Induction of somatostatin immunoreactivity in cultured ciliary ganglion neurons by activin in choroid cell-conditioned medium.

We have previously shown that the expression of somatostatin-like immunoreactivity in cultured ciliary ganglion neurons is stimulated by a macromolecule found in choroid cell-conditioned medium (ChCM). Here, we present the following evidence that this somatostatin-stimulating activity (SSA) is activin: human recombinant activin induces somatostatin-like immunoreactivity in CG neurons; ChCM induces hemoglobin synthesis in K562 cells, a biological activity characteristic of activin; activin A-specific antibodies recognize a protein in ChCM; cultured choroid cells contain activin RNA; and SSA is inhibited by follistatin, a specific activin-binding protein. Thus, activin is likely to be a neurodifferentiation factor for CG neurons in vivo.

Cloning, expression during development, and evidence for release of a trophic factor for ciliary ganglion neurons.

Ciliary ganglion (CG) neurons undergo a period of cell death during development that may be regulated by the limited availability of trophic factor produced by their target tissues. We have previously reported the purification of a ciliary neurotrophic factor from adult chick sciatic nerve that we called growth promoting activity (GPA). Here we demonstrate that GPA can be purified and cloned from embryonic day 15 (E15) chick eyes, which contain all the target tissues of the CG. Our studies show the following: GPA mRNA is induced in embryonic chick eyes during the period of CG neuron cell death; GPA mRNA is expressed specifically in the layer of the eye that contains the targets of the CG and in primary cultures of smooth muscle cells isolated from the choroid layer of the eye; and biologically active GPA is released from cells transfected with a GPA cDNA.

Target tissue influence on somatostatin expression in the avian ciliary ganglion.

Activin as a neurodifferentiation factor. Our studies of neurotransmitter expression have focused on the expression of neuropeptide transmitters in the avian ciliary ganglion (CG) and have examined the influence of choroidal vascular smooth muscle cells in regulating the differential expression of somatostatin in the CG. In these activities we have identified activin A as a potential target-derived neurodifferentiation factor that can stimulate somatostatin expression in cultured CG neurons. In cultured CG neurons, activin can stimulate the expression of somatostatin in choroid neurons, the pattern of neurotransmitter expression found in vivo, and in the ciliary neurons that would normally not express somatostatin. In vivo, mRNA transcripts of the cActR-IIA appear to be expressed by both choroid and ciliary CG neurons. This suggests that activin might serve as an instructive factor in controlling neuropeptide phenotype. For activin to serve as an instructive factor requires that activin be produced by choroid smooth-muscle target cells. Indeed, activin mRNA and activin-like immunoreactivity are found in choroid cells, in vitro. However, the lack of somatostatin expression by ciliary neurons suggests that activin is not produced by their targets, the iris and ciliary body. This simple view is countered by the observation that activin A mRNA is also present in the iris and activin-like immunoreactivity is detectable in the iris and ciliary body. Instead, the production of the specific activin inhibitor follistatin in the iris and ciliary body is likely to limit the availability of activin to only those neurites innervating the choroid layer, thus accounting for the differential expression of somatostatin in only the choroid CG neurons. This somewhat more complicated arrangement is similar to the mechanism thought to be employed for primary induction during frog embryogenesis. The observations reviewed here are all consistent with the hypothesized role for activin as a molecule whose availability to neurites in the target regulates neurotransmitter expression. Additional in vivo perturbation experiments are needed to further examine this hypothesis; nevertheless, activin appears as a strong candidate for a target-derived neurotransmitter differentiation factor. Activin's potential roles in differentiation: A wide variety of biological effects have been ascribed to activin. Initially identified and purified as a gonadal hormone stimulating the production and release of FSH from the pituitary, activin is also implicated in the stimulation of erythroid differentiation, as a modulator of follicular granulosa cell differentiation, as a mesodermalizing factor in both amphibian and avian early development, and as a component in establishing left-right axial patterning in the chicken embryo. Activin has also been found to be a survival factor for several neuronal cell lines and for rat embryonic neural retina cells in culture. However, activin is not a survival factor for chicken CG neurons in culture. Our observation that activin may play a function in target-derived control of neuropeptide expression adds yet another aspect to the list of its potential biological functions. In addition, activin shares regions of amino acid sequence identity with members of the TGF-beta superfamily, which includes the TGF-betas, Mullerian inhibitory substance, Drosophila decapentaplegic gene product, dorsalin, bone morphogenetic proteins, inhibin, and glial-derived neurotrophic factor. Interestingly, these are all factors that have effects upon cellular differentiation. Effects of activin on other neurons. Activin A--as well as two other TGF-beta superfamily members, BMP-2 and BMP-6--has been shown to induce expression of mRNAs for several neuropeptides in cultured rat sympathetic neurons. In addition, activin A induces ChAT mRNA in cultured sympathetic neurons. (ABSTRACT TRUNCATED)

Translocation of latex beads after laser ablation of the avian neural crest.

