Polyamine Metabolism in a Rat Brain Tumor Cell Line: Its Relationship to the Growth Rate OLLE HEBY,' LAURENCE J. MARTON,2 CHARLES B. WILSON HUGO M. MARTINEZ Naffziger Laboratories f o r Neurosurgical Research, Department of Neurological Surgery and Department of Biochemistry and Biophysics, University of Calfornia Medical Center, San Francisco, California 941 43
AND
ABSTRACT To investigate whether the metabolism of the polyamines putrescine, spermidine and spermine is related to cellular growth rate, we have measured the activities of Lamithine decarboxylase and S-adenosyl-L-methionine decarboxylase as well as the levels of the polyamines in rat brain tumor cells at various stages of a 7 d a y in vitro growth period and correlated them with the continuous changes in specific growth rate ([dN[t]/dtJ/N[t]). L-Ornithine decarboxylase and S-adenosyl-L-methionine decarboxylase both exhibited their maximal activities a t the time of most rapid growth. A high positive correlation between the activities of these enzymes and the specific growth rate of the tumor cells during the entire growth period was demonstrated statistically. The pattern of fluctuation of the spermidine content during the culture cycle was similar to those of the enzyme activities and likewise showed a high positive correlation with the specific growth rate of the tumor cells during the entire growth period. The putrescine content exhibited a low positive correlation, whereas the spermine content exhibited a somewhat higher, but negative correlation with the specific growth rate. The high correlation between the specific growth rate of the tumor cells and the synthesis of the polyamines indicates that these events are primarily associated with processes involved in cell replication. Putrescine and spermidine are thought to participate in the regulation of cellular growth rate; a high content may augment, and a low content may restrain, cellular growth rate.
Cells that continuously traverse the cell cycle, and cells that have been stimulated to do so from a quiescent state, exhibit much higher activities of the enzymes in the polyamine biosynthetic pathway and contain much greater amounts of the polyamines putrescine, spermidine and spermine than do noncycling cells (Cohen, '71; Tabor and Tabor, '72; Bachrach, '73; Heby et al., '74). Accordingly, i t has been found that hepatomas usually exhibit higher activities of L-ornithine decarboxylase, the enzyme that catalyzes the formation of putrescine, and contain greater amounts of the polyamines than does normal liver, and that fast-growing hepatomas usually display higher activities of L-ornithine decarboxylase and contain greater amounts of putrescine than do slow-growing hepatomas (Williams-Ashman et al., '72a). However, for some hepatomas it was not possible to relate the polyamine metabolism to J . CELL. PHYSIOL..86. 511-522.
their growth rate. This apparent lack of correlation may be explained by the fact that metabolic events normally vary with the phase of growth, and that the hepatomas may have been analyzed at noncomparable phases. Also, possible differences in cellular origin or diurnal pattern (Echave Llanos and Nash, '70; Nash and Echave Llanos, '71; Hayashi et al., '72) may be important when comparing metabolic patterns of hepatomas with different growth rates. Whether the continuous changes in growth rate, which occur during the growth of an individual tumor, can be related to changes in the metabolism of the polyamines has not been studied in detail. Some information reReceived Jan. 22, ' 7 5 . Accepted Apr. 8, '75. 1 Correspondence: Dr. Olle Heby, Institute of Zoophysiology, University of Lund, Helgonavagen 3, S-223 62 Lund, Sweden. Also Department of Clinical Pathology and Laboratory Medicine.
*
51 1
512
0.HEBY, L. J . MARTON, C. B. WILSON A N D H . M. MARTINEZ
garding the possible relationship between growth rate and polyamine metabolism in individual tumors has been obtained from studies of the Ehrlich ascites carcinoma grown in vivo (Andersson and Heby, '72; Heby and Russell, '73). The L-ornithine decarboxylase activity (Heby and Russell, '73) and the putrescine concentration (Andersson and Heby, '72) were both found to decline as the growth rate of the tumor decreased with increasing tumor mass. To further evaluate a n y positive correlation between growth rate and cellular polyamine content and/or rate of synthesis we have studied the polyamine metabolism of a rat brain tumor cell line grown in vitro. The growth curve, which exhibits an S-shaped pattern with a n initial lag phase, followed by a n exponential phase and finally a stationary phase, thus displaying both a phase of increasing and a phase of decreasing growth rate, was utilized to mathematically derive the specific growth rate. The specific growth rate was then correlated with the activities of the enzymes in the polyamine biosynthetic pathway and with the concentrations of the polyamines at various stages of growth. MATERIALS A N D METHODS
Rat brain tumor cells The tumor cells used in this study were provided by W. H. Sweet, P. T. Kornblith, J..R. Messer and B. 0. Whitman of the Massachusetts General Hospital, Boston, Massachusetts. The brain tumor was induced in CD Fischer rats by weekly intravenous injections of N-methylnitrosourea. Tissue culture and preservation methods have been described by Benda et al. ('71). Upon receipt of frozen cells from the above source, they were suspended in Eagle's basal medium (BME) supplemented with L-glutamine (:398 mg/l), fetal calf serum (10%), and antibiotics (penicillin, 80.5 IU/ ml; streptomycin, 80.5 IU/ml), and grown by the monolayer method for continuous cell culture described by Wilson et al. ('66) and Barker et al. ('72). Experimental protocol A suspension of 1 X 1 0 6 tumor cells in 15 ml of BME with Earle's balanced salt solution supplemented with L-glutamine (292 mg/l), fetal calf serum (lo%), BME vitamin mixture (1 % ), BME essential ami-
no acid mixture (1 % ), and penicillin-streptomycin mixture ( 1 % ) was pipetted into plastic Falcon flasks (75 cm'; 250 ml). The cell cultures were grown in a National COP-incubator (5% COs: 95% air; 37°C). The cells were harvested at 24-hour intervals by trypsinization as follows. The growth medium was decanted and the cell monolayer was washed twice with 2 ml of 0.25% trypsin solution prepared with Caz+- and Mg*+-free Hanks balanced salt solution containing 0.02 % Na2EDTA, both washes being discarded. Only a thin film of trypsin solution was allowed to remain on the cell monolayer after the second washing. The culture vessels were incubated at 37°C until the cell monolayer detached from the plastic surface ( < 10 min). The cells were then transferred into a centrifuge tube with 5-10 ml of Ca"+- and Mg'i-free Hank's balanced salt solution. Cell clumps were dispersed with a capillary pipette. To determine the number of cells per Falcon flask, an aliquot of the dispersed cell suspension was counted in a hemocytometer. The cell suspension was then centrifuged at 900 X g for ten minutes a t 4 ° C and the cell pellet obtained was utilized immediately for enzyme assay or stored at - 20°C for subsequent polyamine analysis. For each experimental series, cells from 5- or 6-day cultures were trypsinized as described above, suspended in growth medium to a concentration of 1 X 106 cells per 15 milliliters, and plated into Falcon flasks. Preparation of tumor cell extracts for enzyme assays For the assay of L-ornithine decarboxylase activity, cellular pellets from single Falcon flasks were homogenized with a Lab-Line Ultratip Labsonic System (LabLine Instruments, Inc., Melrose Park, Ill.) equipped with a microtip, in 200-400 p1 of ice-cold homogenization medium (100 mM glycyl-glycine buffer, pH 7.2, containing 0.1 mM Na2EDTA and 5 mM dithiothreitol). The homogenates were centrifuged in a Sorvall RC2-B centrifuge at 45,000 X g for 90 minutes at 2°C. The resulting supernatant fraction was used for the assay of L-ornithine decarboxylase activity. For the assay of S-adenosyl-Lmethionine decarboxylase activity, cells
POLYAMINE METABOLISM A N D GROWTH RATE
513
from three Falcon flasks were combined, D isomer was assumed not to be decarsonicated and centrifuged as described boxylated by ornithine decarboxylase (Pegg above. The supernatant fraction was used and Williams-Ashman, '68a). for the assay of S-adenosyl-L-methionine Determination ofputrescine-activated decarboxylase activity. S-adenos yl-L-methionine decarboxylase activity Determination of L-ornithine decarboxylase ac tivity The reaction was followed by measureThe reaction was followed by measuring ment of the release of l-LC02 from (-) 4Cin the the release of ' T O 2 from DL-ornithine-1- S-adenosyl-L-methionine-carboxyl-l presence of saturating levels (2.5 mM) of 14C in the presence of saturating levels (2 mM) of L-ornithine mainly as described putrescine mainly as described by Pegg by Janne and Williams-Ashman ('71). The and Williams-Ashman ('68b, '69) and Copreaction mixture contained 100 pmoles of poc et al. ('71). In the presence of puglycylglycine buffer (pH 7.2), 2 pmoles of trescine there is a close correspondence L-ornithine, 0.5 pCi of DL-ornithine-1- between the release of 14C02 from (-)-S14C monohydrochloride (specific activity, adenosyl-L-methionine-carboxyl-14Cand the 7.66 mCi/mmole), 5 pmoles of dithiothrei- incorporation of radioisotope from putrestol, 0.2 pmoles of pyridoxal 5-phosphate cine-1,4-I4C into spermidine (Pegg and and 100 pl of tumor extract in a total Williams-Ashman, '68b). The reaction mixvolume of 1.00 ml. The reactions were ture contained 100 pmoles of glycylglycarried out in Corex centrifuge tubes (18 cine buffer (pH 7.2), 2.5 pmoles of pux 102 mm) equipped with rubber stop- trescine, 5 pmoles of dithiothreitol, 0.2 pers. The Won released during the 60- pCi of (-)-S-adenosyl-L-methionine-carminute incubation at 37°C was trapped boxyl-I4C(specific activity, 7.7 mCi/mmole) in 0.1 ml of 1 M Hyamine hydroxide con- and 100 pl of the tumor cell extract in a tained in a polypropylene well attached total