- Email: [email protected]
tumor promoter periodate
ARCHIVES
Vol.
OF BIOCHEMISTRY
285, No. 2, March,
AND
BIOPHYSICS
pp. 382-387,
1991
COMMUNICATION Reversible Oxidative Activation and Inactivation Mitogen/Tumor Promoter Periodate Rayudu
Gopalakrishna*
J and Wayne
of Protein
Kinase C by the
B. Anderson?
*Department of Pharmacology and Nutrition, California 90033; and t Laboratory of Cellular
School of Medicine, University of Southern California, Los Angeles, Oncology, National Cancer Institute, Bethesda, Maryland 20892
Received
November
September
10, 1990, and in revised
form
21, 1990
latory domain and the ATP-binding site of the catalytic domain, and modifies the vicinal thiols present within these sites. This results in the formation of intramolecular disulfide bridge(s) within the regulatory domain or catalytic domain leading to either reversible activation or inactivation of PKC, respectively. Thus, oxidant mitogenttumor promoters such as periodate may be able to bypass normal transmembrane signalling systems to directly activate pathways involved in cellular regulation.
The oxidant mitogenjtumor promoter, periodate, was used to selectively modify either the regulatory domain or the catalytic domain of protein kinase C (PKC) to induce oxidative activation or inactivation of PKC, respectively. Periodate, at micromolar concentrations, modified the regulatory domain of PKC as determined by the loss of ability to stimulate kinase activity by Ca’+/ phospholipid, and also by the loss of phorbol ester binding. This modification resulted in an increase in Caa+/ phospholipid-independent kinase activity (oxidative activation) . However, at higher concentrations (> 100 FM) periodate also modified the catalytic domain, resulting in complete inactivation of PKC. The oxidative modification induced by low periodate concentrations (<0.5 mM) was completely reversed by a brief treatment with 2 mM dithiothreitol. In this aspect, the modification induced by periodate was different from that of the previously reported irreversible modification of PKC induced by HzOz. However, the inactivation of PKC induced by periodate at concentrations >l mM was not reversed by dithiothreitol. Among the phospholipids and ligands of the regulatory domain tested, only phosphatidylserine protected the regulatory domain from oxidative modification. In the presence of phosphatidylserine, the catalytic site was selectively modified by periodate, resulting in formation of a form of PKC that exhibited phorbol ester binding but not kinase activity. Both reversible and irreversible oxidative activation and inactivation of PKC also were observed in intact cells treated with periodate. Taken together these results suggest that periodate, by virtue of having a tetrahedral structure, binds to the phosphate-binding regions present within the phosphatidylserine-binding site of the regu-
Protein kinase C (PKC)’ may be activated by different mechanisms in response to various cellular transmembrane signals. Besides the reversible activation of PKC by Ca2+, phospholipids (PL) , diacylglycerol, unsaturated fatty acids, and phorbol esters, this enzyme also can be irreversibly activated by proteolytic removal of the regulatory domain to form the so called M-kinase (1-5). Recently we reported the irreversible activation of PKC by selective oxidative modification of the regulatory domain with H202 to generate a Ca’+/phospholipidindependent form of the enzyme (6). However, oxidative modification of the catalytic domain resulted in complete inactivation of the kinase (7). Thus, oxidants may play a role in both the up-regulation and the down-regulation of PKC. Terpenoid tumor promoters such as phorbol esters and mezerein have been shown to directly bind to PKC to reversibly activate this kinase ( 1-3). Other tumor promoters such as benzoyl peroxide, m-chlorobenzoic acid, and m-periodate are oxidants (6-10)) although their mechanism of action in tumor promotion is not well defined. Since PKC can be oxidatively activated by selective oxidative modification of the regulatory domain, it is conceivable that oxidant tumor promoters may act, in part, by inducing the oxidative activation of PKC to alter
A preliminary report of this work was presented Meeting of the American Society for Biochemistry ology, New Orleans, LA, June 4-7,199O. r To whom correspondence and reprint requests
’ Abbreviations bis(B-aminoethyl phosphatidylserine; itol.
at the 81st Annual and Molecular Bishould
be addressed.
0 1991
Academic
Press,
Inc.
used: PKC, protein kinase C; EGTA, ethylen glycol ether) N,N’-tetraacetic acid; PL, phospholipids; PS, PDBu, phorbol12,13-dibutyrate; DTT, dithiothre0003-9861/91
382 All
$3.00
Copyright 0 1991 by Academic Press, Inc. rights of reproduction in any form reserved.
PERIODATE-INDUCED
OXIDATIVE
MODIFICATION
some biological responses in a manner similar to phorbol esters. Thus, we have assessed the ability of m-periodate, an agent which is not only a tumor promoter but also is a mitogen for some cell types (9, 11, 12)) to alter PKC activity. In this communication we show that m-periodate can modify vicinal thiols present within the regulatory domain or the catalytic domain to induce either reversible activation or inactivation of PKC, respectively. MATERIALS
AND
METHODS
PKC was purified by phenyl-Sepharose hydrophobic chromatography as described previously (6). Enzyme having an activity of 600 units/ mg protein was employed in the studies presented. Consideration of proteases present in the PKC preparation, as well as N-chloramines present in the water, has been dealt with in previous reports (6, 7). Oxidatiue modification of PKC. 2-Mercaptoethanol present in the PKC preparation was removed by dialysis (7). Since the PKC protein concentration used was low, oxidative modifications were conducted in polystyrene test tubes to reduce nonspecific adsorption of PKC to the test tube. To the PKC preparation (0.5 to 1 unit) were added 1 mM EGTA or 5 mM CaClz, along with other ligands such as ATP (0.1 mM), Mg2+ (8 mM), and PS, as indicated, in a total volume 0.25 ml. These samples were brought to room temperature (20°C) in a waterbath. Oxidative modification was initiated by the addition of the indicated concentration of sodium m-periodate and the samples were further incubated for 5 min. To the treated samples, 250 ~1 of 10 mM Tris-HCl, pH 7.5, was added and the samples were placed on ice. To the samples containing high concentrations (20.1 mM) of periodate, bovine serum albumin (0.1 mg/ml) was added, and the samples were then subjected to the centrifuge column technique to remove periodate ( 13). Both PKC activity (in the presence and absence of Ca’+/PL) (14) and [3H] phorbol 12,13-dibutyrate ( [ 3H] PDBu) binding (2) were determined immediately. Other experimental details are given in the figure legends.
RESULTS
AND
DISCUSSION
Enzymes that possess binding sites for ligands with a phosphate group have been shown to have vicinal thiols at such binding sites ( 15). Periodate, by virtue of its tetrahedral structure, is able to bind to phosphate-binding sites of this type and modify the vicinal thiols to form an intramolecular disulfide bridge ( 15 ) . These oxidation-susceptible sites can be protected from periodate by appropriate phosphate-containing ligands ( 15). Since periodate can selectively modify phosphate residuebinding sites in protein, it has been used in this study to selectively modify either the ATP-binding site in the catalytic domain, or the PS-binding site3 in the regulatory domain of PKC, by protecting the second site with appropriate ligand. Exposure of isolated PKC to low concentrations (
OF
PROTEIN
KINASE
C
383
p 0.6
16
5
14 s
$0.6
12B
E 4 e $ 0.4 c
10 c” G 6 E 0
3 .E s 0.2 z P
Ia
0
6: 2 0.05
0.1
0.2
0.3 0.40.5
Periodate
1.0
2
3
4
5
( mM)
FIG. 1. Effect of periodate treatment on the kinase activity and PDBu binding of PKC. Purified PKC in the presence of 1 mM EGTA was treated with various concentrations of periodate for 5 min at 2O”C, and then the periodate was removed by the centrifuge column technique ( 13). High concentrations of periodate, left remaining with the treated PKC preparations, can subsequently oxidize the ATP present in the kinase reaction mixture. Then the oxidized ATP formed can bind to the ATP-binding site of PKC and can inactivate the kinase activity. PKC activity was measured using histone (Sigma type III-S) as the substrate (6). Both forms of protein kinase activities were expressed in units, where 1 unit of enzyme transfers 1 nmol of phosphate to histone Hl per minute at 30°C. Specific phorbol ester binding was determined with [3H] PDBu as the ligand (2). (0) PKC activity in the presence of Ca*+/PL; (0) PKC activity in the absence of Ca*+/PL; (A) PDBu binding.
