The tip of the Dictyostelium discoideum pseudoplasmodium as an organizer

Transcription

1 /. Embryol. exp. Morph. Vol. 33, 1, pp , Printed in Great Britain The tip of the Dictyostelium discoideum pseudoplasmodium as an organizer By JONATHAN RUBIN 1 AND ANTHONY ROBERTSON 1 From Department of Biophysics and Theoretical Biology, University of Chicago SUMMARY We have extended Raper's original work on the organizing ability of the tip of the Dictyostelium discoideum slug. Our new results are that tips from all multicellular (pseudoplasmodial) stages act as organizers; that the structure organized by a tip depends on the developmental stage of the cells responding to the tip's signal; that tips from all stages release a qualitatively similar signal which is continuous, most probably a gradient of c-amp; and that the signal from fruiting-body tips appears stronger than that from conus tips. We discuss these results with reference to the control of morphogenetic movement and patterned differentiation and point out that the D. discoideum tip is analagous to a classical organizer, whose signal is interpreted according to the state of determination of cells in its field of influence. INTRODUCTION Very little is known about the systems controlling cellular behaviour during development. One of the most important concepts is that of the organizer (Spemann, 1938), a group of cells in an embryo which can define a developmental axis and can behave autonomously in grafting experiments. Although much of the phenomenology of organizer behaviour has been described, the mechanism of organizer action is not understood at all. However, Raper (1940) pointed out that the tip of the Dictyostelium discoideum slug, the migratory phase of the pseudoplasmodium, behaved like a classical organizer. More recently Farnsworth (1973) has measured the time for the determination of a new tip when the original one has been removed from a pseudoplasmodium. (Full reviews of D. discoideum biology are to be found in Raper (1940), Shaffer (1962), Bonner (1967), Gerisch (1968), Ashworth (1971), Robertson & Cohen (1972).) As D. discoideum provides a multicellular system sharing many of the properties, in particular differentiation and regulation (see Wolpert (1969) for a review), of Metazoan embryos, we decided to investigate tip function further and thereby examine the mode of action of a primitive organizer. A fortunate feature of D. discoideum is that the aggregative signal is now quite well understood and is 1 Authors' address: Department of Biophysics and Theoretical Biology, University of Chicago, 920 East 58th Street, Chicago, Illinois 60637, U.S.A. 15-2

2 228 J. RUBIN AND A. ROBERTSON apparently retained to control at least some aspects of development throughout the morphogenetic cycle. We now provide a brief review of those features of the aggregative signalling system relevant to the experiments reported in this paper. There is good, albeit circumstantial, evidence that the molecule used as an aggregative signal (acrasin) is cyclic adenosine monophosphate (c-amp) (Konijn, van de Meene, Bonner & Barkley, 1967; Konijn, Barkley, Chang & Bonner, 1968). D. discoideum amoebae release c-amp during aggregation, as well as a phosphodiesterase (PDE) (Chang, 1968) which destroys the c-amp. High c-amp levels can inhibit cell movement and aggregation (Konijn, 1972), as can high extracellular beef heart PDE activity (Robertson, 1974). c-amp causes chemotaxis of aggregative amoebae and aggregation can be controlled by an artificial source of c-amp (Robertson, Drage & Cohen, 1972). However, the experiments and conclusions in this paper do not depend on the identification of acrasin with c-amp. During interphase, the period between the end of feeding and the beginning of aggregation, amoebae successively become chemotactically sensitive to c-amp, capable of relaying a c-amp signal and then, for a few amoebae, capable of releasing acrasin signals autonomously (Cohen & Robertson, 1972). By the time autonomous signalling cells have emerged all the amoebae in a field entering interphase at the same time are responsive both chemotactically and by being able to relay a c-amp signal. There are well-defined, different, threshold c-amp concentrations for chemotaxis and relaying (Konijn et ah 1968; Cohen & Robertson, 1971 a, b; Robertson, 1974; Robertson & Cohen, 1974). As the times of emergence of the chemotactic and relaying competences are well defined it is possible to use a field of amoebae in interphase as a sensitive monitor of the waveform and amplitude of the signal released by a suspected c-amp source. For example, a periodic artificial signal source, in a field of amoebae 4 h into interphase, evokes periodic chemotactic movements towards the signal source (Robertson, et ah 1972), while a continuous signal evokes continuous movement (Robertson, 1974). In a field approximately 6 h old both periodic and continuous signals evoke periodic signal relaying if the relaying threshold concentration is exceeded. Amoebae close to the source move either periodically or continuously, while amoebae receiving a relayed signal move periodically. In this paper we show that the D. discoideum tip acts as an organizer throughout morphogenesis from the late aggregate stage, and that the signal it releases is continuous. MATERIALS AND METHODS Amoebae A stock of D. discoideum, NC-4, was obtained from Professor K. B. Raper. The food bacteria were Aerobacter aerogenes.

