Cell, Vol. 61, 1113-l

120, June

15, 1990, Copyright

0 1990 by Cell Press

Phospholipase C-Mediated Hydrolysis of Phosphatidlycholine Is an Important Step in PDGFStimulated DNA Synthesis Pilar Larrodera,’ Maria E. Cornet, Maria T. Diaz-Meco,’ Monica Lopez-Barahona, In& Diaz-Laviada,’ Per Henrik Guddal,t Terje Johansen,t and Jorge Moscat’ * Medicina y Cirugia Experimental Hospital General “Gregorio Maration” Dr. Esquerdo 46 28007 Madrid Spain t Institute of Medical Biology University of Tromso 9001 Tromso Norway

Summary Recent evidence suggests the involvement of phosphatidylcholine (PC) hydrolysis both in the control of normal cell growth and in transformation. We show here that the simple exogenous addition of Bacillus cereus PC-hydrolyzing phospholipase C (PC-PLC) is sufficient to elicit a potent mitogenic response in Swiss 3T3 fibroblasts by a mechanism that is independent of protein kinase C. Our results on the additivity and synergism between 6. cereus PC-PLC, PDGF, and insulin in the mitogenic response indicate that this novel phosphollpid degradative pathway may be important in the mitogenic signaling cascade activated by PDGF. Introduction Most of the studies on the role of phospholipid metabolism in the control of cell proliferation have been focused on the phosphoinositide (PI) turnover. This pathway is initiated by the hydrolysis by PI-phospholipase C (PLC) of phosphatidylinositol 4,5-bisphosphate to generate at least two second messengers (Berridge, 1987): inositol 1,4,5trisphosphate and diacylglycerol (DAG). The former participates, in collaboration with its metabolic product (inositol 1,3,4,5tetrakisphosphate), in the control of intracellular Ca2+ levels (Berridge and Irvine, 1989), whereas the latter is an important activator of protein kinase C (PKC) (Nishizuka, 1984). This kinase has been shown to be a key enzyme in mitogenic signaling (Nishizuka, 1986). Thus, overexpression of PKC in fibroblastic cell lines confers to these cells a reduced dependence of serum and enhanced tumorigenicity (Housey et al., 1988; Persons et al., 1988). Therefore, DAG can be considered an important intermediate in signal transduction pathways regulating cell growth and transformation. A number of reports have suggested the existence of PIindependent signal transduction pathways involving the hydrolysis of phosphatidylcholine (PC), which may be an important source of DAG (Fisher and Mueller, 1968; Mufson et al., 1981; Grove and Schimmel, 1982; Guy and Mur-

ray, 1982). Later studies have confirmed an increasing number of agonists as potent stimulants of a PC-hydrolyzing PLC activity (Bocckino et al., 1985; Besterman et al., 1986; Irving and Exton, 1987; Rosoff et al., 1988; Slivka et al., 1988; Diaz-Meco et al., 1989). The involvement of this novel source of DAG in transformation is suggested by the fact that certain oncogene products (like Ha-ras ~21) are potent stimulants of this phospholipid degradative pathway (Lacal et al., 1987; Moscat et al., 1989a; Price et al., 1989). The fact that protooncogenes, which give rise to oncogenes, are critically involved in mitogenic signal transduction explains the ability of transformed cell lines to proliferate in the absence of mitogens (Weissman and Aaronson, 1985; Falco et al., 1988) and suggests the possible involvement of PLCmediated degradation of PC in the control of the mitogenic response to growth factors. We demonstrate here that a reproducible but delayed increase in PC-PLC activity-measured as increases in intracellular phosphocholine (PCho) levels-occurs following stimulation of quiescent Swiss 3T3 fibroblasts by platelet-derived growth factor (PDGF). More importantly, our results strongly indicate that the hydrolysis of PC may be sufficient to account for a substantial part of the mitogenie activity of PDGF Furthermore, the ability of PC-PLC to induce DNA synthesis in Swiss 3T3 cells is independent of the presence of PKC. Results Activation of PLC-Mediated Hydrolysis of PC by PDGF To examine the effect of PDGF addition to quiescent fibroblasts on PLC-mediated hydrolysis of PC, Swiss 3T3 fibroblasts were labeled with [14C]choline and stimulated with PDGF (10 nglml) for different times. Under these conditions, the release of PCho into the intracellular milieu was potently stimulated. Results shown in Figure 1A indicate that a significant increase in this parameter was detected by 4 hr and that maximal PCho release was observed 8 hr after agonist addition. Concomitant with the increase in intracellular PCho, a significant elevation of DAG levels was detected in [U-14C]glycerol-labeled Swiss 3T3 fibroblasts (Figure lB), which strongly suggested that the rise in PCho was most probably due to the PLC-catalyzed hydrolysis of PC. Another potential source of PCho and DAG could be the phospholipase D (PLD)-mediated degradation of PC, which would give rise to phosphatidic acid (PA) and choline. Dephosphorylation of PA by phosphatidate phosphohydrolase and a choline kinase-mediated phosphorylation of choline could theoretically account for the increases in DAG and PCho levels observed after PDGF challenge. Although from our results this possibility cannot be completely ruled out, it appears to be unlikely since no changes were found either in PA or in choline levels after PDGF addition to cell cultures. Furthermore, no alterations were

Cl?ll 1114

Table 1. Synergistic Stimulation Release by PDGF and Insulin

Treatment None Insulin (10 &ml) PDGF (1 nglml) PDGF (1 @ml)+ Insulin (10 @ml)

01.

