ABSTRACT
There is some evidence that synthesis in single cells of Tetrahymena occurs with very nearly constant rate when traced from one division to the next.
Overall exponential growth over a number of cell cycles may be the statistical result of the fact that centres—of which each one is in control of synthetic processes occurring with constant rates—multiply by discrete doublings, i.e. exponentially, in close relation with the process of cell division.
While this doubling occurs there may be a provisional block to further increase in respiratory rate.
In the post-exponential growth-phase cells may differentiate and some may grow too big for their synthetic machinery, as—after a lag—it becomes reactivated upon transfer to fresh nutrient medium. For this reason overall exponential growth over a number of cell cycles may pertain only to a fraction of the cell with which the experiment was started.
INTRODUCTION
For single-cell cultures of Paramecium grown in capillaries, Giese (1945) found exponential multiplication to a 32-64-cell stage. This inspired the hope in the writer that synchronism of divisions in single-cell cultures of Tetrahymena might be maintained over a number of cell cycles, so that accurately recorded growth curves for such cultures would supply information about the activity of the growing cell in relation to the cell cycle.
The experiments were performed with pure cultures of the ciliate Tetrahymena piriformis, Lwoff’s strain (experiments 1,2, and 4) or Kidder’s strain W (experiment 3). For each experiment 1 cell, taken from a 2–4 days’ old culture (which had passed beyond the exponential growth phase), was grown bacteria-free in a microrespirometer on a small amount (0·05–0·155 µd.) of nutrient medium (2 per cent, proteose-peptone (Difco) + 1 per mille liver fraction L (the Wilson Laboratories) in tap-water).
By convention, the rate of respiration was taken as a measure of the amount of respiring material. The resulting curves (O2/min. = Vt, plotted against hours) are therefore to be dealt with as growth curves, and the slopes of the curves indicate rate of synthesis. All conclusions derived are preliminary; they should in due time be checked, with more experiments and also other measures of growth.
METHOD
Space prevents a detailed description of the method (see Fig. 1,1 and II) used for respiration measurements. It will be published in Experimental Cell Research. It combines the breaking pipette technique of handling single cells in small volumes (Holter, 1943) with the use of the broken-off tip of the pipette as a Cartesian diver respirometer. Fig. 1,1 will give some idea of the way the diver is made and charged. Sterile conditions are extremely easy to maintain. The diver’s equilibrium pressure is adjusted and measured using a new ‘sensitive manometer’
FIG. 1,1. The tip of a glass capillary is pulled in a microflame to form what is to be the diver respirometer. The pointed tip should be about 5-7 mm. long, the wider piece about 1 cm., and the narrow channel about 1 cm. Inner dimensions of the order 50 µ, 500 µ, and 50 µ, respectively. To select the right capillary from which to make the diver one should aim at a tube which will float in N/10 NaOH when filled with air to about half its length. After transfer to the vacuum bottle shown below in the figure the diver is made buoyant. Suction is applied so that first the nutrient medium filling the narrow channel, then air, is sucked out of the diver. When normal pressure is re-established, air has become replaced by alkali. This is repeated in steps so that buoyancy is approximated.
FIG. 1, II. The diver floats in N/10 NaOH in a, thus freely exposed to the pressure prevailing in the air-space b which extends through c and d to the water surface e. At the beginning of the experiment the pressure in b is adjusted by mouth through n, m, I, with d in the position shown in the figure. With m closed, the manometer k is useful for the fine regulation of the initial flotation pressure in b. n is a CO2-trap, I is an air-brake.
During actual measurements d is closed to I, m, n and to k, but open to e.


In the present studies h in the equation has been small compared to B. It should be noticed, however, as an interesting and promising feature of the present method, that x varies inversely with (B+h—e). Therefore the sensitivity of the system should increase greatly at low pressures and should approach infinity when h approaches negative values equal to B—e.
