Protein Kinase Cα Is Required for Vanilloid Receptor 1 Activation

Activation of vanilloid receptor (VR1) by protein kinase C (PKC) was investigated in cells ectopically expressing VR1 and primary cultures of dorsal root ganglion neurons. Submicromolar phorbol 12,13-dibutyrate (PDBu), which stimulates PKC, acutely activated Ca2+ uptake in VR1-expressing cells at pH 5.5, but not at mildly acidic or neutral pH. PDBu was antagonized by bisindolylmaleimide, a PKC inhibitor, and ruthenium red, a VR1 ionophore blocker, but not capsazepine, a vanilloid antagonist indicating that catalytic activity of PKC is required for PDBu activation of VR1 ion conductance, and is independent of the vanilloid site. Chronic PDBu dramatically down-regulated PKCα in dorsal root ganglion neurons or the VR1 cell lines, whereas only partially influencing PKCβ, -δ, -ε, and −ζ. Loss of PKCα correlated with loss of response to acute re-challenge with PDBu. Anandamide, a VR1 agonist in acidic conditions, acts additively with PDBu and remains effective after chronic PKC down-regulation. Thus, two independent VR1 activation pathways can be discriminated: (i) direct ligand binding (anandamide, vanilloids) or (ii) extracellular ligands coupled to PKC by intracellular signaling. Experiments in cell lines co-expressing VR1 with different sets of PKC isozymes showed that acute PDBu-induced activation requires PKCα, but not PKCε. These studies suggest that PKCα in sensory neurons may elicit or enhance pain during inflammation or ischemia.

are under active investigation. Acute application of phorbol esters sensitizes knee joint sensory nerves (10), and enhances neuronal ionic currents activated by a noxious thermal stimulus (11,12), but these studies did not connect PKC activation to VR1 activation. PKC also seems to be an important mediator of bradykinin-and epinephrine-induced nociceptor sensitization and hyperalgesia (13)(14)(15)(16).
Thus, PKC appears to be an effector in multiple signal transduction cascades in primary afferent neurons. However, PKC is not a "single entity," but a family of at least 11 serine-threonine kinases with different protein substrate specificities (17)(18)(19). Whereas there are functional differences between PKC isozymes, it is not exactly known which isozymes contribute to nociceptive processing. One approach to demonstrate specific involvement of a particular PKC isozyme is targeted gene deletion. Knockout of the PKC␥ isozyme in mice suggested that this enzyme, which is normally highly expressed in specific second order neurons in spinal inner lamina II, was involved in the proper processing of incoming afferent nociceptive signals (20 -23).
Another approach to identify specific isozyme participation is selective down-regulation of PKC isotypes by exposure to active phorbol esters. Acute treatment (i.e. 1 M PDBu for 10 min) not only activates classical and novel PKC isozymes but also induces enzyme translocation to the plasma membrane. Upon chronic exposure (ϳ24 h) this leads to isozyme-selective proteolysis and differential down-regulation. Either classical PKC isotypes, activated by diacylglycerol and Ca 2ϩ , or novel PKC isotypes, activated by diacylglycerol only, can be specifically down-regulated by micromolar doses of PDBu, but to different degrees (24). In primary sensory neurons prepared from 2-dayold rats, Cesare et al. (12) reported expression of only 5 isoforms, PKC␤ I , -␤ II , -␦, -⑀, and -. Upon treatment with bradykinin, an endogenous algesic peptide, only PKC⑀ was determined to translocate to the plasma membrane in the DRG cultures (12). In contrast to the isozyme pattern detected in sensory neurons from newborns, PKC␣, an additional isozyme was positively identified in DRG cultures prepared from 3-month-old rats, which may regulate VR1 in adults (25). Bradykinin was also found to release VR1 from phosphatidylinositol 4,5-bisphosphate-mediated inhibition, a mechanism proposed for pain-specific channel opening (26).
The intracellular signaling pathways upstream of PKC isozymes in nociceptive afferents still represent open questions. Receptors for histamine, peptides, different allergens, and neuronal growth and survival factors and eicosanoids, which are connected via phospholipases (i.e. phospholipases C and D) to PKC signaling, are all good candidates to modulate VR1 activity of sensory neurons. Nerve growth factor (NGF), abundantly synthesized de novo after spinal cord injury and inflammation, was noted to couple to PKC. Administration of NGF to neonatal and adult rats in vivo was reported to produce hyperalgesia through activation of PKC (27). Axonal growth and differentiation experiments in cultured PC-12 rat pheochromocytoma cells pointed toward the novel PKC⑀, a Ca 2ϩinsensitive isotype, as an effector enzyme for NGF-mediated intracellular signaling (28,29). NGF pretreatment also was noted to potentiate CAP-induced inward currents in voltageclamped DRG neurons, suggesting either synergy or additively (30). Moreover, other neurotrophins, such as neurotrophins-3 can influence the central terminals of their respective afferent neurons after injury (31). Among several mRNAs, NGF can induce transcription of VR1 in dissociated cultures of adult DRG, as well as VR1 protein levels in undamaged DRG neurons after partial nerve injury (32,33). Together these observations suggest that NGF may engender hyperalgesia either by de novo transcription and translation or by direct, PKC-mediated activation of VR1. The regulation of VR1 by PKC is also supported by the PKC⑀Ϫmediated increase in VR1 ion currents in DRG neurons and in cells ectopically expressing recombinant vanilloid receptor (8). In addition, or as an alternative to direct signaling via PKC, NGF was observed to act as a liberator of VR1 from phosphatidylinositol 4,5-bisphosphate-mediated blockade (26).
