Nucieoside diphosphokinase, an enzyme with step changes in activity ...

Nucieoside diphosphokinase, an enzyme with step changes in activity during the cell cycle of the fission yeast Schizosaccharomyces pombe II. Dissociation of the steps from the DNA-division cycle after induction synchronization

J. CREANOR and J. M. MITCHISON* Department of Zoology, University of Edinburgh, West Mains Road, Edinburgh EH9 3JT, Scotland •Author for correspondence

Summary Synchrony was induced in cultures of the mitotic mutant cdc2.33 of Schizosaccharomyces pombe by shifting up an asynchronous culture to the restrictive temperature for a period of 3-5-4*5h and then shifting down to the permissive temperature. The resulting synchronous divisions had short cycle times, down to 50% of the normal cycle. The oscillatory control of nucieoside diphosphokinase activity was also synchronized by the shift-down and the activity rose in a step pattern. Unlike the situation in the normal cycle, this step pattern was dissociated from the shortened cell cycle and had a

longer period and different phase relations. It may be that the normal entrainment or coupling between the cell cycle and the activity control fails if the cell cycle is too short. The period of the activity control (equal to the protein doubling time at the restrictive temperature) appears to be temperaturecompensated.

Introduction

the disadvantage in many cases of distorting the relations between cell growth and division. It has been widely used in a variety of cellular systems (Mitchison, 1971) and the treatment has often been to impose a block on passage through the DNA-division cycle for a period and then to release the block. The rationale is that all the cells are eventually held at the block point and then divide synchronously when they are released. In the case of S. pombe, the block can be imposed either by treatment with a chemical inhibitor (Mitchison & Creanor, 1971) or by raising a cdc mutant to the restrictive temperature (Benitez et al. 1980; King & Hyams, 1982). We have used the latter method in these experiments, since it gives better synchrony. In both yeast and other cellular systems, cell growth continues during the period of the block so the cells are larger than usual at the time of release and also vary in size. It is a common feature of this kind of induction synchrony that the cell cycles of these oversize cells are shorter than usual and only lengthen to the normal cycle time as the cells resume their normal size at division. An explanation of this, at any rate in S. pombe, is that the normal size control for mitosis is cryptic in the oversize cells and their cycle is controlled by a requirement to spend a minimum time in traversing G2 (Fantes, 1977, 1984).

In the first of these two papers, we showed that the activity of nucieoside diphosphokinase (NDPK) showed a step pattern during the cell cycle of Schizosaccharomyces pombe with a sharp doubling at the beginning of the cycle in selection synchronized cultures (Creanor & Mitchison, 1986). These activity steps persisted with normal cell cycle timing after a block to the DNAdivision cycle imposed by shifting the mitotic mutant cdc2.32 to the restrictive temperature. Similar results were found with changes in the rate of CO2 production, though the timing of the persistent changes was somewhat less than the normal cycle timing (Novak & Mitchison, 1986). These results together with some earlier ones (reviewed by Mitchison, 1989) show that oscillatory controls of some cell cycle events can persist after the main periodic events of the DNA-division cycle have been stopped. In a further study of the relations between the NDPK control and the DNA-division cycle, we have investigated the situation after induction synchrony. This method of producing synchronous cultures is fundamentally different from selection synchrony, since all the cells of an initially asynchronous culture are forced into synchrony by some treatment. It has the advantage of high yield but Journal of Cell Science 93, 185-189 (1989) Pnnted in Great Britain © The Company of Biologists Limited 1989

Key words: cdc2.33 mutant, cell cycle, enzyme activity, fission yeast, nucieoside diphosphokinase, Schizosaccharomyces pombe.

185

Materials and methods

40

Strains, medium, cell collection and cell counting These have been described (Creanor & Mitchison, 1986). The only mutant used was cdc2.33.

