Differential Phosphorylation and Turnover of Nuclear Acidic Proteins during the Cell Cycle of Synchronized HeLa Cells*

SUMMARY The phosphorylation of non-histone nuclear proteins has been studied in HeLa S-3 cells during synchronous growth. The uptake of [32P]orthophosphate into individual protein bands separated on sodium dodecyl sulfate polyacrylamide gels varies at different stages in the cell cycle while the banding patterns themselves are remarkably constant. Rates of phosphate uptake into most major phosphoprotein species are increased during the early Gl and early S phases and are minimal during the late G2 to M period. The turnover of previously incorporated phosphoryl groups during a cold chase following a 23-hour exposure to [“‘PI-orthophosphate shows that some proteins lose their phosphoryl residues much more rapidly than do others. The half-lives vary from 5 to 12 hours, with an over-all half-life of 6.7 hours. In contrast, the 14C specific activity of the proteins falls to 50% of the original activity in about 25 b.ours. The retention of [32P]phosphate is a more sensitive index of differential phosphoryl group turnover in different nuclear acidic proteins than is the uptake of [32P]phosphate in short term labeling experiments. If the nuclear phosphoproteins a role in differential transcription during the cell cycle, changes in phosphorylation would appear to be more significant than changes in relative concentrations of individual protein species.

The phosphorylation of non-histone nuclear proteins has been studied in HeLa S-3 cells during synchronous growth. The uptake of [32P]orthophosphate into individual protein bands separated on sodium dodecyl sulfate polyacrylamide gels varies at different stages in the cell cycle while the banding patterns themselves are remarkably constant. Rates of phosphate uptake into most major phosphoprotein species are increased during the early Gl and early S phases and are minimal during the late G2 to M period.
The turnover of previously incorporated phosphoryl groups during a cold chase following a 23-hour exposure to ["'PIorthophosphate shows that some proteins lose their phosphoryl residues much more rapidly than do others. The halflives vary from 5 to 12 hours, with an over-all half-life of 6.7 hours. In contrast, the 14C specific activity of the proteins falls to 50% of the original activity in about 25 b.ours. The retention of [32P]phosphate is a more sensitive index of differential phosphoryl group turnover in different nuclear acidic proteins than is the uptake of [32P]phosphate in short term labeling experiments.
If the nuclear phosphoproteins play a role in differential transcription during the cell cycle, changes in phosphorylation would appear to be more significant than changes in relative concentrations of individual protein species. The experiments to be described deal with changes in the metabolic activities of nuclear phosphoproteins during the replication cycle of synchronously growing mammalian cells. Interest in this class of proteins stems from observations that they show strong indications of involvement in the positive control of gene activity (e.g. 1-13).
The question arises as to whether the nuclear phosphoproteins are altered at times of gene activation and repression during the cell cycle. It is known, for example, that the synthesis of RNA in synchronously dividing cells is suppressed during mitosis (14- 16), while the synthesis of particular messenger RNAs (e.g. histone mRNAs) is restricted to the late Gl and early S phases of the cycle (17)(18)(19).
A number of observations have been made of synthesis and phosphorylation of nuclear acidic proteins in HeLa cells (20-23). The synthesis and accumulation of acidic nuclear proteins goes on throughout the cell cycle of continuously dividing populations of HeLa S-3 cells (20). In synchronously dividing cell populations there is an increased rate of synthesis and accumulation of these proteins which precedes the onset of DNA synthesis (21). The amount of nuclear protein synthesized, transported, and retained in the acidic chromosomal protein fraction is greater immediately after mitosis and later in Gl than in the S or G2 phases of the cell cycle (22). Preliminary studies of phosphorylation of HeLa nuclear proteins during the cell cycle indicate that the rate of phosphorylation is maximal in the early S phase and decreases in the late S and G2 phases when RNA synthesis is also reduced (23). The suppression of nuclear acidic protein phosphorylation during M has recently been reported by Platz et al. (24).
The present study is a detailed analysis of the uptake and turnover of phosphate in the phenol-soluble nuclear acidic proteins of synchronously and nonsynchronously growing HeLa S-3 cells.

