Experimental reconsideration of the utility of serum starvation as a method for synchronizing mammalian cells

Accurate cell-size determinations support the prediction that serum starvation and related whole-culture methods cannot synchronize cells. Theoretical considerations predict that whole-culture methods of synchronization cannot synchronize cells. Upon serum starvation, the fraction of cells with a G1-phase amount of DNA increased, but the cell-size distribution is not narrowed. In true synchronization, the cell-size distribution should be narrower than the cell-size distribution of the original culture. In contrast, cells produced by a selective (i.e. non-whole-culture) method have a specific DNA content, a narrow size distribution, and divide synchronously. The general theory leading to the conclusion that whole-culture methods for synchronization do not work implies that one can generalize these serum-starvation results to other cell lines and other whole-culture methods, to conclude that these methods do not synchronize cells.     

Rejoinder: whole-culture synchronization cannot, and does not, synchronize cells

There have been numerous proposals suggesting that whole-culture methods - in which all cells in a growing culture are treated identically - can synchronize cells. An explicit defense of these methods has been presented (Spellman and Sherlock, this issue, pp. 270-273, ). Here, this defense of whole-culture 'synchronization' is subjected to a critical evaluation leading to the conclusion that whole-culture synchronization cannot synchronize cells - at all. Whole-culture methods cannot produce a set of cells that reflects the size and genome composition of cells of any particular cell-cycle age during the normal cell cycle. Thus, in addition to the well-recognized problem of artifacts, it is proposed that experiments using whole-culture treatments (usually starvation or inhibition methods) are not suitable for cell-cycle analysis because these methods do not produce a synchronized culture.     

Reply: whole-culture synchronization - effective tools for cell cycle studies

Studies of gene expression during the eukaryotic cell cycle in whole-culture synchronized cultures have been published using many methodologies. These procedures alter the state of the cell cycle for a population of cells, rather than purifying a population of cells that are in the same state. Criticism of these methods (e.g. see Cooper, this issue, pp. 266-269, ) suggests that these studies are flawed, and posits that such methodologies cannot be used to study the cell cycle because they alter the size and age distributions of the cultures. We believe that whole-culture cell cycle studies work even though they alter the size and age distributions: these cells still progress through the cell cycle and although we do not suggest that the methods are perfect, we will explain how these microarray studies have successfully identified cell cycle regulated genes and why these results are biologically meaningful.     

Nocodazole does not synchronize cells: implications for cell-cycle control and whole-culture synchronization

It has been predicted that nocodazole-inhibited cells are not synchronized because nocodazole-arrested cells with a G2-phase amount of DNA would not have a narrow cell-size range reflecting the cell size of some specific, presumably G2-phase, cell-cycle age. Size measurements of nocodazole-inhibited cells now fully confirm this prediction. Further, release from nocodazole inhibition does not produce cells that move through the cell cycle mimicking the passage of normal unperturbed cells through the cell cycle. Nocodazole, an archetypal whole-culture synchronization method, can inhibit growth to produce cells with a G2-phase amount of DNA, but such cells are not synchronized. Cells produced by a selective (i.e., non-whole-culture) method not only have a specific DNA content, but also have a narrow size distribution. The current view of cell-cycle control that is based on methods that are not suitable for cell-cycle analysis must therefore be reconsidered when results are based on whole-culture synchronization.This work was supported by the National Science Foundation (grant MCB–0323346) and (in part) by the National Institutes of Health (University of Michigan’s Cancer Center, support grant 5 P30 CA46592). G.I., M.T., and P. B. are associated with the Undergraduate Research Opportunity Program of the University of Michigan, which also supported this research.     

