ACTION OF A SLOWLY HYDROLYSABLE CYCLIC AMP ANALOGUE ON DEVELOPING

Transcription

1 J. Cell Set. 35, (1978) 321 Printed in Great Britain Company of Biologists Limited ACTION OF A SLOWLY HYDROLYSABLE CYCLIC AMP ANALOGUE ON DEVELOPING CELLS OF DICTYOSTELIUM C. ROSSIER, G. GERISCH, D. MALCHOW Biozentrum der Universitat Basel, 4056 Basel, Switzerland DISCOIDEUM AND F. ECKSTEIN Max-Planck-Institut fur Experimentelle Medizin, 34 GGttingen, Germany SUMMARY Adenosine 3',s'-cyclic phosphorothioate (camp-s) is a cyclic AMP (camp) analogue which is only slowly hydrolysed by phosphodiesterases of Dictyostelium discoideum. The affinity of camp-s to camp receptors at the cell surface is only one order of magnitude lower than that of camp. camp-s can replace camp as a stimulant with respect to all receptor-mediated responses tested, including chemotaxis and the induction of camp pulses. camp-s does not affect growth of D. discoideum but it blocks cell aggregation at a uniform concentration of 5 x 1 o~ 7 M in agar plate cultures of strain NC-4 as well as its axenically growing derivative, Ax-2. Another wild-type strain, of D. discoideum, v-12, is able to aggregate on agar plates supplemented with 1 mm camp-s. The development of Polysphondylium pallidum and P. violaceum is also highly camp-s resistant. In Ax-2 both differentiation from the growth phase to the aggregation-competent stage and chemotaxis are camp-s sensitive, whereas in v-12 only chemotaxis is inhibited, v-12 can still form streams of cohering cells and fruiting bodies when chemotaxis is inhibited by camp-s. Whereas camp induces differentiation into stalk cells at concentrations of io" 3 or io~* M, camp-s has the same effect in strain v-12 at the much lower concentration of io~' M. INTRODUCTION Cyclic AMP acts as a chemotactic factor during cell aggregation of Dictyostelium discoideum (Konijn, van de Meene, Bonner & Barkley, 1967). In strain Ax-2, an axenically growing descendant of the wild-type strain NC-4, periodic pulses of camp stimulate cell differentiation from the growth phase to the aggregationcompetent stage (Gerisch, Fromm, Huesgen & Wick, 1975 a; Darmon, Brachet & Pereira da Silva, 1975). A third effect of camp is the induction of stalk cells. Normally stalk cells are formed following aggregation as one of the cell types constituting the fruiting body. Induction of stalk cell differentiation from non-aggregated cells requires high concentrations of camp (Bonner, 1970) and is more effective in the wild-type clone V-12/M2 than in NC-4 (Town, Gross & Kay, 1976). Two constituents of the cell surface interact with extracellular camp: receptors and phosphodiesterase. The latter regulates the lifetime of camp in the extracellular space. In order to activate receptors without degradation of the ligand, a slowly hydrolysable analogue of camp, adenosine 3',5'-cyclic phosphorothioate (camp-s) can be used (Malchow, Robinson & Eckstein, 1976). In this paper we show that camp

2 322 C. Rossier, G. Gerisch, D. Malchow and F. Eckstein can be replaced by the analogue not only as a chemotactic factor (Konijn, 1972) but also as an inducer of camp pulses and of stalk cell differentiation. A uniform concentration of camp-s in the medium makes the cells insensitive to camp pulses. Thus the analogue can be used to investigate the importance of camp signals for the development of different strains and species. camp pulses are periodically generated by the cells (Gerisch & Wick, 1975; Roos, Scheidegger & Gerisch, 1977). Here we show for D. discoideum that the wild-type strain NC-4 and its axenically growing descendant, Ax-2, differ from wild-type v-12 with respect to camp-s sensitivity of cell development. This indicates differences between these strains in the control systems of development. camp-s can also be used to investigate the role of chemotaxis in cell aggregation. At high, uniform camp-s concentrations chemotaxis is suppressed, so that aggregation can be studied in its absence. MATERIALS AND METHODS Strains. The D. discoideum strains used were NC-4, Ax-2, and clones Mi and M2 of strain v-12. The Polysphondylium strains were P. pallidum WS 320 and P. violaceum 83. The latter was an own isolate from forest soil in Hessen, Germany. The other strains are from Dr K. B. Raper, Madison. Mutant Wag-6 of Ax-2 was a gift of Dr John Ashworth, Colchester, the camp phosphodiesterase-negative Escherichia coli K 12 SB 2252 (Met", phosphodiesterase") of Dr Saul Roseman, Baltimore. Chemicals. camp-s (triethylammonium salt) was synthesized as described by Eckstein, Simonson & Bar (1974). Inhibitor of extracellular phosphodiesterase was obtained from culturefluidof Dictyostelium purpureum and partially purified by DEAE cellulose chromatography. Beef heart phosphodiesterase, camp, cgmp, AMP-S and 8-bromo-cAMP were purchased from Boehringer Mannheim. 8-azido-cAMP was obtained from P-L Biochemicals, Calcofluor White ST from the American Cyanamid Company. Cell cultures. Cells were cultivated at 23 C either on agar plates with E. coli SB 2252 or in suspension. For agar plates 2 ml nutrient agar containing o-i % bacteriological peptone (Oxoid), o-i % glucose, and 2 % Bacto-agar (Difco) in 17 mm phosphate buffer ph 6-o were used in dishes of 10 cm 1 surface area. Non-nutrient agar contained only the buffer with 2 % agar. The effect of beef heart phosphodiesterase on development was tested on 2 ml nutrient agar containing 1 mm MgCU in 17 mm Soerensen phosphate buffer, ph 70, and phosphodiesterase with an activity of /tmol camp per min at 25 C. For suspension cultures strain Ax-2 clone 214 was grown in the medium of Watts & Ashworth (1970) supplemented with 18 % maltose and harvested at a density of not more than 5 x io 6 cells per ml. Strain V-12/IVI2 and mutant Wag-6 were grown in a suspension of 1 x io 10 washed cells of E. co/ib/rpermlin 17 mm phosphate buffer, ph6-o (Gerisch, i960). For use as growthphase cells V-12/M2 was harvested when 4 to 6 x io 8 bacteria per ml were left. Wag-6 cells were used at 8 to 10 h after washing them free of bacteria. For the study of differentiation to aggregation-competence cells were washed free of nutrients and resuspended in the phosphate buffer at a density of 1 x 10' per ml. Ax-2 clone 214 acquired full aggregation-competence after 6 to 8 h, V-12/M2 after 4 to 5 h. camp-s effects on the acquisition of aggregation-competence and on aggregation. Acquisition of aggregation-competence was investigated by adding camp-s to cells cultivated in suspension, washed and resuspended in phosphate buffer. After a period of shaking, control cells and cells treated with camp-s were washed in phosphate buffer. camp-s was removed by washing 3 times and the cells were plated at a density of 5 x io 6 cells per cm* on non-nutrient agar. For the study of camp-s effects on cell aggregation 6 ml 17 mm phosphate buffer, ph 6-o, with or without camp-s, containing 6 x io' aggregation-competent cells, were placed in Petriperm dishes of 5 cm diameter. Petriperm dishes (Heraeus, 6450 Hanau, Germany) contain a teflon bottom permeable to oxygen. Cell behaviour was analysed using a National NV time-lapse video recorder.

