Group IV nociceptors develop axonal chemical sensitivity during neuritis and following treatment of the sciatic nerve with vinblastine

METHODS

The University of New England Animal Care and Use Committee approved these experiments in compliance with National Institutes of Health guidelines for the care and use of laboratory animals. A total of 40 adult female Sprague-Dawley rats (175–250 g; Charles River Laboratories, Wilmington, MA) were used in this study. Female rats were selected for consistency with previous reports, because of the subjective perception that there is less connective tissue within the dorsal roots, and because there has been no difference found between sexes in our studies with similar parameters (Bove et al. 2003; Dilley and Bove 2008a; Dilley et al. 2013).

Dorsal Root Recordings

Surgery.

Neuritis induction and vinblastine treatment were the same as previously published (Bove et al. 2003; Dilley and Bove 2008a). Rats were anesthetized with isoflurane in pure oxygen. The fur over the left posterior thigh was clipped, the skin cleaned with surgical scrub, and the area draped with sterile plastic. A small incision was made posterior to the femur, and the muscles were separated to expose the sciatic nerve, which was then cleared of connective tissue for ~10 mm, using only epineurial fascia to mobilize the nerve. In some animals, a 4 × 4 × 10-mm piece of Gelfoam saturated with ~150 µl of CFA (emulsified 1:1 with sterile buffer; Sigma) was gently wrapped to surround the nerve. In other animals, a cone was formed with Parafilm and placed around the nerve before cotton wool soaked in 0.1 mM vinblastine in sterile buffer was placed around the nerve (Dilley and Bove 2008a; Fitzgerald et al. 1984). After 15 min, the vinblastine was removed and copiously rinsed with sterile buffer. The incision was closed in layers with 4-0 nylon sutures, and the rats were moved to a clean cage for recovery. Because we have not seen differences between surgical sham and nonoperated groups using these methods (Bove et al. 2003; Dilley and Bove 2008b), and because we have two distinct procedures, we used unoperated rats as controls. Because vinblastine is anti-inflammatory (Norris et al. 1977), it can be considered a control group for the inflammation caused by the application of CFA.

Electrophysiology and experimental protocol.

Rats were anesthetized to areflexia with isoflurane and prepared for electrophysiology 3–4 days following vinblastine treatment or 6–7 days following the induction of neuritis. These end points were chosen because they are times when the agents show their greatest effects and because they are consistent with our previous reports (Dilley and Bove 2008a, 2008b; Dilley et al. 2013). Body temperature was maintained at 37°C using a feedback-controlled thermal pad with a rectal probe (FCH) and a circulating warm water flexible pad (Gaymar) folded over the upper body. Electrophysiological methods and neuronal isolation were performed as previously described (Fig. 1; also see Bove et al. 2003). A laminectomy was performed from L2 to L5 to expose the spinal cord. The skin was glued to a metal ring and the pool filled with 37°C mineral oil. The dura mater was incised, and the L5 and L4 dorsal roots were cut at the dorsal root entry zone. Dorsal roots were draped over a bipolar stimulating electrode and placed on a small glass plate. Fine filaments (~6–12 µm) were separated from the dorsal roots using honed forceps and were draped over a bipolar recording electrode made with fine gold wires. The distance between the stimulating and recording electrodes ranged from 11 to 14 mm. Electric stimuli were delivered to the dorsal root at intensity suprathreshold for C-fiber activation (0.05–0.1 ms, 20–30 V) using an isolated constant-voltage stimulator (Grass) to identify neurons and determine conduction latencies. Action potentials were amplified, bandpass filtered (50–5,000 Hz), and monitored with an oscilloscope. Neuronal activity was digitized and recorded with Spike2 software (Cambridge Electronic Designs, Cambridge, UK) for off-line analysis. Conduction velocities were determined by dividing the conduction distance by the latency of individual units. When clear single-neuron waveforms were obtained, receptive fields (RFs) in deep structures (group IV) were searched for distal to the knee by using noxious stimuli applied with the fingers and/or forceps. Cutaneous RFs were distinguished from deep RFs by using forceps to carefully maneuver and pinch the skin only, as previously described (Bove and Light 1995). If the responsiveness moved with the skin, it was concluded that the RF was within the skin. Only group IV neurons with high-threshold RFs were recorded further (Fig. 2, A and B). Group III neurons are far less common in recordings from the dorsal roots (Bove et al. 2003) and were not encountered. Neurons with cutaneous RFs were not characterized, because they have been shown to not develop axonal mechanical sensitivity.

Fig. 1.

