Genetic basis of sex-specific resistance to neuro-oncogenesis in (BDIX × BDIV) F2 rats

The identification of cancer susceptibility- and resistance-mediating genes is an essential prerequisite for prevention and early diagnosis of malignant tumors. Model organisms are helpful to identify variant alleles involved in pathways affecting individual cancer risk. BDIX and BDIV rats of both sexes are highly susceptible and resistant, respectively, to the development of N-ethyl-N-nitrosourea (ENU)-induced malignant peripheral nerve sheath tumors (MPNST), predominantly in the trigeminal nerves. Nevertheless, female (BDIV × BDIX) F₂ intercross rats have a lower MPNST incidence and a longer latency time than males. Six of seven autosomal gene loci (Mss1-Mss7) controlling genetic susceptibility and resistance in (BDIV × BDIX) F₂ hybrids exert allele- and sex-specific effects on tumor incidence and/or latency time of variable strength. Homozygous BDIV alleles at Mss4 or Mss7 located on rat chromosomes 6 and 10, respectively, are sufficient to cause almost complete resistance to ENU-induced MPNST development in female F₂ rats regardless of the genotype of the other locus. Both loci display only weak effects on male cancer risk. Survival curves of ENU-treated F₂ females depleted of animals with homozygous BDIV alleles at Mss4 and Mss7 are not significantly different from those of males, suggesting that these loci account mainly for the excess tumor resistance observed in female F₂ rats. By haplotype analysis Mss4 and Mss7 could be narrowed down to 20 and 12 Mb, respectively, providing a basis for the positional identification of candidate genes.     

The deaf and the dumb: the biology of ErbB-2 and ErbB-3

ErbB-2 (also called HER2/neu) and ErbB-3 are closely related to the epidermal growth factor receptor (EGFR/ErbB-1), but unlike EGFR, ErbB-2 is a ligandless receptor, whereas ErbB-3 lacks tyrosine kinase activity. Hence, both ErbB-2 and ErbB-3 are active only in the context of ErbB heterodimers, and ErbB-2. ErbB-3 heterodimers, which are driven by neuregulin ligands, are the most prevalent and potent complexes. These stringently controlled heterodimers are repeatedly employed throughout embryonic development and dictate the establishment of several cell lineages through mesenchyme-epithelial inductive processes and the interactions of neurons with muscle, glia, and Schwann cells. Likewise, the potent combination of signaling pathways engaged by the heterodimers drives an aggressive phenotype of tumors of secretory epithelia, including breast and lung cancers. This review highlights recent structural insights into the mechanism of ligand-induced heterodimer formation, and concentrates on signaling pathways employed by ErbB-2 and ErbB-3 in normal and in malignant cells.     

Molecular, genetic and biochemical characterization of lactate dehydrogenase-A enzyme activity mutations in Mus musculus

Four independent heterozygous lactate dehydrogenase (LDH) mutations with approximately 60% of wild-type enzyme activity in whole blood have been recovered. The mutant line Ldh1 ^a2Neu proved to be homozygous lethal, whereas for the three lines Ldh1 ^a7Neu, Ldh1 ^a11Neu, and Ldh1 ^a12Neu homozygous mutants with about 20% residual activity occurred in the progeny of heterozygous inter se matings. However, the number of homozygous mutants was less than expected, suggesting an increased lethality of these animals. Various physicochemical and kinetic properties of LDH are altered. Exons of the Ldh1 gene were PCR amplified and sequenced to determine the molecular lesion in the mutant alleles. Ldh1 ^a2Neu carried an A/T → G/C transition in codon 112 (in exon 3), resulting in an Asn → Asp substitution; Asn112 is part of the helix αD, which is involved in the coenzyme-binding domain. Ldh1 ^a7Neu contained an A/T → C/G transversion within the codon for residue 194 in exon 4, causing an Asp → Ala substitution, which may affect the arrangement of the substrate-binding site. Three base substituions were discovered for the mutation Ldh1 ^a11Neu in exon 7: the transition C/G → T/A, a silent mutation, and two transversions C/G → A/T and C/G → G/C, both missense mutations, which led to the amino acid replacements Ala319 → Glu and Thr321 → Ser, respectively, located in the αH helix structure of the COOH tail of LDHA. We suggest that the mutation is the result of a gene conversion event between Ldh1 ^a wild-type gene and the pseudogene Ldh1-ps. The alteration Ile → Thr of codon 241 in exon 6 caused by the base pair change T/A → C/G was identified in the mutation Ldh1 ^a12Neu; IIe241 is included in the helix α2G, a structure that is indirectly involved in coenzyme binding. Each of the sequence alterations has a potential impact on the structure of the LDHA protein, which is consistent with the decreased LDH activity and biochemical and physiological alterations. Received: 3 July 1997 / Accepted: 30 September 1997     