3 Developing cells of Dictyostelium discoideum 323 Rate of camp-s hydrolysis. Hydrolysis of camp and camp-s by extracellular phosphodiesterase was assayed with io~ a M substrate according to Eckstein et al. (1974), using the enzyme of the inhibitorless mutant aggr 50-2 of D. discoideum. Hydrolysis by cell surface phosphodiesterase was determined with aggregation-competent cells of Ax-2 by paper chromatography according to Malchow, Nagele, Schwarz & Gerisch (1972). Stalk cell differentiation. Aggregation-competent cells of V-12/M2 were plated on nonnutrient agar supplemented with camp-s. Cellulose was stained by the fluorescent brightener Calcofluor White ST according to Harrington & Raper (1968). Other methods. Chemotaxis was assayed by the use of microcapillaries with an outer tip width of about 0-2 /im filled with a solution of either camp or camp-s (Gerisch, Hiilser, Malchow & Wick, 19756). Changes in light-scattering after application of camp and camp-s were recorded according to Gerisch & Hess (1974) The rise of the intracellular camp concentration in response to camp-s was measured according to Gerisch & Wick (1975). Binding of camp to cell surface receptors was determined as described by Malchow & Gerisch (1974). Extracellular camp phosphodiesterase was assayed according to Riedel, Gerisch, Miiller & Beug (1973) in the extracellular fluid extracted from agar by centrifugation (Gerisch, 1976). One unit of Dictyostelium phosphodiesterase was denned as the amount of enzyme which hydrolysed 1 nmol camp per minute at 23 C and ph 7'4 at saturating substrate concentrations. One inhibitor unit inactivated one phosphodiesterase unit under standard conditions (Riedel et al. 1973)- RESULTS Action of camp-s as an agonist Chemotaxis. As reported by Konijn (1972), camp-s acts as a chemotactic agent on aggregating cells of D. discoideum at 100-fold higher concentrations than are required for camp. In order to compare the effects of camp and camp-s directly, microcapillaries were filled with either camp or camp-s and attraction of aggregation-competent cells was observed. The strain used was V-12/M2. Chemotaxis was obtained with io~ 3 to IO^M camp-s and, under the same conditions, with io" 4 to io~ 6 M camp. io~ B M camp-s or icr 6 M camp did not cause significant attraction of the cells. When 2 capillaries, one filled with icr 5 M CAMP and the other with IO^M camp-s were inserted into the same field of cells, both attracted the cells similarly, with a slightly stronger effect of the camp capillary. Induction of camp pulses. Stimulation of intact cells by camp pulses results in the activation of adenylate cyclase (Roos & Gerisch, 1976) and in the release of camp in form of a pulse (Gerisch et al ; Shaffer, 1975; Roos, Nanjundiah, Malchow & Gerisch, 1975). This response enables the cells to relay a signal over an aggregation territory in the form of a propagated wave (Arndt, 1937; Shaffer, 1962). In order to demonstrate the efficacy of camp-s in inducing camp pulses, cells of strain Ax-2 were stimulated in an agitated suspension, and the transient increase of the intracellular camp concentration was determined as a measure of the response to camp-s (Fig. 1). The camp-s induced cyclic AMP peaks had a half-width of about 60 seconds, in agreement with those induced by camp (Gerisch et al. 1977). Changes of light scattering in cell suspensions. The responsiveness of cells to camp can be conveniently quantitated by recording light-scattering changes (Gerisch & Hess, 1974). Aggregation-competent cells show a biphasic response: a sharp first peak of decreased light-scattering and a broader second one (Fig. 2 A). Similar responses were elicited by camp-s; sometimes the second peak was extended and modu-

