Changes in neuropeptide expression in the trigeminal ganglion following inferior alveolar nerve section in the ferret

Neuropeptide expression in the trigeminal ganglion

Pergamon

PII: S0306-4522(00)00508-X

Neuroscience Vol. 102, No. 3, pp. 655±667, 2001 655 q 2001 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306-4522/01 $20.00+0.00

www.elsevier.com/locate/neuroscience

CHANGES IN NEUROPEPTIDE EXPRESSION IN THE TRIGEMINAL GANGLION FOLLOWING INFERIOR ALVEOLAR NERVE SECTION IN THE FERRET C. ELCOCK,* F. M. BOISSONADE and P. P. ROBINSON Department of Oral and Maxillofacial Surgery, School of Clinical Dentistry, Claremont Crescent, Shef®eld S10 2TA, UK

AbstractÐChanges in neuropeptide expression in afferent nerve ®bres may play a role in the persistent sensory abnormalities that can be experienced following trigeminal nerve injuries. We have therefore studied changes in the expression of the neuropeptides substance P, calcitonin gene-related peptide, enkephalin, galanin, neuropeptide Y and vasoactive intestinal polypeptide in the trigeminal ganglion following peripheral nerve injury. In anaesthetised adult female ferrets, the left inferior alveolar nerve was sectioned and recovery allowed for three days, three weeks or 12 weeks prior to perfusion± ®xation. During a second procedure, one week prior to perfusion, the inferior alveolar nerve was exposed and an injection made central to the injury site using a mixture of 4% Fluorogold and 4% isolectin B4 conjugated to horseradish peroxidase to identify cell bodies with axons in the inferior alveolar nerve and cells with unmyelinated axons within this population, respectively. Control animals received tracer injection alone. After harvesting the tissue, sagittal sections were taken from both the right and left ganglia and immunohistochemical staining was used to reveal the presence of peptides and isolectin B4±horseradish peroxidase tracer. Within the Fluorogold-labelled population, cell counts revealed a signi®cant reduction in the proportion of substance P-containing cells at three days (P ˆ 0.0025), three weeks (P ˆ 0.0094) and three months (P ˆ 0.0149) after nerve section, and a signi®cant reduction in the proportion of calcitonin gene-related peptide-containing cells at three days (P ˆ 0.0003) and three weeks (P ˆ 0.007). No signi®cant changes were seen in the expression of the other peptides, or at other time periods. A signi®cant reduction in the number of isolectin B4±horseradish peroxidase-positive cells (with unmyelinated axons) was seen at three days (P ˆ 0.0025), three weeks (P ˆ 0.0074) and three months after the injury (P ˆ 0.0133). These results demonstrate a signi®cant reduction in the expression of some neuropeptides in the early stages after inferior alveolar nerve section. Some of the results differ markedly from those reported previously in other systems, and may be related to the speci®c nerve studied, species variations or differences between spinal and trigeminal nerves. q 2001 IBRO. Published by Elsevier Science Ltd. All rights reserved. Key words: nerve injury, inferior alveolar nerve, trigeminal ganglion, neuropeptides, Fluorogold, isolectin B4.

the alteration in afferent impulse barrage, which can lead to variations in central synaptic excitability. 22 Changes in the chemical substances transported from the periphery occur, such as the failure of transport of nerve growth factor, 68 and changes in central control mechanisms are reported. 23 There is some evidence that neuropeptides have a role in the initiation of injury-induced sensory disorders. In uninjured animals and man a range of neuropeptides have been implicated in the modulation and transmission of nociceptive input. Substance P (SP), calcitonin generelated peptide (CGRP) and galanin (GAL) have all been found in capsaicin-sensitive afferents. 45,81,82 SP and CGRP are thought to be important in nociceptive transmission (for review, see Ref. 102), and GAL is believed to play a role in analgesia. 90 Neuropeptide Y (NPY) and enkephalin (ENK) are also known to play a role in the modulation of nociceptive inputs, 9,42 and vasoactive intestinal polypeptide (VIP) has an excitatory effect on dorsal horn neurons. 76 Injury-induced changes in the expression of both neuropeptides and their receptors have been shown in the cell bodies of spinal sensory axons, 24,36,46,63,79,91,92

Recovery of sensation following a peripheral nerve injury may be complicated by the development of a range of sensory abnormalities such as dysaesthesia, allodynia, hyperaesthesia and hyperalgesia. The aetiology of these disorders remains uncertain, but an extensive series of both central and peripheral changes have been shown to occur within the affected part of the nervous system, and some of these changes are likely to contribute to the sensory disturbances. The peripheral changes include the development of spontaneous activity and mechanical sensitivity of the damaged axons, abnormal interactions between damaged axons, and sensitivity to circulating and locally released catecholamines (for review, see Ref. 21). The central changes may result partly from *Corresponding author. Tel.: 144-114-271-7849/7885/7891; fax: 144-114-271-7843. E-mail address: c.elcock@shef®eld.ac.uk (C. Elcock). Abbreviations: CGRP, calcitonin gene-related peptide; CY3, indocarbocyanine; DRG, dorsal root ganglion; ENK, enkephalin; FG, Fluorogold; FITC, ¯uorescein isothiocyanate; GAL, galanin; HRP, horseradish peroxidase; IAN, inferior alveolar nerve; NPY, neuropeptide Y; PBS, phosphate-buffered saline; SP, substance P; VIP, vasoactive intestinal polypeptide. 655

656

C. Elcock et al.

Fig. 1. A diagram showing the course of the IAN axons, indicating the sites of nerve injury and tracer injection.

suggesting complex adaptive responses. 107 The functional consequences of this plasticity in neuropeptide phenotype are not fully understood, but it appears clear that these changes will affect the transmission of peptiderelated nociceptive information in the CNS (for review, see Ref. 37). Damage to the peripheral branches of the trigeminal nerve is a very common clinical problem, often occurring as a complication of routine surgical procedures. For example, the removal of impacted lower third molar teeth results in damage to the inferior alveolar nerve (IAN) with an incidence of approximately 4%. 41 This results in anaesthesia or paraesthesia of the lower lip and chin, the ipsilateral lower teeth and an area of oral mucous membrane. Approximately 0.5±1% of patients are left with permanent sensory disturbances, 48,74,77 such as hyperaesthesia and pain. 33,34 As removal of impacted third molars is a very common clinical procedure it represents a signi®cant problem, for which at present there is little treatment available. For this reason, our study was directed speci®cally toward the outcome of injuries to the IAN. Recent evidence suggests that individual nerves respond to injury in a speci®c manner. 13,14,87 The IAN is unusual as it is a branch of a cranial nerve and consists mainly of somatic afferents and a few sympathetic efferents, unlike the mixed spinal nerves which have often been studied previously. Hoffmann and Matthews 35 have suggested that pain is less likely to develop following injury to branches of the trigeminal nerve as there is a lower proportion of sympathetic ®bres. Other studies have also shown a lower proportion of non-myelinated ®bres in the IAN 29,38 when compared with cutaneous spinal nerves. In addition, the anatomical location of the nerve within a bony canal means that the cut ends of the nerve do not retract following section 39 and may also suffer pressure effects caused, for example, by oedema. We hypothesised that these features would affect the neurochemical changes which occur following injury. The type of nerve injury may also have an effect on the likelihood of development of a persistent sensory disturbance. Sunderland 85 suggested that more severe injuries

would produce a greater degree of sensory disturbance and Robinson 74 con®rmed this by showing that patients who had sustained a crush injury to a branch of the IAN were less likely to experience persistent sensory disorders than those in whom the nerve had been sectioned. Animal studies have revealed that behavioural changes develop after sectioning the IAN and allowing it to regenerate, 44,58,83 and this would also be consistent with the presence of sensory disturbances. We therefore chose this type of injury for our study, in which we sectioned the IAN, allowed regeneration and examined neurochemical expression in the cell bodies of the trigeminal ganglion. As the trigeminal ganglion contains many cell bodies of axons which travel in other branches of the trigeminal nerve, retrograde labelling was needed to identify cells linked to the injured IAN axons. We also used a second label to identify those cell bodies which had unmyelinated axons, 94 as there is evidence of differential changes in different ®bre populations. 3,4,17 We examined peptide expression three days, three weeks and three months following injury. Our survival periods were chosen to include the short and longer term changes that may occur after injury, and to permit regeneration back to the periphery in the longest of our recovery groups. 27 The intervals also matched those used in a parallel series of electrophysiological studies 14 and other observations on peptide changes at the site of nerve injury. 54 EXPERIMENTAL PROCEDURES