MATERIALS AND METHODS
Cell Culture and Synchronization-HeLa S-3 cells' were maintained in suspension culture at 2 to 6 x lo5 cells per ml by daily dilution with fresh Joklik-modified minimal essential medium (Grand Island Biological Co., Grand Island, N. Y.) containing 10% fetal calf serum and supplemented with 2.5 units per ml of penicillin G, 2.5 pg per ml of streptomycin, and 20 units per ml of mycostatin. Synchronization was obtained by the double thymidine block method (25,26), exposing the cells to 2 mM thymidine for 14 hours, to normal medium for the following 9 hours, and to 2 mM thymidine for an additional 14 hours. Cells were harvested by centrifugation at 1,500 x g for 4 min and resuspended in onefifth of the original volume of thymidine-free medium. The time of release from the thymidine block is taken as starting at this resuspension.
The cells were centrifuged and resuspended in the original volume of thymidine-free medium.
The cell cycle was monitored by measurements of cell concentration, mitotic index, and [aH]thymidine incorporation rate. The cell number was measured at different times in the cell cycle using a Coulter counter.
For determination of the mitotic index, the cells were fixed in 3: 1 ethanol-acetic acid and stained with 1 y0 (w/v) crystal violet in water.
The number of mitotic figures was scored in at least 500 cells at, each time point.
Rates of thymidine incorporation into DNA were determined by incubating l-ml aliquots of the cell suspension in the presence of 5 PCi of [3H]thymidine of specific activity 20 Ci per mmole (New England Nuclear, Inc., Boston, Mass.) for 30 min at 37". Cold medium, 3.0 ml, was added at, the end of the incubation period and the cells were collected by centrifugation at 2,000 x g for 3 min. The cells were resuspended in 3 ml of 5% trichloroacetic acid and transferred to a Millipore filter (pore size 0.45 cl). The incubation tube was rinsed twice with 3-ml aliquots of 5y0 trichloroacetic acid and the washings passed through the filter. The filter was finally washed with 10 ml of 70% ethanol (v/v) and dried at 60" for 30 min. Bray's scintillation liquid (27), 20 ml, was added to the dried filters and aH activity was measured in a Packard Tri-Carb model 3375 scintillation spectrometer. Isotopic Labeling of Nuclear Proteins-At different times after removal of the thymidine block, 4 to 8 x 10' cells were harvested by centrifugation at 1,500 x g for 4 min and gently resuspended in 5 ml of culture medium containing 2 mCi of carrier-free ["*PIorthophosphate (New England Nuclear, Inc., Boston, Mass.). After 15 min incubation at 37" the cells were centrifuged as before and the cell pellet was frozen by immersing the tube in acetone at -70".
The cells were stored at -80". Nuclei were isolated within the next 24 hours.
Estimates of turnover in the nuclear phosphoprotein fraction were based on measurements of isotope retention in the nuclear proteins cf cells which had been prelabeled with [@P]orthophosphate and [14C]leucine. In studies of asl' and 14C turnover in synchronously dividing cells, 1 liter of cell suspension containing 4 to 6 x 10" cells per ml was taken after the first thymidine block and exposed to 25 mCi of [32P]orthophosphate and 0.50 mCi of L-[ U-*4C]leucine of specific activity 316 mCi per mmole (Schwarz-Mann, Inc., Orangeburg, N. Y.) for 9 hours in the thymidinefree medium and for 14 hours of the second thymidinc block. The cells were harvested by centrifugation, washed in nonradioactive medium, and resuspended in 1 liter of thymidine-free medium for the cold chase experiments.
At the indicated times after removal of the second thymidine block, loo-ml aliquots of the suspension were centrifuged at 1,500 x g for 4 min. The cell pellet was resuspended in 5 ml of medium for transfer to smaller tubes and centrifuged at 1,500 x g for 3 min. The cell pellet was frozen and stored at -80" prior to isolation of the nuclei. For turnover studies in nonsynchronized cultures, 1 liter of cells was incubated in the presence of 25 mCi of [32P]phosphate and 160 PCi of [14C]leucine for 23 hours. The cells were washed in isotope-free medium and resuspended in 1 liter of fresh medium for the cold chase experiments.