volume of 1.00 ml. The tubes were to the rubber stopper. By the injection of incubated for 30 minutes at 37°C. Incu1 ml of 40% trichloroacetic acid through bations were carried out and radioactivity the rubber stopper the reactions were halt- was measured as described above for the ed and bound "C02was released from the L-ornithine decarboxylase assay. The rereaction mixture. A further incubation for lease of 'TO2 from the substrate was pro20 minutes at 37°C was allowed to ensure portional to the time of incubation over that all 14C02 released during the enzymic the 30 minute initial time period and to reaction was trapped in the Hyamine hy- the amount of soluble tumor extract droxide. The wells and their contents were added. then removed and placed in plastic countPolyamine analysis ing vials containing 2 ml of ethanol and Cellular pellets obtained from single 10 ml of Omnifluor toluene (0.4% Omnifluor in scintillation grade toluene). Radio- Falcon flasks were sonicated in 4% 5-sulactivity was assayed with a Beckman LS fosalicylic acid (approximately 200 pl/106 250 liquid-scintillation spectrometer. The cells). The homogenate was kept at 0°C counting efficiency for I4C was in the range for one hour and then centrifuged at 8000 of 90-95%. Sufficient counts were record- X g in an Eppendorf Micro Centrifuge ed to obtain a S.D. of the count rate of for five minutes. A 50 pl aliquot of the less than -+ 2 % . Corrections were applied supernatant was analyzed, utilizing a Durfor radioactivity of non-enzymic control rum D-500 amino acid analyzer, by a samples (100 p1 of the homogenization modification of the method previously demedium replacing the tumor cell extract) scribed (Marton et al., '74). Durrum (DC-PA cation exchange resin, in all enzyme experiments. Release of "Confrom the substrate in the L-ornithine a sulfonated polystyrene polymer with a decarboxylase assay was proportional to 12% cross linkage and a bead diameter the time of incubation over an initial 60- of 10 k 1 pm, was used instead of the minute period and to the amount of soluble Durrum DC-4A resin. The resin was packed tumor extract added. In calculating the to a height of 8 cm in a stainless steel rate of decarboxylation of L-ornithine the 1.75 mm internal diameter column. Two
514
0. HEBY, L. J. MARTON, C. B. WILSON A N D H . M. MARTINEZ
buffers were used to effect the separation of the polyamines. Buffer 1: A sodium citrate (0.20 M)-sodium chloride (0.35 M) mixture was prepared by dissolving 58.83 g of sodium citrate X 2 H20, 20.45 g of sodium chloride, 5.00 ml of thiodiglycol, and 1.00 ml of liquefied phenol in distilled water to a final volume of one liter and by adding concentrated HC1 until pH 5.09 +- 0.02 was obtained. The total Na+ concentration of this buffer was 0.95 M . Buffer 2: A sodium citrate (0.35 M)-sodium chloride (2.00 M) mixture was prepared by dissolving 102.90 g of sodium citrate X 2 HzO, 116.88 g of sodium chloride, 5.00 ml of thiodiglycol, and 1.00 ml of liquefied phenol in distilled water to a final volume of one liter and by adding concentrated HC1 until pH 4.68 +- 0.02 was obtained. The total Na+ concentration of this buffer was 3.05 M . The buffers were filtered through Millipore filters (47 mm diameter, 0.45 Fm pore size). Elution and separation of the polyamines in the tumor cell extracts was accomplished in 28 minutes as follows: Buffer 1 was directed to the column for five minutes to elute the amino acids and then Buffer 2 was used for 23 minutes to elute putrescine, spermidine and spermine. The buffer flow rate was 18.5 ml/hr and the flow rate of the ninhydrin reagent solution was 9.5 ml/hr. Column temperature was held isothermally at 66°C. Column pressure generated during analysis was approximately 1600 pounds per inch' gauge. Full scale deflection on the recorder was set at 0.1 O.D. units at 590 nm. Following each analysis the column was regenerated with a solution of 0.2 M NaOH containing 0.67 mM Na, EDTA and was then equilibrated with Buffer 1.
chloride, (-)-S-adenosyl-L-methionine-carboxyl-1W-iodide.monohydrate and Hy amine hydroxide were purchased from New England Nuclear Corp., Boston, Massachusetts. The preparations of DL-ornithine-1'Cmonohydrochloride contained variable amounts of 14C02, which increased the blank values unless removed. Therefore, prior to its use in the L-ornithine decarboxylase assays, the labeled amino acid was dissolved in 0.1 M HC1 and then lyophilized. The residue was dissolved in 0.01 M HC1 and stored at -20°C until used. After this treatment the nonenzymic control incubations (i.e., the complete incubation mixture minus enzyme) gave 75100 cpm over the 60 minute incubation period with the usual specific radioactivity of added L-ornithine (0.125 pCi per ymole of L-ornithine on the assumption that the label was equally distributed between the D- and L-isomer in the racemic mixture). Dithiothreitol, pyridoxal 5-phosphate, Lornithine and the hydrochlorides of putrescine, spermidine and spermine were purchased from Calbiochem, La Jolla, California. High-pressure liquid chromatographic analysis indicated that the specimens of putrescine, spermidine and spermine were uncontaminated by each other. Thiodiglycol and ninhydrin reagent solution were purchased from Pierce Chemical Co., Rockford, Illinois. Liquefied phenol was purchased from Matheson, Coleman and Bell, Norwood, Ohio. RESULTS
Figure 1 shows the growth data for the rat brain tumor cell line when it is grown in vitro. The growth curve exhibits an Sshaped pattern, similar to that obtained for other cell lines grown in vitro, with an initial lag phase, followed by an exponential increase in cell number with a douChemicals bling time of approximately 24 hours, and BME with Earle's balanced salt solution, finally a stationary phase. When the cells 200 mM L-glutamine solution, fetal calf are seeded at a density of 1.3 X lo4 cells/ serum, BME vitamin mixture ( X loo), cm2, they double about three to four times BME essential amino acid mixture ( X loo), and reach a stationary phase after about and penicillin-streptomycin mixture (5000 five days. The cell density remains approxILT each/ml) were obtained from Microbio- imately constant at 1.5 X 105 cells/cm* logical Associates, Bethesda, Maryland. thereafter. The decreasing growth rate that Hank's balanced salt solution, free of Ca2+ is apparent after day 3 may be due to a and Mg2+, was purchased from Grand decreasing growth fraction, a lengthenIsland Biological Company, Santa Clara, ing of the cell cycle, an increasing cell California. DL-Ornithine-l-'4C-monohydro-loss (due to cell death or exfoliation) or a
POLYAMINE METABOLISM AND GROWTH RATE
combination of these factors. Even though there is no apparent increase in cell number from day 5 through 7, the rat brain tumor cells do not stop dividing completely. Rather, a slow proliferation is balanced by a loss of cells into the culture medium
v
3t
/
2L/l 1
I
I
I
,
I
TIME AFTER PLATING (days)
1
Fig. 1 Number of cells per Falcon flask at various times after plating 1.0 X 106 rat brain tumor cells. Symbols represent means ( f S.E.M.) of values obtained in seven independent experiments.
515
leading to a constant cell number (Heby, unpublished results). Figure 2 shows the growth curve and a continuous representation of the growth rate of the tumor cells. This growth curve was obtained by a least squares fit of the growth data (fig. 1) with a fifth degree polynomial, P(t). The derivative curve of this polynomial was taken as a representation of the specific growth rate ( [dN [ t ] / dtj /N it] ). N [tj is the number of cells in the population at time t. The maximal growth rate, i.e., the inflection point of the growth curve and the peak of the derivative curve, occurs approximately two days after plating. Figures 3A and 4 A show the changes in activity of the first two enzymes in the polyamine biosynthetic pathway, L-ornithine decarboxylase and S-adenosyl-L-methionine decarboxylase. Figure 5A shows the changes in cellular polyamine content during the 7-day tumor growth period studied. The activities of L-ornithine decarboxylase (fig. 3A) and S-adenosyl-L-methionine decarboxylase (fig. 4A), as well as the cellular content of spermidine (fig. 5A) all exhibited similar patterns, with maxima during exponential growth and decreasing
TIME AFTER PLATING (days)
Fig. 2 Time course of population growth and continuous representation of the growth rate throughout the growth period of the rat brain tumor cells. Growth data is represented by 0 as log of cell number. The values plotted are the means of seven independent experiments (same as i n fig. 1). The growth curve is a fifth degree polynomial, P(t), fitted to the growth data, and the time derivative, dP/dt, of this polynomial represents the specific growth rate, [dN(t)/ dt]/N(t). of the tumor cells. N [t] is the number of cells in the population at time t. Both curves are computer drawn and the growth curve weighted to include the experimental error.
516
0.HEBY, L. J. MARTON, C. B. WILSON AND H. M. MARTINEZ
8 CORRELATION COEFFICIENT: 0.922
7-
65-
4-
3-
21-
i
OI
Q!
i
i
i
i
(day-')
TIME AFTER PLATING (days)
Fig. 3 Relationship between the L-ornithine decarboxylase activity and the specific growth rate of the rat brain tumor cells. (A) Enzyme activity (mean k S.E.M., n = %9) a t various times of growth. The growth curve is superimposed (-----). (B) Specific growth rate linearly fitted to the enzyme activity. Correlation coefficient = 0.922. The curve is computer drawn.
A ~
T
-
B
7-
CORRELATION COEfflClENl: 0.958
654-
3210
I
1
2
I
m -
3
4
5
(SPECIFIC GROWTH RATEtl)
6 X
35
(day-')
TIME AFTER PLATING (days)
Fig. 4 Relationship between the S-adenosyl-L-methionine decarboxylase activity and the specific growth rate of the rat brain tumor cells. (A) Enzyme activity (mean S.E.M.. n = 3) at various times of growth. Cells from three Falcon flasks were pooled for each determination. The growth curve is superimposed (- - - - -). (B) Specific growth rate linearly fitted to the enzyme activity. Correlation coefficient = 0.958. The curve is computer drawn.
