Studies on the changes of c-fos protein in spinal cord and neurotransmitter in dorsal root ganglion of the rat with an experimental peripheral neuropathy

Animal models for human chronic pain syndromes have been developed and widely used for pain research. One of these neuropathic pain models by Kim and Chung (1992) has many advantages for operation and pain elicitation. In this neuropathic model we have examined the c-fos protein, substance P, CGRP immunoreactivity in dorsal root ganglia and dorsal horn. 50 Sprague-Dawley rats were used for this study. L5 and L6 spinal nerves were ligated tightly to produce the neuropathic pain model. After 2, 4, 8, 16, and 24 hours and 1 week of surgery, rats were anesthetized and sacrificed by perfusion. After confirmation of the roots transected by the surgery, the L5 and L6 dorsal root ganglions and spinal cord were removed and processed for immunohistochemistry. All tissue sections were immunohistochemically stained for substance P, CGRP and c-fos using the peroxidase-antiperoxidase (PAP) method. The number of immunostained substance P and CGRP dorsal root ganglion cells and c-fos immunoreactive dorsal horn cells were counted and analyzed statistically with Mann-Whitney U test. The results are as follows. The number of c-fos protein immunoreactive neurons in the superficial layer of dorsal horn were increased markedly 2 hours after operation, and gradually decreased to normal level 1 week after operation. The number of c-fos protein immunoreactive neurons in the deep layer of the dorsal horn gradually increased to a peak 24 hours after operation, then decreased to the normal level 1 week after operation. The number of substance P and CGRP immunoreactive L5 and L6 dorsal root ganglion neurons were decreased markedly 1 week after the pain model operation. In conclusion, after neuropathic pain model operation, c-fos proteins were immediately expressed in the superficial layer of spinal dorsal horn, thereafter c-fos proteins in the deep layer of spinal dorsal horn were expressed. CGRP and substance P immunoreactive neurons in DRG were decreased markedly 1 week after neuropathic pain model operation. These decrements do not coincide with the other chronic pain models, which show great increases in these pain transmitting substances. Therefore, the relationship between pain and c-fos, SP and CGRP should be investigated further.     

Distribution and origin of extrinsic nerve fibers containing calcitonin gene-related peptide, substance P and galanin in the rat upper rectum.

The distributions of nerve fibers containing calcitonin gene-related peptide (CGRP), substance P (SP) and galanin (GAL) were examined in the rat rectum of mutants rats, aganglionic rats (AGRs), which completely lack the intramural nerve cells in the large intestine, and of their normal littermates. The origin of extrinsic peptide-containing nerve fibers was examined using retrograde tracing combined with immunohistochemistry in normal rats. In the rectum of normal rats, CGRP-, SP- and GAL-immunoreactive varicose fibers were observed throughout all layers of the rectal wall, and immunoreactive nerve cells were present in the enteric ganglia of colchicine-treated rats. In the aganglionic rectum of AGR, a rich supply of CGRP-immunoreactive fibers was observed in the mucosa, around the blood vessels, and in the submucous and intermuscular spaces. SP- and GAL-immunoreactive fibers in the aganglionic rectum showed a similar distribution to CGRP-immunoreactive fibers but were less dense. These results suggest that most of CGRP-positive fibers in the rectum are extrinsic whereas a large part of SP- or GAL-positive fibers are intrinsic. Fluoro-gold injected into the upper rectum of normal rat labelled nerve cells (less than 10% of total ganglion cells) in the lumbar (L1 and L2) and lumbosacral (L6 and S1) dorsal root ganglia. More than half of nerve cells in the dorsal root ganglia (L6 and S1) projecting to the rectum were immunoreactive for CGRP, and less than 10% were immunoreactive for SP or GAL. Comparison of serial sections of the dorsal root ganglion revealed that about half of the CGRP-immunoreactive cells were also positive for SP or GAL. These results indicate that SP- or GAL-positive neurons projecting to the rectum are scarce in the dorsal root ganglia. The present investigation suggests that CGRP-containing nerves are visceral afferents forming a major component of the sensory innervation of the rat rectum, and SP- and GAL-containing nerves which share their extrinsic origins appear to form a lesser proportion of the sensory innervation.     

