Lysophospholipids reaction mixtures (0.2 ml) contained 5 pmol of Tris/HCl (pH 7.51, 2 ...... McPhail, L. C., Clayton, C. C., and Snyderman, R. (1984) Science 224,. 2. .... Jacobson, R. D., Virag, I., and gkene, J. H. P. (1986) J. Neurosci. 6 , 1843-.
THEJOURNALOF BIOLOGICALCHEMISTRY 0 1988 by The American Society for Biochemistry and Molecular Biology, Inc.
Vol. 263,No. 14, Issue of May 15,pp. 6865-6871, 1988 Printed in U.S.A.
Regulation of Protein KinaseC by Lysophospholipids POTENTIAL ROLE IN SIGNAL TRANSDUCTION* (Received for publication, August 6, 1987)
Kazuhiko Oishi, Robert L. Raynor, Paul A. Charp, and J. F. Kuo From the Department of Pharmacology, Emory University School of Medicine, Atlanta, Georgia 30322
Certain lysophospholipids, lysophosphatidylcholine acid) have been shown recently to be able to stimulate PKC’ (lyso-PC) in particular, stimulated protein kinase C at activity in the presence or absence of Ca2+or PS (1-5). The low concentrations (30 PM). Protein kinase C existence of yet anothersignal transduction pathway involved stimulation by lyso-PC required the presence of phos- in PKC activation, in addition to the well recognized diacylphatidylserine (PS) and Ca2+and was associated with glycerol system coupled to hydrolysis of polyphosphoinosia decreased K, for PS and increased K,, for Caz+of the tides by phospholipase C (6). One pathway through which enzyme. Cardiolipin and phosphatidic acid could par- fatty acids are generated is hydrolysis of membrane phosphotially substitute for PS in supporting the stimulatory lipids by phospholipase Az. Another group of products genereffect of lyso-PC!.Lyso-PC also biphasically regulated ated by this reaction is lysophospholipids. We reported reprotein kinase C activated by diolein. Of several syn- cently that ALP, a lyso-PC derivative possessing antineoplasthetic lyso-PC preparations tested, the oleoyl, myris- tic properties (for a recentreview, see Ref.7), was an inhibitor toy1 and palmitoyl derivatives were most active. Data of PKC (8)and counteracted the effects of a tumor-producing from the Triton:X-lOO mixed micellar assay indicated phorbol ester (9). We investigated possible effects of lysophospholipids on that 1.4 and 14..0 mol of lyso-PC/micelle produced a maximal stimulation anda complete abolishment of the PKC activity and cellular functions so that a second messenstimulation of protein kinase C, respectively. Protein gerrole of these lipid metabolites can besuggested.We kinase C stimulation by lyso-PC, with a pH optimum reported here that certain lysophospholipids, lyso-PC in parof about 7.5, was observed for phosphorylation of his- ticular, were effective stimulators and inhibitorsof PKC. tone H1,myelin basic protein, and the35-and 47-kDa EXPERIMENTALPROCEDURES proteins from the rat brain, but not for that of other histone subfractionsandprotamine. Lyso-PC acted Materials-Various lyso-PC preparations, lysophosphatidylinosisynergistically withdiacylglycerol in stimulating pro- tol, lysophosphatidylglycerol, lyso-PA, lysophosphatidylethanolamtein kinaseC, whereas thestimulation by lyso-PC was ine, lyso-PAF, cardiolipin, PS, PC, PA, diolein, sphingosine, various additive to thatby oleic acid. Protein kinase C inhibi- histone subfractions, protamine, myelinbasic protein, tamoxifen, CAMP, Pipes, and MES were purchased from Sigma; tors (alkyllysophospholipid, sphingosine, tamoxifen, calmodulin, oleic acid (free acid) was from Aldrich; polymyxin B was from The and polymyxin E:) inhibited more potently the protein Upjohn Co. ALP was kindly provided by Dr. Paul Munder, Maxkinase C activity stimulatedby PS/Ca2+/lyso-PCthan Plank Institut for Immunology (Friedberg, Federal Republic of Gerthat stimulated b:y PS/Ca2+.The stimulatory and inhib-many). MLC kinase and MLC from rabbit skeletal muscle were kind itory effects of lyso-PC were not observed for myosin gifts of Dr. James T. Stull, University of Texas HealthScience Center light chain kinase and CAMP-dependent protein ki- (Houston, TX). Treatment of Agents-PS and oleic acid were dissolved in chloronase, indicatingII specificity of its actions. form, ALP in isopropyl alcohol and all other lipids (unless otherwise The present findings suggested that lyso-PC, likely indicated) in a mixture of methyl alcohokchloroform (l:l, v/v). Apderived from me:mbrane P C by the action of phospho- propriate aliquots of the solutions, with or without PS, were dried lipase Az, might :play a role in signal transductionvia under streams of N, gas, followed by sonication in appropriate vola dual regulationof protein kinase C, and that itcould umes of 25 mM Tris/HCl (pH 7.5) containing 10 mM MgCl, for 30 s further modulate the enzyme and hence the cellular at room temperature. PA, lyso-PA, lysophosphatidylinositol and poactivity by interplaying withdiacylglycerol and unsat- lymyxin B (dissolved in water), and tamoxifen (dissolved in dimethyl sulfoxide) were diluted to appropriate concentrations with the same urated fatty acid, the two other classes of cellular solution. Aliquots of agents prepared above were added directly to mediators also sh,ownto be activators of protein kinase the incubation mixtures for PKC assays. The final concentrations of
C.
