EXPERIMENTAL
CELL
RESEARCH
199,
162-168
(19%‘)
Regulation of Adenylate Cyclase in Electropermeabilized Dictyostelium ciiscoideum Cells CORD.SCHOEN,~TACOBRUIN,'JOSC.ARENTS,ANDROELVANDRIEL~ E. C. Slater Institute
for Biochemical
Research, University
of Amsterdam,
In Dictyostelium discoideum cells the enzyme adenylate cyclase is functionally coupled to cell surface receptors for CAMP. Coupling is known to involve one or more G-proteins. Receptor-mediated activation of adenylate cyclase is subject to adaptation. In this study we employ an electropermeabilized cell system to investigate regulation of D. discoideum adenylate cyclase. Conditions for selective permeabilization of the plasma membrane have been described by C. D. Schoen, J. C. Arents, T. Bruin, and R. Van Driel (1989, Exp. Cell Res. 181, 51-62). Only small pores are created in the membrane, allowing exchange of exclusively low molecular weight substances like nucleotides, and preventing the loss of macromolecules. Under these conditions functional protein-protein interactions are likely to remain intact. Adenylate cyclase in permeabilized cells was activated by the CAMP receptor agonist 2’-deoxy CAMP and by the nonhydrolyzable GTP-analogue GTP-yS, which activates G-proteins. The time course of the adenylate cyclase reaction in permeabilized cells was similar to that of intact cells. Maximal adenylate cyclase activity was observed if CAMP receptor agonist or GTP-analogue was added just before cell permeabilization. If these activators were added after permeabilization adenylate cyclase was stimulated in a suboptimal way. The sensitivity of adenylate cyclase activity for receptor occupation was found to decay more rapidly than that for G-protein activation. Importantly, the adenylate cyclase reaction in permeabilized cells was subject to an adaptation-like process that was characterized by a time course similar to adaptation in vivo. In vitro adaptation was not affected by CAMP receptor agonists or by G-protein activation. Evidently electropermeabilized cells constitute an excellent system for investigating the positive and negative regulation of D. o 1992 Academic PWS, I~C. discoideum adenylate cyclase.
INTRODUCTION Starvation of Dictyostelium gers a specific developmental
discoideum amoebae trigprogram. Within 5 to 7 h
’ Present address: Research Institute for Plant Protection, Binnenhaven 12,6709 PD Wageningen, The Netherlands. ’ Present address: Academic Medical Centre, Meibergdreef 15, 1105 AZ Amsterdam, The Netherlands. 3 To whom reprint requests should be addressed. 0014.4827/92 $3.00 Copyright 0 1992 by Academic Press, All rights of reproduction in any form
Plantage Muidergracht
The Netherlands
after the onset of starvation, amoebae start to aggregate into multicellular structures. This aggregation process is controlled by extracellular CAMP signals [ 1,2]. CAMP binds to highly specific cell-surface CAMP receptors which are functionally coupled to several enzymes, including adenylate cyclase and guanylate cyclase, reviewed in Refs. [3-71. Cyclic GMP that is produced by guanylate cyclase remains mostly intracellular and is probably involved in transduction of the chemotactic signal [8-131. Most of the CAMP produced by adenylate cyclase is secreted, thereby relaying the CAMP signal to neighboring cells [ 141. In intact D. discoideum cells regulation of CAMP receptor-coupled adenylate cyclase activity has been studied in detail [ 15-181. The enzyme is activated within 30 s after stimulation of cells with CAMP. The activity reaches a maximum after 1 to 2 min and then rapidly declines to prestimulation levels. The decrease in adenylate cyclase activity is due to adaptation. Subsequently, extracellular CAMP is hydrolized by phosphodiesterases. The cells undergo deadaptation with a halftime of 3-4 min and become responsive to CAMP again [17]. Despite detailed knowledge of the physiology of adenylate cyclase stimulation the molecular details of the sensory transduction pathway from the cell surface receptor to adenylate cyclase in D. discoideum and the molecular mechanism of the adaptation process have not been elucidated yet. Binding of CAMP to D. discoideum cells is a complex phenomenon. The chemotactic CAMP receptor occurs in several interconvertible forms. In isolated membranes, interconversions between different receptor forms have been found to be induced by guanine nucleotides [ 19-221. This is analogous to what has been found for many vertebrate receptor systems and suggests that the CAMP receptor is coupled to one or more regulatory guanine nucleotide binding proteins, G-proteins. The existence of G-proteins in D. discoideum has recently been confirmed [6, 23, 241. In vertebrate cells, G-proteins transduce the signal from hormone receptor to adenylate cyclase and other effector systems. The effect on adenylate cyclase can be either stimulatory (G,) or inhibitory (Gi) [25-271. Results of Van Haastert et al. [28] and Theibert and Devreotes [29] suggest that in D. discoi162
Inc. reserved.
12, 1018 TV Amsterdam,
REGULATION
OF Dietyostelium
deum Gi and/or G,-like proteins regulate adenylate cyclase activity. Thus far, CAMP-induced activation of adenylate cyclase in D. discoideum could only be observed when intact cells were stimulated and the adenylate cyclase activity was assayed immediately after cell lysis. It has up to now not been possible to stimulate the enzyme via the CAMP receptor in broken cell preparations. The rationale of our present study is to look for a method that preserves cellular structures and functional proteinprotein interactions and, at the same time, allows manipulation and analysis of signal transduction processes in general and of the adenylate cyclase system in particular. To this end we have adapted the electropermeabilization method developed by Schoen et al. [30]. In this report we describe the effect of receptor occupation and G-protein activation on adenylate cyclase activity in electropermeabilized D. disco&urn cells. We show that receptor occupation and G-protein activation can induce adenylate cyclase activation in permeable cells. Furthermore, an adaptation-like process is observed in permeable cells, which results in enzyme inactivation with approximately the same time course as seen in vivo. MATERIALS
AND METHODS
Materials. 2’.Deoxyadenosine 3’,5’-monophosphate (2’-deoxyCAMP), and adenosine 3’,5’-monophosphate (CAMP) were obtained from Boehringer (Mannheim, FRG); (2’,8’-3H)cAMP (0.9 TBq.mol-I), 2’-deoxy-D-(U-i4C)glucose (11.1 TBq.moll’) and 3H,0 (3.34 GBq.moll’) were purchased from Amersham (UK). Poly-carbonate filters with 5-pm pore size were from Nuclepore Corp. (Pleasanton, CA). Culture conditions. D. discoideum strain NC4(H) was cultured on standard medium (SM) agar [31] in combination with Escherichia coli B/r. Cells were harvested in the late logarithmic phase and washed free of nutrients in 15 mM potassium phosphate buffer (pH 6.2) (Pibuffer) three times by centrifugation at 150g for 4 min. Washed cells were resuspended in Pi-buffer, spread evenly on nonnutrient agar plates (1.5% agar in Pi-buffer) at a density of 1.5 to 4.0 X lo6 cells/ cm2, and starved at 6°C for 16-20 h to obtain aggregation-competent cells 1321. Aggregation-competent cells were harvested, washed, and resuspended at a density of 2 X 10’ cells/ml in 20 mM MES’-NaOH buffer (pH 6.0). Before use the cell suspension was aerated continuously. Electropermeabilization of cells and preparation of a cell homogenate. Cells in 20 mM MES-NaOH (pH 6.0) were resuspended at a density of 2 X 10’ cells/ml in standard electropermeabilization buffer containing: 20 mM Hepes-NaOH (pH 6.6), 100 mM sucrose, 20 mM K-acetate, and 1 mM Mg-acetate. One milliliter of the cell suspension was pipetted into the electropermeabilization chamber between a pair of parallel platinum electrodes (each with a surface area of 1 cm*) positioned 1 cm apart. Permeabilization was
4 Abbreviations used: Hepes, N-(2-hydroxyethyl)piperazine-W2ethanesulfonic acid, MES, 2-(N-morpholino)ethanesulfonic acid; EGTA, ethylene glycol bis @-amino ethyl ether)-N,N,N’,N’-tetraacetic acid; EDTA, ethylenediaminotetraacetic acid; GTPyS, guanosine 5’-O-[3-thio] triphosphate; AppNHp, adenosine 5’-[&y,-imido] triphosphate.