Synchronization in the cell cycle by inhibitors of DNA replication induces histone H2AX phosphorylation: an indication of DNA damage

Several methods to synchronize cultured cells in the cell cycle are based on temporary inhibition of DNA replication. Previously it has been reported that cells synchronized this way exhibited significant growth imbalance and unscheduled expression of cyclins A and B1. We have now observed that HL-60 cells exposed to inhibitors of DNA replication (thymidine, aphidicolin and hydroxyurea), at concentrations commonly used to synchronize cell populations, had histone H2AX phosphorylated on Ser-139. This modification of H2AX, a marker of DNA damage (induction of DNA double-strand breaks; DSBs), was most pronounced in S-phase cells, and led to their apoptosis. Thus, to a large extent, synchronization was caused by selective kill of DNA replicating cells through induction of replication stress. In fact, similar synchronization has been achieved by exposure of cells to the DNA topoisomerase I inhibitor camptothecin, a cytotoxic drug known to target S-phase cells. A large proportion of the surviving cells 'synchronized' by DNA replication inhibitors at the G1/S boundary had phosphorylated histone H2AX. Inhibitors of DNA replication, thus, not only selectively kill DNA replicating cells, induce growth imbalance and alter the machinery regulating progression through the cycle, but they also cause DNA damage involving formation of DSBs in the surviving ('synchronized') cells. The above effects should be taken into account when interpreting data obtained with the use of cells synchronized by inhibitors of DNA replication.     

Nuclear DNA polymerases and the HeLa cell cycle.

Purified nuclei of HeLa S3 cells contain two DNA-dependent DNA polymerases that have distinct physical and enzymatic properties. We have investigated the variations in their activity during the cell cycle of a synchronized culture. Cells were synchronized by a double thymidine block, harvested at various phases of the cycle, and the two DNA polymerases were purified partially by DEAE-cellulose and phosphocellulose chromatography. The activity of DNA polymerase I (low molecular weight, N-ethylmaleimide-insensitive) remains essentially constant throughout the cycle. The activity of DNA polymerase II (high molecular weight, N-ethylmaleimide-sensitive), however, increases during G1 to mid-S and declines, 7- to 10-fold between late-S and G2. Addition of cycloheximide (60 mug/ml) to cultures 12 hours after the release from thymidine block abolishes the rise in the activity of DNA polymerase II. Cycloheximide also reduced the activity of DNA polymerase I by 60%. Addition of hydroxyurea (1mM) at 1 hour after release has no effect on the activity of either enzyme. We conclude that in HeLa cells, DNA polymerase I and II are distinct enzymes, that DNA polymerase II probably functions in DNA replication and is probably induced in response to stimuli for DNA biosynthesis.     

Synchronization of human diploid fibroblasts at multiple stages of the cell cycle

Because of the scarcity of techniques for synchronizing the growth of cultured human diploid fibroblasts at multiple stages within the cell cycle, efforts were expended in this report to establish a set of protocols that would permit synchronization of cells at several different points throughout the cycle. The protocols that were developed to synchronize the growth of HSF-24 and HSF-55 cells, human foreskin-derived fibroblast cultures, were modifications of procedures employed to synchronize the growth of cultured rodent cells. Optimization of synchrony induction was directed by consideration of both the biochemical properties of the synchronized populations (determined via three-parameter flow cytometric measurements of DNA, RNA, and protein contents) and their kinetic behavior following reversal of the synchronization-inducing blockade (determined via combined flow cytometric analysis of DNA content, [3H]thymidine autoradiography, and measurement of increase in cell number). The conditions judged to yield the best results for studying events associated with production of a G0 block or for maintaining cells for prolonged periods in G0 were those in which the cells were grown to confluency in D-MEM supplemented with 10% fetal bovine serum. Procedures producing the best results for studying processes associated with the G0 to G1 transition, G1 events, and operations accompanying the transition from G1 to S, employed subconfluent growth for 48 h in alpha-MEM + 0.1% fetal bovine serum (alpha-MEM0.1F) followed by resuspension in alpha-MEM containing 10% fetal bovine serum (alpha-MEM10F). When the goal was to obtain cells in which to study very early S-phase events, satisfactory results were achieved by combining a 48-h period of subconfluent growth in alpha-MEM0.1F, followed by treatment for 24 h in alpha-MEM10F containing 5 micrograms/ml aphidicolin. For study of events occurring in mid- to late-cycle, acceptable results were achieved by combining a 48-h block in alpha-MEM0.1F with resuspension for 24 h in alpha-MEM10F containing 10(-3) M hydroxyurea followed by resuspension in drug-free alpha-MEM10F. The best results were obtained with these latter synchronization procedures (i.e., low-serum/high-serum + APC or HU/high serum) when the fetal calf serum was replaced with heat-inactivated calf serum. The success achieved in synchronizing the growth of these human diploid fibroblasts compared favorably/exceeded the results obtained with synchronized cultures of Chinese hamster ovary cells.     

