Protein Metabolism during Growth of Vero Cells GLORIA T-Y. LEE AND DEAN L. ENGELHARDT Department of Microbiology, Columbia University, College of Physicians and Surgeons, 701 West 168thStreet, New York, New York, 10032
ABSTRACT Protein synthesis and degradation were studied throughout a growth cycle of Vero cells. The rate of protein synthesis, measured a s the rate of amino acid incorporation, reached a maximum a t the mid-exponential phase and declined to 10-30%of the maximum in the stationary phase. The rate of protein degradation, measured a s the release of radioactive amino acids from uniformly labelled cellular proteins, did not vary in the growth cycle. The amount of protein per cell, measured by a n isotopic method, remained constant when normalized to account for the variation in the proportion of actively dividing cells in the cell population during the growth cycle. Cellular protein was determined using this method since i t was found t h a t the chemical determination of the amount of protein in the monolayer was not accurate during the early stage of the growth cycle. This was due to a significant amount of serum protein adsorbed to the cells. In this study we were able to show that, in Vero cells, protein synthetic activity is correlated with the rate of cell division, and variations in the rate of synthesis alone are sufficient to meet the changing requirements for cellular protein in a growth cycle. The overall metabolism of protein in cells has been the subject of considerable interest (Foster and Pardee, '69; Meisler, '73b; Pine '72; Rechcigl, '71; Salzman, '59). Studies in this area have revealed that the amount of protein in a cell is determined by the dynamic relationship between synthetic and degradative processes within the cell and export from the cell. This paper describes an animal cell culture system in which a balanced state of protein metabolism (Campbell, '57; Maabe and Kjeldgaard, '66) was achieved during rapid, exponential growth and in the stationary phase. The approach employed was to permit the cells to grow a t least nine generations before reaching the stationary phase. Such a long growth cycle permitted the cells t o recover from pertubation caused by the initial transfer process, and to approach equilibrium with regard to biosynthetic activity, the major variable, before attaining the phase of decelerating cell division. In this system we were able to analyze the inputs (synthesis and uptake) and the outputs (degradation, cell division, and leakage) of cellular protein and to establish the relationship between these parameters and cell proliferation. J. CELL. PHYSIOL.. 92: 293-302.
MATERIALS AND METHODS
Vero cells were purchased from the American Type Culture collection. Dulbecco's modified Eagle's medium (DME) and leucine, isoleucine, and histidine-deleted DME were from Grand Island Biological Co. Calf serum was from Flow Laboratories. O-methyl-L-threonine and L-histidinol dihydochloride were from Calbiochem. L-leucine ['TI (U) (312 mCi/mmole) was from SchwarzIMann. L-leucine [4,5-3H1(5Ci/mmole) was from New England Nuclear. Plastic tissue culture ware was from Lux Scientific Co.
Cell growth and maintenance Vero cells were grown on tissue culture dishes in DME supplemented with 10% calf serum a t 37°C in 5% COz, and 100%humidity. The cells were detached from the growth surface with trypsin (0.05%w/v) in phosphate buffered s a l i n e (PBS) (150mM NaC1, 4mMKC1, 9.5mM phosphate, pH 7.2) plus 0.5 mM ethylenediaminetetraacetic acid (EDTA) and subcultured every four or five days. The cells were routinely screened for mycoplasma Received Aug. 30, "76. Accepted Jan. 19, '77. )This research was supported by a grant from The National Science Foundation BMS 74-02202 A01.
293
294
GLORIA T-Y.LEE A N D DEAN L. ENGELHARDT
by plating on mycoplasma growth agar and by the method described by Todaro e t al. ('71). Cells in early exponential phase were used for each growth experiment. During the course of the various growth cycles reported in this paper the cells employed were subject to daily medium change. Cell number was determined by the following two methods: (1)For cells at. a density higher than 2 x lo4 cells/cm2, the cell number was obtained by counting the cells in suspension with a hemocytometer. ( 2 ) For cells a t a density lower than 2 X lo4 cells/cm2, the cells were fixed and stained with crystal violet (0.1% (wiv) in 20% ethanol) and counted directly on the petri dish with the aid of a grid. Cellular protein was determined by Lowry's procedure (Lowry et al., '51) as modified by Oyama and Eagle ('56) or by an isotopic technique as described in the RESULTS section. When protein was determined by the procedure of Oyama and Eagle ('561, growth medium was removed, the cell layer was washed three times with PBS, and dissolved in alkaline copper solution. Petri dishes with growth medium only were employed to determine the background amount of protein (-0.3 yg/cm2) that adsorbed non-specifically to the growth surface. This background number was subtracted from all values reported. For cells grown in the presence of amino acid analogues the respective amino acid was deleted from the DME. Note, however, that trace amounts were supplied by the serum (Meisler, '73b).
rial was constant for a t least two hours after the addition of the label.
Specific activity of leucyl-tRNA: To measure the specific activity of leucyltRNA, cells were pulsed with 3H-Leua t 5 yCi/ ml in DME (0.08 mM leucine) plus 10% calf serum for 30 minutes. At the end of the pulse, the cells were washed with cold PBS, scraped off the petri dish, lysed with RSB (10 mM Tris-HC1, pH 7.4, 10 mM NaC1, 1.5 mM MgC1,) supplemented with 0.5%NP-40. Nuclei were removed by centrifugation. The supernatant was adjusted to 0.4 M LiCl, 20 mM Sodium Acetate, pH 5.3, and extracted with phenolchloroform-isoamyl alcohol (50:50: 1). The RNA was precipitated three times with 2.5 volumes of ethanol to remove trace amounts of free leucine. The amount of leucyl-tRNA was then determined as described by Baenziger e t al. ('74). Specifically, the aminoacyl bond was hydrolyzed a t pH 10 for ten minutes a t 37". The amount of leucine released from leucyl-tRNA was determined by a Beckman amino acid analyzer. Leucine was quantitated by measuring chart peak area relative to that of a 10-nmole leucine standard. Radioactivity in the sample applied to the amino acid analyzer was determined in a scintillation spectrophotometer.
Cellular protein degradation
The procedure used was based on that of Poole and Wibo ('73) with the following modifications. Cells growing on a 60-mm petri dish were labelled for 24 hours with 0.5 yCi/ Radioactive labelling experiments All of these experiments were performed 18 ml '%-leucine in DME plus 10% calf serum. hours after the last medium change. To mea- Medium was then removed, and the cells were sure the incorporation of radioactive leucine washed three times with prewarmed noninto trichloroacetic acid (TCA)-precipitable radioactive medium and overlaid with 6 ml of material in vivo, the growth medium was re- this medium. After two hours (a time suffiplaced with prewarmed labelling medium cient to reduce the intracellular TCA-soluble (regular growth medium, unless specified, radioactivity to a minimal level), medium was with either I4C or 3H-leucine). At the end of again removed and another 6 ml of non-radioas the 30-minute pulse, the medium was re- active medium was added. This was defined time T=O. One-half milliliter was removed moved, and the monolayer was washed three every 30 minutes, and the radioactivity in times with ice-cold PBS. The cells were solubilized with 0.025 M Na,CO",, 0.05 M this medium was assayed in Bray's ('60) soluNaOH for 15 minutes a t room temperature. tion. The resultant value was corrected for TCA was added t o a final concentration of 10% the volume change of the medium throughout (w/v), and the precipitate was collected on a the experiment. Total TCA-insoluble radioglass fiber filter (GF/A-Whatman). Radio- activity at T=O was determined in parallel activity was then determined in a scintilla- plates. The TCA-precipitable radioactivity retion spectrophotometer. The rate of incorpo- leased during the course of the experiment ration of leucine into TCA-precipitable mate- was also determined in parallel plates.
