Immunocytochemical localization of nerve growth factor in mouse salivary glands

Reversible Conversion among Subtypes of Salivary Gland Duct Cells as Identified by Production of a Variety of Bioactive Polypeptides

Four major kallikreins (mK1, mK22, mK9, and mK13) were identified in the mouse submandibular gland (SMG). mK1, a true tissue kallikrein, was used as a protein marker to identify different types of SMG granular convoluted tubule (GCT) cells along with epidermal growth factor (EGF), nerve growth factor (NGF), and renin. Kallikrein mK1 was localized in a very small number (~5%) of GCT cells, which were scattered throughout the GCT, indicating that the majority of GCT cells are mK1-negative. Among mK1-positive cells, particularly strong signals were observed in a small number of narrow cells, recognized as slender granular cells (SG cells, Type IV), in the GCT. After postnatal development of the SMG, GCT cells are no longer uniform based on the bioactive substances (mK1, EGF, NGF, and renin) that they produce and secrete. GCT cells were classified into four subtypes, Types I–IV, and it became clear that these subtypes are complicatedly and reversibly converted by the endocrine hormones 5α-dihydrotestosterone (DHT) and triiodothyronine (T₃). Duct segments with similar morphology or hormone dependency were recognized in the sublingual and parotid glands. The presence of duct cells with such characteristics is therefore a common feature of the three major salivary glands of rodents.

Immunocytochemical study of granular duct cells with a hormonally enhanced granular cell phenotype in the mouse parotid gland

In the parotid glands (PGs) of intact male mice (12 weeks of age, ICR strain), immunofluorescence labels for a true tissue kallikrein, mK1, and for nerve growth factor (NGF) were recognized through the subluminal edges of the striated duct (SD) segments and interlobular duct segments. Because of their small size, secretory granules were not detectable by light microscopy in any of the duct cells. Full-fledged granular cells, containing large secretory granules that were visible by light microscopy, were induced in the SD segments of male mice after the injection of 5alpha-dehydrotestosterone (DHT) and triiodothyronine (T(3)), given either alone or in combination every other day for 2 weeks. A stronger effect was detected in the mice that were concomitantly injected with DHT and T(3), and more abundant, fully developed granular cells appeared in the SD segments of these mice. These full-fledged granular cells were immunoreactive for mK1, NGF, and epidermal growth factor, but not for renin. The present results indicate that some of the SD cells with small granules in the mouse PG can develop a granular cell phenotype, producing more kinds of growth factors, as a result of the actions of androgen and thyroid hormone.

Hypophysectomy and hormonal therapy modulate mK1-immunoreactive duct cells in the mice sublingual glands

The immunocytochemical localization of a true tissue kallikrein, mK1, in mouse sublingual glands (SLGs) was examined following hypophysectomy and hormonal replacement therapy. In the glands of intact mice (14 weeks of age), mK1 was detected in the striated ducts (SDs). Full-fledged granular cells were scattered in the SDs of male mice (but not in those of female mice), showing a cellular mosaic distribution of mK1 with some being positive and others being negative. mK1 was also detected in transitional-type granular cells, though the secretory granules were too small and scarce to be visible by a light microscopy. Hypophysectomy in male mice resulted in the atrophy and loss of secretory granules in many SD cells. Granulation recovered after the repeated injection of 5alpha-dihydrotestosterone (DHT), 3,5,3'-triiodo-L: thyronine (T3), and dexamethasone (Dex), given either alone or in combination to the hypophysectomized mice. The concomitant injection of DHT and T3, with or without Dex, resulted in the reappearance of the full-fledged granular cells, only some of which were mK1-positive. Electron microscopy revealed mK1 to be present exclusively in the secretory granules of these mK1-positive cells, and no ultrastructural differences were observed between mK1-positive and mK1-negative full-fledged granular cells. These results show that the differentiation of the granular cell phenotype in the mouse SLG duct system requires the concomitant action of androgen and thyroid hormone and retards mK1 synthesis.