(Fig. 1, II) which is really not a manometer at all but a burette (II, e, f, g, h) by means of which accurately measured volumes of fluid can be withdrawn from a closed air space (II, b) in which regulated pressure changes are thereby created. The amount withdrawn is read by the movements of the air bubble g. The diver floats in a small amount of alkaline medium in a pocket (II, d), open to the air space. The ‘manometer’ can be adjusted to the nearest 1125 mm. water pressure 4.10−6 atm. The diver is sensitive to around this limit, and since its gas volume is 0 5·1 µl. the sensitivity of the method is of the order 2–4. 10−6 µl. O2 uptakes were measured over about 10-minute periods, but were plotted for 1-minute periods in the middle of the interval for which they were measured. The high sensitivity and the high ratio (order 105): oxygen capacity/sensitivity of this diver makes it excellently suited for long-continued growth studies. For further technical information, see the legend to Fig. 1,1 and II.
RESULTS
The experimental results are shown in Figs. 2,3, and 4 and in Table 1. In Figs. 2 and 4 the ordinate scale is logarithmic, in Fig. 3 it is equidistant. In Fig. 2 the upper set of curves (I) are control experiments (by Mr. H. Thormar) on single cells multiplying in capillaries, not themselves used as respirometers; the same nutrient medium was used as in the respiration experiments, but made up with double glass distilled water; salt mixture as in Kidder’s medium A (Kidder & Dewey, 1951). Cell numbers were reckoned to increase by one each time two daughter cells were observed to separate, and in the two experiments all divisions were very carefully observed (except for a short period of time in the final part of the one experiment, curve dotted). In the graph, log cell numbers are plotted against hours of development at 27°. Each step on the curves corresponds to the logarithm of 2 and thus to a doubling of cell numbers.
Single Cells of Tetrahymena piriformis, multiplying on Proteose-Peptone + Liver Extract in Tap-water Characteristics of Growth Curves 1,2, and 3, FIG. 3, after Correction for the Blanks

In single-cell cultures the periodicity of cell division is maintained over a number of generations. Logarithmic ordinate scales omitted, because each experiment requires its own scale. Each step on curves corresponds to a doubling in numbers. Generation times (g.t.): for I: 108 and 113; for II: 133, 122, 152, and 152 min.
In single-cell cultures the periodicity of cell division is maintained over a number of generations. Logarithmic ordinate scales omitted, because each experiment requires its own scale. Each step on curves corresponds to a doubling in numbers. Generation times (g.t.): for I: 108 and 113; for II: 133, 122, 152, and 152 min.
Rate of respiration plotted on an equidistant scale against hours. Of each pair of vertical lines intersecting the curves, the first one indicates onset of division activity, the second indicates that the daughter cells have separated. All experiments start with 1 cell; the number of cells recovered at the end of the experiment is indicated.
Rate of respiration plotted on an equidistant scale against hours. Of each pair of vertical lines intersecting the curves, the first one indicates onset of division activity, the second indicates that the daughter cells have separated. All experiments start with 1 cell; the number of cells recovered at the end of the experiment is indicated.
The next following set of curves (II, Fig. 2) indicates how cell numbers were observed to increase in the four respiration experiments, shown in the curves 1–4, Figs. 3 and 4. In this case, with each new division, cell numbers were assumed to double over the period from the time the first cell was observed to begin furrowing till the last pair of division cells had separated in the diver. By the end of the third or the fourth cycle this assumption was checked by actual cell counts; and the expected number of cells (8 or 16) was obtained.
Same experiments as in Fig. 3, plotted as logarithms against time. See text.
For each experiment (I, II, Fig. 2) two straight lines were made to fit the upper and the lower ‘edges’ of the stepwise curves. The mean of the slopes for each set of two lines gives the generation time, g.t., which was found to vary from 108 to 152 minutes in the experiments reported. The straight-line fit indicates that the data are interpreted to show that cellular multiplication was approximately exponential in all cases.
The first method of following cell multiplication is based on the observation of a definite morphological event, which in the single cell occurs infinitely fast. The second method is based on the observation of the complete cytoplasmic fission, in each cell occupying about 8 per cent, of the cycle. As expected, the resulting curves show more abrupt jumps with the first than with the second method. All curves are interpreted to show that, starting with a single cell, a rather nice periodicity of divisions is maintained over at least 3–4 generations.
