Plastic Change of N-type Ca Channel Expression after Preconditioning Is Responsible for Prostaglandin E²-induced Long-lasting Allodynia

All experiments were conducted under the ethical guidelines for the study of experimental pain in conscious animals, 15and the protocol of the pain behavioral studies described here was approved by the Animal Care Committee of Tokyo Medical and Dental University.

We used male Sprague–Dawley rats weighing 250–300 g (Japan Clea, Sizuoka, Japan) for behavioral and immunoblot analysis and heterozygous alleles of N-type channel knockout (Ca^v2.2 +/−) mice 14weighing 28–35 g for 5-bromo-4-chloro-3-indolyl β-d-galactopyranoside (X-Gal) staining. The animals were housed in a temperature- and humidity-controlled environment with a 12-h light–dark cycle and free access to food and water.

The current study used a recently introduced experimental system with some modification. In the original study, long-lasting hypersensitivity to proinflammatory mediators (100 ng/5 μl prostaglandin E²[PGE²]) after the resolution of carrageenan (1%; 5 μl)–induced acute inflammation was evaluated by the Randall–Selitto test for the nociceptive flexion reflex. 16In our preliminary study, however, it was difficult to detect reproducible mechanical hypersensitivity under a similar condition. Therefore, we used greater amounts of carrageenan (2%; 100 μl) and PGE²(400 ng/20 μl) and evaluated a long-lasting mechanical hypersensitivity using von Frey hairs (Stoelting Company, Wood Dale, IL). Carrageenan is a classic agent causing experimental inflammation, and the induced inflammatory pain is considered relevant to clinically important inflammatory pain states. 17,18PGE²is one of the best-characterized proinflammatory mediators known to be present in increased concentration at sites of inflammation and is shown to induce hyperalgesia and allodynia by a direct action on primary afferent nociceptors. 1,2,19 

All behavioral experiments were performed in a soundproof room during the light cycle (7:00 am to 7:00 pm). Rats were acclimatized to the experimental room for at least 1 h before the experiments.

Tactile allodynia was evaluated with calibrated von Frey hairs by measuring the tactile stimulus producing a 50% likelihood of a withdrawal response (50% gram threshold), which was calculated by using the up–down paradigm of Dixon as described in detail before. 20A stimulus strength of 15.1 g was adopted as the upper limit of testing.

All subcutaneous and intrathecal injections were performed using a 30-gauge needle during a few minutes of anesthetization with halothane, 2% (Takeda Chemical Industries, Ltd., Osaka, Japan). In this study, rats and the Ca^v2.2 +/− mice received a solution, 2%, of λ-carrageenan (Sigma Chemical Company, St. Louis, MO) in saline in the plantar surface of the hind paw (100 μl for rats and 20 μl for mice). The injection side was selected randomly. Control experiments were conducted by injection of the same volume of physiologic saline solution. Seven days after carrageenan or saline injection, rats received either GVIA (0.03, 0.1, or 0.3 μg) or AgaIVA (0.01, 0.03, or 0.1 μg) intrathecally in a total volume of 10 μl 30 min before PGE²injection (400 ng/20 μl) into the pretreated plantar surface of the hind paw. Each animal received no more than a single dose of the Ca channel blockers. Saline (10 μl) was administered intrathecally in the control experiments. Although the effects of GVIA were known to be relatively slow in onset, 21this interval was sufficient to establish maximal inhibitory effects of GVIA in our preliminary study.

AgaIVA and GVIA (Peptide Institute, Inc., Osaka, Japan) and PGE²(Cayman Chemical Company, Ann Arbor, MI) were made up as concentrated stock solutions in MilliQ water (Nihon Millipore, Tokyo, Japan) or 10% ethanol in saline for PGE², divided into aliquots, and stored at −20°C. Aliquots were thawed and diluted to the desired concentrations in saline immediately before use. The doses of these drugs were determined according to previous studies 22–25and our preliminary study.

