Brown, K. D., Blay, J., Irvine, R. F., Heslop, J. P., and Berridge,. M. J. (1984) Biochem. Biophys. Res. Commun. 123,377-384. 14. Hepler, J. R., Nakahata, N., ...
Communication
THEJOURNALOF BIOLOGICAL CHEMISTRY Vol. 263 No. 11, Issue of April 15 pp. 5030-5033 1988 0 1988 by The American Societ; for Biochemistry and Mblecular Biolo;, Inc. Printed in U.S.A.
Transforming Growth Factor,6 Modulates Epidermal Growth Factor-induced Phosphoinositide Metabolism andIntracellular Calcium Levels”
growth arrest and differentiation (3, 4), depending on the system studied. In Rat-1 cells, TGFP treatment results in phenotypic changes typical of transformation including acquisition of a transformed morphology, rapid acidification of the culture medium, and stimulation of growth in confluent TGFP inhibitsthe induction of the transingene by EGF in Rat-1 cells (5) and increases EGF-stimulated (hceived for publication, August 11, 1987) glucose (6) and amino acid (7) transport in other fibroblast cell lines. Leslie L. Muldoon, Karin D. Rodland, and The mechanisms by which TGFP exerts these various efBruce E. Magun+ fects have yet to be elucidated. It is unclear whether the TGFP From the Department of Cell Biology and Anatomy, Oregon modulation of EGF-mediated events occurs at thelevel of the Health Sciences Uniuersity, Portland, Oregon 97201 EGF receptor or more distally. TGFP hasbeen shown to alter either the number or affinity of EGF receptors (8, 9), but Transforming growth factor8 (TGFB) alters the cel- some EGF-mediated events, such as phosphorylation of the lular response to epidermal growth factor (EGF) in a ribosomal protein S6, are unaffected by TGFP treatment(10). number ofsystems, but the underlying mechanisms for In the present report we investigated the effect of TGFB on these alterations are largely unknown. We have examined second messenger formation in Rat1 cells fol- the generation of potential second messengers. One of the earliest actions of many peptide mitogens is to lowing treatment with EGF and/or TGFB to determine whether the ability of TGFB to potentiate some EGF- induce the hydrolysis of inositol phospholipids resulting in stimulated processes might be mediated by TGFfl-in- the production of the second messengers inositol trisphosducedalterations in the signal transductionmecha- phate (IP3), inositol tetrakisphosphate (IP3, and diacylglycnism. Incubation of serum-deprived confluent Rat- 1 erol. Production of IPSis generally thought to promote mocells with 10 ng/ml TGFB resulted in a markedeleva- bilization of Ca2+from stores in the endoplasmic reticulum tion of cellular inositol trisphosphate and inositoltet- ( l l ) , while IP, has been implicated recently in the translocarakisphosphate levels, which were maximal at 4 h and tion of extracellular Ca2+to calcium pools in the endoplasmic maintained for at least 8 h. The effect of TGFBon reticulum via the opening of calcium gates in the plasma levels of inositol trisphosphate and inositol tetrakis- membrane (12). phosphate was blocked by actinomycin D, suggesting The role of phosphoinositide turnover in EGF-stimulation that RNA synthesis was required for theTGFb effect. is unclear. EGF stimulation of phosphoinositide turnover is While EGF stimulation induced a rapid and transient not observed in mouse 3T3 cells (13), but this cell line is (5 min) rise in inositol phosphate levels in controlcells, poorly responsive to EGF. EGF has been found to increase the EGF effect was considerablyincreased,bothin phosphoinositide turnoverand intracellular Ca2+ in A431 magnitude and duration, TGFB by treatment. Measure- human carcinoma cells (14), but EGF is toxic, rather than mentof intracellular free Ca2+with fura-2 demon- mitogenic, to this cell type. Furthermore, Moolenaar et