ADENYLATE
163
CYCLASE
achieved by exposing the cells to seven brief, square wave pulses of 500 ps (rise and fall times approximately 50 ns) of 2 kV, at 0°C under continuous stirring [30]. During these seven discharges, applied with 3-s intervals, the temperature increased only 1 to 2’C. The extent of permeabilization was measured by ethidium bromide fluorescence. Nonpermeable cells exclude ethidium bromide and were not fluorescent, whereas permeable cells accumulate the dye [30]. Under standard conditions more than 98% of the cells were permeable. Cell homogenates were prepared by forcing cells through a 5-pm pore-size polycarbonate filter 128, 331. More than 98% of the cells were broken by this procedure as judged by phase contrast microsCOPY. Measurement of cell permeability. After electropermeabilization the permeability of cells was determined quantitatively by measuring the rate of influx of a radioactive, non-permeant, non-metabolisable substance, i.e., [‘?]deoxyglucose. The total intracellular volume was determined by measuring the uptake of the permeant molecule 3H,0. The permeant and the nonpermeant substances were mixed together and added to the cell suspension. At different time points the intraand extracellular compartments were separated by centrifugation of the cells through a layer of silicone oil [30,34]. From the radioactivity in the pellet, the intracellular volume accessible by [r4C]glucose was calculated. Corrections were made for radioactive label adhering to the outside of the cells during centrifugation through silicone oil. Measurement of adenylate cyclase activity. Adenylate cyclase activity was measured by a method described earlier [35]. The incubation mixture (500 ~1) contained 40 mM Hepes-NaOH (pH 7.7), 10.5 mM K-acetate, 3 mM EDTA, 100 mM sucrose, 20 mM dithiothreitol, 2 mM 3-isobutyl-l-methylxanthine, 2.5 mM phosphoenol-pyruvate, 1 U/ml pyruvate-kinase, and unless stated otherwise 0.5 mM ATP and 6 mM MgCl,. The assay was initiated by the addition of 250 ~1 of cell homogenate or permeabilized cells (2 X 10’ cells or cell equivalents per ml) and was carried out at 20°C under vigorous agitation. At various time points samples were quenched in perchloric acid (final concentration 1 IV) to measure CAMP levels. Quenched samples were centrifuged for 5 min at 10,OOOgand the perchloric acid extracts were neutralized by the addition of a solution containing KHCO, and KOH. Neutralized extracts were centrifuged for 2 min at 10,OOOgand CAMP concentrations were measured by an isotope dilution assay using bovine CAMP-dependent protein kinase as a CAMP binding protein [36]. Zntracelluhr ATP concentration. Freshly permeabilized D. discoideum cells (10’ cells/ml) were incubated for 1 min at 20°C with different concentrations of ATP (0,0.5,10, or 20 mM ATP). Subsequently, the intra- and extracellular compartments were separated by centrifugation (30 s at 10,OOOg) of the cells through a layer of silicone oil [30, 341 into 0.4 M HClO,. The HClO, extracts were directly neutralized and centrifuged (2 min at 10,OOOg).The ATP level in the supernatant, reflecting the amount of ATP originally accumulated by the cells was assayed by the method of Williamson and Corky [37]. The assay mixture (final volume of 2.2 ml) contained 50 mM Triethanolamine-HCl (pH 7.4), 4 mM EDTA, 12.5 mM MgCl,, 50 mM KCl, 1 mM glucose, 1 mM NADP, and 0.35 mU glucose 6-phosphate dehydrogenase. The assay was initiated by the addition of 100 ~1 of the neutralizedperchloric acid extract and was carried out at 20°C. After 10 min, 20 ~1 hexokinase (0.5 U) was added and fluorescence of NADPH (excitation 316366 nm and emission 300-400 nm) was measured after the reaction was complete. The intracellular ATP level was calculated from these data using standard solutions of known ATP concentration. Corrections were made for ATP adhering to the outside of the cells during centrifugation through silicone oil.
RESULTS Electropermeabilization of Dictyosielium Cells Conditions for electropermeabilization of D. discoideum cells have been described in detail by Schoen et al.
164
SCHOEN
ET AL
Adenylate
0.00 0
1
2
3 Time
4
5
6
(min)
FIG. 1. Permeability of electropermeabilized cells. Aggregationcompetent cells were electropermeabilized under standard conditions by 7 discharges of 2 kV and 0.5 ms (0, +). Control cells (0) were treated in the same way but did not receive high voltage pulses. Permeabilized cells were either incubated directly (0) or after 2 min (+) with [%]deoxyglucose (2 mM, 2 X lo4 cpm/ml) and aH,O (2 X lo5 cpm/ml) at 20°C in adenylate cyclase reaction mixture. Control cells were incubated under the same conditions directly after a mock-electropermeabilization. The intracellular [3H] and [‘%]radioactivities were determined after separating the cells from the extracellular medium by centrifugation through a layer of silicone oil [30,34]. Corrections were made for adhering extracellular material during centrifugation. The ratio of intracellular, tritiated water (having access to all cellular compartments) and intracellular [Wldeoxyglucose (entering the cell exclusively through plasma-membrane pores created by electropermeabilization) has been plotted as a function of the time after permeabilization. The arrow marks the “C/3H ratio in the extracellular medium, representing the theoretical maximum that would be reached if [WJdeoxyglucose would have access to the same cellular compartments as ‘H,O.
Cyclase Activity
in Electropermeabilized
Cells
During initial experiments we noticed that near-physiological Mg’+-ATP concentrations (around the 1 m&f) in the medium resulted in little or no adenylate cyclase activity in permeabiiized cells. Much higher concentrations were required. We have, therefore, measured the ATP concentration inside the permeabilized cells as a function of the ATP level in the medium. Table 1 shows that an extracellular concentration of 20 mMis required to obtain a physiological ATP concentration inside the permeable cell (about 0.8 mM). The drop in concentration across the permeable cell membrane is probably due to highly active ATPases. Apparently, the rate of influx of ATP is low compared to the rate of ATP hydrolysis inside the permeable cell, creating locally a lower than ambient concentration. Figure 2a (open circles) shows the CAMP synthesis of permeable cells immediately after electropermeabilization. Adenylate cyclase activity was completely dependent on a high extracellular ATP concentration (20 mM). No adenylate cyclase activity was observed at 0.5 mM ATP (solid circles in Fig. 2a). This indicates that CAMP is produced by permeable cells, rather than by some residual intact cells. Two important observations can be made on cells permeabilized under standard condition. First, after about 2 min the adenylate cyclase reaction terminates. This time course is similar to that of the CAMP response of intact cells (Fig. 2b, solid circles). In vivo termination of the adenylate cyclase reaction is due to an adaptation process [17, 181. This process appeared to be still operative in permeable cells. The rate of adaptation in vivo and in permeabilized cells is about the same (t,., of about 1 min). In permeabilized
TABLE Intracellular Function
1
ATP Concentration in Permeable of the ATP Concentration in the
Cells
as a
Medium
ATP
[30]. For this study we have selected electropermeabilization conditions for which cells become accessible for low molecular weight substances, such as nucleotides, but retain their intracellular macromolecules. Furthermore, their organelles remain intact [30]. Figure 1 shows that after permeabilization about 75% of the total cellular volume, i.e., the volume that is accessible for the permeant molecule 3H,0 becomes accessible for the nonpermeant substance [14C]deoxyglucose. The remaining 25% probably reflects the intact organelles and vacuoles in the cell. After 2 min the plasmamembrane pores are still open, although the rate of influx of [‘“Cldeoxyglucose has decreased slightly.
ATP concentration in the medium (mm Intact cells Permeable Permeable Permeable Permeable
cells cells cells cells
0 0 0.5 10 20
Intracellular concentration
ATP (mM)
0.75 0.10 0.09 0.21 0.80
Note. Directly after electropermeabilization D. disco&urn cells were incubated for 1 min in adenylate cyclase reaction medium containing different MgZ+-ATP concentrations (0, 0.5, 10, or 20 mM) at 20°C. The intracellular ATP level was determined after separation of the intra- and extracellular compartments by centrifugation of the cells through a layer of silicone oil [30, 341 into 0.4 h4 HClO,. The ATP concentrations were measured as indicated under Materials and Methods. The results were reproduced once.
REGULATION
OF Dictyostelium
ADENYLATE
165
CYCLASE
8C I-
a
120
b
c
a
I
0 0
50
100
150
200
Time
250
300
350
400
450
500
(seconds)
0
50
100
150
200
Time
250
300
350
400
450
500
(seconds)
FIG. 2. Cyclic AMP synthesis in permeabilized cells, intact cells, and a cell homogenate. (a) Permeabilized cells. Cyclic AMP synthesis in permeabilized cells is shown as a function of time after the initiation of the adenylate cyclase reaction at t = 0, immediately after electropermeabilization. Aggregation-competent cells were electropermeabilized by 7 pulses of 2 kV and 0.5 ms in the presence of 10 gM 2’-deoxy CAMP. After permeabilization the adenylate cyclase reaction was started immediately in the presence of 0.5 mM Mg’+-ATP (0) or 20 mM Mg’+-ATP (0). The adenylate cyclase activity in cells permeabilized under more harsh conditions (20 pulses of 2 kV and 0.5 ms) in the presence of 2’-deoxy CAMP (10 pM) was measured in the presence of 20 mM Mg’+-ATP (Cl). These latter cells were more permeable than cells treated under standard conditions, since they had lost most soluble cytoplasmatic proteins (data not shown). The results shown are the average of two independent experiments. (b) Intact cells and cell homogenate. Aggregation-competent D. discoideum cells (10s cells/ml) were stimulated at as a phosphodiesterase inhibitor (0). A 20°C with the CAMP receptor agonist P’-deoxy CAMP (10 KM) in the presence of 20 mM dithiothreitol cell homogenate was prepared by forcing a suspension of aggregation-competent cells (10s cells/ml) through a 5-pm pore-size Nuclepore filter. The homogenate was incubated in adenylate cyclase reaction mixture containing 0.5 mM Mg’+-ATP (0). The results shown are the means and standard deviations of triplicate determinations of an experiment reproduced twice.
