Sensory Afferent Nerves Research Articles

Modulation of T lymphocyte function by neuropeptides. Evidence for their role as local immunoregulatory elements.

Abstract In mucosa-bearing organs with inherent lymphoid populations, classical modes for control of the immune response may be augmented by products of extrinsic sensory afferent nerve endings which arborize through the lamina propria compartment containing large numbers of T and B lymphocytes. Therefore, we sought to determine the role of neuropeptides (substance P, vasoactive intestinal peptide, and somatostatin) in immune response regulation by using a homogeneous line of T lymphocytes (AO40.1 hybrid), whose activation is driven by a specific Ag (OVA) and where the end point (IL-2 release) could not be contributed to by accessory or other cells. IL-2 was quantitated by the rate of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) metabolism with the use of a murine CD4+ IL-2-dependent T lymphocyte line, and dose-response effects of each neuropeptide were examined over a broad concentration range (10(-14)-10(-6) M) encompassing that regarded as physiologic. Vasoactive intestinal peptide stimulated IL-2 release at low concentrations with a marked effect at 10(-14) M that gradually returned to control levels by 10(-7) M. Somatostatin was associated with a substantial augmentation of AO40.1 T lymphocyte IL-2 release at 10(-10) to 10(-8) M concentrations, whereas substance P demonstrated a stimulatory effect only at high concentrations (10(-9) to 10(-6) M). Concomitant [3H]thymidine uptake studies suggested that changes in cell proliferation or viability did not account for neuropeptide-induced effects in our system. With several exceptions, similar results were found with mitogen (Con A)-stimulated AO40.1 cells and human colonic lamina propria mononuclear cells. It was concluded that the three study neuropeptides, over a broad range of concentrations, have profound stimulatory (and occasionally inhibitory) effects upon the function of a cloned T lymphocyte hybrid cell responding to specific Ag and that these events may reflect those of Ag-driven mucosal T lymphocytes exposed to neuropeptides in vivo.

Calcitonin gene-related peptide mediates capsaicin-induced neuroendocrine responses in rat antrum

Background: Calcitonin gene-related peptide (CGRP) is a 37-amino acid peptide localized to primary sensory afferent nerves in the rat stomach. The actions of CGRP in regulating antral neuroendocrine function were examined in vitro through the use of capsaicin, an agent capable of evoking neuropeptide release from peripheral sensory nerve endings. These results were compared with the effects of exogenous CGRP and CGRP antagonist, CGRP8−37. Methods: Rat antral mucosal/submucosal fragments were incubated in either static or dynamic perifusion experiments. Media were assayed for gastrin, somatostatin, CGRP, and acetylcholine. Results: Capsaicin, like exogenous CGRP, stimulated antral somatostatin release and inhibited both gastrin release and acetylcholine discharge. Low dose capsaicin (1 × 10−5 mol/L) caused significant stimulation of CGRP release: 33 ± 0.2 vs. 14 ± 1 pg/ mL protein; P < 0.001. Tetrodotoxin blocked capsaicin-induced inhibition of acetylcholine release and prevented partially capsaicin-mediated stimulation of CGRP release. The CGRP receptor antagonist CGRP8−37 prevented capsaicin-induced D-cell stimulation and inhibition of G-cell secretion and cholinergic discharge. Conclusions: The effects of capsaicin-induced changes in antral D- and G-cell secretion and acetylcholine discharge are due primarily to release of CGRP. Antral CGRP release from primary sensory afferent nerve terminals may act as a local effector substance to regulate antral neuroendocrine function.

Effects of FR 113680 and FK 224, novel tachykinin receptor antagonists, on cigarette smoke-induced rat tracheal plasma extravasation

We examined the effects of novel tachykinin antagonists, FR 113680 (N α-[N α-(N α-acetyl-L-threonyl)-N 1-formyl-D-tryptophyl]-N-methyl-N-phenylmethyl-L-[phenylalaninamide) and FK 224 (N-[N 2-[N-[N-[N-[2,3-didehydro-N-methyl-N-[N-[3-(2-phenthylphenyl)-propionyl]-L-threonyl]-tyrosyl]-L-leucynyl]-D-phenylalanyl]-L-allo- threonyl]-L-asparaginyl]-L-asparaginyl]-L-serinev-lactone) on rat tracheal plasma extravasation induced by cigarette smoke. Intravenous injection of FK 224 (0.032–3.2 mg kg 1) inhibited rat tracheal plasma extravastion induced by cigarette smoke and capsaicin. FR 113680 (32 mg kg −1 i.v.) also significantly inhibited cigarette smoke-induced plasma extravasation, whereas D-chlorpheniramine maleate, FPL 55712, atropine sulfate and indomethacin had no effect. Tracheal plasma extravasation induced by substance P (SP) and neurokinin A (NKA), but not histamine, was also reduced by intravenous administration of FR 113680 and FK 224. These findings suggest that cigarette smoke stimulates primary afferent sensory nerves, releases tachykinins and evokes plasma extravasation in rat trachea.