4 324 C. Rossier, G. Gerisch, D. Malchoto and F. Eckstein lated by several flat waves, as shown in Fig. 2 A. When the cells were repeatedly stimulated by camp-s this was no longer observed and the response patterns became indistinguishable from those induced by camp. The amplitude of the responses increased gradually with increasing size of the stimuli until, with io" 6 M camp-s, a maximal response was obtained. This was similar in height to the maximal response to camp pulses. Half-maximal height of the 10 20' Trme, min Fig. 1. Peaks of the intracellular camp concentration in response to stimulation by camp-s. The camp concentrations are calculated as fimol/l. of densely packed cell sediment. First 2 arrows, 1 x io~ 7 M CAMP-S; third arrow, 2XIO"'M CAMP-S. responses was obtained with concentrations between io~ 9 and io" 8 M camp-s. Using camp, the concentration which induced half-maximal responses was 3 x io~ 9 M (Gerisch & Hess, 1974). When io~ 7 M camp-s was applied alternately with io~ 7 and io~* M camp, the strength of the responses to camp-s was in between. These results show that camp-s can fully replace camp as an inducer of light-scattering responses. The camp-s concentrations required are not more than 10 times higher than those of camp. In the experiment shown in Fig. 2B, 1 x IO~ 7 M camp-s was repeatedly applied at intervals of 3 min. As a result, the subsequent responses did not reach the height of the first one. After a series of camp-s applications the response to camp proved to be smaller than normal. Then a series of camp pulses was given. Increase of the responses to camp, as well as to camp-s, showed that the cells returned to a sensitive state. These results indicate that camp-s partially desensitized the cells for a period longer than 3 min, whereas camp did not cause a desensitization of similar length. This is interpreted to indicate that camp-s competed for camp by binding to the same receptors on the cell surface; and that camp-s, unlike camp, was not substantially hydrolysed during the interval between 2 pulses. The following experiments confirmed these conclusions.

5 Developing cells of Dictyostelium discoideum Time min r i camp 1 \ camp-s camp 111 l\* ( \ Time mm Fig. 2. Light-scattering changes of aggregation-competent cells in response to io~' M camp or camp-s. The cell concentration was 2 x io 8 /ml. A: CAMP induced a biphasic response typical for aggregation-competent cells. The first camp-s stimulus induced a normal, fast light-scattering change, followed by an unusually long subsequent response. On repeated stimulation by camp-s the response pattern changed into the one which is normally observed with camp. B : Repeated stimulation by camp induced peaks of equal height in contrast to a camp-s stimulus which caused a decline in the following responses. Sensitivity was recovered within 20 min during which test pulses of camp were given.

6 326 C. Rossier, G. Gerisch, D. Malchow and F. Eckstein Binding of camp-s to cell surface receptors Competition of camp-s with camp for receptor sites was shown by incubating cells with i-2 x IO^M camp plus various concentrations of camp-s in the range of 5 x icr 8 to 3 x icr 5 M. High phosphodiesterase activity on the cell surface makes the addition of phosphodiesterase inhibitors necessary for camp-binding studies with wild-type cells. Either cgmp or dithiothreitol can be used as inhibitor (Malchow & Gerisch, 1974; Green & Newell, 1975; Henderson, 1975). Inhibitors can be avoided, however, when a mutant, Wag-6, with low phosphodiesterase activity is used (Malchow et al. 1976). Non-linear kinetics have been obtained for camp binding to cell surface receptors (Green & Newell, 1975). The same was true for camp-s where the apparent inhibitor constants became higher with increasing camp-s concentrations. As a measure of affinity we have used therefore K^, which defines half-maximal binding independent of a specific type of kinetics. AssumingK^ (CAMr>) = 2 x io~ 7 M (Gerisch & Malchow, 1976), the K 0. 6 values for camp-s were within the limits of icr 6 and IO~ 5 M, as calculated from the inhibition of camp binding. The best estimate was -^o-s (cajrp-s) = 2 x icr 8 M. In these experiments cells of mutant Wag-6 were used in the absence of any phosphodiesterase inhibitor. The results indicate a dissociation constant of the receptor-camp-s complex which is only 10 times higher than that of the receptor-camp complex, in accord with the biological activity of camp-s. Rate of camp-s hydrolysis Extracellular phosphodiesterase. V max of camp and camp-s was determined using extracellular phosphodiesterase of aggr 50-2, a mutant of D. discoideum which lacks an inhibitor of this enzyme (Gerisch, 1976). At an initial substrate concentration of IO~ 2 M, the rate of camp hydrolysis was approximately 90-fold higher than it was for camp-s (Fig. 3). The 2 diastereomers of camp-s were hydrolysed with similar velocity. This was checked by NMR (Eckstein et al. 1974) which showed unchanged ratio of the diastereomers at the point of 37% hydrolysis > 1 Fig. 3. Hydrolysis of camp (O) and camp-s (A) by extracellular phosphodiesterase of D. discoideum at 37 C, ph 76. For camp-s the amount of enzyme was twice as high as it was for camp. Substrate concentrations were IO~*M.