' 2

0

0

2

' 4

4

' . 6 8 Timz(tr)

6

8

Time

[hr]

' IO

IO

' 12

12

31 14

14

I El0

(c)

60 40 20 1000 :.ii.i;il 0 2

Figure 1. DNA Synthesis, Fibroblasts in Response

4

6 8 IO 12 14 16 18 Time (hr)

and PCho and DAG Production to PDGF

in Swiss 3T3

(A) [Methyl-%]choline-labeled quiescent Swiss 3T3 fibroblasts were either untreated (8) or stimulated with 10 nglml PDGF (A) for different times, after which intracellular PCho levels were determined as described under Experimental Procedures. (6) [U-%]glycerol-labeled Swiss 3T3 fibroblasts were either untreated (m) or stimulated with 10 nglml PDGF (A) for different times, after which DAG levels were determined as described under Experimental Procedures. (C) Quiescent cells were stimulated with 10 nglml PDGF for different times and de novo DNA synthesis was measured by determination of 13H)thymidine incorporation, as described. Control value was 7200 f 250 dpm/well and the maximal response was 150,000 -c 14,000 dpm/well. Results are the mean + SD of three independent experiments with incubations in duplicate.

detected in extracellular PCho or choline levels after the PDGF challenge (J. M., P. L., C. J. Molloy, T Fleming, and S. A. Aaronson, unpublished data). PDGF-stimulated production of PCho cannot be accounted for by the action of choline kinase on basal cho-

of DNA Synthesis

and PCho

DNA Synthesis (DPM x 10-3)

PCho Levels (DPM x Ws)

722 10 f 3 150 f 12 250 * 15

2 2 9.8 10.0

f 0.2 * 0.5 * 0.8 * 1.8

Quiescent 3T3 fibroblasts were stimulated with PDGF, insulin, or PDGF plus insulin. DNA synthesis was measured after 24 hr of stimulation, and PCho release was determined after 12 hr. Results are the mean f SD of three independent experiments with incubations in duplicate.

line, because if this were the case a concomitant decrease of choline should parallel the rise of PCho after PDGF stimulation. Such a decrease in choline levels was not observed in our experiments (data not shown). Therefore, all of these results strongly suggest that PCho production most probably derived from PLC-catalyzed breakdown of PC. To support this notion further, we also measured PCho synthesis in Swiss 3T3 fibroblasts either untreated or treated for different times with PDGF (10 nglml) by labeling cells with [methyl-14C]choline for 30 min prior to termination of the reactions. It is noteworthy that by using this experimental approach, the choline-containing phospholipid pool was not significantly labeled and, more importantly, no changes were detected in PCho labeling in PDGF-treated Swiss 3T3 fibroblasts as compared with controls (data not shown). Although PCho release in response to PDGF is a relatively late event, it takes place before any detectable increase in DNA synthesis. Significant 13H]thymidine incorporation was observed by 12 hr, and maximal values were achieved between 16 and 18 hr after PDGF challenge (Figure 1C). A number of experiments indicate that PCho release is specific to the action of certain growth factors and not simply related to the magnitude of the mitogenic response. Thus, insulin, which by itself is unable to promote DNA synthesis in fibroblasts, greatly potentiates the mitogenic effect of PDGF (Falco et al., 1988). Our results (Table 1) show that although insulin working via the IGF-1 receptor (Falco et al., 1988) synergistically enhances the mitogenic activity of PDGF, it neither activates PCho release nor modulates the ability of PDGF to stimulate this pathway. Therefore, these results demonstrate that PC-PLC activation is restricted to the action of PDGF, is independent of the mechanism activated by insulin, and is not simply related to the magnitude of the mitogenic response. Exogenous Addition of B. cereus PC-PLC Promotes DNA Synthesis and Cell Proliferation Since the activation of a PC-hydrolyzing PLC is observed in PDGFstimulated (see above) and oncogenetransformed (Lacal et al., 1987; Price et al., 1989) cells, it was conceivable that this enzymatic activity would be able to activate

;:v;phatidylcholine-Phospholipase

0 I 0 rX h 0

C Signaling

12, ii, E IO. :. 1 \t ---------f -----_---___---_ + .: : .d-* 0.0 5” 10 15” al 25” M 35” 40 4s* 50’ 553 60’ PC-PLC [U/ml]

Tim? (min) Figure 2. Time Course B. cereus PC-PLC

of PCho

Release

in Response

to Addition

of

[Methyl-‘4C]choline-labeled quiescent Swiss 3T3 fibroblasts were either untreated (81) or treated with 1 U/ml 6. cereus PC-specific PLC (A), and reactions were terminated at different times. Intracellular PCho release was then determined as described. Results are the mean + SD of three independent experiments with incubations in duplrcate.

DNA synthesis by itself. In the following experiments we examined whether the exogenous addition of PC-PLC was sufficient to elicit a mitogenic response in serum-starved Swiss 3T3 fibroblasts. For this purpose, we used a highly purified PC-hydrolyzing PLC from B. cereus that has been characterized extensively (Johansen et al., 1988; Little, 1988). Results shown in Figure 2 clearly indicate that the addition of 1 U/ml 8. cereus PC-PLC to quiescent [14C]choline-labeled fibroblast cultures activated the prompt hydrolysis of PC. This occurred without any detectable effect on inositol-containing phospholipids or sphingomyelin and with little or no change in the level of phosphatidylethanolamine or phosphatidylserine (Table 2). However, the hydrolysis of PC was concomitant with a dramatic increase in DAG levels (data not shown). One unit of B. cereus PC-PLC activated PC turnover to an extent similar to that produced by a saturating concentration of PDGF (10 nglml; compare Figure 1A with Figure 2). Interestingly, this concentration of enzyme was able to promote DNA synthesis (Figure 3A) at a magnitude 80% of that induced by a maximal dose (10 rig/ml) of PDGF. The mitogenic activity of exogenous B. cereus PC-PLC correlated excellently with its PC-hydrolyzing activity (Figures 3A and 38). As a further proof of the specificity of this effect, the following experiment was carried out. B. cereus PC-PLC was added to quiescent cell cultures either alone