Brain lipid binding protein in axon-Schwann cell interactions and peripheral nerve tumorigenesis

Loss of axonal contact characterizes Schwann cells in benign and malignant peripheral nerve sheath tumors (MPNST) from neurofibromatosis type 1 (NF1) patients. Tumor Schwann cells demonstrate NF1 mutations, elevated Ras activity, and aberrant epidermal growth factor receptor (EGFR) expression. Using cDNA microarrays, we found that brain lipid binding protein (BLBP) is elevated in an EGFR-positive subpopulation of Nf1 mutant mouse Schwann cells (Nf1(-/-) TXF) that grows away from axons; BLBP expression was not affected by farnesyltransferase inhibitor, an inhibitor of H-Ras. BLBP was also detected in EGFR-positive cell lines derived from Nf1:p53 double mutant mice and human MPNST. BLBP expression was induced in normal Schwann cells following transfection with EGFR but not H-Ras12V. Furthermore, EGFR-mediated BLBP expression was not inhibited by dominant-negative H-Ras, indicating that BLBP expression is downstream of Ras-independent EGFR signaling. BLBP-blocking antibodies enabled process outgrowth from Nf1(-/-) TXF cells and restored interaction with axons, without affecting cell proliferation or migration. Following injury, BLBP expression was induced in normal sciatic nerves when nonmyelinating Schwann cells remodeled their processes. These data suggest that BLBP, stimulated by Ras-independent pathways, regulates Schwann cell-axon interactions in normal peripheral nerve and peripheral nerve tumors.     

Composition and Metabolism of Gangliosides in Rat Peripheral Nervous System During Development

Abstract: Ganglioside composition of rat trigeminal nerve was studied during development in order to understand the changes that occur as a result of cellular differentiation in the nerve. The ganglioside composition of the trigeminal nerve was entirely different from that of brain. The major gangliosides in adult trigeminal nerve were GM₃, GD₃, and LM₁ (sialosyl-lactoneotetraosylceramide or sialosylparagloboside). The structure of LM₁ and other gangliosides was established by enzymatic degradation and by analysis of the products of acid hydrolysis. At 2 days after birth, when the Schwann cells were immature, GM₃ and GD₃ were the major gangliosides in the nerve, 50 and 18 mol %, respectively. As the nerve developed and Schwann cells proliferated and myelinated the axons, the mol % of GM₃ and GD₃ reduced and that of LM₁ steadily increased. Polysialogangliosides did not change drastically with nerve development. The rate of deposition of LM₁ in the nerve with age was very similar to that of myelin marker lipids, cerebrosides, and sulfatides; thus, deposition appears to be localized mainly in the rat nerve myelin. LM₁ also had long-chain fatty acids 22:0 and 24:0, which are not usually found in CNS gangliosides. The ganglioside pattern of the rat trigeminal nerve was very similar to that of rat sciatic nerve, but was different from that of rabbit and chicken sciatic nerve. The activity of the two key enzymes involved in the metabolism of GM₃, viz., CMP-N-acetylneuraminic acid:lactosylceramide sialyltransferase and UDP-N-acetylgalactosamine:GM₃-N-acetylgalactosaminyltransferase, was also studied during development of the nerve and brain. The developmental profiles of both enzymes were consistent with the amounts of GM₃ present in the nerve.     

