Phorbol 12-Myristate 13-Acetate-dependent Protein Kinase Cδ-Tyr311 Phosphorylation in Cardiomyocyte Caveolae*

Protein kinase Cδ (PKCδ) activation is generally attributed to lipid cofactor-dependent allosteric activation mechanisms at membranes. However, recent studies indicate that PKCδ also is dynamically regulated through tyrosine phosphorylation in H2O2- and phorbol 12-myristate 13-acetate (PMA)-treated cardiomyocytes. H2O2 activates Src and related Src-family kinases (SFKs), which function as dual PKCδ-Tyr311 and -Tyr332 kinases in vitro and contribute to H2O2-dependent PKCδ-Tyr311/Tyr332 phosphorylation in cardiomyocytes and in mouse embryo fibroblasts. H2O2-dependent PKCδ-Tyr311/Tyr332 phosphorylation is defective in SYF cells (deficient in SFKs) and restored by Src re-expression. PMA also promotes PKCδ-Tyr311 phosphorylation, but this is not associated with SFK activation or PKCδ-Tyr332 phosphorylation. Rather, PMA increases PKCδ-Tyr311 phosphorylation by delivering PKCδ to SFK-enriched caveolae. Cyclodextrin treatment disrupts caveolae and blocks PMA-dependent PKCδ-Tyr311 phosphorylation, without blocking H2O2-dependent PKCδ-Tyr311 phosphorylation. The enzyme that acts as a PKCδ-Tyr311 kinase without increasing PKCδ phosphorylation at Tyr332 in PMA-treated cardiomyocytes is uncertain. Although in vitro kinase assays implicate c-Abl as a selective PKCδ-Tyr311 kinase, PMA-dependent PKCδ-Tyr311 phosphorylation persists in cardiomyocytes treated with the c-Abl inhibitor ST1571 and c-Abl is not detected in caveolae; these results effectively exclude a c-Abl-dependent process. Finally, we show that 1,2-dioleoyl-sn-glycerol mimics the effect of PMA to drive PKCδ to caveolae and increase PKCδ-Tyr311 phosphorylation, whereas G protein-coupled receptor agonists such as norepinephrine and endothelin-1 do not. These results suggest that norepinephrine and endothelin-1 increase 1,2-dioleoyl-sn-glycerol accumulation and activate PKCδ exclusively in non-caveolae membranes. Collectively, these results identify stimulus-specific PKCδ localization and tyrosine phosphorylation mechanisms that could be targeted for therapeutic advantage.

Protein kinase C-␦ (PKC␦) 2 is a serine/threonine kinase that plays a vital role in signaling pathways that regulate cardiac contraction, ischemic preconditioning, and the pathogenesis of cardiac hypertrophy and failure (1). Traditional models of PKC␦ activation have focused on lipid cofactor-dependent mechanisms that localize PKC␦ in its active conformation to membranes. However, recent studies identify regulatory phosphorylation at Thr 505 in the "activation loop" segment (a region flanked by the highly conserved DFG and APE sequences) as an additional mechanism that contributes to the dynamic control of PKC␦ activity in some cellular contexts. Although activation loop phosphorylation is a stable modification that plays a critical role to structure other PKC isoforms in a favorable conformation for catalysis, PKC␦ is catalytically active even without activation loop Thr 505 phosphorylation. Rather, PKC␦-Thr 505 phosphorylation plays a unique role as a dynamically regulated process that is increased by PMA and certain agonist-activated receptors and contributes to the control of PKC␦ activity and/or substrate specificity (1)(2)(3)(4)(5)(6).
PKC␦ also is phosphorylated at Tyr 311 and Tyr 332 , two tyrosine residues that are unique to the hinge region of PKC␦ (and not conserved in other PKC isoforms). Our recent studies focused on mechanisms that regulate PKC␦-Tyr 311 phosphorylation, showing that oxidative stress resulting from H 2 O 2 treatment leads to the release of PKC␦ from membranes and a global increase in PKC␦-Tyr 311 phosphorylation in both soluble and particulate subcellular compartments. H 2 O 2 -dependent PKC␦-Tyr 311 phosphorylation is via a PP1-sensitive mechanism; this is presumed to reflect a role for Src or a related PP1-sensitive Src family kinase (SFK), because these enzymes are activated in H 2 O 2 -treated cardiomyocytes. PMA also increases PKC␦-Tyr 311 phosphorylation via a PP1-sensitive pathway. However, PMA-dependent PKC␦-Tyr 311 phosphorylation is via a different signaling pathway that is confined to the membrane fraction and is not associated with a global increase in SFK activity. In vitro studies suggest that PMA increases PKC␦-Tyr 311 phosphorylation by inducing a conformational change that renders PKC␦ a better substrate for phosphorylation by Src (7,8). According to this formulation, PMA promotes PKC␦-Tyr 311 phosphorylation by delivering the enzyme in an active conformation to a Src-enriched membrane fraction. Because caveolae have been identified as signaling domains for SFKs in other cell types, and our previous studies demonstrated that PMA delivers PKC␦ to the caveolae com-partment (9), this study examines the role of caveolae as platforms for cross-talk between PKC␦ and SFKs in cardiomyocytes. Insofar as PKC␦-Tyr 311 phosphorylation is predicted to be associated with a coordinate increase in PKC␦ phosphorylation at Tyr 332 (a site that also is a target for in vitro Src-dependent phosphorylation (7)), and PKC␦-Tyr 332 phosphorylation may impart functionally important properties (as a consensus binding sequence for the SH2 domain of the adapter protein Shc (10) or to influence PKC␦ proteolytic cleavage, which contributes to the induction of apoptosis (11)), the mechanisms that regulate PKC␦-Tyr 332 phosphorylation also were examined.
Cell Culture-Cardiomyocytes were isolated from the hearts of 2-day-old Wistar rats by a trypsin dispersion procedure using a differential attachment procedure to enrich for cardiomyocytes followed by irradiation as described previously (2,12,13). The yield of cardiomyocytes typically is 2.5-3 ϫ 10 6 cells per neonatal ventricle. Cells were plated on protamine sulfatecoated culture dishes at a density of 5 ϫ 10 6 cells/100-mm dish. Experiments were performed on cultures grown for 5 days in minimal essential medium (Invitrogen) supplemented with 10% fetal calf serum and then serum deprived for the subsequent 24 h. SYF and Src ϩ cells obtained from ATCC were cultured in Dulbecco's modified Eagle's medium containing 10% fetal calf serum and 100 g/ml hygromycin at 37°C, in a 5% CO 2 atmosphere.