Experimental schematic and nociceptor isolation. A: recording methods schematic. The sciatic nerves were treated with complete Freund’s adjuvant (CFA) or vinblastine (shaded portion of nerve), and recordings were later performed through the dorsal root from singularly identifiable group IV nociceptors. The key characterization in this experiment was to inject the nerve with a combination of inflammatory mediators and evaluate ectopic sensitivities. Injections and mechanical stimuli were delivered within the treated zone. DRG, dorsal root ganglia; Rec, recording; Stim, electrical stimuli. B: to identify that a receptive field (RF) is from a specific neuron, the dorsal root was electrically stimulated (A) while the RF was mechanically stimulated. When the electrical stimulus occurs during the refractory period of the axon, it fails to evoke an action potential (arrow). C: recording of noxious mechanical stimulation of a group IV neuron’s RF, applied using the fingers (arrowheads indicate application and removal of the stimulus). D: mechanical stimulation of the sciatic nerve, using a soft silicone probe, during neuritis, at the treated site, evoking a response from the axon of the neuron being recorded.

Fig. 2.

Responses of group IV nociceptors to chemical stimuli. Representative responses of group IV nociceptors to chemical stimulation (arrows) of their natural RFs (A, C, and F) and of their axons passing through the sciatic nerve (B, E, and G). A shows a robust response to inflammatory mediators injected into the RF, but not when applied to the axon of the same neuron (C). C and E (CFA) and F and G (vinblastine) show responses to injections to both RFs and their axons. D shows a response to vehicle injection into the nerve of a mechanically sensitive axon. Note immediate response rather than the delay in E. There were no responses in C–F due to the injections after the time frames of the graphs.

In early experiments, we determined that the latency in response to topical application of “inflammatory soup” [IS; bradykinin, serotonin, histamine, and prostaglandin, all at a concentration of 10 ^{M in observation medium (OM;} Light et al. 2008)] was 30–60 min. We attributed this to the diffusion barrier and positive intrafascicular pressure that the perineurium presents (Peltonen et al. 2013). We attempted to disrupt this barrier using sodium deoxycholate (Todd et al. 2000a, 2000b), but this did not reduce the response latency. Because these recordings are often time-limited, we chose to inject substances subperineurally using a 30-gauge bent needle.

Once peripheral RFs were identified, the following protocol was followed: 1) recording of ongoing activity for 5+ min; 2) mechanical stimulation of the RF; 3) exposure and mechanical stimulation of the nerve, accomplished by pressing the nerve between a Silastic probe and a hooked spatula; 4) subperineural injection of 100–150 µl of OM into the nerve with 5+ min of recording; 5) subperineural injection of 100–150 µl of IS into the nerve with 5+ minutes of recording; and 6) injection of IS into the distal mechanically sensitive RF with 5+ minutes of recording.

The mechanical stimulation of the nerves was previously described in detail (Bove et al. 2003). In brief, forces of 4 N can be applied with this probe; in these experiments, as in previous experiments, the forces used on the nerve were limited to 2 N or less (measured earlier on an electronic scale) on a footprint of 6–8 mm. As in our previous studies, using the probes in this manner does not interrupt conduction of action potentials from axons in passage, as confirmed by mechanically activating the natural RF following nerve stimulation. When the ectopic mechanically sensitive RF responds, it does so similarly to the natural RF. These observations confirm that the responses are not injury discharge.

Because it was unknown whether previous exposure to noxious chemicals would induce sensitivity changes, we limited our data collection to the first group IV neurons that were appropriately identified. This primarily resulted in one neuron per experiment; however, in three experiments, two neurons with distinctively identifiable waveforms were recorded from the same filament.

Immunohistochemistry

Model and surgery.

We hypothesized that slowing the flow of channels and receptors would facilitate functional expression on the axons. We previously reported that partial tight ligation of the sciatic nerve led to kinesin accumulation and that this accumulation was attenuated during neuritis and by vinblastine application (Dilley et al. 2013). These observations supported the prediction that these treatments reduce fast axoplasmic flow. We repeated this approach to evaluate kinetics of the histamine H₃ receptor (H3R). Sixteen rats were operated on as described above, with the use of four rats per group (CFA and control, vinblastine and control). During each surgery, we added a partial (~50%) tight ligation of the sciatic nerve, using 7-0 nylon suture. Nonligated axons were present in all nerves and served as a control, although they were not used as a comparison group.

CFA and vinblastine rats had the agent applied as described above to cover 8–10 mm of the nerve proximal to and including the ligation. Control rats had their sciatic nerve partially ligated with no other procedure performed.

