INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Sensitization of spinal cord dorsal horn neurons, including spinothalamic tract (STT) cells, to peripheral cutaneous stimuli due to many types of injuries is thought to contribute to secondary hyperalgesia and allodynia. Several experiments done previously by our group have shown that sensitization of STT neurons is dependent on activation of excitatory amino acid (EAA) and neurokinin (NK) receptors (Dougherty and Willis 1991; Dougherty et al. 1992-1994; Neugebauer et al. 1999). Experimentally, sensitization of STT cells induced by intradermal injection of capsaicin usually lasts about 1.5-2 h (Dougherty and Willis 1992; Lin et al. 1996a). EAAs and SP cannot alone account for such a prolonged change in central neuronal function since the activation of EAA and NK₁ receptors is likely to produce depolarizations, such as those underlying wind-up, that last only for seconds to minutes. It has been well documented that signal transduction cascades are activated during the development of central sensitization following tissue injury (Coderre 1992; Coderre and Yashpal 1994; Garry et al. 1994; Lin et al. 1996a, 1997a, 1999a; Meller and Gebhart 1993; Meller et al. 1994) and participate in the production of a prolonged hypersensitivity of dorsal horn neurons. Mechanisms underlying this process include an important contribution of N-methyl-D-aspartate (NMDA) receptors (and consequent Ca²⁺ influx when NMDA channels open) to central sensitization (MacDermott et al. 1986; Sorkin et al. 1999); the role of several excitatory G-protein-coupled receptors, including NK₁ receptors and metabotropic glutamate receptors, in central sensitization has also been demonstrated (Dougherty et al. 1994; Neugebauer et al. 1999, 2000; Womack et al. 1988). Influx of Ca²⁺ and activation of metabotropic glutamate receptors are well known to initiate intracellular signal transduction cascades.

The cAMP transduction cascade is associated with several G-protein receptors found in the spinal cord such as the prostaglandin (Hingtgen et al. 1995), calcitonin gene-related peptide (Bushfield et al. 1993; Santicioli et al. 1995; Sun and Benhishin 1995), NK1 (Satoh et al. 1992; Smith et al. 1992), and metabotropic glutamate (Schoepp and Johnson 1993; Schoepp et al. 1992) receptors. Activation of these receptors would result in binding of adenylate cyclase to the G-protein-linked receptors. If the receptors are associated with a stimulatory G protein, adenylate cyclase converts ATP to cAMP, resulting in an increase in formation of cAMP in the cell (Sassone-Corsi 1998). The increase in cAMP then activates protein kinase A (PKA), resulting in phosphorylation of proteins involved in neurotransmitter release (Hell et al. 1995) or ion channels, such as EAAs or Ca²⁺ channels (Blackstone et al. 1994; Hell et al. 1995; Sculptoreanu et al. 1995).

In the spinal cord, the role of PKA in central sensitization of dorsal horn neurons induced by intradermal injection of capsaicin has been suggested by a behavioral study (Sluka 1997; Sluka and Willis 1997). cAMP content in the dorsal horn following peripheral inflammation has been measured in a few studies, but the results are conflicting (Garry et al. 1994; Igwe and Ning 1994; Przewlocka et al. 1991). In a series of studies by our group (Lin et al. 1996a,b, 1997a, 1999a-c), evidence was obtained that the protein kinase C (PKC) and nitric oxide (NO)/protein kinase G (PKG) transduction cascades in the spinal cord are activated when STT neurons are sensitized following intradermal injection of capsaicin, and the elevation either of PKC or NO/PKG within the spinal cord can sensitize STT cells to mechanical stimulation. Therefore this study examined the possible role of activation of the PKA signal transduction cascade in the processing of nociceptive signals by recording STT neuron responses to mechanical stimuli. Adenylate cyclase was activated by infusing forskolin, an activator of adenylate cyclase (Laurenza et al. 1987), into the spinal cord. These data have appeared in abstract form (Lin et al. 1997b).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

General preparation and anesthesia

All experiments were approved by the animal care and use committee at our institution.