crease in Ca2+/PL-independent activity was 30 to 45% of the initial Ca2+/PL-stimulated activity of the unmodified control enzyme. The data indicate that the Ca’+/PL-independent enzyme has a specific activity lower than that of the Ca’+/PLdependent enzyme. Treatment of isolated PKC with higher concentrations (>l mM) of periodate resulted in the complete loss of both Ca’+/PL-stimulated activity and PDBu binding, along with the loss of Ca’+/PL-independent kinase activity. These results suggest that higher concentrations of periodate induced uncontrolled modification of PKC, which resulted in the complete inactivation of the enzyme. Since PKC is more susceptible to oxidative modification induced by H,Oz in the presence of Ca’+ (6)) the extent of PKC modification induced with periodate in the presence and absence of Ca2+ was compared. The degree of activation or inactivation of PKC induced by periodate in the presence of 5 mM Ca2+ was of the same magnitude as that obtained by modification induced in the absence of Ca’+, but it was observed at a two- to threefold lower concentration of periodate when carried out in the presence of Ca2+ (data not shown). This supports our earlier suggestion that oxidation-susceptible sites in both the regulatory and the catalytic domains may be more accessible in the Ca2+bound conformation of PKC (6). Effect of thiol reagents in reversing periodate-induced modifications. Since periodate is specific for vicinal thiols at low (PM) concentrations and can induce disulfide bridge formation among the vicinal thiols (15), it was of interest to determine whether the observed oxidative activation or inactivation of PKC induced by periodate might be reversed by thiol reagents. Initially PKC was treated with increasing concentrations of periodate for 5 min in the absence of thiol compounds and the oxidatively mod-
384
GOPALAKRISHNA
AND TABLE
Effect
of DTT
on the Periodate-Induced Without
Protein Periodate
Cas+/PL
0.81 0.50 0.42 0.22 0.11 0.02
I
Changes in Protein
Kinase
Activity
No Ca’+/PL
With Protein PDBu
binding
kinase
Ca’+/PL
Binding
DTT
10.4 3.0 0.9 0.4 N.D. N.D.
treatment
activity No Ca’+/PL
’ On the basis of previous studies of periodate-induced modification of other proteins( 18), and also on the present studies of PKC modification by low concentrations of periodate, the observed reversible changes in PKC might be mediated by vicinal thiol modification. However, further studies are required to determine the number of disulfide bridges that are formed and their location in the primary sequence of PKC. Because of limited amounts of PKC that can be purified to homogeniety and also because of complications due to the presence of various isoenzymes of PKC present within the purified preparations, such studies are beyond the scope of this investigation.
PDBu
binding
(pm4 0.04 0.04 0.05 0.04 0.04 0.02
0.87 0.73 0.66 0.60 0.39 0.02
10.8 8.6 7.9 7.1 4.0 N.D.
Note. Two sets of PKC samples were treated with the indicated concentrations of periodate for 5 min at 20°C. DTT then was added to one set of the treated samples, and both sets of samples were incubated on ice for 1 h. Then, protein binding, remaining in the treated samples, were determined. N.D. not detectable.
ified enzyme then was treated with DTT (5 mM) or 2-mercaptoethanol (10 mM) for 1 to 12 h at 4°C. The results presented in Table I indicated that DTT treatment for 1 h reversed by up to 80% the effect induced by 50 pM periodate. Concentrations of DTT as low as 0.1 mM can reverse the periodate-induced effect by 30% within 5 min. This ready reversibility of the periodate-induced oxidative modification of PKC suggests that this modification may involve disulfide bridge formation. Although other oxidatively modified amino acid residues such as methionine sulfoxides can be reduced by DTT, this requires treatment with substantially higher (20 mM) concentrations of DTT and at higher temperatures (100°C) for prolonged periods of time.’ In this respect the modification of PKC induced by periodate was different from that of our previously described irreversible modification of PKC induced by H202 (6). However, the inactivation of PKC induced by periodate at >1 mM concentration was not reversed by DTT even with incubation in the presence of DTT for 12 h. This suggests that periodate at low concentrations can induce reversible oxidative activation or inactivation of PKC, whereas at high concentrations periodate further modifies other oxidation susceptible amino acid residues in the PKC leading to irreversible (not readily reversible) oxidative inactivation of the enzyme. Protection of the catalytic site. Attempts were made to selectively protect the catalytic site from oxidative modification by treating PKC with periodate in the presence of the substrate ATP/Mg2+. In the presence of both ATP (0.1 mM) and Mg2+ (8 mM) , modification of PKC required slightly higher concen-
of PKC
(units)
bmol) 0.04 0.30 0.37 0.16 0.09 0.02
and in PDBu
treatment
activity
(units)
bM) 0 0.05 0.1 0.5 1.0 4.0
kinase
DTT
ANDERSON
(5 mM final concentration) kinase activity and PDBu
trations of periodate. However, ATP/Mg2+ did not significantly protect the kinase activity from inactivation by periodate, either in the presence or in the absence of Ca2+ (data not shown). In this respect, periodate-induced oxidative modification of PKC differs from H,Oa-induced oxidative modification of PKC, where it previously was reported that ATP/Mg’+ could protect the catalytic site from H ,O x (6). It may be that periodate, at higher concentrations, can oxidize the ribose residue of ATP (16) and the oxidized ATP then can bind to and modify the ATP-binding site to inactivate PKC. Protection of regulatory domain. As shown in Fig. 2A, PKC rapidly lost its ability to bind PDBu upon treatment with periodate at concentrations as low as 50 pM in the presence of 5 mM Ca’+. However, in the presence of PS (20 pg/ml), no significant loss of PDBu binding was observed even with treatment at concentrations of periodate up to 500 NM. Nevertheless, at higher concentrations of periodate (>l mM) , PS was not able to protect the protein to retain PDBu binding. While PS was able to protect the PKC from low (
Ca2+
mM)
concentrations
present.
However,
in the
of phosphate
absence
of Ca’+,
did significantly
high
(20
reduce the
PERIODATE-INDUCED
OXIDATIVE
MODIFICATION
I :
I
.\
\ No PS
Periodate (mM)
I
I
,
I
#
1
I
I
I
I
I
2
4
6
8
10
12
14
16
18
20
22
Phosphatidylserine
@g/ml)
FIG. 2. Effect of phosphatidylserine on the periodate-induced loss of PDBu binding. A, PKC, in the presence and absence of PS (20 pg/ml) , was treated with various concentrations of periodate. B, PKC was treated in the presence of various concentrations of PS with a fixed concentration of periodate (0.5 mM) . These modifications were carried out in the presence of 5 mM CaCl*. PKC samples initially were treated with PS for 5 min at 20°C and then periodate was added, and the samples were then further incubated for an additional 5 min.
effect of low (PM) concentrations of periodate concentrations of periodate. In analogy with other enzymes involved in phosphorylated compounds ( 15)) PKC may in the binding sites for phosphate residues However, periodate may have a higher affinity site in the regulatory domain. In our previous
but not high (mM) binding of various have vicinal thiols of PS and ATP.’ for the PS-binding study, we showed
s In the primary sequence of PKC isoenzymes, the Cl constant region present within the regulatory domain contains a tandem repeat of a cysteine-rich sequence that is analogous to the “Zn*+-finger” found in some metalloproteins and DNA-binding proteins (30-32). The PS micelle may bind in close proximity to this cysteine-rich Cl region. The C3 constant region of the PKC isoenzymes has a putative ATP-binding site with one cysteine residue. The other putative ATP-binding site is present in the C4 constant region, which has four to six cysteine residues (30-32). In the tertiary structure of PKC, some of the cysteinyl residues may be in sufficient proximity to form disulfide bridges.