Rats were deeply anesthetized with intraperitoneal sodium pentobarbital (45 mg/kg) at 7 days after carrageenan or saline injections, and L4–L6 DRGs (ipsilateral to the injections) or the spinal cord was dissected. After dissection, rats were killed by further injection of excess doses of sodium pentobarbital (100 mg/kg). Pooled DRGs from at least eight rats were homogenized in 700 μl lysis buffer, pH 8.0, containing 150 mm NaCl, 1 mm EDTA, and protease inhibitors (0.2 mm phenylmethylsulfonyl fluoride, 1 μg/ml aprotinin, 1 μg/ml pepstatin A, 1 μg/ml leupeptin, 0.1 mg/ml benzamidine). The 100,000 g  membrane fraction was dissolved in sample buffer (10 mm Tris/HCl, pH 6.8, containing 0.005% Coomassie brilliant blue-G, 1.5% sodium dodecyl sulfate, 2 m urea, and 10 mm dithiothreitol), and the proteins (150 μg per lane) were resolved on a 5% polyacrylamide gel. Protein concentrations were determined with a protein assay kit (Bio-Rad Laboratories, Hercules, CA). The proteins were transferred onto a polyvinylidene difluoride membrane (Immobilon P; Millipore, Billerica, MA), and the blot was probed with a rabbit polyclonal anti-α^1Bantibody (Exalpha Biologicals, Watertown, MA) by using the ECL system (Amersham Bioscience, Piscataway, NJ). The N-type channel protein levels from seven independent experiments using four different pooled samples were quantified using the analysis software NIH image 1.62 (National Institutes of Health, Bethesda, MD). After immunoblotting, the membranes were stained with SYPRO Orange (Molecular Probes, Eugene, OR) and scanned with a FluorImager 595 (Molecular Dynamics, Sunnyvale, CA), and the total transferred proteins were quantified with IMAGE QUANT software (Molecular Dynamics) to normalize the N-type channel protein band intensities.

For each set of experiments, the normalized band intensity of the saline control sample was set as 100, and then the ratio of the normalized band intensity of the carrageenan-treated sample to the saline control sample was calculated. For negative controls, antibody was excluded or incubated with α^1Bantigen peptide (Alomone Laboratories, Jerusalem, Israel).

Seven days after carrageenan injection, Ca^v2.2 +/− mice were anesthetized with sodium pentobarbital and subjected to fixation by transcardial perfusion with paraformaldehyde, 4%, in phosphate-buffered saline. X-Gal staining was performed on spinal cord sections as reported previously. 26Briefly, after perfusion fixation, the spinal cord was dissected, rinsed with phosphate-buffered saline, and then stained with X-Gal solution (Sigma Chemical Company). Because the gene for Escherichia coli β-galactosidase with a nuclear localization signal was inserted in frame in the N-type channel gene cacna 1b , the β-gal activity monitored by X-Gal staining is considered to represent the native expression of the gene encoding the N-type channel. 14Nuclear localization is useful for excluding the contribution of the channel protein, which exists on the terminals of primary sensory neurons. Thus, this method is more advantageous than immunoblot analysis using dissected spinal cord samples, which inevitably include the central terminal component of primary sensory neurons. The levels of staining intensity in the spinal dorsal horn were quantified with NIH image and calculated as the ratio of the carrageenan-treated side to the contralateral side.

Experimental data in the figures are mean ± SEM. The immunoblot and X-Gal staining data were tested according to the Student t  test. For behavioral analyses, we used the Mann–Whitney U test for comparisons between two groups and the Steel–Dwass test for multiple comparisons. P < 0.05 was considered statistically significant.

We observed that a 100-μl carrageenan (2%) challenge resulted in swelling, erythema, and reduced threshold of paw withdrawal to mechanical stimuli with von Frey filaments in rats. These alterations began 30–60 min after injection, reaching a maximum in 2–4 h, and resolved within 3 to 4 days. Consistent with findings of numerous previous reports, this routinely used dose of carrageenan 27,28produced only short-term hyperalgesia and allodynia from which the rats could fully recover.

Intradermal injection of PGE²at the same site where carrageenan had been injected 7 days earlier resulted in a prolonged mechanical allodynia lasting more than 4 h (fig. 1). In contrast, PGE²produced only transient allodynia lasting less than 1 h (acute allodynia) in saline-injected control rats (fig. 1). Furthermore, the mean values of 50% thresholds of the early (30 min) and late (4 h) phases of PGE²-induced allodynia in carrageenan-preconditioned rats were significantly lower than those for their respective control derived from saline-preconditioned rats (fig. 1). PGE²injection into the carrageenan-preconditioned paw also evoked long-lasting allodynia in the contralateral hind paw (“mirror-image” pain), although the mean values of 50% thresholds were less marked (fig. 2). The 50% threshold values between early and late phases of the PGE²-evoked allodynia were not different on ipsilateral and contralateral sides, respectively. There was no allodynia in the contralateral hind paw of saline-preconditioned rats.

