Calcitonin gene-related peptide (CGRP)-immunoreactive nerves in the tracheal epithelium of rats: an immunohistochemical study by means of whole mount preparations.

The airway mucosa is known to be densely supplied with several types of peptidergic nerve fibers. The distribution of the nerve fibers in a large extent of the mucosa, however, has been difficult to visualize by observation of conventional thin sections. In the present study, whole mount preparations of the rat tracheal mucosa were designed to demonstrate calcitonin gene-related peptide (CGRP)-containing nerves which are most predominant among peptidergic nerves in the trachea. The mucosal sheet of the trachea was separated from the cartilaginous base by means of dispase, a protease, to be processed to immunohistochemistry for CGRP. In the cartilaginous portion, thick nerve bundles immunoreactive for CGRP ran transversely between the cartilage rings, sending terminal branches toward the epithelium. In the membranous portion, immunoreactive nerve bundles were thinner than those in the cartilaginous portion and ran longitudinally. A very dense nerve plexus of CGRP-immunoreactive fibers was demonstrated to extend in the basal part of the epithelium; every epithelial cell appeared to contact more than one nerve fiber. Immunohistochemistry of the whole mount preparations also clearly demonstrated the distribution and entire shape of broncho-pulmonary paraneurons scattered in the epithelium. Ultrastructurally, the CGRP-immunoreactive intraepithelial nerves were found to lack myelin and Schwann sheaths, and to run through the bases of the epithelial cells; they were most frequently surrounded by the cytoplasmic processes of the epithelial cells. The present study demonstrates that immunohistochemistry of whole mount preparations is a useful tool to morphologically analyze neuronal and paraneuronal distribution throughout the tracheal epithelium.     

Co-localization of calcitonin gene-related peptide- and substance P-like immunoreactivity in mucosal intra-epithelial nerves in the toad colon.

Mucosal intra-epithelial nerve terminals have been found in the colon of the toad, Bufo marinus. Co-localized in the fibres are calcitonin gene-related peptide (CGRP)-like immunoreactivity (LI) and substance P (SP)-LI. The intraepithelial nerves are likely to be terminals of primary afferent sensory neurones. Fibres with vasoactive intestinal peptide (VIP)-LI or neuropeptide Y (NPY)-LI also supply the mucosa but do not penetrate the epithelium.     

Calcitonin gene-related peptide (CGRP)-immunoreactive nerve terminals in the whole mount preparations of the dog urethra.

To visualize the entire shape of the intraepithelial nerve fibers, whole mount preparations of the dog urethra were produced and immunostained with an antiserum against CGRP, one of the predominant substances contained in the nerves. The immunoreactive nerves in the lamina propria were smooth (non-beaded) in appearance and weak in immunoreaction. Within the epithelium, they displayed typical beaded profiles and were intense in immunoreaction. The intraepithelial fibers branched and wound into an extensive network with wide meshes ("reticular terminal"). The bead-like swellings included large ones resembling Herring bodies in hypophyseal neurosecretory fibers. Another type of nerve terminal, consisting of fine and weakly immunopositive fibers, was also found in the epithelium. These branched in dendritic or in dense bouquet-like fashion, occupying smaller areas ("bouquet-like terminals). Vesicular swellings often characterized these terminals, though they were smaller and more uniform in size and far less in their amount of immunoreactive substance than were the swellings in the reticular terminals. Both types of nerve terminals originated from the same nerve trunk. The connection between the reticular and bouquet-like terminals, which may presumably represent secretory and receptive parts, respectively, morphologically supports the possible occurrence of an axon reflex in the urethral CGRP neurons. Our whole mount preparations, when doubly stained with CGRP and serotonin antibodies, further revealed the CGRP-positive reticular terminals being closely associated with serotonin- or CGRP-immunoreactive paraneurons dispersed in the epithelium.     

An immunohistochemical study on neurons and paraneurons of the pre- and post-natal chicken lung.