Adult male monkeys (Macaca fascicularis, 1.9-3.9 kg) were initially tranquilized with ketamine (10 mg/kg im). Anesthesia was induced with a mixture of halothane, nitrous oxide, and oxygen followed by -chloralose (60 mg/kg iv) and maintained by sodium pentobarbital (5 mg · kg¹ · h¹ iv). Animals were paralyzed with pancuronium (0.4-0.5 mg/h iv) and ventilated artificially to maintain the end-tidal CO₂ between 3.5 and 4.5% and the arterial oxygen saturation between 96 and 100%. Core body temperature was regulated at approximately 37°C using a thermostatically controlled heating blanket. The electrocardiogram was also recorded to monitor the heart rate and function. The level of anesthesia was frequently checked during the experiment by examining pupillary size and reflexes and observation of end-tidal CO₂ level. A laminectomy exposed the lumbar enlargement. A craniotomy was performed for stereotaxic placement of a stimulating electrode into the ventroposterolateral (VPL) nucleus of the thalamus. The stereotaxic coordinates were: A, 8 mm; L, 8 mm; 16-19 mm from the cortex surface. To ensure correct placement, the electrode was initially used to record the potentials evoked by electrical stimulation of the contralateral dorsal columns and responses to cutaneous stimulation of the contralateral hindlimb.

Placement of microdialysis fibers

For drug application, two to three microdialysis fibers (Spectrum Scientific, 18-kDa cutoff; 150 µm ID; wall thickness, 9 µm) were positioned in the lumbar enlargement in areas most responsive to stimulation of the lower hindlimb, as described previously (Lin et al. 1999a-c). The fibers were coated with a thin layer of silicone rubber (3140RTN, Dow Corning) except for a 1-mm-wide zone that was positioned in the gray matter of the ipsilateral spinal dorsal horn. Each fiber was pulled through the spinal cord just below the dorsal root entry zone using a stainless steel pin cemented in the fiber lumen. The fibers were placed in different segments of the lumbar enlargement (L₅-L₇), no less than 10 mm apart. The microdialysis fibers were connected to a syringe seated in a Harvard infusion pump and were continuously perfused with artificial cerebrospinal fluid (ACSF) (Sorkin et al. 1988) at a rate of 5 µl/min. The ACSF was oxygenated and equilibrated to pH 7.4 with a mixture of 95% O₂-5% CO₂. In the beginning of the experiments, ACSF was pumped through the fiber for at least 1.5 h to wash out substances released during fiber insertion.

Administration of drugs

The following drugs (concentrations in the dialysis fluid given in parentheses) were administered by microdialysis at 5 µl/min for 20-30 min through the fiber closest to the recorded STT neuron: forskolin (an activator of brain membrane adenylate cyclase, 5 mM) (Hackman et al. 1997; Laurenza et al. 1987); 1,9-dideoxy-forskolin (D-forskolin, an inactive isomer of forskolin used as a control, 5 mM); N-[2-((p-bromocinnamyl)amino)ethyl]-5-isoquinolinesulfonamine (H89, an inhibitor of PKA, 0.01 mM) (Chijiwa et al. 1990; Sluka et al. 1997); 8-bromoguanosine-3',5'-cyclophosphate sodium (8-bromo-cGMP, a membrane permeable analogue of cGMP, 10 mM) (Lin et al. 1997a). The concentrations of drugs in the dialysate were presumed to be approximately two to three orders of magnitude higher than the concentrations that reach neurons in the dorsal horn. Our group has studied the diffusion across the microdialysis fiber in vitro of several similar-sized drugs with quite different chemical properties. The concentration ratio across the microdialysis fiber for all these drugs was between 1 and 4% (Dougherty et al. 1992; Sluka et al. 1993). All drugs were dissolved in ACSF and pH was corrected to 7.2-7.4.

Recording of STT neurons

STT neurons were recorded extracellularly in the lumbar enlargement of the spinal cord using low-impedance (3-5 M) carbon filament electrodes. Cells were searched for within 250-750 µm of the edge of a dialysis fiber to ensure that the drugs administered by microdialysis would reach the neuron in a short period of time at a sufficient concentration. Single STT neurons were isolated following antidromic activation from the contralateral VPL nucleus by square-wave current pulses (0.3 Hz, 0.75 mA, 200 µs). Criteria for antidromic activation (Trevino et al. 1973) were as follows: constant latency of the evoked spike; ability to follow high-frequency (333-500 Hz) stimulation; and collision of orthodromic spikes with antidromic spikes. The extracellularly recorded signals were amplified and displayed on analog and digital storage oscilloscopes. Signals were also fed into a window discriminator, the output of which was processed by an interface (CED 1401) connected to a Pentium PC. Peristimulus rate histograms were constructed on-line with Spike-2 software. Digital records of single-unit activity were also stored for off-line analysis. Throughout the experiment, spike size, and configuration were continuously monitored on the digital oscilloscope and with the use of Spike-2 software to confirm that the same neuron was recorded (see Fig. 1, insets) and that the relationship of the recording electrode to the neuron remained constant.