Preparation of Soluble and Particulate or Detergent-insoluble Fractions-Soluble and particulate fractions were prepared according to methods published previously (7). In brief, cells were washed with phosphate-buffered saline and then transferred to ice-cold homogenization buffer (20 mM Tris-HCl, pH 7.5, 2 mM EDTA, 2 mM EGTA, 6 mM ␤-mercaptoethanol, 50 g/ml aprotinin, 48 g/ml leupeptin, 5 M pepstatin A, 1 mM phenylmethylsulfonyl fluoride, 0.1 mM sodium vanadate, and 50 mM sodium fluoride), lysed by sonication, and centrifuged at 100,000 ϫ g for 1 h. The supernatant was saved as the soluble fraction and the particulate fraction was solubilized in SDS-PAGE sample buffer.
Preparation of Caveolae Membranes-Fractions enriched in the muscle-specific caveolin-3 isoform were prepared according to a detergent-free purification scheme described previously (9,14). All steps were carried out at 4°C. Briefly, cells from five 100-mm diameter dishes were washed twice with ice-cold phosphate-buffered saline and then scraped into 0.5 M sodium carbonate, pH 11.0 (0.5 ml/dish). Cells from five dishes were combined (total volume, 2.5 ml) for each preparation. The extract was sequentially disrupted by homogenization with a loose-fitting Dounce homogenizer (10 strokes), a Polytron tissue grinder (three 10-s bursts), and a tip sonicator (three 20-s bursts). The homogenate was then adjusted to 40% sucrose by adding an equal volume of 80% sucrose prepared in MES-buffered saline (25 mM MES, pH 6.5, and 0.15 M NaCl), placed on the bottom of an ultracentrifuge tube, overlaid with a 5-30% discontinuous sucrose gradient (3 ml of 5% sucrose and 4 ml of 35% sucrose, both in MES-buffered saline containing 0.25 M sodium carbonate), and centrifuged at 260,000 ϫ g for 16 to 18 h in a SW40 rotor (Beckman Coulter, Palo Alto, CA). After centrifugation, 13 1-ml fractions were collected. A pooled caveolae fraction (fractions 4 -5, containing all of the buoyant caveolin-3 immunoreactivity and 0.5-1% total starting cell protein), a pooled fraction 8 -13 (F8 -13, which contains the bulk of the cellular material including the cytosol and most of the particulate membrane fraction), and the insoluble pellet (P, which is solubilized in SDS-PAGE sample buffer) were subjected to SDS-PAGE and immunoblotting. The caveolin-3-enriched membrane fraction isolated according to this method is biochemically distinct from the surrounding phospholipid bilayer and is operationally defined as caveolae in this study. However, it is important to note that this buoyant membrane fraction undoubtedly contains both true caveolae (specialized lipid raft membranes that contain caveolin and form invaginations at, or vesicles close to, the surface membrane) and morphologically featureless lipid rafts (that coexist and may even associate with caveolae (15)). Biochemical methods to separate these distinct membrane subdomains and experiments to resolve their discrete cellular functions are beyond the scope of this study.
Immunoprecipitation and Immunoblot Analysis-Immunoblotting on lysates or immunoprecipitated PKC␦ was according to methods described previously or the manufacturer's instructions (2,7,9). All anti-PKC antibodies (including the phosphosite specific antibodies (PSSAs) that specifically recognize PKC␦ phosphorylation at Thr 505 , Tyr 311 , and Tyr 332 ) have been validated (2). Of note, the PKC␦-Tyr(P) 311 antibody is a highly specific reagent that can be used to track PKC␦ phosphorylation (at endogenous levels of enzyme expression) in experiments on cell lysates. The anti-PKC␦-Tyr(P) 332 antibody is less specific and requires immunoprecipitation for studies of endogenous PKC␦ phosphorylation. In each figure, each panel represents the results from a single gel (exposed for a uniform duration); detection was with enhanced chemiluminescence.
For peptide mapping studies, proteins were separated by SDS-PAGE, transferred to nitrocellulose, and the band corresponding to PKC␦ was excised from the membrane, cut into small pieces, and treated for 30 min at 37°C with polyvinylpyrrolidone (0.5%, w/v) in acetic acid (100 mM), followed by 5 water washes (to remove the acid) and a 10-min incubation at room temperature in the dark with 100 mM iodoacetate to carboxymethylate PKC␦. Membrane pieces were then washed three times with water and twice with 50 mM ammonium bicarbonate and incubated overnight at 37°C in 60 l of a buffer containing 42 mM ammonium bicarbonate, 17 M HCl, and 10 g of sequencing grade trypsin. Digested peptides were eluted from the mem-brane by sonication, lyophilized, and the residue was reconstituted in 0.1% trifluoroacetic acid and fractionated by RP-HPLC on a Vydac semimicro C 18 column (2.1 ϫ 250 mm). Peptides were eluted with a linear gradient from 0.1% trifluoroacetic acid in water to 0.1% trifluoroacetic acid in acetonitrile over 140 min at a flow rate of 1 ml/min. The eluant was monitored at 220 nm and fractions were collected every 30 s for Cherenkov counting. Radioactive peptides of interest were submitted to the Howard Hughes Medical Institute/Columbia University Protein Chemistry Core Facility for sequencing by MALDI-MS.

RESULTS
Distinct Agonist-dependent PKC␦ Phosphorylation Patterns in Cardiomyocytes-We previously reported that treatment with PMA, the ␣ 1 -adrenergic receptor (␣ 1 -AR) agonist norepinephrine (NE), and H 2 O 2 results in distinct PKC␦ phosphorylation patterns in cardiomyocytes. Fig. 1 extends these studies to examine PKC␦ regulation by endothelin-1 (ET-1) and PDGF. Fig. 1A shows that PKC␦ is recovered from resting cardiomyocytes with little-to-no Thr 505 or Tyr 311 phosphorylation. NE and ET-1 increase PKC␦ phosphorylation at Thr 505 . In each case, PKC␦-Thr 505 phosphorylation is maximal at 5 min (the first time point sampled in the experiments) and sustained for at least another 10 min of continuous stimulation (Fig. 1B). NEand ET-1-dependent increases in PKC␦-Thr 505 phosphorylation are similar in magnitude, and comparable with the stimulatory effect of PMA (Fig. 1, A and C). We previously reported that PMA induces a rapid and sustained increase in PKC␦-Thr 505 phosphorylation that falls only as PKC␦ protein abundance decreases due to down-regulation during chronic stimulation (2). Fig. 1B shows that the NE-dependent increase in PKC␦-Thr 505 phosphorylation also is sustained for at least 60 min, whereas the ET-1 response wanes with stimulation intervals greater than 30 min. PDGF also increases PKC␦-Thr 505 phosphorylation, but this response is quite modest in magnitude when compared with the considerably more robust increases in PKC␦-Thr 505 phosphorylation induced by NE, ET-1, or PMA. PDGF-dependent PKC␦-Thr 505 phosphorylation is detected at 5 min and wanes as the stimulation interval is prolonged to Ͼ30 min (Fig. 1, B and C).
