- Email: [email protected]
Transforming Growth Factor β Increases the Activity of Phosphatidate Phosphohydrolase-1 in Rat Hepatocytes
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
230, 365–369 (1997)
RC965965
Transforming Growth Factor b Increases the Activity of Phosphatidate Phosphohydrolase-1 in Rat Hepatocytes Mark C. Dixon, Stephen J. Yeaman, Loranne Agius,* and Christopher P. Day*,1 Departments of *Medicine and Biochemistry and Genetics, Medical School, University of Newcastle, Framlington Place Newcastle upon Tyne NE2 4HH, United Kingdom
Received November 25, 1996
Phosphatidic acid (PA) is a potent second messenger arising from growth factor-induced stimulation of phospholipase D which hydrolyses phosphatidylcholine. PA is hydrolysed to diacylglycerol by PA phosphohydrolase (PAP) which exists in two forms: PAP-1 and PAP-2. In rat hepatocyte cultures, overnight (20h) incubation with transforming growth factor (TGF) b (1ng/ml) increased PAP-1 activity two-fold. This effect was concentration and time dependent and was greatest at low cell density. The TGFb effect on PAP-1 was additive to stimulation induced by dexamethasone but not by glucagon and it reversed the inhibition by insulin. Epidermal growth factor had no effect on PAP-1 activity. None of the above hormones or growth factors affected the subcellular distribution of PAP-1. Stimulation of PAP-1 by TGFb may be involved in mediating some of its biological effects. q 1997 Academic Press
Hepatocyte proliferation is controlled by the action of both growth stimulatory and growth inhibitory effectors. Studies on hepatocyte cultures have identified four complete hepatic mitogens - hepatocyte growth factor (HGF), epidermal growth factor (EGF), transforming growth factor a (TGFa) and acidic fibroblast growth factor (aFGF), while transforming growth factor b (TGFb) appears to be the most potent inhibitor of hepatocyte proliferation [1]. The signal transduction pathways initiated by some of these mitogens include activation of the Ras/Raf/MEK/MAP kinase signal cascade [2] which causes activation of nuclear transcription factors such as those encoded by the proto-oncogenes c-fos and c-jun [3]. The precise mechanism(s) whereby TGFb inhibits mitogen-induced proliferation remains unclear, although in the case of EGF, it may involve inhibition of EGF-mediated Ras activation [4]. 1 To whom correspondence should be addressed. Fax: /44 91 222 0723.
It has been demonstrated that several growth factors in different cell types, including EGF in hepatocytes, activate phospholipase D (PLD) in the plasma membrane which catalyses the hydrolysis of phosphatidylcholine (PC) to generate phosphatidic acid (PA) [5]. PA is now emerging as an important mitogenic signal acting either via the Ras/Raf cascade [6-8] or through a specific PA-dependent protein kinase [9]. PA is hydrolysed by the enzyme PA phosphohydrolase (PAP) to produce diacylglycerol (DAG), which is also a key second messenger acting via stimulation of protein kinase C (PKC) [10]. Clearly, PAP occupies a critical position in the signal transduction cascade of growth factors since its activity will determine the balance between two important second messengers, PA and DAG. PAP activity exists in two enzymic forms; one is sensitive to inhibition by N-ethylmaleimide (NEM), is dependent on Mg2/ for its activity in vitro and is present predominantly in the free state in the cytosol (PAP-1), while the other is insensitive to NEM and Mg2/ and is present in the plasma membrane (PAP-2) [11,12]. PAP-2 appears to be the most appropriately located for a role in signal transduction, although PAP-1 is capable of translocating to the membrane compartment in response to a variety of stimuli [13] and may also be involved in cell signalling in addition to its established role in lipid synthesis[12]. The principal aim of this study was to examine the effects of EGF and TGFb, two growth factors with opposing effects on hepatocyte proliferation, on the activities and distribution of both forms of PAP in rat hepatocyte cultures. The results demonstrate that TGFb stimulates PAP-1 activity in a time and concentration dependent manner. MATERIALS AND METHODS Porcine TGFb1 was from British Bio-technology Products Ltd. Abingdon, Oxon, England. Sources of the other reagents were as described previously [14,15]. Hepatocytes were isolated by a collagenase perfusion of whole liver of male Wistar rats (150-200g) fed on 0006-291X/97 $25.00
365
Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 5965
/
6916$$$621
12-18-96 14:48:51
bbrcg
AP: BBRC
Vol. 230, No. 2, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
standard laboratory chow ad libitum [14]. Hepatocytes were suspended in Minimum Essential Medium and cultured in 24-well plates at cell densities of 41104 cells/cm2 (low density) or 81104 cells/cm2 (high density) [15]. The cells were allowed to attach for 4h at 377C in a humidified 95% air/5% CO2 atmosphere before the medium was replaced with serum-free medium. For studies on their short-term (õ6h) effects, TGFb and/or EGF were added to the media from 1100 final concentration stock solutions after overnight (14-16h) cell culture. For studies on the longer term effects, growth factors and hormones, were added to the culture medium immediately after the initial 4h attachment phase and the cells were cultured for a further 20h. All experimental conditions were in triplicate. Phosphatidate phosphohydrolase assay. After incubation for the required time, the medium was aspirated, the cells were washed twice in 150mM NaCl and then sonicated in 300ml of extraction buffer (0.25M sucrose, 200mM DTT, 50mM b-glycerophosphate (pH 7.4) and 0.1mM EDTA). PAP activities were estimated by measuring the release of [32P]Pi from [32P]PA as previously described [11], except that the two forms were distinguished by their in vitro dependence on Mg2/ rather than their sensitivity to N-ethylmaleimide (NEM). Final assay incubations contained 1mM EDTA { 3mM MgCl2 . This method allowed both forms of activity to be measured in the same extract, and removed the need to incubate the cell extracts with and without NEM at 377C for 15 mins [12] which may lead to changes in enzyme activities. Preliminary experiments showed that assaying PAP activity in the absence of Mg2/ ions gave identical results to assays in their presence after pre-incubating the enzyme fraction with NEM (NEM-insensitive activity 95{5% of total PAP activity measured in the absence of Mg2/). One Unit of PAP activity was defined as the release of one nanomole of Pi per minute and was expressed per mg of protein estimated by the method of Bradford [16]. PAP-1 translocation. After overnight (20h) incubation with dexamethasone (10nM) the culture medium was replaced with medium supplemented with 2% fatty acid free BSA and either 2mM oleic acid and/or growth factors/hormones as indicated. After 1 hour the medium was removed and the cells were washed twice in 150mM NaCl. Digitonin buffer (0.05mg/ml digitonin, 300mM sucrose, 3mM Hepes, 2mM DTT, pH 7.2) was then added for 8 min and the plates were shaken gently. The digitonin fraction was then removed and the remaining cell matrix was extracted by sonication in the extraction buffer described above. PAP activities were assayed in both fractions as described above. All results are expressed as the mean{SEM of the number of independent cell cultures indicated. Unless stated statistical analysis was by the Student’s paired t-test.
