Development of GABA-immunoreactive neuron patterning in the spinal cord

Roles of cerebrospinal fluid-contacting neurons as potential neural stem cells in the repair and regeneration of spinal cord injuries

Cerebrospinal fluid-contacting neurons (CSF-cNs) represent a distinct group of interneurons characterized by their prominent apical globular protrusions penetrating the spinal cord's central canal and their basal axons extending towards adjacent cells. Identified nearly a century back, the specific roles and attributes of CSF-cNs have just started to emerge due to the historical lack of definitive markers. Recent findings have confirmed that CSF-cNs expressing PKD2L1 possess attributes of neural stem cells, suggesting a critical function in the regeneration processes following spinal cord injuries. This review aims to elucidate the molecular markers of CSF-cNs as potential neural stem cells during spinal cord development and assess their roles post-spinal cord injury, with an emphasis on their potential therapeutic implications for spinal cord repair.

Cerebrospinal fluid-contacting neurons: multimodal cells with diverse roles in the CNS

The cerebrospinal fluid (CSF) is a complex solution that circulates around the CNS, and whose composition changes as a function of an animal's physiological state. Ciliated neurons that are bathed in the CSF - and thus referred to as CSF-contacting neurons (CSF-cNs) - are unusual polymodal interoceptive neurons. As chemoreceptors, CSF-cNs respond to variations in pH and osmolarity and to bacterial metabolites in the CSF. Their activation during infections of the CNS results in secretion of compounds to enhance host survival. As mechanosensory neurons, CSF-cNs operate together with an extracellular proteinaceous polymer known as the Reissner fibre to detect compression during spinal curvature. Once activated, CSF-cNs inhibit motor neurons, premotor excitatory neurons and command neurons to enhance movement speed and stabilize posture. At longer timescales, CSF-cNs instruct morphogenesis throughout life via the release of neuropeptides that act over long distances on skeletal muscle. Finally, recent evidence suggests that mouse CSF-cNs may act as neural stem cells in the spinal cord, inspiring new paths of investigation for repair after injury.

The Genetic Programs Specifying Kolmer-Agduhr Interneurons

Kolmer-Agduhr (KA) cells are a subgroup of interneurons positioned adjacent to the neurocoele with cilia on the apical surface protruding into the central canal of the spinal cord. Although KA cells were identified almost a century ago, their development and functions are only beginning to be unfolded. Recent studies have revealed the characteristics of KA cells in greater detail, including their spatial distribution, the timing of their differentiation, and their specification via extrinsic signaling and a unique combination of transcription factors in zebrafish and mouse. Cell lineage-tracing experiments have demonstrated that two subsets of KA cells, named KA' and KA" cells, differentiate from motoneuronal progenitors and floor-plate precursors, respectively, in both zebrafish and mouse. Although KA' and KA" cells originate from different progenitors/precursors, they each share a common set of transcription factors. Intriguingly, the combination of transcription factors that promote the acquisition of KA' cell characteristics differs from those that promote a KA" cell identity. In addition, KA' and KA" cells exhibit separable neuronal targets and differential responses to bending of the spinal cord. In this review, we summarize what is currently known about the genetic programs defining the identities of KA' and KA" cell identities. We then discuss how these two subgroups of KA cells are genetically specified.

Light on a sensory interface linking the cerebrospinal fluid to motor circuits in vertebrates

The cerebrospinal fluid (CSF) is circulating around the entire central nervous system (CNS). The main function of the CSF has been thought to insure the global homeostasis of the CNS. Recent evidence indicates that the CSF also dynamically conveys signals modulating the development and the activity of the nervous system. The later observation implies that cues from the CSF could act on neurons in the brain and the spinal cord via bordering receptor cells. Candidate neurons to enable such modulation are the cerebrospinal fluid-contacting neurons (CSF-cNs) that are located precisely at the interface between the CSF and neuronal circuits. The atypical apical extension of CSF-cNs bears a cluster of microvilli bathing in the CSF indicating putative sensory or secretory roles in relation with the CSF. In the brainstem and spinal cord, CSF-cNs have been described in over two hundred species by Kolmer and Agduhr, suggesting an important function within the spinal cord. However, the lack of specific markers and the difficulty to access CSF-cNs hampered their physiological investigation. The transient receptor potential channel PKD2L1 is a specific marker of spinal CSF-cNs in vertebrate species. The transparency of zebrafish at early stages eases the functional characterization of pkd2l1⁺ CSF-cNs. Recent studies demonstrate that spinal CSF-cNs detect spinal curvature via the channel PKD2L1 and modulate locomotion and posture by projecting onto spinal interneurons and motor neurons in vivo. In vitro recordings demonstrated that spinal CSF-cNs are sensing pH variations mainly through ASIC channels, in combination with PKD2L1. Altogether, neurons contacting the CSF appear as a novel sensory modality enabling the detection of mechanical and chemical stimuli from the CSF and modulating the excitability of spinal circuits underlying locomotion and posture.

