Vol. 123, No. 2 Printed in U.SA.
JOURNAL OF BACTERIOLOGY, Aug. 1975, p. 497-504 Copyright 0 1975 American Society for Microbiology
Nuclear and Mitochondrial Deoxyribonucleic Acid Replication During Mitosis in Saccharomyces cerevisiae ELISSA P. SENA,I* JULIET W. WELCH, HARLYN
0.
HALVORSON,
AND
SEYMOUR FOGEL
Genetics Department, University of California, Berkeley, California 94720* and Rosensteil Basic Medical Sciences Research Center, Brandeis University, Waltham, Massachusetts 02154
Received for publication 4 March 1975
To study nuclear and mitochondrial deoxyribonucleic acid (DNA) synthesis during the cell cycle, a 15N-labeled log-phase population of Saccharomyces cerevisiae was shifted to 14N medium. After one-half generation, the cells were centrifuged on a sorbitol gradient in a zonal rotor to fractionate the population according to cell size and age into fractions representing the yeast cell cycle. DNA samples isolated from the zonal rotor cell samples were centrifuged to equilibrium in CsCl in an analytical ultracentrifuge to separate the nuclear and mitochondrial DNA components. The amount of 14N incorporated into each 15N-labeled DNA species was measured. The extent of nuclear DNA replication per sample was obtained by measuring the amount of hybrid DNA. The percentage of hybrid nuclear DNA increased from 6 to 68% and then decreased to 44% during the cell cycle. Upon ultracentrifugation, mitochondrial DNA banded as a unimodal peak in all zonal rotor samples. Mitochondrial DNA replication could be ascertained only by the 14N level in each mitochondrial peak and not, as with nuclear DNA, by hybrid DNA level. In contrast to the nuclear incorporation pattern, the 14N percentage in mitochondrial DNA remained effectively constant during the cell cycle. Comparison of the data to theoretical distributions showed that nuclear DNA was replicated discontinuously during the cell cycle, whereas mitochondrial DNA was replicated continuously throughout the entire mitotic cycle.
Among different phylogenetic groups, the relationship between the initiation and termination of nuclear and mitochondrial deoxyribonucleic acid (DNA) synthesis is not constant. Whereas nuclear DNA synthesis is confined within a restricted segment of the normal mitotic cycle, mitochondrial DNA synthesis does not appear to be so restricted. Some organisms synthesize mitochondrial DNA throughout the mitotic cycle; others display a periodicity for its synthesis. Following is a review of the temporal distribution patterns for mitotic mitochondrial DNA synthesis in various organisms. Mitochondrial DNA is synthesized at a relatively constant rate during the entire cell cycle in Physarum (3, 6, 10, 14) and Tetrahymena (5, 21, 22). Mouse fibroblast mitochondrial DNA replicates throughout the cell cycle, although maximal replication appears to occur in early S-phase (19). Depending on the cell synchrony procedure, HeLa cells either replicate mitochondrial DNA in S-phase and G2 or constantly during the cell cycle (23). Euglena gracilis mitochondrial DNA replicates during S-phase I Present address: Biology Department, Case Western Reserve University, Cleveland, Ohio 44106.
(4). Trypanosome kinetoplast DNA replication occurs in midinterphase, partially overlapping S-phase (27). The kinetics of yeast mitochondrial DNA replication appears to vary among genera and within species strains. Earliest studies on an aerobic yeast, Kluyvromyces lactis, indicated that mitochondrial DNA replication was discontinuous during the cell cycle and occurred immediately before S-phase (26). Conflicting conclusions regarding the timing of mitochondrial DNA synthesis in the facultative anaerobe, S. cerevisiae, were published. Cottrell and Avers (8) reported that S. cerevisiae mitochondrial DNA synthesis occurred synchronously before S-phase; however, Williamson and Moustacchi (30) observed continuous synthesis of mitochondrial DNA during mitosis. Continual mitochondrial DNA synthesis, in a diploid, was observed for at least two-thirds of the S. cerevisiae mitotic cycle by Sena (E. P. Sena, Ph.D. thesis, Univ. of Wisconsin, Madison, 1972). Wells (28) concluded that S. cerevisiae mitochondrial DNA is synthesized concomitantly with nuclear DNA, but, Dawes and Carter (9) suggest that mitochondrial DNA synthesis occurs in G2 when parental 497
498
SENA ET AL.
and daughter cells are about equal size. As for S. cerevisiae life cycle phases other than mitosis, mitochondrial DNA synthesis is continual during the mating reaction, first zygotic bud initiation (E. P. Sena, J. Welch, and S. Fogel, manuscript in preparation), and also during early meiosis in both a/a euploid diploids (24) and a/a disomic haploids (17). This paper reports data concerning both nuclear and mitochondrial DNA replication during the S. cerevisiae mitotic budding cycle. DNA replication was monitored by estimating 14N incorporation into '5N-labeled DNA. A 15N-labeled logarithmically growing population shifted to 14N medium for one-half a cell generation was fractionated on a sorbitol gradient in a zonal rotor, a procedure which separates logphase cells according to size and age into fractions spanning the mitotic cycle. The extent of 14N incorporation into the nuclear and mitochondrial DNA in each mitotic cycle fraction was determined. MATERIALS AND METHODS Strain. A haploid single-spore isolate of S. cerevisiae diploid a 5032A x a 5032B (from the Berkeley collection) containing a tryptophan marker was used: a 6B trp-4. This strain was chosen because it readily releases its buds during log phase in liquid minimal medium; after sonic treatment, all cells are either unbudded or possess a single bud. Media. The defined liquid minimal medium (YNB) for culture growth contained (per liter of water); 1.45 g of yeast nitrogen base without amino acids or ammonium sulfate (Difco) supplemented with 50 mg of (15NH4)2SO4 (International Chemical and Nuclear Corp.; 99 atom%), 500 mg of glucose, and 30 mg of L-tryptophan. Cells were grown at 30 C. Density label shift. About 1.8 x 107 cells, grown in 100 ml of YNB for eight generations, were transferred to 3 liters of YNB in a 6-liter flask and grown 16 h on a rotory shaker. When the cell concentration reached 4.3 x 10 cells/ml, 90 ml of 10% (wt/vol) (14NH4)2S04 was added. After an additional 1 h of growth (one-half generation), direct addition of ice to the flask cooled the culture to 4 C. The cells were harvested by centrifugation at 4 C, washed once with distilled water, and resuspended in 20 ml of cold distilled water. Cell population fractionation. The 20-ml cell suspension was sonically treated at 50 W for 10 s to disperse cell clumps (Heat Systems Sonifier) and immediately layered onto a sterile 8 to 35% sorbitol gradient in an MSE zonal rotor (1,300 ml) at 4 C and centrifuged at 1,200 rpm for 10 min. Halvorson and co-workers (11) have published a detailed description of zonal rotor methodology for yeast cell separations. The use of a displacement pump and an automatic fraction collector aided the collection of 87 15-ml fractions from the gradient. Consecutive fractions
J. BACTERIOL.
containing cells were pooled, washed, and centrifuged to yield nine experimental cell samples. Cell pellets were frozen at -10 C. DNA isolation. Cell samples, containing approximately 5 x 10' cells, were suspended in 2 ml of SES (0.15 M NaCl, 0.1 M ethylenediaminetetraacetic acid, and 1 M sorbitol, pH 8), 0.2 ml of mercaptoethanolamine (0.3 M), 0.05 ml of Glusulase (Endo Laboratories), and incubated 1 h at 37 C. The samples were then centrifuged and washed twice with SES. Cell pellets were vortexed in 0.5 ml of 0.1 M NaCl and 0.02 M ethylenediaminetetraacetic acid, pH 7, after which 0.1 ml of a detergent mixture (2) was added, mixed gently, and incubated at 37 C for 10 min. After addition of 45 mg of solid NaCl and 0.2 ml of previously heated (80 C for 10 min) ribonuclease solution (0.5 mg of pancreatic ribonuclease per ml [Calbiocheml and 15 U of T, ribonuclease per ml [Worthington Biochemicals Corp. ]), the mixture was incubated at 37 C for 1 h. The solution was then dialyzed against SSC (0.15 M NaCl and 0.015 M sodium citrate, pH 7) for 2 to 3 h at 25 C. Analytical ultracentrifugation in CsCl. The method used has been described by Meselson et al. (20). Enough solid CsCl and SSC were added to DNA samples containing 2 to 5 ,g of yeast DNA and about 1.3 gg of Micrococcus lysodeikticus (M. aureus var. Iysodeikticus) DNA to bring the refractive index to 1.400 and volume to 0.6 to 0.7 ml. After determining the refractive index (25 C) of each sample with an Atago refractometer, the samples were centrifuged for 18 to 24 h at 44,770 rpm in a Spinco model E ultracentrifuge. Exposures at 30 and 60 s of each DNA sample were taken using an ultraviolet light source and Kodak commercial film. The ultraviolet photographic DNA banding patterns were converted to absorbance profiles with a Joyce-Loebl microdensitometer at 1:5, 1:10, and 1:20 arm ratios. Chemical DNA measurement. DNA content per cell was determined by the 3,5-diaminobenzoic acid dihydrochloride (DABA) assay. The following DABA procedure is a modification of two currently used methods (15; C. Milne, M.S. thesis, University of Washington, Seattle, 1972). Washed-cell samples were carefully counted in a hemocytometer; unbudded and budded cells were counted as 1 cell unit. Duplicates of each cell sample were assayed. Approximately 108 cells were placed in tapered, siliconecoated centrifuge tubes (12) and suspended in 0.2 ml of 1 N NaOH at 25 C for 16 to 24 h. After neutralizing the base with 0.2 ml of 1 N HCI, the tubes were placed in ice and 0.5 ml of 10W. trichloroacetic acid was added at 0 C. After a 5-min incubation, the samples were centrifuged at 0 C to pellet the cells and the supernatant liquid was carefully removed with a Pasteur pipette. The samples were then treated as follows. They were washed once with 0.5 ml of 51'7 cold trichloroacetic acid, twice with 0.1 M potassium acetate in 95%7c ethanol, and once with absolute ethanol. A fresh Norite A-DABA (99(' , Aldrich) suspension (0.4 g of DABA/ml) was prepared and filtered through a 0.3-Am cellulose filter (Millipore Corp.) at 25 C. Before 100 jul of filtered DABA reagent
VOL. 123, 1975
NUCLEAR AND MITOCHONDRIAL DNA REPLICATION
was added, the cell pellets were thoroughly dried in a boiling-water bath (10 to 15 min). After incubating the samples at 60 C for 30 min with DABA reagent, 3 ml of 1 N HCl was added. The fluorescence of each sample was measured in an Aminco-Bowman spectrophotofluorometer (at 423 nm excitation and 519 nm emission) and compared to calf thymus DNA (Type V, Sigma Chemical Co.) standards. Analysis of percentage of nuclear and mitochondrial DNA per sample. The 1:10 or 1:20 arm ratio microdensitometer tracings were used to calculate the mitochondrial DNA percentage, or l"N: 5N- and '5N: 4N-labeled nuclear DNA per sample. Overlapping nuclear DNA bands were resolved by eye. Care was taken to retain the gaussian shape of the components. Tracings of each DNA band were cut and weighed to determine their percentage per sample. Only photographs of DNA samples within the linear range of the film were used.