RESULTS Effect of cell density on PAP activities. In hepatocyte cultures the expression of certain enzymes is enhanced at low cell density whilst that of other enzymes is greater at high cell density [17]. Since PAP-1 activity is very low in cells cultured without glucocorticoids but is markedly induced by dexamethasone [12,18,19], the effect of cell density on PAP activities was determined with and without overnight (20h) incubation with dexamethasone. Dexamethasone (10nM) increased PAP-1 activity from 0.78{0.16 mU/mg to 6.88{1.12 mU/mg at low cell density and from 0.88{0.31 mU/mg to 6.79{0.60 mU/mg at high cell density (nÅ4 experiments; p ú 0.05 for difference between cell densities). Increasing the concentration of dexamethasone to 100nM did not further increase PAP-1 activity and nei-
FIG. 1. Effects of [TGFb] and incubation time on PAP-1 activity in rat hepatocyte cultures. Hepatocyte monolayers were incubated for 20h at a cell density of 41104 cells/cm2 with 10nM dexamethasone except where indicated. In B TGFb (1ng/ml) was added at the time indicated prior to stopping the incubations. No TGFb was added to cells at zero time. Results from three independent experiments are shown. The effects of dose (with and without dexamethasone) and time were significant (põ0.02, ANOVA).
ther concentration of dexamethasone had any effect on PAP-2 activity. The effect of TGFb and EGF on PAP activities. Long-term incubation of hepatocytes with TGFb increased the activity of PAP-1 in a concentration and time dependent manner (Figure 1). An increase in activity was observed after 20 hours incubation with 50pg/ml TGFb, with no further increase in effect above 200pg/ml. The magnitude of the TGFb effect was similar in the absence of dexamethasone although, as expected, the absolute activities were much lower (Figure
366
AID
BBRC 5965
/
6916$$$621
12-18-96 14:48:51
bbrcg
AP: BBRC
Vol. 230, No. 2, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 1
Effects of TGFb on PAP Activities and Cell Protein Content at High and Low Cell Density PAP-1 (mU/mg)
PAP-2 (mU/mg)
Protein (mg/well)
Density
Low
High
Low
High
Low
High
Control TGFb (1 ng/ml)
6.42 { 0.7 11.58 { 1.4
5.63 { 0.6 7.81 { 0.7
1.95 { 0.1 1.98 { 0.2
2.06 { 0.1 2.07 { 0.3
0.28 { 0.04 0.24 { 0.03
0.71 { 0.06 0.62 { 0.06
Note. All incubations were supplemented with 10 nM dexamethasone and were for 20 h. Each value represents the mean { SEM of 3 independent experiments.
1A). The effect of 1ng/ml TGFb was apparent at 2 hours and by 20 hours the activity had approximately doubled (Figure 1B). The effect of TGFb was cell densitydependent, with a greater stimulation observed at low compared to high cell density (Table 1). Incubation with TGFb produced no significant change in cell protein content and had no effect on PAP-2 activity. EGF (10nM) had no effect on either PAP activity and did not influence the effect of TGFb on PAP-1 activity (not shown). This concentration of EGF was chosen as it had previously been shown to be the optimum required for both its mitogenic and metabolic effects on rat hepatocytes [20]. Short-term (5-30min) incubation of hepatocytes with either TGFb or EGF had no effect on the activity of either form of PAP (not shown). To investigate the possible mechanisms of the increase in PAP-1 activity produced by TGFb, its effect was examined in the presence of glucagon and insulin which have been shown to stimulate and inhibit PAP activity respectively [18,19]. The stimulation of PAP-1 by TGFb was slightly but not significantly lower than with glucagon but the combined effect of TGFb and glucagon was lower than with glucagon alone (Figure 2A). In the presence of dexamethasone, insulin decreased PAP-1 activity, in both the absence and presence of glucagon (Figure 2B), however, TGFb overcame the insulin inhibition so that the activity in the presence of insulin and TGFb was not significantly different from the control (Figure 2C). The effect of TGFb and EGF on subcellular distribution of PAP. The effect of short-term incubation of hepatocytes with growth factors and hormones on the subcellular distribution of PAP-1 activity is shown in Figure 3. In control cells 60% of PAP-1 was in the soluble (cytosolic) fraction and 40% in a bound state (matrix fraction). Oleic acid increased the proportion of bound PAP-1 in agreement with previous studies[21]. However the growth factors and hormones had no effect. In addition overnight incubation with each of the growth factors and hormones had no effect on either basal PAP-1 distribution or on the translocation induced by oleic acid (results not shown). PAP-2 activity was confined to the matrix fraction under all conditions.
