Changes in Cyclic AMP Receptor Properties During Adaptation in Dictyostelium Discoideum

Changes in cyclic AMP receptor properties during adaptation in Dictyostelium discoideum

M. E. E. LUDERUS', M. J. SPIJKERS and R. VAN DRIEL

E.C. Slater Institute for Biochemical Research, University of Amsterdam, Plantage Muidergracht 12, 1018 TVAmsterdam, The \etliet1ands

•Author for correspondence

Summary

In developing Dictyostelium discoideum cells, receptor binding completely, presumably by acting binding of cyclic AMP to the chemotactic receptor via a G protein. The guanine nucleotides reduced has been shown to oscillate. These oscillations the affinity of the receptor at all time-points of the represent cycles of activation, adaptation and de- oscillation cycle to the minimal, i.e. adapted, level. adaptation of the cyclic AMP receptor system. We conclude that the cyclic process of activation, We studied the molecular basis of these oscilla- adaptation and de-adaptation in D. discoideum, at tory changes in cyclic AMP receptor binding. We cyclic AMP receptor level, involves changes in developed a rapid method of lysing cells during the receptor-G protein interaction. During adaptation, course of the oscillations. This method guaranteed the affinity of the cyclic AMP receptor decreases good preservation of ligand binding properties of and the receptor becomes insensitive to guanine the cyclic AMP receptor. nucleo tides. We found that oscillations in cyclic AMP binding resulted from changes in receptor affinity. The total number of receptors did not significantly change during oscillations. Our experiments also showed Key words: chemotactic cyclic AMP receptor, adaptation, D. that both GTP and GDP abolished oscillations in discoideum, G protein, oscillations.