In the current study, we present evidence suggesting a direct role of PKC in VR1 activation. In particular, full expression of PKC␣ was found to be essential for acute activation of VR1 with PDBu, a synthetic, stable agonist of classic and novel PKCs. This action was examined in primary DRG cultures as well as NIH 3T3 and HeLa cells stably transfected to express VR1⑀, a C-terminal ⑀-epitope-tagged VR1 recombinant (4). Each system has a different, complementary composition of PKC isozymes, which permitted discrimination of the isoforms that affected VR1 function. NIH 3T3 cells were determined to express classical PKC␣, novel PKC␦ and -⑀, and atypical PKC isoforms, respectively (34). In contrast, PKC⑀, a previously proposed upstream regulator of VR1 was not expressed in HeLa and was not required for activation of ectopically expressed VR1. Our findings in DRG cultures, VR1⑀-NIH 3T3 and VR1⑀-HeLa cell lines, strongly suggest that PKC␣ can sensitize and/or activate VR1 in the absence of endogenous ligands, including anandamide (ANA), an endovanilloid/eicosanoid agonist of VR1 in acidosis (5). In VR1⑀-HeLa cells, another permanent line we established, acute PDBu induced ionophore activity of VR1 even in the absence of PKC⑀. ANA activates VR1 in an additive fashion with PKC, yet is independent of PKC, because ANA activation is retained after chronic PDBu-mediated down-regulation of PKC␣. Similar to ANA (5), the PKC activation of VR1 requires a drop of extracellular pH Ͻ 6.5, indicating that the protonated VR1 conformer serves as a better substrate of PKC. Our observations emphasize the conditional requirements for the actions of endovanilloids and PKC and their capacity to work in a concerted, additive fashion on the nociceptive nerve terminal.

EXPERIMENTAL PROCEDURES
Preparation of VR1⑀ Expressing NIH 3T3 Cell Line-C-terminaltagged chimeric rat VR1⑀ was prepared in the p⑀MTH vector plasmid vector (35), as described previously (4). To develop cell lines permanently expressing VR1⑀, we used mouse NIH 3T3 fibroblasts and human HeLa adenocarcinoma-derived cells. The parental lines were determined to have low levels of Ca 2ϩ transport that were not altered by treatment with vanilloids, ANA, or PDBu. To avoid toxicity that occurs with VR1 overexpression, VR1⑀ was expressed in the cell lines using only the basal activity of the metallothionein promoter, as described earlier (5). For the purpose of clarity, in sentences where PKC⑀ is discussed in conjunction with VR1⑀ (e.g. VR1⑀-HeLa) we refer to it as VR1 (e.g. VR1-HeLa).
DRG Culture-DRG neuron-enriched cultures were prepared from embryonic rats (E16) (4). Briefly, DRGs were dissected and then pro-cessed in fresh dissection medium (Lebowitz medium, Invitrogen) until plated in Dulbecco's modified Eagle's medium. The Dulbecco's modified Eagle's medium contained 20 mM HEPES (to prevent acidification and stabilize pH at 7.4), 7.5% fetal bovine serum, 7.5% horse serum, 5 mg/ml uridine supplemented with 2 mg/ml 5-fluoro-2Ј-deoxyuridine, and 40 ng/ml NGF to inhibit cell division and promote neuronal survival and differentiation, respectively. Surfaces for cell culture were coated with poly-D-lysine and laminin. Cells were seeded on 25-mm glass coverslips or on multiwell microtiter plates. Cultures were selected in this medium for 1 week, at which point well differentiated neurons dominated the population. Primary DRG cultures in this stage were used in radioactive and fluorescent video microscopy Ca 2ϩ uptake assays.