10

Enzyme assays In the majority of the experiments, NDPK activity was measured by the transfer of labelled phosphate from ATP to TDP. In five experiments the activity was measured by the transfer of phosphate from ATP to labelled UDP. Both these assays have been described (Creanor & Mitchison, 1986) as 'Assay 1' and 'Assay 3' and are referred to as such in the present figure legends. In addition, two experiments were carried out using an assay of phosphate transferred from ATP to labelled TDP (referred to as Assay 4). This was similar to Assay 3 except for the following points. (1) Labelled TDP was generated from labelled T T P using hexokinase + glucose. 100 ^tl of a mixture containing 555 kBq of [>wef/ry/-3H]thymidine 5'-triphosphate (l-55TBqmmol~ ), 21 units of hexokinase, l'6mM-glucose, 2-5 mM-MgCl2 and 0-25 mM-KCl, all in Hepes buffer (10 mM at pH7-5), were incubated for 2h at 36°C and the hexokinase was then inactivated by placing in a boiling water bath. (2) To each cell sample were added 15 fji of Hepes buffer (9mM at p H 7 5 ) containing 2 jil of the [ 3 H]TDP mixture made as above, 21 mMATP, 6-6mM-TDP, 22mM-MgCl2, 2-2 mM-KCl, 2mM-mercaptoethanol and 2-5% isoamyl alcohol. (3) The PEI plates were run with 0-75 M-sodium formate. In the figure legends, 1 enzyme unit (e.u.) = lnmol PO 4 transferred min~ cell" . This is an opportunity for correcting an error in the figure legends of an earlier paper (Creanor & Mitchison, 1986): 'e.u.' should be 'e.u.ml" '. Using the definition of e.u. above, the correct numerical values, in e.u. m P 1 are: Fig. 1, 2-9; Fig. 2A, 042; Fig. 2C, 1-2; Fig. 2E, 2-3; Fig. 4A, 1-3; Fig. 4B, 0-90; Fig. 4C, 1-3.

Results The procedure of the main experiments was to take a normal asynchronous culture of the mitotic mutant cdcZ.33 growing at the permissive temperature and in early exponential phase (about 106 cells ml~ ), and shift it up to the restrictive temperature for a period of usually 4h. During this period, there was no cell division and no increase in cell number, apart from a short time after the shift-up. The cells, however, continued to grow and NDPK activity followed an exponential curve of increase with only minor fluctuations around it (Creanor & Mitchison, 1986). At the end of the period, the culture was shifted down to the permissive temperature and samples were taken thereafter for NDPK assay and cell counting. The restrictive temperature was 35 °C and the permissive temperature was 29CC in most of the experiments, although in two early ones it was 25 °C. A shift from 35 °C to 29 °C produced less perturbation in a wildtype control (see below) than the larger shift from 35 °C to25°C. The results from one of the early experiments with a shift down to 25 °C are shown in Fig. 1. It is included both because it was carried out for a longer time (7 h) after the shift down than in most later experiments and because it brings out a point in timing that will be 186

jf. Creanor andj. M. Mitchison

Time (h)