T(Iat)- and B(Iab)-cell alloantigens determined by theH-2 linkedI region in mice

The specificity of an antiserum directed againstI region associated (Ia) antigens is described. The serum was raised in (DBA/1×B10.D2)F₁ mice against lymphocytes of AQR mice, differing from the responder for theI region only. The serum reacts with Ia antigens expressed on B cells (Iab) as well as with Ia antigens expressed on T cells (Iat). Absorption studies indicate that B cells possess at least two Ia antigens, and one of these is shared by T cells. However, this shared antigen is not present on the surface of lymphocytes of thymectomized mice. Analysis of the strain distribution of Iab and Iat antigens revealed that the Iab antigens are present on lymphocytes of mice carrying theIA ^k subregion and that the Iat antigens are present on lymphocytes of mice carryingI region genes of theH-2 ^k haplotype located between theIA andIB subregions. This conclusion is based on the analysis of the antiserum's reactivity with T and B cells of the strains B10.A(2R), B10.A(4R) and B10.HTT: the serum reacts with B and T cells of B10.A(2R) but only with B cells of B10.A(4R) mice and only weakly with T cells of B10.HTT mice.Abbreviations ALG antimouse lymphocyte globulin from rabbits - B cells bone marrow derived lymphocytes - B10 C57BL/10Sn mice - D1D2F₁ (DBA/1×B10.D2)F₁ hybrid mice - GVHR graft-vs-host reaction - Ia I region associated antigen - Iab on B cells - Iat on T cells - MLR mixed lymphocyte reaction - T cells thymus-derived lymphocytes - Thy-1 thymus antigen 1, formerly called theta - Tx-Lyc lymphocytes of thymectomized, ALG treated, lethally irradiated and anti-Thy-1 treated bone marrow reconstituted mice - 2R B10.A(2R)/SgSn mice - 4R B10.A(4R) mice     

Mitogenic factor from inbred guinea pigs. II. Properties of the factor

Thymectomized adult rats which have been heavily irradiated and reconstituted with syngeneic bone marrow cells rapidly regain the ability to defend themselves against a primary infection with the intracellular bacterial parasite, Listeria monocytogenes. They do so by a cell-mediated immunological mechanism as evidenced by the protective immunity transferred adoptively by thoracic duct lymphocytes or peritoneal exudate cells from donors infected with this organism. But peritoneal exudate cells from thymus-derived donors convey only a fraction of the immunity transmitted by exudate cells from similarly infected intact rats. Since thymectomized irradiated animals can mobilize their cellular defenses more effectively when they are injected with a modest number of thoracic duct lymphocytes, an effect that cannot be duplicated with a massive infusion of bone marrow, it is argued that thymusdependent lymphocytes or T cells have an influential role in the development of cellular resistance to infection.     

Aberrant cAMP Metabolism in NF1 Malignant Peripheral Nerve Sheath Tumor Cells

Malignant peripheral nerve sheath (MPNST) cell lines derived from patients with neurofibromatosis type 1 (NF!) were found to have basal cAMP levels which are two-fold higher than cAMP levels in normal human adult Schwann cells (nHSC). PCR analysis also revealed that normal adult human Schwann cells express mRNA for types Ill, IV, and IX adenylyl cyclase (AC) while NF1 MPNST cells express AC mRNA of types II, V, and VIII in addition to expressing all the isoforms of normal adult human Schwann cells. Further PCR analysis revealed that NF1 MPNST lines express mRNA for EP2 and EP4 prostaglandin receptors whereas nHSC only express mRNA for the EP2 receptor. Exogenous prostaglandins alone or in combination with PDGF BB induced greater increases in cAMP levels and proliferation of NF1 MPNST cells compared to nHSC. We conclude that aberrant cAMP signaling in NF1 MPNST cells contributes to tumor formation in NF1 patients.     

