Cell cycles in the sea urchin embryo The early cell division cycles of ...

J. Cell Sci. Suppl. 12, 129-144 (1989) Printed in Great Britain © The Company of Biologists Limited 1989

129

Calcium-induced chromatin condensation and cyclin phosphorylation during chromatin condensation cycles in ammonia-activated sea urchin eggs R A J N I K A N T P A T E L 1, J E R E M Y T W I G G 1, IA N C R O S S L E Y 1, R O Y G O L S T E Y N 2 a n d M IC H A E L W H I T A K E R 1

1Department of Physiology, University College London, Gower Street, London WC1E 6BT, UK 2Department of Biochemistry, University of Cambridge, Tennis Court Road, Cambridge, UK

Summary Ammonia-activated sea urchin eggs undergo repeated cycles of D N A synthesis, nuclear envelope breakdown (NEB) and chromatin condensation. No mitotic spindle forms, nor do the eggs undergo cytokinesis. Ammonia-activated eggs exhibit a form of the cell cycle in which the nuclear cycle proceeds without segregation of the chromatin into daughter cells. We discuss here experiments that demonstrate that intracellular free calcium concentration controls the S phase-M phase transition in ammonia-activated eggs, as it does in fertilized embryos. Cyclins are proteins that are synthesized throughout the cell cycle and destroyed abruptly during each round of chromatin condensation. We find that cycles of cyclin phosphorylation and destruction occur coincident with chromatin condensation in ammonia-activated eggs. Cyclin phosphorylation also occurs in eggs treated with the tumour promoter, phorbol myristate acetate (PMA). There is no accompanying NEB or chromatin condensation, however, and the nucleus is insensitive to exogenously-generated calcium transients. These latter data indicate that cyclin synthesis and phosphorylation is not a sufficient condition for calcium-induced NEB in sea urchin embryos. PMA must fail to induce one of the necessary cell cycle initiation signals. We suggest that the missing signal is the activation of the cell cycle control protein p34“ /i2, which we have shown to be phosphorylated at fertilization and which is phosphorylated in ammonia-activated eggs.

Introduction

Cell cycles in the sea urchin embryo The early cell division cycles of sea urchin embryos are rapid and synchronous. The unfertilized egg is arrested in interphase. The first cell cycle after fertilization has a short Gi phase, but subsequent cell cycles proceed as rapidly alternating rounds of S phase and M phase until the mid-blastula stage. During these early cell cycles there is no significant increase in mass: these cells are specialized to increase in number and rapidly duplicate and segregate the zygote genome. The nuclear cycle is driven by the cytoplasm, that is, by translational and post-translational control mechanisms (Patel et al. 1989). Cell size-related cell cycle control is absent. The early sea urchin embryo offers a cell cycle pared down to its most basic features. It is a good place to start when asking the question: what are the minimal essential requirements for a competent cell cycle? Key words: calcium, cyclin, chromatin condensation.