DISCUSSION The results of this study demonstrate: (i) TGFb stimulates the activity of PAP-1 in a dose and time-dependent manner that is greater at low than at high cell density; (ii) no hormone or growth factor tested influences subcellular distribution; (iii) effects of hormones on PAP activity are wholly restricted to PAP-1. As far as we are aware this is the first demonstration of any peptide growth factor affecting PAP activity. The time course of the TGFb-induced stimulation of PAP-1 activity is similar to that shown previously for dexamethasone, glucagon and cAMP analogues [18,19] and suggests an effect via increased enzyme synthesis or decreased degradation rather than acute activation via covalent modification or a G-protein linked mechanism. Unlike glucagon or insulin, TGFb has no effect on either basal or stimulated levels of cAMP [22], however, the TGFb-response elements in the promoter regions of several TGFb-regulated genes include elements that are also cAMP responsive [23] suggesting a common regulatory endpoint for the effect of TGFb and cAMP on PAP gene transcription. Further evidence for this possibility is provided by a recent study demonstrating that TGFb mimics the effects of cAMP-elevating agents on mRNA levels of the enzyme phosphoenolpyruvate carboxykinase [24]. Does this stimulation of PAP-1 by TGFb play any role in mediating the known biological effects of TGFb? First, concerning the established role of PAP-1 in glycerolipid synthesis, while there has been no direct demonstration of a TGFb effect on lipid metabolism [23], high TGFb levels following sub-total hepatectomy [1] may contribute to the observed increase in PAP activity and associated fatty liver [25,26]. Second, concerning the postulated role of PAP-1 in cell signalling, the stimulating effect of TGFb on PAP-1 could have potentially important effects on the pattern of lipid- derived second messengers generated in response to agonists that act via PLD, such as EGF and PDGF [5,27]. Therefore the ability of TGFb to inhibit EGF-stimulated hepatocyte proliferation might be due, at least in part, to an increase in PAP-1 activity and a subsequent reduction in the concentration of the mitogenic second messenger,
367
AID
BBRC 5965
/
6916$$$621
12-18-96 14:48:51
bbrcg
AP: BBRC
Vol. 230, No. 2, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
c-myc proto-oncogene [28] which is induced by PA via activation of the Ras/Raf/MEK/MAP kinase signal cascade [29,30]. The concentrations of TGFb required to increase PAP activity are of the same order as those required for inhibition by TGFb of EGF-induced hepatocyte proliferation [31], and the greater stimulation of PAP activity by TGFb at low compared to high cell density is also consistent with a role for this effect in growth-related functions [17]. A further important biological effect of TGFb is to increase the transcription of genes encoding extracellular matrix (ECM) proteins [23]. The recent demonstration, therefore, that DAG and PKC are involved in the signal cascade leading to ECM gene transcription [32] suggests another example whereby the stimulation of PAP-1 by TGFb (via an increase in DAG concentration) might be important in mediating its biological effects. Clearly, the involvement of PAP-1 in cell signalling requires that it participates in the dephosphorylation of PA formed by the PLD pathway at the plasma membrane. This would be facilitated by the translocation of PAP-1 from the cytosol to the plasma membrane either in response to PA accumulation, as has been shown for microsomal and mitochondrial membranes [33,34] or by growth-factor induced phosphorylation as occurs with EGF-induced translocation of PLC-g1 [35]. In this study, however, the lack of PAP-1 translocation in response to EGF, which would be expected to stimulate PLD [5] and increase PA at the plasma membrane, does not support this hypothesis. Translocation to the plasma membrane is not a prerequisite for involvement of PAP-1 in cell signalling if PA, once formed, is transported to the endoplasmic reticulum membranes for
FIG. 2. Effects of TGFb (1ng/ml), glucagon (100nM) and insulin (10nM) on PAP-1 activity. Hepatocyte monolayers were cultured for 20h at a cell density of 4 1 104 cells/cm2 with 10nM dexamethasone and additions as indicated. In A and B each point represents the mean { SEM of 3 independent experiments; in C, 4 experiments. *põ0.01 versus control, **põ0.05 versus glucagon alone, ***põ0.001 versus either insulin or TGFb alone.
PA. Since the mitogenic effects of PA probably operate via increased Ras/Raf activity [6,7], this effect of TGFb may explain its ability to block the EGF-mediated activation of Ras [4] and to decrease the expression of the
FIG. 3. Effects of growth factors and hormones on PAP-1 subcellular distribution. Hepatocyte monolayers were cultured for 20h at a cell density of 4 1 104 cells/cm2 with 10nM dexamethasone prior to the additions indicated (TGFb 1ng/ml, glucagon 100nM, insulin 10nM, EGF 10nM and oleic acid 2mM). PAP activities were assayed in both fractions as described in the methods section. Results represent means { SEM of 4 independent experiments and are expressed as % of total cellular PAP-1 activity (controls 10.8 { 0.9 mU/mg).