Saline diuresis and natriuresis in unanesthetized dogs: a missing atrial factor?

Recent studies in our laboratory indicated that a blunted (40-50%) renal excretory response to isotonic intravenous saline loads occurred in conscious, renal-denervated dogs after 70% of the atrial mass was removed. The blunted responses could not be explained by differences in the responses of arterial pressure, renal nerve activity, or by measured changes of plasma immunoreactive atrial natriuretic peptide (iANP), arginine vasopressin (AVP), plasma renin activity (PRA), or aldosterone (Aldo). The present study was designed to determine whether the central nervous system (CNS) was the source of an unidentified substance, which could account for the blunting of the urine excretory response seen in the atrial-resected dogs. Renal denervation was performed in all dogs to eliminate alterations in efferent renal sympathetic nerve activity derived from reflexes activated during volume expansion. Cardiac denervation (CDX) was used to eliminate sensory cardiac afferent nerve activity to the CNS. A group of five renal-denervated dogs was given an isotonic volume load (400 ml/30 min) before and after complete CDX. Plasma AVP was fixed at normal plasma levels of 3 pg/ml by continuous intravenous infusion. Na and H2O excretion were not different in renal-denervated dogs compared with combined renal and cardiac denervation during the 5 h after the saline load. Plasma AVP and Aldo were unchanged with the volume loads, although PRA rose gradually over the 5 h after the saline loads. Plasma iANP increased transiently in the combined renal and cardiac-denervated state rising from a control of 65-120 pg/ml at the end of the load period.(ABSTRACT TRUNCATED AT 250 WORDS)

Ankle joint (artc. intertarsalis) receptors in the domestic fowl

The physiological responses of joint capsule sensory receptors in the ankle joint of the chicken were studied by recording the electrical activity from single sensory afferent nerve fibres dissected from the parafibular nerve. All units included in this study were sensitive to mechanical stimulation of the joint capsule and were classified with respect to nerve conduction velocity, receptive field size and response threshold. Rapidly adapting mechanoreceptors formed 23% of the receptors present in the sample and responded to the mechanical probe by giving a single response or a short burst of activity at the onset of stimulation. The majority of units identified showed a slowly adapting response and on the basis of conduction velocity were divided into group IV (CV 2.5 m/s), group III (CV 2.5–20 m/s) and those units conducting over 20 m/s. Group IV units had single spot-like receptive fields and mechanical thresholds ranging from 0.6 to 60 g. Group III units could be divided into two groups based on receptive field size. One group had spot-like receptive fields 1–2 mm in diameter, whereas in the other group the receptive fields were larger (over 2×3 mm). The large receptive field units had significantly faster conduction velocities, lower mechanical thresholds and steeper stimulus-response curves than small receptive field units. In response to movement of the joint very few of the receptors, whether rapidly or slowly adapting, were found to be excited by moving the joint by hand in the middle of its physiological range. The physiological properties of these avian mechanoreceptive fine afferent units suggest that while some are activated by normal joint movement and non-noxious local mechanical stimulation of the joint capsule, others have nociceptive functions. The receptors present in the ankle joint were correlated with the anatomical structures found in the avian joints and their physiological properties compared and contrasted to the joint receptors found in mammals.

