Cytokine-induced nuclear factor kappa B activation promotes the survival of developing neurons.

Ciliary neurotrophic factor (CNTF), leukemia inhibitory factor (LIF), cardiotrophin-1 (CT-1), and interleukin 6 (IL-6) comprise a group of structurally related cytokines that promote the survival of subsets of neurons in the developing peripheral nervous system, but the signaling pathways activated by these cytokines that prevent neuronal apoptosis are unclear. Here, we show that these cytokines activate NF-kappaB in cytokine-dependent developing sensory neurons. Preventing NF-kappaB activation with a super-repressor IkappaB-alpha protein markedly reduces the number of neurons that survive in the presence of cytokines, but has no effect on the survival response of the same neurons to brain-derived neurotrophic factors (BDNF), an unrelated neurotrophic factor that binds to a different class of receptors. Cytokine-dependent sensory neurons cultured from embryos that lack p65, a transcriptionally active subunit of NF-kappaB, have a markedly impaired ability to survive in response to cytokines, but respond normally to BDNF. There is increased apoptosis of cytokine- dependent neurons in p65(-/)- embryos in vivo, resulting in a reduction in the total number of these neurons compared with their numbers in wild-type embryos. These results demonstrate that NF-kappaB plays a key role in mediating the survival response of developing neurons to cytokines.

GDNF is an age-specific survival factor for sensory and autonomic neurons.

Glial cell line-derived neurotrophic factor (GDNF) promotes the survival of two populations of CNS neurons: motoneurons and midbrain dopaminergic neurons. To see whether GDNF promotes the survival of PNS neurons, we studied embryonic chicken autonomic and sensory neurons in culture. We show that GDNF promotes the survival of sympathetic, parasympathetic, proprioceptive, enteroceptive, and small and large cutaneous sensory neurons. Whereas sympathetic, parasympathetic, and proprioceptive neurons become less responsive to GDNF with age, enteroceptive and cutaneous sensory neurons become more responsive. GDNF mRNA is expressed in the tissues innervated by these neurons, and developmental changes in its expression in several tissues mirror the changing responses of the innervating neurons to GDNF. These results show that GDNF promotes the survival of multiple PNS and CNS neurons and suggest that GDNF may be important for regulating the survival of various populations of neurons at different stages of their development.

Target-derived GFRalpha1 as an attractive guidance signal for developing sensory and sympathetic axons via activation of Cdk5.

Immobilized and diffusible molecular cues regulate axon guidance during development. GFRalpha1, a GPI-anchored receptor for GDNF, is expressed as both membrane bound and secreted forms by accessory nerve cells and peripheral targets of developing sensory and sympathetic neurons during the period of target innervation. A relative deficit of GFRalpha1 in developing axons allows exogenous GFRalpha1 to capture GDNF and present it for recognition by axonal c-Ret receptors. Exogenous GFRalpha1 potentiates neurite outgrowth and acts as a long-range directional cue by creating positional information for c-Ret-expressing axons in the presence of a uniform concentration of GDNF. Soluble GFRalpha1 prolongs GDNF-mediated activation of cyclin-dependent kinase 5 (Cdk5), an event required for GFRalpha1-induced neurite outgrowth and axon guidance. Together with GDNF, target-derived GFRalpha1 can function in a non-cell-autonomous fashion as a chemoattractant cue with outgrowth promoting activity for peripheral neurons.

Peripheral target-specific influences on embryonic neurite growth vigor and patterns.

