Protein Kinase C Isoform-specific Differences in the Spatial-Temporal Regulation and Decoding of Metabotropic Glutamate Receptor1a-stimulated Second Messenger Responses*

Metabotropic glutamate receptors (mGluRs) coupled via Gq to the hydrolysis of phosphoinositides stimulate Ca2+ and PKCβII oscillations in both excitable and non-excitable cells. In the present study, we show that mGluR1a activation stimulates the repetitive plasma membrane translocation of each of the conventional and novel, but not atypical, PKC isozymes. However, despite similarities in sequence and cofactor regulation by diacyglycerol and Ca2+, conventional PKCs exhibit isoform-specific oscillation patterns. PKCα and PKCβI display three distinct patterns of activity: 1) agonist-independent oscillations, 2) agonist-stimulated oscillations, and 3) persistent plasma membrane localization in response to mGluR1a activation. In contrast, only agonist-stimulated PKCβII translocation responses are observed in mGluR1a-expressing cells. PKCβI expression also promotes persistent increases in intracellular diacyglycerol concentrations in response to mGluR1a stimulation without affecting PKCβII oscillation patterns in the same cell. PKCβII isoform-specific translocation patterns are regulated by specific amino acid residues localized within the C-terminal PKC V5 domain. Specifically, Asn-625 and Lys-668 localized within the V5 domain of PKCβII cooperatively suppress PKCβI-like response patterns for PKCβII. Thus, redundancy in PKC isoform expression and differential decoding of second messenger response provides a novel mechanism for generating cell type-specific responses to the same signal.

The spatial-temporal patterning of Ca 2ϩ release from intracellular stores contributes to the regulation of a diverse array of cellular responses including insulin secretion, sustained activation of mitochondrial function, and the selective activation of transcription factors required for fine-tuning gene expression during inflammatory immune responses (1)(2)(3)(4). Receptor-stimulated oscillations in intracellular-free Ca 2ϩ are observed in both excitable and non-excitable cells following the activation of either G protein-coupled receptors or receptor tyrosine kinases (5). The synchronization of Ca 2ϩ oscillations at the cel-lular level involves at least two modes of Ca 2ϩ signaling: repetitive baseline Ca 2ϩ spikes and sinusoidal-type Ca 2ϩ oscillations (6). These Ca 2ϩ oscillations are faithfully recapitulated by the repetitive redistribution of protein kinase C (PKC) 1 between the cytosol and plasma membrane (7)(8)(9)(10). The molecular mechanism(s) underlying the repetitive baseline plasma membrane translocation of PKC are best characterized for Group I metabotropic glutamate receptors (mGluRs). PKC␤II and PKC␥ oscillations are regulated by mGluR-stimulated oscillations in diacyglycerol (DAG), inositol 1,4,5-triphosphate (InsP 3 ), and Ca 2ϩ release from intracellular stores (8,9). Glutamate-activated Ca 2ϩ and/or PKC oscillations are observed in most cell systems including immature neuronal cultures, developing neocortex, astrocytes, and mGluR-transfected heterologous cell cultures (8,9,(11)(12)(13)(14).
PKC isoforms are classified into three groups based on structural properties and cofactor requirements and exhibit specific in vivo activity as well as spatial organization within the cell (15)(16)(17). The activation and plasma membrane localization of the conventional PKC isoforms (␣, ␤I, ␤II, and ␥) is regulated by Ca 2ϩ and DAG, whereas the activity and subcellular localization of the novel PKC isoforms (␦, ⑀, , and ) is regulated by DAG, but not Ca 2ϩ (17,18). The atypical PKCs do not respond to either Ca 2ϩ or DAG (17,18). Although PKC structure/function has been studied extensively, it is unknown whether PKC subtype-specific differences in Ca 2ϩ -and/or DAG-binding affinities contribute to PKC isoform-specific activity and/or spatialtemporal distribution within cells.