Nerve injury The experiments were carried out on 22 adult female ferrets (Mustela putorius furo) weighing 0.75±1.5 kg, and all procedures were carried out under UK Home Of®ce licence regulations and approval. The animals were fed on a carnivorous diet with free access to water. In 16 animals, under anaesthesia (ketamine, 23.5 mg/kg, and xylazine, 1.2 mg/kg, i.m.), the left IAN was exposed at the level of the lower third premolar via an extraoral incision along the lower border of the mandible. Bone overlying the mandibular canal was removed with dental burs (see Fig. 1), the nerve was separated from the surrounding blood vessels and was sectioned using micro-scissors. The ends of the nerve were replaced in apposition, the tissues closed in layers and an antibiotic (ampicillin, 15 mg/kg, i.m.) administered. The

657

Neuropeptide expression in the trigeminal ganglion

animals were left to recover for either three days (®ve animals), three weeks (®ve animals) or three months (six animals). A further group of six animals was used as controls. Retrograde tracing Seven days before the end of the recovery period, retrograde tracers were injected into the IAN to enable identi®cation of cell bodies within the trigeminal ganglion which had axons in the damaged nerve. Fluorogold (FG) was used to label all cell body types and isolectin B4 conjugated to horseradish peroxidase (HRP) was used to identify a subpopulation of these cell bodies with unmyelinated axons. 94 Under anaesthesia (ketamine, 23.5 mg/kg, and xylazine, 1.2 mg/kg, i.m.), the IAN was reexposed approximately 5 mm central to the initial injury, via the same extra-oral approach and by a second area of bone removal over the mandibular canal (see Fig. 1). A small incision was made in the epineurium with a razor chip, and tracer (5 ml) consisting of 2% FG (Fluorochrome Inc., USA) and 4% isolectin B4±HRP (Sigma Aldrich, UK) in sterile distilled water was slowly injected into the nerve over a period of 5±10 min, using a centrally directed microdialysis needle (o.d. 0.164 mm; World Precision Instruments, UK) placed beneath the epineurium and connected to a Hamilton syringe. After closure of the surgical site, an antibiotic (ampicillin, 15 mg/kg, i.m.) was again administered. In the three-day survival group, the tracer was injected four days before the nerve injury, allowing a three-day recovery from the injury before perfusion. The six control animals, also under anaesthesia (ketamine, 23.5 mg/kg, and xylazine, 1.2 mg/kg, i.m.), received retrograde tracer injections alone and were perfused seven days later, as described below. Immunohistochemistry At the end of the recovery period the animals were transcardially perfused under deep anaesthesia (sodium pentobarbitone, 42 mg/kg, i.p.) with phosphate-buffered saline (PBS) prewash followed by 4% paraformaldehyde at 48C. The right and left trigeminal ganglia were then removed and post®xed in paraformaldehyde for 3 h at 48C before being placed in 30% sucrose solution overnight, at the same temperature. Sagittal frozen sections, 14 mm thick, were cut from the left trigeminal ganglion and mounted on glass slides coated with poly-d-lysine (Sigma Aldrich, UK). The 14-mm sections were collected onto eight sets of slides, such that each section in a set was 112 mm from the adjacent section in that set. (Six of the eight sets were used for peptide labelling procedures.) Peptide and lectin labelling were revealed in the ganglion tissue using immunocytochemistry as follows. Following a 1-h prewash in 10% normal goat serum (Vector Laboratories) at room temperature, three sets were incubated in a moisture chamber for 18±24 h at 48C with primary antibodies raised in goat against one of SP (1:800), CGRP (1:600) or ENK (1:1600; Genosys Biotechnologies Inc., Cambridge, UK). A further three sets were given a 1-h prewash in 10% normal donkey serum (Jackson ImmunoResearch Laboratories, Inc., PA, USA) and then incubated in antibodies raised in donkey against one of GAL (1:1600; Peninsula Laboratories, Europe Ltd), NPY (1:800; Genosys Biotechnologies Inc., Cambridge, UK) or VIP (1:1000; Peninsula Laboratories, Europe Ltd.). All of the antibodies were diluted in PBS containing 0.1% Triton X-100 and either 5% normal goat serum (SP, CGRP and ENK) or 5% normal donkey serum (GAL, NPY and VIP). The sections were washed for 2 £ 10 min in PBS before being incubated with ¯uorescently labelled secondary antibodies for 90 min at room temperature. CGRP, SP and ENK in conjunction with HRP were revealed by secondary antibodies conjugated to ¯uorescein isothiocyanate (FITC) and indocarbocyanine (CY3), respectively, as follows. Sections were incubated in a mixture of goat anti-rabbit immunoglobulin G conjugated to FITC (CGRP 1:20, SP and ENK 1:30; Vector Laboratories) and goat anti-HRP conjugated to CY3 (1:200, Jackson ImmunoResearch Laboratories Inc., PA, USA), diluted in PBS containing 0.1% Triton X-100 and 5% normal goat serum. GAL, NPY and VIP were

revealed by incubation in donkey anti-rabbit immunoglobulin G conjugated to CY3 (1:200, Jackson ImmunoResearch Laboratories, Inc., PA, USA) diluted in PBS containing 0.1% Triton X-100 and 5% normal donkey serum. FG is visible without immunological processing. Finally, the sections were washed for 2 £ 5 min in PBS and coverslipped with Vectashield (Vector Laboratories). Immunohistochemical controls were performed by preabsorption of the primary antibodies with their respective haptens (10 nm/ml). Cell counts The sections were viewed under a Zeiss Axioplan Fluorescent microscope. Sections cut from the right ganglia (14 mm) were examined for any contralateral spread of the tracer. For the left ganglia, a camera lucida attachment was used to count the total number of FG-positive cells in three alternate sections within each set (i.e. at 224-mm intervals), with a minimum cell count of 200 FG-positive cells per set. Using drawings of the labelled cells and viewing the sections through appropriate ®lters, the numbers of double (FG and neuropeptide) and, if applicable, triple (FG, neuropeptide and lectin) labelled cells were then counted. Approximate cell sizes of FG, neuropeptide and lectin cells were also determined. These were taken from cell pro®les with visible nuclei, measured through the nucleus using an eyepiece micrometer. One-way ANOVA was employed to compare the numbers of labelled cells in the four groups of animals. This analysis was undertaken for neuropeptide labelling in FG- and lectin-positive cells, and for the total number of lectin-labelled cells counted. When this test showed a signi®cant difference between the four groups, pairwise comparisons were carried out to see where the differences lay using Student's t-test with Bonferroni correction. In all, three pairwise comparisons were carried out following each ANOVA to determine any signi®cant differences between the control group and each of the three groups of injured animals. RESULTS

All of the animals recovered well from the surgery and none showed any clear behavioural signs of sensory disturbance or pain, as we have noted previously in our studies in ferrets. 15 Fluorogold-labelled cells FG-positive cells were seen in discrete regions of the ipsilateral ganglia, within the area of the mandibular division (Figs 2A, D and 3A, C, E), and no FG cells were observed in any of the contralateral ganglia. It was possible to count an adequate number of FGpositive cells from three alternate sections in each animal and the number ranged from 233 to 1920 (mean 687) per set. Table 1 shows the mean number of cells (and S.E.M.) counted for both FG and lectin for each peptide, in each group of animals. Although all of the FG-positive cells seen were counted, not all of these had a visible nucleus (see Figs 2A, D and 3A, C, E). However, as the counted sections were at 224-mm intervals, there was no possibility of counting any cell more than once. The cell sizes of those positive cells where a nucleus was visible ranged from approximately 25 to 50 mm in diameter. Preabsorption of the antibodies with their respective haptens abolished all peptide immunoreactivity, con®rming the speci®city of staining (see Fig. 2F for SP).

