Alteration of Receptor/G-protein Interaction by Putative Endogenous Protein Kinase Activity in Dictyostelium discoideum Membranes

Membranes of Dictyostelium discoideum cells were incubated under phosphorylation conditions and washed, and the effects on cAMP binding to chemotactic receptors in the absence and presence of guanosine 5’-0-(3-thiotriphosphate) (GTPrS) were investigated. Most experiments were done with adenosine 5’-0-(3-thiotriphosphate) (ATPrS), which is a good substrate for many kinases, but the product, protein phosphoro-thioate, is not easily hydrolyzed by phosphatases. Pretreatment of membranes under phosphorylating conditions with MgATPrS alters the site heterogeneity of the CAMP-binding forms, without a significant effect on the total number of binding sites. A similar effect was induced by GTPrS under nonphosphorylation conditions. The effects of MgATPrS were rapid (tlk = 1 min), irreversible, and not induced by M$+ or ATPrS alone or by magnesium adenylyl imidodiphosphate and magnesium adenylyl (8,r-methy1ene)diphosphate. MgATP induced a smaller inhibition than MgATPrS, which was potentiated by the addition of exogenous CAMP-dependent protein kinase. The effect of MgATP was rapidly reversible; reversibility was reduced by the phosphatase inhibitor NaF. These results suggest that the effects of MgATPrS are mediated by an endogenous protein kinase. The major 35S-thiophosphorylated band detected by sodium dodecyl

Membranes of Dictyostelium discoideum cells were incubated under phosphorylation conditions and washed, and the effects on cAMP binding to chemotactic receptors in the absence and presence of guanosine 5'-0-(3-thiotriphosphate) (GTPrS) were investigated.Most experiments were done with adenosine 5'-0-(3thiotriphosphate) (ATPrS), which is a good substrate for many kinases, but the product, protein phosphorothioate, is not easily hydrolyzed by phosphatases.
Pretreatment of membranes under phosphorylating conditions with MgATPrS alters the site heterogeneity of the CAMP-binding forms, without a significant effect on the total number of binding sites.A similar effect was induced by GTPrS under nonphosphorylation conditions.
MgATP induced a smaller inhibition than MgATPrS, which was potentiated by the addition of exogenous CAMP-dependent protein kinase.The effect of MgATP was rapidly reversible; reversibility was reduced by the phosphatase inhibitor NaF.These results suggest that the effects of MgATPrS are mediated by an endogenous protein kinase.
The major 35S-thiophosphorylated band detected by sodium dodecyl sulfate-polyacrylamide gel electrophoresis was a protein with M, = 36,000.The phosphorylation of a protein with the molecular weight of the cAMP receptor (Mr = 40,000-45,000) was not observed.
Extracellular cAMP functions as a signal molecule in Dictyostelium discoideum during chemotaxis (I), morphogenesis (2), and cell differentiation (3).cAMP is detected by highly specific surface receptors, which results in several cellular responses such as the activation of adenylate and guanylate cyclase (these and other responses have been reviewed (4-6)).The stimulation of guanylate and adenylate cyclase terminates within a few seconds and a few minutes, respectively, even when cAMP remains present a t constant levels (7)(8)(9)(10).Desensitization of adenylate cyclase stimulation has been studied extensively by Dinauer et al. (11,12).
Recent results (13,14) suggest the presence of two subpopulations of cell surface cAMP receptors, A and B sites, that * This work was supported by the C. and C. Huygens Fund which is subsidized by the Netherlands Organization for the Advancement of Pure Scientific Research (Zuiver Wetenschappelijk Onderzoek).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.in the stimulation of adenylate and guanylate cyclase, respectively.Binding of cAMP to both subpopulations is complex, showing interconversions of binding states in vivo (15,16) which are promoted by guanine nucleotides i n uitro (16)(17)(18)(19).This may suggest the involvement of guanine nucleotide regulatory proteins in the transduction pathways of adenylate and guanylate cyclase.This view is further supported by the recent observations (20,21) that guanosine triphosphates stimulate adenylate cyclase i n uitro.
