A fine structural study of divalent cation-mediated epithelial union with connective tissue in human oral mucosa

Porcine buccal mucosa as in vitro model: effect of biological and experimental variables

Porcine buccal mucosa has been used as an in vitro model to assess the potential of delivering a molecule via the transbuccal route. However, permeation studies across porcine buccal mucosa show high variability due to various experimental and biological factors. The variability associated with the use of different mucosal regions, tissue storage conditions and tissue processing methods on drug permeation was investigated in this study. The permeability of model diffusants was significantly higher in the region behind the lip when compared to the cheek region because the latter has a thicker epithelium. Porcine buccal mucosa retained its integrity in Kreb's bicarbonate ringer solution at 4 degrees C for 24 h while many other storage conditions resulted in loss of epithelial integrity. Separation of the epithelium from underlying connective tissue either surgically or by heat treatment resulted in an epithelial thickness of approximately 150 microm. Separation of epithelium from the underlying connective tissue by heat treatment did not adversely affect its permeability and integrity characteristics. Investigation of these important biological and experimental variables provides guidance for conducting in vitro transbuccal permeation studies.

Oxidative stress-related markers and langerhans cells in a hairless rat model exposed to UV radiation

Biomarkers related to the oxidative stress in blood and epidermis and the number of Langerhans cells were determined in hairless rats after acute irradiation with 1.54, 1.93, or 2.41 J/cm2 of ultraviolet (UV) light and chronic exposure to 13 suberythemal UV doses of 1.1 J/cm2 for 2 mo. After acute UV irradiation, in epidermis, the thiobarbituric acid-reactive substances (TBARS) content increased at the highest UV dose, whereas the activities of glutathione S-transferase and catalase rose and the oxidized glutathione (GSSG) content diminished at all UV doses. In erythrocytes, glutathione S-transferase activity increased at the two lowest UV doses, glutathione peroxidase activity rose at all UV doses, and catalase activity increased after the highest UV dose. In plasma, the TBARS content and the reduced glutathione (GSH)/GSSG ratio increased at the highest UV dose; the number of Langerhans cells decreased at all UV doses. Linear Pearson correlation analysis revealed many relationships between different biomarkers, and multiple linear regression analysis indicated that the number of Langerhans cells was predicted by epidermal GSSG and catalase (R2 = .64) and by erythrocytic glutathione peroxidase and GSSG (R2 = .72). After suberythemal UV radiation, in epidermis, the GST activity and the content of GSH and GSSG increased; in erythrocytes, the GST activity decreased and the GSH/GSSG ratio increased. Thus, the hairless rat appears to be a useful model for studying the oxidative stress-related mechanisms after UV radiation, which are involved in the loss of the immune capacity mediated by Langerhans cells, even at suberythemal doses.

Genetic and epigenetic alterations of the blood group ABO gene in oral squamous cell carcinoma

Loss of histo-blood group A and B antigen expression is a frequent event in oral carcinomas and is associated with decreased activity of glycosyltransferases encoded by the ABO gene. We examined 30 oral squamous cell carcinomas for expression of A and B antigens and glycosyltransferases. We also examined DNA from these tumors for loss of heterozygosity (LOH) at markers surrounding the ABO locus at chromosome 9q34, for loss of specific ABO alleles, and for hypermethylation of the ABO promoters. Loss of A or B antigen expression was found in 21 of 25 tumors (84%) and was a consistent feature of tumors lacking expression of A/B glycosyltransferases. LOH at 9q34 was found in 7 of 27 cases (26%), and one case showed microsatellite instability. Among 20 AO/BO cases, 3 showed loss of the A/B allele and 3 showed loss of the O allele. Analysis of the proximal ABO promoter by methylation-specific PCR and melting curve analysis showed hypermethylation in 10 of 30 tumors (33.3%), which was associated with loss of A/B antigen expression. ABO promoter hypermethylation was also found in hyperplastic or dysplastic tissues adjacent to the tumors, suggesting that it is an early event in tumorigenesis. Collectively, we have identified molecular events that may account for loss of A/B antigen expression in 67% of oral squamous cell carcinomas.

EGTA treatment of human airways in vitro unmasks M1/MUC5AC mucin in submucosal glands

Mucin staining can be used to evaluate secretory activity of human airways. However, mucin epitopes may be masked by physicochemical properties of the secretions. The aim of this investigation was to examine the effects of the calcium chelator, ethyleneglycol-bis-(beta-aminoethylether)-N,N,N',N'-tetraacetic acid (EGTA) on the detection of M1/MUC5AC mucin in isolated human bronchial preparations. Immunohistochemical investigation and immunoradiometric assays with anti-M1 monoclonal antibodies (Mabs) were used to detect M1/MUC5AC mucin derived from bronchial preparations with an intact surface epithelium, or in tissues where the epithelium had been removed (rubbed preparations). The Mabs labelled both epithelial goblet cells and submucosal glandular cells in EGTA (4 mM)-exposed bronchial preparations, while only goblet cells were stained in EGTA (0.4 mM)-exposed tissues. The quantities of M1/MUC5AC mucin detected in either the bronchial fluids derived from EGTA (4 mM)-exposed intact and rubbed preparations or in bronchial fluids treated with EGTA (4 mM) were significantly increased by two-fold when compared with untreated control values (p<0.001). In addition, lactate dehydrogenase (LDH) activity and protein measurements were unaltered during exposure of human airways to EGTA (4 mM) suggesting that this treatment did not affect tissue viability. These results provide evidence that ethyleneglycol-bis-(beta-aminoethylether)-N,N,N',N'-tetraacetic acid (4 mM) facilitates the detection of M1/MUC5AC mucin by altering the physicochemical properties of respiratory mucin, thereby exposing epitopes with which anti-M1 monoclonal antibodies are reactive. This will allow more accurate measurement of secretory activity in human airways in vitro.

Gross morphology and ultrastructure of dendritic cells in the normal human decidua

Five normal placentae of normal pregnancy and delivery were used to study the gross morphology and ultrastructure of the dendritic cells in the normal human decidua. Zinc iodide osmium (ZIO) mixture was prepared. Small pieces of the placenta were processed for light microscopy and electron microscopy. For light microscopy, the small pieces of placenta were incubated in 20 mM PBS-EDTA solution, ph 7.4 at 37 degrees C to detach the basal plate. The basal plate pieces were incubated in ZIO. A wholemount preparation of the basal plate demonstrated the whole profile and gross morphology of the dendritic cell. For electron microscopy, the placenta pieces were fixed in 3% glutaraldehyde in 0.1 M phosphate buffer, ph 7.4, washed with phosphate buffer, put in ZIO mixture, washed in distilled water, dehydrated in graded ethanol, cleared in propylene oxide, and embedded in resin. Ultra thin sections of the ZIO blocks were cut using a diamond knife and stained with lead citrate. Ultrastructure of the dendritic cell presented multiple cytoplasmic processes, lobulated or round or oval, heterochromatic or euchromatic nucleus, mitochondria, free ribosomes, and pieces of rough endoplasmic reticulum, but no Birbeck granules.