In the present study, we explore whether mGluR1a-stimulated oscillations in intracellular DAG and/or Ca 2ϩ concentrations regulate conventional and novel PKC isoform activity in an identical manner. We find that, although all conventional and novel PKC isoforms oscillate in response to mGluR1a activation, conventional PKC isoforms exhibit isozyme-specific translocation response patterns that we classify as either PKC␤I-or PKC␤II-like responses. Specifically, we have identified two discrete amino acid residues localized within the V5 domain of PKC␤II that function to suppress PKC␤I-like responses for PKC␤II. Thus, structural differences in the PKC V5 domain allow conventional PKC isoforms to differentially decode and influence receptor-stimulated DAG and Ca 2ϩ signals. Consequently, the expression of multiple conventional PKC isozymes in either the same cell or within different cells provides a novel mechanism by which cell type-specific responses to an identical signal may be established.

MATERIALS AND METHODS
Materials-Restriction enzymes were obtained from Promega and New England Biolabs Inc. The pcDNA3.1/Amp expression vector was acquired from Invitrogen. DsRed2-C1, pEGFP-C1, pEGFP-C2, and pEGFP-C3 expression vectors were purchased from Clontech. The QuikChange TM Site-directed Mutagenesis kit was from Stratagene. The human universal Quick-Clone TM cDNA library was obtained from Clontech. Human embryonic kidney cells (HEK 293) were from American Type Culture Collection (ATCC). Fetal bovine serum was from Hyclone Laboratories Inc. Gentamicin, minimal essential medium, and 0.05% Trypsin containing 0.5 mM EDTA were acquired from Invitrogen. The calcium indicator, Oregon Green 488 BAPTA-1 AM, was obtained from Molecular Probes. Quisqualate was from Tocris Cookson Inc. All other biochemical reagents were purchased from Sigma, Fisher Scientific, and VWR.
Plasmid Constructs-To construct EGFP-tagged PKC␣, PKC␤I, PKC␥, PKC␦, PKC⑀, PKC, PKC, PKC, and PKC/ the cDNA for each of the PKC isoforms were first amplified by PCR from the human universal Quick-Clone TM library (Clontech). The PCR products generated were digested with the appropriate restriction enzymes and subcloned into the appropriate pEGFP-C1, pEGFP-C2, and pEGFP-C3 vectors (Clontech). The PKC␤I cDNA was also cloned into the BglII-XbaI sites of the vector DsRed2-C1 (Clontech). The construction of EGFP-PKC␤II (28), and EGFP-PLC␦1 PH domain were previously described (8). The EGFP-PKC␦ C1 domain was a generous gift from Dr. Sergio Grinstein. PKC␣/␤II, ␤II/␣, ␤I/␤II, and ␤II/␤I chimeras were constructed by "two-step PCR" as described previously (19) and the resulting PCR products subcloned into either the pEGFP-C1 or pEGFP-C3 vector. PKC␣ and ␤II point and deletion mutants were constructed using the QuikChange TM Site-directed Mutagenesis kit (Stratagene). Sequence integrity of all PCR-generated products was confirmed by automated DNA sequencing.
Cell Culture and Transfection-HEK 293 cells were maintained in minimal essential medium supplemented with 10% (v/v) fetal bovine serum and 100 g/ml gentamicin at 37°C in a humidified atmosphere containing 5% CO 2 . Cells used in each of the experiments were transfected using a modified calcium phosphate method as described previously (20). Following transfection (ϳ18 h), cells were incubated with fresh medium and allowed to recover 8 h and allowed to grow an additional 18 h before any experimentation. In all experiments, cells were transfected with 10 g of pcDNA3.1 plasmid cDNA containing FLAG-mGluR1a with and without 1-5 g of each PKC construct expressed in either pEGFP or DsRed2 expression vectors.
