Electrophysiological properties of isthmic neurons in frogs revealed by in vitro and in vivo studies

The frog nucleus isthmi (parabigeminal nucleus in mammals) is a visually responsive, cholinergic and anatomically well-defined group of neurons in the midbrain. It shares reciprocal topographic projections with the ipsilateral optic tectum (superior colliculus in mammals) and strongly influences visual processing. Anatomical and biochemical information indicates the existence of distinct neural populations within the frog nucleus isthmi, which raises the question: are there electrophysiological distinctions between neurons that are putatively classified by their anatomical and biochemical properties? To address this question, we measured frog nucleus isthmi neuron cellular properties in vitro and visual response properties in vivo. No evidence for distinct electrophysiological classes of neurons was found. We thus conclude that, despite the anatomical and biochemical differences, the cells of the frog nucleus isthmi respond homogeneously to both current injections and simple visual stimuli.     

Changes in signal transmission by neurons of the cat lateral geniculate nucleus by acute deafferentation and early degeneration.

1. The responses of single principal cells of the cat lateral geniculate nucleus (LGN) were recorded extracellulary from the optic radiation (OR) axons or intracellularly from the LGN. Electrical stimuli at different frequencies were applied to the optic tract (OT) to test the transneuronal and the synaptic signal transmission in the LGN. 2. The effect of acute deafferentation (by photocoagulation of the retinal receptive field) or of synaptic degeneration induced 2-4 days prior to the recording time on the LGN neuron signal transfer was studied. Immediately after deafferentation, the synaptic signal transfer by LGN neurons exhibits signs of hyperexcitability leading to multiple neuronal discharges. This acute deafferentation hyperexicitability is probably caused mainly by the disapperance of lateral inhibition mediated by LGN interneurons. The deafferentation hyperexcitability disappeared during electrical stimulation of the OT at frequencies greater than 10/sec. 3. With progressing degeneration of the synaptic terminals during the 2nd to 4th day after interruption of the optic nerve axoplasmic flow, the synaptic signal transfer by LGN neurons gradually deteriorates and ceases at the end of the fourth day. The signs of this deterioration (larger temporal scatter, increased exhaustability and reduced upper frequency limit of the transneuronal signal transmission and gradual reduction of the EPSP amplitude in D-neurons) were quantitatively investigated. 4. The neurophysiological data obtained at different levels of synaptic terminal degeneration are well correlated with morphological changes found within the degenerating synaptic terminals.     

Glial activation and TNFR-I upregulation precedes motor dysfunction in the spinal cord of mnd mice

Mice homozygous for the spontaneous motor neuron degeneration mutation (mnd) show at the age of 8 months a marked impairment of the motor function and accumulation of lipofuscin granules in the cytoplasm of almost all neurons of the central nervous system.We previously reported a significant increase in GFAP protein levels in the lumbar spinal cord homogenates by western blot analysis and upregulation of TNF, a proinflammatory cytokine, in the motor neurons of lumbar spinal cord of mnd mice, already in a presymptomatic stage (4 months of age). In the present study, using immunohistochemical analysis, we performed a time course in mnd mice (1, 4 and 9 months of age) evaluating the expression and the distribution of astroglial and microglial cells and the expression of both TNF receptors, TNFR-I and TNFR-II. We observed a marked increase in astroglial and microglial cells and in TNFR-I immunoreactivity already at the 4th month. Since motor neuron dysfunction occurs in mnd mice in the absence of evident loss of spinal motor neurons, the present results indicate that the activation of microglial cells and astrocytes is independent from neuronal degeneration. The role of TNF and TNFR-I on motor neurons is still to be demonstrated.     

A quantitative histological study of Golgi II neurons and pale cells in different cerebellar regions of the adult and ageing mouse brain

Two types of medium to large sized neurons are present in the granular layer of the mouse cerebellum. One type has a large nucleus with a prominent nucleolus and a moderate amount of cytoplasm containing Nissl substance. This type corresponds to the classical Golgi II neuron. The second type has a much smaller nucleus (mean diameter 8.4 microns) with a darkly staining nuclear envelope which is almost invariably deeply indented by cytoplasmic intrusions. The nucleolus is smaller and less conspicuous than in Golgi II neurons. These neurons are identical to the pale cells described by Altman and Bayer (1977). The numbers of both types of neuron were estimated in the spinocerebellum, lobus simplex and nodulus in mice aged 6, 15, 22, 25, 28 and 31 months. There was no significant variation in the number of either Golgi II neurons or pale cells with age in any part of the cerebellum. The number of Golgi II neurons per mm3 was similar in all parts of the cerebellum (mean 3560 mm3). This was identical to the mean number of pale cells per mm3 in the spinocerebellum and pontocerebellum but in the nodulus pale cells were much more numerous (mean 41,170 per mm3). It is postulated that pale cells are small Golgi II neurons.     

