Regulation of Protein Phosphorylation in Dictyostelium discoideum ALISON ANSCHUTZ, HONG-DUCK UM, YOUNG-PING TAO, AND CLAUDETTE KLEIN E.A. Doisy Department of Biochemistry and Molecular Biology, 8t. Louis University Medical School, St. Louis, Missouri We have examined the phosINTRODUCTION ABSTRACT phorylation of the cyclic adenosine 3 ’ 5 ’ monoMetazoan development involves the coordination phosphate (CAMP)cell surface chemotactic recepand integration of varied signals which regulate cell tor and a 36 kDa membrane-associated protein growth, motility, and differentiation. A well-docu(p36) in Dictyostelium discoideum. The activity of mented response to those signals is that of changes CAR-kinase, the enzyme responsible for the phosin protein phosphorylationldephosphorylation. Memphorylation of the cAMP receptor, was studied in brane bound signal receptors are phosphorylated in replasma membrane preparations. It was found that, sponse to ligand binding and the second messengers as in intact cells, the receptor was rapidly phosgenerated function, at least in part, by the activation of phorylated in membranes incubated with [ Y ~ ~ P ] specific protein kinases [see Sommercorn and Krebs, adenosine triphosphate (ATP) but only in the pres1988 for review]. The interaction of second messenger ence of CAMP. This phosphorylation was not obpathways provides a n extensive network by which served in membranes prepared from cells which did physiological processes can be regulated and coordinot display significant cAMP binding activity. cAMP nated to ensure appropriate cell growth and developcould induce receptor phosphorylation at low conment. With the growing recognition that phosphorylacentrations, while cyclic guanosine 3 ’ 5 ’ monotionldephosphorylation represents a fundamental and phosphate (cGMP)could elicit receptor phosphorpowerful mechanism by which the activities of specific ylation only at high concentrations. Neither ConA, cellular components can be regulated, the list of such Ca2+, or guanine nucleotides had an effect on kinases and their sub-types continues to grow. The CAR-kinase. It was also abserved that 2-deoxy wide categories of tyrosine and serinelthreonine kicAMP but not dibutyryl cAMP induced receptor nases include receptor kinases and the enzymes reguphosphorylation. The data suggest that the ligand lated by Ca2+, cyclic adenosine 3’5’ monophosphate occupied form of the cAMP receptor is required for (CAMP),cyclic guanosine 3 ’ 5 ’ monophosphate (cGMP). CAR-kinase activity. Although the receptor is rapProtein phosphatases which can specifically act on eiidly dephosphorylated in vivo, we were unable to ther phosphorylated tyrosine or serinelthreonine resiobserve its dephosphorylation in vitro. In contrast, dues have also been characterized in a number of sysp36 was rapidly dephosphorylated. Also, unlike the tems. Categories of such enzymes include type-1 and cAMP receptor, the phosphorylation of p36 was type 2-B which are serinelthreonine phosphatases that found to be regulated by the addition of guanine can be distinguished by their regulation by second mesnucleotides. Guanosine diphosphate (GDP) ensengers such a s cAMP and Ca2+ [see Cohen, 1988 for hanced the phosphorylation while guanosine review]. The majority of the intracellular targets for triphosphate (GTP)decreased the radiolabeling of many of these enzymes have not been identified [Blackp36 indicating that GTP con compete with ATP for shear et al., 1988; Shenoliker, 19881. Thus the elucidathe nucleotide triphosphate binding site of p36 kition of specific kinases, andlor phosphatases, and their nose. This was verified using radiolabeled GTP as targets remains a goal of many investigators. In this the phosphate donor. Competition experiments manuscript, we will describe the phosphorylation that with GTPyS, ATP, GTP, CTP, and uridine triphosresults from the activation of CAR-kinase [Meier and phote (UTP) indicated that the phosphate donor site Klein, 19881 and p36 kinase [Anschutz et al., 19891 and of p36 kinase is relatively non-sepcific. The mechonism(s) by which GDP functions to alter p36 phosphorylation and the physiological significance of this event are currently under investigation. Received for publication April 30, 1990

Key words: GDP-dependent, chemotactic receptor, CAR-kinase


Address reprint requests to Dr. Claudette Klein, E.A. Doisy Department of Biochemistry and Molecular Biology, St. Louis University Medical School, St. Louis, MO 63119.