al. strated that TGFB treatment markedly increased the (15) found that theincreased free Ca2+observed in A431 cells EGF-stimulated rise in free Ca2+andincreasedthe following EGF treatment was due to Ca2+influx from extraduration of the response. The positive effects of TGFB on EGF stimulation could not be explained on the basis cellular sources ratherthan IP3-mediated mobilization of of increased EGF binding to cells. We conclude that intracellular calcium. We sought to clarify the role of phosTGFB treatment can both activate phosphatidylinositol phoinositide turnover in the EGF stimulation of Rat-1 cells, turnover independently and also sensitize Rat-1 cells which respond to EGF with a full round of mitogenesis (16). In addition we sought to determine if the TGFP modulation to stimulation by EGF. of EGF-stimulatedevents couldbe due to alterationsin second messenger formation. Transforming growth factor (TGFP)’ is a 25-kDapolypeptide purified from a number of normal and neoplastic cell sources. Treatment of cells with TGFp alters the responses to epidermal growth factor (EGF) in a complex manner to promote anchorage independence and proliferation (1, 2), or
* This work wassupported by United States Public Health Service Grants CA-39360 and CA-39181. 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 18 U.S.C. Section 1734 solely to indicate this fact. 3 To whom correspondence should be addressed. The abbreviations used are: TGFj3,transforming growth factor j3, EGF, epidermal growth factor; IP3, inositol trisphosphate; IP,, inositol tetrakisphosphate; HPLC,high pressure liquid chromatography; DMEM, Dulbecco’s modified Eagle’s medium; EGTA, [ethylenebis(oxyethylenenitrilo)]tetraacetic acid.
EXPERIMENTALPROCEDURES
Measurement of Cellular Inositol Phosphate Leuek-Rat-1 cells grown to confluency in 10-cm dishes were rinsed in Dulbecco’s modified Eagle’s medium (DMEM) and brought to quiescence by serum deprivation in DMEM for 36 h in the presence of 3 pCi/ml my0-[2-~H]inositol(Du Pont-New England Nuclear). The cells were then rinsed free of unincorporated label and incubated overnight in DMEM. In other experiments in which the cells were not changed to [3H]inositol-free medium, the results were equivalent (data not shown). Prior to harvesting, all cells were exposed to 100 mM LiCl for 20 min. While this concentration of Li‘ is higher than is usually reported, we found that 100 mM LiClwas necessary in order to observe consistent changes in IP3 and IPI levels. Following experimental treatment as indicated, cells were rinsed twice with ice-cold phosphate-buffered saline and were then extractedin 3 ml of ice-cold L. L. Muldoon, K. D. Rodland, and B. E. Magun, unpublished observations.
5030
TGFP and Phosphoinositide Metabolism
5031
1 M formic acid for 20 min. The extract was diluted to 0.1 M formic . _ 700 I acid, 0.2 M ammonium formate by the addition of NH40H. IPSand 600 140 IP, were separated on Dowex-1 (formate form)ion exchange resin in c 120 columns with a 1-ml bed volume by a modification of the method of 5 500 Berridge et al. (17). The columns were rinsed exhaustively with M 0.1 0 100 . 400 formic acid, 0.5 M ammonium formateto remove both inositol mon- " x 80 c ophosphate and bisphosphate. IP,was then eluted with4 ml of 0.1 M '5 300 + formic acid,0.85 M ammonium formate followed by extensive rinsing .60 U with the IP, buffer until only background levelsof radioactivity were E 200 40 present in the eluate. IP, was then eluted with 0.1 M formic acid, 1.1 .-T) 0 100 20 M ammonium formate. The eluted volumes were added to 16 ml of 'x ScintiVerse (Fisher):TritonX-100 (5:l) and were counted in a Beck- 10.GL' 0 2 4 6 8 man LS3801 scintillationcounter. The identity of the eluted fractions I ? 