|Home | About | Journals | Submit | Contact Us | Français|
Neuropathic pain resulting from chronic constriction injury (CCI) is critically linked to sensitization of peripheral nociceptors. Voltage gated sodium channels are major contributors to this state and their expression can be upregulated by nerve growth factor (NGF). We have previously demonstrated that neurotrophin-3 (NT-3) acts antagonistically to NGF in modulation of aspects of CCI-induced changes in trkA-associated nociceptor phenotype and thermal hyperalgesia. Thus, we hypothesized that exposure of neurons to increased levels of NT-3 would reduce expression of Nav1.8 and Nav1.9 in DRG neurons subject to CCI. In adult male rats, Nav1.8 and Nav1.9 mRNAs are expressed at high levels in predominantly small to medium size neurons. One week following CCI, there is reduced incidence of neurons expressing detectable Nav1.8 and Nav1.9 mRNA, but without a significant decline in mean level of neuronal expression, and similar findings observed immunohistochemically. There is also increased accumulation/redistribution of channel protein in the nerve most apparent proximal to the first constriction site. Intrathecal infusion of NT-3 significantly attenuates neuronal expression of Nav1.8 and Nav1.9 mRNA contralateral and most notably, ipsilateral to CCI, with a similar impact on relative protein expression at the level of the neuron and constricted nerve. We also observe reduced expression of the common neurotrophin receptor p75 in response to CCI that is not reversed by NT-3 in small to medium sized neurons and may confer an enhanced ability of NT-3 to signal via trkA, as has been previously shown in other cell types. These findings are consistent with an analgesic role for NT-3.
Neuropathic pain and its associated syndromes - including hyperalgesia and spontaneous pain - have become well characterized over the last decade. Complicit in the development of hyperalgesia is the sensitization of nociceptors, with ectopic discharge of the nociceptor being a contributing factor in spontaneous pain. It is well established that both the activation threshold of a neuron and the potential for spontaneous firing is regulated by sodium channels (Hodgkin and Huxley, 1952; Catterall, 1995). This hyperexcitability of sensory neurons following nerve injury has been associated with altered expression and the redistribution of voltage gated sodium channels to the tips of the injured axons and/or the neuromas (Devor et al., 1989; England et al., 1994, 1996; Amir et al., 1999).
The tetrodotoxin-resistant (TTX-R) sodium channels Nav1.8 and Nav1.9 are proposed to play an active role in the generation of neuropathic pain syndromes. They are associated with nociceptive neurons (Fang et al., 2002; Djourhi et al., 2003), and have been detected primarily in small diameter, but also in medium and large diameter DRG neurons (Akopian et al., 1996; Black et al., 1996; Rush et al., 1998; Cummins et al., 1999; Dib-Hajj et al., 1999; Renganathan et al., 2000; Hong and Wiley, 2006).
The development of hyperalgesia and allodynia in nerve ligation models is reduced with Nav1.8 antisense treatment (Porreca et al., 1999; Lai et al., 2002; Joshi et al., 2006), while ectopic firing of neurons is related to Nav1.8 protein expression (Novakovic et al., 1998; Gold et al., 2003). Evidence of a direct role of Nav1.9 in the development of thermal and mechanical hyperalgesia is lacking in rats with neuropathic pain states (Porecca et al., 1999; Priest et al., 2005). Although still debated (Hillsley et al., 2006), Nav1.9 may play a significant role in the increased excitability of nociceptive axons during inflammation (Herzog et al., 2001; Baker et al., 2003; Rush and Waxman, 2004; Priest et al., 2005).
Reported alterations in the expression of both Nav1.8 and Nav1.9 in response to chronic constriction injury (CCI) are conflicting (Novakovic et al., 1998; Dib-Hajj et al., 1999; Decosterd et al., 2002), and the channels respond differently to inflammation (Tanaka et al., 1998; Black et al., 2004; Gould et al., 2004). This may be due to the complexity of the CCI model which produces preferential axotomy of large sized DRG neurons and exposure of the remaining intact axons to an inflammatory environment (Bennett and Xie, 1988; Kajander and Bennett, 1992) with increased levels of nerve growth factor (NGF) that increase injury-induced expression of these channels (Dib-Hajj et al., 1998; Fjell et al., 1999).