PROTEIN PHOSPHORYLATION IN DICTYOSTELIUM their corresponding phosphatases. The substrates for these kinases are very specific and are, respectively, the cAMP cell surface chemotactic receptor and a 36 kDa membrane-associated protein in D . discoideum.

MATERIALS AND METHODS Conditions for the growth and starvation of Ax-2 amoebae, and their in vivo labeling with [32P]-P04 have been described [Lubs-Haukeness and Klein, 19821. Preparation of plasma membrane enriched fractions and the conditions of incubation for the phosphorylation of the cAMP receptor [Meier and Klein, 19881 or p36 [Anschutz et al., 19891 were carried out as described in those references. Products were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (Laemmli, 1970) and the labeled proteins identified by autoradiography on XR-5 film (Kodak). Radionucleotides were purchased from New England Nuclear. All other compounds were obtained from Sigma Chemicals. RESULTS Photoaffinity-labeling of intact aggregation competent amoebae has demonstrated the presence of two molecular weight forms of the cell surface cAMP receptor [Juliani and Klein, 1981; Devreotes and Sherring, 1985; Lubs-Haukeness et al., 19853. The higher molecular weight form is induced by cAMP stimulation and has been referred to as either p47 [Juliani and Klein, 1981; Lubs-Haukeness et al., 19851 or the D form [Devreotes and Sherring, 19851. Correspondingly, the lower molecular weight form has been referred to a s p45 or R, respectively. For the purposes of this discussion, we shall employ our original nomenclature and refer to p45 and p47. In vivo [32P]-P04labeling experiments have indicated that p47 represents a more highly phosphorylated form of the receptor [LubsHaukeness and Klein, 1982; Lubs-Haukeness et al., 1985; Klein et al., 19871 and that this increased phosphorylation is correlated with the phenomenon of desensitization [Lubs-Haukeness et al., 1985; Klein et al., 19871. Our approach to characterizing CAR-kinase, the enzyme responsible for the phosphorylation of the receptor [Meier and Klein, 19881, has been to study its regulation in plasma membrane preparations. When the plasma membrane-enriched fraction prepared from aggregation competent cells is incubated with [y32P]-adenosinetriphosphate (ATP), a number of proteins are phosphorylated. One does not observe, however, the presence of a radiolabeled protein of 47 kDa, p47, unless the incubation is performed in the presence of cAMP (Fig. 1, compare lanes 3, 4). The phosphorylation of this protein is not observed in membranes prepared from vegetative cells or from cells that have been starved for only a short period of time and do not display significant cAMP binding activity (lanes 1, 2).






Fig. 1. CAR-kinase in plasma membranes. Plasma membranes were prepared from vegetative (lanes 1, 2) and aggregation competent (lanes 3, 4) cells. Phosphorylation of membrane proteins was carried out for 2 minutes a t 21°C in the absence (lanes 1, 3) or presence (lanes 2, 4) of M CAMP.The reaction was stopped by the addition of sample buffer and the phosphorylated proteins analyzed by SDS-PAGE and autoradiography. The arrow points to the position of p47.