3 Tip ofdictyostoliumpseudopiasmodium as organizer 229 Plating, growth, harvesting and standardization Spores from culminated sorocarps were collected in a drop of distilled water held in a bacterial inoculating loop. The spores were then transferred to a test tube containing 3 ml of A. aerogenes suspension in growth broth, made up as follows: yeast extract, 0-5 g/1.; peptone, 5 g/1.; dextrose, 5 g/1.; KH 2 PO 4,2-25 g/1.; K 2 HPO 4.12H 2 O, 1-5 g/1.; MgSO 4.7H 2 O, 0-5 g/1. Four drops of spore-bacteria medium were transferred to growth plates containing 2 0 agar with 0-5 % peptone and 0-5 % dextrose made up in KK 2 buffer. The amoebae were allowed to grow vegetatively on the bacteria for 48 h at a constant temperature of 23 C. While the amoebae were still in the vegetative phase, they were harvested in chilled KK 2 buffer, made up as follows: KH 2 PO 4, 2-31 g/1.; K 2 HPO 4.12H 2 O, 1-3 g/1.; MgSO 4.7H 2 O, 0-5 g/1. The amoebae were washed free of bacteria by suspending them in buffer and centrifuging at 650 g for 2 min. This process was repeated three times. The final pellet of amoebae was then suspended in 5 ml of KK 2 buffer, and the concentration of amoebae was determined with a haemocytometer. It was adjusted to 5 x 10 7 cells/ml, and 0-5 ml of this suspension was plated out on a non-nutrient agar plate. The amoebae were then synchronized according to Bonner (1967) by returning the plates of harvested amoebae to an incubator at 23 C for 2 h and then refrigerating them at 4 C until needed. The plates were then placed in the 23 C incubator for the desired incubation period. Transplantation Amoebae on some growth plates were not harvested but were allowed to continue morphogenesis beyond the vegetative phase to provide material for the transplant experiments. Glass micropipettes drawn into sharp points were used to effect tip transfers. The removal of tips was accomplished by placing the pipette at the base of the tip when there was a clear morphological difference between the tip and the remainder of the structure (i.e. early fruiting body or stage 17 or 18 of Farnsworth) (Farnsworth, 1973). If there was no clear morphological distinction between tip and body of the structure, as in a slug, the removal was done as near to the front end of the structure as possible. The transfer was then made to the recipient structure. In the later stages of the life-cycle (Farnsworth stages 8-20) the tip was pushed into the recipient until it broke the surrounding slime sheath, if present, and was securely attached. Care was taken to deform the receiving structure as little as possible in the process. When transferring a tip into a field of interphase amoebae, the tip was placed into the field as gently as possible in order to avoid injury to the tip. All of the transplants were done under a Nikon Dissecting Microscope.

