ifn-γa induces a phospholipase d dependent triphasic activation of protein kinase c in endothelial...

Vol. 189, No. 3, 1992 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNlCATlOFjS

December 30, 1992 Pages 1732-1738

IFN-y INDUCES A PHOSPHOLIPASE D DEPENDENT TRIPHASIC ACTIVATION OF PROTEIN KINASE C

IN ENDOTHELIAL CELLS

Pirkko Mattila’ and Risto Renkonen

Department of Bacteriology and Immunology,

University of Helsinki, Helsinki, Finland

Received November 19, 1992

The IFN-rlinked PKC activation in endothelial cells was analysed. It was shown that IFN-r activates PKC in three transient and separate cycles within the first 60 m inutes after IFN-y stimulation. Before each PKC activation there was an increase in DAG level. Ip3, phosphocholine and choline productions were measured to determine the origin of DAG. Neither of the PLC products, IP3 or phosphocholine, were released after IFN-r stimulation. On the other hand the PLD products choline and PA were released before all the three activation cycles of PKC. a 1992 Academic Press, Inc

Today it is known that the hydrolysis of inositol phospholipids is not the only mechanism leading to activation of protein kinase C (PKC). Recent studies have suggested that there are additional signal transduction pathways to provide diacylglycerol (DAG), which is essential for the PKC activation (1,2,3,4,5). One of these additional pathways is the phosphatidylcholine (PC) pathway. PC may be hydrolyzed by either phospholipase C (PLC) or by phospholipase D (PLD) coupled to phosphatidic acid phosphatase (PAPase) to produce DAG. The primary product of PLD is phosphatidic acid (PA), which is readily hydrolyzed to DAG by phosphatidic acid phosphatase (PAPase). There are also some reports indicating that PA itself could directly activate PKC (6). Interferons (IFNs) can turn on or upregulate at least some 20 genes and likewise downregulate several others (7). We have previously shown that activation of PKC is linked to IFN-y-induced major histocompatibility

'To whom correspondence should be addressed at Department of Bacteriology and Immunology, University of Helsinki, Helsinki, Finland. Fax: 358-O-4346382.

0006-29 1 X/92 $4.00 Copyright 0 I992 by Academic Press, Inc. All rights of reproduction in any form reserved. 1732

Vol. 189, No. 3, 1992 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

complex class II -antigen expression on cultured rat heart endothelial cells (89). Here we have analysed the IFN-y linked PKC activation in greater detail and shown that IFN-y activates PKC in three transient and separate cycles within the first 60 m in after stimulation and that this activation of PKC is dependent of DAG released by PLD.

EXPERIMENTAL PROCEDURES

Cell c-8 4-12- day-old DA rats were sacrificed, hearts were removed and endothelial cells were isolated, cultured and identified as earlier described (89). In some experiments the cells growing in flasks or on petri dishes were stimulated with 100 U/ml recombinant rat IFN-y (a gift from Dr P.H. v.d. Meide, Rijswijk, The Netherlands). PmhinkinaseC~~nassay Protein kinase C activity associated to membrane and cytosolic fractions after IFNy stimulation was measured as previously described (9). Phwbozes&?rbinc&lgassiLy Cells on m icrotiter plates (2x104 cells/well) were left untreated or treated in triplicates at 37 OC with 100 U/ml IFN-y for various time periods. The phorbol ester binding assay was done as described by Schutze et. al. (10). ~H~LXa&&ceml prvdudian Cells were labeled to equilibrium with 20 @ i/m l (l(3)-3H)glycerol(12 mCi/ml, Amersham) for 72 hr. Cells were extracted and analysed using thin layer chromatography as described (11). @%?mdheandphospw~~E~nepraduction Cells on 35mm dishes were labeled with 1.25 $i/ml of (methyl- 14C)Choline chloride ( 55 mCi/mmol, Amersham). After 48 hrs the medium was removed and cells were cultured for an additional 24 hrs in MEM plus 10 % FCS without (14C)Choline. The aqueous phase of the Bligh-Dyer extract of nontreated and IFN-y treated (14C)Choline labelled cells was analysed for radioactive choline and phosphoryl choline by TLC according to Yavin (12) using (14C)Choline and phospho(methyl- I4C)choline (Amersham) as standards. Ddenni~nof~~~acid Endothelial cells were labelled for 72 h with 20 l&i/ml (sH)glycerol(12 mCi/ m l, Amersham). The organic phase of methanol:PBS:chloroform extract was dried under vacuum and analysed for (3H)PA by TLC. The running solvent was the organic phase of ethyl acetateliso-octane/acetic acid/water (13:2:3:10), which separated PA from other phospholipids and neutral lipids (13). I~toltrp~(Ip3)dete~&m Inositoltriphosphate determinations were done using radioimmunoassay (Amersham RIA kit) according to the manufacturer’s instructions.