*
517
POLYAMINE METABOLISM AND GROWTH RATE
- 40 - 30 7
- 20 Q
I
2 -10 -9
CL
1:
$
-5 -4
5 Lu u
-8
x
2
3-
-3 2CORRELATION COEFFICIENT: 0.293
-2
1L
2I
I
4I
I
6I
0
I
1-
I
8I (SPECIFIC GROWTH R A T E t I ) x 3.5
TIME AFTER PLATING (days)
(day-’)
C
l9 o/
CORRELATION COEFFICIENT: -0.475
I
- 4
1
2
3
4
5
6
(SPECIFIC GROWTH RATE tl) x 3 5 (day-’)
Fig. 5 Relationship between the polyamine content and the specific growth rate of the rat brain tumor cells. ( A ) Polyamine content (mean t S.E.M., n = 7) at various times of growth. 0 = Putrescine. 0 = spermidine, 0 = spermine. The growth curve is superimposed. Specific growth rate linearly fitted to the putrescine content (B), the spermidine content ( C ) , and the spermine content (D). Correlation coefficients = 0.293,0.951, and - 0.475, respectively. The curves are computer drawn.
518
0 . HEBY, L. J. MARTON, C. B. WILSON AND H . M. MARTINEZ
activities and content as the growth curve approached the plateau and the rate of cell division decreased. The cellular putrescine content (fig. 5A) showed a different pattern with a peak just prior to the stationary phase, and the spermine content (fig. 5A) showed an inverse pattern with a slightly decreased level during exponential growth. The patterns of changes obtained for the activities of L-ornithine decarboxylase and S-adenosyl-L-methionine decarboxylase and for the spermidine content, resembled that of the specific growth rate (fig. 2). To further evaluate the correlation between growth rate and polyamine content and the rate of polyamine synthesis, the specific growth rate ( [dN [ t ]/dt] /N[ t ]) was linearly fitted to these parameters. Figures 3B, 4B and 5B-D show the linear relationships of the enzyme activities and the polyamine contents versus the specific growth rate. The correlation coefficients for the relationships to the specific growth rate were 0.922, 0.958, 0.293, 0.951 and -0.475 for L-ornithine decarboxylase, S-adenosyl-L-methionine decarboxylase, putrescine, spermidine and spermine, respectively. Thus, the amounts of putrescine and spermidine synthesized, as measured by the enzyme activities, are directly related to the growth rate of the tumor cells. The fact that there was a negative relations hip between the spermine content and the specific growth rate means that as the rate of cell replication increases the spermine content decreases. DISCUSSION
The polyamine biosynthetic pathway involves four sequential enzyme reactions catalyzed by L-ornithine decarboxylase (Pegg and Williams-Ashman, '68a), S-adenosyl-L-methionine decarboxylase (Pegg and Williams-Ashman, '68b, '69), spermidine synthase (Janne et al., '71; Raina and Hannonen, '71) and spermine synthase (Hannonen et al., '72). L-Ornithine decarboxylase plays a key role in the biosynthesis of the polyamines since decarboxylation of L-ornithine is the only pathway for putrescine formation in mammals (Williams-Ashman et al., '72b). It may regulate the disposition of L-ornithine visa-,is other metabolic pathways such as the formation of citrulline and glutamate (Williams-Ashman et al., '69; Weber et al.,
'72; Tomino et al., '74). S-Adenosyl-Lmethionine decarboxylase catalyzes the formation of 5'-deoxy-5'-S-(3-methylthiopropylamine) sulfonium adenosine whose propylamine moiety is then conjugated with a putrescine molecule, through the catalysis of spermidine synthase, yielding spermidine. In a similar manner a propylamine moiety is conjugated with a spermidine molecule, through the catalysis of spermine synthase, yielding spermine. In addition to serving a s a substrate in the spermidine synthetic reaction, putrescine stimulates the activity of S-adenosyl-L-methionine decarboxylase, thereby initiating the propylamine transfer reaction (Williams-Ashman et al., '72b). The intracellular levels of the polyamines are determined in large measure by their rates of synthesis. The high cellular putrescine content in the rat brain tumor cells may be explained by the observation that the activity of L-ornithine carboxylase is a t least 100-fold higher than that of Sadenosyl-L-methionine decarboxylase during the entire growth period. In most mammalian tissues, both normal and neoplastic, the putrescine concentration is quite low (Williams-Ashman et al., '72a,b; Heby et al., '73; Marton et al., '74) and constitutes only a fraction of the total polyamine content, but in the rat brain tumor cell line it exhibits a high level which amounts to 6@-90% of the spermidine content, and exceeds the spermine content by 2 0 4 0 % during the in vitro growth period. The high putrescine level in the brain tumor cells is not due to contamination with other substances; and normal mammalian tissues, when analyzed with our procedure, exhibit putrescine levels which are consistent with the low levels reported by others (Marton et al.,'74). The fact that the cellular content of spermine is always lower than that of putrescine and spermidine may be due to inhibition of spermine synthesis by the high cellular putrescine level, inasmuch as putrescine is a potent competitive inhibitor of spermine synthase (Hannonen et al., '72). Even though spermine is a competitive inhibitor of L-ornithine decarboxylase (Pegg and WilliamsAshman, '68a), because of its low cellular content and since i t is a rather weak inhibitor, it is not likely to markedly decrease the putrescine synthesis. Elevated activities of the enzymes in the
POLYAMINE METABOLISM AND GROWTH RATE
polyamine biosynthetic pathway and elevated levels of the polyamines have been observed in many models of stimulated growth (Cohen, '71; Tabor and Tabor, '72; Bachrach, '73; Heby et al., '74), as well as in tumor cell populations when compared to their normal counterparts (Williams-Ashman et al., '72a; Heby et al., '73; Bachrach et al., '74). An elevated putrescine concentration seems to be a general characteristic of cells not only with a high proliferation rate, but of cells that have undergone neoplastic transformation, e.g., virus-transformed cells have been found to have a higher putrescine content than their normal non-infected counterparts despite the fact that they have similar growth rates (Bachrach et al., '74). It was also noted that the intracellular putrescine level increased during the virusinduced transformation and that it remained elevated in the transformed cells even upon subculturing. The results of several recent investigations indicate the probable existence of a positive correlation between tumor cell growth rate and the activities of the enzymes in the polyamine biosynthetic pathway, as well as the endogenous concentrations of the polyamines (Anderson and Heby, '72; Williams-Ashman et al., '72a; Heby and Russell, '73). The present study, however, is the first demonstration of any statistically significant correlations. The activities of L-ornithine decarboxylase and S-adenosyl-L-methionine decarboxylase, as well as the cellular content of spermidine, showed high linear correlations with the specific growth rate of the rat brain tumor cells when grown in culture. Putrescine and spermine, however, did not show significant linear correlations with the specific growth rate. In fact, spermine showed a negative correlation coefficient, implying that the average cellular spermine content decreases rather than increases with increasing specific growth rate. In accordance with this observation, i t has been found that the concentration of spermine decreases in the liver after partial hepatectomy, i.e., as the cells are stimulated to traverse the cell cycle from a quiescent state (Janne, '67; Heby and Lewan, '71). The decrease in spermine concentration noted in regenerating liver, which is a much more pronounced decrease than the one observed in the present study, may be
519
at least partially attributable to tissue hypertrophy. In the present study the decrease in cellular spermine content cannot be due to hypertrophy because spermine was determined on a per cell basis. In the rat brain tumor cell line the decrease in spermine content may be attributed to conversion of spermine into spermidine. In fact Siimes ('67) showed that radioactive spermine was converted into spermidine with a maximal velocity at the time of liver regeneration when the decrease in the endogenous concentration of spermine occurred. In addition to the possibility that the decrease in cellular spermine content could be due to conversion of spermine into spermidine, spermine synthesis may be inhibited by the accumulation of endogenous putrescine which occurs during the same period of growth. As previously mentioned, putrescine has been found to be a potent competitive inhibitor 0s spermine synthase (Hannonen et al., '72). The observation that the activity of Lornithine decarboxylase in hepatomas increases with increasing growth rate (Williams-Ashman et al., '72a) whereas the activities of L-ornithine carbamyltransferase, i.e., citrulline synthesis (Weber et al., '72) and L-ornithine transaminase, i.e., glutamate synthesis (Tomino et al., '74) decrease, demonstrates that L-ornithine is preferentially channeled into polyamine synthesis with increasing growth rate. Weber et al. (72) have suggested that the increased utilization of L-ornithine for polyamine synthesis may confer a selective advantage to the hepatoma cells and that it might contribute to the increased growth rate. The fact that the peak values of the L-ornithine decarboxylase and S-adenosylL-methionine decarboxylase activities and the cellular spermidine content coincide with the inflection point of the growth curve, i.e., the time of maximum growth rate of the tumor cells, and the fact that these parameters show a high positive correlation with the specific growth rate of the tumor cells throughout their in vitro growth period, implies that the polyamines may be essential to the process of cell replication. However, these results do not necessarily indicate that the major function of the polyamines relates to cell division per se. The parasynchronous culture of the rat brain tumor cells does not allow