Neurotrophin-3 (NT-3) can prevent the development and maintenance of thermal hyperalgesia, and in general, acts in an antagonistic fashion to NGF in the CCI model of neuropathic pain (Wilson-Gerwing et al., 2005; Wilson-Gerwing and Verge, 2006). This study investigates whether NT-3 is capable of modulating levels of expression of the sodium channel isoforms Nav1.8 and Nav1.9 following CCI in a manner consistent with an analgesic role for NT-3.
All animal procedures were conducted in accordance with the National Institutes of Health policy on the use of animals in research and the University of Saskatchewan animal care committee guidelines (protocol 19920164). A total of 32 young adult male Wistar rats (Charles River Laboratories, Wilmington, MA) weighing 250–300 g were used. Animals were anesthetized for surgery with sodium pentobarbital (Somnitol, 65 mg/kg; MTC Pharm, Cambridge, Ontario, Canada). Pre- and post-operative (for 24 h) subcutaneous injections of buprenorphine (Temgesic, 0.1–0.2 mg/kg) were given to alleviate any post-operative discomfort. To examine the effect of NT-3 on the expression of Nav1.8 and Nav1.9, 40 rats were used: 17 underwent 7 d unilateral CCI of the sciatic nerve (Bennett and Xie, 1988), 3 received sham CCI surgeries whereby the sciatic nerve was exposed but not ligated, 17 received 7 d unilateral CCI with intrathecal infusion of NT-3 for the duration of the injury, and 3 received 7 d unilateral CCI with sham pump implantation whereby the dorsal roots were exposed, the dura opened as with the CCI + NT-3 procedure, but no pump was implanted.
NT-3 (generously supplied by Regeneron Pharmaceuticals, Tarrytown, NY) was delivered intrathecally for 7 d via mini-osmotic pumps (model 2001; Alza, Cupertino, CA) inserted at the lumbar sacral junction as per Verge et al. (1989a) at a concentration and rate of 600 ng/μl/hr (Karchewski et al., 2002) in a solution of PBS containing rat serum albumin (1 mg/ml), streptomycin (100 U/ml), and penicillin (100 U/ml). This dose of NT-3 was the minimum dose found to selectively reverse injury-associated gene expression in injured trkC-expressing neurons (Verge et al., 1996; Jongsma Wallin et al., 2001; Karchewski et al., 2002). At the conclusion of the experiments, rats were killed, and tissue was dissected and processed for in situ hybridization and/or immunohistochemistry as described below. Previous studies have demonstrated a lack of influence ipsilateral and contralateral to injury when vehicle is infused intrathecally (Verge et al., 1989a; Verge et al., 1995; Jongsma Wallin et al., 2001; Wilson-Gerwing et al., 2005).
Deeply anesthetized animals were perfused via the aorta with 0.1 M PBS, pH 7.4, followed by 4% paraformaldehyde in 0.1M PBS. The right and left L4 and L5 DRG were rapidly dissected, postfixed for 1 hour in the same fixative, and cryoprotected in 20% sucrose in 0.1M PBS overnight. Paired experimental and control tissues were mounted in the same cryomold (to ensure processing under identical conditions), covered with OCT compound (Tissue Tek; Miles Laboratories, Elkhart, IN, USA) and frozen in cooled isopentane. Transverse sections were cut at 6 μm on a Micron cryostat (Zeiss, Canada), thaw mounted onto Probe-On+ slides (Fisher Scientific, Edmonton, AB, Canada) and stored with desiccant at − 20 °C until hybridization.
Prior to hybridization, slides were air dried, fixed in 4% paraformaldehyde, and washed in 1X PBS. Sections were then treated with proteinase K (20 μg/ml) containing 10 ml 1M Tris-HCl (pH 7.6), 2 ml 0.5 M EDTA, 200 μl proteinase K stock (20 mg/ml) and 188 μl ddH20, rinsed in 1 X PBS, and post-fixed in 4% paraformaldehyde. Slides were then rinsed and dehydrated in ascending alcohols.