After 3 days (vinblastine) and 6 days (CFA), treated and control rats were terminally anesthetized with pentobarbital sodium (200 mg/kg intraperitoneally) and perfused transcardially with heparinized 0.1 M phosphate-buffered saline. The affected segments of their sciatic nerves were removed and immersion-fixed in 4% paraformaldehyde for 4 h while pinned straight and were then placed in 30% sucrose for cryoprotection. Nerves were sectioned longitudinally at 10 µm with a cryostat and processed for H3R immunoreactivity using standard methods. H3R was chosen because histamine is known to be involved in nociceptor activation and sensitization (Mense and Schmidt 1974; Pongratz et al. 2002; Zhang et al. 2007), and a well-characterized antibody was available. Sections were incubated in rabbit anti-H3R antibodies [30 μg/ml, catalog no. H7038; Sigma Aldrich (Cannon et al. 2007; Chen et al. 2015)] overnight at 4°C, rinsed, and incubated with donkey anti-rabbit IgG DyLight 488 (diluted 1:200; catalog no. 711-485-152; Jackson ImmunoResearch) for 60 min at room temperature. The pancreas was used as a positive control for the anti-H3R (Nakamura et al. 2014); sections incubated without the primary antibody displayed no positive signal.

Imaging.

Sections were photographed and montaged at ×20 magnification with the use of a Nikon upright microscope fitted with a Prior motorized stage and a Ds-Qi1 MC camera, both controlled by NIS Elements software (Nikon). A background correction image was taken using the appropriate fluorescent slide (Chroma USA) and was applied to all images.

Image analysis.

Montaged images were renamed using random numbers (by G. M. Bove) and analyzed in one session by the same person (by R. M. Govea) using NIS Elements. A region of interest (ROI) was defined as the part of the ligated nerve starting 20–30 µm proximal to the ligation and extending another 500 µm proximally (Fig. 3A). ROIs were analyzed from four to six sections per nerve. The positive signal within the ROI was subjectively determined using the thresholding tool within NIS Elements (Fig. 3A), used to select pixels of light intensities that matched those chosen by the experimenter as “positive.” We measured the fraction of the total ROI area that was above threshold for each section and refer to it as the “H3R signal.”

Fig. 3.

Immunohistochemical quantification method and results. A: regions of interest consisted of the area of nerve between 20 and 30 μm (small double arrow) and 500 μm proximal (large double arrow) to the sutures (asterisks). B: same image as A, showing histamine H₃ receptor (H3R) signal 5 days after ligation surgery. C and D: representative samples from inflamed and vinblastine-treated nerves, respectively, showing little H3R signal. Double asterisk indicates nonligated axons exposed to vinblastine, showing a relative lack of signal. Scale bar (for all panels), 100 μm. E: nerves treated with either CFA or vinblastine showed less H3R signal than nontreated nerves (control nerves, ligated but not treated); n = 4 per group.

Dorsal Root Recordings

Electrophysiology and experimental protocol.

Rats were anesthetized to areflexia with isoflurane and prepared for electrophysiology 3–4 days following vinblastine treatment or 6–7 days following the induction of neuritis. These end points were chosen because they are times when the agents show their greatest effects and because they are consistent with our previous reports (Dilley and Bove 2008a, 2008b; Dilley et al. 2013). Body temperature was maintained at 37°C using a feedback-controlled thermal pad with a rectal probe (FCH) and a circulating warm water flexible pad (Gaymar) folded over the upper body. Electrophysiological methods and neuronal isolation were performed as previously described (Fig. 1; also see Bove et al. 2003). A laminectomy was performed from L2 to L5 to expose the spinal cord. The skin was glued to a metal ring and the pool filled with 37°C mineral oil. The dura mater was incised, and the L5 and L4 dorsal roots were cut at the dorsal root entry zone. Dorsal roots were draped over a bipolar stimulating electrode and placed on a small glass plate. Fine filaments (~6–12 µm) were separated from the dorsal roots using honed forceps and were draped over a bipolar recording electrode made with fine gold wires. The distance between the stimulating and recording electrodes ranged from 11 to 14 mm. Electric stimuli were delivered to the dorsal root at intensity suprathreshold for C-fiber activation (0.05–0.1 ms, 20–30 V) using an isolated constant-voltage stimulator (Grass) to identify neurons and determine conduction latencies. Action potentials were amplified, bandpass filtered (50–5,000 Hz), and monitored with an oscilloscope. Neuronal activity was digitized and recorded with Spike2 software (Cambridge Electronic Designs, Cambridge, UK) for off-line analysis. Conduction velocities were determined by dividing the conduction distance by the latency of individual units. When clear single-neuron waveforms were obtained, receptive fields (RFs) in deep structures (group IV) were searched for distal to the knee by using noxious stimuli applied with the fingers and/or forceps. Cutaneous RFs were distinguished from deep RFs by using forceps to carefully maneuver and pinch the skin only, as previously described (Bove and Light 1995). If the responsiveness moved with the skin, it was concluded that the RF was within the skin. Only group IV neurons with high-threshold RFs were recorded further (Fig. 2, A and B). Group III neurons are far less common in recordings from the dorsal roots (Bove et al. 2003) and were not encountered. Neurons with cutaneous RFs were not characterized, because they have been shown to not develop axonal mechanical sensitivity.

Fig. 1.