View larger version (25K): [in this window] [in a new window]   Fig. 1. Rate histograms show the responses of a spinothalamic tract (STT) neuron to cutaneous mechanical stimuli before and when the spinal dorsal horn was perfused with forskolin by microdialysis. Top: baseline responses to cutaneous stimuli (brush, press, pinch). Second row: responses recorded during forskolin infusion. press and pinch, but not brush, responses were enhanced. Third and bottom rows: responses obtained 0.5 and 1.5 h after the end of drug infusion, respectively. Horizontal lines above histograms show times of application of the stimuli to five preselected sites in the receptive field. Insets: the expanded spikes before, during, and after forskolin.

Experimental protocol

The background activity and responses of an STT cell, once identified and isolated, to innocuous and noxious mechanical stimuli were recorded and the receptive field on the skin of the hindlimb was mapped. Cells were characterized by their responses to the following stimuli applied to the most responsive sites in the receptive field: brush (response to innocuous brushing of the skin with a soft hair brush in a stereotyped manner), press (response to firm pressure on a fold of skin by a large arterial clip that exerted a force of 1,005 g/8 mm², which is near threshold for pain), and pinch (response to a distinctly painful stimulus by a small arterial clip that exerted a force of 2,660 g/4 mm²; this stimulus did not cause overt damage to the skin). Cells were divided into three classes: low threshold (LT), wide dynamic range (WDR) and high threshold (HT), as described previously (Chung et al. 1986). Most cells included in this study were WDR STT neurons that responded consistently to innocuous stimuli and that were strongly activated by noxious stimuli. However, three HT cells were observed. These were excited primarily by noxious stimuli. Each type of stimulus was delivered at five test points chosen to span the receptive field. Each stimulus was applied for 10 s followed by a 10-s pause before the next test site was stimulated. To minimize a potential "human factor" bias, the experimenter who applied the mechanical stimuli did not observe the oscilloscope or the computer monitor and was unaware of the response magnitude.

CELLS TESTED DURING FORSKOLIN AND D-FORSKOLIN ADMINISTRATION. Forskolin was infused into the dorsal horn through a microdialysis fiber after control responses were recorded, and the effects of activation of PKA on responses of STT cells to peripheral cutaneous stimuli were tested while forskolin was being infused for 20-30 min. Drug within the spinal cord was then washed out with ACSF for 1-1.5 h, after which the responses were recorded again. For control purposes, an inactive isomer of forskolin, D-forskolin, was administered while recording from different STT cells than those used for the above experiments.

CELLS TESTED TO EXAMINE THE BLOCKADE OF THE FORSKOLIN EFFECT BY H89. H89 was infused into the dorsal horn for 20-30 min before forskolin was administered to test if the effect produced by forskolin could be blocked by H89. Additionally, the selectivity of H89 in blocking forskolin-induced effects at the dose used in this study (0.01 mM) was examined by testing if H89 could affect the sensitization of STT cells resulting from activation of PKG (Lin et al. 1997a, 1999b). To do this, 8-bromo-cGMP was administered after the spinal cord was pretreated with H89 for 20-30 min by microdialysis.