Oligonucleotide probes complementary to and selective for Nav1.8 mRNA [complementary to bases 640–687 (Akopian et al., 1996)], Nav1.9 mRNA [complementary to bases 2811–2858 (Dib-Hajj et al., 1998)] and p75 mRNA [complementary to bases 873–920 (Radeke et al., 1987)] were synthesized (University of Calgary DNA services, Alberta, Canada). The probes were checked against the GenBank database (NIH) to ensure no greater than 60% homology was found to sequences other than the cognate transcript. The probes were labeled at the 3′-end with α-[35S]dATP (New England Nuclear, Boston, MA, USA) using terminal deoxynucleotidyl-transferase (Amersham Pharmacia Biotech, Piscataway, NJ, USA) in a buffer containing 10 mM CoCl2, 1 mM dithiothreitol, 300 mM Tris base and 1.4 M-potassium cacodylate (pH 7.2), and purified through Bio-Spin® Disposable Chromatograph Columns (Bio-Rad laboratories, Hercules, CA, USA) containing 200 mg of NENSORB™ PREP Nucleic Acid Purification Resin (NEN®, Boston, MA, USA). Dithiothreitol was added to a final concentration of 10 nM.
Hybridization was carried out according to published procedures (Dagerlind et al., 1992) on a minimum of 5 slides/probe from each of the experimental and control groupings. Briefly, the sections were hybridized at 43 °C for 14–18 hours in a buffer containing 50% formamide (Sigma Aldrich, Oakville, ON, Canada), 4X SSC (1X SSC – 0.15 M NaCl, 0.015 sodium citrate), 1X Denhart’s solution (0.02% bovine serum albumin and 0.02% Ficoll), 1% sarcosyl (N-laurylsarcosine), 0.02 M phosphate buffer (pH 7.0), 10% dextran sulphate, 500 μg/ml heat-denatured sheared salmon sperm DNA, 200 mM dithiothreitol and 107 cpm/ml of probe. After hybridization, the slides were washed for 4 × 15 mins in 1X SSC at 55°C, dehydrated in ascending alcohols, processed for radioautography as per Karchewski et al., 2002 and exposed for 7 to 10 days before developing in D-19 (Kodak, Rochester, NY, USA).
The specificity of hybridization signal for the Nav 1.8 and Nav 1.9 probes was confirmed as described in Wilson-Gerwing et al., 2005.
All slides from the 16 groupings of 7 d experimental and control animals were analyzed qualitatively and relative changes in hybridization signal from one group to another noted for sections mounted on the same slide to avoid bias due to the variance in hybridization signal observed from slide to slide. Representative slides were selected for further quantitative analysis. These slides had a similar number of neurons in all DRG sections. Photomontages were prepared and individual neurons with a visible nucleus were identified. Using a 40X light objective and a 2X optivar with an interactive computer-assisted image-analysis system (Richardson et al., 1989), the cross-sectional area of individual neurons and the percentage of cytoplasmic area covered by silver grains was measured for each neuron with a visible nucleus in the DRG section. The area per grain was kept constant for all neurons and a correction for grain overlap was made to obtain a parameter linearly related to density of silver grains (Richardson et al., 1989). Software for the image analysis system was Northern Eclipse, Version 7.0 (Empix Imaging, Mississauga, ON, Canada) and supplemented with Microsoft Office Excel 2003 (Microsoft Corporation, Redmond, WA) and Prism 4.0 (Graph Pad Software, San Diego, CA). Cells were considered labeled if they had more than five times background levels of silver grains, as determined by averaging grain densities over defined areas of the neuropil devoid of positively labeled cell bodies. This criterion for determining labeled neuronal profiles correlates well with the identification of labeled versus unlabeled neurons as determined manually using a 63X oil immersion objective.
Analysis was performed in each instance on all neurons with a nucleus present in the section being quantified: for Nav1.8 this represents 12 DRG sections or 2236 neuronal profiles (Intact: n = 3 animals, CCI: n = 3 animals, Intact + NT-3: n = 3 animals, CCI+NT-3: n = 3 animals); while for Nav1.9 this represents 12 DRG sections or 2025 neuronal profiles (Intact: n = 3 animals, CCI: n = 3 animals, Intact + NT-3: n = 3 animals, CCI+NT-3: n = 3 animals). The contralateral intact DRG was used as an intact control as previous research from our lab has demonstrated that CCI does not induce bilateral hyperalgesia (Wilson-Gerwing et al., 2005) and we have not discerned any qualitative differences between sham-operated or naive DRG and contralateral intact DRG with respect to Nav1.8 or Nav1.9 expression.