cells the total amount of CAMP that is synthesized is one-third of that produced in vivo. This indicates that the rate of CAMP synthesis in permeable cells is less than that in intact cells. The adaptation-like reaction is completely lost in cell homogenates (Fig. 2b, open circles) and by cells electroporated under more harsh condition (20 instead of 7 discharges of 2 kV and 0.5 ms) (Fig. 2a, open squares), inducing large pores in the cell membrane that result in a loss of cytoplasmic proteins (data not shown). A second observation is that the initial rate of CAMP synthesis in cells, permeabilized under standard conditions, was considerably higher than that in homogenates and in cells more extensively permeabilized. It can be concluded that in standard permeabilized cells interactions between adenylate cyclase and inhibiting regulatory components are still intact. These interactions disappear if cells are damaged to an extent that intracellular macromolecules are lost. Regulation
of Adenylate
Cyclase
In D. discoideum cells adenylate cyclase activity is regulated by the cell surface receptor for CAMP and one or more G-proteins (see [6] for an overview). Therefore, we investigated the effect of the CAMP receptor agonist 2’-deoxy CAMP and of GTP$S, an activator of G-proteins, on CAMP synthesis in permeable cells. Figure 3 shows that in the presence of only ATP (i.e., no agonist or activator of G-proteins) the initial adenylate cyclase
activity in permeable cells was low (3 pmol/min/lOs cells). In contrast, under the same conditions in a cell homogenate the activity is more than fourfold higher (14 pmol/min/108 cell equivalents (Fig. 2b)). This result suggests that adenylate cyclase in permeabilized cells interacts with one or more inhibitory components. Stimulation with 10 PM 2’-deoxy CAMP resulted in a 7 (k1.9, N = 6)-fold increase in initial adenylate cyclase activity in permeabilized cells (Figs. 3a and Fig. 3b, bar 2). The receptor agonist was considerably more effective if added immediately before electroporation, compared to addition within 10 s after permeabilization (Fig. 3b, bar 5). The stimulatory effect of 2’-deoxy CAMP had disappeared completely if it was added 1 min after cell permeabilization (not shown). Evidently, a rapid uncoupling of receptor from the adenylate cyclase occurs after permeabilization. GTPyS (0.1 mA4) further potentiated the stimulatory effect of 2’-deoxy CAMP (Fig. 3b, bar 3). GTPrS added alone, either before, or up to 1 min after permeabilization had a lower stimulatory effect than if added in combination with receptor occupation (Fig. 3a and Fig. 3b bars 7 and 8). Remarkably, if GTP+ was present, 2’deoxy CAMP had a stimulatory effect also if added after electroporation (Fig. 3b, bar 6). We conclude from this that occupation of a GTP-binding protein stabilizes the functional receptor-adenylate cyclase interaction in electropermeabilized cells. If GDP (0.1 n-N) was used
166
SCHOEN
ET AL. 12 11
b
10 9 a i P
7
5 'P
6
._E' iz
5 4 3 2
J 100
150
200
Time
250
300
350
400
(seconds)
450
12345678 Incubations
FIG. 3.
Regulation of adenylate cyclase activity in permeabilized cells by 2’-deoxy CAMP and GTP+yS. D. discoideum cells were electropermeabilized under standard conditions by 7 discharges of 2 kV. Subsequently, the permeabilized cells were incubated in adenylate cyclase reaction medium in the presence of 20 mM Mg’+-ATP. (a) Cyclic AMP synthesis by cells permeabilized in the presence of the CAMP receptor agonist 2’-deoxy CAMP and/or the nonhydrolysable GTP-analog GTPyS, which is able to activate G-proteins. Cyclic AMP synthesis in permeabilized cells is shown as a function of time after the initiation of the adenylate cyclase reaction at t = 0, immediately after electropermeabilization. The electropermeabilization process was carried out in the presence of: (0) neither 2’-deoxy CAMP nor GTPyS (control); (0) 0.1 mM GTP-yS; (+) 10 pM 2’-deoxy CAMP; (m) 10 pM 2’-deoxy CAMP plus 0.1 mM GTPyS; (A) 10 pM 2’-deoxy CAMP plus 0.1 mM GDP. The results shown are the average of two independent experiments. (b) Timing of the effects of 2’-deoxy CAMP and GTP+. The stimulatory effect of 2’-deoxy CAMP (10 pM) and GTP-yS (0.1 mM) on CAMP synthesis in permeabilized cells is shown. 2’.Deoxy CAMP and/or GTPyS were added either immediately before permeabilization or immediately afterwards. The adenylate cyclase reaction was started within 10 s after permeabilization. Reactions were stopped after 1 min and CAMP concentrations were determined. Reaction rates are given relative to the rate in the absence of 2’-deoxy CAMP and GTPyS (bar 1). Below we indicate by “before” that the addition was made immediately before permeabilization, and by “after” that the addition was made immediately after permeabilization (within 10 s). Bar 1, reaction without addition; bar 2, 2’-deoxy CAMP before; bar 3,2’-deoxy CAMP before and GTP+ before; bar 4,2’-deoxy CAMP before and GDP (0.1 mM) after; bar 5, P’-deoxy CAMP after; bar 6, 2’-deoxy CAMP after and GTPrS before; bar 7, GTPrS before; bar 8, GTP-rS added after 1 min incubation at 0°C after permeabilization (in this case the adenylate cyclase reaction was also started 1 min after permeabilization. The results shown are the mean and standard deviations of experiments carried out in triplicate and reproduced twice.
instead of GTPrS the stimulatory effect of occupation of the CAMP receptor was almost completely suppressed (Fig. 3b, bar 4). This is compatible with the idea that GTPyS in this system exerts its effect on adenylate cyclase via a stimulatory G-protein. Importantly, neither 2’-deoxy CAMP, nor GTPrS had a significant effect on the rate of the adaptation-like effect on adenylate cyclase activity in permeabilized cells (Fig. 3a). This indicates that switching-off of the enzyme is an autonomous process, not affected itself by receptor occupation or activation of a G-protein. DISCUSSION
Extracellular CAMP binds to highly specific receptors on the surface of D. discoideum cells and activates, among others, the enzyme adenylate cyclase. After about 2 min the rate of CAMP synthesis in intact cells has returned to the basal level due to adaptation [ 171. A large body of evidence suggests that one or more G-proteins are involved in the coupling between receptor and adenylate cyclase [19-22, 28, 291. Biochemical studies
into the molecular mechanism of receptor-adenylate cyclase coupling are hampered by the fact that in such in vitro systems the linkage between receptor, G-protein, and enzyme are rapidly lost. In this study we investigated the properties of adenylate cyclase in permeabilized cells. Schoen et al. [30] have developed an electropermeabilization procedure for D. discoideum cells. Permeabilized cells retain cytoplasmic proteins and are accessible for small molecules like nucleotides and hexoses. Evidence has been presented that only the plasma membrane is permeabilized, whereas organelle membranes remain intact [30]. The advantage of such permeabilized cell system is obvious. The functional interactions between enzymes and regulatory, macromolecular components are not disturbed by dilution effects, and substrates and other low molecular weight substances can be introduced into or removed from the cell. Several conclusions can be drawn from the results presented here. First, CAMP receptor occupation (by the agonist 2’-deoxy CAMP) and G-protein activation (by GTPrS) stimulated the adenylate cyclase activity in
REGULATION
OF Dictyostelium