4 230 J. RUBIN AND A. ROBERTSON Filming The filming was done in a constant-temperature room ( C.) with a Bolex 16 mm movie camera and a Nikon CFMA camera drive. The filming microscope was a Nikon Apophot using transmitted light. The exposure was set automatically. Frame rates varied from 2 to 6 per minute. The most common magnification used was \% (5 x eyepiece, 1-2 x plan objective and - x relay lens). Experiments were recorded on Kodak 16 mm 4X reversal film. Some transplants of tips into fields of interphase amoebae were recorded with a Concord video-tape recorder, VTR-648, at a rate of 1\ frames/sec and played back at the normal rate of 60 frames/sec. The images were displayed on a Concord Mr-900 monitor. Statistical experiments One set of transfers was done to provide large numbers of results for statistical purposes. The transplants were performed as previously described. However, after each tip transplant was completed the recipient and transplant were transferred to a plain agar plate, allowed to develop, and observed to see whether two fruiting bodies had formed. If one fruiting body was less than one-quarter the size of the other the experiment was scored as a failure, as we had observed that tips which fell off and regulated to make independent fruiting bodies never attained that size. Timing and period measurement Time and periods were determined by projecting the films with a Traid Selecta-frame projector, Model 16N/LS, equipped with an automatic frame counter. Times were also measured with a stop-watch. RESULTS A. Grafts of tips into pseudoplasmodia (i) Tips from all multicellular stages were transplanted to each multicellular stage of the life-cycle of D. discoideum. Each tip was taken from the donor structure and placed on the recipient. Each experiment was filmed by time-lapse microphotography. Line drawings from films of each category of experiment are shown in Fig. 1, and a typical graft is shown in the photograph. The dynamics of a transplant were independent of the source of the tip or of the stage of the recipient structure. In a successful transplant the tip began to move within 15 min. It appeared to be orienting itself with respect to the recipient. If the recipient was an upright structure, such as a conus, the tip often began to slide downwards. However, within 30 min the tip stopped its movement and began to set up its own field of influence in the recipient. This event was manifested by a ring or indentation which formed between the tip plus the field of cells it had pulled in and the field of cells still under the control of the original structure's tip. The average time for the indentation to appear was min

5 Tip of Dictyostelium pseudoplasmodium as organizer 231 Transplant Conns onto (plan) con us Slug onto (plan) slug Fruiting body onto (plan) slug Slug onto (elevation) con us Slug onlo (elevation) fruit inn bodv Fruiting body onto (elevation) fruiting bodv Fruiting body onto (plan) con us rn COIHIS onto (elevation) fruit inn bodv Conus onto (plan) slue Time after transplant (h) Two Result at 4 h Two slugs Two slugs Two slugs Two slugs Two fruiting bodies Two fruiting bodies Two slugs fruiting bodies Two slugs Fig. 1. Line drawings, taken directly from single frames of our films, showing pseudoplasmodial organization by grafted tips, which are shaded. from the time of grafting. This corresponds well to the time of field formation, approximately 34 min, measured by Farnsworth (1973). The indentation completely demarcated the structure under the control of the recipient's tip from the newly formed structure under control of the transplanted tip (see the photograph). Once it had formed the two structures acted independently of each other. The recipient continued its developmental cycle independently of the grafted tip and the cells which it had lost. The transplanted tip and the new cells which it had attracted also proceeded through the standard developmental cycle. In all cases in this group of experiments, the new structure produced by the grafted tip was at the same stage as the recipient structure. If, for example, a slug tip was transplanted on to a fruiting body, the slug tip and the cells it had pulled in immediately culminated and fruited without going through the normal migratory phase. In contrast, if a fruiting-body tip was transplanted on to a conus, the two structures formed were slugs, both of which migrated, culminated and fruited. The slug formed by the fruiting-body tip was morphologically indistinguishable from any other slug. Its migration was normal, and

6 232 J. RUBIN AND A. ROBERTSON Table 1. Results of transplants of conus tips to fruiting bodies, fruit ing-body tips to conuses, conus tips to conuses, and fruit ing-body tips to fruiting bodies Each set of transplants is compared to a control consisting of a sham operation. In each case, the difference between the control and transplant was highly significant by the x 2 test {P < 10~ 6 ). The table also shows the results of transplanting conus and fruiting-body tips into fields of amoebae scored as their ability to cause directed movement for 30 min, compared with the control transplantation of fruiting-body bases. The difference is significant. (S = success, F = failure as % of total.) Conus F.B. Controls Recipient S F S F S F Conus F.B it fruited as a normal slug would. Several examples of all of the transplants in the above experiments were recorded on film. (ii) A subset of experiments (Table 1) was then selected in order that large numbers of transplants could be performed to determine the proportion of experiments in which there was successful field organization by grafted tips and to see whether there were any quantitative differences between tips. The subset chosen included conus to conus, conus to fruiting body, fruiting body to conus, and fruiting body to fruiting body. As a control, sham operations were performed. In these, all manipulations were performed as in a normal transplant, except that no tip was transferred. x* tests showed that in every case in Table 1 the difference between the experiments and controls was highly significant (P < 10" 6 ). Table 1 shows that tips transplanted in these experiments always had a significant success rate in organizing fields when compared with the controls. It must be remembered that the criterion for success that we adopted gives us a lower limit for organizing ability and that when the experiments were filmed throughout we could always see some evidence of organizing activity by grafted tips, in particular the formation of an indentation between the grafted and host tips. The controls, however, show that interference with a pseudoplasmodium often leads to regulation and the formation of an extra tip, after a delay of at least B. Transplantation of tips into fields of amoebae Tips of conuses, slugs and fruiting bodies were transplanted into fields of interphase amoebae. The tips were placed into fields of amoebae that were 4, 6 and 8 h past refrigeration - that is, 6, 8 and 10 h into interphase if one includes the 2 h incubation period before refrigeration. As controls, fruiting-body bases and the middle portions of slugs were also transplanted into fields. A summary of the results is shown in Table 2.