To determine whether alterations in the mean labeling index were significant, a nonparametric ANOVA test was employed (Kruskal-Wallis Test) since it could not be assumed that our data followed a Gaussian distribution. Following the Kruskal-Wallis test, Dunn’ Multiple Comparison test was used to determine significant differences between specific groups of data (p < 0.001). All statistical calculations were performed using Prism 4.0 (GraphPad Software, San Diego, CA).
Transverse 10 μm sections were cut on the cryostat, thaw-mounted onto Probe-ON+ slides (Fisher Scientific), and processed for immunohistochemistry. For Nav1.8: sections were washed in 0.1 M PBS, then permeabilized with 0.3% Triton X-100 in 0.1 M PBS for 45 minutes at room temperature. Sections were blocked overnight at 4 °C in 10% goat serum and 0.3% Triton X-100 in 0.1 M PBS, then incubated overnight at 4 °C with rabbit anti-Nav 1.8 Affinity Purified Polyclonal Antibody (1:200; Chemicon International, Temecula, CA, USA) diluted with 10% goat serum and 0.3% Triton X-100 in 0.1 M PBS. Sections were visualized with Alexa Fluor® 488 F(ab′)2 fragment of goat anti-rabbit IgG (H+L) (1:250; Molecular Probes, Eugene, OR, USA) in 2% goat serum in 0.1 M PBS for 2 hours at room temperature. Slides were washed and coverslipped with 50% glycerol/50% PBS. For Nav1.9: sections were washed in 0.1 M PBS, then permeabilized with 0.3% Triton X-100 in 0.1 M PBS for 45 minutes at room temperature. Sections were blocked overnight at 4 °C in 10% goat serum, 3% BSA, and 0.3% Triton X-100 in 0.1 M PBS, then incubated overnight at 4 °C with rabbit anti-Nav1.9 Affinity Purified Polyclonal Antibody (1:100; Chemicon International, Temecula, CA, USA) diluted with 10% goat serum, 3% BSA, and 0.3% Triton X-100 in 0.1 M PBS. Sections were visualized with Alexa Fluor® 546 goat anti-rabbit IgG (H+L), F(ab′)2 fragment conjugate (1:250; Molecular Probes, Eugene, OR, USA) in 2% goat serum for 2 hours at room temperature. Slides were washed and coverslipped with 50% glycerol/50% PBS. Control sections were processed in the same manner, but without the primary antibody. Results were visualized using a Zeiss Axioscope 50 microscope equipped with incident-light fluorescence optics and a digital camera.
Analysis of sections processed for in situ hybridization to detect neuronal expression of Nav1.8 mRNA revealed that in DRG contralateral to CCI (Intact), detectable hybridization signal was observed in 70.3% of neurons analyzed (Table 1). Consistent with previous reports (Akopian et al., 1996; Black et al., 1996; Cummins et al., 1999; Dib-Hajj et al., 1999), we found that expression is observed primarily in small to medium sized (<35 μm in diameter) neurons. These neurons represent 60.7% of all neurons analyzed, while only 9.6% were larger sized (>35 μm in diameter) neurons (Table 1).
Seven days after CCI, there is a reduction in the percentage of DRG neurons expressing detectable levels of Nav1.8 (from 70.3% to 55.0%; Figure 1; Table 1). However, the mean labeling index for these neurons was not significantly changed following CCI (from 18.53 +/− 0.9414 to 17.50 +/− 1.008; p<0.001; Figure 2), suggesting that the neurons that express Nav1.8 do so at slightly higher levels in response to CCI as was apparent for a subpopulation of small to medium size neurons. The decreased incidence of expression, consistent with the findings of Dib-Hajj et al.(1999), occurs primarily in the small to medium diameter neurons (Table 1). The smaller decrease in incidence of expression in the larger cells (presumably those axotomized by CCI) is in agreement with that observed by Decosterd et al. (2002).