In vivo, phosphorylation of the cAMP receptor is very rapid and can be observed within a few seconds of cell stimulation with cAMP [Lubs-Haukeness and Klein, 19821. Upon hydrolysis of the ligand, the receptor is also rapidly dephosphorylated [Lubs-Haukeness and Klein, 1982; Lubs-Haukeness et al., 19851. Phosphorylation of the cAMP receptor in plasma membrane preparations is also very rapid and can be detected within 5-10 seconds of incubation. However, we have not yet observed a dephosphorylation of the receptor in membrane preparations. Reconstituting membrane and cytosolic fractions, as well a s manipulating the membrane incubation conditions, have not allowed us to demonstrate a receptor phosphatase. That no receptor phosphatase activity is present, however, indicates that cAMP functions to increase receptor phosphorylation by activating CAR-kinase as opposed to inhibiting the dephosphorylation of the receptor. The ability of varied treatments to induce receptor phosphorylation was monitored in membrane preparations and compared to their effectiveness when intact cells were similarly treated. The results in both cases were identical and are summarized in Table 1. CAMP, even a t concentrations a s low a s lo-? M, effectively induces receptor phosphorylation while cGMP, which can bind to the cAMP receptor when present at high concentrations, can only elicit receptor phosphorylation a t those high levels. 5’AMP, the cAMP metabolite produced by the action of the cAMP phosphodiesterase and which does not bind to the receptor, is ineffective. The Ca2+ ionophore, A32187, has been shown to influence the development of the cells and, in so doing, will indirectly affect the expression of the cell surface cAMP binding activity which is also develcpmentally regulated [Brachet and Klein, 1977; Juliani and Klein, 19771. This compound, however, does not bind to the



TABLE 1. Effectors of Receptor Phosphorylation


Effect Intact cellsa Membranesa

P36 PHOSF'HORYLATIm Relative Chenges ezI1R.kArr

lop7 M c~~~ 10-4 M 1 0 -M ~ 1 0 -M ~ 1 0 -M ~

c~~~ c~~~

c~~~ 5' AMP M folic acid lop4 M Ca2+ M A23187 100 g/ml ConA 10- PM 2-deoxycAMP M dibutvrvlcAMP a



phosphorylation occurred; -


no phosphorylation.

receptor itself, and correspondingly, does not induce receptor phosphorylation. ConA has been shown to induce membrane perturbations and indirectly alter the number of cAMP receptors on the cell surface [Darmon and Klein, 19761. The addition of Ca2' to the cAMP binding assay has been shown to result in a n increase in receptor activity [Juliani and Klein, 19771. Folic acid is also a chemoattractant but its effects are mediated via a different receptor [Wurster and Butz, 19801. None of these treatments induce the phosphorylation of the cAMP chemotactic receptor. The data indicate that the regulation of CAR-kinase in membrane preparations is a n accurate reflection of that which occurs in intact cells. Additionally, they demonstrate that binding of cAMP to the receptor is necessary for i t to be phosphorylated by CAR-kinase. To eliminate the possibility that cAMP was required to activate a CAMPdependent protein kinase (A-kinase) that could be present in our membrane preparations, the ability of 2-deoxy cAMP and dibutyryl cAMP to stimulate receptor phosphorylation was examined. The former compound, which is a ligand for the receptor and not for A-kinase [Van Haastert and Kein, 1983; Oyama and Blumberg, 19861elicits receptor phosphorylation while the latter, which does not bind to the receptor but which does activate A-kinase [Van Haastert and Kein, 1983; Oyama and Blumberg, 19861 does not. Taken together, the data provide strong evidence that the ligand-occupied form of the cAMP receptor is required for CAR-kinase activity, a requirement faithfully replicated in our membrane preparations. The activity of CAR-kinase, as expressed in membrane preparations, is not affected by the addition of guanine nucleotides to the incubation mixture, in particular guanosine diphosphate (GDP) or guanosine triphosphate (GTP). Thus i t would not appear that the enzyme is regulated by a G-protein that may be present in the membrane preparation. The finding that GTP does not affect the level of receptor phosphorylation would also argue that it is not a good competitor for the nucleotide triphosphate binding site of CAR-kinase. In








Fig. 2. Effectors of p36 phosphorylation. Phosphorylation of p36 in the presence or absence of 2 x M of the indicated compounds was determined by SDS-PAGE and autoradiography. The reaction was carried out for increasing time periods to assure its linearity. Autoradiograms were analyzed by densitometry and plotted as arbitrary units.