We examined axon-target interactions in cocultures of embryonic rat trigeminal, dorsal root, nodose, superior cervical ganglia or retina with a variety of native or foreign peripheral targets such as the whisker pad, forepaw, and heart explants. Axon growth into these peripheral target tissues was analyzed by the use of lipophilic tracer DiI. Embryonic day 15 dorsal root and trigeminal axons grew into isochronic normal and foreign cutaneous targets. Both axon populations avoided the same age heart tissue, but grew profusely into younger (embryonic day 13) or older (postnatal) heart explants. In contrast, embryonic day 15 superior cervical or nodose ganglion axons grew heavily into the same age heart and forepaw explants and to a lesser extent into the whisker pad explants. Embryonic day 15 retinal axons grew into all three peripheral targets used in this study. Primary sensory and sympathetic axons, but not retinal axons, formed target-specific patterns in the whisker pad and forepaw explants. DiI-labeling and immunostaining of primary sensory neurons in coculture revealed that these neurons retain their bipolar characteristics, and express class-specific markers such as parvalbumin, calcitonin gene-related peptide and TrkA receptors. In the whisker pad explants, axons positive for all three markers were seen to form patterns around the follicles. Our results indicate that developing peripheral targets can attract and support axon growth from a variety of sources. Whereas neurotrophins play a major role in attracting and supporting survival of subpopulations of sensory neurons, other substrate-bound or locally released molecules must regulate sensory neurite growth into specific peripheral and central targets.

The origin and development of the vagal and spinal innervation of the external muscle of the mouse esophagus.

Retrograde and anterograde tracing and immunohistochemical techniques were used to examine the origin of the extrinsic innervation, and the development of the vagal innervation to the mouse esophagus. Cholinergic nerve terminals were localised using an antiserum to the vesicular acetylcholine transporter and cholinergic cell bodies were localised using an antiserum to choline acetyltransferase. Cholinergic nerve terminals, which also contained calcitonin gene-related peptide, were present at the motor end plates in the external (striated) muscle of the esophagus. Following injection of Fast Blue into subdiaphragmatic or cervical levels of the esophagus, the only retrogradely-labelled cholinergic nerve cell bodies that also contained calcitonin gene-related peptide were found in the nucleus ambiguus. Neurons in the dorsal motor nucleus of the vagus, the nodose ganglia and dorsal root ganglia gave rise to a number of different types of nerve terminals within the myenteric plexus. Retrogradely-labelled neurons in the dorsal motor nucleus of vagus contained cholinergic markers only, nitric oxide synthase only or cholinergic markers plus nitric oxide synthase, retrogradely-labelled neurons in the dorsal root ganglia contained calcitonin gene-related peptide only, and a small number of retrogradely-labelled neurons in the nodose ganglia contained tyrosine hydroxylase. The development of the vagal innervation to the esophagus was examined following application of DiI to the vagus nerve of fixed mouse embryos. Anterogradely-labelled nerve fibres, which arose from both nodose ganglia and the medulla, were already present in the esophagus of embryonic day 12 (E12) mice. Some of the DiI-labelled vagal nerve fibres were present in among the smooth muscle cells of the external muscle layer prior to their transdifferentiation to striated muscle. We conclude that the neurons in the nucleus ambiguus that project to the esophagus differ from other extrinsic neurons in their chemistry as well as their targets within the esophagus. The development of the extrinsic innervation precedes the transdifferentiation of the external muscle to striated muscle, raising the possibility that, during development, smooth muscle of the esophagus is innervated transiently by vagal neurons.

Effects of nerve growth factor on sensory neurons in the chick embryo: a stereological study.

Chicken embryos on days 6-13 of incubation received injections of purified beta NGF (80 micrograms/day) for 3 or 4 days and were then killed. Sensory ganglia were fixed and taken for embedding and sectioning. A stereological method based on unfolding of cell-diameter frequencies was used to determine the number of neurons of different size in the spinal, trigeminal and nodose ganglia. The total volume of the ganglia was also determined. NGF induced increases in diameter of the neural crest-derived dorsomedial (DM) neurons in spinal and trigeminal ganglia. Injected NGF did not influence ventrolateral (VL) neurons of neural crest origin in the spinal ganglia nor the ventrolateral neurons of placodal origin in the trigeminal ganglion. The volumes of spinal and trigeminal ganglia increased by 50 and 100%, respectively. The volume of the nodose ganglion and the total number and size of the placodal nodose neurons were unaffected by NGF. The results demonstrate a clear difference in the response to NGF in vivo between smaller and larger sensory neurons.

Identification of a chicken homologue in the Brn-3 subfamily of POU-transcription factors.