Down-regulation and Recovery of PKC-To down-regulate PKC isozymes, DRG neurons and VR1 expressing NIH 3T3 and HeLa cell lines were chronically treated with 2 M PDBu in complete medium for 24 h. To determine the effect of chronic PDBu on PKC protein levels, individual PKCs were analyzed with isozyme-specific antibodies. The influence of PKC down-regulation on acute ANA and PDBu-induced Ca 2ϩ uptake was studied in short-term (10 min) experiments, as described above. Recovery of PKC␣ from the down-regulated state (2 M PDBu, 24 h) was monitored in confluent cultures of VR1⑀-NIH3T3 cells. Cells were chronically treated in 96-well plates with PDBu (2 M, 24 h), one set of 12 wells at a time, then changed back into complete growth medium without PDBu and cultured over the 15-day period of the recovery experiment, changing the growth medium every 3rd day. Western blotting experiments were carried out on nontreated cells and cells chronically treated on the 6th, 7th, 13th, and 15th day before cell harvest. In a parallel set of cultures, the effect of recovery of PKC␣ was studied using acute PDBu-induced Ca 2ϩ uptake as the VR1-dependent end point.
Ca 2 Transport-45 Ca 2ϩ uptake experiments were carried out on 1-week-old primary DRG cultures (3 ϫ 10 4 cells/well) and on established VR1⑀-NIH 3T3 and HeLa cell lines (3 ϫ 10 4 cells/well), which were seeded 1 day before use. Immediately before the assay, cells were adapted to room temperature (24°C) for 5 min in Hanks' balanced salt solution (pH 6.0), supplemented with 10 M Ca 2ϩ and 0.1 mg/ml bovine serum albumin (HCB). 45 Ca 2ϩ uptake was performed for 10 min at 24°C in HCB using 0.2 Ci of 45 Ca 2ϩ as radioactive tracer in a 200-l final volume. To determine the pH dependence of the 45 Ca 2ϩ uptake, HCB was buffered with 20 mM Tris-HCl, adjusted to the indicated pH with 1 M MES (HCBTM). To stop 45 Ca 2ϩ uptake, cells were rapidly changed back into 1 ml of HCB, washed two additional times with 1 ml of HCB, and then lysed in 200 l/well of RIPA buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% Triton X-100, 0.5% deoxycholate, 0.1% SDS, 5 mM EDTA) for 30 min. Aliquots of the solubilized cell extracts were counted in a liquid scintillation counter.
Fluorescent Video Microscopic Determination of [Ca 2ϩ ] i -For determination of [Ca 2ϩ ] i an Olympus BX 60 microscope equipped with a low-light fluorescence intensifier system was used, as described earlier (36). DRG cultures were preloaded with 5 M fluo-4 AM dye for 30 min at 34°C, and then washed three times in HCB to remove excess dye and kept in the dark for at least 15 min before starting the experiments. Recordings were carried out in a closed imaging chamber (PH-2, Warner Instrument Corp., Hamden CT), which was perfused with a pump (Miniplus 3, Gilson, France). The emitted fluorescence intensity was calculated from images taken at 10-s intervals by the synapse 3.6e software, an image acquisition and analysis program, as recommended by the manufacturer (Synergy Research, Silver Spring, MD). To monitor vanilloid-induced changes in [Ca 2ϩ ] i , after 1 min of baseline recording, DRG neuron cultures were perfused with 1 M CAP in HCBTM for 30 s, then the medium was changed back to drug-free HCBTM.
Preparation of VR1 Specific Antibody and Western Blotting-The hydrophilic C-terminal fragment of rat VR1 (Met 682 -Lys 838 ), extended with a C-terminal (His) 6 tag was amplified by the polymerase chain reaction employing GGAAGGATTTCAGAATTCATGGGTGAGACCGT-CAACAAGATT as forward primer and TGTCGACCTCGACTCAATGA-TGATGATGATGATGTTTCTCCCCTGGGACCATGGAATC as reverse primer. The PCR-amplified cDNA fragment incorporated the EcoRI and SalI restriction sites from the forward and reverse primers, respectively, and these sites were used to insert the fragment into the pMA-Lc2x (New England BioLabs) prokaryotic expression vector. The peptide fragment was cloned in-frame and downstream of the coding region for maltose-binding protein (MBP). Escherichia coli cells were transformed with the MBP-VR1C(His) 6 construct and production of the fusion protein was induced by 1 mM isopropyl-1-thio-␤-D-galactopyranoside for 3 h. The MBP-VR1C(His) 6 chimeric protein was double affinity purified using the MBP and His tags, consecutively, as recommended by the manufacturers (New England BioLabs and Novagen). The affinity purified MBP-VR1C(His) 6 protein was used to immunize two rabbits. After 8 weeks of immunizations at biweekly intervals, sera were tested at 1:500 dilutions in Western blotting experiments using SDS extracts prepared from neuron-enriched DRG cultures of E16 embryonic rats.
