Showing papers in "The Journal of Comparative Neurology in 1979"

The connections of the septal region in the rat.

Abstract: The efferent, afferent and intrinsic connections of the septal region have been analyzed in the rat with the autoradiographic method. The lateral septal nucleus, which can be divided into dorsal, intermediate and ventral parts, receives its major input from the hippocampal formation and projects to the medial septal-diagonal band complex. The ventral part of the nucleus also sends fibers through the medial forebrain bundle to the medial preoptic and anterior hypothalamic areas, to the lateral hypothalamic area and the dorsomedial nucleus, to the mammillary body (including the supramammillary region), and to the ventral tegmental area. The medial septal nucleus/diagonal band complex projects back to the hippocampal formation by way of the dorsal fornix, fimbria, and possibly the cingulum. Both nuclei also project through the medial forebrain bundle to the medial and lateral preoptic areas, to the lateral hypothalamic area, and to the mammillary complex. The medial septal nucleus also sends fibers to the midbrain (the ventral tegmental area and raphe nuclei) and to the parataenial nucleus of the thalamus, while the nucleus of the diagonal band has an additional projection to the anterior limbic area. Ascending inputs to the medial septal nucleus/diagonal band complex arise in several hypothalamic nuclei and in the brainstem aminergic cell groups. The posterior septal nuclei (the septofimbrail and triangular nuclei) receive their major input from the hippocampal formation, and project in a topographically ordered manner upon the habenular nuclei and the interpeduncular nuclear complex. The bed nucleus of the stria terminalis receives its major input from the amygdala (Krettek and Price, '78); but other afferents arise from the ventral subiculum, the ventromedial nucleus, and the brainstem aminergic cell groups. The principal output of the bed nucleus is through the medial forebrain bundle to the substantia innominata, the nucleus accumbens, most parts of the hypothalamus and the preoptic area, the central tegmental fields of the midbrain, the ventral tegmental area, the dorsal and median nuclei of the raphe, and the locus coeruleus. The bed nucleus also projects to the anterior nuclei of the thalamus, the parataenial and paraventricular nuclei, and the medial habenular nucleus, and through the stria terminalis to the medial and central nuclei of the amygdala, and to the amygdalo-hippocampal transition area.

Efferent connections of the habenular nuclei in the rat

Abstract: The efferent connections of the medial (MHb) and lateral (LHb) habenular nuclei in the rat were demonstrated autoradiographically following small injections of tritiated amino acids localized within various parts of the habenular complex. Comparison of individual cases led to the following conclusions.

An autoradiographic examination of corticocortical and subcortical projections of the mediodorsal‐projection (prefrontal) cortex in the rat

Abstract: The efferent connections of the cortex projected upon by the mediodorsal thalamic nucleus (MD-projection cortex) have been re-examined autoradiographically in the rat following the microelectrophoretic injection of 3H-proline-leucine into different parts of the medial and sulcal MD-projection cortex. Contrary to previous negative findings, the present experiments revealed a system of extensive corticocortical projections and indicated that different areas of the MD-projection cortex have distinctive patterns in both their corticocortical and subcortical projections. Thus, cell of Brodmann's area 32 send axons to the retrosplenial cortex, area 29d, the peri- and entorhinal cortices, and the presubiculum. Both supragenual and more posterior regions of area 24 project to the retrosplenial cortex and area 29d, but only the posterior portion projects additionally to the entorhinal area and presubiculum. The cortical targets of axons from the sulcal MD-projection cortex are mainly the anterior part of the piriform cortex and, for the posterolateral part of the sulcal cortex (insular area), the retrosplenial area, lateral entorhinal area, and presubiculum. While the medial and sulcal divisions of the MD-projection cortex project upon one another, the medial-to-sulcal projection is in general denser than its reciprocal. Earlier findings of projections to subcortical structures affiliated with the limbic system such as midline thalamic nuclei, hypothalamus, and paramedian mesencephalic tegmentum are confirmed, and appear to originate primarily in area 32 and the insular part of the sulcal cortex. The corticothalamic projection to MD in general terms reciprocates the organization seen in the thalamocortical projection from the various subnuclei within MD. Previously undocumented projections are demonstrated mainly from area 32 of the medial cortex and the insular part of the sulcal cortex to the lateral and basal amygdaloid nuclei, the medial part of the lateral septal nucleus, the nucleus accumbens, and the deep layers of the olfactory tubercle; the medial part of the lateral habenular nucleus receives a projection from areas 32 and 24. Projections to the pretectal area and superior colliculus appear to originate from all parts of the medial MD-projection cortex, but are markedly denser when the posterior part of area 24 is injected. The distribution of this corticotectal projection shows a highly characteristic configuration in which areas of high grain concentration surround areas of lower grain concentration.

Spinal termination of functionally identified primary afferent neurons with slowly conducting myelinated fibers.

Abstract: Single primary afferent myelinated fibers from cutaneous receptors of cat and monkey were functionally identified by recording from the spinal cord with micropipettes filled with horseradish peroxidase (HRP). Relatively slowly conducting fibers (less than 40 m/sec) from high threshold mechanoreceptors (mechanical nociceptors) and two types of low threshold mechanoreceptor (D-hair and field) were selected for staining. Iontophoresis of the HRP and subsequent histochemical reaction stained the axons recorded from and their collaterals, including terminations, for several millimeters. The termination patterns in the two species proved essentially identical. Ipsilaterally, the mechanical nociceptor fibers terminated principally in the dorsal horn's marginal zone and in the ventral parts of the nucleus proprius (lamina V in the cat). Some of these nociceptors also had terminals in the midline just dorsal to the central canal, contralaterally in the marginal zone, and at the base of the opposite nucleus proprius. In contrast, the D-hair primary afferent axons terminated in the dorsal part of the nucleus proprius overlapping into the innermost portion of the substantia gelatinosa. The field receptor fibers terminated predominantly in the middle part of the nucleus proprius. These results suggest that there is a highly specialized central projection of primary afferent endings which is related to sensory function and not to fiber diameter. The marginal zone and most dorsal parts of the substantia gelatinosa receive direct projections from cutaneous nociceptors but do not have direct input from cutaneous receptors transmitting activity initiated by innocuous stimulation.