7 Developing cells of Dictyostelium discoideum 327 The K m of camp-s was determined from the inhibition of camp hydrolysis. For camp we have previously obtained K m values between 2 and 4X icr 6 M for the extracellular phosphodiesterase (Riedel, Malchow, Gerisch & Nagele, 1972). AssumingX,,, (camp) = 2 x icr'm, the calculated K t value for camp-s was 2 x IO~ 3 M. Cell surface phosphodiesterase. Interaction coefficients (n H ) less than 1 have been obtained for the kinetics of cell surface phosphodiesterase (Malchow, Fuchila & Nanjundiah, 1975). Accordingly, the apparent K t values increased with increasing camp-s concentrations. Using a camp-concentration of 8 x io" 8 M and assuming ^o-5(camp) = 7><IO~ 7 M (Malchow et al. 1975), the calculated K^ for camp-s was 2 x io~* M. Effects of camp-s on morphogenesis in agar plate cultures Inhibition of cell aggregation in strains NC-4 and Ax-2. The above results indicate that for D. discoideum cells camp-s is an agonist of the chemoreceptors which is much more slowly hydrolysed than camp. Because of these properties camp-s is suitable to maintain fairly constant agonist concentrations throughout development of D. discoideum. The drug was dissolved in the agar on which the cells were grown, together with a phosphodiesterase-negative strain of E. coli. In the wild-type strain NC-4, ce " aggregation was completely inhibited by 1 x io" 6 M camp-s (Table 1). In Ax-2, a strain selected from NC-4 f r axenic growth (Watts & Ashworth, 1970), aggregation was similarly sensitive. It was completely inhibited for 3 days by 5 x io" 7 M CAMP-S. Growth was not inhibited under the same conditions even with io~* M CAMP-S (Table 1). After 6 days Ax-2 cells aggregated and formed minute fruiting bodies in cultures to which icr 8 to io~ 4 M camp-s had been added. During aggregation the formation of streams of cohering cells was still completely suppressed. camp-s resistant cell aggregation in strain v-12 end in other species. The wild-type strains v-12 and NC-4 nave been independently isolated from natural sources by Dr K. B. Raper. In v-12 and in the clone M2 derived from it aggregation was markedly less sensitive to camp-s than in NC-4 or Ax-2. V-12/M2 aggregated normally in the presence of io~* M CAMP-S. With io" 4 and io~ 3 M CAMP-S streams of aggregated cells were formed, but they were irregularly arranged. Time-lapse recording revealed that the cells were highly motile but apparently without chemotactic orientation. They aggregated by random contacts. The streams of cohering cells moved with free ends without orientation towaids aggregation centres. Often they slid along each other in opposite directions. These observations indicate that the strain V-12/M2 still develops, in the presence of camp-s, the cell adhesion system that functions in aggregation. This seems to be sufficient for development to continue in the absence of chemotaxis up to the final stage of fruiting body formation. M2 is a fast-developing strain which in suspension cultures acquires aggregationcompetence 4 to 5 h after the end of growth (Beug, Katz & Gerisch, 1973). Streams of aggregating cells tend to be shorter and fruiting bodies smaller than in NC-4. These properties are, however, not necessarily linked to camp-s resistance. Mi, another clone of v-i2, forms long streams and needs 8 to 9 h for the acquisition of aggregation-

8 C. Rossier, G. Gerisch, D. Malchow and F. Eckstein

9 Developing cells of Dictyostelium discoideum. 329 competence (Gerisch, 1962). It also proved to be largely resistant to camp-s. Mi still formed short streams of aggregating amoebae and small fruiting bodies in the presence of io~ 8 to IO~*M camp-s. Aggregates arose, however, at a greater distance from the giowing edge of the colony than in controls. Fruiting body formation was accordingly delayed. At IO~*M camp-s they were still missing after 3 days of culture (Table 1). Polysphondylium violaceum and P. pallidum, 2 species which aggregate in response to chemotactic factors other than camp (Wurster, Pan, Tyan & Bonner, 1976), aggregated normally in the presence of io" 6 or io~* M camp-s. At the latter concentration, often unbranched fruiting bodies were formed. The effect of camp-s depends on its lifetime As described above, the inhibitory effect of camp- S on cell aggregation in strain Ax-2 was partially reversed after prolonged culture for 6 days. This reversal was apparently due to the slow hydrolysis of camp-s by extracellular phosphodiesterase. The effect of camp-s was enhanced by an inhibitor of this enzyme. The inhibitor was partially purified from the culture fluid of D. purpweum. When supplemented with 50 units of the inhibitor per ml agar, io" 8 M camp-s completely inhibited development for at least 10 days. The inhibitor alone had no detectable effect. On the other hand, short streams and fruiting bodies were formed within 4 days when together with io" 6 M camp-s, beef heart phosphodiesterase was added. In the absence of camp-s, the phosphodiesterase had no detectable effect. (For exact conditions, see Materials and Methods.) These results imply that inhibition of development is not due to the product of camp-s hydrolysis, adenosine 5'-phosphorothioate (AMP-S). This was confirmed by the absence of any detectable effect of authentic AMP-S. The highest concentration tested was io" 4 M. Since camp-s was used as the triethylammonium salt, io" 4 M triethylamine was also tested; it did not impair development. Development of strain Ax-2 on nutrient agar plates was not affected by camp or by its analogues 8-bromo and 8-azido cyclic AMP at concentrations of up to 1 x io" 3 M. These analogues interact with cyclic AMP receptors, although with lower affinity than camp, and they are hydrolysable by phosphodiesterase of D. discoideum (Konijn, 1973; Malchow, Fuchila & Jastorff, 1973; and unpublished results). Presumably they were substantially hydrolysed during the growth phase by extracellular phosphodiesterase. When camp was applied at higher concentrations (2 or 4 x io" 3 M) the cells still aggregated into streams, but fruiting bodies were aberrant, as described by Nestle & Sussmann (1972). However, when its hydrolysis was slowed down, camp caused substantial delay of aggregation. In controls, streams of cohering cells appeared at a distance of 1-2 mm behind the growing edge of a colony. When io" 3 M camp plus 50 units per ml of phosphodiesterase inhibitor from D. purpureum had been added to the agar, the minimal distance was 5 mm. This inhibition was enhanced by the addition of cgmp, a competitive inhibitor of extracellular as well as cell-surface phosphodiesterase M camp plus io~ 3 M cgmp plus 50 units per ml of inhibitor from D. purpureum strongly retarded aggregation, so that streams were formed only