295
PROTEIN METABOLISM DURING GROWTH OF VERO CELLS
I00
80
5 2 60 2 2 LL
0 40
E
W
a
20
01
-r i 5
10
1
5 I
I
' ' ' 11 0
I
I
' '
'
15 '
1
'
I;
0
TIME (DAYS) Fig. 1 The rate of amino acid incorporation during the growth cycle. The cells were plated at (A) 1 X 10' cells/cm2; (B) 1 X lo3cells/cm2. The saturation density ranges from 9 X lo5to 1.2 X lo6cells/cm2in different experiments. This is taken as 100%.Increases in cell number (- 0 -1 were plotted on a logarithmic scale (lefthand ordinate). 'C-leucine was used t o measure the in vivo rate of amino acid incorporation (-@ -) (righthand ordinate). In panel A, 1FCi/ml of 'C-leucine in DME (0.8mM leucine) plus 10%calf serum was the labelling medium. 876 CPM/105cells is the 100%determination with a labelling period of 30 minutes. In panel B, 0.5 FCi/rnl of IC-leucine in DME (0.08 mM leucine) plus 10%calf serum was the labelling medium. 4600 CPM/105 cells is the 100%determination with a labelling period of two hours. Each number represents the average of triplicate determinations. The variation is within 15%.
3H-acetylated serum
RESULTS
Protein synthesis in Vero cells during Free amino groups in serum proteins were growth acetylated with 3H-acetic anhydride (AmerI t has been reported by Engelhardt and Sarsham, 500 mCi/mmole) according to the procedure described by Roseman and Doffner ('56). noski ('75) that the relative activity or Free label was removed through extensive di- amount of various components of the protein alysis against PBS, and finally against DME. synthetic machinery in Vero cells fluctuates in the growth cycle. It was not clear from The 3H-acetylatedserum (specific activity 3.6 these data, however, at which stage of the x lo3 cpm/pg protein) was diluted 10-fold growth cycle these changes occurred, since with unlabelled calf serum, before being used. the cells were not grown for time periods long To determine the adsorption of serum protein enough to permit a clear separation of the onto newly plated cells, cells were plated in various growth phases. Accordingly, cells DME with 10%acetylated calf serum. Twenty were grown such that the cell number inhours later the plates were washed three creased 100-fold (6.6 divisions per cell), and times with cold PBS. The material remaining the incorporation of leucine into TCA-precipon the plates was solubilized and counted as itable material was monitored throughout described above in the section on radioactive this period. The results, presented in figure lA, demonstrated that under these growth labelling.
296
GLORIA T-Y. LEE AND DEAN L. ENGELHARDT
I
0
60
I
120
I
180
I
I
240
TIME (MINI Fig. 2 The time course of release of radioactive leucine from prelahelled cells. The procedure of Poole and Wibo, ('731, was followed with the following modifications. The release of radioactive leucine was measured 2-6 hours after the removal of radioactive leucine in exponential phase cells ( - 0 -1; 2-5 hours (- 0-) and 15-18 hours ( - A - ) in stationary phase cells. Each point is the average of three determinations. The variation in individual determinations is no more than 20%.Radioactivity released is expressed as the percentage of TCA-insoluble radioactive material remaining on the plate a t the time when samples were first taken Le., 2 hours after the removal of label). Exponential phase cells were a t a density of 3 X 10' ceIls/cm'. Stationary phase cells were a t a density of 8 X lo5 cells/cm2.
conditions a t least three and, perhaps four phases of the growth cycle could be defined (Monod, '58; Skehan, '76). They were the phase of accelerating rate of cell division (phase 2) on day 1, the exponential phase (phase 3) on days 2-6, the phase of decelerating rate of division (phase 4) on days 7-9, and the stationary phase (phase 5) on days 1011. The lag phase (phase 1)has never been observed. The measurement, of the rate of leucine incorporation into TCA-precipitable material, reported in figure lA, showed a maximum early in the growth cycle. This maximum occurred on day 2 or 3 in repeated experiments. Thereafter, the rate of leucine incorporation declined throughout the rest of the growth cycle, and reached a minimum when cells entered the stationary phase. While these data demonstrated that the decline in leucine incorporation occurred before cells had entered phase 4, they did not indicate whether the maximum occurred in phase 2 or phase 3. For example, i t was possible t h a t the maximum could have occurred as cells were in phase 2 and the observed low rate on day 1 in this particular experiment (fig. 1A) was due to the slow recovery of the cells from
the transfer process. In order to distinguish phase 2 from phase 3 more clearly, a growth cycle of more than nine cell divisions was attempted. In this experiment prolonged phases 2 and 3 were observed. Again the incorporation of leucine into TCA-precipitable material was followed throughout this period. The results are presented in figure 1B. In this growth cycle, phase 2 occurred on days 1 to 7, phase 3 on days 8 to 14, and phases 4 and 5 on days 15 to 20. The incorporation of leucine reached the highest rate on days 10 to 12, the mid-exponential phase. Thus the maximal rate of incorporation occurred during the period of maximal rate of cell division. The amount of leucine incorporation into TCA-precipitable material is a valid measure of the amount of protein synthesized if the specific activity of the leucine entering proteins remains invariant through the growth cycle. To explore this question, the specific activity of leucyl-tRNA in cells in exponential phase and stationary phase was determined and found to be 194 CPMhmole and 160 CPMhmole respectively. The difference in the specific activity of leucyl-tRNA was not sufficient to account for the 3- t o 10-fold de-
PROTEIN METABOLISM DURING GROWTH OF VERO CELLS
297
crease in the amount of leucine incorporated, which was constantly observed in repeated experiments. Thus, we conclude that the protein synthetic activity fluctuated during the cell growth cycle. I t reaches a maximum in the mid-exponential phase and approaches a minimum in the stationary phase.
degradation was observed under these conditions. These results indicated that reutilization of amino acids did not occur, under these conditions, a t a level sufficient to affect our measurement of the rate of cellular protein degradation. Thus, we conclude that the rate of protein degradation in Vero cells does not vary during the growth cycle.
Protein degradation during growth The rate of protein degradation was also determined for cells in different phases of the growth cycle. One measure of protein degradation in cultured cells is the release of amino acids from prelabelled cellular protein into the growth medium (Poole and Wibo, '73). When protein degradation was determined with this procedure, as indicated in figure 2, cells in mid-exponential phase (at the point of maximal protein synthesis) and cells in the stationary phase (at the point of minimal protein synthesis) had similar rates. The observed values were 1.06% HR-' for growing cells and 1.11%HR-' for stationary phase cells. Less than 1%of the released radioactivity was TCA-insoluble. To test for the possibility of reutilization of labelled amino acids, the concentration of leucine in the growth medium was varied from 0.16 mM t o 4.0 mM, and also the rate of degradation was measured in the presence of histidinol or cycloheximide where the rate of protein synthesis had been reduced to less than 3% of the untreated control. The results of these various experiments are presented in table 1. Little difference in the rate of protein
Amount of protein per cell during growth
TABLE 1
The effectofprotein synthesis inhibitors and varying leucine concentration on protein degradation Rate of protein degradation X per hour
Control 0.16 mM leucine 4.0mM leucine 2 mM L-histidinol 30 pgiml cycloheximide
1.36 1.49 1.36 1.37 1.35
'
Amino acid incorporation %,ofcontrol
100
-
2.78 2.23
' The rate of protein degradation was determined a s described in MATERIALS AND METHODS. Cells were at 4 X 10' cells/cm2, and were incubated with respective media 30 minutes prior to the start of degradation measurement. The release of radioactive amino acid was determined over a 4-hour period. 2DME plus 10%calf serum, leucine concentration was 0.8 mM. "repared in DME, lacking L.histidine, with lo'% calf serum. iCells in parallel plates were labeled with "c-leu (0.5pCi/ml with 0.08 mM Leu in DME) for the time period of degradation measurement; 12,110 cpmiplate for control was taken as 100%.