The repopulation of murine Langerhans cells after depletion by mild heat injury

We have developed a model of focal Langerhans cell depletion by mild heat injury and used it to investigate the mechanisms of Langerhans cell repopulation in the injured epidermis. The possibility whether repopulation occurred by recruitment of precursor cells from the circulation or dermis or, alternatively, by migration from the surrounding normal epidermis into the injured area was considered. Repopulation was studied by evaluating the pattern of Langerhans cell reappearance and calculating the rate of change in the density. Heat injury followed by whole-body irradiation with shielding of the injured skin was used to assess repopulation in the absence of bone marrow precursors. Using tritiated thymidine autoradiography, we also investigated whether the newly arrived Langerhans cells (be they from circulating precursors or surrounding normal epidermis) actually divide. The results showed that heat injury completely eliminated the Langerhans cells within the area delineated by the injury. Two hours after injury, the Langerhans cells were fragmented and 2 days later, they could not be detected. Regeneration of the epidermis occurred 2 days after injury and Langerhans cells reappeared scattered somewhat sparsely in the centre of the lesion on day 3. These cells were small and slender, bearing one or two short dendrites. As the dendrites increased in number and in length, the cells became similar morphologically and phenotypically to normal Langerhans cells. The rate of repopulation increased dramatically between days 5 and 7 and reached normal density on day 11. The pattern of Langerhans cell repopulation in the injured area and the lack of repopulation in the irradiated animals indicated that repopulation occurs by immigration of precursors from the circulation or dermis. There was no indication of migration of Langerhans cells from surrounding normal epidermis. Lastly, the newly arrived Langerhans cells failed to divide at the site of injury.

Pattern of cytokine receptors expressed by human dendritic cells migrated from dermal explants

Different reasons account for the lack of information about the expression of cytokine receptors on human dendritic cells (DC): (a) DC are a trace population; (b) the proteolytic treatment used to isolate DC may alter enzyme-sensitive epitopes; and (c) low numbers of receptors per cell. In the present work the expression of cytokine receptors was analysed by flow cytometry on the population of dermal DC (DDC) that spontaneously migrate from short-term culture dermal explants. DDC obtained after dermal culture were CD1alow, CD1b+, CD1c+, human leucocyte antigen (HLA)-DR+, CD11chigh, CD11b+ and CD32+. The DC lineage was confirmed by ultrastructural analysis. DDC expressed interleukin (IL)-1R type 1 (monoclonal antibody (mAb) hIL-1R1-M1; and 6B5); IL-1R type 2 (mAb hIL-1R2-M22); IL-2R alpha chain (mAb anti-Tac; and hIL-2R-M1) and IL-2R gamma chain (mAb 3B5; and AG14C). DDC did not stain for IL-2R beta chain using four mAbs recognizing two different epitopes of IL-2R beta (mAb 2R-B; Mik-beta 1; and CF1; Mik-beta 3, respectively). DDC were also positive for the cytokine binding chains (alpha chains) of IL-3R (mAb 9F5); IL-4R (mAb hIL-4R-M57; and S456C9); and IL-7R (mAb hIL-7R-M20; and R3434). DDC showed low levels of IL-6R alpha chain (mAb B-F19; B-R6; and B-E23) and its signal transducer gp130 (mAb A2; and B1). DDC strongly expressed interferon-gamma receptor (IFN-gamma R) (mAb GIR-208) and were negative for IL-8R (mAb B-G20; and B-F25). All DDC were highly positive for granulocyte-macrophage colony-stimulating factor receptor (GM-CSFR) alpha chain (mAb hGM-CSFR-M1; SC06; SC04, and 8G6) and to a lesser extent for the common beta chain of GM-CSFR, IL-3R and IL-5R (mAb 3D7). On the other hand, reactivity was not found for granulocyte colony-stimulating factor receptor (G-CSFR) (mAb hGCSFR-M1) nor macrophage colony-stimulating factor receptor (M-CSFR) (mAb 7-7A3-17) confirming the DC lineage of DDC. As previously reported for lymphoid DC, DDC expressed tumour necrosis factor receptort (TNFR) 75000 MW (mAb utr-1; hTNFR-M1; and MR2-1) but lacked TNFR 55000 MW (mAb htr-9; MR1-1; and MR1-2). In summary, DDC express receptors for a broad panel of cytokines, even receptors for cytokines whose effects on DC are still unknown (i.e. IL-2R alpha gamma; IL-6R alpha/gp 130; IL-7R alpha gamma).

Proteoglycan changes in carcinogen (4NQO)-treated rat tongue mucosa

The purpose of this study was to undertake preliminary analyses of the extracellular proteoglycans in carcinogen [4-nitroquinoline N-oxide (4NQO)]-treated rat tongue mucosa. Experimental rats were exposed to twice-weekly applications of 4NQO in propylene glycol for six months, after which the animals were killed. Control and 4NQO-treated tissues were subjected to sequential aqueous extractions of proteoglycans under associative and dissociative conditions, followed by alkaline cleavage of protein-glycosaminoglycan linkages to yield a glycosaminoglycan residue. Tissues subjected to 4NQO applications contained smaller proportions of proteoglycans which were readily soluble under associative and dissociative conditions. Proportionately more proteoglycan remained strongly associated with other intercellular tissue components, being released only by alkaline cleavage. These biochemical alterations in preinvasive 4NQO-treated epithelium and connective tissues, together with an observed associated change in water retention by the connective tissue, occurred prior to actual neoplastic invasion and suggest differences in macromolecular conformation and orderliness. We hypothesize that these changes are related to the phenomenon of neoplastic epithelial invasion.

