Control of cell division in Paramecium tetraurelia

Experimental

Control Effects

Cell Research 167 (1986) 191-202

of Cell Division

of Abrupt

Changes

Macronuclear ADA

S.-L. CHING

in Paramecium

tetraurelia

in Nutrient Level on Accumulation DNA and Cell Mass

and JAMES

of

D. BERGER”

Department of Zoology, University of British Columbia Vancouver,

B.C.,

Canada

V6T 2A9

In the cell cycle of Paramecium there are three points of interaction between cell growth-related processes and the processes of macronuclear DNA replication and cell division: initiation of DNA synthesis, regulation of the rates of growth and DNA accumulation, and initiation of cell division. This study examines the regulation of the latter two processes by analysis of the response of each to abrupt changes in nutrient level brought about either by transferring dividing cells from a steady-state chemostat culture to medium with unlimited food, or by transferring well-fed dividing cells to exhausted medium. The rates of DNA accumulation and cell growth respond quickly to changes in nutrient level. The amounts of these cell components accumulated during the cell cycle following a shift in nutrient level are typical of those occurring during equilibrium growth under postshift conditions. Commitment to division occurs at a fixed interval prior to fission that is similar in well-fed and nutrient-limited cells. Initiation of cell division in Paramecium is associated with accumulation of a threshold DNA increment, whose level is largely independent of nutritive conditions. The amount of DNA accumulated during the cell cycle varies with nutritional conditions because the rates of growth and DNA accumulation are affected by nutrient level; slowly growing cells accumulated relatively little DNA during the fixed interval between commitment to cell division and fission. 0 1986Academlc PESS, IN.

Cells growing under steady-state conditions maintain a relatively constant cell mass and DNA content through coordination of the two fundamental components of the cell cycle: cell growth, and DNA replication and division [20, 401. The mechanisms that coordinate cell growth with DNA synthesis and division are responsible both for the maintenance of a constant variance in cell mass, and for the adjustment of cell size which occurs following changes in nutritional conditions [25,40]. The first process is required so that the variation introduced at each fission by unequal distribution of cell components to daughter cells is attenuated, and the second so that an approximately constant gene concentration (gene dosage per unit cell mass) is maintained under different growth conditions. Since cell growth rather than DNA replication and division is the rate-limiting factor for progression through the cell cycle for most cell types [40], there must be one (or more) cell cycle event that is growth dependent and which leads to coordination between DNA replication and growth. * To whom offprint requests should be addressed. 13-868341

Copynght 0 1986 by Academic Press, Inc. All rights of reproduction in any form reserved 0014-4827186 $03.00

192 Ching and Berger In ciliates the situation is complicated by the occurrence of a polygenomic macronucleus which is the seat of virtually all transcriptional activity [26], and which divides by an amitotic mechanism that partitions parental macronuclear material to daughter cells with only rough equality [lo, 381. Consequently, Paramecium must possess mechanisms that regulate macronuclear DNA content SO that a constant variance in macronuclear gene dosage is maintained [lo], and that ensure that macronuclear gene dosage is coordinated with the cytoplasmic mass of the cell [8, 191. Regulation of cell mass and macronuclear DNA content is brought about by three related processes controlling initiation of DNA synthesis (IDS), regulation of the rates of protein and DNA accumulation, and initiation of cell division. First, initiation of macronuclear DNA synthesis (IDS) is closely associated with cell mass accumulation in normal cells [27-291. The observation that changes in cell mass, gene dosage and nutritional level affect the timing of IDS when they occur in the parental cell cycle but not when they occur at fission or during the Gl period [ 11, 291 suggests that the timing of IDS is established prior to fission, during the previous cell cycle. Second, the rates of growth and DNA accumulation are constrained throughout the cell cycle by their mutual dependence on macronuclear gene dosage and cell mass [27]. These variables interact to control the cellular rate of protein synthesis [6]. Third, cells become committed to division during the latter part of the cell cycle. This second cell cycle control point occurs at 0.73kO.02 in the cell cycle in well-fed Paramecium cells. It occurs well before the end of the macronuclear S period [30], which continues until the start of macronuclear division at about 0.9 in the cell cycle [3, 381. Well-fed Paramecium cells accumulate a fixed DNA increment during the cell cycle. This occurs even if the cells begin the cell cycle with greatly increased or decreased DNA content or with cell mass which is either greater or less than in normal cells [4, 5, 1 I]. These observations suggest that commitment to division may be dependent on the attainment of a fixed macronuclear DNA increment. However, recent observations indicate that the level of the ultimute DNA increment is not a precondition for cell division, since blockage of DNA synthesis late in the cell cycle does not block or retard cell division and leads to the production of daughter cells with significantly reduced DNA content 1291. The foregoing observations suggest that if accumulation of a fixed DNA increment is a precondition for commitment to cell division, it is not the ultimate DNA increment that is critical, but the DNA increment at some earlier point in the cell cycle, at or prior to the point of commitment to cell divison. If this hypothesis is true, then one would expect the magnitude of the DNA increment at the point of commitment to cell division to be similar in cells growing under different conditions, even if their ultimate DNA increments differed significantly. This study tests this hypothesis by analysing the quantitative correlates of commitment to cell division under various nutritive conditions and shows that the Exp.