Confocal Microscopy-Following transfection with plasmid cDNAs encoding EGFP-PKC constructs and mGluR1a, cells were re-seeded on collagen-coated 15-mm glass-cover slips designed for use in a perfusion system (Warner Instrument Corporation). All experiments were con-ducted at 37°C, and prior to visualization or additional treatments the cells were perfused with at least 5 ml of HEPES-buffered salt solution (1.2 mM KH 2 PO 4 , 5 mM NaHCO 3 , 20 mM HEPES, 11 mM glucose, 116 mM NaCl, 4.7 mM KCl, 1.2 mM MgSO 4 , 2.5 mM CaCl 2 , pH 7.4). Cellular InsP 3 and DAG levels were measured indirectly through the use of the EGFP-PLC␦1 PH domain (8) and EGFP-PKC␦ C1 domain reporter constructs (9). Confocal-based Ca 2ϩ imaging was performed by preloading cells for 30 min with 10 mM Oregon Green 488 BAPTA-1 AM prior to receptor activation according to the manufacturer's specifications (Molecular Probes). Confocal microscopy was performed on a Zeiss LSM-510 laser-scanning microscope using Zeiss 63ϫ 1.4 numerical aperture oil immersion lens. Enhanced GFP and Oregon Green 488 BAPTA fluorescence was visualized with excitation at 488 nm and emission 515-540 nm emission filter set. DsRed2 fluorescence was visualized with excitation at 543 nm and emission 590 -610 nm filter set. Fluorescent signals were collected sequentially every 6.8 -12.5 s using the Zeiss LSM software time scan function.
Data Analysis-PKC translocation time courses and Ca 2ϩ , DAG, and InsP 3 responses were recorded as time series of 150 -300 confocal images for each experiment. Image analysis was performed using the Zeiss LSM-510 physiology analysis software and was defined as the relative change in cytoplasmic fluorescence intensity over time in a 5-m-diameter region of interest. All time course data were plotted using GraphPad Prism. The statistical significance of the data presented in Fig. 6 was analyzed using a non-parametric SIGN test.

PKC Isozyme-specific Plasma Membrane Translocation
Responses -Because individual conventional (␣, ␤I, ␤II, and ␥), novel (␦, ⑀, , and ), and atypical (/ and ) PKC isoforms may exhibit differences in their receptor-activated translocation profiles, we have examined the translocation response patterns for each PKC isozyme following mGluR1a agonist activation. We find that each of the conventional PKC isoforms displays the capacity to oscillate between the cytosol to the plasma membrane in response to mGluR1a activation. Shown in Fig. 1 is an agonist-stimulated oscillatory response pattern that is exhibited by all four conventional PKC isoforms. This agoniststimulated transient, but repetitive, translocation of enzyme to the plasma membrane is the only response pattern that is observed for PKC␤II (PKC␤II-like). However, PKC␣ and PKC␤I display additional translocation responses (PKC␤I-like) that are never observed for PKC␤II. Although some cell to cell variability is observed, PKC␤I-like translocation patterns can be categorized into three distinct patterns that are illustrated in Fig. 2. First, GFP-PKC␣ and GFP-PKC␤I exhibit the capacity to oscillate between the cytosol and plasma membrane in the absence of mGluR1a activation ( Fig. 2A). In these cells, mGluR1a agonist activation serves to increase the oscillation frequency of the entire cellular complement of GFP-tagged PKC␣ and PKC␤I ( Fig. 2A). The removal of the agonist results in the return of the oscillation frequency to the pre-agoniststimulated oscillatory rate ( Fig. 2A, PKC␣). These agonistindependent oscillations are only observed following mGluR1a expression, indicating that they occur in response to spontaneous mGluR1a activity. Second, in response to mGluR1a activation, the entire pool of either GFP-PKC␣ or GFP-PKC␤I translocates from the cytosol to the plasma membrane where it remains persistently localized (at steady state) until the agonist is removed by perfusion (Fig. 2B). Third, mGluR1a activation stimulates the translocation of the entire GFP-PKC␣ and GFP-PKC␤I pools to the plasma membrane, and a fraction of the enzyme returns to the cytosol and subsequently oscillates in the presence of the agonist (Fig. 2C). PKC␥ also exhibits very weak PKC␤I-like responses (13/49 cells), but unlike what is observed for the other conventional PKC isoforms the most common PKC␥ response pattern is a single transient plasma membrane translocation (20/49 cells) (data not shown). For PKC␣, PKC␤I, and PKC␤II transient translocation responses were rarely observed (Ͻ 5% of cells imaged). Taken together, these observations suggest that PKC␣, PKC␤I, and PKC␥ exhibit the capacity to decode subtly different changes in second messenger responses that are not recognized by PKC␤II. Alternatively, the expression of PKC␣, PKC␤I, and PKC␥ may lead to multiple distinct second messenger responses to mGluR1a activation.
