Regulation of choline kinase activity and phosphatidylcholine ...

THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1985 by The American Society of Biological Chemists, Inc.

Vol. 260,No. 10, Issue of May 25, pp. 6OO6-6011 1985 Printed in L~.s.A.

Regulation of Choline KinaseActivity and Phosphatidylcholine Biosynthesis by Mitogenic Growth Factorsin 3T3 Fibroblasts* (Received for publication, June 7, 1984,and in revised form, January 10, 1985)

Craig H. Warden$# and Morris Friedkinn From the $Department of Medicine, Divisionof Cardwlogy, University of California, Los Angeles, Los Angeles, California90024 and the (IDepartment of Biology, University of California, San Diego, La Jolla, California 92093

The regulation of choline kinase activity by fetal with growth factors increased phosphorylation of choline, in bovine serum and the regulation of phosphatidylcho- vitro choline kinase activity, phosphocholine and phosphoethline biosynthesis by choline kinase have been investi- anolamine pools, andPC synthesis (Warden et al., 1980; gated in 3T3 fibroblasts. Treatmentof quiescent 3T3 Warden and Friedkin, 1984). These results suggested that fibroblasts with serumwas shown in previous work to choline kinase could regulate PC biosynthesis in 3T3 cells. If increase phosphocholine pool size and phosphatidyl- choline kinase is aregulatory enzyme for PC synthesis in 3T3 choline biosynthesis.We now report that treatmentof cells, then it should show several additional properties. First, 3T3 cells with serum increased intracellular choline it should be a crossover point. Enzymes whose intracellular kinase activity by 2-3-fold with a concomitant 2-3- substrate concentrationsdecrease while their product concenfold decrease of intracellular free choline concentrations. Initial rates of choline transport were the same trations increase following an experimental manipulation are in quiescent and serum-treatedcells, whereas choline crossover points (Rolleston, 1972). Identification of crossover kinase activity was 2-%fold higher in serum-treated points gives a reliable indication of the in vivo sites of metacells. As aconsequence, free choline concentrations bolic regulation. Thus, intracellular concentrationsof choline were 2-3-fold lower in serum-stimulatedcells than in and phosphocholine were studied to determine if choline control quiescentcells. Phosphocholine turnoverrates kinase is at a crossover point for PC synthesis following serum were increased 2-fold by serum treatment both as a stimulation of 3T3 cells. Second, phosphocholine turnover was measured to confirm that increased phosphocholine conconsequence of a serum-dependent increase of phosphocholine pools andas a result of a serum-dependentcentrations do in fact increase conversion of phosphocholine to CDP-choline. Finally, activation of intracellular choline lowering of thephosphocholine half-life. Thus,the overall response of 3T3 cells to serum stimulation in- kinase by serum was measured since in vitro and in vivo cluded decreased choline pools and increased choline enzyme activities are notalways correlated (Vance et al., 1980; kinase activity, phosphocholine poolsize, phosphocho- Wohlhueter and Plagemann, 1981; Rozengurt et al., 1978). line turnover, and phosphatidylcholine biosynthesis. EXPERIMENTALPROCEDURES

Choline is incorporated into phosphatidylcholine (PC’) by a three-step pathway first described by Kennedy and Weiss (1956). Choline kinase catalyzes the first committed step, phosphorylation of choline to form phosphocholine. Phosphocholine andCTParethen converted to CDP-choline by CTP:phosphocholine cytidylyltransferase, and finally CDPcholine is incorporated into PC by choline phosphotransferase. Biosynthesis of phosphatidylcholine is clearly regulated by cytidylyltransferase, which catalyzes the rate-limiting stepfor PC synthesis (Vance and Choy, 1979). Activity of the cytidylyltransferase is regulated by phosphorylation (Pelech and Vance, 1982), translocation from cytosol to microsomes (Lim et al., 1983),increased cytoplasmic CTP concentrations (Choy et al., 1980), and increased concentrations of phosphocholine (Vigo and Vance, 1981;Paddon et al., 1982). In previous work, we have shown that stimulation of quiescent 3T3 fibroblasts

* This work was supported by United States Public Health Service Grant CA 11449 from the National Cancer Institute. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18U.S.C. Section 1734 solelyto indicate this fact. 5 To whom correspondence should be addressed. ’ The abbreviations used are: PC, phosphatidylcholine; DMEM, Dulbecco’s modified Eagle’s medium; MEM, Eagle’s minimal essential medium.