PDGF increases PKC␦ phosphorylation at Tyr 311 , whereas NE and ET-1 treatments for 5 min (Fig. 1A) or selected time points up to 60 min of incubation (Ref. 8 and data not shown) do not lead to a detectable increase in PKC␦-Tyr 311 phosphorylation. The magnitude of the PDGF-dependent increase in PKC␦-Tyr 311 phosphorylation is comparable with the stimulatory effect of PMA (Fig. 1, A and C). However, PDGF and PMA responses are not necessarily mediated by the identical signaling mechanism. Using an anti-Src-Tyr(P) 416 PSSA that selectively recognizes the activation loop phosphorylated/activated forms of Src and related SFKs, Fig. 1B shows that PDGF induces a modest increase in Src activity. In contrast, PMA promotes PKC␦-Tyr 311 phosphorylation without increasing Src-Tyr 416 phosphorylation ( Fig. 1B Fig. 1B shows that H 2 O 2 induces a robust increase in PKC␦-Tyr 311 phosphorylation (to a level that exceeds PKC␦-Tyr 311 phosphorylation in PDGF-or PMA-treated cardiomyocytes) and that the H 2 O 2 response is associated with a massive increase in Src-Tyr 416 phosphorylation.
PKC␦ partitions between the soluble and particulate fractions of resting cardiomyocytes (Fig. 1D). PMA, NE, and ET-1 translocate PKC␦ protein from the soluble to the particulate fraction and increase PKC␦-Thr 505 phosphorylation exclusively in the particulate fraction ( Fig. 1D and data not shown). The PMA-dependent increase in PKC␦-Tyr 311 phosphorylation also is detected exclusively in the particulate fraction. In contrast, treatment with H 2 O 2 (which does not translocate PKC␦ to the particulate fraction, but rather releases PKC␦ from the particulate fraction) leads to an increase in PKC␦-Tyr 311 phosphorylation in both the soluble and particulate fractions. Fig. 1D also shows that H 2 O 2 does not increase PKC␦-Thr 505 phosphorylation and NE does not increase PKC␦-Tyr 311 phosphorylation.
Tyr 332 has been identified as another major phosphorylation site on heterologously overexpressed PKC␦ in H 2 O 2 -treated COS-7 cells (16). Because preliminary studies demonstrated that the anti-PKC␦-Tyr(P) 332 antibody used in our studies is not sufficiently sensitive or specific to be used directly in immunoblotting studies on cell lysates, PKC␦ was immunoprecipitated from resting and agonist-treated cardiomyocyte cultures followed by immunoblot analysis with anti-PKC␦-Tyr(P) 332 , anti-PKC␦-Tyr(P) 311 , a general anti-Tyr(P) antibody (to track total PKC␦ tyrosine phosphorylation), and an anti-PKC␦ protein antibody (to validate equal protein recovery and loading). Fig. 2A shows that high concentrations of H 2 O 2 promote PKC␦-Tyr 311 and -Tyr 332 phosphorylation in association with an increase in Src-Tyr(P) 416 immunoreactivity (used as a surrogate strategy to track SFK activity). In contrast, PMA increases PKC␦-Tyr 311 phosphorylation without detectably activating SFKs or increasing PKC␦-Tyr 332 phosphorylation; NE does not increase Src phosphorylation at Tyr 416 or PKC␦ phosphorylation at either tyrosine residue.
We previously used an immunoprecipitation strategy with antibodies that discriminate individual SFKs (Src, Lyn, and Fyn) followed by immunoblotting with the anti-Src-Tyr(P) 416 PSSA to compare the time course for Src activation and PKC␦-Tyr 311 phosphorylation. These previous studies established that Src activation precedes PKC␦-Tyr 311 phosphorylation; H 2 O 2 induces a similar rapid and robust increase in Src, Lyn, and Fyn activity that is maximal by 2 min, whereas the H 2 O 2 -dependent increase in PKC␦-Tyr 311 phosphorylation is detectable at 2 min and increases further as the incubation interval is prolonged to 5-30 min (8). Fig. 2B (17,18)). The steep concentration-response curves for H 2 O 2 -dependent SFK activation identified in these experiments suggest an amplification mechanism. The results could reflect oxidative-inactivation of an antioxidant enzyme or an H 2 O 2 response localized to mitochondria (where oxidative stress responses are amplified as a result of ROS-induced ROS release once H 2 O 2 exceeds a threshold concentration (19,20)).