Data analysis

Responses evoked by mechanical stimulation were analyzed off-line from peristimulus time histograms using Spike-2 software after subtraction of background activity. Responses to the mechanical stimuli applied to five points across the receptive field were summed to yield a total discharge rate for each type of stimulus. Total discharge rates of responses (brush, press, and pinch) and background activity were calculated for each cell under control conditions, during drug infusion and after the drug washout period. Statistical significance was tested using ANOVA with repeated measures and post hoc paired t-tests for differences from the baseline levels. A value of P < 0.05 was considered significant. All values are given as the means ± SE.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Recordings were made from 43 identified STT neurons in the lumbar enlargements (L₅-L₇) of 26 monkeys. Of these, 40 were classified as WDR cells and 3 as HT cells. These neurons were recorded at depths of 860-1,972 µm from the dorsal surface of the spinal cord. Thus these neurons were presumed to be in laminae I-V (see Lin et al. 1999a). Thirty-six cells were used to examine the effects of forskolin and to examine if forskolin acted through the PKA pathway by testing if pretreatment of the spinal cord with H89 could specifically block the forskolin-induced responses. These cells included 33 WDR neurons and 3 HT cells. Seven WDR cells were used for control experiments, in which D-forskolin was administered.

Changes in the responses of STT neurons to mechanical stimuli produced by spinal administration of forskolin

In a total of 23 STT neurons (including 2 HT cells), responses to innocuous and noxious mechanical stimuli were tested while forskolin was infused into the spinal dorsal horn by microdialysis. A consistent increase in responses to press and pinch stimuli was seen in all STT cells tested when forskolin was administered; but in most STT cells tested, the responses to brush stimuli were not obviously changed when forskolin was given. Neurons were considered responsive when changes in responses to mechanical stimuli were increased more than 30% from the baseline value. According to this criterion, an increase in responses to brush stimuli was seen in only eight of the STT cells tested (34.8%). Figure 1 shows a typical experiment on one WDR neuron. The responses to press and pinch stimuli were enhanced when forskolin was infused (2nd row), and the increases in the responses outlasted the drug infusion period for more than half an hour of wash out (3rd row). The effect of forskolin recovered partially when the drug had been washed out for approximately 1.5 h (bottom). Forskolin produced no obvious effect on the responses to brush stimuli (left). The statistical analysis of grouped data showed a significant increase in both press- and pinch-evoked responses (3rd and 4th groups in Fig. 2A). Because forskolin produced inconsistent effects on responses to brush stimuli, obvious changes in the brush response in some individual cells were not sufficient to cause the population response to reach statistical significance (2nd group in Fig. 2A). In the same cells (n = 23), the effect of forskolin on background activity was also tested. An increase in background activity was seen in only nine cells (39.1%). Of these, an increase in response to brush stimuli was observed in seven cells. This means that, for those STT cells whose responses to brush stimuli were enhanced, the background activity in most of them (7 of 8) was also increased following forskolin infusion. A statistical significance for the grouped data for the effect on background activity was obtained only at 30 min after the beginning of wash out. (1st group in Fig. 2A).

View larger version (31K): [in this window] [in a new window]   Fig. 2. A and B: bar graphs summarize, respectively, the effects of infusion of forskolin (n = 23) or D-forskolin (n = 7) on the averaged background activity and responses of cells to mechanical stimuli. BKG, background activity. *P < 0.05; **P < 0.01, ***P < 0.001, compared with baseline.

In seven additional cells, an inactive isomer of forskolin, D-forskolin, was administered through a microdialysis fiber in the same way as for forskolin. No obvious changes were observed either in responses to mechanical stimuli or in background activity (Fig. 2B).

Blockade of forskolin-induced enhancement of responses to mechanical stimuli when the spinal cord was pretreated with a PKA inhibitor

Because forskolin is thought to be capable of activating adenylate cyclase leading to an increase in cAMP level, thus activating PKA, we propose that the enhancement of responses to mechanical stimuli induced by forskolin could be due to the activation of PKA. The following experiments examined whether a blockade of PKA could selectively interfere with the enhanced responses produced by forskolin. In one group of nine STT cells (including 1 HT neuron), the spinal dorsal horn was pretreated with a PKA inhibitor, H89, before forskolin was infused. Figure 3 consists of rate histograms for a representative STT cell that showed a blockade of forskolin-induced enhancement by H89. H89 pretreatment itself produced no significant changes in background activity or in the responses of the cell to peripheral cutaneous stimuli (2nd row). However, the enhanced responses to mechanical stimuli produced by forskolin infusion were nearly completely blocked (3rd row). The grouped data summarized in Fig. 4 show that H89 effectively blocked the forskolin-induced enhancement. An identical result was obtained from one HT cell that was included in this group.