Experimental schematic and nociceptor isolation. A: recording methods schematic. The sciatic nerves were treated with complete Freund’s adjuvant (CFA) or vinblastine (shaded portion of nerve), and recordings were later performed through the dorsal root from singularly identifiable group IV nociceptors. The key characterization in this experiment was to inject the nerve with a combination of inflammatory mediators and evaluate ectopic sensitivities. Injections and mechanical stimuli were delivered within the treated zone. DRG, dorsal root ganglia; Rec, recording; Stim, electrical stimuli. B: to identify that a receptive field (RF) is from a specific neuron, the dorsal root was electrically stimulated (A) while the RF was mechanically stimulated. When the electrical stimulus occurs during the refractory period of the axon, it fails to evoke an action potential (arrow). C: recording of noxious mechanical stimulation of a group IV neuron’s RF, applied using the fingers (arrowheads indicate application and removal of the stimulus). D: mechanical stimulation of the sciatic nerve, using a soft silicone probe, during neuritis, at the treated site, evoking a response from the axon of the neuron being recorded.

Fig. 2.

Responses of group IV nociceptors to chemical stimuli. Representative responses of group IV nociceptors to chemical stimulation (arrows) of their natural RFs (A, C, and F) and of their axons passing through the sciatic nerve (B, E, and G). A shows a robust response to inflammatory mediators injected into the RF, but not when applied to the axon of the same neuron (C). C and E (CFA) and F and G (vinblastine) show responses to injections to both RFs and their axons. D shows a response to vehicle injection into the nerve of a mechanically sensitive axon. Note immediate response rather than the delay in E. There were no responses in C–F due to the injections after the time frames of the graphs.

In early experiments, we determined that the latency in response to topical application of “inflammatory soup” [IS; bradykinin, serotonin, histamine, and prostaglandin, all at a concentration of 10 ^{M in observation medium (OM;} Light et al. 2008)] was 30–60 min. We attributed this to the diffusion barrier and positive intrafascicular pressure that the perineurium presents (Peltonen et al. 2013). We attempted to disrupt this barrier using sodium deoxycholate (Todd et al. 2000a, 2000b), but this did not reduce the response latency. Because these recordings are often time-limited, we chose to inject substances subperineurally using a 30-gauge bent needle.

Once peripheral RFs were identified, the following protocol was followed: 1) recording of ongoing activity for 5+ min; 2) mechanical stimulation of the RF; 3) exposure and mechanical stimulation of the nerve, accomplished by pressing the nerve between a Silastic probe and a hooked spatula; 4) subperineural injection of 100–150 µl of OM into the nerve with 5+ min of recording; 5) subperineural injection of 100–150 µl of IS into the nerve with 5+ minutes of recording; and 6) injection of IS into the distal mechanically sensitive RF with 5+ minutes of recording.

The mechanical stimulation of the nerves was previously described in detail (Bove et al. 2003). In brief, forces of 4 N can be applied with this probe; in these experiments, as in previous experiments, the forces used on the nerve were limited to 2 N or less (measured earlier on an electronic scale) on a footprint of 6–8 mm. As in our previous studies, using the probes in this manner does not interrupt conduction of action potentials from axons in passage, as confirmed by mechanically activating the natural RF following nerve stimulation. When the ectopic mechanically sensitive RF responds, it does so similarly to the natural RF. These observations confirm that the responses are not injury discharge.

Because it was unknown whether previous exposure to noxious chemicals would induce sensitivity changes, we limited our data collection to the first group IV neurons that were appropriately identified. This primarily resulted in one neuron per experiment; however, in three experiments, two neurons with distinctively identifiable waveforms were recorded from the same filament.

Surgery.

Neuritis induction and vinblastine treatment were the same as previously published (Bove et al. 2003; Dilley and Bove 2008a). Rats were anesthetized with isoflurane in pure oxygen. The fur over the left posterior thigh was clipped, the skin cleaned with surgical scrub, and the area draped with sterile plastic. A small incision was made posterior to the femur, and the muscles were separated to expose the sciatic nerve, which was then cleared of connective tissue for ~10 mm, using only epineurial fascia to mobilize the nerve. In some animals, a 4 × 4 × 10-mm piece of Gelfoam saturated with ~150 µl of CFA (emulsified 1:1 with sterile buffer; Sigma) was gently wrapped to surround the nerve. In other animals, a cone was formed with Parafilm and placed around the nerve before cotton wool soaked in 0.1 mM vinblastine in sterile buffer was placed around the nerve (Dilley and Bove 2008a; Fitzgerald et al. 1984). After 15 min, the vinblastine was removed and copiously rinsed with sterile buffer. The incision was closed in layers with 4-0 nylon sutures, and the rats were moved to a clean cage for recovery. Because we have not seen differences between surgical sham and nonoperated groups using these methods (Bove et al. 2003; Dilley and Bove 2008b), and because we have two distinct procedures, we used unoperated rats as controls. Because vinblastine is anti-inflammatory (Norris et al. 1977), it can be considered a control group for the inflammation caused by the application of CFA.