Repeated androgen and thyroid hormone injection modulates the morphology of hormone-responsive duct cells in the mouse parotid gland

When the parotid glands of normal male and female ICR mice (12 weeks of age) were examined under a light microscope, no granular cells were seen in the duct system. However, transmission electron microscopy revealed that, in both sexes, many striated duct cells contained a few electron-dense secretory granules in their subluminal cytoplasm and had formed so-called granular striated tubules (GSTs) in some of the striated duct segments. These secretory granules were not large enough to be visible with a light microscope. Fully fledged granular cells, containing large secretory granules visible with a light microscope, could be induced in the GST segments of the glands of males by injection with 5alpha-dihydrotestosterone (DHT), triiodothyronine (T(3)), and dexamethasone (Dex), given alone or in combination every other day for 2 weeks. Dex alone showed no effect on the GSTs in this study. Both DHT and T(3), either individually or with Dex, were moderately effective, inducing a few scattered fully fledged granular cells. A stronger effect was detected after concomitant injection of DHT and T(3), with or without Dex, with more abundant fully developed granular cells appearing in the GST segments. Electron microscopy revealed that these granular cells had abundant large secretory granules in their apical two-thirds, a basal nucleus, and modest basal infoldings. By contrast, the effect of the same hormones was very weak in the glands of females, and even the concomitant injection of DHT and T(3), with or without Dex, rarely induced fully fledged granular cells. These results indicate a close similarity between the ductal systems of the major salivary glands of the mouse, in terms of some of the striated duct segments containing secretory granules, being under the same multihormonal regulation, and being sexually dimorphic.

Evidence of a phenotypically determined ductal cell lineage in mouse salivary glands

The submandibular salivary gland of mice contains a parenchymal element, the granular duct, which matures peripubertally from the striated ducts. Granular duct cells also differentiate from intercalated ducts in the adult mouse submandibular gland. Using preproNGF-A as a signature protein of mature granular duct cells, this study inquired if phenotypic determination might have occurred earlier than the first signs of cellular differentiation. Results from RT-PCR indicate the presence of preproNGF-A transcripts at all postnatal stages of development of the submandibular glands, as well as in adult sublingual glands. The preproNGF-A transcript was also detected prenatally as early as embryonic day 17 in the submandibular/sublingual complex. Using an antibody directed specifically against the "pre" peptide, immunocytochemistry showed preproNGF-A localized in the granular ducts and striated ducts of the adult submandibular gland. In addition preproNGF-A was detected throughout the first order branches of the intercalated duct system. In the neonatal gland, preproNGF-A was found in the large tubules that differentiate to the striated ducts. The early appearance of preproNGF-A in the histological lineage that sequentially gives rise to striated ducts and then to granular ducts suggests that this lineage is phenotypically determined as early as birth. An undifferentiated stage of the phenotypically determined lineage also appears to be retained in the intercalated duct system to provide progenitors for subsequent differentiation in the adult gland. Throughout development of the sublingual gland, preproNGF-A was detectable in the striated ducts or in their predecessors, suggesting that they may also represent a phenotypically determined cell lineage similar to that of the submandibular gland.

Satellite-cell-derived nerve growth factor and neurotrophin-3 are involved in noradrenergic sprouting in the dorsal root ganglia following peripheral nerve injury in the rat

Injury to a peripheral nerve induces in the dorsal root ganglia (DRG) sprouting of sympathetic and peptidergic terminals around large-diameter sensory neurons that project in the damaged nerve. This pathological change may be implicated in the chronic pain syndromes seen in some patients with peripheral nerve injury. The mechanisms underlying the sprouting are not known. Using in situ hybridization and immunohistochemical techniques, we have now found that nerve growth factor (NGF) and neurotrophin-3 (NT3) synthesis is upregulated in satellite cells surrounding neurons in lesioned DRG as early as 48 h after nerve injury. This response lasts for at least 2 months. Quantitative analysis showed that the levels of mRNAs for NT3 and NGF increased in ipsilateral but not contralateral DRG after nerve injury. Noradrenergic sprouting around the axotomized neurons was associated with p75-immunoreactive satellite cells. Further, antibodies specific to NGF or NT3, delivered by an osmotic mini-pump to the DRG via the lesioned L5 spinal nerve, significantly reduced noradrenergic sprouting. These results implicate satellite cell-derived neurotrophins in the induction of sympathetic sprouting following peripheral nerve injury.

