The adrenergic differentiation of sympathetic neurons is controlled by complex interactions with the embryonic environment. To provide a basis for the experimental analysis of these interactions, the expression of the adrenergic marker enzyme tyrosine hydroxylase (TH) was analyzed by immunohistochemistry and in situ hybridization in sympathetic ganglia and the adrenal medulla of chick embryos. In parallel, the developmental expression of the transcription factors cPhox-2 and Cash-1 was analyzed by in situ hybridization. TH protein was first detectable during the third day of development (stage 19) in cells of the primary sympathetic strands. A few hours earlier (stage 18), TH mRNA could be found by in situ hybridization. At the very same time and location, mRNA for the transcription factor cPhox-2 was first observed. In contrast, mRNA for the transcription factor Cash-1 was detected much earlier, at stage 15, dorsal to the aorta where the primary sympathetic ganglia form. High TH mRNA levels are maintained during later embryonic development (stage 35) in both sympathetic ganglia and adrenal chromaffin cells. In contrast, cPhox-2 and Cash-1 mRNA are selectively reduced in chromaffin cells and sympathetic ganglia, respectively. The results show that TH and cPhox-2 are early markers expressed in sympathetic ganglia. Their coordinated expression points towards an inductive event possibly occurring close to the aorta and leading to the expression of an adrenergic phenotype. Cash-1 is detected significantly earlier, suggesting that its expression is induced by a separate event.

We have examined the regulation of transcription factor gene expression and phenotypic markers in developing chick sympathetic neurons. Sympathetic progenitor cells first express the bHLH transcriptional regulator Cash-1 (a chicken achaete-scute homologue), followed by coordinate expression of Phox2, a paired homeodomain protein, and GATA-2, a zinc finger protein. SCG10, a pan-neuronal membrane protein, is first detected one stage later, followed by the catecholaminergic neurotransmitter enzyme tyrosine hydroxylase (TH). We have used these markers to ask two questions: (1) is their expression dependent upon inductive signals derived from the notochord or floor plate?; (2) does their sequential expression reflect a single linear pathway or multiple parallel pathways? Notochord ablation experiments indicate that the floor plate is essential for induction of GATA-2, Phox2 and TH, but not for that of Cash-1 and SCG10. Taken together these data suggest that the development of sympathetic neurons involves multiple transcriptional regulatory cascades: one, dependent upon notochord or floor plate-derived signals and involving Phox2 and GATA-2, is assigned to the expression of the neurotransmitter phenotype; the other, independent of such signals and involving Cash-1, is assigned to the expression of pan-neuronal properties. The parallel specification of different components of the terminal neuronal phenotype is likely to be a general feature of neuronal development.

Transcriptional regulation of the gene encoding the cell adhesion receptor NCAM (neural cell adhesion molecule), a putative effector molecule of a variety of morphogenetic events, is likely to involve important regulators of morphogenesis. Here we identify two mouse homeodomain proteins that bind to an upstream regulatory element in the Ncam promoter: Cux, related to Drosophila cut and human CDP, and Phox2, a novel protein with a homeodomain related to that of the Drosophila paired gene. In transient transfection experiments, Cux was found to be a strong inhibitor of Ncam promoter activity, and this inhibition could be relieved by simultaneously overexpressing Phox2. These results suggest that the Ncam gene might be a direct target of homeodomain proteins and provide a striking example of regulatory cross-talk between homeodomain proteins of different classes. Whereas the expression pattern of Cux/CDP includes many NCAM-negative sites, Phox2 expression was restricted to cells also expressing Ncam or their progenitors. The localisation data thus strongly reinforce the notion that Phox2 plays a role in transcriptional activation of Ncam in Phox2-positive cell types. In the peripheral nervous system, Phox2 was strongly expressed in all ganglia of the autonomic nervous system and more weakly in some cranial sensory ganglia, but not in the sensory ganglia of the trunk. Phox2 transcripts were detected in the primordia of sympathetic ganglia as soon as they form. Phox2 expression in the brain was confined to spatially restricted domains in the hindbrain, which correspond to the noradrenergic and adrenergic nuclei once they are identifiable. All Phox2-expressing components of the peripheral nervous system are at least transiently adrenergic or noradrenergic. In the developing brain, Phox2 was expressed at all known locations of (nor)adrenergic neurones and of their precursors. These results suggest that Phox2, in addition to regulating the NCAM gene, may be part of the regulatory cascade that controls the differentiation of neurons towards this neurotransmitter phenotype.