Isolation of Non-h&one Nuclear Proteins-Nuclei and chromatin (from mitotic cells) were isolated by a modification of the method of Hancock (28) ; 4 to 8 x 10' cells were suspended in 5 ml of 80 mM NaCl-20 mM EDTA containing 1% Triton X-100, pH 7.2, and homogenized with 20 strokes in a wide clearance Dounce-type glass homogenizer (Kontes, Inc., Vineland, N. J.). The homogenate was centrifuged at 2,000 x g for 5 min to pellet, nuclei and intact cells, and the pellet was resuspended and rehomogenized as before. After centrifugation at 2,000 x g for 5 min, the pellet was again homogenized and centrifuged to sediment the nuclei.
At this point the nuclei were free of cyto-plasmic or whole cell contamination as judged by phase-contrast microscopy.
The protein to DNA ratio of the final nuclear pellet was about 2.8: 1.
The nuclear phosphoprotein fraction was prepared by the method of Shelton and Allfrey (29) as modified by Teng et al. (2). The nuclei were extracted twice with 0.14 M NaCl and twice with 0.25 N HCl.
The residue was washed once with 1: 1 chloroform-methanol containing 0.2 N HCl and once with 2:l chloroform-methanol containing 0.2 N HCl. The distribution of proteins in the various extracts is summarized in Table I. The residue was then resuspended in 0.1 M Tris-HCl, pH 8.4, containing 0.01 M EDTA and 0.14 M 2-mercaptoethanol.
The suspension was mixed gently with an equal volume of phenol (saturated with the buffer) and allowed to stand for 14 hours at 2". The mixture was homogenized briefly and centrifuged at 12,000 x g for 10 min. The aqueous phase was collected, reextracted with an equal volume of phenol, and centrifuged as before.
The combined phenol extracts containing the nuclear phosphoproteins were dialyzed against a series of urea-containing buffers (2) to restore the proteins to the aqueous phase.
Electrophoretic Analysis of Non-Histone Nuclear Proteins-The fractions to be analyzed by polyacrylamide gel electrophoresis were dialyzed overnight against 0.01 M sodium phosphate buffer, pH 7.4, containing 0.1% SDS? and 0.14 M 2-mercaptoethanol.
The proteins were separated by electrophoresis in 10% polyacrylamide gels containing 0.1 y0 SDS, prepared as described previously (2). Electrophoresis was carried out at 6 ma per tube until a pyronin Y marker reached the bottom of the tube. About 10 hours was usually required.
The gels were stained with 1 To fast green in 7% acetic acid-35% methanol for 12 hours and then destained in 7 y0 acetic acid-35 y0 methanol. Densitometric analysis at 615 nm of the stained gels was carried out, in a Gilford spectrophotometer equipped with a model 2410 S linear transport device. Estimates of the molecular weights of individual protein bands were based on mobility versus molecular weight plots for proteins of known molecular weight, using as standards horse heart cytochrome c (mol wt 12,000), myoglobin (mol wt 17,800), alcohol dehydrogenase (mol wt 36,000), ovalbumin (mol wt 45,000), bovine serum albumin (mol wt 67,000), and y-globulin (mol wt 160,000), all measured under identical electrophoretic conditions. For measurement of isotope distributioq in different protein bands the gels were swollen by immersion in 7% acetic acid, frozen by contact with solid COz, and sliced transversely in l-mm slices. Each slice was solubilized by incubation with 200 ~1 of 50% HzOz at 60" for 12 hours. Scintillation liquid, 15 ml, (Omnifluor (New England Nuclear, Inc.) 4 g per liter of 1: 1 methyl Cellosolve-toluene) was added and 3zP and 14C activities were measured by scintillation spectrometry. Pulse Labeling of ATP Pools with [a2P]Orthophosphate-At different times after removal of the thymidine block, cells were harvested by centrifugation and gently resuspended in 5 ml of culture medium containing 2 mCi of carrier-free [a2P]orthophosphate.
After 15 min incubation at 37", the cells were centrifuged as before and the cell pellet was frozen by immersing the tube in acetone at -70".
