Peptide 19 in the rat superior cervical ganglion.

Peptide 19 is a 7.6 kDa polypeptide which can bind to calmodulin and inhibit calcium-calmodulin signaling. In this study, peptide 19-immunoreactivity was examined in the rat superior cervical ganglion. In the ganglion, 54.8% of postganglionic sympathetic neuron profiles were immunoreactive for peptide 19. These neuron profiles were small- to medium-sized and measured 87-845 microm(2) (mean+/-SD = 343+/-111 microm(2)). Double immunofluorescence method revealed that 99.9% of peptide 19-containing neurons had neuropeptide Y in the superior cervical ganglion. Retrograde neuronal tracing and immunohistochemical studies also demonstrated that peptide 19 was common in postganglionic sympathetic neurons which innervated the facial skin and masseter but not the submandibular gland; 55.6% and 75.2% of cutaneous and muscular neuron profiles, respectively, contained peptide 19. Only 9.8% of glandular neurons were immunoreactive for peptide 19. These findings indicate that the content of peptide 19 in superior cervical ganglion neurons depends on their cell sizes and peripheral projections. On the other hand, colchicine injection into the superior cervical ganglion decreased the number of peptide 19-positive neurons (30.7%) compared to saline injection (53.3%). In contrast, the treatment induced nicotine adenine dinucleotide phosphate diaphorase activity in 12.7% of postganglionic sympathetic neurons. Double stain demonstrated that 56.3% of nicotine adenine dinucleotide phosphate diaphorase-positive neurons co-expressed peptide 19. These findings indicate that colchicine treatment causes decrease of peptide 19 expression and increase of nitric oxide synthase activity.

Expression of human beta-defensin -1, -2, and -3 in non-inflamed pseudocyst, mucoceles.

OBJECTIVES AND DESIGN: The expressions of human beta defensin-1 (HBD-1), -2 (HBD-2) and -3 (HBD-3) in non-inflamed pseudocysts such as mucoceles were investigated immunohistochemically in this study. MATERIALS AND METHODS: Mucocele specimens were obtained from 21 patients. The expression of HBDs was studied immunohistochemically by using antibodies directed against HBD-1, -2, and -3. Statistical analyses were carried out on serial sections stained with antibodies. RESULTS: Cells expressing HBDs were found in mucoceles. The expression of HBD-2 was observed in floating cells in all the specimens, whereas HBD-1 and HBD-3-expressing cells were detected in 93% and 73% of the mucoceles, respectively. The HBD-2 signal was the most intense and the HBD-3 signal intensity was weaker than that of HBD-1. HBDs were expressed in neutrophils and in other floating cells. Interestingly, the signal intensity and the population of positive cells located close to the centers of cysts were higher than those located in the peripheral areas of cysts. CONCLUSION: The expression of HBDs was found even in non-inflamed pseudocysts such as mucoceles. These results suggest that an unknown mechanism not involved in biophylaxis for the expression of HBDs may exist.

The number of nociceptors in the trigeminal ganglion but not proprioceptors in the mesencephalic trigeminal tract nucleus is reduced in dystonin deficient dystonia musculorum mice.

The trigeminal ganglion (TG) and mesencephalic trigeminal tract nucleus (Mes5) were investigated in wild type and dystonia musculorum (dt) mice to study the effect of dystonin deficiency on primary sensory neurons in the trigeminal nervous system. At postnatal day 14, the number of TG neurons was markedly decreased in dt mice when compared to wild type mice (43.1% reduction). In addition, dystonin disruption decreased the number of sensory neurons which bound to isolectin B4, and contained calcitonin gene-related peptide or high-affinity nerve growth factor receptor TrkA. Immunohistochemistry for caspase-3 demonstrated that dystonin deficiency induced excess cell death of TG neurons during the early postnatal period. In contrast, Mes5 neurons were barely affected in dt mice. These data together suggest that dystonin is necessary for survival of nociceptors but not proprioceptors in the trigeminal nervous system.