Fig. 1. NDPK activity in induction synchrony of cdc2.33. Asynchronous culture shifted up from the permissive temperature (25 °C) to the restrictive temperature (35 °C) for 4-5 h and then returned to the permissive temperature at time zero. Curve A, NDPK activity; assay 3 (incubation temperature 14-5°C), 10a.u. (arbitrary units) = 2-5 e.u. ml"1. Curve B, cell numbers; 1 a.u. = 1x10* cells ml"1. discussed later. There were three synchronous divisions shown by the steps in the cell number curve. The degree of synchrony decreased at each division and the cycle time increased from llOmin between the mid-points of the fir9t and second steps to 140 min between the midpoints of the second and third steps. This is considerably less than the nonmal cycle time of 220 min at this temperature. The NDPK activity curve also showed two and a half steps, but they were less steep than the number steps and had a different timing. The mid-point of the first activity step was 24 min later than the mid-point of the first number step and the difference between the midpoints of the first and second activity steps was 128 min. The activity steps were therefore dissociated from the number steps both in frequency and in phase. Out of a total of 18 experiments, the enzyme assays in four of them were so variable that no clear pattern could be seen. Of the remaining 14, 12 showed patterns similar to Fig. 1, with one or two NDPK activity steps preceded by plateaux. This pattern occurred with three different permissive temperatures (25, 28 and 29°C) and with the three different assay methods. In the majority of the experiments, the permissive temperature was 29°C, and three of these are illustrated in Fig. 2. Only one cell number curve is shown (Fig. 2D), since there was very little variation between these curves at the same temperature. The activity curves were much more variable in the timing and the extent of the steps. In most of the experiments at 29°C, the measurements were terminated at 5-6 h after the shift down and it was impossible to determine the mid-point of the second activity step, since the activity was still rising at this time. The timing of this step had therefore to be given by its start, which is much less accurate than giving it by the mid-point, especially when there is variability in the assay. The same point applies to the fourth step in cell number in the longer experiments. Optical density is an approximate measure of total biomass. Fig. 2E shows some minor fluctuations

40r

•| 10

Time (h) Fig. 2. NDPK activity in induction synchrony of cdc2.33. Asynchronous cultures shifted up from the permissive temperature (29°C) to the restrictive temperature (35°C) for 4h and then returned to the permissive temperature at time zero. Curve A, NDPK activity; assay 1 (incubation temperature 20°C), 10a.u. (arbitrary units) = 10e.u. ml" ; initial cell number = 1X106 cells ml" 1 . Curve B, NDPK activity; assay 4 (incubation temperature 10°C), 10a. u. = 3-0e.u.ml" 1 ; initial cell number= 1X106 cells ml" 1 . Curve C, NDPK activity; assay 1 (incubation temperature 20°C), la.u. = S-9e.u.ml" 1 . Curve D, cell numbers for curve C; 1 a.u. = 1 X 106 cells ml" 1 . Curve E, optical density (595 nm); initial cell number = 0-7X106cells ml" 1 .

around an exponential line of increase with a doubling time of 184 min. This is the same as the doubling time, in optical density, of a normal asynchronous culture at this temperature. This value is greater than the cycle time, since the cells are becoming smaller at each division. Collected data for the experiments at 29°C are given in Table 1. Although the cell cycle times were lengthening, they had still not reached the cycle time of 174 min in a normal asynchronous culture of this mutant at this temperature. The length of the first three cycles were,

respectively, 50%, 60% and 74% of the normal cycle time. The small standard errors in the cell number steps show that induction synchronization is highly reproducible. In contrast, the timing of the activity steps was much more variable with standard errors five to ten times greater. It could be that a late first step was followed by an equally late second step. If so, the standard error of the time between the steps in a series of experiments should be small. This method of presenting the data, in contrast to the timing map in the preceding paper, shows that the standard error is large. So not only is the position of the activity steps variable but also the difference between them. Much, but not perhaps all, of this variability comes from variability in the activity assays. The average positon of the first step (at 143-9min) is between the first and second number step and is at 0-63 of this cycle. Similar shift-down experiments were done with wildtype cells, as a control for the effect of the temperature shift. Fig. 3 shows that a shift from 35 °C to 29°C produced only small fluctuations round a curve of exponential increase in both number and activity. The fluctuations were somewhat larger when the permissive temperature was 25 °C rather than 29°C. Discussion We showed in an earlier paper (Creanor & Mitchison, 1986) that steps in NDPK activity, under the control of what can be loosely called an 'oscillator', persist after a DNA-division block in a selection synchronized culture. This dissociation of the activity steps from the events of the DNA-division cycle is also shown in the present paper. In the normal cycle, the activity step is, on average, at 0-1 of the cycle, whereas after induction synchrony the first activity step is at 0-63 of the shortened cycle. More important, the period between the first two activity steps is considerably greater than the shortened cell cycles. So it is clear that under these conditions there must be separate timing controls for the activity steps and for cell division.