When NT-3 is infused for the duration of the 7 day CCI, we find that the relative level of expression of Nav1.8 mRNA is significantly decreased both ipsilateral (CCI + NT-3) and contralateral (Intact + NT-3) to CCI as compared to the control DRG (CCI, Intact; Figure 1). In NT-3 infused DRG contralateral to CCI (Intact + NT-3) the percentage of neurons expressing detectable levels of Nav1.8 is reduced from 70.3% to 52.4% (Figure 1; Table 1). The mean labeling index was significantly decreased from 18.53 +/− 0.9414 in the intact state to 14.34 +/− 0.7077 in the NT-3 treated group (p<0.001; Figure 2). A greater attenuation of neuronal Nav1.8 expression was observed in the NT-3 treated DRG ipsilateral to CCI (CCI + NT-3). Again, the incidence of a neuron expressing detectable levels of Nav1.8 mRNA was reduced from 55.0% to 33.4% (Figure 1, Table 1). This was accompanied by a dramatic and significant decrease in the mean labeling index from 17.50 +/− 1.008 to 8.166 +/− 1.598 (p<0.001; Figure 2). The decrease in the relative level of Nav1.8 expression is most apparent in those small to medium sized sensory neurons in both ipsilateral and contralateral DRG (Figure 1; Table 1). There is also a decreased incidence of expression observed in some larger sized neurons albeit to a lesser extent (Figure 1; Table 1).
Analysis of sections processed for in situ hybridization to detect neuronal expression of Nav1.9 mRNA revealed that in DRG contralateral to CCI (Intact), detectable hybridization signal occurred over 72.9% of neurons. As observed for Nav1.8 mRNA, this was localized primarily over small to medium sized neurons representing 59.0% of all neurons analyzed, while only 13.9% of the neurons analyzed were large size neurons that express detectable levels of this message (Figure 3; Table 2). These findings are consistent with those previously described (Black et al., 1996; Cummins et al., 1999; Dib-Hajj et al., 1999).
Seven days after CCI, there is a reduction in the percentage of neurons expressing detectable levels of Nav1.9 from 60.9% to 32.3% (Figure 3), with the reduced incidence of expression occurring in both the small to medium and large diameter sensory neurons (Table 2). These findings are consistent with the decreased incidence of expression observed in the small sized DRG neurons observed by Dib-Hajj et al. (1999), and the decrease in the large sized neurons (presumably those axotomized by CCI) reported by Decosterd et al. (2002). The mean labeling index for these neurons, however, was not significantly decreased following CCI (33.29 +/− 7.145 in the intact to 27.23 +/− 6.756 following CCI; p<0.001; Figure 4) and is probably reflective of the increased expression observed in a subpopulation of small to medium size neurons.
Consistent with our Nav1.8 mRNA results, NT-3 infusion for the duration of the seven day CCI effected a marked reduction in the incidence and relative levels of Nav1.9 mRNA expression observed in neurons both contralateral (Intact + NT-3) and ipsilateral (CCI + NT-3) to CCI and (Figure 3). In NT-3 infused DRG contralateral to CCI (Intact + NT-3), the mean labeling index was significantly decreased from 33.29 +/− 7.145 in the intact state to 17.52 +/− 3.621 in the NT-3 treated group (p<0.001; Figure 4). There was also a decrease in as the percentage of neurons expressing detectable levels of Nav1.9 from 72.9% to 49.4% (Figure 3; Table 2). A more dramatic attenuation of Nav1.9 expression was observed in the NT-3 treated DRG ipsilateral to CCI (CCI + NT-3). The mean labeling index was significantly decreased with NT-3 treatment from 27.23 +/− 6.756 to 8.080 +/− 0.9123 (p<0.001; Figure 4), while the percentage of neurons expressing detectable levels of Nav1.9 mRNA was also reduced from 60.9% to 32.3% (Table 2). In the NT-3 treated DRG both contralateral (Intact + NT-3) and ipsilateral (CCI + NT-3) to CCI, the decreased levels of expression are most prominent in the small to medium sized neuronal populations.
To examine whether the ability of NT-3 to attenuate Nav1.8 and Nav1.9 message was reflected with a similar decrease in protein expression of these two ion channels, sections from the same ganglia were processed immunohistochemically to detect either Nav1.8 or Nav1.9 protein. Qualitative analysis of the immunofluorescence revealed that the patterns of protein expression were similar to those observed for mRNA. In the intact state Nav1.8 protein was observed most prominently in the smaller DRG neurons, with some medium to large sized neurons also expressing low to moderate levels (Figure 5). Similarly, Nav1.9 immunofluorescence signal was observed predominantly in the small to medium sized DRG neurons. Relative levels of Nav1.8 protein remain largely unchanged 7 days following CCI (Figure 5). Levels of Nav1.9 protein appear slightly elevated following CCI (Figure 6) with some small neurons expressing Nav1.9 and at higher levels than observed in the intact state.