Materials-ANA (1:4 in a soybean oil:water emulsion), was purchased from Tocris and PDBu from Calbiochem. 45 Ca was purchased from ICN. PKC isozyme-specific antibodies were obtained as follows: PKC␣, clone M6 mouse monoclonal from Upstate Biotechnology; PKC␦ and -⑀ affinity purified polyclonals from Invitrogen; and PKC from Calbiochem. Affinity purified anti-RAKIGQGTKAPEEKTANTISK, a specific PKC␤ peptide antibody was prepared in rabbits and characterized as described earlier (37).

RESULTS
To study PKC-dependent up-regulation of VR1, permanent cell lines ectopically expressing an ⑀-epitope-tagged recombinant of VR1 were established in both NIH 3T3 (4) and HeLa cells. VR1 specificity and pH dependence of acute (10 min) PDBu-induced 45 Ca 2ϩ transport were studied in the parental, immortalized NIH 3T3 mouse fibroblast and VR1⑀-NIH 3T3 cell lines at pH 5.5 and 7.5, as indicated (Fig. 1). In the parental line, the basal Ca 2ϩ uptake at pH 7.5 was about 4 times higher than at 5.5, and addition of PDBu to the assay medium had little or no effect on the basal uptake at either pH. Expression of recombinant VR1 elevated the baseline compared with parental NIH 3T3 cells at either pH 5.5 or 7.5. At pH 6.5, 7.0, and 7.5 only minor increases in calcium uptake occurred with ascending PDBu concentrations. However, the net change for VR1⑀-NIH 3T3 cells was not different from the parental cell line at pH 7.5 (open triangles), indicating that these elevations are not dependent on the expression of VR1. The marked transition in ionophore activity occurs with PDBu treatment when the VR1⑀-NIH 3T3 cells are at pH 5.5. Addition of PDBu at submicromolar concentrations increased Ca 2ϩ transport in a concentration-dependent manner between 3-and 4-fold over baseline as determined in repeated experiments ( Figs. 1 and 2). An increase in Ca 2ϩ uptake could be detected at around 20 nM and the EC 50 of PDBu-induced 45 Ca 2ϩ uptake at pH 5.5 was ϳ100 nM, and reached a plateau at concentrations above 1 M (Figs. 1 and 2a).
To further study the PKC specificity of PDBu-induced, VR1mediated 45 Ca 2ϩ uptake at pH 5.5, bisindolylmaleimide (BIS), an inhibitor directed to the catalytic ATP-binding site of PKC, was co-incubated with increasing concentrations of PDBu (Fig.  2a). BIS (25 M) added together with progressively increasing concentrations of PDBu completely inhibited the inducible portion of the 45 Ca 2ϩ uptake, but had only slight or no effect on baseline uptake in VR1⑀-NIH 3T3 cells. Studies with VR1 selective inhibitors showed that acute PDBu-induced 45 Ca 2ϩ transport was nearly completely inhibited by 10 M ruthenium red (a blocker of the VR1 ionophore), but not with a high concentration (25 M) of capsazepine (CPZ, a competitive antagonist at the vanilloid ligand-binding site) (Fig. 2b).  45 Ca 2ϩ uptake assays were carried out as described under "Experimental Procedures." At pH 5.5, the basal Ca 2ϩ uptake was lower than at pH 7.5, but PDBu clearly activated the Ca 2ϩ transport in a dose-dependent manner (2-4-fold determined in independent experiments), and reached saturation at 1 M (EC 50 ϳ 150 nM). In contrast, no strong induction with PDBu was noted in the parental NIH 3T3 cells at either pH 5.5 or 7.5. In comparison with the parental cells at pH 7.5, PDBu failed to produce a net activation of Ca 2ϩ uptake at pH 6.5, 7.0, and 7.5 in the stable VR1expressing cells. Each point on the graph is the average of triplicate determinations. Experiments were repeated two additional times in triplicate with similar results. Western blotting with specific antibodies was employed to analyze expression of VR1 and different PKC isozymes in DRG cultures and VR1⑀-NIH 3T3 cells (Fig. 3). The DRG cultures from E16 embryonic rats were enriched for neurons by a 1-week treatment with 5-fluoro-2Ј-deoxyuridine to eliminate dividing cells. Cultures were also treated with NGF to promote neuronal differentiation. Before Western blotting, the DRG cultures were morphologically and functionally characterized with phase-contrast and fluorescence video microscopy. After 1 week, the DRG cultures typically contained large (25-30 m in diameter), medium (ϳ15-25 m), and small size (7-10 m) neurons in close to equal ratio. The CAP-induced increase of intracellular [Ca 2ϩ ] i in the DRG cultures is one of the markers we used to identify of VR1-expressing nociceptors. After loading cultures with fluo-4, a Ca 2ϩ -sensitive fluorescent dye, cells within the latter two populations (ϳ30% of the total) responded to 1 M CAP with elevated green fluorescence (Fig. 3, a versus b).