658

C. Elcock et al.

Fig. 2. (A±C) Photomicrographs of trigeminal ganglia showing triple labelling in a control animal (on the same section viewed through different ®lters). (A) FG-positive cells identifying IAN cell bodies. (B) FITC-positive SP-containing cells. (C) CY3-positive lectin-containing cells. The straight arrow indicates a triple-labelled cell (FG-, SP- and lectin-positive). The curved arrow indicates a cell labelled with FG alone. (D, E) Photomicrographs of trigeminal ganglia showing a reduction in SP labelling three days after section injury (same section viewed through different ®lters). (D) FG labelling. (E) FITC-positive SP-containing cells. The straight arrow indicates a double-labelled cell. (F) Photomicrograph of trigeminal ganglia showing preabsorption control for SP. No labelled cells are evident. Scale bar ˆ 100 mm.

Neuropeptide expression in the trigeminal ganglion

Fig. 3. (A, B) Photomicrographs of trigeminal ganglia in a control animal. (A) FG-positive cells. (B) CY3-positive GAL-containing cells. The straight arrow indicates a double-labelled cell. (C, D) Photomicrographs of trigeminal ganglia in an animal three weeks after section injury. (C) FG-positive cells. (D) FITC-positive ENK-containing cells. The curved arrow indicates a double-labelled cell. (E, F) Photomicrographs of trigeminal ganglia in an animal three days after section injury. (E) FG-positive cells. (F) CY3positive NPY-containing cells. The open arrow indicates a double-labelled cell. Scale bar ˆ 100 mm (A, B, E, F), 50 mm (C, D).

659

660

C. Elcock et al. Table 1. The mean number of peptide cells counted per number of Fluorogold cells and per number of lectin cells SP

CGRP

ENK

GAL

NPY

Control FG Lectin

167/640 (^46) 26/44 (^16)

171/657 (^37) 32/48 (^15)

22/721 (^10) 2/32 (^14)

22/756 (^12) N/A

6/620 (^1) N/A

Three days FG Lectin

52/816 (^18) 2/5 (^3)

87/756 (^25) 5/8 (^4)

13/801 (^2) 0/4 (^2)

11/830 (^5) N/A

18/850 (^15) N/A

Three weeks FG Lectin

81/720 (^39) 3/8 (^5)

112/916 (^14) 6/10 (^5)

71/794 (^43) 3/12 (^7)

8/721 (^3) N/A

8/947 (^2) N/A

Three months FG Lectin

80/603 (^29) 7/14 (^9)

129/579 (^26) 6/14 (^11)

18/603 (^7) 3/19 (^12)

14/578 (^4) N/A

2/542 (^0) N/A

Standard errors are given in parentheses. Data for VIP cells are not shown due to the very low number of cells double-labelled with FG in any animal.

FG-positive cells containing SP were present in all groups of animals (Fig. 2B, E) and the majority of these were small, with diameters of approximately 25 mm. In the control animals, a mean of 26.2% of the FG-positive cells contained SP (Fig. 4A), although there was wide variation between animals (range 7.3±56.6%). Comparison between the four groups revealed signi®cant differences in peptide expression (ANOVA, P ˆ 0.0229). Following nerve section, the proportions were signi®cantly reduced to 6.4% (range 1.6±11.1%, P ˆ 0.0025; Fig. 2E), 10.3% (range 2.45±16.1%, P ˆ 0.0094) and 12.3% (range 4.4±21.2%, P ˆ 0.0149), at three days, three weeks and three months, respectively. CGRP was present in a mean of 26.7% (range 20.5± 42.9%) of FG-positive cells in the control group (Fig. 4B). Both small and larger diameter cells stained positively for this peptide (approximately 25±50 mm) in all animals. Comparison between the four groups revealed signi®cant differences in peptide expression (ANOVA, P ˆ 0.0013). At three days and three weeks following nerve section, the proportion had signi®cantly reduced to 11.7% (range 8±19.2%, P ˆ 0.0003) and 13.1% (range 9.5±18.5%, P ˆ 0.0007), respectively. Three months following injury, this proportion had risen to 22.2% (range 11.5±27.3%, P ˆ 0.1018). ENK was found in 2.5% (range 0.56±5.6%) of FGpositive cells in the control animals (Fig. 3D), and in 1.8% (range 0.38±3.1%), 6.9% (range 2.6±14.5%) and 3.0% (range 0.85±8.3%) of these cells at three days, three weeks and three months following section, respectively (Fig. 4C). None of the injury groups were signi®cantly different from the control group (ANOVA, P ˆ 0.0696) and most of the positive cells were small (approximately 25 mm in diameter). Although low in number, the majority of ENK-positive cells seen in the ganglion tissue were also FG positive. A large number of small cells (approximately 25 mm in diameter) stained positively for GAL in all regions of the ganglion, but the majority of these did not contain FG in any of the groups (Fig. 3B). The proportion of

FG-positive cells containing GAL was 2.4% (range 0.44±4.8%) in the control group (Fig. 4D), and this proportion was not signi®cantly altered at three days (1.2%, range 0.33±1.84%), three weeks (1%, range 0± 2.01%) or three months (2.2%, range 0.22±5.25%) after injury (ANOVA, P ˆ 0.2971). The proportion of FG-positive cells expressing NPY was also low (Figs 3F and 4E). However, unlike the distribution of GAL, very few labelled cells were seen in other regions of the ganglion. Only 1.0% (range 0.47±1.86%) of FG-positive cells in the control group were NPY positive. This was similar at three days (1.1%, range 0±4.26%) and three weeks (0.9%, range 0.35±1.34%) following section, but was reduced to 0.3% at three months (range 0±0.9%). None of the injury groups were signi®cantly different from the control group (ANOVA, P ˆ 0.4126). All NPYlabelled cells were large (30±50 mm) and there was also positive labelling of terminals associated with vasculature. Very few VIP-positive cells were seen in any of the ganglia, and these were generally not FG positive. Those seen were small in diameter (approximately 25 mm). There was, however, a considerable amount of terminal labelling for VIP associated with vasculature. In view of the very low positive cell numbers, meaningful counts were not possible and were therefore not undertaken. Lectin-labelled cells All cells positively stained for the lectin tracer (i.e. with unmyelinated axons) were small in diameter (approximately 25 mm) (Fig. 2C) and the number of FG-positive lectin-containing cells was very low in all groups (range 0±106). When data on the proportion of lectin-labelled cells within the FG population were pooled from all of the peptide-labelled sections, a signi®cant difference was revealed between the groups (ANOVA, P ˆ 0.0209). In the control group, the

661

Neuropeptide expression in the trigeminal ganglion

Fig. 4. The mean (^S.E.M.) proportion of FG-containing cells which also stained positively for peptides at each time period (3d, three days; 3w, three weeks; 3m, three months) following nerve injury (T-section). (A) SP. (B) CGRP. (C) ENK. (D) GAL. (E) NPY. *P , 0.05, injury group vs control.

proportion of lectin-labelled cells was 6.4% (range 0± 19.3%), and this was signi®cantly reduced at three days (0.6%, range 0±2.86%, P ˆ 0.0025), three weeks (1.4%, range 0±6.4%, P ˆ 0.0074) and three months (2.2%, range 0±8.25%, P ˆ 0.0133) following injury (Fig. 5A). Peptide expression within the population of lectinlabelled cells was higher than for the whole FG-positive population. The mean proportion of SP-labelled cells ranged from 57.2% in controls to 35.5% at three days, 46.1% at three weeks and 47.3% at three months (Fig. 5B) but, because of the variation between animals in each group and the very small numbers of lectin-positive cells, these differences were not signi®cant (ANOVA, P ˆ 0.8856). Similarly, there was CGRP labelling of 67.7% of cells in controls, 64.9% at three days, 54.1% at three weeks and 46.2% at three months (Fig. 5C). Despite this apparent reduction in CGRP expression, there was no signi®cant difference between groups because of interanimal variation (ANOVA, P ˆ 0.8020). The majority of lectin-positive cells did not stain positively for ENK. Again, there was wide variation between animals, and there were no signi®cant changes in expression between controls (6.2%) and at three days (0%), three weeks (23.8%) or three months (16.8%) (ANOVA, P ˆ 0.5684; Fig. 5D).