The indirect immunoperoxidase (PAP) method was used on chicken lung specimens from embryos ranging in age from 6 days to hatching, chicks and adult chickens of up to 6 months. The ontogenesis and distribution of neurons and paraneurons containing immunoreactivities for serotonin (5HT), bombesin, vasoactive intestinal polypeptide (VIP), substance P (SP) and galanin were investigated. Serotonin-immunoreactive paraneurons were first detected in the pulmonary mesenchyma of 8-day-old embryos, while in the 12-day-old embryos the following neurons and paraneurons were first detected in their respective locations: serotonin-immunoreactive paraneurons in the bronchial epithelium; VIP- and galanin-immunoreactive ganglionic cells and SP-immunoreactive nerve fibres in the intrapulmonary ganglia. At hatching, serotonin-immunoreactive paraneurons in the epithelium of the air capillaries and air sacs, and bombesin-immunoreactive paraneurons in the epithelium of the primary bronchus, VIP-, galanin- and SP-immunoreactive nerve fibres in the lamina propria of the primary and secondary bronchi and in the pulmonary septa could also be shown. Some serotonin-immunoreactive small paraneurons were also found in the intrapulmonary ganglia. In the adult specimens, VIP-, galanin- and SP-immunoreactive nerve fibre networks were observed throughout the primary bronchus wall and in the lung septa. In intrapulmonary ganglia, VIP- and galanin-immunoreactive neurons and serotonin-immunoreactive small paraneurons could be more numerously demonstrated. Moreover, bombesin paraneurons occurred in the epithelium of primary and secondary bronchi, and serotonin-immunoreactive paraneurons were found in the epithelia of the bronchi and air sacs and in some pluricellular bodies in the lamina propria of the air sac ostia.     

Neurogenic inflammation in the rat trachea. II. Identity and distribution of nerves mediating the increase in vascular permeability

Summary This study addresses the question of whether increased vascular permeability, which is a prominent feature of neurogenic inflammation in the respiratory tract, is mediated by sensory axons that end near venules in the airway mucosa. In these experiments, neurogenic inflammation was produced in the tracheal and bronchial mucosa of atropine-treated Long-Evans rats by electrical stimulation of the left or right superior laryngeal nerve and/or cervical vagus nerve. The particulate tracer Monastral blue was injected intravenously to localize the sites of increased vascular permeability, and microspectrophotometry was used to measure the amount of extravasated Monastral blue in the trachea and thereby quantify the increase in vascular permeability. In some rats, selective denervations were made to locate the cell bodies of neurons that mediate the increase in vascular permeability; in others, fluorescence immunohistochemistry and quantitative electron microscopic methods were used to determine which structures in the tracheal mucosa are innervated by these neurons. The study revealed that the vagally mediated increase in vascular permeability was sudden, transient (half-life=2.4 min) and restricted to venules. Stimulation of the left or right superior laryngeal nerve increased the permeability of venules in the extrathoracic trachea, whereas stimulation of either vagus nerve increased vascular permeability in the intrathoracic trachea and bronchi. All nerves had bilateral effects in the trachea, but the vagus nerves had largely unilateral effects in the bronchi. Neurons that mediated the increase in venular permeability had their cell bodies in the jugular (superior sensory) ganglion of the vagus nerve or rostral portion of the nodose (inferior sensory) ganglion. Preganglionic autonomic vagal neurons in the brain stem were not essential for this increase in venular permeability. Few nerves identifiable by substance P-immunohistochemistry or electron microscopy were located near the affected venules, and no nerves were within 1 m of the walls of venules. However, the epithelium and arterioles of the airway mucosa were densely innervated. All intraepithelial nerves were within 0.1 m of epithelial cells, and at least two-thirds of nerves near arterioles were within 1 m of the vessel walls. We conclude that the increase in venular permeability associated with neurogenic inflammation in the trachea and bronchi of rats is mediated by sensory axons that travel in the vagus nerves and superior laryngeal nerves. We question whether tachykinins from the sensory nerves mediate the increase in vascular permeability through a direct action on venules, and raise the possibility that these nerves evoke the release from epithelial cells of mediators that contribute to the increase in vascular permeability.     