The frozen cells were then homogenized in 4 ml of ice-cold 0.6 N HC104, centrifuged, and re-extracted with 4 ml of 0.2 N HC104. The extracts were combined and titrated to pH 7.0 with 5 N KOH.
After 12 hours at 4", the KClO, precipitate was removed by centrifugation, and the 2 The abbreviations used are: SDS, sodium dodecyl sulfate; cyclic AMP, CAMP, adenosine 3':5'-monophosphate. supernatant was filtered through Whatman No. 3 paper. Aliquots of the clear supernatant were subjected to chromatography on Dowex 1 (formate) columns (0.6 x 5.0 cm) as described by Hurlbert (30). After application of the sample, the columns were washed with distilled water and with 1.0 M ammonium formate.
The ATP fraction was eluted with 2 M ammonium formate-0.75 M formic acid, and the ATP was further purified by chromatography on Whatman No. 3 paper in ethanol-l M ammonium acetate, pH 7.3-water, 66.5 : 30 : 3.5 (v/v/v). Spots corresponding in RF to authentic ATP standards were cut out and the ATP eluted and analyzed by measuring absorbance at 259 nm and azP activity.
Measurement of Cyclic AMP-dependent Protein Kinase Activity-The levels of cyclic AMP-dependent protein kinase activity were measured in high speed supernatant fractions obtained after homogenization of HeLa cells in 0.32 M sucrose containing 0.1% Triton X-100, using a Potter-Elvehjem homogenizer (0.15.mm clearance) at 2,000 rpm with 20 up-and-down strokes. The homogenate was centrifuged for 60 min at 100,000 x g, and aliquots of the supernatant containing 20 pg of protein were incubated at 31" with 100 pg of histone Fl as substrate and [y-azP]ATP in the presence or absence of 10m6 M cyclic AMP as described previously (31,32). The reaction was stopped by the addition of 5% trichloroacetic acid containing 0.25% sodium tungstate.
The protein precipitate was washed and assayed for a2P activity as described (32).
Chemical Analyses-Protein was determined by the method of Lowry et al. (33) using bovine serum albumin as a standard for nuclear acidic protein determinations, and calf thymus histone as a standard for histone analyses.
DNA was determined by the diphenylamine reaction as modified by Burton (34), using highly polymerized calf thymus DNA as a standard.
The amino acid compositions of the phosphoprotein fractions were determined by ion exchange chromatography after the method of Spackman et al. (35). The Beckman amino acid analyzer 120B was modified for a 7-fold increase in sensitivity by insertion of a Honeywell expanded range card. For determination of alkali-labile phosphate, the proteins were first dialyzed exhaustively against distilled water and hydrolyzed in 1.0 N NaOH at 100" for 5 min. Inorganic phosphate released into the supernatant was analyzed as the phosphomolybdate complex after acidification with 0.1 ml of 4 N HCl-1 N HzS04 and precipitation of the protein with 0.1 ml of 0.1 M silicotungstic acid in 0.1 M HZS04. To the clear supernatant were added 1.5 ml of 5% ammonium-molybdate in 4 N HzS04. The phosphomolybdate complex was extracted in 2.5 ml of 1:l isobutyl alcohol-benzene, reduced with SnC12, and measured at 660 nm (2,36).

AND DISCUSSION
Composition of HeLa Nuclei (or Chromatin) at Di$erent Stages in Cell Cycle-Nuclear protein and DNA contents have been examined in synchronously growing HeLa S-3 cells at different times after release from a double thymidine block (25,26). The degree of synchrony obtained by this procedure is illustrated in Fig. 1, which plots three parameters of growth: cell number, mitotic index, and rate of [aH]thymidine incorporation into DNA. The S phase, as measured by the rate of DNA labeling, begins immediately on release of the cells from the thymidine block and lasts for 5 to 6 hours. Maximal rates of DNA synthesis occur at 3 hours. By 8 hours the majority of the cells has entered mitosis.
New cells are first evident at about 7% hours; by 12 hours the population has almost doubled. The timing of these events is highly reproducible, and was determined for each of the isotope-labeling experiments described in this communication.