Differentiation of embryonic chick sympathetic neurons in vivo: ultrastructure,and quantitative determinations of catecholamines and somatostatin

Summary The ultrastructural and transmitter development of lumbar sympathetic ganglia was studied in embryonic day-6 through-18 chick embryos. At embryonic day 6, ganglia are populated by two morphologically distinct types of neuronal cells and Schwann cell precursors. The neuronal populations basically comprise a granule-containing cell and a developing principal neuron. Granule-containing cells have, an irregularly shaped or oval nucleus with small clumps of chromatin attached to the inner nuclear membrane and numerous large (up to 300 nm) membrane-limited granules. Developing principal neurons display a more rounded vesicular nucleus with evenly distributed chromatin, prominent nucleoli, more developed areas of Golgi complexes, and rough endoplasmic reticulum and large dense-core vesicles up to 120 nm in diameter. There are granule-containing cells with fewer and smaller granules which still display the nucleus typical for granule-containing cells. These granule-containing cells may develop toward developing principal neurons or the resting state of granule-containing cells found in older ganglia. Both granule-containing cells and developing principal neurons proliferate and can undergo degeneration. At embryonic day 9 there are far more developing principal neurons than granule-containing cells. Most granule-containing cells have very few granules. Mitotic figures and signs of cell degeneration are still apparent. Synapse-like terminals are found on both developing principal neurons and granule-containing cells. Ganglionic development from embryonic day 11 through 18 comprises extensive maturation of developing principal neurons and a numerical decline of granule-containing cells. Some granule-containing cells with very few and small granules still persist at embryonic day 18. The mean catecholamine content per neuron increases from 0.044 femtomol at embryonic day 7 to 0.22 femtomol at embryonic day 15. Concomitantly, there is a more than 6-fold increase in tyrosine hydroxylase activity. Adrenaline has a 14% share in total catecholamines at embryonic day 15. Somatostatin levels are relatively high at embryonic day 7 (1.82 attomol per neuron) and are 10-fold reduced by embryonic day 15. Our results suggest the presence of two morphologically distinct sympathetic neuronal precursors at embryonic day 6: one with a binary choice to become a principal neuron or to die, the other one, a granule-containing cell, which alternatively may develop into a principal neuron, acquire a resting state or die.     

The distribution of cholinergic neurons in the human central nervous system

Choline acetyltransferase (ChAT), the enzyme responsible for the biosynthesis of acetylcholine, is presently the most specific marker for identifying cholinergic neurons in the central and peripheral nervous systems. The present article reviews immunohistochemical and in situ hybridization studies on the distribution of neurons expressing ChAT in the human central nervous system. Neurons with both immunoreactivity and in situ hybridization signals of ChAT are observed in the basal forebrain (diagonal band of Broca and nucleus basalis of Meynert), striatum (caudate nucleus, putamen and nucleus accumbens), cerebral cortex, mesopontine tegmental nuclei (pedunculopontine tegmental nucleus, laterodorsal tegmental nucleus and parabigeminal nucleus), cranial motor nuclei and spinal motor neurons. The cerebral cortex displays regional and laminal differences in the distribution of neurons with ChAT. The medial septal nucleus and medial habenular nucleus contain immunoreactive neurons for ChAT, which are devoid of ChAT mRNA signals. This is probably because there is a small number of cholinergic neurons with a low level of ChAT gene expression in these nuclei of human. Possible connections and speculated functions of these neurons are briefly summarized.     