DISCUSSION

Retrograde cellular labelling Our results showed successful retrograde labelling of two populations of axons in the IAN, both following injury and in controls. The FG-labelled cells were numerous, indicating good uptake of the tracer from the injection site, and the cell sizes were similar to those seen in previous studies on the trigeminal ganglion in the cat. 39 However, the proportion of lectin-positive cells (up to 19%) was very low compared to previous studies. This tracer has been shown to label a subpopulation of dorsal root ganglion (DRG) cells with unmyelinated axons 94 and has little overlap with cells which label for RT97, which has been shown to label only A-®bres. 50 Previous studies report varying proportions of DRG cells to be lectin positive. Wang et al. 94 showed that 51% of rat L5 DRG cells, which have axons in the sciatic nerve, were labelled with lectin tracer. Other workers 66 have reported that approximately 30% of rat L4 DRG cells label for an alternative lectin, soybean agglutinin, which is reported to stain the same subpopulation of cells as lectin B4. 84 Wang et al. 94 included in their counts cells which were only weakly stained, and this may explain the difference between their data and those of

662

C. Elcock et al.

Fig. 5. (A) The mean (^S.E.M.) proportion of lectin-labelled cells (cells with unmyelinated axons) within the FG cell population. (B±D) The mean proportion of lectin-containing cells which also stained positively for peptides, at each time period (3d, three days; 3w, three weeks; 3m, three months) following nerve injury (T-section). (B) SP. (C) CGRP. (D) ENK. *P , 0.05, injury group vs control.

Plenderleith et al. 66 The low proportion of lectin-positive cells seen in our experiments may be due to differences in uptake, transport or lectin binding in the ferret from that in the rat, 94 or it may be related to speci®c differences between the location and ®bre population of the sciatic nerve and the IAN (see below). It could also be due to a difference between cranial and spinal nerves. In support of the latter, Ambalavanar and Morris 5 demonstrated lectin binding in the rat trigeminal ganglion in only 24±29% of cells. It is unlikely that our low proportion of lectin-labelled cells re¯ects a low number of unmyelinated axons within the IAN, as almost equal proportions of myelinated and unmyelinated ®bres have been reported in the IAN of the cat. 29,38 An alternative explanation for the low proportion of lectin-positive cells seen in our study would be the method and timing of tracer injection, but this seems unlikely as we used similar methods to those reported in previous studies. 72,73,94 It has been shown that unmyelinated primary afferents are composed of peptide- and non-peptide-containing afferents 43 and, in accordance with this, that lectinpositive cells are composed of peptide- and nonpeptide-containing subpopulations. 80 There is, however, much debate over the use of lectin as a label to identify speci®c cell types. Silverman and Kruger 80 found that only 10% of lectin-labelled DRG cells stained positively for CGRP, whereas Wang et al. 94 found that 59% of rat L5 DRG cells that were lectin positive were positive for CGRP and 17% also stained positively for SP. This also contrasts with the results of Ambalavanar and Morris, 5 who found that, in the rat trigeminal ganglion, 23% of the lectin-labelled cells stained positively for CGRP and 15% stained positively for SP. Our study in the ferret revealed that a high proportion of the lectin-labelled cells in the trigeminal ganglion contained peptide, with

67.7% labelling for CGRP and 57.2% for SP. In addition, we also noted a signi®cant level of co-localisation of ENK in the lectin-labelled cells, and our results would therefore indicate that this was clearly a predominantly peptidergic population. This would be supported by a study on rats which revealed that, after pretreatment with colchicine, all of the lectin-positive cells in the DRG were also labelled for SP. 84 Our results revealed a signi®cant decrease in the proportion of lectin-positive cells after injury and the proportion of cells remained low over the study period. This observation is consistent with previous reports on the reduction of lectin binding and tracing in the rat dorsal horn following ligation 59 or section 86 of the sciatic nerve. Peyronnard et al. 65 have also shown that injured nerves may take up and transport HRP less ef®ciently than intact nerves. The reduction in lectin-positive cells could also be consistent with the loss of unmyelinated axons, which has been reported central to inferior alveolar 28 or sciatic nerve injury, 4 and loss of the small cells in the associated ganglia. 1,18,69,70 We are not able to establish which of these possibilities explains the reduced lectin labelling found after injury in our study.

Neuropeptide expression Lectin cells. Within the lectin-positive cell population, no signi®cant changes were seen in peptide expression after nerve injury. However, in our study, signi®cant changes would have been hard to detect, as the number of cells in this population was small and wide variations were seen in different animals. Increasing the number of animals studied may have revealed differences. Peptide expression in such cells after injury has not been

Neuropeptide expression in the trigeminal ganglion

investigated previously and so we are unable to make comparisons with published studies. Fluorogold cells. In contrast, our results for the FGlabelled population revealed that changes in neurochemical expression do occur in the trigeminal ganglion following nerve injury. Although there was still variation in peptide expression in different animals, the numbers in each group were suf®cient to reveal signi®cant differences. Some of the changes are similar to those reported previously in DRG cells following spinal nerve injury (e.g. Refs 8 and 89), but others differ markedly, and the results for each peptide will be discussed individually. Substance P and calcitonin gene-related peptide. The number of cells expressing SP was reduced from 26.2% to 6.4% at three days following nerve section; the proportion increased slightly from this level at three weeks and three months, but was still signi®cantly reduced in comparison to the controls. Previous studies on the saphenous and sural cutaneous limb nerves in the rat report a reduction to between 1% and 6% for SP in the DRG following nerve ligation, 8,79,89 and this persisted for up to 12 weeks after injury. Our results also revealed reductions in the proportion of cells expressing CGRP, from 26.7% to 11.7% at three days and to 13.1% at three weeks after nerve section. However, by three months after injury, the proportion was not signi®cantly different from that in controls. Other studies on cutaneous limb nerves have also revealed a reduction in CGRP after injury (from 37.8% to 17.5%), which persisted for up to six weeks. 25,63,64,90 These two neuropeptides are known to be co-localised with excitatory amino acids such as glutamate 10,57 and to have excitatory or modulatory roles in neuronal transmission (for review, see Ref. 102). They are released in the spinal cord after C-®bre activation 49,60,78,101 and a reduction in the cell proportion expressing this peptide might imply a change in the potential for excitation of postsynaptic neurons. The mechanism underlying the injury-induced changes in SP and CGRP expression is not clear; there could be a decrease in the cellular production of peptide, an increase in its transport or a selective death of the peptide-producing cells (for review, see Ref. 32), although the latter seems unlikely in view of the return to normal CGRP expression over time. Twelve weeks after IAN section in the ferret, some regeneration of axons back to the periphery will have occurred 27 and the restoration of CGRP expression may be indicative of the return to function of the injured axons. It is also interesting to note that the changes in SP and CGRP expression are, in some ways, similar. As these peptides are co-localised in a subpopulation of neurons, 51,52 it is possible that these changes may be occurring in the same group of cells. However, we are unable to determine this from our data. Galanin. Our experiments revealed only low levels of expression of GAL in the normal ferret trigeminal ganglion, with generally a reduction from these levels after IAN section. Previous studies have demonstrated that, in