Table 2. Specificity

of 6. cereus

None 8. cereus

PC-PLC

(1 U/ml)

001

01

I

IO

30

50

PC-PLC [U/ml]

Figure 3. Dose Response of the Effect Synthesis and PCho Release

of 8. cereus

PC-PLC

on DNA

(A) Quiescent cells were stimulated with different concentrations of B. cereus PC-PLC, and DNA synthesis was determined as described under Experimental Procedures. Control value was 7100 f 500 dpm/well and maximal stimulation was 92,000 f 300 dpm/well. (B) [Methyl-14C~cholinslabeled quiescent cells were treated with different concentrations of B. cereus PC-PLC for 5 min, after which PCho release was determined as described. Results are the mean k SD of three independent experiments with incubations in duplicate.

or in the presence of a potent neutralizing affinity-purified anti-B. cereus PC-PLC antibody. Results shown in Table 3 indicate that PLC-catalyzed PCho release was completely inhibited in the presence of the neutralizing antibody. Concomitant with the inhibition of the enzymatic activity by the antibody, a complete abolition of the ability of B. cereus PC-PLC to induce DNA synthesis was observed. These data support the specificity of the mitogenic effect of PC-PLC.

PC-PLC Phospholipid

Addition

0

Levels

(Percent

of Control)

PC

PE

PS

SM

PIPS

100 75 + 3

100 93 + 2

100 9s + 2

100 100 + 3

100

100 + 5

Swiss 3T3 fibroblasts were labeled with different precursors as described under Experimental Procedures. Cells were treated with 1 U/ml B. cereus PC-PLC for 30 min, reactions were stopped, and lipids were extracted and fractionated as described. Results are expressed as the percent of control values and are the mean f SD of three independent experiments performed in duplicate. Control values for different phospholiplds were as follows: PC, 549,000 dpm/well; SM (sphingomyelin), 39,000 dpm/well; PE (phosphatidyfethanolamine), 445,000 dpm/wetl; PS (phosphatidylserine), 93,000 dpm/well; PIPS (polyphosphoinositides), 150,000 dpm/well.

Cdl 1116

Table 3. Inhibition of B. cereus PC-PLC-Stimulated PCho Release and DNA Synthesis by a Neutralizing Anti-B. cereus PC-PLC Antibody Treatment None Antibody PC-PLC (1 U/ml) Antibody + PC-PLC (1 U/ml)

DNA Synthesis (DPM x 1O-3)

PCho Levels (DPM x lo+)

7-c2 6k2 80 f 9 8*3

2 2 14 3

f f f f

0.2 0.3 0.8 0.2

Swiss 3T3 fibroblasts were stimulated with 1 U/ml B. cereus PC-PLC in either the absence or presence of 25 ug/ml neutralizing antibody, and DNA synthesis and PCho levels were determined as described. Results are the mean f SD of three independent experiments with incubations in duplicate.

To investigate whether the capability of 6. cereus PCPLC to induce DNA synthesis correlated with its potential ability to activate cell proliferation, the following experiment was carried out. Cells were incubated for 8 days as described previously (Falco et al., 1988) either with B. cereus PC-PLC or with PDGF in the absence or presence of insulin (10 hg/ml). As shown in Table 4, both mitogens support the sustained growth of Swiss 3T3 fibroblasts in the presence of high concentrations of insulin. The Activation of DNA Synthesis by PC-PLC Occurs More Rapidly Than by PDGF The fact that activation of PC-PLC by PDGF occurs at least 4 hr after agonist addition indicates that PC hydrolysis is a late event in the mitogenic regulatory cascade triggered by PDGF. In keeping with this notion, if PC-PLC is an important element in the mitogenic pathway activated by PDGF, and taking into account that PC hydrolysis occurs immediately following the addition of 6. cereus PCPLC, this enzyme should not only induce DNA synthesis (as we demonstrated above) but should also do so faster than PDGF. This would indicate that all of the steps activated by PDGF prior to the hydrolysis of PC would be bypassed by the addition of the purified enzyme. The feasibility of this model was assessed as follows. Quies-

Table 4. B. cereus 3T3 Fibroblasts

PC-PLC

Induces

Insulin (10 whl)

Additions None PC-PLC PDGF

(1 U/ml) (10 nglml)

+ +

Proliferation

of Swiss

Fold Increase Cell Numbers 1.0 1.5 2.0 6.3 2.2 10.6

* f f f f

in

0.3 0.2 0.3 0.2 0.7

Swiss 3T3 fibroblasts were seeded (15,000 cells/well) in 24well plates with DMEM supplemented with transferrin and NaaSeOs containing the indicated stimulants. The medium was replaced with fresh medium every other day. Cells were counted on day 9 after seeding. Results are the mean f SD of three independent experiments with incubations in duplicate. The number of cells in control wells after 9 days in quiescent medium was 19,500 * 2,000 per well.