10 330 C. Rossier, G. Gerisch, D. Malchow and F. Eckstein at a minimal distance of 14 mm from the boundary of the colony. cgmp alone did not affect development under our conditions, probably due to substantial hydrolysis. A paradoxical effect was observed when io^8 M camp-s was combined with io" 5 M camp. After 3 days of culture many slugs and small fruiting bodies had been formed. As mentioned above, camp-s alone completely inhibited aggregation, io" 4 or icr 3 M camp were not as effective as io~* M CAMP in partially suppressing the inhibitory action of camp-s. The camp effect was at least partly due to an increase of extracellular phosphodiesterase activity and, in consequence, to the shortening of the lifetime of camp-s. Table 2 shows that camp-s alone caused an increase of the phosphodiesterase activity. This increase was, however, stronger in combination with IO~ 5 M camp. Table 2. Effects of camp on morphogenesis and extracellular phosphodiesterase activity in the presence and absence of camp-s Cone, M Units of f camp-s o o 0 o o io-«10-' icr* 10"* camp o ' io-» JO" 5 IO- 1 io-»j o IO"* IO- 5 IO- 4 JO" 3 Morphology Streams of aggregating cells, slugs and fruiting bodies No fggregation No aggregation Aggregates without streams. Slugs and small fruiting bodies No aggregation No accrecation phosphodiesterase/mj. 0 I 2 I V4 ) Strain Ax-2 was grown for 3 days on 5 ml glucose-peptone agar with E. coli (PD~) in Petri dishes of 35 mm diameter. Phosphodiesterase activity was determined in the fluid pressed out of the rgar by centrifugation; it thus represents the averege over the whole thickness of the agar. IO II Separation of cell development end chemotaxis The development of cell aggregates can be dissected into 2 steps: the acquisition of aggregation-competence and the actual assembly of the cells. These processes can be separated using agitated suspensions where the cells acquire aggregationcompetence. However, the cells cannot aggregate by chemotaxis unless they are transferred onto a supporting surface on which they can migrate. Aggregation inhibitors like camp-s can be added either to the suspension culture and then washed off, or they can be applied afterwards when aggregation-competent cells are spread onto a surface. In the first case the sensitivity of differentiation towards aggregationcompetence can be tested, in the latter case the sensitivity of chemotaxis.

11 Developing cells of Dictyostelium discoideum 13 * S&Jl A^ *r\ ^^J vise--ips3 Fig. 4. camp-s inhibits the acquisition of aggregation-competence. Cells were washed and plated on non-nutrient agar after shaking for 8 h in A, phosphate buffer ph 6-o, B, the same with 2 x io~* M camp-s, and c, 1 x io" 6 M camp-s. Photographs were taken 0-5 h after plating. In B only few chains of connected, elongated cells were formed, in c almost none.

12 332 C. Rossier, G. Gerisch, D. Malchow and F. Eckstein camp-s suppresses the acquisition of aggregation-competence in strain Ax-2. After 8 h of starvation in phosphate buffer Ax-2 cells became fully aggregation-competent. When plated on non-nutrient agar they immediately elongated and aggregated into streams (Fig. 4A). When ZXIO^M camp-s was present during the 8-h period and then washed off, only few streams were formed on agar within a period of min (Fig. 4B). Most of the cells remained single and did not elongate. After incubation for 8 h in 1 x io~* M camp-s no streams appeared and only few cells were elongated (Fig. 4c). 1 x io" 8 M AMP-S, the product of camp-s hydrolysis, had no effect on the acquisition of aggregation-competence. Cells to which 2 x io" 6 M camp had been added after the removal of nutrient medium, aggregated without delay when they were washed and plated on agar after a period of 8 h. Under the same conditions, a slight inhibitory effect of 1 x IO^M camp on cell development was observed: in one of 2 experiments, washed 8-h cells required 1 h befoie they started to aggregate by forming streams. After incubation for 8 h with camp-s the cells were not permanently prevented from development. Slugs were produced overnight from cells plated on agar after the removal of camp-s. At the same time control cells had already formed fruiting bodies. During the first 2 h after the removal of nutrient medium the adenylate cyclase activity of Ax-2 cells remains low (Klein, 1976), and for the first 1 to 2 h of development camp pulses appear to have no influence on the acquisition of aggregationcompetence (Darmon et al. 1975). In the above experiments camp-s was added at the beginning of this period. In order to investigate if camp is able to control cell differentiation already during these first 2 h of development, IO^M camp-s was added to aliquots of suspension cultures either immediately after the removal of nutrient medium, or 2 h later. In both cases the cells were washed free of camp-s at 8 h after the removal of nutrient medium. In contrast to the samples which received camp-s immediately, those to which camp-s was added at 2 h showed no complete suppression of cell differentiation. Within 30 min short streams of aggregated cells were formed. This experiment was performed 3 times with similar results. Cells of strain v-12 acquire aggregation-competence in the presence of camp-s. Cells of V-12/M2 were grown in suspension cultures on E. coli and washed free of bacteria during the exponential growth phase, io" 6 or IO^M camp-s was added immediately after washing, and development of the cells compared with control cells. This was done by washing of the cells after various periods, and plating them in Petriperm dishes. In the presence of camp-s as well as in controls, the cells became aggregation-competent within 4 h. This was indicated by their ability to assemble into streams of adhering cells and to form normally developing aggregates. Chemotaxis is inhibited by a uniform concentration of camp-s. In both strains Ax-2 and V-12/M2, the chemotactic orientation of aggregating cells was suppressed by 2 x io~ 6 M camp-s. This was tested by subjecting aggregation-competent cells to camp-s in Petriperm dishes. Aggregation in the presence and absence of camp-s was followed by time-lapse video recording. In the controls chemotactic orientation of single cells towards cell aggregates was clearly observed. Large aggregates were