RESULTS To study nuclear and mitochondrial DNA synthesis during the cell cycle, a "IN-labeled log-phase population of S. cerevisiae was shifted to 14N medium. After a time lapse equal to one-half generation (60 min), the cells were centrifuged on a sorbitol gradient in a zonal rotor to fractionate the population according to cell size and age (25). This fractionation does not perturb the cell's biochemical balance and yields cell samples sufficient for DNA isolation. Consecutive gradient fractions were pooled into nine samples, containing at least 4 x 108 cells/sample, to facilitate subsequent DNA isolation. The cell number per sample varied as shown in Fig. 1. The distribution of cell number within the rotor is very similar to a coulter counter cell size distribution plot of the unfractionated log-phase population (Fig. 1). The close identity between the two distributions strongly suggests that zonal rotor fractionation separated cells according to their size. Further evaluation of fractionation was obtained from the distributions of DNA content per cell and percent of budded cells across the rotor (Fig. 2). Sample 1 contained 94% unbudded cells, budding increased in samples 2 through 5 and stabilized at 86% budded cells in samples 6 through 9. DNA content per haploid cell unit increased from 18.5 fg to 34 fg (1 femtogram = 10-15 g) across the rotor. Since ex-
cellent morphological separation was achieved, and DNA synthesis displayed kinetics commonly observed for a synchronous yeast population (31), we assumed that the successive rotor fractions adequately represent the continuum of the yeast cell cycle. Chemical estimations of total DNA per cell, such as those in Fig. 2, cannot accurately dis-
499
Window 5
25
15 : 1:2:3:4:5:6.7:8.8
9
2 a
I°
E co
0
x
u)
\O
1 * * ** \-* **...0 :/
0:::\::.
./ 1::.
*v
40
20
Fraction
FIG. 1. A comparison of cell number per zonal a size distribution plot of a log-phase S. cerevisiae culture. An exponentially growing population of S. cerevisiae was fractionated on a 8 to 35% sorbitol gradient in an MSE zonal rotor as described. Consecutive tubes containing cells were pooled into nine samples. The number of cells per pooled sample is plotted at each sample midpoint (0). The size distribution was generated by a coulter counter fitted with a 100-,gm orifice and a particle size distribution plotter with 25 windows (0). rotor sample to
tinguish nuclear from mitochondrial DNA replication during the cell cycle. However, estimates of nuclear and mitochondrial DNA synthesis can be obtained from the amount of '4N incorporated into each "5N-labeled DNA species. To determine 14N incorporation
into '5N nuclear and mitochondrial DNA fractions, DNA samples were centrifuged to equilibrium in CsCl in an analytical ultracentrifuge to separate the nuclear and mitochondrial DNA components. The buoyant densities of both nuclear and mitochondrial DNA peaks were calculated from microdensitometer tracings of their ultraviolet photographs. In all, 11 DNA samples were analyzed: (i) the initial log-phase '5N-labeled population; (ii) the unfractionated population after onehalf generation growth in '4N medium; and
(iii) the nine samples obtained from the zonal rotor. The "IN-labeled cells contained nuclear DNA of buoyant density 1.715 g/cm5 and mitochondrial DNA of density 1.699 g/cm' (Fig. 3). After the cell population had replicated for a period equal to one-half a cell generation in '4N medium, hybrid ('5N: "4N) nuclear DNA (p = 1.708 g/cm) appeared and the mitochondrial DNA peak shifted to 1.696 g/cm3. Figure 4 shows the densitometer tracings obtained from the nine zonal rotor samples. The average mitochondrial peak density for all nine samples
500
SENA ET AL.
30 0
z 0
20 20
~ ~I
J. BACTERIOL. 100
-) 50 0o
co .o~
40
Fraction
FIG. 2. The percentage of budding cells and DNA per cell unit for each zonal rotor sample. Duplicate DNA measurements (0) were obtained with the fluorometric DABA method described. A hemocytometer was used to determine the percentage of budding cells (0) per sample.
was 1.696 g/cm3, though systematic variation in the amount of heavy (15N: 15N) and hybrid (l5N: 4N)-labeled nuclear DNA is apparent in these samples. The extent of nuclear DNA replication per sample during the one-half generation incorporation of 14N, was obtained by measuring the amount of hybrid DNA. Figure 5 shows that the percentage of hybrid nuclear DNA increased from 6 to 68% and then decreased to 44% during the cell cycle. Since heavy and hybrid peaks were not seen in any zonal rotor mitochondrial DNA sample, mitochondrial DNA replication could be ascertained only by the 14N level in each mitochondrial peak. In contrast to the nuclear incorporation pattern, the '4N percentage in mitochondrial DNA remained effectively constant during the cell cycle (Fig. 6).
DISCUSSION The usefulness of the zonal rotor procedure for cell cycle studies relies upon a successful fractionation across the rotor. If the consecutive zonal rotor samples represent increasing cell age in the cell cycle, they should display the bud initiation distribution and a discontinuous DNA synthetic pattern observed in a synchronous population. Since zonal rotor samples analyzed here and elsewhere (25) satisfy these criteria (Fig. 1, 2), we conclude that they represent adequately the yeast cell budding cycle. Therefore, they may be used to study the timing of DNA replication during the cell cycle. Nuclear DNA synthesis. Although a discontinuous nuclear DNA synthetic pattern based on measurements of total DNA per cell across the mitotic cycle is well documented, studies on nuclear DNA replication during the yeast cell cycle based on density-shift techniques are reported only by Sena (Ph.D. thesis, 1972). To
assist the analysis of the observed hybrid nuclear DNA distribution across the zonal rotor, a theoretical hybrid nuclear DNA distribution across the cell cycle was generated. This theoretical distribution predicts that: (i) unbudded cells, in G2 at the time of label shift, should not contain hybrid nuclear DNA; (ii) cells entering S-phase during 14N addition should show increasing levels of hybrid nuclear DNA; (iii) cells completing the entire S-phase during the density shift should contain only hybrid DNA; and (iv) those completing S-phase when '4N was
U
CD co
VI .
1-IN
h0 140 Lq P-
O
Density g/cc FIG. 3. Densitometer tracings of DNA from exponentially growing cultures of S. cerevisiae. A cell culture grown 12 generations in (I5NH4)2SO4 containing medium was shifted for one-half generation to ( 14NH4)2SO4 medium. DNA from cell samples removed before (upper tracing) and after the shift (lower tracing) were centrifuged to equilibrium in CsCI as described. The density of heavy-labeled nuclear DNA (15N: 15N) is 1.715 g/cm 3, whereas heavylabeled mitochondrial DNA is 1.699 g/cms. The hybrid nuclear band (15NW1"N) density is 1. 708 g/cm 3, whereas the mitochondrial peak is 1.696 g/cm3. M. lysodeikticus DNA (1.731 g/cm,) was used as marker.
NUCLEAR AND MITOCHONDRIAL DNA REPLICATION
VOL. 123, 1975 4
C0 5
U.