Rods and cones in the mouse retina. I. Structural analysis using light and electron microscopy

Abstract: Rods and cones of the C57BL/6J mouse retina have been examined by light and electron microscopy to distinguish the structural features of the two photoreceptor types. By light microscopy, cone nuclei are conspicuously different from rod nuclei in 1-2 micrometer plastic sections. Cone nuclei have an irregularly shaped clump of heterochromatin that appears in single sections to be one to three clumps, whereas rod nuclei are more densely stained and have one large, central clump of heterochromatin. Cone nuclei make up approximately 3% of the photoreceptor nuclei in both the central and peripheral retina at all ages examined up to 267 days. Cone nuclei are confined to the outer half of the outer nuclear layer, and more than 50% of the cone nuclei lie adjacent to the outer limiting membrane. By electron microscopy, cones in the mouse retina meet virtually every morphological criterion of mammalian cones. The outer segments are conically shaped. Many, if not all of the outer segment discs are continuous with the outer plasma membrane, whereas almost all of the rod discs are not. Cone outer segments are only about half the length of the rod outer segments, and they are contacted by long, villous pigment epithelial cell processes. The cone inner segment diameter is greater than the outer segment diameter, and the accumulation of mitochondria present at the apical end of the inner segment forms a more conspicuous ellipsoid than in rods. The internal fiber or axon of the cone is larger in diameter than that of the rod, and it terminates in a large synaptic pedicle with multiple ribbon synapses, whereas the rod terminal is a smaller spherule with only a single ribbon synaptic complex.

Ascending projections to the inferior colliculus

Abstract: Cells that send ascending projections to the inferior colliculus were identified following injections of horseradish peroxidase into the colliculus. Labelled cells were found in all subcollicular auditory nuclei. Virtually all cells of the ipsilateral ventral nucleus of the lateral lemniscus and medial superior olive appear to project to the colliculus. Very few cells in these nuclei were labelled on the contralateral side. Heavy labelling on the contralateral side was found in the dorsal nucleus of the lateral lemniscus and cochlear nucleus, with less labelling being found ipsilaterally in these nuclei. The lateral superior olive was approximately evenly labelled on the two sides, with about half the cells from each side projecting to each colliculus. Cells in all periolivary cell groups were labelled, with most being found adjacent to the medial superior olive. An effort was made to identify individual cell types that were labelled and some 24 cell types were identified. In the cochlear nucleus there were marked differences between cell types in the extent of their labelling. Topographic projections matched previously described tonotopic organization of the colliculus and all major subcollicular nuclei except the ventral nucleus of the lateral lemniscus. A description of the cells in the nucleus is provided.

The origin of descending pathways in the dorsolateral funiculus of the spinal cord of the cat and rat: further studies on the anatomy of pain modulation.

Abstract: There is considerable evidence that the dorsolateral funiculus (DLF) of the spinal cord contains descending pathways critical for both opiate and brainstem stimulation-produced analgesia. To obtain a comprehensive map of brainstem neurons projecting to the spinal cord via the DLF, large injections of horseradish peroxidase (HRP) were made into the lumbosacral spinal cord of cat and rat. These injections were made caudal to midthoracic lesions which spared only a single DLF or ventral quadrant (VQ); thus only those neurons whose axons descended in the spared funiculus would be labelled. Cells with descending axons in the VQ were concentrated in the medullary nucleus raphe pallidus and obscurus, nucleus retroambiguus and in various subregions of the reticular formation including the nucleus reticularis ventralis, gigantocellularis, magnocellularis, pontis caudalis and pontis oralis. Significant numbers of neurons were also found in medial and lateral vestibular nuclei and in several presumed catecholamine-containing neurons of the dorsolateral pons. In the rat, but not in the cat, considerable numbers of cells are present in the mesencephalic reticular formation, just lateral to the periaqueductal gray. In both species, some cells were found in the paraventricular nucleus of the hypothalamus. Brainstem cells projecting in the DLF were concentrated in the nucleus raphe magnus and in the adjacent nucleus reticularis magnocellularis, ipsilateral to the spared funiculus. Significant numbers of cells were found in the dorsolateral pons, differing somewhat in their distribution from those projecting in the VQ. DLF-projecting cells were also present in the ipsilateral Edinger-Westphal nucleus and periaqueductal grey contralateral red nucleus of the midbrain and in the ipsilateral hypothalamus. Smaller projections from other sites are described. These results are discussed in terms of the differential contribution of several brainstem neuronal groups, including the serotonergic nucleus, raphe magnus, the ventromedial reticular formation of the medulla, and various catecholamine-containing neurons of the dorsolateral pontine tegmentum to the analgesia produced by opiates and electrical brain stimulation.

Afferent projections to the ventral tegmental area of Tsai and interfascicular nucleus: a horseradish peroxidase study in the rat.

Abstract: Using the retrograde transport of horseradish peroxidase (HRP), a study has been made of projections to the ventral tegmental area of Tsai (VTA) and related dopaminergic cell groups (A 10). In order to minimise the possibility of damage to fibres of passage, a technique was evolved for the microiontophoresis of HRP such that minimal current strengths and durations were applied. In addition to a sham injection, control injections were also made to the medial leminiscus, red nucleus, deep tegmental decussations, mesencephalic reticular formation and brachium conjunctivum. Following HRP injections confined to the areas of the VTA containing the dopamine cell groups, labelled neurons appeared in prefrontal cortex, dorsal bank of rhinal sulcus, nucleus accumbens, bed nucleus of stria terminalis, amygdala, diagonal band of Broca, substantia innominata, magnocellular preoptic area, medial and lateral preoptic areas, anterior, lateral and postero-dorsal hypothalamus, lateral habenular, nucleus parafascicular nucleus of thalamus, superior colliculus, nucleus raphe dorsalis, nucleus raphe magnus and pontis, dorsal and ventral parabrachial nuclei, locus coeruleus and deep cerebellar nuclei. Regions containing catecholamine goups A 1, A 5, A 6, A 7, A 9, A 13 and the serotonin group B 7 corresponded to the topography of labeled cell groups. Injections of HRP to the interfascicular nucleus resulted in labeling predominantly confined to the medial habenular and median raphe nuclei. The results are discussed in relation to the known connections of these regions. Other regions of the brain labelled by VTA injections are assessed in relation to control injections and the limitations of the HRP technique. A review of the organisation of some of these afferents in relation to the known cortical-subcortical-mesencephalic projection systems, suggests that the VTA is in a position to receive information from a massively convergent system derived ultimately from the entire archi-, paleo-, and neo-cerebral cortices. In addition A 10 dopaminergic neurons are known to project to restricted regions of both pre-frontal and entorhinal cortices, which themselves also receive massively convergent association cortico-cortical connections. It would appear reasonable to propose that these neurons perform a correspondingly important integrative function.