663

normal rats, the number of DRG cells expressing GAL is low, 36,82,90 but there is a dramatic up-regulation following sciatic nerve injury. 36,47,89 This is in marked contrast to our data. However, studies on other species, such as the guinea-pig 75 and cat, 6 have revealed only a moderate increase, much less dramatic than that reported in the rat. 36,89 GAL is known to inhibit transmission in the dorsal horn 36,97,103 and has been shown to have analgesic actions. 67,90 It seems likely that these effects would be enhanced following sciatic nerve injury, but our results suggest that it is unlikely that GAL has this role following injury to the ferret IAN. Vasoactive intestinal polypeptide. Our experiments revealed very few cells containing VIP, either in the normal trigeminal ganglion or after nerve injury. This is consistent with previous reports of VIP in the DRG and trigeminal ganglion of normal rats. 7,79 However, it is in stark contrast to the dramatic and persistent upregulation in VIP expression found after nerve injury, 7,31,79 where it has been reported to be found in cells in which the expression of other neuropeptides has been depleted. 79 In support of our ®ndings, however, there is reported to be no up-regulation of VIP in the DRG (and dorsal horn) of the monkey, 106 and in the DRG of the guinea-pig, following sciatic injury. 75 The injury used in Atkinson and Shehab's rat study 7 involved section of the whole of the mandibular division of the trigeminal nerve, close to the trigeminal ganglion. As it has been reported that more severe cellular changes occur when the axonal injury is closer to the nerve cell body, 53 the difference may result from the severity of the injury. The effect of the cellular up-regulation of VIP in the rat is unknown, although the peptide is believed to have an excitatory role and is released in the spinal cord following C-®bre activation in vivo. 100 However, in view of reports of its prolonged expression, reparative functions have also been proposed. In view of our ®ndings, it is unlikely that VIP has such a role following injury to the ferret IAN. Enkephalin. Few cells in the normal trigeminal ganglion contained ENK in our study and there were no signi®cant changes in expression after injury. ENK expression has not been investigated previously in trigeminal ganglion cells after nerve injury. However, in agreement with our results, no change in the level of expression of ENK in the DRG has been reported after sciatic nerve injury, 106 although there was a downregulation of ENK receptors. 104 ENK is known to play a role in modulating nociceptive inputs. 9,16,26,95,99 It is found in descending inhibitory systems 61 and is known to be present in the trigeminal nuclear complex of the ferret. 12 Our data would indicate that ENK originating within trigeminal ganglion neurons does not play a role in the modulation or development of nerve injury-induced pain. Neuropeptide Y. Even fewer cells were seen to express NPY in the trigeminal ganglion in our study, with no change after injury. Two other studies have reported a

664

C. Elcock et al.

marked up-regulation of NPY expression following trigeminal nerve injury in rats. 30,93 However, in one study this up-regulation was in young rats 30 and the site of injury was close to the mandibular foramen, and the other involved injury to different branches of the rat mandibular nerve. 92 Injury in neonates or young adults has been reported to result in more marked changes than in the adult 2,19 (also see Refs 71 and 96) and, as discussed above, more severe changes are reported when the injury is closer to the nerve cell bodies. 53 These factors may explain why our results differ from those of these other studies. However, one other study on the sciatic nerve of the monkey showed a less dramatic increase in NPY expression after injury, 106 more comparable with our data. NPY is known to play a role in modulating nociceptive inputs 42,62 and is believed to have an analgesic role. Thus, the increase in expression reported in previous studies may indicate a mechanism for a reduction in nociceptive input occurring as a consequence of nerve injury. However, our data suggest that it is unlikely to play such a role in the ferret following IAN injury. Other inferior alveolar nerve changes after injury Other changes occur following injury to the IAN which may lead to, or combine with, the central neurochemical changes which we have described. Our previous electrophysiological studies have shown the development of ectopic neural discharge from myelinated afferents in the IAN following injury by ligation, 14 constriction or nerve section, 15 and the highest level of activity occurs at the early stages (three days) after injury. In parallel immunocytochemical studies, we have shown an accumulation of the neuropeptides SP, CGRP, ENK, GAL, NPY and VIP 11,54 in the IAN at the site of injury, and this accumulation follows a similar time course to that of the ectopic discharge. 11,54 Thus, we have postulated that the peripheral neuropeptide accumulation may play a role in the development of the ectopic discharge. Our present data show that the peripheral neuropeptide accumulation occurs at the same time as a reduction in central expression, possibly implying that both changes are due to an increase in peripheral transport rather than changes in production. Previous studies have demonstrated changes in axonal transport following peripheral nerve injury. 88 Whatever the mechanism of the peptide changes, it is also possible that they may affect central transmission in the trigeminal sensory nuclear complex. The brainstem tissue from these animals has also been harvested and we are currently investigating changes in peptide expression within the central terminals. Comparison with non-trigeminal nerves As indicated above, one possible reason for the differences in peptide expression found in our investigation and in previous studies may be related to the IAN itself and its unusual anatomy. As described in our introduction, it is a branch of a cranial nerve carrying sensory afferent ®bres and a few sympathetic efferents and it lies within a bony canal. It is markedly different from a limb

nerve, such as the sciatic, which is a mixed nerve lying within soft tissues. The support of the bony canal around the IAN prevents retraction after injury, maintains apposition of the nerve stumps and increases the potential for regeneration. 40 Other studies have also provided evidence that responses to injury in the trigeminal nerve differ from those reported in limb nerves. For example, electrophysiological studies of ours 14 and others 87 have shown differences in the nature of ectopic neural discharge. In addition, the injury-induced sprouting of sympathetic ®bres around cells in the DRG after sciatic nerve injury 20,56,98 does not appear to be paralleled in the trigeminal ganglion after IAN injury. 13 These factors, coupled with the lower proportion of sympathetic ®bres carried in the trigeminal nerve branches, may explain the lower incidence of causalgia and other sympathetically evoked pain syndromes seen in the orofacial region. 35,55,87 In contrast, however, some other studies undertaken on trigeminal nerve branches have revealed responses comparable with those seen in DRG cells. As discussed above, Wakisaka et al. 93 and Fristad et al. 30 found an upregulation of NPY in the trigeminal ganglion after injury to different branches of the mandibular nerve, and Fristad et al. 31 found an up-regulation of VIP. In addition, Zhang et al. 105 found changes in the trigeminal ganglion following nerve injury which closely resembled those seen in the DRGs, with increases in expression of GAL, NPY and VIP (and their respective mRNAs), amongst others. 108 Interestingly, all of these experiments were undertaken in rats and the nerve injury was carried out further centrally than in our study. In support of the suggestion that species differences are important in the changes observed after injury, studies in the monkey 106 showed only a few NPY-containing cells in the DRG after sciatic nerve axotomy, and no VIP-containing cells, unlike the studies by the same authors in the rat. 105 In addition, studies in the guinea-pig 75 have also demonstrated no up-regulation of VIP after peripheral nerve injury and less dramatic increases in GAL expression have been seen in the cat6 than reported in other species. CONCLUSION

Within the ferret trigeminal system, changes in peptide expression following injury are markedly different from those reported previously in other systems. The reasons for these differences are not clear, but they may be related to the ®bre types within this nerve, its anatomical location within the mandibular canal, species variations or differences between spinal and trigeminal nerves. As all of these peptides play a role in the processing of nociceptive information, it is likely that the injury-induced changes will be important in the development of neuropathic pain. A clearer understanding of these changes may permit the development of therapeutic regimes to treat persistent trigeminal sensory disturbances. AcknowledgementsÐWe would like to thank Mrs Sarah Bodell and Mrs Adele Long for their excellent technical assistance with this work. This work was supported by the Medical Research Council (UK).