Immunocytochemistry of B-50 (GAP-43) in the spinal cord and in dorsal root ganglia of the adult cat

Summary The distribution of the neural-specific growth associated protein B-50 (GAP-43), which persists in the mature spinal cord and dorsal root ganglia, has been studied by light and electron microscopic immunohistochemistry in the cat. Throughout the spinal cord, B-50 immunoreactivity was seen confined to the neuropil, whereas neuronal cell bodies were unreactive. The most conspicuous immunostaining was observed in the dorsal horn, where it gradually decreased from superficial laminae (I–II) toward more ventral laminae (III–V), and in the central portion of the intermediate gray (mainly lamina X). In these regions, the labelling was localized within unmyelinated, small diameter nerve fibres and axon terminals. In the rest of the intermediate zone (laminae VI–VIII), B-50 immunoreactivity was virtually absent. The intermediolateral nucleus in the thoracic and cranial lumbar cord showed a circumscribed intense B-50 immunoreactivity brought about by the labelling of many axon terminals on preganglionic sympathetic neurons. In motor nuclei of the ventral horn (lamina IX), low levels of B-50 immunoreactivity were present in a few axon terminals on dendritic and somal profiles of motoneurons. In dorsal root ganglia, B-50 immunoreactivity was mainly localized in the cell bodies of small and medium-sized sensory neurons. The selective distribution of persisting B-50 immunoreactivity in the mature cat throughout sensory, motor, and autonomie areas of the spinal cord and in dorsal root ganglia suggests that B-50-positive systems retain in adult life the capacity for structural and functional plasticity.     

Opiate and histamine H1 receptors are present on some substance P-containing dorsal root ganglion cells

Using combined immunohistochemical and receptor binding techniques, substance P-containing sensory neurones of rhesus monkey cervical dorsal root ganglia were examined for the presence of opiate or histamine (H1) receptors. Serial sets of three sections were examined sequentially for substance P-containing neurones, opiate receptors using [3H]etorphine binding and histamine (H1) receptors using [3H]mepyramine binding. Of 3484 dorsal root ganglion cells, 513 contained substance P. Of 30 randomly chosen substance P-positive neurones, 6 possessed opiate receptors and 7 histamine receptors, including 4 neurones with both sets of binding sites. The results are interpreted to suggest that both nociceptors and non-nociceptive sensory inputs may be biochemically heterogeneous and that a simple correlation between substance P content and a particular receptive field profile is unlikely.     

Histochemical characterization of sensory neurons with high-affinity receptors for nerve growth factor

Summary Approximately one half of the neurons in the lumbar dorsal root ganglion of adult rats display high-affinity receptors for nerve growth factor (NGF). To ascertain which types of sensory neurons are potentially responsive to NGF, adjacent cryostat sections of rat dorsal root ganglia were processed either for NGF-receptor using radioautography or by one of four histochemical procedures. Histograms of the densities of neuronal labelling by radioiodinated NGF were examined for subpopulations of lumbar sensory neurons with thiamine monophosphatase enzyme activity or with immunoreactivity for calcitonin gene-related peptide (CGRP), substance P, or somatostatin. Virtually all neurons with strong CGRP immunoreactivity had high-affinity NGF binding sites, although some neurons with faintly positive CGRP immunoreactivity lacked such NGF binding. A subpopulation of large neurons, approximately 5% of the total, had dense labelling by¹²⁵I-NGF but were not stained by this immunohistochemical technique for CGRP. Of the three major populations of small neurons those with substance P immunoreactivity were consistently and heavily labelled by radioiodinated NGF whereas those with somatostatin immunoreactivity or thiamine monophosphatase activity were not specifically labelled by radioautography. For these primary sensory neurons in mature rats the genes for substance P and CGRP seem to be strongly expressed only in neurons capable of responding to NGF. On the other hand, neurons containing somatostatin and thiamine monophosphatase invariably lack high-affinity NGF receptors.     

Time-dependent differences in the increase in GAP-43 expression in dorsal root ganglion cells after peripheral axotomy.