Investigation of spinal cerebrospinal fluid-contacting neurons expressing PKD2L1: evidence for a conserved system from fish to primates

Over 90 years ago, Kolmer and Agduhr identified spinal cerebrospinal fluid-contacting neurons (CSF-cNs) based on their morphology and location within the spinal cord. In more than 200 vertebrate species, they observed ciliated neurons around the central canal that extended a brush of microvilli into the cerebrospinal fluid (CSF). Although their morphology is suggestive of a primitive sensory cell, their function within the vertebrate spinal cord remains unknown. The identification of specific molecular markers for these neurons in vertebrates would benefit the investigation of their physiological roles. PKD2L1, a transient receptor potential channel that could play a role as a sensory receptor, has been found in cells contacting the central canal in mouse. In this study, we demonstrate that PKD2L1 is a specific marker for CSF-cNs in the spinal cord of mouse (Mus musculus), macaque (Macaca fascicularis) and zebrafish (Danio rerio). In these species, the somata of spinal PKD2L1⁺ CSF-cNs were located below or within the ependymal layer and extended an apical bulbous extension into the central canal. We found GABAergic PKD2L1-expressing CSF-cNs in all three species. We took advantage of the zebrafish embryo for its transparency and rapid development to identify the progenitor domains from which pkd2l1⁺ CSF-cNs originate. pkd2l1⁺ CSF-cNs were all GABAergic and organized in two rows—one ventral and one dorsal to the central canal. Their location and marker expression is consistent with previously described Kolmer–Agduhr cells. Accordingly, pkd2l1⁺ CSF-cNs were derived from the progenitor domains p3 and pMN defined by the expression of nkx2.2a and olig2 transcription factors, respectively. Altogether our results suggest that a system of CSF-cNs expressing the PKD2L1 channel is conserved in the spinal cord across bony vertebrate species.

Genome wide analysis of Silurana (Xenopus) tropicalis development reveals dynamic expression using network enrichment analysis

Development involves precise timing of gene expression and coordinated pathways for organogenesis and morphogenesis. Functional and sub-network enrichment analysis provides an integrated approach for identifying networks underlying development. The objectives of this study were to characterize early gene regulatory networks over Silurana tropicalis development from NF stage 2 to 46 using a custom Agilent 4×44K microarray. There were >8000 unique gene probes that were differentially expressed between Nieuwkoop-Faber (NF) stage 2 and stage 16, and >2000 gene probes differentially expressed between NF 34 and 46. Gene ontology revealed that genes involved in nucleosome assembly, cell division, pattern specification, neurotransmission, and general metabolism were increasingly regulated throughout development, consistent with active development. Sub-network enrichment analysis revealed that processes such as membrane hyperpolarisation, retinoic acid, cholesterol, and dopamine metabolic gene networks were activated/inhibited over time. This study identifies RNA transcripts that are potentially maternally inherited in an anuran species, provides evidence that the expression of genes involved in retinoic acid receptor signaling may increase prior to those involved in thyroid receptor signaling, and characterizes novel gene expression networks preceding organogenesis which increases understanding of the spatiotemporal embryonic development in frogs.

Contexts for dopamine specification by calcium spike activity in the CNS

Calcium-dependent electrical activity plays a significant role in neurotransmitter specification at early stages of development. To test the hypothesis that activity-dependent differentiation depends on molecular context, we investigated the development of dopaminergic neurons in the CNS of larval Xenopus laevis. We find that different dopaminergic nuclei respond to manipulation of this early electrical activity by ion channel misexpression with different increases and decreases in numbers of dopaminergic neurons. Focusing on the ventral suprachiasmatic nucleus and the spinal cord to gain insight into these differences, we identify distinct subpopulations of neurons that express characteristic combinations of GABA and neuropeptide Y as cotransmitters and Lim1,2 and Nurr1 transcription factors. We demonstrate that the developmental state of neurons identified by their spatial location and expression of these molecular markers is correlated with characteristic spontaneous calcium spike activity. Different subpopulations of dopaminergic neurons respond differently to manipulation of this early electrical activity. Moreover, retinohypothalamic circuit activation of the ventral suprachiasmatic nucleus recruits expression of dopamine selectively in reserve pool neurons that already express GABA and neuropeptide Y. The results are consistent with the hypothesis that spontaneously active neurons expressing GABA are most susceptible to activity-dependent expression of dopamine in both the spinal cord and brain. Because loss of dopaminergic neurons plays a role in neurological disorders such as Parkinson's disease, understanding how subpopulations of neurons become dopaminergic may lead to protocols for differentiation of neurons in vitro to replace those that have been lost in vivo.