RESULTS

We have previously shown that IFN-y treatment of cultured rat heart endothelial cells led to MHC class II upregulation in a PKC-dependent manner (8,9). To further analyse the IFN-y-induced PKC activation we

1733


used two different approaches to measure PKC activation. In the first assay we measured the ability of phorbol dibutyrate (PBt2) to bind to its receptor, which is PKC. PBt2 is known to bind primarily to PKC translocated to cell membranes (10). The PBt2 binding assay revealed a triphasic PKC activation curve, where the peaks of transient PKC activation were seen in 8, 28 and 44 m inutes the maximum increase being 45 % above the control level (FIG.l). In the classical membrane translocation assay the cytosolic and membrane fractions were isolated from cells stimulated with IFN-y for different time periods (O-60 m in) and histone phosphorylation activity was measured. Activity in membrane fraction increased 25 % in 5 m inutes (from 2,5 to 3,2 pmol/min/pg prot) with a concomitant decrease in cytosolic fraction (FIG. 2). Within the first 60 m in three transient PKC activations were seen at the timepoints 5,20 and 40 m inutes. Basically these two different methods of measuring PKC activation gave the same results.

DAG is the physiological stimulator of cellular PKC. Therefore we measured DAG release in endothelial cultures after IFN-y stimulation. Three transient peaks of DAG release were found at 8, 28, and 48 m in after IFN-y stimulation (FIG. 3). Maximal counts varied between 300-1300 cpm in three different experiments.

-_ 0 4 8 1216aD24~88264)4446iS666)

time bin) Fipure

0 0 6 1015!al!zLma5~466055a 2 time (mid

IF’N-y (100 U/ml) induced (%-I)PB& binding. Cells were treated with IFN-y for O-60 m in, incubated with (3H)PBt2, lysed and counted. Nonspecific binding in the presence of 250-fold PMA was subtracted. Two experiments out of three are shown.

lciamd IFN-y (100 U/ml) induced PKC kranslocatioa Cells were incubated with IFN-y for different time periods (O-60 min). Afterwards membrane and cytosolic fractions were isolated and PKC activity in fractions was measured by histone phosphorylation activity. The data shown are one out of three experiments giving similar results.

1734

Vol. 189, No. 3. 1992 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

-c DAG

10 : .-..a... PA ol'l~l'l~l~i'l~l'l'l~l~l~l'l'l'l

0 4 8 1216P24283236404446S668D time @in)

Fiaure 3 IFN-y induced (3H)DAG release fi-om cell phospholipids. (3H)DAG was measured by TLC from the organic phase of methanol:PBS:chloroform extract. 1-stearoyl-2-(l-I’k!)arachidonyl-sn-glycerol used as standard was visualized by fluorography and sample fractions co-chromatographed with standard were scraped off the plate and counted. The data shown are one out of three experiments giving similar results. IFN-y induced (14Ckholine production. The aqueous phase of the methanol:PBS:chloroform extract of nontreated and IFN-y-treated cells was analysed for radioactive choline by TLC using (14C)choline as standard. This data are representative of three similar experiments. IFN-y induced (3H)PA production. (3H)PA was measured by TLC from the organic phase of methanol:PBS:chloroform extract using unlabelled L-cr phosphatidic acid (P-arachidonoyl-y-stearoyl) as standard that was visualized by exposure to iodine. The data are representative of three experiments giving the same results.

