Direct and indirect roles of Fgf3 and Fgf10 in innervation and vascularisation of the vertebrate hypothalamic neurohypophysis.

The neurohypophysis is a crucial component of the hypothalamo-pituitary axis, serving as the site of release of hypothalamic neurohormones into a plexus of hypophyseal capillaries. The growth of hypothalamic axons and capillaries to the forming neurohypophysis in embryogenesis is therefore crucial to future adult homeostasis. Using ex vivo analyses in chick and in vivo analyses in mutant and transgenic zebrafish, we show that Fgf10 and Fgf3 secreted from the forming neurohypophysis exert direct guidance effects on hypothalamic neurosecretory axons. Simultaneously, they promote hypophyseal vascularisation, exerting early direct effects on endothelial cells that are subsequently complemented by indirect effects. Together, our studies suggest a model for the integrated neurohemal wiring of the hypothalamo-neurohypophyseal axis.

The endocannabinoid receptor, CB1, is required for normal axonal growth and fasciculation.

Endocannabinoids are retrograde neurotransmitters, which act upon the presynaptically located, G-protein coupled receptor CB1, to modulate synaptic transmission in the adult brain. Recently, however, a number of lines of evidence have suggested that endocannabinoid signalling may play an important role in early neuronal development. In this study, we show that the CB1 receptor has a wide expression pattern in the developing nervous system and that its expression follows neuronal differentiation in the embryo from the earliest stages. We also show that the enzymes involved in 2-AG synthesis are expressed in an overlapping manner at these stages. We further show that interfering with CB1 function using a pharmacological inhibitor causes problems in axon pathfinding and fasciculation. Similarly, CB1 gene knock down in the zebrafish by morpholino injection results in defects in axonal growth and fasciculation in these embryos. Thus CB1 function is required in the early embryo for axonal growth and fasciculation.

Immunocytochemical developmental patterns of the thoracolumbar sympathetic chain in the chick and a comparison with its adrenal counterpart.

The immunocytochemical development of the thoracolumbar sympathetic ganglion and its adrenal counterpart was studied in the chick from days 3.5 to 12 of incubation, using antibodies to 17 separate antigens, including antibodies to pan-neuroendocrine markers, catecholamine-synthesizing and proprotein-processing enzymes, and neuropeptides. Some of the antigens studied (Go protein-alpha subunit, thyrosine hydroxylase, and galanin) were strongly expressed from the first days of development, whereas others (chromogranin-A, chromogranin-B, 7B2 protein, and somatostatin) showed a diverse immunoreactive expression at different stages. Three different patterns were found in the development of both adrenal medulla and thoracolumbar sympathetic ganglion. In the first (chromogranin-A and B, Go protein-alpha subunit, tyrosine hydroxylase, HNK-1, and galanin), virtually all medullary and thoracolumbar sympathetic ganglion cells were strongly immunostained from day 4 onward. Except for HNK-1, chromogranin-A and B, there was a steady increase in immunoreactive cells for all the remaining antigens up to day 12. In the second (7B2 protein, proprotein convertase 2, and secretogranin II), full antigenic expression was reached in medullary and thoracolumbar sympathetic ganglion cells by day 10. In the third pattern (proprotein convertase 3, somatostatin, dopamine-beta-hydroxylase, neuron-specific enolase, vasoactive intestinal polypeptide, and met-enkephalin), differences in immunoreactivity were observed between the medullary and thoracolumbar sympathetic ganglion cells.

Origins of the neurovascular bundle: interactions between developing nerves and blood vessels in embryonic chick skin.

Growth cones of nerves and endothelial cells of blood vessels are closely analogous in their migratory behavior, and they are both set a similar task during the early development of a limb. Both must invade the mesenchyme to form ramifying networks of large nerves and vessels. Both systems must densely pervade certain regions of the developing limb, such as muscle rudiments, and both form dense cutaneous plexuses at precisely the same depth beneath the epidermis. Moreover, adult tissues show many examples of neurovascular bundles in which nerves and blood vessels run closely parallel and branch in a correlated fashion, suggesting some interdependence during development. We have examined the interrelationship between developing nerves and blood vessels in chick wing skin because it allows a particularly convenient two-dimensional analysis of the two systems which can be revealed simultaneously in the same preparation by injection of Indian ink combined with silver-staining. We show that nerves do not use blood vessels as pathways along which to crawl, but that there are two other ways in which neurovascular associations arise: in some situations nerves and blood vessels follow the same route because they are responding independently to the same mesenchymal cues; and in some situations nerves induce blood vessels to remodel around them.