130

R. Patel et al.

Ammonia activation of sea urchin eggs Treatment of unfertilized sea urchin eggs with sea water containing ammonium ions is one of several methods of parthenogenetically activating sea urchin eggs (Loeb, 1919; Harvey, 1956). Ammonia activation stimulates protein and DNA synthesis. DNA synthesis alternates with rounds of chromatin condensation (Mazia, 1974; Mazia and Ruby, 1974). While it has not been shown that this DNA synthesis is semi-conservative, it is clear that the quantity of [3H]thymidine doubles with each well-defined round of synthesis; this is accompanied by successive doublings of chromosomes. Ammonia-activated eggs do not undergo mitosis, nor do they cleave, presumably because they lack the centrosome normally donated by the sperm at fertilization (Brandriff et al. 1975). They represent an even simpler example of the nuclear division cycle than fertilized eggs, in that mitosis and cytokinesis are absent. We can compare the results of experiments in fertilized and ammonia-activated eggs to determine which cell cycle controls are essential for the nuclear cycle. There are two crucial cytoplasmic signals at fertilization that are responsible for triggering the initiation of the cell cycle. The first is a transient increase in intracellular free calcium concentration ([Ca2+];) and the second a sustained increase in cytoplasmic pH (pHi). These two signals are each generated by one of the two arms of the polyphosphoinositide messenger system (Whitaker and Steinhardt, 1985; Whitaker, 1989). Ammonia is a weak base and causes an increase in pHi (Shen and Steinhardt, 1978; Winkler and Grainger, 1978). Ammonia activation was used very successfully as a tool to dissect out which of the responses at fertilization was due to the pHi signal, as opposed to the increase in [Ca2+]; (Epel, 1978; Whitaker and Steinhardt, 1985). However, treatment of eggs with ammonia clearly causes something more to happen than a mere change in pH (Whitaker and Steinhardt, 1981; Whitaker and Steinhardt, 1985; Dube et al. 1985): it can induce a calcium transient, for example (Zucker et al. 1978), though this ammonia-induced transient is rather different from the sperm-induced transient in its properties. We shall return to this point in discussing the effects of phorbol myristate acetate, a tumour promoter that mimics some of the effects of stimulating the diacylglycerol-mediated arm of the phosphoinositide messenger system. [Ca2+]i is a cell cycle messenger in sea urchin eggs It has been clear for some time that the increase in [Ca2+]j at fertilization is responsible for initiating the cell cycle in sea urchin eggs. It now seems that a transient increase in [Ca2+]; is essential for progress through the cell cycle at the time of nuclear envelope breakdown (NEB) and chromatin condensation. Measurements of [Ca2+]; using the fluorescent indicator dye fura-2 have shown that [Ca2+], increases transiently at NEB (Poenie et al. 1985; Patel et al. 1989). The increase in [Ca2+]; at NEB can be blocked by using the calcium chelators l,2-bis(2-aminophenoxy)ethane-iV,Ar,Ar',iV'-tetraacetic acid (BAPTA) and EGTA: blocking the in­ crease in [Ca2+]j prevents NEB (Steinhardt and Alderton, 1988; Twiggetal. 1988). NEB and chromatin condensation can be induced precociously by microinjection of

131 Calcium and cyclin in ammonia-activated embryos calcium buffers or the calcium-mobilizing phosphoinositide messenger inositol trisphosphate (Insf^). This is good evidence that the increase of [Ca2+]; at NEB is the trigger that stimulates the onset of mitosis. The immediate target of the calcium transient at NEB is most probably a cell cycle control protein, because preventing protein synthesis after fertilization prevents NEB (Wagenaar, 1983; Twigg et al. 1988). Moreover, the nucleus does not respond to calcium or InsP3 microinjections under these conditions (Twigg et al. 1988; Patel et al. 1989). Cell cycle control proteins Cell cycle regulation appears to be exerted at specific control points (Mitchison, 1971). One of these is situated at the G2/M boundary, just prior to NEB and chromatin condensation. It is thought that regulation of the cell cycle at these control points is brought about by phosphoproteins that are in turn regulated by specific kinases and phosphatases. A phosphorylation/dephosphorylation cascade of this type might seem a daunting thing to analyse and understand. Fortunately, two of the key proteins in this cascade have been reasonably well characterized. The two pivotal proteins are cyclin and p34fA2. These two proteins comprise an activity known as maturation- or mitosis-promoting factor (MPF). MPF is an activity that appears in frog and starfish oocytes as they undergo germinal vesicle breakdown (the meiotic counterpart of NEB [Masui and Markert, 1971; Newport and Kirschner, 1984; Kishimoto and Kanatani, 1976; Kishimoto and Kondo, 1986]). It is also found in mammalian cell cytoplasm when the cell is undergoing mitosis (Sunkara et al. 1979) and in yeast (Kishimoto et al. 1982). It is a cytoplasmic factor that appears at the time of NEB and chromatin condensation and drives the nucleus into mitosis. Cyclins are a family of proteins that show a characteristic pattern of accumulation and destruction in each cell cycle. They were first identified in sea urchin embryos and surf clam and starfish oocytes (Evans et al. 1983; Swenson et al. 1986; Standart et al. 1987) and have since been identified in frog eggs (Minshull et al. 1989) and in yeast (Solomon et al. 1988; Goebl and Byers, 1988). They are synthesized continuously during the cell cycle and are rapidly destroyed once in each cell cycle during mitosis (Evans et al. 1983). One of the two proteins of MPF is a cyclin (Dorée et al. 1989 and Mailer et al. 1989, [both this volume] Draetta et al. 1989; Meijer et al. 1989). p 3 4 «fc2 -g a prote[n identified by analysis of a cell cycle-deficient yeast mutant (Nurse and Bisset, 1981). It is also found in frogs (Dunphy ei al. 1988; Gautier et al. 1988), starfish (Arion et al. 1988; Labbé et al. 1988), sea urchins (Patel et al. 1989) and mammalian cells (Lee and Nurse, 1987). It is the other protein component of MPF (Gautier et al. 1988; Dunphy et al. 1988). It has an intrinsic kinase activity (Simanis and Nurse, 1986), binds to and phosphorylates a protein that may well be a cyclin (Draetta and Beach, 1988; Arion et al. 1988) and is itself phosphorylated in a cell cycle-dependent manner (Simanis and Nurse, 1986; Lee et al. 1988; Draetta et al. 1988). It forms a complex with cyclin (Dunphy et al. 1988; Gautier et al. 1988;