Introduction phosphodiesterase (Malchow et al. 1972). The resulting periodic changes in extracellular cyclic AMP concen- Early in development, Dictyostelium discoideum cells tration synchronize various cyclic AMP-induced cellular acquire the capacity to detect, synthesize and secrete responses. Many cellular responses have in fact been cyclic AMP. This cyclic AMP signal-relay system is the shown to oscillate with the same frequency as the cyclic basis of an intercellular communication system that AMP signal (Aeckerle et al. 1985; Bumann et al. 1986; controls chemotaxis, cell aggregation, differentiation and Gerisch and Hess, 1974; Klein et al. 1977; Malchow et pattern formation (Devreotes, 1982; Sussman, 1982). al. 1978; Wurster et al. 1977). These oscillations are the The signal molecule in this system is extracellular cyclic result of repeated cycles of activation, adaptation and AMP. It is detected by highly specific receptors that are de-adaptation. present on the surface of the cell. Binding of cyclic AMP Regulation of receptor-mediated signal transduction in to these receptors induces several rapid responses, such D. discoideum is similar to that in higher eukaryotes as synthesis and secretion of cyclic AMP (Gerisch and (Janssens, 1987). The cyclic AMP receptor of D. discoid- Wick, 1975), synthesis of cyclic GMP (Mato et al. 1977), eum is thought to have the classical seven transmembrane influx of Ca2+ (Wick et al. 1978) and efflux of K+ and H+ spanning domains that are characteristic of all eukaryotic (Aeckerle et al. 1985; Malchow e< a/. 1978). Most of these G protein-coupled receptors studied so far (Klein et al. responses are subject to adaptation. Adaptation refers to 1988). The cyclic AMP receptor is coupled to two or the mechanism by which the response to extracellular more G proteins (Janssens et al. 1985; Kumagai et al. cyclic AMP is terminated. Adaptation is induced by 1989; Van Haastert et al. 1986). These G proteins prolonged exposure of cells to a constant level of cyclic mediate the activation (Janssens and De Jong, 1988; Van AMP; de-adaptation occurs as the cyclic AMP concen- Haastert et al. 1987), and possibly adaptation (Small et tration declines. al. 1987; Van Haastert et al. 1986), of the enzymes During development, D. discoideum cells secrete adenylate cyclase and guanylate cyclase. Furthermore, pulses of cyclic AMP with a frequency of one pulse per evidence has been presented for a role for the phospho- 5-8 min (Klein et al. 1977). Secreted cyclic AMP is inositol cycle in the signal transduction system of D. hydrolyzed by the extracellular enzyme cyclic AMP- discoideum (Europe-Finner and Newell, 1985; Van Journal of Cell Science 95, 623-629 (1990) Printed in Great Britain © The Company of Biologists Limited 1990 623 Lookeren Campagne et al. 1988). Despite detailed in Dictyostelium; Green and Newell, 1975), 20fm S'AMP), 3 knowledge of the physiology of receptor-mediated re- containing [ H]cyclic AMP at concentrations given in the text. If indicated, 0.1 mM GTPyS, GDP or GMP were present. sponses in both D. discoideum and higher eukaryotes, 3 little is known about the molecular mechanism of acti- Subsequently, the amount of [ H]cyclic AMP bound to the membranes was determined by measuring the radioactivity in vation, adaptation and de-adaptation of these responses. the pellet after 2 min of centrifugation at 10 000 # in a microfuge The synchronous oscillations of the cyclic AMP receptor (sedimentation assay). Non-specific binding was determined by system of D. discoideum make this organism a convenient equilibration in the presence of a large excess (0.5 mM) of model system in which to study these mechanisms. unlabeled cyclic AMP. In the present work we studied the repetitive cycle of Measurements of equilibrium [3H]cyclic AMP binding to the activation, adaptation and de-adaptation in D. discoid- different receptor forms were done according to Van Haastert eum at the level of the cyclic AMP receptor. In cell lysates et al. (1986). Cyclic AMP receptors occur in a limited number prepared at various time-points during the oscillation of well-defined receptor types, called Fast, S and SS (Janssens cycle, we investigated ligand binding and receptor-G et al. 1986; Van Haastert and De Wit, 1984). Each of these protein interaction. Lysates were prepared by a newly receptor types is characterized by its own dissociation rate constant. In isolated membranes, at 20°C, the following dis- developed, simple procedure based on freezing cells in sociation rate constants (£-i) have been measured (Van Haas- liquid nitrogen. The procedure was rapid and guaranteed x tert, 1984): Fast type, k-x=G.2b±QM%~ \ S type, £_,= good preservation of cyclic AMP receptor properties. 