Segmentation of sensory and sympathetic ganglia: interactions between neural crest and somite cells.

The segmental pattern of peripheral ganglia in higher vertebrates is generated by interactions between neural crest and somite cells. Each mesodermal somite is subdivided into at least two distinct domains represented by its rostral and caudal halves. Most migratory pathways taken by neural crest cells in trunk regions of the axis, as well as the outgrowth of motoneuron fibers are restricted to the rostral domain of each somite. Experimental modification of the somites, achieved by constructing a mesoderm composed of multiple rostral half-somites, results in the formation of continuous and unsegmented nerves, dorsal root ganglia (DRG) and sympathetic ganglia (SG). In contrast, both neurites and crest cells are absent from a mesoderm composed of multiple-caudal half somites. However, the mechanisms responsible for gangliogenesis within the rostral half of the somite, appear to be different for DRG and SG. Vertebral development from the somites is also segmental. In implants of either multiple rostral or caudal somite-halves, the grafted mesoderm dissociates normally into sclerotome and dermomyotome. However, the morphogenetic capabilities of each somitic half differ. The lateral vertebral arch is continuous in the presence of caudal half-somite grafts and is virtually absent in rostral half-somite implants. Therefore, the rostrocaudal subdivision of the sclerotome determines the segmental pattern of neural development and is also important for the proper metameric development of the vertebrae.     

Normal segmentation and size of the primary sympathetic ganglia depend upon the alternation of rostrocaudal properties of the somites.

Metameric organization of the dorsal root ganglia (DRG) and ventral roots depends on the alternation of rostrocaudal properties within the somites. In addition, the size of DRG is likely to be regulated by the adjacent mesoderm, because unilateral creation of a paraxial mesoderm with only rostral somitic (RS) halves, leads to the development of non-segmented DRG that are larger and contain more cells than the sum of the contralateral, control DRG. We have now extended our studies of the role of the paraxial mesoderm in the morphogenesis of the peripheral nervous system (PNS) to another metameric PNS component, the sympathetic ganglia (SG). The development of the primary sympathetic chain was studied in chick-quail chimeras with multiple half-somite grafts using quantitative morphometric analysis. In the presence of an exclusively rostral or caudal somitic mesoderm, segmentation of the initially homogeneous primary sympathetic chain into ganglia is prevented. Therefore, the SG, like the DRG and ventral roots, require the normal rostrocaudal alternation of the somitic mesoderm for segmental morphogenesis. On embryonic day 4 (E4), there is a 38% average decrease in the volume of the primary sympathetic chain opposite a RS mesoderm, compared with the primary chain on the unoperated side. This is in contrast to the average increase of 27% in the volume of the DRG opposite the grafted mesoderm in the same embryos. Our results, and classical observations, have led us to propose a model in which the mesoderm controls DRG and SG size by modulating the partition of migrating NC precursors between the anlage of these two ganglion types. According to this model, the reduction in SG volume and concomitant increase in DRG volume observed opposite RS grafts, results from the arrest in the DRG anlage of neural crest cells that normally migrate to the SG.     

Determination of epithelial half-somites in skeletal morphogenesis.

The segmental body plan of vertebrates arises from the metameric organization of the paraxial mesoderm into somites. Each mesodermal somite is subdivided into at least two distinct domains: rostral and caudal. The segmental pattern of dorsal root ganglia, sympathetic ganglia and nerves is imposed by differential properties of either somitic domain. In the present work, we have extended these studies by investigating the contribution of rostral or caudal-half somites to vertebral development using grafts of multiple somite halves. In both rostral and caudal somitic implants, the grafted mesoderm dissociates normally into sclerotome and dermomyotome, and the sclerotome further develops into vertebrae. However, the morphogenetic capabilities of each somitic half differ. The pedicle of the vertebral arch is almost continuous in caudal half-somite grafts and is virtually absent in rostral half-somite implants. Similarly, the intervertebral disk is present in rostral half-somite chimeras, and much reduced or virtually absent in caudal somite chimeras. Thus, only the caudal half cells are committed to give rise to the vertebral pedicle, and only the rostral half cells are committed to give rise to the fibrocartilage of the intervertebral disk. Each vertebra is therefore composed of a pedicle-containing area, apparently formed by the caudal half-somite, followed by a pedicle-free zone, the intervertebral foramen, derived from the rostral somite. These data directly support the hypothesis of resegmentation, in which vertebrae arise by fusion of the caudal and rostral halves of two consecutive somites.     