Transitional epithelial zone of the bovine nasal mucosa

To determine the extent and ultrastructure of epithelium lining the transitional nasal mucosa of the neonate, gnotobiotic calf tissues were prepared for scanning and transmission electron microscopy. Stratified cuboid epithelium of the rostral 40% of the nasal cavity contained few ciliated cells; the next caudal 10–15%, although ciliated, had extensive nonciliated areas. The predominant type of surface cell was nonciliated, had short microvilli, and contained a multilobate nucleus and numerous pinocytotic vesicles. In some areas the surface of these cells presented a cobblestone appearance. Basal cells contained numerous bundles of filaments, ribosomes, and basal vesicles. Caudally, nonciliated columnar cells included a cell type similar to the more rostral cuboid cell, as well as brush cells and immature secretory and ciliated cells. Goblet cells were infrequently observed. Intraepithelial nerve terminals were abundant. Other intraepithelial cells, often difficult to identify owing to varying characteristics, included lymphocytes. Based upon comparisons of this neonatal epithelium with mature epithelium, observed in earlier studies of other mammalian species, the transitional mucosa is believed normally to occupy an extensive area of the nasal cavity.     

Taste cells in the gut and on the tongue. Their common, paraneuronal features

Chemoreception of foodstuff in the gut is performed by endocrine cells dispersed in the gut epithelium. They are bipolar cells extending an apical process to the gut lumen and releasing their messenger substances from the cell base in response to the apical stimuli. The cells share cell-biological features with neurons and are classified as paraneurons. Noteworthy, the gut paraneurons do not seem to be stimulated by monosodium glutamate (MSG) as our dog experiment indicates. Administration of 50 mM MSG into the duodenal loop of anesthetized dogs did not cause changes in the volume and protein output in the pancreatic juice. The gustatory cells in the taste bud show essentially the same structure and function as the gut endocrine cells. A single gustatory cell type (type III) seems to receive different chemical stimuli, whereas different endocrine cell types in the gut react to different stimuli. The gustatory cells possess numerous peptidergic-type granules besides synapse-associated small clear vesicles. The former granules, in the guinea pig and dog, are abundant in the cytoplasm, giving an endocrine-like appearance to the cell. Peptidic signal substances contained in the granules remain to be identified. Comparison of the gustatory cells in the taste bud with the endocrine cells in the gut and with other paraneurons may put into front certain hitherto unexplored structures and functions of the cells.     

Human gallbladder mucosa infrastructure: Evidence of intraepithelial nerve structures

The zinc iodide-osmium tetroxide (ZIO) fixative-staining method was used along with topographical ultrastructure to investigate cholecystectomized human gallbladders under light and electron microscopic techniques. This method delineated neuronal structures which may be involved in controlling the functions of the gallbladder epithelium. Three epithelial cell types were described in the surface epithelium: (1) Co lumnar clear cells; (2) dark, tuft osmiophilic cells; and (3) basal clear cells with electron-dense granules and showing intense ZIOphilic staining properties. While mucous granules were delineated in the first two cell types, the columnar epithelial clear cells are of uncertain function(s) and content but are probably absorptive cells. The small basal clear cells displaying intense ZIOphilia are associated with intraepithelial nerve endings. These nerve structures may have a sensory and/or motor function(s); they were detected throughout the gallbladder epithelial lining and mucosa.     