Immunohistochemical localization of neurotrophins and neurotrophin receptors in human and mouse salivary glands

This study was undertaken to analyze the occurrence of low- (p75) and high-affinity (TrkA, TrkB and TrkC) neurotrophin receptor proteins in human and mouse salivary glands using immunohistochemistry. Furthermore, the presence of neurotrophins was also investigated. The study was carried out on 14 human (4 parotid, 6 submandibular and 4 sublingual glands) and 5 mouse salivary glands, using polyclonal antibodies against Trk proteins. The intensity of immunostaining was calculated automatically and evaluated in arbitrary units of grey levels. In human tissues no immunoreactivity (IR) for the assessed antigens was observed in the serous or mucous acinar cells, although TrkA IR was found in the acini of the submandibular gland. The cells of the intercalated ducts showed p75 IR (sublingual) and TrkA IR (parotid gland). The striated and excretory ducts displayed p75 IR, TrkA IR and TrkC IR in all glands, but TrkB IR was never detected. No neurotrophins were detected. In the mouse glands the ductal cells display IR for p75 (submandibular) and Trks A and C (parotid and submandibular) but not the sublingual gland. Acinar cells of the submandibular gland also show p75 IR. The only neurotrophin found in the mouse salivary glands was NGF (submandibular gland). These results suggest that neurotrophins may be involved in controlling the physiology of epithelial salivary cells.

Immunocytochemical localization of seminal proteins in salivary and lacrimal glands of the rat

Antibodies against 10 different secretory proteins from the accessory sex glands of the male rat were used for immunohistochemical studies of salivary and lacrimal glands from intact and castrated rats, at the light- and electron-microscopic levels. In the parotid gland, secretory acinar cells showed immunoreactivity with antibodies against prostatic binding protein, cystatin-related peptide and acid phosphatase (isoenzyme pI 8.0; 5.6) typical of ventral prostate, and seminal vesicle secretion VI. Western blotting analysis indicated that immunoreactivity against prostatic binding protein was attributable to a subunit, presumably C3. Acid phosphatase pI 5.6 showed a molecular weight of 66 kDa, which is at variance with the prostatic form. Immunoreactivity for secretory transglutaminase, derived from the coagulating gland, was restricted to myoepithelial and stromal cells. In castrated animals, the immunoreactivity of acinar cells was reduced to the background level, whereas stromal transglutaminase immunoreactivity was unaltered. The distribution pattern of immunoreactivity for the proteins mentioned was almost identical in the lacrimal gland. Significant differences were however observed in the immunoreactivity of the inframandibular gland, where serous glandular cells were non-immunoreactive for seminal proteins, with the exception of acid phosphatase isoenzyme pI 8.0. Granules present in the convoluted granular ducts were immunoreactive particularly for acid phosphatase (isoenzyme pI 5.6) but much less for cystatin-related peptide; immunoreactivity was reduced after castration. The straight portion of the inframandibular duct system was immunoreactive for transglutaminase, but no influence of castration was visible. The distribution of immunoreactivity for seminal proteins present in the salivary and lacrimal glands and the pronounced androgen-dependence of their expression point to functional relationships of the respective proteins at both glandular sites.