22 1 1 333 (4.3±0.6)Xl0" s- ; SS type, ^_, = (3.9±0.9)Xl0-- s-s '' Our results indicate that adaptation in D. discoideum, Since no interconversion of these receptor forms occurs during measured at receptor level, involves changes in recep- the course of dissociation, it can be calculated that after a tor-G protein coupling. In the adapted state, the cyclic dissociation period of 10 s essentially no occupied Fast sites are AMP receptor has a reduced affinity for cyclic AMP and present, whereas binding to S sites is reduced to 64%, and is insensitive to guanine nucleotides. binding to SS is reduced to 97 %. After a dissociation period of 120 s, essentially all Fast and S sites are empty, while binding to SS is reduced to 61 %. This is described by the following set of Materials and methods equations: 6(0) = Fast(0) + S(0) + SS(0) Materials 6(10) = (0.64) S(0) + (0.97) SS(0) [5',8-3H]cyclic AMP (l.SSTBqmmol"1) was purchased from 6(120) = (0.61) SS(0) Amersham International (UK), cyclic AMP and dithiothreitol Thus, from measuring the amount of [3H]cyclic AMP bound at from Serva (Heidelberg, Federal Republic of Germany), and equilibrium, 6(0), and at 10s and 120s after the onset of S'AMP, GDP, GMP, and GTPyS from Boehringer (Mann- dissociation, 6(10) and 6(120), it is possible to calculate the heim, Federal Republic of Germany). Nitrocellulose filters occupancy of the three receptor forms at equilibrium (Fast(0), (type BA 85) were from Schleicher & Schiill (Dassel, Federal S(0) and SS(0)). Republic of Germany). Measurement of 6(0), 6(10) and 6(120) was performed as follows. Washed paniculate fraction (200 j/g protein) was equi- Culture conditions and membrane isolation librated in 100 fi\ BM with 10 nM [3H]cyclic AMP for 5 min at D. discoideum cells (strain AX2) were grown in HL-5 medium 20cC. Subsequently, either the total amount of [3H]cyclic AMP (Watts and Ashworth, 1970), that contained maltose instead of bound to the membranes, 6(0), was determined directly by the glucose. Cells were harvested in the late logarithmic growth 6 sedimentation assay (see above), or dissociation of the recep- phase at a density of approximately 5xl0 cellsml~ . For tor-[3H]cyclic AMP complexes was initiated at 20°C by 100- development in suspension, cells were washed once with 10 mM fold dilution with P, buffer. After 10 s and 120 s of dissociation, 3 KH2PO4/Na2HPO4, pH6.5 (P, buffer), resuspended in the 7 1 [ H]cyclic AMP that remained bound to the membranes (6(10) same buffer at 2xl0 cellsml~', and shaken at ISO revs min" at and 6(120), respectively) was determined by filtration through 22°C. When indicated in the text, every 6min a cyclic AMP 0.45 um pore-size nitrocellulose filters, measuring the [3H]cyc- pulse of 5 nM was given, starting after 4 h of development. After lic AMP retained by the membranes on the filters (nitrocellulose 5 h of development, 500-/^1 samples were taken from the filtration assay). suspension every minute, and mixed with 20% (w/v) sucrose plus the following protease inhibitors: 5 mM benzamidine, 100;Ugml~ aprotinin, 50/igml~ trypsin inhibitor and Results 20^gml~' antipain (final concentrations). The cells were lysed immediately by rapid freezing in liquid nitrogen. The sucrose in Lysis method the medium was essential for the preservation of cyclic AMP In synchronously developing D. discoideum cells, cyclic receptor binding. The lysate was stored in liquid nitrogen. For AMP binding to the chemotactic receptor has been measurements of cyclic AMP receptor binding, the lysate was shown to oscillate (Gerischef al. 1979; Klein et al. 1977). thawed and centrifuged in a microfuge at 4°C, for 2min at In this paper we studied the molecular background of 100001*. The pelleted, paniculate fraction was washed twice these oscillations. For this purpose we developed a rapid and resuspended in P, buffer to a concentration of 2 mg protein ml"1 (SX107 cell equivalents ml"1). method of lysing cells during the course of oscillations, which allows accurate investigation of cyclic AMP recep- Cyclic AMP binding assays tor properties and receptor-G protein interaction. The Washed paniculate fraction (200 ug protein) was incubated for lysis method is based on freezing cells in liquid nitrogen in the presence of protease inhibitors and 20% sucrose. 5 min at 0°C in 100 u\ binding medium (BM; 25 mM KH2PO4/ NaOH (pH7.4), 10mM dithiothreitol (an inhibitor of extra- For an adequate use of the method in our investigations, cellular and cell surface-bound cyclic AMP-phosphodiesterase it is important to verify that the lysis method per se does