GABA: a pioneer transmitter that excites immature neurons and generates primitive oscillations

Developing networks follow common rules to shift from silent cells to coactive networks that operate via thousands of synapses. This review deals with some of these rules and in particular those concerning the crucial role of the neurotransmitter gamma-aminobuytric acid (GABA), which operates primarily via chloride-permeable GABA(A) receptor channels. In all developing animal species and brain structures investigated, neurons have a higher intracellular chloride concentration at an early stage leading to an efflux of chloride and excitatory actions of GABA in immature neurons. This triggers sodium spikes, activates voltage-gated calcium channels, and acts in synergy with NMDA channels by removing the voltage-dependent magnesium block. GABA signaling is also established before glutamatergic transmission, suggesting that GABA is the principal excitatory transmitter during early development. In fact, even before synapse formation, GABA signaling can modulate the cell cycle and migration. The consequence of these rules is that developing networks generate primitive patterns of network activity, notably the giant depolarizing potentials (GDPs), largely through the excitatory actions of GABA and its synergistic interactions with glutamate signaling. These early types of network activity are likely required for neurons to fire together and thus to "wire together" so that functional units within cortical networks are formed. In addition, depolarizing GABA has a strong impact on synaptic plasticity and pathological insults, notably seizures of the immature brain. In conclusion, it is suggested that an evolutionary preserved role for excitatory GABA in immature cells provides an important mechanism in the formation of synapses and activity in neuronal networks.

Expression of the glycinergic system during the course of embryonic development in the mouse spinal cord and its co-localization with GABA immunoreactivity

To understand better the role of glycine and gamma-aminobutyric acid (GABA) in the mouse spinal cord during development, we previously described the ontogeny of GABA. Now, we present the ontogeny of glycine-immunoreactive (Gly-ir) somata and fibers, at brachial and lumbar levels, from embryonic day 11.5 (E11.5) to postnatal day 0 (P0). Spinal Gly-ir somata appeared at E12.5 in the ventral horn, with a higher density at the brachial level. They were intermingled with numerous Gly-ir fibers reaching the border of the marginal zone. By E13.5, at the brachial level, the number of Gly-ir perikarya sharply increased throughout the whole ventral horn, whereas the density of fibers declined in the marginal zone. In the dorsal horn, the first Gly-ir somata were then detected. From E13.5 to E16.5, at the brachial level, the density of Gly-ir cells remained stable in the ventral horn, and after E16.5 it decreased to reach a plateau. In the dorsal horn, the density of Gly-ir cells increased, and after E16.5 it remained stable. At the lumbar level, maximum expression was reached at E16.5 in both the ventral and dorsal horn. Finally, the co-localization of glycine and GABA was analyzed, in the ventral motor area, at E13.5, E15.5, and E17.5. The results showed that, regardless of developmental stage studied, one-third of the stained somata co-expressed GABA and glycine. Our data show that the glycinergic system matures 1 day later than the GABAergic system and follows a parallel spatiotemporal evolution, leading to a larger population of glycine cells in the ventral horn.

Ontogenic changes of the spinal GABAergic cell population are controlled by the serotonin (5-HT) system: implication of 5-HT1 receptor family