Desensitization of CAMP-stimulated guanylate cyclase occurs within a few seconds after cAMP addition (9) and may be caused by an impairment of receptor/G-protein interaction (16).Desensitization of adenylate cyclase is much slower and requires the presence of constant cAMP concentrations during several minutes (12).This desensitization process may be caused by a receptor modification, presumably by phosphorylation (22)(23)(24).cAMP also induces the loss of detectable CAMP-binding sites (down-regulation) which occurs with similar kinetics as desensitization of adenylate cyclase, but which requires 10-fold higher cAMP concentrations and recovers more slowly (25).' The role of phosphorylation of signal transducing components in desensitization has been extensively studied in vertebrates (26)(27)(28)(29)(30)(31)(32)(33).These studies make use of the well-characterized functional coupling between receptors, G-proteins, and adenylate cyclase in native membranes or after reconstitution of purified components in artificial membranes.It has been shown that phosphorylation of receptor or G-protein diminishes their functional interaction (29)(30)(31).
Conditions for the functional reconstitution of isolated receptor, G-protein(s), and adenylate cyclase have not yet been found for the components from D. discoideum.Therefore, a study of the role of phosphorylation in transmembrane signaling in this organism is limited to experiments i n uiuo (22)(23)(24) or to receptor/G-protein/adenylate cyclase interaction in isolated membranes.
In this study, the receptor/Gprotein interaction was investigated by detecting the effects of guanine nucleotides on cAMP binding to membranes that were preincubated under conditions where protein kinases are active.The results show that phosphorylation conditions alter the receptor/G-protein interaction, and evidence is presented which suggests that this is due to an endogenous protein kinase.
Culture Conditions and Membrane Isolation-D.discoideum cells (strain NC-4) was grown on a buffered glucose/peptone medium (9).Cells were harvested in 10 mM KH,PO,/Na,HPO,, pH 6.5 (buffer A), freed from bacteria by repeated centrifugations at 100 X g for 2 min, and starved in buffer A by shaking at a density of lo7 cells/ml.After 5-6 h, cells were collected by centrifugation and washed twice with buffer A, and the pellet was resuspended in buffer B (40 mM HEPES/ NaOH, 0.5 mM EDTA, 250 mM sucrose, pH 7.7) to a density of 10' cells/ml.Cells were lysed by pressing them through 3-pm pores of a Nuclepore filter (35).The lysate was centrifuged at 10,000 X g for 5 min, the pellet was washed once with buffer B, and the final pellet was resuspended in buffer B to the equivalent of 2 X 10' cells/ml.
Incubation with ATP and Analogs-Membranes (300 pl) were incubated for 5 min at 20 "C with 3 pl of 100 X concentrated solutions yielding final concentrations of 10 mM NaF, 5 mM MgC12, and 1 mM ATP or one of its analogs.In one experiment, membranes were incubated in the presence of CAMP-dependent protein kinase.This enzyme (1 pmol of holoenzyme in 10 pl) was preactivated with 3 p~ (S,)-CAMPS at 20 "C for 15 min; the amount of enzyme was estimated from its cAMP binding activity as described (34).
Incubations were terminated by addition of 1200 pl of ice-cold buffer A and immediately centrifuged at 0 "C for 2 min at 10,000 X g.The pellet was washed once with 1 ml of buffer A, and the final pellet was resuspended in 600 p1 of buffer A.
CAMP Binding Assay-CAMP binding was measured in a total volume of 100 p1 containing buffer A, 5 nM t3H]cAMP, 10 mM dithiothreitol, GTPyS (30 p~ when present), and 80-pl membranes.The incubation period was 5 min a t 0 "C.Samples were centrifuged for 3 min at 10,000 X g, the supernatant was aspirated, and the pellet was dissolved in 75 p1 of 1 M acetic acid.Radioactivity was determined by liquid scintillation counting after the addition of 1.4 ml of Emulsifier (Packard Instrument Co.).Nonspecific binding was determined in the presence of 1 mM cAMP and was subtracted from all data shown.Typical binding data for control membranes at 5 nM [3H] cAMP are input = 18,000 cpm, specific binding = 3,058 +-112 cpm, and nonspecific binding = 186 * 21 cpm (both n = 3).