Formation of the dorsal root ganglia in the avian embryo: Segmental origin and migratory behavior of neural crest progenitor cells

The segmental origin and migratory pattern of neural crest cells at the trunk level of avian embryos was studied, with special emphasis on the formation of the dorsal root ganglia (DRG) which organize in the anterior half of each somite. Neural crest cells were visualized using the quail-chick marker and HNK-1 immunofluorescence. The migratory process turned out to be closely correlated with somitic development: when the somites are epithelial in structure few labeled cells were found in a dorsolateral position on the neural tube, uniformly distributed along the craniocaudal axis. Following somitic dissociation into dermomyotome and sclerotome labeled cells follow defined migratory pathways restricted to each anterior somitic half. In contrast, opposite the posterior half of the somites, cells remain grouped in a dorsolateral position on the neural tube. The fate of crest cells originating at the level of the posterior somitic half was investigated by grafting into chick hosts short segments of quail neural primordium, which ended at mid-somitic or at intersomitic levels. It was found that neural crest cells arising opposite the posterior somitic half participate in the formation of the DRG and Schwann cells lining the dorsal and ventral root fibers of the same somitic level as well as of the subsequent one, whereas those cells originating from levels facing the anterior half of a somite participate in the formation of the corresponding DRG. Moreover, crest cells from both segmental halves segregate within each ganglion in a distinct topographical arrangement which reflects their segmental origin on the neural primordium. Labeled cells which relocate from posterior into anterior somitic regions migrate longitudinally along the neural tube. Longitudinal migration of neural crest cells was first observed when the somites are epithelial in structure and is completed after the disappearance of the last cells from the posterior somitic region at a stage corresponding to the organogenesis of the DRG.     

Regulation of trunk myogenesis by the neural crest: a new facet of neural crest-somite interactions

It is well established that the somitic mesoderm regulates early stages of neural crest development and further segmentation of crest-derived peripheral ganglia. The possibility that neural crest progenitors feed back on the somites was, however, not explored. Two recent studies provide evidence that the neural crest regulates somite-derived myogenesis by distinct mechanisms.     

Consequences of somite manipulation on the pattern of dorsal root ganglion development

We have investigated dorsal root ganglion formation, in the avian embryo, as a function of the composition of the paraxial somitic mesoderm. Three or four contiguous young somites were unilaterally removed from chick embryos and replaced by multiple cranial or caudal half-somites from quail embryos. Migration of neural crest cells and formation of DRG were subsequently visualized both by the HNK-1 antibody and the Feulgen nuclear stain. At advanced migratory stages (as defined by Teillet et al. Devl Biol. 120, 329-347 1987), neural crest cells apposed to the dorsolateral faces of the neural tube were distributed in a continuous, nonsegmented pattern that was indistinguishable on unoperated sides and on sides into which either half of the somites had been grafted. In contrast, ventrolaterally, neural crest cells were distributed segmentally close to the neural tube and within the cranial part of each normal sclerotome, whereas they displayed a nonsegmental distribution when the graft involved multiple cranial half-somites or were virtually absent when multiple caudal half-somites had been implanted. In spite of the identical dorsal distribution of neural crest cells in all embryos, profound differences in the size and segmentation of DRG were observed during gangliogenesis (E4-9) according to the type of graft that had been performed. Thus when the implant consisted of compound cranial half-somites, giant, coalesced ganglia developed, encompassing the entire length of the graft. On the other hand, very small, dorsally located ganglia with irregular segmentation were seen at the level corresponding to the graft of multiple caudal half-somites. We conclude that normal morphogenesis of dorsal root ganglia depends upon the craniocaudal integrity of the somites.     

Organization and differentiation of the segmentary plate in common toad embryo Bufo bufo L., following heat shock (author's transl)

A heat shock greatly disturbs the shaping and the growth of embryos with 4 or 5 somites. One part of a segmentary plate with a size including four control somites stays inorganized, whereas posterior segmentation gets on normally. The differentiation of the unsegmented somitic block (fusing myoblasts, myogenesis) is the same as the one of the previous somites; the dermatome and the sclerotome issued from the unsegmented block are not cut up. Ultrastructural observations of treated segmentary plates show that the somitic block undergoes important changes. The non-segmented somitic mesoderm of the control animals obtains a loose aspect; its constitutive cells separated by lacunae are brought into contact by the means of filopods and junctions. 18 hours after the heat shock, the cells of the unsegmented block keep close to each other and are tightened together by many "pseudo-desmosomes", the intercellular spaces are reduced to small meatus. The nuclei of the somitic cells are clear and provided with a restricted amount of chromatine spots; dark granular nuclear like chromatin material is thrown out of the nucleus. In embryos fixed 40 hours after heat shock, several large plurinucleated cells, are formed in the non-segmented somitic block. Inside these plurinucleated cells, there is synthesis of myofilaments and differentiation of myofibrilles. Numerous intercellular lacunae are formed, some yet apparent "pseudo-desmosomes" form "pinchings" between the membranes of muscular cells and therefore confine series of intercellular lacunae.     