R. Patel et al. Dorée, 1989 [this volume]; Draetta et al. 1989; Meijer et al. 1989); when activated by a cyclin, p34rA'2 is an active histone kinase (Arion et al. 1988; Draetta and Beach, 1988). High levels of histone kinase activity are found in close temporal correlation with mitosis (Meijer and Pondaven, 1988; Meijer et al. 1989). 132

Interactions between calcium, cyclin and p34cdc2 Calcium, cyclin and p34cdc2 must cooperate in some way to bring about NEB and chromatin condensation during each cell cycle in early sea urchin embryos. We describe below some experiments on ammonia-activated sea urchin eggs that shed some light on the relative contribution of each to the onset of mitosis. Results and discussion

Chromatin condensation cycles in ammonia-activated eggs Unfertilized eggs respond to treatment with ammonium chloride and undergo cycles of DNA synthesis and chromatin condensation (Mazia, 1974; Mazia and Ruby, 1974). The chromatin condensation cycles can be followed in the fluorescence microscope by using the vital DNA stain Hoechst 33342. Fig. 1 compares the timing of chromatin condensation cycles in fertilized and ammonia-activated eggs. The cycles in ammonia-activated eggs are markedly slower than in fertilized eggs, but the egg population maintains a good degree of synchrony over two or more chromatin condensation cycles. We wanted to determine whether calcium controlled the chromatin condensation cycle in ammonia activated eggs. In three sets of exper­ iments, we (1) measured [Ca2+]; using the calcium indicator dye fura-2, (2) blocked any increase in [Ca2+]; with the calcium chelator BAPTA and (3) injected the calcium-releasing messenger InsP 3 to determine whether there was a causal relation between the [Ca2+]i transient and NEB and chromatin condensation.

Time (min)

Fig. 1. Eggs were incubated in 20 [am Hoechst 33342 (a vital DNA stain) and fertilized (O------O) or tre a te d w ith 15 m M -am m o nium c h lo rid e-c o n tain in g sea w a te r ( • --------• ) for IS min. Chromatin condensation was assesssed in the fluorescence microscope (exci­ tation: 340-380 nm; emission: 450-490 nm). Lytechinus pictus, 16°C.