Studies on the Afferent and Efferent Renal Nerves following Autotransplantation of the Canine Kidney

The presence of both afferent and efferent renal nerves following renal transplantation was investigated in a canine autotransplant model. The efferent postganglionic sympathetic renal nerves were studied using the glyoxylic acid histofluorescence technique to identify renal tissue adrenergic amines (Grade 0-4). The afferent sensory renal nerves were studied by the systemic blood pressure response to renal arterial injection of capsaicin. In 8 control dogs with native innervated kidneys (Group I), intrarenal injection of capsaicin significantly increased the systemic blood pressure from baseline by 32.4 ± 6.3mm. Hg (p < 0.01). This response was equivalent to the blood pressure increase following injection of capsaicin into the mesenteric artery which was 37.3 ± 9.8mm. Hg. The renal tissue histofluorescence grade in this group was 4. Six dogs were studied two to three weeks after autotransplantation of a solitary kidney (Group II). Intrarenal injection of capsaicin did not increase the systemic blood pressure in these animals. Three dogs in this group had no evidence of renal tissue adrenergic amines by histofluorescence (Grade 0); the remaining two animals had renal tissue histofluorescence grades of 1 and 2. Eight dogs were studied 12 to 35 months after autotransplantation of a solitary kidney (Group III). Intrarenal injection of capsaicin in these animals significantly increased the systemic blood pressure from baseline by 10 ± 1.4mm. Hg (p < 0.001). The renal tissue histofluorescence grade in this group ranged from 1 to 3. These data support the presence of both afferent and efferent renal nerves in the kidney at> one year post-transplant.

Neonatal capsaicin-treatment in mice: effects on pancreatic peptidergic nerves and 2-deoxy- d-glucose-induced insulin and glucagon secretion

It is not known whether sensory nerves are involved in the insulin, glucagon or glucose responses to autonomic nerve activation induced by 2-deoxy- d-glucose (2-DG). We therefore treated mice neonatally with capsaicin which permanently destroys sensory afferent nerve fibers. Immunohistochemistry of the pancreas at 13–14 weeks of age revealed a substantial reduction of calcitonin gene-related peptide (CGRP)-immunoreactive nerves and a partial reduction of substance P-immunoreactive nerves. In contrast, no effect was observed on galanin-immunoreactive nerves. At age 10–12 weeks, the mice were injected intravenously with 2-DG (500 mg/kg). In controls, 2-DG stimulated insulin and glucagon secretion and induced hyperglycemia ( P < 0.01). Capsaicin treatment partially reduced the glucose and glucagon responses to 2-DG ( P < 0.01). In contrast, the insulin response to 2-DG was not affected by capsaicin. It is concluded that the mouse pancreas contains capsaicin-sensitive sensory CGRP- and substance P-immunoreactive nerve fibers, whereas the galanin-immunoreactive nerve fibers are not sensitive to capsaicin. Furthermore, capsaicin-sensitive sensory nerve fibers are partially involved in 2-DG-induced glucagon secretion and hyperglycemia, whereas sensory nerves are not involved in 2-DG-induced insulin secretion.

Neuronal control of the kidney: Contribution to hypertension

The renal nerves contribute to hypertension in experimental models of the disease, and appear to play a role in human hypertension. Several lines of evidence indicate that both in spontaneously hypertensive rats and in deoxycorticosterone acetate--NaCl rats, the full development of hypertension is dependent on renal efferent nerves and their induction of excess sodium retention. Renal sensory (afferent nerve) feedback to the central nervous system does not contribute to either of these forms of hypertension. In contrast, renovascular hypertension in rats and aortic coarctation hypertension in dogs are mediated, at least in part, by overactivity of renal afferent nerves and a resultant increase in systemic sympathetic nervous system activity. These forms of hypertension are not associated with sodium retention, and selective sensory denervation of renal afferent nerves by dorsal rhizotomy and total renal denervation result in similar reductions in hypertension. Surprisingly, the renal nerves do not contribute to dietary NaCl exacerbated hypertension in the spontaneously hypertensive rat, dietary NaCl-induced hypertension in the Dahl NaCl-sensitive rat, or the chronic hypertensive and nephrotoxic effects of cyclosporine A therapy in the rat, despite the finding that in all three forms of hypertension, overactivity of the sympathetic nervous system is prominent. Clinical studies indicate that the renal afferent and efferent nerves contribute to hypertension of different etiologies. Together these data point to the complex role that the renal nerves likely play in human essential hypertension.

Neurogenic component of different models of acute inflammation in the rat knee joint.