OF
PROTEIN
KINASE
C
385
that the oxidation-susceptible site in the catalytic domain is not readily exposed unless PKC is activated (6). In the absence of PS, periodate may bind to vicinal thiols present in the PS-binding site(s) and induce the formation of a disulfide bridge(s) in the regulatory domain. In turn, this may alter the conformation of PKC to an active form similar, but not identical, to that obtained with the binding of Ca2+/PL activators to the regulatory domain. This mechanism might enable the conversion of PKC to a catalytically active form without the requirement for Ca2+ and PL. However, periodate at higher concentrations also may bind to vicinal thiols present in the ATP-binding site that may become accessible with the initial stage of oxidative activation of PKC by periodate. In the event that PS is already bound to PKC prior to the addition of periodate, the periodate would be unable to bind to this site. However, under this condition, periodate could selectively bind to the ATP-binding site that becomes accessible by PS-mediated activation of the enzyme and thereby induce inactivation of PKC. Oxidative activation and inactivation of PKC in intact cells treated with periodate. We examined whether periodate could induce this oxidative modification of PKC within intact B16 melanoma, C-6 glioma, and A 431 cells. Control cells, and cells treated with periodate (1 to 5 mM) for 10 to 120 min, were rapidly harvested and homogenized in the presence of a detergent to obtain total (cytosolic and particulate) PKC. The detergentsolubilized cell extracts then were subjected to DEAE-cellulose chromatography (6). Two peaks of PKC activity were eluted with chromatographic separation of control cell extract (Fig. 3A). Peak I was eluted with 0.1 M NaCl and exhibited Ca’+/PL dependence. Peak II, eluted with 0.25 M NaCl, was found to contain higher amounts of Ca2+/PL-independent activity. As shown in Fig. 3B, treatment of B16 melanoma cells with 3 mM periodate for 10 min resulted in a pronounced decrease in Ca2+/PL-stimulated PKC activity (an 80% loss from peak I and a 40% loss from peak II). However, no increase in Ca”‘/ PL-independent activity was observed in peak II. Previously, we have shown that the Ca’+/PL-independent activity, obtained by HzOz modification of PKC, was eluted in peak II (6). When the periodate-treated cells were homogenized in the presence of 2 mM DTT, Ca2+/PL-stimulated activity was recovered completely in both peaks I and II. This indicated that periodatemodified peak I PKC was present in the cell extract, but either could not be detected, or was not eluted in the peak I and II regions, unless it was reduced prior to application to DEAEcellulose. No increase in kinase activity was observed in the unbound, flow-through protein fraction collected from the DEAE-cellulose column. Furthermore, no Ca2+/PL-stimulated kinase activity was observed by treating the eluted fractions (peak I and II) with DTT. This suggested that the modified PKC might remain tightly bound to the DEAE-cellulose when the column was eluted with 0.25 M NaCl. Attempts to elute this modified form of PKC with high salt (0.5 M NaCl) were complicated by the elution in this fraction of proteins such as calmodulin and S-100 which have been shown to inhibit PKC activity (17). Therefore, attempts were made to elute the tightly bound modified PKC after first reducing it with DTT. Thus, the DEAE-cellulose column was eluted further with 0.25 M NaCl in the presence of 2 mM DTT. This fraction, designated as peak III, was eluted at a slow flow rate to allow the reduction of the enzyme.
An
increase
PDBu binding)
in Ca’+/PL-independent
was noted in this fraction
activity
(but
with chromatographic
no
386
GOPALAKRISHNA A. Control
1.8
1.2
(no periodate) II
m
1
1 A
z 0.6 : .c Y .E 0,
C. Periodate
Fraction
i
treatment
followed
8 number
10
by DTT
12
14
FIG. 3. DEAE-cellulose chromatography of native and oxidatively modified forms of PKC isolated from untreated and periodate-treated B16 melanoma with periodate.
cells. A, PKC activity profile from control cells not treated B, PKC activity profile from cells treated with 3 mM
periodate for 10 min and then extracted; the extract was chromatographed initially in the absence of DTT. C, PKC activity profile from cells treated with periodate (3 mM, 10 min) but extracted and incubated with 2 mM DTT for 1 h at 4°C prior to chromatographic separation in the presence of 0.1 mM DTT. B16 melanoma cells were grown in 150-
cm* flasks in Dulbecco’s modified Eagle’s medium containing 10% fetal calf serum. Confluent cells were treated with periodate in serum-free medium. The cells were then homogenized in buffer A (20 mM TrisHCl, pH 7.5/l mM EDTA/0.5 mM phenylmethylsulfonyl fluoride/20 pM leupeptin/0.15 mM pepstatin A) containing 1% Nonidet-P40 and the homogenate was centrifuged at 13,000g for 10 min. The detergentsolubilized cell extract (5 ml; approx. 25 mg protein) was applied to a l-ml DEAE-cellulose (DE 52, Whatman) column (0.8 X 2 cm) previously equilibrated with buffer B (20 mM Tris-HCl, pH 7.5/l mM EDTA/ 20 j,tM leupeptin/0.15 mM pepstatin A). After the column was washed with 5 ml of buffer B, the bound proform of PKC was eluted with 3 ml of 0.1 M NaCl in buffer B (peak I). Peak II activity was then eluted with 2.5 ml of 0.25 M NaCl in buffer B. The tightly bound protein kinase was eluted with 0.25 M NaCl along with 5 mM DTT (peak III). At this stage the flow rate was reduced to ensure optimal reduction of PKC. Fractions of 0.5 ml were collected and the PKC activity present in these fractions was measured in the presence of Ca*‘/PL (0) or 1 mM EGTA
with no added Ca2+/PL(0).
separation of the extract obtained from periodate-treated cells (Fig. 3B). However, the increase in Ca*+/PL-independent activity found in this fraction was only 20% of the Ca*+/PL-stimulated activity lost from peak I. This low recovery of PKC activity in this fraction may be due to intrinsically low catalytic
AND ANDERSON activity of the oxidatively activated PKC, or it may be due to a low recovery of tightly bound PKC even after reduction by DTT. Even though fraction III was left in the presence of DTT for 2 to 12 h after elution, no reversal of the modification occurred as determined by the lack of appearance of Ca*+/PL-stimulated activity. In contrast, periodate modification induced in the purified form of PKC was readily reversed with DTT. Furthermore, the modification of PKC which occurred in periodate-treated cells also was readily reversed when the crude homogenates were treated with DTT. However, during elution from the DEAEcellulose column, the presence of DTT partially reversed the modification of PKC, which resulted in the elution of the enzyme from the column as peak III. The presence of DTT was not able to completely convert the oxidatively modified protein to the native proform of PKC. Taken together this data suggests that the PKC modification which occurred in intact cells may be similar, but not identical, to the modification observed with the purified enzyme. The PKC modification which occurs in periodate-treated cells may involve the formation of multiple disulfide bridges, and reversal of these modifications may require other enzymes or cofactors present in the crude homogenate. This reversal of PKC modification in crude extracts prepared from periodate-treated cells also was obtained with 10 mM 2mercaptoethanol. Although purified PKC was modified with low (PM) concentrations of periodate, with intact cells high concentrations (1 to 5 mM) of periodate were required to induce the oxidative modification of PKC. It may be possible that the concentration of periodate reached inside the cell is quite low (PM) and thus induces only the reversible modification of PKC. Prolonged treatment of cells with periodate for 2 to 3 h resulted in a 30 to 40% decrease in PKC activity, which was not restored by treatment with 2 mM DTT. This down-regulation with prolonged periodate treatment may be due to increased uptake of periodate into the cells, or due to increased proteolytic susceptibility of the oxidatively modified PKC. Significance of periodate-induced PKC oxidative activation. Conceivably, readily reducible oxidative modification of PKC may occur in cells under normal or pathophysiological conditions in response to oxidative stress. Recently, activation of PKC by redoxy-cycling quinones was reported in hepatocytes’( 18). An increase in phosphorylation of proteins was reported in periodate-treated lymphocytes (19). However, the protein kinase (s) responsible for this increase in endogenous protein phosphorylation has not yet been identified. Mercapto compounds such as DTT and 2-mercaptoethanol are often added to buffers used in cell homogenization and in column chromatography. The oxidative changes that occur in PKC by disulfide bridge formation may not be noticed if these reducing agents are present in the buffers. Also because of the increased susceptibility of oxidatively modified proteins to proteolysis (20, 21) , oxidative modification also may increase the turnover of PKC in the cell. Alternatively, the oxidative modification of PKC may be reversed by reducing enzyme systems such as those of ‘Although the modification of thiol groups was implicated in this study (18)) quinones induced additional stimulation of the Ca’+/PLdependent kinase activity without having any effect on Ca’+/PL-independent PKC activity and phorbol ester binding. Thus, our study on periodate-induced Ca’+/PL-independent activation of PKC is clearly different from that of the quinone-induced modification of PKC.