Do Peanut Agglutinin Receptors on Somites Control the Behavior of Neural Cells?

Peanut agglutinin (PNA) receptors are expressed in the caudal halves of sclerotomes in chick embryos after 3 days of incubation (stages 19–20 of Hamburger & Hamilton). The neural crest cells forming dorsal root ganglia (DRG) and motor nerves appear to avoid PNA positive regions and concentrate into rostral halves of sclerotomes. To investigate the role of PNA receptors in gangliogenesis and nerve growth, we examined PNA binding ability in quail sclerotomes and in chick-quail chimeric embryos made by transplanting quail somites to chick embryos, comparing the development of DRG, motor nerves and sclerotomes. PNA did not bind to any part of the somites of 4.5-day quail embryos, although dorsal root ganglia and motor nerves appeared only in the rostral halves of sclerotomes as in chick embryos. Moreover, in spite of no PNA binding ability of the transplanted quail somite in 4.5-day chick-quail chimeric embryos, DRG and motor nerves derived from chick tissues appeared only in the rostral halves of the sclerotomes derived from these somites. Thus, both quail and chick neural crest cells and motor nerves recognized the difference between the rostral and caudal halves of sclerotomes of quail embryos in the absence of PNA binding ability, indicating that PNA binding site on somite cells does not support the selective neural crest migration and nerve growth.     

Axolotls with an under- or oversupply of neural crest can regulate the sizes of their dorsal root ganglia to normal levels

How animals adjust the size of their organs is a fundamental, enduring question in biology. Here we manipulate the amount of neural crest (NC) precursors for the dorsal root ganglia (DRG) in axolotl. We produce embryos with an under- or over-supply of pre-migratory NC in order to find out if DRG can regulate their sizes during development. Axolotl embryos are perfectly suitable for this research. Firstly, they are optimal for microsurgical manipulations and tissue repair. Secondly, they possess, unlike most other vertebrates, only one neural crest string located on top of the neural tube. This condition and position enables NC cells to migrate to either side of the embryo and participate in the regulation of NC cell distribution. We show that size compensation of DRG in axolotl occurs in 2 cm juveniles after undersupply of NC (up-regulation) and in 5 cm juveniles after oversupply of NC (down-regulation). The size of DRG is likely to be regulated locally within the DRG and not via adaptations of the pre-migratory NC or during NC cell migration. Ipsi- and contralateral NC cell migration occurs both in embryos with one and two neural folds, and contralateral migration of NC is the only source for contralateral DRG formation in embryos with only one neural fold. Compensatory size increase is accompanied by an increase in cell division of a DRG precursor pool (PCNA+/SOX2−), rather than by DRG neurons or glial cells. During compensatory size decrease, increased apoptosis and reduced proliferation of DRG cells are observed.     

3D reconstructions of quail–chick chimeras provide a new fate map of the avian scapula

Limbed vertebrates have functionally integrated postcranial axial and appendicular systems derived from two distinct populations of embryonic mesoderm. The axial skeletal elements arise from the paraxial somites, the appendicular skeleton and sternum arise from the somatic lateral plate mesoderm, and all of the muscles for both systems arise from the somites. Recent studies in amniotes demonstrate that the scapula has a mixed mesodermal origin. Here we determine the relative contribution of somitic and lateral plate mesoderm to the avian scapula from quail-chick chimeras. We generate 3D reconstructions of the grafted tissue in the host revealing a very different distribution of somitic cells in the scapula than previously reported. This novel 3D visualization of the cryptic border between somitic and lateral plate populations reveals the dynamics of musculoskeletal morphogenesis and demonstrates the importance of 3D visualization of chimera data. Reconstructions of chimeras make clear three significant contrasts with existing models of scapular development. First, the majority of the avian scapula is lateral plate derived and the somitic contribution to the scapular blade is significantly smaller than in previous models. Second, the segmentation of the somitic component of the blade is partially lost; and third, there are striking differences in growth rates between different tissues derived from the same somites that contribute to the structures of the cervical thoracic transition, including the scapula. These data call for the reassessment of theories on the development, homology, and evolution of the vertebrate scapula.     