624 M. E. E. Ludirus et al. In Fig. 1, we compared equilibrium binding characteristics of cyclic AMP receptors present in: (1) the membrane fraction of cells lysed by the new method of liquid nitrogen freezing; (2) intact cells; and (3) membranes isolated by the method of nitrogen cavitation described earlier (Janssens et al. 1985, 1986). The cyclic AMP binding curves of the receptors in the two membrane preparations were similar, with respect to both receptor affinity and receptor number. The affinity of the receptors in the membrane preparations was comparable to the affinity of receptors on intact cells. The number of receptors in the membranes was, however, somewhat smaller than the number of receptors on intact cells. This 0 was presumably due to receptor degradation that oc- 6 2 4 6 8 10 12 14 4 curred during membrane isolation. Competition binding Bound (molecules/cell equivalent, xlO~ ) experiments with various cyclic AMP analogues (Van Fig. 1. Comparison of [3H]cyclic AMP binding to intact Haastert and Kien, 1983) revealed that the specificity of cells and isolated membranes. Intact cells (•) or crude the cyclic AMP binding shown in Fig. 1, was identical to membranes, isolated by either liquid nitrogen lysis (D) (see the specificity of binding to the chemotactic cyclic AMP Materials and methods) or nitrogen cavitation (O) (Janssens receptor (results not shown). et al. 1985, 1986), were equilibrated in binding medium with In Fig. 2, we examined oscillations in cyclic AMP different concentrations of [3H]cyclic AMP (between 0.1 nM 3 receptor binding in membranes derived from cells lysed and 350 nM). The amount of [ H]cyclic AMP that specifically by freezing in liquid nitrogen. At different time points bound to the cells or membranes was determined by the during the oscillation cycle we lysed cells and measured sedimentation assay. The results are shown in a Scatchard 3 3 plot. Each point is the mean of a determination in triplicate [ H]cyclic AMP receptor binding (at 10nM [ H]cyclic of an experiment reproduced twice. AMP) in the paniculate fraction. Fig. 2A shows that the amount of [3H]cyclic AMP that bound to the receptor oscillated with a period of 6 min. The exact period varied from culture to culture, but was always between 6 and 8 min. The rates of increase and subsequent decrease in oscillatory cyclic AMP binding behaviour can be characterized by half-times of rise and decline of 2 min and 1 min, respectively. Minimal and maximal binding values differed by a factor of about 2. In membranes isolated from cells that were synchronized by entrainment with a series of 5 nM cyclic AMP pulses, oscillations had the same frequency as the applied pulses. Each cyclic AMP pulse, within 0.5 min, induced a drop in [3H]cyclic AMP binding. This periodic decrease was not caused by simple 20 0 isotope dilution, since the samples were washed free of 3 Time (min) unlabeled cyclic AMP before [ H]cyclic AMP binding was measured. Fig. 2. Binding of [3H]cyclic AMP to crude membranes We conclude that the method of liquid nitrogen lysis isolated from oscillating cells. A. Cells were starved in preserves the binding characteristics of the cyclic AMP suspension at a density of 2xlO7 cells ml" . Starting after 5h receptor at each time point in the oscillation cycle. The of starvation, 500-^*1 samples were taken from the suspension method is thus suitable for investigating cyclic AMP with time intervals of 1 min, mixed with 20% sucrose (final receptor properties and receptor-G protein interaction in concentration) plus protease inhibitors, and immediately D. discoideum during free-running oscillations. lysed by freezing in liquid nitrogen. After thawing, the paniculate fraction of the lysed cells was isolated by Binding properties of cyclic AMP receptors during centrifugation and washed with potassium phosphate buffer. Subsequently, equilibrium [3H]cyclic AMP binding (at 10 nM oscillations [3H]cyclic AMP) to this crude membrane fraction was Several classes of chemotactic cyclic AMP receptors have measured by the sedimentation assay. B. The same procedure been detected on D. discoideum cells (Van Haastert and as described for A was followed, except that, starting after 4h De Wit, 1984; Van Haastert et al. 1986). They differ of starvation, cells were stimulated with a 5 nM cyclic AMP from each other in dissociation rate constant and/or pulse every 6 min (arrows). apparent affinity for cyclic AMP. The majority of the receptors are called Fast receptors; the remaining recep- not affect receptor properties. We therefore compared tors are designated S and SS receptors. At 20°C, Fast ligand binding properties (Fig. 1) and oscillation charac- receptors are characterized by a dissociation rate constant l 2 1 teristics (Fig. 2) of cyclic AMP receptors in D. discoid- k-x=QMs~ \ S receptors by *_, = 4.3X 10~ s" and SS 3 1 eum cells before and after cell lysis. receptors by ^_! = 3.9x 10~ s" . SS and S receptors have