Among the many transcription factors thus far identified several are found to be expressed almost exclusively in the nervous system. The Brn-3 subfamily of POU-transcription factors constitutes a highly conserved group of such factors showing expression predominantly in sensory neurons. We now describe the nucleotide sequence and proposed amino acid sequence of a chicken homologue to the murine and human Brn-3 genes. Furthermore we characterise the early embryonic expression pattern of this chicken Brn-3 gene and show it to be expressed in peripheral sensory ganglia as well as in retinal ganglion cells. Based on these findings we conclude that the chicken homologue to the murine and human Brn-3a genes has been cloned. We have begun to examine possible regulatory pathways of Brn-3a by stimulating chick embryonic peripheral ganglia with trophic factors and assaying resulting levels of Brn-3a with a quantitative PCR approach. Trigeminal and dorsal root ganglia stimulated in culture by NGF and NT-3 embryonic day 9 (E9) produce neurites without raising the Brn-3a mRNA levels.

Immunohistochemical demonstration of cytokeratin in human embryonic neurons arising from placodes.

Sensory neurons of the olfactory, trigeminal, facial, vestibulo-cochlear, glossopharyngeal and vagal nerves, and neurons migrating along the olfactory nerve to the brain have special anlagen, made up of placodes located in the epithelial layer. To investigate the characteristic phenotype of placode-derived neurons, immunohistochemical analysis of intermediate filaments was conducted on formalin-fixed human embryonic tissues. Neurons arising from placodes including luteinizing-hormone releasing hormone (LHRH) neurons migrating from the olfactory placode to the brain had immunoreactivity to antibodies specific to cytokeratin, AE1 and CAM5.2 during the embryonic stage. However, this immunoreactivity disappeared during the late embryonic to the post-embryonic stage and was not observed in the roots of these nerves in the post-natal stage. Immunoreactivity was detected in both the somata and processes, and the distribution differed from that described in rodent brain neurons. With this exception, no other human peripheral neurons, including spinal dorsal root ganglia, had immunoreactivity with anti-cytokeratin antibodies throughout the entire developmental stage. Although the cephalic neural crest also directly generates neurons to most of the cranial sensory ganglia, we could not find any evidence that it contributed to the genesis of cytokeratin-positive embryonic neurons. We concluded that cytokeratin is an intermediate filament common to human embryonic neurons of cephalic placodal origin and that this immunohistochemical marker may be useful in analyzing the developmental sequence of several congenital diseases involving the cranial nerves, such as Moebius syndrome and Goldenhar syndrome.

Formation of synapses on the growth cones in human embryonic inferior ganglion of the vagus.

Ultrastructural study was performed on inferior ganglia of the vagus in human embryos aged 7and 8 weeks (developmental stages 18 to 23, 44 to 56 days). The growth cones are observed between the bundles of axons of the inferior ganglia of the vagus. Many primitive synapses (protosynapses) between dendritic and axonal growth cones are observed.

Heterotopic transplantation of presumptive placodal ectoderm changes the fate of sensory neuron precursors.

The placode-derived cranial sensory neurons of the vestibular and nodose ganglia in avian embryos exhibit differences in neurite growth rate and the duration of neurotrophin-independent survival in vitro that arise prior to gangliogenesis and target contact (Davies, A. M. (1989) Nature 337, 553-555; Vogel, K. S. and Davies, A. M. (1991) Neuron 7, 819-830). To ascertain the state of commitment of presumptive placodal ectoderm to differentiate into neurons of the vestibular or nodose type, we performed heterotopic transplantation of labelled presumptive placodal ectoderm at E1.5 in the chicken embryo. We then assayed transplant-derived neurons for hindbrain innervation patterns, neurite growth and survival at E3.5. We show that presumptive placodal ectoderm is not determined to give rise to neurons of the vestibular or nodose phenotype at E1.5. Explantation of presumptive placodal ectoderm at E1.5 showed that this ectoderm is also not specified to differentiate into neurons at this stage. In addition, we demonstrate that non-neurogenic ectoderm from the trunk can give rise to nodose-type neurons when transplanted heterotopically to the nodose region.