Activation of microglia and p38 mitogen-activated protein kinase in the dorsal column nucleus contributes to tactile allodynia following peripheral nerve injury.

The activation of glial cells in the CNS has been suggested to be involved in abnormal pain sensation after peripheral nerve injury. Previous studies demonstrated phosphorylation of p38 mitogen-activated protein kinase (MAPK) in spinal cord glial cells after peripheral nerve injury, and such phosphorylation has been suggested to be involved in the development of neuropathic pain. The aim of this study was to examine the dorsal column nuclei for phosphorylation of p38 MAPK following peripheral nerve injury and to explore a possibility of its contribution to neuropathic pain. Immunohistochemical labeling for phosphorylated p38 (p-p38) MAPK was performed in histological sections of the rat spinal cord and medulla oblongata after the fifth lumbar (L5) spinal nerve ligation (SNL). The number of p-p38 MAPK-immunoreactive (IR) cells was significantly increased in the L5 dorsal horn and the gracile nucleus ipsilateral to the injury at days 3-21 after SNL. Double immunofluorescence labeling with cell-specific markers revealed that p-p38 MAPK-IR cells co-expressed OX-42, suggesting their microglial identity. Increased immunofluorescence labeling for OX-42 indicated that microglial cells were activated by SNL in the L5 dorsal horn and the gracile nucleus ipsilateral to the injury. Continuous infusion of a p38 MAPK inhibitor into the cisterna magna for 14 days beginning on the day of SNL suppressed the development of tactile allodynia, but not thermal hyperalgesia induced by nerve injury. These results demonstrate that SNL activates p38 MAPK pathway in microglia in the gracile nucleus as well as in the spinal cord dorsal horn. Activation of p38 MAPK in medullary microglia may contribute to the pathogenesis of neuropathic pain.

Brain-derived neurotrophic factor-immunoreactive neurons in the rat vagal and glossopharyngeal sensory ganglia; co-expression with other neurochemical substances.

Immunohistochemistry for brain-derived neurotrophic factor (BDNF) was performed on the rat vagal and glossopharyngeal sensory ganglia. In the jugular, petrosal and nodose ganglia, 56.1+/-5.5%, 52.4+/-9.4% and 80.0+/-3.0% of sensory neurons, respectively, were immunoreactive for BDNF. These neurons were small- to medium-sized and observed throughout the ganglia. In the solitary tract nucleus, the neuropil showed BDNF immunoreactivity. A double immunofluorescence method demonstrated that BDNF-immunoreactive neurons were also immunoreactive for calcitonin gene-related peptide (CGRP), P2X3 receptor, the capsaicin receptor (VR1) or vanilloid receptor 1-like receptor (VRL-1) in the jugular (CGRP, 43.5%; P2X3 receptor, 51.1%; VR1, 71.7%; VRL-1, 0.5%), petrosal (CGRP, 33.2%; P2X3 receptor, 58.4%; VR1, 54.2%; VRL-1, 23.3%) and nodose ganglia (CGRP, 1.8%; P2X3 receptor, 49.1%; VR1, 70.7%; VRL-1, 11.5%). The co-expression with tyrosine hydroxylase was also detected in the petrosal (2.9%) and nodose ganglia (2.2%). However, BDNF-immunoreactive neurons were devoid of parvalbumin in these ganglia. The present findings suggest that BDNF-containing vagal and glossopharyngeal sensory neurons have nociceptive and chemoreceptive functions.

The reduction of proprioceptors in the mesencephalic trigeminal tract nucleus after neonatal masseteric nerve transection; effect of brain-derived neurotrophic factor.