Determination of somite cells: independence of cell differentiation and morphogenesis

Somites are mesodermal structures which appear transiently in vertebrates in the course of their development. Cells situated ventromedially in a somite differentiate into the sclerotome, which gives rise to cartilage, while the other part of the somite differentiates into dermomyotome which gives rise to muscle and dermis. The sclerotome is further divided into a rostral half, where neural crest cells settle and motor nerves grow, and a caudal half. To find out when these axes are determined and how they rule later development, especially the morphogenesis of cartilage derived from the somites, we transplanted the newly formed three caudal somites of 2.5-day-old quail embryos into chick embryos of about the same age, with reversal of some axes. The results were summarized as follows. (1) When transplantation reversed only the dorsoventral axis, one day after the operation the two caudal somites gave rise to normal dermomyotomes and sclerotomes, while the most rostral somite gave rise to a sclerotome abnormally situated just beneath ectoderm. These results suggest that the dorsoventral axis was not determined when the somites were formed, but began to be determined about three hours after their formation. (2) When the transplantation reversed only the rostrocaudal axis, two days after the operation the rudiments of dorsal root ganglia were formed at the caudal (originally rostral) halves of the transplanted sclerotomes. The rostrocaudal axis of the somites had therefore been determined when the somites were formed. (3) When the transplantation reversed both the dorsoventral and the rostrocaudal axes, two days after the operation, sclerotomes derived from the prospective dermomyotomal region of the somites were shown to keep their original rostrocaudal axis, judging from the position of the rudiments of ganglia. Combined with results 1 and 2, this suggested that the fate of the sclerotomal cells along the rostrocaudal axis was determined previously and independently of the determination of somite cell differentiation into dermomyotome and sclerotome. (4) In the 9.5-day-old chimeric embryos with rostrocaudally reversed somites, the morphology of vertebrae and ribs derived from the explanted somites were reversed along the rostrocaudal axis. The morphology of cartilage derived from the somites was shown to be determined intrinsically in the somites by the time these were formed from the segmental plate. The rostrocaudal pattern of the vertebral column is therefore controlled by factors intrinsic to the somitic mesoderm, and not by interactions between this mesoderm and the notochord and/or neural tube, arising after segmentation.     

Hox genes and morphological identity: axial versus lateral patterning in the vertebrate mesoderm

The successful organization of the vertebrate body requires that local information in the embryo be translated into a functional, global pattern. Somite cells form the bulk of the musculoskeletal system. Heterotopic transplants of segmental plate along the axis from quail to chick were performed to test the correlation between autonomous morphological patterning and Hox gene expression in somite subpopulations. The data presented strengthen the correlation of Hox gene expression with axial specification and focus on the significance of Hox genes in specific derivatives of the somites. We have defined two anatomical compartments of the body based on the embryonic origin of the cells making up contributing structures: the dorsal compartment, formed from purely somitic cell populations; and the ventral compartment comprising cells from somites and lateral plate. The boundary between these anatomical compartments is termed the somitic frontier. Somitic tissue transplanted between axial levels retains both original Hox expression and morphological identity in the dorsal compartment. In contrast, migrating lateral somitic cells crossing the somitic frontier do not maintain donor Hox expression but apparently adopt the Hox expression of the lateral plate and participate in the morphology appropriate to the host level. Dorsal and ventral compartments, as defined here, have relevance for experimental manipulations that influence somite cell behavior. The correlation of Hox expression profiles and patterning behavior of cells in these two compartments supports the hypothesis of independent Hox codes in paraxial and lateral plate mesoderm.     

Developmental Morphology of the Head Mesoderm and Reevaluation of Segmental Theories of the Vertebrate Head: Evidence from Embryos of an Agnathan Vertebrate, Lampetra japonica