This study was performed to investigate whether different models of acute joint inflammation showed a neurogenic component and to establish whether this is mediated through sensory afferent or sympathetic efferent nerve fibres. Intra-articular injection of 2% carrageenan, 20 micrograms substance P, 1% formalin, and 2% urate all produced an inflammatory response. Prior surgical denervation of the joint significantly inhibited this response in the carrageenan and formalin models, but not the others. Pretreatment of the joint with 1% capsaicin (about one week previously) significantly reduced the inflammatory response in all models except formalin. In animals pretreated long term with reserpine (to deplete sympathetic nerve endings of their neurotransmitters) significant reductions occurred in the inflammatory responses to substance P and urate. Intraarticular injection of compound 48/80 produced a marked inflammatory response, which was only significantly reduced by capsaicin pretreatment. These results suggest that both the formalin and carrageenan models of inflammation depend to some extent on the integrity of the sensory innervation of the joint, and thus have a neurogenically mediated component to the inflammatory process they generate. In these models there seems to be little contribution from sympathetic efferent fibres. Each model of inflammation showed a different pattern of response to the pretreatments, suggesting that the mediators of the inflammatory process may differ in each case.

Inhibition of platelet aggregation by capsaicin. An effect unrelated to actions on sensory afferent neurons

The effects of capsaicin on the ability of platelets to aggregate in response to thrombin, platelet-activating factor or calcium ionophore (A23187) were examined. At concentrations previously shown to activate sensory afferent neurons, capsaicin markedly inhibited the responsiveness of platelets to the three agonists. The effects of capsaicin on platelet aggregation were reversible. and could be observed if capsaicin was added after platelets had begun to aggregate in response to the agonist. Capsaicin did not affect the shape change which occurs in response to the agonists, a process which is calcium-independent. These results demonstrate that capsaicin, at concentrations which are frequently used to ‘selectively’ activate sensory afferent neurons, is also capable of affecting the function of the platelet. Such non-specific effects of capsaicin must be considered when this substance is used as a pharmacological probe of sensory afferent nerve function.

PGI 2-induced activation and sensitization of articular mechanonociceptors

The effects of PGE 2, PGI 2 and the stable PGI 2 analogue cicaprost on the afferent discharge of ankle joint mechanonociceptors were studied in the anaesthetized rat. Close-arterial injection of PGI 2 (0.01–0.1 μg) or cicaprost (0.05–5 μg) caused both sensitization to mechanical stimulation and excitation of the majority of mechanonociceptors, whereas PGE 2 (0.03–3 μg) had only weak effects on a small number of nociceptive units. These results suggest the existence of specific PGI 2 sensitive receptors (IP receptors) on rat sensory afferent nerves, and support the hypothesis that in the rat endogenous PGI 2 plays an important role in the lowering of nociceptive thresholds in inflamed joints.

Response Properties of Periodontal Mechanosensitive Neurons in the Trigeminal Spinal Tract Nucleus of the Cat

Periodontal mechanosensitive (PM) units were recorded from the trigeminal spinal tract nucleus (Vst) of the cat. The Vst is divided into three subnuclei: oralis (Vo), interpolaris (Vi), and caudalis (Vc). The receptive fields of PM units in Vo and Vi were arranged in a dorsoventral sequence in the mandibular to maxillary divisions, and those in Vc were arranged in a mediolateral sequence. The majority of Vo units were single-tooth ones, whereas more than half the Vi units and all the Vc ones were multitooth units. The PM units in each subnucleus were predominantly responsive to canine tooth stimulation. Most of the PM units in Vo and Vi gave sustained responses to pressure applied to the tooth, were directionally selective, and were most actively excited by canine tooth stimulation in the caudomedial or rostrolateral direction. Vc units, however, were transient. The threshold intensity for firings by canine tooth stimulation was less than 0.05 N. These findings indicate that only the response properties of PM units in the rostral part of Vst resemble those of the trigeminal main sensory nucleus neurons and primary afferent nerves.

Opioid-induced respiratory depression and analgesia may be mediated by different subreceptors.

Use of selective delta opioid antagonists provide evidence that the delta receptor within the brain seems an integrated part in the mediation of respiratory depression induced by a potent analgesic like fentanyl. Low doses of the delta antagonists RX-8008M (3-6 micrograms/kg) as well as ICI 174,864 (3-6 micrograms/kg) reversed fentanyl-related respiratory depression (arterial blood gases) in the unanesthetized canine. Opioid-induced blockade of afferent sensory nerve volleys (amplitude height of the somatosensory-evoked potential) could be reversed only by a high dose (9 micrograms/kg) of RX-8008M. Depression of amplitude height of the SEP could not be reversed by ICI 174,864 over the whole dose range (3-6-9 micrograms/kg). In comparison, naloxone (1-5-10 micrograms/kg) not only reversed depression of PaO2, it also reversed the blockade of afferent sensory nerve impulses in the low (5-micrograms/kg)-dose range. A highly selective delta antagonist may have a therapeutic value in reversing opioid-related respiratory depression, resulting in little or no attenuation of analgesia.