Proposal for the existence of a nasogastric reflex in humans, as a potential cause of upper gastrointestinal symptoms

The nose is a gateway of air from the environment to the body and with its rich innervation from the olfactory and trigeminal nerves plays a critical role as a sensor in both human beings and primitive animals. Irritation of the nasal or paranasal mucosa may initiate a severe bradycardia, apnea, and vasoconstriction and increase the pulmonary airflow resistance. However, the interaction between nasal mucosa and the upper gastrointestinal tract is more often than not neglected in the clinical literature. We propose that a nasogastric reflex might exist with its afferent and efferent loops being the trigeminal and vagus nerves, respectively. The central connection of these loops is located at the pontomedullary level. The sensory inputs from the nasal mucosa to the general somatic afferent component of the brainstem including the pontine and medullary trigeminal nucleuses may induce the neighboring nucleus of the solitary tract and dorsal motor nucleus of the vagus. This initiates, via the efferent fibers of the vagus nerve, the manifestations of the vagal stimulation. The presence of a nasogastric reflex may warrant considerations as diseases of nose and paranasal sinuses may be the cause upper gastrointestinal symptmatology.     

Histogenesis of the mucosa of the descending colon in mouse fetuses

The present study clarify intraepithelial events occurring during the formation of intestinal crypts in the descending colon. In early stage, the mucosa is lined with a simple columnar epithelium consisting of undifferentiated cells which are mutually joined by the junctional complex surrounding the intestinal lumen. The epithelium becomes thickened and stratified, and secondary junctional complexes are formed in the abluminal part of the epithelium. An intracellular cavity opens to the focal tight junction of the secondary junctional complex to form an intraepithelial (intercellular) cavity surrounded by the tight junction. The intraepithelial cavity grows by the fusion of intracellular cavities and vesicles as well as by the division of surrounding cells. Intraepithelial cavities fuse with each other and with the main intestinal lumen. This, together with mesenchymal invasion, results in the formation of intestinal crypts lined with a simple columnar epithelium, where various cell types begin to differentiate. These processes seem essentially similar to those occurring in the developing small intestine except that villi, not crypts, are formed in the small intestine. It is suggested that the formation site and elongation of the tight junction provide for the location and fusion of intraepithelial cavities and eventually result in the formation of villi or crypts, alternatively.     

Electron microscope observations on the formation of primitive villi in rat small intestine with special reference to intercellular junctions.

Formation and fusion of intraepithelial cavities have long been considered an essential process in the histogenesis of the intestinal mucosa. By electron microscope observation of thin sections and freeze-fracture replicas of the small intestine of rat fetuses, we first demonstrated the initial steps in the formation of intraepithelial cavities: A focal tight junction (macula occludens) was formed in the abluminal part of the epithelium, after which the membrane of an intracellular cavity was fused with that of the focal tight junction to form an intercellular (intraepithelial) cavity enclosed by a zonula occludens. The present study also revealed that gap junctions appeared and enlarged simultaneously with the formation of primitive villi and differentiation of absorptive cells. These gap junctions gradually came to be confined in the epithelium of intervillous regions where proliferation and differentiation of epithelial cells took place. Absorptive cells in villi rarely had gap junctions. These results suggest that tight and gap junctions play important roles in the histogenesis of the intestinal mucosa, and in the proliferation and differentiation of epithelial cells.     