Infusion of NT-3 effected a decrease in the levels of Nav1.8 protein detected in neurons in DRG both contralateral (Intact + NT-3) and ipsilateral to CCI (CCI + NT-3) (Figure 5). On the other hand, infusion of NT-3 had only a modest effect on relative levels of Nav1.9 protein expression in neurons contralateral to CCI (Intact + NT-3) as opposed to the decrease in both the relative levels and the incidence of neurons expressing detectable Nav1.9 protein in DRG ipsilateral to CCI (CCI + NT-3) (Figure 6).
Following nerve injury, voltage gated sodium channels are highly localized/redistributed to the tips of the injured axons and/or neuromas despite a lower level of expression in the cell body of these neurons (Devor et al., 1989; England et al., 1994, 1996; Amir et al., 1999). Thus, we asked whether the dramatic reduction in sodium channel expression effected by NT-3 in neurons ipsilateral to CCI is also reflected in a reduced localization of these channels to the neuroma at the constriction sites formed by the ligatures, unlike that observed for CCI alone. The common sciatic nerves at the level of the CCI with or without NT-3 infusion were processed to detect either Nav1.8 or Nav1.9 protein. Consistent with previous reports (Devor et al., 1989; England et al., 1994, 1996; Amir et al., 1999), both Nav1.8 and Nav1.9 protein levels were more highly localized to the constriction sites 7 days following CCI (CCI; Figure 7 bottom). Infusion of NT-3 attenuated this redistribution as evidenced by a reduction in relative levels of Nav1.8 and Nav1.9 protein at the constriction sites of the CCI + NT-3 treated nerves (Figure 7 bottom). This bolsters our proposal that although CCI can effect a redistribution of sodium channels to the injured axon tips, the extremely low transcript levels following NT-3 treatment result in protein levels that are not sufficient to sustain this response. Alternatively, exogenous NT-3 may also dampen the signals that effect this translocation.
While channel localization was elevated at the ligature sites following CCI, relative levels of expression in the uninjured contralateral sciatic nerve (Intact) for both of these sodium channels is only low (Nav1.8 ) to moderate (Nav1.9; Figure 7 top). In accordance with our findings for the DRG, infusion of NT-3 effected a subtle decrease on the relative levels of Nav1.8 protein expression observed in the intact nerves (Intact + NT-3) (Figure 7 top). However, infusion of NT-3 did not result in any perceptible changes in the levels of Nav1.9 protein expression in the uninjured nerve (Intact + NT-3) (Figure 7 top).
Our previous studies have demonstrated that DRG neurons co-express two isoforms of trkA, one that selectively binds and is activated by NGF and one with a 6 amino acid insert in the extracellular domain that binds and is activated by both NGF and NT-3 (Karchewski et al., 1999; Barker et al., 1993; Clary and Reichardt 1994). In other cell models, the common neurotrophin receptor, p75, inhibits NT-3 signaling through trkA (Mischel et al., 2001). As the ability of NT-3 to attenuate expression of Nav1.8 and Nav1.9 is always most pronounced ipsilateral to CCI we examined whether expression of p75 might be altered in response to CCI and NT-3 in a manner consistent with that which might confer an enhanced ability of NT-3 to signal via trkA.
Qualitative examination of p75 mRNA expression in DRG subjected to one week unilateral CCI with or without NT-3 infusion begun at the time of injury, revealed that in intact DRG neurons p75 is heterogenously expressed at moderate to high levels across all size ranges of neurons (Figure 8), consistent with that previously reported (Verge et al., 1992, Karchewski et al., 1999). CCI results in a dramatic reduction in neuronal p75 expression (Figure 8). NT-3 infusion did not discernibly alter p75 expression in small and medium size neurons but did result in increased neuronal expression of p75 mRNA within a subpopulation of primarily large size neurons ipsilateral to CCI (Figure 8). This is consistent with our previously reported data showing that NT-3 can reverse the downregulation in p75 expression that occurs in spinal nerve injured neurons, but only in those large size neurons expressing trkC (Verge et al., 1996).