Finally, we used a pharmacologic approach as an initial strategy to identify the kinase pathway(s) that promote PKC␦-Tyr 311 and -Tyr 332 phosphorylation. Fig. 2, A and C, show that agonist-dependent PKC␦-Tyr 311 and -Tyr 332 (and Src-Tyr 416 ) phosphorylation are fully abrogated by PP1. Although a recent study attributed H 2 O 2 -dependent PKC␦-Tyr 332 phosphorylation to the epidermal growth factor receptor (which is recovered in a multiprotein complex with PKC␦ from H 2 O 2 -treated COS-7 cells (21)), agonist-dependent PKC␦-Tyr 311 and -Tyr 332 phosphorylation is preserved in cardiomyocytes treated with AG1478 (an epidermal growth factor receptor inhibitor, Fig.  2C); control experiments validate the efficacy of AG1478 treatment showing that AG1478 abrogates epidermal growth factordependent activation of ERK (Fig. 2C). Agonist-dependent PKC␦-Tyr 311 and -Tyr 332 phosphorylation is also preserved in cardiomyocytes treated with GF109203X (a general PKC inhibitor) or PP3 (a structurally similar, but functionally inactive, PP1 analogue that serves as a negative control; Fig. 2A and data not shown). These results implicate Src or a related PP1-sensitive SFK in the PKC␦-Tyr 311 and -Tyr 332 phosphorylation pathway in cardiomyocytes. This conclusion gains further support from studies in SYF cells (that lack the major SFKs, Src, Yes, and Fyn) and Src ϩ cells (SYF cells engineered to re-express Src). In SYF cells, a high concentration of H 2 O 2 leads to a robust increase in PKC␦ tyrosine phosphorylation (identified with anti-PKC␦-Tyr(P) 332 antibodies in PKC␦ pull-downs (Fig. 3A) and with anti-PKC␦-Tyr(P) 311 in cell lysates (Fig. 3B)); PMA also induces a modest increase in PKC␦-Tyr 311 phosphorylation in Src ϩ cells (Fig. 3B). In contrast, H 2 O 2 -dependent PKC␦-Tyr 311 phosphorylation is detected only at very low levels, and PMA-dependent PKC␦-Tyr 311 phosphorylation is not detected in SYF cells. Control experiments establish that the defect in PKC␦ tyrosine phosphorylation in H 2 O 2 -treated SYF cells is not due to a generalized defect in H 2 O 2 -dependent responses, because H 2 O 2 -dependent activation of ERK is similar in SYF and Src ϩ cells.
The observation that SYF cells support a low level of H 2 O 2dependent PKC␦-Tyr 311 phosphorylation was surprising. Further studies suggest that this might be attributable to Lyn, which has been implicated as a PKC␦ kinase (10) and is detected at similar low levels (exclusively in a detergent-insoluble fraction) of SYF and Src ϩ cells (Fig. 3C). Lyn also is detected, as multiple molecular species with different electrophoretic mobilities, in both the soluble and insoluble fractions of neonatal rat cardiomyocytes. Lyn immunoreactivity in SYF and Src ϩ cells co-migrates with the slowest migrating form of Lyn in the insoluble fraction of cardiomyocytes. Studies with the anti-Src-Tyr(P) 416 PSSA show a similar amount of active enzyme (comigrating with Lyn protein) in the insoluble fractions of H 2 O 2treated SYF and Src ϩ cells. The increase in anti-Src-Tyr(P) 416 immunoreactivity due to Src expression in Src ϩ cells is detected exclusively in the soluble fraction. Collectively, these studies indicate that SYF cells contain some Lyn protein/activity (i.e. they are not completely devoid of SFK activity), introducing a note of caution regarding the use of SYF cells to resolve the cellular actions of SFKs. Nevertheless, whereas SYF cells support a low level of PKC␦-Tyr 311 phosphorylation, studies in Src ϩ cells suggest that Src itself is the major cellular PKC␦-Tyr 311 kinase. The results also establish that PKC␦ tyrosine   (9). Because caveolae function as platforms to facilitate cross-talk between signaling molecules (and SFKs partition to caveolae membranes in non-cardiomyocyte models (22)), we tested the hypothesis that PMAdependent PKC␦-Tyr 311 phosphorylation requires PKC␦ translocation to SFK-enriched caveolae membranes. Fig. 4A shows that PKC␦ is detected at low levels, and PKC⑀ is largely excluded, from resting caveolae (tracked by immunoblotting for the muscle-specific caveolin-3 isoform that is used as a recovery marker in these experiments). The bulk of the PKC isoform immunoreactivity is recovered in the heavy gradient fractions (F8 -13 fraction). PMA translocates PKC␦ (and PKC⑀) to the caveolae fraction. Our previous estimates place ϳ15-25% of total PKC isoform immunoreactivity (for both PKC␦ and PKC⑀) in the caveolae fraction of PMA-treated cardiomyocytes (9). Because the bulk of the PKC protein immunoreactivity remains in the F8 -13 fraction of PMA-treated cardiomyocytes, a reciprocal PMA-dependent decrease in PKC protein immunoreactivity in the F8 -13 fraction is not resolved for technical reasons.
Initial studies using a generic anti-Tyr(P) antibody showed that PMA treatment leads to the appearance of a single band in caveolae (but not F8 -13 or pellet fractions) that co-migrates with PKC␦ (Fig.  4A). Studies with the more specific anti-PKC␦-Tyr(P) 311 PSSA show that a Tyr 311 -phosphorylated form of PKC␦ is detected in resting cardiomyocytes and is confined to the heavy F8 -13 gradient fractions and that PMA increases PKC␦-Tyr 311 phosphorylation exclusively in the caveolae fraction. The PMAdependent increment in PKC␦-Tyr(P) 311 immunoreactivity is blocked by PP1, but not GF109203X (a general inhibitor of all PKC isoforms) or Gö6976 (a compound that inhibits conventional PKC isoforms or PKD, but does not block the catalytic activity of PKC␦ or other novel PKC isoforms, Fig. 4B). PP1, GF102903X, and Gö6976 do not interfere with PMAdependent PKC␦ translocation to caveolae. Fig. 4A shows that caveolae contain considerable amounts of (albeit not all) Src immunoreactivity and virtually all of the cellular Fyn, Lyn, and Yes proteins. These SFK subcellular local- ization patterns are presumed to be mediated by lipid modifications; Fyn, Lyn, and Yes are dually acylated and confined to caveolae, whereas Src is myristoylated (but lacks the cysteine necessary for palmitoylation) and partitions to both caveolae and non-caveolae fractions (22). SFK activity (tracked with the anti-Src-Tyr(P) 416 antibody) is detected exclusively in the caveolae fraction of resting cardiomyocytes and is not altered by PMA treatment. Collectively, these studies are consistent with the notion that PKC␦ localization to caveolae (in an active conformation and in close proximity with SFKs) is sufficient to support PKC␦-Tyr 311 phosphorylation by resident SFKs (that retain a low level of activity in this subcellular compartment). According to this model, caveolae function as platforms to facilitate cross-talk between PKC␦ and Src. Fig. 4B also shows that PMA increases PKC␦-Thr 505 phosphorylation. The PMAdependent increment in PKC␦-Thr 505 phosphorylation is detected in both caveolae and F8 -13 fractions; it is abrogated by GF109203X, but not by Gö6976 or PP1. Fig. 4C shows that H 2 O 2 increases PKC␦-Tyr(P) 311 immunoreactivity (in association with an increase in Src-Tyr(P) 416 immunoreactivity) in both caveolae and non-caveolae fractions. These results are consistent with studies in Fig. 1C, showing that H 2 O 2 induces a global increase in PKC␦-Tyr 311 phosphorylation in both soluble and particulate fractions; the bulk of the cellular protein (all of the cytosol and most of the particulate membrane material in the cell) is recovered in the F8 -13 fraction, and not caveolae (a highly specialized membrane fraction that comprises only ϳ0.5-1% of total cellular protein). H 2 O 2 does not alter PKC␦ or SFK partitioning between caveolae and the F8 -13 fraction. Of note, a direct comparison between H 2 O 2 and PMA responses in caveolae is revealing (Fig.  4D). Although H 2 O 2 and PMA treatment lead to similar increases in PKC␦-Tyr 311 phosphorylation when studies are performed on cell extracts or soluble/particulate fractions (Fig.  1 We previously demonstrated that cyclodextrin (a membrane-impermeable cholesterol-binding drug) can be used to extract approximately two-thirds of total cellular cholesterol and redistribute caveolin-3 (and other resident caveolae proteins) from buoyant membranes to heavy sucrose fractions, without inducing any gross morphological toxicity or major changes in spontaneous automaticity (14). Fig. 4E shows that cyclodextrin treatment also leads to a defect in PMA-dependent PKC␦-Tyr 311 phosphorylation, whereas PMA-dependent PKC␦-Thr 505 phosphorylation and H 2 O 2 -dependent PKC␦-Tyr 311 phosphorylation (which are detected in both caveolae and F8 -13 fractions) persist in cyclodextrin-treated cardiomyocytes. Control experiments show that this effect requires cholesterol depletion, as it is not observed when treatment is with cyclodextrincholesterol complexes (that do not deplete cellular cholesterol, excluding a nonspecific effect of the cyclodextrin treatment). Collectively, these results identify a critical role for cholesterol-/Src-enriched caveolae in the PMA-dependent PKC␦-Tyr 311 phosphorylation pathway.