Nasal sensory nerve populations responding to histamine and capsaicin

BACKGROUND: Inflammation of the nasal mucosa leads to sneezing, nasal itch, rhinorrhea, and nasal blockage. Many of these symptoms are likely the result of nasal trigeminal sensory nerve stimulation by inflammatory mediators. Nasal challenge with the C-fiber stimulant capsaicin causes a different set of symptoms than those evoked by histamine, suggesting that these 2 stimuli may activate separate subpopulations of nasal sensory nerves. OBJECTIVE: To investigate the trigeminal sensory nerves innervating the guinea pig nasal mucosa and to address specifically the hypothesis that histamine and capsaicin activate distinct subgroups of these nerves. METHODS: Guinea pig trigeminal neurons (retrogradely labeled from the nasal mucosa) were assessed for their responses to histamine and capsaicin by studying changes in the intracellular free calcium concentration, and assessed for substance P immunoreactivity. RESULTS: Only 60% of the nasal-specific trigeminal sensory neurons were found to be capsaicin-sensitive. Histamine stimulated only a subset (<40%) of these capsaicin-sensitive neurons. No nasal-specific capsaicin-insensitive neurons responded to histamine, although about 10% of trigeminal ganglion neurons per se responded to histamine but not capsaicin. Immunohistochemical analysis revealed that most (about 60%) of the sensory neurons innervating the nasal mucosa did not express the neuropeptide substance P, including nearly all large-diameter neurons, but also a significant number of small-diameter neurons (presumably C-fiber neurons). CONCLUSION: Nasal neurons are not homogenous with respect to chemosensitivity or substance P content. It is likely that this heterogeneity in nasal afferent nerves underlies the differences in nasal responses to specific inflammatory mediators associated with the allergic reaction.     

Pulmonary neuroepithelial bodies are innervated by vagal afferent nerves: an investigation with in vivo anterograde DiI tracing and confocal microscopy

 The pulmonary airway and alveolar epithelia contain distinctly innervated clusters of basally granulated cells: the neuroepithelial bodies. In the past, morphological criteria and the results of selective vagotomy have led to the interpretation that their innervation is sensory. Consequently, they are regarded as receptor organs. As a further test of this hypothesis, the present investigation set out to label vagal sensory nerve fibres to the lungs by anterograde neural tracing, and to establish the relationship between these fibres and the neuroepithelial bodies. A fluorescent neural tracer was injected unilaterally into the left or right nodose ganglion of adult rats. After suitable survival times, thick frozen sections of lung tissue were studied with laser scan confocal microscopy. Sensory nerve fibres were seen to run in the airway walls and occasionally penetrated the epithelium, where they formed complex terminals. The resulting intraepithelial sensory end organs showed a close morphological resemblance to the neuroepithelial bodies. Subsequently, electron microscopic investigation of such identified structures revealed the typical ultrastructural characteristics of neuroepithelial bodies: corpuscular cells containing dense cored secretory vesicles and contacted by mitochondria-rich nerve endings. We conclude that anterograde tracing of sensory nerves from the nodose ganglion confirms the receptor nature of the pulmonary neuroepithelial bodies, which may correspond to a subpopulation of the irritant and C-fibre receptors. Accepted: 14 October 1997     

Distribution of SP- and CGRP-like immunoreactive nerve fibers in the lower respiratory tract of neonatal foals: evidence for loss during development

The lungs of neonatal foals contain many nerves immunoreactive for substance P and calcitonin gene-related peptide. These nerves are closely associated with the epithelium, bronchial and pulmonary vessels and the airway smooth muscle of all intrathoracic airways, including non-cartilaginous bronchioles. Activation of sensory nerves in the respiratory epithelium could thus potentially affect, via local axon reflexes, vascular and respiratory smooth muscle in neonatal equine airways. Nerves immunoreactive for these peptides are much more widely distributed within the lung than in adult horses; they may thus play a trophic role before birth, or contribute to the post-natal adaptation to breathing.     