Diagnostic Accuracy of PET/CT-Guided Percutaneous Biopsies for Malignant Peripheral Nerve Sheath Tumors in Neurofibromatosis Type 1 Patients

Background

Malignant peripheral nerve sheath tumors (MPNST) are one of the most frequent causes of death in patients with neurofibromatosis type 1 (NF1). Early detection is crucial because complete surgical resection is the only curative treatment. It has been previously reported that an ¹⁸F-fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) image with a T/L (Tumor/Liver) SUV_max ratio > 1.5 provides a high negative predictive value; however, it is not specific enough to make a NF1-related MPNST diagnosis. A formal proof of malignant transformation from a histological analysis is necessary before surgical excision because the procedure can cause mutilation. The objective of the present work was to investigate the effectiveness of and complications associated with PET/CT-guided percutaneous biopsies for an NF1-related MPNST diagnosis.

Methods

PET/CT-guided percutaneous biopsy procedures performed on 26 NF1 patients with a clinical suspicion of MPNST and a suspect lesion from a PET/CT scan (T/L SUV_max ratio > 1.5) were retrospectively evaluated. The localization of the suspected malignant site was determined using PET/CT. A stereotactic (ultrasonic and CT control) core biopsy technique was used with a local anesthesia.

Results

The first PET/CT-guided percutaneous biopsies enabled a pathological diagnosis for all of the patients (no "inconclusive " results were obtained), and no secondary procedures were needed. Among the 26 patients, the histopathological results from the biopsy were malignant in 17 cases and benign (BPNST with atypical cells) in nine cases. No complications from the diagnostic procedure were observed. A surgical resection was performed in 18 patients (seven benign and 11 malignant biopsies), removing the fine needle biopsy scar. In addition, six locally advanced/metastatic MPNST were treated with chemo/radiotherapy, and two BPNST had no progression after a follow-up of 14 and 39 months, respectively. The PET/CT-guided percutaneous biopsy gave 25 accurate diagnoses and one false negative (BPNST with atypical cells on the biopsy and MPNST on the operated tumor), resulting in a diagnostic accuracy rate of 96%. This false negative case may be explained by the high heterogeneity of the tumor: benign areas were contiguous with the malignant ones and associated with inflammation.

Conclusions

PET/CT-guided percutaneous biopsies are an effective and relatively non-traumatic procedure for diagnosis of NF1-related MPNST. It is the most reliable approach for early detection of MPNST.     

Studies of nonmigrating long-lived lymphocytes in rats. V. Presence and reactions to antigens after neonatal thymectomy

Neonatally thymectomized and sham-operated rats were given courses of hind foot pad injections of tritiated thymidine after immune stimuli, and the popliteal lymph nodes were examined for the presence of persistently labeled lymphocytes 32 days later. A larger proportion of labeled small lymphocytes was found in the nodes of thymectomized rats than in the nodes of sham-operated rats, indicating that a significant portion of the long-lived nonmigrating cells are B cells. In sham-operated rats given a course of tritiated thymidine after an immune stimulus with a Salmonella flagellin, labeled blasts appeared 3 days after secondary challenges with both the same or a serologically distinct flagellin. In neonatally thymectomized rats, a blastogenic response occurred after challenge with the same flagellin but not after challenge with the non-specific flagellin. Thus, in normal rats there are also long-lived nonmigrating T cells in primed lymph nodes, and these are the cells responsible for the nonspecific blasto-genesis seen in earlier experiments.     