Since the overall level of cellular protein is determined by the synthesis and degradation of protein in the cell, the amount of cellular protein was determined throughout t h e growth cycle to establish the correlation among these parameters. The amount of protein per cell was measured by labelling cellular protein with a radioactive amino acid. To do this, cells were grown in medium with radioactive leucine (0.5 &i/ml) for four to five generations to label the cellular protein uniformly. These prelabelled cells were then used for a growth experiment in medium with radioactive leucine at the same specific activity. The results are presented in figure 3. In figure 3A, i t is seen that the amount of protein per cell (measured as the amount of radioactive leucine per cell) was lowest on day 1,then began to increase until i t reached the maximum on day 5. In repeated experiments this maximum was reached three to five days after plating. Thereafter, the average amount of cellular protein decreased to a level approximately 60%that of the maximum as the cells entered stationary phase. Also, i t is shown in figure 3A that as the cells progressed towards stationary phase, the decrease of protein per cell from the maximal point was compatible with the predicted decline of cellular protein determined by the decrease of the proportion of cells that were actively dividing. When the amount of cellular protein was normalized to take account of the predicted division cycle-dependent fluctuations in the amount of cellular protein for this period, i t became approximately constant. (See fig. 3B legend for detailed discussion.) The results are presented in figure 3B. Furthermore, i t was our observation (Engelhardt, unpublished data) that Vero cells in the stationary phase had approximately 70% as much DNA as they did in the exponential phase (measured by the diphenylamine reaction). Thus, the protein per DNA ratio of exponential phase cells (e.g., day 5: fig. 3) was approximately the same as that of stationary
298
GLORIA T-Y. LEE AND DEAN L. ENGELHARDT
-
-100 - 80
-
f2
-60 X
5
-
LL
0
-40
-
5LL # W
n
-20
-
CB *
Z QT
ocn
2
I 5
10 TIME (DAYS)
0 15
Fig. 3 Measurement of cellular protein with 3H-leucine. (A) Cells were plated a t 5 X lo3cells/cm2on 35 mm petri dishes. Saturation density (1.85 X lo6 cells/cm2) was taken as 100%.Increases in cell number (-0-1 were plotted on a logarithmic scale (left-handordinate). The amount of protein per cell ( - A - ) was expressed as the percentage of the highest point on the curve (719.1 CPM/105 cells) (right-hand ordinate). The daily proportionate increment of cellular protein in the cell population was cakulated from the following formula: CPmb-cPma . -1 cpma +0.429(cpmb-cpma) t
where cpm, and c p are~ t he radioactivity per plate of two consecutive determinations and 0.429 (cpmb-cpq,) is the average increase of cellular protein a t the midpoint of this time-interval, assuming the increase in protein per cell occurs exponentially; t is the time in days between these two determinations; 1.138 day-' is the value for day 2-3 and is used a s 100%.Each increment value is compared with this number, and is presented as percentage (-0 -1. (B) The average amount of radioactivity per cell was normalized for the change in cell size due to the variation in the proportion of dividing cells in the population. As cells traverse from early G , into mitosis, the protein content in the cell increases from 1to 2, (Killander and Zetterberg, '65). In the exponential phase of t h e growth cycle, all t he cells are dividing. Since cell growth is exponential (Stanners andTill, '60),the average calculated amount of protein per cell for such a population is 1.429. In stationary phase, the majority of cells are in GI (Engelhardt and Mao, '76). The average amount of protein per cell is 1 or slightly over 1. We use the value 1for t h e sake of simple calculation. Therefore, during the growth cycle, the expected value for cellular protein can be calculated as the proportion of dividing cells (determined by normalizing the proportionate increase in cell numbers to that of exponential phase cells) multiplied by 1.429plus the proportion of nondividing cells multiplied by 1. This number decreases proportionally to the decreased rate of cell proliferation from 1.429 to 1 during the phase of declining rate of cell division. The amount of cellular protein obtained from figure 3A was normalized to this number. The ratio is plotted in this figure.
299
PROTEIN METABOLISM DURING GROWTH OF VERO CELLS
phase cells (e.g., day 15: fig. 3). Therefore, both theoretical derivation and actual measurement support the conclusion that the amount of protein per cell, after being corrected for division cycle-dependent fluctuation, does not vary significantly from the midexponential through the stationary phase of the growth cycle. Since the isotopic method measured specifically the accumulation of cellular protein in the monolayer, i t was possible to derive the daily proportionate increment of cellular protein from the increment of cellular protein during the time-interval between two consecutive measurements and the amount of cellular protein a t the mid-point of this time-interVal. (See fig. 3A legend for details.) Since the rate of protein degradation remained constant, this parameter also represents a measure of the protein synthetic activity of a cell population. The results of this determination are presented in fig. 3A. Again, the maximum was reached between day 2 and day 3, when the cells were in the mid-exponential phase of the growth cycle. This is consistent with our observation of the rate of protein synthesis measured as the rate of radioactive amino acid incorporation.
I
TABLE 2
The increase ofprotein in the monolayer in thepresence of protein synthesis inhibitors
Inhibitor 2 mM histidinol 15 mM 0-methyl threonine None
Amino acid incorporation % of control
Proteinl106 cells y%. of control .’
6.5 12.4 100
258
265 245
I The cells were plated in the presence of inhibitor for 24 hours at 1 x 10‘ cellslcmYin medium deficient in the respective amino acid. (MATERIALS AND METHODS.) The cells were grown in respective medium containing ‘C-leucine for 24 hours. One hundred percent is 6.15 x 10‘ cpml106 cells. 3 The cells plated at zero time had 310 p g protein per loficells. This number is taken as 100%.
during the 24-hour period after plating did not affect the increase in protein in the monolayer. Second, when protein degradation was measured in untreated cells between 5 and 9 hours and between 21 and 25 hours after plating, rates of 0.9% HR-’ and 0.8% HR-’, respectively, were obtained compared to 1.3%HR-’ of the same batch of cells measured 72 hours after plating. This 30% reduction in the rate of protein degradation was not sufficient to account for a %fold increase of cellular protein during this initial 24-hour The attachment of serum protein to period. the cells To test the possibility that the increased When the amount of cellular protein was amount of protein in the monolayer was measured a s the amount of protein in the actually due to the attachment of the serum monolayer using the procedure described by proteins in the growth medium to the cells, Oyama and Eagle (‘561,as much as a 3-fold in- cells were plated a t three different densities crease was frequently observed during the 24- in the presence of 3H-acetylated serum. The hour period after plating. (See, for example, in radioactivity remaining with the monolayer table 2 the data reported for the entry marked a t 20 hours after plating was then deter“none.”) This increase was due neither to de mined. The results are reported in table 3. It novo protein synthesis nor to a cessation of can be seen that a considerable amount of protein degradation as was shown by the following experiments. First, cells were plated in TABLE 3 the presence of histidinol or O-methyl-threoThe attachment of ’ff-labelledserum protein to newly-plated nine (Vaughn and Hanson, ’731, under condicells tions that reduced protein synthesis during pg of cellular protein ’ (no. of p g of serum protein the first 24 hours to less than 13%of the uncells) I plate at 0 hours after remaining I plate at treated control. The amount of protein in the 20 hours after plating plating monolayer was determined. The results are 3.36 (2 X 10‘) 51.5 presented in table 2. There was no difference 16.8 (1 X lo5) 91.7 in the amount of protein present in the mono33.6 (2 X lo5) 117.2 layer between the cells plated in the presence ’ The amount of cellular protein was determined by the Lowry proand in the absence of protein synthesis inhibi- cedure after the cells (in suspension) used for plating had been washed three times with cold PBS. Sixty millimeter plates were used. tors. (Recall, i t has been shown above t h a t 2The concentration of serum protein in growth medium is 9.6 mgl histidinol does not affect the rate of protein ml with a specific activity of 360 cpmlpg. I X lo6cpm of this medium degradation. Similar results were also ob- was used in each plate. The amount of serum protein remaining was derived from the radioactivity remaining on the plate and the specific tained with 0-methyl-threonine [Lee, unpub- activity of the serum protein. Each value is ao average of triplicate lished data].) Thus blocking protein synthesis plates. ~~
~~~
~
~~
300
GLORIA T-Y.LEE AND DEAN L. ENGELHARDT
serum protein was attached to the freshly plated cells. Thus we conclude that when the procedure of Oyama and Eagle is employed to determine the amount of protein in the monolayer, the observed increase in the newly plated culture is primarily due to the attachment of serum proteins to the cells. DISCUSSION
provided a specific determination of the cellular protein. We have shown that the procedure of Oyama and Eagle ('561, that measures the gross amount of protein in the monolayer, yielded a value that may well contain significant amounts of serum protein early in the growth cycle. Thus i t is impossible to achieve a precise determination of cellular protein during this period with this method. This may explain the numerous reports which describe a decline of the amounts of cellular protein during the growth cycle (Baenziger et al., '74; Foster and Pardee, '69; Kimbal et al., '71; Kimball et al., '74; Kruse e t al., '67; Meisler, '73a; Oyama and Eagle, '56; Salzman, '59). Finally, we have devoted considerable effort to analyzing the nature of the signal in the growth medium that leads t o this characteristic decrease in protein synthesis as Vero cells entered the stationary phase. We have found that it is not affected if the concentration of essential amino acids in the growth meduim is varied from 20-200% of the concentration present in DME, nor is it affected by growing the cells in medium buffered with 0.037% (w/ v) bicarbonate (pH range of 6.6-7.0) or in 0.37% (w/v) bicarbonate (pH range of 7.6-7.9). Also it is not affected when the calcium concentration in the medium is varied from 1.8 mM to 5.4 mM (Lee, unpublished data). It is however, directly dependent on the presence of serum in the growth medium (Hassell and Engelhardt, '73; Hassell and Engelhardt, '76).