Changes in cytokeratins following treatment of hamster cheek pouch epithelia with hyperplastic or neoplastic agents

The effects of four different hyperplastic agents and of the carcinogen DMBA on cytokeratin expression in hamster cheek pouch epithelia were compared. Reversible hyperplasia was produced by the application of either oil of turpentine, vitamin A or TPA. No hyperplastic changes were produced by application of EPP. Apart from the transient appearance of a 45 kDa cytokeratin in one group treated with vitamin A, the immunohistochemical staining patterns and immunoblot profiles of cytokeratins from cheek pouches treated with each of the hyperplastic agents were identical to controls. Following application of DMBA, the cytokeratins stained with increased intensity in the spinous and granular cell layers. This was associated with increased amounts of 42-56 kDa cytokeratins and decreased production of 62-75 kDa cytokeratins. Monoclonal antibody AE1 detected a 45 kDa cytokeratin in extracts of DMBA-treated epithelia that was not detected in untreated epithelial extracts. Monoclonal antibody AE3 detected an additional 54 kDa cytokeratin band in extracts of DMBA-treated epithelia. These cytokeratin changes were present in preneoplastic epithelia and maintained in neoplastic epithelia.

S100 positive dendritic cells in human lung tumors associated with cell differentiation and enhanced survival

Antigen presenting S100 positive dendritic cells have been quantified in normal trachea, lung, bronchial lymph nodes, 130 lung tumors and in 100 lymph nodes regional to tumor. Dendritic cells are rarely seen as intraepithelial components of the normal bronchial mucosa, but are more commonly observed in the perivascular lymphoid tissue of the submucosa and in the alveolar septae of normal lung parenchyma (6 +/- 4.85 cells/HPF). The density of these dendritic cells is marked in histologically normal bronchial lymph nodes. Bronchioalveolar (Alveolar II), well and moderately differentiated squamous cell carcinomas contain the highest density of S100 positive dendritic cells, while small cell lung cancer and poorly differentiated squamous cell carcinoma show the lowest density. Regional lymph nodes to lung tumors with lymphocytic predominance and active germinal centres show the highest density of dendritic cells, while unstimulated lymph nodes contain the lowest number of S100 positive dendritic cells. Tumor infiltrating lymphocytes are marked in and around lung tumors with the higher density of dendritic cells. Survival of patients whose tumors contain high density of S100 positive dendritic cells is more favourable compared to tumors with low density of these cells. It is concluded that the density of the antigen presenting S100 dendritic cells in lung tumors is related to subtype, and tumor differentiation. A high dendritic cell density is associated with enhanced patient survival.

Immunohistochemical detection of sialyl Le(x) antigen on mucosal Langerhans cells of human oral mucosa following neuraminidase pretreatment

Histochemical assessment of selected carbohydrate sequences on Langerhans cells of human oral mucosa was made by combined use of enzyme digestion and immunostaining with monoclonal antibodies against specific carbohydrate structures. In both frozen sections and epithelial sheets without the enzyme pretreatment, mucosal Langerhans cells, identified by positive staining with anti-CD1a and HLA-DR antibodies, did not express any carbohydrate antigens on their surface. In contrast, following neuraminidase pretreatment of both types of material, the fucosylated type 2 chain (Le(x)) became detectable on Langerhans cells, indicating that sialic acid is the terminal residue of this sequence. Other enzymes were ineffective in this apparent unmasking, and the staining patterns of the other related carbohydrate sequences (Le(y)+, Le(a), Le(b)) remained unaffected by pretreatment with any of the enzymes used. These findings suggest that the mucosal Langerhans cells possess a unique carbohydrate chain, the sialyl fucosylated type 2 sequence (sialyl Le(x) antigen).

Density and distribution of Langerhans cells in the human uterine cervix

Using a combination of immunohistochemistry and planimetric observation we report the density and distribution of Langerhans cells (LCs) in epithelium from the normal ectocervix and from the normal transformation zone (TZ). The density of LCs increased towards the basement membrane. Dendritic LCs aspect was more prominent in the superficial layers.

Elastic fibers in the duct system of the rat submandibular salivary gland

The submandibular salivary gland originates from the floor of the mouth whose mucosa contains elastic fibers. Therefore, such fibers were sought in the duct system of the derivative organ. In adult rats, light microscopy has indeed revealed fine, circumferential, elastic fibers near the basement membrane of the duct epithelium. In the larger extralobular ducts, they were separated from several layers of longitudinal elastic fibers by a capillary-rich zone sparse in elastic fibers except for fine angular ones. More peripherally, larger angular-appearing fibers were frequently present near the submandibular parasympathetic ganglia in the duct wall. As duct diameter decreased, elastic fiber size and number diminished. Intralobularly, the smaller striated ducts, granular and intercalated ducts, and acini generally lacked such fibers. Electron microscopy of the extraglandular portion of the main duct revealed fibrils extending from both fibroblasts and elastic fibers that were close to the epithelium. Microfibrils coursed from them toward the lamina densa. Anchoring filaments joined the lamina densa to the basal plasma membrane of the epithelium. Elastic fibers also appeared to connect to both capillaries and collagen via finer intermediate structures. These associations might permit dynamic interactions of fibroblasts, fibers, smaller fibrillar components, vasa, and the duct epithelium. This interplay could occur during feeding and grooming when tongue protrusion and neck extension stretch the submandibular duct and the gland itself. As a result, the tensile forces engendered could modify cell geometry and the calibers of the larger ducts' lumens and intercellular spaces, thus affecting the flow and composition of salivary secretion.

A scanning electron microscopic study of rete ridges in the cheek pouch of normal young, adult and aged Syrian hamsters

Cheek pouches were removed from BIO 87.20 male hamsters 4 weeks, 8 months or 18 months of age. One pouch from each animal was routinely processed for light microscopy. The epithelium and connective tissue of the remaining pouches were separated with EDTA and routinely processed for scanning electron microscopy. Two types of rete ridge were found: simple, consisting of a single epithelial ridge or blunt projection and complex, made up of multiple interconnecting ridges. The simple type predominated at 4 weeks (90%) but the complex type predominated at 8 (58%) and 18 months (73%). Although the 8- and 18-month pouches were significantly larger than the 4-week ones, the total number of rete ridges was similar at all ages. This, taken in association with the shift in type with age, suggests that the simple rete ridges develop into the complex ones. With increasing age more of the complex rete ridges took on the characteristics of touch domes. These changes in the shape of the rete ridges with age need to be taken into account in studies on experimentally altered cheek pouches so that they are not mistaken as pathological changes.

A light and scanning electron microscope study of epithelial thickenings and rete ridges in the adult hamster cheek pouch

Cheek pouches from 8 adult male BIO strain hamsters were removed; half of them were routinely processed for light microscopy. The epithelium and connective tissue of the remaining pouches were separated using EDTA. These tissues were routinely processed for scanning electron microscopy and their separated surfaces examined; after this, some were reprocessed for light microscopy. A total of 1200 sections were examined by light microscopy and of these, 414 had at least one area showing rete ridges. Scanning electron microscopy confirmed the presence of isolated areas of either single or multiple rete ridges on both the separated epithelial and connective tissue surfaces. The presence of isolated areas of rete ridge formation in normal adult hamster cheek pouch mucosa needs to be taken into account in studies on experimentally altered cheek pouches so that the ridges are not interpreted as being part of a pathological change.