O
7
i:;
8
6
D
9
r-
/D
clear DNA across the cell cycle. Little is known about DNA precursor pool fluctuations during the mitotic cycle, therefore, accurate predictions of "4N incorporation lag into nuclear DNA during this experiment cannot be made at this time. Perhaps future experiments will shed some light on these observa'tions. Mitochondrial DNA Synthesis. In contrast to the nuclear 14N incorporation pattern which displayed a large fluctuation across the rotor, mitochondrial DNA "4N incorporation was relatively constant through the cell cycle (Fig. 6). The line depicted in Fig. 6 was calculated by the method of least squares. It does not differ significantly from the line drawn through the mean of all the mitochondrial incorporation values (which has a slope of zero). Figure 7 compares these data with "4N levels expected across the cell cycle for four different mitochondrial DNA replication patterns: (i) mitochondrial DNA replication in GI only; (ii) mitochondrial DNA replication in S-phase only; (iii) mitochondrial DNA replication in G, only; or (iv) mitochondrial DNA replication throughout the cell cycle. The three confined and discontin-
Density g/cc FIG. 4. Densitometer tracings of DNA isolated from nine pooled zonal rotor cell samples. An exponentially growing "5N-labeled cell population shifted to "N medium for one-half generation was fractionated into nine samples on a sorbitol gradient in a zonal rotor. DNA isolated from each sample was centrifuged to equilibrium in CsCI. Densitometer tracings of the ultraviolet photographs of each sample are shown. For reference, sample I contained 95% unbudded cells. Budding increased in samples 2 through 6, and maximized at 85% budded cells in samples 7 through 9 (see Fig. 2).
added should contain decreasing levels of DNA hybrid. Accordingly, the theoretical hybrid nuclear DNA distribution (Fig. 5) should be bell shaped. Whereas the actual hybrid nuclear DNA levels show a distribution similar to the theoretical expectation, only the early cell cycle values correspond with those predicted. Since the theoretical hybrid nuclear DNA distribution was calculated assuming no "4N incorporation lag into 15N DNA, and there was a one-fourth generation lag ( "5N-labeled cells grown for onehalf generation in "4N medium should contain 66% hybrid nuclear DNA, however, the unfractionated cell population contained only 40% hybrid nuclear DNA), the discrepancy between predicted and observed values probably reflects the "4N incorporation lag into "5N-labeled nuclear DNA. Possibly, the fractionation of the cell population on the zonal rotor unmasked the lag distribution for "4N incorporation into nu-
501
0
z
a
0\
c;
z
z
0 5010
uzz
P* 20
40
Fraction
FIG. 5. Distribution of hybrid nuclear DNA during the cell cycle. Shown is the percentage of "5N: 4N
nuclear DNA per zonal sample (0) versus the hybrid values predicted across the rotor (0). The sample percentages were estimated according to the procedure outlined. The theoretical hybrid DNA distribution expected for an exponentially growing cell population shifted from "N to 14N for one-half generation before fractionation on a zonal rotor was calculated based on the following assumptions: (i) fraction number was proportional to cell age; (ii) nuclear DNA was replicated at a constant rate only in samples 2 through 5 (or 0.13 to 0.47 of the cell cycle); and (iii) there was no 14N incorporation lag into nuclear DNA at any portion of the cell cycle. Therefore, cells at the end of the cycle in fraction 44 were at approximately the same age as cells presently in fraction 26 (0.50 generation younger) when 14N was added. Since nuclear DNA was not synthesized between fractions 26 to 44, no 14N incorporation in fraction 44 is predicted. Other theoretical points were calculated in the
same manner.
502
SENA ET AL.
J. BACTERIOL.
higher than the observed values suggesting
z
30 0
0
20~
40
Fraction
0-0
FIG. 6. Percentage of 14N in mitochondrial DNA in zonal rotor samples. The 14N content of each mitochondrial DNA peak was calculated from the buoyant density of each sample assuming 1.699 g/cm 3 as 100% I N and 1.684 g/cm3 as 100%07o "N DNA. The standard error for each calculation is included.
.X Z
30-
10
20
40
Fraction
FIG. 7. 14N incorporation into mitochondrial DNA during the cell cycle. The amount of 14N in each zonal rotor sample (-) (from Fig. 6) is compared with the pattern predicted by the four major temporal patterns of mitochondrial DNA replication during the cell cycle. (i) Mitochondrial DNA replication during G, only (O); (ii) during S-phase only (0); (iii) in G2 only (A); or (iv) during the complete cell cycle (x). The following assumptions were made for these theoretical distributions. (i) Fraction number was proportional to cell age; (ii) nuclear DNA replication occurred in fractions 2 through 5; (iii) mitochondrial DNA was replicated completely and at a constant rate only during the cell cycle stage in question; and (iv) there was no 14N incorporation lag into mitochondrial DNA at any portion of the cell cycle. For example, cells in fraction 44 were about the age of cells in fraction 26 (0.50 generation younger) when the medium was changed, and were already through G1 and S, and 3.9% of G2. If mitochondrial DNA were replicated only in G1 or during S-phase, fraction 44 should contain 0% 14N. If replication occurred only in G2, 48% 14N was expected; while 33% 14N was predicted if mitochondrial DNA replication were continuous throughout the cell cycle. The other theoretical values were calculated in the same manner.
uous replication patterns predict wide fluctuations in "4N incorporation during the cell cycle, while continuous mitochondrial DNA replication incorporates a constant amount of' 14N throughout the cell cycle. Our data (Fig. 6, 7) appears to fit the continuous model. However, the 14N levels predicted for each sample are
a
14N incorporation lag into "5N-labeled mitochondrial DNA in every cell fraction. The lag seen in each fraction is about the same as the incorporation lag of one-fourth of a generation in the unfractionated cell population. ('5Nlabeled cells grown for one-half generation in "4N medium should contain 66% '5N: 14N or in this case 3367c '4N, whereas only 20% 14N was
observed.)
Could the 14N percentage in mitochondrial DNA across the rotor be spurious, in that '4N was incorporated during the zonal rotor isolation procedure? Because the mitochondrial DNA from the zonal samples averaged the same amount of' 14N as the initial unfractionated cell population, extensive 14N incorporation during zonal fractionation did not occur. (A similar experiment using cells frozen before zonal fractionation showed the same mitochondrial DNA 14N incorporation pattern across the rotor [Sena, Ph.D. thesis, 1972].) We believe the constant amount of' label incorporation represents continuous synthesis of mitochondrial DNA during the cell cycle. To insure that mitochondrial DNA was not selectively lost from the zonal samples during DNA isolation or diluted out during cell growth in 14N, the percentage of' mitochondrial DNA per sample was determined. The sample percentages and those predicted for a continuous pattern of mitochondrial DNA synthesis during the cell cycle are in agreement (Fig. 8). The fluctuations between the two curves are well within the standard error common for densitometer tracing analysis. Therefore the 14N incorporation data do not reflect either selective mitochondrial DNA loss during isolation or cell growth. However, it is still conceivable that small fluctuations in the mitochondrial DNA replication rate during the cell cycle could go undetected by our methods. Since the error in DNA density measurements is _40.001 g/cm3 (Fig. 6), rate fluctuations up to ±30'Yc during one-half' generation would fall within these error limits and be undetected. If we assume that mitochondrial DNA has a molecular weight of 50 >, 106 (13), unbudded cells, in the present work, contain about 29 molecules of' mitochondrial DNA. Continual, relatively constant, 14N incorporation into this mitochondrial DNA population could occur in two basic ways. First, the replication time for each individual molecule (not known at present) could span the cell cycle. Second, certain molecules in the mitochondrial DNA population (for instance, in dif'ferent mitochondria)
VOL. 123, 1975
NUCLEAR AND MITOCHONDRIAL DNA REPLICATION
503
content, malate dehydrogenase and cytochrome oxidase activity in synchronous cells, Cottrell 1AC e 14 and Avers (8) concluded that mitochondrial V DNA was synthesized immediately before the S-period. Since malate dehydrogenase activity is a composite of both cytoplasmic and mitochondrial enzyme activity, and is coded by the 20 40 -0-0 nucleus (18), periodic enzyme activity during Fraction the cell cycle does not necessarily reflect periFIG. 8. Mitochondrial percentage per zonal rotor sample. Shown is the percentage mitochondrial DNA odic mitochondrial growth. A similar argument per sample measured as described (0). The standard applies for cytochrome oxidase; although the error for each measurement is included. The predicted enzyme is localized in mitochondria, it is compercentages (0) are calculated assuming (i) cell age is posed of both nuclear and mitochondrially proportional to fraction number; (ii) nuclear DNA coded components (7, 16). Furthermore, no data synthesis occurred at a constant rate only in samples concerning the actual amount or percentage 2 through 5 (or 0.13 to 0.47 of the cell cycle); and (iii) mitochondrial DNA during the cell cycle was mitochondrial DNA replication occurred at a constant presented by these authors. Accordingly, we rate throughout the cell cycle. The predictions are suggest that their conclusions concerning the obtained by multiplying the ratio between the calculated femtograms of mitochondrial DNA per sample temporal mitochondrial DNA replication patand the calculated femtograms of total DNA per tern are at best tentative or somewhat premasample by 100. The initial amount of mitochondrial ture in view of the data reported. By measuring DNA in sample 1 is assumed to be 2.4 fg/unbudded the 3H and 32p label incorporation into "4Ccell, or 13%o (Fig. 8) of 18.5 fg/cell (Fig. 2). For labeled nuclear and mitochondrial DNA during example, cells in fraction 26 have completed all the cell cycle, Wells (28) concluded that mitonuclear DNA synthesis but only 0.54 of mito- chondrial DNA was synthesized concomitantly chondrial DNA synthesis, therefore they contain with nuclear DNA. However, closer scrutiny of 3.7/35.9 x 100 or 10.3%o mitochondrial DNA. All the the results indicates that mitochondrial DNA other calculations were done similarly. incorporated some label throughout the cell cycle. It is possible that his particular strain could replicate at discrete intervals, but at the does show an increased level of mitochondrial total population level replication would span DNA synthesis in S-phase. It may be emphathe entire mitotic cycle. Because mitochondrial sized that our data does not rule out this DNA replication in S. cerevisiae is accompanied possibility, but it is also true that not all by a dispersive mechanism (Sena, Ph.D. thesis, synthesis occurs in S-phase. Additionally, if all 1972; 29) and no discrete 15N:'5N or 15N:14N mitochondrial DNA is synthesized during Speaks are observed from mitochondrial DNA phase, as concluded by Wells, the mitochonwhich has incorporated 14N when the DNA is drial DNA percentage should remain relatively banded isopycnically, our data cannot discrimi- constant during the cell cycle, and not increase nate between these alternatives. That we see no in the later cell cycle fractions, as was pub15N: 15N shoulder on any mitochondrial peak lished. We feel there are alternative conclusions (even when the CsCl gradients are over-loaded to the data presented. A novel approach to for the nuclear component) suggests that label studying DNA synthesis was taken by Dawes dispersion is very rapid. Because all the zonal and Carter (9). They used the level of nisamples displayed a unimodal peak (Fig. 4), the trosoguanidine-induced cytoplasmically inherunderlying dispersive mechanism(s) (sister- ited oligomycin- and erythromycin-resistant strand exchanges, intramolecular exchanges, mutants to probe mitochondrial DNA replicaintermolecule exchanges, and/or repair mech- tion during the cell cycle. The basic assumption anisms) probably occur throughout the cell in this approach is that nitrosoguanidine mutagenesis occurs principally at the replication cycle. A continuous mode of mitochondrial DNA fork. However, because nitrosoguanidinesynthesis during mitosis in S. cerevisiae has induced mutations in eukaryotes may arise now been observed in three different yeast independently of DNA replication (1), the data strains (Sena, Ph.D. thesis, 1972; 30; this pa- of Dawes and Carter may not only indicate per). Can these conclusions be reconciled with DNA replication, but also other cell cycle pheother conflicting temporal distribution patterns nomena such as DNA repair etc. The previous discussion indicates that nureported for mitochondrial DNA synthesis during mitosis? By measuring whole cell DNA clear and mitochondrial DNA replication are z
Q
504
SENA ET AL.