Cell

Rrs

167 (1986)

Cell cycle in Paramecium

193

interval between commitment to division and fission is relatively constant, as is the DNA increment at the point of commitment to division. The duration of the macronuclear S period. the relative position of the point of commitment to cell division, and the final DNA increment, however, vary significantly with changes in nutrient conditions. MATERIALS Stocks

and Culture

AND

METHODS

of Paramecium

Paramecium fetraureiia Sonneborn 1975 [35], strain d4-1001 was grown in phosphate-buffered Cerophyl medium with Enterobacter aerogenes as the food source [34]. Strain d4-1001 was derived by mutation from the wild-type stock, 51-S, and carries the recessive temperature-sensitive mutation ccl, which blocks macronuclear DNA synthesis and cell cycle progression at 34.4”C 123, 281. The procedures used in preparation of exhausted medium and initiation of chemostat cultures have been described previously [II]. Synchronous cell samples were obtained by manual selection of dividing cells during a 5-min interval from either exponential cultures or from the chemostat. When dividers were collected from the chemostat culture, the total period during which groups of dividers were selected from a sample did not exceed 15 min after the cells were removed from the chemostat.

Nutritional

Shifts

Nutritional enrichment (shift-up) was done by transferring newly divided cells from a steady-state chemostat culture to medium with excess food. Nutritional shift-down was accomplished by transferring well-fed dividing cells to exhausted medium. Control cells were allowed to grow exponentially in the presence of excess food.

Estimation of the Position of the Point of Commitment to Cell Division The relative position of the point of commitment within the cell cycle was estimated by the residual cell method of Howell et al. [14]. Groups of 50 ccliccl cells from an asynchronous (culture were collected and were immediately shifted to the restrictive temperature in the presence of excess nutriment. After an interval corresponding approximately to the duration of the cell cycle under permissive conditions, the cells were counted. The relative location of the point of commitment to division within the cell cycle was estimated from the fraction of the cells which divided at the restrictive temperature. Cells which were shifted to restrictive conditions prior to the point of commitment to division were unable to divide, while those past that point divided. The location of the point to commitment to cell division within the cell cycle was also estimated using samples of synchronous well-fed cells. At various points within the cell cycle groups of 40 ccl cells were shifted to the restrictive temperature. Two hours after the mean time of division in a control sample, the number of experimental cells which had divided was determined. Those which had divided were assumed to be past the point of commitment to cell division at the time at which the sample was transferred to restrictive conditions.

Cytological Procedures Fixation and fluorochrome staining of cells by the Acriflavin-Primulin method of Cornelisse 8-r Ploem [ 121 was carried out as described previously [27]. Acriflavin stains DNA and Primulin stains protein. Microflourimetry was used to estimate the DNA and protein content of individual cells 1271. DNA and protein content of cells are expressed as percentages of the values for well-fed post-fission control cells: Thus a normal newly-divided cell would contain 100 units of protein and 100 units of DNA.

Statistical Procedures Statistical procedures were used as described calculated according to Beers [l]. Sample means

by Sokal & Rohlf are shown with their

[33]. Propagated standard errors. Erp.