Regulatory function of sorbose-resistance gene C in sugar transport of Neurospora

Summary A sorbose-resistant double mutant sor ^r A-10/sor ^r C-17 produces larger colonies in sorbose containing test-medium than the respective single mutants; wildtype colonies remain very small. Resistance of the single mutants was shown to be connected with a decreased rate of sorbose-uptake into their conidia; however, sorbose uptake of the double mutant had not been measured. To check, whether the improved performance of the double mutant on test medium is correlated with a further decrease of sorbose uptake in this strain, studies on the uptake of fructose, sorbose and deoxyglucose by ungerminated conidia of the two single mutants, the double mutant and the wildtype were conducted, using C¹⁴-marked sugars, the millipore filter technique, and conidia either untreated or pretreated with 1% sorbose for 4 hours.If sorbose uptake is referred to that of fructose as basis of calculations, as in the earlier studies, the sorbose uptake by cells of the double mutant is smaller than that of both single mutants for conidia not pretreated with sorbose (Fig. 7a). However, for conidia pretreated with sorbose, this correlation does not hold. Rather, cells of the double mutant take up less sorbose than those of the C-mutant, but as much or slightly more than those of the A-mutant (Fig. 7b). If sorbose uptake is referred to that of deoxyglucose for an independent point of reference, cells of the double mutant take up less sorbose than those of the C-mutant, but much more than those of the A-mutant. This holds for untreated and sorbose pretreated cells (Fig. 5 a and b). These data rule out a correlation between colony size and transport defect for at least one of the strains used here, i.e. the C-mutant.The following data suggest a new interpretation: In contrast to the earlier findings with germinated conidia, ungerminated untreated cells of the C-mutant take up much more fructose and sorbose than those of the wildtype (Fig. 3 a and 1a). The uptake of fructose by cells of the C-mutant can not be improved by sorbose pretreatement (Fig. 3 b), but in both wildtype and A-mutant it is increased (Figs. 1b and 2b). Uptake of deoxyglucose was nearly equal for all three strains either untreated or pretreated. Untreated cells of the A-mutant take up as much sorbose as those of the wildtype (Figs. 2 a and 1 a). On pretreatment their sorbose uptake remains nearly constant (Figs. 2b), in contrast to wildtype cells, where it increases drastically and without an increase of fructose uptake by an equivalent amount (Fig. 1b).The new interpretation suggests that gene C is of the regulator type. Mutation of it in the C-strain used here has lead to the simultaneous de-repression of a system for fructose and sorbose uptake. Deoxyglucose uptake is not served by this system. Gene A is a structural gene, harbouring the information for the inducible synthesis of a carrier or permease specifically engaged in sorbose uptake. It is not under the controll of gene C.This interpretation is supported by results on untreated cells of the double mutant. However, fructose uptake of such cells is roughly equal to that of C-mutant cells (Fig. 6a) and sorbose uptake is less (Fig. 5a). Hence, a secondary effect of the A-gene, i.e. on the amount of de-repression of sorbose uptake by mutation in gene C, is indicated.     

Characterization of two electrophoretic lactate dehydrogenase-A mutants inMus musculus

Two lactate dehydrogenase (LDH) mutations were recovered independently among offspring of ethylnitrosourea-treated male mice by screening for alterations of isoelectric focusing pattern in liver homogenates. Investigations of physicochemical and kinetic properties of the mutant enzymes indicated that the mutant traits resulted from point mutations at theLdh-1 structural locus. Therefore, the new alleles were designatedLdh-1 ^a-m5Neu andLdh-1 ^a-m6Neu, respectively. Both mutant alleles code for proteins which exhibit an altered stability to heat, in addition to changes in isolectric focusing pattern and a reduction in anodal electrophoretic mobility. While LDH-A^a-m5Neu proteins are markedly less heat stable, LDH-A^a-m6Neu proteins are more heat stable than the wild-type enzyme. Furthermore, a small elevation ofK _m for pyruvate, a slightly reduced inhibition by high pyruvate concentrations, and a slight acidic shift of the pH activity profile distinguish LDH-A^a-m6Neu from both wild-type and LDH-A^a-m5Neu enzymes. Significant alterations of LDH activity were detected in some tissues from LDH-A^a-m5Neu individuals but not in those from LDH-A^a-m6Neu animals. Erythrocytes and blood of LDH-A^a-m5Neu mutants revealed activity levels which were reduced by approximately 6 and 13% compared with those of wild types in heterozygous and homozygous individuals, respectively. In addition, an elevation of approximately 6% in LDH activity was found in skeletal muscle in homozygous mutants. Consistent with the unaltered or only slightly altered LDH activity in tissues, the genetic as well as the physiological characterization yielded no easily detectable effects from either mutation on metabolism or fitness of the affected individuals.This research was supported in part by Contract BI6-156-D from the Commission of the European Communities.     