Calcium and cyclin in ammonia-activated embryos

133

Fig. 2. [Ca2+]j was measured in a single egg after microinjection of the calcium indicator dye fura-2 (50 flm: for methods, see Swann and Whitaker, 1986). Addition of ammonium chloride (15 mM in artificial sea water at pH 9 for 15 min) caused a transient increase in [Caz+]j to 170 nM. 46 min after addition of ammonium chloride, a second, larger increase in [Ca2+]; was detected. NEB occurred 3 min after the onset of the second [Ca2+]; transient. L. pictus, 16°C.

A transient increase in [Ca2+]i occurs at NEB in ammonia-treated eggs In all of five experiments, a calcium transient that slightly preceded NEB and chromatin condensation was measured in fura-2-injected, ammonia-treated eggs. A typical experiment is shown in Fig. 2. The transient had a mean amplitude of 0 .8 4 ± 0 .0 8 , um (mean and s . e . m .) and a duration of 8± lm in. NEB occurred 5 -1 0 min after the onset of the transient. These experiments indicate that a calcium transient immediately precedes NEB in ammonia-activated eggs, just as it does in fertilized eggs (Poenie et al. 1985; Patel et al. 1989). The calcium chelator BAPTA blocks NEB when microinjected into ammoniaactivated eggs NEB in fertilized eggs is prevented by microinjection of the calcium chelator BAPTA to a final concentration of 2.3 mM (Twigg et al. 1988). We injected this concentration of BAPTA into two batches of ammonia-activated eggs and found that it completely prevented NEB in all of 20 eggs. BAPTA injection does not substantially alter the resting [Ca2+]; in eggs, but prevents the [Ca2+]j transient (not shown). This suggests that it is the calcium transient itself, rather than a permissive, resting level of [Ca2+]j that is required for NEB. Microinjection of InsPj causes precocious chromatin condensation in ammoniaactivated eggs The phosphoinositide messenger InsP 3 releases calcium from intracellular stores in sea urchin eggs (Whitaker and Irvine, 1984) and causes precocious NEB and chromatin condensation when injected into fertilized sea urchin eggs (Twigg et al. 1988). Fig. 3 shows two experiments in which InsP 3 was microinjected into

134

R. Patel et al.

Time after NH4C1 addition (min)

Fig. 3. InsP 3 was microinjected (pipette concentration, 100 final concentration 0.5—1 ,mm) into ammonia-activated eggs around the time of NEB. Three experiments on different egg batches are shown. 50 % NEB has been normalized to 65 min after ammonia addition. The time course of NEB in the three ammonia-treated egg batches is shown (,A,0). 90% (8/9) of InsP 3 injections (A) led to precocious NEB when injected 55 min after ammonia activation, 40 % (2/ 5) gave precocious NEB 50 min after activation and one injection at 48 min was ineffective. Similar concentrations of InsP3 co-injected with the calcium chelator EGTA (final concentration 0.5-1 m M ) were ineffective (V ). These data indicate that the nucleus becomes sensitive to calcium at 55 min after ammonia activation, 10min before the egg population undergoes spontaneous NEB. L. pictus, 16 °C.

ammonia-activated eggs. Eggs injected with InsP3 rapidly underwent NEB and chromatin condensation in advance of uninjected controls. Co-injection of the calcium chelator EGTA prevented InsP3-induced NEB, indicating that InsP3induced calcium release (Twigg et al. 1988) caused the premature initiation of the mitotic phase of the nuclear cycle. Onset of sensitivity to InsFj injection In fertilized eggs, the nucleus is initially insensitive to microinjection of calcium or Ins-P3 (Twigg et al. 1988). Indeed, this is to be expected, since the nucleus shows no sign of NEB or chromatin condensation during the calcium transient at fertilization. A process occurs that sensitizes the nucleus to increases in [Ca2+]j. In fertilized eggs, the nucleus is sensitized at 45-50 min after fertilization, some 20 min before NEB would normally occur. The sensitizing factor appears to be a protein synthesized after fertilization. We think this for two reasons: first, because eggs treated with the protein synthesis inhibitor emetine do not undergo NEB and are insensitive to calcium and InsP3 injections; second, because the nucleus becomes sensitized to calcium at just the time at which the cell cycle (in this context, subsequent NEB) ceases to be sensitive to emetine. In fertilized eggs, a sufficient amount of the sensitizing protein is synthesized by 50 min after fertilization. The situation in ammonia-activated eggs is similar. Fig. 3 shows that the nucleus becomes sensitized to InsP3 injection 50-55 min after ammonia activation,