Fig. 3 shows the four respiration experiments (1–4) and also two experiments (both given triangular symbols) with divers filled as usual, but without Tetra-hymena cells. They serve as blanks and also as controls of sterility. With the possible exception of curve 4 (not included in Table 1, because the scattering is of the same order as some of the phenomena to be observed), all experiments seem to indicate respiratory periodicity correlating with the cell cycles.
DISCUSSION
With reference to Fig. 3, the first point to be stressed is that in most cases the curves can be interpreted as showing that the rate of respiration increases linearly with time, usually beginning with the onset of the cytoplasmic fission (periods of fission indicated by frames), continuing through the process of fission, and ending 15-30 minutes prior to the onset of the next fission. This observation of constant rates of synthesis between divisions should be checked by more experiments. As the results stand, they suggest that although we deal with microbial cultures in their exponential phase of growth, synthesis between divisions is not an exponential function of mass, but rather a simple function of the number of individuals present. Therefore, growth between divisions does not— as might be inferred from the words logarithmic or exponential—seem to be an autocatalytic event.
Again with reference to Fig. 3, the next observation to be stressed is a tendency (for an exception see curve 3) for the rate of synthesis to double just prior to the onset of the cytoplasmic fission. In other words, although the rate of synthesis appears to be governed by the number of cellular units, the important thing may not be the number of cells per se, but rather the number, in the cells, of undefined centres doubling before each cytoplasmic fission. Such centres governing the rate of synthesis could be the macronucleus or it could be other cellular organelles. Tetrahymena piriformis is known not to contain micronuclei and the macronucleus is known to divide amitocally, prior to and during the cytoplasmic division (Furgason, 1940). Prior to amitosis ‘reorganization bands’ have been observed to move across the macronucleus.
Fig. 3 indicates a third point which should be stressed, that prior to doubling of the rate of synthesis there seems to be a respiratory plateau, or maybe a slight dip, lasting 15-30 minutes. Whatever the ‘synthetic centres’ are, they seem to have the alternative of duplicating themselves (in the respiratory plateau) or of controlling synthetic processes in the whole cell.
With special reference to Fig. 4, the overall shape of the growth curves should be discussed.
When plotting the logarithms of the total respiratory rate (vt = O2 / min.) against hours of growth the four curves result which in Fig. 4 are designated 9·5.10−6 (experiment 1), 9·5.10−6 (experiment 2), 4·5.10−6 (experiment 3), and 4.10−6 (experiment 4). Thus the curves are labelled according to the rate of oxygen uptake (µl. O2/min.) which the initial cell showed before growth began in the respirometer (see Table 1). The position of all curves on the logarithmic ordinate scale is arbitrary; we are concerned only with the slopes.
In experiment 2, perhaps also in experiment 3, there is a definite lag phase before growth is resumed on the fresh medium, supplied with the introduction of the cell into the diver. It may be significant that this lag (in experiment 2) is broken by a cell division, because prior to each new division—even in the fast multiplying culture—there is a short lag, the respiratory plateau. Whether justified or not, for the initial parts of the other experiments (1,3, and 4) the curves have been drawn in accordance with the findings in experiment 2.
Inspecting, throughout their course, the two curves which are both designated by 9 5.10”6, the respiratory periodicity shows up rather clearly. Disregarding the periodicity, however, it is apparent that the overall trend of both curves deviates from linearity. In other words, although multiplication is logarithmic (Fig. 2), synthesis, even in its overall course, does not appear to be so.
However, in their main trend the two curves can be transformed into straightline logarithmic curves if we accept that not the total initial respiratory rate (v0) but only part of it (v0 -a) shows logarithmic growth with a doubling time equal to g.t.