We have previously reported that administration of exogenous NT-3 at the time of CCI prevents the development of thermal hyperalgesia and attenuates expression of one of the molecules and associated signalling pathways complicit in this behavior (Wilson-Gerwing et al., 2005). In concert with these findings, we now show that exogenous NT-3 at the time of injury significantly decreases levels of expression of two TTX-R sodium channels, Nav1.8 and Nav1.9, and thus, presumably mitigates the hyperexcitable state associated with this pathology.
In support of a role for Nav1.8 in the development of thermal hyperalgesia, it has been shown that if Nav1.8 is blocked by antisense oligonucleotides, the development of thermal hyperalgesia is reduced in both the spinal nerve ligation and CCI models (Porreca et al., 1999; Lai et al., 2002; Joshi et al., 2006). However, if antisense oligonucleotides to Nav1.9 or genetic ablation of Nav1.9 are employed, there appears to be no effect on thermal or mechanical hypersensitivity in the neuropathic rat (Porreca et al., 1999; Priest et al., 2005; Amaya et al., 2006). Interestingly, it appears that Nav1.9 does play a role in hypersensitivity produced by the application of inflammatory mediators to the peripheral terminals of the nociceptors (Amaya et al., 2006). It has been proposed that Nav1.9 plays a crucial role in setting the resting membrane potential of a neuron and that an increased density of this channel, such as is seen with the accumulation of voltage gated sodium channels at the tips of the injured neurons (Devor et al., 1989; England et al., 1994, 1996), may hyperpolarize the neuron (Herzog et al., 2001). It thus appears that the decreased expression of Nav1.8 by exogenous NT-3 likely plays an important role in preventing the development of thermal hyperalgesia, while the decreased expression of Nav1.9 may prevent hyperexcitablity and/or repetitive firing of the neuron by increasing the resting membrane potential of these neurons. The ability of NT-3 to modulate sodium channel expression is not limited to Nav1.8 and Nav1.9. In preliminary studies, we have obtained similar findings for Nav1.3 (TDW-G and VMKV, unpublished findings), a TTX-sensitive sodium channel whose expression is elevated in response to nerve injury (Lindia et al., 2005). The effect of NT-3 is also not just a global nonselective affect that inhibits translation, as we have recently found somatostatin expression not to be altered by NT-3 treatment (TDW-G and VMKV, unpublished findings), and also observed expression of other markers to be upregulated in injured neurons expressing trkC, including expression of trkC (Verge et al., 1996; Jongsma Wallin et al., 2001; Karchewski et al., 2002)
Inflammatory mediators (including NGF) are capable of increasing the expression voltage gated sodium channels (Dib-Hajj et al., 1998; Fjell et al., 1999; Gould et al., 2000; Fang et al., 2005; Amaya et al., 2006). It has also been postulated that the NGF regulation of Nav1.8 is limited by availability of its cognate receptor trkA (Fang et al., 2005). It therefore seems plausible that the ability of NT-3 to downregulate expression of trkA (Wilson-Gerwing and Verge, 2006) and to act in an antagonistic fashion to the pro-inflammatory effects of NGF (Wilson-Gerwing et al., 2005) may also underlie its ability to attenuate expression of the TTX-R sodium channel Nav1.8 and possibly Nav1.9. NT-3 can mitigate many aspects of trkA-associated nociceptive phenotype in vivo, including effecting a notable reduction in expression of trkA, NGF high-affinity binding sites, BDNF, substance P, and PACAP in the nociceptive subpopulation (Jongsma Wallin et al., 2001; Karchewski et al., 2002; Gratto and Verge, 2003). Up to half of trkA neurons express a low level of the cognate NT-3 receptor, trkC, while all express the common neurotrophin receptor p75 (Karchewski et al., 1999). It is unlikely that NT-3 signaling by trkC greatly influences the anti-nociceptive responses observed in intact neurons, because the downregulation of BDNF, substance P, and PACAP occurs predominantly in the subpopulation of trkA neurons that lack trkC and express the neuropeptide substance P (Jongsma Wallin et al., 2001; Karchewski et al., 2002; Gratto and Verge, 2003). The ability of NT-3 to differentially regulate phenotype in trkC versus trkA neurons may be linked to its differential influence on activated MAPK signaling in these two populations. In preliminary work, we find that NT-3 treated neurons expressing only trkA have a significantly reduced level of activated/nuclear localized ERK1/2 signaling relative to control, while those expressing only trkC have elevated levels of activated/nuclear localized ERK1/2, with no net effect observed for neurons expressing both trkA and trkC (Wilson-Gerwing and Verge, 2005).