View larger version (29K): [in this window] [in a new window]   Fig. 3. Rate histograms show that the enhanced effects of forskolin on responses of an STT cell to peripheral cutaneous stimuli were blocked by pretreatment of the spinal cord with N-[2-((p-bromocinnamyl)amino)ethyl]-5-isoquinolinesulfonamine (H89). Top: baseline responses. Second row: responses recorded during H89 infusion. Third row: responses when forskolin was infused immediately after H89 infusion. Bottom: 0.5 h after the end of forskolin infusion. Horizontal lines above histograms represent times of application of mechanical stimuli to 5 preselected sites in the receptive field.

View larger version (27K): [in this window] [in a new window]   Fig. 4. Bar graphs summarize, respectively, the effects of H89 pretreatment within the spinal cord on the enhanced responses of STT cells to peripheral cutaneous stimuli produced by forskolin or 8-bromo-cGMP. The left set of bars in each panel represent the forskolin-induced responses without H89 pretreatment [forskolin (artificial cerebrospinal fluid), n = 23]. The middle set of bars in each panel represents forskolin-induced responses with H89 pretreatment [forskolin (H89), n = 9]. The right set of bars in each panel shows the responses produced by 8-bromo-cGMP infusion with H89 pretreatment [8-Bromo-cGMP (H89), n = 6]. *P < 0.05, **P < 0.01, ***P < 0.001, compared with the baseline.

H89 has been shown to inhibit PKA selectively with a Ki value of 48 nM, but it can also produce a weak inhibitory action on other kinases (Chijiwa et al. 1990). For instance, a higher dose of H89 may inhibit PKG at a Ki value of 480 nM, which is only an order of magnitude different from the Ki value for PKA (Chijiwa et al. 1990). Therefore we chose to test the PKG pathway in a control experiment to determine if H89 blockade of the forskolin-induced effect was specific for PKA. The PKG pathway was activated by intra-spinal infusion of 8-bromo-cGMP, which has been shown to sensitize STT cells in our previous work (Lin et al. 1997a), after the spinal cord was pretreated with same dose of H89. As expected, 8-bromo-cGMP administration enhanced the responses to all of mechanical stimuli. When the spinal cord was pretreated with the same dose of H89, the changes were not prevented (Fig. 4).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study demonstrated that application of forskolin, an activator of neuronal membrane adenylate cyclase, which in turn, activates PKA, into the spinal dorsal horn produced long-lasting changes in responsiveness of STT neurons. The effects included a substantial increase in responses of all STT cells tested to press and pinch stimuli lasting more than 1 h after infusion and a slight increase in background activity. In most cells, responses to brush stimuli remained unchanged by the forskolin administration. Several control experiments were done to test the selectivity of the forskolin effect: infusion of an inactive isomer, D-forskolin, failed to affect the background activity or responses of STT cells; pretreatment of the spinal cord with a PKA inhibitor, H89, at a dose of 0.01 mM (Chijiwa et al. 1990; Sluka et al. 1997) blocked completely the forskolin-induced effect; and the same dose of H89 did not affect the enhanced responses of STT cells to mechanical stimuli induced by activation of PKG by spinal infusion of 8-bromo-cGMP. These experiments imply that the effect of forskolin was due to PKA activation.

cAMP-dependent protein kinase (PKA) has been implicated in prolonged changes in synaptic efficacy, notably long-term potentiation (LTP) in the hippocampus (Frey et al. 1993; Huang et al. 1995). The initiation of LTP involves NMDA, non-NMDA, and metabotropic glutamate receptors, influx of Ca²⁺ ions through EAA and voltage-gated Ca²⁺ channels, and activation of a number of signal transduction cascades, including the PKA cascade (see review by Dingledine et al. 1999). Therefore it is presumed that the central sensitization of dorsal horn neurons resembles LTP by functioning in the spinal cord as a transient "memory" of a painful event. Several lines of evidence have shown a role of PKA in the enhanced responsiveness of spinal dorsal horn neurons: a dose-dependent hyperalgesia and allodynia can be induced when the spinal dorsal horn is perfused with 8-bromo-cAMP (Sluka 1997); secondary hyperalgesia and allodynia produced by intradermal injection of capsaicin can be reduced in a dose-dependent manner by posttreatment of the spinal cord with a PKA inhibitor (Sluka 1997); injection of cAMP or the catalytic subunit of PKA enhances the responses of dorsal horn neurons to glutamate-gated ion channel activation (Cerne et al. 1992, 1993); inhibition of PKA reverses the sensitization of dorsal horn cells to mechanical stimuli produced by intradermal injection of capsaicin (Sluka et al. 1997). Thus activation of the PKA pathway is one of the mechanisms responsible for central sensitization.