Scatchard analysis of CAMP-binding (Fig. 1) was determined by incubating membranes with different cAMP concentrations at 20 "C.The binding reaction was terminated by centrifugation of membranes through silicon oil (16).
Phosphorylation of Receptor in Viuo-Aggregative cells were harvested in 15 mM Tris/HCI, pH 7.0, washed twice, and resuspended in this buffer at a density of 10' cells/ml.Cells (200 p l ) were incubated for 30 min with 1 mCi/ml orth~[~~P]phosphate and 5 mM caffeine.Then, the suspension was divided into two equal portions.One portion was stimulated with 10 mM dithiothreitol and 1 p M CAMP; the other portion was not stimulated.After 15 min, the reactions were terminated by the addition of 1 ml of 95% saturated ammonium sulfate, and the suspension was incubated for 5 min at 0 "C and centrifuged for 5 min at 10,000 X g.The pellet was washed once with 1 ml of 95% saturated ammonium sulfate, and the final pellet was resuspended in 1 ml of receptor buffer (22), which induced cell lysis.Membranes were centrifuged for 15 min at 10,000 X g, and the pellet was resuspended in 100 pl of sample buffer.SDS-PAGE was carried out as described in the previous section.

RESULTS
The aim of the present study was to investigate the effects of endogenous protein kinase activity on cell-surface CAMP receptors and on the interaction between receptor and putative G-protein(s).D. discoideum cells contain high levels of protein phosphophosphatase activity (36).To by-pass this phosphatase activity, a derivative of ATP was used, ATPyS, which is a relatively good substrate of many protein kinases.However, the product, protein phosphorothioate, is a poor substrate for protein phosphophosphatases (reviewed in Ref. 37).As a result, the phosphorylation reaction becomes irreversible.
Scatchard Analysis of CAMP Binding-D.discoideum cells contain different forms of CAMP-binding sites.A form designated as Bss has high affinity for cAMP (Kd -13 nM) and shows a very slow rate of dissociation (h,? = 150 s).BS has the same high affinity for CAMP, but the complex dissociates 10- fold faster (tu = 15 s).The major part of the receptors are fast dissociating (Fast, tLh -1 s) and are composed of a t least two forms, AH and AL, with high (Kd = 60 nM) and low (& = 450 nM) affinity, respectively (15,16).These binding forms are also present in membranes from D. discoideum cells; their proportioning is affected by guanine nucleotides, resulting in an increase of AL and a decrease of the other forms.
The influence of pretreatment, of membranes with MgATPyS on the binding heterogeneity was investigated.Membranes were incubated with 1 nM ATPyS and 5 mM M$+ for 5 min at 20 "C and then extensively washed at 0 "C.The results are shown in Fig. 1.The binding of cAMP to BS or Bss is apparently homogeneous, whereas binding to Fast is clearly composed of at least two components.The kinetic data of BS and Bss were calculated by linear regression analysis, whereas the data of the fast dissociating components were calculated by a hyperbolic regression analysis.
The obtained kinetic data are presented in Table I.The site distribution of control membranes is 26% A H with Kd = and 5% Bss with Kd = 7.7 nM.GTPyS strongly alters this distribution: about 98% of the sites have the properties of AL; A" and Bss are reduced to 1%, whereas BS is virtually absent ( ~0 . 2 % ) .
It should be noted that the low abundance of AH complicates the determination of its affinity.
cAMP binding to MgATPyS-treated membranes in the absence of GTPyS strongly resembles the binding to control membranes in the presence of GTPyS.Binding of cAMP to A", BS, and Bss is strongly decreased with a concomitant increase of binding to AL.It should be noted that binding to BS is still detectable in MgATPyS-treated membranes.As in control membranes, this form is undetectable after the addition of GTPyS.