7 Tip ofdictyostelium pseudoplasmodium as organizer 233 Table 2. Summary of results of transplants into fields of amoebae (F.B. = fruiting body. A dash indicates that the category is not applicable to a particular graft.) Graft F.B. tip Conus tip F.B. base Slug middle No. of experiments Attraction (%) Initial continuous movement (%) Periodic relaying (%) Time to periodic relaying (min) Time to pseudoplasmodium formation (min) Direct culmination (%) (i) Fruit ing-body tips into fields Ninety-three transplants were performed. In 87 of these chemotactic responses persisted for more than 30 min from the time of transplantation. The responses were independent of the age of the field, demonstrating that all fields had achieved both chemotactic and relaying competences. When a tip was placed into an interphase field, amoebae were immediately attracted toward the tip, moving continuously, and in straight lines. No pulsations were observed during this initial period. This implies that the attracting signal was continuous. The initiation of periodic movements occurred from 40 min to 4 h after transplantation. This happened in 26 % (23/87) of those experiments in which tips attracted amoebae for more than 30 min. The average time for the initiation of periodic movements was 87-8 ± 39-5 min. The number of periodic movements associated with each tip was quite variable, ranging up to 12. After the initial pulse, the pulse period quickly decreased to about 10 min (see Fig. 2). Once pulsations began and the field of amoebae showed periodic movements towards the tip, streams formed. The streams either began at the boundary of the tip or up to 50 /im away from the boundary. Thus the tips caused periodic signal relaying, even though their own signal was continuous, and the threshold for signal relaying may be exceeded at up to 50 jum from the tip. Approximately 2\ h after transplantation the tip and the cells it had attracted formed a pseudoplasmodium (Fig. 3). In 96 % of the 48 cases which were observed throughout this period, this was a conus which then transformed into a slug. In two cases out of 48, however, a fruiting body formed directly, as can happen in normal morphogenesis (Newell, Telser & Sussman, 1969).

8 234 J. RUBIN AND A. ROBERTSON 160 i J I Pulse number Fig. 2. Periods of waves propagated from fruiting-body tips transplanted into fields of amoebae , Hours r-l Fig. 3. Histogram of times to pseudoplasmodium formation after transplantation of fruiting-body tips into fields of amoebae. (ii) Conus tips into fields Eighty-four per cent (78/93) of conus tips attracted cells. The amoebae first responded with continuous movement toward the tip and occasionally (5/78) with later pulsations and stream formation. Twenty-six per cent (20/78) of the transplants culminated without slug formation; the difference in times to 10

9 Tip ofdictyostelium pseudoplasmodium as organizer Hours Fig. 4. Histogram of times for pseudoplasmodium formation after transplantation of conus tips into fields of amoebae LI Hlini in m m Interval length (min) Fig. 5. Periods of waves propagated from slug middles transplanted into fields of amoebae. culmination between the conus tips and the fruiting-body tips was significant (P < 0-05). The mean time to slug formation was 87 min (Fig. 4) - much shorter than that for fruiting-body tips (P < 0-05). (iii) Base of fruiting bodies To check that the tip behaves differently from an inert object placed into the field of interphase amoebae, the ability of bases of fruiting bodies to attract cells was investigated. Using the same criterion of continuous directed movement toward the object for 30 min only 5 cases out of 61 base transplants were