Relationship between changes of the steroid receptor and synchronization in human endometrial adenocarcinoma cells in vitro

It has been reported that the response of target cells to steroid hormone (SH) stimulation may depend on their position in the cell cycle. The DNA and RNA contents of malignant cells of the endometrium cultured in vitro were measured using flow cytometry (FCM). We also measured estrogen receptor (ER) and progesterone receptor (PR) levels of cells at different positions in the cell cycle. The G1 and S phases of the cell cycle were investigated using cells synchronized by sodium n-butyrate (G1 block), methotrexate (S block), and excess thymidine (S block). For DNA measurements, the cells were stained with propidium iodide following RNase treatment. For RNA measurements (double-stranded RNA) the cells were treated with DNase. We found that S phase synchronization by methotrexate was 136.2% of control (100%). Using the excess thymidine block and release procedure, the S phase fraction was 185.1% of control. G1 phase synchronization by sodium n-butyrate was 134% of control. The estrogen receptor level in G1 phase synchronized cells increased to 5.94 fmol/micrograms DNA in the cytosol and 12.35 fmol/micrograms DNA in the nuclear fraction. These levels represent a sevenfold total increase over that of the control estrogen receptor level. Cells in S phase showed no significant increase in estrogen receptor levels over control cells. Based on this study, the functional increase of the steroid receptor was most significant in the G1 phase.     

Inhibition of DNA synthesis and cell division by a cell surface sialoglycopeptide

We have isolated and purified a cell surface sialoglycopeptide (SGP) from bovine cerebral cortex cells that previously was shown to be a potent inhibitor of cellular protein synthesis. The following studies were carried out to characterize the potential ability of the SGP to inhibit DNA synthesis and to arrest cell division. Treatment of exponentially proliferating Swiss 3T3 cells with the SGP inhibitor resulted in a marked inhibition of thymidine incorporation within 24 h. When the SGP was removed from inhibited cultures, a sharp rise in 3H-thymidine incorporation followed within 3-4 h that peaked well above that measured in exponentially growing cultures, suggesting that the inhibitory action of the SGP was reversible and that a significant proportion of the arrested cells was synchronized in the mitotic cycle. In addition to DNA synthesis, the inhibitory action of the SGP was monitored by direct measurement of cell number. Consistent with the thymidine incorporation data, the SGP completely inhibited 3T3 cell division 20 h after its addition to exponentially growing cultures. Upon reversal there was a delay of 15 h before cell division resumed, when the arrested cells quickly doubled. Most, if not all, of the growth-arrested cells appeared to have been synchronized by the SGP. The SGP inhibited DNA synthesis in a surprisingly wide variety of target cells, and the relative degree of their sensitivity to the inhibitor was remarkably similar. Cells sensitive to the SGP ranged from vertebrate to invertebrate cells, fibroblast and epitheliallike cells, primary cells and established cell cultures, as well as a wide range of transformed cell lines.     