Cyclic AMP receptor properties during adaptation 625 0.12- 1500-

1000-

12 14 Bound (molecules/cell equivalent, xlO~4)

Fig. 3. Scatchard analysis of cyclic AMP receptor binding at oscillation minima and oscillation maxima. Crude membranes Fast (H+L) were isolated, by the method of liquid nitrogen lysis (see Materials and methods), from oscillating cells at time-points Fig. 4. Binding of [3H]cyclic AMP to the different cyclic at which cells showed either maximal (O) or minimal (•) AMP receptor forms at oscillation minima and maxima. cyclic AMP binding (see Fig. 2). Membranes were washed Crude membranes were isolated, by the method of liquid with potassium phosphate buffer and equilibrium [3H]cyclic nitrogen lysis, from oscillating cells at time-points at which AMP binding (at [3H]cyclic AMP concentrations between total [3H]cyclic AMP binding (measured at 10nM [3H]cyclic 0.1 nM and 350 nM) was determined by the sedimentation AMP) was either maximal (open bars) or minimal (filled assay. Each point is the mean of a determination in triplicate bars) (see Figs 2 and 3). Membranes were washed and of an experiment reproduced twice. equilibrated with 10 nM [3H]cyclic AMP in binding medium at 20°C. The amount of [3H]cyclic AMP that specifically bound to the membranes was either measured directly by the a similar affinity for cyclic AMP: A"d=6-13nM. Fast 3 H L sedimentation assay, or dissociation of the [ H]cyclic AMP receptors occur in two affinity forms, Fast and Fast , receptor complexes was initiated by 100-fold dilution with having a A'd of 60 and 450nM, respectively. These potassium phosphate buffer. After 10 s and 120 s of different receptor forms have also been detected in cell dissociation, [3H]cyclic AMP that remained bound to the lysates and isolated membranes of D. discoideum (Jans- membranes was measured by the nitrocellulose filtration sens et al. 1985, 1986; Van Haastert, 1984; Van Haastert assay. Binding to Fast, S and SS receptor forms at etal. 1986). equilibrium was calculated as described in Materials and Using the method of liquid nitrogen lysis, we investi- methods. Shown here are the means and standard deviations gated the binding properties of cyclic AMP receptors of determinations in triplicate of an experiment reproduced three times. during the course of oscillations. Thus far, measurements of cyclic AMP receptor binding to intact cells during oscillations have given controversial results. Gerisch and observed for the Fast sites. Here the difference was a Hess (1974) have found that the affinity of the cyclic factor of about 2. Binding to the S form, on the contrary, AMP receptor was lower at oscillation minima than at was essentially the same at oscillation minima and max- oscillation maxima. Devreotes and Sherring (1985), on ima. the other hand, did not observe significant differences in In conclusion, oscillations in total equilibrium affinity. The lysis method that we used here has the [3H]cyclic AMP binding, measured at 10nM [3H]cyclic advantage that the state of the receptor is 'frozen' at the AMP, are composed of oscillations in binding to SS and moment of sampling. In this way spontaneous alterations Fast receptor forms. During oscillations, the overall in receptor properties after sampling, which are likely to receptor affinity shows periodic changes, whereas the occur in intact cells, are prevented. Fig. 3 shows that the total receptor number remains constant. oscillations in equilibrium [3H]cychc AMP receptor binding (Fig. 2) were the result of periodic changes in Effect of guanine nucleotides on oscillations in cyclic receptor affinity, in agreement with the findings of AMP receptor binding Gerisch and Hess (1974). The number of receptors did Janssens et al. (1985, 1986) have demonstrated that GTP not significantly change during oscillations. The average and GDP, but not GMP, reduce cyclic AMP receptor affinity of cyclic AMP receptors at oscillation minima was binding in isolated membranes of D. discoideum by considerably lower than the affinity of receptors at lowering the affinity of the receptor for cyclic AMP. The oscillation maxima. We investigated the nature of the guanine nucleotides are believed to exert this effect by reduced affinity in more detail and determined the binding to a G protein that interacts with the chemotactic relative occupancy of SS, S and Fast receptor forms at receptor. oscillation minima and maxima. Fig. 4 shows that, after 3 We investigated the effect of different guanine nucleo- equilibration with IOIIM [ H]cyclic AMP, about three tides on cyclic AMP receptor binding in crude mem- times fewer SS receptors were occupied at oscillation branes of D. discoideum, isolated at various time-points minima than at oscillation maxima. A similar result was during autonomous oscillations. Fig. 5 shows that GMP