Table 1. Timing of steps after induction synchronization in min after shift-down of cdc2.JJ from 35 °C to 29° C Mean

Standard error of mean

Cell numbers Time of 1st step (MP) Time of 2nd step (MP) Time of 3rd step (MP) Time of 4th step (SS)

88-9 176-4 279-9 379-5

0-8 1-8 2-8 3-5

12 12 9 4

Cell cycle times Between 1st and 2nd step (MPs) Between 2nd and 3rd step (MPs) Between 3rd and 4th step (SSs)

87-8 104-9 128-3

1-7 1-9 3-9

12 9 4

NDPK activity Time of 1st step (SS) Time of 1st step (MP) Time of 2nd step (SS) Time between 1st and 2nd step (SSs)

113-8 143-9 245-2 138-5

12 4 10-6 6-4 16-5

8 8 6 6

MP, mid-point (half rise point) of step; SS, start of step.

Nucleoside diphosphokinase in yeast cell cycle

187

10,-

.. A . - • - * •

.o

< 2

0

1

2

3 Time (h)

4

5

Fig. 3. NDPK activity in control with wild-type cells. Asynchronous culture of wild-type cells (972 h ) shifted from 29°C to 35°C for 4h and then returned to 29°C at time zero. Curve J4, NDPK activity; assay 4 (incubation temperature 10°C), 10a.u. (arbitrary units) = 2-5e.u. ml" 1 . Curve B, cell numbers; 1 a.u. = l x 106 cells.

There must, however, be some initial connection between the two controls for the following reason. Activity steps continue in cells at the restrictive temperature but these are smoothed out in an asynchronous culture to give a continuous increase in activity (Creanor & Mitchison, 1986). The immediate effect of the shiftdown to the permissive temperature is not only to synchronize the cell cycles but also to synchronize the activity steps and to produce the pattern of plateaux and steps. Some early signal must pass to the activity control from the cell cycle control as it is released from the cdc2.33 block. Thereafter the controls function independently for a time. Why they do so is mysterious, although there may be an analogy in the behaviour of two oscillators coupled by entrainment (Tyson, 1976). The coupling is effective if the free-running periods of the oscillators are fairly near each other but it fails if they are too far apart. It may be that the cell cycles after the shiftdown are too short to permit entrainment of the activity control. An interesting point emerges from the timing between the free-running steps, i.e. the period of the oscillator. After a DNA-division block at 35 °C, it is, on average 138 min, which is close to the doubling time for protein/ cell (143 min) in this mutant (Creanor & Mitchison, 1986). It is the same, on average (138-5 min), after induction synchrony at 29°C, where the normal cell cycle time is 174 min. It is also nearly the same (128 min) in the experiment at 25 °C shown in Fig. 1. This is only a single measurement of a variable quantity but it is even more divergent from the normal cycle time of 220 min at this temperature. It appears therefore that the free-running control is temperature-compensated whereas the cycle time and most aspects of metabolism and growth (e.g. optical density) are temperature-dependent, as would be expected in yeast. The timings at 29°C and 25 °C could be regarded as a relic of the timing at the restrictive temperature but it is still surprising that it is maintained for several hours at the lower temperatures. Temperature compensation is a well-known phenomenon in circadian 188

J. Creanor and jf. M. Mitchison

rhythms and there are reasons why evolution should favour a mechanism linked to true night and day irrespective of temperature (like a real clock). This mechanism, however, is poorly understood and may well be complicated (Edmunds, 1988). It is not clear why it should occur with one process in yeast. Nevertheless there is some evidence for temperature compensation in timers running for less than 24 h ('ultradian'). One part of the division control is temperature-compensated in Chlaviydomonas and Chlorella (Donnan et al. 1985), and in Thalassiosira (Heath & Spencer, 1985). There are also indications of temperature-compensated rhythms of tyrosine aminotransferase activity in synchronized cultures of Tetrahymena (Michel & Hardeland, 1985) and of respiration rate in Acanthamoeba (Lloyd et al. 1982). A final note of caution, however, should be struck about comparisons between induction synchrony and a DNA-division block after selection synchrony. In the former case not only is the DNA-division cycle still running but also the cells start off at a considerably larger size. We thank Peter Fantes and Bela Novak for helpful discussions, and Yvonne Bisset for expert technical assistance. This work was supported by grants from the Science and Engineering Research Council and the Wellcome Trust.