Calbindin D28k-immunoreactive afferent nerve endings in the laryngeal mucosa

The distribution of the calbindin D28k in the laryngeal sensory structures was studied by immunohistochemistry, immunoelectronmicroscopy, and double immunofluorescence with calretinin-immunoreactivity. Moreover, origin of the nerve endings were observed using retrograde tracer, fast blue. Immunoreactivity for calbindin D28k was found in the various types of nerve endings in the larynx, namely, laminar nerve endings, nerve endings associated with the taste buds, intraepithelial nerve endings, and endocrine cells. The laminar endings with calbindin D28k-immunoreactivity were observed in the subepithelial connective tissue. In some endings, terminals were expanded. The laminar endings were also observed in the perichondrium of the epiglottic cartilage. In the epiglottic and arytenoid epithelia, thick nerve fibers with calbindin D28k-immunoreactivity ascending to taste buds and intragemmal nerve fibers were also observed. Within the epithelial layer, intraepithelial free nerve endings with calbindin D28k-immunoreactivity were observed. Furthermore, diffuse endocrine cells were observed within the laryngeal epithelium. By immunoelectron microscopy, immunoreaction products in the endings mentioned above were localized in the cytoplasm of the axon terminals and nerve fibers which contained with numerous mitochondria. Out of the 100 laminar endings, 18 endings were immunopositive for both calbindin D28k and calretinin, 33 were positive for calbindin D28k but negative for calretinin, and 49 were positive for only calretinin in the double immunofluorescence microscopy. The nerve fibers associated with the taste buds and the free nerve endings, which immunostained for calbindin D28k, were not stained with antibody against calretinin. After injection of the fast blue in the laryngeal mucosa, fast blue-labeled cells were mainly observed in the nodose ganglia. Of the total number of labeled cell in the nodose and dorsal root ganglia at the level C1 to Th2, 65.1% occurred in nodose ganglia (572/879, n = 6). In the nodose ganglia, 79.7% of labeled cells (456/572) were immunoreacted for calbindin D28k. The distribution of calbindin D28k-immunoreactivity may be differnt from that of calretinin. It is suggested that calbindin D28k have regulatory role on intracellular calcium concentration in the laryngeal sensory corpuscles.     

GRP (gastrin-releasing peptide)-containing nerves in the rat stomach and stress-induced depletion of their synaptic vesicles. An electron microscope study.

Gastrin-releasing peptide (GRP)-containing nerves are extremely numerous in the mucosa of the gastric body. Our previous study demonstrated that depletion of GRP from the nerves occurred in close relation to ulceration in the stomach. The present study deals with the ultrastructure of the GRP-immunoreactive nerves under normal conditions and its changes induced by conditions of stress. The immunoreactivity for GRP was recognized selectively in large cored vesicles in the swollen axoplasm of the nerves. The same axoplasm further revealed immunonegative small clear vesicles which were believed to contain acetylcholine. In materials where the GRP-immunoreactive nerves markedly decreased in number, both large and small synaptic vesicles were depleted from the nerves. These findings suggest that GRP and acetylcholine coexist in single nerves in the oxyntic mucosa, and that by nerve stimulation, they are coreleased into the lamina propria.     

Histological changes in radial forearm skin flaps in the oral cavity

We reported previously that skin flaps transplanted to the oral cavity in reconstructive surgery for oral cancer frequently acquired the gross appearance of buccal mucosa. The changes were shown to be reactive in nature. The "changed" flaps generally had a heavier infiltration of leukocytes in the dermis and appeared to have thicker epithelium. The present study quantifies these parameters, as well as the numbers of intraepithelial leukocytes. The flaps that had acquired the gross appearance of oral mucosa had significantly thicker epithelium, larger numbers of dermal leukocytes, and more intraepidermal inflammatory cells per unit length than flaps that retained the gross appearance of thin skin. No correlation was found between these changes and radiotherapy.     

Immunolocalization of neurotransmitter-synthesizing enzymes and neuropeptides with associated receptors in the photophores of the hatchetfish, Argyropelecus hemigymnus Cocco, 1829