626 M. E. E. Luderus et al. quently, several cellular responses to cyclic AMP stimuli show an oscillator}' behaviour. These include the transient activation of the enzymes adenylate cyclase (Klein et al. 1977) and guanylate cyclase (Wurster et al. 1977), and changes in cyclic AMP receptor properties (Gerisch et al. 1979; C. Klein et al. 1977; P. Klein et al. 1985«). The oscillations involve cycles of alternating activation, adaptation and de-adaptation (Devreotes and Steck, 1979; Dinauer et al. 1980; Gerisch et al. 1979). The temporal correlation between oscillations in adenylate cyclase (AC) and guanylate cyclase (GC) activity, cyclic AMP receptor binding, and light-scattering responses has been studied in intact cells (Gerisch et al. 1979; Klein et al. 1977; Wurster et al. 1977). Oscillations in AC and GC activity and receptor binding appeared to run approxi- 12 16 mately synchronously, peaks in GC activity occurring Time (min) slightly in advance of peaks in AC activity and receptor

3 binding. The state of reduced receptor binding correlates Fig. 5. Effect of guanine nucleotides on [ H]cyclic AMP with adaptation of AC and GC responses. Oscillations in binding to crude membranes isolated from oscillating cells. Cells were starved in suspension at a density of equilibrium receptor binding have thus been interpreted 2X107 cells ml"1. After 5 h of starvation, samples (500 ^1) as cycles of adaptation and de-adaptation of the cyclic were withdrawn from the suspension every minute, mixed AMP receptor system, oscillation minima and maxima with 20% sucrose plus protease inhibitors, and lysed by representing the adapted and the de-adapted state, re- freezing in liquid nitrogen. After thawing, the paniculate spectively. fraction was isolated by centrifugation and washed with In this paper we investigate the molecular basis of the 50 mM potassium phosphate buffer. Subsequently, equilibrium [3H]cyclic AMP binding (at 10 nM [3H]cyclic cyclic changes in cyclic AMP receptor properties. We AMP) was measured in the presence of either 0.1 mM GMP developed a simple and rapid procedure for lysing cells (O), 0.1 mM GDP (A) or 0.1 mM GTPyS (•), or in the without changing the state of the cyclic AMP receptor. absence of guanine nucleotides (0). This lysis procedure was based on freezing cells in liquid nitrogen under well-defined ionic and osmolaric conditions. (0.1 mM) did not change kinetics or amplitude of the We found that oscillations in equilibrium [3H]cyclic oscillations in [3H]cyclic AMP binding. In contrast, both AMP binding were the result of changes in the affinity of GTPyS (a non-hydrolysable GTP analogue; 0.1 mM) and the receptor for cyclic AMP. The total number of GDP (0.1 mM) lowered [3H]cyclic AMP receptor bind- 3 receptors remained approximately constant. We analyzed ing (measured at 10nM [ H]cyclic AMP) at all time- the oscillations in receptor affinity in more detail by points in the oscillation cycle to one and the same level. 3 The nucleotides thus completely abolished oscillations in studying [ H]cyclic AMP binding to the different kinetic receptor binding. The level of [3H]cyclic AMP binding forms of the receptor. The majority of the chemotactic in the presence of either GTPyS or GDP was somewhat cyclic AMP receptors of D. discoideum, approximately below the average binding measured at oscillation 95 % of the total receptor number, are fast-dissociating receptors. They are called Fast, and can be subdivided minima in the absence of guanine nucleotide. Scatchard 1 1 into a high-affinity form, Fast " (A!c)=60nM) and a low- analysis (not shown) revealed that GTPyS reduced the L affinity of the receptor at both oscillation minima and affinity form, Fast (A'd=450nM). The remaining cyclic oscillation maxima to one and the same value. This AMP receptors consist of two types of slowly dissociat- affinity value was somewhat below the receptor affinity at ing, high-affinity receptors, designated S and SS oscillation minima in absence of guanine nucleotide. (A'd=6-13nM) (Van Haastert and De Wit, 1984; Van GTPyS did not alter the total number of cyclic AMP Haastert et al. 1986). We found that the oscillations in receptors. GDP had a similar effect to that of GTPyS equilibrium receptor binding (Fig. 2) were composed of (results not shown). oscillations in binding to SS and Fast receptor forms. Binding to the S form did not change during oscillations. We conclude that the periodic decrease in [3H]cyclic We conclude that the adaptation process is characterized AMP receptor affinity that occurs in vivo, in oscillating by a decreasing occupancy of Fast and SS receptors. cell cultures, can be mimicked in vitro, after cell lysis, by Since the total number of cyclic AMP receptors remained addition of GTP or GDP. approximately constant during oscillations, the reduced occupancy of Fast receptors in adapted cells cannot be Discussion due to a loss of binding sites. Instead, a conversion of Fast1"1 to FastL receptors may have occurred. Reduction Developing D. discoideum cells periodically synthesize of binding to the relatively small population of SS and secrete pulses of the chemoattractant cyclic AMP receptors, on the other hand, may be explained either by with a period length of 5-8 min (Gerisch and Hess, 1974; a conversion of SS sites to faster-dissociating forms, or by Gerisch and Wick, 1975: Roos et al. 1977). Conse- inactivation of SS sites.