Cell Culture-Swiss mouse 3T3 fibroblasts were cultured in 100mm Falcon brand tissue culture dishes in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum, 100 units/ml penicillin, and 100 pg/ml streptomycin in a humidified atmosphere of 5% CO, and 95% air at 37 “C. Cells subcultured into 30-mm Nunc brand dishes in DMEM with 10% serum were utilized for experiments 7-14 days after reaching confluence. Assay of Intracellular Choline Kinase Activity-In order to measure intracellular choline kinase activity it is necessary to: 1) know the specific activity of free intracellular choline, 2) separate choline from its labeled products; and 3) show that intracellular choline phosphorylation is linear with time. 1)When 3T3 cells are exposed to labeled choline, radioactivity in free intracellular choline increases for 5 min and then remains a t a constant level for at least 3 h (Figs. lA and 3A). Thus, after 5 min of exposure to labeled choline, the specific activity of intracellular choline can be approximated by the specific activity of labeled choline in the culture medium. This approximation will be accurate as long as 3T3 cells are not generating significant amounts of unlabeled choline by degradation of pre-existing PC. Intracellular phosphocholine specific activity was identical with the specific activity of choline in the labeling medium after exposing 3T3 cells to labeled choline for 60 min (Warden and Friedkin, 1984);thus, there is no dilution of intracellular labeled choline by unlabeled cellderived choline. 2) 3T3 fibroblasts do not make betaine (Warden and Friedkin, 1984). Thus, the products of choline metabolism in 3T3 cells are confined to the products of the choline kinase reaction: phosphocholine, CDP-choline, PC, lysolecithin, and glycerophosphorylcholine. To assay intracellular choline kinase activity, it is necessary to measure the radioactivity present in all of these products, since they will all be produced during exposures to labeled choline. Formation of choline kinase products can be quantitated from two assays. First, acid-soluble choline, phosphocholine, CDP-choline, and

6006

6007

Regulation of CholineKinaseActivity glycerophosphorylcholine are separated from uptake into acid-insoluble material (PC andlysolecithin) by extraction of labeled cells with 5% trichloroacetic acid or with 0.4 M HC10,. Second, choline can be separated from acid-soluble choline kinase products by thin-layer chromatography (Warden and Friedkin, 1984). Finally, the quantity of choline kinase products formed is calculated by adding together uptake into acid-insoluble material and all acid-soluble radioactivity other than choline. 3) Figs. 1 and 3can be utilized to estimate intracellular choline kinase activity since formation of choline kinase products is linear with time. However, measurements of the rate of formation of choline kinase products, as described above, will lead to underestimates of the true rate of intracellular choline kinase activity, since it has been reported that cells prelabeled with choline release label back into the culture medium during chase periods (Vance et al., 1980).Thus, theactual rate of intracellular choline kinase activity could be determined by adding together the combined choline kinase products which are cell-associated, measured as described above, and the choline kinase products which were released into the medium. When 3T3 cells were prelabeled with ("C]choline for 3 h a t 37 "C, the rate of release of label into medium, measured at 15-min intervals during a 90-min chase, was approximately 12, 14 or 26 pmol/mg of cell protein/min for cells exposed to culture medium, insulin, or fetal bovine serum, respectively (Fig. 6). In an analogous experiment, Vance et al. (1980) showed that phosphocholine accounted for 80% of the label lost from HeLa cells. Even if one assumes that all of the label lost from 3T3 cells is phosphocholine, rather thancholine, then the rate of loss of label from 3T3 cells is at most 12"25% of the rate of formation of cell-associated choline kinase products at 36 or 7 pM choline (Figs. 1 and 3 and Table I). Since we routinely measured cellassociated choline kinase products, true intracellular choline kinase activity was underestimated by approximately 12-25%; however, serum increased formation of all choline kinase products by 2-fold, both those which remain cell-associated and those released into the culture medium. Thus, use of cell-associated choline kinase products in our calculations has no qualitative effect on the results. Other Materials and Methods-Measurement of uptake into acidsoluble and acid-insoluble material and thin-layer chromatography of acid-soluble material have been described by Warden and Friedkin (1984). Protein was determined by the method of Lowry et al. (1951) with saline-washed cells. Crystalline bovine insulin was a gift from Bill Bromer of Eli Lilly & Co. [l'C,SH]choline was purchased from Amersham Corp. Dulbecco's modified Eagle's medium and Eagle's minimum essential medium were obtained from Grand Island Biological Co. RESULTS

a

60

10

TIME(MIN)

