Molecular Pain Activation of Protein Kinase C in the Spinal Cord Produces Mechanical Hyperalgesia by Activating Glutamate Receptors, but Does Not Mediate Chronic Muscle-induced Hyperalgesia

Background: Protein kinase C (PKC) in the spinal cord appears to mediate chronic injury-induced pain, but not acute nociceptive pain. Muscle insult results in increased release of glutamate spinally, and hyperalgesia that is reversed by spinal blockade of NMDA and non-NMDA glutamate receptors. Therefore, we hypothesized that spinal activation of PKC 1) mediates the late phase of hyperalgesia 1 week after muscle insult, and 2) produces mechanical hyperalgesia through activation of NMDA and non-NMDA glutamate receptors.


Background
Protein kinase C activation involves translocation from the cytosol to binding domains at cell membranes of dorsal horn neurons of the spinal cord [1,2]. There are at least twelve isoforms of PKC. Several of these isoforms are concentrated in the superficial laminae of the dorsal horn, an anatomical indication that these PKC isoforms play a potential role in nociceptive signaling. In particular, PKCβ I, PKCβ II, and PKCα are found in cell bodies within the superficial dorsal horn, where PKCγ is primarily found in cell bodies in lamina IIii [2,3]. PKC is involved in many aspects of cellular sensitization, including modulation of channel conductivity by phosphorylation, increased trafficking of receptors to the cell membrane, and release of excitatory neurotransmitters [4][5][6][7][8][9].
Activation of PKC with phorbol esters in the spinal cord decreases heat and mechanical withdrawal thresholds, increases glutamate release in the spinal cord, and sensitizes spinothalamic tract and other dorsal horn neurons [9][10][11][12]. PKC is involved in animal models of both neuropathic and inflammatory pain. In rats with neuropathic pain produced by spinal nerve ligation or sciatic nerve ligation the mechanical hyperalgesia is reversed by intrathecally administered PKC inhibitors and reduced in PKCγ knockout mice when compared to wild-type mice [3,[13][14][15]. Similarly, spinal blockade of PKC reverses the hyperalgesia induced by subcutaneous formalin, pancreatitis, thermal injury, cutaneous capsaicin, diabetic neuropathy and subcutaneous bee venom [10,[16][17][18][19][20]. In PKCγ knockout mice acute responses to thermal and mechanical stimuli are similar to wild-type mice [3], suggesting a role for PKC in more chronic injury-induced pain, but not in acute nociceptive pain.
Activation of cAMP (cyclic adenosine monophosphate) spinally activates intracellular pathways that results in sensitization of spinal neurons and mechanical hyperal-gesia. A decrease in mechanical paw withdrawal threshold produced by intramuscular injections of acid or capsaicin is reversed by spinal blockade of cAMP-PKA pathway in the spinal cord 24 h, but not 1 week, following muscle insult [21,22]. These studies demonstrate a role of the cAMP-PKA pathway in the early phase of development, but not in the later phase. We further show an increased release of glutamate in the spinal dorsal horn, and that blockade of NMDA and non-NMDA glutamate receptors 1 week after muscle insult [23,24]. Since PKC appears to mediate more chronic pain conditions we hypothesized that activation of PKC mediates the late phase of hyperalgesia 1 week after muscle insult. We further hypothesized that activation of PKC in the spinal cord produces mechanical hyperalgesia through activation of NMDA and non-NMDA glutamate receptors.

PDBU-Dose
Two intramuscular injections of pH 4 saline 5 days apart produced a bilateral decrease in mechanical withdrawal thresholds 24 h and 1 week after the second injection. Spinal blockade of PKC with GF109023X (0.05-0.5 nmol/10 µl) had no effect on the decreased mechanical withdrawal threshold when delivered intrathecally 24 h or 1 week after the second injection of acidic saline compared to vehicle controls (Fig. 3).