Cell

errors

h'rs

were

167 llY86)

194 Ching and Berger Table 1. Cell cycle variables following Expt no

nutritional

shifts u

Initial DNA”

Initial protein’

Final DNAh

Final protein’

Increase in DNA

Increase in protein

CT (exp.)

GT (control)

66.4kl.7 56.22 1.6 44.9f3.0 72.424.2 ND 58.6k3.2 56.4f5.6 66.4k3.4 54.2k1.6 47.3f1.7 48.0t2.5 65.1fl.7 43.5k3.5 55.3k1.7 76.913.4 88.2k4.7 52.6f1.6 49.7kl.l

28.0-+0.8 41.7+1.0 ND 32.920.9 34.3k1.8 44.6kl.6 23.7kO.8 29.7f1.3 24.6+1.5 ND 32.022.2 ND ND 43.5+0.8 67.9i1.5 52.8+ I .3 32.3fl.O 39.620.8

165.2+ I .O 139.8f0.9 126.4k3.2 191.113.7 ND 144.9f2.5 162.524.2 157.2+_3.6 140..5+3.0 164.4k4.8 132.2k3.0 194.822.1 142.4zL7.8 ND 196.523.3 190.0+3.4 133.4t2.9 144.6k2.6

146.6+2. I 152.6t1.4 ND 137.8+0.7 134.6k2.4 143.0+2.0 106.3+0.8 154.8k3.2 124.6f2.5 ND 116.4+2.0 ND ND 171.8f0.8 198.2t0.9 ND ND 144.2f1.2

98.8i2.0 83.6k1.4 81.4f4.3 118.7k5.6 ND 86.3k4.0 106.1f7.0 90.8k4.9 86.3k3.5 117.1+5.1 84.3k3.9 129.722.7 98.9k8.6 ND 119.6t4.7 101.X+5.8 80.8i3.3 94.9f2.7

118.6f2.7 110.9tl.7 ND 104.il.l 100.3+3.0 98.5k2.6 79.3f1.2 125.1Yc3.5 100.0+2.5 ND 84.4k2.9 ND ND 128.3+1.1 130.3&1.8 ND ND 104.6fl.5

ND ND 12.9 13.3 12.3 12.4 11.4 14.5 ND 15.3 13.9 13.7 13.5 13.8 11.0 15.9 15.1 15.0

ND ND ND ND 6.0 5.0 5.0 5.8 ND 7.0 6.5 ND ND 6.3 6.3 5.8 6.3 6.1

58.Ok3.0 17

39.723.2 14

98.723.9 16

107.1+4.7 12

13.6kO.4 6.0+0.2 I5 II

59.1+5.0

33.851.5

A. Shift-up

83-36 83-37 83-41 8342 83-43 83-44 8345 8348 83-49 84-3 845 84-7 84-11 8415 84-16 84-18 84-19 84-25 Mean ?SE N

B. Shift-down

84-34

100.0+3.8

100.0+4.2

159.1+3.2

133.8fl.6

6.3

5.7

o All values except generation times are expressed as percentages of control. CT is the generation time in hours. ND, Not determined. ’ Mean values from newly divided cells taken from pre-shift cultures. In the case of part A, ceils were from chemostat cultures at or near equilibrium. In B. cells were from well-fed exponential growth phase cultures. ’ Mean values from newly divided cells collected at the end of the first cell cycle after nutritional shift. In A, cells had completed the cell cycle in fresh medium with excess food. In B, cells had completed the cell cycle in exhausted medium.

RESULTS Nutritional

Limitation

Effect on mean cell mass and macronuclear DNA content. Chemostat cultures were started from a well-fed cell population. The mean DNA and protein contents of newly divided cells were examined at regular intervals for at least one week following the initation of a chemostat culture. As the chemostat culture approached equlibrium the mean DNA content and the mean cell mass were reduced to stable values significantly lower than those in well-fed cells. Population doubling time in the chemostat cultures ranged from 26 to 30 h vs 5.5 h for well-fed cells. Exp.