Aging in the vestibular nuclear complex of the male golden hamster (Mesocricetus auratus): anatomic and morphometric study

To study the effects of senescence on the vestibular nuclear complex twenty brainstems from male golden hamsters between 3 and 27 months-old were used and the possible variations in the number of neurons, neuronal morphology and nuclear volume were studied. The neuron profiles were drawn with a camera lucida and Abercrombie's method was used to estimate the total number of neurons. The test of Kolmogorov-Smirnov with the correction of Lilliefors was used to evaluate the fit of our data to a normal distribution and a regression analysis was done to decide if the variation of our data with age was statistically significant. The results of the present study are relevant only for male animals and the effect of senescence could be different in female vestibular nuclear complex. Aging affects the volume of the superior and lateral vestibular nuclei, as well as the nuclear neuronal diameter of the medial vestibular nucleus, but no significant neuronal loss has been appreciated in vestibular nuclear complex related with age. During the aging process we have observed that the distribution of neurons within the vestibular nuclei of the golden hamster does not show important changes and most of their morphometric parameters do not vary significantly.     

Destructive Changes in the Neuronal Structure of the FVB/N Mouse Retina

We applied a series of selective antibodies for labeling the various cell types in the mammalian retina. These were used to identify the progressive loss of neurons in the FVB/N mouse, a model of early onset retinal degeneration produced by a mutation in the pde6b gene. The immunocytochemical studies, together with electroretinogram (ERG) recordings, enabled us to examine the time course of the degenerative changes that extended from the photoreceptors to the ganglion cells at the proximal end of the retina. Our study indicates that photoreceptors in FVB/N undergo a rapid degeneration within three postnatal weeks, and that there is a concomitant loss of retinal neurons in the inner nuclear layer. Although the loss of rods was detected at an earlier age during which time M- and S-opsin molecules were translocated to the cone nuclei; by 6 months all cones had also degenerated. Neuronal remodeling was also seen in the second-order neurons with horizontal cells sprouting processes proximally and dendritic retraction in rod-driven bipolar cells. Interestingly, the morphology of cone-driven bipolar cells were affected less by the disease process. The cellular structure of inner retinal neurons, i.e., ChAT amacrine cells, ganglion cells, and melanopsin-positive ganglion cells did not exhibit any gross changes of cell densities and appeared to be relatively unaffected by the massive photoreceptor degeneration in the distal retina. However, Muller cell processes began to express GFAP at their endfeet at p14, and it climbed progressively to the cell’s distal ends by 6 months. Our study indicates that FVB/N mouse provides a useful model with which to assess possible intervention strategies to arrest photoreceptor death in related diseases.     

Effects of neonatal capsaicin administration on the numbers and volumes of neurons in left and right T10 dorsal root ganglia in the rat

The long-term effects of neonatal capsaicin were studied in left and right dorsal root ganglia (T10) from control and capsaicin-treated groups of Wistar rats. At 12 hours post partum, 5 females per group were injected subcutaneously with capsaicin or vehicle solution and killed at 6 months of age. Tissues were perfusion-fixed, embedded in resin and serially sectioned. A Nissl stain was used to distinguish between A and B neurons and systematic random sampling schemes were employed to obtain stereological estimates of numbers of neurons and mean volumes of their perikarya. Numbers were calculated from ganglion volumes (estimated via the Cavalieri principle) and neuron packing densities (estimated using physical disectors). Mean perikaryal volumes were calculated from packing densities and volume densities (estimated by point counting). Data were analysed to isolate main and interaction effects of neuron subtype, laterality and treatment. There was no evidence of lateral asymmetry or interaction effects. Control ganglia contained 3320 (coefficient of variation, CV, 8%) neurons. Most (73%) were B cells with a mean volume of 13,100 µm³ (CV 17%) of which the nucleus accounted for 1,800 µm³ (CV 18%). About 22% were A cells with a mean volume of 79,800 µm³ (CV 24%) and a nucleus of 6,100 µm³ (CV 26%). After capsaicin, over half the original population of cells was destroyed and B cell loss was significantly greater than that of A cells (about 80% of all cells lost were B cells). The mean size of A cells was greater after capsaicin due to selective loss of smaller cells and a greater volume of cytoplasm. B cell perikaryal volume was not affected but nuclear volume declined. The findings show that capsaicin destruction of peripheral sensory neurons is bilaterally symmetrical. In general, smaller neurons are selectively destroyed but this operates differently in A and B cells. It is size-dependent in A cells but size-independent (possibly random) in B cells.     