In Vitro PKC␦ Tyrosine Phosphorylation-Cell-based studies show that PMA selectively increases PKC␦ phosphorylation at Tyr 311 (but not Tyr 332 ) in caveolae that contains multiple SFKs; pharmacologic studies implicate Src or a related PP1-sensitive enzyme as the PMA-dependent PKC␦-Tyr 311 kinase. Because previous in vitro kinase assays identified an effect of Src to phosphorylate PKC␦ at both Tyr 311 and Tyr 332 (7), and PP1 is an equipotent inhibitor of Src, related SFKs, and PDGFRs (23), we performed in vitro kinase assays to determine whether other SFKs or the PDGFR might promote PKC␦-Tyr 311 phosphorylation without inducing a coordinate increase in PKC␦ phosphorylation at Tyr 332 . Fig. 5A shows that Lyn, Fyn, and Yes mimic the effect of Src to increase PKC␦ phosphorylation at both Tyr 311 and Tyr 332 (i.e. in vitro PKC␦ tyrosine phosphorylation by individual SFKs cannot be distinguished). In contrast, an active fragment of recombinant human PDGFR␤ that undergoes robust in vitro autophosphorylation triggers only a very modest increase in PKC␦ phosphorylation at Tyr 311 . Studies with the anti-PKC␦-Tyr(P) 332 PSSA suggest that this is not associated with a coordinate increase in PKC␦-Tyr 332 phosphorylation. However, conclusions regarding the role of PDGFRs as a PKC␦-Tyr 332 kinase are somewhat tenuous, because these immunoblotting experiments are undermined to some degree by the imperfect specificity of the anti-PKC␦-Tyr(P) 332 PSSA; this PSSA recognizes the autophosphorylated PDGFR. Although this should not entirely preclude the identification of a PDGFR-dependent increase in PKC␦-Tyr 332 phosphorylation (because the PDGFR and PKC␦ have somewhat different mobilities and are resolved in these studies), this negative conclusion is considered more tentative. Nevertheless, the physiologic significance of a very modest PDGFR-dependent increase in PKC␦-Tyr 311 phosphorylation (in the context of the very high levels of PDGFR autocatalytic activity) is uncertain.
FAK and JAK2 also have been identified as enzymes that cooperate with Src and localize to caveolae in other cell types (24). Fig. 5B shows that a pool of cellular FAK constitutively localizes to the cardiomyocyte caveolae with little-to-no phosphorylation at Tyr 397 (a target for an autophosphorylation reaction that accompanies FAK activation (25)). Oxidative stress increases FAK abundance and FAK-Tyr 397 phosphorylation in caveolae; FAK-Tyr 397 phosphorylation also is detected at lower levels in the heavy F8 -13 fraction. Of note, H 2 O 2 -dependent FAK translocation to caveolae and FAK-Tyr 397 phosphorylation are inhibited by PP1, indicating that Src is required for optimal/efficient FAK-Tyr 397 autophosphorylation in cardiomyocyte caveolae. This is presumed to reflect a docking interaction between the Src SH2 domain and Tyr 397 -phosphorylated FAK, which has been implicated in FAK activation by certain stimuli (i.e. integrin engagement but not VEGF or G protein-coupled receptor activation (25)(26)(27)(28)). The observation that FAK localizes to (and is activated in) caveolae via a PP1sensitive mechanism provided the rationale to consider a role for FAK as a putative PKC␦-Tyr 311 kinase. Fig. 5C shows that FAK and JAK2 (another non-receptor tyrosine kinase that localizes to caveolae in other cell types) undergo robust autophosphorylation reactions but neither enzyme displays PKC␦-Tyr 311 or -Tyr 332 kinase activity.
Finally, we considered a role for c-Abl that constitutively localizes to the detergent-insoluble lipid raft fraction in certain cell types, is moderately sensitive to the direct inhibitory effects of PP1, and also can be inhibited by PP1 indirectly (because it is downstream from Src in certain signaling pathways). A recent study implicated c-Abl activity in a cellular PKC␦-Tyr 311 phosphorylation pathway (29), but the direct in vitro effects of c-Abl (and a role for c-Abl in PKC␦-Tyr 332 phosphorylation) were not considered. Fig. 6A shows that c-Abl mimics the effect of Src to phosphorylate PKC␦ at Tyr 311 , but c-Abl does not act as a PKC␦-Tyr 332 kinase. Of note, our previous studies showed that Src-dependent PKC␦-Tyr 311 phosphorylation is via a PMA-dependent mechanism (presumably reflecting a role for lipid cofactors to induce a conformational change that renders PKC␦ a better substrate for phosphorylation by Src). Fig. 6 extends the analysis by showing that c-Abl also phosphorylates PKC␦ in a lipid cofactor-dependent manner and that PKC␦ tyrosine phosphorylation (either by Src or c-Abl) is similar in assays with PS/PMA, PS/DAG, or PS/diC8 (a membrane-permeant DAG analogue typically used in cell based studies); PS alone does not support Src-or c-Abldependent PKC␦ tyrosine phosphorylation. Collectively, these results indicate that: 1) c-Abl mediates PKC␦ phosphorylation at Tyr 311 , but not Tyr 332 and 2) that DAG analogues can substitute for PMA and render PKC␦ a substrate for tyrosine phosphorylation by either Src or c-Abl.