Time (hr) Figure 4. Time Course eus PC-PLC or PDGF

of DNA Synthesis

in Cells Stimulated

by B. cer-

Quiescent Swiss 3T3 fibroblasts were treated either with 1 U/ml PCPLC (m) or with 10 nglml PDGF (A) in the presence of 13H]thymidine (2 uCi/ml). Reactions were terminated at 3 hr intervals, and de nova DNA synthesis was determined as described. Results are the mean r SD of three independent experiments with incubations in duplicate.

cent Swiss 3T3 fibroblasts were stimulated with either PDGF (10 nglml) or 8. cereus PC-PLC (1 U/ml), and [3H]thymidine incorporation was determined every 3 hr after the addition of the respective stimuli. Results shown in Figure 4 demonstrate that B. cereus PC-PLC induces mitogenesis more rapidly than PDGF. Thus, whereas maximal [3H]thymidine incorporation was detected between 9 and 12 hr after the addition of B. cereus PC-PLC, the maximal response to PDGF was first observed 18 hr after PDGF challenge. Additivity and Synergistic Relationships of the Mitogenic Response Activated by PC-PLC, PDGF, and Insulin If the activation by PDGF of PLC-mediated degradation of PC is important in the mitogenic signaling cascade of this growth factor, the fact that both PDGF and B. cereus PCPLC activate the same phospholipid metabolic pathway suggests that mammalian PC-PLC and PDGF work through the same mitogenic signaling cascade. If this is true, the mitogenic response elicited by the simultaneous addition to cell cultures of maximal doses of PDGF (10 nglml) and B. cereus PC-PLC (1 U/ml) should not differ from the response produced by a maximal dose of PDGF alone. As shown in Figure 5A, this is actually the case. One possible explanation for the nonadditivity of maximal concentrations of PDGF and B. cereus PC-PLC could be that cells under these conditions are fully stimulated to DNA synthesis. Results from Figure 5A clearly indicate that this is not the case, since 10% serum is able to promote DNA synthesis to a higher level than that induced by the simultaneous addition of maximal doses of PDGF and B. cereus PC-PLC. On the other hand, it has been shown (see above) that high concentrations of insulin synergistically enhance the mitogenic properties of PDGF (Table 1). Because PC-PLC appears to be involved in the mitogenic signaling cascade of PDGF, insulin should be able to increase synergistically the [3H]thymidine incorporation induced by PC-PLC. This

~:;;phatidylcholine-Phospholipase

C Signaling

I

0.1

1

5

!ier

10 ,

46

PD.3 [rig/ml]

PMA Figure

0 01

0.1

05

Effects

1.0

of B. cereus

+

of PMA Treatment

on PKC Levels

Lysates containing 100 pg of total cellular protein were resolved by electroblotting as described in Experimental Procedures. PKC was visualized using a monoclonal anti-PKC antibody. Lane 1, control; lane 2, treated with PMA (500 @ml) for 24 hr. Molecular weight markers are shown on the right.

PC-PLC [U/ml] Figure 5. Synergistic and Nonadditive PDGF, and Insulin on DNA Synthesis

6. Effect

-

PC-PLC,

(A) Quiescent cells were treated with different concentrations of PDGF in the absence (open bars) or presence (stippled bars) of 1 U/ml B. cereus PC-PLC or were treated with 10% serum (black bar) and incubated with 2 t&i/ml 13H]thymidine. After 24 hr of incubation, reactions were stopped and DNA synthesis was determined. (B) Quiescent cells were treated with different concentrations of B. cereus PC-PLC in either the absence (open bars) or presence (stippled bars) of insulin (10 pg/ml). DNA synthesis under these conditions was determined as described above. Results are the mean * SD of three independent experiments with incubations in duplicate.

prediction is confirmed by the results in Figure 56. Thus, although insulin alone induces little or no change of the proliferative state of quiescent Swiss 3T3 fibroblasts, it greatly augments the mitogenic response elicited by the addition of different concentrations of B. cereus PC-PLC. The Mitogenic Response Activated by PC-PLC Is Independent of PKC DAG, which is a potent activator of PKC, is produced as a consequence of the PLC-mediated hydrolysis of PC. Therefore, PKC may be important for the mitogenic properties of PC-PLC. To address this possibility, PKC was down-regulated in Swiss 3T3 fibroblasts by treating cell cultures with phorbol myristate acetate (PMA) (500 nglml) for 24 hr. This treatment completely removed PKC from Swiss 3T3 fibroblasts, as determined by Western blot analysis with a specific monoclonal anti-PKC antibody (Figure 6). The antibody used recognizes the a subtype of PKC, which is the sole PKC isotype present in Swiss 3T3 fibroblasts (Rose-John et al., 1988). Interestingly, results shown

in Table 5 indicate that the mitogenic response produced by different concentrations of 6. cereus PC-PLC is not inhibited in cells with down-regulated PKC. This strongly suggests that PKC activation is not a required step in the mitogenic cascade triggered by PC-PLC. The fact that PMA was unable to promote DNA synthesis in down-regulated cells confirms the lack of any residual PKC in depleted cells (Table 5). Furthermore, when the phosphorylation of p80-a very well established substrate for PKC in Swiss 3T3 fibroblasts (Blackshear et al., 1985; Rodriguez-Pefia and Rozengurt, 1986)-was determined in response to PMA in PKC down-regulated cells, no response was found, although a potent activation of that parameter was observed in control cells (data not shown).

Table 5. Effect PC-PLC-Induced

of PKC Down-Regulation DNA Synthesis

Additions

PMA

None PC-PLC

(0.1 U/ml)

PC-PLC

(1 .O U/ml)

PMA (100 nglml)

on B. cereus

DNA Synthesis (DPM x 1O-3) 7k2 6+ 38 -c 31 * 82 + 76 f 48 + 622

1 3 5 8 7 6

Quiescent Swiss 3T3 fibroblasts were either untreated or treated with PMA (500 nglml) for 24 hr. Afterward, cells were stimulated with B. cereus PC-PLC or PMA, and DNA synthesis was determined. Results are the mean f SD of three independent experiments with incubations in duplicate.