The effect of neonatal masseteric nerve transection on primary proprioceptors was examined in the mesencephalic trigeminal tract nucleus (Mes5) of the rat. At 72 h to 21 days after the injury, the number of Mes5 neurons decreased on the side ipsilateral to the transection. The means+/-SD of percentage proportion of ipsilateral/contralateral neurons at 72 h and 21 days were 69.9+/-7.5% and 58.2+/-14.6%, respectively. The application of brain-derived neurotrophic factor to the proximal stump of the masseteric nerve delayed the loss of Mes5 neurons at 72 h after the injury; the mean numbers+/-SD of ipsilateral and contralateral Mes5 neurons in injured animals with BDNF application was 553.6+/-61.9 and 558.4+/-55.3, respectively. Saline application had no effect on the injury-induced loss of Mes5 neurons; i.e., the mean numbers+/-SD of ipsilateral and contralateral Mes5 neurons were 367.3+/-72.5 and 543+/-33.5, respectively. These findings indicate that trigeminal primary proprioceptors are sensitive to the neonatal injury. The survival of proprioceptors during early postnatal period is probably dependent upon brain-derived neurotrophic factor in the trigeminal nervous system.

Dystonin deficiency reduces taste buds and fungiform papillae in the anterior part of the tongue.

The anterior part of the tongue was examined in wild type and dystonia musculorum mice to assess the effect of dystonin loss on fungiform papillae. In the mutant mouse, the density of fungiform papillae and their taste buds was severely decreased when compared to wild type littermates (papilla, 67% reduction; taste bud, 77% reduction). The mutation also reduced the size of these papillae (17% reduction) and taste buds (29% reduction). In addition, immunohistochemical analysis demonstrated that the dystonin mutation reduced the number of PGP 9.5 and calbindin D28k-containing nerve fibers in fungiform papillae. These data together suggest that dystonin is required for the innervation and development of fungiform papillae and taste buds.

Brain-derived neurotrophic factor-immunoreactive primary sensory neurons in the rat trigeminal ganglion and trigeminal sensory nuclei.

Immunohistochemistry for brain-derived neurotrophic factor (BDNF) was performed on the rat trigeminal ganglion (TG). The immunoreactivity (IR) was detected in 46% of TG neurons. These neurons were mostly small- or medium-sized (range, 149.7-1246.3 microm2; mean +/- SD = 373.4 +/- 151.6 microm2). A double immunofluorescence method also revealed that 54% of BDNF-immunoreactive (IR) neurons were immunoreactive for calcitonin-gene-related peptide. In addition, 93% of BDNF-IR TG neurons contained vanilloid receptor subtype 1. However, the co-expression of BDNF and vanilloid receptor 1-like receptor was very rare (less than 1%). In the trigeminal sensory nuclei, laminae II of the medullary dorsal horn was abundant in presumed BDNF-IR axon terminals. Such profiles were also detected in the dorsolateral part of the subnucleus oralis. The retrograde tracing and immunohistochemical methods demonstrated that BDNF-IR was common among cutaneous TG neurons (47%) but not tooth pulp TG neurons (13%). The present study indicates that BDNF-IR TG neurons have unmyelinated axons and project to the superficial medullary dorsal horn. It is likely that BDNF-containing neurons in both the trigeminal and spinal sensory systems have similarities in morphology and function. However, the content of BDNF in TG neurons probably depends on their peripheral targets. BDNF seems to convey nociceptive cutaneous input to the trigeminal sensory nuclei.

ASIC3-immunoreactive neurons in the rat vagal and glossopharyngeal sensory ganglia.