Activity requirements of epidermolytic toxin from Staphylococcus aureus studied by an in vitro assay

The activity of epidermolytic toxin from Staphylococcus aureus was studied in vitro using discs of neonatal mouse skin. By assessing the loss of skin integrity as a function of toxin dose and time, it was possible to put the assay on a semi-quantitative basis. Epidermolysis occurred without any change in rate from pH 3.8 to 8.7, and at an increasing rate in the temperature range of 0-37 degrees C. Activity was observed even at the lowest temperature. More than 30 inhibitors of energy metabolism, central metabolic pathways, receptor binding or proteolysis, individually failed to prevent epidermolysis and it is suggested that intoxication cannot be dependent on receptor-mediated endocytosis. Five metal-ion chelators inhibited epidermolysis, due to an effect on the tissue rather than on the toxin. Using X-ray fluorescence and atomic absorption spectroscopy, it was shown that epidermolytic toxins do not contain any essential metal ions. Some transition metals, but not Ca2+ or Mg2+, prevented the chelator-dependent inhibition of epidermolysis.

A biochemical study of glycosaminoglycans in the palatal rugae of the monkey (Macaca fascicularis)

Glycosaminoglycans (GAG) were extracted from the connective tissue of the palatal rugae, separated by electrophoresis and compared with the results obtained for the remaining palatal mucosal and gingival connective tissues. The GAG content of the rugae (3.01 mg/g defatted dry weight) was higher than in the remaining palatal mucosa (2.33 mg/g defatted dry weight) or gingiva (1.68 mg/g defatted dry weight). Dermatan sulphate was the predominant GAG in both the palatal rugae (48% of total GAG) and the remaining palatal mucosa (50%) followed by hyaluronic acid (33 and 31% respectively). The results do not support previous histochemical observations in which the rugae appeared to be rich in hyaluronic acid.

Glucose metabolism and protein synthesis in stratified squamous epithelia from young and old mice

Glucose metabolism in regionally distinct epidermis and oral epithelia of young adult and aged mice was studied using 5-3H glucose and differentially labelled 14C-glucose. The results obtained with 3H-glucose indicate there is active utilization of glucose by all the epithelia examined. However, an enhanced utilization of glucose was observed in old ear epidermis when compared to that in young mice. The measurements of respiratory 14CO2 in the young and old epidermis indicated that aerobic glycolysis was significantly greater in ear epidermis and buccal epithelium from old than from young mice. Pentose phosphate activity was significantly reduced in palate epithelium from old animals when compared with that of young animals. Incorporation of radioactively labelled leucine and histidine revealed some regional differences between the epithelia examined. However, no age-associated differences were observed in any of the tissues. It is concluded that the observed metabolic changes reflect regionally specific adaptations to local factors rather than representing a programmed biological event.

Diffusion of [2-14C]diazepam across isolated hairless mouse stratum corneum/epidermal tissues

The objective of this study was to develop a method of preparing mouse stratum corneum/epidermal (SCE) tissue without the dermis for use in drug diffusion studies. The diffusion of radiolabeled diazepam across this new preparation has been studied and the effect of the dermis on diffusion evaluated. Incubation of large pieces of mouse skin in a 20mM EDTA, 15 mM sodium phosphate buffer, pH 7.2, in normal saline for 3-4 h at 37 degrees C resulted in a tissue which easily separated at the epidermal-dermal junction. The resulting tissue contains stratum corneum and epidermis, which are the same layers used in studies with human skin in vitro. The EDTA treatment did not effect diffusion of [2-14C]diazepam across whole mouse skin (SCE and dermis) used as controls. The rate of drug diffusion was greater across SCE than SCE and dermis, however, 0.48-1.12 micrograms/cm2/h versus 0.11-0.52 microgram/cm/h, respectively. The permeability coefficients for mouse SCE ranged from 1.92-4.48 X 10(-2) cm/h. The lag times and diffusion coefficients were 0.36-0.91 h and 0.1-0.6 X 10(-6) cm2/h, respectively. The presence of the dermis decreased the diffusion rate or flux of diazepam. The dermis appears to accumulate drug until it is saturated and then the drug diffuses into the receiving chamber.

Developmental expression of C3 receptor on murine epidermal Langerhans cells during ontogeny

The developmental expression of C3 receptor, an important surface marker of murine epidermal Langerhans cells (LCs), was quantitatively studied using an immunohistochemical technique on epidermal sheets and then compared with developmental expression of Ia antigen and membrane ATPase. Anti-Mac-1 monoclonal antibody associated with CR3 was used for detecting C3 receptor and proved positive for LCs by immunoelectron microscopy. Mac-1 positive (Mac-1+) cells showed quite a different distribution from those of ATPase+ and Ia+ cells. Almost the same number of Mac-1+ and ATPase+ cells were present during the embryonic period. The number of Mac-1+ cells gradually decreased from day 1 to day 5 of postnatal life, after which they increased again. Using the double-labeling technique on epidermal sheets at day 1 of postnatal life, it was shown that Ia+ cells possessed membrane ATPase activity and some Mac-1+ cells expressed Ia antigen. On days 4 and 7 of postnatal life all Mac-1+ cells expressed Ia antigen. These findings suggest that Mac-1 antigen observed during the embryonic period gradually fades after birth and is re-expressed after day 5 of postnatal life.

Age-associated changes in Langerhans cells of murine oral epithelium and epidermis

Oral mucosa and skin of older individuals are immunologically less responsive to a range of allergens, but it is not known whether this is due to changes in the number of Langerhans cells or to impaired cell function. EDTA-separated epithelial sheets from the cheek and palate mucosa, and from ear aN< footpad skin of three-month-old and 24-month-old C57BL/6NNia mice were stained for ATPase, beta-glucuronidase activity and Iab-surface antigen to demonstrate Langerhans cells. The general distribution of such cells was unchanged with age, but those in epithelia from the old mice were more varied in shape, with irregular celL bodies and more elongated dendritic processes. The numerical density of Langerhans cells in old mice was reduced by 30-59 per cent compared with that in young mice.