During the development of the nervous system, the acquisition of the GABA neurotransmitter phenotype is crucial for neural networks operation. Although both intrinsic and extrinsic signals such as transcription factors and growth factors have been demonstrated to govern the acquisition of GABA, few data are available concerning the effects of modulatory transmitters expressed by axons that progressively invade emerging neuronal networks. Among such transmitters, serotonin (5-HT) is a good candidate because serotonergic axons innervate the entire CNS at very early stages of development. We have shown previously that descending 5-HT slows the maturation of inhibitory synaptic transmission in the embryonic mouse spinal cord. We now report that 5-HT also regulates the spatiotemporal changes of the GABAergic neuronal population in the mouse spinal cord. Using a quantitative confocal study performed on acute and cultured spinal cords, we find that the GABAergic population matures according to a similar rostrocaudal temporal gradient both in utero and in organotypic culture. Moreover, we show that 5-HT delays the appearance of the spinal GABAergic system. Indeed, in the absence of 5-HT descending inputs or exogenous 5-HT, the GABAergic population matures earlier. In the presence of exogenous 5-HT, the GABA population matures later. Finally, using a pharmacological approach, we show that 5-HT exerts its action via the 5-HT1 receptor family. Together, our data suggest that, during the course of the embryonic development, 5-HT descending inputs delay the maturation of lumbar spinal motor networks relative to brachial networks.

Activated notch disrupts the initial patterning of dopaminergic spinal cord neurons

Dopaminergic spinal cord neurons differentiate in the ventral spinal cord in a nonrandom dispersed pattern. To test whether Notch signaling was involved in generating the pattern of this neuron population as with others, we overexpressed a constitutively active form of Xenopus Notch (XotchDeltaE) in developing frog embryos. Overexpression was targeted to half the spinal cord by injecting activated Notch RNA into one blastomere at the two-cell stage. Injected animals showed morphological differences on the injected side including reduced numbers of dopaminergic spinal cord neurons. This is consistent with a role for Notch signaling in establishing the fate of this population in the developing spinal cord. At a later stage of development, dopaminergic neurons continued to differentiate on both sides of the spinal cord, but the difference between experimental and control columns remained constant. This is consistent with transient activation of Notch disrupting the fate of the earliest (primary) but not later (secondary) dopaminergic neurons. The precursors to secondary neurons appear to be refractory to Notch signaling at earlier stages of development.

Development of projection-specific interneurons and projection neurons in the embryonic mouse and rat spinal cord

Interneurons and projection neurons in the lumbar spinal cord of mouse and rat embryos were labeled retrogradely with fluorescent dextran amines from a distance of one segment from the segment of origin [lumbar segment (L) 2]. Six classes with specific axonal projections (ipsilateral ascending, descending, and bifurcating, and commissural ascending, descending, and bifurcating) were identified by differential labeling in both species and followed from embryonic day (E)12 to birth in the mouse. Neurons with shorter projections (intrasegmental interneurons) were not studied. We show that the four nonbifurcating neuron classes occupy characteristic, partially overlapping domains in the transverse plane, indicating a systematic pattern of migration and settlement related to axon trajectories. The number of neurons in each of the nonbifurcating classes increased steadily during development. Bifurcating neurons represented a minor fraction of the total throughout development and had relatively scattered positions within the ipsilateral and commissural neuron domains. Combination of retrograde tracing and immunohistochemistry for the inhibitory neurotransmitter gamma-aminobutyric acid (GABA) showed that none of the spinal neurons in the six projection-specific classes was GABA positive, suggesting that all GABA-positive spinal neurons, including previously described GABA-positive commissural neurons, are unlikely to have projections exceeding one or two segments in either direction.

Distribution and development of glutamic acid decarboxylase immunoreactivity in the spinal cord of the dogfishScyliorhinus canicula(elasmobranchs)

The adult distribution and development of gamma-aminobutyric acid (GABA)-synthesizing cells and fibers in the spinal cord of the lesser spotted dogfish (Scyliorhinus canicula L.) was studied by means of immunohistochemistry using antibodies against glutamic acid decarboxylase (GAD). Complementary immunostaining with antibodies against GABA, tyrosine hydroxylase (TH), and HuC/HuD (members of the Hu/Elav family of RNA-associated proteins) and staining with a reduced silver procedure ("en bloc" Bielschowski method), Nissl, and hematoxylin were also used. In adults, GAD-immunoreactive (GAD-ir) cells were observed in the ventral horns, in the spinal nucleus of the dorsal horn, at the base of the dorsal horns, and around the central canal, where some GAD-ir cells were cerebrospinal fluid-contacting (CSF-c). In addition, a few GAD-ir cells were observed in the lateral funiculus between the ventral horn and the marginal nucleus. The adult spinal cord was richly innervated by GAD-ir fibers. Large numbers of GAD-ir fibers and boutons were observed in the dorsal and ventral horns and also interstitially in the dorsal, lateral, and ventral funiculi. In addition, a rich GAD-ir innervation was observed in the marginal nucleus of the spinal cord. In the embryonic spinal cord, GAD-ir cells develop very early: The earliest cells were observed in the very thin mantle/marginal layer of stage 22 embryos in a short length of the spinal cord. At stages 25 and 26, several types of GAD-ir cells (commissural and noncommissural) were distinguished, and two of these cells were of CSF-c type. At stages 28, 30, and 31, the GAD-ir populations exhibited a marked longitudinal columnar organization. Double-immunolabeling experiments in embryos showed the presence of two different GAD-ir CSF-c cell populations, one ventral that is simultaneously TH-ir and other more dorsal that is TH-negative. By stage 33 (prehatching), GAD-expressing cells are present in virtually all loci, as in adults, especially in the ventral horn and base of the dorsal horn. The present results for the lesser spotted dogfish suggest an important role for gamma-aminobutyric acid in sensory and motor circuits of the spinal cord.