Neuropeptide expression in the trigeminal ganglion

665

REFERENCES

1. Aldskogius H. and Arvidsson J. (1978) Nerve cell degeneration and death in trigeminal ganglion of adult rat following peripheral nerve injury. J. Neurocytol. 7, 229±250. 2. Aldskogius H., Arvidsson J. and Grant G. (1985) The reaction of primary sensory neurons to peripheral nerve injury with particular emphasis on transganglionic changes. Brain Res. 10, 27±46. 3. Aldskogius H. and Risling M. (1981) Effect of sciatic neurectomy on neuronal number and size distribution in the L7 ganglion of kittens. Expl Neurol. 74, 597±604. 4. Aldskogius H. and Risling M. (1983) Preferential loss of unmyelinated L7 dorsal root axons following sciatic nerve resection in kittens. Brain Res. 289, 358±361. 5. Ambalavanar R. and Morris R. (1992) The distribution of binding by isolectin I-B4 from Griffonia simplicifolia in the trigeminal ganglion and brainstem trigeminal nuclei in the rat. Neuroscience 47, 421±429. 6. Arvidsson U., Ulfhake B., Cullheim S., Bergstrand A., Theodorsson E. and HoÈkfelt T. (1991) Distribution of 125I-Galanin binding sites, immunoreactive galanin, and its coexistence with 5-hydroxytryptamine in the cat spinal cord: biochemical, histochemical, and experimental studies at the light and electron microscopic level. J. comp. Neurol. 308, 115±138. 7. Atkinson M. E. and Shehab S. A. S. (1986) Peripheral axotomy of the rat mandibular trigeminal nerve leads to an increase in VIP and decrease of other primary afferent neuropeptides in the spinal trigeminal nucleus. Regul. Pept. 16, 69±82. 8. Baranowski A. P., Priestley J. V. and McMahon S. (1993) Substance P in cutaneous primary sensory neuronsÐa comparison of models of nerve injury that allow varying degrees of regeneration. Neuroscience 55, 1025±1036. 9. Basbaum A. I. and Fields H. L. (1984) Endogenous pain control systems: brainstem spinal pathways and endorphin circuitry. A. Rev. Neurosci. 7, 309±332. 10. Battaglia G. and Rustioni A. (1988) Coexistence of glutamate and substance P in dorsal root ganglion neurons of the rat and monkey. J. comp. Neurol. 277, 302±312. 11. Bird E. V., Boissonade F. M. B. and Robinson P. P. (1998) Neuropeptide expression in the inferior alveolar nerve following transection injury. J. dent. Res. 77, 107. 12. Boissonade F. M., Sharkey K. A. and Lucier G. E. (1993) Trigeminal nuclear complex of the ferret: anatomical and histochemical studies. J. comp. Neurol. 329, 291±312. 13. Bongenhielm U., Boissonade F. M., Westermark A., Robinson P. P. and Fried K. (2000) Sympathetic nerve sprouting fails to occur in the trigeminal ganglion after peripheral nerve injury in the rat. Pain 82, 283±288. 14. Bongenhielm U. and Robinson P. P. (1996) Spontaneous and mechanically evoked afferent activity originating from myelinated ®bres in ferret inferior alveolar nerve neuromas. Pain 67, 399±406. 15. Bongenhielm U. and Robinson P. P. (1998) Afferent activity from myelinated inferior alveolar nerve ®bres in ferrets after constriction or section and regeneration. Pain 74, 123±132. 16. Budai D. and Fields H. L. (1998) Endogenous opioid peptides acting at mu-opioid receptors in the dorsal horn contribute to midbrain modulation of spinal nociceptive neurons. J. Neurophysiol. 79, 677±687. 17. Carter D. A. and Lisney S. J. W. (1987) The numbers of unmyelinated and myelinated axons in normal and regenerated rat saphenous nerves. J. neurol. Sci. 80, 163±171. 18. Cavanaugh M. V. (1951) Quantitative effects of the peripheral innervation on nerve and spinal ganglion cells. J. comp. Neurol. 94, 181±219. 19. Chiaia N. L., Hess P. R. and Rhoades R. W. (1987) Preventing regeneration of infraorbital axons does not alter the ganglionic or transganglionic consequences of neonatal transection of this trigeminal branch. Devl Brain Res. 36, 75±88. 20. Chung K., Kim H. J., Na H. S., Park M. J. and Chung J. M. (1993) Abnormalities of sympathetic innervation in the area of an injured peripheral nerve in a rat model of neuropathic pain. Neurosci. Lett. 162, 85±88. 21. Devor M. (1994) The pathophysiology of damaged nerves. In Textbook of Pain (eds Wall P. D. and Melzack R.), Chap. 4, pp. 79±100. Churchill Livingstone, Edinburgh. 22. Devor M., Basbaum A. I., Bennett G. J., Blumberg H., Campbell J. N., Dembowsky K. P., Guilbaud G., JaÈnig W., Koltzenburg M., Levine J. D., Otten U. H. and Portenoy R. K. (1991) Group report: mechanisms of neuropathic pain following peripheral nerve injury. In Towards a New Pharmacotherapy of Pain (eds Basbaum A. I. and Besson J.-M.), pp. 417±440. Wiley, Chichester. 23. Devor M. and Wall P. D. (1981) Plasticity in the spinal cord sensory map following peripheral nerve injury in rats. J. Neurosci. 1, 679±684. 24. Doughty S. E., Atkinson M. E. and Shehab S. A. (1991) A quantitative study of neuropeptide immunoreactive cell bodies of primary afferent neurones following rat sciatic nerve peripheral axotomy. Regul. Pept. 35, 59±72. 25. Dumoulin F. L., Raivich G., Streit W. J. and Kreutzberg G. W. (1991) Differential regulation of calcitonin gene-related peptide (CGRP) in regenerating rat facial nucleus and dorsal root ganglion. Eur. J. Neurosci. 3, 338±342. 26. Fields H. L. (1988) Sources of variability in the sensation of pain. Pain 33, 195±200. 27. Foster E. and Robinson P. P. (1994) The effect of nerve injury on the incidence and distribution of branched pulpal axons in the ferret. J. dent. Res. 73, 1803±1810. 28. Fried K. and Erdelyi G. (1982) Inferior alveolar nerve regeneration and incisor pulpal reinnervation following intramandibular neurotomy in the cat. Brain Res. 244, 259±268. 29. Fried K. and Hildebrand C. (1982) Axon number and size distribution in the developing feline inferior alveolar nerve. J. neurol. Sci. 53, 169±180. 30. Fristad I., Heyeraas K. J. and Hals Kvinnsland I. (1996) Neuropeptide Y expression in trigeminal ganglion and mandibular division of the trigeminal nerve after inferior alveolar nerve axotomy in young rats. Expl Neurol. 142, 276±286. 31. Fristad I., Jacobsen E. B. and Hals Kvinnsland I. (1998) Coexpression of vasoactive intestinal polypeptide and substance P in reinnervating pulpal nerves and in trigeminal ganglion neurones after axotomy of the inferior alveolar nerve axotomy in the rat. Archs oral Biol. 43, 183±189. 32. Grant G. and Arvidsson J. (1975) Transganglionic degeneration in trigeminal primary sensory neurons. Brain Res. 95, 265±279. 33. Gregg J. M. (1990) Studies of traumatic neuralgia in the maxillofacial region: symptom complexes and response to microsurgery. J. oral maxillofac. Surg. 48, 135±140. 34. Gregg J. M. (1990) Studies of traumatic neuralgia in the maxillofacial region: surgical pathology and neural mechanisms. J. oral maxillofac. Surg. 48, 228±237. 35. Hoffmann K. D. and Matthews M. A. (1990) Comparison of sympathetic neurons in orofacial and upper extremity nerves: implications for causalgia. J. oral maxillofac. Surg. 48, 720±726.