Because the immunoblotting studies are biased by the availability of PSSAs, and previous reports in the literature using a mutagenesis strategy mapped c-Abl-dependent PKC␦ phosphorylation to Tyr 512 (a highly conserved kinase domain tyrosine residue corresponding to Tyr 204 in PKA, a residue that participates in intramolecular interactions that radiate throughout the PKA structure and are critical for the transition from inactive/apo to active/closed conformations of the enzyme (30,31)), we also subjected c-Abl-and Src-phosphorylated PKC␦ to peptide mapping analysis. Radiolabeled PKC␦ was purified by SDS-PAGE, blotted to nitrocellulose, excised from the membrane, digested with trypsin, and peptide fragments were separated by RP-HPLC (Fig. 6B). The four peaks, containing radioactive peptide fragments detected in assays with PKC␦ ϩ Src or PKC␦ ϩ c-Abl (but not in assays with PKC␦ alone), were sequenced by MALDI-TOF MS to determine sites for PKC␦ tyrosine phosphorylation. Peak 1 which was common to RP-HPLC chromatograms from in vitro kinase assays with PKC␦ and either Src or c-Abl (and was not detected when in vitro kinase assays were performed with PKC␦ alone) contained a phospho-Tyr 311 fragment (as well as non-phosphorylated forms of peptides containing Tyr 64 and Tyr 52 , other sites on PKC␦ that have been reported to be phosphorylated under certain stimulatory conditions). Peaks 2 and 3 were detected in RP-HPLC chromatograms from in vitro kinase assays with PKC␦ and Src, but not c-Abl. These peaks contained the oxidized and reduced forms of a phospho-Tyr 332 -containing fragment (note, differences in oxidation status arise during sample preparation and should not be construed as evidence that this modification occurs in cells). Finally, peak 4 was specific for RP-HPLC chromatograms from in vitro kinase assays with PKC␦ and c-Abl (but not Src). Unfortunately, repeated attempts to sequence peptide fragments in this peak were not successful. Of note, because we did not achieve complete coverage of the PKC␦ sequence, and importantly never recovered Tyr 512 -bearing peptide fragments, a role for c-Abl as a possible PKC␦-Tyr 512 kinase is not excluded. Nevertheless, these peptide mapping and immunoblotting studies provide consistent evidence that in vitro Src-dependent PKC␦ tyrosine phosphorylation maps to Tyr 311 and Tyr 332 and that c-Abl phosphorylates PKC␦ at Tyr 311 , but not Tyr 332 .
Although the activity profile of c-Abl conforms to the prediction for a PMA-activated PKC␦-Tyr 311 kinase, a role for c-Abl is excluded by additional studies showing that: 1) c-Abl immunoreactivity is readily detected in the heavy F8 -13 gradient fractions, but c-Abl is not detected in cardiomytocyte caveolae (Fig.  6C) and 2) ST1571 (also known as Gleevec, a potent c-Abl inhibitor that also inhibits c-kit and PDGFR activity, but does not inhibit the activity of non-receptor tyrosine kinases such as Src (32)) attenuates the H 2 O 2 -activated PKC␦-Tyr 311 phospho-rylation pathway, but ST1571 does not block PMA-dependent PKC␦-Tyr 311 phosphorylation (Fig. 6D). Studies of PDGFR activation (which is directly inhibited by ST1571) validate the efficacy of the ST1571 treatment protocol; PDGFRdependent PKC␦-Tyr 311 phosphorylation is blocked by ST1571, PP1, and AG1295 (a quinoxaline compound that acts as a selective inhibitor of PDGFRs, but does not inhibit Src (33)). These results indicate that PDGF promotes PKC␦-Tyr 311 phosphorylation via a pathway that requires PDGFR activity and perhaps also Src (which is activated downstream of the PDGFR signaling pathway and is the better in vitro PKC␦-Tyr 311 kinase). Collectively, these results indicate that c-Abl and SFKs cooperate to increase PKC␦ tyrosine phosphorylation during oxidative stress, but PMA-dependent PKC␦-Tyr 311 phosphorylation is via a c-Abl-independent mechanism. Fig. 1 show that NE treatment leads to the translocation of PKC␦ from the cytosol to the particulate fraction and increased PKC␦-Thr 505 phosphorylation in the particulate fraction. However, experiments depicted in Fig. 7A provide surprising evidence that NE and ET-1 do not mimic the effect of PMA to drive PKC␦ or PKC⑀ to caveolae membranes. NE and ET-1 also do not increase PKC␦-Thr 505 phosphorylation in the caveolae fraction. NE-and ET-1-dependent increases in PKC␦-Thr 505 phosphorylation are detected exclusively in the F8 -13 fraction. Importantly, the effect of NE to promote PKC␦-Thr 505 phosphorylation is blocked when DAG accumulation is prevented by the phospholipase C inhibitor U73123; control experiments show that U73123 does not block PMA-dependent PKC␦-Thr 505 or -Tyr 311 phosphorylation, excluding a nonspecific effect of U73123. These results indicate that NE activates PKC␦ through a phospholipase C-dependent mechanism involving the hydrolysis of membrane phosphoinositides and the generation of DAG. The observation that NE and ET-1 do not drive PKC␦ to caveolae suggests either: 1) GPCR-dependent lipid cofactor generation is defective in caveolae (i.e. ␣ 1 -adrenergic and ET-1 receptors increase DAG exclusively in non-caveolae membranes) or 2) NE and ET-1 increase DAG in caveolae but a local increase in DAG is not sufficient to anchor PKC␦ and PKC⑀ to caveolae membranes.