Previous studies from this laboratory (M.E. Bronner-Fraser, 1982, Dev. Biol. 91, 50-63) have demonstrated that latex beads translocate ventrally after injection into avian embryos during the phase of neural crest migration, to settle in the vicinity of neural-crest-derived structures. In order to examine the role of host neural crest cells in the ventral translocation of implanted beads, latex beads have been injected into regions of embryos from which the neural crest cells have been ablated using a laser microbeam. Prior to their migratory phase, neural crest cells reside in the dorsal portion of the neural tube. Laser irradiation of the dorsal neural tube was used to reproducibly achieve either partial or complete ablation of neural crest cells from the irradiated regions. The effectiveness of the ablation was assessed by the degree of reduction in dorsal root ganglia, a neural crest derivative. Because of the rapidity and precision of this technique, it was possible to selectively remove neural crest cells without significantly altering other embryonic structures. The results indicate that, after injection of latex beads into the somites of embryos whose neural crest cells were removed by laser irradiation, the beads translocate ventrally in the absence of the endogenous neural crest.

Activin receptor mRNA expression by neurons of the avian ciliary ganglion.

Previous studies have suggested that activin may serve as a neurodifferentiation factor regulating somatostatin expression in neurons of the avian ciliary ganglion (CG). As one aspect of examining the role of activin in CG development, we inquired whether any of the known activin receptors are expressed by developing CG neurons in vivo. In addition, we examined whether activin A mRNA is expressed in the choroid layer and iris of the chicken eye. Oligonucleotide primers were designed for the chicken activin receptor type IIA (cActR-IIA), type IIB (cActR-IIB), and activin A. In reverse-transcription-polymerase chain reaction (rtPCR), an appropriately sized product was amplified from CG cDNA using primers to the cActR-IIA but not the cActR-IIB. Sequencing confirmed the identity of the PCR product as a fragment of the cActR-IIA. It thus appears that mRNA for the type IIA but not the type IIB activin receptor is expressed in the chicken CG. An antisense strand digoxigenin-labeled riboprobe complimentary to a 358-bp portion of the cActR-IIA kinase region hybridized to cells within cryostat sections of embryonic CG. From E6.5-E18, hybridization of this probe appears to be specific for cells with a neuronal morphology. Using rtPCR with activin A-specific primers we detected activin mRNA in the choroid layer of E14 and E19 eyes, and from the iris at E14. Our results are consistent with a role for activin as a neurodifferentiation factor in vivo, and imply that within the CG, the cActR-IIA is specifically expressed by neurons, and that activin A is expressed in the targets of these neurons.

Stimulation of somatostatin expression in developing ciliary ganglion neurons by cells of the choroid layer.

An important component of neuronal development is the matching of neurotransmitter expression with the appropriate target cell. We have examined how peptide transmitter expression is controlled in a simple model system, the avian ciliary ganglion (CG). This parasympathetic ganglion contains 2 distinct types of neurons: choroid neurons, which project to vasculature in the eye's choroid layer and use somatostatin as a co-transmitter with ACh, and ciliary neurons, which innervate the ciliary body and iris and use ACh but no known peptide co-transmitter. We have found that the earliest developmental stage in which neurons with somatostatinlike immunoreactivity (SOM-IR) are consistently found in vivo is stage 30 (embryonic day 6.5), a time shortly after the extension of neurites to targets in the eye's choroid layer. In cell culture, CG neurons expressed SOM-IR in co-culture with choroid cells, but not when cultured with striated muscle myotubes or with ganglion non-neuronal cells. No significant differences in neuronal survival or in ChAT activity were observed under these different co-culture conditions, which suggests that somatostatin expression is independently regulated. The stimulation of somatostatin expression was also specific in that other neuropeptides commonly found in autonomic neurons [neuropeptide Y (NPY), substance P (SP), vasoactive intestinal polypeptide (VIP)] were not induced in the presence of choroid cells. The ability to stimulate SOM-IR was not contact dependent because a macromolecule of greater than or equal to 10 kDa in choroid-conditioned medium (ChCM) was found to stimulate somatostatin expression in a dosage-dependent fashion. The somatostatin-stimulating activity induced SOM-IR in more than 90% of CG neurons, as well as in retrogradely labeled ciliary neurons, which would not normally express SOM-IR. Thus, the expression of somatostatin in cultured CG neurons is regulated by a macromolecule produced by cells in the choroid layer, a target normally innervated in vivo by CG neurons expressing somatostatin.