permeable cells severalfold (Fig. 3). In the presence of both activators the activity in permeable cells approached that of adenylate cyclase in a cell homogenate (Fig. 2). In the absence of 2’-deoxy CAMP and GTPrS only a low, basal level of adenylate cyclase activity was observed. Evidently, the adenylate cyclase activity is suppressed in non-stimulated intact and in permeabilized cells. The inhibitory effect disappeared upon cell breakage. Since GTP$S did stimulate adenylate cyclase fivefold in permeable cells (Fig. 3), it is likely that a nonactivated G-protein is involved in adenylate cyclase inhibition. The coupling between CAMP receptor and adenylate cyclase was rapidly lost after permeabilization. If receptors were occupied immediately before permeabilization a sevenfold stimulation of CAMP production was observed. In contrast, if we first permeabilized the cells and subsequently, within 10 seconds, added 2’-deoxy CAMP, only a slight stimulation of adenylate cyclase was observed (Fig. 3), an effect that was completely gone after 1 min. This situation is quite similar to that observed by Theibert and Devreotes [29] for cell homogenates. Importantly, if GTP$S was present at the moment of permeabilization, receptor occupation after permeabilization was still observed (Fig. 3). This shows that GTP$S prevents, or at least slows down the uncoupling of adenylate cyclase from the receptor. In permeable cells the CAMP response is about 3 times less than in uiuo (Fig. 2). This is due mainly to a lower rate of CAMP synthesis in permeabilized cells. Whether this is caused by a suboptimal activation of the enzyme or a reduced V,, in permeabilized cells remains to be investigated. Despite this quantitative difference the time course of CAMP synthesis in uiuo and in permeable cells is strikingly similar. This points to yet another important aspect of the adenylate cyclase reaction in permeabilized cells. In permeabilized cells adenylate cyclase became completely inactive after about 2 min (Figs. 2 and 3). In contrast, in cell homogenates the enzyme activity was constant for at least 10 min, showing that the enzyme is intrinsically stable under in vitro conditions. Furthermore, we were able to prove that the cells remained permeable for the substrate for at least 2 min (Fig. l), indicating that the reaction did not stop due to lack of ATP. The rate of inactivation was similar to that of the adaptation process in uiuo. These results strongly suggest that permeabilized cells still have the ability to show adaptation, a property lost completely in cell homogenates. Interestingly, we have found that also guanylate cyclase undergoes adaptation in permeabilized cells5 with a time course that is similar to that observed for guanylate cyclase in uiuo, i.e., within 15 s
ADENYLATE
167
[38]. Neither 2’-deoxy CAMP, nor GTPyS had any effect on the adaptation of adenylate cyclase in the permeable cells. This suggests that adaptation is an autonomous process that is triggered if adenylate cyclase becomes active. Again, we came to the same conclusion with respect to in vitro adaptation of the guanylate cyclase reaction in permeabilized cells’ [38]. Van Haastert et al. [28] have carried out similar experiments using a crude cell membrane fraction. Their results also show a stimulatory effect of receptor agonists and of GTPrS, although the effects on membranes are less pronounced than we found for permeable cells. This is most likely due to the fact that functional protein-protein interactions are better preserved in permeabilized cells than in highly diluted homogenates. Van Haastert et al. [28] have presented evidence that under conditions that favor protein phosphorylation an inhibitory G-protein can be activated, although again the effects are only small. Our experimental conditions (presence of Mg2+ -ATP) are likely to induce protein phosphorylation. However, we did not observe any enhancement by GTP$S of the adaptation-like effect in permeabilized cells (Fig. 3). This indicates that inactivation of adenylate cyclase in permeable cells is regulated differently from enzyme in crude membranes. Nevertheless, we cannot exclude that protein phosphorylation is instrumental in the adaptation-like process that we observed. Summarizing, our work shows that electropermeabilized D. discoideum cells constitute an excellent system to study the molecular mechanism of activation of adenylate cyclase and inactivation of the enzyme due to adaptation. Particularly, the latter process cannot be investigated at all in broken cell preparations. Evidently, in permeabilized cells the protein-protein interactions that are involved in adenylate cyclase inactivation remain functional, whereas they are lost after homogenization due to dilution. ACKNOWLEDGMENTS We thank Dr. M. E. E. Luderus for critical reading of the manuscript. These investigations were supported in part by the Foundation for Fundamental Biological Research (BION) which is subsidized by the Netherlands Organization for the Advancement of Scientific Research (N.W.O.).
REFERENCES 1. 2. 3. 4.
5 C. D. Schoen, J. C. Arents, C. C. G. M. Schulkes, and R. Van Driel, 1991, manuscript submitted.
CYCLASE
5.
Bonner, J. T., Barkley, D. S., Hall, E. M., Konijn, T. M., Masoin, I. W., O’Keefe, G., and Wolfe, B. B. (1969) Deu. Biol. 20, 72-87. Konijn, T. M., Van de Meene, J. G. C., Bonner, J. T., and Barkley, D. S., (1967) Proc. Natl. Acad. Sci. USA 58, 1152-1154. Gerisch, G. (1982) Annu. Reu. Physiol. 44,535-552. Van Haastert, P. J. M., and Konijn, T. M. (1982) Mol. Cell. Endocrinol. 26, 1-17. Devreotes, P. N. (1983) Adu. Cyclic Nucl. Res. 15,55-96.
168
SCHOEN
6. Van Haastert,
P. J. M., Janssens, P. M. W., and Erneux, C. (1991) Eur. J. Biochm. 195, 289-303. 7. Janssens, P. M. W., and Van Haastert, P. J. M. (1987) Microbial. Reu. 51,396-418. 8. Van Haastert, P. J. M., Van Lookeren Campagne, M. M., and Kesbeke, F. (1984) Biochim. Biophys. Acta 756,67-71. 9. Ross, F. M., and Newell, P. C. (1981) J. Gen. Microbial. 127,
339-350. 10. Van Haastert,
P. J. M., Ross, F. M., and Van Lookeren Campagne, M. M. (1982) FEBS Lett. 147,149-152. 11. Liu, G., and Newell, P. C. (1988) J. Cell Sci. 90, 123-129. 12. Malchow, D., Boehme, R., and Rahmsdorf, H. J. (1981) J. Bio-
chem.117,213-218. 13. Mato, J. M., Krens, F. A., Van Haastert, 14. 15. 16. 17. 18. 19.
20. 21.
P. J. M., and Konijn, T. M. (1977) Proc. N&l. Acad. Sci. USA 74, 2348-2351. Schaffer, B. M. (1975) Nature (London) 255, 549-552. Devreotes, P. N., and Steck, T. L. (1979) J. Cell. Biol. 80, 300309. Gerisch, G., Malchow, D., Roos, W., and Wick, U. (1979) J. Exp. Biol. 81, 33-47. Dinauer, M. C., Steck, T. L., and Devreotes, P. N. (1980) J. Cell. Biol. 86,545-553. Dinauer, M. C., Steck, T. L., and Devreotes, P. N. (1980) J. Cell. Biol. 86, 554-561. Janssens, P. M. W., Van der Geer, P. J. L., Arents, J. C., and Van Driel, R. (1985) Mol. Cell. Biochem. 67, 119-124. Janssens, P. M. W., Arents, J. C., Van Haastert, P. J. M., and Van Driel, R. (1986) Biochemistry 25, 1314-1320. Van Haastert, P. J. M. (1984) Biochem. Biophys. Res. Cornmum
124,597-604. 22. Van Haastert, P. J. M., De Wit, R. J. W., Janssens, P. M. W., Kesbeke, F., and De Goede, J. (1986) J. Biol. Chem. 261,69046911. Received September
23, 1991
ET AL.
23. Kumagai, Devreotes,
A., Pupillo, M., Gundersen, R., Miake-Lye, P. N., and Firtel, R. A. (1989) Cell 57, 265-275.
24. Kumagai,
A., (1991) J. Biol. 25. Gilman, A. G. 26. Gilman, A. G. 27. Lefkowitz, R. 4993-4996.
R.,
Hadwiger J. A., Pupillo, M., and Firtel, R. A. Chem. 266, 1220-1228. (1984) Cell 36,577-579. (1990) Biol. Med. Signal Transduction 24,51-57. J., and Caron, M. G. (1988) J. Biol. Chem. 263,
28. Van Haastert, 29.
P. J. M., Snaar-Jagalska, B. E., and Janssens, P. M. W. (1987) Eur. J. Biochem. 162, 251-258. Theibert, A., and Devreotes, P. N. (1986) J. Biol. Chem. 261, 15,121-15,126.
30. Schoen, C. D., Arents, J. C., Bruin, T., and Van Driel, R. (1989) Exp. Cell Res. 181, 51-62. 31. Sussman, R., and Sussman, M. (1967) Biochem. Biophys. Res. Commun. 29,53-55.
32. Van Haastert, P. J. M. (1984) J. Gen. Microbial. 130,2559-2564. 33. Das, 0. P., and Henderson, E. J. (1983) Biochim. Biophys. Actu 736,45-56. 34. Gerisch, G., Wick, U. (1975) Biochem. Biophys. Rex Commun. 65,364-370. 35. Otte, A. P., Plomp, M. J. E., Arents, J. C., Janssens, P. M. W., and Van Driel, R. (1986) Differentiation
32, 185-191.
36. Tovey, K. C., Oldham, K. G., and Whelan, J. A. M. (1974) Clin. Chim. Acta 56,221-234.
37. Williamson,
J. R., and Corky, B. E. (1969) in Methods in Enzymology (Lowenstein, J. M., Ed.), pp. 443-513, Academic Press, New York. 38. Schulkes C. C. G. M., Schoen, C. D., Arents, J. C., and Van Driel, R. (1991) in Biological Signal Transduction (Ross, E. M., and Wirtz, K. W., Eds), NATO ASI series H: Cell biology Vol. 52., pp. 497-509, Springer-Verlag, Berlin/Heidelberg.