Discussion
The current study demonstrates that activation of PKC in the spinal cord produces a decrease in mechanical withdrawal threshold that is prevented by blockade of PKC with GF109023X. This PKC-mediated decrease in mechanical withdrawal threshold is consistent with previous data showing a similar decrease in mechanical withdrawal threshold by spinal activation of PKC [9,10]. We further show that the decreased mechanical withdrawal threshold produced by spinal activation of PKC is prevented by spinal blockade of NMDA receptors and AMPA/ kainate receptors. There are several possibilities that could explain how PKC activation produces its effects through glutamate receptors. First, PKC could produce increased release of glutamate into the spinal cord resulting in continued activation of glutamate receptors. Spinal activation of PKC increases release of glutamate in vivo [9] and in vitro [12], and formalin-induced release of glutamate is prevented by blockade of PKC [17]. Further, spinal activation of PKC enhances responses of dorsal horn neurons to NMDA and AMPA agonists [12,25,26]. Thus increased glutamate release and increased activation of glutamate receptors are likely results of activation of PKC.
Second, PKC could phosphorylate the NMDA or AMPA/ kainite receptor to result in increase channel conductance or potentiation of glutamate gated currents that would be manifested as an increased excitability of the neuron [27][28][29][30]. Activation of PKC sensitizes dorsal horn neurons exhibited as an increased spontaneous firing, increased response to peripherally applied mechanical stimuli [31], and an increased response to glutamate and NMDA [12]. Further, PKC decreases Mg2+ affinity in the NMDA receptor pore which increases the probability of the channel opening [26]. Indeed, in animal models of pain there is an increase in phosphorylation of the NR1 subunit and GlurR1 subunit that likely depends on activation of PKC [32][33][34][35][36].
Third, phosphorylation of NMDA or AMPA receptor subunits could increase trafficking and insertion of receptors into the cell membrane and synapse. Phosphorylation of NR1 by PKC increases surface expression of NMDA receptors, increases delivery of NMDA receptors to the surface and to the synapse [28,37,38]. The AMPA receptor subunits are also phosphorylated by PKC [27,29]. However, the functional role of phosphorylation of AMPA receptor subunits in receptor trafficking has not been tested to date. Thus, increased phosphorylation of glutamate receptors could result in an increased channel conductance and an increased number of receptors available synaptically resulting in increased excitation of the nociceptive spinal neurons.
Lastly, PKC decreases efficacy of inhibitory neurotransmitters on spinothalamic tract neurons, which would be manifested as an increased excitation. For example intradermal injection of capsaicin reduces the inhibition of spinothalamic tract neurons normally produced by electrical stimulation of the periaqueductal gray or by GABA; this loss of inhibition is prevented by spinal blockade of PKC [39]. Similarly, the inhibition of calcium channels by µ-opioid agonists in the spinal dorsal horn of rats is prevented by inhibition of PKC [40]. Thus, increased PKC activity reduces normal inhibition within the spinal cord.
Thus, spinal activation of PKC could result in increased glutamate release, increased channel conductance of glutamate receptors, increased number of glutamate receptors in the membrane, and decreased inhibition. Taken together this would result in increased excitability of neurons that is manifested as decreased withdrawal thresholds to noxious stimuli.
Surprisingly, spinal blockade of PKC had no effect on the decreased mechanical withdrawal threshold induced by repeated intramuscular acid injections suggesting PKC Time course of effects of intrathecal pretreatment with AP5 or NBQX on the decreased withdrawal threshold induced by PDBu, for 3 nmol dose (top graph) and for the 10 nmol dose (bottom graph)  injury [44]. Taken together, it appears that PKC is involved in long-term hyperalgesia, and that PKCγ and PKCβ II may be important in the processing of nociceptive information in chronic hyperalgesia. However, blockade of PKC in the spinal cord had no effect on the acidinduced model of muscle pain. This model is unique; there is no detectable injury to the peripheral muscle after injection of acidic saline and hyperalgesia is maintained by changes in the central nervous system [45]. Thus, differences could be related to the dependence on continued primary afferent input to the spinal cord. Differences could also result from the type of tissue activated: i.e. cutaneous vs. deep somatic tissue. Prior studies utilize animal models with cutaneous inflammation, i.e. formalin, capsaicin, paw carrageenan, paw CFA, cutaneous thermal injury, or nerve injury which likely includes cutaneous afferent damage. The mechanisms responsible for cutaneous pain are distinctly different from those related to muscle pain. In support, we previously show that capsaicin injected into cutaneous tissue results in short lasting (hours) mechanical hyperalgesia while capsaicin injected into muscle or joint results in long-lasting mechanical hyperaglesia (days to weeks) [21]. Formalin injected into skin of the back resulted in fos expression in laminae I-V; whereas formalin injected into the paraspinal muscles did not result in fos staining in laminae II [46]. Formalin injection into the muscle, when compared to the skin, resulted in greater fos expression in brain areas known to mediate descending facilitation of nociception, i.e. amygdala, and the ventrolateral periaqueductal gray II [46]. Thus different anatomical pathways could mediate differences between muscle pain and those in other cutaneous pain models.
The current study showed that 2 or the 3 PKC inhibitors tested were ineffective against PDBu, an activator of PKC. GF109023X has been shown to be more efficacious in inhibition if substance P-induced nocifensive behaviours compared to chelerythrine chloride [47]. In the current study, GF109023X significantly reduced the effects of 10 nmol PDBu; while chelerythrine chloride and NPC15437 had no effect on the reduction in withdrawal thresholds produced by NPC15437. It is possible, based on the differences in efficacy of the PKC inhibitors, that an effect could be observed if tested against lower doses of PDBu. Of note, the majority of the previous studies did not test the PKC inhibitor utilized against the agonist. The doses of PKC inhibitors utilized in the current study, were similar or higher than those utilized previously [47][48][49][50]. However, dosing with chelerythrine chloride and NPC15437 were limited by solubility for in vivo delivery of the antagonist to the spinal cord intrathecally. Our prior study using microdialysis was able prevent the effects of activation of PKC with NPC15437 [10]. This different methodology may be able to deliver a higher effective dose closer to the neurons than a single intrathecal bolus since it utilizes continued slow diffusion into the spinal cord parenchyma.
Although the current data support that blockade of glutamate receptors prevents the onset of PKC-induced hyperalgesia, it is possible that these glutamate receptor antagonists were analgesic on their own and thus masking the effects of PDBu. One limitation to testing mechanical thresholds with von Frey filaments is that we cannot detect an analgesic effect as the cut-off for withdrawal is essentially our baseline response before injury. The doses of AP5 and NBQX used in the current study, however, have no significant effects on motor behaviors 15-120 minutes after delivery of the drug intrathecally [23,51,52]. Further, intrathecal treatment using AP5 (10 nmol), similar doses to those in used in the current study, had no effect on the hyperalgesia that develops after plantar incision [52,53]. These data argue against the analgesic potential of these antagonists at the doses utilized, and thus support our conclusions that PDBu produces its effects through activating NMDA and non-NMDA receptor antagonists.