13 Developing cells of Dictyostelium discoideum 333 Fig. s. Stalk cells formed in the presence of i x io~' M camp-s (A) or I X IO~* M camp-s (B). Photographs were taken after 7 days on non-nutrient agar. A, phase-contrast, B, fluorescence of cellulose stained with Calcofluor White ST. CEU 35

14 334 C- Rossier, G. Gerisch, D. Malchow and F. Eckstein formed, consisting of streams about 400 /tm long which were radially oriented around aggregation centres. In the presence of camp-s streams were still formed, but they were shorter and, duiing the earlier stages of aggregation, irregularly oriented. The movement of single cells was apparently not directed by chemotaxis; the cells entered aggregates by random collisions. At later stages of aggregation the tightly cohering cells in the streams assembled into rounded cell masses. Thereby the cells moved towards the centres. Guidance by cell-cell contact rather than chemotaxis was probably responsible for orientation of the cells within the streams. Addition of io~ 5 M camp-s to cells which had already started to aggregate resulted in the disruption of streams into groups of no-longer-elongated but still cohering cells. Although these cells exerted intense pseudopodial activity, they showed no evidence for chemotactic orientation as judged by time-lapse video recording. Induction of stalk cells For the induction of stalk cells camp conccntiations of the order of IO" 3 M are required (Bonner, 1970; Town et al. 1976). The effective concentrations of camp-s were substantially lower. Cells of strain V-12/M2 were cultivated in suspension up to the aggregation-competent stage, and plated on non-nutrient agar at a density of 2 x 10 s cells per cm 2. On agar containing io" 6 or io" 6 M camp-s, many vacuolated, thick-walled cells were formed from single cells or small cell groups (Fig. 5 A). The stalk cell character was proved by staining the walls of these cells with Calcofluor White ST, a fluorescent indicator of cellulose (Harrington & Raper, 1968) (Fig. 5B). Under the above conditions fruiting bodies were still formed in the presence of io~ 6 or io~ i M camp-s, but they sometimes lacked spore heads or had only minute ones. Often aggregates were converted into piles of stalk-like cells, rather than into fruiting bodies. DISCUSSION In this paper adenosine 3',5'-cyclic phosphorothioate (camp-s) is used as a probe for cyclic-amp-dependent processes in the development of D. discoideum. Cyclic AMP itself is rapidly hydrolysed by phosphodiesterases present at the cell surface or in the extracellular medium. Because of the combination of low V max and high K m the thioanalogue is about times more slowly hydrolysed than camp at substrate concentrations in the micromolar range. These are the biologically relevant concentrations. camp-s binds to camp receptors on the cell surface as demonstrated by its competition with camp binding. Accordingly, camp-s acts as an agonist for chemotaxis and as an inducer of camp pulses two responses known to be mediated by cell surface receptors. The chemotactic activity of camp-s, together with its low affinity for cell-surface phosphodiesterase, demonstrates that hydrolysis of a compound is not a requirement for its chemotactic activity, although phosphodiesterase at the cell surface might improve the detection of a concentration gradient by the cellular recognition system (Nanjundiah & Malchow, 1976). It is not clear whether the induction of stalk cell differentiation by camp is also

15 Developing cells of Dictyostelium discoideum 335 mediated by cell surface receptors since concentrations in the millimolar range are required (Bonner, 1970; Town et al. 1976). This could be interpreted to mean that stalk cell induction is mediated by a rise of the intracellular camp pool due to diffusi6n of camp through the plasma membrane. In contrast, camp-s induces stalk cells at a concentration of 1 x io~ 6 M. It follows that a high concentration is not required if the stalk cell inducer has a longer lifetime. Since, according to unpublished results, the permeation of camp-s through the plasma membrane is negligible, it is likely that stalk cell differentiation is induced by continued action of the agent on cell surface receptors. Because of its extended lifetime camp-s will block reactions which depend on a change of the camp concentration in time. Some of these reactions can be conveniently recorded by light-scattering measurements. As expected, camp-induced lightscattering changes were suppressed by previous application of camp-s. Investigation was then made to see if the longer lifetime of camp-s results in an extended response to a sudden increase of its concentration. Sometimes this was the case, as shown in Fig. 2A. Most often, however, the light-scattering responses to camp-s were as short as those induced by camp (Fig. 2B). This was also true for a pulse of intracellular camp which, when induced by camp-s, had the same half-width of 1 min as a pulse induced by camp (Gerisch et al. 1977). These results confirm previous reports that there are responses to camp which are subject to rapid adaptation. Light-scattering changes, the activation of adenylate cyclase, and probably also chemotaxis (Gerisch et al ) are based on the detection of changes of camp concentrations in time. Other responses are based on the detection of steady concentrations. This is the case for the regulation of extracellular phosphodiesterase activity (Gerisch, 1978). In a later paper we will show that the use of camp-s enables one to easily distinguish between these different types of response. It has been demonstrated in strain Ax-2 that camp is periodically released in the form of pulses before the cells acquire full aggregation-competence. Such pulses stimulate the differentiation of cells to the aggregation-competent stage, as shown by application of camp pulses to cells which do not spontaneously differentiate (Gerisch et al. 1975a; Darmon et al. 1975). By contrast, a steadily elevated concentration of camp has little promoting (Darmon et al. 1975) or even a contrary effect (Gerisch et al a). A continuous influx of camp is also known to block the spontaneous periodic light-scattering changes known to be linked to the generation of camp pulses (Gerisch & Hess, 1974). Since camp-s is more suitable than camp for maintaining non-fluctuating agonist concentrations, we have studied differentiation to the aggregation-competent stage under the influence of camp-s. camp-s inhibited this differentiation in strain Ax-2 (Fig. 4). The inhibition is probably due to interference with the generation and recognition of spontaneous camp pulses. The finding that strain V-12/M2 acquires aggregation competence in the presence of camp-s would then suggest that in this strain periodic camp signals are not necessary. Thus expression of the characters of aggregation-competent cells appears to be constitutive with respect to camp signals in v-12, and inducible in Ax-2.