The degree of synchronization, as calculated by the method of Engelberg (37,38) is 63% for the cells shown in Fig. 1. On the basis of this data, we consider the G2 period of the cell cycle to extend from 5% to 735 hours, while the Gl phase of the following cycle is taken as the period from 8% to 12 hours.
In order to monitor changes in nuclear proteins during the cell cycle, we have counted the number of cells in the population and analyzed for DNA content and protein distribution in different nuclear extracts.
The results are summarized in Table I. A comparison of the figures for DNA content of the original cell suspension (Column 3) and the DNA recovered in the nuclear (or chromatin) fractions (Column 4) shows that recovery is very high at all stages in the cell cycle, with an average recovery of 86.1 f 1.7%. This minimizes the risks of artifact due to the selection of a small or variable fraction of the nuclei for analysis.
There are significant variations in the protein to DNA ratio of the nuclei during synchronous cell growth, ranging from 2.19 to 3.27 to 1 (Column 6). This ratio drops in the late S phase and increases during the Gl period of the following cycle. The changes are largely due to varying proportions of the non-histone nuclear proteins, because the histone to DNA ratios are not appreciably altered during the cell cycle but remain constant at about 1.1 to 1. This is consistent with observations that histone and DNA synthesis proceed concomitantly throughout the S phase of HeLa cells (19,28,39).
In the fractionation procedure employed to separate nuclear proteins about 40% of the total nuclear protein is removed when the nuclei are washed twice in 0.14 M NaCl (Table I, Column 8).
The acid-soluble proteins, mainly histones, comprise another 4057, of the total nuclear proteins.
Little loss of protein occurs in the chloroform-methanol-HCl washes; the bulk of the extracted material comprises lipids and phospholipids.
The residual proteins, which are then extracted in phenol, comprise about 13% of the total protein in the isolated nuclei. This fraction includes many of the nuclear phosphoproteins. was incubated for 23 hours. An aliquot containing 4 X 10' cells was mixed with 4 X lOa unlabeled cells for isolation of nuclei and extraction of the nuclear proteins as described under "Materials and Methods." The protein content, alkali-labile phosphorus content, and a2P activity were determined for each fraction.
* The phosphoriis content per fig of protein and the specific a2P activity per mg of protein are not given for the chloroform-methanol extract because of the high contamination by radioactive phospholipids in this fraction.
The amount of phenol-soluble protein per nucleus varies during the cell cycle, as judged by the phosphoprotein to DNA ratio at different times after release from the thymidine block. This ratio falls from 0.42 to 1 in the early S phase to 0.28 to 1 in G2 and M (Table I, Column 12). It follows that a substantial increase in the acidic protein complement of the nucleus must occur in the prereplicative phase of the cycle. It should be noted that the high recoveries of protein obtained (approximately 9S'r0) minimize the possibilities of artifact due to differential extraction of proteins depending upon stages in the cell cycle. As a further check, we have compared the recovery of phenol-soluble proteins from metaphase chromosomes (isolated from cells blocked in mitosis by exposure to vinblastine sulfate for 16 hours) and from nuclei obtained from an unsyn chronized cell population.
No indications of differential extractability were obtained. The distribution of alkali-labile phosphorus in the different nuclear subfractions is shown in Table II. It can be seen that the phenol-soluble proteins, representing only 13 to 14y0 of the total nuclear protein, contain about one-third of the total phosphorus and also account for about one-third of the total [Vphosphate incorporation in a long term (23.hour) labeling period.
The proteins extractable in 0.14 IVY NaCl contain only 7% of the total V incorporated.
The HCl-soluble protein fraction, mainly histones, contains about half of the total counts. It should be noted that the proteins of the phenol-soluble fraction have both the highest specific azl' activity and the highest phosphorus content of the nuclear fractions analyzed.

Characterization of Nuclear Phosphoprotein
J'raction-The nuclear phosphoprotein fraction comprises a heterogeneous mixture of proteins differing in molecular weight, amino acid composition, and degree of phosphorylation.