Calcium and cyclin in ammonia-activated embryos

135

Time (min)

Fig. 4. The protein synthesis inhibitor emetine (100 fm) was added to eggs 45 or 65 min (V) after ammonia activation. The former treatment prevented NEB ( • ------• ) while the latter treatment (A ------A) was without effect, NEB occurring at the same time as in controls (O------O). The experiments indicate that continued protein synthesis is necessary until 20-30min before NEB. L. pictus, 16°C.

10-15 min before NEB. Fig. 4 shows an experiment in which the sensitivity of NEB to the protein synthesis inhibitor was tested. Emetine added 45 min after activation (35 min before NEB in controls) blocks NEB; emetine added 20min later does not. These experiments suggest that a protein is synthesized in ammonia-activated eggs that is required to sensitize the nucleus to calcium, and that adequate levels of this protein are made only 10-15 min before NEB occurs, as is the case in fertilized eggs. Cyclin synthesis in ammonia-activated eggs The above experiments suggest that the nuclear cycle in ammonia-activated eggs resembles that in fertilized eggs quite closely and that calcium regulates the onset of the mitotic phase in ammonia-activated eggs as it does in fertilized eggs. It has been reported, however, that in another important respect there was a difference between ammonia-activated and fertilized eggs: the cell cycle protein cyclin did not appear to undergo rounds of destruction in ammonia-activated eggs (Evans et al. 1983). This seemed to us an anomaly, since cyclin was clearly a component of the regulation of mitosis onset in other cells; it might suggest, for example, that cyclin was involved in regulating spindle formation rather than the nuclear events of NEB and chromatin condensation. We therefore looked for cyclin synthesis in ammonia-activated eggs. We found [3oS] methionine incorporation into a 47K (K = 1 0 3Mr) protein that disappeared during the mitotic phase of the chromosome condensation cycle. This protein comigrated on SDS-polyacrylamide gels with the 47K cyclin found in fertilized eggs. Well-synchronized ammonia-activated eggs contain a cyclin that is destroyed periodically during the mitotic phase of the cell cycle, just as in fertilized eggs (Fig. 5).

R. Patel et al.

136 3

4

7

5 1 Bill tBT

8

9 a

10

11

n

I

P -cyclin -

cyclin-

—

—

_ .

. . . .

i

12

13

1WT

f f *

4M»

**< *

14 *

_ •* -

R ibonucleotide, reductase

45

60

90

75

Tim e after fertilization (min)

1

2

3

4

P-cyclin -► cyclin -►

5

6

7

8

9 10

11 12 13

14 15

135

165

16

i 45

75

105

Tim e after NH4CI activation (min)

Fig. S. The appearance and disappearance of cyclin in fertilized and ammonia-activated eggs. Eggs were incubated with [ S] methionine and run on polyacrylamide gels. Upper panel: the pattern of protein synthesis after fertilization. P-cyclin, cyclin and the small subunit of ribonucleotide reductase are indicated. P-cyclin and cyclin decrease during mitosis (70min), while, in contrast, ribonucleotide reductase labelling increases steadily throughout the experiment. Lower panel: the pattern of protein synthesis after ammonia activation during two rounds of the nuclear cycle. P-cyclin begins to predominate over cyclin by 75 min post-activation. 100% NEB occurred at 85 min. P-cyclin disappears and cyclin re-appears by 95 min. P-cyclin increases again in lanes 15 & 16 (*) at 175 min. By 185 min, a second round of NEB had occurred in 50 % of the eggs.