Cyclic AMP receptor properties during adaptation 627 The population of Fast receptors is thought to mediate by a cyclic AMP receptor that displays a reduced affinity activation of AC, while SS and S receptors have been for its ligand and is insensitive to guanine nucleotides. proposed to be coupled to GC (Kesbeke and Van Whether this receptor state is caused by occupation of a G Haastert, 1985; Van Haastert et al. 1986). Consequently, protein by guanine nucleotide, or has other causes such as it is likely that the reduced occupancy of Fast and SS sites transient protein modification, remains to be elucidated. at oscillation minima correlates with adaptation of AC and GC, respectively. In agreement with this idea, Van Haastert (1987) has found that adaptation of the GC References response in batch cultures of D. discoideum is correlated AECKERLE, S., WURSTER, B. AND MALCHOW, D. (1985). Oscillations with a reduction in binding to the SS receptor form. + and cAMP-induced changes of the K concentration in In isolated membranes of D. discoideum, the guanine Dictyostelium discoideum. EMBOJ. 4, 9-43. nucleotides GTP and GDP have been shown to reduce BUMANN, J., MALCHOW, D. AND WURSTER, B. (1986). Oscillations of 2+ the affinity of the chemotactic receptor for cyclic AMP Ca concentration during the cell differentiation of Dictyostelium discoideum. Their relation to oscillations in cAMP and other (Janssens et al. 1985, 1986). The nucleotides are believed components. Differentiation 31, 85-91. to bind to a G protein at the cytoplasmic side of the DE LEAN, A., STADEL, J. M. AND LEFKOWITZ, R. J. (1980). A plasma membrane. We found that addition of GTPyS (a ternary complex model explains the agonist-specific binding non-hydrolysable GTP-analogue), or GDP, to cell properties of the adenylate cyclase-coupled /3-adrenergic receptor. lysates prepared from oscillating cells, completely abol- J. bwl. Chem. 255, 7108-7117. DEVREOTES, P. N. (1982). Chemotaxis. In The development of ished oscillations in receptor binding. In the presence of Dictyostelium discoideum (W. Loomis, ed.), pp. 117-168. either guanine nucleotide, the overall receptor affinity at Academic Press, NY, London. all time points of the oscillation cycle was reduced to the DEVREOTES, P. N. AND SHERRING, J. A. (1985). Kinetics and same value. This affinity value was close to the affinity concentration dependence of reversible cAMP-induced measured at oscillation minima in the absence of guanine modification of the surface cAMP receptor in Dictvostelium. J. biol. Chem. 260, 6378-6384. nucleotides. Thus, in lysates prepared from adapted DEVREOTES, P. N. AND STECK, T. L. (1979). Cyclic 3',5' AMP relay cells, i.e. at oscillation minima, GTPyS and GDP had in Dictyostelium discoideum. II. Requirements for the initiation hardly any effect on receptor affinity. On the other hand, and termination of the response. J. Cell Biol. 80, 300-309. in lysates from de-adapted cells, i.e. isolated at oscillation DINAUER, M. C, STECK, T. L. AND DEVREOTES, P. N. (1980). maxima, both nucleotides strongly reduced receptor Cyclic 3',5'AMP relay in Dictyostelium discoideum. IV. Recovery of the cAMP signaling response after adaptation to cAMP.,7. Cell affinity. We conclude that, at cyclic AMP receptor level, Biol. 86, 545-553. the process of cellular adaptation can be mimicked in EUROPE-FINNER, G. N. AND NEWELL, P. C. (1985). Inositol 1,4,5- vitiv by addition of GTP or GDP. trisphosphate induces cyclic GMP formation in Dictyostelium discoideum. Biochem. biophys. Res. Commun. 130, 1115-1122. The simplest interpretation of our results is that GERISCH, G. AND HESS, B. (1974). Cyclic AMP-controlled adaptation of cyclic AMP-induced responses in D. dis- oscillations in suspended Dictyostelium cells: their relation to coideum involves occupation of a G protein by GTP or morphogenetic cell interactions. Pivc. natn. Acad. Sci. U.SA. 71, GDP, resulting in alteration of the receptor-G protein 2118-2122. interaction, possibly by uncoupling (De Lean et al. GERISCH, G., MALCHOW, D., ROOS, W. AND WICK, U. (1979). Oscillations of cyclic nucleotide concentrations in relation to the 1980). Alternatively, it is possible that the reduction of excitability of Dictyostelium cells. J. exp. Biol. 81, 33-47. receptor affinity, which occurs during adaptation, is a GERISCH, G. AND WICK, U. (1975). Intracellular oscillations and consequence of the modification of receptor or G protein, release of cAMP from Dictvostelium cells. Biochem. biophvs. Res. e.g. phosphorylation. In this latter model, our present Comm. 65, 364-370. GREEN, A. A. AND NEWELL, P. C. (1975). Evidence for the existence results imply that the putative modification affects the of two types of cAMP binding sites in aggregating cells of interaction between receptor and G protein as does Dictyostelium discoideum. Cell 6, 129-136. binding of GTP or GDP. JANSSENS, P. M. W. (1987). Did vertebrate signal transduction Evidence has been presented in support of a role for mechanisms originate in eukaryotic microbes? Trends biochem. Sci. receptor modification in the adaptation process of D. 12, 456-459. JANSSENS, P. M. W., ARENTS, J. C, VAN HAASTERT, P. J. M. AND discoideum. Extracellular cyclic AMP has been shown to VAN DRIEL, R. (1986). Forms of the chemotactic adenosine 3'-5'- induce a reversible modification of the chemotactic recep- cyclic phosphate receptor in isolated Dictyostelium discoideum tor in vivo that alters its electrophoretic mobility in membranes and interconversions induced by guanine nucleotides. SDS-PAGE (Devreotes and Sherring, 1985; C. Kleins Biochemistry 25, 1314-1320. JANSSENS, P. M. W. AND DE JONG, C. C. C. (1988). A magnesium- al. 1985; P. Klein et al. 19856). This modification dependent guanylate cyclase in cell-free preparations of appears to be associated with an increased level of Dictvostelium discoideum. Biochem. biophvs. Res. Commun. 150, phosphorylation of the receptor and was proposed to be 405-411. the biochemical mechanism of adaptation (P. Klein et al. JANSSENS, P. M. W., VAN DER GEER, P. L., ARENTS, J. C. AND VAN DRIEL, R. (1985). Guanine nucleotides modulate the function 1985a,b), analogous to what has been found for other of chemotactic receptors in Dictvostelium discoideum. Molec. cell eukaryotic systems (Sibley et al. 1987). Interestingly, Biochem. 67, 119-124. both the kinetics and the cyclic AMP dose dependency of KESBEKE, F. AND VAN HAASTERT, P. J. M. (1985). Selective down- this receptor modification correlate closely with those of regulation of cell surface cAMP-binding sites and cAMP-induced the periodic decrease in receptor affinity that we describe responses in Dictvostelium discoideum. Biochim. biophvs. Ada 847, 33-39. here. KLEIN, C, BRACHET, P. AND DARMON, M. (1977). Periodic changes Summarizing, our results indicate that in D. discoid- in adenylate cyclase and cAMP-receptors in Dictyostelium eum, at receptor level, the adapted state is characterized discoideum. FEBS Lett. 76, 145-147.