10 236 J. RUBIN AND A. ROBERTSON Hours Fig. 6. Histogram of times for pseudoplasmodium formation after transplantation of slug centres into fields of amoebae. successful. In all of these five cases, an independent autonomous centre had formed close to the base. This is close to the proportion ofconus tip transplants, but significantly (P <^ 0-001) smaller than that of fruiting-body tip transplants, showing periodic signalling. We therefore conclude that fruiting-body tips tend to cause periodic signal relaying, but that conus tips do not. This implies that conus tips secrete less attractant than do fruiting-body tips, as the threshold for signal relaying is higher than that for chemotaxis (see Introduction). (iv) Mid-portion of slugs We placed 41 mid-portions of slugs into fields of amoebae 8 h into interphase. The mid-portions were approximately 100 jumm diameter. All attracted amoebae periodically. The first propagated waves occurred as early as 10 min from the beginning of filming - that is, approximately 20 min after transplantation. The distribution of recorded periods is shown in Fig. 5. After a mean time of 4 h 50 min (Fig. 6) all slug centres had regulated to produce conuses which then proceeded through the normal developmental cycle to become fruiting bodies containing some amoebae attracted from the field. The differences in mean times between the production of slugs from fruiting-body tips (P < 0-01) or conus tips (P <t 0-001) and slug middles placed in interphasefields are highly significant. They are partially accounted for by the extra time required for regulation by the slug middles to produce new tips - about 1 h (Farnsworth, 1973).

11 Tip o/dictyostelium pseudoplasmodium as organizer 237 Fig. 7. Typical result of grafting a fruiting-body tip into an early fruiting body, 1 h after grafting. Two structures are beginning to separate; the indentations demarcating the fields of the two tips are clearly visible. The frame width is 0-5 mm. The grafted tip is on the right. DISCUSSION 1. General remarks Although there has been no complete review, it is now well established in the literature that tips have special properties. We shall mention these briefly and then discuss the relevance and significance of our own results. Raper showed that the tip controls slug migration (Raper, 1940) and that extra tips grafted into the side of a slug would take over cells posterior to the graft site provided that the original polarity of the grafted tips was maintained. He concluded that the tips acted as though they were organizers, controlling cell movement and defining the developmental axis of the slug. Farnsworth has shown that the presence of a tip on a pseudoplasmodium inhibits further tip formation. The inhibition is released within approximately \ h - the tip determination time - when a tip is removed. Both Raper's and Farnsworth's results support the analogy between tips and classical organizers. Several authors have shown that tip cells have histochemical and ultrastructural properties not shared by other cells in the pseudoplasmodium (Bonner,

12 238 J. RUBIN AND A. ROBERTSON Chiquoine & Kolderie, 1955; Hohl & Hamamoto, 1969; Maeda & Takeuchi, 1969; Gregg & Badman, 1970; Miiller & Hohl, 1973; Maeda & Maeda, 1974), and that the cells that first form the tip remain in it until the end of morphogenesis (Takeuchi, 1969; Farnsworth, 1973). The tip forms the anterior 10% or even less, of the pseudoplasmodium. Recently Maeda & Maeda (1974) have found that the anterior (light) cells in D. discoideum also contain up to ten times more c-amp than the posterior (heavy) cells, though their light fraction certainly includes pre-stalk cells posterior to the tip proper. Bonner (1949) had demonstrated that for both slugs and fruiting bodies, the front released more acrasin than the remainder of the pseudoplasmodium. Finally, it has been shown that cells in isolated slug fragments de-differentiate, and then re-differentiate (Gregg, 1965; Sakai, 1973) but only after there has been sufficient time for regulation to produce a new tip. 2. The equivalence of tips In Section A of the results we showed that tips from all pseudoplasmodial stages act as organizers when grafted into all pseudoplasmodial stages and into fields of amoebae before aggregation. The structures organized depend on the developmental stage of cells in the recipient pseudoplasmodium. Therefore, tips from all stages have qualitatively similar organizing abilities, and we can infer that their organizing signals are qualitatively similar. This is analogous to the interpretation by cells of the signal from a classical organizer such as the dorsal lip of the amphibian gastrula (cf. Spemann, 1938) or the hypostome of hydra (Wolpert, Hicklin & Hornbruch, 1971). It appears that organizers in general produce a stage-independent signal which is interpreted by responding cells according to their capacities or state of determination (Wolpert, 1969; Robertson & Cohen, 1972). 3. Quantitative differences between tips Our experimental results in Section B show that conus tips are less likely to initiate signal relaying and periodic cell movement, when grafted into fields of sensitive amoebae, than are fruiting-body tips, although both kinds of tip cause chemotaxis. As the threshold for chemotaxis is lower than that for signal relaying (Robertson & Cohen, 1974) we may conclude that the signal from a conus tip is weaker than that from a fruiting-body tip. These results are therefore consistent with the possibility that tips release c-amp. However, conus tips organize new slugs more quickly than do fruiting-body tips. We do not know why. It is possible that the conus tip secretes slime but that the fruiting-body tip does not. The secretion of slime may expedite slug formation (Shaffer, 1962). 4. The tips' signal The continuous movements of interphase amoebae towards tips show that the tips release continuous signals. In experiments to be published in the next