Cdl 1118

A number of recent studies have unveiled the existence of a novel source of DAG in mammalian cells activated by different agonists (Besterman et al., 1988; Irving and Exton, 1987; Rosoff et al., 1988; Slivka et al., 1988; DiazMeco et al., 1989). This pathway is mediated by an as yet poorly characterized PC-hydrolyzing PLC activity. Since we and others have shown that the product of the ras oncogene, ras ~21, potently activates this novel phospholipid degradative mechanism (Lacal et al., 1987; Moscat et al., 1989a; Price et al., 1989), this enzymatic activity could be important in the control of cell growth and transformation. Further support for this notion comes from studies showing the activation of the phosphodiesterase degradation of PC in response to several growth factors (Besterman et al., 1988; Muir and Murray, 1987; Pessin and Raben, 1989). The results reported here indicate that following the addition of human recombinant B-homodimer PDGF to Swiss 3T3 fibroblasts, a delayed and dramatic elevation in intracellular PCho levels is observed prior to any detectable increase in DNA synthesis. In contrast to the results of Besterman et al. (1988) with 3T3 L-l cells stimulated by PDGF from an unspecified source, we did not detect any increase in intracellular PCho levels until 4 hr after PDGF challenge. The recent discovery of more than one PDGF receptor (Matsui et al., 1989) together with the fact that different cell lines were used in both studies could account for this discrepancy. On the other hand, as far as we know, the results shown here are the first report of long-term effects of a growth factor on PLC-mediated hydrolysis of PC. Several authors have demonstrated that besides PLC, a PLD specific for PC can also be responsible for the activation of PC turnover in cells stimulated by several agonists (Bocckino et al., 1987; Agwu et al., 1989; Billah et al., 1989a, 1989b; Cook and Wakelam, 1989). Although from our results the action of a PLD cannot be completely ruled out in PDGF-stimulated Swiss 3T3 fibroblasts, such a possibility seems unlikely. Thus, the fact that addition of B-homodimer PDGF to Swiss 3T3 fibroblasts activates the release of PCho and DAG without detectable change in choline or PA levels strongly supports the notion that PCPLC is the main route activated by B-homodimer PDGF in Swiss 3T3 fibroblasts. One attractive hypothesis could be that different agonists activate distinct mechanisms leading to PC turnover. Further work is necessary to evaluate a possible role for PC-PLD in mitogenic pathways. To examine the possible importance of PC-PLC activation in the mitogenic signaling cascade triggered by PDGF, we carried out a series of experiments aimed to investigate the potential mitogenic activity of a highly purified PC-hydrolyzing PLC from B. cereus. This enzyme has been characterized extensively (Johansen et al., 1988; Little, 1988), and we show that its addition to fibroblast cultures leads to the specific hydrolysis of PC. Our results also clearly demonstrate that the simple exogenous addition of this enzyme is able to induce a potent mitogenic response in quiescent Swiss 3T3 fibroblasts. Preliminary data obtained by an independent experimental strategy

confirm the mitogenic effect of B. cereus PC-PLC: overexpression of this enzyme in Swiss 3T3 cells stably transfected within a plasmid containing the PC-PLC gene from B. cereus, under the control of a dexamethasoneinducible enhancer/promoter, leads to a potent mitogenic response associated with the generation of PCho (T. J., M. T D.-M., and J. M., unpublished data). The magnitude of the mitogenic effect elicited by a maximal dose of 6. cereus PC-PLC is 80% of that produced by a saturating concentration (10 nglml) of PDGF. Smith et al. (1989) have recently demonstrated that microinjection of a PI-specific PLC into NIH 3T3 fibroblasts produces a mitogenic response similar to that described here by B. cereus PC-PLC. The fact that both PLCs cannot account for the full mitogenic potential of PDGF suggests that this growth factor may also be activating PLC-independent pathways that may be important for an optimal proliferative response. On the other hand, the time course of the effect of 8. cereus PC-PLC-activated DNA synthesis differs significantly from that produced by PI-PLC (Smith et al., 1989). Thus, whereas the peak of maximal mitogenic activity elicited by PI-PLC coincides with that produced by serum, the maximal DNA synthesis in response to B. cereus PC-PLC is detected -9 hr before the maximal response triggered by PDGF, which is considered to be the main mitogen in serum. Interestingly, the 9 hr interval between both maxima is very similar to the lag seen between the maximum of PDGF-induced PCho release and the activation, by this growth factor, of maximal DNA synthesis (compare Figure 4 with Figures 1A and 1C). Therefore, the fact that B. cereus PC-PLC mimics the mitogenic activity of PDGF in a time course that is in agreement with the kinetics of induction of PCho and DNA by this growth factor is consistent with a model whereby PC-PLC is important in the control of cell proliferation. Yet whereas PI-PLC activation is an early step in the mitogenic signaling cascade, PC-PLC activation is a late event in this pathway. Further support of this notion comes from our results concerning the additivity and synergism between B. cereus PC-PLC, PDGF, and insulin in the mitogenic response. Thus, the fact that 6. cereus PC-PLC mimics the ability of PDGF-induced DNA synthesis to be synergized by insulin indicates that PC-PLC is actually activating the same route as PDGF. Furthermore, the lack of additivity of the presence of a maximal dose of B. cereus PC-PLC on the mitogenic response elicited by a saturating concentration of PDGF strongly indicates that PLC-mediated PC hydrolysis is actually an important step in the PDGF mitogenic signaling cascade. The mechanism whereby B. cereus PC-PLC activates the release of intracellular PCho and induces DNA synthesis in Swiss 3T3 fibroblasts remains to be clarified. The use of the specific antibody described here, in immunofluorescence experiments underway in our laboratory, will help explain how the exogenous addition of this enzyme is able to mimic the increase in intracellular PCho levels triggered by PDGF. On the other hand, since DAG is a metabolic product of PC-PLC action, a feasible hypothesis to explain the mechanisms used by this enzyme to activate DNA synthesis should consider PKC activation as a