ASIC3-immunoreactivity (ir) was examined in the rat vagal and glossopharyngeal sensory ganglia. In the jugular, petrosal and nodose ganglia, 24.8%, 30.8% and 20.6% of sensory neurons, respectively, were immunoreactive for ASIC3. These neurons were observed throughout the ganglia. A double immunofluorescence method demonstrated that many ASIC3-immunoreactive (ir) neurons co-expressed calcitonin gene-related peptide (CGRP)- or vanilloid receptor subtype 1 (VRL-1)-ir in the jugular (CGRP, 77.8%; VRL-1, 28.0%) and petrosal ganglia (CGRP, 61.7%; VRL-1, 21.5%). In the nodose ganglion, however, such neurons were relatively rare (CGRP, 6.3%; VRL-1, 0.4%). ASIC3-ir neurons were mostly devoid of tyrosine hydroxylase in these ganglia. However, some ASIC3-ir neurons co-expressed calbindin D-28k in the petrosal (5.5%) and nodose ganglia (3.8%). These findings may suggest that ASIC3-containing neurons have a wide variety of sensory modalities in the vagal and glossopharyngeal sensory ganglia.

Aspartate-immunoreactive primary sensory neurons in the mouse trigeminal ganglion.

Aspartate-immunoreactivity (ir) was examined in the mouse trigeminal ganglion (TG). The ir was detected in 34% of TG neurons and their cell bodies were of various sizes (mean +/- S.D. = 1,234 +/- 543 microm(2)). A triple immunofluorescence method revealed the co-expression of aspartate with calcitonin gene-related peptide (CGRP) and parvalbumin; 22% and 14% of aspartate-immunoreactive (ir) neurons were also immunoreactive for CGRP and parvalbumin, respectively. The co-expression of aspartate with both CGRP and parvalbumin was very rare in the TG. By retrograde tracing method, half and 66% of TG neurons which innervate the vibrissa and palate, respectively, contained aspartate-ir. The co-expression of aspartate with CGRP was more common among palatal neurons (36%) compared to vibrissal neurons (22%). Aspartate-ir neurons which co-expressed parvalbumin-ir were numerous in the vibrissa (17%) but not in the palate (4%). These findings may suggest that the function of aspartate-containing TG neurons is correlated with their peripheral receptive fields.

Immunohistochemical localization of gamma and beta subunits of epithelial Na+ channel in the rat molar tooth pulp.

The distribution of gamma and beta subunits of epithelial Na(+) channel (ENaC), markers for low-threshold mechanoreceptors in peripheral tissues, was examined in the tooth pulp. In the root pulp, gammaENaC- and betaENaC-immunoreactive (IR) nerve fibers showed a thick smooth appearance. These nerve fibers ascended toward the pulp horn and formed subodontoblastic nerve plexuses. Immunoelectron microscopic method revealed that 63% of axons were immunoreactive for gammaENaC in the root pulp. Virtually all myelinated axons showed gammaENaC-IR (97%), whereas unmyelinated axons were mostly devoid of it (12%). These findings suggest that myelinated tooth pulp nociceptors respond to mechanical stimuli.

Calretinin-containing neurons which co-express parvalbumin and calbindin D-28k in the rat spinal and cranial sensory ganglia; triple immunofluorescence study.

The co-expression of calretinin with parvalbumin and calbindin D-28k was examined in the rat cranial and spinal sensory ganglia by triple immunofluorescence method. In the trigeminal and nodose ganglia, 9% and 5% of calretinin-immunoreactive neurons, respectively, also contained both parvalbumin- and calbindin D-28k immunoreactivity. These neurons had large cell bodies. In the trigeminal ganglion, they were restricted to the caudal portion. Such neurons were evenly distributed throughout the nodose ganglion. The co-expression could not be detected in the dorsal root, jugular or petrosal ganglia. Nerve fibers which co-expressed all the three calcium-binding proteins were observed in the inferior alveolar nerve but not the infraorbital nerve or palate. In the periodontal ligament, these nerve fibers formed Ruffini-like endings. These findings suggest that (1) the co-expression in trigeminal neurons is intimately related to their peripheral receptive fields; (2) the three calcium-binding proteins (calretinin, parvalbumin, calbindin D-28k) co-expressed in the trigeminal neurons may have mechanoreceptive function in the periodontal ligament.

Expression of osteoclast differentiation factor and osteoclastogenesis inhibitory factor in rat osteoporosis induced by immunosuppressant FK506.