Increased Number of Dorsal Root Ganglion Neurons in Vitamin-E-Deficient Rats

Quantitative and morphometric observations were carried out on neurons of L₃-L₆ dorsal root ganglia (DRGs) in control and vitamin-E-deficient rats at different ages. Controls were fed a standard diet and sacrificed at 1 or at 5 months of age; deficient rats were fed a diet without vitamin E from 1 to 5 months of age and then sacrificed. No significant difference in total number of neurons was found, but an increase in neuron sizes, a decrease in nucleus-cytoplasm ratio, and a more circular neuron shape were found in controls with increasing age (from 1 to 5 months). In L₃-L₆ DRGs of vitamin-E-deficient rats (5 months of age), a higher number of neurons was found than in those of either young or adult controls. Moreover, some morphometric characteristics of neurons in the deficient rats were similar to those of neurons in 1-month-old controls. The findings suggest that vitamin E deficiency can trigger events resulting in appearance of new neurons, possibly anticipating phenomena that normally occur in aging.     

The role of target size in neuronal survival.

A loss of about half of the trochlear motor neurons occurs during the course of normal development in duck and quail embryos. The role of the size of the target muscle in controlling the number of surviving motor neurons was examined by making motor neurons innervate targets either larger or smaller in size than their normal target. In one experiment the smaller trochlear motor neuron pool of the quail embryo was forced to innervate the larger superior oblique muscle of the duck embryo. This was accomplished by grafting the midbrain of a quail embryo in the place of the midbrain of a duck embryo. Results indicated that no additional quail trochlear motor neurons were rescued in spite of a considerable increase in target size. In another experiment the larger trochlear motor neuron pool of the duck embryo was made to innervate the smaller superior oblique muscle of the quail embryo. This resulted in loss of some additional neurons; however, the number of surviving motor neurons was not proportionate to the reduction in target size. These experiments failed to provide support for the hypothesis that the size of the target muscle controls the number of surviving motor neurons. Although contact with target is necessary for survival of neurons, factors other than the number or size of target cells are involved in the control of motor neuron numbers during development.     

Profile of adult rat sensory neuron loss, apoptosis and replacement after sciatic nerve crush

Following permanent transection of the adult rat sciatic nerve, sensory neuron apoptosis in the contributing L4 and L5 dorsal root ganglia can be observed for at least 6 months afterwards. To establish the profile of any sensory neuron apoptosis and loss over time when axonal regeneration is allowed, serial sections of L4 and L5 ganglia were examined and the neurons counted using a stereological technique 1, 2 and 3 months after crushing the right sciatic nerve at mid-thigh level. Our results show that an identical degree of sensory neuron loss and apoptosis occurs 1 month after crush as at 1 month after permanent transection. However, at 3 months no neurons undergoing apoptosis could be observed and no significant loss could be detected in the ipsilateral ganglia when compared to unoperated controls. One explanation was a neuronal replacement mechanism, which was investigated by administering bromodeoxyuridine to rats for 1 month after sciatic nerve transection or crush, prior to detection using immunohistochemistry on sections of their ganglia after 2 months. The presence of bromodeoxyuridine in the nuclei of occasional cells that would be counted as neurons on the basis of size and morphology indicates that a process of apparent neurogenesis may underlie the profile of sensory neuron loss after axotomy.     

Trophic effects of androgen: development and hormonal regulation of neuron number in a sexually dimorphic vocal motor nucleus.

In Xenopus laevis, the laryngeal motor nucleus (n. of cranial nerves IX-X) is part of a sexually differentiated, androgen sensitive neuromuscular system devoted to vocalization. Adult males have more n. IX-X neurons than females; however, during development of n. IX-X, the rate of neurogenesis does not appear to differ between the sexes. In this study, we explored the role of naturally occurring cell death in the development of this nucleus and asked whether cell death might be involved in establishing the sex difference in neuron number. Counts of n. IX-X neurons reveal that at tadpole stage 56, males and females have similar numbers of n. IX-X neurons, but by stage 64 male neuron numbers are greater. This sex difference arises owing to a greater net loss of neurons in females-males lose approximately 25% of their n. IX-X neurons between stages 56 and 64, while females lose approximately 47%. Sexual differentiation of n. IX-X neuron number coincides with a period of developmental cell death, as evidenced by terminal transferase-mediated dUTP nick-end labeling and the presence of pyknotic nuclei in n. IX-X. A role for gonadal hormones in controlling cell number was examined by treating tadpoles with exogenous androgen and determining the number of n. IX-X neurons at stage 64. Dihydrotestosterone (DHT) treatment from the beginning of the cell death period (stage 54) until stage 64 had no effect on the number of n. IX-X neurons in males but did significantly increase n. IX-X neuron number in females. This increase was sufficient to abolish the sex difference normally observed at stage 64. Although DHT induced increases in female neuron number, it did not induce increases in cell proliferation or addition of newly born neurons to n. IX-X. DHT may therefore have increased neuron number by protecting cells from death. We conclude that androgens can influence the survival of n. IX-X neurons during a period of naturally occurring cell death, and that this action of androgen is critical to the development of sex differences in n. IX-X neuron number.     