dimethyl sulfoxide in the incubation mixtures, if present, were less than 0.2% (v/v), which was without effect on the activity of PKC. Preparation and Assay of PKC-PKC was purified from pig brain extracts through the stepof phenyl-Sepharose affinity step (10). The enzyme, without contamination of other proteinkinases, was assayed essentially as described elsewhere (11, 12). Briefly, the standard
Unsaturated free fatty acids (suchas oleic and arachidonic * The present work was supported by United States Public Health Service Research Grants HL-15696, CA-36777, and NS-17608. The costs of publication of this article were defrayed in part by the payment of page charges. This articlemust therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The abbreviations used are: PKC, protein kinase C; MLC, myosin light chain; PS, phosphatidylserine; PC, phosphatidylcholine; PA, phosphatidic acid; PAF, platelet-activating factor (1-0-alkyl-2-acetylglycero-3-phosphocholine);ALP, alkyllysophospholipid (1-0-octadecyl-2-0-methyl-glycero-3-phosphocholine);Pipes, 1,l-piperazinediethanesulfonic acid; MES, 2-(N-morpholino)ethanesulfonic acid; EGTA, [ethylenebis(oxyethylenenitrilo)]tetraacetic acid; SDS, sodium dodecyl sulfate.
6865
PKCLysophospholipids Regulation by
6866
reaction mixtures (0.2 ml) contained 5 pmol of Tris/HCl (pH 7.51, 2 pmol of MgC12, 2 pg of PS, 40 pg of histone H1 (lysine-rich), either 0.8 pmol of EGTA or 0.04 pmol ofCaC12, 1 nmol of [y3'PJATP (containing about 1 X IO6 cpm), appropriate amounts of the agents to be tested, and appropriate amounts of the enzyme. The reaction, started with the addition of the radioactive ATP, was carried out at 30 'C for 5 min. Homogenous PKC was prepared from rat brain extracts through the polylysine step according to the procedures of Huang et al. (13). PKC assay using mixed micellar assay system was also performed essentially according to the method of Hannun et al. (14). The assay system contained 0.3% Triton X-100, 6 mol % of PS (8.4 molecules/ micelle), 0.5 mol % of diolein (0.7 molecule/micelle), 200 p M CaCL and varying mol % of lyso-PC as indicated. Other Methods-Phosphorylation of endogenous proteins in the calmodulin-depleted soluble fraction of rat cerebral cortex was carried out as described (15). CAMP-dependent protein kinase was partially purified from bovine heart extracts (161,and MLC kinase was purified to homogeniety from rabbit skeletal muscle extracts (17). CAMPdependent protein kinase using mixed histone as substrate (16) and MLC kinase using MLC as substrate (17) were assayed as described in the references cited. The activities of CAMP-dependent protein kinase and MLC kinase as well as PKC were linear with respect to the amounts of the enzymes and thetime of incubation. [y-S2P]ATP was prepared by the method of Post and Sen(18).Free Ca" concentrations were estimated by the EGTA-Ca" buffer system (19).
activated by a low concentration (0.3 pg/ml) of diolein (Fig.
1B). The modes of stimulation by lyso-PC were investigated further. Analysis of the activation kinetics revealed that lysoPC acted by increasing the V,, accompanied by a decreased KOfor PS from 19.9 to 3.6 pg/ml (Fig. 2 . 4 ) and an increased K, for Ca2*from 19.4 to 39.4 pM (Fig. 2 B ) . Unless otherwise indicated, lyso-PC (oleoyl) was used in experiments shown in Figs. 1 and 2 and in all subsequent studied reported herein. It was found that the acyl groups of lyso-PC were important in determining its ability to stimulate PKC (Fig. 3). The relative effectiveness of the fatty acid groups in lyso-PC, in a decreasing order, was oleoyl (C,,,,) = myristoyl (C14.0)= palmitoyl (Cleo) > lauroyl (C,,,,) = heptadecanoyl (C,,,,) > decanoyl (Cleo) = caproyl (Ceo) = stearoyl the last three species of lyso-PC were practically inactive. Of the naturally occurring lyso-PC, the relative effectiveness according to sources of the preparations, in a decreasA
B 25
20 .
RESULTS
15
Although without effect on PKC activity in the absence of PS, lyso-PC stimulated further the enzyme activated by PS in the presence ofCaC1, (Fig. lA).The extent to which the enzyme was stimulated was positively related to the PSconcentration. The effects of lyso-PC were biphasic in that it stimulated the enzyme at low concentrations but conversely inhibited it athigh concentrations. It was further noted that the extent of enzyme inhibition by lyso-PC was inversely related to the PS concentration (Fig. L4).Similar biphasic effects of lyso-PC were also observed using a homogenous PKC preparation or crude extracts of rat brains (data not shown). The findings clearly indicated that regulation of PKC by lyso-PC involved itsinteractions with PS vesicle. The biphasic regulation by lyso-PC was also noted for PKC activated by high concentrations (1 and 3 pg/ml) of diolein, whereas lyso-PC was strictly inhibitorywhen the enzyme was
A
75
I
1)
10
100
PS (}lg/ml)
1000
-
0
1
10
100
1000
Ca2* (pM1
FIG. 2. PKC activation as a function of PS or Ca" concentration. PKC was assayed under the standard conditions except for varying concentrations of PS with a fixed concentration (200 p M ) of CaClz ( A )and varying concentrations of Ca2+with a fixed concentration (10 pg/ml) of PS ( B ) . The free Ca2' concentrations ( B ) were determined as described under "Methods." When present, the concentration of lyso-PC was 8 pglml. Double-reciprocal plots of the data are also shown (insets). The values presented are means of duplicate assays, with assay errors less than 3%. Similar results were obtained in two separate sets of experiments.