Conclusion
Activation of the PKC pathway in the spinal cord produces mechanical hyperalgesia that is prevented by blockade of PKC, NMDA, or AMPA/kainite receptors. However, activation of PKC in a model of non-inflammatory muscle pain has no role in the maintenance of the hyperalgesia.

Methods
All experiments were approved by the animal care and use committee at the University of Iowa and are in accordance with NIH guidelines and the International Association for the Study of pain. A total of 123 male Sprague-Dawley rats (250-450 g; Harlan, St. Louis, MO) were utilized in these studies.

Intrathecal catheter placement
Animals were placed under general anesthesia using 2% halothane. Catheters (32 G polyeurethane tubing, 10 cm; ReCathCo, Allison Park, PA) were attached to PE10 tubing (0.28 mm ID, polyeurethane, 10 cm; Becton Dickson and Company, Sparks, MD) and threaded through a 23 G needle. The 23 G needle was inserted into the intrathecal space between L5/L6 until a tail flick was elicited confirming placement intrathecally. The tubing was then advanced 4 cm so the tip was located in the lumbar enlargement. The catheter was secured to the fascia, the PE10 tubing was threaded out between the shoulder blades, and the incision was closed with silk sutures. Animals were allowed to recover from anesthesia under observation prior to testing 3-5 days later.

Mechanical withdrawal threshold
Before behavioral testing, rats were placed in clear Plex-iglas™ chambers on a wire mesh elevated surface to acclimate for 30-40 min. Withdrawal thresholds from mechanical stimuli of von Frey filaments of ascending bending force (from 10.0 to 494.6 mN) were applied twice to the plantar surface of the bilateral hind paws. A positive response was defined as withdrawal from the von Frey filament. Confirmation of threshold was tested by examining the filament above and below the withdrawal response. Test-retest reliability was previously determined for this method (r 2 = 0.7; p = 0.007) [54]. A decrease in mechanical withdrawal threshold is interpreted as mechanical hyperalgesia.

Drug administration
Drugs were administered to awake animals through intrathecal (i.t.) catheters. A 30 G drug delivery needle attached to a length of PE50 tubing was affixed to the end of a 50 µl Hamilton syringe for drug administration through i.t. catheters. Drugs were delivered as a 10 µl bolus. Following the completion of all experiments correct catheter placement was confirmed by intrathecal injection of 10 µl of 2% lidocaine followed by 10 µl methylene blue dye. A catheter was considered correctly placed if there was loss of sensory and motor function after lidocaine and methylene blue dye staining included the L4-L6 spinal segments.

Data analysis
Paw withdrawal thresholds were compared across time, and between groups, using statistical analysis with a repeated measures analysis of variance (ANOVA). Post hoc testing between individual groups was done using a Tukey's test. Significance was determined at p ≤ 0.05. Data is presented as the mean +/-S.E.M.
Publish with Bio Med Central and every scientist can read your work free of charge