Each of the novel PKC isoforms also oscillate in response to mGluR1a activation (Fig. 3). The response for each of the novel PKC isozymes is similar to the third PKC␤I-like response pattern outlined above (Fig. 2C). Following the agonist-dependent translocation of the entire pool of GFP-PKC, a fraction of each of the novel GFP-PKC isoforms remains localized to the plasma membrane, and the remaining enzyme oscillates in the contin-ued presence of agonist (Fig. 3). The atypical PKC (/ and ) do not respond to mGluR1a activation (data not shown). Because the conventional PKCs exhibit the greatest behavioral complexity and diversity to mGluR1a activation, we focused subsequent experimentation on the conventional PKC isoforms.
PKC␤I-dependent Alterations in mGluR1a-stimulated Second Messenger Responses-It is possible that the expression of either PKC␣ or PKC␤I leads to alterations in the patterning of mGluR1a-stimulated second messenger responses and that this may underlie the multiplicity of PKC␤I-like response patterns. To address this possibility, we examined the patterning of red fluorescent protein tagged-PKC␤I (DsRed2-PKC␤I) responses at the same time as we measured changes in intracellular DAG, InsP 3 , and Ca 2ϩ concentrations. DAG responses were measured using a GFP-PKC␦ C1 domain construct, which translocates from the cytosol to the plasma membrane in response to increases in intracellular DAG concentrations (9). We found that mGluR1a-stimulated GFP-PKC␦ C1 domain translocation patterns were synchronized exactly with DsRed2-PKC␤I membrane translocation responses (Fig. 4A). Following mGluR1a activation, the GFP-PKC␦ C1 domain either oscillated between the cytosol and plasma membrane in synchrony with DsRed2-PKC␤I (Fig. 4A,  upper panel) or accumulated with DsRed2-PKC␤I at the plasma membrane (Fig. 4A, lower panel).
InsP 3 responses were measured using a GFP-PLC␦1 PH domain construct, which under basal conditions was localized at the plasma membrane due to its interaction with membrane phosphatidylinositol 4,5-bisphosphate (PIP 2 ) (8). Following mGluR1a-stimulated PIP 2 hydrolysis and InsP 3 formation the GFP-PLC␦1 PH domain was released from the plasma membrane and redistributed to the cytosol (Fig. 4B). Identical to what was observed for DAG responses, the GFP-PLC␦1 PH domain either oscillated between the plasma membrane and cytosol at the same frequency at which DsRed2-PKC␤I translocated from the cytosol to plasma membrane (Fig. 4B, upper  panel), or the GFP-PLC␦1 PH domain accumulated in the cytosol with the same time course as DsRed2-PKC␤I accumulated at the plasma membrane (Fig. 4B, lower panel). Thus, the spatial-temporal localization of PKC␤I at the plasma membrane was coordinated with alterations in both intracellular DAG and InsP 3 concentrations.
Changes in intracellular Ca 2ϩ concentrations were measured using the Ca 2ϩ indicator dye Oregon Green 488 BAPTA-1 AM.
We found that oscillatory DsRed2-PKC␤I responses were synchronized with Ca 2ϩ oscillations (Fig. 4C, upper panel). However, unlike what was observed for DAG and InsP 3 responses, Ca 2ϩ oscillations persisted when DsRed2-PKC␤I was persistently localized to the plasma membrane (Fig. 4C, lower panel). Taken together, our observations indicate that expression of PKC␤I alters the patterning of DAG and InsP 3 formation, but not the patterning of Ca 2ϩ release from intracellular stores in response to mGluR1a activation. This is different from cells expressing PKC␤II where only synchronized oscillatory DAG, InsP 3 , and Ca 2ϩ responses are observed following mGluR1a activation (data not shown and Ref. 8).