60

"lQ TIME(MIN)

FIG.1. Time course of choline uptake and phosphorylation at 7 PM choline. Confluent quiescent cultures of 3T3 fibroblasts, grown on 30-mm Nunc brand dishes, were washed twice with DMEM to remove residual serum and were then incubated in 2 ml of DMEM or DMEM containing 10%fetal bovine serum (A),5 gg/ml insulin (B), alone (0).Six hours later, thesecultures and a fourth set of dishes of quiescent cells (V)were washed twice with DMEM and then incubated with ["Clcholine (58 Ci/mol, 7 p ~ in) MEM at 37 "C. After 1, 2.5, 5, 10, or 60 min, culture dishes were washed, acid-soluble radioactivity was extracted with 0.4 M HC104,HClO, was neutralized with KOH, and choline was separated from phosphocholine, CDP-choline, and glycerophosphorylcholine,as described by Warden and Friedkin (1984). Acid-insoluble uptake was measured as reported by Warden and Friedkin (1984). All data reported are the average of duplicate determinations. A, free intracellular labeled choline (nanomoles/ milligram of cell protein); B , phosphorylated acid-soluble radioactivity, i.e. label in phosphocholine, CDP-choline, and glycerophosphorylcholine (nanomoles/milligram of cell protein); C, per cent of acidsoluble radioactivity which is phosphorylated (per centof acid-soluble radioactivity in phosphocholine plus CDP-choline plus glycerophosphorylcholine); D,formation of combined choline kinase products. Acid-insoluble uptake after 60 min of exposure of [14C]cholinewas 242, 208,552, and 727 pmol/mg of cell protein for quiescent cells or cells treated with DMEM, 5 pg/ml insulin, or 10%serum, respectively. Formation of choline kinase products, in nanomoles/milligram of cell protein, was calculated as described under "Experimental Procedures."

respectively. Uptake into acid-insoluble material measures Time Course of Choline Phosphorylation-Stimulation of PC biosynthesis (Warden and Friedkin, 1984). Uptake into PC synthesis by growth factors can be measured by exposing acid-insoluble material and the data in Fig. 1B were utilized quiescent 3T3 cells to growth factors or DMEM for 6 h, to measure formation of the products of the choline kinase washing, treating with labeled choline for 60 min at 37 "C, reaction (Fig. 1D).Formation of choline kinase products was and then measuring uptake into acid-soluble and acid-insol- linear with time in serum and in insulin-stimulated cells, with uble material (Warden and Friedkin, 1984). In order to dem- correlation coefficients exceeding 0.99. In quiescent and onstrate thatintracellular choline phosphorylation was linear DMEM-treated cells, the rate of choline phosphorylation with time,this protocol was altered for the experiment shown reached steady-state levels after a 5-10-min lag (Fig. 1D). in Fig. 1 by labeling previously stimulated cells with ["C] Decreases of choline pool size in cells exposed to 8.4 ~ L M choline for the times shown. Free intracellularlabeled choline [14C]choline wereclearly evident and virtually maximal after increased rapidly for 5 min andinall cases had reached 60 min of stimulation with serum or insulin (Fig. 2). In maximal levels by 10 min (Fig. l.4). Control quiescent cells another experiment, choline pools decreased most markedly or cells exposed to DMEM for 6 h had %fold Iarger choline after 15-30 min of exposure to serum (data not shown). pools than cells stimulated with 10% fetal bovine serum or Choline Poolsand Phosphorylation at a Concentration above insulin (Fig. lA).Incorporation of radioactivity into acid- the K,,, for Transport-The experiments shown in Figs. 1and soluble products of choline kinase was approximately linear 2 were conducted with 7 or 8.4 p~ choline in the labeling over the 60-min exposure to [14C]choline(Fig. 1B). The data medium. When the initial rates of choline transport were from Fig. 1, A and B, were utilized to calculate the per cent measured by incubating quiescent 3T3 cells with several conof acid-soluble material which is phosphorylated (Fig. IC). In centrations of labeled choline for 60 s, the K, for transport cells exposed to serum or insulin, choline phosphorylation was found to be 16 k 2 p M (n = three experiments, *S.D., proceeded rapidly and was nearly maximal by 10 min, while data not shown). When cells were exposed to 36 PM labeled quiescent and DMEM-treated cells showed a strikingly dif- choline, intracellular free choline pools werethe same from 1, ferent pattern of linear time-dependent increases (Fig. IC). 2, and 3 h; however, serum-stimulated cells showed a 2-fold At the 60-min time point, uptake intoacid-insoluble material decrease of choline pools compared with DMEM- or with was 242, 208, 552, and 727 pmol/mg of cell protein for quies- insulin-treated cells (Fig. 3A). Amounts of phosphorylated cent cells or cells treated with DMEM, insulin, or serum, acid-soluble material formed were3-fold higher in serum-