Immunosuppressant drugs are currently required by transplant recipients for the remainder of their lives, despite the many adverse effects associated with these therapies. Acute osteoporosis is one such effect, and a reproducible osteoporosis model has been established through the administration of the immunosuppressant drug FK506 in rats. The cause of this osteoporosis has been shown to be abnormal osteoclast proliferation, altering the process of bone remodeling. However, the reasons why FK506 induces osteoclast proliferation and whether this process is mediated by cytokine changes or an increase in bone resorption factors have been unclear. An investigation was therefore conducted focusing on the recent discoveries of osteoclast differentiation factor (ODF) and osteoclastogenesis inhibitory factor (OCIF). These factors led to elucidation of the osteoclast differentiation-maturation mechanism. An osteoporosis model was produced in rats utilizing intramuscular FK506 injection (1 mg/kg) for 28 consecutive days. Trabecular bone resorption was observed inferior to enchondral ossification in the FK506 group, and tartrate resistant acid phosphatase (TRAP) staining revealed a clear increase in osteoclasts at the site of enchondral ossification, relative to the control group. Real-time PCR and in situ hybridization (ISH) demonstrated minimal differences in OCIF expression between control and the treatment groups. However, Real-time PCR revealed clearly increased ODF expression in the treatment group. ODF expression was also shown to be increased in the treatment group using ISH. This was histologically consistent with a region of osteoclast proliferation inferior to enchondral ossification. The results of this study support the hypothesis that FK506-mediated osteoporosis occurs by action of the drug on osteoclasts, promoting expression of ODF messenger ribonucleic acid (mRNA) and thus prompting osteoclast differentiation and maturation.

Expression of connective tissue growth factor/hypertrophic chondrocyte-specific gene product 24 (CTGF/Hcs24) during fracture healing.

Localization and expression of connective tissue growth factor/hypertrophic chondrocyte-specific gene product 24 (CTGF/Hcs24) during fracture healing in mouse ribs were investigated. In situ hybridization demonstrated that CTGF/Hcs24 mRNA was remarkably expressed, especially in hypertrophic chondrocytes and proliferating chondrocytes, in the regions of regenerating cartilage on days 8 and 14 after fracture. CTGF/Hcs24 mRNA was also expressed in proliferating periosteal cells in the vicinity of the fracture sites on days 2 and 8, and in cells in fibrous tissue around the callus on day 8. Northern blot analysis showed that expression of CTGF/Hcs24 mRNA was 3.9 times higher on day 2 of fracture healing than that on day 0. On day 8, it reached a peak of 8.6 times higher than that on day 0. It then declined to a lower level. Immunostaining showed that CTGF/Hcs24 was localized in hypertrophic chondrocytes and proliferating chondrocytes in the regions of regenerating cartilage, and in active osteoblasts in the regions of intramembranous ossification. Although CTGF/Hcs24 was abundant in the proliferating and differentiating cells (on days 8 and 14), immunostaining decreased as the cells differentiated to form bone (on day 20). CTGF/Hcs24 was also detected in cells in fibrous tissue, vascular endothelial cells in the callus, and periosteal cells around the fracture sites. These results suggest that CTGF/Hcs24 plays some role in fracture healing.

Effect of Brn-3a deficiency on parvalbumin-, calbindin D-28k-, calretinin- and calcitonin gene-related peptide-immunoreactive primary sensory neurons in the trigeminal ganglion.