Effect of PKC␤I Expression on PKC␤II Plasma Membrane Translocation Responses-PKC␤I and PKC␤II isoforms are thought to be regulated in the same manner by DAG (17). Therefore, if differences in PKC␤I versus PKC␤II response patterns are solely the consequence of PKC␤I expression-dependent alterations in DAG formation or differences in mGluR1a expression levels between cells, GFP-PKC␤II should exhibit PKC␤I-like translocation patterns in cells co-expressing DsRed2-PKC␤I. When co-expressed together in HEK 293 cells, we observe two distinct DsRed2-PKC␤I and GFP-PKC␤II responses to mGluR1a activation: 1) DsRed2-PKC␤I and GFP-PKC␤II exhibit synchronized oscillatory plasma membrane translocation responses (Fig. 5A); and 2) DsRed2-PKC␤I accumulates at the plasma membrane, whereas GFP-PKC␤II continues to oscillate between the plasma membrane and cytosol (Fig. 5B). These observations suggest that, although PKC␤I expression alters the patterning of mGluR1a-stimulated DAG and InsP 3 response patterns, PKC␤II is apparently insensitive to PKC␤I-induced changes in DAG formation. Moreover, the differences in PKC␤I versus PKC␤II membrane translocation patterns observed in the same cell indicates that differences in mGluR1a expression levels between cells cannot account for the different PKC␤I-like translocation patterns.
Molecular Determinants for Isozyme-specific Translocation Response Patterns-The patterning of conventional PKC isoform responses to mGluR1a activation are sub-classified as either PKC␤I-like (agonist-independent oscillations, agoniststimulated oscillations, and persistent plasma membrane localization) or PKC␤II-like (only agonist-stimulated oscillations). We have used these definitions to characterize the structural determinants underlying differences in conventional PKC isozyme response patterns. PKC␤I and PKC␤II differ by only 53 amino acids comprising the alternatively spliced V5 domains of the kinases (Fig. 6A). Furthermore, the exchange of the last 54 amino acid residues of PKC␣ with the corresponding residues from PKC␤II generates a PKC␣/␤II 620 -673 chimera that displays a PKC␤II-like response pattern (Fig. 6B). Thus, PKC isoform-specific response patterns must be regulated by amino acid residues localized within V5 domains of conventional PKC isoforms.
Sequence alignment of PKC␣, PKC␤I, and PKC␤II reveals considerable carboxyl-terminal sequence conservation (Fig.  6A). The last 13 amino acids of the V5 domain exhibit the greatest sequence disparity (Fig. 6A). By swapping either the last 13 amino acids of PKC␤II for the last 15 amino acid residues of PKC␣ (PKC␤II/␣ 657-672) or deleting the last 13 amino acid residues from the carboxyl-terminal of PKC␤II (PKC␤II-S660D), we create PKC␤II chimeras with PKC␤I-like response patterns (Fig. 6B). Serial truncation analysis of PKC␤II between amino acid residues 660 and 672 reveals that PKC␤II-like responses are lost if the final six (PKC␤II-L667⌬) but not the final three (PKC␤II-E670⌬) PKC␤II amino acids are deleted (Fig. 6C). The deletion of Lys-668 -Glu-670 (PKC␤II-KPE⌬) from PKC␤II also establishes a PKC␣-like response pattern for PKC␤II (Fig. 6C). The mutation of Lys-668, Pro-669, and Glu-670 individually to glycine residues reveals that only Lys-668 is required to maintain PKC␤II-like responses and to suppress PKC␤I-like response patterns (Fig.  6D). When expressed together in HEK 293 cells, we find that DsRed2-PKC␤I and GFP-PKC␤II-K668G exhibit identical response patterns to mGluR1a activation (Fig. 7).