6008

Regulation of Choline Kinase Activity

parable results were obtained from a second experiment at 36 p~ choline. When data for choline pools at all time points from two experiments at 36 pM choline were combined, average intracellular choline pools were 1530f 240, 1380 -t 180, and 760 -+ 220 pmol/mg of protein & S.D. (n = 5 ) for cells treated with DMEM, insulin, or serum, respectively. Insulinandserum-stimulated cells had, respectively, 2- or 4-fold greater uptake into acid-insoluble material than controls at all time points (Fig.3C). Formation of all choline kinase products, calculated by combining data in Fig.3, B and C (Fig. 3D), was linear throughout the 3-h exposure to ["CC] choline. Intracellular choline kinase activity, estimated by combining the data from all of the time points, was 181 f 9.5 (n = 3), 82 ? 2.6 (n = 3), and 61 f 5.7 (n = 6) pmol/mg of L I I 1 1 cell protein/min -+ S.D. for cells exposed to serum, insulin, or 1 2 3 DMEM, respectively. Increased Choline Phosphorylation Decreases Choline Pools TIME(H0URS) FIG.2. Choline pool size is rapidly decreased by serum and in Serum-stimulated cells-Figs. 1-3 clearly show that intrainsulin. Confluent quiescent 3T3cells (V) were first exposed to cellular choline concentrations are reduced by serum. These DMEM (O), to DMEM plus 5 pg/ml insulin (a),or to 10% fetal data suggest that increased choline kinase activity is responbovine serum (A) at 37 "C for 1, 2, or 3 h. They were then washed sible for the decreased choline concentrations in serum-stimwith MEM, exposed to MEM with ["Clcholine (58 Ci/mol, 8.4 p h t ) ulated cells. Intracellular choline concentrations in 3T3 cells for 30 min at 37 "C, extracted with HClO4, and analyzed for radioactivity in choline as described for Fig. 1. Data are the mean of three will depend on rates of choline influx (initial rateof transport) and on the rate at which choline is removed from the intraseparate measurements of each point & S.D. cellular pool by efflux from the cell and phosphorylation by choline kinase. Initial rates of choline transport were linear with time from 15, 30, and 60 s and were approximately the same in quiescent and in serum-stimulated cells (Fig. 4); thus, any modulations of intracellular choline concentrations must depend on alterations of efflux or phosphorylation. Combined data from several experiments clearly show that the rate of intracellular choline kinase activity was2-fold higher in serum-stimulated cells than in quiescent ones (Table I), which suggests that serum decreases choline pools by increasing removal of choline from the intracellular poolby choline kinase. Half-life and Turnover Time of Phosphocholine-The halflife and turnover time of phosphocholine were measured to quantitatethe consequences of increased phosphocholine pools for PC synthesis in serum-stimulated 3T3 cells (Fig. 5 ) . Cells treated with serum had shorter half-lifes for ["Clphosphocholine than cells exposed to DMEM or insulin (Table 1 2 3 1 2 3 11). The radioactive half-life of phosphocholine can be used TIME(H0URS) TIME(H0URS) to calculate a turnover time, which is the amount of time FIG.3. Time course of choline uptake and phosphorylation at 36 PM choline. Confluent quiescent 3T3 cells were first exposed required to synthesize one complete pool of phosphocholine to DMEM (O),to DMEM plus 5 pg/ml insulin (m), or to 10% dialyzed (Table 11). Dividing phosphocholine pool size (Warden and fetal bovine serum (A)at 37 "C for 3 h. They were then switched to Friedkin, 1984) by the turnover time yields a turnover rate, . identical media containing ["C]choline (6.9 Ci/mol, 36 p ~ )Duplicate which is the rateof loss of radioactivity from phosphocholine samples were extracted after 1, 2, and 3 h at 37 "C and analyzed as (Table 11). The phosphocholine turnover rate is increased described for Fig. 1. Data are expressed as nanomoles of ["C]choline/ more than 2-fold in serum-stimulated cells, as a result of milligram of cell protein. Time is given on the abscissa as hours after addition of ["C]choline. A, free intracellular labeled choline; B , increased phosphocholine pool size and a shorter phosphophosphorylated acid-soluble label; C, acid-insoluble radioactivity; D, choline half-life. Thus, increased phosphocholine pools, combined choline kinase products. brought about by serum treatment, may contribute to an increased conversion of phosphocholine to CDP-choline by stimulated cells than in DMEM- and insulin-treated cells cytidylyltransferase, the rate-limiting stepof PC synthesis. (Fig. 3B). Comparison of Figs. 1 and 3 reveals that choline Fig. 6 shows the appearance of radioactivity in culture pools and quantities of phosphorylated acid-soluble label promedium from cells prelabeled with ['4C]choline for the experduced were 2-fold higher in serum- and DMEM-treated cells iment shown in Fig. 5. Clearly, serum-stimulated cells loose exposed to 36 p~ choline than in cells treated with 7 p~ choline. Choline pools of insulin-stimulated cells show differ- radioactivity twice as fast as quiescent or insulin-stimulated ent patterns at 7 or 36 pM choline. Insulin-stimulated cells cells. Thus, comparisons of in vivo choline kinase activity in have 3-fold lower choline pools than DMEM-treated cells at quiescent or serum-stimulated cells (Figs. 1 and 3 and Table low choline concentrations (Figs. lA and 2). However, at 36 I) are notqualitatively affected by our reliance on the radiop~ choline, intracellular choline pools of DMEM- and insulin- activity present in intracellular choline kinase products for stimulated cells are notsignificantly different (Fig. 3A). Com- measurement of in vivo choline kinase activity. 1