Due to the peculiar morphology of its preotic head, lampreys have long been treated as an intermediate animal which links amphioxus and gnathostomes. To reevaluate the segmental theory of classical comparative embryology, mesodermal development was observed in embryos of a lamprey, Lampetra japonica, by scanning electron microscopy and immunohistochemistry. Signs of segmentation are visible in future postotic somites at an early neurula stage, whereas the rostral mesoderm is unsegmented and rostromedially confluent with the prechordal plate. The premandibular and mandibular mesoderm develop from the prechordal plate in a caudal to rostral direction and can be called the preaxial mesoderm as opposed to the caudally developing gastral mesoderm. With the exception of the premandibular mesoderm, the head mesodermal sheet is secondarily regionalized by the otocyst and pharyngeal pouches into the mandibular mesoderm, hyoid mesoderm, and somite 0. The head mesodermal components never develop into cephalic myotomes, but the latter develop only from postotic somites. These results show that the lamprey embryo shows a typical vertebrate phylotype and that the basic mesodermal configuration of vertebrates already existed prior to the split of agnatha-gnathostomata; lamprey does not represent an intermediate state between amphioxus and gnathostomes. Unlike interpretations of theories of head segmentation that the mesodermal segments are primarily equivalent along the axis, there is no evidence in vertebrate embryos for the presence of preotic myotomes. We conclude that mesomere-based theories of head metamerism are inappropriate and that the formulated vertebrate head should possess the distinction between primarily unsegmented head mesoderm which includes preaxial components at least in part and somites in the trunk which are shared in all the known vertebrate embryos as the vertebrate phylotype.     

Signals from trunk paraxial mesoderm induce pronephros formation in chick intermediate mesoderm

We used Pax-2 mRNA expression and Lim 1/2 antibody staining as markers for the conversion of chick intermediate mesoderm (IM) to pronephric tissue and Lmx-1 mRNA expression as a marker for mesonephros. Pronephric markers were strongly expressed caudal to the fifth somite by stage 9. To determine whether the pronephros was induced by adjacent tissues and, if so, to identify the inducing tissues and the timing of induction, we microsurgically dissected one side of chick embryos developing in culture and then incubated them for up to 3 days. The undisturbed contralateral side served as a control. Most embryos cut parallel to the rostrocaudal axis between the trunk paraxial mesoderm and IM before stage 8 developed a pronephros on the control side only. Embryos manipulated after stage 9 developed pronephric structures on both sides, but the caudal pronephric extension was attenuated on the cut side. These results suggest that a medial signal is required for pronephric development and show that the signal is propagated in a rostral to caudal sequence. In manipulated embryos cultured for 3 days in ovo, the mesonephros as well as the pronephros failed to develop on the experimental side. In contrast, embryos cut between the notochord and the trunk paraxial mesoderm formed pronephric structures on both sides, regardless of the stage at which the operation was performed, indicating that the signal arises from the paraxial mesoderm (PM) and not from axial mesoderm. This cut also served as a control for cuts between the PM and the IM and showed that signaling itself was blocked in the former experiments, not the migration of pronephric or mesonephric precursor cells from the primitive streak. Additional control experiments ruled out the need for signals from lateral plate mesoderm, ectoderm, or endoderm. To determine whether the trunk paraxial mesoderm caudal to the fifth somite maintains its inductive capacity in the absence of contact with more rostral tissue, embryos were transected. Those transected below the prospective level of the fifth somite expressed Pax-2 in both the rostral and the caudal isolates, whereas embryos transected rostral to this level expressed Pax-2 in the caudal isolate only. Thus, a rostral signal is not required to establish the normal pattern of Pax-2 expression and pronephros formation. To determine whether paraxial mesoderm is sufficient for pronephros induction, stage 7 or earlier chick lateral plate mesoderm was cocultured with caudal stage 8 or 9 quail somites in collagen gels. Pax-2 was expressed in chick tissues in 21 of 25 embryos. Isochronic transplantation of stage 4 or 5 quail node into caudal chick primitive streak resulted in the generation of ectopic somites. These somites induced ectopic pronephroi in lateral plate mesoderm, and the IM that received signals from both native and ectopic somites formed enlarged pronephroi with increased Pax-2 expression. We conclude that signals from a localized region of the trunk paraxial mesoderm are both required and sufficient for the induction of the pronephros from the chick IM. Studies to identify the molecular nature of the induction are in progress.     

The foot formation stimulating peptide pedibin is also involved in patterning of the head in hydra

Avian neural crest cells migrate on precise pathways to their target areas where they form a wide variety of cellular derivatives, including neurons, glia, pigment cells and skeletal components. In one portion of their pathway, trunk neural crest cells navigate in the somitic mesoderm in a segmental fashion, invading the rostral, while avoiding the caudal, half-sclerotome. This pattern of cell migration, imposed by the somitic mesoderm, contributes to the metameric organization of the peripheral nervous system, including the sensory and sympathetic ganglia. At hindbrain levels, neural crest cells also travel from the neural tube in a segmental manner via three migratory streams of cells that lie adjacent to even-numbered rhombomeres. In this case, the adjacent mesoderm does not possess an obvious segmental organization, compared to the somitic mesoderm at trunk levels. Thus, the mechanisms by which the embryo controls segmentally-organized cell migrations have been a fascinating topic over the past several years. Here, I discuss findings from classical and recent studies that have delineated several of the tissue, cellular and molecular elements that contribute to the segmental organization of neural crest migration, primarily in the avian embryo. One common theme is that neural crest cells are prohibited from entering particular territories in the embryo due to the expression of inhibitory factors. However, permissive, migration-promoting factors may also play a key role in coordinating neural crest migration.     