Compensatory responses and development of the nodose ganglion following ablation of placodal precursors in the embryonic chick (Gallus domesticus).

The nodose ganglion is the distal cranial ganglion of the vagus nerve which provides sensory innervation to the heart and other viscera. In this study, removal of the neuronal precursors which normally populate the right nodose ganglion was accomplished by ablating the right nodose placode in stage 9 chick embryos. Subsequent histological evaluation showed that in 54% of lesioned embryos surviving to day 6, the right ganglion was absent. Most embryos surviving to day 12, however, had identifiable right ganglia. In day 12 embryos, the right ganglion which developed was abnormal, with ganglion volume and ganglion cell diameter reduced by 50% and 20%, respectively, compared to control ganglia. To investigate the source of the neuron population in the regenerated ganglion, we combined nodose placode ablation with bilateral replacement of chick with quail "cardiac" neural crest (from mid-otic placode to somite 3). These cells normally provide only non-neuronal cells to the nodose ganglion, but produce neurons in other regions. At day 9, quail-derived neurons were identified in the right nodose ganglia of these chimeras, indicating that cardiac neural crest cells can generate neurons in the ganglion when placode-derived neurons are absent or reduced in number. On the other hand, we found that "sympathetic" neural crest (from somites 10 to 20) does not support ganglion development, suggesting that only neural crest cells normally present in the ganglion participate in reconstituting its neuronal population. Our previous work has shown that right nodose placode ablation produces abnormal cardiac function, which mimics a life-threatening human heart condition known as long QT syndrome. The present results suggest that the presence of neural crest-derived neurons in the developing right nodose ganglion may contribute to the functional abnormality in long QT syndrome.

Ultrastructural studies on differentiation of nerve cells in the human embryonic and fetal inferior ganglion of the vagus.

Inferior ganglia of the vagus nerve in human embryos and fetuses ranging in C.-R. length from 16 to 220 mm and in age from 7 to 23 weeks were studied by electron microscopy. During the investigated period of development 5 types of cells were distinguished, viz.: 1. primary (apolar) neuroblast, 2. early bipolar neuroblast, 3. intermediate bipolar neuroblast, 4. late bipolar neuroblast, and 5. unipolar neuroblast. The unipolar neuroblasts appear at the end of the embryonic period. In the 9th and 10th week of the fetal period the apolar and early bipolar neuroblasts disappear. In the 23rd week the neuronal cells of the vagus resemble those of adults. The maturation of neurons is associated with 1. structure of nuclear components, 2. cytoplasmic organization with special reference to the rough endoplasmic reticulum, 3. changes of cell shape and development of processes.

Developmental characteristics of the chick nodose ganglion.

Morphometric studies were carried out on the chick nodose ganglion between day 5 of incubation and 2 weeks after hatching. Previous findings showed that ablation of the nodose placode, the locus of precursor cells of nodose ganglion sensory neurons, results in abnormal cardiac function, and that these precursors can be induced to migrate abnormally to the heart and express abnormal phenotypes there, following cardiac neural crest ablation. These results prompted us to investigate further the normal development of nodose ganglion neurons. We find that the major period of neuron generation from placodal precursor cells in the ganglion occurs prior to day 5 of incubation. The loss of more than half of these neurons takes place between embryonic days 5 and 20, while neuron and ganglion sizes increase dramatically. Myelination is not seen at day 12 of incubation, but is present on day 15. Neurons continue to develop after hatching (day 21), reaching their adult size by 2 weeks after hatching. Unexpectedly, we found that the number of neurons increases after hatching, reaching the adult level of 62% more than embryonic day-20 numbers by 2 weeks after hatching. The mechanisms underlying the increase in neuron number after hatching are unexplained and require further investigation.

Morphology and development of the distal ganglion of the vagus nerve (ganglion distale n. vagi) in sheep during the prenatal period.

This study was carried out on 31 sheep fetuses, which were divided into four developmental groups. It was established that the development and morphology of the investigated part of vagus nerve is strongly correlated with the developing organs of the fetuses. The very distinct 'translocation' of the distal ganglion of the vagus nerve in the caudoventral direction was observed in the course of development. There was no influence of sex on the results of investigation.