Sex-dependent loss of projection neurons involved in avian song learning.

In zebra finches, only males sing, and the neural regions controlling song exhibit prominent, hormone-induced sex differences in neuron number. In order to understand how sexual differentiation regulates neuron number within one song nucleus, the lateral magnocellular nucleus of the anterior neostriatum (IMAN), we studied the development of sex differences among IMAN neurons that project to the robust nucleus of the archistriatum (RA). The IMAN is implicated in song learning, and previous ontogenetic studies have indicated that males lose over 50% of their IMAN neurons during the juvenile song learning period. Based on developmental changes in both the extent of androgen accumulation within the IMAN and its appearance in Nissl-stained tissue, it had been hypothesized that IMAN neuron loss was even greater in young females, resulting in sex differences in neuron number. However, this hypothesis has not been tested directly because the Nissl-stained boundaries of the IMAN sometimes are ambiguous in young animals, and are not evident at all in adult females. To circumvent these problems, we employed the retrograde tracer fast blue to study the development of IMAN neurons defined on the basis of their projections to the RA. We find that the number of these IMAN-RA projection neurons is much greater in adult males than in females, and that this sex difference develops during the juvenile period of sexual differentiation and song learning because a significant number of these neurons are lost in females but not in males. With respect to sexual differentiation, we conclude that masculinization (which is stimulated by the hormone estradiol) promotes the retention of IMAN-RA projection neurons. In addition, our results indicate that any loss of IMAN neurons that may occur in young males does not include cells projecting to the RA.     

Sex-dependent loss of projection neurons involved in avian song learning

In zebra finches, only males sing, and the neural regions controlling song exhibit prominent, hormone-induced sex diffences in neuron number. In order to understand how sexual differentiation regulates neuron number within one song nucleus, the lateral magnocellular nucleus of the anterior neostriatum (IMAN), we studied the development of sex differences among IMAN neurons that project to the robust nucleus of the archistriatum (RA). The IMAN is implicated in song learning, and previous ontogenetic studies have indicated that males lose over 50% of their IMAN neurons during the juvenile song learning period. Based on developmental changes in both the extent of androgen accumulation within the IMAN and its appearance in Nissl-stained tissue, it had been hypothesized that IMAN neuron loss was even greater in young females, resulting in sex differences in neuron number. However, this hypothesis has not been tested directly because the Nissl-stained boundaries of the IMAN sometimes are ambiguous in young animals, and are not evident at all in adult females. To circumvent these problems, we employed the retrograde tracer fast blue to study the development of IMAN neurons defined on the basis of their projections to the RA. We find that the number of these IMAN-RA projection neurons is much greater in adult males than in females, and that this sex difference develops during the juvenile period of sexual differentiation and song learning because a significant number of these neurons are lost in females but not in males. With respect to sexual differentiation, we conclude that masculinization (which is stimulated by the hormone estradiol) promotes the retention of IMAN-RA projection neurons. In addition, our results indicate that any loss of IMAN neurons that may occur in young males does not include cells projecting to the RA. © 1992 John Wiley & Sons, Inc.     