We found that the establishment of a PKC␤II-like response pattern in PKC␣ required the exchange of the entire PKC␤II V5 domain. Furthermore, neither the exchange of the last 13 amino acid residues from PKC␤II into PKC␣ (PKC␣/␤II 660 -673) nor the introduction of the KPE motif into PKC␤I established PKC␤II-like responses for either PKC␣ or PKC␤I (Fig. 6,  B and C). Therefore, there must be additional amino acid residues localized within the PKC␤II V5 domain that cooper- ated with Lys-668 to establish a PKC␤II-like response pattern. Sequence alignment of PKC␣, PKC␤I, and PKC␤II indicated that only 4 amino acid residues were not conserved between PKC␤II and either PKC␣ or PKC␤I: Asn-625, His-636, Glu-646, and Arg-649 of PKC␤II (Fig. 6A). Therefore, we mutated each of these residues to the corresponding amino acid residue in PKC␣ and found that only PKC␤II-N625G exhibited PKC␣like behavior patterns (Fig. 6D). In summary, extensive structure-function analysis identified Asn-625 and Lys-668 as essential amino acid residues within the PKC␤II V5 domain required for the establishment of PKC␤II translocation responses. The mutation of either residue releases the suppression of PKC␣-like response behaviors for PKC␤II. DISCUSSION In the present study, we show that each of the conventional (␣, ␤I, ␤II, and ␥) and novel (␦, ⑀, , and ), but not atypical ( and /) PKC isozymes respond to mGluR1a activation by repetitively translocating between the cytosol and plasma membrane. The detailed analysis of conventional PKC isoform translocation responses reveals that PKC␣ and PKC␤I exhibit a variety of unique response patterns to mGluR1a activation that are not observed for PKC␤II. Because PKC␤I and PKC␤II exhibit distinct translocation patterns to the same stimulus, even when expressed in the same cell, it is unlikely these differences can be attributed to differences in mGluR1a expression levels or G protein coupling efficiency. Rather PKC␣ and/or PKC␤I expression alters the patterning of mGluR1astimulated second messenger responses. Furthermore, the sensitivity of conventional PKC isozymes to intracellular DAG and Ca 2ϩ concentrations appears to be regulated by residues localized to the V5 domains of the kinases. Thus, we conclude that the spatial-temporal dynamics of mGluR signaling will not only be determined by the identity of the mGluR isoform that is activated but will also be controlled by the PKC isozyme that is decoding and modulating the mGluR1a-generated second messenger signals. As a consequence, the expression of multiple conventional PKC isozymes in either the same cell or within different cells provides a novel mechanism by which cell typespecific responses to an identical signal may be established.
Both PKC␣ and PKC␤I exhibit agonist-independent oscillations in cells expressing mGluR1a, and this may be related to the observation that mGluR1a exhibits significant basal activity in the absence of agonist (21,22). Intrinsic mGluR1a activity leading to the spontaneous activation of PLC may result in sub-threshold changes in DAG and InsP 3 formation. If PKC␣ and PKC␤I exhibit heightened sensitivity to changes in intracellular DAG concentrations as compared with PKC␤II, this Asterisks indicate PKC␤II mutants displaying differences in translocation pattern compared with the expected wild-type PKC␤II isoform response patterns. A non-parametric SIGN test, p Ͼ 0.07 supports the null hypothesis that the asterisked PKC␤II mutants exhibit no differences in oscillation patterns with the expected patterns observed for PKC␣. may be translated into isoform-specific agonist-independent membrane translocation. The crystal structure of the mGluR ligand-binding domain predicts that in the absence of agonist the ligand-binding domain exists in equilibrium between active and inactive conformations (23). Agonist binding likely stabilizes the active (closed) conformation of the ligand-binding domain at equilibrium, which is then translated as an increase in both the efficacy and frequency of G protein coupling. Similarly, the mutation of an aspartic acid residue at position 854 in the G protein-coupling domain of mGluR1a to an alanine residue creates a receptor that gains the capacity to drive agonistindependent PKC␤II oscillations (8). Conversely, in cells expressing mGluR1b, a mGluR1 splice variant that exhibits reduced spontaneous G protein coupling activity (21,24), PKC␣ and PKC␤I do not exhibit agonist-independent oscillations, but retain the capacity to both oscillate and accumulate at the cell surface in response to mGluR1a activation (data not shown). Taken together, these observations suggest that agonist-independent PKC␣ and PKC␤I oscillations are driven by basal mGluR1a activity.