In saline-preconditioned rats, the PGE²-induced acute allodynia was significantly attenuated by GVIA at the lowest dose used in this study (0.03 μg) and almost completely blocked at 0.1 μg (fig. 2A). In carrageenan-preconditioned rats, however, the early and late phases of PGE²-induced allodynia were not significantly affected by 0.03 μg GVIA, and greater than 0.1 μg GVIA was required to achieve significant inhibition of both allodynia phases (fig. 2A and B).

The early and late phases of PGE²-induced allodynia observed in the contralateral hind paws were almost completely blocked by lower doses of GVIA (fig. 2C and D). At the GVIA doses used in this study, rats behaved apparently normal, and we did not observe any aversive responses reported for this toxin. 22 

To test whether the P/Q-type channel was also involved in this long-lasting allodynia, we investigated the effects of intrathecal injection of AgaIVA on this PGE²-induced long-lasting allodynia. As shown in figure 3A and B, AgaIVA did not attenuate PGE²-induced prolonged ipsilateral allodynia at both time points. AgaIVA also had no effects on the contralateral allodynia at 0.01 and 0.03 μg (fig. 3C and D). At 0.1 μg, however, AgaIVA significantly enhanced the early, but not the late, phase of the contralateral allodynia (fig. 3C and D). PGE²-induced acute allodynia in saline-preconditioned rats was also significantly potentiated at 0.1 μg (fig. 3A). We did not increase the dose of AgaIVA to greater than 0.1 μg, because higher doses induced some anomalous neurologic behaviors (sudden intense irritation and agitation behaviors against environmental stimuli including von Frey stimulation). 22 

We studied possible changes in the level of the N-type channel in the DRG and spinal dorsal horn after carrageenan inflammation. By quantitative Western blotting, we found that the normalized expression level of N-type channel protein in the DRGs from carrageenan-preconditioned rats was significantly higher than that of saline controls (fig. 4). Although we tried hard to quantify the N-type channel protein expression in L4–L6 spinal cord sections from these animals, in many cases, we could not detect the channel protein.

It has been suggested previously that both rats and mice have a similar time course for carrageenan-induced acute inflammation and hyperalgesia. 29We confirmed that both wild-type and Ca^v2.2 +/− mice have the acute inflammatory pain behavior similar to that in rats (unpublished data, K. Yokoyama, M.D., T. Kurihara, Ph.D., and T. Tanabe, Ph.D., Tokyo, Japan [June 4, 2003]). Thus, we investigated the expression level of the N-type channel in the spinal cord of Ca^v2.2 +/− mice, preconditioned by carrageenan 7 days ago, as the β-galactosidase activity monitored by X-Gal staining. X-Gal staining revealed a broad expression of the N-type channel throughout the spinal gray matter, particularly in the dorsal horn (fig. 5A), which is generally consistent with findings of a previously reported immunohistochemical study. 9The expression level of the channel protein ipsilateral to the carrageenan-preconditioned paw was significantly higher than that of the contralateral side (fig. 5B).

To our knowledge, this is the first behavioral and biochemical study addressing the plastic nature of N-type channel expression in a latent hypersensitization state after acute inflammation. Our results are in line with the hypothesis that the up-regulation of the N-type channel in DRGs and the spinal dorsal horn is a cause of the long-lasting hypersensitivity to an inflammatory mediator.

The up-regulation of channel proteins in DRGs or the spinal cord after experimental short-term inflammation has not been reported before. However, the up-regulation of P/Q-type, but not N- or L-type, channel protein has been shown in the trigeminal ganglion after tooth pulp inflammation. 30In the peripheral nerve injury/neuropathic pain models, the up-regulation of N-type channel protein in the spinal cord after chronic constrictive nerve injury has been reported 31; on the other hand, no change of N-type channel expression in the spinal cord after spinal nerve injury 32and injured DRGs at the mRNA level 32–34has been demonstrated. A previous electrophysiologic study showed that densities of high-voltage-activated Ca²⁺currents in DRG neurons were reduced by the peripheral nerve injuries. 8,35,36In the current study, we could not detect N-type channel expression constantly in the rat spinal cord by Western blotting. This might be partly due to a much lower basal expression of the channel protein in the spinal cord than in the DRG. 32The study of N-type channel expression in the Ca^v2.2 +/− mice spinal cord, although the possibility of species differences could not be completely eliminated, further suggested the plastic nature and importance of this channel in a latent hypersensitization state.