Mechanism of intragastric nicotine protection against ethanol-induced gastric injury

To elucidate the mechanism of intragastric nicotine protection against ethanol-induced gastric mucosal injury seen in a previous report and in our preliminary study, the following studies were performed. Rats were pretreated with naloxone (8 mg/kg intraperitoneal, 0.5 hr prior to study) to block opiate receptors; or capsaicin (125 mg/kg subcutaneous 10 days prior to study) to denervate the afferent sensory fibers; or indomethacin (2.5 mg/kg intragastric or 5 mg/kg subcutaneous, 1 hr prior to study) to inhibit endogenous prostaglandin synthesis. At 1-hr intervals, nicotine (4 mg/kg) or vehicle and 40% ethanol were then given intragastrically. Total gastric corpus mucosal lesion length was measured unbiasedly. In separate studies, gastric mucosal blood flow (GMBF) was assessed by hydrogen gas clearance before and after intragastric nicotine or vehicle; luminal mucus volume, gastric juice volume, and acid output were measured 1 hr after either intragastric nicotine or vehicle administration. The results showed that the acute protective effect of intragastric nicotine was associated with a significantly larger luminal mucus volume. It was not blocked by naloxone, capsaicin, or indomethacin. There was no increase in GMBF. The larger gastric residual volume did not account for the protection. We conclude that the mechanism mediating nicotine protection is unique and is independent of opiate receptors, capsaicin-sensitive afferent sensory nerve fibers, endogenous prostaglandin generation, or dilution of the injurious agent. The increase in luminal gastric mucus volume may contribute to the protective effect of intragastric nicotine against gastric mucosal injury produced by 40% ethanol.

Evidence that calcitonin gene-related peptide contributes to the capsaicin-induced relaxation of guinea pig cerebral arteries

Pretreatment with capsaicin caused a depletion of substance P (SP)-, neurokinin A (NKA)- and calcitonin gene-related peptide (CGRP)-like immunoreactivity (-LI) from the trigeminal ganglion, dura mater and cerebral arteries. The effect of capsaicin on isolated basilar arteries of guinea pig resulted in a biphasic relaxant response of histamine precontracted vessels. The first phase of the capsaicin-induced relaxation was absent in capsaicin-treated guinea pigs. Furthermore, repeated administration of capsaicin decreased the first but not the second phase of relaxation, supporting the view that a stored agent was released, while the second phase probably was due to a direct effect of capsaicin per se. The biphasic effect was not modified by the SP antagonist Spantide in a concentration that blocks tachykinin response (3 · 10 −6 M), nor by removal of the endothelium. There was no significant difference in pD 2 values (− log concentration eliciting half maximum relaxation) between relaxations induced by SP, NKA, neurokinin B, neuropeptide K or CGRP in capsaicin pretreated as compared to vehicle-treated animals. These results are in support of the assumption that CGRP is involved in the capsaicin-induced relaxation caused by release of vasoactive agents from sensory afferent nerves.

Serotonin modulates substance P-induced plasma extravasation and vasodilatation in rat skin by an action through capsaicin-sensitive primary afferent nerves

Using a blister model of inflammation in the rat hind footpad, the present study was undertaken to examine the ability of serotonin (5-HT) to modulate an inflammatory reaction manifested as plasma extravasation and vasolidation induced by the neuropeptide substance P (SP). In addition, the role of primary afferent sensory nerve fibres in these modulatory effects was studied in capsaicin pretreated rats. Using a protocol of simultaneous perfusion of amine and peptide over the blister base, no major modulatory effect was observed. On the other hand, using a protocol of sequential perfusion, 5-HT was found to extend the plasma extravasation and vasodilatation responses to SP. 5-HT maintained the plasma extravasation response to SP after cessation of stimulation (during the post-stimulation period). On the other hand, it extended the vasodilatation response to SP during the actual stimulation period by preventing the occurrence of tachyphylaxis. These modulatory effects were absent in capsaicin-pretreated rats. The present study provides evidence for the first time in vivo to suggest that serotonin can modulate an inflammatory response to SP via a mechanism that involves capsaicin-sensitive sensory fibres.