Cyclin phosphorylation during the cell cycle We noticed a band on polyacrylamide gels of [3;>S] methionine-labelled proteins in ammonia-activated eggs that ran at a molecular weight 3—4K greater than cyclin and that increased markedly in intensity at the time of NEB and chromatin condensation. This band increased as the cyclin band decreased before itself decreasing. The higher molecular weight protein could be immunoprecipitated with an anti-cyclin antibody and could be converted to a band that comigrated with cyclin by treatment with acid phosphatase (R. Golsteyn, unpublished experiments). This higher molecular weight band therefore represents a phosphorylated form of cyclin that we shall call P-cyclin. It is also found and behaves similarly in fertilized eggs (Patel et al. 1989). The pattern of synthesis, phosphorylation and destruction of cyclin is shown in Fig. 6. Cyclin and P-cyclin accumulate [35S]methionine in approximately equal amounts until the time of nuclear envelope breakdown, when cyclin falls and Pcyclin rises in intensity. Ten minutes later, the intensity of the P-cyclin band decreases, without any corresponding decrease in the cyclin band. A kinetic analysis suggests that cyclin and P-cyclin exist in equilibrium (not steady state) in an

Calcium and cyclin in ammonia-activated embryos

137

Time (min)

Fig. 6. Eggs were activated with ammonium chloride and incubated with [33S]methionine (30-45/.tCi ml-1). Proteins were separated by electrophoresis on 12.5% SDS-polyacrylamide gels and autoradiographed using Hyperfilm-betamax (Amersham). The bands corresponding to cyclin and P-cyclin were cut out, dissolved in NCS tissue solubilizer and quantitated using liquid scintillation counting. The experiment indicates that cyclin ( • ------• ) and P-cyclin (O------O) accumulate in approximately equal amounts until NEB, when a rapid net conversion of cyclin to P-cyclin occurs. After NEB, P-cyclin levels fall. L. pictus, 16°C.

approximately one-to-one ratio until the time of NEB, when an abrupt increase in the ratio of P-cyclin to cyclin occurs. The increase may be due to stimulated cyclin kinase activity or, equally, to a decreased P-cyclin phosphatase activity, since cyclin and Pcyclin appear to be at equilibrium before the rapid shift takes place. Subsequently the P-cyclin is destroyed. These experiments (and others in fertilized eggs: Patel et al. 1989) suggest that the disappearance of cyclin during the mitotic episode of the cell cycle occurs in two phases. In the first phase, there is a net phosphorylation of cyclin; in the second phase, phosphorylated cyclin is destroyed. It is clear that net cyclin phosphorylation accompanies the cell cycle transition from S phase to M phase but the temporal resolution of the measurements is such that one cannot discern whether net cyclin phosphorylation precedes, accompanies or follows NEB and chromatin conden­ sation. However, since the calcium transient precedes NEB by 5-10 min, it seems safe to conclude that the increase in [Ca2+]i precedes net cyclin phosphorylation, and so may conceivably cause it. Calcium and cyclin as a bistable cell cycle oscillator? The sawtooth oscillations in cyclin levels suggest cyclin as one component of a cell cycle oscillator (see Murray, 1989, this volume). High levels of cyclin may induce NEB and chromatin condensation; low levels may favour chromatin decondensation and S phase. One obvious possibility is that the [Ca2+]j transient at NEB may form the other component of this oscillatory mechanism. On this hypothesis, a critical cytoplasmic concentration of cyclin would trigger the [Ca2+]i transient and the

R. Patel et al.

138 2.0

5 3.