16 336 C. Rossier, G. Gerisch, D. Malchoxv and F. Eckstein The finding that camp-s causes stronger inhibition in Ax-2 when applied immediately after removal of the nutrient medium as compared to 2 h later, suggests a camp-regulated process at the beginning of the differentiation period. This is in accord with previous results of Alcantara & Bazill (1976) on camp-phosphodiesterase effects during the first hours of starvation. camp pulses stimulate cell differentiation not earlier than 1 to 2 h after starvation (Darmon et al. 1975). The early phosphodiesterase and camp-s effects might be due to a developmental process which is regulated by camp but not necessarily by periodic pulses. The resistance of cell aggregation in strain V-12/M2 to camp-s concentrations as high as io" 3 M was unexpected in the light of chemotaxis assays which showed that cells of V-12/M2 respond chemotactically both to camp and camp-s, in the same way as do those of NC-4 used by Konijn (1972). Thus camp-s binds to the chemotaxis receptors of V-12/M2, and high concentrations should block the receptors so that camp gradients become undetectable. The reason why camp-s does not inhibit cell aggregation in V-12/M2 appears to be the ability of the cells to aggregate without chemotactic orientation by random walk and adhesion of collided cells. We thank Professor D. Hiilser, Stuttgart, for cooperation in the chemotaxis assays and Mrs M. Brunner and G. Waser for expert assistance. Our work was supported by a grant of the Schweizerischer NationaJfonds to G. Gerisch and of the Deutsche Forschungsgemeinschaft to D. Malchow. REFERENCES ALCANTARA, F. & BAZILL, G. W. (1976). Extracellular cyclic AMP-phosphodiesterase accelerates differentiation in Dictyostelium discoideum. J. gen. Microbiol. 9a, ARNDT, A. (1937). Untersuchungen iiber Dictyostelium mucoroides Brefeld. Wilhelm Roux Arch. EntwMech. Org. 136, BEUG, H., KATZ, F. E. & GERISCH, G. (1973). Dynamics of antigenic membrane sites relating to cell aggregation in Dictyostelium discoideum. J. Cell Biol. 56, BONNER, J. T. (1970). Induction of stalk cell differentiation by cyclic AMP in the cellular slime mold Dictyostelium discoideum. Proc. natn. Acad. Set. U.S.A. 65, DARMON, M., BRACHET, P. & PEREIRA DA SILVA, L. H. (1975). Chemotactic signals induce cell differentiation in Dictyostelium discoideum. Proc. natn. Acad. Sci. U.S.A. 72, ECKSTEIN, F., SIMONSON, L. P. & BAR, H-P. (1974). Adenosine 3',5'-cyclic phosphorothioate: Synthesis and biological properties. Biochemistry, N.Y. 13, GERISCH, G. (i960). Zellfunktionen und Zellfunktionswechsel in der Entwicklung von Dictyostelium discoideum. I. Zellagglutination und Induktion der Fruchtk6rperpolarit3t. Wilhelm Roux Arch. EntwMech. Org. 152, GERISCH, G. (1962). Zellfunktionen und Zellfunktionswechsel in der Entwicldung von Dictyostelium discoideum. IV. Der Zeitplan der Entwicklung. Wilhelm Roux Arch. EntwMech. Org. 153, GERISCH, G. (1976). Extracellular cyclic-amp phosphodiesterase regulation in agar plate cultures of Dictyostelium discoideum. Cell Differentiation 5, GERISCH, G. (1978). Control circuits in cell aggregation and differentiation of Dictyostelium discoideum. In Mechanisms of Cell Change (ed. J. D. Ebert & T. Okada). New York: Wiley. GERISCH, G., FROMM, H., HUESGEN, A. & WICK, U. (1975a). Control of cell-contact sites by cyclic AMP pulses in differentiating Dictyostelium cells. Nature, Lond. 255, GERISCH, G. & HESS, B. (1974). Cyclic-AMP-controlled oscillations in suspended Dictyostelium cells: Their relation to morphogenetic cell interactions. Proc. natn. Acad. Set. U.S.A. 71, ,