B
r ool
r
250
r
50
25
LL
0
3
10
30
100
LYSO
0
3
10
30
PC ( p / m l )
FIG. 1. PKC regulation by lyso-PC as a function of PS and diolein concentration. A , PKC was assayed under the standard conditions (containing 200 pM CaC12)except for the varying concentrations of PS andIyso-PC. B, PKC was assayed under the standard conditions but using a low concentration (1 p ~ ofCaC12, ) in the presence of varying concentrations of diolein and lyso-PC and a fixed concentration (10 pg/mI) of PS. The values presented are means of triplicate assays. Similar results were also obtained in three separate sets of experiments.
FIG. 3. PKC activation by lyso-PC as related to its fatty acid constituents. PKC was assayed under the standard conditions except for the kinds and concentrations of lyso-PC, as indicated. The data shown are means of duplicate assays, with assay errors less than 2%.The findings were confirmed in two separate sets of experiments.
PKC Regulation by Lysophospholipids ing order, was soybean > egg yolk > bovine brain > bovine liver (inactive). Of several lysophospholipids examined, lyso-PC (oleoyl) was most active; lysophosphatidylglycerol (egg yolk), lysoPAF (hexadecyl), and lysophosphatidylinositol (soybean) were less active; lysophosphatidylethanolamine (bovine brain), lyso-PA (oleoyl), and PC (bovine brain) were inactive (Fig. 4). The optimal concentration for the four active lysophospholipids in stimulatingPKC, surprisingly, was quite similar, i.e. 10 pg/ml. These lysophospholipids, with an exception of lyso-PA, inhibited the enzyme nearly completely at 50-100 pg/ml. PC (Fig. 4) and PAF (hexadecyl) (data not shown), unlike their lyso analogs, were not stimulatory. Because of the diversity of sources and acyl constituents of the lipids used, the observed specificity on PKC stimulationcould only be regarded as tentative. A more rigorous studies employing, for example, synthetic oleoyl-lysophospholipidscould yield more definite data on the structure-activity relationship of the agents. In addition to PS,PA and cardiolipin are alternative phospholipid cofactors for PKC activation (12, 20, 21). It was noted here that they could also substitute for PS to support the stimulatory effect of lyso-PC (Fig. 5). The maximal lysoPC stimulation seen with PA (3-fold) was higher than those seen with PS (2-fold) and cardiolipin (1.5-fold). The optimal concentration of lyso-PC for PKC stimulation for both PS and PA was about 10 pg/ml, which was higher than thevalue (3-5 pg/ml) seen for cardiolipin. We next examined potential effects of substrate proteins on PKC stimulated by lyso-PC compared with that by oleic acid, previously shown to stimulate PKC (1-5). Lyso-PC and oleic acid increased the VmaX about 2.2- and 3.7-fold in histone H1 phosphorylation, respectively (Fig. 6). Lyso-PC and oleic acid also increased the K,,, for histone H1 from about 40 to about 60 and 70 pg/ml, respectively (Fig. 6). It was further observed, using other substrate proteins (atfixed a concentration of 200 pglml), that thestimulatory effect of lyso-PC was unobserved when phosphorylation of histones H2a, H2b, and H3 were measured (Table I). PKC stimulation by oleic acid, on the other hand, was unobserved only for the phosphorylation of histone H2b. Phosphorylation of protamine by PKC was found to be unstimulated oreven inhibited by Ca2+in the 0
0
A A
PC PI L y s o PG LYSD PA
0
1
10
200 c'
? C
?I
150
L
0
aQ 100 +-r
I\
~
\
\
50
0
FIG. 5. SRecificity of RhosRholiDids in suDDorting PKC aCtivation by-lyso-PC. PKC was assayed under t h e standard conditions except for the kinds and concentrations of phospholipids (PS, 25 pg/ml; PA, 25 pg/ml; cardiolipin, 50 pg/ml) and varying concentrations of lyso-PC as indicated. The control activity, seen in the presence of PS but in the absence of lyso-PC, was taken as 100%. The data presented are the means of duplicate assays, with assay errors less than 4%. The findings were confirmed in three separate sets of experiments. "
-
0 None 0 L Y S O PC
s" A
0
0
Oleicacid
100
200
300
400
HISTONE H1 (pglml)
FIG. 6. PKC activation by lyso-PC and oleic acid as a function of histone H 1 concentration. The enzyme was assayed under the standardconditions except for the inclusion of lyso-PC (8 pg/ml) and oleic acid (15 pg/ml) and varying concentrations of histone H1, as indicated. The datapresented are means of duplicate assays, with assay variations less than 3%. Similar results were also obtained in two separate sets of experiments.