1-5

e

Ia 10 o CL)

^ 0.5 0.0

0

30 60 90 120 Time after NH4C1 addition (min) Fig. 7. [Ca2+]j recorded from a single egg microinjected with fura-2 (50/am). The egg was activated by ammonia (15 mM-NH4Cl, pH 9.0) and treated with the protein synthesis inhibitor emetine (100 ^¿m). Despite the absence of protein synthesis, a calcium transient was detected 54 min after ammonia activation. The nuclear envelope did not break down. One of two experiments. L. pictus, 16°C.

[Ca2+]i transient would cause net cyclin phosphorylation, NEB and, ultimately, the

destruction of P-cyclin, until further accumulation of cyclin by continuous cyclin synthesis would cause the cycle to repeat itself. There is evidence that cyclin synthesis is a necessary condition for the nuclear cycle (Murray, 1989, this volume; Minshull et al. 1989). The hypothesis (originally suggested by indirect evidence: Twigg et al. 1988) predicts that preventing cyclin synthesis should prevent the [Ca2+]i transient that accompanies NEB. It does not (Fig. 7); ammonia-activated eggs treated with the protein synthesis inhibitor emetine to block cyclin synthesis still show an increase in [Ca2+]; at the time at which NEB would normally occur. This experiment indicates that the NEB calcium transient is under the control of a timing mechanism that is independent of cyclin. It does not exclude the possibility that net cyclin phosphorylation is caused by calcium. The finding that a premature increase in [Ca2+]j, caused for example, by microinjection of InsP 3 , leads to precocious NEB and chromatin condensation suggests that calcium, not cyclin, is the cell cycle pacemaker at NEB. Do increased levels of cyclin or P-cyclin alone cause NEB? Injection of surf clam cyclin mRNA into Xenopus oocytes triggers germinal vesicle breakdown (GVBD) (Swenson et al. 1986) and exogenous frog cyclin mRNA can drive cycles of chromatin condensation and decondensation in cell-free extracts made from activated Xenopus eggs (Murray and Ivirschner, 1989). These observations imply that cyclin is necessary for stimulating NEB and that cyclin may be the only newly synthesized protein required at NEB. Protein synthesis can be stimulated in unfertilized sea urchin eggs by treatment with phorbol myristate acetate (PMA), a tumour promoter (Swann and Whitaker, 1985). Fig. 8 shows that PMA stimulates synthesis of cyclin. It is also evident that cyclin is phosphorylated to P-cyclin. The

Calcium and cyclin in ammonia-activated embryos

139

Fig. 8. Eggs were treated with PMA (250 nM) and incubated with [35S]methionine. Cyclin and P-cyclin levels were quantitated as described in the legend to Fig. 5. Treatment of eggs with PMA leads to a constant rate of accumulation of both cyclin (9------• ) and P-cyclin (O------O) in approximately equal amounts. The rapid phase of net conversion of cyclin to P-cyclin associated with NEB and seen in fertilized and ammonia-activated eggs does not occur. Nor does NEB. L. pictus, 16°C.

two exist in the one-to-one ratio found in fertilized and ammonia-activated eggs prior to NEB, but cyclin and P-cyclin accumulate together over a period of several hours to levels higher than those found in the former experiment: there is no phase of rapid net conversion of cyclin to P-cyclin. PMA-treated eggs do not undergo NEB when observed over periods of 4—6h. It seems then that, though both cyclin and the [Ca2+]; transient are necessary for NEB and chromatin condensation, neither alone is sufficient. Are they sufficient in combination? We microinjected InsP 3 into eggs treated with PMA to test this. The results of these experiments are shown in Fig. 9A. InsP 3 injection does not lead to NEB in PMA-treated eggs. Nor does the [Ca2+]; transient that occurs when PMAtreated eggs are fertilized (Fig. 9B). The combination of a [Ca2+]; transient and high cyclin levels are not a sufficient condition for NEB. Initiation of the cell cycle and p34cdc2 phosphorylation We imagine that the missing element that prevents [Ca2+] ¡-induced NEB in PMAactivated eggs is an active form of p34cdc2. Unfertilized and fertilized sea urchin eggs contain comparable amounts of p34rrf PMA

- V