The molecular size heterogeneity is indicated by differences in electrophoretic mobility in SDS-polyacrylamide gels. A complex banding pattern is obtained which shows the presence of multiple polypeptide chains ranging in molecular weight from 18,000 to 170,000 (Fig.  2). At least 21 major bands are detectable; each of these, in turn, may include different protein species of similar or identical molecular weights.
A comparison of the banding patterns of the phenol-soluble nuclear proteins at different stages in the cell cycle is presented in Fig. 2 with the findings of Bhorjee and Pederson (40), but, on the whole, it is the uniformity of the protein pattern, rather than its variability, which attracts attention.
This result is in contrast with findings that the nuclear phosphoprotein complement changes appreciably in cells during the course of differentiation (41)(42)(43)(44)(45). It has been noted previously that the synthesis of nuclear acidic proteins proceeds throughout the HeLa cell cycle (20-22), with increasing rates of synthesis in late Gl (21). The reproducibility of the gel patterns at different times, shown here, suggests that the synthesis and accumulation of many of the major nuclear acidic proteins are under close coordinate control.
The amino acid analyses presented in Table III confirm the impression of constant proportionality of the phenol-soluble proteins throughout the HeLa cell cycle, although one would only expect to detect gross differences by this technique.
The average amino acid composition of the nuclear phenol-soluble protein fraction indicates a clear predominance of the acidic amino acids, aspartic and glutamic acid (21 mole %) over the basic amino acids, lysine, arginine, and histidine (15 mole %). This is in accord with findings in other cell types (2  cells were frozen immediately and the nuclear proteins were fractionated as described under "lMaterials and Methods." The specific activity of the phosphoprotein fraction was determined and plotted against time, as shown in Fig. 3. Two peaks of phosphate incorporation are evident. The first occurs early in the S phase (between 136 and 3 hours) and the second peak occurs in early Gl (at about 10 hours).
The rate of phosphate uptake appears to be somewhat greater in S than The possibility that estimates of the rate of nuclear protein phosphorylation might be in error due to injurious effects of the thymidine double block was tested by comparing azl' uptakes in control (unsynchronized) cell cultures and in cultures exposed to 2 mM thymidine for 2 or 4 hours. As can be seen from the data in Table IV, the phosphate uptake into the nuclear proteins of control and thymidine-treated cells is virtually identical. Moreover, no differences in gel electrophoretic patterns could be discerned. Other control experiments have established that cytoplasmic protein fractions do not contribute significantly to the radioactivity of the nuclear fractions we have analyzed. This possibility was tested by preparing the nuclear phosphoproteins from unlabeled cells which were homogenized in the presence of a a2P-labeled postnuclear supernatant fraction from homogenates of cells incubated in the usual way with 2 mCi of ["PIorthophosphate. Less than 4.8 % contamination was observed (Table IV).
The distribution of [a2Plphosphate in different size classes of nuclear phosphoproteins after pulse-labeling for 15 min was determined by radioassay of the multiple bands separated by SDS-polyacrylamide gel electrophoresis (Fig. 4). The staining pattern shown at the bottom of the figure is aligned with the corresponding densitometer tracing in the top panel. The other panels compare the distribution and specific activities of the nuclear phosphoproteins at the indicated stages in the cell cycle.
It is evident that there is a marked heterogeneity in ["'PIphosphate incorporation into the proteins at different regions of the gel. In some cases the peaks of a2P activity do not coincide exactly with the positions of the major protein bands. This strongly suggests that minor bands, not visible because of their low concentration, may be contributing disproportionately to the 32P uptake measurements.
The azP activity of the individual protein bands follows quite closely the cell cycle-dependent changes described for the total phenol-soluble protein fraction. Labeling is greatest in the S and Gl phases and depressed in the G2 to M phase. Some minor differences in the rate of labeling of different bands can be observed, but, on the whole, the pulse-labeling experiments indicate a parallel metabolic response of many diverse nuclear proteins to events occurring at different stages of the cell cycle.
The fluctuation observed in the rates of 3ZP incorporation into the non-histone proteins at different stages in synchronous cell growth does not appear to depend on fluctuations in the specific activity of the cellular ATP pools. The incorporation of [32P]orthophosphate into ATP pools was measured in cells at different stages of the cycle. As the results in Table V show, the specific activity of the cellular ATP pools remains relatively uniform in the course of synchronous growth.