~:~;phatidylcholine-Phospholipase

C Signaling

possible link in the chain of events triggered by that stimulus. Our results, however, show that in cells in which PKC is completely down-regulated, B. cereus PC-PLC elicits a mitogenic response identical to that produced in fibroblasts with normal PKC levels. Therefore, although DAG is generated after the addition of B. cereus PC-PLC, PKC is not required for the mitogenic activity of this PLC. Of note is that recent reports have found that the generation of biological signals associated with DAG production is not always mediated by PKC (van Corven et al., 1989; Hockberger et al., 1989). Whatever the mechanism, the results presented here clearly demonstrate the importance of PDGF-activated PLC-mediated hydrolysis of PC in cell proliferation. Experimental

Procedures

Cell Cultures Swiss 3T3 fibroblasts (passage 123) were purchased from Flow Laboratories and cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% (v/v) fetal bovine serum, 100 U/ml penicillin, 100 pglml streptomycin, and 2 mM L-glutamine. Cells were grown in standard tissue culture flasks in a humidified air:COs (19:l) incubator at 37oC. Cells were made quiescent by incubation for 24 hr in the presence of serum-free medium supplemented with 5 pglml transferrin and 1 nM NasSeOs. The recombinant PDGF B-chain homodimer was from Amersham International and bovine insulin was from Collaborative Research (Waltham, MA). Isolation of PC-PLC from 6. cereus and Preparation of Affinity-Purified Antibody PC-PLC was isolated from cultures of B. cereus SE-1 essentially as described by Myrnes and Little (1980). Following the Johansen et al. (1988) protocol, the enzyme preparation was purified to complete homogeneity as confirmed by SDS-PAGE followed by silver staining. The specific activity of the purified enzyme was 1.5 Ulng (see Little, 1981). A rabbit antiserum was raised against this B. cereus PC-PLC by multiple intradermal injections with 75 pg of this enzyme. Serum was diluted 1:3 in phosphate-buffered saline (PBS) and applied to an Affigel lO(BioRad) column containing immobilized B. cereus PC-PLC. The column was washed with PBS, with PBS with increasing salt (up to 1 M NaCI), and with PBS containing 3 M urea before elution with 4 M urea, 0.5 M NaCl adjusted to pH 3.0 with acetic acid. The affinity purified antibody was eluted directly into 1 M glycine-NaOH [pH 10.51 and dialyzed extensively against a suitable buffer, The sole presence of heavy and light antibody chains in the final preparation was confirmed by SDS-PAGE followed by silver staining. Analysis of Products of PC Hydrolysis Cells were labeled for 48 hr with 2 pCi of [methyl-14C]choline (Amersham International; spec. radioactivity50-60 mCi/mmol)/dish. The last 24 hr of labeling was performed in serum-free medium supplemented as described above. Afterward, cells were treated with the corresponding agonists for different times. Reactions were stopped by removing the supernatants and adding ice-cold methanol to cells. Methanolic cell extracts were fractionated into chloroform and aqueous phases as previously described (Bligh and Dyer, 1959). The presence and levels of intra- and extracellular water-soluble choline metabolites were evaluated in the aqueous phases and extracellular medium, respectively, by thin-layer chromatography (Diaz-Meco et al., 1989) followed by autoradiography of plates in which standards corresponding to the different water-soluble choline metabolites were included. Analysis of DA0 and PA Release, and Levels of Other Phospholipids For determination of DAG production, cells were labeled with 10 nCi of [U-‘x]glycerol (spec. radioactivity 141 mCi/mmol)/dish as described above. For determination of the levels of different phospholipids, cells were labeled as follows: for PC and sphingomyelin, with 10 PCi of [methyl-‘4C]choline (spec. radioactivity 55 mCi/mmol)/well; for phosphatidylethanolamine, with 10 t&i of [2-r4C]ethan-l-olamine