LYSO LYSO
0 L y s o PAF 0
6867
L y s o PE
100
1000
LIPID (Pg/rnl)
FIG. 4. Specificity of lysophospholipids in activating PKC. The enzyme was assayed under the standard conditions except for the kinds and concentrations of lysophospholipids, as indicated. The data presented are themeans of duplicate assays, with assay variation less than 3%. The findings were confirmed in two separate sets of experiments. PI, phosphatidylinositol; PG, phosphatidylglycerol; PAF, platelet-activating factor; PE, phosphatidylethanolamine.
presence of PS; both lyso-PC and oleic acid practically had no effect (Table I). Phosphorylation of myelin basic protein, like that of histone H1, was augmented by lyso-PC and oleic acid (Table I). The data indicated that thenature and extent of PKC stimulationby lyso-PC and oleic acid were influenced by substrateproteins,in addition to phospholipids shown above (Fig. 5), suggesting complex interplays among various reactants in the PKC reaction system. It appeared that an interaction existed between lyso-PC and histone H1, because the concentrations of lyso-PC required for a maximal stimulation increased as the concentrations of the substrate proteins increased (Fig. 7). The effects of lyso-PC were examined further on phosphorylation of cellular proteins. It was observed that lyso-PC stimulated and inhibited,in a concentration-dependent manner, phosphorylation of the 47- and 35-kDa PKC substrate proteins from the soluble fraction of rat brain cortex (Fig. 8). Although phosphorylation of the 35-kDa protein shown in Fig. 8 did not appear to be PS/Ca2+-dependent, itsphosphorylation was clearly stimulated by PS/Ca2+, and, furthermore, this phosphorylation was biphasically regulated by lyso-PC in another set of experiments where the extent of phospho-
PKC Regulation by Lysophospholipids
6868
TABLE I Effects of substrate proteinson PKC stimulated by lyso-PC and oleic acid PKC was assayed under the standardconditions except for the inclusions of various substrate proteins (200 pg/ ml), lyso-PC (8 pg/ml), and oleic acid (15 pg/ml), as indicated. When present, the concentrations of EGTA and CaC12were 4 and 0.2 mM, respectively. PS (10 pg/ml) was present in all incubations.The data shown are mean f S.E. of triplicate assays. PKC activity Substrate acid None (control) Oleic Lyso-PC protein
CaCIZ
EGTA EGTA
CaClz CaCL
EGTA pmol Pfmin
None (basal) Histone H 1 Histone H2a Histone H2b Histone H3 Protamine Myelin basic protein
0.01 f 0.01 0.23 f 5.06 0.03 0.51 f 0.01 0.80 4.07 f 0.09 0.41 f 0.02 0.49 10.41 f 0.26 0.73 f 0.02 9.26
0.01 f 0.01 f 0.02 f 0.01 0.45 11.46 f 0.38 3.86 f 0.01 0.43 6.22 f 0.26 f 0.45
f 0.01 0.02 0.08 f 0.01 16.47 f 0.01 0.48 f 0.17 11.31 f 0.02 0.36
f 1.02 0.01
f 0.01 0.49 f 23.71 0.02 f 0.03 1.83 3.42 f 0.10 10.97 f 0.02 0.82 f 16.53 1.61 f 0.01 12.43
f 0.23 f 0.03 0.70 f 0.32 f 0.01 0.72 17.80 0.32 15.00 f 0.27 1.76
*
13.73 f 0.61 12.44 1.61 f 0.08
0.06 f 0.01 f 0.71 f 0.09 f 0.13 f 0.05 f 1.04 k 0.34
1 2 3 4 5 6 7 8 9 1 0
0
2
5
10
20
50
Lyso PC (pg/rnl)
FIG. 7. PKC activation by lyso-PC as a function of histone H1 concentration. The enzyme was assayed under the standard
35
conditions except for the inclusion of varying concentrations of lysoPC and histone H1, as indicated. The data presented are means of triplicate assays. Similar results were also obtained in another setof experiments.
rylation was less (datanot shown). Lyso-PC was strictly inhibitory to phosphorylation of many other cellular proteins, however (Fig. 8). We also noted that lyso-PC stimulated and inhibited, in a concentration-dependent manner as shown in Figs. 4,7, and8, phosphorylation of some proteins stimulated by PS/Ca’+ in extracts of HL-60 cells.’ The pH optimum for PKC stimulation by lyso-PC was found to be around 6.5-7.5 (Fig. 9). The stimulatory effect of lyso-PC was noted for various buffer systems with a wide range of pH values, including sodium acetate (pH 5.0 and 5.5), MES (pH 5.5, 6.0, and 6.5), and Tris/HCl (pH 7.5, 8.0, 8.5, and 9.0). The effect, however, was less pronounced for Pipes (pH 6.5, 7.0, and 7.5) and was practically nil for potassium phosphate (pH 6.5, 7.0, and 7.5). The stoichiometric data on lyso-PC regulation of PKC was determined using the Triton X-100 mixed micellar system (14). It was noted that about 1 and 10 mol % of lyso-PC, calculated to be 1.4 and 14 molecules of lyso-PC/micelle, produced a maximal stimulation and complete abolishment of the stimulation, respectively (Fig. 10). The molar concentration of 1 mol % lyso-PC (oleoyl) in the mixed micellar system was calculated to be 48 p ~which , washigher than its optimal stimulation concentration (about 10 pg/ml, or 19 p ~ seen under the standard PKC assay system (Figs. 1, 4, and 5). The combined effects of lyso-PC and other PKCactivators M. Shoji and J. F. Kuo, unpublished observations.