References BENITEZ, T., NURSE, P. & MITCHISON, J. M. (1980). Arginase and

sucrase potential in the fission yeast Schizosaccharomyces pombe. J. Cell Sri. 46, 399-431. CREANOR, J. & MITCHISON, J. M. (1986). Nucleoside

diphosphokinase, an enzyme with step changes in activity during the cell cycle of the fission yeast Schizosaccharomyces pombe. I. Persistence of steps after a block to the DNA-division cycle. J. Cell Sri. 86, 207-215. DONNAN, L., CARVILL, E. P., GILLILAND, T. J. & JOHN, P. C. L.

(1985). The cell cycles of Chlamvdomonas and Chlorella. Neto Phytol. 99, 1-40. EDMUNDS, L. N. (1988). Cellular and Molecular Bases of Biological Clocks, chap. 5. New York: Spnnger-Verlag. FANTES, P. A. (1977). Control of cell size and cycle time in Schizosaccharomyces pombe. J. Cell Sri 24, 51-67. FANTES, P. A. (1984). Temporal control of the Schizosaccharomyces pombe cell cycle. In Cell Cvcle Clocks (ed. L. N. Edmunds), pp. 233-252. New York & Basel: Marcel Dekker. HEATH, M. R. & SPENCER, C. P. (1985). A model of the cell cycle and cell division phasing in a marine diatom. J. gen. Microbiol. 131,411-425. KING, S. M. & HYAMS, J. S. (1982). Synchronisation of mitosis in a cell division cycle mutant of Schizosaccharomyces pombe released from temperature arrest. Can. J. Microbiol. 28, 261-264. LLOYD, D., EDWARDS, S. W. & FRY, J. C. (1982). Temperature-

compensated oscillations in respiration and cellular protein content in synchronous cultures of Acanthamoeba castellanii. Proc. natn. Acad. Sci. USA. 79, 3785-3788. MICHEL, U. & HARDELAND, R. (1985). On the chronobiology of

Tetrahymena. III. Temperature compensation and temperature dependence in the ultradian oscillation of tyrosine aminotransferase. J. lnterdiscipl. Cycle Res. 16, 17-23. MITCHISON, J. M. (1971). The Biology of the Cell Cycle, pp. 25-47. Cambridge: Cambridge University Press. MITCHISON, J. M. (1989). Cell cycle growth and periodicities. In Molecular Biology of the Fission Yeast (ed. A. Nasim, P. Young & B. F. Johnson). New York, London: Academic Press, (in press). MrrcmsoN, J. M. & CREANOR, J. (1971). Induction synchrony in the

fission yeast Schizosaccharomyces pombe. Expl Cell Res. 67, 368-374. NOVAK, B. & MITCHISON, J. M. (1986). Change in the rate of CO 2 production in synchronous cultures of the fission yeast Schizosaccharomyces pombe: aperiodiceventthat persists after the DNA-division cycle has been blocked. J. Cell Sci. 86, 191-206.

TYSON, J. J. (1976). Mathematical background group report. In The Molecular Basis of Circadian Rhythms (ed. J. W. Hastings & H-G. Schweiger), p.89. Life Sci. Res. Rep., vol. 1. Berlin: Dahlem Konferenzen. (Received 6 January 1989 - Accepted 8 February 1989)

Nucleoside diphosphokinase in yeast cell cycle

189