VR1 expression was determined immunochemically by a rabbit antibody, raised against MBP-VR1C(His) 6 , a soluble Cterminal fragment of rat VR1. The immune sera recognized a prominent band at 92 Ϯ 4 kDa in DRG extracts (Fig. 3c), which corresponds well with the theoretical molecular mass of rat holo-VR1 (94.9 kDa). In addition to the 94.9-kDa protein, our antibody detected a 114-kDa glycosylated form of VR1 (38), in extracts from VR1⑀-NIH 3T3 cells (Fig. 3d). The expression of different isoforms of PKC in extracts prepared from either control or chronic PDBu-treated DRG cultures or VR1⑀-NIH 3T3 cells was determined by Western blotting employing isotype-specific antibodies. The DRG cultures expressed PKC␣, in addition to PKC␤, -␦, -⑀, and -isozymes; the latter have been reported previously in cultures from neonatal animals (12). Chronic PDBu treatment had the most dramatic effect on the PKC␣ isozyme, which was almost completely down-regulated in both DRG neurons and VR1⑀-NIH 3T3 cells (Fig. 3, c and d). However, chronic PDBu produced only a partial down-regula-tion of the PKC␤, -␦, and -⑀ isoforms in DRG neurons. Similar effects were seen in the VR1⑀-NIH 3T3 cells (Fig. 3, c and d) except that no PKC␤ was measured, which is in accordance with previous observations in the parental NIH 3T3 cell line (34) (Fig. 3d). As expected from a nonphorbol-binding isotype, little or no change was noted in the levels of PKC in either cell system.
The effects of chronic PDBu-induced down-regulation on 45 Ca 2ϩ uptake were studied in a parallel set of DRG cultures as those analyzed above for VR1 and PKC isozyme expression. Chronic PDBu almost completely eliminated the acute PDBuinducible activity as assayed by a short, 10-min re-exposure to PDBu in the presence of 45 Ca 2ϩ (Fig. 4a). Elimination of the acute PDBu effect correlated with the complete down-regulation of PKC␣ in DRG cultures chronically treated with PDBu ( Figs. 3c and 4a). In control DRG cultures, addition of ANA, a vanilloid agonist in acidic conditions, induced 45 Ca 2ϩ uptake at pH 5.5 with similar kinetic parameters as previously characterized (5). After chronic PDBu treatment, the ANA-induced Ca 2ϩ transport remained intact, although shifted slightly into a lower affinity region; an EC 50 of ϳ20 M compared with ϳ10 M was obtained in the experiment shown (Fig. 4b). Parallel experiments repeated in VR1⑀-NIH 3T3 cells also resulted in complete inhibition of acute PDBu-induced 45 Ca 2ϩ uptake (Fig.  5a), but left the ANA-elicited 45 Ca 2ϩ transport intact (Fig. 5b).
NIH 3T3 cells appear to survive growth arrest in the confluent state without loss of viability, and we took advantage of this c, cultures (ϳ250,000 neurons) were extracted in hot SDS sample buffer and proteins were size separated by 10% PAGE, then electroblotted to nitrocellulose membranes. Levels of PKC isozymes were determined by isotype-specific antibodies before and after chronic PDBu treatment (24 h). VR1 was determined by a rabbit antibody raised against the affinity purified MBP-VR1C-(His) 6 antigen, as described. Baseline DRG cultures expressed VR1 and PKC␣, -␤, -␦, -⑀, and -isozymes. Chronic PDBu down-regulated PKC␣ almost completely, however, the levels of PKC␤, -␦, -⑀, and VR1 were only partially affected. d, the same antigens were analyzed in VR1-NIH 3T3 cell lines. In contrast to DRG, PKC␤ is not expressed in VR1-NIH 3T3 cells. Similar to the DRG, ϳ90% loss of PKC␣ was detected with only minor effects, if any, on the other isozymes. Levels of PKC did not change in either DRG or VR1-NIH 3T3 cells which served as an internal control for protein loading. The asterisk stands for the glycosylated form (110 kDa) of VR1 detected in cell lines ectopically expressing VR1. Similar results were obtained in a second, independent experiment. property to examine recovery of PKC sensitivity following PDBu-induced down-regulation. Confluent cultures of VR1⑀-NIH 3T3 cells were treated with 2 M PDBu for 24 h in complete medium. Recovery of PKC␣ was monitored for 2 weeks by Western blot analysis using a monoclonal antibody specific for PKC␣ (Fig. 6, upper panel). No recovery of PKC␣ from the down-regulated state was detected during the following week (days 6 and 7 after chronic PDBu treatment), however, by 12 days a progressive re-appearance of PKC␣ was seen and by day 15 nearly complete replenishment had occurred. A set of acute, 0.2 M PDBu-induced Ca 2ϩ uptake experiments at pH 5.5 were performed in parallel. Coincident with the recovery of PKC␣ by Western blot, the acute PDBu-induced Ca 2ϩ -uptake activity also returned to levels near baseline. These data further support the essential role of PKC in direct activation of VR1.