Cell

Rrs

/67 (1986)

Cell cycle in Paramecium

195

At low growth rates, such as in chemostat cultures with a generation time of 26-30 h, the normal tight coupling between cell mass and macronuclear DNA content was disrupted. Cell mass decreased relative to macronuclear DNA content. The first two columns of table IA (Initial DNA and Initial Protein) show the mean DNA content and cell mass (total protein) of newly divided cells from chemostat cultures at or near equilibrium. The mean post-fission DNA content was typically about 60 units (percent of the mean post-fission value for well-fed cells), while the protein content decreased to lower values. If it is assumed that the cultures were at or near equlibrium, then the DNA increment produced during each cell cycle would also be the same as the mean post-fission value. The observations imply that under conditions of nutritional limitation, the DNA increment is about half as great as that produced in well-fed cells during the course of a cell cycle (100 units) [4, 5, 101. On the other hand, relative cell mass (protein content) decreased to as little as 30 units or less when the growth rate was reduced. Macronuclear DNA contents significantly below 50 units were not observed in these very small cells. These results suggest either that a minimum DNA content of approx. 50 units is necessary to maintain viability at reduced growth rates. or that the production of about 50 units of DNA is a necessary precondition for fission. As discussed below, the latter hypothesis is favored. Timing of initiation of cell division in chemostat cultures. The temperaturesensitive mutant, ccl, was used to map the relative position of the point of commitment to division within the cell cycle in both well-fed and chemostat cells. This mutation leads to rapid inactivation of macronuclear DNA synthesis under restrictive conditions [28] and division is blocked in cells which have not yet reached the point of commitment to cell division [23, 301. Thirteen groups of 50 asynchronous ccl cells from a chemostat culture were collected and immediately placed at the restrictive temperature in the presence of excess food. After a time interval equal to the duration of the recovery cell cycle, the cells were counted. The fractional increase in cell number since the beginning of the heat treatment (0.027+0.014) was used to estimate the relative position of the point of commitment to cell division by the residual cell method. The point of commitment to cell division was estimated to be at 0.96kO.02 in the cell cycle. This value is significantly different from the value of 0.73kO.02 obtained for well-fed cells [30]. However, the absolute duration of the interval between the point of commitment to cell division and fission in chemostat cells (67?40 min) does not differ statistically at the 0.05 level from that in well-fed cells (8727 min). Effects of Nutritional

Shifts on Macronucleur

DNA Content and Cell Muss

Nutritional shift-up. Two groups of 40 dividing cells were selected from a chemostat culture. One group was fixed immediately after fission and was used to determine the initial DNA and protein content. The second group was allowed to progress through the cell cycle in the presence of excess food before being fixed and stained to determine the effects of nutritional shift-up on accumulation of Exp.

Cell

I?es 167 (1986)

196

Ching and Berger

Fig. 1. 0, Mean macronuclear DNA ; 0, percentage of ceils committed to division, as a function of stage in the cell cycle. Vertical range bars denote 95% confidence intervals of the mean. Sample size, 3540 cells. 00

02

04 CELL

0% CYCLE

08 STAGE

10

macronucelar DNA and total protein. Cells accumulated an average of 98.7k3.9 units of DNA and 107.1 zk4.7 units of protein in the first cell cycle after shift-up. While some variation in these values occurred from one experiment to another, this variation was not correlated with the initial DNA or protein content of the cells (table 1 A). The amounts of DNA and protein accumulated during the cell cycle following nutritional shift-up did not differ significantly from the normal values for well-fed cells (100 units). The magnitude of the DNA and protein increments reflect the new conditions following the nutrient shift. During the second post-shift cell cycle, cells accumulated normal amounts of DNA and cell mass. The cell cycle was slightly (1.1320.04) longer than the control cell cycle. The increase in the DNA and protein increments to normal levels during the cell cycle following return to well-fed conditions forms the basis for the regulation of cell mass and DNA content to normal levels. Nutritional shift-down. The shift-up experiments show that the DNA and protein increments are rapidly reset to reflect the new nutrient conditions and suggest that the initial DNA and protein content have little effect on the amounts of DNA and protein accumulated. To test this hypothesis a shift-down experiment was performed. Cells were transferred at division from medium with excess nutrition to exhausted medium, and were fixed after the subsequent fission. During the post-shift cell cycle cells accumulated an average of 59.1 M.0 units of DNA and 33.8f 1.5 units of protein (table 1 B), compared with 100 urits for control cells. As expected, the DNA and protein increments were rapidly reset to reflect the new nutritive conditions. Initial cell mass and DNA content have little or no effect on the amounts of DNA or protein accumulated. Estimation of the Threshold DNA Increment Commitment to Cell Division

Associated

with

Well-fed cells. The observations that the relative location of the point of commitment to cell division in chemostat cells is much later in the cell cycle than it is in well-fed cells, and that the DNA increment accumulated during the course of the cell cycle is about half as great, suggest that the point of commitment to Exp.