PERIODATE-INDUCED
OXIDATIVE
MODIFICATION
thioredoxin/glutaredoxin (22)) to regenerate the proform of PKC. This would act to switch the PKC regulation to the control of transmembrane signals such as Ca’+, diacylglycerol, or arachidonic acid. In this context, it is interesting to note that phosphoinositide-specific phospholipase C has a thioredoxin-like domain ( 23 ) . Terpenoid tumor promoters such as phorbol esters and mezerein induce not only the activation of PKC but also the downregulation of PKC (24, 25). The present observations suggest that the oxidant tumor promoter, periodate, even though it is structurally unrelated to terpenoid tumor promoters, can induce both the activation and the inactivation of PKC, but by a different mechanism. In this regard, phorbol ester treatment of cells has been shown to enhance the generation of reactive oxygen species (26). Other oxidant tumor promoters such as benzoyl peroxide, and m-chlorobenzoic acid, irrespective of structural differences, may induce the oxidative activation and subsequent inactivation of PKC to act, in part, through a mechanism common to the terpenoid tumor promoters. Periodate also can serve as a mitogen for some cell types (9, 11, 12). Although periodate action may involve modification of cell-surface sugar residues (9, 11, 12), it also may have intracellular sites of action. Several growth regulatory factors are known to induce phosphatidylinositol turnover to regulate cellular processes through the generation of intracellular messengers such as inositol polyphosphates, Ca2+, and diacylglycerol. However, recently it was reported that periodate has no effect on phosphatidylinositol turnover in the lymphocytes (27). The present observations suggest that oxidants such as periodate may bypass transmembrane signalling systems to directly activate pathways that are normally regulated by second messengers.
We thank Mr. Stephen Ross and Mr. Jonathan Lee for their excellent technical assistance. This work was supported by USPHS Grant CA 47142 from National Cancer Institute, Grant RT 388 from TobaccoRelated Disease Research Program, University of California, Basic Research Support Grant from PHS, Grant IN-21-30 from American Cancer Society.
KINASE
Y. (1986) Science 233,305-311. E., Kuhn, L. J., and Vandenbark, G. R. ( 1983) Proc. N&l. USA 80, 36-40. N., Leach, K., and Blumberg, P. M. (1984) Proc. N&l. USA 81,607-610. L. C., Clayton, C. C., and Snyderman, R. (1984) Science P. J., Nairn,
A. C., and Kuo,
J. F. (1988)
Science
2,
387
C
6. Gopalakrishna, R., and Anderson, Sci. USA 86,6758-6762.
W. B. (1989)
Proc.
7. Gopalakrishna, 233-237.
W. B. (1987)
FEBS
R., and Anderson,
Natl.
Acad.
Lett.
225,
8. Slaga, T. J., Klein-Szanto, A. J. P., Triplett, L. L., Yotti, P., and Trosko, J. E. (1981) Science 213, 1023-1025. 9. Srinivas, L., and Colburn, N. H. (1984) Curcinogenesti 5,515-519. 10. Cerutti, P. ( 1985) Science 227,375381. 11. Novogrodsky, A., and Katchalski, E. (1971) 300.
FEBS
Z&t.
12. O’Brien, R. L., and Parker, J. W. (1976) Cell 7, 13-20. 13. Gopalakrishna, R., and Nagarajan, B. (1978) Biochem. 70-75.
12, 297-
Med.
22,
14. Gopalakrishna, R., Barsky, S. H., Thomas, T. P., and Anderson, W. B. (1986) J. Biol. Chem. 261,16,438-16,445. 15. Rippa, M., Bellini, T., Signorini, M., and Dallocchio, F. (1981) J. Biol. Chem. 256,450-458. 16. Coleman,
R. F. ( 1983)
Annu.
Reu. Biochem.
17. Albert, K., Wu, W. C. -S., Nairn, Proc. Natl. Acad. Sci. USA 81, 18. Kass, G. E. N., Duddy, 260,499-507.
62,67-91.
A. C., and Greengard,
P. (1984)
3622-3625.
S. K., and Orrenius,
S. (1989)
Biochem.
J.
19. Malinin, G. I., Hornicek, F. J., and Malinin, T. I. (1989) Biochem. Biophys. Res. Commun. 158.590-594. 20. Levine, R. L., Oliver, C. N., Fulks, R. M., and Stadtman, E. R. (1981) Proc. Natl. Acad. Sci. USA 7, 2120-2124. 21. Davies,
K. J. A. (1987)
22. Holmgren,
A. (1989)
J. Biol. J. Biol.
Chem. Chem.
262,9895-9901. 264,
23. Bennet, F. C., Balcarek, (1988) Nature (London)
J. M., Varrichio, 334,268-270.
24. Ballester, 15,199.
0. M. ( 1985)
R., and Rosen,
26. Goldstein, W. ( 1981)
13,963-13,966. A., and Crooke,
J. Biol.
B. D., Witz, G., Amoruso, Cancer Z&t. 11. 257-262.
M.,
30. Kikkawa, U., Ogita, K., Ono, Y., Asaoka, T., Ase, K., Sekiguchi, K., and Nishizuka, 223, 212-216.
260,15,194-
O., and Horecker, D. S., and Troll, T. I., and Malinin,
154,606-612.
C. R., and Bell, R. M. (1985) D. E., Jr. ( 1989)
S. T.
128,331-338.
Stone,
27. Hornicek, F., Reese, D. H., Lo, H. K., Malinin, G. I., ( 1988) Biochem. Biophys. Res. Commun. 28. Hannun, Y. A., Loomis, 260, 10,039-10,043.
Chem.
S., Michetti, M., Sacco, Biophys. Res. Commun.
29. Newton, A. C., and Koshland, 14,909-14,915.
REFERENCES
5. Blackshear, 2957-2967.
PROTEIN
25. Melloni, E., Pontremoli, B. L. (1985) Biochem.
ACKNOWLEDGMENTS
1. Nishizuka, 2. Niedel, J. Acad. Sci. 3. Sharkey, Acad. Sci. 4. McPhail, 224,622-624.
OF
J. Biol.
J. Biol. Chem.
Chem. 264,
Y., Shearman, M., Fujii, Y. (1987) FEBS Z&t.
31. Housey, G. M., O’Brian, A. A., Johnson, M. D., Kirschmeier, P., and Weinstein, I. B. (1987) Proc. N&l. Acad. Sci. USA 84, 10651069. 32. Parker, P. J., Coussens, E., Stabel, S., Waterfield, 853-859.
L., Totty, N., Rhee, L., Young, S., Chen, M. D., and Ullich, A. (1986) Science 233,
Vol.