0 2 4 6 8 was confirmed by HPLC analysis,by the method of Batty et al. (la), Time (hrs) Time (hrs) utilizing a Partasil 10 SAX column. The IP3 peak eluted at 0.68 M ammonium formate, and the IP, peak eluted at 1.2 M ammonium FIG. 1. Effect of TGFB and actinomycin D on inositol phosformate. Inositol phosphate samples obtained from the Dowex column phate levels. Panel A, IP, levels; panel B, IP4levels. Serum-deprived method co-eluted with 'H-labeledstandards (Du Pont-New England -Rat-1cells were treated as follows: A, TGFB; A, TGFB and actinoNuclear) when analyzedby HPLC. mycin D added simultaneously; 0,TGFB with actinomycinD added Measurement of Zntracellulnr Ca" Using Fura-2--Rat-1cells to the medium at 4 h; (Act. D)0, actinomycin D alone; 0, control grown to confluency on 25-mmdiameter glass coverslips were serum- cells (vehicle only). The concentration of TGFB was 10 ng/ml added deprived for48 h in DMEM and then incubated for 30 min with 0.2 in acetonitrile/trifluoroacetic acidvehicle (1 pl/ml). Controlcells p~ fura-2 AM (Molecular Probes)as described by Grynkiewicz et al. were incubated in the acetonitrile/trifluoroacetic acid vehicle alone. (19).The cells wererinsed freeof extracellular dye, and the slide was Actinomycin D was added at 10 pg/ml. The cells were harvested at placed in a Sykes-Moore chamber on a temperature-controlled(37 "C) the indicated times and processed for determination of [3H]inositol microscope stage. Fura-2 fluorescence intensity was measured using phosphates as described under "Experimental Procedures." LiCl(lO0 a Spex microspectrofluorometeralternating between excitation at 340 mM) was added to all plates 20 min prior to harvesting. Each point and 380 nm. The fluorescence ratio, displayed as the ordinate, is represents the mean f S.D. of [,H]IP3 (Panel A) or [,H]IP, counts/ directly related to intracellular Caz' concentration (19). Each trace min (Panel B ) from three experimental plates. Although the values represents measurements recorded from a single field of about 30 given inthis and other figures are expressedas counts/min, quenching cells and is representativeof at least five similar experiments. of all samples counted was similar as determinedby the observed H "'1-EGF Einding Assay--Rat-1 cells grown to confluence on 35- values. mm plates were serum-deprived for 24 h in DMEM. Followingtreatment as indicated, cells(3 plates/point) were rinsed 3 times in DMEM (37 "C) and incubated for 5 min at 37 "C in DMEM containing 10 ng/ml '%I-EGF and 1 mg/mlbovine serumalbumin. Nonspecific binding was determined by incubation in medium containing '"1EGF plusa 1000-fold excess of unlabeled EGF. Materials-TGFB was prepared by the method of Assoian et al. (20) and was then purified to greater than 98% homogeneity by two rounds of reverse phase HPLC in acetonitrile, 0.1% trifluoroacetic acid using a Beckman C3 column. TGFB eluted at 29% acetonitrile. EGF was prepared by the method of Savage and Cohen (21) and was 0 10 20 30 40 50 60 0 10 30 20 40 50 60 (minutes) Time (minutes) Time further purified to homogeneity as described by us(22). The ultrapure EGF was iodinated usingNa'T and chloramine T (23). The specific FIG. 2. Effect of TGFj3 treatment on EGF-stimulated inosiactivity obtainedwas approximately 0.5 mCi/pg. tol phosphate levels. Panel A, IPSlevels; panel B, IP, levels. Rat-1 cells were treated with 10 ng/ml TGFB (0,A) or the acetonitrile/ RESULTS A N D DISCUSSION trifluoroaceticacid TGFB diluent (0)4 h prior to the experiment. At Incubation of serum-deprived confluentRat-1 cells with 10 zero time, 10 ng/ml EGF was added (0,O),and cells were harvested for determinations of [3H]inositol phosphates at the indicated times. mg/ml TGFB resulted in the elevation of the levels of both A indicates TGFO-treated cells which did not