520
0 HEBY, L. J . MARTON, C. B. WILSON AND H. M. MARTINEZ
us to discern the definitive temporal relationships between polyamine synthesis and cell cycle parameters. In fact, though results obtained from growth curves have indicated a positive correlation between the rate of cell division and the rate of incorporation of DNA precursors (Todo et al., '71; Wheeler and Alexander, '72; Zinninger and Little, '73; Rosenblatt and Erbe, '73; Macieira-Coelho, '73), it is known that DNA synthesis precedes cell division by many hours. Studies relating cell cycle kinetics to growth data have shown that the average cell cycle time is shortest during exponential growth, when the relative time spent in S, G2 and M is maximal and in GI minimal (Todo et al., '71; Zinninger and Little, '73). On this basis we interpret our results in terms of an obligatory involvement of the polyamines in events occurring while cells are proceeding through the S, G2, and M phases of the cell cycle. The decrease in the extent of polyamine synthesis observed with decreasing growth rate is probably a reflection of the increased fraction of the overall cell cycle time spent in the GI phase, and of the decreased number of cells actually involved in proliferation. The fact that the fraction of the cell cycle spent in GI increases in the stationary phase while the rate of polyamine synthesis decreases, argues against polyamine synthesis and accumulation occurring during the G I phase of the cell cycle. Further support for this view is given by our recent observations that WI-38 human diploid fibroblasts stimulated to proliferate from a quiescent state (Heby et al., '74), and synchronous cultures of Chinese hamster ovary cells, obtained by mitotic selection of cells in exponential growth (Heby et al., '75, submitted), initiate their synthesis of polyamines in late GI or early S and continue to accumulate polyamines until they divide. Therefore, we believe that the peak of putrescine synthesis which has been observed in early GI in many systems of stimulated growth may be unrelated to the cell's preparation for DNA synthesis and division. It may be part of a hypertrophic response due to release of hormones, or the exposure to some growthstiniulatory factors. ACKNOWLEDGMENTS
We thank Msi. Kathy D. Knebel for her
technical assistance. This work was supported by NIH Center Grant CA-13525 and gifts from the Phi Beta Psi Sorority, the Joe Gheen Medical Foundation, and the Association for Brain Tumor Research. LITERATURE CITED Andersson, G., and 0. Heby 1972 Polyamine and nucleic acid concentrations in Ehrlich ascites carcinoma cells and liver of tumor-bearing mice a t various stages of tumor growth. J. Nat. Cancer Inst., 48: 165-172. Bachrach, U. 1973 Function of naturally occurring polyamines. Academic Press, New York. Bachrach, U., S. Don and H. Wiener 1974 Polyamines in normal a n d in virus-transformed chick embryo fibroblasts. Cancer Res., 34: 1577-1580. Barker, M.. T. Hoshino and C. B. Wilson 1972 Tissue culture of h u m a n brain tumors. I n : The Experimental Biology of Brain Tumors. W. N. Kirsch, E. G. Paoletti and P. Paoletti, eds. Charles C Thomas, Publisher, Springfield, Illinois, pp. 57-84. Benda, P., K. Someda, J. Messer and W. H. Sweet 1971 Morphological and immunochemical studies of rat glial tumors and clonal strains propagated i n culture. J. Neurosurg., 34: 3 1 0 3 2 3 . Cohen, S. S. 1971 Introduction to the Polyamines. Prentice-Hall, Inc., Englewood Cliffs, New Jersey. Coppoc. G . L.. P. Kallio and H . G . WilliamsAshman 1971 Characteristics of S-adenosylL-methionine decarboxylase from various organisms. Int. J, Biochem., 2: 6 7 3 4 8 1 . Echave Llanos. J. M.. and R. E. Nash 1970 Mitotic circadian rhythm in a fast-growing a n d a slow-growing hepatoma: Mitotic rhythm in hepatomas. J. Nat. Cancer Inst.. 44: 581-585. Hannonen, P., J . J a n n e and A. Raina 1972 Partial purification and characterization of spermine synthase from rat brain. Biochim. Biophys. Acta, 289: 2 2 5 2 3 1 . Hayashi, S., Y . Aramaki and T. Noguchi 1972 Diurnal change in ornithine decarboxylase activity of rat liver. Biochem. Biophys. Res. Comm u n . , 46: 7 9 5 4 0 0 . Heby. 0.. J . W. Gray. P. Lindl. L. J. Marton and C. B. Wilson (1975. submitted) Changes in polyamine content during the cell cycle of Chinese hamster ovary cells. Heby, O., a n d L. Lewan 1971 Putrescine and polyamines in relation to nucleic acids in mouse liver after partial hepatectomy. Virchows Arch. Abt. B. Zellpath., 8: 58-66. Heby, O., L. J. Marton, L. Zardi, D. H. Russell and R. Baserga 1975 Accumulation of polyamines after stimulation of cellular proliferation i n h u m a n diploid fibroblasts. I n : The Cell Cycle i n Malignancy and Immunity. J. C. Hampton, ed. NTIS. Springfield. Virginia, pp. 5 0 4 6 . Heby, O., a n d D. H. Russell 1973 Changes i n polyamine metabolism in tumor cells and host tissues during tumor growth a n d after treatment with various anticancer agents. In: Polyamines in Normal a n d Neoplastic Growth. D. H. Russell, ed. Raven Press, New York, pp. 221237. Hebv. 0.. G. P. Sarna. L. .T. Marton. M. Omine. S - Perry a n d D H Russ"el1 1973 Polyamine
POLYAMINE METABOLISM AND GROWTH RATE content of AKR leukemic cells i n relation to the cell cycle. Cancer Res., 33: 295S2964. Janne, J. 1967 Studies o n the biosynthetic pathway of polyamines i n rat liver. Acta Physiol. Scand., Suppl., 300: 1-71. Janne, J., A. Schenone and H. G. Williams-Ashm a n 1971 Separation of two proteins required for synthesis of spermidine from S-adenosylL-methionine a n d putrescine in r a t prostate. Biochem. Biophys. Res. Commun., 42: 758-764. Janne, J., a n d H. G. Williams-Ashman 1971 On the purification of L-ornithine decarboxylase from r a t prostate and effects of thiol compounds o n the enzyme. J . Biol. Chem., 246: 1725-1 732. Macieira-Coelho, A . 1973 Cell cycle analysis. A. Mammalian cells. I n : Tissue Culture. Methods and Applications. P. F. Kruse, Jr. a n d M. K. Patterson, Jr., eds. Academic Press, New York, pp. 4 1 2 4 2 2 . Marton, L. J., 0. Heby, C. B. Wilson a n d P. L. Y. Lee 1974 An automated micromethod for the quantitative analysis of di- a n d polyamines utilizing a sensitive high pressure liquid chromatographic procedure. FEBS Lett., 41 : 99-103. Nash, R. E., and J. M. Echave-Llanos 1971 Twenty-four-hour variations in DNA synthesis of a fast-growing a n d a slow-growing hepatoma: DNA synthesis rhythm i n hepatoma. J. Nat. Cancer Inst., 47: 1007-1012. Pegg, A. E., a n d H. G. Williams-Ashman 1968a Biosynthesis of putrescine i n the prostate gland of the rat. Biochem. J., 108: 533-539. 1968b Stimulation of the decarboxylation of S-adenosylmethionine by putrescine i n mammalian tissues. Biochem. Biophys. Res. Commun., 30: 7 6 8 2 . 1969 On the role of S-adenosyl-L-methionine in the biosynthesis of spermidine by rat prostate. J. Biol. Chem., 244: 682-693. Raina, A,, and P. Hannonen 1971 Separation of enzyme activities catalysing spermidine a n d spermine synthesis i n rat brain. FEBS Lett., 16: 1-4. Rosenblatt, D. S . , a n d R. W. Erbe 1973 Reciprocal changes in the levels of functionally related folate enzymes during the culture cycle i n h u m a n fibroblasts. Biochem. Biophys. Res. Comm u n . , 54: 1627-1633.
521
Siimes. M . 1967 Studies on the metabolism of 1.4-"C-spermidine and 1.4-IIC-spermine in the rat. Acta Physiol. Scand.. Suppl.. 298. 1-66. Tabor, H., and C. W. Tabor 1972 Biosynthesis and metabolism of 1,4-diaminobutane, spermidine, spermine, and related amines. Adv. Enzymol., 36: 2 0 S 2 6 8 . Todo, A , , A. Strife, J . Fried and B. D. Clarkson 1971 Proliferative kinetics of h u m a n hematopoietic cells during different growth phases i n vitro. Cancer Res., 31: 133G1340. Tomino, I., N. Katunuma, H . P. Morris and G. Weber 1974 Imbalance in ornithine metabolism in hepatomas of different growth rates a s expressed i n behavior of L-orthinine:2-oxoacid aminotransferase (ornithine transaminase EC 2.6.1.13).Cancer Res., 34: 6 2 7 4 3 6 . Weber, G., S . F. Queener a n d H. P. Morris 1972 Imbalance i n ornithine metabolism in hepatomas of different growth rates as expressed in behavior of L-ornithine carbamyl transferase activity. Cancer Res., 32: 193S1940. Wheeler, G. P., and J. A. Alexander 1972 Metabolic activities of murine carcinoma 755 at various stages of growth. Cancer Res., 32: 17611768. Willaims-Ashman. H. G., G. L. Coppoc and G . Weber 1972a Imbalance in ornithine metabolism in hepatomas of different growth rates as expressed in formation of putrescine, spermidine, and spermine. Cancer Res., 32. 19241932. Williams-Ashman. H. G . . J . Janne. G. L. Coppoc. M . E . Geroch and A. Schenone 1972b N e w aspects of polyamine biosynthesis in eukaryotic organisms. Advan. Enzyme Regul.. 1 0 : 225245. Williams-Ashman, H. G., A. E. Pegg a n d D. H. Lockwood 1969 Mechanisms and regulation of polyamine and putrescine biosynthesis in male genital glands a n d other tissues of mammals. Advan. Enzyme Regul., 7: 291-323. Wilson, C. B., M. Barker and D. E. Slagel 1966 Tumors of the central nelvous system in monolayer tissue. Arch. Neurol., 15: 275-282. Zinninger, G . F., a n d J. B. Little 1973 Proliferation kinetics of density-inhibited cultures of h u m a n cells, a complex in vitro cell system. Cancer Res., 33: 2343-2348.