contrast to these results, phosphorylation of a 36 kDa membrane protein, p 36, was found to be regulated by the addition of guanine nucleotides. As shown in Figure 2, GDP enhances p36 phosphorylation. Maximum stimulation is observed at approximately 2-4 x lop6 M GDP. The effect of GDP on p36 phosphorylation is quite specific: no other proteins show enhanced radiolabeling upon addition of GDP. Other compounds, including CAMP,are ineffective in modulating p36 phosphorylation. The inability of cAMP to stimulate p36 phosphorylation would argue that CAR-kinase is not responsible for p36 phosphorylation. Additionally, it was noted that GTP actually decreased the radiolabeling of p36 indicating that GTP can compete with ATP for the nucleotide triphosphate binding site of p36 kinase. This has been verified using radiolabeled GTP as the phosphate donor. The ability of p36 kinase to utilize either ATP or GTP as a phosphate donor contrasts with the specificity of CAR-kinase for ATP to phosphorylate the cAMP chemotactic receptor and would again argue against the role of that kinase in p36 phosphorylation. The specificity of the nucleotide triphosphate binding site of p36 kinase was explored in competition experiments and the results are shown in Figure 3. The addition of nucleotide triphosphates to the reaction mixture reduces the extent of radiolabel incorporated into p36 by [ Y ~ ~ P I - A TThe P . relative order of effectiveness is GTPyS > ATP > GTP = CTP = UTP. It would appear from these results that the phosphate donor site of p36 kinase is relatively non-specific. Membrane preparations also contain a phosphatase for p36 (Fig. 4).When p36 is radiolabeled for 2 minutes and excess non-radioactive ATP added, a rapid loss of radiolabel from the protein is observed. By 7 minutes of chase, no radiolabeled p36 is detected. Dephosphorylation of p36 is very specific and does not involve the



P36 Fhosphorylatton using A T P XTP Carpetition =ATP

m G T P


m C T P



~ G I P I E O Q








Fig. 3. Specificity of the phosphate donor site of p36 kinase. Radiolabeling of p36 was performed by incubating membranes with [y3'P]ATP in the presence of the indicated non-radioactive nucleotide triphosphates. Samples were analyzed by SDS-PAGE and autoradiography. Percent phosphorylation refers to the percent of radiolabel incorporated into p36 compared to incubations performed in the absence of added non-radioactive nucleotides.

detectable loss of radiolabel from any other proteins. We also examined if the cAMP receptor could be dephosphorylntcd under these exact conditions. Mcmbranes were incubated with [ Y ~ ~ P I - A and T P cAMP to allow both the receptor and p36 to become phosphorylated. Upon addition of non-radioactive ATP, p36 was rapidly dephosphorylated but the radiolabel in the receptor remained stable. Thus, both the kinase and phosphatase for p36 are unique and specific for this protein. The addition of GDP does not inhibit the phosphatase for p36, indicating that the increased phosphorylation of the protein under those conditions cannot be accounted for by a n inhibition of its dephosphorylation. Currently, we do not know the mechanism(s) by which GDP functions to enhance p36 phosphorylation. It may bind to, and, therefore, directly activate a protein kinase. If so, p36 kinase would represent a novel second-messenger dependent kinase. GDP is a well-known effector of signal transduction, eliciting its effects through a family of guanine-nucleotide binding proteins called G-proteins [reviewed by Bourne, 19881. G-proteins are activated by the exchange of bound GDP for GTP and their subsequent interaction with the appropriate effector molecule. Such released GDP could provide a local source for the activation of p36 kinase. Alternatively, GDP could bind to p36 and render i t a better substrate for p36 kinase. Under such conditions, the enzyme would more likely simulate that of receptor kinases in which the ligand-occupied form of the protein is the preferred substrate. In either case, the finding that GDP is a n effector of protein phosphorylation represents a novel regulatory role for this guanine nucleotide. How GDP functions to alter p36 phosphorylation, and the physi-

Fig. 4. P36 phosphatase in plasma membranes. P36 was phosphorylated with [y3'P]-ATP (lane 1). After a 2 minute incubation, M non-radioactive ATP was added and additional points taken after 1 minute (lane 2), 2 minutes (lane 3), 3 minutes (lane 4), 5 minutes (lane 5), 7 minutes (lane 6), or 10 minutes (lane 7). Samples were analyzed by SDS-PAGE and autoradiography.

ological significance of that event, are under current investigation.