temporally dissociated during mitosis in S. cerevisiae. In addition, mitochondrial DNA replication occurs continuously during all major decision points in the S. cerevisiae life cycle. Since nuclear synthesis is always restricted to a discrete time period while mitochondrial DNA synthesis occurs continuously, their synthesis appear to be controlled by different mech-
J. BACTERIOL. 13. Hollenberg, C. P., P. Borst, and F. F. J. Van Bruggen. 1970. Mitochondrial DNA. V. A 25 closed circular duplex DNA molecule in wild-type yeast mitochondria. 14.
15.
anisms. ACKNOWLEDGMENTS We like to thank R. K. Mortimer for use of the model E analytical ultracentrifuge in Donner Laboratory and the Department of Molecular Biology for use of their Joyce Loebl Microdensitometer. The helpful suggestions of Douglas Campbell are acknowledged. This research was supported in part by Public Health Service research grants GM-17317 to S. Fogel from the National Institute of General Medical Sciences and Al010610 to H. 0. Halvorson from the National Institute of Allergy and Infectious Diseases. Preliminary investigation was carried out at the University of Wisconsin during the tenure of a predoctoral fellowship (E.S.) from the Public Health Service. LITERATURE CITED 1. Auerbach, C., and B. J. Kilbey. 1971. Mutation in eukaryotes. Annu. Rev. Gen. 5:163-218. 2. Blamire, J., D. R. Cryer, D. B. Finkelstein, and J. Marmur. 1972. Sedimentation properties of yeast nuclear and mitochondrial DNA. J. Mol. Biol. 67:11-24. 3. Braun, R., and T. E. Evans. 1969. Replication of nuclear satellite and mitochondrial DNA in the mitotic cycle of Physarum. Biochim. Biophys. Acta 182:511-522. 4. Calvayrac, R., R. A. Butow, and M. Lefort-tran. 1972. Cyclic replication of DNA and changes in mitochondrial morphology during the cell cycle of Euglena gracilis. Exp. Cell Res. 71:422-432. 5. Cameron, I. L. 1966. A periodicity of tritiated-thymidine incorporation into cytoplasmic deoxyribonucleic acid during the cell cycle of Tetrahymena pyriformis. Nature (London) 209:630-631. 6. Charret, R., and J. Andre. 1968. La synthese de l'ADN mitochondrial chez Tetrahymena pyriformis. J. Cell Biol. 39:369-381. 7. Chen, W. L., and F. C. Charalampous. 1969. Mechanism of induction of cytochrome oxidase in yeast. I. Kinetics of induction and evidence for accumulation of cytoplasmic and mitochondrial precursors. J. Biol. Chem. 244:2767-2776. 8. Cottrell, S. F., and C. J. Avers. 1970. Evidence of mitochondrial synchrony in synchronous cultures of yeast. Biochem. Biophys. Res. Commun. 38:973-980. 9. Dawes, I. W., and B. L. A. Carter. 1974. Nitrosoguanidine mutagenesis during nuclear and mitochondrial gene replication. Nature (London) 250:709-712. 10. Guttes, E. W., C. Hanawalt, and S. Guttes. 1967. Mitochondrial DNA synthesis and the mitotic cycle in Physarum polycephalum. Biochim. Biophys. Acta 142:181-194. 11. Halvorson, H. O., B. L. A. Carter, and P. Tauro. 1968. Use of synchronous cultures of yeast to study gene position, p. 462-470. In L. Grossman and K. Moldave (ed.), Methods in enzymology, nucleic acids, vol. 21. Academic Press Inc., New York. 12. Hinegardner, R. T. 1971. An improved fluorometric assay for DNA. Anal. Biochem. 39:197-201.
16.
17. 18. 19.
20.
21.
22. 23.
24.
25.
26.
27. 28.
29.
30.
31.
Structure and genetic complexity. Biochim. Biophys. Acta 209:1-15. Holt, C. E., and E. G. Gurney. 1969. Minor components of the DNA of Physarum polycephalum. Cellular location and metabolism. J. Cell Biol. 40:484-496. Kissane, J. M., and E. Robbins. 1958. The fluorometric measurement of deoxyribonucleic acid in animal tissues with special reference to the central nervous system. J. Biol. Chem. 233:184-188. Kraml, J., and H. R. Mahler. 1967. Biochemical correlates of respiratory deficiency. VII. A precipitating antiserum against cytochrome oxidase of yeast and its use in the study of respiratory deficiency. Immunochemistry 4:213-226. Kuenzi, M. T., and R. Roth. 1974. Timing of mitochondrial DNA synthesis during meiosis in Saccharomyces cerevisiae. Exp. Cell Res. 85:377-382. Longo, G. P., and J. G. Scandalios. 1969. Nuclear gene control of mitochondrial malic defiydrogenase in maize. Proc. Natl. Acad. Sci. U.S.A. 62:104-111. Madreiter, H., C. Mittermayer, and 0. Rainhardt. 1972. 3H-thymidine incorporation into mitochondria of synchronized mouse fibroblasts. Beitr. Pathol. 145:249-255. Meselson, M., F. W. Stahl, and J. Vinograd. 1957. Equilibrium sedimentation of macromolecules in density gradients. Proc. Natl. Acad. Sci. U.S.A. 43:581-588. Parsons, J. A. 1965. Mitochondrial incorporation of tritiated thymidine in Tetrahymena pyriformis. J. Cell Biol. 25:641-646. Parsons, J. A., and R. C. Rustad. 1968. The distribution of DNA among dividing mitochondria of Tetrahymena pyriformis. J. Cell Biol. 37:683-693. Pica-Mattoccia, L., and G. Attardi. 1972. Expression of the mitochondrial genome in HeLa cells IX. Replication of mitochondrial DNA in relationship to the cell cycle in HeLa cells. J. Mol. Biol. 64:465-484. Pinon, R., Y. Salts, and G. Simchen. 1974. Nuclear and mitochondrial DNA synthesis during yeast sporulation. Exp. Cell Res. 83:231-238. Sebastian, J., B. L. A. Carter, and H. 0. Halvorson. 1971. Use of yeast populations fractionated by zonal centrifugation to study the cell cycle. J. Bacteriol. 108:1045-1050. Smith, D., P. Tauro, E. Schweizer, and H. 0. Halvorson. 1968. The replication of mitochondrial DNA during the cell cycle in Saccharomyces lactis. Proc. Natl. Acad. Sci. U.S.A. 60:936-942. Steinert, M., and G. Steinert. 1962. La synthese de l'acide desoxyribonucleic au cours du cycle de division de Trypanosoma mega. J. Protozool. 9:203-211. Wells, J. R. 1974. Mitochondrial DNA synthesis during the cell cycle of Saccharomyces cerevisiae. Exp. Cell Res. 85:278-286. Williamson, D. H., and D. J. Fennell. 1974. Apparent dispersive replication of yeast mitochondrial DNA as revealed by density labelling experiments. Mol. Gen. Genet. 131:193-207. Williamson, D. H., and E. Moustacchi. 1971. The synthesis of mitochondrial DNA in synchronously dividing cultures of Saccharomyces cereuisiae. Biochem. Biophys. Res. Commun. 42:195-201. Williamson, D. H., and A. W. Scopes. 1960. The behavior of nucleic acids in synchronously dividing cultures of Saccharomyces cereuisiae. Exp. Cell Res. 20:338-349.