Peripheral axotomy of primary afferent neurons results in the up-regulation of the growth-associated phosphoprotein GAP-43, by dorsal root ganglion cells. We have studied the temporal sequence of GAP-43 expression in those dorsal root ganglion neurons with unmyelinated axons (the small dark cells) and in those with myelinated axons (the large light cells) after sciatic nerve section in the adult rat. Immunoreactivity for the RT 97 neurofilament epitope, which is detectable only in large light dorsal root ganglion cells, was used to differentiate the two types of dorsal root ganglion cell. Within two days of a sciatic nerve section the number of GAP-43-immunoreactive profiles in the ipsilateral ganglion had increased five-fold and this increase persisted for 80 days post-section. While 50% of the small numbers of GAP-43-positive cells in control ganglia were RT 97 positive, only 8% of the large number of GAP-43-immunoreactive cells four days post-section, were RT 97 positive. By 14 days the number of RT 97-positive/GAP-43-positive cells had increased to 29%. This was paralleled by an increase in GAP-43 immunoreactivity in large diameter profiles at 14 days. The signals that alter GAP-43 expression in unmyelinated (small, RT 97 -ve) and myelinated (large, RT 97 +ve) afferents after peripheral nerve injury appear to operate with different time-courses.     

Fluoride-resistant acid phosphatase-containing neurones in dorsal root ganglia are separate from those containing substance P or somatostatin

The distribution of fluoride-resistant acid phosphatase, substance P and somatostatin were investigated in the dorsal horn of the spinal cord and in dorsal root ganglia. In the dorsal horn, the distribution of fluoride-resistant acid phosphatase closely paralleled that of somatostatin and only partly overlapped with that of substance P. In sensory ganglia, none of the fluoride-resistant acid phosphatase-containing neurones contained either substance P or somatostatin. The results suggest the existence of a population of fluoride-resistant phosphatase-positive sensory neurones which is distinct from neurones containing either of these peptides.     

Afferent fibres from the guinea-pig ureter: size and peptide content of the dorsal root ganglion cells of origin.

A study was undertaken to determine the segmental organization of the dorsal root ganglion cells which give rise to ureteric primary afferent fibres in the guinea-pig. The size-distribution and peptide content of these dorsal root ganglion cells were examined and compared with a sample of all dorsal root ganglion cells from the same ganglia. Afferent fibres to the guinea-pig ureter were found to arise mainly from dorsal root ganglia L2-L3 and S1-S2. A large contralateral component of the afferent innervation of the ureter was found when either the right or the left ureter was injected with tracer. This amounted to approximately 40% of the total labelled cells. The cross-sectional areas of the dorsal root ganglion cells of ureteric afferents were found to be at the smaller end of the size-range for the whole ganglion. Most (90%) of the cells innervating the ureter were immunoreactive for one of the peptides studies, substance P or calcitonin gene-related peptide, and a large proportion (65%) were immunoreactive for both. This was very different for the ganglia as a whole, where only about 50% of the cells were immunoreactive for either of the peptides and only 14% were immunoreactive for both peptides. These results show a bilateral afferent innervation of the ureter by nerve fibres which, in the vast majority, contain substance P and/or calcitonin gene-related peptide.     

Capsaicin sensitivity is associated with the expression of the vanilloid (capsaicin) receptor (VR1) mRNA in adult rat sensory ganglia

A vanilloid receptor (VR1) has recently been cloned and shown to be a target for capsaicin, the excitotoxic component of capsicum peppers (Caterina, M.J., Schumacher, M.A., Tominaga, M., Rosen, T.A., Levine, J.D. and Julius, D., Nature, 389 (1997) 816–824). The effects of capsaicin appear to be selective for a subset of sensory neurones which includes polymodal nociceptors. The present study describes the distribution of VR1 mRNA, together with measurements of capsaicin sensitivity, in sensory nerve ganglia of different embryological origins and a single sympathetic ganglion, the superior cervical ganglion (SCG). In situ hybridisation revealed the expression of VR1 mRNA in small-to-medium-sized neurones of the dorsal root, trigeminal and vagal ganglia. No hybridisation signal was observed in the SCG neurones. This pattern of expression correlated with capsaicin sensitivity measured by whole-cell voltage clamp where, in similar sized cells, over 80% of neurones from dorsal root and vagal ganglia were capsaicin sensitive, but all SCG neurones were insensitive to capsaicin.     

Growth-associated proteins and regeneration-induced gene expression in the aging neuron

Axonal elongation and sprouting during regeneration are retarded with aging but the etiology of this is unclear. We investigated whether this age-associated decline is related to a decline in expression of three different growth-associated proteins (GAPs): alpha(1)-tubulin, neurofilament (NF) light subunit (NF-L) and GAP-43. Northern analysis was performed on L4-L5 dorsal root ganglia (DRG) of young (3 months) and aged (23 months) rats following a sciatic nerve crush and compared to their age-matched controls. The results show that initial mRNA levels of alpha(1)-tubulin and NF-L in the control aged rat DRG were half those of the control young adults, whereas expression of GAP-43 was unchanged. Two weeks after axotomy, the expression of alpha(1)-tubulin and GAP-43 in the aged DRG was induced to the same levels as in the axotomized young adult, and the expression of NF-L decreased proportionately in both age groups. These results indicate that certain neuronal mRNAs, such as alpha(1)-tubulin and NF-L may be maintained at lower levels in aging DRG neurons, whereas others, such as GAP-43 appear to be unaltered. However, during regeneration, the aging DRG neuron appears capable of inducing alpha(1)-tubulin, NF-L and GAP-43 as well as the young adult.     