Electrophysiology and experimental protocol.

Rats were anesthetized to areflexia with isoflurane and prepared for electrophysiology 3–4 days following vinblastine treatment or 6–7 days following the induction of neuritis. These end points were chosen because they are times when the agents show their greatest effects and because they are consistent with our previous reports (Dilley and Bove 2008a, 2008b; Dilley et al. 2013). Body temperature was maintained at 37°C using a feedback-controlled thermal pad with a rectal probe (FCH) and a circulating warm water flexible pad (Gaymar) folded over the upper body. Electrophysiological methods and neuronal isolation were performed as previously described (Fig. 1; also see Bove et al. 2003). A laminectomy was performed from L2 to L5 to expose the spinal cord. The skin was glued to a metal ring and the pool filled with 37°C mineral oil. The dura mater was incised, and the L5 and L4 dorsal roots were cut at the dorsal root entry zone. Dorsal roots were draped over a bipolar stimulating electrode and placed on a small glass plate. Fine filaments (~6–12 µm) were separated from the dorsal roots using honed forceps and were draped over a bipolar recording electrode made with fine gold wires. The distance between the stimulating and recording electrodes ranged from 11 to 14 mm. Electric stimuli were delivered to the dorsal root at intensity suprathreshold for C-fiber activation (0.05–0.1 ms, 20–30 V) using an isolated constant-voltage stimulator (Grass) to identify neurons and determine conduction latencies. Action potentials were amplified, bandpass filtered (50–5,000 Hz), and monitored with an oscilloscope. Neuronal activity was digitized and recorded with Spike2 software (Cambridge Electronic Designs, Cambridge, UK) for off-line analysis. Conduction velocities were determined by dividing the conduction distance by the latency of individual units. When clear single-neuron waveforms were obtained, receptive fields (RFs) in deep structures (group IV) were searched for distal to the knee by using noxious stimuli applied with the fingers and/or forceps. Cutaneous RFs were distinguished from deep RFs by using forceps to carefully maneuver and pinch the skin only, as previously described (Bove and Light 1995). If the responsiveness moved with the skin, it was concluded that the RF was within the skin. Only group IV neurons with high-threshold RFs were recorded further (Fig. 2, A and B). Group III neurons are far less common in recordings from the dorsal roots (Bove et al. 2003) and were not encountered. Neurons with cutaneous RFs were not characterized, because they have been shown to not develop axonal mechanical sensitivity.

Fig. 1.

Experimental schematic and nociceptor isolation. A: recording methods schematic. The sciatic nerves were treated with complete Freund’s adjuvant (CFA) or vinblastine (shaded portion of nerve), and recordings were later performed through the dorsal root from singularly identifiable group IV nociceptors. The key characterization in this experiment was to inject the nerve with a combination of inflammatory mediators and evaluate ectopic sensitivities. Injections and mechanical stimuli were delivered within the treated zone. DRG, dorsal root ganglia; Rec, recording; Stim, electrical stimuli. B: to identify that a receptive field (RF) is from a specific neuron, the dorsal root was electrically stimulated (A) while the RF was mechanically stimulated. When the electrical stimulus occurs during the refractory period of the axon, it fails to evoke an action potential (arrow). C: recording of noxious mechanical stimulation of a group IV neuron’s RF, applied using the fingers (arrowheads indicate application and removal of the stimulus). D: mechanical stimulation of the sciatic nerve, using a soft silicone probe, during neuritis, at the treated site, evoking a response from the axon of the neuron being recorded.

Fig. 2.

Responses of group IV nociceptors to chemical stimuli. Representative responses of group IV nociceptors to chemical stimulation (arrows) of their natural RFs (A, C, and F) and of their axons passing through the sciatic nerve (B, E, and G). A shows a robust response to inflammatory mediators injected into the RF, but not when applied to the axon of the same neuron (C). C and E (CFA) and F and G (vinblastine) show responses to injections to both RFs and their axons. D shows a response to vehicle injection into the nerve of a mechanically sensitive axon. Note immediate response rather than the delay in E. There were no responses in C–F due to the injections after the time frames of the graphs.

In early experiments, we determined that the latency in response to topical application of “inflammatory soup” [IS; bradykinin, serotonin, histamine, and prostaglandin, all at a concentration of 10 ^{M in observation medium (OM;} Light et al. 2008)] was 30–60 min. We attributed this to the diffusion barrier and positive intrafascicular pressure that the perineurium presents (Peltonen et al. 2013). We attempted to disrupt this barrier using sodium deoxycholate (Todd et al. 2000a, 2000b), but this did not reduce the response latency. Because these recordings are often time-limited, we chose to inject substances subperineurally using a 30-gauge bent needle.