Unlike PKC␤II, PKC␣ and PKC␤I exhibit the capacity to be persistently localized at the plasma membrane in response to mGluR1a activation, and this response persists until the cells are perfused with agonist-free medium. Persistent PKC␣ translocation responses have also been reported previously (25) in response to thyrotropin-releasing hormone receptor activation. Constitutive plasma membrane localization of PKC␤II can also be achieved following the mutation of two autophosphorylated amino acid residues, Thr-641 and Ser-660, to alanine residues or by treating cells with PKC inhibitors (8, 26 -28). In contrast, the autophosphorylation of equivalent residues within the carboxyl-terminal variable domain of PKC␣ prolongs its activation (29,30), which may account in part for the persistent localization of the enzyme at the plasma membrane in response to mGluR activation. However, the persistent localization of PKC␣ and PKC␤I at the plasma membrane cannot be completely explained by PKC subtype-regulated differences in autophosphorylation, because the spatialtemporal dynamics of DAG and InsP 3 formation are also altered in these cells. One potential explanation for the observed changes in DAG and InsP 3 formation in PKC␣ and PKC␤I-expressing cells is that the enzymes participating in the regulation of intracellular DAG levels may serve as PKC isoform-specific substrates and that the V5 domain may control substrate-specificity. Consistent with this idea, the activation of endogenous conventional PKC with thymeleatoxin results in PLC␤ 3 phosphorylation and attenuation of oxytocin receptorstimulated phosphatidylinositide turnover (31). However, it is unknown which conventional PKC isoforms contribute to the phosphorylation-dependent inactivation of PLC. An alternative explanation may involve the differential ability of conventional PKC isoforms to associate with membrane anchoring proteins due to structural differences in their V5 domains. For example, the association of PKC␤II with receptor for activated C kinase 1 (RACK1) is regulated by three regions within the V5 domain of PKC␤II (32), and two of these regions encompass the amino acid residues (Asn-625 and Lys-668) that regulate PKC␤II translocation patterns. The relative contributions of PKC autophosphorylation, PLC phosphorylation, and membrane anchoring proteins to the regulation of PKC subtype-specific translocation patterns will require extensive future study.
An important observation made in the present study is that PKC␤I and PKC␤II exhibit distinct activation patterns even when expressed in the same cell. These differences in activity are abolished by either the mutation of Asn-625 or Lys-668 in PKC␤II. These two residues appear to cooperate with one an-other to repress PKC␣-like response patterns for PKC␤II. This indicates the amino acid composition of the V5 domains of otherwise identical PKC isoforms regulates the relative sensitivity of the enzymes to both changes in intracellular DAG and Ca 2ϩ concentrations. Previously, Keranen and Newton (33) demonstrated that the PKC␤I and PKC␤II V5 domains regulate differences in the enzymes Ca 2ϩ -dependent affinity for acidic membranes. Thus, the PKC␤I-like versus PKC␤II-like response patterns may reflect V5 domain-regulated differences in both DAG and Ca 2ϩ affinity. Thus, the periodicity of activation for different conventional PKC isoforms is not only regulated by the duration and strength of DAG and Ca 2ϩ signals, but is also determined by the relative sensitivity of the kinases to alterations in intracellular DAG and Ca 2ϩ concentrations.
In summary, we have discovered that homologous PKC isozymes, even when expressed in the same cell, display distinct patterns of activation in response to the same receptor stimulus. This observation is of fundamental importance to our understanding of cell signaling and clearly illustrates that the expression of multiple kinase isoforms in the same cell does not result in redundancy of cellular function. Rather, our results provide concrete evidence that the stimulation of a single receptor subtype, in a single cell, has the potential to activate distinct patterns of PKC isozyme activation, which may be translated into distinct cellular responses. We speculate that this may be particularly important in the developing and adult nervous system where Ca 2ϩ spikes and PKC translocation responses are linked to both synapse formation and synaptic plasticity required for memory and learning (34,35).