Ultrastructural evidence for the coexistence of calcitonin gene-related peptide and substance P in secretory vesicles of peripheral nerves in the guinea pig

Summary Using double immunogold staining procedures, calcitonin gene-related peptide (CGRP)-like and substance P (SP)-like immunoreactivities were localized at the ultrastructural level to guinea pig trigeminal ganglia, dorsal root ganglia and peripheral nerve fibres associated with the vascular system. CGRP-like and SP-like immunoreactivities were found consistently in large granular secretory vesicles (70–100 nm in diameter), and both peptide immunoreactivities were co-localized to the same vesicle in both sensory ganglion cells and within axons and their terminals in the adventitia and adventitial-medial border of the superior mesenteric artery. These results suggest that CGRP and SP are co-stored and may be released together from peripheral axons in the guinea pig.     

The growth-associated protein GAP-43 appears in dorsal root ganglion cells and in the dorsal horn of the rat spinal cord following peripheral nerve injury

When adult dorsal root ganglion cells are dissociated and maintained in vitro, both the small dark and the large light neurons show increases in the growth-associated protein GAP-43, a membrane phosphoprotein associated with neuronal development and plasticity. Immunoreactivity for GAP-43 appears in the cytoplasm of the cell bodies as early as 3.5 h post axotomy and is present in neurites and growth cones as soon as they develop. At early stages of culture (4 h to eight days) satellite/Schwann cells are also immunoreactive for GAP-43. Neurons in isolated whole dorsal root ganglion maintained in vitro become GAP-43-immunoreactive between 2 and 3 h after axotomy. It takes three days however, after cutting or crushing the sciatic nerve in adult rats in vivo, for GAP-43 immunoreactivity to appear in the axotomized dorsal root ganglion cells. GAP-43 immunoreactivity can be detected in the central terminals of primary afferent neurons in the superficial laminae of the dorsal horn of the lumbar enlargement four days after sciatic cut or crush. The intensity of the GAP-43 staining reaches a peak at 21 days and becomes undetectable nine weeks following crush injury and 36 weeks following sciatic nerve cut. The pattern of GAP-43 staining is identical to the distribution of sciatic small-calibre afferent terminals. Little or no staining is present in the deep dorsal horn, but GAP-43 does appear in the ipsilateral gracile nucleus 22 days after sciatic injury. In investigating the mechanism of GAP-43 regulation, blockade of axon transport in the sciatic nerve with vinblastine (10(-5) M-10(-4) M) or capsaicin (1.5%) was found to produce a pattern of GAP-43 immunoreactivity in the dorsal horn identical to that found with crush, while electrical stimulation of the sciatic nerve had no effect. Axotomy of primary sensory neurons or the interruption of axon transport in the periphery therefore acts to trigger GAP-43 production in the cell body. The GAP-43 is transported to both the peripheral and the central terminals of the afferents. In the CNS the elevated GAP-43 levels may contribute to an inappropriate synaptic reorganization of afferent terminals that could play a role in the sensory disorders that follow nerve injury.     

银杏酮酯对大鼠坐骨神经损伤后生长相关蛋白43表达的影响

目的:研究银杏酮酯对大鼠坐骨神经损伤后生长相关蛋白43(GAP-43)表达的影响。方法:SD大鼠78只,随机分成正常组、损伤对照组与实验组,给予不同处理,后两组切断右侧坐骨神经并缝合。实验组给予银杏酮酯200mg·kg-1.d-1溶于1ml生理盐水中灌胃,损伤对照组给予生理盐水1ml灌胃,正常组不做处理。分别于术后1、3、7、14、21及28d取吻合口远段的神经、相应节段的脊神经节及脊髓,应用免疫组织化学和图像分析的方法研究所取组织中GAP-43蛋白的表达并进行定量分析。结果:实验组坐骨神经、脊神经节及脊髓中GAP-43蛋白免疫阳性区域面积和平均光密度值在术后7、14和21d明显高于对照组。结论:大鼠坐骨神经损伤后用银杏酮酯治疗,在早期可促使坐骨神经及相应节段脊神经节和脊髓组织中的GAP-43蛋白表达增加。     