Activation of VR1 by the endovanilloid ANA and PKC activator PDBu appear to represent two independent and potentially additive mechanisms. The inhibition studies in Fig. 2b show that acute PDBu-induced 45 Ca 2ϩ transport is not blocked with 25 M CPZ, a competitive vanilloid antagonist, but is blocked with 10 M ruthenium red, which inhibits Ca 2ϩ entry through the VR1 ion channel (39). The VR1 specificity of the 45 Ca 2ϩ uptake was analyzed by co-application of either 25 M ruthenium red or 10 M CPZ at ascending concentrations of ANA at pH 5.5. Either ruthenium red or CPZ eliminated the ANA inducible portion of 45 Ca 2ϩ uptake, indicating that ANA has to specifically interact with VR1 to induce cation transport (Fig. 7). Co-application of 1 M PDBu with each concentration of ANA shifted the dose-response curve to the left and augmented 45 Ca 2ϩ uptake in a manner characteristic of additive and independent routes for VR1 activation (Fig. 7).
A previous study (8) suggested PKC⑀ as a unique upstream regulator of VR1 in DRG nociceptors. The present experiments suggest rather that PKC␣ is required for intracellular activation of VR1 in DRG neurons. A remarkable correlation between down-regulation of PKC␣ and loss of acute PDBu-induced, VR1-mediated Ca 2ϩ transport was determined both in embryonic rat DRG cultures and in VR1⑀-NIH 3T3 cells. To examine Experiments were carried out in VR1⑀-NIH 3T3 cells in acidic assay medium (pH 5.5). The maximum effect of ANA induced 45 Ca 2ϩ uptake ϳ5 times above the baseline at pH 5.5. An additive effect was seen with co-incubation of 1 M PDBu and increasing concentrations of ANA (filled triangles). The addition of PDBu elevated the V max of Ca 2ϩ transport, and shifted the EC 50 of ANA from ϳ10 to ϳ4 M in VR1⑀-NIH3T3 cells. As expected, CPZ, a competitive inhibitor of vanilloid ligand binding, and ruthenium red (RR), a selective channel blocker of VR1, almost completely inhibited ANA action. Each point represents the mean Ϯ S.E. of triplicate determinations. Similar results were obtained in two independent experiments also performed in triplicate. the requirement for PKC⑀ in VR1 activation, a permanent VR1-expressing HeLa cell line (VR1⑀-HeLa) was established. This cell line exhibits a different PKC isozyme composition (notably PKC⑀ is not expressed) (40) than either the NIH 3T3 cells or the DRG neurons. The HeLa line stably expressing VR1 showed approximately a 10-fold elevation of 45 Ca 2ϩ uptake above baseline when treated with a maximal dose of CAP (10 M) or resiniferatoxin (10 nM) at pH 5.5 (Fig. 8a). We corroborated the lack of PKC⑀ in the VR1 expressing HeLa cells (40) with isozyme-specific Western blots (Fig. 7b). Similar to parental cells, the VR1-HeLa subclone expresses PKC␣, Ϫ␤, -␦, and -, but not -⑀ (Fig. 8b). As seen in DRG neurons and NIH 3T3 cells, chronic treatment with PDBu almost completely downregulated PKC␣ from VR1Ϫexpressing HeLa cells. At the same time little or no change in the levels of PKC␤, -␦, and -or VR1 were detected (Fig. 8). Coincident with the loss of PKC␣, the VR1-expressing HeLa cells exhibited inhibition of acute PDBuinduced 45 Ca 2ϩ uptake (Fig. 9a), without significant impact on Ca 2ϩ uptake induced by the VR1 agonist ANA (Fig. 9b). As an aside, we note that the VR1-expressing HeLa line showed about 5-fold higher affinity toward ANA and high concentrations of ANA (above 10 M) caused an inhibition of 45 Ca 2ϩ uptake, determined at pH 5.5, which was not seen in VR1expressing NIH 3T3 cells (Figs. 5b and 6 versus 9b). DISCUSSION PKC mediated up-regulation of VR1 was investigated in DRG neurons, NIH 3T3, and HeLa cell lines, with the latter two permanently expressing VR1⑀, a C-terminal ⑀-epitopetagged recombinant of rat vanilloid receptor type 1. We observed that PDBu, a synthetic agonist capable of binding to the regulatory domain of classical and novel PKC isozymes, activated Ca 2ϩ uptake in the absence of any added vanilloid ligand in all three cell preparations. PDBu-evoked channel activation exhibited a strong pH dependence: VR1⑀-NIH 3T3 and VR1⑀-HeLa cells responded to less than 1 M PDBu at