NE and ET-1 Do Not Promote PKC␦ Translocation to Caveolae; NE and ET-1 Selectively Increase PKC␦-Thr 505 (but Not Tyr 311 ) Phosphorylation in Non-caveolae Membranes-Experiments depicted in
The notion that DAG might not be sufficient to anchor PKC␦ to caveolae membranes is based upon recent evidence that the molecular determinants for PKC␦ activation by PMA and DAG differ. PMA binds with high affinity to the PKC␦-C1B domain, whereas DAG anchors full-length PKC␦ to membranes via an interaction with the PKC␦-C1A domain (34). Therefore, the final set of experiments used diC8 to determine whether DAG binding to the C1A domain is sufficient to anchor full-length PKC␦ to the particulate fraction and to caveolae membranes. Fig. 8A shows that diC8 mimics the effects of PMA to translocate PKC␦ and PKC⑀ from the soluble to the particulate fraction; this effect is detected largely as a PMA-/diC8-dependent decrease in soluble enzyme; the already high levels of PKC␦ and PKC⑀ immunoreactivity in the particulate fractions of resting cardiomyocytes obscure a lipid cofactor-dependent increase in PKC abundance in this fraction. diC8 also mimics the effects of PMA to increase PKC␦-Thr 505 and -Tyr 311 phosphorylation in the particulate fraction. Fig. 8B shows that diC8 translocates PKC␦ and PKC⑀ to caveolae; diC8 also increases PKC␦-Tyr 311 phosphorylation (exclusively in caveolae) and PKC␦-Thr 505 phosphorylation (in both caveolae and heavy fractions). The observation that DAG analogues localize PKC␦ (and support PKC␦-Tyr 311 phosphorylation) in caveolae suggests that the defect in NE-/ET-1-dependent PKC␦ and PKC⑀ localization to caveolae (and PKC␦-Tyr 311 phosphorylation in caveolae) is due to a defect in agonist-dependent DAG accumulation in this membrane fraction.

DISCUSSION
The conventional model of PKC activation by lipid cofactors is predicated on the assumption that PKCs are regulated entirely by translocation events that deliver the enzyme in an active conformation to their target substrates in membranes. This model, which assumes that PKC catalytic activity is an inherent and immutable property of the enzyme that is not altered by the activation process, has recently been challenged by studies showing that PKC␦ undergoes a series of highly regulated phosphorylation events at both an activation loop threonine and on tyrosine residues that contribute to the regulation of PKC␦ catalytic function (and particularly PKC␦ substrate specificity (5,6)). 3 These recent findings provide a rationale for studies that examine the mechanisms that regulate PKC␦ phosphorylation in a cellular context. This study identifies stimulusspecific differences in PKC␦ localization and phosphorylation that are likely to have physiologic relevance.
The PKC␦ sequence contains at least nine highly conserved tyrosine residues that have been implicated as sites for regulatory phosphorylation in various cellular contexts. Although the mechanisms and consequences of overall PKC␦ tyrosine phosphorylation have been scrutinized in considerable detail, there is still relatively limited information on the mechanisms that specifically regulate PKC␦ phosphorylation at Tyr 311 and/or Tyr 332 (sites that are the major targets for PKC␦ phosphorylation during oxidative stress and are critical for PKC␦ signaling to certain proapoptotic pathways (35)). We previously demonstrated that H 2 O 2 and PMA promote PKC␦-Tyr 311 phosphorylation via a PP1-sensitive pathway in cardiomyocytes (8 (29)). In contrast, PMA translocates PKC␦ to caveolae and leads to a highly localized increase in PKC␦-Tyr 311 phosphorylation in this subcellular compartment. Although PMA treatment does not activate SFKs, PMA-dependent PKC␦-Tyr 311 phosphorylation also is blocked by PP1; it is not blocked by ST1571 (effectively excluding a role for c-Abl, which is not detected in the caveolae fraction). On the basis of further studies showing that Src acts as an in vitro PKC␦-Tyr 311 kinase only in assays containing lipid cofactors (that induce a conformational change that renders PKC␦ a better substrate for SFKs) and that PMA-dependent PKC␦-Tyr 311 phosphorylation is blocked by cyclodextrin, which disrupts caveolae, our studies suggest that PMA promotes PKC␦-Tyr 311 phosphorylation by delivering PKC␦ in its active conformation to caveolae, in close proximity to SFKs. These studies also expose distinct PKC␦ phosphorylation patterns in caveolae and non-caveolae membranes. PKC␦ accumulates as a dually Thr 505 /Tyr 311 -phosphorylated enzyme in the caveolae fraction of PMA-treated cardiomyocytes, whereas PKC␦ is selectively phosphorylated at Thr 505 in other membrane fractions. These results are consistent with recent findings from Markou et al. (36) showing that fast protein liquid chromatography can be used to resolve PKC␦ from agonisttreated cardiomyocytes into pools of enzyme with distinct phosphorylation patterns and activities. Our study extends the analysis by identifying the unique subcellular compartment of individual PKC␦ subspecies.