368
AID
BBRC 5965
/
6916$$$621
12-18-96 14:48:51
bbrcg
AP: BBRC
Vol. 230, No. 2, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
hydrolysis and resynthesis as previously suggested [36]. Undoubtedly confirmation of a signalling role for PAP-1 awaits the demonstration that changes in its activity can influence the PA/DAG ratio following PLD activation by extracellular agonists. The best evidence thus far in this regard comes from studies with sphingosine, the mitogenic effects of which have been shown to reflect its ability to increase levels of PA, at least in part via inhibition of PAP activity [37,38]. Studies with intact cells have demonstrated that this inhibition is wholly restricted to the membrane-associated form of PAP-1 with no effect on the cytosolic form or on PAP2 activity [39]. ACKNOWLEDGMENTS C.P.D is supported by a Medical Research Council Clinician Scientist Fellowship. M.C.D holds a postgraduate studentship from the Biotechnology and Biological Sciences Research Council.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.
Michalopoulos, G. K. (1990) FASEB J. 4, 176–187. Egan, S. E., and Weinberg, R. A. (1993) Nature 365, 781–783. Davis, R. J. (1993) J. Biol. Chem. 268, 14553–14556. Attisano, L., Wrana, J. L., Lo´pez-Casillas, F., and Massague´, J. (1994) Biochim. Biophys. Acta 1222, 71–80. Bocckino, S. B., Blackmore, P. F., Wilson, P. B., and Exton, J. H. (1987) J. Biol. Chem. 262, 15309–15315. Yu, C-L., Tsai, M-H., and Stacey, D. W. (1988) Cell 52, 63–71. Ghosh, S., Strum, J. C., Sciorra, V. A., Daniel, L., and Bell, R. M. (1996) J. Biol. Chem. 271, 8472–8480. Tsai, M-H., Yu, C-L., and Stacey, D. W. (1990) Science 250, 982– 985. Khan, W. A., Blobe, G. C., Richards, A. L., and Hannun, Y. A. (1994) J. Biol. Chem. 269, 9729–9735. Nishizuka, Y. (1984) Nature 308, 693–698. Day, C. P., and Yeaman, S. J. (1992) Biochim. Biophys. Acta. 1127, 87–94. Jamal, Z., Martin, A., Gomez-Munoz, A., and Brindley, D. N. (1991) J. Biol. Chem. 266, 2988–2996. Cascales, C., Mangiapane, E. H., and Brindley, D. N. (1984) Biochem. J. 219, 911–916. Agius, L., Peak, M., and Alberti, K. G. M. M. (1990) Biochem J. 266, 91–102.
15. Peak, M., and Agius, L. (1994) Eur. J. Biochem. 221, 529–536. 16. Bradford, M. M. (1976) Anal. Biochem. 72, 248–254. 17. Nakamura, T., Yoshimoto, K., Nakayama, Y., Tomita, Y., and Ichihara, A. (1983) Proc. Nat. Acad. Sci. (USA) 80, 7229–7233. 18. Pittner, R. A., Fears, R., and Brindley, D. N. (1985) Biochem. J. 230, 525–534. 19. Pittner, R. A., Fears, R., and Brindley, D. N. (1985) Biochem. J. 225, 455–462. 20. Chowdhury, M. H., and Agius, L. (1987) Biochem. J. 247, 307– 314. 21. Cascales, C., Mangiapane, E. H., and Brindley, D. N. (1984) Biochem. J. 219, 911–916. 22. Thoresen, G. H., Refsnes, M., and Christoffersen, T. (1992) Cancer Res. 52, 3598–3603. 23. Massague´, J., Cheifetz, S., Laiho, M., Ralph, D. A., Weiss, F. M. B., and Zentella, A. (1992) Cancer Surveys 12, 81–103. 24. Thoresen, G. H., and Christoffersen, T. (1994) Cell Biol. International 18, 171–175. 25. Mangiapane, E. H., Lloyd-Davies, K. A., and Brindley, D. N. (1973) Biochem J. 134, 103–112. 26. Tijburg, L. B. M., Gellen, M. J. H., and Van Golde, L. M. G. (1989) Biochim. Biophys. Acta 1004, 1–19. 27. Fukami, K., and Takenawa, T. (1992) J. Biol. Chem. 267, 10988– 10993. 28. Pietenpol, J. A., Holt, J. T., Stein, R. W., and Moses, H. L. (1990) Proc. Nat. Acad. Sci. (USA) 87, 3758–3762. 29. Moolenaar, W. H., Kruijer, W., Tilly, B. C., Verlaan, L., Bierman, A. J., and de Laat, S. W. (1986) Nature 323, 171–173. 30. Howe, L. R., and Marshall, C. J. (1993) J. Biol. Chem. 268, 20717–20720. 31. Baskin, G., Schenker, S., Frosto, T., and Henderson, G. (1991) J. Biol. Chem. 266, 13238–13242. 32. Casini, A., Galli, G., Salzano, R., Ceni, E., Franceschelli, F., Rotella, C. M., and Surrenti, C. (1994) Alcohol 29, 303–314. 33. Martin, A., Hopewell, R., Martin-Sanz, P., Morgan, J. E., and Brindley, D. N. (1986) Biochim. Biophys. Acta 876, 581–591. 34. Freeman, M., and Mangiapane, E. H. (1989) Biochem. J. 263, 589–595. 35. Todderud, G., Wahl, M., Rhee, S. G., and Carpenter, G. (1990) Science 24, 296–298. 36. Michell, R. (1975) Biochim. Biophys. Acta 415, 81–147. 37. Zhang, H., Desai, N. N., Murphey, J. M., and Spiegel, S. (1990) J. Biol. Chem. 265, 21309–21316. 38. Lavie, Y., Piterman, O., and Liscovitch, M. (1990) FEBS Lett. 277, 7–10. 39. Gomez-Munoz, A., Hatch, G. M., Martin, A., Jamal, Z., Vance, D. E., and Brindley, D. N. (1992) FEBS Lett. 301, 103–106.