6009

Regulation of CholineKinaseActivity

TIME.SEC

TIME(H0URS)

FIG. 4 (left). Initial rate of choline transport. Quiescent 3T3 cells (0)or cells first exposed to 10% serum for 6 h (A) were washed twice with choline-free MEM at 37 “C andfurther incubated for 15,30, or 60 s in MEM containing [3H]choline (1430 Ci/mol, 7 p M ) at 37 “C. Washing, trichloroacetic acid extraction, and analysis were as described by Warden and Friedkin (1984). Data shown are the mean f S.D. of three separate determinations. FIG. 5 (right). Half-life of labeled phosphocholine. Confluent quiescent cells were first exposed to DMEM (O),to DMEM plus 5 pg/ml insulin (m), or to 10% dialyzed fetal bovine serum (A) for 3 h. They were then exposed to identical media containing [“Clcholine (6.9 Ci/mol, 36 p M ) for 3 h a t 37 “C. Following a total of 6 h of exposure to DMEM or to DMEM plus insulin or serum, including 3 h of exposure to [“Clcholine, each dish was washed three times and reincubated at 37 ”C in its respective medium, without [“Clcholine. Duplicate samples were then harvested at 15-min intervals, and radioactivity in phosphocholine was determined as described by Warden and Friedkin (1984). Lines were drawn from least squares fits of the data, which are expressed as nanomoles of [“C] phosphocholine/milligram of cell protein. Time is given in the abscissa as hours after removal of [14C]choline.

TABLEI Initial rate of choline transport and intracellular choline kinase activity The initial rate of choline transport at7 p~ choline is from Fig. 4. To determine the initial rate of transport at36 pM choline, quiescent cells or cells first exposed to 10% serum for 6 h were washed twice with choline-free MEM and thenfurther incubated in MEM contain. 60 s, cells were washed ing [14C]choline (6.9 Ci/mol, 36 p ~ ) After and trichloroacetic acid extractions were prepared as described (Warden and Friedkin, 1984). Three independent samples were used for both conditions. Intracellular choline kinase activity a t 7 or 36 p~ choline was calculated for quiescent cells and for cells pre-exposed to DMEM to 5 pg/ml insulin, or to 10% fetal bovine serum for 6 h ath 37 “ C .Cells were exposed to [3H]choline (7 NM, 60 Ci/mol) or to [“C] choline (36 pM, various specific activities) for 60 min at 37 “ c , extracted, and analyzed for determination of intracellular choline kinase activity as described under “Experimental Procedures.” Statistical significance was determined by comparison to quiescent cells. 2840 experiments The data presented are the average of five independent at 7 p~ choline and of two independent experiments at 36 p~ choline, with duplicate determinations in each experiment. 7 p M choline