(spec. radioactivity 55 mCi/mmol)/ml; for phosphatidylserine. with 10 t&i of L-[U-‘%]serine (spec. radioactivity 55 mCi/mmol)/well; for polyphosphoinositides, with 10 pCi of myo-[2-aH]inositol (spec. radioactivity 16.3 Ci/mmol)/well. Labeled compounds were obtained from Amersham International. Afterward, cells were treated with the corresponding agonists, according to the different experiments, and reactions were stopped by removing the supernatants and adding ice-cold methanol to cells. Methanol extracts were fractionated into chloroform and aqueous phases as described above. Organic phases were dried down under Ns, and lipids were fractionated by thin-layer chromatography using the following solvent systems. For the separation of DAG: hexane:diethylether:acetic acid (664O:l) (vol/volhrol) was used. For the fractionation of different phospholipids: chloroform:methanol:ammonia (65:25:4) (vol/vol/vol) was used in the first dimension and chloroform:acetone:methanoI:acetic acid:water (36:40:10:10:5) (vol/vol/vol/ volhrol) was used in the second dimension. For determination of PA production, the following protocol was carried out. Quiescent cells were incubated overnight with phosphate and serum-free culture medium supplemented with 10 f&i of [32P]orthophosphate (Du Pont-New England Nuclear; 9000 CVmmol). Afterward, cells were stimulated and phospholipids were extracted as above. PA levels were determined after separation by thin-layer chromatography with the upper phase of the following solvent system: ethyl acetate:trimethylpentane:acetic acid (90:50:20) (vol/vol/vol). Different lipids were visualized after autoradiography of plates where the corresponding standards were included. PH]Thymidine Incorporation Assays Quiescent cells were incubated with the corresponding stimulants in the presence of 13H]thymidine (2 @/ml) either for different times or for 24 hr according to the experiments. Afterward, de novo DNA synthesis was determined as previously described (Lea1 et al., 1985). Identification of PKC by lmmunoblotting Cell extracts containing 100 pg of total cell protein were obtained from cultures treated as described in Table 5. Following denaturation in SDS sample buffer, proteins were resolved in 10% SDS-PAGE and then transferred electrophoretically onto a polyvinylidene difluoride membrane (Immobilon, Millipore Water Systems, Bedford, MA). Tovisualize PKC, the membrane was incubated as previously described (Moscat et al., 1989b) using a monoclonal anti-PKC antibody (clone MC5, Amersham International). This antibody recognizes the a form of PKC, which is the sole subtype present in Swiss 3T3 fibroblasts (Rose-John et al., 1988). Estimation of PKC Activation by Analysis of Endogenous 80 kd Protein Phosphorylation Phosphorylation of endogenous proteins in response to PMA either in untreated cells or in cells chronically exposed to PMA was performed by two-dimensional gel electrophoresis following a slight modification of the method of Rodriguez-Pefia and Rozengurt (1986) as described previously (Moscat et al., 1989b). Acknowledgments We are greatly indebted to the late Professor Clive Little (University of Tromso) who originally suggested the use of B. cereus PC-PLC in this type of study. The help and advice of Eirik Bjorklid in the preparation of antibodies is gratefully acknowledged. This work was supported in part by Grant PB860590 from Comisidn Interministerial de Ciencia y Tecnologia. M. E. C., M. T D.-M., and M. L.-B. are Fellows from Comunidad de Madrid, Ministerio de Educacibn, and Universidad Complutense, respectively. T J. and R H. G. are postdoctoral and research fellows, respectively, of the Norwegian Research Council for Science and the Humanities. J. M. the is recipient of an Award from Fundaci6n Cientifica Asociaci6n Espaftola Contra el Cancer. The technical assistance of Maria Jestis Sanchez is greatly appreciated. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Received

November

15, 1989; revised

March

19, 1990.

Cdl 1120

References Agwu, D. E., McPhail, L. C., Chabot, M. C., Daniel, L. W., Wykle, R. L., and McCall, C. E. (1989). Choline-linked, a source of phosphatidic acid and diglycerides in stimulated neutrophils. J. Biol. Chem. 264, 14051413. Berridge, teracting

M. J. (1987). lnositol trisphosphate and diacylglycerol: two insecond messengers. Annu. Rev. Biochem. 56, 159-194.

Berridge, signalling.

M. J., and Irvine, R. F (1989). Nature 347, 197-205.

Leal, F, Williams, L. T., Robbins, K. C., and Aaronson, S. A. (1985). Evidence that the v&s gene product transforms by interaction with the receptor for platelet-derived growth factor. Science 230, 327-330. Little, C. (1981). Phospholipase 71, 725-730. Little, C. (1988).

C from Bacillus

Phospholipase

C. Biochem.

census.

Meth. Enzymol.

Sot. Trans.

17 271-273.

and cell

Matsui, T., Heidaran, M., Miki, T., Popescu, N., La Rochelle, W., Kraus, M., Pierce, J., and Aaronson, S. A. (1989). Isolation of a novel receptor for cDNA establishes the existence of two PDGF receptor genes, Science 243, 800-804.

Besterman, J. M., Duronio, V., and Cuatrecasas, P (1986). Rapid formation of diacylglycerol from phosphatidylcholine: a pathway for generation of a second messenger. Proc. Natl. Acad. Sci. USA 63, 67854789.

Moscat, J., Cornet, M. E., Diaz-Meco, M. T., Larrodera, P., LopezAlarion, D.. and Lopez-Barahona, M. (1989a). Activation of phosphatidylcholine-specific phospholipase C in cell growth and oncogene transformation. Biochem. Sot. Trans. 77, 988-991.

Billah, M. M., Pai, J. K., Mullman, T. J., Egan, R. W., and Siegel, M. I. (1989a). Regulation of phospholipase D in HL-60 granulocytes. J. Biol. Chem. 264, 9069-9078.

Moscat, J., Fleming, T. P, Molloy, C. J.. Lopez-Barahona, M., and Aaronson, S. A. (1989b). The calcium signal for BALB/MK keratinocyte terminal differentiation induces sustained alterations in phosphoinositide metabolism without detectable protein kinase C activation. J. Biol. Chem. 264, 11228-11235.

lnositol

phosphates

Billah, M. M., Eckel, S., Mullmann, T. J., Egan, R. W., and Siegel, M. I. (1989b). Phosphatidylcholine hydrolysis by phospholipase D determines phosphatidate and diglyceride levels in chemotactic peptidestimulated human neutrophils. J. Biol. Chem. 264, 17069-17077. Blackshear, F! J., Witters, L. A., Girard, P R.. Kuo, J. F., and Quamo, S. N. (1985). Growth factor-stimulated protein phosphorylation in 3T3 cells: evidence for protein kinase C-dependent and -independent pathways. J. Biol. Chem. 260, 13304-13315. Bligh, E., and Dyer, W. (1959). A rapid method for total lipid extraction and purification. Can. J. Biochem. Physiol. 37, 911-917. Bocckino, S. B., Blackmore, P F, and Exton, J. H. (1985). Stimulation of l,P-diacylglycerol accumulation in hepatocytes by vasopressin, epinephrine, and angiotensin II. J. Biol. Chem. 260, 14201-14207 Bocckino, S. 6.. Blackmore, f? F., Wilson, P W., and Exton, J. H. (1987). Phosphatidate accumulation in hormone-treated hepatocytes via phospholipase D mechanism. J. Biol. Chem. 262, 15309-15315. Cook, S. J., and Wakelam, M. J. 0. (1989). Analysis of the water-soluble products of phosphatidylcholine breakdown by ion-exchange chromatography. Biochem. J. 263, 581-587. Diaz-Meco, M. T, Larrodera, P, Lopez-Barahona. M., Cornet, M. E., Barreno, P G., and Moscat, J. (1989). Phospholipase C-mediated hydrolysis of phosphatidylcholine is activated by muscarinic agonists. Biochem. J. 263, 115-120. Falco, J. P., Taylor, W. G., DiFiore, P. P. Weissman, B. E., and Aaron-. son, S. A. (1988). Interactions of growth factors and retroviral oncogenes with mitogenic signal transduction pathways of BALB/MK keratinocytes. Cncogene 2, 573-578. Fisher, D. B., and Mueller, G. C. (1988). An early alteration in the phospholipid metabolism of lymphocytes by phytohemagglutinin. Proc. Natl. Acad. Sci. USA 60, 1396-1402. Grove, R. I., and Schimmel, S. D. (1982). Effects phorbol 13-acetate on glycerolipid metabolisms Biochim. Biophys. Acta 7l7, 272-280.