The minor variations listed in Table V do not seem significant and are not likely to account for the large differences in the specific activity of the phosphoprotein fraction at different stages. The fluctuations in [32P]phosphate incorporation into the nuclear proteins are more likely to represent alterations in protein kinase activities in nuclei at different stages of the cycle. It is known that the phosphorylation of some acidic nuclear proteins is CAMP-dependent (32,46). Cyclic AMP levels are known to fluctuate throughout the cell cycle of synchronously growing HeLa cells (47) with minimal concentrations in the late G2 to M period.
This suggests that the major changes in nuclear protein phosphorylation in the HeLa cell cycle reflect the changing activities of the CAMP-dependent protein kinases.
Histone Phosphorylation in Cell Cycle-The rate of [32P]orthophosphate incorporation into the histones of synchronized HeLa cells varies throughout the cell cycle as shown in Fig. 5. The peak of phosphorylation is observed at 3 hours, which coincides with the peak of DNA synthesis.
Correlations between histone phosphorylation and DNA synthesis have been noted before (49-52) ; the phosphorylation of histone fraction Fl is known to be dependent on cell cycle position and to be active in the S phase (50)(51)(52).
Some differences exist between the timing of histone phosphorylation and phosphate uptake into the non-histone proteins of the nucleus. The peak of histone phosphorylation occurs somewhat later in the cell cycle than the corresponding peak for the nuclear phenol-soluble proteins (Fig. 3). Moreover, no major peak in histone phosphorylation appears in Gl as it does for the more acidic protein fraction.
These differences argue strongly against the view that the variable rates of phosphorylation of the nuclear acidic proteins simply reflect differences in the specific activities of nuclear ATP pools at different phases of the cycle. Some preliminary experiments have been carried out to determine whether corresponding changes in kinase activity toward histone Fl can be detected at different stages of synchronous growth.
The soluble protein kinase activity of HeLa cell homogenates was compared at different times after release from the thymidine block, both in the presence and absence of cyclic AMP.
The activation of histone Fl phosphorylation by cyclic AMP varies from 3-to 6-fold during the cycle, being highest at early S and lowest in early M (Table VI).
Turnover of Phosphoprotein-Phosphate at Di$erent Stages of Cell Cycle-The retention by nuclear proteins of previously incorporated phosphate groups was compared at different times after a, .z + 0 u 60   L  I  I  I  I  I  I   0   3  6  9  12  15 Hours after releose from thymidine block  teins has also been confirmed by analysis of the alkali-labile phosphate activity in the protein samples. The rate of 32P turnover varies in different nuclear phosphoproteins.
The distribution of [32P]phosphate in different size classes of proteins separated by SDS-polyacrylamide gel electrophoresis is shown in Fig. 7 U, the ratio of z2P activity to '4C activity is plotted against time. The decreasing ratio indicates that phosphate groups in the protein are subject, to replacement without a corresponding degradation of the polypeptide chain.
release of the cells from the thymidine block.
In these experiments, the acidic nuclear proteins were prelabeled in a 23.hour incubation in the presence of [3*P]orthophosphate and [14C]leutine, as described under "Materials and Methods." After washing to remove the radioactive precursors (and thymidine), the cells were incubated in a nonradioactive medium. Aliquots of the suspension were withdrawn at different times during the "cold chase" for preparation and analysis of the nuclear phenolsoluble phosphoprotein fraction. The results summarized in Fig. 6 compare the retention of [Y'lphosphate and [14C]leucine in the total phenol-soluble protein fraction.
After a brief initial period in which specific activities increase slightly, radioactivity is lost following an exponential decay curve. The rate of 32P loss greatly exceeds that of the 14C label; protein phosphate activity declines with an average half-life of 6.7 hours, while the 14C activity of the proteins falls to 50% of the original activity in about 25 hours. These divergent results indicate that the phosphoryl groups in the proteins are subject to removal without a corresponding degradation of the polypeptide chain. The loss of [3*P]phosphate from the pro- experiments. It follows that the retention of [3*P]phosphate is a more sensitive indes of differential phosphoryl group turnover in different nuclear acidic proteins than is the uptake of [32P]phosphate in short term incubations.