36 pM choline

TABLEI1 Half-life, turnover time,and turnover rate for phosphocholine The half-life of radioactivity in phosphocholine was calculated from the experiment shown in Fig. 5 and from a duplicate experiment. Calculated half-lifes from each experiment are shown in parentheses, beneath the average values. Turnover timeswere calculated from the half-lifes, as described by Zilversmit (1960). The turnover rate is the phosphocholine pool size (Table 111 of Warden and Friedkin, 1984) divided bv the turnover time. Half-life of Phosphocholine phosphocholine turnover time

pmolf mg cell proteinfh

h

DMEM Insulin 1.7Serum

1.94 (2.2, 1.7) 1.3 (1.7, 0.91) 1.2 (1.3, 1.2)

Turnover rate

2.8

1220

2.4

1300

while phosphocholine concentrations increase (Warden and Friedkin, 1984). Second, phosphocholine turnover was inInitial rate of choline kinase creased (Fig 5 and Table 11), in part because of increased transport activity transport activity phosphocholine concentrations. Third, intracellular choline ( n = 5) ( n = 4) kinase activity was increased (Table I) as one would expect if pmolfmg cellproteinfmin k S.D. choline kinase activity were regulated i n vivo by serum. 250 f 30 71 f 14 99 f 11 43 k 5.5 Quiescent Choline kinase can regulate PC synthesis by modulating ND“ 35 f 7.2 ND 62 f 7.0 DMEM phosphocholine concentrations. Increased concentrations of 94 f 12 ND ND 48 f 6.6 Insulin phosphocholine could increase PC synthesis by two mechaSerum 220k8.0 160 f 20b 81 f 10 9 5 f 13’ nisms. First, increased phosphocholine could raise cytidylylND, not done. transferase activity if the initial phosphocholine concentrab p < 0.001. tion in quiescent 3T3 cells is below or near the intracellular K, of cytidylyltransferase for phosphocholine. Rates of enDISCUSSION zyme reactions in vivo may be different than rates predicted Choline kinase is clearly a regulated enzyme in 3T3 cells. with Michaelis-Menten kinetics from i n vitro measurements Previous studiesdemonstrated that addition of serum to of kinetic constants (Wohlhueter and Plagemann, 1981). quiescent cultures of 3T3 fibroblasts resulted in enhanced Thus, comparison of intracellular phosphocholine concentracholine phosphorylation, phosphocholine pools, i n vitro cho- tions with i n vitro determined K,,, values of cytidylyltransferline kinase activity, and PC synthesis (Warden et al., 1980; ase for phosphocholine cannot be utilized to prove that Warden and Friedkin, 1984). The current studies strengthen changes of intracellular phosphocholine are regulatory for the hypothesis that choline kinase is regulatory for PC bio- cytidylyltransferase; proof that increased phosphocholine insynthesis in 3T3 cells. First, in serum-stimulated cells, choline creases PC synthesis would require that one have an experikinase was at a crossover point for the intermediates of PC mental method for systematically changing intracellular consynthesis, since choline concentrations decrease (Figs. 1-3) centrations of phosphocholine. Nevertheless, it is useful to

