,),SO, rendered this protein more susceptible to ADP-ribosylation by cholera toxin. (NH4)$04 present in the ADP-ribosylation mixture also resulted in increased ADP-ribosylation of the G protein.
Other investigators have studied the effects of (NH4)2S04 on the adenylate cyclase system of higher eukaryotic cells. In the studies of Neer and Salter (14, 18), (NH,),SO4 facilitated the activation of detergent-solubilized enzyme by nonhydrolyzable analogs of GTP. Ferguson et al. (19) demonstrated that (NH,),SO, was effective in depleting G proteins of tightly bound GDP such that GTP or its analogs could bind to that site. Facilitated binding of GTP would result in the dissociation of the subunits of G, and the subsequent activation of the enzyme by as. This mechanism is not readily applicable to the case of (NH4)*SO4 activation of D. discoideum adenylate cyclase. This enzyme is not activated by guanine nucleotides when membranes are isolated from cells under normal conditions of starvation and aggregation (5, 15). (NH4)2S04 pretreatment of membranes does not allow the enzyme to be activated by GTP or its analogs. In fact, (NH4),S04 eliminated the inhibitory effects of guanine nucleotides. The unusual characteristics of the D. discoideum system can better be explained if one assumes that (N&)&Od impairs a negative effector of the system.
We would like to propose a mechanism by which (NH4)2S04 activates adenylate cyclase in D. discoideum. The model is based upon the association of the catalytic protein, C, and the inhibitory G protein, Gd (or a subunit should future experiments reveal a multimeric structure). In a manner analogous to mammalian systems where, upon binding GTP, a, is able to interact with C, Gd would associate with C in the presence of GTP. To be consistent with the characteristics of other G proteins, one could assume that Gd possesses a GTPase activity which hydrolyzes bound GTP to GDP. Evidence for such activity is not yet available. We have previously shown that D. discoideum adenylate cyclase exhibits greater activity when assayed in the presence of Mn2+. Guanine nucleotide inhibition of adenylate cyclase is greater when the enzyme is assayed in the presence of Mn2+ (15). This observation can be explained if the C-Gd-guanine nucleotide complex could not be activated by Mn2+. We propose that the association of the Gdguanine nucleotide complex and the catalytic protein is persistent, the two proteins dissociating only upon removal of the guanine nucleotide. In such a schema, (NH4),S04 would alter Gd such that bound nucleotide would be cleared. This would result in dissociation of Gd from the catalytic protein.
To explain the stable activation of the enzyme and loss of GTP inhibition, we suggest that (NH4),S04 limits the ability of Gd to reassociate with the enzyme. Such an effect on Gd by (NH,),S04 would also explain the inability of Gd to interact with the cAMP receptor to cause the guanine nucleotidemediated decrease in binding. The (NHJ2S04-induced dissociation of Gd from C could also explain the observation that the salt stimulates ADP-ribosylation of Gd by cholera toxin. In mammalian systems, the preferred substrate for ADPribosylation by cholera toxin is the isolated a subunit. Gd which is dissociated from either the catalytic protein or the chemotactic receptor may be a better substrate for cholera toxin ADP-ribosylation. Data concerning how a G protein inhibits adenylate cyclase in mammalian cells suggest that its actions are indirect: the GTP-induced dissociation of Gi subunits releases / 3 and y proteins which capture free a, proteins such that a, can no longer activate catalytic activity (20). This model for the mechanism by which Gi mediates the inhibition of adenylate cyclase does not explain how the adenylate cyclase of the S49 cyc-variant cell is inhibited by somatostatin and Gpp(NH)p (21-23). The a, protein, as well as the mRNA that codes for it, is missing from this cell variant (24, 25). Thus, inhibition by Gi in cyc-strains can better be explained by an inhibitory action of ai on the catalytic protein (21). Consistent with a model in which Gi functions to inhibit directly adenylate cyclase are the findings of Katada et al. (26) that ai can inhibit partially purified catalytic subunit activity.
The data presented by Neer and Salter (14) for both membrane-bound and detergent-solubilized adenylate cyclase from brain could also suggest that (NH,),SO, releases an inhibitor from the catalytic protein. In those studies, (NH4),S04 in the assay mixture stimulated both basal and Mn2+-stimulated activity %fold at concentrations below those that inhibited enzymatic activity. Furthermore, the Mn2+-stimulated activity was greater in a solubilized preparation that had been subjected to gel filtration in the presence of (NH&SO+ This would suggest that a stable change in the catalytic activity might result from its separation from another component(s). Also, the apparent size of the catalytic protein, assayed in the presence of Mn2+, decreased in the presence of (NHJ,SO4. If some of the catalytic protein in D. discoideurn does exist in an inhibited state, changes in equilibrium between the free and complexed protein could underlie the production of CAMP pulses. In intact cells, formation of the CAMP-receptor complex results in a rapid activation of the enzyme. The enzyme remains active for a short period, approximately 1 min, and rapidly returns to a low basal level. It is intriguing to consider the possibility that enzyme activation reflects the dissociation of Gd from C and that this dissociation may be facilitated by occupied receptors. This activation would be limited in time, due possibly to rapid desensitization which occurs upon cAMP binding to its receptor (3). The resulting effect would be a rapid, transient activation of adenylate cyclase followed by the accumulation of a population of catalytic proteins associated with inhibitory G d .
Changes in the association of the catalytic protein with an inhibitory G protein may not be the sole mechanism by which adenylate cyclase activity is regulated in D. discoideum. It is also possible that cAMP receptors act more directly to stimulate adenylate cyclase in a manner comparable to GB-mediated effects in mammalian cells. It has recently been reported that lysates, prepared from aggregation-competent cells incubated at 0 "C for 30 min prior to lysis, can show higher adenylate cyclase activity in the presence of guanine nucleotide (27). A component(s) which mediates this activation appears to be cytosolic. These findings suggest that several mechanisms may function to regulate D. discoideum adenylate cyclase.
In summary, we have observed a stable activation of adenylate cyclase by (NH,),SO, that is consistent with the attenuation of a G protein interaction with both the catalytic protein and the cAMP receptor. The proposed mechanism for this activation is based on the concept that the G protein of D. discoideum behaves as an inhibitor of the catalytic activity.
It may be that Gd has unique characteristics that differ significantly from those currently known for G. or Gi, which is suggested by the ability of Gd to be modified by both pertussis toxin and cholera toxin.* Alternatively, the catalytic protein of D. discoideum may have properties that differ from the mammalian protein such that its interaction with G protein(s) is not comparable to that of mammalian cells. Studies are continuing to elucidate further the regulation of this enzyme.