13 Tip of Dictyostelium pseudoplasmodium as organizer 239 paper in this series we have found that beef heart PDE, when added to the culture medium, inhibits the signal. These results support the postulate that tips release c-amp continuously, which in turn is consistent with the observed high c-amp concentration in the anterior cells of slugs. If the signal is a c-amp gradient then it can exceed a concentration, for fruiting-body tips, of 10~ 7 M, the approximate threshold for relaying, at a distance of at least 50 jum from the tip border. For conus tips the relaying threshold is not exceeded outside the tip, although it may well be within the tip. In pseudoplasmodia cells are joined closely by intercellular contacts. The amount of signal a tip must release to cause relaying in a pseudoplasmodium will therefore be less than that required to cause relaying by separated cells. 5. Pseudoplasmodial organization The tip forms during late aggregation. Durston (1974) has shown that from this time all autonomously signalling regions close to a tip are brought under its control. As the autonomous sources are periodic we can understand this observation if the tip releases a continuous signal that has a range greater than that of the signal of the most stable autonomous pacemaker, that which propagates spiral waves at the refractory period of the field. The continuous signal from the tip is also consistent with Durston's observation that the period of signals propagated from an aggregate decreases monotonically to 2\ min, once the tip has formed. This happens because the tip supplies a signal continuously above the threshold for signal relaying, and because the refractory period for signal relaying decreases monotonically to 2\ min. This property of the tip may be fundamental to its ability to control pseudoplasmodial organization and movement throughout the rest of the developmental cycle. In this model tips dominate because they can initiate signal relaying at a distance and at the refractory period of the medium (Durston, 1973, 1974). The properties of slug mid-portions are illuminating in this respect. The signal from middles placed in interphase fields is always pulsatile although the period is ill-defined. It is consistent with the idea that cells in the middle, which had been entrained by the refractory period signal caused by the slug tip, can now exhibit autonomous signalling as they are no longer under the tip's influence. The signals begin when the inhibition from the tip has disappeared (Farnsworth, 1973). A further possibility is that the signal gradient formed by the tip controls the differentiation of pre-stalk cells to stalk cells, a view supported by Bonner's observation that c-amp can cause stalk cell differentiation (Bonner, 1970). As we have shown that a tip can be produced, by regulation, from any portion of a pseudoplasmodium, it is probable that its special signalling properties depend on a geometrical feature of its organization rather than on an inherent capacity of its constituent cells.