Intracellular choline Condition

Intracellular kinase

6010

Regulation of Choline Kinase Activity

1-3) is the only pool of free choline in 3T3 cells. In contrast, Sundler et al. (1972) have hypothesized that there are two choline pools in rat liver in vivo, a small one which rapidly equilibrates with extracellular choline and a 6-&fold larger pool which does not equilibrate. Sunder’s two-choline-pool hypothesis does not change our conclusions for two reasons. First, Sundler et al. (1972) did not actually prove that there are two choline pools in rat liver; rather this hypothesis is just one of several explanations which are consistent with their labeling data. Second, we have previously shown that intracellular phosphocholine-specificactivity is identical with the specific activity of extracellular choline after 60 min of exposure to labeled choline (Warden and Friedkin, 1984). Thus, there must be only one active pool of choline which acts as a substrate for choline kinase, and the presence or absence of a second inactive pool of free choline would not FIG. 6. Appearance of labeled choline in culture medium matter for labeling or metabolic studies in 3T3cells. Plasma choline concentrations may be important in vivo during a chase. Data are derived from the experiment described for Fig. 5. Mediumwas taken for scintillation counting a t each time regulators of intracellular choline kinase activity and PC point during the chase period after removal of medium containing synthesis. In Figs. 1 and 3 and Table I it was shown that [“C]choline. Duplicate samples were taken for each time point. Time intracellular choline kinase activity was increased approxiisgiven as minutes after removal of [“C]choline. Thedataare mately 2-fold when [14C]choline concentrations were raised expressed as nanomoles of radioactivity ([“C)choline and labeled products of the choline kinase reaction) released into the medium/ from 7 to 36 pM. Since plasma choline concentrations range from 5 pM in rats to 16pM in humans (Wang andHaubrich, milligram of cell protein. 1975), then an increase of plasma choline concentrations could raise intracellular chojine kinase activity. It is noteworcompare intracellular phosphocholine concentrations with in vitro determined K,,, values of cytidylyltransferase for phos- thy that both Sundler and Akesson (1975)and Pritchard and phocholine since if they areapproximately the same then this Vance (1981)have reported that increased choline concentrawould be consistent with the hypothesis that modulations of tions in culturemedia enhance the rateof PC synthesisin rat phosphocholine can regulate cytidylyltransferase activity. hepatocytes. Moreover, intracellular choline concentrations While it is difficult to determine the exact intracellular con- at 7 ~ L Mexternal choline, estimated as described above for centration of phosphocholine, it is possible to make a rough intracellular phosphocholine, were decreased from approxiestimate by assuming that protein accounts for 10% of the mately 80 PM in quiescent cells to approximately 25 p~ in wet weight of a cell and that 80% of the wet weight is water. serum-stimulated cells; thus, it is clear that intracellular The estimated concentrationsof phosphocholine are 240,410, choline concentrations are notthe same as extracellular cho270, and 630 pM in quiescent cells and in cells treated with line concentration and arealso regulated. Phosphocholine turnover rates provide the best single inDMEM, insulin, or serum, respectively (Warden and Frieddication of the comparability of our experimental values with kin, 1984). The reported K , of cytidylyltransferase for phosdependent phocholine, under optimal conditions in vitro, ranges from previously reported work, as turnover rates are 167 to 1000 p~ (Ansell and Chojnacki, 1969; Choyet al., 1977; upontwo independent experiments which determine first, Feldman et al., 1978). Thus, increased phosphocholine con- phosphocholine pool size (Warden and Friedkin, 1984), and centrations might increase PC synthesis. Second, cytidylyl- second, half-life of radioactivity in phosphocholine (Fig. 5 and transferase is a reversible enzyme; under some conditions it Table 11). In a study of PC biosynthesis in HeLa cells Vance catalyzes thenet formation of CTP and phosphocholine et al. (1980) reported phosphocholine pools of 1820 nmol/g of (Feldman et al., 1978). Thus, increased phosphocholine could cells, or approximately 18 nmol/mg of protein, and ahalf-life increase the rate of PC synthesis by shifting the equilibrium for phosphocholine of2-6 h. His calculated turnover rates of the reversible reaction catalyzed by cytidylyltransferase translate to 1200-2000 pmol/mg of protein/h for mock-intoward the product, CDP-choline. In fact, serum increased fected cells and 4200-6400 pmol/mg of protein/h for poliothe ratio of phosphocholine to CDP-choline by 2-fold (War- virus-infected cells. Our results, which range from 1220pmol/ den and Friedkin, 1984).These results are consistent with the mg of cell protein/h for control cells in DMEM up to 2840 pmol/mg of cell protein/h for cells exposed to serum (Table data in Table I1 in which the phosphochline turnover rate was 2.5-fold higher in serum-stimulated cells than in control 11) are clearly comparable with those of Vance’s group. In thecurrent work, we have focused onthe effects of serum in serum-stimulated cells than in control cells. Nishijima et al. (1984) have found new evidence which and insulin upon choline phosphorylation and PC synthesis; suggests that choline kinase is a regulatory enzyme for PC however, in previous work (Warden and Friedkin, 1984), we synthesis. They found a mutant Chinese hamster ovary cell showed that insulin-like growth factors I and 11, epidermal line with a 3-fold reduction of choline kinase activity and a and fibroblast growth factors, vasopressin, and 12-0-tetradecorresponding %fold reduction of phosphorylcholine, CDP- canoyl phorbol13-acetate all increase uptake into acid-insolcholine, and PCsynthesis. These resultsstrongly suggest that uble material, apparently by activating different parts of the choline kinase is regulatory for PC synthesis in Chinese pathway for PC biosynthesis. The combination of insulin plus 12-0-tetradecanoyl phorbol 13-acetate was particularly pohamster ovary cells. Our measurements of intracellular choline concentrations tent, stimulating uptake intoacid-insoluble material to levels implicitly assumed that there is only one pool of free choline twice as high as seen in serum-stimulated cells. These results in 3T3 cells since we assumed that the intracellular free suggest that purified growth factors, and especially combinacholine pool which equilibrates with extracellular labeled tions of Durified erowth factors,. may- doeverything- that serum choline by 5 min after exposure to radioactive choline (Figs. does: increase choline kinase activity, phosphocholine pools, I