369
AID
BBRC 5965
/
6916$$$621
12-18-96 14:48:51
bbrcg
AP: BBRC
230, 365–369 (1997)
RC965965
Transforming Growth Factor b Increases the Activity of Phosphatidate Phosphohydrolase-1 in Rat Hepatocytes Mark C. Dixon, Stephen J. Yeaman, Loranne Agius,* and Christopher P. Day*,1 Departments of *Medicine and Biochemistry and Genetics, Medical School, University of Newcastle, Framlington Place Newcastle upon Tyne NE2 4HH, United Kingdom
Received November 25, 1996
Phosphatidic acid (PA) is a potent second messenger arising from growth factor-induced stimulation of phospholipase D which hydrolyses phosphatidylcholine. PA is hydrolysed to diacylglycerol by PA phosphohydrolase (PAP) which exists in two forms: PAP-1 and PAP-2. In rat hepatocyte cultures, overnight (20h) incubation with transforming growth factor (TGF) b (1ng/ml) increased PAP-1 activity two-fold. This effect was concentration and time dependent and was greatest at low cell density. The TGFb effect on PAP-1 was additive to stimulation induced by dexamethasone but not by glucagon and it reversed the inhibition by insulin. Epidermal growth factor had no effect on PAP-1 activity. None of the above hormones or growth factors affected the subcellular distribution of PAP-1. Stimulation of PAP-1 by TGFb may be involved in mediating some of its biological effects. q 1997 Academic Press
Hepatocyte proliferation is controlled by the action of both growth stimulatory and growth inhibitory effectors. Studies on hepatocyte cultures have identified four complete hepatic mitogens - hepatocyte growth factor (HGF), epidermal growth factor (EGF), transforming growth factor a (TGFa) and acidic fibroblast growth factor (aFGF), while transforming growth factor b (TGFb) appears to be the most potent inhibitor of hepatocyte proliferation [1]. The signal transduction pathways initiated by some of these mitogens include activation of the Ras/Raf/MEK/MAP kinase signal cascade [2] which causes activation of nuclear transcription factors such as those encoded by the proto-oncogenes c-fos and c-jun [3]. The precise mechanism(s) whereby TGFb inhibits mitogen-induced proliferation remains unclear, although in the case of EGF, it may involve inhibition of EGF-mediated Ras activation [4]. 1 To whom correspondence should be addressed. Fax: /44 91 222 0723.
It has been demonstrated that several growth factors in different cell types, including EGF in hepatocytes, activate phospholipase D (PLD) in the plasma membrane which catalyses the hydrolysis of phosphatidylcholine (PC) to generate phosphatidic acid (PA) [5]. PA is now emerging as an important mitogenic signal acting either via the Ras/Raf cascade [6-8] or through a specific PA-dependent protein kinase [9]. PA is hydrolysed by the enzyme PA phosphohydrolase (PAP) to produce diacylglycerol (DAG), which is also a key second messenger acting via stimulation of protein kinase C (PKC) [10]. Clearly, PAP occupies a critical position in the signal transduction cascade of growth factors since its activity will determine the balance between two important second messengers, PA and DAG. PAP activity exists in two enzymic forms; one is sensitive to inhibition by N-ethylmaleimide (NEM), is dependent on Mg2/ for its activity in vitro and is present predominantly in the free state in the cytosol (PAP-1), while the other is insensitive to NEM and Mg2/ and is present in the plasma membrane (PAP-2) [11,12]. PAP-2 appears to be the most appropriately located for a role in signal transduction, although PAP-1 is capable of translocating to the membrane compartment in response to a variety of stimuli [13] and may also be involved in cell signalling in addition to its established role in lipid synthesis[12]. The principal aim of this study was to examine the effects of EGF and TGFb, two growth factors with opposing effects on hepatocyte proliferation, on the activities and distribution of both forms of PAP in rat hepatocyte cultures. The results demonstrate that TGFb stimulates PAP-1 activity in a time and concentration dependent manner. MATERIALS AND METHODS Porcine TGFb1 was from British Bio-technology Products Ltd. Abingdon, Oxon, England. Sources of the other reagents were as described previously [14,15]. Hepatocytes were isolated by a collagenase perfusion of whole liver of male Wistar rats (150-200g) fed on 0006-291X/97 $25.00
365
Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 5965
/
6916$$$621
12-18-96 14:48:51
bbrcg
AP: BBRC
Vol. 230, No. 2, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
standard laboratory chow ad libitum [14]. Hepatocytes were suspended in Minimum Essential Medium and cultured in 24-well plates at cell densities of 41104 cells/cm2 (low density) or 81104 cells/cm2 (high density) [15]. The cells were allowed to attach for 4h at 377C in a humidified 95% air/5% CO2 atmosphere before the medium was replaced with serum-free medium. For studies on their short-term (õ6h) effects, TGFb and/or EGF were added to the media from 1100 final concentration stock solutions after overnight (14-16h) cell culture. For studies on the longer term effects, growth factors and hormones, were added to the culture medium immediately after the initial 4h attachment phase and the cells were cultured for a further 20h. All experimental conditions were in triplicate. Phosphatidate phosphohydrolase assay. After incubation for the required time, the medium was aspirated, the cells were washed twice in 150mM NaCl and then sonicated in 300ml of extraction buffer (0.25M sucrose, 200mM DTT, 50mM b-glycerophosphate (pH 7.4) and 0.1mM EDTA). PAP activities were estimated by measuring the release of [32P]Pi from [32P]PA as previously described [11], except that the two forms were distinguished by their in vitro dependence on Mg2/ rather than their sensitivity to N-ethylmaleimide (NEM). Final assay incubations contained 1mM EDTA { 3mM MgCl2 . This method allowed both forms of activity to be measured in the same extract, and removed the need to incubate the cell extracts with and without NEM at 377C for 15 mins [12] which may lead to changes in enzyme activities. Preliminary experiments showed that assaying PAP activity in the absence of Mg2/ ions gave identical results to assays in their presence after pre-incubating the enzyme fraction with NEM (NEM-insensitive activity 95{5% of total PAP activity measured in the absence of Mg2/). One Unit of PAP activity was defined as the release of one nanomole of Pi per minute and was expressed per mg of protein estimated by the method of Bradford [16]. PAP-1 translocation. After overnight (20h) incubation with dexamethasone (10nM) the culture medium was replaced with medium supplemented with 2% fatty acid free BSA and either 2mM oleic acid and/or growth factors/hormones as indicated. After 1 hour the medium was removed and the cells were washed twice in 150mM NaCl. Digitonin buffer (0.05mg/ml digitonin, 300mM sucrose, 3mM Hepes, 2mM DTT, pH 7.2) was then added for 8 min and the plates were shaken gently. The digitonin fraction was then removed and the remaining cell matrix was extracted by sonication in the extraction buffer described above. PAP activities were assayed in both fractions as described above. All results are expressed as the mean{SEM of the number of independent cell cultures indicated. Unless stated statistical analysis was by the Student’s paired t-test.