Retinotopic organization of areas 18 and 19 in the cat

Abstract: The location and retinotopic organization of areas 18 and 19 in cat cortex were determined using electrophysiological mapping techniques. These two areas each contain a single representation of the visual hemifield and each has a distinctive cytoarchitecture. The visual hemifield representations in these two areas are nearly mirror images of each other. Compared to area 17, areas 18 and 19 have less cortical surface area, have a lower cortical magnification factor, contain less of the visual field and contain second order instead of first order transformations of the visual hemifield. An unusual asymmetry was found between the representations of the upper and lower visual quadrants not seen before in maps of other areas of cat or other species. A considerable amount of variability in the retinotopic organization of these two areas was found among cats.

Sources of subcortical projections to the superior colliculus in the cat.

Abstract: A comprehensive search for subcortical projections to the cat superior colliculus was conducted using the retrograde horseradish peroxidase (HRP) method. Over 40 different subcortical structures project to the superior colliculus. The more notable among these are grouped under the following categories. Visual structures: ventral lateral geniculate nucleus, parabigeminal nucleus, pretectal area (nucleus of the optic tract, posterior pretectal nucleus, nuclei of the posterior commissure). Auditory structures: inferior colliculus (external and pericentral nuclei), dorsomedial periolivary nucleus, nuclei of the trapezoid body, ventral nucleus of the lateral lemniscus. Somatosensory structures: sensory trigeminal complex (all divisions, but mainly the γ division of nucleus oralis), dorsal column nuclei (mostly cuneate nucleus), and the lateral cervical nucleus. Catecholamine nuclei: locus coeruleus, raphe dorsalis, and the parabrachial nuclei. Cerebellum: medial, interposed, and lateral nuclei, and the perihypoglossal nuclei. Reticular areas: zona incerta, substantia nigra, midbrain tegmentum, nucleus paragigantocellularis lateralis, and the hypothalamus. Evidence is presented that only the parabigeminal nucleus, the nucleus of the optic tract, and the posterior pretectal nucleus project to the superficial collicular layers (striatum griseum superficiale and stratum opticum), while all other afferents terminate in the deeper layers of the colliculus. Also presented is information concerning the rostrocaudal distribution of some of these afferent connections. These findings stress the multiplicity and diversity of inputs to the deeper collicular layers, and more specifically, identify multiple sources of the physiologically well-known representations of the somatic and auditory modalities in the colliculus.

An autoradiographic study of the efferent connections of the lateral hypothalamic area in the rat.

Abstract: The efferent connections of the lateral hypothalamic area (LHA) have been analyzed in a series of 30 rat brains with injections of 3H-amino acids into different parts of the area and the surrounding regions. Our findings indicate that all parts of the LHA contribute ascending and descending fibers to the medial forebrain bundle, and also project medially to certain of the adjoining hypothalamic nuclei. All levels of the LHA appear to send some fibers to a continuous group of structures that extends from the medial septal-diagonal band complex rostrally, through the lateral preoptic and lateral hypothalamic areas to the mammillary complex and the ventral tegmental area caudally. In addition, it is evident that cells at different levels within the LHA may have differential projections. Thus, the anterior and lateral parts of the LHA also appear to project substantially to the anterior hypothalamic area, the ventromedial and dorsomedial hypothalamic nuclei, the parataenial and paraventricular nuclei of the thalamus, and the medial part of the lateral habenular nucleus. Similarly, cells in the tuberal and posterior parts of the LHA project to the central gray, the longest projections from the posterior region reaching as far caudally as the central tegmental field, the parabrachial nucleus, the locus coeruleus, and the superior central and dorsal nuclei of the raphe. Viewed as a whole, the LHA is therefore well-suited to integrate inputs from the limbic system and brainstem and to relay them on the one hand to the medial zone of the hypothalamus and on the other to virtually every structure closely associated with the medial forebrain bundle and to the nuclei of origin of the major ascending monoaminergic systems.

Axonal guidance during embryogenesis and regeneration in the spinal cord of the newt: The blueprint hypothesis of neuronal pathway patterning

Abstract: In our previous studies on studies on spinal cord regeneration in the adult lizard and the newt, we observed that the radial processes of the regenerating ependyma form between them channels which are subsequently invaded by growing neurites. In the present study we compare embryogenesis of the newt spinal cord with regeneration in the adult. Except for minor differences, we observed that the germinal neuroepithelium of the embryo and larva patterns the longitudinal neural tracts in a similar manner. With these facts in mind we propose the blueprint hypothesis which asserts that inherent in the primitive germinal neuroepithelium and its derivative primitive glia is the pattern of the primary neuronal pathways which is expressed in neurogenesis as formed channels or spaces between the processes of the epithelial cells, the surfaces of which contain trace pathways which the growing neurites follow toward their destination. The trace pathways are envisoned as mechanical-chemical itineraries which the neurities follow according to their individual affinities. The hypothesis is compared to extant theories and the limitations in central nervous regeneration of vertebrates is compared.

The origin, distribution and synaptic relationships of substance P axons in rat spinal cord.