Apical accumulation of MARCKS in neural plate cells during neurulation in the chick embryo.

BACKGROUND: The neural tube is formed by morphogenetic movements largely dependent on cytoskeletal dynamics. Actin and many of its associated proteins have been proposed as important mediators of neurulation. For instance, mice deficient in MARCKS, an actin cross-linking membrane-associated protein that is regulated by PKC and other kinases, present severe developmental defects, including failure of cranial neural tube closure. RESULTS: To determine the distribution of MARCKS, and its possible relationships with actin during neurulation, chick embryos were transversely sectioned and double labeled with an anti-MARCKS polyclonal antibody and phalloidin. In the neural plate, MARCKS was found ubiquitously distributed at the periphery of the cells, being conspicuously accumulated in the apical cell region, in close proximity to the apical actin meshwork. This asymmetric distribution was particularly noticeable during the bending process. After the closure of the neural tube, the apically accumulated MARCKS disappeared, and this cell region became analogous to the other peripheral cell zones in its MARCKS content. Actin did not display analogous variations, remaining highly concentrated at the cell subapical territory. The transient apical accumulation of MARCKS was found throughout the neural tube axis. The analysis of another epithelial bending movement, during the formation of the lens vesicle, revealed an identical phenomenon. CONCLUSIONS: MARCKS is transiently accumulated at the apical region of neural plate and lens placode cells during processes of bending. This asymmetric subcellular distribution of MARCKS starts before the onset of neural plate bending. These results suggest possible upstream regulatory actions of MARCKS on some functions of the actin subapical meshwork.

Central retinal area is not the site where ganglion cells are generated first.

The development of retinal ganglion cells (RGC) was studied in the chick from stage 18 to adulthood. Our main objectives were to identify the retinal site where the first RGCs differentiate, to locate this site relative to the optically defined central retinal area, and to map the spatial arrangement of the RGC field at different stages in development. The eyes of the experimental animals were fixed and serially sectioned. The borders of RGC fields were determined from the presence of either ganglion cell perikarya or ganglion cell axons. In seven cases between stages 21 and 26, the borders of the RGC fields were confirmed electron microscopically. The serial sections together with the RGC fields were then reconstructed in three dimensions. The reconstructed retinae were projected onto a plane by using the radially equidistant polar azimuthal projection. First, RGCs appear dorsal to the apex of the optic fissure. Ganglion cell development then initially spreads out symmetrically with respect to the optic fissure. However, from stage 29 on, the nasal half of the retina expands much more than the temporal half. This asymmetrical growth entails that the optic fissure is eventually located in the temporal half of the retina in the mature animal. The RGC fields of the embryonic stages were superimposed on the retina of a visually active animal according to their real size and position. It turned out that the central retinal area was at least 2 mm away from the site where the first RGCs were generated. It is not before stage 28 that the prospective central retinal area is included into the expanding ganglion cell field. The fact that RGCs at the central retinal area are generated 2.5 days later than first RGCs near the apex of the optic fissure has important implications for the formation of the retinotectal projection.

Cell proliferation and cell death in the developing chick inner ear: spatial and temporal patterns.

Morphogenesis of the inner ear is a complex process in which the balance of cell division and death is presumed to play an important role. Surprisingly, there are no reports of a systematic comparison of these two processes in individual ears at different stages of development. This study presents such an analysis for the chicken otocyst at stages 13-29 (embryonic days 2.5-6). To detect proliferating cells, we used exposure to bromodeoxyuridine. To detect apoptotic cells, we used nuclear condensation and fragmentation or terminal dUTP nick-end labeling (TUNEL). The spatial and temporal locations of proliferating and dying cells were mapped across serial sections, revealing regional differences in proliferation within the otocyst epithelium that are more complex than previously reported. In addition, almost all of the previously identified "hot spots" of cell death correspond spatially to regions of reduced cell proliferation. An exception is the ventromedial hot spot of cell death, which is intermingled with proliferating cells when it first appears at stages 19-23 before becoming a cold spot of proliferation. The results further show that the inferior part of the otocyst has a high level of proliferation, whereas the superior part does not. Since the superior part of the otocyst demonstrates outward expansion that is comparable to the inferior part, it appears that regional outgrowth of the otic vesicle is not necessarily coupled to cell proliferation. This study provides a basis for exploring the regulation and function of cell proliferation and cell death during inner ear morphogenesis.

Effect of wound healing and tissue transplantation on the navigation of axons in organ-cultured embryonic chick eyes.