17 Developing cells of Dictyostelium discoideum 337 GERISCH, G., HOLSER, D., MALCHOW, D. & WICK, U. (19756). Cell communication by periodic cyclic-amp pulses. Phil. Trans. R. Soc. Ser. B 272, GERISCH, G. & MALCHOW, D. (1976). Cyclic AMP receptors and the control of cell aggregation in Dictyostelium. In Advances in Cyclic Nucleotide Research, vol. 7 (ed. P. Greengard & G. A. Robison), pp New York: Raven Press. GERISCH, G., MALCHOW, D., ROOS, W., WICK, U. & WURSTER, B. (1977). Periodic cyclic-amp signals and membrane differentiation in Dictyostelium. In Cell Interactions in Differentiation (ed. M. Karkinen-jaaskelainen, L. Saxen & L. Weiss), pp New York and London: Academic Press. GERISCH, G. & WICK, U. (1975). Intracellular oscillations and release of cyclic AMP from Dictyostelium cells. Biochem. biophys. Res. Contmun. 65, GREEN, A. A. & NEWELL, P. C. (1975). Evidence for the existence of two types of campbinding sites in aggregating cells of Dictyostelium discoideum. Cell 6, HARRINGTON, B. J. & RAPER, K. B. (1968). Use of a fluorescent brightener to demonstrate cellulose in the cellular slime molds. Appl. Microbiol. 16, HENDERSON, E. J. (1975). The cyclic adenosine 3',s'-monophosphate receptor of Dictyostelium discoideum. J. biol. Chem. 250, KLEIN, C. (1976). Adenylate cyclase activity in Dictyostelium discoideum amoebae and its changes during differentiation. FEBS Letters, Amsterdam 68, KONIJN, T. M. (1972). Cyclic AMP as a first messenger. In Advances in Cyclic Nucleotide Research, vol. 1 (ed. P. Greengard & G. A. Robison), pp New York: Raven Press. KONIJN, T. M. (1973). The chemotactic effect of cyclic nucleotides with substitutions in the base ring. FEBS Letters, Amsterdam 34, KONIJN, T. M., VAN DE MEENE, J. G. C, BONNER, J. T. & BARKLEY, D. S. (1967). The acrasin activity of adenosine-3',5'-cyclic phosphate. Proc. natn. Acad. Sci. U.S.A. 58, MALCHOW, D., FUCHILA, J. & JASTORFF, B. (1973). Correlation of substrate specificity of campphosphodiesterase in Dictyostelium discoideum with chemotactic activity of camp-analogues. FEBS Letters, Amsterdam 34, 5-9. MALCHOW, D., FUCHILA, J. & NANJUNDIAH, V. (1975). A plausible role for a membranebound cyclic AMP phosphodissterase in cellular slime mold chemotaxis. Biochim. biophys. Ada 385, MALCHOW, D. & GERISCH, G. (1974). Short term binding and hydrolysis of cyclic 3',5'- adenosine monophosphate by aggregating Dictyostelium cells. Proc. natn. Acad. Sci. U.S.A. 71, MALCHOW, D., NAGELE, B., SCHWARZ, H. & GERISCH, G. (1972). Membrane-bound cyclic AMP phosphodiesterase in chemotactically responding cells of Dictyostelium discoideum. Eur.J. Biochem. 28, MALCHOW, D., ROBINSON, D. & ECKSTEIN, F. (1976). Cyclic AMP receptors at the surface of aggregating Dictyostelium discoideum cells. In Surface Membrane Receptors (ed. R. A. Bradshaw, W. A. Frazier, R. C. Merrell, D. I. Gottlieb & R. A. Hogue-Angeletti), pp New York: Plenum Publishing Corporation. NANJUNDIAH, V. & MALCHOW, D. (1976). A theoretical study of the effects of cyclic AMP phosphodiesterases during aggregation in Dictyostelium. J. Cell Sci. 22, NESTLE, M. & SUSSMAN, M. (1972). The effect of cyclic AMP on morphogenesis and enzyme accumulation in Dictyostelium discoideum. Devi Biol. 28, RIEDEL, V., MALCHOW, D., GERISCH, G. & NAGELE, B. (1972). Cyclic AMP phosphodiesterase interaction with its inhibitor of the slime mold, Dictyostelium discoideum. Biochem. biophys. Res. Commun RIEDEL, V., GERISCH, G., MOLLER, E. & BEUG, H. (1973). Defective cyclic-adenosine-3',5'- phosphate-phosphodiesterase regulation in morphogenetic mutants of Dictyostelium discoideum. J. molec. Biol. 74, Roos, W. & GERISCH, G. (1976). Receptor-mediated adenylate cyclase activation in Dictyostelium discoideum. FEBS Letters, Amsterdam 68, Roos, W., NANJUNDIAH, V., MALCHOW, D. & GERISCH, G. (1975). Amplification of cyclic-amp signals in aggregating cells of Dictyostelium discoideum. FEBS Letters, Amsterdam 53,

18 338 C. Rossier, G. Gerisch, D. Malchow and F. Eckstein Roos, W., SCHEIDECGER, C. & GERISCH, G. (1977). Adenylate cyclase activity oscillations as signals for cell aggregation in Dictyostelium discoideum. Nature, Lond. 266, SHAFFER, B. M. (1962). The Acrasina. Adv. Morphogenesis 2, SHAFFER, B. M. (1975). Secretion of cyclic AMP induced by cyclic AMP in the cellular slime mould Dictyostelium discoideum. Nature, Lond. 255, TOWN, C. D., GROSS, J. D. & KAY, R. R. (1976). Cell differentiation without morphogenesis in Dictyostelium discoideum. Nature, Lond. 262, WATTS, D. J. & ASHWORTH, ]. M. (1970). Growth of myxamoebae of the cellular slime mould Dictyostelium discoideum in axenic culture. Biochem. J. 119, WURSTER, B., PAN, P., TYAN, G.-G. & BONNER, J. T. (1976). Preliminary characterization of the acrasin of the cellular slime mold PolysphondyHum violaceum. Proc. natn. Acad. Sci. U.S.A. 73, (Received 23 June 1978)