AND
ABSTRACT To investigate whether the metabolism of the polyamines putrescine, spermidine and spermine is related to cellular growth rate, we have measured the activities of Lamithine decarboxylase and S-adenosyl-L-methionine decarboxylase as well as the levels of the polyamines in rat brain tumor cells at various stages of a 7 d a y in vitro growth period and correlated them with the continuous changes in specific growth rate ([dN[t]/dtJ/N[t]). L-Ornithine decarboxylase and S-adenosyl-L-methionine decarboxylase both exhibited their maximal activities a t the time of most rapid growth. A high positive correlation between the activities of these enzymes and the specific growth rate of the tumor cells during the entire growth period was demonstrated statistically. The pattern of fluctuation of the spermidine content during the culture cycle was similar to those of the enzyme activities and likewise showed a high positive correlation with the specific growth rate of the tumor cells during the entire growth period. The putrescine content exhibited a low positive correlation, whereas the spermine content exhibited a somewhat higher, but negative correlation with the specific growth rate. The high correlation between the specific growth rate of the tumor cells and the synthesis of the polyamines indicates that these events are primarily associated with processes involved in cell replication. Putrescine and spermidine are thought to participate in the regulation of cellular growth rate; a high content may augment, and a low content may restrain, cellular growth rate.
Cells that continuously traverse the cell cycle, and cells that have been stimulated to do so from a quiescent state, exhibit much higher activities of the enzymes in the polyamine biosynthetic pathway and contain much greater amounts of the polyamines putrescine, spermidine and spermine than do noncycling cells (Cohen, '71; Tabor and Tabor, '72; Bachrach, '73; Heby et al., '74). Accordingly, i t has been found that hepatomas usually exhibit higher activities of L-ornithine decarboxylase, the enzyme that catalyzes the formation of putrescine, and contain greater amounts of the polyamines than does normal liver, and that fast-growing hepatomas usually display higher activities of L-ornithine decarboxylase and contain greater amounts of putrescine than do slow-growing hepatomas (Williams-Ashman et al., '72a). However, for some hepatomas it was not possible to relate the polyamine metabolism to J . CELL. PHYSIOL..86. 511-522.
their growth rate. This apparent lack of correlation may be explained by the fact that metabolic events normally vary with the phase of growth, and that the hepatomas may have been analyzed at noncomparable phases. Also, possible differences in cellular origin or diurnal pattern (Echave Llanos and Nash, '70; Nash and Echave Llanos, '71; Hayashi et al., '72) may be important when comparing metabolic patterns of hepatomas with different growth rates. Whether the continuous changes in growth rate, which occur during the growth of an individual tumor, can be related to changes in the metabolism of the polyamines has not been studied in detail. Some information reReceived Jan. 22, ' 7 5 . Accepted Apr. 8, '75. 1 Correspondence: Dr. Olle Heby, Institute of Zoophysiology, University of Lund, Helgonavagen 3, S-223 62 Lund, Sweden. Also Department of Clinical Pathology and Laboratory Medicine.
*
51 1
512
0.HEBY, L. J . MARTON, C. B. WILSON A N D H . M. MARTINEZ
garding the possible relationship between growth rate and polyamine metabolism in individual tumors has been obtained from studies of the Ehrlich ascites carcinoma grown in vivo (Andersson and Heby, '72; Heby and Russell, '73). The L-ornithine decarboxylase activity (Heby and Russell, '73) and the putrescine concentration (Andersson and Heby, '72) were both found to decline as the growth rate of the tumor decreased with increasing tumor mass. To further evaluate a n y positive correlation between growth rate and cellular polyamine content and/or rate of synthesis we have studied the polyamine metabolism of a rat brain tumor cell line grown in vitro. The growth curve, which exhibits an S-shaped pattern with a n initial lag phase, followed by a n exponential phase and finally a stationary phase, thus displaying both a phase of increasing and a phase of decreasing growth rate, was utilized to mathematically derive the specific growth rate. The specific growth rate was then correlated with the activities of the enzymes in the polyamine biosynthetic pathway and with the concentrations of the polyamines at various stages of growth. MATERIALS A N D METHODS
Rat brain tumor cells The tumor cells used in this study were provided by W. H. Sweet, P. T. Kornblith, J..R. Messer and B. 0. Whitman of the Massachusetts General Hospital, Boston, Massachusetts. The brain tumor was induced in CD Fischer rats by weekly intravenous injections of N-methylnitrosourea. Tissue culture and preservation methods have been described by Benda et al. ('71). Upon receipt of frozen cells from the above source, they were suspended in Eagle's basal medium (BME) supplemented with L-glutamine (:398 mg/l), fetal calf serum (10%), and antibiotics (penicillin, 80.5 IU/ ml; streptomycin, 80.5 IU/ml), and grown by the monolayer method for continuous cell culture described by Wilson et al. ('66) and Barker et al. ('72). Experimental protocol A suspension of 1 X 1 0 6 tumor cells in 15 ml of BME with Earle's balanced salt solution supplemented with L-glutamine (292 mg/l), fetal calf serum (lo%), BME vitamin mixture (1 % ), BME essential ami-
no acid mixture (1 % ), and penicillin-streptomycin mixture ( 1 % ) was pipetted into plastic Falcon flasks (75 cm'; 250 ml). The cell cultures were grown in a National COP-incubator (5% COs: 95% air; 37°C). The cells were harvested at 24-hour intervals by trypsinization as follows. The growth medium was decanted and the cell monolayer was washed twice with 2 ml of 0.25% trypsin solution prepared with Caz+- and Mg*+-free Hanks balanced salt solution containing 0.02 % Na2EDTA, both washes being discarded. Only a thin film of trypsin solution was allowed to remain on the cell monolayer after the second washing. The culture vessels were incubated at 37°C until the cell monolayer detached from the plastic surface ( < 10 min). The cells were then transferred into a centrifuge tube with 5-10 ml of Ca"+- and Mg'i-free Hank's balanced salt solution. Cell clumps were dispersed with a capillary pipette. To determine the number of cells per Falcon flask, an aliquot of the dispersed cell suspension was counted in a hemocytometer. The cell suspension was then centrifuged at 900 X g for ten minutes a t 4 ° C and the cell pellet obtained was utilized immediately for enzyme assay or stored at - 20°C for subsequent polyamine analysis. For each experimental series, cells from 5- or 6-day cultures were trypsinized as described above, suspended in growth medium to a concentration of 1 X 106 cells per 15 milliliters, and plated into Falcon flasks. Preparation of tumor cell extracts for enzyme assays For the assay of L-ornithine decarboxylase activity, cellular pellets from single Falcon flasks were homogenized with a Lab-Line Ultratip Labsonic System (LabLine Instruments, Inc., Melrose Park, Ill.) equipped with a microtip, in 200-400 p1 of ice-cold homogenization medium (100 mM glycyl-glycine buffer, pH 7.2, containing 0.1 mM Na2EDTA and 5 mM dithiothreitol). The homogenates were centrifuged in a Sorvall RC2-B centrifuge at 45,000 X g for 90 minutes at 2°C. The resulting supernatant fraction was used for the assay of L-ornithine decarboxylase activity. For the assay of S-adenosyl-Lmethionine decarboxylase activity, cells
POLYAMINE METABOLISM A N D GROWTH RATE
513
from three Falcon flasks were combined, D isomer was assumed not to be decarsonicated and centrifuged as described boxylated by ornithine decarboxylase (Pegg above. The supernatant fraction was used and Williams-Ashman, '68a). for the assay of S-adenosyl-L-methionine Determination ofputrescine-activated decarboxylase activity. S-adenos yl-L-methionine decarboxylase activity Determination of L-ornithine decarboxylase ac tivity The reaction was followed by measureThe reaction was followed by measuring ment of the release of l-LC02 from (-) 4Cin the the release of ' T O 2 from DL-ornithine-1- S-adenosyl-L-methionine-carboxyl-l presence of saturating levels (2.5 mM) of 14C in the presence of saturating levels (2 mM) of L-ornithine mainly as described putrescine mainly as described by Pegg by Janne and Williams-Ashman ('71). The and Williams-Ashman ('68b, '69) and Copreaction mixture contained 100 pmoles of poc et al. ('71). In the presence of puglycylglycine buffer (pH 7.2), 2 pmoles of trescine there is a close correspondence L-ornithine, 0.5 pCi of DL-ornithine-1- between the release of 14C02 from (-)-S14C monohydrochloride (specific activity, adenosyl-L-methionine-carboxyl-14Cand the 7.66 mCi/mmole), 5 pmoles of dithiothrei- incorporation of radioisotope from putrestol, 0.2 pmoles of pyridoxal 5-phosphate cine-1,4-I4C into spermidine (Pegg and and 100 pl of tumor extract in a total Williams-Ashman, '68b). The reaction mixvolume of 1.00 ml. The reactions were ture contained 100 pmoles of glycylglycarried out in Corex centrifuge tubes (18 cine buffer (pH 7.2), 2.5 pmoles of pux 102 mm) equipped with rubber stop- trescine, 5 pmoles of dithiothreitol, 0.2 pers. The Won released during the 60- pCi of (-)-S-adenosyl-L-methionine-carminute incubation at 37°C was trapped boxyl-I4C(specific activity, 7.7 mCi/mmole) in 0.1 ml of 1 M Hyamine hydroxide con- and 100 pl of the tumor cell extract in a tained in a polypropylene well attached total volume of 1.00 ml. The tubes were to the rubber stopper. By the injection of incubated for 30 minutes at 37°C. Incu1 ml of 40% trichloroacetic acid through bations were carried out