Our next aim was to analyse how DAG was released. DAG can be produced either by phosphatidylinositol or phosphatidylcholine specific PLC leading to inositoltrisphosphate (IP3) and phosphocholine release respectively in conjunction with DAG production. We were unable to find any traces of elevated IP3 levels between 30 seconds to 60 minutes in endothelial cell cultures treated with IFN-y (data not shown). Similarly phosphocholine also remained at the basal level (data not shown).

If PLC -pathways are not involved in the IFN-y induced PKC activation in endothelial cells could the PC specific PLD-pathway be involved instead? To answer this question the following requirements should be fulfilled. First, PLD induced choline production should be seen with a concomitant release of phosphatidic acid (PA). Second, the timing of DAG release should be just after PA production.

Trying to fulfill the first requirement we measured choline release aRer IFN-y stimulation. The endothelial cells were prelabelled with l*C-

1735


choline for 48 hours, incubated without the label for 24 h to reduce the intracellular free I4Ccholine, washed and stimulated with IFN-y. Increased levels of choline were transiently released from cell membranes in triphasic manner at 6, 22 and 38 minutes after stimulation (FIG.3). The maximal increase was 150 % over control values and maximal counts in three separate experiments varied between 300- 900 cpm depending on labelling efficiency. Similarly PA was detected in a triphasic manner at timepoints 2, 20 and 40 min after stimulation the maximum increase being 100 % over control values. Maximal counts varied between 100-300 cpm (FIG. 3).

DISCUSSION

IFNs increase DAG release in several cell types. There is an immediate increase in DAG and inositol bis- and tris-phosphates in human fibroblasts within 30 to 60s after exposure to IFNs a, p and y (14). The classical scheme that involves inositol phospholipid hydrolysis after treatment with IFNs as the sole source of DAG has recently been challenged by accumulating evidence that phosphatidylcholine may be an alternative source. It has been shown that IFN-(x and IFN-p do not induce inositol phospholipid turnover in Daudi and HeLa cells (15,16). In HeLa cells increased phosphatidylcholine hydrolysis and phosphocholine production were observed in response to IFN-a (17). PC metabolism may be involved in IFN-y-regulated DAG accumulation also in murine macrophages (18).

Biphasic formation of DAG is observed in bradykinin (BK) -stimulated human fibroblasts (18). A first peak is found at lo-15 s containing DAG released from phosphoinositols and a second peak at lo-13 min containing mainly DAG released by PLD from PC. In cultured vascular smooth- muscle cells angiotensin II also induces a biphasic increase in DAG. The first peak is due to a transient breakdown of inositolphospholipids and the second, sustained phase of DAG formation is due to PC hydrolysis probably via PLD activation (4).

Activation of PLD can occur by PKC-dependent and -independent pathways. A role for PKC in the generation of DAG from PC by the action of PLD has been described in some cell systems (19,201. On the other hand PKC was not essential for the activation of PLD in response to N-formyl- Met-Leu-Phe in neutrophils (21) and also antigen-induced PLD activation in rat mast cells is independent of PKC (22). In bovine pulmonary artery endothelial cells purinergic agonists stimulate PC breakdown by a PLD

1736

Vol. 189, 40. 3, 1992 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

mechanism without any observation of phosphatidylinositol breakdown (23).

The only report of oscillating DAG comes from Wener et. al. who detected oscillating DAG in thrombin activated human platelets (24). Like us, also these authors report that the statistical analysis of the data of DAG production is complicated by the lack of syncrony between the individual experiments e.g. even though the form of the DAG release curve was oscillating in all experiments the timing of peaks and lows were slightly different. These timing differences might reflect the different cell patches used in different experiments.

On the basis of the data shown here we propose a model for IFN-y induced PKC activation and MHC-class II upregulation (8,9) in endothelial cells. We suggest that IFN-y bound to its receptor activates PLD, which hydrolyzes cellular phosphatidylcholine to choline and PA. PA in turn is readily hydrolyzed to DAG by the action of phosphatidic acid phosphatase. The available free DAG then activates PKC.