Differential distribution of ATPase- and T6-positive cells (Langerhans cells) in the limbus and cornea of Hereford and non-Hereford cattle

Epithelial sheets from the limbus, cornea, and third eyelid of Hereford and non-Hereford cattle were examined for the presence of Langerhans cells (LC) using the membrane enzyme ATPase as a marker for LC. The aim of the study was to test the hypothesis that differences in LC density exist between the various ocular epithelia of these animals producing depressed immune surveillance in the case of Hereford cattle. The presence of LC in ocular tissues was confirmed by parallel studies which detected epithelial cells bearing T6, an antigen expressed by human LC. Studies using serial sections demonstrated that T6+ cells also reacted with an anti-human HLA-DR monoclonal antibody. The detection of T6+, DR+ and ATPase+ cells in ocular epithelium in the absence of infiltrating macrophages suggested that LC are present in these tissues. While there were no significant differences in the density of T6+ cells between non-Hereford and Hereford cattle, in the latter ATPase+ cells were significantly fewer in the lateral, medial, and upper limbus.

Langerhans' cell density in the normal exocervical epithelium and in the cervical intraepithelial neoplasia

A morphometric analysis was performed of the Langerhans' cell density in epithelial sheets obtained from normal exocervices and from exocervices with squamous carcinoma. Laminae of exocervices with squamous carcinoma that showed cervical intraepithelial neoplasia (CIN) were classified according to its predominant degree of severity as CIN I, CIN II and CIN III or as normal when no neoplastic changes were found. Laminae with CIN showed a higher Langerhans' cell density than the laminae from normal exocervices and than the normal laminae of exocervices with carcinoma. The magnitude of the increase of Langerhans' cell density and the degree of severity of the neoplastic changes appear as closely related phenomena. An increase of the more ramified types of Langerhans' cell in the laminae containing CIN was found, with the most ramified being more frequent in the most severe lesions.

Blister fluid from epidermolysis bullosa letalis induces dermal-epidermal separation in vitro

Normal human skin was cultured with blister fluid (BF) from 2 patients with epidermolysis bullosa letalis (EB letalis). After 12-24 h incubation, initial separation was observed at the dermal-epidermal junction. Complete separation was seen after 48-72 h incubation. Epidermal and dermal structures themselves were well preserved up to 72 h. Electron microscopy revealed that the separation took place through the lamina lucida. These findings corresponded to in vivo histologic pictures of the skin lesions. The activity in BF was lost by lowering (pH 4) or elevating (pH 9) pH of the media or by prolonged autoclaving (5h). Boiling and dialization did not affect the activity of BF. Certain protease inhibitors, such as alpha 2-macroglobulin, aprotinin (trasylol), soybean trypsin inhibitor, phenylmethylsulfonylfluoride, or 50% fetal bovine serum, inhibited or depressed the separation, while triamcinolone acetonide, n-ethylmaleimide, or EDTA did not. Patient's serum (case 2) and normal human sera partially inhibited cleft formation. It was proposed that in EB letalis some factors in BF, probably protease(s), play an important role in the induction and/or spread of blister.

ATPase Langerhans cell staining: a technique allowing progression from light to electron microscope observation

A technique which enables good visualization of the membranous ATPase activity of epidermal Langerhans cells is described. The method has the advantage of keeping intact most of the ultrastructural details. It may allow the observation, under pathologic conditions, of ultrastructural modifications in ATPase-negative Langerhans cells still recognizable by their Langerhans cell granules.

The in vitro effect of retinol on human gingival epithelium. II. Modulation of Langerhans cell markers and interleukin-1 production

The in vitro influence of retinol on the markers of gingival Langerhans cells (LC) was investigated using an organ culture system. Retinol at a dose of 5 micrograms/ml produced an increase in the density of T6-positive cells within the epithelium which peaked during the first 24 h of culture. LC HLA-DR and ATPase markers were maintained for the same period, while all markers were depressed after 72 h. These effects were not seen in explants cultured in conventional or alcohol-enriched media, in which all markers were lost in an exponential fashion. In addition to modulation of LC markers, retinol treatment also prolonged the expression of HLA-DR antigens by gingival keratinocytes. These findings, together with the augmented production of interleukin-1-like activity by retinol-treated gingival organ cultures suggest that low doses of retinol may alter immune reactions within epithelia via stimulation of both keratinocytes and LC.

Modulation of HLA-DR antigens in the gingival epithelium in vitro by heat-killed Fusobacterium nucleatum and E. coli lipopolysaccharide

The in vitro influence of the periodontopathic organism Fusobacterium nucleatum (FN) on gingival tissue was examined using an organ culture system. Treatment of gingival explants obtained from periodontally diseased sites with suspensions of FN, stimulated the expression of HLA-DR antigens by Langerhans cells (LC) in a dose-dependent fashion, and produced a maintenance of the LC markers T6 and ATPase. Similar effects were seen when E. coli lipopolysaccharide (LPS) was substituted for suspensions of FN. With both FN and LPS the expression of HLA-DR by gingival keratinocytes was maintained throughout the 72-h culture period, despite the cytotoxic effects of these agents. Using a variety of immunohistological techniques and a monoclonal antibody specific for the strain of FN used, it was possible to demonstrate the uptake of FN antigens by LC within the gingival epithelium.

The effect of atrophy, hyperplasia, and keratinization accompanying the estrous cycle on Langerhans' cells in mouse vaginal epithelium

The morphology and numbers of Langerhans' cells vary in epithelia with different patterns of hyperplasia and keratinization. Langerhans' cells stained for ATPase were compared at five phases of the estrous cycle in murine vaginal epithelium. The cells were more dendritic and sparsely distributed with hyperplasia and were less dendritic and more densely distributed with atrophy. Greater numbers of the cells did not accompany keratinization at estrus. Ultrastructurally, three types of Langerhans' cells were discriminated. The first type, active in protein synthesis and phagocytosis, was commonest in sloughing and atrophic epithelium. The second type, containing accumulated and dispersed, electron-dense bodies presumed to be lysosomes, predominated in hyperplastic epithelium. The third, a mature resting cell, was found only after keratinization was complete. This study shows that Langerhans' cells in murine vaginal epithelium vary in morphology and numbers with the epithelial changes of the estrous cycle which may relate to their immunological role, but does not support the contention that their distribution is important for keratinization.