CaC12(200p1\rlJ (pg/ml) PS
-
t
-
0
0 2 2
t
t
t
t
t
t
t
102 22 2 2
0 0 0 0 2 5 10 20 50 FIG. 8. Autoradiograph showing phosphorylation of endogenous proteins in the calmodulin-depleted soluble fraction of rat cerebral cortex. The experimental conditions were essentially Lyso PC(pg/ml) 0
the same as described earlier (15). Briefly, fresh rat cerebral cortex (1.1 g) was homogenized in 1 volume of ice-cold solution A (0.25 M sucrose, 25 mM Tris/HCl (pH 7.5), 10 mM MgClz, 50 mM mercaptoethanol, 0.3 mM EGTA, and 2 mM phenylmethylsulfonyl fluoride), with the use of a glass-Teflon homogenizer. The homogenate was centrifuged a t 105,000 X g for 60 min. An aliquot (1 ml) of the resultant supernatant was loaded onto a DEAE-cellulose column (1 X 12 cm) and was eluted with 20 ml of 250 mM KC1 dissolved in solution B (20 mM Tris/HCl, pH 7.5,50 mM mercaptoethanol). The column was previously equilibrated with solution B. The eluate (calmodulin-deficient cytosol) was used as the source of PKC and its )substrate proteins. Aliquots (containing 100 pg protein) of the eluate were phosphorylated under the standard PKC assay conditions (see “Experimental Procedures”), with additions of PS, CaCl2, and lysoPC, as indicated. EGTA (4 mM) was included in the samples incubated in the absence of added CaClz (lanes 1 and 3).
PKCLysophospholipids Regulation by
6869
TABLEI1 Effects of lyso-PC, diolein, oleic acid, and SDS, present singly or in combination, on PKC activity PKC was assayed under the standard conditions except for the inclusions of the agents, as indicated. The data presented are mean f S.E. of triplicate assays. The numbers in parentheses are percent stimulation compared to the basal value seen in the absence of lysoPC andother additions, which was taken as0% stimulation. PKC activity in the presence of lyso-PC (pg/ml)
PKC activators
4
0
8
pmol P/min I
0 5
6
7
8
9
PH
FIG. 9. Lyso-PC activation of PKC as a function of pH. The enzyme was assayed under the standard conditions except for the presence of lyso-PC (10 pglml) and the kinds of buffers (25 mM), as indicated. The values presented are means of duplicate assays, with assay errors less than 4%. The findings were confirmed in two separate sets of experiments.
v"d 0
0
2
4
6
8
1
0
None Diolein 0.3 pg/ml 3.0 pg/ml Oleic acid 4 pg/ml 8 pg/ml SDS 6 pg/ml 12 pg/ml
10.5 f 0.2 (0)
16.8 f 0.1 (60)
11.4 f 0.5 (9) 16.1 f 0.1 (53)
22.9 f.0.4 (118)" 30.8 f 0.2 (194) 29.7 f 1.4 (183)n32.7 f 1.0 (214)
21.8 f 0.9 (107) 24.9 f 0.5 (137) 36.2 f 0.3 (245) 32.6 f 0.7 (211) 29.5 f.0.3 (182)b 36.0 f 0.6 (244)b 14.3 f 0.4 (36) 20.4 f 0.1 (95)
micellar system. The assay system, essentially according to Hannun et at. (141, contained 0.3% Triton X-100,6 mol % of PS (8.4 molecules/micelle), 0.5 mol % of diolein (0.7 molecule/micelle) and 200 p M CaCL In addition, varying mol % of lyso-PC was also present, as indicated. The data presented are the means of triplicate assays. Similar results were obtained in two separate sets of experiments. One mol % of lyso-PC was calculated to be 1.4 molecules/micelle.
were explored in order to gain further insights into themodes of action of these agents. Lyso-PC, at a suboptimal concentration (4 wg/ml), acted synergistically with diolein, whereas an addition of the effect was noted when lyso-PC was present at an optimal stimulatory concentration of 8 pg/ml (Table 11); a suboptimal (0.3 pg/ml) andan optimalstimulating concentration (3 pg/ml) of diolein were used in these cases. In comparison, the stimulatory effect of lyso-PC at both concentrations was simply additive to the effect produced by a suboptimal concentration (4 pg/ml) of oleic acid; the combined stimulation, however, was lower than the theoretical sum of the effect when an optimal concentration (8 pg/ml) of oleic acid was used (Table 11). SDS, an ionic detergent, also stimulated PKC as did lyso-PC, diolein, and oleic acid, and, furthermore, the combined effect of suboptimal (6 pg/ml) and optimal concentration (12 pg/ml) of SDS with lyso-PC was similar to that seen for oleic acid (Table 11). These findings were consistent with the notion that the manner in which lyso-PC interacted with the PKC.PS.Ca2+ active ternary complex might resemble more to that of oleic acid (or SDS) than thatof diolein. Inhibition of PS/Ca'+-stimulated PKC activity by ALP (8), sphingosine (22), polymyxin B ( 2 3 ) , and tamoxifen (24) has
21.6 f.0.3 (106) 28.8 f 0.3 (174) 19.5 f 0.3 (86Ib 27.2 f 1.0 (159)b
Significantly higher 0, < 0.01) than the theoretical sum of the stimulatory effects of diolein and lyso-PC at theindicated concentrations. bSignificantly lower (p < 0.01) than the theoretical sum of the stimulatory effects of oleic acid (or SDS) and lyso-PC a t the indicated concentrations.
TABLE111 Effects of ALP, sphingosine, polymyxin B, and tamoxifen onPKC activity assayed in the presence or absence of lyso-PC PKC was assayed under the standard incubation conditions except for the additions of the agents, as indicated. The values presented are the mean f S.E. of triplicate assays. The numbers in the parentheses are percent of the respective control values seen in the absence of PKC inhibitors with or without lyso-PC, which were taken as 100%.