In this study, we have shown that the rate of protein degradation in Vero cells did not vary with cell growth, while the rate of protein synthesis, measured by the rate of amino acid incorporation, varied by 3- t o 10-fold a t different growth stages (fig. 1).This fluctuation in protein synthetic activity was reflected also in the derivation of daily proportionate increment of cellular protein in the monolayer over a similar time course (fig. 3A). These data are consistent with the model that the change of protein requirements of Vero cells during a growth cycle are met by fluctuations in the overall rate of protein biosynthesis. The degradation rate of a population of proteins with a short half-life, as described by Poole and Wibo, ('731, was not pursued in our study. This class of proteins represents only a small portion of the total cellular protein, and the variation of its stability does not significantly affect the overall levels of cellular protein. The amount of protein per cell remained constant throughout the growth cycle if the changes related to the cell division cycle were taken into account. The low level of cellular ACKNOWLEDGMENTS protein during the early stage of the growth cycle (fig. 3B) was probably due to the trypsin The authors wish to express their appreciaand EDTA treatment used in transferring the tion t o Edward Kisailus for help in preparing cells. Similar loss of cellular protein during 3H-labelled serum and to Hamish Young for trypsin and EDTA treatment has been report- helpful criticism during the preparation of ed by Maize1 et al. ('75). The recovery from the the manuscript. loss of cellular protein represented 3-4%ofthe LITERATURE CITED total protein accumulated throughout the growth cycle. This figure was obtained by cal- Baenziger, N. L., C. H. Jacobi, R. E. Thach 1974 Regulation of protein synthesis during density-dependent growth culating the increment of protein per cell per inhibition of BHK21/13 cells. J. Biol. Chem., 249: 3483day, then dividing the sum of these incre3488. ments by the total protein accumulated in the Bray, G . A. 1960 A simple efficient liquid scintillation for growth cycle. This repair synthesis representcounting aqueous solutions in a liquid scintillation counter. Anal. Biochem., 1: 279-285. ed a significant proportion of protein synthesis during the first few days of the growth Campbell, A. 1957 Synchronization of cell division. Bact. Revs., 21: 263-272. cycle. For example, it was 29%of the total pro- Engelhardt, D. L., and J . Sarnoski, 1975 Variations in the teins synthesized from day 1 t o day 2, 20% cell-free translating apparatus of cultured animal cells as a function of time during cell growth. J. Cell. Physiol., from day 2 to day 3, 15% from day 3 to day 4, 86: 15-30. 11%from day 4 to day 5. Thereafter, the Engelhardt. D. L., and Mao Jen-hao 1977 A serum factor amount of protein per cell did not increase. requirement for the passage of cultured vero cells The measurement of cellular protein of through G,. J. Cell. Physiol., 90: 307-320. monolayer cell cultures by an isotopic method Foster, D. O., and A. 0. Pardee 1969 Transport of amino
PROTEIN METABOLISM DURING GROWTH OF VERO CELLS acids by confluent and nonconfluent 3T3 and polyoma virus-transformed 3T3 cells growing on glass cover slips. J. Biol. Chem., 244; 2675-2681. Hassell, J. A., and D. L. Engelhardt 1973 Translational inhibition in extracts from serum-deprived animal cells. Biochim. Biophys. Acta, 325: 345-353. Hassell, J. A,, and D. L. Engelhardt 1976 The regulation of protein synthesis in animal cells by serum factors. Biochem., 15: 1375-1381. Killander, D., and A. Zetterberg 1965 Quantitative cytochemical studies on interphase growth. I. Determination of DNA, RNA, and mass content of age determined mouse fibroblasts in uitro and of intercellular variation in generation time. Exptl. Cell Res., 38; 272-284. Kimball, R. F., S. W. Pardue and E. H. Y. Chu 1974 Cell cycle and cell protein content of Chinese hamster cells grown in culture with daily renewal of medium. Exptl. Cell. Res., 84; 111-120. Kimball, R. F., S. W. Pardue, E. Y. H. Chu and J. R. Ortiz 1971 Microphotometric and autoradiographic studies on the cell cycle and cell size during growth and decline of Chinese hamster cell cultures. Exptl. Cell Res., 66: 17-32. Kruse, P. F., Jr., E. Miedema and H. C. Carter 1967 Amino acid utilizations and protein synthesis a t various proliferation rates, population densities, and protein contents of perfused animal cell and tissue cultures. Biochem., 6: 949-955. Lowry, 0. H., N. H. Rosehrough, A. L. Farr and R. J. Randall 1951 Protein measurement with t he Folin phenol reagent. J. Biol. Chem., 193: 265-275. M a a h , O., and N. 0. Kjeldgaard 1966 Control of Macromolecular Synthesis. Benjamin, New York. Maizel, A,, C. Nicolini and R. Baserga 1975 Effect of cell trypsinization on nuclear proteins of WI-38 fibroblasts in culture. J. Cell. Physiol., 86: 71-82. Meisler, A. I. 1973a Studies on contact inhibition of
30 1
growth in the mouse fibroblasts 3T3. I. Changes in cell size and composition during unrestricted growth. J. Cell Sci., 12: 847-859. Meisler, A. I. 1973b Studies on contact inhibition of growth in the mouse fibroblasts 3T3. 11. Effects of amino acid deprivation and serum on growth rate. J. Cell Sci., 12: 861-873. Monod, J. 1958 Recherches sur la Croissance des Cultures Bacteriennes. (These de 1942). Herman e t Cie. Paris. Oyama, V. I., and H. Eagle 1956 Measurement of cell growth in tissue culture with a phenol reagent (FolinExp. Biol. Med., 91: 305-307. Ciocalteu). Proc. SOC. Pine, M. J. 1972 Turnover of intracellular proteins. Ann. Rev. Microbiology, 26: 103.126. Poole, B., and M. Wibo 1973 Protein degradation in cultured cells. J. Biol. Chem., 248: 6221-6226. Rechcigl, M. 1971 Enzyme Synthesis and Degradation in Mammalian Systems. University Park Press, Baltimore. Roseman, S., and I. Daffner 1956 Colorimetric method for determination of glucosamine and galactosamine. Anal. Chem., 28: 1743-1746. Salzman, N. P. 1959 Systematic fluctuations in the cellular protein, RNA, and DNA during growth of mammalian cell cultures. Biochim. Biophys. Acta, 31: 158-163. Skehan, P. 1976 On the regulation of cell growth in culture. Exptl. Cell Res., 97: 184-192. Stanners, C. P., and J. E. Till 1960 DNA synthesis in individual L-strain mouse cells. Biochem. Biophys. Acta, 37: 406-419. Todaro, G. J., S. A. Aaronson and E. Rands 1971 Rapid detection of mycoplasma-infected cell cultures. Exptl. Cell Res., 65: 256-257. Vaughn, M. H., and B. S. Hansen 1973 Control of initiation of protein synthesis in human cells. Evidence for a role of uncharged transfer ribonucleic acid. J . Biol. Chem., 248: 7087-7096.