Regulation of Choline Kinase Activity and PC synthesis. In the future, it should be possible to determine which mitogens affect which parts of the PC biosynthetic pathway. REFERENCES Ansell, G . B., and Chojnacki, T.(1969) Methods Enzymol. 1 4 , 121125 Choy, P. C., Lim, P. H., and Vance, D. E. (1977) J. Biol. Chem. 2 5 2 , 7673-7677 Chov. P. C.. Paddon. H. B.. and Vance. D. E. (1980) J. Biol. Chem. 255,1070-1073 ’ Feldman. D. A.. Kovac. C. R.. Draneinis. P. L.. and Weinhold. P. A. (1978) ’J. Bioi. hem.' 253,4980-40986 ’ Kennedy, E. P., and Weiss, S. B. (1956) J. Biol. Chem. 222,193-214 Lim, P. H., Pritchard, P. H., Paddon, H. B., and Vance, D. E. (1983) Biochim. Biophys. Acta 7 5 3 , 74-82 Lowry, 0.H.,Rosebrough, N. J., Farr, A. L., andRandal1, R. J. (1951) J. Biol. Chem. 193,265-275 Nishijima, M., Kuge, O., Maeda, M., Nakano, A., and Akamatsu, Y. (1984) J. Biol. Chem. 259,7101-7108 Paddon, H. B., Vigo, C., and Vance, D. E. (1982) Biochim. Biophys. Acta 710,112-115

6011

Pelech, S. L., and Vance, D. E. (1982) J. Biol. Chem. 2 5 7 , 1419814202 Pritchard, P. H., and Vance, D. E. (1981) Biochem. J. 196,261-267 Rolleston, F. S. (1972) Curr. Top. Cell. Regul. 5,47-75 Rozengurt, E., Mierzejewski, K.,and Wiggelsworth, N. (1978) J. Cell. Physiol. 97,241-252 Sundler, R., and Akesson, B. (1975) J. Biol. Chem. 250,3359-3367 Sundler, R., Arvidson, G., and Akesson, B. (1972) Biochim. Biophys. Acta 280,559-568 Vance, D. E., and Choy, P. C. (1979) Trends Biochem. Sci. 4 , 145148 Vance, D. E., Trip, E. M., and Paddon, H. B. (1980) J. Biol. Chem. 255, 1064-1069 Vigo, C., and Vance, D. E.(1981) Biochem. J. 2 0 0 , 321-326 Wang, F. L., and Haubrich, D. R. (1975) Anal. Biochem. 63,195-201 Warden, C. H., and Friedkin, M. (1984) Biochim. Biophys.Acta 7 9 2 , 270-280 Warden, C. H., Friedkin, M., and Geiger, P. J. (1980) Biochem. Biophys. Res. Commun. 9 4 , 690-696 Wohlhueter, R. M., and Plagemann, P. G . W. (1981) J. Biol. Chem. 2 6 6 , 869-875 Zilversmit, D. B. (1960) Am. J. Med. 29,832-848