LYSO PC (mol %)
FIG.10. PKC regulation by lyso-PC as assayed using mixed
27.4 f 0.3 (162)
PKC activity PKC inhibitors
Lyso-PC None
(10pg/ml)
pmol P/min
None (control) ALP 10 p M 20 p M
Sphingosine 6 PM 15 p M Polymyxin B, 15 p M Tamoxifen 20 pM 60 p M
3.33 f 0.11 (100)
7.20 f. 0.17 (100)
2.56 f 0.03 (77) 0.98 f. 0.03 (29)
0.96 f. 0.02 (13) 0.53 f 0.01 (7)
1.75 f 0.01 (53) 1.11 f. 0.01 (33) 1.55 f 0.01 (47)
2.91 f 0.09 (40) 1.51 f 0.03 (21) 1.73 f 0.03 (24)
2.42 -C 0.12 (73) 1.42 f 0.02 (43)
1.99 f 0.07 (28) 0.65 f 0.01 (9)
been reported. It was observed here that the enzyme activity that was further stimulated by lyso-PC was more susceptible to inhibition by these agents than the activity in its absence (Table 111). Because each of the agents inhibited PKC competitively with respect to PS (8, 22-24) and lyso-PC stimulated the enzyme by increasing its affinity for PS (Fig. %I), it appeared that interactions of lyso-PC with PS could induce a conformational change on the PKC. PS . Ca2+complex, with which the inhibitors could bind and act more effectively. The action of lyso-PC on various protein kinases were examined. It was observed that thestimulatory effect of lysoPC was specific to PKC, because the activity of MLC kinase or CAMP-dependent protein kinase was not stimulated by it over a wide range of concentration (Fig. 11).The data indicated that a functional interactionof lyso-PC with the hydrophobic domain of calmodulin appeared to be absent in the MLC kinase reaction system. CAMP-dependent protein kinase, however, was inhibited about 60% by a high concentra-
PKC Regulation by Lysophospholipids
6870
1AGONISTS1
250 0 MLCK 0
200 0
\\\\\\\\\
0
8
&.&.\\L\>-\\\\\\\.
PI +PIP,
150
Sphingoglycolipids
v
>
t 2
100
I-
0
a Y
n
IPQ
50 0 0
1
3
10
30
100
DAG
SDhinaosine
FA
L y s o PC
Ca2+ mobilization
Lyso PC (pg/ml)
FIG. 11. Comparative effects of lyso-PC on different protein kinases. PKC, myosin light chain kinase (MLCK), and CAMPdependent protein kinase (APK) were assayed under the standard incubation conditions, except for the presence of varying concentrations of lyso-PC, as indicated. The activity values of the enzymes seen in the absence of lyso-PC were taken as100%.The values presented are means of duplicate assays, with assay errors less than 2%.Similar results were obtained in two separate sets of experiments.
tion (100 pg/ml) of lyso-PC (Fig. 11)and inhibited about20% under the standard PKCactivation conditions, i.e. CaC1, (200 p M ) plus Ps (10 pg/ml) (data not shown).
FIG.12. A scheme depicting membrane signal transduction systems that generate lipid second messengers positively or negatively regulating PKC. PZ, phosphatidylinositol; PZPz,phosphatidylinositol 4,5-bisphosphate; IPS, inositol 1,4,5-trisphosphate; DAG, diacylglycerol;PLC, phospholipase C; P U P phospholipase , AP; FA, fatty acids; +, activation; -, inhibition.
stimulation by lyso-PC required PS and was associated with an increased affinity for PS (Fig. 2 A ) , as seen previously for diacylglycerol(26) and phorbol ester (27). Although both lysoPC and diacylglycerol required Ca2+for their effects, lyso-PC DISCUSSION decreased (Fig. 2B), whereas diacylglycerol (26) and phorbol Several lines of recent evidence have clearly established ester (27) increased the Ca2+ affinity of the enzyme.Obthat phospholipase C acts on membrane inositol phospho- viously, the mechanisms of action of the threeclasses of lipid lipids to generate two separate arms of second messengers, second messengers were different from each other, a conclui.e. diacylglycerol that activates PKC, and inositol 1,4,5-tri- sion which was also partly supported by distinct combination phosphate that mobilizes Ca2+from internal storesleading to effects achieved by diolein plus lyso-PC (Fig. 1B, Table 11) or activation of calmodulin/Ca''-dependent enzymes and per- oleic acid plus lyso-PC (Table 11). It has been reported rehaps PKC as well (for reviews, see Ref. 6). More recently, cently that three distinct forms of rat brain PKC responded sphingosine (22) and several lysosphingolipids (25) have been differently to stimulation by unsaturated fatty acids (5). It is shown to be inhibitors of PKC, suggesting that metabolites of interest to examine whether such a specificity would also of membrane sphingoglycolipids can function as negative be found in stimulation and inhibition of the PKC isoforms effectors and that they may be causally related to the patho- by lysO-PC. The membrane content of PC is much higher than that of physiology of sphingolipidoses due to deficiencies in several metabolizing enzymes for sphingolipids and lysosphingolipids. phosphatidylinositol. Assuming that the activity levels of Another important development in signal transduction asso- phospholipase Az and phospholipase C are comparable, the ciated with membrane lipid catabolism is the reports that potential capacity of the cells to generate fatty acids and lysounsaturated fatty acids (e.g. arachidonic and oleic acid), pre- PC would be accordinglyfar greater than thatfor diacylglycsumably generated by the action of phospholipase A2, can erol. The concentration required for a maximal activation of activate PKC (1-5). The findings are of potential importance PKC