Distribution and origin of nerve fibers in the rat temporomandibular joint capsule

 The distribution and origin of nerve fibers containing neuropeptides and NOS projecting to the temporomandibular joint capsule (TMJ) of the rat were studied by retrograde tracing in combination with immunocytochemistry. Numerous nerve fibers were seen in the TMJ as revealed by the neuronal marker protein gene product 9.5. Nerve fibers containing neuropeptide Y (NPY), vasoactive intestinal peptide (VIP), pituitary adenylate cyclase activating peptide (PACAP), substance P (SP), calcitonin gene-related peptide (CGRP), and nitric oxide synthase (NOS) were seen in the synovial membrane, the joint capsule and entering the articular disc. Injection of the retrograde tracer True Blue (TB) into the TMJ resulted in the appearance of numerous labeled nerve cell bodies in the trigeminal and superior cervical ganglia, and moderate numbers in the nodose, the otic, the sphenopalatine, the stellate and the dorsal root ganglia at levels C2–C5. Most of the TB-labeled cell bodies in the superior cervical and stellate ganglia contained NPY. In the trigeminal ganglion, numerous TB labeled cell bodies contained CGRP and a minor population stored SP, a few cell bodies were seen to store NOS or PACAP. In the sphenopalatine and otic ganglia, TB labeled cell bodies contained NOS or VIP. In the nodose ganglion, labeled cell bodies contained CGRP; other labeled cell bodies harbored NOS. In the cervical dorsal root ganglia, the majority of the labeled cell bodies stored CGRP and smaller populations stored SP and PACAP. Thus, the innervation of the TMJ is complex and many different ganglia are involved. Accepted: 13 October 1997     

Mitochondrial DNA deletions parallel age-linked decline in rat sensory nerve function.

In rats, the function of sensory nerves in the hind limb declines significantly with age. Normally aging rats and rats treated neonatally with capsaicin were studied here. Quantification of vascular response and substance P in young (3 months) and old (24 months) rats showed additive effects of age and capsaicin treatment. The levels in dorsal root ganglion of a particular deletion in mitochondrial DNA (mtDNA(4834)) were about 300-fold higher in old compared to young rats. Capsaicin treatment had no significant effect on mtDNA(4834) abundance. Dorsal root ganglia of old (but not young) rats were found to contain a spectrum of multiple deletions. The abundance of mtDNA(4834) in dorsal root ganglia from individual rats correlated strongly with their decline in vascular function, even where vascular responses were systematically depressed due to prior capsaicin treatment. One possibility is that mitochondrial DNA mutations directly lead to functional decline at mitochondrial and tissue levels. Alternatively, loss of mitochondrial DNA integrity and physiological decline may be consequences of the same factor, such as oxidative stress.     

Distribution of five peptides,three general neuroendocrine markers,and two synaptic-vesicle-associated proteins in the spinal cord and dorsal root ganglia of the adult and newborn dog: An immunocytochemical study

This study describes the immunocytochemical distribution of five neuropeptides (calcitonin gene-related peptide [CGRP], enkephalin, galanin, somatostatin, and substance P), three neuronal markers (neurofilament triplet proteins, neuron-specific enolase [NSE], and protein gene product 9.5), and two synaptic-vesicle-associated proteins (synapsin I and synaptophysin) in the spinal cord and dorsal root ganglia of adult and newborn dogs. CGRP and substance P were the only peptides detectable at birth in the spinal cord; they were present within a small number of immunoreactive fibers concentrated in laminae I–II. CGRP immunoreactivity was also observed in motoneurons and in dorsal root ganglion cells. In adult animals, all peptides under study were localized to varicose fibers forming rich plexuses within laminae I–III and, to a lesser extent, lamina X and the intermediolateral cell columns. Some dorsal root ganglion neurons were CGRP- and/or substance P-immunoreactive. The other antigens were present in the spinal cord and dorsal root ganglia of both adult and newborn animals, with the exception of NSE, which, at birth, was not detectable in spinal cord neurons. Moreover, synapsin I/synaptophysin immunoreactivity, at birth, was restricted to laminae I–II, while in adult dogs, immunostaining was observed in terminal-like elements throughout the spinal neuropil. These results suggest that in the dog spinal cord and dorsal root ganglia, peptide-containing pathways complete their development during postnatal life, together with the full expression of NSE and synapsin I/synaptophysin immunoreactivities. In adulthood, peptide distribution is similar to that described in other mammals, although a relative absence of immunoreactive cell bodies was observed in the spinal cord.     