ABSTRACT Protein synthesis and degradation were studied throughout a growth cycle of Vero cells. The rate of protein synthesis, measured a s the rate of amino acid incorporation, reached a maximum a t the mid-exponential phase and declined to 10-30%of the maximum in the stationary phase. The rate of protein degradation, measured a s the release of radioactive amino acids from uniformly labelled cellular proteins, did not vary in the growth cycle. The amount of protein per cell, measured by a n isotopic method, remained constant when normalized to account for the variation in the proportion of actively dividing cells in the cell population during the growth cycle. Cellular protein was determined using this method since i t was found t h a t the chemical determination of the amount of protein in the monolayer was not accurate during the early stage of the growth cycle. This was due to a significant amount of serum protein adsorbed to the cells. In this study we were able to show that, in Vero cells, protein synthetic activity is correlated with the rate of cell division, and variations in the rate of synthesis alone are sufficient to meet the changing requirements for cellular protein in a growth cycle. The overall metabolism of protein in cells has been the subject of considerable interest (Foster and Pardee, '69; Meisler, '73b; Pine '72; Rechcigl, '71; Salzman, '59). Studies in this area have revealed that the amount of protein in a cell is determined by the dynamic relationship between synthetic and degradative processes within the cell and export from the cell. This paper describes an animal cell culture system in which a balanced state of protein metabolism (Campbell, '57; Maabe and Kjeldgaard, '66) was achieved during rapid, exponential growth and in the stationary phase. The approach employed was to permit the cells to grow a t least nine generations before reaching the stationary phase. Such a long growth cycle permitted the cells t o recover from pertubation caused by the initial transfer process, and to approach equilibrium with regard to biosynthetic activity, the major variable, before attaining the phase of decelerating cell division. In this system we were able to analyze the inputs (synthesis and uptake) and the outputs (degradation, cell division, and leakage) of cellular protein and to establish the relationship between these parameters and cell proliferation. J. CELL. PHYSIOL.. 92: 293-302.
MATERIALS AND METHODS
Vero cells were purchased from the American Type Culture collection. Dulbecco's modified Eagle's medium (DME) and leucine, isoleucine, and histidine-deleted DME were from Grand Island Biological Co. Calf serum was from Flow Laboratories. O-methyl-L-threonine and L-histidinol dihydochloride were from Calbiochem. L-leucine ['TI (U) (312 mCi/mmole) was from SchwarzIMann. L-leucine [4,5-3H1(5Ci/mmole) was from New England Nuclear. Plastic tissue culture ware was from Lux Scientific Co.
Cell growth and maintenance Vero cells were grown on tissue culture dishes in DME supplemented with 10% calf serum a t 37°C in 5% COz, and 100%humidity. The cells were detached from the growth surface with trypsin (0.05%w/v) in phosphate buffered s a l i n e (PBS) (150mM NaC1, 4mMKC1, 9.5mM phosphate, pH 7.2) plus 0.5 mM ethylenediaminetetraacetic acid (EDTA) and subcultured every four or five days. The cells were routinely screened for mycoplasma Received Aug. 30, "76. Accepted Jan. 19, '77. )This research was supported by a grant from The National Science Foundation BMS 74-02202 A01.
293
294
GLORIA T-Y.LEE A N D DEAN L. ENGELHARDT
by plating on mycoplasma growth agar and by the method described by Todaro e t al. ('71). Cells in early exponential phase were used for each growth experiment. During the course of the various growth cycles reported in this paper the cells employed were subject to daily medium change. Cell number was determined by the following two methods: (1)For cells at. a density higher than 2 x lo4 cells/cm2, the cell number was obtained by counting the cells in suspension with a hemocytometer. ( 2 ) For cells a t a density lower than 2 X lo4 cells/cm2, the cells were fixed and stained with crystal violet (0.1% (wiv) in 20% ethanol) and counted directly on the petri dish with the aid of a grid. Cellular protein was determined by Lowry's procedure (Lowry et al., '51) as modified by Oyama and Eagle ('56) or by an isotopic technique as described in the RESULTS section. When protein was determined by the procedure of Oyama and Eagle ('561, growth medium was removed, the cell layer was washed three times with PBS, and dissolved in alkaline copper solution. Petri dishes with growth medium only were employed to determine the background amount of protein (-0.3 yg/cm2) that adsorbed non-specifically to the growth surface. This background number was subtracted from all values reported. For cells grown in the presence of amino acid analogues the respective amino acid was deleted from the DME. Note, however, that trace amounts were supplied by the serum (Meisler, '73b).
rial was constant for a t least two hours after the addition of the label.
Specific activity of leucyl-tRNA: To measure the specific activity of leucyltRNA, cells were pulsed with 3H-Leua t 5 yCi/ ml in DME (0.08 mM leucine) plus 10% calf serum for 30 minutes. At the end of the pulse, the cells were washed with cold PBS, scraped off the petri dish, lysed with RSB (10 mM Tris-HC1, pH 7.4, 10 mM NaC1, 1.5 mM MgC1,) supplemented with 0.5%NP-40. Nuclei were removed by centrifugation. The supernatant was adjusted to 0.4 M LiCl, 20 mM Sodium Acetate, pH 5.3, and extracted with phenolchloroform-isoamyl alcohol (50:50: 1). The RNA was precipitated three times with 2.5 volumes of ethanol to remove trace amounts of free leucine. The amount of leucyl-tRNA was then determined as described by Baenziger e t al. ('74). Specifically, the aminoacyl bond was hydrolyzed a t pH 10 for ten minutes a t 37". The amount of leucine released from leucyl-tRNA was determined by a Beckman amino acid analyzer. Leucine was quantitated by measuring chart peak area relative to that of a 10-nmole leucine standard. Radioactivity in the sample applied to the amino acid analyzer was determined in a scintillation spectrophotometer.
Cellular protein degradation
The procedure used was based on that of Poole and Wibo ('73) with the following modifications. Cells growing on a 60-mm petri dish were labelled for 24 hours with 0.5 yCi/ Radioactive labelling experiments All of these experiments were performed 18 ml '%-leucine in DME plus 10% calf serum. hours after the last medium change. To mea- Medium was then removed, and the cells were sure the incorporation of radioactive leucine washed three times with prewarmed noninto trichloroacetic acid (TCA)-precipitable radioactive medium and overlaid with 6 ml of material in vivo, the growth medium was re- this medium. After two hours (a time suffiplaced with prewarmed labelling medium cient to reduce the intracellular TCA-soluble (regular growth medium, unless specified, radioactivity to a minimal level), medium was with either I4C or 3H-leucine). At the end of again removed and another 6 ml of non-radioas the 30-minute pulse, the medium was re- active medium was added. This was defined time T=O. One-half milliliter was removed moved, and the monolayer was washed three every 30 minutes, and the radioactivity in times with ice-cold PBS. The cells were solubilized with 0.025 M Na,CO",, 0.05 M this medium was assayed in Bray's ('60) soluNaOH for 15 minutes a t room temperature. tion. The resultant value was corrected for TCA was added t o a final concentration of 10% the volume change of the medium throughout (w/v), and the precipitate was collected on a the experiment. Total TCA-insoluble radioglass fiber filter (GF/A-Whatman). Radio- activity at T=O was determined in parallel activity was then determined in a scintilla- plates. The TCA-precipitable radioactivity retion spectrophotometer. The rate of incorpo- leased during the course of the experiment ration of leucine into TCA-precipitable mate- was also determined in parallel plates.