by diacylglycerol (assumed to be diolein, M , = 621) was because they suggest the existence of yet another second about 5 p~ (26), by oleic acid (Mr = 283) was about 50 p~ (Refs. 1-5; also present studies), and by lyso-PC (oleoyl, M , messenger system for PKCactivation, in addition to the phospholipase C/phosphoinositide pathway mentioned above. = 522) was about 20 @M (Figs. 1,4, and 5).Because phosphoIn the present studies, we found that certain lysophospho- lipase A2 can generate two arms of the putative second messengers that directly activate PKC, one of which (lyso-PC) lipids (lyso-PC in particular) were PKC activators, further supporting arole of the phospholipase Az/PC system in signal having an activation potency comparable to that of diacyltransduction. The pathways for membrane lipid metabolism glycerol, coupled with the high membrane content of PC, it is and regulation of PKC are schematically presented below conceivable that the lyso-PC/fatty acid system would be at least as functionally important as the diacylglycerol system. (Fig. 12). One interesting aspect of lyso-PC (and other active lysoLyso-PC andcertainother lysophospholipids regulated PKC activity in a biphasic manner, i.e. they stimulated it at phospholipids) was its dual actions. The transient and suslow concentrations (30 p ~ ) .This is distinct from the ac- a matter of speculation. It is plausible, however, that a low tions of other PKC activators (diacylglyceroland fatty acids) concentration (30 activity even at very high concentrations. PKC activation by p ~ it ) can inhibit PKC stimulated by other activators. It is unclear whether the stimulation or inhibition of PKC afforded fatty acids was independent of PS and Ca2+, althoughgreater a stimulation was noted in their presence (4). In contrast, the by lyso-PC would be morecrucial in the regulation of cellular
PKC Regulation by Lysophospholipids activities; lyso-PC might represent aunique second messenger in that itcan subserve as a positive or negative effector in the signal transduction. If one considers only the phospholipase AJPC system, an equal number of molecules of fatty acid (assuming oleic acid) and lysophospholipid (assuming lysoPC) would be generated from PC hydrolysis. Because lysoPC was a more potent activator of PKC than oleic acid, it is likely that ata low activation state of the cells PKC would be stimulated almost exclusively by lyso-PC, whereas at a high activation state the stimulatory effect of oleic acid would be markedly negated by lyso-PC. In this respect, lyso-PC, in contrast tooleic acid and diacylglycerol, wouldlikely play the predominant role as a negative effector in PKC regulation. It has been shown recently that phospholipase Az,like phospholipase C, is activated by distinct GTP-binding proteins (28) and that phospholipase AP can be activated by physiologic concentrations (submicromolar) of Ca2+ (29). Receptor-mediated activation of phospholipase Az has been reported, for example, in leukemia HL-60 (30) and RBL-2H3 cells (31). Treatment of platelets with thrombin has been shown to cause a number of changes in phospholipid metabolism, including a decreased PC and an increased lyso-PC content (32). Serum stimulationof the serum-starved 3T3-Ll cells, in comparison, resulted in a rapid (within 20 s) decrease in lyso-PC level^.^ Although the changes seen in these cells are rather small (up to30%), the finding suggest that formation of lyso-PC is coupled to thereceptor-coupled phospholipase Az regulation and PC degradation. Further studies using various cell types and agonists would be required to establish lyso-PC as a second messenger, subserving transmembrane signaling via regulation of PKC activity. Perhapscertaincharacteristics of lyso-PC regulation of PKC deserve some comments. The ability of lyso-PC to stimulate PKC in phosphorylating histone H1, myelin basic protein, and two brain proteins, but not the other proteins tested, might be attributed to the unique binding of lyso-PC with these proteins named above, thus facilitating interactions of the lyso-PC-protein complexes with PKC or that of PKC with its activators PS and/or Ca2+.A role of substrate proteins in influencing Ca2+ and phospholipid requirements for PKC activation has been reported (33, 34). Phosphate (a polyanion) could conceivably interfere with such binding of lyso-PC with histone H1 or myelin basic protein(both strongly basic proteins and positively charged), resulting in a loss of lyso-PC stimulation when the enzyme was assayed in phosphate buffer. In addition to substrate proteins, lyso-PC interacted with PS, resulting in changes in the affinity of the enzyme for PS and Ca2+as well as the effects of other PKC activators diolein and oleicacid. PKCinhibition by high concentrations of lyso-PC, regardless of the substrates and activators used, might be due to the nonspecific detergentlike property of lysophospholipids. Further studies, however, are required to determine the precise mechanisms underlying stimulation and inhibition of PKC by lyso-PC. Identities of the two brain PKC substrate proteins whose phosphorylation was stimulated by lyso-PC are unknown. It C. J. Kuo, K. Oishi, and J. F. Kuo, unpublished observations.