Derivation of Langerhans cell granules from cytomembrane

OKT6 monoclonal antibody has been shown to react selectively with Langerhans cell (LC) cytomembrane in human skin. Taking advantage of this property, we examined the stainability of LC granules with OKT6 using immunoperoxidase method at the ultrastructural level. It was found that the delimiting membranes of LC granules, both attached to the cytomembrane and in the cytoplasm, were specifically stained in the epidermal sheet. These findings suggest that LC granules are derived from invagination of LC cytomembrane and retain T6 antigen. We employed two new methods: one is promotion of endocytosis by incubating the epidermal sheet in tissue culture medium with OKT6 at 4 degrees C and further incubation without OKT6 at 37 degrees C, and the other is the use of saponin to facilitate penetration of the secondary antibody through the cytomembrane of LC. In the LC cytoplasm only the delimiting membrane of the granule was stained; other membranous organelles, such as mitochondria and rough endoplasmic reticulum, did not react.

Microthread-like filaments connecting the epithelial basal lamina with underlying fibrillar components of the connective tissue in the rat trachea. A real anchoring device?

The structural relationship between the basal lamina and the underlying reticular tissue was studied, with special attention to the relationship among basal lamina-associated anchoring fibrillar (AF) arcs (Kawanami et al. 1978, 1979) and other fibrillar components, in the epithelium-denuded trachea of the rat. Quantitative analysis of a large number of AF arcs reveals that the majority of the AF arcs has no other fibrillar components of passage. This suggests that most AF arcs do not serve as a real anchoring device, connecting the basal lamina with the underlying reticular tissue, as has so far been suggested by Kawanami et al. (1978). Ruthenium-red staining reveals the presence of a unique meshwork of microthread-like filaments connecting the undersurface of the basal lamina or the AF arcs with the underlying fibrillar components with a remarkable continuity, suggesting that the filaments act as a real anchoring device; these filaments link, instead of the AF arcs, the basal lamina, to the subjacent reticular tissue. Various enzymatic treatments of the filaments indicate that their chemical nature is probably non-collagenous (glyco)protein without glycosaminoglycan moieties.

Variation in basement membrane topography in human thick skin

Samples of human plantar and palmar skin were excised and incubated in 20 mM EDTA after which the epidermis was gently separated from the dermis with the plane of separation occurring in the lamina lucida. Scanning electron microscopic examination of the dermal component revealed the classically described series of regularly spaced grooves and papillae that characterize the epidermal-dermal junction in thick skin. Primary dermal grooves exhibited evenly spaced tunnels that were originally occupied by sweat gland ducts. The basement membrane (basal lamina) in the primary grooves was relatively smooth but did exhibit a flattened, reticulated pattern at high magnifications. The basement membrane of secondary dermal grooves and papillae was in the form of numerous, elevated microridges off of which septae arose at roughly right angles. The surface appearance of the basement membrane in these areas was that of a honeycomb owing to the numerous compartments and recesses formed by the ridges and septae. Degradation of the basement membrane by trypsin demonstrated that the foundation for the highly folded and compartmentalized basement membrane was composed of dermal collagen fibrils, 60-70 nm in diameter, that were arranged in a series of variably sized, interconnected collagen bundles or walls. Epidermal basal cells extended cytoplasmic (foot) processes into two or more compartments, formed by the ridges and septae, which considerably amplified the basement membrane surface available for epidermal attachment. Scanning electron microscopic studies of the epidermal-dermal junction confirm the variable surface character of this interface previously reported by others using sectioned material.(ABSTRACT TRUNCATED AT 250 WORDS)

In-vitro permeability of porcine oral mucosa after epithelial separation, stripping and hydration

The permeability of porcine skin, gingiva, floor of mouth and buccal mucosa was measured in perfusion chambers using isotopically-labelled water and horseradish peroxidase. Values obtained for the permeability of the epithelium of each of these regions, after separation from the connective tissue with EDTA, did not differ significantly from those obtained for the intact tissue; however, the connective tissue alone had a permeability 2-8 times greater than that of the whole tissue. Stripping the surface layers of the floor of mouth mucosa increased its permeability to that of connective tissue. These results indicate that the functional permeability barrier of the oral mucosa, like that of skin, is located in the epithelium and occupies the superficial layers. After exposure to an aqueous environment for up to 67 h, the permeability of skin and keratinized oral mucosa showed similar but slight increases whereas that of non-keratinized mucosa showed a more rapid rise. These differences may reflect the different composition of the intercellular permeability barrier in keratinized and non-keratinized oral tissues.

The keratin polypeptide patterns in heterotypically recombined epithelia of skin and mucosa of adult mouse

Epithelial-mesenchymal interactions were investigated considering both morphologic criteria and keratin polypeptide expression in homotypic and heterotypic recombinants of adult mouse skin and oral mucosa. Two series of cross-recombinants of epithelia with different morphology and keratin patterns were chosen: (a) footpad epidermis/ear dermis and ear epidermis/footpad dermis; (b) palate epithelium/cheek connective tissue and cheek epithelium/palate connective tissue. Homotypic and heterotypic recombinants were prepared after EDTA-separation of the original tissues and then grown on syngeneic mice in subcutaneously prepared protected graft chambers. EDTA-separation is especially suited to completely separate the epidermal-dermal union, and the transplantation procedure used strictly prevents contamination with host epithelium. Five weeks after implantation keratins were analyzed by one and two-dimensional gel electrophoresis and peptide mapping. In both series, homotypic recombination of the tissues did not alter the original morphology and keratin polypeptide composition of the individual epithelial components. Ear epidermis displayed no significant changes in structure or keratin pattern in heterotypic recombinants. Recombined with ear dermis, footpad epidermis showed acquisition of some morphologic features typical for ear epidermis and slight changes in keratin composition which were, however, difficult to interpret due to the normal similarities of footpad keratin with that of ear. In contrast, the heterorecombinants of the palate/cheek series exhibited considerable alterations in their keratin patterns. Either epithelium showed suppression of distinct keratin subunits and de novo expression of subunits characteristic of the epithelium normally associated with the connective tissue component. The keratin patterns of both matches closely resembled each other and represented patterns intermediate between the normal patterns. This partial, however, significant modulation in the expression of differentiation markers was paralleled by similarly directed changes in the architecture of the heterotransplanted tissues, thus indicating that both morphogenesis and cytodifferentiation of certain adult epithelia can be influenced by extrinsic mesenchymal factors.

Dermatitis herpetiformis: preparation of papillary dermis and the effect of proteolytic enzymes on the IgA deposits

A technique has been devised to isolate the thin papillary part of the dermis from punch biopsies. Papillary dermis has been treated with various proteolytic enzymes in order to release or solubilize the granular IgA deposits from the papillary dermis of patients with dermatitis herpetiformis. Incubation of the thin skin preparations with pepsin caused a disappearance of the specific fluorescence with antibodies to human IgA. After peptic digestion small amounts of IgA could be detected in the supernatants. There was some evidence that this amount was larger for preparations from patients with dermatitis herpetiformis than from controls. Corresponding procedures with trypsin, collagenase, or elastase had no detectable effect on the IgA deposits. The experiments with elastase seemed to give support for previous reports on association between the granular IgA deposits and the microfibrils of elastic fibers.