Cell

Res

167 11986)

Cell cycle in Paramecium

Table 2. Estimation

of threshold

macronuclear

DNA

content in well-fed

Control cell

Cells blocked in DNA synthesis after point of commitment to div.

Expt

Post-fission DNA content0

Post-fission DNA content0

82-7 82-9 82-12 83-5

100.0+2.2 1OO.ot3.4 100.0f0.7 100.0+2.1

70.8k2.7 76.Ok3.2 73.3k1.8 77.7k2.6

4lS-tS.8 52.0f7.3 46.2k3.6 55.3k5.6

74.4f1.3

48.8f3.1

Mean&SE

191 cells

Increase in DNA in experimental cells’

n Data from reference [3 11. ’ Increase in DNA is calculated as the prefission DNA content of the experimental cells (twice the post-fission value) less the initial DNA content (post-fission control DNA content).

division may occur at similar levels of DNA accumulation under both conditions. The DNA increment present at the point of commitment to division in well-fed cells was estimated in two ways. First, the mean DNA content of cells at various points during the latter part of the cell cycle was determined and the fraction of the cells committed to cell division at each sample point was ascertained. A plot of both variables against stage in the cell cycle reveals that the mean DNA content at the median point of commitment to cell division was about 163+5 % of the initial value. This corresponds to a DNA increment of about 63+5 units at the point of commitment to cell division (fig. 1). A second approach to determine the DNA increment at the time of commitment to cell division involved blocking DNA synthesis in ccl cells near the mean time of commitment to cell division (about 0.75 in the cell cycle) and then collecting newly divided cells at the subsequent fission. Cells which subsequently reached division were past the point of commitment to division at the time of the temperature shift. Since macronuclear DNA synthesis remained blocked until after fission, the post-fission DNA content of these cells can be used to calculate the DNA increment at the point of commitment to cell division (table 2). The value obtained (48.8f3.1 units) is somewhat less than that obtained from direct measurement of DNA content at the median time of commitment to division. Although there is a consistent difference between the results of the two types of estimate of the DNA increment at the point of commitment, the differences are not significant at the 0.05 level. Since the cells in the second experiment were already past the point of commitment to cell division, the true value of the DNA increment at commitment to cell division should be no higher than the value obtained in this experiment. The basis for the consistent discrepancy between the results of the two experiments is not known. The values obtained for the DNA Exp.

Cell

Res 167 11986)

198

Ching and Berger

increment at the point of commitment to cell division in well-fed cells are close to the mean post-fission DNA content of chemostat cells (table IA, Initial DNA) and suggest that the threshold DNA increment in nutrient-limited cells may be similar to that in well-fed cells. limited (chemostat) cells the point of commitment to cell division lies very close to the end of the cell cycle as shown above. Consequently, the mean post-fission DNA content is very close to the value of the DNA increment at the point of commitment to cell division. In wellfed cells, in contrast, a greater fraction of the S period occurs between the point of commitment to cell division and the end of the S period. This difference, combined with the higher rate of DNA synthesis in well-fed cells leads to the synthesis of approximately half of the final DNA increment between the point of commitment to division and the end of the S period which occurs about 60 min later. If it is assumed that the chemostat cultures of nutritionally-limited cells were at or near equilibrium, then the DNA increment at the point of commitment to cell division can be estimated by multiplying the post-fission DNA content by the fraction of the S period which occurred prior to the point of commitment (at 90 min prior to fission) (table 3). The results (49.4k2.6 units) agree closely with the value estimated for well-fed cells blocked at the point of commitment (48M3.1 units). It thus appears that the DNA increment at the point of commitment to cell division is not reduced significantly in nutrient-limited cells, even though the ultimate DNA increment is only half as great as that in well-fed cells. Nutrient-limited