Abstract: Substance P positive (SP+) immunoperoxidase reaction product has been localized in light and electron microscopic preparations of rat lumbar spinal cord using an immunocytochemical method. SP+ reaction product was found to be highly concentrated in dorsal horn laminae I, II, portions of III, Lissauer's tract, s small nucleus in the dorsal part of the lateral funiculus, and in a longitudinal bundle of fibers just ventral to the central canal. Moderate accumulations of SP+ reaction product were observed in portions of laminae III-V, lamina X and in a narrow zone bordering fasciculus gracilis which expanded in its ventral aspect to include nucleus cornucommissuralis dorsalis and nucleus dorsalis. Remaining portions of the spinal gray matter exhibited extremely sparse staining. Light microscopic observations indicated that SP+ product was concentrated in fine axons and collaterals, which displayed densely-stained varicosities and terminal puncta. These varicosities and puncta were closely associated with neurons and blood vessels. Dorsal rhizotomy produced a marked reduction in the number of SP+ fibers in Lissauer's tract, dorsal horn laminae I-III and the nucleus of the dorsolateral funiculus. However, dorsal root fibers were not the only source of SP+ structures in these regions since some SP+ fibers remained following dorsal root resections. In addition, this residual population of SP+ fibers did not appear to be altered by a combination of dorsal rhizotomy and ipsilateral transverse hemisection of the spinal cord. In conjunction with the results of other investigators, this finding suggests that the SP+ fibers which remain after dorsal rhizotomy are derived from local circuit interneurons of the spinal cord. Electron microscopic observations revealed that certain axons and axon terminals contained concentrations of SP+ product which were associated with large, granular vesicles, while lesser amounts were in the terminal cytoplasm and associated with the exterior surfaces of small, agranular synaptic vesicles. SP+ terminals formed axodendritic, axosomatic and axoaxonal synapses, but the deposition of SP+ product often was not concentrated precisely at the “active sites” of these synaptic junctions. Some SP+ axons also made rudimentary contacts with astrocytic processes including those surrounding blood vessels. Since many of the SP+ terminals are similar in several respects to neuroendocrine terminals, it is possible that substance P may be released at several different types of non-synaptic sites as well as at conventional synaptic junctions. Thus our findings suggest that substance P could participate in a variety of neural functions ranging from those which are limited to synaptic junctions to those which are more generally distributed via an involvement of axonal plexuses as well as the glial, vascular and ventricular systems.

Reexamination of the dorsal root projection to the spinal dorsal horn including observations on the differential termination of coarse and fine fibers.

Abstract: Primary afferent fibers in the lumbar, sacral, and caudal spinal segments of several mammals (rat, cat, monkey) were stained by applying horseradish peroxidase to the proximal part of cut dorsal rootlets and reacting the tissue histochemically after several hours of survival. The stained fibers' pattern of termination in the dorsal horn was similar in all three species, with many bouton-like enlargements in the ipsilateral marginal zone, substantia gelatinosa, and nucleus proprius, as well as a few projections at each level to the dorsal commissure and contralaterally to the ventral border of the nucleus proprius. Partial lesions of dorsal rootlets in monkey revealed that the thin fibers comprising the lateral division end principally in the marginal zone and substantial gelatinosa, while the thick fibers of the medial division terminate in the nucleus proprius and deeper regions, contributing little to the substantia gelatinosa and marginal zone. On the basis of the termination patterns observed for whole and partly sectioned rootlets, the superficial dorsal horn can be divided into at least four regions. (1) The marginal zone (lamina I of cat) appears to receive terminations from intermediate (smaller myelinated) fibers; (2) the outer substantia gelatinosa (outer lamina II) receives many terminations from the very finest afferent fibers; (3) the inner substantia gelatinosa (inner lamina II) receives endings from some of the finest fibers and also from intermediate (smaller myelinated) fibers; and (4) the superficial part of the nucleus proprius (lamina III) receives endings from intermediate and large diameter dorsal root fibers.

The afferent and efferent connections of the ventromedial thalamic nucleus in the rat.

Abstract: The afferent and efferent connections of the ventromedial (VM) nucleus of the thalamus in the rat were studied by experiments using the methods of retrograde cell marking by horseradish peroxidase (HRP) and anterograde fiber tracing by autoradiography. Tritiated amino acids deposited microelectrophoretically into VM label a cortical projection that is distributed to a sharply defined superficial portion of layer I of almost the entire extent of the ipsilateral neocortex. The labeling is most dense at frontal cortical levels, where fibers radiate through the deeper layers to terminate in the outer one-quarter of layer I throughout all neocortex rostral to the genu of the corpus callosum. A lesser number of labeled fibers extends caudally in a supracallosal location to innervate parieto-occipital cortical areas. Labeled collaterals ascend through the cortical layers to reach layer I, where grains in the superficial portion are found in a gradually decreasing rostrocaudal gradient of density that reaches the caudal pole of the hemisphere. Coronal sections at most levels contain a band of labeling in layer I that extends uninterrupted from the callosal sulcus at the midline to the banks of the rhinal sulcus laterally. Caudal retrosplenial and ventral temporal areas appear to be the only sectors of neocortex spared by the ubiquitous projection. Evidence for additional terminal distribution in deeper layers is found only in the dorsal and lateral sectors of the cortex rostral to the genu where sparsely labeled bands appear in layers III and V. The nearly exclusive distribution of VM's cortical afferents to layer I is compared and contrasted with multilaminar distributions of other “unspecific” cortical afferent fibers. HRP injected into VM labels neurons in a variety of structures at levels ranging from the frontal cortex to caudal medulla. Cell labelling in the globus pallidus, deep layers of the superior colliculus, cerebellar nuclei and the substantia nigra, pars reticulata suggest that VM is a point of convergence for several components of the extrapyramidal motor system. The nigrothalamic projection is topographic: medial and lateral districts of the pars reticulata are connected to medial and lateral districts of VM, respectively. A dorsal-ventral association may also obtain. Cell labeling in the prefrontal cortex, the cortex along the rhinal sulcus, the lateral habenular nucleus, tegmental and medullary reticular formations, and parabrachial nuclei indicates that VM also receives projections from more heterogeneous sources.

Rods and cones in the mouse retina. II. Autoradiographic analysis of cell generation using tritiated thymidine

Abstract: The period of cell genesis of rod and cone photoreceptor cells has been determined in the retinas of C57BL/6J mice. Embryonic mice were exposed to a single dose of 3H-thymidine at embryonic day (E) 10--18 by injecting pregnant mice intraperitoneally. Animals at postnatal ages were injected subcutaneously once between postnatal day (P) 0--10. The eyes were removed at one to three months of age. After fixation, they were embedded in glycol methacrylate, sectioned at 1.5 micrometers and prepared for autoradiographic analysis. All of the cone cells are generated over a relatively short time interval during the fetal period. In the posterior retina, the peak of cone cell genesis occurs at E13-E14, and no cones are generated after E16. The rods, by contrast, are generated later and over a longer time period. They first begin to be generated in the posterior retina on E13, but the peak of cell genesis is not reached until the day of birth, and some rods are generated as late as P5. For both rods and cones the peaks of cell genesis in the peripheral retina occur two to three days later than in the posterior retina. The findings demonstrate that rods and cones are developmentally distinct cell types in the mouse retina.