RESULTS Effect of cell density on PAP activities. In hepatocyte cultures the expression of certain enzymes is enhanced at low cell density whilst that of other enzymes is greater at high cell density [17]. Since PAP-1 activity is very low in cells cultured without glucocorticoids but is markedly induced by dexamethasone [12,18,19], the effect of cell density on PAP activities was determined with and without overnight (20h) incubation with dexamethasone. Dexamethasone (10nM) increased PAP-1 activity from 0.78{0.16 mU/mg to 6.88{1.12 mU/mg at low cell density and from 0.88{0.31 mU/mg to 6.79{0.60 mU/mg at high cell density (nÅ4 experiments; p ú 0.05 for difference between cell densities). Increasing the concentration of dexamethasone to 100nM did not further increase PAP-1 activity and nei-
FIG. 1. Effects of [TGFb] and incubation time on PAP-1 activity in rat hepatocyte cultures. Hepatocyte monolayers were incubated for 20h at a cell density of 41104 cells/cm2 with 10nM dexamethasone except where indicated. In B TGFb (1ng/ml) was added at the time indicated prior to stopping the incubations. No TGFb was added to cells at zero time. Results from three independent experiments are shown. The effects of dose (with and without dexamethasone) and time were significant (põ0.02, ANOVA).
ther concentration of dexamethasone had any effect on PAP-2 activity. The effect of TGFb and EGF on PAP activities. Long-term incubation of hepatocytes with TGFb increased the activity of PAP-1 in a concentration and time dependent manner (Figure 1). An increase in activity was observed after 20 hours incubation with 50pg/ml TGFb, with no further increase in effect above 200pg/ml. The magnitude of the TGFb effect was similar in the absence of dexamethasone although, as expected, the absolute activities were much lower (Figure
366
AID
BBRC 5965
/
6916$$$621
12-18-96 14:48:51
bbrcg
AP: BBRC
Vol. 230, No. 2, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 1
Effects of TGFb on PAP Activities and Cell Protein Content at High and Low Cell Density PAP-1 (mU/mg)
PAP-2 (mU/mg)
Protein (mg/well)
Density
Low
High
Low
High
Low
High
Control TGFb (1 ng/ml)
6.42 { 0.7 11.58 { 1.4
5.63 { 0.6 7.81 { 0.7
1.95 { 0.1 1.98 { 0.2
2.06 { 0.1 2.07 { 0.3
0.28 { 0.04 0.24 { 0.03
0.71 { 0.06 0.62 { 0.06
Note. All incubations were supplemented with 10 nM dexamethasone and were for 20 h. Each value represents the mean { SEM of 3 independent experiments.
1A). The effect of 1ng/ml TGFb was apparent at 2 hours and by 20 hours the activity had approximately doubled (Figure 1B). The effect of TGFb was cell densitydependent, with a greater stimulation observed at low compared to high cell density (Table 1). Incubation with TGFb produced no significant change in cell protein content and had no effect on PAP-2 activity. EGF (10nM) had no effect on either PAP activity and did not influence the effect of TGFb on PAP-1 activity (not shown). This concentration of EGF was chosen as it had previously been shown to be the optimum required for both its mitogenic and metabolic effects on rat hepatocytes [20]. Short-term (5-30min) incubation of hepatocytes with either TGFb or EGF had no effect on the activity of either form of PAP (not shown). To investigate the possible mechanisms of the increase in PAP-1 activity produced by TGFb, its effect was examined in the presence of glucagon and insulin which have been shown to stimulate and inhibit PAP activity respectively [18,19]. The stimulation of PAP-1 by TGFb was slightly but not significantly lower than with glucagon but the combined effect of TGFb and glucagon was lower than with glucagon alone (Figure 2A). In the presence of dexamethasone, insulin decreased PAP-1 activity, in both the absence and presence of glucagon (Figure 2B), however, TGFb overcame the insulin inhibition so that the activity in the presence of insulin and TGFb was not significantly different from the control (Figure 2C). The effect of TGFb and EGF on subcellular distribution of PAP. The effect of short-term incubation of hepatocytes with growth factors and hormones on the subcellular distribution of PAP-1 activity is shown in Figure 3. In control cells 60% of PAP-1 was in the soluble (cytosolic) fraction and 40% in a bound state (matrix fraction). Oleic acid increased the proportion of bound PAP-1 in agreement with previous studies[21]. However the growth factors and hormones had no effect. In addition overnight incubation with each of the growth factors and hormones had no effect on either basal PAP-1 distribution or on the translocation induced by oleic acid (results not shown). PAP-2 activity was confined to the matrix fraction under all conditions.