Whether NT-3 mediates it affect on the nociceptive subpopulation by signaling directly through trkA is currently being investigated. While NT3 is best known for its ability to signal through its cognate receptor trkC, it is also able to interact with trkA (Lamballe et al., 1991; Ip et al., 1993). Signaling via the trkA family of receptors may also be modulated by select expression of isoforms of full length trkA receptors. The splice variants of the trkA receptor encode two receptors with or without a 6 amino acid (a.a.) insert in the extracellular binding domain. This insert region does not appear to affect the receptor’s binding specificity or functional response to NGF, but can confer enhanced responsiveness of the receptor to NT-3 (Barker et al., 1993; Clary et al., 1994). We have shown virtually all trkA sensory neurons to co-express both isoforms, thus providing an anatomical substrate for this interaction (Karchewski et al., 1999). In addition, because in other cell types p75 has been shown to inhibit NT-3 signaling via trkA (Mischel et al., 2001), our observed downregulation of neuronal p75 expression following CCI may allow for enhanced signaling via trkA by NT-3. Even though NT-3 infusion resulted in some neurons expressing higher levels of p75, this effect was limited to large size neurons that are likely to be trkC-positive (Verge et al., 1996), while expression in small and medium size neurons consistent with the nociceptive subpopulation remained depressed. An enhanced ability of NT-3-to signal via trkA when p75 levels are decreased may account for the more robust response observed ipsilateral to CCI in this and past studies (Wilson-Gerwing et al., 2005; Wilson-Gerwing and Verge 2006).
Finally, there is some evidence that NT-3 may be able to act through the glial derived neurotrophic factor (GDNF) receptor c-Ret (Kobayashi and Masuoka, 2000) and thus may exert its effects on Nav1.8 and Nav1.9 in this manner. Similar to NGF, GDNF has been shown to upregulate Nav1.8 and Nav1.9 following sciatic nerve transection (Cummins et al., 2000) and to reduce ectopic neuronal discharges (spontaneous pain) after nerve injury (Boucher et al., 2000) – a property of neuropathic pain attributed to Nav1.9.
The use of sodium channel blockers to treat both neuropathic and inflammatory pain in a clinical setting is well known to result in analgesia (Galer, 1995; McQuay et al., 1995; Clayton et al., 1997; Evans et al., 1997; Trezise and Xie, 1997). These include such treatments as topical creams (for example, lidocaine), anticonvulsants, and tricyclic antidepressants (reviewed in Rogers et al., 2006).
It therefore becomes important to ask: Does the decreased expression of Nav1.8 and Nav1.9 also result in a diminished channel activity? It has been demonstrated that a significant downregulation of these two channels also resulted in the significant reduction of the TTX-R sodium current (Dib-Hajj et al., 1999). In addition, phosphorylation of Nav1.8 by p38 MAPK, a signalling pathway implicated in the regulation of TRPV1 (Wilson Gerwing et al., 2005), increases current density within DRG neurons (Hudmon et al., 2008). We have previously reported that exogenous NT-3 greatly dampens p38 MAPK signalling in DRG neurons subjected to CCI (Wilson Gerwing et al., 2005). This, in addition to its ability to attenuate sodium channel expression could have a significant impact on overall channel activity and contributions to the neuropathic pain state.
In conclusion, investigation into the ability of exogenous NT-3 to modulate expression of two TTX-R sodium channels, Nav1.8 and Nav1.9 that are implicated in the generation of neuropathic pain, reveals that NT-3 can effectively antagonize yet another pro-inflammatory aspect of the neuropathic pain state and in doing so presumably alter the electrophysiological properties of these neurons.
This work was supported by a grant from the Canadian Institutes of Health Research to VMKV (Grant #MOP74747). TDW-G is supported by a scholarship from the University of Saskatchewan, Saskatchewan Health Research Foundation – RPP, and the Canadian Institutes of Health Research. We thank Regeneron Pharmaceuticals (Tarrytown, NY) for the generous supply of NT-3 used in this study.