Functional organization of rat olfactory bulb analysed by the 2-deoxyglucose method.

Abstract: The spatial patterns of activity elicited in the rat olfactory bulb under different odor conditions have been analysed using the 2-deoxyglucose (2DG) technique. Rats were injected with 14C-2DG, exposed to controlled environments of amyl acetate, camphor, cage air, dimethyl disulfide, and pure air and autoradiographs prepared by the method of Sokoloff. Amyl acetate was associated with regions of glomerular layer densities in the anterolateral and mid- to posteromedial parts of the bulbar circumference, as previously reported. The extents of the densities increased with increasing concentration. Camphor odor was associated with regions of increased density in the anterodorsal and mid- to posteromedial parts of the bulb. Exposure to cage air produced scattered densities in the posteromedial and posterolateral bulb. Exposure to dimethyl disulfide gave variable results. Pure air was associated with a minimal number of small dense foci. The results with amyl acetate, camphor and cage air suggest that patterns for different odors are distinguishably different but overlapping. The regions of activity are greatest in extent and density with the highest odor concentrations. These define the regions within which more restricted and isolated foci appear at lower concentrations. The results thus provide evidence for the specific role of spatial factors in the neural processing of odor quality and odor concentration.

The GABA neurons and their axon terminals in rat corpus striatum as demonstrated by GAD immunocytochemistry.

Abstract: Glutamic acid decarboxylase (GAD, EC 4.1.1.15), the enzyme which catalyzes the α-decarboxylation of L-glutamate to form the neurotransmitter γ-aminobutyric acid (GABA), was localized immunocytochemically in neurons of rat neostriatum, pallidum and entopeduncular nucleus. A large amount of GAD-positive reaction product was observed in both the pallidum and entopeduncular nucleus in light microscopic preparations and was localized ultrastructurally to axon terminals that surrounded dendrites and large somata. In the neostriatum the relative numbers of GAD-positive axon terminals per unit area were substantially less than in the pallidum. GAD-positive terminals predominantly formed symmetric synapses with somata, dendrites and spines, but a small number of them formed asymmetric synapses with either dendrites or spines. The presence of GAD within these terminals is consistent with results of other investigations which have indicated that the striatopallidal and striatoentopeduncular pathways as well as neostriatal local circuit neurons and/or collaterals from neostriatal projection neurons, use GABA as a neurotransmitter. GAD-positive reaction product was also localized within the somata and dendrites of neostriatal and pallidal neurons in colchicine-injected preparations. The GAD-positive somata in the pallidum were medium-sized neurons and since such cells project to the substantia nigra, our results are in agreement with those from other studies which demonstrate a GABAergic, pallidonigral pathway. In the neostriatum, GAD-positive somata were identified light microscopically as medium-sized neurons with either round or fusiform shapes. Electron microscopic examinations also showed GAD-positive reaction product within the perikaryal and dendritic cytoplasm of these neurons, as well as in dendritic spines. These findings are in accord with the results of studies which have indicated that medium-sized, spinous neurons of the neostriatum give rise to a GABAergic, striatonigral pathway. The significance of GAD localization within these neostriatal neurons is discussed in relation to recent findings which show that substance P is contained within this same class of striatonigral projection neuron.

The parasagittal zonation within the olivocerebellar projection. II. Climbing fiber distribution in the intermediate and hemispheric parts of cat cerebellum.

Abstract: Olivocerebellar fibers from different subnuclei of the rostral inferior olive decussate in the brain stem and terminate as climbing fibers in one or two narrow, longitudinally arranged zones of the cerebellar cortex. These fibers issue collaterals to the central cerebellar nucleus that receives its afferents from the same cortical zone in which the parent fibers terminate. The rostral medial accessory olive projects to zone C2 and sends collaterals to the posterior interposed nucleus. A differentiation can be made between the rostral pole of this subnucleus which projects primarily to the paraflocculus and the ansiform lobule. More caudal areas connect with zone C2 in the anterior lobe and the paramedian lobule. The dorsomedial cell column projects to a lateral zone (zone A2) of lobule IX and more rostrolateral portions of the medial accessory olive supply a still more lateral zone of this lobule. The rostral half of the dorsal accessory olive sends fibers to zones C1 and C3. These fibers issue collaterals to the anterior interposed nucleus. A distinction can be made between the ventrolateral dorsal accessory olive, projecting to lobules II-IV and the ventral folia of the paramedian lobule and the dorsomedial portion of the rostral dorsal accessory olive that connects with lobules V, VI and the dorsal folia of the paramedian lobule. The most rostral part of the dorsal accessory olive provides more fibers into zone C3, more caudally located cells distribute primarily to zone C1. The rostral principal olive is connected with zone D and collateral terminations are found in the lateral cerebellar nucleus. In the paraflocculus the D zone can be divided into subzones D1 and D2. This study further substantiates the similarity in the organization of corticonuclear and olivocerebellar connections. The results are in general agreement with other recent investigations on the olivocerebellar system (Armstrong et al., '74; Brodal et al., '75; Brodal, '76; Hoddevik et al., '76; Brodal and Walberg, '77a,b; Oscarsson, '73, '76; Oscarsson and Sjolund, '77a,b).

An autoradiographic examination of the central distribution of the trigeminal, facial, glossopharyngeal, and vagal nerves in the monkey.

Abstract: The central distributions of primary afferent axons in cranial nerves V, VII, IX, and X have been re-examined autoradiographically after 3H-proline injections into their peripheral ganglia. Fiber-labeling after subtotal injections of the trigeminal ganglion, besides confirming earlier classical descriptions, suggests that trigeminal fibers of the ophthalmic and mandibular (but not maxillary) branches enter the ventrolateral part of the nucleus of the solitary tract (NST). Injection of VII's geniculate ganglion labels fibers which both ascend and descend upon reaching NST. The ascending fibers distribute in a compact and circumscribed zone immediately dorsal to the spinal V nucleus as far rostral as the caudal pole of the principal trigeminal nucleus. The descending fibers distribute to the lateral NST rostral to the level at which X joins the solitary tract. For a short distance caudal to this level, sparse label is confined to a small part of lateral NST ventral to the solitary tract, which corresponds to the zone receiving direct trigeminal afferents. Fiber-labeling after injections of the ganglia of nerves IX and X suggest the following. Although, upon reaching NST, a few fibers of either IX or X ascend as far rostrally as had those of VII, both have a much larger descending component which distributes to more caudal levels of NST. Most of IX's axons appear to end in the lateral NST; only a few travel as far as the obex. Fibers of X, on the other hand, are abundant in the medial and commissural parts of NST. Moreover, only X appears to have a crossed projection in the commissural nucleus and caudal portion of the contralateral NST. A few fibers of vagal origin also appear to enter the area postrema. Whereas fibers of X appear to constitute the solitary tract, few if any fibers of VII or IX travel within that fascicle. A significant descending components of labeled fibers appears in the spinal V tract when the superior ganglion of either IX or X is injected. These fibers distribute mainly in the pars caudalis of the spinal V nucleus and, to a lesser degree, the cuneate nucleus.