666

C. Elcock et al.

36. HoÈkfelt T., Weisenfeld-Hallin Z., Villar M. and Melander T. (1987) Increase of galanin-like immunoreactivity in rat dorsal root ganglion after peripheral axotomy. Neurosci. Lett. 83, 217±220. 37. HoÈkfelt T., Zhang X. and Wiesenfeld-Hallin Z. (1994) Messenger plasticity in primary sensory neurones following axotomy and its functional implications. Trends Neurosci. 17, 22±30. 38. Holland G. R. (1978) Fibre numbers and sizes in the inferior alveolar nerve of the cat. J. Anat., Lond. 127, 343±352. 39. Holland G. R. and Robinson P. P. (1990) Cell counts in the trigeminal ganglion of the cat after inferior alveolar nerve injuries. J. Anat. 171, 179±186. 40. Holland G. R. and Robinson P. P. (1990) The number and size of axons central and peripheral to inferior alveolar nerve injuries in the cat. J. Anat. 173, 129±137. 41. Holland G. R. and Robinson P. P. (1998) Peripheral nerve damage and repair. In Clinical Oral Science (eds Harris M., Edgar M. and Meghji S.), Chap. 22, pp. 274±289. Wright, Oxford. 42. Hua X., Boublik J. H., Spicer M. A., Rivier J. E., Brown M. R. and Yaksh T. L. (1991) The antinociceptive effects of spinally administered neuropeptide Y in the rat; systematic studies on structure±activity relationship. J. Pharmac. exp. Ther. 258, 243±248. 43. Hunt S. P. and Rossi J. (1985) Peptide- and non-peptide-containing unmyelinated primary afferents: the parallel processing of nociceptive information. Phil. Trans. R. Soc. Lond. B 308, 283±289. 44. Jacquin M. F. and Ziegler H. P. (1982) Trigeminal orosensory deafferentation disrupts feeding and drinking mechanism in the rat. Brain Res. 238, 198±204. 45. Jessell T. M., Iversen L. L. and Cuello A. C. (1978) Capsaicin-induced depletion of substance P from primary sensory neurones. Brain Res. 152, 183±188. 46. Jessell T., Tsunoo A., Kanazawa I. and Otsuka M. (1979) Substance P: depletion in the dorsal horn of the rat spinal cord after section of the peripheral processes of primary sensory neurones. Brain Res. 168, 247±259. 47. Kashiba H., Senba E., Ueda Y. and Tohyama M. (1992) Co-localised but target-unrelated expression of vasoactive intestinal polypeptide and galanin in rat dorsal root ganglion neurons after peripheral nerve crush injury. Brain Res. 582, 47±57. 48. Kipp D. P., Goldstein B. H. and Weiss W. W. (1980) Dysaesthesia after mandibular third molar surgery: a retrospective study. J. Am. dent. Ass. 100, 185±192. 49. Kuraishi Y., Hirota N., Sato Y., Hanashima N., Takagi H. and Satoh M. (1989) Stimulus speci®city of peripherally evoked substance P release from rabbit dorsal horn in situ. Neuroscience 30, 241±250. 50. Lawson S. N. and Waddell P. J. (1991) Soma neuro®lament immunoreactivity is related to cell size and ®bre conduction velocity in rat primary sensory neurons. J. Physiol. 435, 41±63. 51. Lee Y., Kawai Y., Shiosaka S., Takami K., Kiyama H., Hillyard C. J., Girgis S., MacIntyre I., Emson P. C. and Tohyama M. (1985) Coexistence of calcitonin gene-related peptide and substance P-like peptide in single cells of the trigeminal ganglion of the rat: immunohistochemical analysis. Brain Res. 330, 194±196. 52. Lee Y., Takami K., Kawai Y., Girgis S., Hillyard C. J., MacIntyre I., Emson P. C. and Tohyama M. (1985) Distribution of calcitonin generelated peptide in the rat peripheral nervous system with reference to its coexistence with substance P. Neuroscience 15, 1227±1237. 53. Lieberman A. R. (1974) Some factors affecting retrograde neuronal responses to axonal lesions. In Essays in the Nervous System (eds Bellairs R. and Gray E. G.), pp. 71±105. Clarendon, Oxford. 54. Long A., Bongenhielm U., Boissonade F. M. B., Fried K. and Robinson P. P. (1998) Neuropeptide immunoreactivity in ligature-induced neuromas of the inferior alveolar nerve in the ferret. Brain Res. 791, 263±270. 55. Matthews B. (1989) Autonomic mechanisms in oral sensations. Proc. Finn. dent. Soc. 85, 365±373. 56. McLachlan E. M., JaÈnig W., Devor M. and Michaelis M. (1993) Peripheral nerve injury triggers noradrenergic sprouting within dorsal root ganglia. Nature 363, 543±546. 57. Merighi A., Polak J. M. and Theodosis D. T. (1991) Ultrastructural visualisation of glutamate and aspartate immunoreactivities in the rat dorsal horn, with special reference to the co-localisation of glutamate, substance P and calcitonin gene-related peptide. Neuroscience 40, 67±80. 58. Miller M. G. (1981) Trigeminal deafferentation and ingestive behaviour in rats. J. comp. physiol. Psychol. 95, 252±269. 59. Molander C., Wang H. F., Riverio-MeliaÂn C. and Grant G. (1996) Early decline and late restoration of spinal cord binding and transganglionic transport of isolectin B4 from Griffonia simplicifolia I after peripheral nerve transection or crush. Rest. Neurol. Neurosci. 10, 123±133. 60. Morton C. R. and Hutchinson W. D. (1990) Morphine does not reduce the intraspinal release of calcitonin gene-related peptide in the cat. Neurosci. Lett. 117, 319±324. 61. Moss M. S., Glazer E. J. and Basbaum A. I. (1993) The peptidergic organisation of the cat periaqueductal grey. I. The distribution of enkephalin-containing neurons and terminals. J. Neurosci. 13, 603±616. 62. Munglani R., Hudspith M. J. and Hunt S. P. (1996) The therapeutic potential of neuropeptide Y. Analgesic, anxiolytic and antihypertensive. Drugs 52, 371±389. 63. Noguchi K., De LeoÂn M., Nahin R. L., Senba E. and Ruda M. A. (1993) Quanti®cation of axotomy-induced alteration of neuropeptide mRNAs in dorsal root ganglion neurons with special reference to neuropeptide Y mRNA and the effects of neonatal capsaicin treatment. J. Neurosci. 35, 54±66. 64. Noguchi K., Senba E., Morita Y., Sato M. and Toyhama M. (1990) a-CGRP and b-CGRP mRNAs are differentially regulated in the rat spinal cord and dorsal root ganglion. Molec. Brain Res. 7, 299±304. 65. Peyronnard J. M., Charron L., Lavoie J., Messier J. P. and Bergouignan F. X. (1988) A comparative study of the effects of chronic axotomy, crush lesion and re-anastomosis of the rat sural nerve on horseradish peroxidase labelling of primary sensory neurons. Brain Res. 443, 295±309. 66. Plenderleith M. B., Cameron A. A., Key B. and Snow P. J. (1989) The plant lectin soybean agglutinin binds to the soma, axon and central terminals of a subpopulation of small diameter primary sensory neurons in the rat and cat. Neuroscience 31, 683±695. 67. Post C., Alari L. and HoÈkfelt T. (1988) Intrathecal galanin increases the latency in the tail-¯ick and hot-plate test in mouse. Acta physiol. scand. 132, 583±584. 68. Raivich G., Hellweg R. and Kreutzberg G. W. (1991) NGF receptor-mediated reduction in axonal NGF uptake and retrograde transport following sciatic nerve injury and during regeneration. Neuron 7, 151±164. 69. Ranson S. W. (1906) Retrograde degeneration in the spinal nerves. J. comp. Neurol. 16, 265±293. 70. Ranson S. W. (1909) Alterations in the spinal ganglion cells following neurotomy. J. comp. Neurol. 19, 125±153. 71. Rhoades R. W., Chiaia N. L., Macdonald G. J. and Jacquin M. F. (1989) Effect of fetal infraorbital nerve transection upon trigeminal primary afferent projections in the rat. J. comp. Neurol. 287, 82±97. 72. Rivero-MeliaÂn C. (1996) The organization of hindlimb muscle nerve projections to the rat spinal cord: a choleragenoid horseradish peroxidase study. J. comp. Neurol. 364, 652±663. 73. Rivero-MeliaÂn C. and Grant G. (1991) Choleragenoid horseradish peroxidase used for studying projections of some hindlimb cutaneous nerves and plantar foot afferents to the dorsal horn and Clarke's column in the rat. Expl Brain Res. 84, 125±132.