Wound closure and repair of embryonic neuroepithelium were studied in organ-cultured embryonic retinae. Eyes from 3 to 4-day-old embryos were cultured after removing pieces of retinal tissue. During the subsequent 24 hours of incubation, the 150 to 200 microns wide holes in the retina closed completely. Histological studies showed that the wound closure was not accomplished by cell migration or cell proliferation, but by an approximation of the wound edges mediated by extracellular matrix fibrils of the vitreous body. The wound contraction facilitated the integration of transplants into the retinal neuroepithelium with a perfect alignment of the implants with the host at the vitreal surface. Within 24 hours, a continuous inner limiting membrane between transplant and host retina was established. The effect of wound healing and tissue transplantation on the navigation of optic axons in the retina was investigated. The wound contraction in the retina caused the optic axons near the lesion site to grow to the wound center, where the axons traversed the retina and formed a neuroma at the ventricular side, resembling the organization of axons at the optic disc. In the transplantation paradigm, axons from the host retina migrated into the transplant and vice versa. However, due to the wound contraction around the transplant, most axons grew into the interface between the transplant and host tissue.

Differential expression of neuron-glia cell adhesion molecule (Ng-CAM) on developing axons and growth cones of interneurons in the chick embryo spinal cord: an immunoelectron microscopic study.

To elucidate the role of neuron-glia cell adhesion molecule (Ng-CAM) in axonal pathway formation of avian spinal interneurons, we have examined the ultrastructural expression of Ng-CAM in the developing spinal cord, by using a preembedding immunocytochemical method. Ng-CAM immunoreactivity was punctate and was restricted to cell surfaces. In accordance with our previous light microscopic observations (Shiga et al., '90), the earliest developing spinal interneurons were Ng-CAM-positive on their cell bodies, axons, and growth cones. Axons and growth cones that were either fasciculated or in contact with each other strongly expressed Ng-CAM, thus indicating the possible involvement of Ng-CAM in fasciculation of axons and in the contact guidance of growth cones along preexisting axons. By using higher resolution immunoelectron microscopy, the present study has also revealed new information on the subcellular localization of Ng-CAM on developing spinal interneurons, neuroepithelial cells, and floor plate cells. Although Ng-CAM immunoreactivity was prominent on both axons and growth cones, these structures were Ng-CAM-negative when they contacted the basal lamina around the spinal cord. By contrast, Ng-CAM was detectable on the surface of both neuroepithelial cells and floor plate cells only when they made contact with the Ng-CAM-positive axons and growth cones of interneurons. These results suggest that the subcellular distribution of Ng-CAM is regulated differentially, depending on the apposing cell surfaces, and that such differential and developmentally regulated expression may contribute to the elongation, fasciculation, and guidance of spinal axons.

Early development of efferent projections from the chick tectum.

The early development of the uncrossed tectobulbar and the crossed tectospinal tracts was studied. These two projections arise from the same structure, the mesencephalon, and develop during the same time period, but follow divergent courses. We have traced the pathways followed by these projections and identified the positions at which axon guidance decisions are made. The first neurons differentiate either side of the entire rostrocaudal extent of the dorsal midline and initiate axons that extend dorsoventrally across the surface of the tectum. At the ventral edge of the tectum these axons turn abruptly and fasciculate to form a caudal descending projection to the hindbrain. These axons extend to the caudal hindbrain and do not project to the periphery along cranial nerve roots. We therefore consider this tract to be the tectobular, rather than the mesencephalic division of the trigeminal. While the tectobulbar projection is still developing, a second wave of axons is initiated, which arises from only the rostral part of the tectum. These axons grow beyond the tectobulbar turn point and continue toward the ventral midline, where they cross the floor plate, before turning caudally at the lateral edge of the main descending hindbrain tract, the ventrolateral tract. We discuss the development of these tracts with reference to possible guidance cues mediating their course.

Maintenance of targeting errors by isthmo-optic axons following the intraocular injection of tetrodotoxin in chick embryos.

We have studied the role of electrical activity in the elimination of axonal targeting errors, which is a normal process in brain development. The experiments were focused on the isthmo-optic nucleus (ION), which, in adults, projects in topographical order on the contralateral retina. During embryogenesis, however, a few isthmo-optic neurons project to the ipsilateral retina, and many project to topographically inappropriate parts of the contralateral one; both kinds of targeting error are known to be eliminated by the deaths of the parent neurons. We injected tetrodotoxin (TTX) intraocularly at embryonic days 13 and 15 and, on the latter, applied a retrograde label to the retina of the same eye. Embryos were fixed at embryonic day 17. In some embryos, the label was a peripherally placed fleck of the carbocyanine dye "diI"; the resulting retrogradely labeled neurons in the contralateral ION were much more widely scattered in the TTX-injected embryos than in controls (errors in topography). In other embryos, the label was a solution of rhodamine-B-isothiocyanate (RITC) injected into the vitreous body; this yielded several ipsilaterally labeled isthmo-optic neurons in the TTX-injected embryos, but virtually none in the controls. The numbers of both kinds of aberrantly projecting neuron approached those previously reported near the beginning of the ION's period of neuronal death. We conclude that electrical activity plays an important role in the elimination of axonal targeting errors in the chick embryo's isthmo-optic system.