Once peripheral RFs were identified, the following protocol was followed: 1) recording of ongoing activity for 5+ min; 2) mechanical stimulation of the RF; 3) exposure and mechanical stimulation of the nerve, accomplished by pressing the nerve between a Silastic probe and a hooked spatula; 4) subperineural injection of 100–150 µl of OM into the nerve with 5+ min of recording; 5) subperineural injection of 100–150 µl of IS into the nerve with 5+ minutes of recording; and 6) injection of IS into the distal mechanically sensitive RF with 5+ minutes of recording.

The mechanical stimulation of the nerves was previously described in detail (Bove et al. 2003). In brief, forces of 4 N can be applied with this probe; in these experiments, as in previous experiments, the forces used on the nerve were limited to 2 N or less (measured earlier on an electronic scale) on a footprint of 6–8 mm. As in our previous studies, using the probes in this manner does not interrupt conduction of action potentials from axons in passage, as confirmed by mechanically activating the natural RF following nerve stimulation. When the ectopic mechanically sensitive RF responds, it does so similarly to the natural RF. These observations confirm that the responses are not injury discharge.

Because it was unknown whether previous exposure to noxious chemicals would induce sensitivity changes, we limited our data collection to the first group IV neurons that were appropriately identified. This primarily resulted in one neuron per experiment; however, in three experiments, two neurons with distinctively identifiable waveforms were recorded from the same filament.

Immunohistochemistry

Image analysis.

Montaged images were renamed using random numbers (by G. M. Bove) and analyzed in one session by the same person (by R. M. Govea) using NIS Elements. A region of interest (ROI) was defined as the part of the ligated nerve starting 20–30 µm proximal to the ligation and extending another 500 µm proximally (Fig. 3A). ROIs were analyzed from four to six sections per nerve. The positive signal within the ROI was subjectively determined using the thresholding tool within NIS Elements (Fig. 3A), used to select pixels of light intensities that matched those chosen by the experimenter as “positive.” We measured the fraction of the total ROI area that was above threshold for each section and refer to it as the “H3R signal.”

Fig. 3.

Immunohistochemical quantification method and results. A: regions of interest consisted of the area of nerve between 20 and 30 μm (small double arrow) and 500 μm proximal (large double arrow) to the sutures (asterisks). B: same image as A, showing histamine H₃ receptor (H3R) signal 5 days after ligation surgery. C and D: representative samples from inflamed and vinblastine-treated nerves, respectively, showing little H3R signal. Double asterisk indicates nonligated axons exposed to vinblastine, showing a relative lack of signal. Scale bar (for all panels), 100 μm. E: nerves treated with either CFA or vinblastine showed less H3R signal than nontreated nerves (control nerves, ligated but not treated); n = 4 per group.

Model and surgery.

We hypothesized that slowing the flow of channels and receptors would facilitate functional expression on the axons. We previously reported that partial tight ligation of the sciatic nerve led to kinesin accumulation and that this accumulation was attenuated during neuritis and by vinblastine application (Dilley et al. 2013). These observations supported the prediction that these treatments reduce fast axoplasmic flow. We repeated this approach to evaluate kinetics of the histamine H₃ receptor (H3R). Sixteen rats were operated on as described above, with the use of four rats per group (CFA and control, vinblastine and control). During each surgery, we added a partial (~50%) tight ligation of the sciatic nerve, using 7-0 nylon suture. Nonligated axons were present in all nerves and served as a control, although they were not used as a comparison group.

CFA and vinblastine rats had the agent applied as described above to cover 8–10 mm of the nerve proximal to and including the ligation. Control rats had their sciatic nerve partially ligated with no other procedure performed.

After 3 days (vinblastine) and 6 days (CFA), treated and control rats were terminally anesthetized with pentobarbital sodium (200 mg/kg intraperitoneally) and perfused transcardially with heparinized 0.1 M phosphate-buffered saline. The affected segments of their sciatic nerves were removed and immersion-fixed in 4% paraformaldehyde for 4 h while pinned straight and were then placed in 30% sucrose for cryoprotection. Nerves were sectioned longitudinally at 10 µm with a cryostat and processed for H3R immunoreactivity using standard methods. H3R was chosen because histamine is known to be involved in nociceptor activation and sensitization (Mense and Schmidt 1974; Pongratz et al. 2002; Zhang et al. 2007), and a well-characterized antibody was available. Sections were incubated in rabbit anti-H3R antibodies [30 μg/ml, catalog no. H7038; Sigma Aldrich (Cannon et al. 2007; Chen et al. 2015)] overnight at 4°C, rinsed, and incubated with donkey anti-rabbit IgG DyLight 488 (diluted 1:200; catalog no. 711-485-152; Jackson ImmunoResearch) for 60 min at room temperature. The pancreas was used as a positive control for the anti-H3R (Nakamura et al. 2014); sections incubated without the primary antibody displayed no positive signal.