Modulation of a peripheral inflammatory response to substance P by locally administered opioid receptor agonists

Using a blister model of inflammation in the rat hind footpad, the present study was undertaken to examine possible peripheral effects of specific μ (DAGO) and δ (DSLET) opioid receptor agonists on an inflammatory response induced by substance P, the putative mediator of neurogenic inflammation. When perfused over the blister base, SP induced both plasma extravasation and vasodilatation responses. These responses were significantly inhibited in the presence of either opioid receptor agonist in a naloxone reversible manner. DSLET inhibited SP responses in a dose dependent manner and was 100 times more potent than DAGO. The role of primary afferent sensory nerve terminals in these modulatory effects was investigated in rats pretreated as neonates with capsaicin. The abilityof DAGO and DSLET to inhibit the inflammatory response in these rats was significantly less than that in controls. The data raises the possibility that the inhibitory effect of the opioid receptor agonists on the inflammatory response might reflect a role for opioids in modulating tachyphylaxis to SP.

Capsaicin selectively reduces airway responses to histamine, substance P and vagal stimulation

Airway responsiveness to histamine, substance P, methacholine and bilateral electrical vagal stimulation was assessed in capsaicin-treated and control guinea pigs. In animals treated with capsaicin (50 mg/kg s.c.) 7 days before experiments, airway responsiveness to histamine, substance P and vagal stimulation was significantly reduced but responsiveness to methacholine was unchanged. The findings suggest that histamine and substance P, but not methacholine, require capsaicin-sensitive sensory afferent nerves for part of their actions.

Cutaneous sensory afferents recorded from the nervus intramandibularis of Gallus gallus var domesticus.

The responses of single sensory afferent nerve fibres were recorded from small nerve bundles of the intramandibular nerve of the chicken following thermal and mechanical stimulation of the beak. Thermoreceptors, nociceptors and mechanoreceptors were identified and their responses characterized. Of the thermoreceptors identified 11 units were classified as cold receptors, which responded to cooling the receptive field by increasing the discharge rate and had conduction velocities in the range 0.83 to 4.4 m/s. Only one warm unit was identified. Two classes of nociceptors were identified: mechano-thermal (polymodal) nociceptors and high threshold mechanical nociceptors. The discharge characteristics and stimulus-response curves of both types were described. While the mechano-thermal nociceptors were exclusively C-fibres (c.v. 0.4 to 1.86 m/s), the high threshold mechanoreceptors contained both C and A delta fibres (c.v. 1 to 5.5 m/s). Thermal response thresholds for the mechano-thermal units ranged from 41 to 50 degrees C with mechanical thresholds of 2 to over 50 g. Mechanical thresholds for the high threshold units ranged from 5 to over 50 g. The mechanoreceptors were either slowly or rapidly adapting. The pattern of response together with stimulus-response curves were presented for the slowly adapting units. Conduction velocities of the slowly adapting units varied from 0.7 to 20 m/s and mechanical threshold from 0.1 to 2 g. On the basis of their response to a vibrating, and a ramp-and-hold mechanical stimulus, the rapidly adapting units were divided into Herbst and Grandry units with only the Herbst units responding accurately to the vibrating stimulus. Both units had fibres conducting in the 50 m/s range with thresholds in the 0.1 to 10 g range. The results are discussed in relation to the receptors found in other avian species and mammalian peripheral sensory afferents.

Neuropeptides and Immunity

Our studies have clearly shown that neuropeptides have a profound effect on immunoglobulin synthesis both in vivo and in vitro. The effects varied according to the neuropeptide added or the tissue from which the lymphocytes were obtained. Substance P caused the most pronounced enhancement of both functions, especially in Peyer's patch cells, where it selectively increased IgA synthesis. Somatostatin was inhibitory, and the effect of vasoactive intestinal peptide varied according to the source of the cells. We have previously shown that neuropeptides also cause mast cell secretion and that only substance P was effective in this regard on intestinal mucosal mast cells. Therefore, we looked for microanatomic relationships between peptidergic nerves and immune effector cells. Mast cells appear to have structural associations with neuropeptides-containing nerves in the intestine. Nerve growth factor, known to promote the growth of sensory afferent and sympathetic nerves, has significant direct effects on mast cells. In vitro, this substance caused enhanced antigen mediated histamine release and, in vivo, extensive mast cell hyperplasia. Also, in humans, we were able to produce increased numbers of mast cell/basophil colonies from peripheral blood in the presence of nerve growth factor.