To rule out the possibility that such different rates of azP turnover in different nuclear phosphoproteins might be an artifact due to prolonged exposure of the cells to thymidine (possibly permitting selective protein degradation during growth arrest), similar experiments were carried out in nonsynchronously growing HeLa S-3 cultures.
The results summarized in Fig. 8 compare the retention of [Y'lphosphate and ['4C]leucine in the protein subfractions soluble in 0.14 M NaCl, 0.25 N HCl, and phenol. As in the synchronously growing cell cultures, the rate of 32P loss from the phenol-soluble proteins exceeds the rate of decline of the 1°C specific activity.
The ratio of 3*P:14C activities consequently decreases rapidly with time during the cold chase (Fig.  8). The kinetics of 3*P turnover appear to be characteristic for the different nuclear subfractions.
In agreement with the results obtained in synchronously growing cells the rate of 3*P turnover varies for different proteins in the phenol-soluble fraction. The distribution of [3sP]phosphate in different protein bands separated by SDS-polyacrylamide gel electrophoresis is shown in Fig. 9 for 3 and 9 hours after commencement of the cold chase. Three major phosphorylated bands (of molecular weights 28,000, 47,000 and 55,000) are evident. The rapid QP turnover in proteins of molecular weight 55,000 and the relatively slow turnover in proteins of molecular weight 28,000 are in agreement with results obtained in the synchronously growing cell populations (see Fig. 7). Fig. 9 also shows the distribution of the [14C]leucine-labeled proteins in the electrophoretic pattern.
It is clear that in some cases, the peaks of 32P activity do not coincide exactly with the positions of the major [14C]leucine-labeled bands.
Phosphate Content of Isolated Nuclear Phosphoproteins-The average phosphorus content of the total nuclear phenol-soluble protein fraction was determined at different times after release from the thymidine block. There are no major differences in alkali-labile phosphorus content of the proteins prepared at different stages of the cycle (Table VII) The proteins were separated by electrophoresis in 0.1% SDS-lo% polyacrylamide gels. The distribution of 32P (O-----O) and 1°C (C---0) activity is shown after 3 hours and after 9 hours in the nonradioactive medium. Note the rapid rate of loss of 32P from the protein band at molecular weight 55,000 and the relatively slower loss from bands at molecular weights 28,000 and 47,000.  The studies of isotope retention in both synchronous and nonsynchronous cell cultures show that individual nuclear proteins exchange their phosphate groups at differing rates. The complexity and diversity of these structural modifications are not evident in the over-all phosphate level of the phenol-soluble fraction.
Based on an average phosphorus content of about 0.1% (by weight), it can be estimated that an average protein of molecular weight 120,000 containing about 73 seryl residues (Table III) might contain only about 4 of these in the phosphorylated form. (Since not all proteins in the phenol extract are equally phosphorylated, some will be more and others less phosphorylated than this average figure indicates.) CONCLUSION Although the relative concentrations of the major nuclear phenol-soluble phosphoproteins appear to be stable, there are clear differences in rates of phospharylation of the nuclear proteins at different stages in the cell cycle. Protein phosphorylatian is most active at periods when RNA synthesis is high (Gl and S) and it is minimal in the G2 to M period when RNA synthesis is suppressed.
These changes are not due to changes in the specific activity of the cellular ATP pools which remain relatively constant.
They do parallel changes in cellular cyclic AMP concentrations and probably reflect alterations in nuclear protein kinase activities at different phases of the cycle.
Phosphate incorporated into the nuclear acidic proteins is not stable, but turns over at rates which differ from one protein species to another.
The phosphate groups are subject to exchange without degradation of the polypeptide chain.
If the nuclear phasphoproteins play a role in the control of transcription during the cell cycle, changes in their phosphorylation would offer a mechanism for modifying their structure and influencing their interactions with other components of the chromosome.
In a nondifferentiating cell population, this may be more important than changes in the relative concentrations of t,he individual protein species.