Immunohistochemical survey of cobalamin-binding proteins

Dietary cobalamin is bound to intrinsic factor (IF) and nonintrinsic factor (R protein) in the gastrointestinal tract. Distribution of IF has previously been described only in the stomach (parietal, chief cells), duodenum (mucous cells), and pancreatic juice. Cells containing R protein have been reported to be more widespread, including mucin-secreting cells of the intestinal tract. To further clarify the cells producing these cobalamin binders in the various organs, we surveyed a number of tissues in four species, with monospecific antisera raised against rat IF and hog R protein, using indirect immunoperoxidase histochemistry. A positive staining pattern was confirmed after blocking with the appropriate protein purified to homogeneity. Strong IF staining was noted over chief cells in the rat stomach, with weaker but substantial staining in the parietal cells as well. Intrinsic factor was found only in parietal cells in all other species examined. R protein was found in parietal cells of all species tested, except in the hog, where mucous cells of the basal crypt stained. The area of the brush border and apical cytoplasm in small intestinal epithelial cells was also positive for both IF and R protein, probably reflecting uptake. Accessory digestive organs, such as salivary gland and pancreas were also positive for both IF and R protein in the secretory ducts. Accessory digestive organs contribute to the secretion of IF as well as R protein, but the physiologic importance of this IF secretion is not clear.

Characterization of antibodies to synthetic nerve growth factor (NGF) and proNGF peptides

Sequence data for the mature nerve growth factor (NGF) protein and its precursor are available from molecular cloning of the NGF gene in several species, including mice, humans, rats, and chickens. Hydrophilicity analysis of the predicted rat and chicken prepro-NGF was carried out to locate putative antigenic determinants. Eight peptides were selected and synthesized based on hydrophilicity profiles. Two peptides represent sequences in the rat (and mouse) pro-NGF, one peptide (our peptide P3) represents a highly conserved region of the mature NGF protein (identical in humans, mice, rats, and chickens), two peptides are specific for the mature chicken NGF, and the remaining three peptides are specific for the mature rat NGF (each with only one amino acid substitution compared with corresponding segments of the mouse NGF). For immunization, the peptides were conjugated to keyhold limpet hemocyanin and used to produce antisera in rabbits. After bleeding, peptide-specific antibodies were purified on affinity columns prepared by coupling each of the synthetic peptides. The different peptide antisera and affinity-purified antibodies then were characterized by enzyme-linked immunoassay (ELISA) and immunohistochemistry of the male mouse submandibular gland, a rich exocrine source of NGF. ELISA analysis showed that all peptide antisera bound two to four orders of magnitude better than normal rabbit serum to a coat of their proper peptide. The higher binding was retained by the purified peptide antibodies compared with normal rabbit immunoglobulin. Specific tests, in which one peptide antiserum was checked against different peptide coats in the ELISA, also showed two to four orders of magnitude higher binding of antibodies to the proper synthetic peptide. The peptide antibodies also were tested for their ability to bind to native mouse beta NGF coated to the immunoplates. Only antibodies raised to the conserved P3 peptide recognized native NGF to an extent similar to that obtained with polyclonal anti-NGF antibodies. Conversely, P3 was well recognized by several different NGF antisera. Immunohistochemically, both peptide antisera against the pro-NGF stained the perinuclear cytoplasm in the basal part of the cells of the granulated convoluted tubules in the mouse submandibular gland.(ABSTRACT TRUNCATED AT 400 WORDS)

Detection of nerve growth factor and its mRNA by separate and combined immunohistochemistry and in situ hybridization in mouse salivary glands

Intense labelling of secretory cells in the male mouse submandibular gland was observed after in situ hybridization using mouse nerve growth factor (NGF) cDNA probes. Under the same conditions, sparse less intensely labelled cells were also found in the sublingual gland. Hybridization to a chicken NGF cDNA probe gave weak labelling on the glands in accordance with a weak cross-hybridization between mouse NGF mRNA and chicken NGF cDNA probes, whereas no labelling was seen using pUC9 DNA as a hybridization probe. A combination of in situ hybridization and immunohistochemistry was also carried out on the same sections of submandibular gland. A good correlation was seen between actively synthesizing and intensely immunoreactive cells in the gland. The technique described here allows the detection of individual cells synthesizing relatively low levels of NGF. The combination of in situ hybridization and immunocytochemistry on the same section should be particularly useful in cases where NGF is transported away from its site of synthesis.