Ontogenic changes of the GABAergic system in the embryonic mouse spinal cord

Numerous studies have demonstrated an excitatory action of GABA early in development, which is likely to play a neurotrophic role. In order to better understand the role of GABA in the mouse spinal cord, we followed the evolution of GABAergic neurons over the course of development. We investigated, in the present study, the ontogeny of GABA immunoreactive (GABA-ir) cell bodies and fibers in the embryonic mouse spinal cord at brachial and lumbar levels. GABA-ir somata were first detected at embryonic day 11.5 (E11.5) exclusively at brachial level in the marginal zone. By E13.5, the number of GABAergic neurons sharply increased throughout the extent of the ventral horn both at brachial and lumbar level. Stained perikarya first appeared in the future dorsal horn at E15.5 and progressively invaded this area while they decreased in number in the presumed ventral gray matter. At E12.5, E13.5 and E15.5, we checked the possibility that ventral GABA-ir cells could belong to the motoneuronal population. Using a GABA/Islet-1/2 double labeling, we did not detect any double-stained neurons indicating that spinal motoneurons do not synthesize GABA during the course of development. GABA-ir fibers also appeared at the E11.5 stage in the presumptive lateral white matter at brachial level. At E12.5 and E13.5, GABA-ir fibers progressively invaded the ventral marginal zone and by E15.5 reached the dorsal marginal zone. At E17.5 and postnatal day 0 (P0), the number of GABA-ir fibers declined in the white matter. Finally, by P0, GABA immunoreactivity that delineated somata was mainly restricted to the dorsal gray matter and declined in intensity and extent. The ventral gray matter exhibited very few GABA-ir cell bodies at this neonatal stage of development. The significance of the migration of somatic GABA immunoreactivity from ventral to the dorsal gray matter is discussed.

Ontogeny of gamma-aminobutyric acid-immunoreactive neurons in the rhombencephalon and spinal cord of the sea lamprey

The development of neurons expressing gamma-aminobutyric acid (GABA) in the rhombencephalon and spinal cord of the sea lamprey (Petromyzon marinus) was studied for the first time with an anti-GABA antibody. The earliest GABA-immunoreactive (GABAir) neurons appear in late embryos in the basal plate of the isthmus, caudal rhombencephalon, and rostral spinal cord. In prolarvae, the GABAir neurons of the rhombencephalon appear to be distributed in spatially restricted cellular domains that, at the end of the prolarval period, form four longitudinal GABAir bands (alar dorsal, alar ventral, dorsal basal, and ventral basal). In the spinal cord, we observed only three GABAir longitudinal bands (dorsal, intermediate, and ventral). The larval pattern of GABAir neuronal populations was established by the 30-mm stage, and the same populations were observed in premetamorphic and adult lampreys. The ontogeny of GABAergic populations in the lamprey rhombencephalon and spinal cord is, in general, similar to that previously described in mouse and Xenopus.

Establishment of a ventral cell fate in the spinal cord

The neural plate is induced during gastrulation when the organizer affects the ectoderm around it. Recent experiments show that axial mesoderm can stimulate formation of specific ventral cell types in the spinal cord, including floor plate, motor neurons, and several types of interneurons. We have eliminated or disrupted axial mesoderm by using a variety of methods to show that ventral columns of intermittent dopaminergic neurons in the frog Xenopus also appear to be induced by axial mesoderm. Inversion of the dorsal-ventral neural axis by splitting the presumptive neural plate in vivo, produced two spinal cords with ectopic dopaminergic neurons. The location and number of neurons suggest that even a brief association with axial mesoderm can specify the identity of the first or primary dopaminergic neurons and that notochord retains the ability to induce cells to become secondary dopaminergic neurons.