Differential thalamic relationships of sensory-motor and parietal cortical fields in monkeys.

Abstract: The thalamic connectivity of areas 3b, 1 and 2 of the first somatic sensory cortex (SI) and of the adjacent areas 4, 3a and 5 has been studied in monkeys with anterograde and retrograde labeling techniques. Anteroposterior sectors of the SI cortex are represented in the thalamic ventrobasal complex by curved lamellae of thalamocortical relay cells extending through the dorso-ventral and anteroposterior dimensions of the ventrobasal complex. Within such a lamella there are clustered aggregations of cells each projecting to a punctate zone of SI. Such cortical zones are less than 1 mm in circumference and are interpreted as comparable to the “columns” of electrophysiological studies. Each clustered aggregation in the ventrobasal complex is of limited mediolateral and dorsoventral extent but extends through much of the anteroposterior dimension of the ventrobasal complex. Punctate zones lying adjacent to one another in the mediolateral dimension of the SI cortex are connected with aggregations of cells lying in adjacent lamellae of the ventrobasal complex. Punctate zones lying anterior or posterior to one another in the anteroposterior dimension of SI are connected with aggregations of cells lying ventral or dorsal to one another in a lamella of the ventrobasal complex. Sectors of SI extending from posterior to anterior across areas 2 and 1 and others extending from posterior to anterior across area 3b, are each represented systematically across the full dorsal to ventral dimension of the ventrobasal complex. This implies at least two separate representations of the body surface: one in areas 2 and 1, and another in area 3b. Within a lamella of the ventrobasal complex, aggregations of cells projecting to areas 2 and 1 are mingled with those projecting to area 3b. Measurements of the sizes of retrogradely labeled cells in brains in which areas 2 and 1 or area 3b were separately injected showed no distinction between cells projecting to the three areas on the basis of size. Experiments combining retrograde cell degeneration due to ablation of area 3b with retrograde labeling after injection of areas 1 and 2 indicated little possibility of collateral projections to the three areas from the same cell. No part of SI, as defined by the most liberal anatomical criteria, is connected with any thalamic nucleus outside the confines of the ventrobasal complex (the caudal division of the ventroposterolateral nucleus and the large-celled part of the ventroposteromedial nucleus) or of the intralaminar complex. Area 3a, as traditionally defined, has connectional relationships that strongly suggest it is a part of the motor cortex, area 4. However, short latency Group I evoked potentials could be elicited from a small part of area 3a lacking layer V giant cells and lying adjacent to area 3b. This part receives its thalamic input from the ventrobasal complex. The data indicate that area 4 is connected with the oral division of the ventroposterolateral nucleus and with the caudal nucleus of the ventral lateral complex. These cellular groupings, however, are only artificially separated and appear to form part of the same thalamic relay nucleus. Experiments on area 5 not only suggest that the posterior boundary of SI should be placed further posterior than is customary, but also suggest that area 5 can be divided into an anterior field related to the anterior nucleus of the pulvinar and a posterior field related to the lateral posterior nucleus.

Restoration of visual cortical plasticity by local microperfusion of norepinephrine.

Abstract: Using a newly developed technique of continuous microperfusion, we obtained further evidence in support of our hypothesis that the neocortical catecholamines (CAs), particularly norepinephrine (NE), are responsible for a high level of cortical plasticity. We used the visual cortical changes in ocular dominance which follow a brief monocular deprivation as a simple and reliable index of cortical plasticity. Local perfusion of kitten visual cortex with 1 mg/ml (4.0 mM) 6-hydroxydopamine (6-OHDA) prevented the effects of monocular deprivation in kittens, thus replicating the results we had obtained using intraventricular injections (Kasamatsu and Pettigrew, '76b, '79). Locally perfused NE at a concentration of of 10(-2) mg/ml (48.6 micron) restored visual cortical plasticity in animals which were no longer susceptible to brief monocular lid-suture. These numbers refer to the concentration of solutions in the cannula/minipump system. The effective concentrations at the site of recording (about 2 mm away) are probably much lower than these. This effect of NE perfusion was seen both in kittens which had received prior 6-OHDA treatment as well as in older animals which had outgrown the susceptible period. In the kittens we obtained as a nearly complete shift in ocular dominance toward the open eye and in the older animals a decrease in binocularity was obtained. The changes were found only in the local region of visual cortex perfused with either NE or 6-OHDA, while nearby cortical regions in the same animals were unaffected. There were no obvious changes in receptive field properties of individual neurons other than ocularity, and externally perfused NE did not itself reduce binocularity in normal animals: the effects of NE described about only occurred when the animal's visual experience was simultaneously altered. These results support the view that NE plays an important role in cortical plasticity.