Studies on the cell cycle of Myxobacter AL-1

Myxobacter AL-1 was synchronized by size fractionation of asynchronous cultures on a 5–20% equivolumetric sucrose gradient in a Ti—15 zonal rotor. The degree of synchronization obtained by this method was determined 1. by a comparison of the size distribution of an asynchronous culture to the distribution of cell numbers in the fractions of the zonal run, 2. by studying the size distributions in the different fractions obtained with the zonal rotor and 3. by following the growth and size distribution of a fraction from a zonal run during one division cycle. The results indicate that this method is suitable for the study of the cell cycle of Myxobacter AL-1.     

Membrane fatty acid changes during the cell cycle of CV-1 cells

Monolayers of CV-1 cells were synchronized at the G1/S boundary of the cell cycle by a 24-h 2 mM thymidine blockade. Uptake of tritiated thymidine indicated that the peak DNA synthesis occurred 6-8 h after release from the block and that cell cycle time was 18-20 h. The fatty acid composition of phospholipids extracted from cells at 0, 7, and 18 h postblockade was measured by gas chromatography. The results indicate cyclic changes in membrane fatty acids with a significant increase in long-chain polyunsaturated fatty acids during the DNA synthesis phase (S phase) of the cell cycle.     

Is whole-culture synchronization biology's 'perpetual-motion machine'?

Whole-culture or batch synchronization cannot, in theory, produce a synchronized culture because it violates a fundamental law that proposes that no batch treatment can alter the cell-age order of a culture. In analogy with the history of perpetual-motion machines, it is suggested that the study of these whole-culture 'synchronization' methods might lead to an understanding of general biological principles even though these methods cannot be used to study the normal cell cycle.     

Light regulation of the cell cycle in Euglena gracilis bacillaris

We have studied the light regulation of the cell division cycle in the photosynthetic alga Euglena gracilis bacillaris. Euglena grown under phototrophic conditions are easily synchronized to a 12 h light-12 h dark regime. By inoculating stationary phase, nondividing cells into fresh media and exposing the diluted cells to either light or darkness, we have determined that initiation of DNA synthesis for the cell division cycle is light dependent. By varying the length of time in light to which synchronized cells are exposed, we have shown that commitment to the cell cycle requires exposure to more than 6 h of light. We propose that this is to allow the accumulation, through photosynthetic electron transport, of an initiating factor that will enable DNA synthesis to begin. Flow cytometry analysis also shows that once cells are committed to the cell cycle, they complete the cycle in the dark, so mitosis is a light-independent step.     

A model of cell cycle control: effects of thymidine on synchronous cell cultures.

Further evidence is presented in support of a model for growth control in which commitment for cell division is determined by an event in the preceding cell cycle. A study was made of conditions affecting synchronous growth following treatment of murine mastocytoma cells with excess thymidine at different phases of the cell cycle. Cells were synchronized by a physical procedure involving velocity sedimentation in a zonal rotor. Pulse treatment of such cultures with thymidine at times corresponding to the S, G2, and M periods had no effect on further growth. However, addition at G1, although having no immediate effect, arrested cell growth in the next cell cycle. This temporal effect may account for the decay of synchrony observed during double thymidine blockade or thymidine-FUdR blockade. When the time interval between two such blocks was 7 hr or less, P815Y cells were arrested after one synchronous division. At this critical time a majority of cells were at, or near, G1. It is suggested that thymidine exerts a hitherto unrecognized effect at the G1 interval.     