Immunohistochemistry for parvalbumin, calbindin D-28k, calretinin and calcitonin gene-related peptide (CGRP) was performed on the trigeminal ganglion and oro-facial tissues in Brn-3a wildtype and knockout mice at embryonic day 18.5 and postnatal day 0. In wildtype mice, the trigeminal ganglion contained abundant parvalbumin-, calbindin D-28k- and CGRP-immunoreactive neurons while the ganglion was almost devoid of calretinin-immunoreactive neurons. In Brn-3a knockout mice, a 63% decrease of parvalbumin-immunoreactive neurons was detected. In contrast, the absence of Brn-3a dramatically increased the number of calbindin D-28k-immunoreactive (3.5-fold increase) and calretinin-immunoreactive neurons (91-fold increase). The number of CGRP-immunoreactive neurons, however, was not altered by the Brn-3a deficiency. Cell size analysis indicated that loss of Brn-3a increased the proportions of small (<100 microm (2)) parvalbumin-, calbindin D-28k- and CGRP-immunoreactive neurons while it decreased those of large (>200 microm(2)) immunoreactive cells. Calretinin-immunoreactive neurons were either small or medium (100-200 microm (2)) in mutant mice. The oro-facial tissues contained parvalbumin-, calbindin D-28k- and CGRP-immunoreactive fibers, but not calretinin-immunoreactive ones in wildtype mice. In Brn-3a knockout mice, the number of parvalbumin-immunoreactive fibers markedly decreased in the infraorbital nerve and parvalbumin-immunoreactive endings disappeared in the vibrissa. In contrast, the number of calbindin D-28k-immunoreactive fibers increased significantly in the infraorbital and mental nerves. In addition, calbindin D-28k-immunoreactive endings appeared in the vibrissa. As well, some fibers showed calretinin-immunoreactivity in the infraorbital nerve of the mutant. However, no obvious change of CGRP-immunoreactive fibers was observed in the oro-facial region of knockout mice. Taken together, our data suggest that Brn-3a deficiency has effects on the expression of neurochemical substances in the trigeminal ganglion.

Immunohistochemical study on expression of alpha-defensin and beta-defensin-2 in human buccal epithelia with candidiasis.

OBJECTIVES AND DESIGN: It has been previously reported that alpha-defensin (HNPs) and beta-defensin-2 (HBD-2) peptides with antifungal and cytotoxic activities can be detected in oral carcinomas and the saliva of patients with oral carcinomas. The present study investigated the presence of HNPs and HBD-2 in oral epithelia with candidiasis. MATERIALS AND METHODS: Tissue sections (4 microm) were prepared from biopsy and surgically removed specimens diagnosed as oral candidiasis (n = 10). The sections were examined immunohistochemically with antibodies directed against HNPs and HBD-2. RESULTS: Tissue sections of oral candidiasis were immunostained with antidefensin antibodies. Neutrophils in the inflamed lamina propria were positively immunostained with anti-HNPs antibody. The cytoplasm of cells in the upper spinous layer, in the lower spinous layer and in the parakeratinized layer of buccal epithelia with candidiasis was immunostained intensely with anti-HBD-2 antibody. In contrast, the expression of HBD-2 in the normal spinous layer was much weaker than that in oral candidiasis. No signals of HNPs were found in normal buccal epithelium. CONCLUSION: Buccal specimens from individuals with oral candidiasis show greater levels of expression of both HNPs and HBD-2. There might be a dual protection manner by defensins against fungal inflammation in infected buccal epithelia locally. Generally, HBD-2 signals have been found everywhere in the buccal epithelium; however, in an infected area, the signal intensity of HBD-2 has increased. HNPs signals have not been found in the normal buccal epithelium; however, HNPs signals have increased when the infection occurred.

Overexpression of connective tissue growth factor/hypertrophic chondrocyte-specific gene product 24 decreases bone density in adult mice and induces dwarfism.

Connective tissue growth factor/hypertrophic chondrocyte-specific gene product 24 (CTGF/Hcs24) is a multifunctional growth factor for fibroblasts, chondrocytes, and vascular endothelial cells. In the present study, we established transgenic (Tg) mice that overproduce CTGF/Hcs24 under the control of mouse type XI collagen promoter. Tg mice could develop and their embryonic and neonatal growth occurred normally. But they showed dwarfism within a few months of birth. X-ray analysis revealed that their bone density was decreased compared with normal mice. The femurs in the hindlimbs in particular showed an apparent low density. These results indicated that overexpression of CTGF/Hcs24 affects certain steps of endochondral ossification. In addition, the testes were much smaller than normal and fertility was affected in Tg mice, indicating that CTGF/Hcs24 may also regulate the embryonic development of the testis.