The tumor suppressor RECK interferes with HER-2/Neu dimerization and attenuates its oncogenic signaling

Our previous study demonstrates that HER-2/Neu oncogene inhibits a matrix metalloproteinase inhibitor and tumor metastasis suppressor RECK to promote metastasis. Conversely, the effect of RECK on the oncogenic function of HER-2/Neu is unknown. Ectopic expression of RECK in 293T cells and HER-2/Neu-overexpressing breast cancer cells shows that RECK and HER-2/Neu are co-localized and these two proteins can be co-immunoprecipitated. RECK inhibits HER-2/Neu receptor dimerization and autophosphorylation, which causes reduction of ERK and AKT kinase activity and down-regulation of HER-2/Neu target genes. RECK expression is reduced in 58.8% of breast cancer tissues and is associated with lymph node invasion supporting its anti-metastatic role. Collectively, we provide the first evidence that RECK can negatively regulate oncogenic activity of HER-2/Neu by inhibiting receptor dimerization.

Structured summary

HER-2/Neuphysically interacts with HER-2/Neu by blue native page (View interaction)HER-2/Neuphysically interacts with RECK by coimmunoprecipitation (View interaction)HER-2/Neu and RECKcolocalize by fluorescence microscopy (View Interaction 1, 2)HER-2/Neuphysically interacts with RECK by anti bait coimmunoprecipitation (View interaction)     

Comparative study of binding of ovine complement factor H with different <Emphasis Type="Italic">Borrelia</Emphasis> genospecies

This study presents the binding of ovine factor H (fH) by various serotypes of Borrelia and simultaneously correlates their complement resistance to sheep serum. Affinity ligand binding assay was employed to study the binding of borrelial proteins to ovine recombinant fH and its truncated forms (short consensus repeat, SCR 7 and SCRs 19–20). From a repertoire of 17 borrelial strains, only two strains showed affinity to sheep fH. A ~28-kDa protein of Borrelia burgdorferi sensu stricto (B. burgdorferi s.s., strain SKT-2) bound full-length fH as well as SCRs 19–20. This fH-binding protein was further identified as complement regulator-acquiring surface protein of B. burgdorferi (BbCRASP-1) by MALDI-TOF analysis. Surprisingly, a ~26-kDa protein of Borrelia bissettii (DN127) showed affinity to full-length fH but not to SCR 7 and SCRs19–20. In complement sensitivity assay, both strains—SKT-2 and DN127—were resistant to normal sheep serum. Significant complement resistance of two Borrelia garinii strains (G117 and T25) was also observed; however, none of those strains was able to bind sheep fH. Our study underscores the need of further exploration of fH-mediated evasion of complement system by Borrelia in domestic animals.     

Use of a biotinylated probe and in situ hybridization for light and electron microscopic localization of P 0 mRNA in myelin-forming Schwann cells

Summary A biotinylated P ₀ glycoprotein cDNA was hybridized in situ to aldehyde-fixed vibratome sections and to aldehyde-fixed thin sections of Lowicryl-embedded trigeminal ganglia of 15 day old rats. Alkaline phosphatase and peroxidase detectors were used for light microscopic (LM) studies and peroxidase or colloidal gold were employed for electron microscopic (EM) detection. In both LM and EM sections, probe was found in cytoplasmic areas of myelinforming Schwann cells that were enriched in granular endoplasmic reticulum, demonstrating that these regions contain P ₀ mRNA. Interestingly, P ₀ mRNA tended to cluster in regions close to the developing myelin sheath. Relatively simple methods are here described for EM detection of mRNA with reasonable tissue preservation and high resolution. These methods may be useful for developmental and disease-related studies of specific mRNAs in mammalian tissues.     