receive EGF. LiCl(100 cellular IPSand IP4 (Fig. 1,panels A and B ) . Accumulation of mM) was added to all plates 20 min prior to harvesting. Each point a 20-min exposure toLiCl was represents the mean k S.D. from three separate plates. the inositol phosphates during maximal at 4 h following TGFD addition and was sustained for at least 8 h (Fig. 1) before declining to control levels by EGF. Under these conditions, maximallevels of IP, and IP, 24 h (data not shown). The elevation of inositol phosphate were observed 10 min after addition of EGF and persisted for levels in response t o TGFB treatment was completely blocked at least 1h (Fig. 2). The extentof the IP, elevation in TGFBby simultaneous addition of actinomycin D (10 pg/ml). An treated Rat-1 cells at 10 and 20 min after EGF addition is established elevation of IPSand IP, produced by exposure to indicative of a synergistic interaction between the two agoTGFB was reduced to control levels 4 h after the addition of nists. actinomycin D (Fig. 1).These datasuggest that a transcribed Neither preincubation withEGF prior to addition of TGFB of the nor simultaneous addition of the two growth factors caused product is required for the induction and maintenance inositol phosphate responseto TGFP. an increase in the magnitude or time course of inositol phosIn order to determine whether the interactions between phate production. Only in TGFP-pretreated cells was there a TGFB and EGF might be explained by changes in EGFpotentiation in the ability of EGF to elevate levels of inositol stimulated phosphoinositide metabolism, we examined the phosphates. Following the 4-h TGFP incubation EGF treateffects of EGF stimulation of inositol phosphate formation in ment increased both IP, and IP, levels to values observed control cells and in cells treatedfor 4 h withTGFpto following serum stimulation, asshown in Table I. maximally increase their IP, and IP, levels. In control unIn order to determine whether the increasedlevels of IP3 treated Rat-1cells, addition of EGF increased thecellular IP, and IP, documented inFigs. 1 and 2 correlated with increases and IP, levels to 2-3-fold control, but the effect was transient in intracellular Ca2+ and to determine thesource of this Ca2+, and of small magnitude (Fig. 2). Treatment of cells with we examined cytosolic free Ca2+ levels utilizing the Ca*+TGFp for 4 h prior to the additionof EGF increased both the sensitive probe fura-2 (19). Whereasaddition of EGF to magnitude and durationof the inositol phosphateresponse to control untreated cells caused a minor and transient change
$
I
5032
TGFP and Phsphinositide Metabolism TABLEI
IPSand IP, levels following stimulation by EGF, !EF& and serum Values shown are mean & S.D., n = 3. Theexperiment was performed as described under “Experimental Procedures” withconditions as follows: control, TGFO diluentfor 4 h; EGF, TGFp diluent for 4 h followed by 10 ng/ml EGF for 5 min; TGFB, 10 ng/ml TGFp for 4 h; TGFp EGF, 10 ng/ml TGFp for 4 h followed by 10 ng/ml EGF for 5 min. Serum: 10%fetal bovine serum (Hyclone)for 5 min.
+
[‘HIIPa
Condition
dpm
:
Control EGF, 5 min TGFB, 4 h TGFp, 4 h, + EGF, 5 min Serum, 5 min 1.1
r
.
A 1 .o
,
.
777 2 147
1857 & 50 2407 f 105 3860 2 46 4196 f 88 .
.
.
X
6 hl
/
40 35
/
30
c
.-c
E
P
“-4
“”””
TGFB-treated
25 2o
I’HIIP, dpm
:
215 f 34 374 2 21 622 f 147 740 f 8
0 0 3
’
00
2
4
6
8
time (hrs)
849 f 55
FIG. 4. Effect of TGFB treatment on binding of ‘Z61-EGFto Rat-1 cells.Rat-1 cells were treated with 10 ng/ml TGFD (0)or the TGFB diluent(0).At the indicated times the cells were incubated for 5 min at 37 “Cin DMEM containing 10 ng/ml ‘“I-EGF and 1mg/ml bovine serum albumin. Dishes were then harvested, and specific’%IEGF bound was determined as described under “Experimental Procedures.”