628 M. E. E. Luderus et al. KLEIN, C, LUBS-HAUKENESS, J. AND SIMONS, S. (1985). cAMP VAN HAASTERT, P. J. M. (1984). Guanine nucleotides modulate cell induces a rapid and reversible modification of the chemotactic surface cAMP binding sites in membranes from Dictyostelium receptor in Dictyostelium discoideum. J. Cell Biol. 100, 715-720. discoideum. Biochem. biophys. Res. Commun. 124, 597-604. KLEIN, P., FONTANA, D., KNOX, B., THEIBERT, A. AND DEVREOTES, VAN HAASTERT, P. J. M. (1987). Kinetics and concentration P. (1985a). cAMP receptors controlling cell-cell interactions in the dependency of cAMP-induced desensitization of a subpopulation of development of Dictyostelium. In Cold Spring Harbor Symposia on surface cAMP receptors in Dictvostelium discoideum. Biochemistiy Quantitative Biology, vol. L, Molecular Biology of Development. 26, 7518-7523. Cold Spring Harbor Laboratory Press, NY. VAN HAASTERT, P. J. M. AND DE WIT, R. J. W. (1984). KLEIN, P., THEIBERT, A., FONTANA, D. AND DEVREOTES, P. N. Demonstration of receptor heterogeneity and affinity modulation (19856). Identification and cyclic AMP-induced modification of the by nonequihbnum binding experiments. The cell surface receptor cyclic AMP receptor in Dictyostelium discoideum. J. biol. Chem. of Dictvostelium discoideum. j. biol. Chem. 259, 13 321-13 328. 260, 1757-1764. VAN HAASTERT, P. J. M., DE WIT, R. J. W., JANSSENS, P. M. W., KLEIN, P. S., SUN, T. J., SAXE III, C. L., KIMMEL, A. R., KESBEKE, F. AND DE GOEDE, J. (1986). G protein mediated JOHNSON, R. L. AND DEVREOTES, P. N. (1988). A chemoattractant interconversions of cell-surface cAMP receptors and their receptor controls development in Dictvosteliuin discoideum. involvement in excitation and desensitization of guanylate cyclase Science 241, 1467-1472. in Dictyostelium discoideum. J. biol. Chem. 261, 6904-6911. KUMAGAl, A., PUPILLO, M., GUNDERSEN, R., MlAKE-LYE, R., VAN HAASTERT, P. J. M. AND KIEN, E. (1983). Binding of cAMP- DEVREOTES, P. N. AND FIRTEL, R. A. (1989). Regulation and derivatives to Dictyostelium discoideum cells. Activation function of Ga protein subunits in Dictvosteliuin. Cell 57, mechanisms of the cell surface cAMP receptor. J. biol. Chem. 258, 265-275. 9636-9642. MALCHOW, D., NAGELE, B., SCHWARZ, H. AND GERISCH, G. (1972). VAN HAASTERT, P. J. M., SNAAR-JAGALSKA, B. E. AND JANSSENS, P. Membrane-bound cyclic AMP phosphodiesterase in M. W. (1987). The regulation of adenylate cyclase by guanine chemotactically responding cells of Dictvostelium discoideum. Eur. nucleotides in Dictvostelium discoideum membranes. Eur. J. J. Biochem. 28, 136-142. Biochem. 162, 251-258. MALCHOW, D., NANJUNDIAH, V. AND GERISCH, G. (1978). pH VAN LOOKEREN CAMPAGNE, M. M., ERNEUX, C, VAN EUK, R. AND oscillations in cell suspensions of Dictyostelium discoideum: their VAN HAASTERT, P. J. M. (1988). Two dephosphorylation pathways relation to cAMP signals. J. Cell Sci. 30, 319-330. of inositol, l,4,5-tri8phosphate in homogenates of the cellular slime MATO, J. M., KRENS, F. A., VAN HAASTERT, P. J. M. AND KONUN, mould Dictyostelium discoideum. Biochem. J. 254, 343-350. T. M. (1977). 3',5'cycliccAMP-dependent 3',5'cychc GMP WATTS, D. J. AND ASHWORTH, A. M. (1970). Growth of accumulation in Dictvostelium discoideum. Proc. natn. Acad. Sci. myxamoebae of the cellular slime mould Dictyostelium discoideum U.SA. 74,2348-2351. in axenic culture. Biochem. J. 119, 171-174. Roos, W., SCHEIDEGGER, C. AND GERISCH, G. (1977). Adenylate WICK, U., MALCHOW, D. AND GERISCH, G. (1978). Cyclic AMP cyclase activity oscillations as signals for cell aggregation in stimulated calcium influx into aggregating cells of Dictyostelium Dictyostelium discoideum. Nature, Land. 266, 259-261. discoideum. Cell Biol. Int. Rep. 2, 71-79. SIBLEY, D. R., BENOVIC, J. L., CARON, M. G. AND LEFKOWTTZ, R. WURSTER, B., SCHUB1GER, K., WlCK, U. AND GERISCH, G. (1977). J. (1987). Regulation of transmembrane signaling by receptor Cyclic GMP in Dictyostelium discoideum. Oscillations and pulses phosphorylation. Cell 48, 913-922. in response to folic acid and cAMP signals. FEBS Lett. 76, SMALL, N. V., EUROPE-FINNER, G. N. AND NEWELL, P. C. (1987). 141-144. Adaptation to chemotactic cyclic AMP signals in Dictyostelium involves the G protein, jf. Cell Sci. 88, 537-545. SuSSMAN, M. (1982). Morphogenetic signaling, cytodifferentiation, and gene expression. In The development of Dictyostelium discoideum (Loomis, F., ed.), pp. 353-385. Academic Press, NY, London. (Received 17 November 19S9 -Accepted 17 January 1990)

Cyclic AMP receptor pivperties during adaptation 629