Imaging.

Sections were photographed and montaged at ×20 magnification with the use of a Nikon upright microscope fitted with a Prior motorized stage and a Ds-Qi1 MC camera, both controlled by NIS Elements software (Nikon). A background correction image was taken using the appropriate fluorescent slide (Chroma USA) and was applied to all images.

Image analysis.

Montaged images were renamed using random numbers (by G. M. Bove) and analyzed in one session by the same person (by R. M. Govea) using NIS Elements. A region of interest (ROI) was defined as the part of the ligated nerve starting 20–30 µm proximal to the ligation and extending another 500 µm proximally (Fig. 3A). ROIs were analyzed from four to six sections per nerve. The positive signal within the ROI was subjectively determined using the thresholding tool within NIS Elements (Fig. 3A), used to select pixels of light intensities that matched those chosen by the experimenter as “positive.” We measured the fraction of the total ROI area that was above threshold for each section and refer to it as the “H3R signal.”

Fig. 3.

Immunohistochemical quantification method and results. A: regions of interest consisted of the area of nerve between 20 and 30 μm (small double arrow) and 500 μm proximal (large double arrow) to the sutures (asterisks). B: same image as A, showing histamine H₃ receptor (H3R) signal 5 days after ligation surgery. C and D: representative samples from inflamed and vinblastine-treated nerves, respectively, showing little H3R signal. Double asterisk indicates nonligated axons exposed to vinblastine, showing a relative lack of signal. Scale bar (for all panels), 100 μm. E: nerves treated with either CFA or vinblastine showed less H3R signal than nontreated nerves (control nerves, ligated but not treated); n = 4 per group.

Data Analysis

Data were analyzed using GraphPad Prism 7 and are means ± SE unless otherwise noted, with statistical significance set at ≤ 0.05. Conduction velocities were analyzed using a one-way ANOVA. Proportions of axonal mechanical and chemical sensitivities, and ongoing activity, were analyzed using χ ^{tests. The H3R signal in treatment and control groups were compared using unpaired} t-tests.

RESULTS

Dorsal root recordings were obtained from a total of 27 neurons in 24 rats (9 naive rats, 9 CFA-treated rats, and 6 vinblastine-treated rats). There was no difference in conduction velocities by group [F(3,24) = 0.77, P = 0.8]. The mean conduction velocity for all recorded neurons was 0.83 ± 0.20 m/s (mean ± SD; Table 1). All neurons had a noncutaneous mechanically responsive RF in the lower limb or foot (Fig. 2B).

Responses to IS

Sixty-six percent of normal neurons responded to injection of IS directly into their mechanically sensitive RFs (Table 1 and Fig. 2D). Following treatment with CFA or vinblastine, 29% and 50% responded to IS injection into their RFs, respectively. There was no statistical difference between these proportions. These results are consistent with previous reports of cutaneous C-fibers (Kessler et al. 1992) and muscle nociceptors (reviewed by Mense 2009). The discharge patterns were similar among the groups, often consisting of a mechanical artifact or response of the axon due to the needle insertion and pressure of the fluid, followed by a true response. The response latencies and durations were highly variable among groups, ranging from 0.5 to 171 s and from 16 to 143 s, respectively.

Axonal Chemical Sensitivity During Neuritis and Following Transient Application of Vinblastine

Following treatment with CFA or vinblastine, 80% and 38% of axons, respectively, responded to injection of IS (Table 1 and Fig. 2, E and G). No normal axons or treated axons responded to intraneural OM, and no normal axons responded to intraneural IS (Fig. 2B). The proportions of responsive axons differed significantly by group [χ ^(2, n = 26) = 11.51, P = 0.005]. The discharge pattern to IS varied, with latencies ranging from 6 to 88 s. The duration of the responses to IS also varied, lasting from 16 to 148 s.

Axonal Mechanical Sensitivity

Axonal mechanical sensitivity was not present in any normal axons but was present in 33% and 50% of axons treated with CFA or vinblastine, respectively (Table 1 and Fig. 1D). Although the contingency test with these data was not statistically significant [χ ^(2, n = 25) = 5.16, P = 0.08], there was a statistically significant trend toward more axonal mechanical sensitivity in treated axons [χ ^(1, n = 25) = 11.51, P = 0.03], and the proportions are consistent with our previous reports (Bove et al. 2003; Dilley and Bove 2008a, 2008b). The normal RFs of all these neurons were responsive to noxious mechanical stimulation (Fig. 2C). There was no sustained discharge after the mechanical stimuli were removed (Fig. 1, C and D).

Ongoing Activity

Few neurons in this series of experiments had ongoing activity (1 of 9 control, 2 of 10 during neuritis, and 0 of 8 after vinblastine treatment). There were no statistical differences between the proportions, which is consistent with our previous reports (Bove and Dilley 2010; Dilley and Bove 2008a).