Expression of Hox 2.1 protein in restricted populations of neural crest cells and pharyngeal ectoderm.

A polyclonal antibody, alpha Hox 2.1a, was used to localize Hox 2.1 protein in presumptive neural crest cells and nodose ganglion of 8.5-10.0 day p.c. mouse embryos. The following results were obtained: (1) The nodose placode, in its epithelial state, first expresses Hox 2.1 protein at 9.0 d.p.c. By 9.5 d.p.c. presumptive migrating neuroblasts between the nodose placode and ganglion primordium also express Hox 2.1 protein. (2) At 9.5 d.p.c., presumptive crest cells lateral to the cephalic cardinal vein and within pharyngeal arches 4 and 6 are immunoreactive for alpha Hox 2.1a. In the arch 6 region, positive cells extend medially to a mesenchymal cell population on the lateral aspect of the foregut wall. (3) At 10.0 d.p.c., Hox 2.1 protein expression in putative crest cells is restricted to the arch 6 cell population. A similar staining pattern is seen using alpha Hox 2.1a with chick embryos. Comparison with the chicken embryo suggests that the Hox 2.1 positive cells in the pharyngeal arch and those on the lateral aspect of the foregut in the mouse embryo correspond to the caudalmost subpopulation of the circumpharyngeal crest (Kuratani and Kirby: Am. J. Anat. 191:215-227, 1991; Anat. Rec. 234:263-280, 1992). These results are consistent with a role for Hox 2.1 in pattern formation in the caudalmost region of the vertebrate head.

Antibodies against mouse nerve growth factor interfere in vivo with the development of avian sensory and sympathetic neurones.

The monoclonal antibody 27/21 directed against mouse nerve growth factor (NGF) interferes in vivo with the survival of sensory dorsal root ganglion (DRG) neurones during the development of the quail embryo: the number of DRG neurones at embryonic day 11 (E11) was reduced by about 30% in embryos treated with the antibody between E3 and E11. Neurone numbers in the nodose ganglion were not affected. The effect of NGF antibodies on sympathetic neurones was assessed by determining the levels of the adrenergic marker enzyme tyrosine hydroxylase. Both total tyrosine hydroxylase activity and protein levels in sympathetic chains were reduced by about 30% in embryos treated with 27/21 antibody but not in embryos treated with a control antibody. The 27/21 antibody cross-reacts with chick NGF-like activity as shown in vitro by the ability of the antibody to partially block the survival activity of chick-embryo-fibroblast-conditioned medium for E9 chick DRG neurones.

Trophic regulation of nodose ganglion cell development: evidence for an expanded role of nerve growth factor during embryogenesis in the rat.

Peripheral sensory neurons are derived from two distinct embryonic tissues, the neural crest and epibranchial placodes. Studies in the chick suggest that embryonic lineage and trophic dependence are interrelated, such that many crest-derived cells depend on NGF for survival during development, whereas placodal derivates, including nodose ganglion neurons, do not (30). It remains controversial, however, whether or not a similar dichotomy exists in mammalian species, in which trophic requirements during early development of placodal ganglia have not been defined. To approach this issue, the present study examined the effects of nerve growth factor (NGF) on neuronal survival in embryonic rat nodose ganglion cultures. Treatment of E13.5-14.5 nodose explants with 20 ng/ml NGF resulted in a four-fold increase in neuronal survival that was blocked by anti-NGF antiserum. Increased neuronal survival and neurite outgrowth were also observed in neuron-enriched dissociated cell cultures; these effects were seen within 12 h of plating, indicating that NGF-responsive neurons or neuroblasts were already present in the ganglion at the time of explantation. This was further supported by immunocytochemical staining of nodose cell bodies in situ with the monoclonal antibody 192-IgG against the NGF receptor (12). These findings indicate that NGF may be important in regulating nodose development during early gangliogenesis in mammals and suggest that NGF plays a more widespread role in peripheral nervous system ontogeny than previously recognized.