JOURNAL OF BACTERIOLOGY, Aug. 1975, p. 497-504 Copyright 0 1975 American Society for Microbiology
Nuclear and Mitochondrial Deoxyribonucleic Acid Replication During Mitosis in Saccharomyces cerevisiae ELISSA P. SENA,I* JULIET W. WELCH, HARLYN
0.
HALVORSON,
AND
SEYMOUR FOGEL
Genetics Department, University of California, Berkeley, California 94720* and Rosensteil Basic Medical Sciences Research Center, Brandeis University, Waltham, Massachusetts 02154
Received for publication 4 March 1975
To study nuclear and mitochondrial deoxyribonucleic acid (DNA) synthesis during the cell cycle, a 15N-labeled log-phase population of Saccharomyces cerevisiae was shifted to 14N medium. After one-half generation, the cells were centrifuged on a sorbitol gradient in a zonal rotor to fractionate the population according to cell size and age into fractions representing the yeast cell cycle. DNA samples isolated from the zonal rotor cell samples were centrifuged to equilibrium in CsCl in an analytical ultracentrifuge to separate the nuclear and mitochondrial DNA components. The amount of 14N incorporated into each 15N-labeled DNA species was measured. The extent of nuclear DNA replication per sample was obtained by measuring the amount of hybrid DNA. The percentage of hybrid nuclear DNA increased from 6 to 68% and then decreased to 44% during the cell cycle. Upon ultracentrifugation, mitochondrial DNA banded as a unimodal peak in all zonal rotor samples. Mitochondrial DNA replication could be ascertained only by the 14N level in each mitochondrial peak and not, as with nuclear DNA, by hybrid DNA level. In contrast to the nuclear incorporation pattern, the 14N percentage in mitochondrial DNA remained effectively constant during the cell cycle. Comparison of the data to theoretical distributions showed that nuclear DNA was replicated discontinuously during the cell cycle, whereas mitochondrial DNA was replicated continuously throughout the entire mitotic cycle.
Among different phylogenetic groups, the relationship between the initiation and termination of nuclear and mitochondrial deoxyribonucleic acid (DNA) synthesis is not constant. Whereas nuclear DNA synthesis is confined within a restricted segment of the normal mitotic cycle, mitochondrial DNA synthesis does not appear to be so restricted. Some organisms synthesize mitochondrial DNA throughout the mitotic cycle; others display a periodicity for its synthesis. Following is a review of the temporal distribution patterns for mitotic mitochondrial DNA synthesis in various organisms. Mitochondrial DNA is synthesized at a relatively constant rate during the entire cell cycle in Physarum (3, 6, 10, 14) and Tetrahymena (5, 21, 22). Mouse fibroblast mitochondrial DNA replicates throughout the cell cycle, although maximal replication appears to occur in early S-phase (19). Depending on the cell synchrony procedure, HeLa cells either replicate mitochondrial DNA in S-phase and G2 or constantly during the cell cycle (23). Euglena gracilis mitochondrial DNA replicates during S-phase I Present address: Biology Department, Case Western Reserve University, Cleveland, Ohio 44106.
(4). Trypanosome kinetoplast DNA replication occurs in midinterphase, partially overlapping S-phase (27). The kinetics of yeast mitochondrial DNA replication appears to vary among genera and within species strains. Earliest studies on an aerobic yeast, Kluyvromyces lactis, indicated that mitochondrial DNA replication was discontinuous during the cell cycle and occurred immediately before S-phase (26). Conflicting conclusions regarding the timing of mitochondrial DNA synthesis in the facultative anaerobe, S. cerevisiae, were published. Cottrell and Avers (8) reported that S. cerevisiae mitochondrial DNA synthesis occurred synchronously before S-phase; however, Williamson and Moustacchi (30) observed continuous synthesis of mitochondrial DNA during mitosis. Continual mitochondrial DNA synthesis, in a diploid, was observed for at least two-thirds of the S. cerevisiae mitotic cycle by Sena (E. P. Sena, Ph.D. thesis, Univ. of Wisconsin, Madison, 1972). Wells (28) concluded that S. cerevisiae mitochondrial DNA is synthesized concomitantly with nuclear DNA, but, Dawes and Carter (9) suggest that mitochondrial DNA synthesis occurs in G2 when parental 497
498
SENA ET AL.
and daughter cells are about equal size. As for S. cerevisiae life cycle phases other than mitosis, mitochondrial DNA synthesis is continual during the mating reaction, first zygotic bud initiation (E. P. Sena, J. Welch, and S. Fogel, manuscript in preparation), and also during early meiosis in both a/a euploid diploids (24) and a/a disomic haploids (17). This paper reports data concerning both nuclear and mitochondrial DNA replication during the S. cerevisiae mitotic budding cycle. DNA replication was monitored by estimating 14N incorporation into '5N-labeled DNA. A 15N-labeled logarithmically growing population shifted to 14N medium for one-half a cell generation was fractionated on a sorbitol gradient in a zonal rotor, a procedure which separates logphase cells according to size and age into fractions spanning the mitotic cycle. The extent of 14N incorporation into the nuclear and mitochondrial DNA in each mitotic cycle fraction was determined. MATERIALS AND METHODS Strain. A haploid single-spore isolate of S. cerevisiae diploid a 5032A x a 5032B (from the Berkeley collection) containing a tryptophan marker was used: a 6B trp-4. This strain was chosen because it readily releases its buds during log phase in liquid minimal medium; after sonic treatment, all cells are either unbudded or possess a single bud. Media. The defined liquid minimal medium (YNB) for culture growth contained (per liter of water); 1.45 g of yeast nitrogen base without amino acids or ammonium sulfate (Difco) supplemented with 50 mg of (15NH4)2SO4 (International Chemical and Nuclear Corp.; 99 atom%), 500 mg of glucose, and 30 mg of L-tryptophan. Cells were grown at 30 C. Density label shift. About 1.8 x 107 cells, grown in 100 ml of YNB for eight generations, were transferred to 3 liters of YNB in a 6-liter flask and grown 16 h on a rotory shaker. When the cell concentration reached 4.3 x 10 cells/ml, 90 ml of 10% (wt/vol) (14NH4)2S04 was added. After an additional 1 h of growth (one-half generation), direct addition of ice to the flask cooled the culture to 4 C. The cells were harvested by centrifugation at 4 C, washed once with distilled water, and resuspended in 20 ml of cold distilled water. Cell population fractionation. The 20-ml cell suspension was sonically treated at 50 W for 10 s to disperse cell clumps (Heat Systems Sonifier) and immediately layered onto a sterile 8 to 35% sorbitol gradient in an MSE zonal rotor (1,300 ml) at 4 C and centrifuged at 1,200 rpm for 10 min. Halvorson and co-workers (11) have published a detailed description of zonal rotor methodology for yeast cell separations. The use of a displacement pump and an automatic fraction collector aided the collection of 87 15-ml fractions from the gradient. Consecutive fractions
J. BACTERIOL.
containing cells were pooled, washed, and centrifuged to yield nine experimental cell samples. Cell pellets were frozen at -10 C. DNA isolation. Cell samples, containing approximately 5 x 10' cells, were suspended in 2 ml of SES (0.15 M NaCl, 0.1 M ethylenediaminetetraacetic acid, and 1 M sorbitol, pH 8), 0.2 ml of mercaptoethanolamine (0.3 M), 0.05 ml of Glusulase (Endo Laboratories), and incubated 1 h at 37 C. The samples were then centrifuged and washed twice with SES. Cell pellets were vortexed in 0.5 ml of 0.1 M NaCl and 0.02 M ethylenediaminetetraacetic acid, pH 7, after which 0.1 ml of a detergent mixture (2) was added, mixed gently, and incubated at 37 C for 10 min. After addition of 45 mg of solid NaCl and 0.2 ml of previously heated (80 C for 10 min) ribonuclease solution (0.5 mg of pancreatic ribonuclease per ml [Calbiocheml and 15 U of T, ribonuclease per ml [Worthington Biochemicals Corp. ]), the mixture was incubated at 37 C for 1 h. The solution was then dialyzed against SSC (0.15 M NaCl and 0.015 M sodium citrate, pH 7) for 2 to 3 h at 25 C. Analytical ultracentrifugation in CsCl. The method used has been described by Meselson et al. (20). Enough solid CsCl and SSC were added to DNA samples containing 2 to 5 ,g of yeast DNA and about 1.3 gg of Micrococcus lysodeikticus (M. aureus var. Iysodeikticus) DNA to bring the refractive index to 1.400 and volume to 0.6 to 0.7 ml. After determining the refractive index (25 C) of each sample with an Atago refractometer, the samples were centrifuged for 18 to 24 h at 44,770 rpm in a Spinco model E ultracentrifuge. Exposures at 30 and 60 s of each DNA sample were taken using an ultraviolet light source and Kodak commercial film. The ultraviolet photographic DNA banding patterns were converted to absorbance profiles with a Joyce-Loebl microdensitometer at 1:5, 1:10, and 1:20 arm ratios. Chemical DNA measurement. DNA content per cell was determined by the 3,5-diaminobenzoic acid dihydrochloride (DABA) assay. The following DABA procedure is a modification of two currently used methods (15; C. Milne, M.S. thesis, University of Washington, Seattle, 1972). Washed-cell samples were carefully counted in a hemocytometer; unbudded and budded cells were counted as 1 cell unit. Duplicates of each cell sample were assayed. Approximately 108 cells were placed in tapered, siliconecoated centrifuge tubes (12) and suspended in 0.2 ml of 1 N NaOH at 25 C for 16 to 24 h. After neutralizing the base with 0.2 ml of 1 N HCI, the tubes were placed in ice and 0.5 ml of 10W. trichloroacetic acid was added at 0 C. After a 5-min incubation, the samples were centrifuged at 0 C to pellet the cells and the supernatant liquid was carefully removed with a Pasteur pipette. The samples were then treated as follows. They were washed once with 0.5 ml of 51'7 cold trichloroacetic acid, twice with 0.1 M potassium acetate in 95%7c ethanol, and once with absolute ethanol. A fresh Norite A-DABA (99(' , Aldrich) suspension (0.4 g of DABA/ml) was prepared and filtered through a 0.3-Am cellulose filter (Millipore Corp.) at 25 C. Before 100 jul of filtered DABA reagent
VOL. 123, 1975
NUCLEAR AND MITOCHONDRIAL DNA REPLICATION
was added, the cell pellets were thoroughly dried in a boiling-water bath (10 to 15 min). After incubating the samples at 60 C for 30 min with DABA reagent, 3 ml of 1 N HCl was added. The fluorescence of each sample was measured in an Aminco-Bowman spectrophotofluorometer (at 423 nm excitation and 519 nm emission) and compared to calf thymus DNA (Type V, Sigma Chemical Co.) standards. Analysis of percentage of nuclear and mitochondrial DNA per sample. The 1:10 or 1:20 arm ratio microdensitometer tracings were used to calculate the mitochondrial DNA percentage, or l"N: 5N- and '5N: 4N-labeled nuclear DNA per sample. Overlapping nuclear DNA bands were resolved by eye. Care was taken to retain the gaussian shape of the components. Tracings of each DNA band were cut and weighed to determine their percentage per sample. Only photographs of DNA samples within the linear range of the film were used.