The relation between protein accumulation and cell cycle traverse of human NHIK 3025 cells in unbalanced growth

Human NHIK 3025 cells, synchronized by mitotic selection, were given 2 mM thymidine, which inhibited DNA synthesis without reducing the rate of protein accumulation. After removal of the thymidine the cells proceeded towards mitosis and cell division, with an S duration 2 hours shorter than, but a G2 and M duration nearly identical to that of the control cells. If cycloheximide (1.25 m?M) was present together with thymidine, no net protein accumulation took place during the treatment, and the subsequent duration of S, G2, and M was similar to that of the untreated cells. The shortening of S seen after treatment with thymidine alone would therefore indicate that the rate of DNA synthesis depended on the amount of some preaccumulated protein. The postreplicative period in thymidine-treated cells was lengthened by cycloheximide treatment although the protein content had already been doubled. This suggests that proteins required for the traverse of this part of the cell cycle might have to be synthesized after completion of DNA replication. Shortly after removal of thymidine, the rate of protein accumulation declined markedly, indicating the existence of some mechanism for negative control of cell mass. In addition, the daughters of thymidine-treated cells had their cell cycle shortened by 2 hours. As a result, the cells had returned to balanced growth already in the first cell cycle following the induction of unbalanced growth. In conclusion, our experiments suggest that NHIK 3025 cells might require a minimum time in order to traverse the cell cycle, which is independent of cell mass.     

Screening of five specific cell cycle inhibitors using a T Cell lymphoma cell line synchrony/release assay

To obtain different cell populations at specific cell cycle stages, we used a cell culture synchronization protocol. Effects of five different cell cycle inhibitors acting throughout the cell cycle were examined by DNA flow cytometric analysis of a synchrony/release lymphoma cell line (CEM). The screening synchronized protocol showed that staurosporine, mimosine and aphidicolin are reversible G1 phase inhibitors that act at different times. Staurosporine acted in early G1, exhibited the strongest cytotoxic effect, and induced apoptosis. Mimosine and aphidicolin acted in late G1 and at the G1/S boundary, respectively. Hydroxyurea arrested CEM cells in early S phase, but later than the aphidicolin arrest point. Nocodazole synchronized CEM cells in M phase. All the inhibitors examined in this study can be used to synchronize cells at different phases of the cell cycle and were reversible with little toxicity except for staurosporine which is highly toxic. Because the regulatory mechanism of the cell cycle is disrupted by their effects on protein synthesis, however, these drugs must be used with caution.     

Regulation of mating in the cell cycle of Saccharomyces cerevisiae

The capacity of haploid a yeast cells to mate (fuse with a haploid strain of alpha mating type followed by nuclear fusion to produce a diploid cell) was assessed for a variety of temperature-sensitive cell division cycle (cdc) mutants at the permissive and restrictive temperatures. Asynchronous populations of some mutants do not mate at the restrictive temperature, and these mutants define genes (cdc 1, 4, 24, and 33) that are essential both for the cell cycle and for mating. For most cdc mutants, asynchronous populations mate well at the restrictive temperature while populations synchronized at the cdc block do not. Populations of a mutant carrying the cdc 28 mutation mate well at the restrictive temperature after synchronization at the cdc 28 step. These results suggest that mating can occur from the cdc 28 step, the same step at which mating factors arrest cell cycle progress. The cell cycle interval in which mating can occur may or may not extend to the immediately succeeding and diverging steps (cdc 4 and cdc 24). High frequency mating does not occur in the interval of the cell cycle extending from the step before the initiation of DNA synthesis (cdc 7) through DNA synthesis (cdc 2, 8, and 21), medial nuclear division (cdc 13), and late nuclear division (cdc 14 and 15).     

Membranes of animal cells. V. Biosynthesis of the surface membrane during the cell cycle

KB cells grown in suspension culture were synchronized by using a double thymidine block. At various times throughout the life cycle aliquots of cells were pulsed with ¹⁴C-L-leucine, ¹⁴C-D-glucosamine and ¹⁴C-choline for one hour periods. Surface membranes, cell particulates and soluble proteins were isolated and their ¹⁴C specific activities were determined. It was found that there was a marked increase in the rate of incorporation into surface membrane just after division. The pattern of incorporation was the same for all three isotopic precursors. The rate of incorporation of isotopic precursors into soluble proteins was constant throughout the cycle. Some increase in rate of incorporation of isotope into the particulate fraction was observed during division.