Changes in lamination and neuronal survival in the isthmo-optic nucleus following the intraocular injection of tetrodotoxin in chick embryos.

We have studied how the development of the isthmo-optic nucleus (ION) is affected by electrical activity in the ION's axonal target territory, the contralateral retina. Electrical activity was blocked or reduced in the retina for various periods by tetrodotoxin injected intraocularly in different doses. The effects on the morphology of the retina appear to have been minor. During the ION's period of naturally occurring neuronal death (embryonic days 12 to 17), the injections substantially reduced this neuronal death and disrupted the development of lamination in the contralateral ION; there was also a lesser reduction in neuronal death in the ipsilateral ION. The dose of tetrodotoxin required to affect lamination was lower than that affecting neuronal death. Thus, the effects on neuronal death and on lamination were independent, since either could occur without the other. These effects were mediated by retrograde signals (probably two or more) from the eye; they occurred too early for the alternative anterograde route via the optic tectum (which projects to the ION) to be responsible. After embryonic day 17, the ION's response to intraocular tetrodotoxin changes abruptly from increased survival to total and rapid degeneration.

The development of brain stem projections to the spinal cord in the chicken embryo.

Recent studies of the development of brain stem projections to the spinal cord in the chicken embryo, with an emphasis on axon pathway selection, are reviewed. Neurons from medullary to mesencephalic levels project to the spinal cord along specific fiber tracts. Coherent, segregated neuron groups can be defined on the basis of which tract and which side of the brain stem they project on. The choice of axon trajectory is, therefore, correlated with neuron position. During development, these trajectory-defined brain stem groups project to the spinal cord in a stereotyped sequence. Early stages of this sequence reveal a potential homology between the reticulospinal systems of avians and lower vertebrates. The possibility that neuron position may be involved in determining axon pathway choice of brain stem projections is supported by complementary studies on vestibuloocular projections. The boundaries between vestibuloocular, vestibulospinal, and reticulospinal neuron groups coincide with rhombomere boundaries and boundaries between longitudinal cell columns. Axon trajectory-specific domains are, therefore, correlated with segmental and mediolateral patterns of differential gene expression.

The segmental precision of the motor projection to the intercostal muscles in the developing chicken embryo. A differential labelling study using fluorescent tracers.

Each skeletal muscle in the vertebrate is innervated by a group of motoneurons called a motoneuron pool. Retrograde labelling of single motoneuron pools has suggested that the arrangement of motoneuron pools innervating different limb muscles does not change during the embryonic period when more than 50% of the motoneurons die. In this study we retrogradely labelled neighbouring intercostal motoneuron pools differentially with latex microspheres or dextran amines coupled to fluorescent dyes. We then mapped the positions of the differentially labelled motoneurons in whole-mount preparations using a computer-aided drawing system. While the intercostal motoneuron pools are clearly segregated even at early stages, there is some intermingling at the rostral and caudal ends. We used a logistic regression to determine the extent of segmental overlap, and to facilitate a quantitative comparison of the overlap at different stages. Statistical analysis shows that the overlap (expressed as the percentage of the length of the overlapping motoneuron pools) decreases modestly during the period of motoneuron death. Computer simulations suggest that this decrease does not result from random motoneuron death alone; one alternative possibility is selective death of motoneurons in the overlap zone. Occasional "rogue" motoneurons, that is, motoneurons of one pool that scatter into the neighbouring pool, are still present at the end of the period of cell death, representing a potential source of "noise" in the establishment of segmental patterns of connectivity.

A simple method for screening photoelectric dyes towards their use for retinal prostheses.

Photoelectric dyes absorb light and convert photon energy to electric potentials. To test whether these dyes could be used for retinal prostheses, a simple in vitro screening system was developed. Retinal neurons were cultured from the eyes of chick embryos at the 10-day embryonic stage, at which time no retinal photoreceptor cells have yet developed. Intracellular calcium elevation was observed with Fluo-4 in cultured retinal neurons before and after photoelectric dye was applied at varying concentrations to the culture medium. Five of 7 photoelectric dyes tested in this in vitro system induced intracellular calcium elevation in cultured chick retinal neurons. The intracellular calcium elevation generated by the 5 photoelectric dyes was blocked by extracellular calcium depletion in the case of all 5 dyes, and, except for one dye, by the presence of voltage-gated calcium channel blockers. The photoelectric dyes absorbed light under an inverted microscope and stimulated retinal neurons. This simple in vitro system allows the screening of photoelectric dyes which can be used for retinal prostheses.