CELL
RESEARCH
199,
162-168
(19%‘)
Regulation of Adenylate Cyclase in Electropermeabilized Dictyostelium ciiscoideum Cells CORD.SCHOEN,~TACOBRUIN,'JOSC.ARENTS,ANDROELVANDRIEL~ E. C. Slater Institute
for Biochemical
Research, University
of Amsterdam,
In Dictyostelium discoideum cells the enzyme adenylate cyclase is functionally coupled to cell surface receptors for CAMP. Coupling is known to involve one or more G-proteins. Receptor-mediated activation of adenylate cyclase is subject to adaptation. In this study we employ an electropermeabilized cell system to investigate regulation of D. discoideum adenylate cyclase. Conditions for selective permeabilization of the plasma membrane have been described by C. D. Schoen, J. C. Arents, T. Bruin, and R. Van Driel (1989, Exp. Cell Res. 181, 51-62). Only small pores are created in the membrane, allowing exchange of exclusively low molecular weight substances like nucleotides, and preventing the loss of macromolecules. Under these conditions functional protein-protein interactions are likely to remain intact. Adenylate cyclase in permeabilized cells was activated by the CAMP receptor agonist 2’-deoxy CAMP and by the nonhydrolyzable GTP-analogue GTP-yS, which activates G-proteins. The time course of the adenylate cyclase reaction in permeabilized cells was similar to that of intact cells. Maximal adenylate cyclase activity was observed if CAMP receptor agonist or GTP-analogue was added just before cell permeabilization. If these activators were added after permeabilization adenylate cyclase was stimulated in a suboptimal way. The sensitivity of adenylate cyclase activity for receptor occupation was found to decay more rapidly than that for G-protein activation. Importantly, the adenylate cyclase reaction in permeabilized cells was subject to an adaptation-like process that was characterized by a time course similar to adaptation in vivo. In vitro adaptation was not affected by CAMP receptor agonists or by G-protein activation. Evidently electropermeabilized cells constitute an excellent system for investigating the positive and negative regulation of D. o 1992 Academic PWS, I~C. discoideum adenylate cyclase.
INTRODUCTION Starvation of Dictyostelium gers a specific developmental
discoideum amoebae trigprogram. Within 5 to 7 h
’ Present address: Research Institute for Plant Protection, Binnenhaven 12,6709 PD Wageningen, The Netherlands. ’ Present address: Academic Medical Centre, Meibergdreef 15, 1105 AZ Amsterdam, The Netherlands. 3 To whom reprint requests should be addressed. 0014.4827/92 $3.00 Copyright 0 1992 by Academic Press, All rights of reproduction in any form
Plantage Muidergracht
The Netherlands
after the onset of starvation, amoebae start to aggregate into multicellular structures. This aggregation process is controlled by extracellular CAMP signals [ 1,2]. CAMP binds to highly specific cell-surface CAMP receptors which are functionally coupled to several enzymes, including adenylate cyclase and guanylate cyclase, reviewed in Refs. [3-71. Cyclic GMP that is produced by guanylate cyclase remains mostly intracellular and is probably involved in transduction of the chemotactic signal [8-131. Most of the CAMP produced by adenylate cyclase is secreted, thereby relaying the CAMP signal to neighboring cells [ 141. In intact D. discoideum cells regulation of CAMP receptor-coupled adenylate cyclase activity has been studied in detail [ 15-181. The enzyme is activated within 30 s after stimulation of cells with CAMP. The activity reaches a maximum after 1 to 2 min and then rapidly declines to prestimulation levels. The decrease in adenylate cyclase activity is due to adaptation. Subsequently, extracellular CAMP is hydrolized by phosphodiesterases. The cells undergo deadaptation with a halftime of 3-4 min and become responsive to CAMP again [17]. Despite detailed knowledge of the physiology of adenylate cyclase stimulation the molecular details of the sensory transduction pathway from the cell surface receptor to adenylate cyclase in D. discoideum and the molecular mechanism of the adaptation process have not been elucidated yet. Binding of CAMP to D. discoideum cells is a complex phenomenon. The chemotactic CAMP receptor occurs in several interconvertible forms. In isolated membranes, interconversions between different receptor forms have been found to be induced by guanine nucleotides [ 19-221. This is analogous to what has been found for many vertebrate receptor systems and suggests that the CAMP receptor is coupled to one or more regulatory guanine nucleotide binding proteins, G-proteins. The existence of G-proteins in D. discoideum has recently been confirmed [6, 23, 241. In vertebrate cells, G-proteins transduce the signal from hormone receptor to adenylate cyclase and other effector systems. The effect on adenylate cyclase can be either stimulatory (G,) or inhibitory (Gi) [25-271. Results of Van Haastert et al. [28] and Theibert and Devreotes [29] suggest that in D. discoi162
Inc. reserved.
12, 1018 TV Amsterdam,
REGULATION
OF Dietyostelium
deum Gi and/or G,-like proteins regulate adenylate cyclase activity. Thus far, CAMP-induced activation of adenylate cyclase in D. discoideum could only be observed when intact cells were stimulated and the adenylate cyclase activity was assayed immediately after cell lysis. It has up to now not been possible to stimulate the enzyme via the CAMP receptor in broken cell preparations. The rationale of our present study is to look for a method that preserves cellular structures and functional proteinprotein interactions and, at the same time, allows manipulation and analysis of signal transduction processes in general and of the adenylate cyclase system in particular. To this end we have adapted the electropermeabilization method developed by Schoen et al. [30]. In this report we describe the effect of receptor occupation and G-protein activation on adenylate cyclase activity in electropermeabilized D. disco&urn cells. We show that receptor occupation and G-protein activation can induce adenylate cyclase activation in permeable cells. Furthermore, an adaptation-like process is observed in permeable cells, which results in enzyme inactivation with approximately the same time course as seen in vivo. MATERIALS
AND METHODS
Materials. 2’.Deoxyadenosine 3’,5’-monophosphate (2’-deoxyCAMP), and adenosine 3’,5’-monophosphate (CAMP) were obtained from Boehringer (Mannheim, FRG); (2’,8’-3H)cAMP (0.9 TBq.mol-I), 2’-deoxy-D-(U-i4C)glucose (11.1 TBq.moll’) and 3H,0 (3.34 GBq.moll’) were purchased from Amersham (UK). Poly-carbonate filters with 5-pm pore size were from Nuclepore Corp. (Pleasanton, CA). Culture conditions. D. discoideum strain NC4(H) was cultured on standard medium (SM) agar [31] in combination with Escherichia coli B/r. Cells were harvested in the late logarithmic phase and washed free of nutrients in 15 mM potassium phosphate buffer (pH 6.2) (Pibuffer) three times by centrifugation at 150g for 4 min. Washed cells were resuspended in Pi-buffer, spread evenly on nonnutrient agar plates (1.5% agar in Pi-buffer) at a density of 1.5 to 4.0 X lo6 cells/ cm2, and starved at 6°C for 16-20 h to obtain aggregation-competent cells 1321. Aggregation-competent cells were harvested, washed, and resuspended at a density of 2 X 10’ cells/ml in 20 mM MES’-NaOH buffer (pH 6.0). Before use the cell suspension was aerated continuously. Electropermeabilization of cells and preparation of a cell homogenate. Cells in 20 mM MES-NaOH (pH 6.0) were resuspended at a density of 2 X 10’ cells/ml in standard electropermeabilization buffer containing: 20 mM Hepes-NaOH (pH 6.6), 100 mM sucrose, 20 mM K-acetate, and 1 mM Mg-acetate. One milliliter of the cell suspension was pipetted into the electropermeabilization chamber between a pair of parallel platinum electrodes (each with a surface area of 1 cm*) positioned 1 cm apart. Permeabilization was
4 Abbreviations used: Hepes, N-(2-hydroxyethyl)piperazine-W2ethanesulfonic acid, MES, 2-(N-morpholino)ethanesulfonic acid; EGTA, ethylene glycol bis @-amino ethyl ether)-N,N,N’,N’-tetraacetic acid; EDTA, ethylenediaminotetraacetic acid; GTPyS, guanosine 5’-O-[3-thio] triphosphate; AppNHp, adenosine 5’-[&y,-imido] triphosphate.