Morphological features of functionally defined neurons in the marginal zone and substantia gelatinosa of the spinal dorsal horn

Abstract: Functional characteristics of spinal neurons located in the marginal zone (lamina I) and substantia gelatinosa (lamina II) were compared to their structural features by intracellularly staining the source of unitary potentials with horseradish peroxidase (HRP) in unanesthetized, spinal cats. The responses of postsynaptic units to graded electrical volleys in intact dorsal roots and to physiological stimulation revealed that the peripheral excitatory input to neurons of the region is dominated by slowly conducting afferent fibers; often, the input to a given element is largely from a particular class of receptors. One type commonly seen received its principal peripheral excitation from low threshold mechanoreceptors with Aδ or C afferent fibers. Mechanoreceptive elements often exhibited a marked, prolonged habituation and many were not excited by afferent volleys. Other units were predominantly excited by nociceptors with myelinated or unmyelinated fibers, or by thermoreceptors with unmyelinated fibers. A few units (principally the thermoreceptive) showed substantial ongoing activity which was modulated by sensory stimulation, but most had little or none. The HRP staining revealed neuronal morphology in fine detail. No relationship between neuronal configuration and physiological response was discerned. Soma location was not always linked to afferent input, although the cell bodies of nonciceptive and thermoreceptive neurons tended to be in lamina I or outer lamina II (SGo) while those of the inocuous mechanoreceptive neurons tended to be in inner lamina II (SGi). The locus of a neuron's major dendritic arborization was more closely related to the source(s) of peripheral excitation. Cells excited by nociceptors with myelinated fibers had major dendritic projections in the marginal zone. Cells excited by nociceptors or thermoreceptors with unmyelinated fibers had important dendritic branching in the SGo. Innocuous mechanoreceptive neurons had primary dendritic aranching in the SGo. Innocuous mechanoreceptive neurons had primary dendritic arborization in the SGi when the input derived from unmyelinated fibers, or in the SGi and extending into the outer nucleus proprius (lamina III) when the afferent drive came from Aδ fibers. These findings support the concept that laminae I and II constitute a major termination region for thin primary afferent fibers, myelinated fibers from nociceptors ending principally in lamina I and unmyelinated fibers from nociceptors, thermoreceptors, and mechanoreceptors terminating predominantly in lamina II. Substantial integrative and distributive functions can be expected of such an afferent termination zone.

Anatomical organization of the primary visual cortex (area 17) of the cat. A comparison with area 17 of the macaque monkey.

Abstract: Golgi and axonal transport techniques have been used to examine the organization of neurons within primary visual cortex, area 17, of the cat. This organization has been compared to that of the primate cortical area 17 as described in previous studies and it is discussed in relationship to the distribution of afferents from the dorsal lateral geniculate nucleus (dLGN). The visual cortex of the cat and monkey show strong similarities in the laminar positions of neurons projecting extrinsically and also in the restriction of spiny stellate neurons to a central lamina (lamina 4) receiving input from the dLGN. However, lamina 4B in the monkey, which contains spiny stellate neurons but does not receive direct input from the dLGN, has no direct counterpart in cat area 17. Axon projections of spiny stellate neurons in the other divisions of lamina 4 differ in cat and monkey: the small, closely packed neurons in the lowermost division of lamina 4 (4B in the cat, 4Cbeta in the monkey) project chiefly within lamina 4 in the cat whereas in the monkey they have a strong projection to lamina 3. In the cat, spiny stellate neurons of lamina 4A project upon lamina 3 whereas in the monkey those in the apparently equivalent zone, 4Calpha, project upon lamina 4B. Most non-spiny stellate neurons examined have precisely organized interlaminar axonal projections which differ from the axon trajectories of neighboring spiny neurons.

The distribution of oxytocin- and neurophysin-stained fibers in the spinal cord of the rat and monkey.

Abstract: Immunohistochemical methods have been used to describe the distribution of neurophysin I- and oxytocin-stained fibers in the spinal cord. In albino rats of either sex, such fibers descend through the dorsal part of the lateral funiculus to the caudal end of the cord. Fibers leave the lateral funiculus to innervate the marginal zone of the dorsal horn at all levels, and the intermediolateral column at thoracic, lumbar and sacral levels. Some fibers course medially through the intermediate gray to end in the central gray, which is innervated at all levels, and also appears to be innervated by fibers descending in the central gray itself. A semiquantitative analysis shows that the density of terminal fields in the intermediolateral column, marginal zone, and central gray varies significantly at different levels. The intermediolateral column is preferentially innervated between T1-T3, T9-T11, and T13-L2. The nucleus intercalatusspinalis and the dorsal commissural nucleus (between T13 and L2), both of which contain preganglionic neurons, are densely innervated. In the monkey (Macaca fascicularis), neurophysin I- and oxytocin-stained fibers descend through the lateral funiculus to the filum terminale, where some of them enter the pia-arachnoid surrounding the filum. The intermediolateral column is not innervated between T1-T3, is sparsely innervated at sacral levels, and is moderately innervated between T4 and L3; the greatest density of fibers occurs at L3. At most levels, few if any fibers are found in the central gray and marginal zone; the central gray is sparsely innervated between T8 and L2 at S2, and between Co3-Co6, and the marginal zone contains a small number of fibers in the region of S2 and Co6. Neurophysin I- and oxytocin-stained fibers were also found in the intermediolateral column, marginal zone, and central gray of homozygous Brattleboro, spontaneously hypertensive, and Egyptian sand rats. The results suggest that the paraventriculo-spinal pathway is particularly related to specific groups of sympathetic and parasympathetic preganglionic neurons, and to the marginal zone, whick is involved in the relay of ascending nociceptive information through the spinothalamic tract.

Differential origins of spinothalamic tract projections to medial and lateral thalamus in the rat.

Abstract: In an earlier study (Giesler et al., 1976), antidromically identified cells of origin of the rat spinothalamic tract (STT) were found to be widely distributed throughout the lumbar dorsal horn and intermediate gray zone (IGZ). Interestingly, it appeared that STT neurons located within the ventral dorsal horn and IGZ, which appeared to respond to stimuli delivered to subcutaneous tissue, tended to be activated only from midline thalamic structures. In contrast, STT neurons within nucleus proprius and the marginal zone, which responded to noxious and innocuous cutaneous stimuli, were activated from lateral thalamic structures. The present study has investigated this apparent dichotomy in thalamic afferents using the technique of retrograde transport of horseradish peroxidase (HRP). Single and multiple small injections of HRP were made in various thalamic regions. Tissue was reacted with diaminobenzidine (DAB) or o-dianisidine (OD). Our data provide evidence that DAB histochemistry yields a more accurate reflection of the HRP concentrations surrounding an injection site which are necessary for visible retrograde transport. On the other hand, OD was found to be the more useful technique for examining labeled spinal cord neurons. With this method, 3 to 5.3 times more STT neurons were seen and far greater morphological detail was evident. Multiple HRP infusions which filled an entire hemi-thalamus labeled large numbers of cells within the spinal extensions of the dorsal column nuclei, the lateral cervical nucleus, the ventral horns and IGZ bilaterally within C1 and C2, and the nucleus proprius and IGZ at all levels. STT neurons were also seen, but in lesser numbers within the marginal zone. Labeling in spinal segments below C3 was, with a single exception, exclusively contralateral. The lumbar enlargement was seen to contribute a greater number of STT neurons than the cervical enlargement. Injections into posterior medial thalamic structures produced bilateral labeling in the ventral horns and IGZ in upper cervical segments. Within the cervical and lumbar enlargements, labeled neurons were restricted to ventral dorsal horn and to IGZ contralaterally. With the exception of the ventral horn cells of C1 and C2 where a marked reduction in labeling occurred, more anterior medial thalamic injections produced labeling within the identical areas of the gray matter and in numbers comparable to those produced by more posterior medial thalamic injections. In contrast, injections into lateral thalamic areas yielded dense labeling within the caudal extensions of the dorsal column nuclei and lateral cervical nucleus. Labeled cells were seen in the marginal zone and nucleus proprius at all levels. Together with our earlier findings, these data suggest that the rat STT is composed of two components which vary in their cells of origin, their terminations and, very likely, their functions.