Extracellular calcium ion depletion induces release of glycoproteins by canine trachea

The effect of depletion of extracellular Ca2+ on the release by canine trachea in vitro of [14C]glucosamine-labeled trichloroacetic-phosphotungstic acid-precipitable glycoproteins was evaluated. Incubation in Ca2+,Mg2+-free medium containing ethyleneglycol-bis(beta-aminoethylether)-N,N'-tetraacetic acid (EGTA, 10(-3) to 10(-5) M) induced a time-dependent increase in base-line release of high molecular weight, radiolabeled glycoprotein that was not blocked by 10(-5) M atropine, propranolol, or phentolamine. Maximal increase was proportional to EGTA concentration and was augmented by preincubation with methacholine chloride. The secretory response was reversed by reincubation of explants in medium containing Ca2+, Sr2+, or Ba2+ but not Mg2+. Assessment of explants denuded of epithelium or cartilage indicated that the secretory response to depletion of Ca2+ originated in the tracheal submucosa, probably the submucosal glands. Quantitative autoradiographic analyses showed, however, that incubation in Ca2+,Mg2+-free medium had no effect on radiolabel release from mucous or serous cells in the tracheal submucosal glands. Increased radiolabeled glycoprotein release in Ca2+,Mg2+-free medium was accompanied by exfoliation of the surface epithelium, though the two effects were apparently unrelated. The secretory response was not due to cell lysis or increased release of radiolabeled glycosaminoglycans, and the finding that the specific activity (ratio of bound radiolabel to protein content) of the glycoprotein released was not changed in Ca2+,Mg2+-free medium showed that it was not due to a change in the rate of glycoprotein synthesis. A model is proposed in which depletion of Ca2+ increases the rate of flow of mucus from the duct lumens of the tracheal glands.

Human mucosal Langerhans cells: postmortem identification of regional variations in oral mucosa

Modified ATPase histochemistry was used to identify and count Langerhans cells (LC) in autopsy tissue from 8 oral mucosal sites, 8-20 h postmortem. The specificity of the ATPase method was confirmed on serial sections with indirect immunofluorescence using monoclonal antibodies OKT6 and antihuman Ia. Average LC counts on ATPase-stained epithelial sheets from each of the 8 sites ranged from 160-550 LC/mm2. Nonkeratinized mucosae of the soft palate, ventral tongue, lip, and floor of the mouth had the highest counts (mean +/- SD, 508 +/- 110 LC/mm2, n = 24), and keratinized mucosae of the hard palate and gingiva had the lowest counts (201 +/- 97 LC/mm2; n = 8). LC frequency was variable in 2 sites: In the dorsal tongue, LC occurred on only one side of filiform papillae and were absent from regularly recurring areas of interpapillary epithelium. In the cheek mucosa, LC clustered around connective tissue papillae and their numbers showed marked individual variation (130-650 LC/mm2). The number of LC in nonkeratinized oral mucosa is approximately the same as in skin, but keratinized oral mucosa has fewer LC. The frequency of oral mucosal LC varies inversely with the degree of keratinization. There are regions of the oral mucosa that have no LC.

Langerhans cells in squamous exocervical carcinoma: a quantitative and ultrastructural study

The gross morphology, density, distribution, and ultrastructure of the Langerhans cell (LC) in the human exocervix in cases of cervical intraepithelial neoplasia (CIN) and incipient carcinoma were investigated. Two zinc-iodide-osmium (ZIO) procedures were applied to epithelial sheets and to tissue sections for light and electron microscopy. The ZIO methods as well as conventional electron microscopy revealed the presence of LCs in CIN, in the invasive prongs, and in the lymphoid infiltrate of the chorion. The epithelial sheets demonstrated a LC density 3.5 times higher than we have previously reported for the normal exocervix. The region around the external os displayed the highest LC density and the most advanced grades of CIN. The LCs were classified into six types according to the number of processes and secondary branches. The most ramified types were more abundant in the neoplastic exocervix than in the normal exocervix. The LC in all locations of the exocervix with carcinoma showed hypertrophy of the rough and smooth endoplasmic reticula and a large number of filaments. The RER displayed dilated cisternae containing electron-dense material. Specialized contacts between LCs and lymphocytes were an outstanding finding. These findings suggest that the LC might play an important role in the neoplastic process.

Human upper epidermal cytoplasmic antibodies are directed against keratin intermediate filament proteins

Upper cytoplasmic (U-Cyt) antibodies are directed against cytoplasmic antigens found in keratinocytes in the upper layers of the epidermis. Until now, they have been defined by indirect immunofluorescence and are known to occur in the sera of patients with cutaneous diseases such as bullous dermatoses, basal cell carcinomas, and melanomas. An increased incidence of U-Cyt antibodies has also been reported in the sera of patients with noncutaneous diseases, such as pulmonary neoplasms. They have been found in addition in the sera of some normal individuals. In this study we have identified keratin intermediate filaments (KIF) as antigens U-Cyt antibodies are directed against. KIF proteins were prepared, separated by polyacrylamide gel electrophoresis, transblotted to nitro-cellulose strips, and used as substrates for antibody binding. Sera containing U-Cyt antibodies by indirect immunofluorescence also had antibodies that were directed against high molecular weight (65,000, 63,000, 61,500) KIF proteins. When KIF proteins were separated according to their charge and their molecular weight by two-dimensional gel electrophoresis and transblotted, the anti-KIF protein antibodies bound to virtually all charge isomers of the KIF proteins at the respective molecular weight. The antibody titers measured using the transblotting technique were 10 to 160 times higher than those found by indirect immunofluorescence. To determine whether U-Cyt antibodies were directed against KIF, a series of absorption and elution experiments were performed. Absorption of test sera with purified KIF removed both U-Cyt antibodies and anti-KIF protein antibodies. Absorption with another type of intermediate filament derived from fibroblasts, vimentin, did not remove U-Cyt or anti-KIF protein antibodies. Absorbed U-Cyt and anti-KIF protein antibodies were both eluted from the same KIF preparation and shown to bind to U-Cyt antigens by indirect immunofluorescence and KIF proteins by transblotting. Absorption of a serum containing U-Cyt antibodies, anti-nuclear antibodies, and anti-basement membrane zone antibodies with purified KIF resulted in the removal of the U-Cyt antibodies but not the other types of antibody. In addition, all test sera, even those that lacked U-Cyt antibodies, were found to have low-titer antibodies against KIF proteins by the transblotting technique. These data indicate that KIF proteins bear antigens to which U-Cyt antibodies are directed and that low titer antibodies against KIF proteins may be much more common than previously appreciated.