OF BIOCHEMISTRY
285, No. 2, March,
AND
BIOPHYSICS
pp. 382-387,
1991
COMMUNICATION Reversible Oxidative Activation and Inactivation Mitogen/Tumor Promoter Periodate Rayudu
Gopalakrishna*
J and Wayne
of Protein
Kinase C by the
B. Anderson?
*Department of Pharmacology and Nutrition, California 90033; and t Laboratory of Cellular
School of Medicine, University of Southern California, Los Angeles, Oncology, National Cancer Institute, Bethesda, Maryland 20892
Received
November
September
10, 1990, and in revised
form
21, 1990
latory domain and the ATP-binding site of the catalytic domain, and modifies the vicinal thiols present within these sites. This results in the formation of intramolecular disulfide bridge(s) within the regulatory domain or catalytic domain leading to either reversible activation or inactivation of PKC, respectively. Thus, oxidant mitogenttumor promoters such as periodate may be able to bypass normal transmembrane signalling systems to directly activate pathways involved in cellular regulation.
The oxidant mitogenjtumor promoter, periodate, was used to selectively modify either the regulatory domain or the catalytic domain of protein kinase C (PKC) to induce oxidative activation or inactivation of PKC, respectively. Periodate, at micromolar concentrations, modified the regulatory domain of PKC as determined by the loss of ability to stimulate kinase activity by Ca’+/ phospholipid, and also by the loss of phorbol ester binding. This modification resulted in an increase in Caa+/ phospholipid-independent kinase activity (oxidative activation) . However, at higher concentrations (> 100 FM) periodate also modified the catalytic domain, resulting in complete inactivation of PKC. The oxidative modification induced by low periodate concentrations (<0.5 mM) was completely reversed by a brief treatment with 2 mM dithiothreitol. In this aspect, the modification induced by periodate was different from that of the previously reported irreversible modification of PKC induced by HzOz. However, the inactivation of PKC induced by periodate at concentrations >l mM was not reversed by dithiothreitol. Among the phospholipids and ligands of the regulatory domain tested, only phosphatidylserine protected the regulatory domain from oxidative modification. In the presence of phosphatidylserine, the catalytic site was selectively modified by periodate, resulting in formation of a form of PKC that exhibited phorbol ester binding but not kinase activity. Both reversible and irreversible oxidative activation and inactivation of PKC also were observed in intact cells treated with periodate. Taken together these results suggest that periodate, by virtue of having a tetrahedral structure, binds to the phosphate-binding regions present within the phosphatidylserine-binding site of the regu-
Protein kinase C (PKC)’ may be activated by different mechanisms in response to various cellular transmembrane signals. Besides the reversible activation of PKC by Ca2+, phospholipids (PL) , diacylglycerol, unsaturated fatty acids, and phorbol esters, this enzyme also can be irreversibly activated by proteolytic removal of the regulatory domain to form the so called M-kinase (1-5). Recently we reported the irreversible activation of PKC by selective oxidative modification of the regulatory domain with H202 to generate a Ca’+/phospholipidindependent form of the enzyme (6). However, oxidative modification of the catalytic domain resulted in complete inactivation of the kinase (7). Thus, oxidants may play a role in both the up-regulation and the down-regulation of PKC. Terpenoid tumor promoters such as phorbol esters and mezerein have been shown to directly bind to PKC to reversibly activate this kinase ( 1-3). Other tumor promoters such as benzoyl peroxide, m-chlorobenzoic acid, and m-periodate are oxidants (6-10)) although their mechanism of action in tumor promotion is not well defined. Since PKC can be oxidatively activated by selective oxidative modification of the regulatory domain, it is conceivable that oxidant tumor promoters may act, in part, by inducing the oxidative activation of PKC to alter
A preliminary report of this work was presented Meeting of the American Society for Biochemistry ology, New Orleans, LA, June 4-7,199O. r To whom correspondence and reprint requests
’ Abbreviations bis(B-aminoethyl phosphatidylserine; itol.
at the 81st Annual and Molecular Bishould
be addressed.
0 1991
Academic
Press,
Inc.
used: PKC, protein kinase C; EGTA, ethylen glycol ether) N,N’-tetraacetic acid; PL, phospholipids; PS, PDBu, phorbol12,13-dibutyrate; DTT, dithiothre0003-9861/91
382 All
$3.00
Copyright 0 1991 by Academic Press, Inc. rights of reproduction in any form reserved.
PERIODATE-INDUCED
OXIDATIVE
MODIFICATION
some biological responses in a manner similar to phorbol esters. Thus, we have assessed the ability of m-periodate, an agent which is not only a tumor promoter but also is a mitogen for some cell types (9, 11, 12)) to alter PKC activity. In this communication we show that m-periodate can modify vicinal thiols present within the regulatory domain or the catalytic domain to induce either reversible activation or inactivation of PKC, respectively. MATERIALS
AND
METHODS
PKC was purified by phenyl-Sepharose hydrophobic chromatography as described previously (6). Enzyme having an activity of 600 units/ mg protein was employed in the studies presented. Consideration of proteases present in the PKC preparation, as well as N-chloramines present in the water, has been dealt with in previous reports (6, 7). Oxidatiue modification of PKC. 2-Mercaptoethanol present in the PKC preparation was removed by dialysis (7). Since the PKC protein concentration used was low, oxidative modifications were conducted in polystyrene test tubes to reduce nonspecific adsorption of PKC to the test tube. To the PKC preparation (0.5 to 1 unit) were added 1 mM EGTA or 5 mM CaClz, along with other ligands such as ATP (0.1 mM), Mg2+ (8 mM), and PS, as indicated, in a total volume 0.25 ml. These samples were brought to room temperature (20°C) in a waterbath. Oxidative modification was initiated by the addition of the indicated concentration of sodium m-periodate and the samples were further incubated for 5 min. To the treated samples, 250 ~1 of 10 mM Tris-HCl, pH 7.5, was added and the samples were placed on ice. To the samples containing high concentrations (20.1 mM) of periodate, bovine serum albumin (0.1 mg/ml) was added, and the samples were then subjected to the centrifuge column technique to remove periodate ( 13). Both PKC activity (in the presence and absence of Ca’+/PL) (14) and [3H] phorbol 12,13-dibutyrate ( [ 3H] PDBu) binding (2) were determined immediately. Other experimental details are given in the figure legends.
RESULTS
AND
DISCUSSION
Enzymes that possess binding sites for ligands with a phosphate group have been shown to have vicinal thiols at such binding sites ( 15). Periodate, by virtue of its tetrahedral structure, is able to bind to phosphate-binding sites of this type and modify the vicinal thiols to form an intramolecular disulfide bridge ( 15 ) . These oxidation-susceptible sites can be protected from periodate by appropriate phosphate-containing ligands ( 15). Since periodate can selectively modify phosphate residuebinding sites in protein, it has been used in this study to selectively modify either the ATP-binding site in the catalytic domain, or the PS-binding site3 in the regulatory domain of PKC, by protecting the second site with appropriate ligand. Exposure of isolated PKC to low concentrations (
OF
PROTEIN
KINASE
C
383
p 0.6
16
5
14 s
$0.6
12B
E 4 e $ 0.4 c
10 c” G 6 E 0
3 .E s 0.2 z P
Ia
0
6: 2 0.05
0.1
0.2
0.3 0.40.5
Periodate
1.0
2
3
4
5
( mM)
FIG. 1. Effect of periodate treatment on the kinase activity and PDBu binding of PKC. Purified PKC in the presence of 1 mM EGTA was treated with various concentrations of periodate for 5 min at 2O”C, and then the periodate was removed by the centrifuge column technique ( 13). High concentrations of periodate, left remaining with the treated PKC preparations, can subsequently oxidize the ATP present in the kinase reaction mixture. Then the oxidized ATP formed can bind to the ATP-binding site of PKC and can inactivate the kinase activity. PKC activity was measured using histone (Sigma type III-S) as the substrate (6). Both forms of protein kinase activities were expressed in units, where 1 unit of enzyme transfers 1 nmol of phosphate to histone Hl per minute at 30°C. Specific phorbol ester binding was determined with [3H] PDBu as the ligand (2). (0) PKC activity in the presence of Ca*+/PL; (0) PKC activity in the absence of Ca*+/PL; (A) PDBu binding.