The role of SF/HGF and c-Met in the development of skeletal muscle.

Hypaxial skeletal muscles develop from migratory and non-migratory precursor cells that are generated by the lateral lip of the dermomyotome. Previous work shows that the formation of migratory precursors requires the c-Met and SF/HGF genes. We show here that in mice lacking c-Met or SF/HGF, the initial development of the dermomyotome proceeds appropriately and growth and survival of cells in the dermomyotome are not affected. Migratory precursors are also correctly specified, as monitored by the expression of Lbx1. However, these cells remain aggregated and fail to take up long range migration. We conclude that parallel but independent cues converge on the migratory hypaxial precursors in the dermomyotomal lip after they are laid down: a signal given by SF/HGF that controls the emigration of the precursors, and an as yet unidentified signal that controls Lbx1. SF/HGF and c-Met act in a paracrine manner to control emigration, and migratory cells only dissociate from somites located close to SF/HGF-expressing cells. During long range migration, prolonged receptor-ligand-interaction appears to be required, as SF/HGF is expressed both along the routes and at the target sites of migratory myogenic progenitors. Mice that lack c-Met die during the second part of gestation due to a placental defect. Rescue of the placental defect by aggregation of tetraploid (wild type) and diploid (c-Met-/-) morulae allows development of c-Met mutant animals to term. They lack muscle groups that derive from migratory precursor cells, but display otherwise normal skeletal musculature.

Severe neuropathies in mice with targeted mutations in the ErbB3 receptor.

Neuregulins and their specific receptors, members of the ErbB family of tyrosine kinases, have been implicated in the control of growth and development of Schwann cells, specialized cells that wrap around nerve axons to provide electrical insulation. Here we use gene targeting to generate mice that lack ErbB3, a high-affinity neuregulin receptor. Homozygous erbB3 mutant embryos lack Schwann-cell precursors and Schwann cells that accompany peripheral axons of sensory and motor neurons. The initial development of motor neurons and sensory neurons of dorsal root ganglia occurs as it should, but at later stages most motor neurons (79%) and sensory neurons in dorsal root ganglia (82%) undergo cell death in erbB3 mutant embryos. Degeneration of the peripheral nervous system in erbB3 mutant pups is thus much more severe than the cell death in mice that lack neurotrophins or neurotrophin receptors. We also show that ErbB3 functions in a cell-autonomous way during the development of Schwann cells, but not in the survival of sensory or motor neurons. Our results indicate that sensory and motor neurons require factors for their survival that are provided by developing Schwann cells.

Isoform-specific expression and function of neuregulin.

Neuregulin (also known as NDF, heregulin, ARIA, GGF or SMDF), induces cell growth and differentiation. Biological effects of neuregulin are mediated by members of the erbB family of tyrosine kinase receptors. Three major neuregulin isoforms are produced from the gene, which differ substantially in sequence and in overall structure. Here we use in situ hybridization with isoform-specific probes to illustrate the spatially distinct patterns of expression of the isoforms during mouse development. Ablation of the neuregulin gene in the mouse has demonstrated multiple and independent functions of this factor in development of both the nervous system and the heart. We show here that targeted mutations that affect different isoforms result in distinct phenotypes, demonstrating that isoforms can take over specific functions in vivo. Type I neuregulin is required for generation of neural crest-derived neurons in cranial ganglia and for trabeculation of the heart ventricle, whereas type III neuregulin plays an important role in the early development of Schwann cells. The complexity of neuregulin functions in development is therefore due to independent roles played by distinct isoforms.