acidic pH (pH 5.5), but not at pH 6.5 or higher (up to pH 7.5). We previously observed that ANA activated VR1 45 Ca 2ϩ uptake in cells ec-topically expressing VR1 or in primary DRG cultures only when the assay medium was acidic (pH Ͻ 6.5) (5). Protonation of VR1 in an acidic environment does not directly cause channel opening (6) or Ca 2ϩ uptake (Fig. 1), but increases the affinity of the receptor for vanilloid and eicosanoid ligands (5). This observation prompted us and others to suggest that VR1 exists in two distinct conformations with H ϩ serving as an allosteric switch (5,41,42). Between pH 6.5 and 7.5, protons dissociate from VR1, perhaps in the pore loop region (43,44), yielding a conformer with lower affinity toward vanilloids and vanilloid-like eicosanoids such as ANA (5,6). However, the protonated form of VR1 (between pH 6.5 and 5.5) displays an increased affinity toward vanilloid ligands, which coincides with a decreased capacity for cation transport (5). The fact that PDBu induces 45 Ca 2ϩ transport through VR1 when the medium is acidic is consistent with the hypothesis that the protonated conformer of VR1 is also the preferred substrate of PKC. Sequence analysis of rat and human VR1 indicates 18 intracellular sites that conform to PKC consensus motifs for potential phosphorylation. This highlights the need for determining the exact phosphorylation site(s) involved in VR1 regulation.
Studies with selective inhibitors of VR1 showed that PDBu can induce 45 Ca 2ϩ transport in the presence of 25 M CPZ, a competitive inhibitor of vanilloid ligand binding (Fig. 2b). In contrast to CPZ, ruthenium red, a blocker of the VR1 ion channel (39) can prevent 45 Ca 2ϩ uptake activated by both vanilloid agonists or PKC (Figs. 2b and 6). In addition, results with BIS indicate that the catalytic activity of PKC is required for VR1-mediated 45 Ca 2ϩ uptake (Fig. 2). BIS eliminates the effect of acute PDBu when they are co-applied to VR1⑀-NIH 3T3 and VR1⑀-HeLa cells, suggesting that phorbol esters activate VR1 only indirectly, via intact catalytic activity of PKC. In contrast to a previous report (26), but in concert with findings of others, our PKC down-regulation experiments also indicate that PDBu does not directly interact with VR1 (45-47). However, from the literature, it was not clear which isozymes are expressed and mediate VR1 activation in sensory ganglia neu- rons. Data from different studies are not completely in accord, consequently, different conclusions about which PKC isozyme regulates VR1 have been reached (12,21,25). In our DRG cultures the presence of VR1 was detected with physiological methods using a CAP-induced increase in [Ca 2ϩ ] i , as measured by fluo-4, and biochemically by Western blotting. In addition to VR1, among the PKC isoforms, we detected the presence of PKC␣, a result similar to that reported for 3-month-old rats (25), which is in contrast to what was reported for cultures prepared from 2-day-old animals (12). The ability to positively detect PKC␣ may be related to differences in antibodies used or in methodology of the primary culture systems (Fig. 3).
Activation of VR1 by second messenger systems appears to occur through multiple signaling pathways involving PKA and PKC (7,8). Although PKC⑀ has been proposed as an upstream regulator of VR1 by others (8,15,16), the expression of PKC⑀ does not appear to be a necessary component for acute PDBuinduced VR1 activation in DRG neurons, and clearly not in VR1-HeLa cells that lack PKC⑀ (Figs. 3 and 9a). The results of chronic PDBu down-regulation and recovery experiments support a key role for the PKC␣ isozyme in activation of VR1. Following chronic PDBu treatment (24 h, 2 M PDBu), PKC␣ was almost completely down-regulated in DRG neuron cultures and NIH 3T3 and HeLa cell lines ectopically expressing VR1, whereas the levels of other isozymes either did not change, or as in the case of PKC⑀, decreased only partially. The coincident recovery of PKC␣ and PDBu-induced acute Ca 2ϩ uptake, 2 weeks after the down-regulation, further emphasizes the essential role of this isozyme in VR1 expressing cells.