crease in Ca2+/PL-independent activity was 30 to 45% of the initial Ca2+/PL-stimulated activity of the unmodified control enzyme. The data indicate that the Ca’+/PL-independent enzyme has a specific activity lower than that of the Ca’+/PLdependent enzyme. Treatment of isolated PKC with higher concentrations (>l mM) of periodate resulted in the complete loss of both Ca’+/PL-stimulated activity and PDBu binding, along with the loss of Ca’+/PL-independent kinase activity. These results suggest that higher concentrations of periodate induced uncontrolled modification of PKC, which resulted in the complete inactivation of the enzyme. Since PKC is more susceptible to oxidative modification induced by H,Oz in the presence of Ca’+ (6)) the extent of PKC modification induced with periodate in the presence and absence of Ca2+ was compared. The degree of activation or inactivation of PKC induced by periodate in the presence of 5 mM Ca2+ was of the same magnitude as that obtained by modification induced in the absence of Ca’+, but it was observed at a two- to threefold lower concentration of periodate when carried out in the presence of Ca2+ (data not shown). This supports our earlier suggestion that oxidation-susceptible sites in both the regulatory and the catalytic domains may be more accessible in the Ca2+bound conformation of PKC (6). Effect of thiol reagents in reversing periodate-induced modifications. Since periodate is specific for vicinal thiols at low (PM) concentrations and can induce disulfide bridge formation among the vicinal thiols (15), it was of interest to determine whether the observed oxidative activation or inactivation of PKC induced by periodate might be reversed by thiol reagents. Initially PKC was treated with increasing concentrations of periodate for 5 min in the absence of thiol compounds and the oxidatively mod-
384
GOPALAKRISHNA
AND TABLE
Effect
of DTT
on the Periodate-Induced Without
Protein Periodate
Cas+/PL
0.81 0.50 0.42 0.22 0.11 0.02
I
Changes in Protein
Kinase
Activity
No Ca’+/PL
With Protein PDBu
binding
kinase
Ca’+/PL
Binding
DTT
10.4 3.0 0.9 0.4 N.D. N.D.
treatment
activity No Ca’+/PL
’ On the basis of previous studies of periodate-induced modification of other proteins( 18), and also on the present studies of PKC modification by low concentrations of periodate, the observed reversible changes in PKC might be mediated by vicinal thiol modification. However, further studies are required to determine the number of disulfide bridges that are formed and their location in the primary sequence of PKC. Because of limited amounts of PKC that can be purified to homogeniety and also because of complications due to the presence of various isoenzymes of PKC present within the purified preparations, such studies are beyond the scope of this investigation.
PDBu
binding
(pm4 0.04 0.04 0.05 0.04 0.04 0.02
0.87 0.73 0.66 0.60 0.39 0.02
10.8 8.6 7.9 7.1 4.0 N.D.
Note. Two sets of PKC samples were treated with the indicated concentrations of periodate for 5 min at 20°C. DTT then was added to one set of the treated samples, and both sets of samples were incubated on ice for 1 h. Then, protein binding, remaining in the treated samples, were determined. N.D. not detectable.
ified enzyme then was treated with DTT (5 mM) or 2-mercaptoethanol (10 mM) for 1 to 12 h at 4°C. The results presented in Table I indicated that DTT treatment for 1 h reversed by up to 80% the effect induced by 50 pM periodate. Concentrations of DTT as low as 0.1 mM can reverse the periodate-induced effect by 30% within 5 min. This ready reversibility of the periodate-induced oxidative modification of PKC suggests that this modification may involve disulfide bridge formation. Although other oxidatively modified amino acid residues such as methionine sulfoxides can be reduced by DTT, this requires treatment with substantially higher (20 mM) concentrations of DTT and at higher temperatures (100°C) for prolonged periods of time.’ In this respect the modification of PKC induced by periodate was different from that of our previously described irreversible modification of PKC induced by H202 (6). However, the inactivation of PKC induced by periodate at >1 mM concentration was not reversed by DTT even with incubation in the presence of DTT for 12 h. This suggests that periodate at low concentrations can induce reversible oxidative activation or inactivation of PKC, whereas at high concentrations periodate further modifies other oxidation susceptible amino acid residues in the PKC leading to irreversible (not readily reversible) oxidative inactivation of the enzyme. Protection of the catalytic site. Attempts were made to selectively protect the catalytic site from oxidative modification by treating PKC with periodate in the presence of the substrate ATP/Mg2+. In the presence of both ATP (0.1 mM) and Mg2+ (8 mM) , modification of PKC required slightly higher concen-
of PKC
(units)
bmol) 0.04 0.30 0.37 0.16 0.09 0.02
and in PDBu
treatment
activity
(units)
bM) 0 0.05 0.1 0.5 1.0 4.0
kinase
DTT
ANDERSON
(5 mM final concentration) kinase activity and PDBu
trations of periodate. However, ATP/Mg2+ did not significantly protect the kinase activity from inactivation by periodate, either in the presence or in the absence of Ca2+ (data not shown). In this respect, periodate-induced oxidative modification of PKC differs from H,Oa-induced oxidative modification of PKC, where it previously was reported that ATP/Mg’+ could protect the catalytic site from H ,O x (6). It may be that periodate, at higher concentrations, can oxidize the ribose residue of ATP (16) and the oxidized ATP then can bind to and modify the ATP-binding site to inactivate PKC. Protection of regulatory domain. As shown in Fig. 2A, PKC rapidly lost its ability to bind PDBu upon treatment with periodate at concentrations as low as 50 pM in the presence of 5 mM Ca’+. However, in the presence of PS (20 pg/ml), no significant loss of PDBu binding was observed even with treatment at concentrations of periodate up to 500 NM. Nevertheless, at higher concentrations of periodate (>l mM) , PS was not able to protect the protein to retain PDBu binding. While PS was able to protect the PKC from low (
Ca2+
mM)
concentrations
present.
However,
in the
of phosphate
absence
of Ca’+,
did significantly
high
(20
reduce the
PERIODATE-INDUCED
OXIDATIVE
MODIFICATION
I :
I
.\
\ No PS
Periodate (mM)
I
I
,
I
#
1
I
I
I
I
I
2
4
6
8
10
12
14
16
18
20
22
Phosphatidylserine
@g/ml)
FIG. 2. Effect of phosphatidylserine on the periodate-induced loss of PDBu binding. A, PKC, in the presence and absence of PS (20 pg/ml) , was treated with various concentrations of periodate. B, PKC was treated in the presence of various concentrations of PS with a fixed concentration of periodate (0.5 mM) . These modifications were carried out in the presence of 5 mM CaCl*. PKC samples initially were treated with PS for 5 min at 20°C and then periodate was added, and the samples were then further incubated for an additional 5 min.
effect of low (PM) concentrations of periodate concentrations of periodate. In analogy with other enzymes involved in phosphorylated compounds ( 15)) PKC may in the binding sites for phosphate residues However, periodate may have a higher affinity site in the regulatory domain. In our previous
but not high (mM) binding of various have vicinal thiols of PS and ATP.’ for the PS-binding study, we showed
s In the primary sequence of PKC isoenzymes, the Cl constant region present within the regulatory domain contains a tandem repeat of a cysteine-rich sequence that is analogous to the “Zn*+-finger” found in some metalloproteins and DNA-binding proteins (30-32). The PS micelle may bind in close proximity to this cysteine-rich Cl region. The C3 constant region of the PKC isoenzymes has a putative ATP-binding site with one cysteine residue. The other putative ATP-binding site is present in the C4 constant region, which has four to six cysteine residues (30-32). In the tertiary structure of PKC, some of the cysteinyl residues may be in sufficient proximity to form disulfide bridges.