Anatomical and functional studies of the autonomic innervation of the photophores of luminescent fishes are scarce. The present immunohistochemical study demonstrated the presence of nerve fibers in the luminous epithelium and lens epithelium of the photophores of the hatchet fish, Argyropelecus hemigymnus and identified the immunoreactive elements of this innervation. Phenylethanolanine N-methyltransferase (PNMT) and catecholamine (CA)-synthesizing enzymes were detected in nerve varicosities inside the two epithelia. Neuropeptides were localized in neuropeptide Y (NPY) and substance P (SP)- and its NK11 receptor-immunopositive nerves in the lens epithelium. Neuropeptides were also localized in non-neural cell types such as the lens cells, which displayed immunoreactivities for pituitary adenylate cyclase activating peptide (PACAP) and their receptors R-12 and 93093-3. This reflects the ability of the neuropeptide-containing nerves and lens cells to turn on and off the expression of selected messengers. It appears that the neuropeptide-containing nerves demonstrated in this study may be sensory. Furthermore, neuronal nitric oxide synthase-immunopositive axons associated with photocytes in the luminous epithelium have previously been described in this species. Whereas it is clear that the photophores receive efferent (motor) fibers of spinal sympathetic origin, the origin of the neuropeptide sensory innervation remains to be determined. The functional roles of the above neuropeptides or their effects on the bioluminescence or the chemical nature of the terminals, either sensory or postganglionic neurons innervating the photophores, are still not known.     

Localization of telomerase hTERT protein and hTR in benign mucosa, dysplasia, and squamous cell carcinoma of the cervix

Telomerase has been detected by telomerase repeat amplification protocol (TRAP) assay in cervical dysplasia and squamous cell carcinoma but not in most normal cervical tissues. In the present study, the cellular localization of the protein catalytic subunit of telomerase (hTERT) and the RNA component (hTR) were investigated by a sensitive immunohistochemical technique and by in situ hybridization, respectively. hTERT protein was detected in all diagnostic categories of cervical specimens. hTERT was localized predominantly to the lower suprabasal levels of normal squamous mucosa but was detected throughout virtually all levels of the lesional epithelium in low-grade squamous intraepithelial lesions (LSILs), high-grade squamous intraepithelial lesions (HSILs), and squamous cell carcinoma (SCC). Telomerase expression correlated with hTERT detection in SCC and HSIL but was not detected by TRAP assay in most samples of normal mucosa or LSIL. The distribution of hTR correlated with the localization of hTERT in HSIL and SCC but was restricted to the basal and suprabasal cell layers in normal mucosa and LSIL.     

Immunohistochemical characterization of intraepithelial and subepithelial mononuclear cells of the upper airways.

The upper airway is the first site of exposure to inhaled antigens and the site of initiation of mucosal immunity to certain antigens; however, the intraepithelial lymphoid populations of this region have not been well characterized. We studied 6-mu frozen tissue sections from tonsils, adenoids, and nasal mucosae using immunohistochemistry and a panel of antibodies to mononuclear antigens to determine whether nasal mucosa contained distinctive populations of mononuclear cells. Intraepithelial lymphocytes (IELs) of nasal mucosa were CD3+, CD8+, and mainly CD5+. Tonsil and adenoid both showed diffuse CD8+ IELs; clusters of CD4+ IELs were associated with B cells within the crypt epithelium. All nasal IELs were uniformly negative for Leu8 (homing receptor analog of Mel14). Scattered Leu8-positive cells were present within tonsil and adenoid crypt epithelium only. Nasal IELs rarely expressed HML1 and were often CD7-, whereas the majority of tonsillar and adenoidal IELs were HML1+ and variably CD7+. In nasal mucosa and in deep submucosa of tonsil and adenoid, 80 to 90% of T cell receptor expression was of alpha/beta type. There was a concentration of gamma/delta T cell receptor-positive cells in intraepithelial and subepithelial zones of tonsil and adenoid, with areas of up to 30% gamma/delta T cell receptor positivity. A population of intraepithelial dendritic cells was identified in all three tissues expressing mononuclear phagocyte system antigens CD14 and KiM1P, but lacking CD1a. Virtually no B cells and no organized subepithelial lymphoid tissue were identified in nasal mucosa. Nasal mucosal lymphoid tissue seems to differ from that of endodermally derived mucosae, tonsil, and adenoids to share similarities with both mucosa-associated lymphoid tissue and peripheral lymph nodes.