Presynaptic dendrite cells and two other classes of neurons in the superficial layers of the superior colliculus of the chimpanzee

In the superior colliculus of chimpanzee, three classes of neurons can be identified by ultrastructural criteria. They are 1) marginal cells located in the stratum zonale, 2) collicular relay cells in the stratum griseum superficiale and 3) presynaptic dendrite (PSD) cells, i.e., neurons with presynaptic specializations in soma and/or dendrites. PSD cells are the smallest neurons in the stratum griseum superficiale; they have a relatively large, deeply infolded nucleus and a small rim of cytoplasm rich in free ribosomes. PSD cells are sufficiently different from the two other classes of neurons to be reliably identified at the ultrastructural level. They closely resemble presynaptic neurons as described in the lateral geniculate nucleus of other mammalian species. Presynaptic dendrites in continuity with PSD cells are rich in organelles, especially ribosomal cluster, and establish en passage contact with other dendrites. Another type of presynaptic dendrite, poor in organelles, except for bundles of microtubules, could not be traced back to its parent neuron. Homo- or heterogeneity of PSD cells is discussed. No amxon was traced from a PSD cell.     

Unit responses in area 5b of the suprasylvian gyrus to stimulation of the ventral posterolateral thalamic nucleus

Unanesthetized cats were immobilized with D-tubocurarine. Single unit responses in area 5b of the suprasylvian gyrus to stimulation of the ventral posterolateral thalamic nucleus were recorded extracellularly. Of the total number of neurons tested, 32% were excited and 3% inhibited. In 65% of neurons the responses were mixed, most of them being predominantly excitatory. Repetitive stimulation of the ventral posterolateral nucleus (6–9/sec) frequently intensified the excitatory component of the responses. Sometimes inhibition, present in the response to a single stimulus, was replaced by increased excitation. However, the same response as to a single stimulus frequently appeared in response to each consecutive stimulus of a series. Stimulation of the ventral posterolateral nucleus had a mainly excitatory effect on neurons in area 5b. Stimulation of the dorsal lateral nucleus, on the other hand, inhibited their activity. This antagonism could also be observed on the same neuron. It was concluded from the short latent periods of the orthodromic responses (3–6 msec) and from the antidromic responses of the cortical neurons to stimulation of the ventral posterolateral nucleus that this nucleus has direct two-way connections with the cortex of area 5b.     

Neuron loss and addition in developing zebra finch song nuclei are independent of auditory experience during song learning

In zebra finches early auditory experience is critical for normal song development. Young males first listen to and memorize a suitable song model and then use auditory feedback from their own vocalizations to mimic that model. During these two phases of vocal learning, song-related brain regions exhibit large, hormone-induced changes in volume and neuron number. Overlap between these neural changes and auditory-based vocal learning suggests that processing and acquiring auditory input may influence cellular processes that determine neuron number in the song system. We addressed this hypothesis by measuring neuron density, nuclear volume, and neuron number within the song system of normal male zebra finches and males deafened prior to song learning (10 days of age). Measures were obtained at 25, 50, 65, and 120 days of age, and included four song nuclei: the hyperstriatum ventralis pars caudalis or higher vocal center (HVc), Area X, the robust nucleus of the archistriatum (RA), and the lateral magnocellular nucleus of the anterior neostriatum (IMAN). In both HVc and Area X, nuclear volume and neuron number increased markedly with age in both normal and deafened birds. The volume of RA also increased with age and was not affected by early deafening. In IMAN, deafening also did not affect the overall age-related loss of neurons, although at 25 days neuron number was slightly less in deafened than in normal birds. We conclude that while the addition and loss of neurons in the developing song system may provide plasticity essential for song learning, these changes do not reflect learning.(ABSTRACT TRUNCATED AT 250 WORDS)     

The role of target size in neuronal survival

A loss of about half of the trochlear motor neurons occurs during the course of normal development in duck and quail embryos. The role of the size of the target muscle in controlling the number of surviving motor neurons was examined by making motor neurons innervate targets either larger or smaller in size than their normal target. In one experiment the smaller trochlear motor neuron pool of the quail embryo was forced to innervate the larger superior oblique muscle of the duck embryo. This was accomplished by grafting the midbrain of a quail embryo in the place of the midbrain of a duck embyro. Results indicated that no additional quail trochlear motor neurons were rescued in spite of a considerable increase in target size. In another experiment the larger trochlear motor neuron pool of the duck embryo was made to innervate the smaller superior oblique muscle of the quail embryo. This resulted in loss of some additional neurons; however, the number of surviving motor neurons was not proportionate to the reduction in target size. These experiments failed to provide support for the hypothesis that the size of the target muscle controls the number of surviving motor neurons. Although contact with target is necessary for survival of neurons, factors other than the number or size of target cells are involved in the control of motor neuron numbers during development. © 1992 John Wiley & Sons, Inc.