295
PROTEIN METABOLISM DURING GROWTH OF VERO CELLS
I00
80
5 2 60 2 2 LL
0 40
E
W
a
20
01
-r i 5
10
1
5 I
I
' ' ' 11 0
I
I
' '
'
15 '
1
'
I;
0
TIME (DAYS) Fig. 1 The rate of amino acid incorporation during the growth cycle. The cells were plated at (A) 1 X 10' cells/cm2; (B) 1 X lo3cells/cm2. The saturation density ranges from 9 X lo5to 1.2 X lo6cells/cm2in different experiments. This is taken as 100%.Increases in cell number (- 0 -1 were plotted on a logarithmic scale (lefthand ordinate). 'C-leucine was used t o measure the in vivo rate of amino acid incorporation (-@ -) (righthand ordinate). In panel A, 1FCi/ml of 'C-leucine in DME (0.8mM leucine) plus 10%calf serum was the labelling medium. 876 CPM/105cells is the 100%determination with a labelling period of 30 minutes. In panel B, 0.5 FCi/rnl of IC-leucine in DME (0.08 mM leucine) plus 10%calf serum was the labelling medium. 4600 CPM/105 cells is the 100%determination with a labelling period of two hours. Each number represents the average of triplicate determinations. The variation is within 15%.
3H-acetylated serum
RESULTS
Protein synthesis in Vero cells during Free amino groups in serum proteins were growth acetylated with 3H-acetic anhydride (AmerI t has been reported by Engelhardt and Sarsham, 500 mCi/mmole) according to the procedure described by Roseman and Doffner ('56). noski ('75) that the relative activity or Free label was removed through extensive di- amount of various components of the protein alysis against PBS, and finally against DME. synthetic machinery in Vero cells fluctuates in the growth cycle. It was not clear from The 3H-acetylatedserum (specific activity 3.6 these data, however, at which stage of the x lo3 cpm/pg protein) was diluted 10-fold growth cycle these changes occurred, since with unlabelled calf serum, before being used. the cells were not grown for time periods long To determine the adsorption of serum protein enough to permit a clear separation of the onto newly plated cells, cells were plated in various growth phases. Accordingly, cells DME with 10%acetylated calf serum. Twenty were grown such that the cell number inhours later the plates were washed three creased 100-fold (6.6 divisions per cell), and times with cold PBS. The material remaining the incorporation of leucine into TCA-precipon the plates was solubilized and counted as itable material was monitored throughout described above in the section on radioactive this period. The results, presented in figure lA, demonstrated that under these growth labelling.
296
GLORIA T-Y. LEE AND DEAN L. ENGELHARDT
I
0
60
I
120
I
180
I
I
240
TIME (MINI Fig. 2 The time course of release of radioactive leucine from prelahelled cells. The procedure of Poole and Wibo, ('731, was followed with the following modifications. The release of radioactive leucine was measured 2-6 hours after the removal of radioactive leucine in exponential phase cells ( - 0 -1; 2-5 hours (- 0-) and 15-18 hours ( - A - ) in stationary phase cells. Each point is the average of three determinations. The variation in individual determinations is no more than 20%.Radioactivity released is expressed as the percentage of TCA-insoluble radioactive material remaining on the plate a t the time when samples were first taken Le., 2 hours after the removal of label). Exponential phase cells were a t a density of 3 X 10' ceIls/cm'. Stationary phase cells were a t a density of 8 X lo5 cells/cm2.
conditions a t least three and, perhaps four phases of the growth cycle could be defined (Monod, '58; Skehan, '76). They were the phase of accelerating rate of cell division (phase 2) on day 1, the exponential phase (phase 3) on days 2-6, the phase of decelerating rate of division (phase 4) on days 7-9, and the stationary phase (phase 5) on days 1011. The lag phase (phase 1)has never been observed. The measurement, of the rate of leucine incorporation into TCA-precipitable material, reported in figure lA, showed a maximum early in the growth cycle. This maximum occurred on day 2 or 3 in repeated experiments. Thereafter, the rate of leucine incorporation declined throughout the rest of the growth cycle, and reached a minimum when cells entered the stationary phase. While these data demonstrated that the decline in leucine incorporation occurred before cells had entered phase 4, they did not indicate whether the maximum occurred in phase 2 or phase 3. For example, i t was possible t h a t the maximum could have occurred as cells were in phase 2 and the observed low rate on day 1 in this particular experiment (fig. 1A) was due to the slow recovery of the cells from
the transfer process. In order to distinguish phase 2 from phase 3 more clearly, a growth cycle of more than nine cell divisions was attempted. In this experiment prolonged phases 2 and 3 were observed. Again the incorporation of leucine into TCA-precipitable material was followed throughout this period. The results are presented in figure 1B. In this growth cycle, phase 2 occurred on days 1 to 7, phase 3 on days 8 to 14, and phases 4 and 5 on days 15 to 20. The incorporation of leucine reached the highest rate on days 10 to 12, the mid-exponential phase. Thus the maximal rate of incorporation occurred during the period of maximal rate of cell division. The amount of leucine incorporation into TCA-precipitable material is a valid measure of the amount of protein synthesized if the specific activity of the leucine entering proteins remains invariant through the growth cycle. To explore this question, the specific activity of leucyl-tRNA in cells in exponential phase and stationary phase was determined and found to be 194 CPMhmole and 160 CPMhmole respectively. The difference in the specific activity of leucyl-tRNA was not sufficient to account for the 3- t o 10-fold de-
PROTEIN METABOLISM DURING GROWTH OF VERO CELLS
297
crease in the amount of leucine incorporated, which was constantly observed in repeated experiments. Thus, we conclude that the protein synthetic activity fluctuated during the cell growth cycle. I t reaches a maximum in the mid-exponential phase and approaches a minimum in the stationary phase.
degradation was observed under these conditions. These results indicated that reutilization of amino acids did not occur, under these conditions, a t a level sufficient to affect our measurement of the rate of cellular protein degradation. Thus, we conclude that the rate of protein degradation in Vero cells does not vary during the growth cycle.
Protein degradation during growth The rate of protein degradation was also determined for cells in different phases of the growth cycle. One measure of protein degradation in cultured cells is the release of amino acids from prelabelled cellular protein into the growth medium (Poole and Wibo, '73). When protein degradation was determined with this procedure, as indicated in figure 2, cells in mid-exponential phase (at the point of maximal protein synthesis) and cells in the stationary phase (at the point of minimal protein synthesis) had similar rates. The observed values were 1.06% HR-' for growing cells and 1.11%HR-' for stationary phase cells. Less than 1%of the released radioactivity was TCA-insoluble. To test for the possibility of reutilization of labelled amino acids, the concentration of leucine in the growth medium was varied from 0.16 mM t o 4.0 mM, and also the rate of degradation was measured in the presence of histidinol or cycloheximide where the rate of protein synthesis had been reduced to less than 3% of the untreated control. The results of these various experiments are presented in table 1. Little difference in the rate of protein
Amount of protein per cell during growth
TABLE 1
The effectofprotein synthesis inhibitors and varying leucine concentration on protein degradation Rate of protein degradation X per hour
Control 0.16 mM leucine 4.0mM leucine 2 mM L-histidinol 30 pgiml cycloheximide
1.36 1.49 1.36 1.37 1.35
'
Amino acid incorporation %,ofcontrol
100
-
2.78 2.23
' The rate of protein degradation was determined a s described in MATERIALS AND METHODS. Cells were at 4 X 10' cells/cm2, and were incubated with respective media 30 minutes prior to the start of degradation measurement. The release of radioactive amino acid was determined over a 4-hour period. 2DME plus 10%calf serum, leucine concentration was 0.8 mM. "repared in DME, lacking L.histidine, with lo'% calf serum. iCells in parallel plates were labeled with "c-leu (0.5pCi/ml with 0.08 mM Leu in DME) for the time period of degradation measurement; 12,110 cpmiplate for control was taken as 100%.