Neuropeptide expression in the trigeminal ganglion

667

74. Robinson P. P. (1988) Observations on the recovery of sensation following inferior alveolar nerve injuries. Br. J. oral maxillofac. Surg. 26, 177±189. 75. Rydh-Rinder M., Holmberg K., Elfvin L.-G., Wiesenfeld-Hallin Z. and HoÈkfelt T. (1996) Effects of peripheral axotomy on neuropeptides and nitric oxide synthase in dorsal root ganglia and spinal cord of the guinea pig: an immunohistochemical study. Brain Res. 707, 180±188. 76. Salt T. E. and Hill R. H. (1983) Neurotransmitter candidates of somatosensory primary afferent ®bres. Neuroscience 10, 1083±1103. 77. Sandstedt P. and SoÈrensen S. (1995) Neurosensory disturbances of the trigeminal nerve: a long-term follow-up of traumatic injuries. J. oral maxillofac. Surg. 53, 498±505. 78. Saria A., Gamse R., Petermann J., Fischer J. A., Theodorsson-Norheim E. and Lundberg J. M. (1986) Simultaneous release of several tachykinins and calcitonin gene-related peptide from rat spinal cord slices. Neurosci. Lett. 63, 310±314. 79. Shehab S. A. S. and Atkinson M. E. (1986) Vasoactive intestinal polypeptide increases in areas of the dorsal horn of the spinal cord from which other neuropeptides are depleted following peripheral axotomy. Expl Brain Res. 62, 422±430. 80. Silverman J. D. and Kruger L. (1990) Selective neuronal glycoconjugate expression in sensory and autonomic ganglia: a relation of lectin reactivity to peptide and enzyme markers. J. Neurocytol. 19, 789±801. 81. Sko®tsch G. and Jacobowitz D. M. (1985) Calcitonin gene-related peptide coexists with substance P in capsaicin sensitive neurons and sensory ganglia of the rat. Peptides 6, 747±754. 82. Sko®tsch G. and Jacobowitz D. M. (1985) Galanin-like immunoreactivity in capsaicin sensitive sensory neurons and ganglia. Brain Res. Bull. 15, 191±195. 83. Stern C. W. and Kolunie J. M. (1991) Trigeminal lesions and maternal behaviour in Norway rats: I. Effects of cutaneous rostral snout denervation on maintenance of nuturance and maternal aggression. Behav. Neurosci. 105, 984±997. 84. Streit W. J., Schulte B. A., Balentine J. D. and Spicer S. S. (1986) Evidence for glycoconjugate in nociceptive primary sensory neurons and its origin from the Golgi complex. Brain Res. 377, 1±17. 85. Sunderland S. (1991) Nerve injuries and their repair, a critical appraisal. In Nerve Injury and Sensory Function, Chap. 33, pp. 305±331. Churchill Livingstone, London. 86. Tajti J., Fischer J., Knyihar-Csillik E. and Csillik B. (1988) Transganglionic regulation and ®ne structural localisation of lectin-reactive carbohydrate epitopes in primary sensory neurons of the rat. Histochemistry 88, 213±218. 87. Tal M. and Devor M. (1992) Ectopic discharge in injured nerves: comparison of trigeminal and somatic afferents. Brain Res. 579, 148±151. 88. Tonra J. R., Curtis R., Wong V., Cliffer K. D., Park J. S., Timmes A., Nguyen T., Lindsay R. M., Acheson D. and DiStefano P. S. (1998) Axotomy upregulates the anterograde transport and expression of brain-derived neurotrophic factor by sensory neurons. J. Neurosci. 18, 4374±4383. 89. Villar M. J., CorteÂs R., Theodorsson E., Wiesenfeld-Hallin Z., Schalling Z., Fahrenkrug J., Emson P. C. and HoÈkfelt T. (1989) Neuropeptide expression in dorsal root ganglion cells and spinal cord after peripheral nerve injury with special reference to galanin. Neuroscience 33, 587±604. 90. Villar M. J., Wiesenfeld-Hallin Z., Xu X. J., Theodorsson E., Emson P. C. and HoÈkfelt T. (1991) Further studies on Galanin-, Substance P-, and CGRP-like immunoreactivities in primary sensory neurons and spinal cord: effect of dorsal rhizotomies and sciatic nerve lesions. Expl Neurol. 122, 29±39. 91. Wakisaka S., Kajander K. C. and Bennett G. J. (1991) Increased neuropeptide Y (NPY)-like immunoreactivity in rat sensory neurons following peripheral axotomy. Neurosci. Lett. 124, 200±203. 92. Wakisaka S., Kajander K. C. and Bennett G. J. (1992) Effects of peripheral nerve injuries and tissue in¯ammation on the levels of neuropeptide Y-like immunoreactivity in rat primary afferent neurones. Brain Res. 598, 349±352. 93. Wakisaka S., Takikita S., Sasaski Y., Kato J., Tabata M. J. and Kurisu K. (1993) Cell size-speci®c appearance of neuropeptide Y in the trigeminal ganglion following peripheral axotomy of different branches of the mandibular nerve of the rat. Brain Res. 620, 347±350. 94. Wang H., Rivero-MeliaÂn C., Robertson B. and Grant G. (1994) Transganglionic transport and binding of the isolectin B4 from Griffonia simplicifolia I in rat primary sensory neurons. Neuroscience 62, 539±551. 95. Watkins L. R. and Mayer D. J. (1982) Organization of endogenous opiate and nonopiate pain control systems. Science 216, 1185±1192. 96. White F. A., Bennett-Clarke C. A., Macdonald G. J., En®ejian H. L., Chiaia N. L. and Rhoades R. W. (1990) Neonatal infraorbital nerve transection in the rat: comparison of effects on substance P immunoreactive primary afferents and those recognised by the lectin Bandierea simplicifolia-I. J. comp. Neurol. 300, 249±262. 97. Wiesenfeld-Hallin Z., Villar M. J. and HoÈkfelt T. (1989) The effects of intrathecal galanin and C-®ber stimulation on the ¯exor re¯ex in the rat. Brain Res. 486, 205±213. 98. Xie Y., Zhang J., Petersen M. and LaMotte R. H. (1995) Functional changes in dorsal root ganglion cells after chronic nerve constriction in the rat. J. Neurophysiol. 73, 1811±1820. 99. Xu J. Y., Fujimoto J. M. and Tseng L. F. (1992) Involvement of supraspinal epsilon and mu opioid receptors in inhibition of the tail-¯ick response induced by etorphine in the mouse. J. Pharmac. exp. Ther. 263, 246±252. 100. Yaksh T. L., Abay E. O. and Go V. L. W. (1982) Studies on the location and release of cholecystokinin and vasoactive intestinal polypeptide in the rat and cat spinal cord. Brain Res. 242, 279±290. 101. Yaksh T. L., Jessell T. M., Gamse R., Mudge R. and Leeman S. E. (1980) Intrathecal morphine inhibits substance P release from mammalian spinal cord in vivo. Nature 286, 155±156. 102. Yaksh T. L. and Malmberg A. B. (1994) Central pharmacology of nociceptive transmission. In Textbook of Pain (eds Wall P. D. and Melzack R.), Chap. 9, pp. 165±200. Churchill Livingstone, Edinburgh. 103. Yanagisawa Y.-M., Yagi N., Otsuka M., Yanaihara C. and Yanaihara N. (1986) Inhibitory effects of galanin on the isolated spinal cord of the newborn rat. Neurosci. Lett. 70, 278±282. 104. Zhang X., Bao L., Arvidsson U., Elde R. and HoÈkfelt T. (1998) Localisation and regulation of the delta-opioid receptor in dorsal root ganglia and spinal cord of the rat and monkey: evidence for association with the membrane of large dense-core vesicles. Neuroscience 82, 1225±1242. 105. Zhang X., Ji R.-R., Arvidsson J., Lundberg J. M., Bartfai T., Bedecs K. and HoÈkfelt T. (1996) Expression of peptides, nitric oxide synthase and NPY receptor in trigeminal and nodose ganglia after nerve lesions. Expl Brain Res. 111, 393±404. 106. Zhang X., Ju G., Elde R. and HoÈkfelt T. (1993) Effect of peripheral nerve cut on neuropeptides in dorsal root ganglia and the spinal cord of monkey with special reference to galanin. J. Neurocytol. 22, 342±381. 107. Zhang X., Shi T., Holmberg K., Landry M., Huang W., Xiao H., Ju G. and HoÈkfelt T. (1997) Expression and regulation of neuropeptide Y Y2 receptor in sensory autonomic ganglion. Proc. natn. Acad. Sci. USA 94, 729±734. 108. Zhang X., Wiesenfeld-Hallin Z. and HoÈkfelt T. (1994) Effect of peripheral axotomy on expression of neuropeptide Y receptor mRNA in rat lumbar dorsal root ganglia. Eur. J. Neurosci. 6, 43±57. (Accepted 17 October 2000)