ACKNOWLEDGMENTS This research was supported by research grants from the National Science Foundation and the American Cancer Society. A.A. is a predoctoral trainee on the NIH training grant HL07050.

REFERENCES Anschutz A, Howlet A, Klein C (1989): GDP stimulates the phosphorylation of a 36-kDa membrane protein in Dictyostelium discoideum. Proc Natl Acad Sci USA 86:3665-3668. Blackshear PJ, Nairn AG, Keeo JF (1988): Protein kinases 1988: A current perspective. FASEB J 14:2957-2969. Bourne H (1988): Do GTPases direct membrane traffc secretion? Cell 53:669-67 1. Brachet P, Klein C (1977): Cell responsiveness to cAMP during the aggregation phase of Dictyostelium drscoideum. Differentiation 8:

1-8. Cohen PFRS (1988): Protein phosphorylation and hormone action. Proc R SOCLond B234:115-144. Darmon M, Klein C (1976): Binding of concanavalin A and its effect on the differentiation of Dictyosteliurn discodeurn. Biochem J 154: 743-750. Devreotes PN, Sherring JA (1985):Kinetics and concentration depen-



dence of reversible CAMP-induced modification of the surface cAMP receptor in Dictyostelium. J Biol Chem 260:6378-6384. Juliani MH, Klein C (1977): Calcium ion effects on cyclic adenosine 3’5’-monosphosphate binding to the plasma membrane of Dictyostelium discoideum. Biochim Biophis Acta 497:369-376. Juliani MH, Klein C (1981): Photoafinity labeling of the cell surface adenosine 3’:5’-monophosphate receptor of Dictyostelium discozdeum and its modification in down-regulated cells. J Biol Chem 256:613-619. Klein C, Lubs-Haukeness J , Simons S (19851: cAMP induces a rapid and reversible modification of the chemotactic receptor in Dictyosteliurn discoideum. J Cell Biol 100:715-720. Klein P, Knox B, Borleis J, Devreotes P (1987): Purification of the surface cAMP receptor in Dictyostelium. J Biol Chem 262:352-357. Laemmli UK (1970): Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-685. Lubs-Haukeness J, Klein C (1982):Cyclic nucleotide-dependent phosphorylation in Dictyostelium discoideum amoebae. J Biol Chem 257:12204-12208.

Lubs-Haukeness J , Simons S, Klein C (1985): cAMP induces a rapid and reversible modification of the chemotactic receptor in the D. discoideum. J Cell Bio 100:715-720. Meier K, Klein C (1988): An unusual protein kinase phosphorylates the chemotactic receptor of Dictyostelium discoideum. Proc Natl Acad Sci USA 85:2181-2185. Oyama M, Blumberg D (19861: Interaction of cAMP with the cellsurface receptor induces cell-type-specific mRNA accumulation in Dictyostelium discoideum. Proc Natl Acad Sci USA 83:4819-4823. Shenoliker S (1988): Protein phosphorylation: Hormones, drugs and bioregulation. FASEB J 12:2753-2764. Sommercorn J , Krebs EG (1988):In Zappia V,Galletti P, Porta R, and Wold F (eds): “Advances in Post-Translational Modifications of proteins and Aging.” New York: Plenum Press pp 403-415. Van Haastert PJM, Kein E (1983): Binding of cAMP derivatives to Dictyostelium discoideum cells. J Biol Chem 258:9636-9642. Wurster B, Butz U (1980):Reversible binding of the chemoattractant folic acid to cells of Dictyostelium discoideum. Eur J Biochem 109: 613-618.

Regulation of protein phosphorylation in Dictyostelium discoideum.

We have examined the phosphorylation of the cyclic adenosine 3':5' monophosphate (cAMP) cell surface chemotactic receptor and a 36 kDa membrane-associ...
513KB Sizes 0 Downloads 0 Views