The results of Fig. 1 and Table I indicate that GTPyS or pretreatment with MgATPyS predominantly alters the proportioning of binding forms, whereas the total number of binding sites and the affinities are not strongly altered.The effect of MgATPyS is a shift of high to low affinity binding forms, which is most easily detected by equilibrium binding of low cAMP concentrations.
Kinetics of Modification of CAMP Binding by MgATPyS-Membranes were preincubated for various time periods with MgATPyS at 20 "C and washed, and cAMP binding was FIG. 1. Scatchard analysis of cAMP binding.Membranes were incubated with or without 5 mM MgC1, and 1 mM ATPyS for 5 min at 20 "C and washed at 0 "C, and cAMP binding was detected at 20 "C in the absence or presence of 30 y~ GTPyS.Binding equilibrium was reached after 75 s; at this moment, excess cAMP (0.1 mM) was added.Membranes were centrifuged through silicon oil just before or 10 or 120 s after the addition of excess CAMP.Membrane-associated radioactivity was determined, and binding to the Fast (A), BS ( B ) , and Bss (C) forms was calculated using Equation a-c.These curves were subjected to hyperbolic (A) or linear ( B and C) curve fitting.
The kinetic data are shown in Table I. 0 and 0, control cells; A and A, MgATPyS-treated cells; 0 and A, and 0 and A, cAMP binding in the absence or presence of GTPyS, respectively.The binding of 10 nM cAMP is equal to 1.5 pmol of cAMP/mg of protein and to 75,000 binding sites/cell equivalent.cAMP bound, nM GTPyS (Fig. 2).MgATPyS induces a gradual loss of cAMP binding activity when detected in the absence of GTPyS (Fig. 2).Binding of cAMP in the presence of GTPyS was slightly increased after a short incubation with MgATPyS and was decreased after longer incubation periods with MgATPyS.As a result, the percent inhibition of cAMP binding by GTPyS decreased from 60 to 10% with tL,> = 1 min.
Modification of cAMP Binding Requires Phosphorylating Conditions-Membranes were preincubated with 1 mM ATPyS and/or 5 mM Mg2+ for 5 min at 20 "C.Then, membranes were washed at 0 "C, and cAMP binding was detected in the absence or presence of 30 PM GTPyS (Table 11).
Control membranes, which were incubated at 20 "C in the absence of ATPyS or divalent cations, showed a 75% inhibition of cAMP binding by 30 FM GTPyS.Membranes treated with either ATPyS or M$+ had identical CAMP binding activity as control membranes.In contrast, treatment of membranes with M$+ and ATPyS resulted in a 60% inhibition of cAMP binding in the absence of GTPyS and essentially no inhibition of binding in the presence of GTPyS.
Analogs of ATP which cannot donate a phosphate group in phosphorylation reactions, i.e.MgAppNHp and MgAppCHp,

FIG. 2. Kinetics of MgATPyS-induced modification of cAMP binding.
Membranes were incubated for 10 min at 20 "C; the last part of the incubation period was with 5 mM MgC1, and 1 mM ATPyS (Thus, the incubation at t = 2 min was for the first 8 min without and then for 2 min with MgATPyS; this protocol allows the simultaneous processing of all samples.)Membranes were washed, and cAMP binding was detected in the absence (0) or presence (0) of 30 ~L M GTPyS.The percent inhibition by GTPyS is indicated (*).did not alter CAMP binding (Table 11).The replacement of ATPyS by ATP yielded qualitatively similar results.However, the effect was less pronounced than with MgATPyS, possibly due to rapid reversion of the reaction by protein phosphophosphatase.The addition of an exogenous protein kinase act.ivitymay shift the equilibrium of phosphorylation/ dephosphorylation to higher levels of phosphorylation.We have used mammalian CAMP-dependent protein kinase because it has been shown that this enzyme alters receptor/(;protein/adenylate cyclase interaction in vertebrate cells (26,32).It was observed that t.he effect of MgATP is significantly (p < 0.05, paired t test) potentiated by exogenous protein kinase (Table 11).