and radioactivity the rubber stopper the reactions were halt- was measured as described above for the ed and bound "C02was released from the L-ornithine decarboxylase assay. The rereaction mixture. A further incubation for lease of 'TO2 from the substrate was pro20 minutes at 37°C was allowed to ensure portional to the time of incubation over that all 14C02 released during the enzymic the 30 minute initial time period and to reaction was trapped in the Hyamine hy- the amount of soluble tumor extract droxide. The wells and their contents were added. then removed and placed in plastic countPolyamine analysis ing vials containing 2 ml of ethanol and Cellular pellets obtained from single 10 ml of Omnifluor toluene (0.4% Omnifluor in scintillation grade toluene). Radio- Falcon flasks were sonicated in 4% 5-sulactivity was assayed with a Beckman LS fosalicylic acid (approximately 200 pl/106 250 liquid-scintillation spectrometer. The cells). The homogenate was kept at 0°C counting efficiency for I4C was in the range for one hour and then centrifuged at 8000 of 90-95%. Sufficient counts were record- X g in an Eppendorf Micro Centrifuge ed to obtain a S.D. of the count rate of for five minutes. A 50 pl aliquot of the less than -+ 2 % . Corrections were applied supernatant was analyzed, utilizing a Durfor radioactivity of non-enzymic control rum D-500 amino acid analyzer, by a samples (100 p1 of the homogenization modification of the method previously demedium replacing the tumor cell extract) scribed (Marton et al., '74). Durrum (DC-PA cation exchange resin, in all enzyme experiments. Release of "Confrom the substrate in the L-ornithine a sulfonated polystyrene polymer with a decarboxylase assay was proportional to 12% cross linkage and a bead diameter the time of incubation over an initial 60- of 10 k 1 pm, was used instead of the minute period and to the amount of soluble Durrum DC-4A resin. The resin was packed tumor extract added. In calculating the to a height of 8 cm in a stainless steel rate of decarboxylation of L-ornithine the 1.75 mm internal diameter column. Two
514
0. HEBY, L. J. MARTON, C. B. WILSON A N D H . M. MARTINEZ
buffers were used to effect the separation of the polyamines. Buffer 1: A sodium citrate (0.20 M)-sodium chloride (0.35 M) mixture was prepared by dissolving 58.83 g of sodium citrate X 2 H20, 20.45 g of sodium chloride, 5.00 ml of thiodiglycol, and 1.00 ml of liquefied phenol in distilled water to a final volume of one liter and by adding concentrated HC1 until pH 5.09 +- 0.02 was obtained. The total Na+ concentration of this buffer was 0.95 M . Buffer 2: A sodium citrate (0.35 M)-sodium chloride (2.00 M) mixture was prepared by dissolving 102.90 g of sodium citrate X 2 HzO, 116.88 g of sodium chloride, 5.00 ml of thiodiglycol, and 1.00 ml of liquefied phenol in distilled water to a final volume of one liter and by adding concentrated HC1 until pH 4.68 +- 0.02 was obtained. The total Na+ concentration of this buffer was 3.05 M . The buffers were filtered through Millipore filters (47 mm diameter, 0.45 Fm pore size). Elution and separation of the polyamines in the tumor cell extracts was accomplished in 28 minutes as follows: Buffer 1 was directed to the column for five minutes to elute the amino acids and then Buffer 2 was used for 23 minutes to elute putrescine, spermidine and spermine. The buffer flow rate was 18.5 ml/hr and the flow rate of the ninhydrin reagent solution was 9.5 ml/hr. Column temperature was held isothermally at 66°C. Column pressure generated during analysis was approximately 1600 pounds per inch' gauge. Full scale deflection on the recorder was set at 0.1 O.D. units at 590 nm. Following each analysis the column was regenerated with a solution of 0.2 M NaOH containing 0.67 mM Na, EDTA and was then equilibrated with Buffer 1.
chloride, (-)-S-adenosyl-L-methionine-carboxyl-1W-iodide.monohydrate and Hy amine hydroxide were purchased from New England Nuclear Corp., Boston, Massachusetts. The preparations of DL-ornithine-1'Cmonohydrochloride contained variable amounts of 14C02, which increased the blank values unless removed. Therefore, prior to its use in the L-ornithine decarboxylase assays, the labeled amino acid was dissolved in 0.1 M HC1 and then lyophilized. The residue was dissolved in 0.01 M HC1 and stored at -20°C until used. After this treatment the nonenzymic control incubations (i.e., the complete incubation mixture minus enzyme) gave 75100 cpm over the 60 minute incubation period with the usual specific radioactivity of added L-ornithine (0.125 pCi per ymole of L-ornithine on the assumption that the label was equally distributed between the D- and L-isomer in the racemic mixture). Dithiothreitol, pyridoxal 5-phosphate, Lornithine and the hydrochlorides of putrescine, spermidine and spermine were purchased from Calbiochem, La Jolla, California. High-pressure liquid chromatographic analysis indicated that the specimens of putrescine, spermidine and spermine were uncontaminated by each other. Thiodiglycol and ninhydrin reagent solution were purchased from Pierce Chemical Co., Rockford, Illinois. Liquefied phenol was purchased from Matheson, Coleman and Bell, Norwood, Ohio. RESULTS
Figure 1 shows the growth data for the rat brain tumor cell line when it is grown in vitro. The growth curve exhibits an Sshaped pattern, similar to that obtained for other cell lines grown in vitro, with an initial lag phase, followed by an exponential increase in cell number with a douChemicals bling time of approximately 24 hours, and BME with Earle's balanced salt solution, finally a stationary phase. When the cells 200 mM L-glutamine solution, fetal calf are seeded at a density of 1.3 X lo4 cells/ serum, BME vitamin mixture ( X loo), cm2, they double about three to four times BME essential amino acid mixture ( X loo), and reach a stationary phase after about and penicillin-streptomycin mixture (5000 five days. The cell density remains approxILT each/ml) were obtained from Microbio- imately constant at 1.5 X 105 cells/cm* logical Associates, Bethesda, Maryland. thereafter. The decreasing growth rate that Hank's balanced salt solution, free of Ca2+ is apparent after day 3 may be due to a and Mg2+, was purchased from Grand decreasing growth fraction, a lengthenIsland Biological Company, Santa Clara, ing of the cell cycle, an increasing cell California. DL-Ornithine-l-'4C-monohydro-loss (due to cell death or exfoliation) or a
POLYAMINE METABOLISM AND GROWTH RATE
combination of these factors. Even though there is no apparent increase in cell number from day 5 through 7, the rat brain tumor cells do not stop dividing completely. Rather, a slow proliferation is balanced by a loss of cells into the culture medium
v
3t
/
2L/l 1
I
I
I
,
I
TIME AFTER PLATING (days)
1
Fig. 1 Number of cells per Falcon flask at various times after plating 1.0 X 106 rat brain tumor cells. Symbols represent means ( f S.E.M.) of values obtained in seven independent experiments.
515
leading to a constant cell number (Heby, unpublished results). Figure 2 shows the growth curve and a continuous representation of the growth rate of the tumor cells. This growth curve was obtained by a least squares fit of the growth data (fig. 1) with a fifth degree polynomial, P(t). The derivative curve of this polynomial was taken as a representation of the specific growth rate ( [dN [ t ] / dtj /N it] ). N [tj is the number of cells in the population at time t. The maximal growth rate, i.e., the inflection point of the growth curve and the peak of the derivative curve, occurs approximately two days after plating. Figures 3A and 4 A show the changes in activity of the first two enzymes in the polyamine biosynthetic pathway, L-ornithine decarboxylase and S-adenosyl-L-methionine decarboxylase. Figure 5A shows the changes in cellular polyamine content during the 7-day tumor growth period studied. The activities of L-ornithine decarboxylase (fig. 3A) and S-adenosyl-L-methionine decarboxylase (fig. 4A), as well as the cellular content of spermidine (fig. 5A) all exhibited similar patterns, with maxima during exponential growth and decreasing
TIME AFTER PLATING (days)
Fig. 2 Time course of population growth and continuous representation of the growth rate throughout the growth period of the rat brain tumor cells. Growth data is represented by 0 as log of cell number. The values plotted are the means of seven independent experiments (same as i n fig. 1). The growth curve is a fifth degree polynomial, P(t), fitted to the growth data, and the time derivative, dP/dt, of this polynomial represents the specific growth rate, [dN(t)/ dt]/N(t). of the tumor cells. N [t] is the number of cells in the population at time t. Both curves are computer drawn and the growth curve weighted to include the experimental error.
516
0.HEBY, L. J. MARTON, C. B. WILSON AND H. M. MARTINEZ
8 CORRELATION COEFFICIENT: 0.922
7-
65-
4-
3-
21-
i
OI
Q!
i
i
i
i
(day-')
TIME AFTER PLATING (days)
Fig. 3 Relationship between the L-ornithine decarboxylase activity and the specific growth rate of the rat brain tumor cells. (A) Enzyme activity (mean k S.E.M., n = %9) a t various times of growth. The growth curve is superimposed (-----). (B) Specific growth rate linearly fitted to the enzyme activity. Correlation coefficient = 0.922. The curve is computer drawn.
A ~
T
-
B
7-
CORRELATION COEfflClENl: 0.958
654-
3210
I
1
2
I
m -
3
4
5
(SPECIFIC GROWTH RATEtl)
6 X
35
(day-')
TIME AFTER PLATING (days)
Fig. 4 Relationship between the S-adenosyl-L-methionine decarboxylase activity and the specific growth rate of the rat brain tumor cells. (A) Enzyme activity (mean S.E.M.. n = 3) at various times of growth. Cells from three Falcon flasks were pooled for each determination. The growth curve is superimposed (- - - - -). (B) Specific growth rate linearly fitted to the enzyme activity. Correlation coefficient = 0.958. The curve is computer drawn.