OF
PROTEIN
KINASE
C
385
that the oxidation-susceptible site in the catalytic domain is not readily exposed unless PKC is activated (6). In the absence of PS, periodate may bind to vicinal thiols present in the PS-binding site(s) and induce the formation of a disulfide bridge(s) in the regulatory domain. In turn, this may alter the conformation of PKC to an active form similar, but not identical, to that obtained with the binding of Ca2+/PL activators to the regulatory domain. This mechanism might enable the conversion of PKC to a catalytically active form without the requirement for Ca2+ and PL. However, periodate at higher concentrations also may bind to vicinal thiols present in the ATP-binding site that may become accessible with the initial stage of oxidative activation of PKC by periodate. In the event that PS is already bound to PKC prior to the addition of periodate, the periodate would be unable to bind to this site. However, under this condition, periodate could selectively bind to the ATP-binding site that becomes accessible by PS-mediated activation of the enzyme and thereby induce inactivation of PKC. Oxidative activation and inactivation of PKC in intact cells treated with periodate. We examined whether periodate could induce this oxidative modification of PKC within intact B16 melanoma, C-6 glioma, and A 431 cells. Control cells, and cells treated with periodate (1 to 5 mM) for 10 to 120 min, were rapidly harvested and homogenized in the presence of a detergent to obtain total (cytosolic and particulate) PKC. The detergentsolubilized cell extracts then were subjected to DEAE-cellulose chromatography (6). Two peaks of PKC activity were eluted with chromatographic separation of control cell extract (Fig. 3A). Peak I was eluted with 0.1 M NaCl and exhibited Ca’+/PL dependence. Peak II, eluted with 0.25 M NaCl, was found to contain higher amounts of Ca2+/PL-independent activity. As shown in Fig. 3B, treatment of B16 melanoma cells with 3 mM periodate for 10 min resulted in a pronounced decrease in Ca2+/PL-stimulated PKC activity (an 80% loss from peak I and a 40% loss from peak II). However, no increase in Ca”‘/ PL-independent activity was observed in peak II. Previously, we have shown that the Ca’+/PL-independent activity, obtained by HzOz modification of PKC, was eluted in peak II (6). When the periodate-treated cells were homogenized in the presence of 2 mM DTT, Ca2+/PL-stimulated activity was recovered completely in both peaks I and II. This indicated that periodatemodified peak I PKC was present in the cell extract, but either could not be detected, or was not eluted in the peak I and II regions, unless it was reduced prior to application to DEAEcellulose. No increase in kinase activity was observed in the unbound, flow-through protein fraction collected from the DEAE-cellulose column. Furthermore, no Ca2+/PL-stimulated kinase activity was observed by treating the eluted fractions (peak I and II) with DTT. This suggested that the modified PKC might remain tightly bound to the DEAE-cellulose when the column was eluted with 0.25 M NaCl. Attempts to elute this modified form of PKC with high salt (0.5 M NaCl) were complicated by the elution in this fraction of proteins such as calmodulin and S-100 which have been shown to inhibit PKC activity (17). Therefore, attempts were made to elute the tightly bound modified PKC after first reducing it with DTT. Thus, the DEAE-cellulose column was eluted further with 0.25 M NaCl in the presence of 2 mM DTT. This fraction, designated as peak III, was eluted at a slow flow rate to allow the reduction of the enzyme.
An
increase
PDBu binding)
in Ca’+/PL-independent
was noted in this fraction
activity
(but
with chromatographic
no
386
GOPALAKRISHNA A. Control
1.8
1.2
(no periodate) II
m
1
1 A
z 0.6 : .c Y .E 0,
C. Periodate
Fraction
i
treatment
followed
8 number
10
by DTT
12
14
FIG. 3. DEAE-cellulose chromatography of native and oxidatively modified forms of PKC isolated from untreated and periodate-treated B16 melanoma with periodate.
cells. A, PKC activity profile from control cells not treated B, PKC activity profile from cells treated with 3 mM
periodate for 10 min and then extracted; the extract was chromatographed initially in the absence of DTT. C, PKC activity profile from cells treated with periodate (3 mM, 10 min) but extracted and incubated with 2 mM DTT for 1 h at 4°C prior to chromatographic separation in the presence of 0.1 mM DTT. B16 melanoma cells were grown in 150-
cm* flasks in Dulbecco’s modified Eagle’s medium containing 10% fetal calf serum. Confluent cells were treated with periodate in serum-free medium. The cells were then homogenized in buffer A (20 mM TrisHCl, pH 7.5/l mM EDTA/0.5 mM phenylmethylsulfonyl fluoride/20 pM leupeptin/0.15 mM pepstatin A) containing 1% Nonidet-P40 and the homogenate was centrifuged at 13,000g for 10 min. The detergentsolubilized cell extract (5 ml; approx. 25 mg protein) was applied to a l-ml DEAE-cellulose (DE 52, Whatman) column (0.8 X 2 cm) previously equilibrated with buffer B (20 mM Tris-HCl, pH 7.5/l mM EDTA/ 20 j,tM leupeptin/0.15 mM pepstatin A). After the column was washed with 5 ml of buffer B, the bound proform of PKC was eluted with 3 ml of 0.1 M NaCl in buffer B (peak I). Peak II activity was then eluted with 2.5 ml of 0.25 M NaCl in buffer B. The tightly bound protein kinase was eluted with 0.25 M NaCl along with 5 mM DTT (peak III). At this stage the flow rate was reduced to ensure optimal reduction of PKC. Fractions of 0.5 ml were collected and the PKC activity present in these fractions was measured in the presence of Ca*‘/PL (0) or 1 mM EGTA
with no added Ca2+/PL(0).
separation of the extract obtained from periodate-treated cells (Fig. 3B). However, the increase in Ca*+/PL-independent activity found in this fraction was only 20% of the Ca*+/PL-stimulated activity lost from peak I. This low recovery of PKC activity in this fraction may be due to intrinsically low catalytic
AND ANDERSON activity of the oxidatively activated PKC, or it may be due to a low recovery of tightly bound PKC even after reduction by DTT. Even though fraction III was left in the presence of DTT for 2 to 12 h after elution, no reversal of the modification occurred as determined by the lack of appearance of Ca*+/PL-stimulated activity. In contrast, periodate modification induced in the purified form of PKC was readily reversed with DTT. Furthermore, the modification of PKC which occurred in periodate-treated cells also was readily reversed when the crude homogenates were treated with DTT. However, during elution from the DEAEcellulose column, the presence of DTT partially reversed the modification of PKC, which resulted in the elution of the enzyme from the column as peak III. The presence of DTT was not able to completely convert the oxidatively modified protein to the native proform of PKC. Taken together this data suggests that the PKC modification which occurred in intact cells may be similar, but not identical, to the modification observed with the purified enzyme. The PKC modification which occurs in periodate-treated cells may involve the formation of multiple disulfide bridges, and reversal of these modifications may require other enzymes or cofactors present in the crude homogenate. This reversal of PKC modification in crude extracts prepared from periodate-treated cells also was obtained with 10 mM 2mercaptoethanol. Although purified PKC was modified with low (PM) concentrations of periodate, with intact cells high concentrations (1 to 5 mM) of periodate were required to induce the oxidative modification of PKC. It may be possible that the concentration of periodate reached inside the cell is quite low (PM) and thus induces only the reversible modification of PKC. Prolonged treatment of cells with periodate for 2 to 3 h resulted in a 30 to 40% decrease in PKC activity, which was not restored by treatment with 2 mM DTT. This down-regulation with prolonged periodate treatment may be due to increased uptake of periodate into the cells, or due to increased proteolytic susceptibility of the oxidatively modified PKC. Significance of periodate-induced PKC oxidative activation. Conceivably, readily reducible oxidative modification of PKC may occur in cells under normal or pathophysiological conditions in response to oxidative stress. Recently, activation of PKC by redoxy-cycling quinones was reported in hepatocytes’( 18). An increase in phosphorylation of proteins was reported in periodate-treated lymphocytes (19). However, the protein kinase (s) responsible for this increase in endogenous protein phosphorylation has not yet been identified. Mercapto compounds such as DTT and 2-mercaptoethanol are often added to buffers used in cell homogenization and in column chromatography. The oxidative changes that occur in PKC by disulfide bridge formation may not be noticed if these reducing agents are present in the buffers. Also because of the increased susceptibility of oxidatively modified proteins to proteolysis (20, 21) , oxidative modification also may increase the turnover of PKC in the cell. Alternatively, the oxidative modification of PKC may be reversed by reducing enzyme systems such as those of ‘Although the modification of thiol groups was implicated in this study (18)) quinones induced additional stimulation of the Ca’+/PLdependent kinase activity without having any effect on Ca’+/PL-independent PKC activity and phorbol ester binding. Thus, our study on periodate-induced Ca’+/PL-independent activation of PKC is clearly different from that of the quinone-induced modification of PKC.