The role of several sequential PKC activations is somewhat confusing. However, it is described by us and others (9,25) that IFN-7 has to be present practically most of the 24h incubation time for MHC-class II upregulation to occur. This fact suggests that it is possible that new PKC activations have to happen in order not to interrupt the IFN-y -induced signal. From this point of view activation of PLD instead of PLC seems logical, because phosphatidylinositol hydrolysis yields a relatively small amount of DAG for a short period of time compared to phosphatidylcholine hydrolysis (6).

ACKNOWLEDGMENTS

This work was supported in part by grants from the Research and Science Foundation of Farmos, Turku, Finland, Alfred Kordelin Foundation, Paul0 Foundation, Cancer Foundation and Finnish Academy, Helsinki, Finland.

REFERENCES

1.

2.

3. 4.

Besterman, J.M., Duronio, V. and Cuatrecasas, P. (1986) Proc. Natl.Acad. Sci. USA 83,67&j-6789. Fisher, G.J., Henderson, P.A., Voorhees, J.J. and Baldassare J.J. (1991) J. Cell. Phys. 146,309-317. Irving, H.R. and Exton, J.H. (1987) J. Biol. Chem. 262,3440-3443. Lassegue, B., Alexander, R.W., Clark M. and Griendling, K.K. (1991) Biochem. J. 276,19-25.

1737


5.

;: 8.

9.

10.

11. 12. 13.

14. 15.

16.

17.

18.

19.

20.

21.

22.

23. 24.

25.

Billah, M .M., Pai, J-K., Mullmann, T.J., Egan R.W. and Siegel M .I. (1989) J. Biol. Chem. 264,9069-9076. Exton, J.H. (1990) J. Biol. Chem. 265,1-4. Pfeffer, L.M. and Tan, Y.H. (1991) TIBS 16: 321-323. Mattila, P.M., Hayry, P. and Renkonen, R. (1989) Febs. Lett. 250, 362366. Mattila, P.M. and Renkonen, R. (1991) Stand. J. Immunol. 34, 153160. Schutze, S., Nottrott, S, Pfizenmaier, K. and Kriinke, M . (1990) J. Immunol. 144,2604-2608. Rosoff, P.M., Savage, N. and Dinarello, C. (1988) Cell. 54, 73-81. Yavin, E. (1976) J. Biol. Chem. 251,1392-1397. Blitterswijk, W.J., Hilkmann, H., de Widt, J. and van der Bend, R.L. (1991) J. Biol. Chem. 266,10344-10350 a. Yap, W.H., Teo, T.S. and Tan, Y.H. (1986) Science 234, 355-358 a. Pfeffer, L.M., Strulovici, B. and Saltiel, A.R. (1990) Proc. Natl. Acad. Sci. USA 87,6537-6541. Yap, W.H., Teo, T.S., McCoy, E. and Tan, Y.H. (1986) Proc. Natl. Acad. Sci. USA 83,7765-7769 b. Sebaldt, R.J., Prpic, V., Hollenbach, P.W., Adams, D.O. and Uhing, R.J. (1990) J. Immunol. 145, 684-689. Blitterswijk, W.J., Hilkmann, H., de Widt, J. and van der Bend, R.L.J. (1991) Biol. Chem. 266, 10337-10343 b. Cabot, M .C., Clement, J.W., Zhang, Zu-chuan and Cao, Hui-ting. (1989) Febs. Lett. 245,85-90. Martinson, E.A., Trilivas, I. and Brown, J.H. (1990) J. Biol. Chem. 265,22282-22287. Reinhold, S.L., Prescott, S.M., Z immerman G.A. and McIntyre T.M. (1990) FASEB J. 4,208-214. Yamada, K., Kanaho, Y., M iura, K. and Nozawa, Y. (1991) Biochem. Biophys. Res. Commun. 175, 159-164. Martin, T.W. and M ichaelis, K. (1989) J. Biol. Chem. 264,8847-8856. Werner, M .H., Bielawska, A.E. and Hannun, Y.A. (1992) Biochem. J. 282,815-820. Gumina, R.J., Feire-Moar, J., DeYoung, L., Webb, D.R and Devens, B.H. (1991) Cell. Immunol. 138,265-279.

1738

ifn-γa induces a phospholipase d dependent triphasic activation of protein kinase c in endothelial...

Documents