Since the overall level of cellular protein is determined by the synthesis and degradation of protein in the cell, the amount of cellular protein was determined throughout t h e growth cycle to establish the correlation among these parameters. The amount of protein per cell was measured by labelling cellular protein with a radioactive amino acid. To do this, cells were grown in medium with radioactive leucine (0.5 &i/ml) for four to five generations to label the cellular protein uniformly. These prelabelled cells were then used for a growth experiment in medium with radioactive leucine at the same specific activity. The results are presented in figure 3. In figure 3A, i t is seen that the amount of protein per cell (measured as the amount of radioactive leucine per cell) was lowest on day 1,then began to increase until i t reached the maximum on day 5. In repeated experiments this maximum was reached three to five days after plating. Thereafter, the average amount of cellular protein decreased to a level approximately 60%that of the maximum as the cells entered stationary phase. Also, i t is shown in figure 3A that as the cells progressed towards stationary phase, the decrease of protein per cell from the maximal point was compatible with the predicted decline of cellular protein determined by the decrease of the proportion of cells that were actively dividing. When the amount of cellular protein was normalized to take account of the predicted division cycle-dependent fluctuations in the amount of cellular protein for this period, i t became approximately constant. (See fig. 3B legend for detailed discussion.) The results are presented in figure 3B. Furthermore, i t was our observation (Engelhardt, unpublished data) that Vero cells in the stationary phase had approximately 70% as much DNA as they did in the exponential phase (measured by the diphenylamine reaction). Thus, the protein per DNA ratio of exponential phase cells (e.g., day 5: fig. 3) was approximately the same as that of stationary
298
GLORIA T-Y. LEE AND DEAN L. ENGELHARDT
-
-100 - 80
-
f2
-60 X
5
-
LL
0
-40
-
5LL # W
n
-20
-
CB *
Z QT
ocn
2
I 5
10 TIME (DAYS)
0 15
Fig. 3 Measurement of cellular protein with 3H-leucine. (A) Cells were plated a t 5 X lo3cells/cm2on 35 mm petri dishes. Saturation density (1.85 X lo6 cells/cm2) was taken as 100%.Increases in cell number (-0-1 were plotted on a logarithmic scale (left-handordinate). The amount of protein per cell ( - A - ) was expressed as the percentage of the highest point on the curve (719.1 CPM/105 cells) (right-hand ordinate). The daily proportionate increment of cellular protein in the cell population was cakulated from the following formula: CPmb-cPma . -1 cpma +0.429(cpmb-cpma) t
where cpm, and c p are~ t he radioactivity per plate of two consecutive determinations and 0.429 (cpmb-cpq,) is the average increase of cellular protein a t the midpoint of this time-interval, assuming the increase in protein per cell occurs exponentially; t is the time in days between these two determinations; 1.138 day-' is the value for day 2-3 and is used a s 100%.Each increment value is compared with this number, and is presented as percentage (-0 -1. (B) The average amount of radioactivity per cell was normalized for the change in cell size due to the variation in the proportion of dividing cells in the population. As cells traverse from early G , into mitosis, the protein content in the cell increases from 1to 2, (Killander and Zetterberg, '65). In the exponential phase of t h e growth cycle, all t he cells are dividing. Since cell growth is exponential (Stanners andTill, '60),the average calculated amount of protein per cell for such a population is 1.429. In stationary phase, the majority of cells are in GI (Engelhardt and Mao, '76). The average amount of protein per cell is 1 or slightly over 1. We use the value 1for t h e sake of simple calculation. Therefore, during the growth cycle, the expected value for cellular protein can be calculated as the proportion of dividing cells (determined by normalizing the proportionate increase in cell numbers to that of exponential phase cells) multiplied by 1.429plus the proportion of nondividing cells multiplied by 1. This number decreases proportionally to the decreased rate of cell proliferation from 1.429 to 1 during the phase of declining rate of cell division. The amount of cellular protein obtained from figure 3A was normalized to this number. The ratio is plotted in this figure.
299
PROTEIN METABOLISM DURING GROWTH OF VERO CELLS
phase cells (e.g., day 15: fig. 3). Therefore, both theoretical derivation and actual measurement support the conclusion that the amount of protein per cell, after being corrected for division cycle-dependent fluctuation, does not vary significantly from the midexponential through the stationary phase of the growth cycle. Since the isotopic method measured specifically the accumulation of cellular protein in the monolayer, i t was possible to derive the daily proportionate increment of cellular protein from the increment of cellular protein during the time-interval between two consecutive measurements and the amount of cellular protein a t the mid-point of this time-interVal. (See fig. 3A legend for details.) Since the rate of protein degradation remained constant, this parameter also represents a measure of the protein synthetic activity of a cell population. The results of this determination are presented in fig. 3A. Again, the maximum was reached between day 2 and day 3, when the cells were in the mid-exponential phase of the growth cycle. This is consistent with our observation of the rate of protein synthesis measured as the rate of radioactive amino acid incorporation.
I
TABLE 2
The increase ofprotein in the monolayer in thepresence of protein synthesis inhibitors
Inhibitor 2 mM histidinol 15 mM 0-methyl threonine None
Amino acid incorporation % of control
Proteinl106 cells y%. of control .’
6.5 12.4 100
258
265 245
I The cells were plated in the presence of inhibitor for 24 hours at 1 x 10‘ cellslcmYin medium deficient in the respective amino acid. (MATERIALS AND METHODS.) The cells were grown in respective medium containing ‘C-leucine for 24 hours. One hundred percent is 6.15 x 10‘ cpml106 cells. 3 The cells plated at zero time had 310 p g protein per loficells. This number is taken as 100%.
during the 24-hour period after plating did not affect the increase in protein in the monolayer. Second, when protein degradation was measured in untreated cells between 5 and 9 hours and between 21 and 25 hours after plating, rates of 0.9% HR-’ and 0.8% HR-’, respectively, were obtained compared to 1.3%HR-’ of the same batch of cells measured 72 hours after plating. This 30% reduction in the rate of protein degradation was not sufficient to account for a %fold increase of cellular protein during this initial 24-hour The attachment of serum protein to period. the cells To test the possibility that the increased When the amount of cellular protein was amount of protein in the monolayer was measured a s the amount of protein in the actually due to the attachment of the serum monolayer using the procedure described by proteins in the growth medium to the cells, Oyama and Eagle (‘561,as much as a 3-fold in- cells were plated a t three different densities crease was frequently observed during the 24- in the presence of 3H-acetylated serum. The hour period after plating. (See, for example, in radioactivity remaining with the monolayer table 2 the data reported for the entry marked a t 20 hours after plating was then deter“none.”) This increase was due neither to de mined. The results are reported in table 3. It novo protein synthesis nor to a cessation of can be seen that a considerable amount of protein degradation as was shown by the following experiments. First, cells were plated in TABLE 3 the presence of histidinol or O-methyl-threoThe attachment of ’ff-labelledserum protein to newly-plated nine (Vaughn and Hanson, ’731, under condicells tions that reduced protein synthesis during pg of cellular protein ’ (no. of p g of serum protein the first 24 hours to less than 13%of the uncells) I plate at 0 hours after remaining I plate at treated control. The amount of protein in the 20 hours after plating plating monolayer was determined. The results are 3.36 (2 X 10‘) 51.5 presented in table 2. There was no difference 16.8 (1 X lo5) 91.7 in the amount of protein present in the mono33.6 (2 X lo5) 117.2 layer between the cells plated in the presence ’ The amount of cellular protein was determined by the Lowry proand in the absence of protein synthesis inhibi- cedure after the cells (in suspension) used for plating had been washed three times with cold PBS. Sixty millimeter plates were used. tors. (Recall, i t has been shown above t h a t 2The concentration of serum protein in growth medium is 9.6 mgl histidinol does not affect the rate of protein ml with a specific activity of 360 cpmlpg. I X lo6cpm of this medium degradation. Similar results were also ob- was used in each plate. The amount of serum protein remaining was derived from the radioactivity remaining on the plate and the specific tained with 0-methyl-threonine [Lee, unpub- activity of the serum protein. Each value is ao average of triplicate lished data].) Thus blocking protein synthesis plates. ~~
~~~
~
~~
300
GLORIA T-Y.LEE AND DEAN L. ENGELHARDT
serum protein was attached to the freshly plated cells. Thus we conclude that when the procedure of Oyama and Eagle is employed to determine the amount of protein in the monolayer, the observed increase in the newly plated culture is primarily due to the attachment of serum proteins to the cells. DISCUSSION
provided a specific determination of the cellular protein. We have shown that the procedure of Oyama and Eagle ('561, that measures the gross amount of protein in the monolayer, yielded a value that may well contain significant amounts of serum protein early in the growth cycle. Thus i t is impossible to achieve a precise determination of cellular protein during this period with this method. This may explain the numerous reports which describe a decline of the amounts of cellular protein during the growth cycle (Baenziger et al., '74; Foster and Pardee, '69; Kimbal et al., '71; Kimball et al., '74; Kruse e t al., '67; Meisler, '73a; Oyama and Eagle, '56; Salzman, '59). Finally, we have devoted considerable effort to analyzing the nature of the signal in the growth medium that leads t o this characteristic decrease in protein synthesis as Vero cells entered the stationary phase. We have found that it is not affected if the concentration of essential amino acids in the growth meduim is varied from 20-200% of the concentration present in DME, nor is it affected by growing the cells in medium buffered with 0.037% (w/ v) bicarbonate (pH range of 6.6-7.0) or in 0.37% (w/v) bicarbonate (pH range of 7.6-7.9). Also it is not affected when the calcium concentration in the medium is varied from 1.8 mM to 5.4 mM (Lee, unpublished data). It is however, directly dependent on the presence of serum in the growth medium (Hassell and Engelhardt, '73; Hassell and Engelhardt, '76).