0.9
the observed changes in the inositol phosphate levels. The effect of EGF required the presence of available external Caz+, 4 0.8 as no increase in intracellular Ca2+occurred in the presence n of either 5 mM EGTA (Fig. 3C) or 5 mM Coz+in the external medium (data not shown). Despite the ability of TGFp to 0.7 L 0 1 2 3 4 5 6 7 8 9 1 0 elevate inositol phosphate levels, the resting Ca2+levels in 0 1.1 cells pretreated with TGFP were no different than in controls, co as determined by the fura-2 A340/A380excitation ratios (averr) age baseline = 0.86, corresponding to 120 nM). This effect \ ’.O may be due to steady-state buffering of intracellular Ca2+by Ca2+-bindingproteins and by activation of Ca2+-ATPases. S 0.9 In view of reports that TGFp may exert its modulation of EGF-induced processes by its ability to alterbinding of EGF 0 to its receptor (8, 9), binding of [’“I]-EGF to Rat-1 cells was -it 0.8 measured in the presence and absence of TGFp. Since the m L i increased production of inositol phosphates induced by EGF in TGFB-treated cells occurs within 5 min of EGF addition, 0 1 2 3 4 5 6 7 8 9 1 0 -w altered binding of EGF should be detectable within that time 0 frame if such binding is responsible for the increased response l.l to EGF. Following incubation in the presence of TGFp for C varying times, the association of T - E G F with the surfaces 1.0 of Rat-1 cells was measured during a 5-min exposure to 10 ng/ml EGF at 37 “C. As shown in Fig.4, there was little difference in the binding of ‘T-EGF to cells treated for 4 h with TGFp when compared to the untreated control cells. The cells treated with TGFp demonstrated a slightly decreased binding of EGF beginning 2 h after TGFp addition. The results indicate that the TGFP-induced sensitization of , sdum , , J 0.7 phosphoinositide generation to the effects of EGF observed 0 1 2 5 4 5 6 7 8 9 1 0 in Rat-1 cells is not due to increased binding of EGF to its Time (min) receptor. FIG. 3. Measurement of intracellular free Cas+ using furaTGFphas been found to be both a stimulatorand an 2 microfluorometry. EGF (100 ng/ml) or fetal bovine serum (Hy- inhibitor of cell growth, depending on the system studied (1clone, final concentration 5%)were added to fura-2 loaded Rat-1 cells 4). In Rat-1cells TGF@ stimulates growth in confluent monoat the times indicatedby the arrows. Panel A, control untreated cells. layers but causes a transient inhibition of proliferation when Panel B, cells were preincubated with 10 ng/ml TGFp for 4 h prior to loading with fura-2. Allexperimentalmanipulations were per- added to exponentially growing cultures.2 In the presence of formed in the presence of TGFB. Panel C, cells were treated with TGFB, EGF-stimulated DNA synthesis is initially slowed but TGFp as in panel B , but theassay was performed in DMEM contain- then surpasses the level of DNA synthesis stimulated by EGF ing 5 mM EGTA. alone.3 Our data in Rat-1 cells are consistentwith previous results in Ca2+in control cells (Fig. 3A),the EGF-stimulated increase from other laboratories which haveshown that although EGF was of larger magnitude and more prolonged when cells were treatment ofA431 cells results in an increase in inositol pretreated with TGFB for 4 h (Fig. 3B). In all these experi- phosphate levels (14), the EGF-stimulated increase in intraments a concentration of 100 ng/ml EGF was used to optimize cellular Ca2+ is dependent on the presence of Ca2+ in the the fura-2 response by increasing the association rate for S. Planck, personal communication. EGF. Changes in free Ca2+levels correlated temporally with
=!