DISCUSSION The results of this study demonstrate: (i) TGFb stimulates the activity of PAP-1 in a dose and time-dependent manner that is greater at low than at high cell density; (ii) no hormone or growth factor tested influences subcellular distribution; (iii) effects of hormones on PAP activity are wholly restricted to PAP-1. As far as we are aware this is the first demonstration of any peptide growth factor affecting PAP activity. The time course of the TGFb-induced stimulation of PAP-1 activity is similar to that shown previously for dexamethasone, glucagon and cAMP analogues [18,19] and suggests an effect via increased enzyme synthesis or decreased degradation rather than acute activation via covalent modification or a G-protein linked mechanism. Unlike glucagon or insulin, TGFb has no effect on either basal or stimulated levels of cAMP [22], however, the TGFb-response elements in the promoter regions of several TGFb-regulated genes include elements that are also cAMP responsive [23] suggesting a common regulatory endpoint for the effect of TGFb and cAMP on PAP gene transcription. Further evidence for this possibility is provided by a recent study demonstrating that TGFb mimics the effects of cAMP-elevating agents on mRNA levels of the enzyme phosphoenolpyruvate carboxykinase [24]. Does this stimulation of PAP-1 by TGFb play any role in mediating the known biological effects of TGFb? First, concerning the established role of PAP-1 in glycerolipid synthesis, while there has been no direct demonstration of a TGFb effect on lipid metabolism [23], high TGFb levels following sub-total hepatectomy [1] may contribute to the observed increase in PAP activity and associated fatty liver [25,26]. Second, concerning the postulated role of PAP-1 in cell signalling, the stimulating effect of TGFb on PAP-1 could have potentially important effects on the pattern of lipid- derived second messengers generated in response to agonists that act via PLD, such as EGF and PDGF [5,27]. Therefore the ability of TGFb to inhibit EGF-stimulated hepatocyte proliferation might be due, at least in part, to an increase in PAP-1 activity and a subsequent reduction in the concentration of the mitogenic second messenger,
367
AID
BBRC 5965
/
6916$$$621
12-18-96 14:48:51
bbrcg
AP: BBRC
Vol. 230, No. 2, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
c-myc proto-oncogene [28] which is induced by PA via activation of the Ras/Raf/MEK/MAP kinase signal cascade [29,30]. The concentrations of TGFb required to increase PAP activity are of the same order as those required for inhibition by TGFb of EGF-induced hepatocyte proliferation [31], and the greater stimulation of PAP activity by TGFb at low compared to high cell density is also consistent with a role for this effect in growth-related functions [17]. A further important biological effect of TGFb is to increase the transcription of genes encoding extracellular matrix (ECM) proteins [23]. The recent demonstration, therefore, that DAG and PKC are involved in the signal cascade leading to ECM gene transcription [32] suggests another example whereby the stimulation of PAP-1 by TGFb (via an increase in DAG concentration) might be important in mediating its biological effects. Clearly, the involvement of PAP-1 in cell signalling requires that it participates in the dephosphorylation of PA formed by the PLD pathway at the plasma membrane. This would be facilitated by the translocation of PAP-1 from the cytosol to the plasma membrane either in response to PA accumulation, as has been shown for microsomal and mitochondrial membranes [33,34] or by growth-factor induced phosphorylation as occurs with EGF-induced translocation of PLC-g1 [35]. In this study, however, the lack of PAP-1 translocation in response to EGF, which would be expected to stimulate PLD [5] and increase PA at the plasma membrane, does not support this hypothesis. Translocation to the plasma membrane is not a prerequisite for involvement of PAP-1 in cell signalling if PA, once formed, is transported to the endoplasmic reticulum membranes for
FIG. 2. Effects of TGFb (1ng/ml), glucagon (100nM) and insulin (10nM) on PAP-1 activity. Hepatocyte monolayers were cultured for 20h at a cell density of 4 1 104 cells/cm2 with 10nM dexamethasone and additions as indicated. In A and B each point represents the mean { SEM of 3 independent experiments; in C, 4 experiments. *põ0.01 versus control, **põ0.05 versus glucagon alone, ***põ0.001 versus either insulin or TGFb alone.
PA. Since the mitogenic effects of PA probably operate via increased Ras/Raf activity [6,7], this effect of TGFb may explain its ability to block the EGF-mediated activation of Ras [4] and to decrease the expression of the
FIG. 3. Effects of growth factors and hormones on PAP-1 subcellular distribution. Hepatocyte monolayers were cultured for 20h at a cell density of 4 1 104 cells/cm2 with 10nM dexamethasone prior to the additions indicated (TGFb 1ng/ml, glucagon 100nM, insulin 10nM, EGF 10nM and oleic acid 2mM). PAP activities were assayed in both fractions as described in the methods section. Results represent means { SEM of 4 independent experiments and are expressed as % of total cellular PAP-1 activity (controls 10.8 { 0.9 mU/mg).
368
AID
BBRC 5965
/
6916$$$621
12-18-96 14:48:51
bbrcg
AP: BBRC
Vol. 230, No. 2, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
hydrolysis and resynthesis as previously suggested [36]. Undoubtedly confirmation of a signalling role for PAP-1 awaits the demonstration that changes in its activity can influence the PA/DAG ratio following PLD activation by extracellular agonists. The best evidence thus far in this regard comes from studies with sphingosine, the mitogenic effects of which have been shown to reflect its ability to increase levels of PA, at least in part via inhibition of PAP activity [37,38]. Studies with intact cells have demonstrated that this inhibition is wholly restricted to the membrane-associated form of PAP-1 with no effect on the cytosolic form or on PAP2 activity [39]. ACKNOWLEDGMENTS C.P.D is supported by a Medical Research Council Clinician Scientist Fellowship. M.C.D holds a postgraduate studentship from the Biotechnology and Biological Sciences Research Council.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.