A monoclonal antibody (SN1) identifies a subpopulation of avian sensory neurons whose distribution is correlated with axial level.

We have produced a monoclonal antibody, designated SN1, which binds to the surfaces of a subpopulation of avian sensory neurons, but not to other neurons of the peripheral or central nervous systems. The proportion of SN1(+) neurons in brachial and lumbosacral dorsal root ganglia (DRG), which innervate the wings and legs respectively, is low (30-40%), compared to the proportion (80-90%) in the lower thoracic DRG. SN1 immunoreactive fibers project to laminae I and II of the spinal cord dorsal horn, and are seen in the skin, but not the deeper tissues of older embryos. On the basis of the time of appearance, axial level-dependent distribution, and the central and peripheral projections of SN1(+) neurons, we suggest that they are cutaneous afferents that depend on interaction with peripheral targets to differentiate.

Timing and regulation of trkB and BDNF mRNA expression in placode-derived sensory neurons and their targets.

The sensory neurons of the vestibular and nodose ganglia of the chicken embryo have nearby and distant targets, respectively. In vitro studies have shown that these neurons survive independently of neurotrophins when their axons are growing to their targets and become dependent on brain-derived neurotrophic factor (BDNF) for survival when their axons reach the vicinity of their targets. Although the timing of BDNF dependence is principally controlled by an intrinsic timing mechanism in the neurons, the onset of dependence can be accelerated by BDNF exposure toward the end of the phase of neurotrophin independence. We have used quantitative reverse transcription/polymerase chain reaction to study the expression of transcripts coding for BDNF and the catalytic isoform of its receptor tyrosine kinase, TrkB, in these neurons and their targets at different stages of development. We show that the peripheral and central target tissues of these neurons express BDNF mRNA prior to the arrival of sensory axons. Vestibular neurons express trkB mRNA before nodose neurons, which accords with the earlier response of vestibular neurons to BDNF. In culture, early nodose neurons start expressing trkB mRNA after 36 h incubation, which is 36 h before these neurons become dependent on BDNF for survival. Although BDNF does not affect the timing and level of trkB mRNA expression during the first 48 h in vitro, it increases the level of trkB mRNA after this time. The timing of BDNF-induced elevation of trkB mRNA correlates with the period during which BDNF exposure accelerates the onset of BDNF dependence in nodose neurons. These results suggest that the timing of BDNF dependence in developing sensory neurons is due in part to expression of catalytic TrkB and demonstrate that a BDNF autocrine loop is not required for the survival of sensory neurons during the earliest stages of their development.

TrkB expression and early sensory neuron survival are independent of endogenous BDNF.

Sensory neurons initially survive independently of neurotrophins in culture during the stage of development when their axons are growing to their targets. Because mRNAs encoding brain-derived neurotrophic factor (BDNF) and its receptor tyrosine kinase TrkB are detectable in subsets of sensory neurons from the earliest stages of their development, we investigated whether a BDNF autocrine loop is responsible for sustaining the survival of these neurons during this early stage in their development. Low-density dissociated cultures of nodose and dorsal root ganglion neurons were established from wild type and BDNF(-/-) mouse embryos at this stage and were grown in defined medium without added neurotrophins. Wild type and BDNF-deficient neurons survived equally well under these conditions, indicating that a BDNF autocrine loop does not play a role in sustaining the survival of sensory neurons during the earliest stages of their development. As sensory axons approach their targets, TrkB expression increases in a subset of neurons that becomes dependent on BDNF produced by other cells. Because numerous studies have shown that neurotrophins, including BDNF, increase expression of their receptors, we investigated whether endogenous BDNF is required for the increase in TrkB expression observed during stage of development. Quantitative reverse transcriptase-polymerase chain reaction (RT-PCR) showed that the developmental increase in TrkB mRNA expression occurred normally in the sensory ganglia of BDNF(-/-) embryos. Taken together, our studies of sensory neuron development in BDNF-deficient embryos have demonstrated that endogenous BDNF is neither required for the early survival of these neurons nor for the induction of TrkB expression.