TABLE I1 Modification of CAMP binding by ATP analogs
Rwersibility of Modification of cAMP Rinding-Membranes were preincubated with MgATP or MgATPyS, washed, and resuspended in buffer at 0 "C.CAMP binding was detected immediately at 5 min or at 45 min after resuspending the membranes (Fig. 8).Membranes treat.edwith MgATPyS showed the usual decrease of CAMP binding, which did not reverse.The effect of MgATP was almost reversed a t 5 min and completely reversed at 45 min after resuspending treated membranes in phosphat,e buffer.The reversibility was considerably decreased in the presence of the phosphatase inhibitor NaF (38).This and duplicate experiments suggest that the effects of MgATP reverse with half-times of about 3 and 20 min in the absence and presence of NaF, respectively.
Phosphoplation of Receptor-It has been shown that CAMP induces the phosphorylation of the CAMP receptor in vivo, which is associated with a shift of the apparent molecular weight of the receptor from 40,000 to 43,000 (22)(23)(24).The CAMP-induced shift of the mobility of a '"P-phosphorylated protein is shown in Fig. 4. The question was addressed whether MgATPyS or MgATP may phosphorylate the receptor in vitro.Therefore, membranes were incubated with [y- %]ATP$3 or [y-""PIATP, and phosphorylated proteins were separated by SDS-PAGE.The results (Fig. 4) reveal that the major phosphorylated protein has a molecular weight of 36,000 and clearly does not co-migrate with the phosphorylated protein whose mobility changes in stimulated cells.It should be noted that cells contain a phosphorylated protein which co-migrates with the protein that is phosphorylated in vitro. ..
The binding of CAMP to membranes is modified by guanine nucleotides (16)(17)(18)(19), which may suggest that the receptor(s) interact with a G-protein(s).This view is substantially supported by the recent findings that guanosine triphosphates activate adenylate cyclase in vitro (20,21).
In this report, experiments are shown which probe the hypothesis that phosphorylation reactions modify the functional interaction between cell-surface CAMP receptors and a putative guanine nucleotide regulatory protein in D. discoideum membranes.An ATP derivative, ATPyS, has been used.
It has been reported that this analog is a potent substrate in many kinase reactions, but that the product, protein phosphorothioate, is not easily hydrolyzed by phosphatases (37).Therefore, little attention has to be paid to protein phosphophosphatase activity which is high in D. discoidam (36).The main results of this study are the following.1) MgATPyS rapidly (t1,% -1 min) and irreversibly alters cAMP binding.The number of binding sites is slightly (-15%) reduced.In control membranes, GTP@ induces a transition of high affinity binding to a low affinity form.This low affinity binding form is induced by pretreatment of membranes with MgATPyS, and GTPyS does not strongly affect cAMP binding any further.2) These effects are not induced by Mg' or ATPyS alone or by MgAppNHp or MgAppCHzp.3) MgATP induces qualitatively similar results as MgATPyS, which are, however, less pronounced. 4)The effect of MgATP is potentiated by exogenous CAMP-dependent protein kinase and is readily reversible.Reversibility is slowed down by NaF, a phosphatase inhibitor (38).5 ) The major phosphorylated protein in membranes is a protein with M, = 36,000.
These results suggest that the effects of MgATPyS and MgATP on cAMP binding are mediated by endogenous protein kinase activity.The substrate of this reaction could be the cAMP receptor, the G-protein, or any other component that affects the receptor/(;-protein interaction.The major phosphorylated protein in isolated membranes has an apparent molecular weight of 36,000 and does not co-migrate on SDS-PAGE with the phosphorylated receptor.This suggests that the receptor is not phosphorylated in vitro or that phosphorylation is below the detection limit.A definite proof whether phosphorylation of the receptor, G-protein, or another component is responsible for the observed effects can be given only after a functional analysis of purified phosphorylated and nonphosphorylated components that are reconstituted in artificial membranes.Unfortunately, conditions for the isolation and reconstitution of signal transducing components that have been used in vertebrates (31) are not effective in D. discoideum, and no alternative conditions have been found yet.3 Recently, it has been shown that guanosine triphosphate stimulates adenylate cyclase of D. discoideum in vitro (20,21).