ADENYLATE
163
CYCLASE
achieved by exposing the cells to seven brief, square wave pulses of 500 ps (rise and fall times approximately 50 ns) of 2 kV, at 0°C under continuous stirring [30]. During these seven discharges, applied with 3-s intervals, the temperature increased only 1 to 2’C. The extent of permeabilization was measured by ethidium bromide fluorescence. Nonpermeable cells exclude ethidium bromide and were not fluorescent, whereas permeable cells accumulate the dye [30]. Under standard conditions more than 98% of the cells were permeable. Cell homogenates were prepared by forcing cells through a 5-pm pore-size polycarbonate filter 128, 331. More than 98% of the cells were broken by this procedure as judged by phase contrast microsCOPY. Measurement of cell permeability. After electropermeabilization the permeability of cells was determined quantitatively by measuring the rate of influx of a radioactive, non-permeant, non-metabolisable substance, i.e., [‘?]deoxyglucose. The total intracellular volume was determined by measuring the uptake of the permeant molecule 3H,0. The permeant and the nonpermeant substances were mixed together and added to the cell suspension. At different time points the intraand extracellular compartments were separated by centrifugation of the cells through a layer of silicone oil [30,34]. From the radioactivity in the pellet, the intracellular volume accessible by [r4C]glucose was calculated. Corrections were made for radioactive label adhering to the outside of the cells during centrifugation through silicone oil. Measurement of adenylate cyclase activity. Adenylate cyclase activity was measured by a method described earlier [35]. The incubation mixture (500 ~1) contained 40 mM Hepes-NaOH (pH 7.7), 10.5 mM K-acetate, 3 mM EDTA, 100 mM sucrose, 20 mM dithiothreitol, 2 mM 3-isobutyl-l-methylxanthine, 2.5 mM phosphoenol-pyruvate, 1 U/ml pyruvate-kinase, and unless stated otherwise 0.5 mM ATP and 6 mM MgCl,. The assay was initiated by the addition of 250 ~1 of cell homogenate or permeabilized cells (2 X 10’ cells or cell equivalents per ml) and was carried out at 20°C under vigorous agitation. At various time points samples were quenched in perchloric acid (final concentration 1 IV) to measure CAMP levels. Quenched samples were centrifuged for 5 min at 10,OOOgand the perchloric acid extracts were neutralized by the addition of a solution containing KHCO, and KOH. Neutralized extracts were centrifuged for 2 min at 10,OOOgand CAMP concentrations were measured by an isotope dilution assay using bovine CAMP-dependent protein kinase as a CAMP binding protein [36]. Zntracelluhr ATP concentration. Freshly permeabilized D. discoideum cells (10’ cells/ml) were incubated for 1 min at 20°C with different concentrations of ATP (0,0.5,10, or 20 mM ATP). Subsequently, the intra- and extracellular compartments were separated by centrifugation (30 s at 10,OOOg) of the cells through a layer of silicone oil [30, 341 into 0.4 M HClO,. The HClO, extracts were directly neutralized and centrifuged (2 min at 10,OOOg).The ATP level in the supernatant, reflecting the amount of ATP originally accumulated by the cells was assayed by the method of Williamson and Corky [37]. The assay mixture (final volume of 2.2 ml) contained 50 mM Triethanolamine-HCl (pH 7.4), 4 mM EDTA, 12.5 mM MgCl,, 50 mM KCl, 1 mM glucose, 1 mM NADP, and 0.35 mU glucose 6-phosphate dehydrogenase. The assay was initiated by the addition of 100 ~1 of the neutralizedperchloric acid extract and was carried out at 20°C. After 10 min, 20 ~1 hexokinase (0.5 U) was added and fluorescence of NADPH (excitation 316366 nm and emission 300-400 nm) was measured after the reaction was complete. The intracellular ATP level was calculated from these data using standard solutions of known ATP concentration. Corrections were made for ATP adhering to the outside of the cells during centrifugation through silicone oil.
RESULTS Electropermeabilization of Dictyosielium Cells Conditions for electropermeabilization of D. discoideum cells have been described in detail by Schoen et al.
164
SCHOEN
ET AL
Adenylate
0.00 0
1
2
3 Time
4
5
6
(min)
FIG. 1. Permeability of electropermeabilized cells. Aggregationcompetent cells were electropermeabilized under standard conditions by 7 discharges of 2 kV and 0.5 ms (0, +). Control cells (0) were treated in the same way but did not receive high voltage pulses. Permeabilized cells were either incubated directly (0) or after 2 min (+) with [%]deoxyglucose (2 mM, 2 X lo4 cpm/ml) and aH,O (2 X lo5 cpm/ml) at 20°C in adenylate cyclase reaction mixture. Control cells were incubated under the same conditions directly after a mock-electropermeabilization. The intracellular [3H] and [‘%]radioactivities were determined after separating the cells from the extracellular medium by centrifugation through a layer of silicone oil [30,34]. Corrections were made for adhering extracellular material during centrifugation. The ratio of intracellular, tritiated water (having access to all cellular compartments) and intracellular [Wldeoxyglucose (entering the cell exclusively through plasma-membrane pores created by electropermeabilization) has been plotted as a function of the time after permeabilization. The arrow marks the “C/3H ratio in the extracellular medium, representing the theoretical maximum that would be reached if [WJdeoxyglucose would have access to the same cellular compartments as ‘H,O.
Cyclase Activity
in Electropermeabilized
Cells
During initial experiments we noticed that near-physiological Mg’+-ATP concentrations (around the 1 m&f) in the medium resulted in little or no adenylate cyclase activity in permeabiiized cells. Much higher concentrations were required. We have, therefore, measured the ATP concentration inside the permeabilized cells as a function of the ATP level in the medium. Table 1 shows that an extracellular concentration of 20 mMis required to obtain a physiological ATP concentration inside the permeable cell (about 0.8 mM). The drop in concentration across the permeable cell membrane is probably due to highly active ATPases. Apparently, the rate of influx of ATP is low compared to the rate of ATP hydrolysis inside the permeable cell, creating locally a lower than ambient concentration. Figure 2a (open circles) shows the CAMP synthesis of permeable cells immediately after electropermeabilization. Adenylate cyclase activity was completely dependent on a high extracellular ATP concentration (20 mM). No adenylate cyclase activity was observed at 0.5 mM ATP (solid circles in Fig. 2a). This indicates that CAMP is produced by permeable cells, rather than by some residual intact cells. Two important observations can be made on cells permeabilized under standard condition. First, after about 2 min the adenylate cyclase reaction terminates. This time course is similar to that of the CAMP response of intact cells (Fig. 2b, solid circles). In vivo termination of the adenylate cyclase reaction is due to an adaptation process [17, 181. This process appeared to be still operative in permeable cells. The rate of adaptation in vivo and in permeabilized cells is about the same (t,., of about 1 min). In permeabilized
TABLE Intracellular Function
1
ATP Concentration in Permeable of the ATP Concentration in the
Cells
as a
Medium
ATP
[30]. For this study we have selected electropermeabilization conditions for which cells become accessible for low molecular weight substances, such as nucleotides, but retain their intracellular macromolecules. Furthermore, their organelles remain intact [30]. Figure 1 shows that after permeabilization about 75% of the total cellular volume, i.e., the volume that is accessible for the permeant molecule 3H,0 becomes accessible for the nonpermeant substance [14C]deoxyglucose. The remaining 25% probably reflects the intact organelles and vacuoles in the cell. After 2 min the plasmamembrane pores are still open, although the rate of influx of [‘“Cldeoxyglucose has decreased slightly.
ATP concentration in the medium (mm Intact cells Permeable Permeable Permeable Permeable
cells cells cells cells
0 0 0.5 10 20
Intracellular concentration
ATP (mM)
0.75 0.10 0.09 0.21 0.80
Note. Directly after electropermeabilization D. disco&urn cells were incubated for 1 min in adenylate cyclase reaction medium containing different MgZ+-ATP concentrations (0, 0.5, 10, or 20 mM) at 20°C. The intracellular ATP level was determined after separation of the intra- and extracellular compartments by centrifugation of the cells through a layer of silicone oil [30, 341 into 0.4 h4 HClO,. The ATP concentrations were measured as indicated under Materials and Methods. The results were reproduced once.
REGULATION
OF Dictyostelium
ADENYLATE
165
CYCLASE
8C I-
a
120
b
c
a
I
0 0
50
100
150
200
Time
250
300
350
400
450
500
(seconds)
0
50
100
150
200
Time
250
300
350
400
450
500
(seconds)
FIG. 2. Cyclic AMP synthesis in permeabilized cells, intact cells, and a cell homogenate. (a) Permeabilized cells. Cyclic AMP synthesis in permeabilized cells is shown as a function of time after the initiation of the adenylate cyclase reaction at t = 0, immediately after electropermeabilization. Aggregation-competent cells were electropermeabilized by 7 pulses of 2 kV and 0.5 ms in the presence of 10 gM 2’-deoxy CAMP. After permeabilization the adenylate cyclase reaction was started immediately in the presence of 0.5 mM Mg’+-ATP (0) or 20 mM Mg’+-ATP (0). The adenylate cyclase activity in cells permeabilized under more harsh conditions (20 pulses of 2 kV and 0.5 ms) in the presence of 2’-deoxy CAMP (10 pM) was measured in the presence of 20 mM Mg’+-ATP (Cl). These latter cells were more permeable than cells treated under standard conditions, since they had lost most soluble cytoplasmatic proteins (data not shown). The results shown are the average of two independent experiments. (b) Intact cells and cell homogenate. Aggregation-competent D. discoideum cells (10s cells/ml) were stimulated at as a phosphodiesterase inhibitor (0). A 20°C with the CAMP receptor agonist P’-deoxy CAMP (10 KM) in the presence of 20 mM dithiothreitol cell homogenate was prepared by forcing a suspension of aggregation-competent cells (10s cells/ml) through a 5-pm pore-size Nuclepore filter. The homogenate was incubated in adenylate cyclase reaction mixture containing 0.5 mM Mg’+-ATP (0). The results shown are the means and standard deviations of triplicate determinations of an experiment reproduced twice.