Metamorphic changes in GABA immunoreactivity in the brainstem of the bullfrog, Rana catesbeiana

We examined immunoreactivity for gamma-aminobutyric acid (GABA) in auditory and vestibular brainstem nuclei of the bullfrog, Rana catesbeiana, across metamorphosis, a developmental period featuring significant anatomical and functional remodeling of the nervous system. In the early larval period, GABA-immunoreactive cell somata were visible in the vestibular nucleus complex and the torus semicircularis, as well as in the spinal cord, cerebellum and optic tectum. Fiber bundles such as the medial longitudinal fasciculus and the lemnsical pathways also exhibited intense label at these early stages. In contrast, only diffuse neuropil label was visible in the dorsolateral nucleus and the superior olivary nucleus at the same stages. This diffuse immunoreactivity became progressively more reduced over larval development, and stained somata were visible in these medullary nuclei by metamorphic climax stages. In the torus semicircularis, the numbers of labeled somata in both the developing laminar and principal nuclei increased over metamorphic development, and became progressively more organized into distinct layers. The adult pattern of GABA-like immunoreactivity in the auditory brainstem was reached by metamorphic climax stages, coincident with the maturation of the opercularis system, and preceding the final development of the external tympanum and the tympanic conduction pathway. The relatively earlier maturation of vestibular, compared to auditory, areas in the medulla might reflect the behavioral importance of vestibular-mediated motor reactions during tadpole life. The distribution of GABA in auditory brainstem nuclei in both developing and adult frogs is comparable to that observed in mammals and birds.

Cerebrospinal fluid-contacting neurons in the rat spinal cord, a gamma-aminobutyric acidergic system expressing the P2X2 subunit of purinergic receptors, PSA-NCAM, and GAP-43 immunoreactivities: light and electron microscopic study

Cerebrospinal fluid-contacting neurons (CSFcNs) occur in various brain regions of lower vertebrates. In mammals, they are restricted to medullospinal areas, and little is known about their projection sites. In the present work, we investigated some morphofunctional characteristics of such neurons in the rat spinal cord by light and electron microscopic immunocytochemistry. CSFcNs expressing the P2X(2) subunit of purinergic receptors were present throughout the spinal cord, though more numerous at lower thoracolumbar and sacral levels. These neurons coexpressed GAD and the polysialylated neural cell adhesion molecule (PSA-NCAM), a marker of cellular plasticity. From low thoracic levels downward, tiny amyelinic axons (less than 200 nm in diameter) were tightly packed in bundles, which ran along the ependyma and extended ventrally, eventually concentrating against the walls of the ventral median fissure. In addition to P2X(2), GAD, gamma-aminobutyric acid (GABA), and PSA, these axons expressed GAP-43 immunoreactivity. Moreover, they were labelled along their entire lengths with antibodies against synaptotagmin and synaptophysin, but these failed to reveal intraspinal terminal fields. Taken together, our observations indicate the presence in the rat spinal cord of a highly plastic system of GABAergic CSFcNs that express the P2X(2) subunit of purinergic receptors. The function of this original system remains open to question. In these neurons, the P2X(2) receptors may confer a sensitivity to ATP either present in the CSF or released by nearby neurons of the central autonomic area.

Thyroid hormone promotes neurogenesis in the Xenopus spinal cord

Three phases of neurogenesis can be recognized during Xenopus spinal cord development. An early peak during gastrulation/neurulation is followed by a phase of low level neurogenesis throughout the remaining embryonic stages and a later peak at early larval stages. We show here that several genes known to be essential for early neurogenesis (X-NGNR-1, XNeuroD, XMyT1, X-Delta-1) are also expressed during later phases of neurogenesis in the spinal cord, suggesting that they are involved in regulating spinal neurogenesis at later stages. However, additional neuronal determination genes may be important during larval stages, because X-NGNR-1 shows only scant expression in the spinal cord during larval stages. Thyroid hormone treatment of early larvae promotes neurogenesis in the spinal cord, where thyroid hormone receptor xTRalpha is expressed from early larval stages onward and results in precocious up-regulation of XNeuroD, XMyT1, and N-Tubulin expression. Similarly, thyroid hormone treatments of Xenopus embryos, which were coinjected with xTRalpha and the retinoid X receptor xRXRalpha, repeatedly resulted in increased numbers of neurons, whereas unliganded receptors repressed neurogenesis. Our findings show that thyroid hormones are sufficient to up-regulate neurogenesis in the Xenopus spinal cord.