Cholinergic neurones acquire adrenergic neurotransmitters when transplanted into an embryo.

During development, cells become progressively restricted, until they reach their final phenotype. Differentiation was originally thought to be irreversible, but phenotypic plasticity has been observed in a variety of cell types, for example sympathetic neurones, the limb blastema and some glial cell types. A detailed description of the individual steps that lead to expression or reversal of phenotype is essential to understand the molecular events underlying cell differentiation. We examined whether ciliary neurones acquire adrenergic properties when exposed to a permissive embryonic environment. Cholinergic neurones were selectively labelled with a retrogradely transported marker and injected into chick embryos during active neural crest migration. Four to five days after injection, some of the labelled neurones were found in 'adrenergic sites' and had developed catecholamine histofluorescence. The cells had thus accumulated adrenergic neurotransmitters even after differentiation into cholinergic neurones. This result shows that neurotransmitter plasticity occurs in cholinergic neurones and suggests that the neurotransmitter phenotype can be modified by the embryonic environment.

Combined vital dye labelling and catecholamine histofluorescence of transplanted ciliary ganglion cells.

We have utilized the carbocyanine dye, DiI, to label suspensions of dissociated ciliary ganglion cells removed from 6 to 12 day old quail embryos. Some of the cells were injected into the trunk somites of 2.5-3 day old chick embryos along pathways where neural crest cells migrate to form sensory and sympathetic ganglia, aortic plexuses and the adrenal medulla; the remainder of the cells were cultured to check their viability and the persistence of the DiI label. Embryos were incubated for 1-8 days post-injection, fixed in 4% paraformaldehyde/0.25% glutaraldehyde and processed for cryostat sectioning. DiI-labelled cells were readily identifiable in culture and in sections of embryos at all stages examined. Several cell types were identified, based on their morphology and soma size. These included cells with large cell bodies and bright DiI-labelling that appeared to be neurons and smaller, more weakly labelled cells that appeared non-neuronal. The latter presumably had divided several times, accounting for their reduced levels of dye. Many of the DiI-labelled cells were found in and around neural crest-derived sympathetic ganglia, aortic plexuses and adrenomedullary cords, but were rarely observed in dorsal root ganglia. The aldehyde fixative (Faglu mixture) used in this study reacts with catecholamines to form a bright reaction product in adrenergic cells including those in the sympathetic ganglia and the adrenal medulla. The catecholamine biproduct and the DiI in the same cell can easily be viewed with different fluorescent filter sets. A variable number of the DiI-labelled cells in these adrenergic sites contained catecholamines. Cells derived from younger 6 day ciliary ganglion dissociates exhibited detectable catecholamine neurotransmitters earlier and more frequently than those derived from 8 day embryos. The presence of cells exhibiting both bright DiI and catecholamine fluorescence is consistent with previous indications that post-mitotic ciliary ganglion neurons can undergo phenotypic conversion from cholinergic to adrenergic when transplanted to the trunk environment.

Activin type II receptors in embryonic dorsal root ganglion neurons of the chicken.

Activin induces neuropeptide expression in chicken ciliary ganglion neurons. To determine if activin might also influence neuropeptide expression in developing sensory neurons, we examined whether type II activin receptors are expressed during embryonic development of the chicken dorsal root ganglia (DRG), and also examined the effects of activin on neuropeptide expression in cultured DRG neurons. Using reverse transcription polymerase chain reaction (rtPCR), we detected mRNAs for both the activin receptors type IIA (ActRIIA) and type IIB (ActRIIB) in DRG from embryonic day 7 through posthatch day 1. With in situ hybridization, we found that morphologically identifiable neurons express mRNAs for both ActRIIA and ActRIIB. With developmental age, a subset of neurons that hybridizes more intensely with riboprobes to these receptor mRNAs becomes evident. A similar pattern of expression is observed with immunocytochemical staining using antisera against activin type II receptors. To examine whether embryonic DRG cells respond to activin we treated dissociated cultures of DRG with activin A and assessed the expression of vasoactive intestinal peptide (VIP) and calcitonin gene related peptide (CGRP) mRNAs using semiquantitative rtPCR. Activin treatment results in an increase in VIP mRNA, but does not affect CGRP mRNA levels. These observations indicate that neurons in the embryonic chicken DRG can respond to activin and suggest that activin has the potential to play a role in the development and function of DRG sensory neurons.