RESULTS To study nuclear and mitochondrial DNA synthesis during the cell cycle, a "IN-labeled log-phase population of S. cerevisiae was shifted to 14N medium. After a time lapse equal to one-half generation (60 min), the cells were centrifuged on a sorbitol gradient in a zonal rotor to fractionate the population according to cell size and age (25). This fractionation does not perturb the cell's biochemical balance and yields cell samples sufficient for DNA isolation. Consecutive gradient fractions were pooled into nine samples, containing at least 4 x 108 cells/sample, to facilitate subsequent DNA isolation. The cell number per sample varied as shown in Fig. 1. The distribution of cell number within the rotor is very similar to a coulter counter cell size distribution plot of the unfractionated log-phase population (Fig. 1). The close identity between the two distributions strongly suggests that zonal rotor fractionation separated cells according to their size. Further evaluation of fractionation was obtained from the distributions of DNA content per cell and percent of budded cells across the rotor (Fig. 2). Sample 1 contained 94% unbudded cells, budding increased in samples 2 through 5 and stabilized at 86% budded cells in samples 6 through 9. DNA content per haploid cell unit increased from 18.5 fg to 34 fg (1 femtogram = 10-15 g) across the rotor. Since ex-
cellent morphological separation was achieved, and DNA synthesis displayed kinetics commonly observed for a synchronous yeast population (31), we assumed that the successive rotor fractions adequately represent the continuum of the yeast cell cycle. Chemical estimations of total DNA per cell, such as those in Fig. 2, cannot accurately dis-
499
Window 5
25
15 : 1:2:3:4:5:6.7:8.8
9
2 a
I°
E co
0
x
u)
\O
1 * * ** \-* **...0 :/
0:::\::.
./ 1::.
*v
40
20
Fraction
FIG. 1. A comparison of cell number per zonal a size distribution plot of a log-phase S. cerevisiae culture. An exponentially growing population of S. cerevisiae was fractionated on a 8 to 35% sorbitol gradient in an MSE zonal rotor as described. Consecutive tubes containing cells were pooled into nine samples. The number of cells per pooled sample is plotted at each sample midpoint (0). The size distribution was generated by a coulter counter fitted with a 100-,gm orifice and a particle size distribution plotter with 25 windows (0). rotor sample to
tinguish nuclear from mitochondrial DNA replication during the cell cycle. However, estimates of nuclear and mitochondrial DNA synthesis can be obtained from the amount of '4N incorporated into each "5N-labeled DNA species. To determine 14N incorporation
into '5N nuclear and mitochondrial DNA fractions, DNA samples were centrifuged to equilibrium in CsCl in an analytical ultracentrifuge to separate the nuclear and mitochondrial DNA components. The buoyant densities of both nuclear and mitochondrial DNA peaks were calculated from microdensitometer tracings of their ultraviolet photographs. In all, 11 DNA samples were analyzed: (i) the initial log-phase '5N-labeled population; (ii) the unfractionated population after onehalf generation growth in '4N medium; and
(iii) the nine samples obtained from the zonal rotor. The "IN-labeled cells contained nuclear DNA of buoyant density 1.715 g/cm5 and mitochondrial DNA of density 1.699 g/cm' (Fig. 3). After the cell population had replicated for a period equal to one-half a cell generation in '4N medium, hybrid ('5N: "4N) nuclear DNA (p = 1.708 g/cm) appeared and the mitochondrial DNA peak shifted to 1.696 g/cm3. Figure 4 shows the densitometer tracings obtained from the nine zonal rotor samples. The average mitochondrial peak density for all nine samples
500
SENA ET AL.
30 0
z 0
20 20
~ ~I
J. BACTERIOL. 100
-) 50 0o
co .o~
40
Fraction
FIG. 2. The percentage of budding cells and DNA per cell unit for each zonal rotor sample. Duplicate DNA measurements (0) were obtained with the fluorometric DABA method described. A hemocytometer was used to determine the percentage of budding cells (0) per sample.
was 1.696 g/cm3, though systematic variation in the amount of heavy (15N: 15N) and hybrid (l5N: 4N)-labeled nuclear DNA is apparent in these samples. The extent of nuclear DNA replication per sample during the one-half generation incorporation of 14N, was obtained by measuring the amount of hybrid DNA. Figure 5 shows that the percentage of hybrid nuclear DNA increased from 6 to 68% and then decreased to 44% during the cell cycle. Since heavy and hybrid peaks were not seen in any zonal rotor mitochondrial DNA sample, mitochondrial DNA replication could be ascertained only by the 14N level in each mitochondrial peak. In contrast to the nuclear incorporation pattern, the '4N percentage in mitochondrial DNA remained effectively constant during the cell cycle (Fig. 6).
DISCUSSION The usefulness of the zonal rotor procedure for cell cycle studies relies upon a successful fractionation across the rotor. If the consecutive zonal rotor samples represent increasing cell age in the cell cycle, they should display the bud initiation distribution and a discontinuous DNA synthetic pattern observed in a synchronous population. Since zonal rotor samples analyzed here and elsewhere (25) satisfy these criteria (Fig. 1, 2), we conclude that they represent adequately the yeast cell budding cycle. Therefore, they may be used to study the timing of DNA replication during the cell cycle. Nuclear DNA synthesis. Although a discontinuous nuclear DNA synthetic pattern based on measurements of total DNA per cell across the mitotic cycle is well documented, studies on nuclear DNA replication during the yeast cell cycle based on density-shift techniques are reported only by Sena (Ph.D. thesis, 1972). To
assist the analysis of the observed hybrid nuclear DNA distribution across the zonal rotor, a theoretical hybrid nuclear DNA distribution across the cell cycle was generated. This theoretical distribution predicts that: (i) unbudded cells, in G2 at the time of label shift, should not contain hybrid nuclear DNA; (ii) cells entering S-phase during 14N addition should show increasing levels of hybrid nuclear DNA; (iii) cells completing the entire S-phase during the density shift should contain only hybrid DNA; and (iv) those completing S-phase when '4N was
U
CD co
VI .
1-IN
h0 140 Lq P-
O
Density g/cc FIG. 3. Densitometer tracings of DNA from exponentially growing cultures of S. cerevisiae. A cell culture grown 12 generations in (I5NH4)2SO4 containing medium was shifted for one-half generation to ( 14NH4)2SO4 medium. DNA from cell samples removed before (upper tracing) and after the shift (lower tracing) were centrifuged to equilibrium in CsCI as described. The density of heavy-labeled nuclear DNA (15N: 15N) is 1.715 g/cm 3, whereas heavylabeled mitochondrial DNA is 1.699 g/cms. The hybrid nuclear band (15NW1"N) density is 1. 708 g/cm 3, whereas the mitochondrial peak is 1.696 g/cm3. M. lysodeikticus DNA (1.731 g/cm,) was used as marker.
NUCLEAR AND MITOCHONDRIAL DNA REPLICATION
VOL. 123, 1975 4
C0 5
U.