In vitro formation of the Merkel cell‐neurite complex in embryonic mouse whiskers using organotypic co‐cultures

A Merkel cell‐neurite complex is a touch receptor composed of specialized epithelial cells named Merkel cells and peripheral sensory nerves in the skin. Merkel cells are found in touch‐sensitive skin components including whisker follicles. The nerve fibers that innervate Merkel cells of a whisker follicle extend from the maxillary branch of the trigeminal ganglion. Whiskers as a sensory organ attribute to the complicated architecture of the Merkel cell‐neurite complex, and therefore it is intriguing how the structure is formed. However, observing the dynamic process of the formation of a Merkel cell‐neurite complex in whiskers during embryonic development is still difficult. In this study, we tried to develop an organotypic co‐culture method of a whisker pad and a trigeminal ganglion explant to form the Merkel cell‐neurite complex in vitro. We initially developed two distinct culture methods of a single whisker row and a trigeminal ganglion explant, and then combined them. By dissecting and cultivating a single row from a whisker pad, the morphogenesis of whisker follicles could be observed under a microscope. After the co‐cultivation of the whisker row with a trigeminal ganglion explant, a Merkel cell‐neurite complex composed of Merkel cells, which were positive for both cytokeratin 8 and SOX2, Neurofilament‐H‐positive trigeminal nerve fibers and Schwann cells expressing Nestin, SOX2 and SOX10 was observed via immunohistochemical analyses. These results suggest that the process for the formation of a Merkel cell‐neurite complex can be observed under a microscope using our organotypic co‐culture method.     

Fine Structure of Subepithelial “Free” and Corpuscular Trigeminal Nerve Endings in Anterior Hard Palate of the Rat

Axonally transported protein labeled many trigeminal nerve endings in subepithelial regions of the anterior hard palate of the rat. Sensory endings were most numerous in the lamina propria near the tips of the palatal rugae where large connective tissue and epithelial papillae interdigitated. Two kinds of sensory ending were found there: “free” endings, and a variety of corpuscular endings. The “free” sensory endings consisted of bundles of unmyelinated axons separated from the connective tissue by relatively unspecialized Schwann cells covering part or all of their surface and a completely continuous basal lamina; they were commonly found running parallel to the epithelium or near corpuscular endings. The corpuscular sensory endings all had a specialized nerve form, specialized Schwann cells, and axonal fingers projecting into the corpuscular basal lamina or connective tissue. There were at least four distinct types of corpuscular ending: Ruffini-like endings were found among dense collagen bundles, and they had a flattened nerve ending with a flattened Schwann lamella on either side. Meissner endings had an ordered stack of flattened nerve terminals with flattened Schwann cells and much basal lamina within and around the corpuscle. Simple corpuscles were single nerve endings surrounded by several layers of concentric lamellar Schwann processes. Glomerular endings were found in lamina propria papillae or encircling epithelial papillae; they were a tangle of varied neural forms each of which had apposed flattened Schwann cells, and a layer of basal lamina of varied thickness. Fibroblasts often formed incomplete partitions around Meissner and simple corpuscles.

The axoplasm of all kinds of subepithelial sensory endings contained numerous mitochondria and vesicles, as well as occasional multivesicular bodies and lysosomes; the axoplasm of all endings was pale with few microtubules and neurofilaments. The specialized lamellar Schwann cells had much pinocytotic activity. Four kinds of junctions were found between the corpuscular sensory endings and the lamellar Schwann cells: (1) symmetric densities that resemble desmosomes; (2) asymmetric densities with either the neuronal or glial membrane more dense; (3) neural membrane densities adjacent to Schwann parallel inner and outer membrane densities; and (4) sites of apparent Schwann endocytosis associated with neural blebs. The “free” sensory endings only made occasional desmosome-like junctions with their Schwann cells.

These observations are discussed in relation to possible mechanosensory transduction mechanisms, with particular attention to axoplasmic structure, axonal fingers, and neural and nonneural cell associations.