Another important finding of this study was that the carrageenan-conditioning stimulus could induce PGE²-evoked prolonged allodynia not only in the conditioned hind paw but also in the contralateral (noninflamed) hind paw. This result is consistent with previous findings of recurrent inflammation 4,5and neuropathic pain models. 37,38A recent study suggested the role of spinal glia and proinflammatory cytokines in “mirror-image (contralateral)” neuropathic pain. 39Although the precise mechanism causing pain on the contralateral side remains to be explored, our results may suggest that the spinal N-type channel is involved in “mirror-image” pain.

It has been reported that the activation of protein kinase Cε in the sensitized peripheral neurons might be involved in this long-lasting hypersensitivity. 16Another recent report suggested a possible functional link between protein kinase Cε activation and N-type channel up-regulation. 40There is evidence indicating that protein kinase Cε is one of the major protein kinase C isozymes expressed in sensory neurons and is important for nociceptor regulation. 41,42Therefore, an increase or decrease of some factors in inflamed hind paw and spinal cord during acute inflammation would eventually enhance the activity of protein kinase Cε, leading to the up-regulation of the N-type channel.

Recently, several N-type channel blockers have been developed as promising analgesics for a variety of pain conditions. 43In particular, ziconotide (a synthetic form of ω-conotoxin MVIIA) has been evaluated in a number of clinical trials (mainly for neuropathic and cancer pain) via  intrathecal administration. Although ziconotide is generally tolerated, it has been reported to produce significant adverse effects such as bradycardia, nausea, ataxia, and so on. 44Recently, CVID, a more selective N-type channel blocker, was shown to have a wider therapeutic window than ω-conotoxin MVIIA in the treatment of a chronic inflammatory pain 45(adjuvant-induced mechanical hyperalgesia). Our results suggest that the efficacy of N-type channel blockers might be reduced in the chronic inflammatory pain conditions preceded by the acute inflammation. In this context, we showed recently that the induction level of VDCC α²δ subunit expression might determine the sensitivity of gabapentin analgesia among the different neuropathic pain models. 34 

In contrast to the N-type channel blocker, the P/Q-type channel blocker had no inhibitory actions on any type of the PGE²-induced allodynia. At the highest dose in this study, AgaIVA rather exerted stimulatory actions on the contralateral allodynia in carrageenan-conditioned rats and the acute allodynia in saline-conditioned rats. The doses used in this study (0.01–0.1 μg: roughly corresponding to 2–20 pmol) appear to be within the commonly used range. Malmberg and Yaksh 22showed that less than 10 pmol of intrathecal AgaIVA had potent inhibitory effects of formalin-induced nociceptive behavior. Sluka 23,24demonstrated that intraspinal preadministration of AgaIVA within the similar range reduced the secondary hyperalgesic and allodynic behavior induced by intradermal capsaicin or acute joint inflammation. Thus, the P/Q-type channel plays a role in spinal pain transmission but does not seem to be involved in the allodynia induced by PGE²administration. Although the reason for a lack of antiallodynic action of AgaIVA is not known at present, the stimulatory effect is partly consistent with the observations suggesting that the P/Q-type channel is preferentially involved in inhibitory neurotransmission in the spinal dorsal horn in the normal state. 6,13 

In summary, PGE²produced prolonged mechanical allodynia in both hind paws (more prominent on the ipsilateral side) after apparent resolution of unilateral carrageenan-induced acute inflammation. On the other hand, PGE²induced only short-lasting mechanical allodynia in saline-preconditioned paws of control rats. An N-type channel blocker, GVIA, but not a P/Q-type channel blocker, AgaIVA, inhibited PGE²-induced allodynia. High doses of GVIA were necessary in carrageenan-preconditioned rats, which might be caused by the up-regulation of N-type channel expression in DRGs and the spinal cord ipsilateral to the carrageenan pretreatment. Thus, clarifying the mechanism of up-regulation of N-type channel protein would be important as a future study.

The authors are grateful to Z. David Luo, M.D., Ph.D. (Assistant Professor, Department of Anesthesiology, University of California, San Diego, CA), for his helpful advice on Western blotting and Hironao Saegusa, Ph.D., and Yoshihiro Matsuda, Ph.D. (Department of Pharmacology and Neurobiology, Graduate School of Medicine, Tokyo Medical and Dental University, Tokyo, Japan), for providing useful suggestions on X-gal staining.