cells. In nutritionally

DISCUSSION Regulation

of Macronuclear

DNA

Content

and Cell Size

Cell growth in Paramecium occurs more or less parallel with the increase in DNA content and a close correlation between DNA content and cell mass (measured as protein content) is maintained under normal growth conditions [9, 201. Under more extreme conditions, the relative increase in DNA content may be greater or less than the increase in protein content. The decrease in protein content relative to DNA content observed in nutrient-limited cells in this study is probably a consequence of the reduced rate of protein synthesis, combined with rates of protein turnover or degradation which are not proportionately decreased. The greater than normal protein increment following nutrient shift-up is presumably the result of cell growth without DNA accumulation during the greatly extended Gl interval following shift-up [ 111. These observations suggest that the DNA increment is more stringently controlled than the cell mass increment. In Paramecium regulation of macronuclear DNA content and cell mass take place via incremental mechanisms [4, 5, 8, 10, 271. In such regulative systems all cells synthesize a standard amount (increment) of DNA or cell mass during each cell cycle, regardless of the initial values of these variables [ 131. This study shows Exp.

Cell

Rev 167 (1986)

Cell cycle in Paramecium Table. 3. Estimation

of threshold DNA content in chemostat

199

cells

Expt

Initial DNA content

Estimated generation time (h)

Estimated duration of S phase (h)b

Fraction of S phase prior to commitmenV

Estimated DNA threshold levef

84-3 84-5 84-7 84-l 1 84-15 84-19 84-25 Mean

47.3k1.7 47.9k2.5 65.1k1.7 43.5k3.5 55.351.7 52.6k1.6 49.7fl.l 51.6k2.7

33.3 31.3 33.3 27.0 29.4 27.8 29.4 30.2+ 1.O

24.5 23.0 24.5 19.8 21.6 20.4 28.9 23.211.7

0.96 0.96 0.96 0.95 0.95 0.95 0.97 0.96f0.003

45.4. 45.8 62.4 41.3 52.7 50.0 47.9 49.422.6

a Reciprocal of dilution rate. b (0.75.generation time) -0.5 h. ’ (Duration of S phase - 1.0 h)/(duration of S phase). d Initial DNA content fraction of S phase prior to commitment.

that the magnitudes of the DNA and cell mass increments vary under different nutritive conditions, and that the levels of the DNA and cell mass increments are rapidly reset following changes in nutrient level. During the first cell cycle after a change in nutrient level the DNA and cell mass increments are similar to those produced during equilibrium growth under post-shift conditions. The mechanisms regulating DNA content and cell mass are growth rate-dependent and the rates of DNA accumulation and cell growth are similarly regulated [6, 271. Control of Cell Division This study suggests that the maintenance of a constant DNA increment in both well-fed and nutrient-limited cells is a consequence of the requirement for fixed DNA increment as a precondition for commitment to cell division in Paramecium. Consequently, the difference in the final DNA increments between well-fed and nutrient-limited cells is a consequence of differences in the rate of DNA accumulation during the fixed interval between commitment to division and tission. Cells with a low rate of DNA accumulation accumulate less DNA during this interval than do cells with a high rate of DNA synthesis. In well-fed cells, commitment to division occurs about two-thirds through the S period, while in nutrient-limited cells commitment to division occurs very near the end of the S period. Consequently, in cells with a low rate of DNA accumulation the final DNA increment is only slightly greater than the DNA increment required for commitment to cell division. Thus the minimum DNA content of newly-divided chemostat cells is very similar to the estimated threshold DNA increment required for commitment to cell division. Similar observations have been made on aged P. tetraureliu cells in which the rate of DNA synthesis decreases and the duration of the S period is extended [32]. The mean macronuclear DNA content Exp.

Cell

Hes 167 (1986)