Inflammation-induced increase in the density of neuropeptide-immunoreactive nerve endings in rat skeletal muscle

The density of substance P (SP)-, calcitonin gene-related peptide (CGRP)- and vasoactive intestinal polypeptide (VIP)-immunoreactive (ir) nerve endings was quantitatively evaluated in intact and inflamed gastrocnemius-soleus muscle of the rat. In persistently inflamed muscle (12 days after a single injection of Freund’s adjuvant into the muscle), the density of SP-ir fibres was significantly increased. CGRP- and VIP-ir fibres displayed an insignificant increase in density. The density of fibres ir for nerve growth factor (NGF) and for growth-associated protein 43 (GAP-43/B-50), a marker for axonal sprouting, regeneration and synaptic reorganisation, increased significantly in persistently inflamed muscle. The data are consistent with the established contribution of NGF on the expression of SP and GAP-43 in afferent neurones under the influence of a persistent inflammation. Received: 8 September 1997 / Accepted: 12 February 1998     

Substance P, neurofilament, peripherin and SSEA4 immunocytochemistry of human dorsal root ganglion neurons obtained from post-mortem tissue: a quantitative morphometric analysis

Summary Immunocytochemical studies on lumbar dorsal root ganglia obtained at routine postmortem 24–36 h after death were carried out, and neuronal cross-sectional areas measured. The subjects were elderly (76–81 years), of both sexes, had died from heart attack or haemorrhage, and had no clinical evidence of clinical neuropathy or of disease known to be associated with neuropathy. The data were consistent between ganglia from the three subjects. There were striking similarities with data from other species. Two populations of cell profiles with overlapping size distributions were distinguished with an anti-neurofilament antibody, neurofilament-rich (45% of cell profiles) with a large mean area and neurofilament-poor with a smaller mean area. Anti-substance P and anti-peripherin antibodies both labelled a population with a small mean area, with extensive co-localization between them. There were also some differences between these human dorsal root ganglia and dorsal root ganglia from some other species. More neuronal profiles were labelled for substance P in humans (44%) than in rat (20%). More neuronal profiles were labelled for SSEA4 (stage specific embryonic antigen 4) in human (40.5%) than in rat dorsal root ganglia (10%), and the SSEA4-positive profiles were relatively smaller in human than in rat. No selective accumulation of lipofusin in profiles of large cells was apparent. This study also shows that quantitative morphometric analysis of immunocytochemically labelled dorsal root ganglion neuronal profiles can be carried out successfully on human sensory ganglia obtained at post-mortem. This is the first demonstration of the two main subgroups of dorsal root ganglia neurones with neurofilament-rich and poor somata in human tissue. The size distributions of neurons with neurofilament, substance P and peripherin are consistent with these neuronal populations having similar functional properties to those described in other species. From the known sensory and fibre loss with aging, it is speculated that the loss of some large diameter neurones with myelinated fibres and low mechanical thresholds, might account for the high percentage of neurones expressing substance P.     

Substance P-containing primary sensory neurons projecting to the inferior mesenteric ganglion: evidence from combined retrograde tracing and immunohistochemistry

The origin of substance P-containing fibres in the inferior mesenteric ganglion of the guinea-pig was studied by combining retrograde tracing and indirect immunofluorescence techniques. 'True Blue' or propidium iodide was injected into the inferior mesenteric ganglion. Four days later the animals were perfused with formalin and the dorsal root ganglia L2 and L3 were dissected out. Sections of the spinal ganglia were cut on a cryostat and analysed in a fluorescence microscope with appropriate filter combinations and subsequently processed for indirect immunofluorescence using antiserum to substance P. Several ganglion cells containing both dye (True Blue or propidium iodide) and substance P-like immunoreactivity were observed, strongly supporting the view that at least some of the substance P fibres in the inferior mesenteric ganglion are of sensory nature. Taken together with physiological results from other workers, the present findings suggest that substance P can be released from peripheral, sensory nerve endings, inducing changes in the membrane potential of sympathetic neurons.