cells the total amount of CAMP that is synthesized is one-third of that produced in vivo. This indicates that the rate of CAMP synthesis in permeable cells is less than that in intact cells. The adaptation-like reaction is completely lost in cell homogenates (Fig. 2b, open circles) and by cells electroporated under more harsh condition (20 instead of 7 discharges of 2 kV and 0.5 ms) (Fig. 2a, open squares), inducing large pores in the cell membrane that result in a loss of cytoplasmic proteins (data not shown). A second observation is that the initial rate of CAMP synthesis in cells, permeabilized under standard conditions, was considerably higher than that in homogenates and in cells more extensively permeabilized. It can be concluded that in standard permeabilized cells interactions between adenylate cyclase and inhibiting regulatory components are still intact. These interactions disappear if cells are damaged to an extent that intracellular macromolecules are lost. Regulation
of Adenylate
Cyclase
In D. discoideum cells adenylate cyclase activity is regulated by the cell surface receptor for CAMP and one or more G-proteins (see [6] for an overview). Therefore, we investigated the effect of the CAMP receptor agonist 2’-deoxy CAMP and of GTP$S, an activator of G-proteins, on CAMP synthesis in permeable cells. Figure 3 shows that in the presence of only ATP (i.e., no agonist or activator of G-proteins) the initial adenylate cyclase
activity in permeable cells was low (3 pmol/min/lOs cells). In contrast, under the same conditions in a cell homogenate the activity is more than fourfold higher (14 pmol/min/108 cell equivalents (Fig. 2b)). This result suggests that adenylate cyclase in permeabilized cells interacts with one or more inhibitory components. Stimulation with 10 PM 2’-deoxy CAMP resulted in a 7 (k1.9, N = 6)-fold increase in initial adenylate cyclase activity in permeabilized cells (Figs. 3a and Fig. 3b, bar 2). The receptor agonist was considerably more effective if added immediately before electroporation, compared to addition within 10 s after permeabilization (Fig. 3b, bar 5). The stimulatory effect of 2’-deoxy CAMP had disappeared completely if it was added 1 min after cell permeabilization (not shown). Evidently, a rapid uncoupling of receptor from the adenylate cyclase occurs after permeabilization. GTPyS (0.1 mA4) further potentiated the stimulatory effect of 2’-deoxy CAMP (Fig. 3b, bar 3). GTPrS added alone, either before, or up to 1 min after permeabilization had a lower stimulatory effect than if added in combination with receptor occupation (Fig. 3a and Fig. 3b bars 7 and 8). Remarkably, if GTP+ was present, 2’deoxy CAMP had a stimulatory effect also if added after electroporation (Fig. 3b, bar 6). We conclude from this that occupation of a GTP-binding protein stabilizes the functional receptor-adenylate cyclase interaction in electropermeabilized cells. If GDP (0.1 n-N) was used
166
SCHOEN
ET AL. 12 11
b
10 9 a i P
7
5 'P
6
._E' iz
5 4 3 2
J 100
150
200
Time
250
300
350
400
(seconds)
450
12345678 Incubations
FIG. 3.
Regulation of adenylate cyclase activity in permeabilized cells by 2’-deoxy CAMP and GTP+yS. D. discoideum cells were electropermeabilized under standard conditions by 7 discharges of 2 kV. Subsequently, the permeabilized cells were incubated in adenylate cyclase reaction medium in the presence of 20 mM Mg’+-ATP. (a) Cyclic AMP synthesis by cells permeabilized in the presence of the CAMP receptor agonist 2’-deoxy CAMP and/or the nonhydrolysable GTP-analog GTPyS, which is able to activate G-proteins. Cyclic AMP synthesis in permeabilized cells is shown as a function of time after the initiation of the adenylate cyclase reaction at t = 0, immediately after electropermeabilization. The electropermeabilization process was carried out in the presence of: (0) neither 2’-deoxy CAMP nor GTPyS (control); (0) 0.1 mM GTP-yS; (+) 10 pM 2’-deoxy CAMP; (m) 10 pM 2’-deoxy CAMP plus 0.1 mM GTPyS; (A) 10 pM 2’-deoxy CAMP plus 0.1 mM GDP. The results shown are the average of two independent experiments. (b) Timing of the effects of 2’-deoxy CAMP and GTP+. The stimulatory effect of 2’-deoxy CAMP (10 pM) and GTP-yS (0.1 mM) on CAMP synthesis in permeabilized cells is shown. 2’.Deoxy CAMP and/or GTPyS were added either immediately before permeabilization or immediately afterwards. The adenylate cyclase reaction was started within 10 s after permeabilization. Reactions were stopped after 1 min and CAMP concentrations were determined. Reaction rates are given relative to the rate in the absence of 2’-deoxy CAMP and GTPyS (bar 1). Below we indicate by “before” that the addition was made immediately before permeabilization, and by “after” that the addition was made immediately after permeabilization (within 10 s). Bar 1, reaction without addition; bar 2, 2’-deoxy CAMP before; bar 3,2’-deoxy CAMP before and GTP+ before; bar 4,2’-deoxy CAMP before and GDP (0.1 mM) after; bar 5, P’-deoxy CAMP after; bar 6, 2’-deoxy CAMP after and GTPrS before; bar 7, GTPrS before; bar 8, GTP-rS added after 1 min incubation at 0°C after permeabilization (in this case the adenylate cyclase reaction was also started 1 min after permeabilization. The results shown are the mean and standard deviations of experiments carried out in triplicate and reproduced twice.
instead of GTPrS the stimulatory effect of occupation of the CAMP receptor was almost completely suppressed (Fig. 3b, bar 4). This is compatible with the idea that GTPyS in this system exerts its effect on adenylate cyclase via a stimulatory G-protein. Importantly, neither 2’-deoxy CAMP, nor GTPrS had a significant effect on the rate of the adaptation-like effect on adenylate cyclase activity in permeabilized cells (Fig. 3a). This indicates that switching-off of the enzyme is an autonomous process, not affected itself by receptor occupation or activation of a G-protein. DISCUSSION
Extracellular CAMP binds to highly specific receptors on the surface of D. discoideum cells and activates, among others, the enzyme adenylate cyclase. After about 2 min the rate of CAMP synthesis in intact cells has returned to the basal level due to adaptation [ 171. A large body of evidence suggests that one or more G-proteins are involved in the coupling between receptor and adenylate cyclase [19-22, 28, 291. Biochemical studies
into the molecular mechanism of receptor-adenylate cyclase coupling are hampered by the fact that in such in vitro systems the linkage between receptor, G-protein, and enzyme are rapidly lost. In this study we investigated the properties of adenylate cyclase in permeabilized cells. Schoen et al. [30] have developed an electropermeabilization procedure for D. discoideum cells. Permeabilized cells retain cytoplasmic proteins and are accessible for small molecules like nucleotides and hexoses. Evidence has been presented that only the plasma membrane is permeabilized, whereas organelle membranes remain intact [30]. The advantage of such permeabilized cell system is obvious. The functional interactions between enzymes and regulatory, macromolecular components are not disturbed by dilution effects, and substrates and other low molecular weight substances can be introduced into or removed from the cell. Several conclusions can be drawn from the results presented here. First, CAMP receptor occupation (by the agonist 2’-deoxy CAMP) and G-protein activation (by GTPrS) stimulated the adenylate cyclase activity in
REGULATION
OF Dictyostelium