Michalopoulos, G. K. (1990) FASEB J. 4, 176–187. Egan, S. E., and Weinberg, R. A. (1993) Nature 365, 781–783. Davis, R. J. (1993) J. Biol. Chem. 268, 14553–14556. Attisano, L., Wrana, J. L., Lo´pez-Casillas, F., and Massague´, J. (1994) Biochim. Biophys. Acta 1222, 71–80. Bocckino, S. B., Blackmore, P. F., Wilson, P. B., and Exton, J. H. (1987) J. Biol. Chem. 262, 15309–15315. Yu, C-L., Tsai, M-H., and Stacey, D. W. (1988) Cell 52, 63–71. Ghosh, S., Strum, J. C., Sciorra, V. A., Daniel, L., and Bell, R. M. (1996) J. Biol. Chem. 271, 8472–8480. Tsai, M-H., Yu, C-L., and Stacey, D. W. (1990) Science 250, 982– 985. Khan, W. A., Blobe, G. C., Richards, A. L., and Hannun, Y. A. (1994) J. Biol. Chem. 269, 9729–9735. Nishizuka, Y. (1984) Nature 308, 693–698. Day, C. P., and Yeaman, S. J. (1992) Biochim. Biophys. Acta. 1127, 87–94. Jamal, Z., Martin, A., Gomez-Munoz, A., and Brindley, D. N. (1991) J. Biol. Chem. 266, 2988–2996. Cascales, C., Mangiapane, E. H., and Brindley, D. N. (1984) Biochem. J. 219, 911–916. Agius, L., Peak, M., and Alberti, K. G. M. M. (1990) Biochem J. 266, 91–102.
15. Peak, M., and Agius, L. (1994) Eur. J. Biochem. 221, 529–536. 16. Bradford, M. M. (1976) Anal. Biochem. 72, 248–254. 17. Nakamura, T., Yoshimoto, K., Nakayama, Y., Tomita, Y., and Ichihara, A. (1983) Proc. Nat. Acad. Sci. (USA) 80, 7229–7233. 18. Pittner, R. A., Fears, R., and Brindley, D. N. (1985) Biochem. J. 230, 525–534. 19. Pittner, R. A., Fears, R., and Brindley, D. N. (1985) Biochem. J. 225, 455–462. 20. Chowdhury, M. H., and Agius, L. (1987) Biochem. J. 247, 307– 314. 21. Cascales, C., Mangiapane, E. H., and Brindley, D. N. (1984) Biochem. J. 219, 911–916. 22. Thoresen, G. H., Refsnes, M., and Christoffersen, T. (1992) Cancer Res. 52, 3598–3603. 23. Massague´, J., Cheifetz, S., Laiho, M., Ralph, D. A., Weiss, F. M. B., and Zentella, A. (1992) Cancer Surveys 12, 81–103. 24. Thoresen, G. H., and Christoffersen, T. (1994) Cell Biol. International 18, 171–175. 25. Mangiapane, E. H., Lloyd-Davies, K. A., and Brindley, D. N. (1973) Biochem J. 134, 103–112. 26. Tijburg, L. B. M., Gellen, M. J. H., and Van Golde, L. M. G. (1989) Biochim. Biophys. Acta 1004, 1–19. 27. Fukami, K., and Takenawa, T. (1992) J. Biol. Chem. 267, 10988– 10993. 28. Pietenpol, J. A., Holt, J. T., Stein, R. W., and Moses, H. L. (1990) Proc. Nat. Acad. Sci. (USA) 87, 3758–3762. 29. Moolenaar, W. H., Kruijer, W., Tilly, B. C., Verlaan, L., Bierman, A. J., and de Laat, S. W. (1986) Nature 323, 171–173. 30. Howe, L. R., and Marshall, C. J. (1993) J. Biol. Chem. 268, 20717–20720. 31. Baskin, G., Schenker, S., Frosto, T., and Henderson, G. (1991) J. Biol. Chem. 266, 13238–13242. 32. Casini, A., Galli, G., Salzano, R., Ceni, E., Franceschelli, F., Rotella, C. M., and Surrenti, C. (1994) Alcohol 29, 303–314. 33. Martin, A., Hopewell, R., Martin-Sanz, P., Morgan, J. E., and Brindley, D. N. (1986) Biochim. Biophys. Acta 876, 581–591. 34. Freeman, M., and Mangiapane, E. H. (1989) Biochem. J. 263, 589–595. 35. Todderud, G., Wahl, M., Rhee, S. G., and Carpenter, G. (1990) Science 24, 296–298. 36. Michell, R. (1975) Biochim. Biophys. Acta 415, 81–147. 37. Zhang, H., Desai, N. N., Murphey, J. M., and Spiegel, S. (1990) J. Biol. Chem. 265, 21309–21316. 38. Lavie, Y., Piterman, O., and Liscovitch, M. (1990) FEBS Lett. 277, 7–10. 39. Gomez-Munoz, A., Hatch, G. M., Martin, A., Jamal, Z., Vance, D. E., and Brindley, D. N. (1992) FEBS Lett. 301, 103–106.
369
AID
BBRC 5965
/
6916$$$621
12-18-96 14:48:51
bbrcg
AP: BBRC