200

Ching and Berger

approaches a minimum value which is about 50% of that in young well-fed cells [31]. In P. caudatum the mean DNA content decreases to much lower levels [37] suggesting that the mechanisms regulating DNA content in the two species differ somewhat. These observations and the present study both suggest that accumulation of a threshold DNA increment is an essential precondition for initiation of cell division in P. tetraurelia. The point of commitment to cell division in Paramecium is closely associated with initiaiton of oral morphogenesis and with incipient micronuclear mitosis [El, as it is in Tetrahymena [El. In Tetrahymena, oral morphogenesis and cell division also occupy a fixed interval at the end of the cell cycle [21, 361. This study shows that this is also true for Paramecium, despite the approximately fivefold difference in the generation times of well-fed and chemostat cells. Growth-related control of initiation of cell division also occurs in Tetrahymena, but involves increase in cell mass rather than macronuclear DNA content as in Paramecium. Initiation of oral morphogenesis occurs when cells reach a size about 1.8 times the post-fission value. This critical size is independent of generation time and growth rate [21]. Some evidence indicates that the ribosomal RNA increment is the critical variable for initiation of cell division [16-181. Cell Cycle Control

in Paramecium

In contrast to yeasts [2, 2.51 and Tetrahymena [38] which have a well defined regulatory point just prior to IDS, the major cell cycle control points in Paramecium occur late in the cell cycle. Both commitment to cell division and commitment to meiosis in the next cell cycle are firmly linked and occur late in the cell cycle [9]. The timing of IDS is also established prior to the preceding fission [ 11, 291. The association of the point of commitment to meiosis with the point of commitment to IDS in other organisms [25, 391 suggests that the timing of IDS may also be set at or just prior to the point of commitment to meiosis in Paramecium. On this assumption we propose that a regulatory function similar to ‘start’ in yeasts [2, 251 might act late in the cell cycle in Paramecium and be involved both with commitment to cell division in the present cell cycle and with commitment to the DNA replication in the next [l I]. This is not unprecedented; the ‘start’ function is involved both with commitment to DNA replication and commitment to cell division in yeasts [22, 241. Although commitment to DNA synthesis and commitment to cell division seem to occur at similar positions within the cell cycle, from the perspective of the control of a single cell cycle they are widely separated (fig. 2). The first regulatory events of the cell cycle take place prior to the preceding fission, and cell cycles are functionally overlapped. The cell cycle of Paramecium consists of three functional segments. The initial segment begins at the point of commitment to IDS and the vegetative division pathway. This control point occurs in the latter part of the preceding cell cycle. Its precise location is unknown, but by analogy with yeasts or Tetrahymena it is likely to be located close to the point of commitment to meiosis. In Paramecium Exp.

Cell

Res 167 (1986)

Cell cycle in Paramecium -cycle --I,-

1-A

075

cycle 100

0’25

Tlmlng of IDS set

2m

o’s0

IDS

Of5

1’00

CommltmQnt Comlnitment

FixQdDNA

-----____1

SEGMENT

DNA contant nutrient level in cycle 1 prior to sotpolnt

to Diwsron to MQIOSIS

-

Increment

1

varlabk duration dQpQnds on cell mass

201

--

SEGMENT vartable duration depends on rataof DNA accumulation

2 -

SEGMENT flxQd

duratton.

3

Fig. 2. Cell cycle model for Paramecium.The relative duration of cycle stages is shown for norma1 well-fed cells. See text for explanation.

cell

this point is closely associated with the point of commitment to cell division and occurs at approx. 0.75 in the cell cycle [9]. The duration of this initial segment of the cell cycle is variable and responds in the expected ways to changes in cell mass, macronuclear gene dosage and nutrient level [l 1, 27, 291. However, once the timing of IDS is established, changes in these variables are without significant effect on the timing of IDS. The second segment of the cell cycle begins with initiation of DNA synthesis and extends to the point of commitment to cell division. These two control points are linked by accumulation of macronuclear DNA, which is the growth-related function most strongly associated with commitment to cell division. The duration of the S period may vary considerably, unlike the situation in yeast or in mammalian cells in which the period from initiation of DNA synthesis to division is relatively fixed [25, 401. Extension of the macronuclear S period in Paramecium is especially noticeable in nutrientlimited [I I] or aged [32] cells, or in cells with drastically reduced gene dosage [5, 71. In such cases the S period may be five or more times the normal length. The third segment of the cell cycle is of fixed duration (about 90 min) and extends from the point of commitment to cell division until fission. The foregoing analysis suggests that cell cycles in Paramecium are functionally overlapped, with the initiation of the next cell cycle occurring prior to the completion of the preceding cell cycle. Such an arrangement is not unprecendented: in budding yeasts the relative location of ‘start’ within the cell cycle can be brought forward into the preceding cell cycle to a point just after completion of nuclear division [18]. However, in Paramecium the point of commitment to IDS is probably prior to nuclear division and even before the completion of macronuclear DNA synthesis in the preceding cell cycle. It therefore becomes important to determine the precise location of the point of commitment to IDS in the Exp.