permeable cells severalfold (Fig. 3). In the presence of both activators the activity in permeable cells approached that of adenylate cyclase in a cell homogenate (Fig. 2). In the absence of 2’-deoxy CAMP and GTPrS only a low, basal level of adenylate cyclase activity was observed. Evidently, the adenylate cyclase activity is suppressed in non-stimulated intact and in permeabilized cells. The inhibitory effect disappeared upon cell breakage. Since GTP$S did stimulate adenylate cyclase fivefold in permeable cells (Fig. 3), it is likely that a nonactivated G-protein is involved in adenylate cyclase inhibition. The coupling between CAMP receptor and adenylate cyclase was rapidly lost after permeabilization. If receptors were occupied immediately before permeabilization a sevenfold stimulation of CAMP production was observed. In contrast, if we first permeabilized the cells and subsequently, within 10 seconds, added 2’-deoxy CAMP, only a slight stimulation of adenylate cyclase was observed (Fig. 3), an effect that was completely gone after 1 min. This situation is quite similar to that observed by Theibert and Devreotes [29] for cell homogenates. Importantly, if GTP$S was present at the moment of permeabilization, receptor occupation after permeabilization was still observed (Fig. 3). This shows that GTP$S prevents, or at least slows down the uncoupling of adenylate cyclase from the receptor. In permeable cells the CAMP response is about 3 times less than in uiuo (Fig. 2). This is due mainly to a lower rate of CAMP synthesis in permeabilized cells. Whether this is caused by a suboptimal activation of the enzyme or a reduced V,, in permeabilized cells remains to be investigated. Despite this quantitative difference the time course of CAMP synthesis in uiuo and in permeable cells is strikingly similar. This points to yet another important aspect of the adenylate cyclase reaction in permeabilized cells. In permeabilized cells adenylate cyclase became completely inactive after about 2 min (Figs. 2 and 3). In contrast, in cell homogenates the enzyme activity was constant for at least 10 min, showing that the enzyme is intrinsically stable under in vitro conditions. Furthermore, we were able to prove that the cells remained permeable for the substrate for at least 2 min (Fig. l), indicating that the reaction did not stop due to lack of ATP. The rate of inactivation was similar to that of the adaptation process in uiuo. These results strongly suggest that permeabilized cells still have the ability to show adaptation, a property lost completely in cell homogenates. Interestingly, we have found that also guanylate cyclase undergoes adaptation in permeabilized cells5 with a time course that is similar to that observed for guanylate cyclase in uiuo, i.e., within 15 s
ADENYLATE
167
[38]. Neither 2’-deoxy CAMP, nor GTPyS had any effect on the adaptation of adenylate cyclase in the permeable cells. This suggests that adaptation is an autonomous process that is triggered if adenylate cyclase becomes active. Again, we came to the same conclusion with respect to in vitro adaptation of the guanylate cyclase reaction in permeabilized cells’ [38]. Van Haastert et al. [28] have carried out similar experiments using a crude cell membrane fraction. Their results also show a stimulatory effect of receptor agonists and of GTPrS, although the effects on membranes are less pronounced than we found for permeable cells. This is most likely due to the fact that functional protein-protein interactions are better preserved in permeabilized cells than in highly diluted homogenates. Van Haastert et al. [28] have presented evidence that under conditions that favor protein phosphorylation an inhibitory G-protein can be activated, although again the effects are only small. Our experimental conditions (presence of Mg2+ -ATP) are likely to induce protein phosphorylation. However, we did not observe any enhancement by GTP$S of the adaptation-like effect in permeabilized cells (Fig. 3). This indicates that inactivation of adenylate cyclase in permeable cells is regulated differently from enzyme in crude membranes. Nevertheless, we cannot exclude that protein phosphorylation is instrumental in the adaptation-like process that we observed. Summarizing, our work shows that electropermeabilized D. discoideum cells constitute an excellent system to study the molecular mechanism of activation of adenylate cyclase and inactivation of the enzyme due to adaptation. Particularly, the latter process cannot be investigated at all in broken cell preparations. Evidently, in permeabilized cells the protein-protein interactions that are involved in adenylate cyclase inactivation remain functional, whereas they are lost after homogenization due to dilution. ACKNOWLEDGMENTS We thank Dr. M. E. E. Luderus for critical reading of the manuscript. These investigations were supported in part by the Foundation for Fundamental Biological Research (BION) which is subsidized by the Netherlands Organization for the Advancement of Scientific Research (N.W.O.).
REFERENCES 1. 2. 3. 4.
5 C. D. Schoen, J. C. Arents, C. C. G. M. Schulkes, and R. Van Driel, 1991, manuscript submitted.
CYCLASE
5.
Bonner, J. T., Barkley, D. S., Hall, E. M., Konijn, T. M., Masoin, I. W., O’Keefe, G., and Wolfe, B. B. (1969) Deu. Biol. 20, 72-87. Konijn, T. M., Van de Meene, J. G. C., Bonner, J. T., and Barkley, D. S., (1967) Proc. Natl. Acad. Sci. USA 58, 1152-1154. Gerisch, G. (1982) Annu. Reu. Physiol. 44,535-552. Van Haastert, P. J. M., and Konijn, T. M. (1982) Mol. Cell. Endocrinol. 26, 1-17. Devreotes, P. N. (1983) Adu. Cyclic Nucl. Res. 15,55-96.
168
SCHOEN
6. Van Haastert,
P. J. M., Janssens, P. M. W., and Erneux, C. (1991) Eur. J. Biochm. 195, 289-303. 7. Janssens, P. M. W., and Van Haastert, P. J. M. (1987) Microbial. Reu. 51,396-418. 8. Van Haastert, P. J. M., Van Lookeren Campagne, M. M., and Kesbeke, F. (1984) Biochim. Biophys. Acta 756,67-71. 9. Ross, F. M., and Newell, P. C. (1981) J. Gen. Microbial. 127,
339-350. 10. Van Haastert,
P. J. M., Ross, F. M., and Van Lookeren Campagne, M. M. (1982) FEBS Lett. 147,149-152. 11. Liu, G., and Newell, P. C. (1988) J. Cell Sci. 90, 123-129. 12. Malchow, D., Boehme, R., and Rahmsdorf, H. J. (1981) J. Bio-
chem.117,213-218. 13. Mato, J. M., Krens, F. A., Van Haastert, 14. 15. 16. 17. 18. 19.
20. 21.
P. J. M., and Konijn, T. M. (1977) Proc. N&l. Acad. Sci. USA 74, 2348-2351. Schaffer, B. M. (1975) Nature (London) 255, 549-552. Devreotes, P. N., and Steck, T. L. (1979) J. Cell. Biol. 80, 300309. Gerisch, G., Malchow, D., Roos, W., and Wick, U. (1979) J. Exp. Biol. 81, 33-47. Dinauer, M. C., Steck, T. L., and Devreotes, P. N. (1980) J. Cell. Biol. 86,545-553. Dinauer, M. C., Steck, T. L., and Devreotes, P. N. (1980) J. Cell. Biol. 86, 554-561. Janssens, P. M. W., Van der Geer, P. J. L., Arents, J. C., and Van Driel, R. (1985) Mol. Cell. Biochem. 67, 119-124. Janssens, P. M. W., Arents, J. C., Van Haastert, P. J. M., and Van Driel, R. (1986) Biochemistry 25, 1314-1320. Van Haastert, P. J. M. (1984) Biochem. Biophys. Res. Cornmum
124,597-604. 22. Van Haastert, P. J. M., De Wit, R. J. W., Janssens, P. M. W., Kesbeke, F., and De Goede, J. (1986) J. Biol. Chem. 261,69046911. Received September
23, 1991
ET AL.
23. Kumagai, Devreotes,
A., Pupillo, M., Gundersen, R., Miake-Lye, P. N., and Firtel, R. A. (1989) Cell 57, 265-275.
24. Kumagai,
A., (1991) J. Biol. 25. Gilman, A. G. 26. Gilman, A. G. 27. Lefkowitz, R. 4993-4996.
R.,
Hadwiger J. A., Pupillo, M., and Firtel, R. A. Chem. 266, 1220-1228. (1984) Cell 36,577-579. (1990) Biol. Med. Signal Transduction 24,51-57. J., and Caron, M. G. (1988) J. Biol. Chem. 263,
28. Van Haastert, 29.
P. J. M., Snaar-Jagalska, B. E., and Janssens, P. M. W. (1987) Eur. J. Biochem. 162, 251-258. Theibert, A., and Devreotes, P. N. (1986) J. Biol. Chem. 261, 15,121-15,126.
30. Schoen, C. D., Arents, J. C., Bruin, T., and Van Driel, R. (1989) Exp. Cell Res. 181, 51-62. 31. Sussman, R., and Sussman, M. (1967) Biochem. Biophys. Res. Commun. 29,53-55.
32. Van Haastert, P. J. M. (1984) J. Gen. Microbial. 130,2559-2564. 33. Das, 0. P., and Henderson, E. J. (1983) Biochim. Biophys. Actu 736,45-56. 34. Gerisch, G., Wick, U. (1975) Biochem. Biophys. Rex Commun. 65,364-370. 35. Otte, A. P., Plomp, M. J. E., Arents, J. C., Janssens, P. M. W., and Van Driel, R. (1986) Differentiation
32, 185-191.
36. Tovey, K. C., Oldham, K. G., and Whelan, J. A. M. (1974) Clin. Chim. Acta 56,221-234.
37. Williamson,
J. R., and Corky, B. E. (1969) in Methods in Enzymology (Lowenstein, J. M., Ed.), pp. 443-513, Academic Press, New York. 38. Schulkes C. C. G. M., Schoen, C. D., Arents, J. C., and Van Driel, R. (1991) in Biological Signal Transduction (Ross, E. M., and Wirtz, K. W., Eds), NATO ASI series H: Cell biology Vol. 52., pp. 497-509, Springer-Verlag, Berlin/Heidelberg.