PERIODATE-INDUCED
OXIDATIVE
MODIFICATION
thioredoxin/glutaredoxin (22)) to regenerate the proform of PKC. This would act to switch the PKC regulation to the control of transmembrane signals such as Ca’+, diacylglycerol, or arachidonic acid. In this context, it is interesting to note that phosphoinositide-specific phospholipase C has a thioredoxin-like domain ( 23 ) . Terpenoid tumor promoters such as phorbol esters and mezerein induce not only the activation of PKC but also the downregulation of PKC (24, 25). The present observations suggest that the oxidant tumor promoter, periodate, even though it is structurally unrelated to terpenoid tumor promoters, can induce both the activation and the inactivation of PKC, but by a different mechanism. In this regard, phorbol ester treatment of cells has been shown to enhance the generation of reactive oxygen species (26). Other oxidant tumor promoters such as benzoyl peroxide, and m-chlorobenzoic acid, irrespective of structural differences, may induce the oxidative activation and subsequent inactivation of PKC to act, in part, through a mechanism common to the terpenoid tumor promoters. Periodate also can serve as a mitogen for some cell types (9, 11, 12). Although periodate action may involve modification of cell-surface sugar residues (9, 11, 12), it also may have intracellular sites of action. Several growth regulatory factors are known to induce phosphatidylinositol turnover to regulate cellular processes through the generation of intracellular messengers such as inositol polyphosphates, Ca2+, and diacylglycerol. However, recently it was reported that periodate has no effect on phosphatidylinositol turnover in the lymphocytes (27). The present observations suggest that oxidants such as periodate may bypass transmembrane signalling systems to directly activate pathways that are normally regulated by second messengers.
We thank Mr. Stephen Ross and Mr. Jonathan Lee for their excellent technical assistance. This work was supported by USPHS Grant CA 47142 from National Cancer Institute, Grant RT 388 from TobaccoRelated Disease Research Program, University of California, Basic Research Support Grant from PHS, Grant IN-21-30 from American Cancer Society.
KINASE
Y. (1986) Science 233,305-311. E., Kuhn, L. J., and Vandenbark, G. R. ( 1983) Proc. N&l. USA 80, 36-40. N., Leach, K., and Blumberg, P. M. (1984) Proc. N&l. USA 81,607-610. L. C., Clayton, C. C., and Snyderman, R. (1984) Science P. J., Nairn,
A. C., and Kuo,
J. F. (1988)
Science
2,
387
C
6. Gopalakrishna, R., and Anderson, Sci. USA 86,6758-6762.
W. B. (1989)
Proc.
7. Gopalakrishna, 233-237.
W. B. (1987)
FEBS
R., and Anderson,
Natl.
Acad.
Lett.
225,
8. Slaga, T. J., Klein-Szanto, A. J. P., Triplett, L. L., Yotti, P., and Trosko, J. E. (1981) Science 213, 1023-1025. 9. Srinivas, L., and Colburn, N. H. (1984) Curcinogenesti 5,515-519. 10. Cerutti, P. ( 1985) Science 227,375381. 11. Novogrodsky, A., and Katchalski, E. (1971) 300.
FEBS
Z&t.
12. O’Brien, R. L., and Parker, J. W. (1976) Cell 7, 13-20. 13. Gopalakrishna, R., and Nagarajan, B. (1978) Biochem. 70-75.
12, 297-
Med.
22,
14. Gopalakrishna, R., Barsky, S. H., Thomas, T. P., and Anderson, W. B. (1986) J. Biol. Chem. 261,16,438-16,445. 15. Rippa, M., Bellini, T., Signorini, M., and Dallocchio, F. (1981) J. Biol. Chem. 256,450-458. 16. Coleman,
R. F. ( 1983)
Annu.
Reu. Biochem.
17. Albert, K., Wu, W. C. -S., Nairn, Proc. Natl. Acad. Sci. USA 81, 18. Kass, G. E. N., Duddy, 260,499-507.
62,67-91.
A. C., and Greengard,
P. (1984)
3622-3625.
S. K., and Orrenius,
S. (1989)
Biochem.
J.
19. Malinin, G. I., Hornicek, F. J., and Malinin, T. I. (1989) Biochem. Biophys. Res. Commun. 158.590-594. 20. Levine, R. L., Oliver, C. N., Fulks, R. M., and Stadtman, E. R. (1981) Proc. Natl. Acad. Sci. USA 7, 2120-2124. 21. Davies,
K. J. A. (1987)
22. Holmgren,
A. (1989)
J. Biol. J. Biol.
Chem. Chem.
262,9895-9901. 264,
23. Bennet, F. C., Balcarek, (1988) Nature (London)
J. M., Varrichio, 334,268-270.
24. Ballester, 15,199.
0. M. ( 1985)
R., and Rosen,
26. Goldstein, W. ( 1981)
13,963-13,966. A., and Crooke,
J. Biol.
B. D., Witz, G., Amoruso, Cancer Z&t. 11. 257-262.
M.,
30. Kikkawa, U., Ogita, K., Ono, Y., Asaoka, T., Ase, K., Sekiguchi, K., and Nishizuka, 223, 212-216.
260,15,194-
O., and Horecker, D. S., and Troll, T. I., and Malinin,
154,606-612.
C. R., and Bell, R. M. (1985) D. E., Jr. ( 1989)
S. T.
128,331-338.
Stone,
27. Hornicek, F., Reese, D. H., Lo, H. K., Malinin, G. I., ( 1988) Biochem. Biophys. Res. Commun. 28. Hannun, Y. A., Loomis, 260, 10,039-10,043.
Chem.
S., Michetti, M., Sacco, Biophys. Res. Commun.
29. Newton, A. C., and Koshland, 14,909-14,915.
REFERENCES
5. Blackshear, 2957-2967.
PROTEIN
25. Melloni, E., Pontremoli, B. L. (1985) Biochem.
ACKNOWLEDGMENTS
1. Nishizuka, 2. Niedel, J. Acad. Sci. 3. Sharkey, Acad. Sci. 4. McPhail, 224,622-624.
OF
J. Biol.
J. Biol. Chem.
Chem. 264,
Y., Shearman, M., Fujii, Y. (1987) FEBS Z&t.
31. Housey, G. M., O’Brian, A. A., Johnson, M. D., Kirschmeier, P., and Weinstein, I. B. (1987) Proc. N&l. Acad. Sci. USA 84, 10651069. 32. Parker, P. J., Coussens, E., Stabel, S., Waterfield, 853-859.
L., Totty, N., Rhee, L., Young, S., Chen, M. D., and Ullich, A. (1986) Science 233,