6871
is of considerable interest if the 35-kDa protein should be identified as lipocortins, the endogenous phospholipase AQ inhibitors shown recently to be phosphorylated by PKC (35). A change in the activity of lipocortins caused by lyso-PCaugmented PKC-catalyzed phosphorylation would constitute an intriguing feedback mechanism having a profound biological implication. The 47-kDa substrate protein from brain might be related to GAP-43/pp46, a phosphoprotein shown to be a component of growth cones of extending axons (3638), where PKC has been shown to be heavily localized as determined immunocytochemically (39). REFERENCES 1. McPhail, L.C., Clayton, C.C., and Snyderman, R. (1984) Science 2 2 4 , 622-625 2. Murakami, K., and Routenberg, A. (1985) FEES Lett. 1 9 2 , 189-193 3. Leach, K. L., and Blumberg, P. M. (1985) Cancer Res. 4 6 , 1958-1963 4. Murakami, K., Chan, S. Y., and Routtenberg,A. (1986) J. Biol. Chem. 2 6 1 , 15424-15429 5. Sekiguchi, K., Tsukada, M., Ogita, K., Kikkawa, U., and Nishizuka, Y. (1987) Bwchem. Biophys. Res. Commun. 1 4 5 , 797-802 6. Nishizuka, Y. (1984) Nature 308,693-698 7. Berdel, W. E., Andressen, R., and Munder, P. G. (1985) in Phospholipids and Cellular Regulation (Kuo, J. F., ed) Vol. 2, pp. 41-74, CRC Press,
Roca Rnton. ~ . " , FT. --
8. Helfman, D. M., Barnes, K. C., Kinkade, J. M., Jr., Vogler, W. R., Shoji, M., and Kuo, J. F. (1983) Cancer Res. 43,2955-2961 9. Kiss. Z.. Deli. E.. Voeler. W. R.. and Kuo. J. F. (1987) . . Biochem. B ~ O D ~ V S . Res. Commun.'145, 661-666 ' 10. Giy@, P. R.,Mazzei, G. J., and Kuo, J. F. (1986) J. Biol. Chem. 261,370-
. _
~~
d 13
11. Kuo, J. F., Andersson, R. G. G., Wise, B. C., Mackerlova, L., Katoh, N., Shoji, M., and Wrenn, R. W. (1980) Proc. Natl. Acad. Sci. U. S. A . 77, 7039-7043 12. Wise, B. C., Raynor, R. L., and Kuo, J. F. (1982) J. Biol. Chem. 267,84818488 13. Huang, K.-P., Nakabayashi, H., and Huang, F. L. (1986) Proc. Natl. Acad. Sci. U. S. A . 83,8535-8539 14. Hannun, Y. A,, Loomis, C. R., and Bell, R. M. (1985) J. Biol. Chem. 2 6 0 , 10039-10043 15. Wrenn. R. W.. Katoh.. N... Wise.. B. C... and Kuo. J. F. (1980) J. Biol. Chem. 255; 12042112046 16. Shoji, M., Patrick, J. G., Davis, D. W., and Kuo, J. F. (1977) Biochem. J. 161,213-221 17. Blumenthal, D. K., and Stull, J. T. (1980) Biochemistry 19,5008-5014 18. Post, R. L., and Sen, A. K. (1967) Methods Enzymol. 1 0 , 773-775 19. Solaro, R. J., and Shiner, J. S. (1976) Circ. Res. 39,8-14 20. Takai, Y., Kishimoto, A., Iwasa, Y., Kawahara, Y., Mori, T., and Nishizuka, Y. (1979) J. Biol. Chem. 264,3692-3695 21. Schatzman, R. C., Raynor, R. L., Fritz, R. B., andKuo, J. F. (1983)Biochem. J. 209,435-443 22. Hannun, Y.A,, Loomis, C. R., Merrill, A. H., Jr., and Bell, R. M. (1986) J. Biol. Chem. 261,12604-12609 23. Mazzei, G. J., Katoh, N., and Kuo, J. F. (1982) Biochem. Biophys. Res. Commun. 1 0 9 , 1129-1133 24. Su. H.-D.. Mazzei. G. J.. Vonler.W. R.. and Kuo. J. F. (1985) . , Ewchem. Pharmdcol. 34,3649-3653" ' 25. Hannun, Y., and Bell, R. M. (1987) Science 2 3 5 , 670-674 26. Kishimoto, A,, Takai, Y., Mori, T., Kikkawa, U., and Nishizuka, Y. (1980) J . Bwl. Chem. 256,2273-2276 27. Castagna, M., Takai, Y., Kaibuchi, K., Sano, K., Kikkawa, U., and Nishizuka, Y. (1981) J. Biol. Chem. 2 6 6 , 7847-7851 28. Burch, R. M., Luini, A., and Axelrod, J. (1986) Proc. Natl. Acad. Sci. U. S. A . 8 3 , 7201-7205 29. Loeb, L. A., and Gross, R. W. (1986) J. Bid. Chem. 2 6 1 , 10467-10470 30. Billah, M. M., and Siegel, M. I. (1987) Biochem. Biophys. Res. Commun. 144,683-691 31. Garcia-Gil, M., and Siraganian, R. P. (1986) J. Imrnunol. 136, 259-263 32. Broekman, M. J., Ward, J. W., and Maecus, A. J. (1980) J . Clin. Inuest. 6 6 , 275-283 33. Bazzi, M. D., and Nelsestuen, G. L. (1987) Biochemistry 2 6 , 1974-1982 34. Bazzi, M. D., and Nelsestuen, G. L. (1987) Biochemistry 2 6 , 5002-5008 35. Khanna, N. C., Tokuda, M., and Waisman, D. M. (1987) Cell Calcium 8,
_"
21 7-228 ".
36. Katz, F., Ellis, L., and Pfennin er, K. H.(1985) J . Neurosci. 6, 1402-1411 37. Jacobson, R. D., Virag, I., and gkene, J. H. P. (1986) J. Neurosci. 6 , 18431855 38. Meiri, K. F., Pfenninger, K. H., and Willard, M. B. (1986) Proc. Natl. Acad. Sci. U. S. A . 83,3537-3541 39. Gifard, P. R., Wood, J. G., Freschi, J. E., and Kuo, J. F. (1988) Deu. Biol., In
press