*
517
POLYAMINE METABOLISM AND GROWTH RATE
- 40 - 30 7
- 20 Q
I
2 -10 -9
CL
1:
$
-5 -4
5 Lu u
-8
x
2
3-
-3 2CORRELATION COEFFICIENT: 0.293
-2
1L
2I
I
4I
I
6I
0
I
1-
I
8I (SPECIFIC GROWTH R A T E t I ) x 3.5
TIME AFTER PLATING (days)
(day-’)
C
l9 o/
CORRELATION COEFFICIENT: -0.475
I
- 4
1
2
3
4
5
6
(SPECIFIC GROWTH RATE tl) x 3 5 (day-’)
Fig. 5 Relationship between the polyamine content and the specific growth rate of the rat brain tumor cells. ( A ) Polyamine content (mean t S.E.M., n = 7) at various times of growth. 0 = Putrescine. 0 = spermidine, 0 = spermine. The growth curve is superimposed. Specific growth rate linearly fitted to the putrescine content (B), the spermidine content ( C ) , and the spermine content (D). Correlation coefficients = 0.293,0.951, and - 0.475, respectively. The curves are computer drawn.
518
0 . HEBY, L. J. MARTON, C. B. WILSON AND H . M. MARTINEZ
activities and content as the growth curve approached the plateau and the rate of cell division decreased. The cellular putrescine content (fig. 5A) showed a different pattern with a peak just prior to the stationary phase, and the spermine content (fig. 5A) showed an inverse pattern with a slightly decreased level during exponential growth. The patterns of changes obtained for the activities of L-ornithine decarboxylase and S-adenosyl-L-methionine decarboxylase and for the spermidine content, resembled that of the specific growth rate (fig. 2). To further evaluate the correlation between growth rate and polyamine content and the rate of polyamine synthesis, the specific growth rate ( [dN [ t ]/dt] /N[ t ]) was linearly fitted to these parameters. Figures 3B, 4B and 5B-D show the linear relationships of the enzyme activities and the polyamine contents versus the specific growth rate. The correlation coefficients for the relationships to the specific growth rate were 0.922, 0.958, 0.293, 0.951 and -0.475 for L-ornithine decarboxylase, S-adenosyl-L-methionine decarboxylase, putrescine, spermidine and spermine, respectively. Thus, the amounts of putrescine and spermidine synthesized, as measured by the enzyme activities, are directly related to the growth rate of the tumor cells. The fact that there was a negative relations hip between the spermine content and the specific growth rate means that as the rate of cell replication increases the spermine content decreases. DISCUSSION
The polyamine biosynthetic pathway involves four sequential enzyme reactions catalyzed by L-ornithine decarboxylase (Pegg and Williams-Ashman, '68a), S-adenosyl-L-methionine decarboxylase (Pegg and Williams-Ashman, '68b, '69), spermidine synthase (Janne et al., '71; Raina and Hannonen, '71) and spermine synthase (Hannonen et al., '72). L-Ornithine decarboxylase plays a key role in the biosynthesis of the polyamines since decarboxylation of L-ornithine is the only pathway for putrescine formation in mammals (Williams-Ashman et al., '72b). It may regulate the disposition of L-ornithine visa-,is other metabolic pathways such as the formation of citrulline and glutamate (Williams-Ashman et al., '69; Weber et al.,
'72; Tomino et al., '74). S-Adenosyl-Lmethionine decarboxylase catalyzes the formation of 5'-deoxy-5'-S-(3-methylthiopropylamine) sulfonium adenosine whose propylamine moiety is then conjugated with a putrescine molecule, through the catalysis of spermidine synthase, yielding spermidine. In a similar manner a propylamine moiety is conjugated with a spermidine molecule, through the catalysis of spermine synthase, yielding spermine. In addition to serving a s a substrate in the spermidine synthetic reaction, putrescine stimulates the activity of S-adenosyl-L-methionine decarboxylase, thereby initiating the propylamine transfer reaction (Williams-Ashman et al., '72b). The intracellular levels of the polyamines are determined in large measure by their rates of synthesis. The high cellular putrescine content in the rat brain tumor cells may be explained by the observation that the activity of L-ornithine carboxylase is a t least 100-fold higher than that of Sadenosyl-L-methionine decarboxylase during the entire growth period. In most mammalian tissues, both normal and neoplastic, the putrescine concentration is quite low (Williams-Ashman et al., '72a,b; Heby et al., '73; Marton et al., '74) and constitutes only a fraction of the total polyamine content, but in the rat brain tumor cell line it exhibits a high level which amounts to 6@-90% of the spermidine content, and exceeds the spermine content by 2 0 4 0 % during the in vitro growth period. The high putrescine level in the brain tumor cells is not due to contamination with other substances; and normal mammalian tissues, when analyzed with our procedure, exhibit putrescine levels which are consistent with the low levels reported by others (Marton et al.,'74). The fact that the cellular content of spermine is always lower than that of putrescine and spermidine may be due to inhibition of spermine synthesis by the high cellular putrescine level, inasmuch as putrescine is a potent competitive inhibitor of spermine synthase (Hannonen et al., '72). Even though spermine is a competitive inhibitor of L-ornithine decarboxylase (Pegg and WilliamsAshman, '68a), because of its low cellular content and since i t is a rather weak inhibitor, it is not likely to markedly decrease the putrescine synthesis. Elevated activities of the enzymes in the
POLYAMINE METABOLISM AND GROWTH RATE
polyamine biosynthetic pathway and elevated levels of the polyamines have been observed in many models of stimulated growth (Cohen, '71; Tabor and Tabor, '72; Bachrach, '73; Heby et al., '74), as well as in tumor cell populations when compared to their normal counterparts (Williams-Ashman et al., '72a; Heby et al., '73; Bachrach et al., '74). An elevated putrescine concentration seems to be a general characteristic of cells not only with a high proliferation rate, but of cells that have undergone neoplastic transformation, e.g., virus-transformed cells have been found to have a higher putrescine content than their normal non-infected counterparts despite the fact that they have similar growth rates (Bachrach et al., '74). It was also noted that the intracellular putrescine level increased during the virusinduced transformation and that it remained elevated in the transformed cells even upon subculturing. The results of several recent investigations indicate the probable existence of a positive correlation between tumor cell growth rate and the activities of the enzymes in the polyamine biosynthetic pathway, as well as the endogenous concentrations of the polyamines (Anderson and Heby, '72; Williams-Ashman et al., '72a; Heby and Russell, '73). The present study, however, is the first demonstration of any statistically significant correlations. The activities of L-ornithine decarboxylase and S-adenosyl-L-methionine decarboxylase, as well as the cellular content of spermidine, showed high linear correlations with the specific growth rate of the rat brain tumor cells when grown in culture. Putrescine and spermine, however, did not show significant linear correlations with the specific growth rate. In fact, spermine showed a negative correlation coefficient, implying that the average cellular spermine content decreases rather than increases with increasing specific growth rate. In accordance with this observation, i t has been found that the concentration of spermine decreases in the liver after partial hepatectomy, i.e., as the cells are stimulated to traverse the cell cycle from a quiescent state (Janne, '67; Heby and Lewan, '71). The decrease in spermine concentration noted in regenerating liver, which is a much more pronounced decrease than the one observed in the present study, may be
519
at least partially attributable to tissue hypertrophy. In the present study the decrease in cellular spermine content cannot be due to hypertrophy because spermine was determined on a per cell basis. In the rat brain tumor cell line the decrease in spermine content may be attributed to conversion of spermine into spermidine. In fact Siimes ('67) showed that radioactive spermine was converted into spermidine with a maximal velocity at the time of liver regeneration when the decrease in the endogenous concentration of spermine occurred. In addition to the possibility that the decrease in cellular spermine content could be due to conversion of spermine into spermidine, spermine synthesis may be inhibited by the accumulation of endogenous putrescine which occurs during the same period of growth. As previously mentioned, putrescine has been found to be a potent competitive inhibitor 0s spermine synthase (Hannonen et al., '72). The observation that the activity of Lornithine decarboxylase in hepatomas increases with increasing growth rate (Williams-Ashman et al., '72a) whereas the activities of L-ornithine carbamyltransferase, i.e., citrulline synthesis (Weber et al., '72) and L-ornithine transaminase, i.e., glutamate synthesis (Tomino et al., '74) decrease, demonstrates that L-ornithine is preferentially channeled into polyamine synthesis with increasing growth rate. Weber et al. (72) have suggested that the increased utilization of L-ornithine for polyamine synthesis may confer a selective advantage to the hepatoma cells and that it might contribute to the increased growth rate. The fact that the peak values of the L-ornithine decarboxylase and S-adenosylL-methionine decarboxylase activities and the cellular spermidine content coincide with the inflection point of the growth curve, i.e., the time of maximum growth rate of the tumor cells, and the fact that these parameters show a high positive correlation with the specific growth rate of the tumor cells throughout their in vitro growth period, implies that the polyamines may be essential to the process of cell replication. However, these results do not necessarily indicate that the major function of the polyamines relates to cell division per se. The parasynchronous culture of the rat brain tumor cells does not allow