O
7
i:;
8
6
D
9
r-
/D
clear DNA across the cell cycle. Little is known about DNA precursor pool fluctuations during the mitotic cycle, therefore, accurate predictions of "4N incorporation lag into nuclear DNA during this experiment cannot be made at this time. Perhaps future experiments will shed some light on these observa'tions. Mitochondrial DNA Synthesis. In contrast to the nuclear 14N incorporation pattern which displayed a large fluctuation across the rotor, mitochondrial DNA "4N incorporation was relatively constant through the cell cycle (Fig. 6). The line depicted in Fig. 6 was calculated by the method of least squares. It does not differ significantly from the line drawn through the mean of all the mitochondrial incorporation values (which has a slope of zero). Figure 7 compares these data with "4N levels expected across the cell cycle for four different mitochondrial DNA replication patterns: (i) mitochondrial DNA replication in GI only; (ii) mitochondrial DNA replication in S-phase only; (iii) mitochondrial DNA replication in G, only; or (iv) mitochondrial DNA replication throughout the cell cycle. The three confined and discontin-
Density g/cc FIG. 4. Densitometer tracings of DNA isolated from nine pooled zonal rotor cell samples. An exponentially growing "5N-labeled cell population shifted to "N medium for one-half generation was fractionated into nine samples on a sorbitol gradient in a zonal rotor. DNA isolated from each sample was centrifuged to equilibrium in CsCI. Densitometer tracings of the ultraviolet photographs of each sample are shown. For reference, sample I contained 95% unbudded cells. Budding increased in samples 2 through 6, and maximized at 85% budded cells in samples 7 through 9 (see Fig. 2).
added should contain decreasing levels of DNA hybrid. Accordingly, the theoretical hybrid nuclear DNA distribution (Fig. 5) should be bell shaped. Whereas the actual hybrid nuclear DNA levels show a distribution similar to the theoretical expectation, only the early cell cycle values correspond with those predicted. Since the theoretical hybrid nuclear DNA distribution was calculated assuming no "4N incorporation lag into 15N DNA, and there was a one-fourth generation lag ( "5N-labeled cells grown for onehalf generation in "4N medium should contain 66% hybrid nuclear DNA, however, the unfractionated cell population contained only 40% hybrid nuclear DNA), the discrepancy between predicted and observed values probably reflects the "4N incorporation lag into "5N-labeled nuclear DNA. Possibly, the fractionation of the cell population on the zonal rotor unmasked the lag distribution for "4N incorporation into nu-
501
0
z
a
0\
c;
z
z
0 5010
uzz
P* 20
40
Fraction
FIG. 5. Distribution of hybrid nuclear DNA during the cell cycle. Shown is the percentage of "5N: 4N
nuclear DNA per zonal sample (0) versus the hybrid values predicted across the rotor (0). The sample percentages were estimated according to the procedure outlined. The theoretical hybrid DNA distribution expected for an exponentially growing cell population shifted from "N to 14N for one-half generation before fractionation on a zonal rotor was calculated based on the following assumptions: (i) fraction number was proportional to cell age; (ii) nuclear DNA was replicated at a constant rate only in samples 2 through 5 (or 0.13 to 0.47 of the cell cycle); and (iii) there was no 14N incorporation lag into nuclear DNA at any portion of the cell cycle. Therefore, cells at the end of the cycle in fraction 44 were at approximately the same age as cells presently in fraction 26 (0.50 generation younger) when 14N was added. Since nuclear DNA was not synthesized between fractions 26 to 44, no 14N incorporation in fraction 44 is predicted. Other theoretical points were calculated in the
same manner.
502
SENA ET AL.
J. BACTERIOL.
higher than the observed values suggesting
z
30 0
0
20~
40
Fraction
0-0
FIG. 6. Percentage of 14N in mitochondrial DNA in zonal rotor samples. The 14N content of each mitochondrial DNA peak was calculated from the buoyant density of each sample assuming 1.699 g/cm 3 as 100% I N and 1.684 g/cm3 as 100%07o "N DNA. The standard error for each calculation is included.
.X Z
30-
10
20
40
Fraction
FIG. 7. 14N incorporation into mitochondrial DNA during the cell cycle. The amount of 14N in each zonal rotor sample (-) (from Fig. 6) is compared with the pattern predicted by the four major temporal patterns of mitochondrial DNA replication during the cell cycle. (i) Mitochondrial DNA replication during G, only (O); (ii) during S-phase only (0); (iii) in G2 only (A); or (iv) during the complete cell cycle (x). The following assumptions were made for these theoretical distributions. (i) Fraction number was proportional to cell age; (ii) nuclear DNA replication occurred in fractions 2 through 5; (iii) mitochondrial DNA was replicated completely and at a constant rate only during the cell cycle stage in question; and (iv) there was no 14N incorporation lag into mitochondrial DNA at any portion of the cell cycle. For example, cells in fraction 44 were about the age of cells in fraction 26 (0.50 generation younger) when the medium was changed, and were already through G1 and S, and 3.9% of G2. If mitochondrial DNA were replicated only in G1 or during S-phase, fraction 44 should contain 0% 14N. If replication occurred only in G2, 48% 14N was expected; while 33% 14N was predicted if mitochondrial DNA replication were continuous throughout the cell cycle. The other theoretical values were calculated in the same manner.
uous replication patterns predict wide fluctuations in "4N incorporation during the cell cycle, while continuous mitochondrial DNA replication incorporates a constant amount of' 14N throughout the cell cycle. Our data (Fig. 6, 7) appears to fit the continuous model. However, the 14N levels predicted for each sample are
a
14N incorporation lag into "5N-labeled mitochondrial DNA in every cell fraction. The lag seen in each fraction is about the same as the incorporation lag of one-fourth of a generation in the unfractionated cell population. ('5Nlabeled cells grown for one-half generation in "4N medium should contain 66% '5N: 14N or in this case 3367c '4N, whereas only 20% 14N was
observed.)
Could the 14N percentage in mitochondrial DNA across the rotor be spurious, in that '4N was incorporated during the zonal rotor isolation procedure? Because the mitochondrial DNA from the zonal samples averaged the same amount of' 14N as the initial unfractionated cell population, extensive 14N incorporation during zonal fractionation did not occur. (A similar experiment using cells frozen before zonal fractionation showed the same mitochondrial DNA 14N incorporation pattern across the rotor [Sena, Ph.D. thesis, 1972].) We believe the constant amount of' label incorporation represents continuous synthesis of mitochondrial DNA during the cell cycle. To insure that mitochondrial DNA was not selectively lost from the zonal samples during DNA isolation or diluted out during cell growth in 14N, the percentage of' mitochondrial DNA per sample was determined. The sample percentages and those predicted for a continuous pattern of mitochondrial DNA synthesis during the cell cycle are in agreement (Fig. 8). The fluctuations between the two curves are well within the standard error common for densitometer tracing analysis. Therefore the 14N incorporation data do not reflect either selective mitochondrial DNA loss during isolation or cell growth. However, it is still conceivable that small fluctuations in the mitochondrial DNA replication rate during the cell cycle could go undetected by our methods. Since the error in DNA density measurements is _40.001 g/cm3 (Fig. 6), rate fluctuations up to ±30'Yc during one-half' generation would fall within these error limits and be undetected. If we assume that mitochondrial DNA has a molecular weight of 50 >, 106 (13), unbudded cells, in the present work, contain about 29 molecules of' mitochondrial DNA. Continual, relatively constant, 14N incorporation into this mitochondrial DNA population could occur in two basic ways. First, the replication time for each individual molecule (not known at present) could span the cell cycle. Second, certain molecules in the mitochondrial DNA population (for instance, in dif'ferent mitochondria)
VOL. 123, 1975
NUCLEAR AND MITOCHONDRIAL DNA REPLICATION
503
content, malate dehydrogenase and cytochrome oxidase activity in synchronous cells, Cottrell 1AC e 14 and Avers (8) concluded that mitochondrial V DNA was synthesized immediately before the S-period. Since malate dehydrogenase activity is a composite of both cytoplasmic and mitochondrial enzyme activity, and is coded by the 20 40 -0-0 nucleus (18), periodic enzyme activity during Fraction the cell cycle does not necessarily reflect periFIG. 8. Mitochondrial percentage per zonal rotor sample. Shown is the percentage mitochondrial DNA odic mitochondrial growth. A similar argument per sample measured as described (0). The standard applies for cytochrome oxidase; although the error for each measurement is included. The predicted enzyme is localized in mitochondria, it is compercentages (0) are calculated assuming (i) cell age is posed of both nuclear and mitochondrially proportional to fraction number; (ii) nuclear DNA coded components (7, 16). Furthermore, no data synthesis occurred at a constant rate only in samples concerning the actual amount or percentage 2 through 5 (or 0.13 to 0.47 of the cell cycle); and (iii) mitochondrial DNA during the cell cycle was mitochondrial DNA replication occurred at a constant presented by these authors. Accordingly, we rate throughout the cell cycle. The predictions are suggest that their conclusions concerning the obtained by multiplying the ratio between the calculated femtograms of mitochondrial DNA per sample temporal mitochondrial DNA replication patand the calculated femtograms of total DNA per tern are at best tentative or somewhat premasample by 100. The initial amount of mitochondrial ture in view of the data reported. By measuring DNA in sample 1 is assumed to be 2.4 fg/unbudded the 3H and 32p label incorporation into "4Ccell, or 13%o (Fig. 8) of 18.5 fg/cell (Fig. 2). For labeled nuclear and mitochondrial DNA during example, cells in fraction 26 have completed all the cell cycle, Wells (28) concluded that mitonuclear DNA synthesis but only 0.54 of mito- chondrial DNA was synthesized concomitantly chondrial DNA synthesis, therefore they contain with nuclear DNA. However, closer scrutiny of 3.7/35.9 x 100 or 10.3%o mitochondrial DNA. All the the results indicates that mitochondrial DNA other calculations were done similarly. incorporated some label throughout the cell cycle. It is possible that his particular strain could replicate at discrete intervals, but at the does show an increased level of mitochondrial total population level replication would span DNA synthesis in S-phase. It may be emphathe entire mitotic cycle. Because mitochondrial sized that our data does not rule out this DNA replication in S. cerevisiae is accompanied possibility, but it is also true that not all by a dispersive mechanism (Sena, Ph.D. thesis, synthesis occurs in S-phase. Additionally, if all 1972; 29) and no discrete 15N:'5N or 15N:14N mitochondrial DNA is synthesized during Speaks are observed from mitochondrial DNA phase, as concluded by Wells, the mitochonwhich has incorporated 14N when the DNA is drial DNA percentage should remain relatively banded isopycnically, our data cannot discrimi- constant during the cell cycle, and not increase nate between these alternatives. That we see no in the later cell cycle fractions, as was pub15N: 15N shoulder on any mitochondrial peak lished. We feel there are alternative conclusions (even when the CsCl gradients are over-loaded to the data presented. A novel approach to for the nuclear component) suggests that label studying DNA synthesis was taken by Dawes dispersion is very rapid. Because all the zonal and Carter (9). They used the level of nisamples displayed a unimodal peak (Fig. 4), the trosoguanidine-induced cytoplasmically inherunderlying dispersive mechanism(s) (sister- ited oligomycin- and erythromycin-resistant strand exchanges, intramolecular exchanges, mutants to probe mitochondrial DNA replicaintermolecule exchanges, and/or repair mech- tion during the cell cycle. The basic assumption anisms) probably occur throughout the cell in this approach is that nitrosoguanidine mutagenesis occurs principally at the replication cycle. A continuous mode of mitochondrial DNA fork. However, because nitrosoguanidinesynthesis during mitosis in S. cerevisiae has induced mutations in eukaryotes may arise now been observed in three different yeast independently of DNA replication (1), the data strains (Sena, Ph.D. thesis, 1972; 30; this pa- of Dawes and Carter may not only indicate per). Can these conclusions be reconciled with DNA replication, but also other cell cycle pheother conflicting temporal distribution patterns nomena such as DNA repair etc. The previous discussion indicates that nureported for mitochondrial DNA synthesis during mitosis? By measuring whole cell DNA clear and mitochondrial DNA replication are z
Q
504
SENA ET AL.