A number of neurotransmitter receptors are involved in the induction of central sensitization, including non-NMDA and NMDA (ionotropic receptors) (Dickenson and Sullivan 1987; Dougherty et al. 1992; Neugebauer et al. 1993), NK1 and metabotropic glutamate (G-protein-linked) receptors (Dougherty et al. 1994; Neugebauer et al. 1995, 1999; Sluka et al. 1997), as are high-voltage calcium channels (Chaplan et al. 1995; Malmberg and Yaksh 1994). Activation of Ca²⁺ influx through NMDA or Ca²⁺ channels or increased intracellular release of Ca²⁺ by activation of G-protein-linked channels could lead to increased accumulation of cAMP intracellularly and activation of PKA. Protein phosphorylation would be evoked to regulate the functions of receptors, such as glutamate receptors (Dingledine et al. 1999). It has been demonstrated that serine/threonine phosphorylation of NMDA receptors is mediated by PKC and PKA on different serine residues (Leonard and Hell 1997; Tingley et al. 1997). We have recently reported that the NR1 subunit of NMDA receptors in STT neurons is phosphorylated on Ser-897 following intradermal injection of capsaicin (Zou et al. 2000). Therefore combined with the above information, our study supports the view that the end result of PKA activation could be an increase in the effectiveness of ionotropic channels and/or increased neurotransmitter release (Blackstone et al. 1994; Hell et al. 1995), which would contribute to central sensitization, even though the current experiments were unable to determine if the site of drug action is presynaptic on primary afferent terminals or on interneurons presynaptic to STT cells.

We have found from the current study that when PKA was activated by spinal infusion of forskolin, the responses of STT neurons to noxious mechanical stimuli were significantly enhanced in all neurons, but responses to innocuous mechanical stimuli generally were not changed. In previous work, we showed that activation of PKC in the dorsal horn by a spinally administered phorbol ester enhanced the responses of STT cells to innocuous brushing but not to noxious compression of the skin (Lin et al. 1996b), and administration of 8-bromo-cGMP or SIN-1 (a NO donor) enhanced both types of responses (Lin et al. 1997a, 1999a). Thus a series of studies, including the present one, seems to indicate that activation of different protein kinase cascades produce different effects on the responsiveness of STT neurons. PKC may play a major role in the process of sensitization of dorsal horn neurons chiefly to innocuous stimuli, which contributes to allodynia. PKG may be involved in the mediation of central sensitization that produces both allodynia and hyperalgesia. In contrast, activation of the PKA cascade seems to play a major role in mediation of sensitization of dorsal horn neurons to noxious stimuli, which contributes to secondary hyperalgesia. It has been shown in previous work by our group that blockade of spinal NMDA receptors by AP7 did not affect the responses of STT cells to brushing stimuli but substantially reduced the responses to pinch stimuli (Dougherty et al. 1992), suggesting that NMDA receptors are involved in the responses of STT neurons to noxious but not to innocuous mechanical stimuli. Anatomical data obtained recently showed that the NR1 subunits of NMDA receptors in STT cells were phosphorylated on the PKA-modulated site, Ser-897, following intradermal injection of capsaicin (Zou et al. 2000). Thus activation of PKA would predominantly upregulate the NMDA receptors in STT cells by phosphorylation of NR1 subunits to enhance the function of NMDA receptors that underlies the mechanism of central sensitization of dorsal horn neurons to noxious stimuli. However, the data obtained from this study cannot exclude completely a possible role of PKA in mediating secondary allodynia because we have found that the responses to brush stimuli were increased in about 35% of STT cells tested following forskolin administration, and the background activity in most of these cells was increased as well. Behavioral experiments have indicated that activation of the cAMP cascade in the spinal cord contributes to both mechanical hyperalgesia and allodynia (Sluka 1997). In future experiments, we will investigate if there is a functional change in spinal inhibitory and excitatory receptors when PKA is activated to elucidate further the role of the PKA in central sensitization of dorsal horn neurons.