14 240 J. RUBIN AND A. ROBERTSON We are very grateful for the assistance of David Drage and Diane Wonio, who performed some of the first grafting experiments, and to Morrel Cohen and Susan Weiter for their critical reading of the manuscript. This work was supported by grant number GB from the National Science Foundation to the University of Chicago. A.R. is an Alfred P. Sloan Fellow (1973-5). REFERENCES ASHWORTH, J. M. (1971). Cell development in the cellular slime mould, Dictyostelium discoideum. Symp. Soc. exp. Biol. 25, BONNER, J. T. (1949). The demonstration of acrasin in the later stages of development of the slime mold Dictyostelium discoideum. J. exp. Zool. 110, BONNER, J. T. (1967). The Cellular Slime Molds. Princeton University Press. BONNER, J. T., CHIQUOINE, A. D. & KOLDERIE, M. Q. (1955). A histochemical study of differentiation in the cellular slime molds. /. exp. Zool. 130, BONNER, J. T. (1970). Induction of stalk cell differentiation by cyclic AMP in the cellular slime mold Dictyostelium discoideum. Proc. natn. Acad. Sci. U.S.A. 65, CHANG, Y. Y. (1968). Cyclic 3'-5' adenosine monophosphate phosphodiesterase produced by the slime mould Dictyostelium discoideum. Science, N. Y. 160, COHEN, M. H. & ROBERTSON, A. (1971 a). Wave propagation in the early stages of aggregation of cellular slime molds. /. theor. Biol. 31, COHEN, M. H. & ROBERTSON, A. (19716). Chemotaxis and the early stages of aggregation in cellular slime molds. /. theor. Biol. 31, COHEN, M. H. & ROBERTSON, A. (1972). Differentiation for aggregation in the cellular slime molds. In Cell Differentiation (ed. R. Harris, P. Allin and D. Viza), pp Copenhagen: Munksgaard. DURSTON, A. J. (1973). Aggregation fields of Dictyostelium discoideum as excitable media. J. theor. Biol. 42, 3, DURSTON, A. J. (1974). Pacemaker activity during aggregation in Dictyostelium discoideum. Devi Biol. 37, FARNSWORTH, P. (1973). Morphogenesis in the cellular slime mould Dictyostelium discoideum: the formation and regulation of aggregate tips and the specification of developmental axes. /. Embryol. exp. Morph. 29, GERISCH, G. (1968). Cell aggregation and differentiation in Dictyostelium. Curr. Top. in Devi Biol. 3, GREGG, J. H. (1965). Regulation in the cellular slime molds. Devi Biol. 12, GREGG, J. H. & BADMAN, W. S. (1970). Morphogenesis and ultrastructure in Dictyostelium. Devi Biol. 22, HOHL, H. R. & HAMAMOTO, S. T. (1969). Ultrastructure of spore differentiation in Dictyostelium: theprespore vacuole. /. Ultrastruc. Res. 26, KONIJN, T. M. (1972). Cyclic AMP and cell aggregation in the cellular slime molds. Acta Protozool. 11, KONIJN, T. M., BARKLEY, D. S., CHANG, Y. Y. & BONNER, J. T. (1968). Cyclic AMP: a naturally occurring acrasin in the cellular slime molds. Am. Nat. 102, 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, 3. MAEDA, Y. & MAEDA, M. (1974). Heterogeneity of the cell population of the cellular slime mold Dictyostelium discoideum before aggregation and its relation to the subsequent location of the cells. Expl Cell Res. 84, MAEDA, Y. & TAKEUCHI, I. (1969). Cell differentiation and fine structures in development of cellular slime molds. Development, Growth & Differentiation 11 (3), MiiLLER, N. & HOHL, H. R. (1973). Pattern formation in Dictyostelium discoideum, temporal and spatial distribution of prespore vacuoles. Differentiation 1, NEWELL, P. C, TELSER, A. & SUSSMAN, M. (1969). Alternative developmental pathways determined by environmental conditions in the cell slime mold. /. Bad. 100,

15 Tip of Dictyostelium pseudoplasmodium as organizer 241 RAPER, K. B. (1940). Pseudoplasmodium formation and organization in Dictyostelium discoideum. J. Elisha Mitchell scient. Soc. 56, ROBERTSON, A. (1974). Information handling at the cellular level: intercellular communication in slime mould development. In The Biology of Brains. London: Institute of Biology. ROBERTSON, A. & COHEN, M. H. (1972). Control of developing fields. A. Rev. Biophys. Bioeng. 1, ROBERTSON, A. & COHEN, M. H. (1974). Quantitative analysis of the development of the cellular slime molds: II. Some Mathematical Questions in Biology 6 (in the Press). ROBERTSON, A., DRAGE, D. J. & COHEN, M. H. (1972). Control of aggregation in Dictyostelium discoideum by an external periodic pulse of cyclic adenosine monophosphate. Science, N. Y. 175, SAKAI, Y. (1973). Cell type conversion in isolated pre-stalk and pre-spore fragments of the cellular slime mold Dictyostelium discoideum. Development, Growth & Differentiation 15, SHAFFER, B. M. (1962). The Acrasina, Part I. Adv. Morphogen. 2, SPEMANN, H. (1938). Embryonic Development and Induction. Yale University Press. (Reprinted in 1967 by Harper, New York.) TAKEUCHI, I. (1969). Establishment of polar organization during slime mold development. In Nucleic Acid Metabolism, Cell Differentiation and Cancer Growth (ed. E. V. Cowdrey & S. Seno), pp Oxford: Pergamon Press. WOLPERT, L. (1969). Positional information and the spatial pattern of cellular differentiation. /. theor. Biol. 25, WOLPERT, L., HICKLIN, J. & HORNBRUCH, A. (1971). Positional information and pattern regulation in regeneration of Hydra. Symp. Soc. exp. Biol. 25, {Received 2 July 1974, revised 5 August 1974) 16 i M B 33