The two experiments can thus be described by the formula In experiments 1 and 2 (v0-a) approximates 6–4.10−6, and 4·5.10−6 (µl. O2/min.) respectively; and the non-growing constant fraction (a) seems to represent nearly one-third and one-half respectively of the total initial rate. The characteristics of the curves representing experiments 1, 2, and 3 are given in Table 1. Section I of that table gives the slopes of the straight lines between divisions (Fig. 3) and demonstrates the basis on which a doubling in rate with each division has been suggested (paragraph 1 of discussion). Section II of the table demonstrates the total gain, Av, in respiratory rate per generation. Except for the last period in experiment 3, Av is also accepted as showing doubling with each new generation. Sections III and IV show that—due to the existence of the non-growing fraction a—the cells get smaller and smaller with each generation, as measured by the amount of O2 they consume per unit time. In experiments 1 and 2, and in agreement with Ormsbee (1942), this was checked by the visual observation that the cells get progressively smaller in the first generations.Experiments 3 and 4, so far omitted from this discussion, leave us in doubt whether in these two experiments the volume of nutrient medium (not measured, but in both cases smaller than in experiments 1 and 2) ever permitted unrestricted growth to take place, that is, whether (vt -a) ever grew exponentially with time. In experiment 3 (section II, Table 1) Av/generation doubled from the first generation to the next, but remained nearly constant from the second to the third. In the same experiment the rate of synthesis (Av /hour, section I, Table 1) nearly doubled with each generation. The slope over the third generation may have been slightly lower than that corresponding to a doubling from the previous generation, although hardly significantly so. This situation suggests that when growth becomes limited by the volume of nutrient medium available, the first thing to happen is that the respiratory plateau (paragraph 3 of discussion) increases in duration at the expense of the preceding period of straight-line increase in respiring mass. Experiment 4 suggests that this develops to a point where the respiratory rate fluctuates up and down in broad waves, with little overall increase, resembling the situation prevailing in dividing eggs (Zeuthen, 1950 a, b, reviewed in 1949,1951; Scholander et al., 1952).
The uncertainties mentioned, greater for experiment 4 than for experiment 3, prevent the determination of the value of a in equation (1) in these two experiments. However, in Fig. 4 the dotted lines below the two curves representing experiments 3 and 4 indicate which overall slope in each case would have corresponded to a doubling in mass over the first generation. This slope is higher than the one found for the total respiration. Hence the cell with which the diver was first charged in experiments 3 and 4 may not have been capable of growing with its whole body. By analogy with the conclusions derived from experiments 1 and 2 the curves for experiments 3 and 4 will give some idea of the possible value of a also in experiments 3 and 4.
Tetrahymena cells grown in mass culture from stationary phase cells are known in the early phase of growth progressively to reduce to about 60 per cent, of the initial weight and again, after the end of the exponential phase, to grow bigger. Ultimately, big stationary phase cells can be recovered (Ormsbee, 1942). Late in the stationary phase cells differentiate, e.g. they become highly polymorphic with respect to shape and size of cell body and cell nucleus.
In the four respiration experiments reported the initial level of respiration varied from cell to cell, although for three of them (curves 1, 2, and 4) cells were from the same clone. As has been demonstrated, the progeny of each cell retains constant growth characteristics (see curves, Fig. 3 and 4, and Table 1), probably till the end of the exponential phase. Thus, with respect to growth capacities, the individuality of the initial cell was transmitted to several subsequent generations. Accordingly the present results, together with those of Ormsbee, may indicate that in Tetrahymena differentiation is limited to postexponential phases of growth. The outcome of this fifth point to be discussed is that one aspect of differentiation in stationary phase cultures is that cells may grow too big for the synthetic machinery which they possess and which becomes reactivated, after a lag, upon transfer to fresh nutrient medium.
The comparison made above between a dividing egg and a multiplying ciliate may turn out to be quite superficial. Nevertheless, this presentation has served the purpose of demonstrating, on a less complex system, which factors may be involved in producing respiratory rhythmicity in the cleaving egg. While the egg represents a closed system, Tetrahymena may grow on fully defined media, as demonstrated by Kidder and collaborators (1951) and by Elliot (1950, 1951), with all the possibilities for controlling limiting factors which lie therein. It is planned to extend the present studies to cells growing on defined media, but so far we have met difficulties because on such media there has been a low degree of synchronism of divisions.