of 12-O-tetradecanoylin cultured myoblasts.

Guy, G. R.. and Murray, A. W. (1982). Tumor promoter stimulation of phosphatidylcholine turnover in HeLa cells. Cancer Res. 42, 19801985. Hockberger, P., Toselli, M., Swandulla, D., and Lux, H. D. (1989). A diacylglycerol analogue reduces neuronal calcium currents independently of protein kinase C activation. Nature 336, 340-342. Housey, G. M., Johnson, M D., Hsiao, W. L. W., O’Brian, C. A., Murphy, J. F’., Kirschmeier, f?, and Weinstein, I. B. (1988). Overproduction of protein kinase C causes disordered growth control in rat fibroblasts. Cell 52, 343-354. Irving, H. R., and Exton, J. H. (1987). Phosphatidylcholine in rat liver membranes. J. Biol. Chem. 262, 3440-3443.

breakdown

Johansen, T., Holm, T., Guddal, P. H., Sletten, K., Haugli, F. B., and Little, C. (1988). Cloning and sequencing of the gene encoding the phosphatidylcholine-preferring phospholipase C of Bacillus cereus. Gene 65, 293-304. Lacal, J. C., Moscat, J., and Aaronson, S. A. (1987). Novel source of 1,Bdiacylglycerol elevated in cells transformed by Hams oncogene. Nature 330, 269-272.

Mufson, R. A., Okin, E., and Weinstein, I. B. (1981). stimulate the rapid release of choline from prelabelled genesis 2, 1095-1102.

Phorbol esters cells. Carcino-

Muir, J. G.. and Murray, A. W. (1987). Bombesin and phorbol esters stimulate phosphatidylcholine hydrolysis by phospholipase C: evidence for a role of protein kinase C. J. Cell. Physiol. 730, 382-391. Myrnes, B. J., and Little, C. (1980). A simple purification scheme yielding crystalline phospholipase C from Bacillus cemus. Acta Chem. Stand. [B] 32, 375-377. Nishizuka, Y. (1984). The role of protein kinase C in cell surface transduction and tumor promotion. Nature 308, 693-698. Nishizuka, Y. (1986). Studies Science 233, 305-312.

and perspectives

of protein

signal

kinase

C.

Persons, D. A., Wilkison, W. 0.. Bell, R. M., and Finn, 0. J. (1988). Altered growth regulation and enhanced tumorigenicity of NIH 3T3 fibroblasts transfected with protein kinase C-l cDNA. Cell 52, 447-458. Pessin, M. S., and Raben, D. M. (1989). Molecular species of analysis of 1,2-diglycerides stimulated by a-thrombin in cultured fibroblasts. J. Biol. Chem. 264, 87298738. Price, B. D., Morris, J. D. H., Marshall, C. J., and Hall, A. (1989). Stimulation of phosphatidylcholine hydrolysis, diacylglycerol release, and arachidonic acid production by oncogenic ras is a consequence of protein kinase C activation. J. Biol. Chem. 264, 18838-16643. Rodriguez-Peria, A., and Rozengurt, E. (1986). Phosphorylation of an acidic mol. wt. 80,000 cellular protein in a cell-free system and intact Swiss 3T3 cells: a specific marker of protein kinase C activity. EMBO J. 5, 77-83. Rose-John, S., Dietrich, A., and Marks, F. (1988). Molecular cloning of mouse protein kinase C (PKC) cDNA from Swiss 3T3 fibroblasts. Gene 74, 485-471. Rosoff, I? M., Savage, N., and Dinarello, C. A. (1988). stimulates diacylglycerol production in T lymphocytes mechanism. Cell 54, 73-81.

Interleukin-I by a novel

Slivka, S. R., Meier, K. E., and Insel, P A. (1988). a, -adrenergic receptors promote phosphatidylcholine hydrolysis in MDCK-Dl cells. J. Biol. Chem. 263, 12242-12246. Smith, M. R., Ryu, S., Suh, F’.. Rhee, S., and Kung, H. (1989). S-phase induction and transformation of quiescent NIH 3T3 cells by microinjection of phospholipase C. Proc. Natl. Acad. Sci. USA 86, 3659-3663. van Corven, E. J., Groenink, A., Jalink, K., Eichholtz, T., and Moolenaar, W. H. (1989). Lysophosphatidate-induced cell proliferation: identification and dissection of signaling pathways mediated by G proteins. Cell 59, 45-54. Weissman, B., and Aaronson, S. oncogene families supplant the of BALBIMK-2 keratinocytes and ferentiation program. Mol. Cell.

A. (1985). Members of the src and ras epidermal growth factor requirement induce alterations in the terminal difBiol. 5. 3388-3396.