In this study, we have shown that the rate of protein degradation in Vero cells did not vary with cell growth, while the rate of protein synthesis, measured by the rate of amino acid incorporation, varied by 3- t o 10-fold a t different growth stages (fig. 1).This fluctuation in protein synthetic activity was reflected also in the derivation of daily proportionate increment of cellular protein in the monolayer over a similar time course (fig. 3A). These data are consistent with the model that the change of protein requirements of Vero cells during a growth cycle are met by fluctuations in the overall rate of protein biosynthesis. The degradation rate of a population of proteins with a short half-life, as described by Poole and Wibo, ('731, was not pursued in our study. This class of proteins represents only a small portion of the total cellular protein, and the variation of its stability does not significantly affect the overall levels of cellular protein. The amount of protein per cell remained constant throughout the growth cycle if the changes related to the cell division cycle were taken into account. The low level of cellular ACKNOWLEDGMENTS protein during the early stage of the growth cycle (fig. 3B) was probably due to the trypsin The authors wish to express their appreciaand EDTA treatment used in transferring the tion t o Edward Kisailus for help in preparing cells. Similar loss of cellular protein during 3H-labelled serum and to Hamish Young for trypsin and EDTA treatment has been report- helpful criticism during the preparation of ed by Maize1 et al. ('75). The recovery from the the manuscript. loss of cellular protein represented 3-4%ofthe LITERATURE CITED total protein accumulated throughout the growth cycle. This figure was obtained by cal- Baenziger, N. L., C. H. Jacobi, R. E. Thach 1974 Regulation of protein synthesis during density-dependent growth culating the increment of protein per cell per inhibition of BHK21/13 cells. J. Biol. Chem., 249: 3483day, then dividing the sum of these incre3488. ments by the total protein accumulated in the Bray, G . A. 1960 A simple efficient liquid scintillation for growth cycle. This repair synthesis representcounting aqueous solutions in a liquid scintillation counter. Anal. Biochem., 1: 279-285. ed a significant proportion of protein synthesis during the first few days of the growth Campbell, A. 1957 Synchronization of cell division. Bact. Revs., 21: 263-272. cycle. For example, it was 29%of the total pro- Engelhardt, D. L., and J . Sarnoski, 1975 Variations in the teins synthesized from day 1 t o day 2, 20% cell-free translating apparatus of cultured animal cells as a function of time during cell growth. J. Cell. Physiol., from day 2 to day 3, 15% from day 3 to day 4, 86: 15-30. 11%from day 4 to day 5. Thereafter, the Engelhardt. D. L., and Mao Jen-hao 1977 A serum factor amount of protein per cell did not increase. requirement for the passage of cultured vero cells The measurement of cellular protein of through G,. J. Cell. Physiol., 90: 307-320. monolayer cell cultures by an isotopic method Foster, D. O., and A. 0. Pardee 1969 Transport of amino
PROTEIN METABOLISM DURING GROWTH OF VERO CELLS acids by confluent and nonconfluent 3T3 and polyoma virus-transformed 3T3 cells growing on glass cover slips. J. Biol. Chem., 244; 2675-2681. Hassell, J. A., and D. L. Engelhardt 1973 Translational inhibition in extracts from serum-deprived animal cells. Biochim. Biophys. Acta, 325: 345-353. Hassell, J. A,, and D. L. Engelhardt 1976 The regulation of protein synthesis in animal cells by serum factors. Biochem., 15: 1375-1381. Killander, D., and A. Zetterberg 1965 Quantitative cytochemical studies on interphase growth. I. Determination of DNA, RNA, and mass content of age determined mouse fibroblasts in uitro and of intercellular variation in generation time. Exptl. Cell Res., 38; 272-284. Kimball, R. F., S. W. Pardue and E. H. Y. Chu 1974 Cell cycle and cell protein content of Chinese hamster cells grown in culture with daily renewal of medium. Exptl. Cell. Res., 84; 111-120. Kimball, R. F., S. W. Pardue, E. Y. H. Chu and J. R. Ortiz 1971 Microphotometric and autoradiographic studies on the cell cycle and cell size during growth and decline of Chinese hamster cell cultures. Exptl. Cell Res., 66: 17-32. Kruse, P. F., Jr., E. Miedema and H. C. Carter 1967 Amino acid utilizations and protein synthesis a t various proliferation rates, population densities, and protein contents of perfused animal cell and tissue cultures. Biochem., 6: 949-955. Lowry, 0. H., N. H. Rosehrough, A. L. Farr and R. J. Randall 1951 Protein measurement with t he Folin phenol reagent. J. Biol. Chem., 193: 265-275. M a a h , O., and N. 0. Kjeldgaard 1966 Control of Macromolecular Synthesis. Benjamin, New York. Maizel, A,, C. Nicolini and R. Baserga 1975 Effect of cell trypsinization on nuclear proteins of WI-38 fibroblasts in culture. J. Cell. Physiol., 86: 71-82. Meisler, A. I. 1973a Studies on contact inhibition of
30 1
growth in the mouse fibroblasts 3T3. I. Changes in cell size and composition during unrestricted growth. J. Cell Sci., 12: 847-859. Meisler, A. I. 1973b Studies on contact inhibition of growth in the mouse fibroblasts 3T3. 11. Effects of amino acid deprivation and serum on growth rate. J. Cell Sci., 12: 861-873. Monod, J. 1958 Recherches sur la Croissance des Cultures Bacteriennes. (These de 1942). Herman e t Cie. Paris. Oyama, V. I., and H. Eagle 1956 Measurement of cell growth in tissue culture with a phenol reagent (FolinExp. Biol. Med., 91: 305-307. Ciocalteu). Proc. SOC. Pine, M. J. 1972 Turnover of intracellular proteins. Ann. Rev. Microbiology, 26: 103.126. Poole, B., and M. Wibo 1973 Protein degradation in cultured cells. J. Biol. Chem., 248: 6221-6226. Rechcigl, M. 1971 Enzyme Synthesis and Degradation in Mammalian Systems. University Park Press, Baltimore. Roseman, S., and I. Daffner 1956 Colorimetric method for determination of glucosamine and galactosamine. Anal. Chem., 28: 1743-1746. Salzman, N. P. 1959 Systematic fluctuations in the cellular protein, RNA, and DNA during growth of mammalian cell cultures. Biochim. Biophys. Acta, 31: 158-163. Skehan, P. 1976 On the regulation of cell growth in culture. Exptl. Cell Res., 97: 184-192. Stanners, C. P., and J. E. Till 1960 DNA synthesis in individual L-strain mouse cells. Biochem. Biophys. Acta, 37: 406-419. Todaro, G. J., S. A. Aaronson and E. Rands 1971 Rapid detection of mycoplasma-infected cell cultures. Exptl. Cell Res., 65: 256-257. Vaughn, M. H., and B. S. Hansen 1973 Control of initiation of protein synthesis in human cells. Evidence for a role of uncharged transfer ribonucleic acid. J . Biol. Chem., 248: 7087-7096.