520
0 HEBY, L. J . MARTON, C. B. WILSON AND H. M. MARTINEZ
us to discern the definitive temporal relationships between polyamine synthesis and cell cycle parameters. In fact, though results obtained from growth curves have indicated a positive correlation between the rate of cell division and the rate of incorporation of DNA precursors (Todo et al., '71; Wheeler and Alexander, '72; Zinninger and Little, '73; Rosenblatt and Erbe, '73; Macieira-Coelho, '73), it is known that DNA synthesis precedes cell division by many hours. Studies relating cell cycle kinetics to growth data have shown that the average cell cycle time is shortest during exponential growth, when the relative time spent in S, G2 and M is maximal and in GI minimal (Todo et al., '71; Zinninger and Little, '73). On this basis we interpret our results in terms of an obligatory involvement of the polyamines in events occurring while cells are proceeding through the S, G2, and M phases of the cell cycle. The decrease in the extent of polyamine synthesis observed with decreasing growth rate is probably a reflection of the increased fraction of the overall cell cycle time spent in the GI phase, and of the decreased number of cells actually involved in proliferation. The fact that the fraction of the cell cycle spent in GI increases in the stationary phase while the rate of polyamine synthesis decreases, argues against polyamine synthesis and accumulation occurring during the G I phase of the cell cycle. Further support for this view is given by our recent observations that WI-38 human diploid fibroblasts stimulated to proliferate from a quiescent state (Heby et al., '74), and synchronous cultures of Chinese hamster ovary cells, obtained by mitotic selection of cells in exponential growth (Heby et al., '75, submitted), initiate their synthesis of polyamines in late GI or early S and continue to accumulate polyamines until they divide. Therefore, we believe that the peak of putrescine synthesis which has been observed in early GI in many systems of stimulated growth may be unrelated to the cell's preparation for DNA synthesis and division. It may be part of a hypertrophic response due to release of hormones, or the exposure to some growthstiniulatory factors. ACKNOWLEDGMENTS
We thank Msi. Kathy D. Knebel for her
technical assistance. This work was supported by NIH Center Grant CA-13525 and gifts from the Phi Beta Psi Sorority, the Joe Gheen Medical Foundation, and the Association for Brain Tumor Research. LITERATURE CITED Andersson, G., and 0. Heby 1972 Polyamine and nucleic acid concentrations in Ehrlich ascites carcinoma cells and liver of tumor-bearing mice a t various stages of tumor growth. J. Nat. Cancer Inst., 48: 165-172. Bachrach, U. 1973 Function of naturally occurring polyamines. Academic Press, New York. Bachrach, U., S. Don and H. Wiener 1974 Polyamines in normal a n d in virus-transformed chick embryo fibroblasts. Cancer Res., 34: 1577-1580. Barker, M.. T. Hoshino and C. B. Wilson 1972 Tissue culture of h u m a n brain tumors. I n : The Experimental Biology of Brain Tumors. W. N. Kirsch, E. G. Paoletti and P. Paoletti, eds. Charles C Thomas, Publisher, Springfield, Illinois, pp. 57-84. Benda, P., K. Someda, J. Messer and W. H. Sweet 1971 Morphological and immunochemical studies of rat glial tumors and clonal strains propagated i n culture. J. Neurosurg., 34: 3 1 0 3 2 3 . Cohen, S. S. 1971 Introduction to the Polyamines. Prentice-Hall, Inc., Englewood Cliffs, New Jersey. Coppoc. G . L.. P. Kallio and H . G . WilliamsAshman 1971 Characteristics of S-adenosylL-methionine decarboxylase from various organisms. Int. J, Biochem., 2: 6 7 3 4 8 1 . Echave Llanos. J. M.. and R. E. Nash 1970 Mitotic circadian rhythm in a fast-growing a n d a slow-growing hepatoma: Mitotic rhythm in hepatomas. J. Nat. Cancer Inst.. 44: 581-585. Hannonen, P., J . J a n n e and A. Raina 1972 Partial purification and characterization of spermine synthase from rat brain. Biochim. Biophys. Acta, 289: 2 2 5 2 3 1 . Hayashi, S., Y . Aramaki and T. Noguchi 1972 Diurnal change in ornithine decarboxylase activity of rat liver. Biochem. Biophys. Res. Comm u n . , 46: 7 9 5 4 0 0 . Heby. 0.. J . W. Gray. P. Lindl. L. J. Marton and C. B. Wilson (1975. submitted) Changes in polyamine content during the cell cycle of Chinese hamster ovary cells. Heby, O., a n d L. Lewan 1971 Putrescine and polyamines in relation to nucleic acids in mouse liver after partial hepatectomy. Virchows Arch. Abt. B. Zellpath., 8: 58-66. Heby, O., L. J. Marton, L. Zardi, D. H. Russell and R. Baserga 1975 Accumulation of polyamines after stimulation of cellular proliferation i n h u m a n diploid fibroblasts. I n : The Cell Cycle i n Malignancy and Immunity. J. C. Hampton, ed. NTIS. Springfield. Virginia, pp. 5 0 4 6 . Heby, O., a n d D. H. Russell 1973 Changes i n polyamine metabolism in tumor cells and host tissues during tumor growth a n d after treatment with various anticancer agents. In: Polyamines in Normal a n d Neoplastic Growth. D. H. Russell, ed. Raven Press, New York, pp. 221237. Hebv. 0.. G. P. Sarna. L. .T. Marton. M. Omine. S - Perry a n d D H Russ"el1 1973 Polyamine
POLYAMINE METABOLISM AND GROWTH RATE content of AKR leukemic cells i n relation to the cell cycle. Cancer Res., 33: 295S2964. Janne, J. 1967 Studies o n the biosynthetic pathway of polyamines i n rat liver. Acta Physiol. Scand., Suppl., 300: 1-71. Janne, J., A. Schenone and H. G. Williams-Ashm a n 1971 Separation of two proteins required for synthesis of spermidine from S-adenosylL-methionine a n d putrescine in r a t prostate. Biochem. Biophys. Res. Commun., 42: 758-764. Janne, J., a n d H. G. Williams-Ashman 1971 On the purification of L-ornithine decarboxylase from r a t prostate and effects of thiol compounds o n the enzyme. J . Biol. Chem., 246: 1725-1 732. Macieira-Coelho, A . 1973 Cell cycle analysis. A. Mammalian cells. I n : Tissue Culture. Methods and Applications. P. F. Kruse, Jr. a n d M. K. Patterson, Jr., eds. Academic Press, New York, pp. 4 1 2 4 2 2 . Marton, L. J., 0. Heby, C. B. Wilson a n d P. L. Y. Lee 1974 An automated micromethod for the quantitative analysis of di- a n d polyamines utilizing a sensitive high pressure liquid chromatographic procedure. FEBS Lett., 41 : 99-103. Nash, R. E., and J. M. Echave-Llanos 1971 Twenty-four-hour variations in DNA synthesis of a fast-growing a n d a slow-growing hepatoma: DNA synthesis rhythm i n hepatoma. J. Nat. Cancer Inst., 47: 1007-1012. Pegg, A. E., a n d H. G. Williams-Ashman 1968a Biosynthesis of putrescine i n the prostate gland of the rat. Biochem. J., 108: 533-539. 1968b Stimulation of the decarboxylation of S-adenosylmethionine by putrescine i n mammalian tissues. Biochem. Biophys. Res. Commun., 30: 7 6 8 2 . 1969 On the role of S-adenosyl-L-methionine in the biosynthesis of spermidine by rat prostate. J. Biol. Chem., 244: 682-693. Raina, A,, and P. Hannonen 1971 Separation of enzyme activities catalysing spermidine a n d spermine synthesis i n rat brain. FEBS Lett., 16: 1-4. Rosenblatt, D. S . , a n d R. W. Erbe 1973 Reciprocal changes in the levels of functionally related folate enzymes during the culture cycle i n h u m a n fibroblasts. Biochem. Biophys. Res. Comm u n . , 54: 1627-1633.
521
Siimes. M . 1967 Studies on the metabolism of 1.4-"C-spermidine and 1.4-IIC-spermine in the rat. Acta Physiol. Scand.. Suppl.. 298. 1-66. Tabor, H., and C. W. Tabor 1972 Biosynthesis and metabolism of 1,4-diaminobutane, spermidine, spermine, and related amines. Adv. Enzymol., 36: 2 0 S 2 6 8 . Todo, A , , A. Strife, J . Fried and B. D. Clarkson 1971 Proliferative kinetics of h u m a n hematopoietic cells during different growth phases i n vitro. Cancer Res., 31: 133G1340. Tomino, I., N. Katunuma, H . P. Morris and G. Weber 1974 Imbalance in ornithine metabolism in hepatomas of different growth rates a s expressed i n behavior of L-orthinine:2-oxoacid aminotransferase (ornithine transaminase EC 2.6.1.13).Cancer Res., 34: 6 2 7 4 3 6 . Weber, G., S . F. Queener a n d H. P. Morris 1972 Imbalance i n ornithine metabolism in hepatomas of different growth rates as expressed in behavior of L-ornithine carbamyl transferase activity. Cancer Res., 32: 193S1940. Wheeler, G. P., and J. A. Alexander 1972 Metabolic activities of murine carcinoma 755 at various stages of growth. Cancer Res., 32: 17611768. Willaims-Ashman. H. G., G. L. Coppoc and G . Weber 1972a Imbalance in ornithine metabolism in hepatomas of different growth rates as expressed in formation of putrescine, spermidine, and spermine. Cancer Res., 32. 19241932. Williams-Ashman. H. G . . J . Janne. G. L. Coppoc. M . E . Geroch and A. Schenone 1972b N e w aspects of polyamine biosynthesis in eukaryotic organisms. Advan. Enzyme Regul.. 1 0 : 225245. Williams-Ashman, H. G., A. E. Pegg a n d D. H. Lockwood 1969 Mechanisms and regulation of polyamine and putrescine biosynthesis in male genital glands a n d other tissues of mammals. Advan. Enzyme Regul., 7: 291-323. Wilson, C. B., M. Barker and D. E. Slagel 1966 Tumors of the central nelvous system in monolayer tissue. Arch. Neurol., 15: 275-282. Zinninger, G . F., a n d J. B. Little 1973 Proliferation kinetics of density-inhibited cultures of h u m a n cells, a complex in vitro cell system. Cancer Res., 33: 2343-2348.