Neuritis and Vinblastine Reduce H3R Transport

In sections of sciatic nerve that underwent partial ligation with no treatment, there was a robust accumulation of the H3R proximal to the ligation (Fig. 3B). In sections of sciatic nerve that underwent partial ligation and treatment with either CFA or vinblastine, a reduction in the accumulation of H3R was clearly visible (Fig. 3, C–E). When quantified, the differences were statistically significant for the vinblastine experiment (P < 0.05; Fig. 3E). This reduction of accumulation is indicative of reduced axoplasmic flow induced by the treatments.

DISCUSSION

Using single neuronal recordings, we have demonstrated that whereas normal group IV nociceptor axons are chemically insensitive, CFA-induced neuritis and treatment with vinblastine induce ectopic axonal sensitivity to chemicals normally found in the inflammatory milieu (bradykinin, histamine, serotonin, and prostaglandin). We also have shown that these same conditions impair the axonal transport of H3R.

We chose the combination and concentration of chemicals for our IS because many studies have been published using this combination (Becerra et al. 2017; Kessler et al. 1992; Lang et al. 1990). The initial descriptions of this combination (Steen et al. 1995) were based on concentrations found in various tissues, and the chemicals were applied to skin nociceptors. We do not know the concentrations and proportions of these chemicals that are present in the CFA model. Our observations of little to no ongoing activity in the CFA model but consistent responses to intraneural application of IS suggest that the concentrations of these chemicals inside the nerve are too low to evoke activity. Our combined observations remain consistent with the concept that inflammatory mediators are not required to lead to ectopic sensitivities, but that reduced axonal transport, such as that induced by vinblastine, is sufficient.

Although we and others have shown that CFA-induced neuritis reduces axonal transport (Armstrong et al. 2004; Dilley et al. 2013), there remains limited insight on the mechanism of this phenomenon. Our previous studies (Bove et al. 2003; Dilley and Bove 2008a) and the current report show a parallel between axonal transport and ectopic sensitivities but are methodologically unable to directly correlate the phenomena. Our reports of similar effects of neuritis and vinblastine on kinesin transport suggest a common mechanism of reduced axonal transport, independent of the effects of inflammation, especially because vinblastine is anti-inflammatory (Norris et al. 1977). Although we hypothesize that reduced axonal transport facilitates the insertion of functional receptors into the axonal membrane, such as H3R, we cannot test this prediction using the methods presented.

Clinical Implications

We have shown that neuritis and transient vinblastine application lead to ectopic axonal sensitivity to a mixture of endogenous chemicals. This finding has possible clinical diagnostic relevance for cases presenting with ongoing radiating pain, especially when combined with our previous reports of ectopic axonal mechanical sensitivity induced by the same means (Bove et al. 2003; Dilley and Bove 2008a). Inflammation and mechanical pressure are interrelated (Schmid et al. 2013), reduce axonal transport (Armstrong et al. 2004; Dahlin et al. 1984; Dilley et al. 2013; Gallant 1992), and can result in ectopic mechanical and, as we have shown in the present study, chemical sensitivity. This could manifest as movement-induced and ongoing nociception, respectively. The site of mechanical pressure can often be appreciated with current diagnostic imaging methods, but similar methods are limited in terms of revealing neuritis. Clinicians can use “neurodynamic tests” that have been designed to specifically move and tension the major nerves of the limbs (Butler 2000; Shacklock 2005) and have shown moderate reliability in identifying the involved nerve by reproducing the presenting symptoms (Greening et al. 2005; Schmid et al. 2009). Palpation of an involved nerve is relatively straightforward and can lead to the identification of the site of pathophysiology, again by reproducing the symptoms (Greening et al. 2005; Schmid et al. 2009). In cases of deep radiating pain, clinicians are advised to search for areas along the entire path of the involved nerve for tenderness, which could identify a site of inflammation and lead to an accurate diagnosis.

GRANTS

This work was supported by a discretional fund (to G. M. Bove) by the University of New England College of Osteopathic Medicine, by the National Center for Complementary and Integrative Health (NCCIH) R01AT009350 (to G. M. Bove and M. F. Barbe), and by an Institutional Development Award (IDeA) from the National Institute of General Medical Sciences of the National Institutes of Health under grant number P20GM103643.

DISCLAIMERS

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

G.M.B. conceived and designed research; R.M.G. and G.M.B. performed experiments; R.M.G. and G.M.B. analyzed data; R.M.G., M.F.B., and G.M.B. interpreted results of experiments; R.M.G. and G.M.B. prepared figures; R.M.G. and G.M.B. drafted manuscript; R.M.G., M.F.B., and G.M.B. edited and revised manuscript; R.M.G., M.F.B., and G.M.B. approved final version of manuscript.