The somitic level of origin of embryonic chick hindlimb muscles

Studies of avian chimeras made by transplanting groups of quail somites into chick embryos have consistently shown that the muscle cells of the hindlimb are derived from the adjacent somites, however, the pattern of cell distribution from individual somites to individual hindlimb muscles has not been characterized. I have mapped quail cell distribution in the chick hindlimb after single somite transplantation to determine if cells from an individual somite populate discrete limb muscle regions and if there is a spatial correspondence between a muscle's somitic level of origin and the known spinal cord position of its motoneuron pool. At stages 15-18 single chick somites or equivalent lengths of unsegmented somitic mesoderm adjacent to the prospective hindlimb region were replaced with the corresponding tissue from quail embryos. At stages 28-34, quail cell distribution was mapped within individual thigh muscles and shank muscle regions. A quail-specific antiserum and Feulgen staining were used to identify quail cells. Transplants from somite levels 26-33 each gave rise to consistent quail cell labeling in a unique subset of limb muscles. The anteroposterior positions of these subsets corresponded to that of the transplanted somitic tissue. For example, more anterior or anteromedial thigh muscles contained quail cells when more anterior somitic tissue had been transplanted. For the majority of thigh muscles studied and for shank muscle groups, there was also a clear correlation between somitic level of origin and motoneuron pool position. These data are compatible with the hypothesis that motoneurons and the muscle cells of their targets share axial position labels. The question of whether motoneurons from a specific spinal cord segment recognize and consequently innervate muscle cells derived from the same axial level during early axon outgrowth is addressed in the accompanying paper (C. Lance-Jones, 1988, Dev. Biol. 126, 408-419). Quail cell distribution was also mapped in chick embryos in which quail somites or unsegmented mesoderm had been placed 2-3 somites away from their position of origin. In all cases donor somitic tissues contributed to muscles in accord with their host position. These results indicate that muscle cell precursors within the somites are not specified to migrate to a predetermined target region.     

Requirement of a neural tube signal for the differentiation of neural crest cells into dorsal root ganglia

The influence of the neural tube on early development of neural crest cells into sensory ganglia was studied in the chick embryo. Silastic membranes were implanted between the neural tube and the somites in 30-somite-stage embryos at the level of somites 21-24, thus separating the early migrated population of neural crest cells from the neural tube. Neural crest cells and peripheral ganglia were visualized by immunofluorescence using the HNK-1 monoclonal antibody and several histochemical techniques. Separation of crest cells from the neural tube caused the selective death of the neural crest cells from which dorsal root ganglia (DRG) would have developed. Complete disappearance of HNK-1 positive cells was evident already 10 hr after silastic implantation, before early differentiation sensory neurons could have reached their peripheral targets. In older embryos, DRG were absent at the level of implantation. In contrast, the development of ventral roots, sympathetic ganglia and adrenal gland was normal, and so was somitic differentiation into cartilage and muscle, while morphogenesis of the vertebrae was perturbed. To overcome the experimentally induced crest cell death, the silastic membranes were impregnated with a 3-day-old embryonic chick neural tube extract. Under these conditions, crest cells which were separated from the tube survived for a period of 30 hr after operation, compared to less than 10 hr in respective controls. The extract of another tissue, the liver, did not protract survival of DRG progenitor cells. Among the cells which survived with neural tube extract, some even succeeded in extending neurites; nevertheless, in absence of normal connections with the central nervous system (CNS) they finally died. Treatment of silastic implanted embryos with nerve growth factor (NGF) did not prevent the experimentally induced crest cell death. These results demonstrate that DRG develop from a population of neural crest cells which depends for its survival and probably for its differentiation upon a signal arising from the CNS, needed as early as the first hours after initiation of migration. Recovery experiments suggest that the subpopulation of crest cells which will develop along the sensory pathway probably depends for its survival and/or differentiation upon a factor contained in the neural tube, which is different from NGF.