This study provides novel evidence that H 2 O 2 promotes PKC␦ phosphorylation at both Tyr 311 and Tyr 332 , but PMA stimulation leads to a selective increase in PKC␦ phosphorylation at Tyr 311 (and not Tyr 332 ). Because the pharmacologic studies implicate a PP1-sensitive kinase in the PMA-dependent PKC␦-Tyr 311 phosphorylation pathway, and in vitro kinase assays implicate Src as a dual PKC␦-Tyr 311 /Tyr 332 kinase, we performed additional in vitro kinase assays to identify an enzyme that selectively phosphorylates PKC␦ at Tyr 311 , but not Tyr 332 . These studies identified a unique role for c-Abl as a selective Tyr 311 kinase; other tyrosine kinases induced either a coordinate increase in PKC␦ phosphorylation at both Tyr 311 and Tyr 332 (other SFKs such as Lyn, Fyn, and Yes) or exhibited little to no PKC␦-Tyr 311 /Tyr 332 kinase activity (PDGFR, FAK, and JAK2). Although our results suggest that c-Abl may contribute to PKC␦-Tyr 311 phosphorylation during oxidative stress (and there is ample evidence that c-Abl-dependent PKC␦ phosphorylation at Tyr 311 , and perhaps other sites such as Tyr 512 , may contribute to the catalytic or pro-apoptotic pathways of this enzyme (31,37)), results pertaining to c-Abl do not resolve the dilemma regarding the identity of the PMA activation/PP1sensitive PKC␦-Tyr 311 kinase in cardiomyocyte caveolae. PMAdependent PKC␦-Tyr 311 phosphorylation is not blocked by ST1571 and c-Abl is not detected in cardiomyocyte caveolae fractions. Rather, several other mechanisms should be considered. First, whereas the literature has focused primarily on the kinases that promote PKC␦ tyrosine phosphorylation, phosphatases that control PKC␦ dephosphorylation may also regulate this process. Second, PKC phosphorylation may be controlled through protein-protein interactions. Lee et al. (38) recently identified heat shock protein 25 (HSP25) as a PKC␦ binding partner that inhibits PKC␦ tyrosine phosphorylation and PKC␦-dependent cellular responses. Studies to date suggest that HSP25 inhibits PKC␦ by interacting with the V5 domain, an accessible binding surface that engages in long range intramolecular interactions that influence C1 domain interactions with lipid cofactors, PKC interactions with membranes, and the conformation of the catalytic pocket (regulating catalysis). The PKC␦ hinge region represents another relatively exposed surface that also might participate in protein-protein interactions that regulate enzyme function and also should be considered in future studies. Finally, it is worth noting that in vitro kinase assays examined Src-dependent phosphorylation of human PKC␦, whereas the cell-based studies were performed in rat cardiomyocytes. Although the residues immediately surrounding Tyr 332 are conserved across species, the remainder of the V3 region is considerably more variable. Recent studies also implicate alternative splicing as a mechanism that further increases PKC␦ V3 domain structural diversity by introducing inserts into the caspase-3 cleavage site of both rodent (DIL2DNNGTY 332 (39)) and human (DMQ2DNSGTY 334 (40)) V3 domains. These inserts disrupt the caspase cleavage site, prevent proteolytic PKC␦ activation, and protect cells from proapoptotic stimuli. An additional role for V3 domain inserts (which are in close proximity to V3 domain phosphorylation sites) to influence Src-dependent PKC␦ tyrosine phosphorylation is possible and will be considered in future studies.
This study exposes a mechanism for differential activation of PKC␦ by PMA and G protein-coupled receptor agonists such as NE and ET-1. We and others previously reported that NE and ET-1 translocate PKC␦ to membranes and promote PKC␦-Thr 505 phosphorylation. This study shows that NE-dependent PKC␦-Thr 505 phosphorylation requires PLC activity (presumably reflecting a role for DAG to stabilize PKC␦ at membranes). However, NE and ET-1 do not deliver PKC␦ to Src-enriched caveolae membranes or increase PKC␦-Tyr 311 phosphorylation. These agonist-specific differences in PKC␦ regulation were surprising. The observation that DAG analogues mimic the effect of PMA to recruit PKC␦ to caveolae and increase PKC␦-Tyr 311 (and Thr 505 ) phosphorylation in vivo in cardiomyocytes, and that DAG analogues effectively substitute for PMA in vitro in kinase assays, suggests that ␣ 1 -ARs do not increase PKC␦ tyrosine phosphorylation because they do not promote DAG accumulation in caveolae membranes. This could suggest that the density of ␣ 1 -ARs or their downstream signaling partners is limiting in cardiomyocytes caveolae. In this regard, it is worth noting that caveolae were originally implicated as the source of hormone-sensitive phosphatidylinositol bisphosphate pools, but this conclusion has recently been challenged (41,42). The role of caveolae as signaling domains for ET-1 and ␣ 1 -adrenergic receptors also remains uncertain. Although ET-1 receptors are reported to co-localize with caveolin-1 in a variety of non-cardiomyocyte models (43,44), ET-1 receptors apparently do not colocalize with caveolin-3 in cardiomyocytes (45). Similarly, there is evidence that ␣ 1 -ARs are recovered in phosphatidylinositol bisphosphate-enriched caveolae membranes (and regulate phosphatidylinositol bisphosphate levels) in rat heart, but the consensus of most recent studies is that ␣ 1 -ARs localize primarily to intracellular membranes (46 -49). These inconsistencies in the literature could reflect differences in the localization patterns for native and heterologously overexpressed G protein-coupled receptors or cell specific differences in G protein-coupled receptor localization (that may or may not be attributable to differences in receptor interactions with caveolin-1 and the muscle-specific caveolin-3 isoform) and will require further studies. Finally, an ␣ 1 -AR-dependent increase in DAG accumulation that is offset by the simultaneous activation of a diacyglycerol kinase (an enzyme that catalyzes the conversion of DAG to phosphatidic acid and locally depletes DAG) also might limit PKC␦ activation in caveolae. In fact, there is recent evidence that NE activates diacyglycerol kinaseleading to feedback inhibition of PKC in the caveolae/ raft fraction of rat mesenteric small arteries (48). The relevance of this finding to PKC regulation in cardiomyocytes, where a different diacyglycerol kinase isoform (namely diacyglycerol kinase-) negatively regulates hypertrophic signaling responses to G protein-coupled receptor agonists requires further study (50,51).
In summary, this study identifies distinct PKC␦ phosphorylation/localization mechanisms in cardiomyocytes treated with PMA, NE, and H 2 O 2 . Our studies suggest that models of PKC␦ function must be revised to allow for stimulus-specific PKC␦ signaling "modes" that trigger functionally distinct signaling responses in cells. In this manner, growth factors (that promote DAG generation in membranes) can trigger PKC␦-mediated events that are membrane-delimited and distinct from events induced by the Tyr 311 /Tyr 332 -phosphorylated form of PKC␦ that accumulates during oxidative stress. According to this model, PKC␦ acts both as an enzyme with highly regulated catalytic function (that is calibrated by phosphorylations at the activation loop and on tyrosine residues) and perhaps also by protein-protein interactions that regulate the efficiency/fidelity of signal transduction in cells. PKC␦ signaling modes that differ, according to the inciting stimulus, would provide an attractive explanation for the diverse (and at times contradictory) cellular responses that have been attributed to PKC␦ in the heart, where PKC␦ inhibitors have emerged as clinically important cardioprotective agents but PKC␦ activation also is reported to afford cardioprotection during ischemic preconditioning. These findings are not necessarily contradictory if the diverse and opposing effects of PKC␦ are mediated by distinct molecular forms of the enzyme (with different phosphorylation patterns) in specific subcellular localizations. This model would suggest that stimulus-specific differences in PKC␦ activation loop and tyrosine phosphorylation, that lead to dynamic changes in the enzymology and signaling function of PKC␦ in cells, might be targeted for therapeutic advantage.