temporally dissociated during mitosis in S. cerevisiae. In addition, mitochondrial DNA replication occurs continuously during all major decision points in the S. cerevisiae life cycle. Since nuclear synthesis is always restricted to a discrete time period while mitochondrial DNA synthesis occurs continuously, their synthesis appear to be controlled by different mech-
J. BACTERIOL. 13. Hollenberg, C. P., P. Borst, and F. F. J. Van Bruggen. 1970. Mitochondrial DNA. V. A 25 closed circular duplex DNA molecule in wild-type yeast mitochondria. 14.
15.
anisms. ACKNOWLEDGMENTS We like to thank R. K. Mortimer for use of the model E analytical ultracentrifuge in Donner Laboratory and the Department of Molecular Biology for use of their Joyce Loebl Microdensitometer. The helpful suggestions of Douglas Campbell are acknowledged. This research was supported in part by Public Health Service research grants GM-17317 to S. Fogel from the National Institute of General Medical Sciences and Al010610 to H. 0. Halvorson from the National Institute of Allergy and Infectious Diseases. Preliminary investigation was carried out at the University of Wisconsin during the tenure of a predoctoral fellowship (E.S.) from the Public Health Service. LITERATURE CITED 1. Auerbach, C., and B. J. Kilbey. 1971. Mutation in eukaryotes. Annu. Rev. Gen. 5:163-218. 2. Blamire, J., D. R. Cryer, D. B. Finkelstein, and J. Marmur. 1972. Sedimentation properties of yeast nuclear and mitochondrial DNA. J. Mol. Biol. 67:11-24. 3. Braun, R., and T. E. Evans. 1969. Replication of nuclear satellite and mitochondrial DNA in the mitotic cycle of Physarum. Biochim. Biophys. Acta 182:511-522. 4. Calvayrac, R., R. A. Butow, and M. Lefort-tran. 1972. Cyclic replication of DNA and changes in mitochondrial morphology during the cell cycle of Euglena gracilis. Exp. Cell Res. 71:422-432. 5. Cameron, I. L. 1966. A periodicity of tritiated-thymidine incorporation into cytoplasmic deoxyribonucleic acid during the cell cycle of Tetrahymena pyriformis. Nature (London) 209:630-631. 6. Charret, R., and J. Andre. 1968. La synthese de l'ADN mitochondrial chez Tetrahymena pyriformis. J. Cell Biol. 39:369-381. 7. Chen, W. L., and F. C. Charalampous. 1969. Mechanism of induction of cytochrome oxidase in yeast. I. Kinetics of induction and evidence for accumulation of cytoplasmic and mitochondrial precursors. J. Biol. Chem. 244:2767-2776. 8. Cottrell, S. F., and C. J. Avers. 1970. Evidence of mitochondrial synchrony in synchronous cultures of yeast. Biochem. Biophys. Res. Commun. 38:973-980. 9. Dawes, I. W., and B. L. A. Carter. 1974. Nitrosoguanidine mutagenesis during nuclear and mitochondrial gene replication. Nature (London) 250:709-712. 10. Guttes, E. W., C. Hanawalt, and S. Guttes. 1967. Mitochondrial DNA synthesis and the mitotic cycle in Physarum polycephalum. Biochim. Biophys. Acta 142:181-194. 11. Halvorson, H. O., B. L. A. Carter, and P. Tauro. 1968. Use of synchronous cultures of yeast to study gene position, p. 462-470. In L. Grossman and K. Moldave (ed.), Methods in enzymology, nucleic acids, vol. 21. Academic Press Inc., New York. 12. Hinegardner, R. T. 1971. An improved fluorometric assay for DNA. Anal. Biochem. 39:197-201.
16.
17. 18. 19.
20.
21.
22. 23.
24.
25.
26.
27. 28.
29.
30.
31.
Structure and genetic complexity. Biochim. Biophys. Acta 209:1-15. Holt, C. E., and E. G. Gurney. 1969. Minor components of the DNA of Physarum polycephalum. Cellular location and metabolism. J. Cell Biol. 40:484-496. Kissane, J. M., and E. Robbins. 1958. The fluorometric measurement of deoxyribonucleic acid in animal tissues with special reference to the central nervous system. J. Biol. Chem. 233:184-188. Kraml, J., and H. R. Mahler. 1967. Biochemical correlates of respiratory deficiency. VII. A precipitating antiserum against cytochrome oxidase of yeast and its use in the study of respiratory deficiency. Immunochemistry 4:213-226. Kuenzi, M. T., and R. Roth. 1974. Timing of mitochondrial DNA synthesis during meiosis in Saccharomyces cerevisiae. Exp. Cell Res. 85:377-382. Longo, G. P., and J. G. Scandalios. 1969. Nuclear gene control of mitochondrial malic defiydrogenase in maize. Proc. Natl. Acad. Sci. U.S.A. 62:104-111. Madreiter, H., C. Mittermayer, and 0. Rainhardt. 1972. 3H-thymidine incorporation into mitochondria of synchronized mouse fibroblasts. Beitr. Pathol. 145:249-255. Meselson, M., F. W. Stahl, and J. Vinograd. 1957. Equilibrium sedimentation of macromolecules in density gradients. Proc. Natl. Acad. Sci. U.S.A. 43:581-588. Parsons, J. A. 1965. Mitochondrial incorporation of tritiated thymidine in Tetrahymena pyriformis. J. Cell Biol. 25:641-646. Parsons, J. A., and R. C. Rustad. 1968. The distribution of DNA among dividing mitochondria of Tetrahymena pyriformis. J. Cell Biol. 37:683-693. Pica-Mattoccia, L., and G. Attardi. 1972. Expression of the mitochondrial genome in HeLa cells IX. Replication of mitochondrial DNA in relationship to the cell cycle in HeLa cells. J. Mol. Biol. 64:465-484. Pinon, R., Y. Salts, and G. Simchen. 1974. Nuclear and mitochondrial DNA synthesis during yeast sporulation. Exp. Cell Res. 83:231-238. Sebastian, J., B. L. A. Carter, and H. 0. Halvorson. 1971. Use of yeast populations fractionated by zonal centrifugation to study the cell cycle. J. Bacteriol. 108:1045-1050. Smith, D., P. Tauro, E. Schweizer, and H. 0. Halvorson. 1968. The replication of mitochondrial DNA during the cell cycle in Saccharomyces lactis. Proc. Natl. Acad. Sci. U.S.A. 60:936-942. Steinert, M., and G. Steinert. 1962. La synthese de l'acide desoxyribonucleic au cours du cycle de division de Trypanosoma mega. J. Protozool. 9:203-211. Wells, J. R. 1974. Mitochondrial DNA synthesis during the cell cycle of Saccharomyces cerevisiae. Exp. Cell Res. 85:278-286. Williamson, D. H., and D. J. Fennell. 1974. Apparent dispersive replication of yeast mitochondrial DNA as revealed by density labelling experiments. Mol. Gen. Genet. 131:193-207. Williamson, D. H., and E. Moustacchi. 1971. The synthesis of mitochondrial DNA in synchronously dividing cultures of Saccharomyces cereuisiae. Biochem. Biophys. Res. Commun. 42:195-201. Williamson, D. H., and A. W. Scopes. 1960. The behavior of nucleic acids in synchronously dividing cultures of Saccharomyces cereuisiae. Exp. Cell Res. 20:338-349.