.-
TGFD and Phosphoinositide Metabolism extracellular medium (15). Our work extends these results to Rat-1 cells which display a full mitogenic response to EGF (16) and further indicates that, at least in Rat-1 cells, the ability of EGF to alterinositol phosphate and Ca2+levels is highly dependent on the presence of TGFP. These results provide the first evidence, to our knowledge, of second messenger modulation by TGFP. The data also suggest that the ability of EGF tomodulate the phosphatidylinositol signaling system may reflect the ability of TGFP to synergize with EGF in various cell systems. Since the ability of TGFB to increase EGF-stimulated phosphoinositide metabolism in Rat-1 cells cannot be explained by an increased binding of EGF to cell surfaces following TGFP exposure, TGFPmust affect the signal transduction pathway at asitedistal tothe EGF receptor. The synergistic interaction between TGFP and EGF in stimulatingphosphatidylinositol metabolism in Rat-1 cells may account for the variety of cellular activities observed in response to treatment with EGF and TGFP. Acknowledgments-Our thanks to Dr. D. McCarron and Dr. R. Bukoski of the Department of Medicine for the use of the Spex microspectrofluorimeter and for helpful discussions. We thank Jean Aschenbach and Thanh-Hoai Dinh for their excellent technical assistance and Julie Robinson for secretarial assistance. REFERENCES 1. Moses, H. L., Tucker, R. F., Leof, E. B., Coffey, R. J., Jr., Halper, J., and Shipley, G. D. (1985) Cancer Cells 3/Growth Factors and Transformation, pp. 65-71, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 2. Shipley, G. D., Tucker, R. F., and Moses, H. L. (1985) Proc. Natl. Acud. Sci. U. S. A. 82,4147-4151 3. Tucker, R. F., Shipley, G. D., Moses, H. L., and Holley, R.W. (1984) Science 2 2 6 , 705-707
5033
4. Anzano, M. A., Roberts, A. B., Meyers, C . A., Komoriya, A., Lamb, L. C., Smith, J. M., and Sporn, M. B. (1982) Cancer Res. 42,4776-4778 5. Matrisian, L.M., Leroy, P., Rautmann, G., Gesnel, M., and Breathnach, R. (1986) Mol Cell. Bid. 6 , 1679-1686 6. Inman, W. H., and Colowick, S. P. (1985) Proc. Natl. Acad. Sci. U. S. A. 8 2 , 1346-1349 7. Boerner, P., Resnick, R. J., and Racker, E. (1985) Proc. Natl. Acud. S C ~V. . S. A. 8 2 , 1350-1353 8. Massague, J. (1985) J. Cell Biol. 100,1508-1514 9. Assoian, R. K. (1985) J.Biol. Chem. 2 6 0 , 9613-9617 10. Like, B., and Massague, J. (1986) J. Biol.Chem. 2 6 1 , 1342613429 11. Berridge, M. J., and Irvine, R. F. (1984) Nature 3 1 2 , 315-321 12. Irvine, R. F., and Moor, R. M. (1986) Biochem. J. 240,917-920 13. Brown, K. D., Blay, J., Irvine, R. F., Heslop, J. P., and Berridge, M.J. (1984) Biochem. Biophys. Res. Commun. 123,377-384 14. Hepler, J. R., Nakahata, N., Lovenberg, T. W., DiGuiseppi, J., Herman, B., Earp, H. S., and Harden, T. K. (1987) J. Biol. Chem. 2 6 2 , 2951-2956 15. Moolenaar, W. H., Aerts, R. J., Tertoolen, Id. G. J., and deLaat, S. W. (1986) J. Bid. Chem. 261,279-284 16. Matrisian, L. M., Bowden, G. T., and Magun, B. E. (1981) J. Cell. Physiol. 108,417-425 17. Berridge, M. J., Heslop, J. P., Irvine, R. F., and Brown, K. D. (1984) Biochem. J . 2 2 3 , 1-7 18. Batty, I. R., Nahorski, S. R., and Irvine, R. F. (1985) Biochem. J. 232,211-215 19. Grynkiewicz, G., Poenie, M., and Tsien, R. Y. (1985) J . Biol. Chem. 260,3440-3450 20. Assoian, R.K., Komoriya, A., Meyers, A., Miller, D.M., and Sporn, M. (1983) J. Biol. Chem. 2 5 8 , 7155-7160 21. Savage, C.R., Jr., and Cohen, S. (1972) J . Biol.Chem. 2 4 7 , 7609-7611 22. Matrisian, L. M., Larsen, B. R., Finch, J. S., and Magun, B. E. (1982) Anal. Biochem. 125,339-351 23. Magun, B. E.,Planck, S. R., and Wagner, H. N., Jr. (1982) J. Cell Biochem. 2 0 , 259-276