Previously, HeLa, a human carcinoma cell line, was found not to express PKC⑀ (40) and we confirmed its absence in our clonal VR1-HeLa cell line by Western blot analysis (Fig. 8). The loss of PKC␣ correlated with loss of the acute activation by PDBu on 45 Ca 2ϩ uptake in all three cell cultures. However, direct activation through the vanilloid agonist route remained intact because stimulation by ANA exhibited little alteration (Figs. 4, 5, and 9). These experiments indicate an important role of PKC␣ isozyme on VR1 function through intracellular signaling pathways. Down-regulation results in DRG and VR1 expressing cell lines suggest that substantially reduced levels of PKC␣ (Figs. 3, 4, and 8) compromise, whereas the absence of PKC⑀ does not affect acute response to PDBu (Fig. 8a), again emphasizing an essential role for PKC␣ in VR1 activity. In contrast to the behavioral alteration noted in PKC⑀ mutant mice, results at the cellular levels with PDBu support a more direct role of PKC␣ in VR1 activation. A PKC⑀-derived, selective inhibitor peptide decreased the mechanical hyperalgesia only incompletely, produced by epinephrine injection into the hindpaw of normal rats. Together with results of others (8,16,26) it seems that PKC⑀ might be involved in multiple nociceptive signaling pathways, including epinephrine-induced hyperalgesia in mechanosensor neurons, which lack VR1 expression (1). The partially attenuated pain phenotype noted in PKC⑀ knock-out mouse nociception models does not rule out a function for other PKC isotypes and PKC⑀ may serve as an alternative upstream regulator of VR1 in nociceptors (15) in addition to PKC␣. Taken together, these data suggest that, in addition to heat, at least two independent signaling pathways can regulate the ionophore function of VR1 and hence nociceptive and inflammatory pain signaling. The first pathway utilizes vanilloid ligands or eicosanoid/endovanilloid/endocannabinoid compounds such as ANA, leukotriene B (4), 12-and 15-(S)-hydroperoxyeicosatetraenoic acids, and 5-and 15-(S)hydroxyeicosatetraenoic acids (5,6). The second pathway uses PKC, and likely functions as an intracellular second messenger-coupled effector for extracellular inflammatory agents. Both pathways are conditionally dependent on pH and we hypothesize that the protonated form of VR1 is the preferred substrate for PKC. The two pathways produce additive effects on VR1-mediated 45 Ca 2ϩ uptake. However, they can be dissociated, as shown by the selective effect of chronic PDBu downregulation on acute activation of VR1; ANA/vanilloid signaling is preserved, whereas signaling through the PKC pathway is eliminated. Thus, intracellular signaling by PKC isozymes is not necessary for ligand-induced, direct activation of VR1, but PKC isozymes can open the VR1 ionophore in the absence of extracellular vanilloid-like ligands. The importance of the PKC␣ isoform in VR1 regulation, in comparison with the ⑀ isoform, was evident in chronic PDBu down-regulation experiments carried out with VR1-NIH 3T3 cells and DRG neurons. Furthermore, in the VR1-HeLa cell line, acute PDBu was as effective as in our other VR1 expressing preparations, despite its lack of the PKC⑀ isoform.
ANA and PDBu co-incubation experiments highlight the independence of the two pathways as well as the potential for mutual interaction of nociceptive stimuli to reinforce and sustain pain signaling. The number of different signal transducers that are upstream of different PKC isozymes in primary afferent endings is still an open question. Various agonists implicated in pain, including NGF, bradykinin, and substance P have receptors in nociceptors and are coupled to different phospholipases C and D isozymes. These lipases, upon activation, produce diacylglycerol at the plasma membrane (19,48,49), FIG. 9. To demonstrate the role of PKC␣ in activation of VR1, experiments were carried out with the same set of VR1-HeLa cells characterized in Fig. 7, which are deficient in PKC⑀. a, chronic PDBu inhibited the acute effect of PDBu on VR1-mediated 45 Ca 2ϩ uptake. b, no inhibition on ANA-induced short-term 45 Ca 2ϩ transport was seen with a 24-h PDBu exposure but a reduced affinity is obtained similar to that observed in VR1-NIH 3T3 cells. Note also the inverse concentration response in this cell line. Similar results were obtained in a second independent experiment. which is able to activate PKC generally, and in particular the classic PKC␣ in sensory neurons, as suggested here, or novel PKC⑀ isotypes proposed previously by others (12,15,16,29,50).
In conclusion, PKC-induced activation of VR1 is a novel regulatory pathway, which deserves further investigation. Similar to endovanilloid ligand-induced ionophore activity of VR1, a drop in pH below 6.5 is necessary for PDBu-evoked calcium uptake. PKC⑀ was proposed previously as one of the likely isozymes to activate VR1 by transmembrane signaling. Here we positively identify PKC␣, in thoroughly characterized DRG neuron cultures from embryonic rats, as a necessary factor for acute PDBu-induced activation of VR1. The PKC⑀ isoform does not appear to be absolutely necessary for VR1 activation as shown by VR1-HeLa, a permanent cell line not expressing PKC⑀. An interesting aspect of PKC isozyme-specific regulation of VR1 is that certain isotypes may be therapeutic targets for novel pain treatments. Specific inhibition or elimination of PKC␣ in the periphery may result in effective pain management in nerve injuries where upstream regulators of PKC isozymes are overproduced and may contribute to intractable pain states.