Neuronal types in the chicken's statoacoustic ganglion.

In the hearing organs of vertebrates, bipolar sensory (afferent) neurons innervate the hair cells (sensory epithelium) by their dendritic (peripheral) processes. In the mammalian cochlea, two neuronal types (I and II) innervate the inner and outer hair cells, respectively. By comparison, the correspondence between neuronal and hair cell types is uncertain in the avian auditory lagena (cochlea). The objective of the present study was to identify and map the relative distributions of neuronal types in the chicken's statoacoustic ganglion (SAG) by means of morphometric analysis and immunoreactivity to neuro-specific enolase (NSE) and neuro-filament protein 200 kDa (NFP-200).

Emergence of hearing in the chicken embryo.

It is commonly held that hearing generally begins on incubation day 12 (E12) in the chicken embryo (Gallus domesticus). However, little is known about the response properties of cochlear ganglion neurons for ages younger than E18. We studied ganglion neurons innervating the basilar papilla of embryos (E12-E18) and hatchlings (P13-P15). We asked first, when do primary afferent neurons begin to encode sounds? Second, when do afferents evidence frequency selectivity? Third, what range of characteristic frequencies (CFs) is represented in the late embryo? Finally, how does sound transfer from air to the cochlea affect responses in the embryo and hatchling? Responses to airborne sound were compared with responses to direct columella footplate stimulation of the cochlea. Cochlear ganglion neurons exhibited a profound insensitivity to sound from E12 to E16 (stages 39-42). Responses to sound and frequency selectivity emerged at about E15. Frequency selectivity matured rapidly from E16 to E18 (stages 42 and 44) to reflect a mature range of CFs (170-4,478 Hz) and response sensitivity to footplate stimulation. Limited high-frequency sound transfer from air to the cochlea restricted the response to airborne sound in the late embryo. Two periods of ontogeny are proposed. First is a prehearing period (roughly E12-E16) of endogenous cochlear signaling that provides neurotrophic support and guides normal developmental refinements in central binaural processing pathways followed by a period (roughly E16-E19) wherein the cochlea begins to detect and encode sound.

The innervation of dorsoventrally reversed chick wings: evidence that motor axons do not actively seek out their appropriate targets.

In normal chick embryos the extensor (dorsal) muscles are innervated by motoneurones lying laterally in the motor horn, while flexor muscles are supplied by more medially placed motoneurones. After reversal of the dorsoventral axis of the forelimb prior to innervation in most cases the opposite pattern is found, the extensors innervated by medial and flexors by lateral motor neurones. In a minority of cases the normal innervation pattern is obtained. Three hypotheses are discussed, two involving specific target affinity between motor axon and target and one involving passive deployment of axons to targets. We conclude that our results favour the latter hypothesis but that we cannot exclude a short-range specific signal.

Development of cholinergic chronotropic control in chick (Gallus gallus domesticus) embryos.

In chick (Gallus gallus domesticus) embryos, instantaneous heart rate begins to fluctuate with the appearance of rapid, transient decelerations at around the end of the second week of incubation. Previously, it was shown that instantaneous heart rate decelerations were eliminated by administration of atropine and concurrently heart rate baseline was elevated in late embryos. Because the previous study lacked statistical treatment and there has been recent controversy over the development of tonic vagal control of the heart, we reexamine the hypothesis that transient decelerations of instantaneous heart rate are mediated by vagus nerve and the vagal tone begins to appear at around the end of the second week of incubation. Atropine administration tests were conducted for sixty-seven 11- to 14-day-old and 18-day-old embryos in total. Heart rate decelerations appeared sporadically in three out of ten 12-day-old embryos, but the difference of mode heart rate before and after administration of atropine was not significant. Seven out of nine 13-day-old embryos and all nine 14-day-old embryos showed heart rate decelerations and the difference of mode heart rate before and after atropine administration was significant. In late (18-day-old) embryos, magnitude and frequency of instantaneous heart rate decelerations further increased with additional appearance of transient, irregular accelerations. Administration of varying doses of atropine completely eliminated the heart rate decelerations and elevated the heart rate baseline more markedly than in young embryos, indicating the maturation of vagal tone late in incubation.