Antigens specified by the Tla locus are expressed on the surface of murine Langerhans cells

A monoclonal antibody against the murine thymus leukemia antigen TL, was employed to demonstrate the presence of the antigen on the surface of dendritic cells in murine epidermis of Tla-positive strains, B.10A and A.TH. Immunofluorescence and immunoperoxidase staining of EDTA-separated epidermal sheets demonstrated dendritic cells with a distribution pattern and density comparable to that noted for anti-IAk staining. Tla-negative mouse strains such as A.TL, C3H/HeJ, and C57BL/6 did not show any staining of dendritic epidermal cells. Epidermal cell suspensions similarly contained 2-4% cells with discrete surface staining with anti-TL antibody. Capping was noted in these cells. Once again positive results were noted only in appropriate Tla-positive strains. Control staining was carried out in all cases on frozen sections of thymii from mice. Thymocytes in the cortical zones and some dendritic cells at the corticomedullary junction were stained. TL antigen in mouse appears to be analogous to T-6 antigen previously detected on human Langerhans cells.

Proteoglycans of human gingival epithelium and connective tissue

Proteoglycans extracted from separated specimens of healthy human gingival epithelium and from connective tissue have been purified. The epithelial proteoglycans fractionated as a single included peak on Sepharose 4B-CL and contained heparan sulphate and dermatan sulphate glycosaminoglycans. The connective-tissue proteoglycans separated into three major populations on Sepharose 4B-CL, one of which was excluded from this gel under associative conditions (0.5 M-sodium acetate, pH 7.4). Subsequent fractionation of the excluded material under dissociative conditions (4 M-guanidinium chloride/0.05 M-sodium acetate, pH 7.4) revealed an absence of any aggregate formation of molecules within this population. The connective-tissue proteoglycans contained heparan sulphate, dermatan sulphate and chondroitin 4-sulphate, the proportions of which varied with the molecular size of the proteoglycans. Amino acid analysis of the protein cores of gingival-epithelial and connective-tissue proteoglycans revealed differences that were similar to the differences described between other types of proteoglycans such as those from skin.

EDTA separation and ATPase Langerhans cell staining in the mouse epidermis

Sheets of ethylenediaminetetraacetic acid (EDTA)-separated epidermis were examined using scanning electron, transmission electron, and light microscopy; sheets were also examined after staining for adenosine triphosphatase (ATPase) activity. Staining was improved by longer incubation with EDTA and by elimination of Trismal buffer as a tissue rinse. EDTA-separated epidermis showed better retention of ultrastructural integrity when washed with phosphate-buffered saline. The ATPase staining procedures described in this present study are ultrastructurally specific for the Langerhans cell.

Differential distribution of Langerhans cells in organ culture of human skin

Epidermal Langerhans cells (LCs) function as the antigen-presenting cells in such cutaneous cell-mediated immune responses as contact hypersensitivity and in the mixed epidermal cell-lymphocyte reaction. They have also been implicated in the immune response in skin allograft rejection. Since organ culture of thyroid and pancreas has been shown to prolong allograft survival, presumably through the loss of antigen-presenting cells, we examined the effect of skin explant culture on LC survival. Human skin explants were placed in organ culture and examined serially as whole mounts of epidermis for the presence of LCs as judged by ATPase activity, and OKT-6 and HLA-DR antigens. Although we observed morphologic changes and an absolute reduction in the number of positively stained cells, culture for up to 28 days failed to deplete explants of these cells. Langerhans cells were also sought in the epidermal outgrowths that develop peripheral to the original explants. They were never seen in the area beyond 0.3 mm from the explant edge. Organ culture of skin thus provides a means to explore the contribution of LCs to skin allograft rejection by comparing the immunogenicity of epidermal portions of the explant with the epidermal outgrowth.

Langerhans cells of the ocular surface

Recent evidence has been collected by several investigators defining a distinct population of dendritic cells (Langerhans cells) of mesenchymal origin residing in the epidermal surfaces of many mammalian species. These cells play a dominant role in the processing of antigens presented through cutaneous surfaces and carry a Class II histocompatability antigen felt to be of central importance in the afferent arm of allograft rejection. They also possess many of the characteristics of macrophages active in the efferent arm of immunologic responses. An equivalent subset of dendritic cells (Langerhans cells) in ocular surface epithelium of the human, mouse rat, and guinea pig has been identified by enzyme histochemistry, immunofluorescence, and electron microscopy. Ocular surface Langerhans cells proliferate in the setting of corneal inflammation (remote and recent) and are depleted by topical and systemic corticosteroids. Ocular surface Langerhans cells may play a central role in ocular contact hypersensitivity, corneal allograft rejection, and ocular surface immune surveillance.

Distribution of histidine-rich basic protein, a possible keratin matrix protein, in rat oral epithelium

The indirect fluorescent-antibody technique was used to locate histidine-rich basic protein, filaggrin. In the newborn, immunofluorescence was seen in the cornified layers and in keratohyalin granules throughout the mouth using antibody specific for epidermal filaggrin, a distribution similar to that in epidermis where it is thought that filaggrin functions as the keratin matrix protein. In the adult immunofluorescence was in keratohyalin granules of palate, buccal and tongue epithelium but in the stratum corneum was limited to the soft palate with weak, patchy areas in the densely keratinized epithelium of the hard palate and tongue. Immunofluorescence was delineated at the boundary between the soft and hard palates. A protein apparently identical to epidermal filaggrin was identified in extracts of newborn palate by its mobility on sodium dodecyl sulphate-polyacrylamide gels and subsequent reaction with the antibody used for immunofluorescent studies. This protein was not detected in extracts of adult oral epithelia. Both newborn and adult tissues contained high mol. wt cross-reactive protein, suggestive of the filaggrin-precursor protein extractable from keratohyalin granules. The distribution of filaggrin was consistent with its function as a keratin matrix protein in the newborn oral epithelium and some less densely keratinized regions of the adult. However, in the adult mouth, filaggrin is not detectable in the stratum corneum of the most densely keratinized regions. Thus, the protein must be lost or its antigenic sites altered with maturation of the animal, depending on the type and extent of keratinization.