We observed only a modest degree of stimulation (70%).Interestingly, however, this stimulation was lost in membranes derived from desensitized cells (adenylate and guanylate cyclases were desensitized, and receptors were partly down-regulated).In addition, when membranes from control cells were preincubated with MgATPyS, the 70% stimulation of adenylate cyclase by GTPyS was converted to a 30% inhibition.These observations suggest that MgATPyS alters the receptor/G-protein as well as the G-protein/adenylate cyclase interactions.MgATPyS does not alter the number of binding sites.Interestingly, 40% of the binding sites are lost in a reversible manner upon further incubation of MgATPyStreated membranes with cAMP and micromolar concentrations of Ca2'.4These observations may suggest that the action of MgATPyS could be related to the mechanism of downregulation of cAMP receptors.
The present and previous (42) results that a putative endogenous protein kinase activity modifies receptor/G-protein interactions allow the investigation of many potentially interesting questions such as whether occupation of the receptor P. M. W. Janssens and R. Van Driel, personal communication.P. J. M. V a n Haastert, unpublished results.
or G-protein affects the rate of phosphorylation and dephosphorylation and concerning the identities of the kinase, the phosphatase, and their substrates.The notion that ATPyS irreversibly and ATP reversibly alter the cAMP binding activity could facilitate these experiments.
Control and MgATPy S-treated membranes were incubated with different concentrations [3H]cAMP in the absence and presence of 30 p~ GTPyS.Binding of [3H]cAMP was detected at equilibrium (b(O)), a t 10 s after the onset of dissociation (b(lO)), and at 2 min after the onset of dissociation (b(120)).Binding to the different receptor forms was calculated with the following set of equations (see Ref. 16).b(0) = Fast + BS + Bss (la) b(10) = 0.65 BS + 0.95 Bss (1b) b(120) = 0.56 Bss (IC) Membranes were preincubated in the absence (control) or presence of 5 m ~ M P , 1 mM nucleotides, and preactivated CAMP-dependent protein kinase (cAK) as indicated for 5 min at 20 "C.Then, membranes were extensively washed, and cAMP binding was measured in the absence or presence of 30 y~ GTPyS.The results shown are the means f S.D. of three independent experiments.

FIG. 3 .
FIG. 3. Reversibility of MgATPyS-and MgATP-induced modification of cAMP binding.Membranes were incubated with 5 mM MgCl? and 1 mM ATP or ATPyS for 5 min a t 20 "C, washed, and resuspended in buffer A in the presence or absence of 10 mM NaF. cAMP binding was detected in the absence (open bars) or presence (hatched bars) of 30 p~ GTPyS at 0, 5, and 45 min after resuspension.The results are the means and standard deviations of triplicate determinations of an experiment reproduced twice.

FIG. 4 .
FIG. 4. Protein phosphorylation in vivo and in vitro.Membranes were incubated for 5 rnin with [y-"P]ATP (lane I ) or [y-,'"S] ATPyS (lane 2, see "Experimental Procedures").Cells were incubated with ortho["'P]phosphate and caffeine.Control cells (lane 3) and cells a t 15 min after stimulation with 1 p M cAMP and 10 mM dithiothreitol (lane 4 ) are shown.Proteins were separated by SDS-PAGE and stained with Coomassie Blue.The gel was incubated with EN-HANCE, dried, and autoradiographed.A densitometric scan of the autoradiograms is shown below.The result shown is representative for two (lanes 3 and 4 ) or four (lanes 1 and 2 ) independent experiments.Arrows indicate the apparent molecular masses in kilodaltons.

TABLE I
Effect of MgATPyS pretreatment on CAMP binding heterogeneity Membranes were preincubated in the absence or presence of MgATPyS.cAMPbinding was performed in the absence or presence or GTPyS.The data are derived from the results presented in Fig.1.- detected in the absence or presence of 30 ~L M