Cell

Re.r 167 (1986)

202

Ching and Berger

cell cycle, and to determine whether the period between commitment to IDS and fission functions as part of the Gl interval.

Paramecium

This study was supported by grant A-6300 of NSERC Canada to J. D. Berger. We thank Dr H. W. Brock for helpful comments about the manuscript.

REFERENCES 1. Beers, Y, Introduction to the theory of error. Addison-Wesley. Cambridge, Mass. (1953). 2. Beach, D, Durkaz, B & Nurse, P, Nature 300 (1982) 706. 3. Berger, J D, J protozool 18 (1971) 419. 4. - Exp ceil res 114 (1978) 253. 5. - J protozool 26 (1979) 18. 6. - Exp cell res 142 (1982) 261. 7. - Can j zoo1 60 (1982) 2501. 8. - The microbial cell cycle (ed P Nurse & E Streiblova) p. 191. CRC Press, Boca Raton, Fla. (1984). 9. - Exp cell res 166 (1986) 475. 10. Berger, J D & Schmidt, H J, J cell biol 76 (1978) 116. 11. Ching, A S-L & Berger, J D, Exp ceil res 167 (1986) 177. 12. Comelisse, C J & Ploem, J S, J histochem cytochem 24 (1976) 72. 13. Fantes, P & Nurse, P, The cell cycle (ed P C L John) p. 11. Cambridge Univ Press, Cambridge (1981). 14. Howell, S H, Blashko, W J & Drew, C M, J cell biol 67 (1975) 126. 15. Jauker, F, J cell biol 67 (1975) 901. 16. Jauker, F & Rinaldy, A, Exp cell res 143 (1983) 163. 17. Jauker, F, VII International congress of protozoology. Abstracts p. 143 (1985). 18. Johnston, G C & Singer, R A, Exp cell res 149 (1983) 1. 19. Kimball, R F, Exp cell res 48 (1967) 378. 20. Mitchison, J M, The biology of the cell cycle. Cambridge University Press (197 I). 21. Nelson, E M, Frankel, J & Martel, E, Dev biol 88 (1981) 27. 22. Nurse, P & Bissett, Y, Nature 292 (1981) 558. 23. Peterson, E L & Berger, J D, Can j zoo1 54 (1976) 2089. 24. Piggott, J R, Rai, R & Carter, B L A, Nature 298 (1982) 391. 25. Pringle, J R & Hartwell, L H, The molecular biology of the yeast Succharomyces (ed J N Strathem, E W Jones, E W & J R Broach) p. 97. Cold Spring Harbor Laboratory. Cold Spring Harbor, N.Y. (1981). 26. Raikov, I B, The protozoan nucleus. 2nd ed. Springer Verlag New York (1982). 27. Rasmussen, C D & Berger, J D, J cell sci 57 (1982) 315. 28. - Exp cell res 155 (1984) 593. 29. - Exp cell res 165 (1986) 53. 30. Rasmussen, C D, Ching, A S-L & Berger, J D, J protozool 32 (1985) 366. 31. Schwartz, V & Meister, H, Arch Protistenk 117 (1973) 85. 32. Smith-Sonnebom, J & Klass, M, J cell biol 61 (1974) 591. 33. Sokal, R R & Rohlf, F J, Biometry. Freeman, San Francisco (1969). 34. Sonnebom, T M, Methods in cell physio14 (1970) 241. 35. - Trans Am microsc sot 94 (1975) 155. 36. Suhr-Jessen, P B, Stewart, J M & Rasmussen, L, J protozool 24 (1977) 299. 37. Takagi, Y & Kanazawa, N, J cell sci 54 (1982) 137. 38. Tucker, J B, Beisson, J & Roche, D L J, J cell sci 44 (1980) 135. 39. Wolfe, J, Dev biol 54 (1976) 116. 40. Yanishevsky, R M & Stein, G H, Int rev cytol 69 (1981) 223. Received April 30, 1986 Revised version May 30, 1986

Exp.

Cell

Res 167 (1986)