Two modes of cutaneous reinnervation following peripheral nerve injury

Abstract: The return of sensation to the foot following sciatic nerve crush injury was analyzed behaviorally and electrophysiologically in the rat. Functional recovery begins within four days. Its early phase is accounted for by expansion of the functional distribution of intact neighboring fibers of the saphenous n. It occurs even if the sciatic n. is ligated, and it disappears with section of the saphenous n. Accompanying this functional expansion we began to encounter in electrophysiological recordings from the saphenous n., fibers with unusually large receptive fields (RF's) extending onto the plantar surface of the foot, well beyond their limits in intact rats. All of the expanded RF's were high threshold mechanoreceptors. On about the twentieth day after crushing, the regenerating sciatic n. began to make a functional contribution. This was seen by return of sensation to zones not invaded by the saphenous n. and by the onset of sensation in rats in which the saphenous n. had previously been ligated. With return of the sciatic n. the expanded distribution of the saphenous n. went back to its original boundaries. Correspondingly, we could no longer find expanded saphenous n. RF's. We conclude that cutaneous reinnervation begins with the collateral expansion of high threshold afferents from intact neighboring nerves. This alien innervation is later replaced upon regeneration of the original nerve.

Origin and organization of brainstem catecholamine innervation in the rat.

Abstract: The catecholamine (CA) innervation of the rat brainstem was studied by biochemical analysis of discrete nuclei or areas and by glyoxylic acid-formaldehyde freeze dry fluorescence histochemistry. CA assays demonstrate that the highest norepinephrine (NE) content in brainstem is present in the trigeminal motor nucleus, nucleus tractus solitarius, dorsal motor nucleus of the vagus and nucleus raphe dorsalis. Bilateral locus coeruleus (LC) lesions do not significantly alter NE content in these nuclei but do decrease NE content in the superior and inferior colliculi, medial geniculate body, interpeduncular nucleus, pontine nuclei and the main sensory trigeminal nucleus (60-75%). Dopamine (DA) and epinephrine (E) are found in significant concentration in only a few of the nuclei examined. Fluorescence histochemical analysis indicates that two groups of NE axons innervate rat brainstem. LC neuron axons with a distinctive morphology principally innervate sensory and association nuclei of the brainstem. These disappear completely after bilateral LC lesions. The second group of axons originates from lateral and dorsal tegmental NE cell groups. Primary motor and visceral nuclei are densely innervated by fine and thick axons from these groups. Lesions of LC do not alter the NE innervation in any of the nuclei which contain axons of the second group. These results indicate that the brainstem NE innervation is divided into two major systems. The locus coeruleus complex innervates mainly primary sensory and association nuclei whereas the lateral tegmental NE neurons innervate primary motor and visceral nuclei. Although some overlap is present, the LC and lateral tegmental NE systems predominantly innervate separate and functionally distinct areas of the brainstem. DA and E neurons provide a very minor component of the brainstem CA innervation.

Preservation of binocularity after monocular deprivation in the striate cortex of kittens treated with 6-hydroxydopamine.

Abstract: The results of single unit recordings from Area 17 of monocularly deprived kittens were compared with similar ones from littermates who had been monocularly lid-sutured for the same period of time, but who had in addition been given intraventricular injections of 6-hydroxydopamine (6-OHDA) to deplete brain catecholamines. This visual cortices of all catecholamine-depleted kittens showed high proportions of binocular neurons, in contrast to the control group, a majority of whose visual cortical neurons were driven exclusively by the non-deprived eye. Preservation of binocularity in 6-OHDA-treated kittens was dose-related. Even after a 1 to 2-week period of lidsuture which reduced binocularity to 20% in controls, normal proportions of binocular neurons (greater than 75%) were preserved if the cumulative dose had been 10 mg 6-OHDA or more. The density of single neurons sampled from electrode tracks through the cortex of drug-treated kittens was high and did not differ significantly from controls. Neurons were isolated every 100 micron on the average. There was some indication that the drug's effect in preventing an ocular dominance shift disappears by six weeks following cessation of 6-OHDA treatment. This reversal of the physiological effects in cortex is preceded by recovery from the behavioral manifestations of 6-OHDA treatments. Binocularity was only slightly increased in a kitten who received large doses of 6-OHDA after a period of monocular deprivation. This observation, together with control recordings from normal kittens and adults treated with 6-OHDA, indicates that the direct effects of 6-OHDA on cortical neurons' response properties play a minor role in comparison to its effects in reducing the sensitivity of the cortex to monocular deprivation. The overwhelming majority of cortical neurons in 6-OHDA-treated kittens remained normal in receptive field properties after a period of monocular deprivation. These data support the hypothesis that catecholamines are required for the maintenance of visual cortical plasticity during the critical period.

Cell death of motoneurons in the chick embryo spinal cord. IV. Evidence that a functional neuromuscular interaction is involved in the regulation of naturally occurring cell death and the stabilization of synapses.

Abstract: Embryos immobilized with neuromuscular blocking agents for differing periods between 4.5 and 9 days of incubation had an increased number of motoneurons in the brachial and lumbar lateral motor columns. Treatment with alpha-cobratoxin (alpha-CTX) on days 4--9, for instance, was able to prevent virtually all natural cell death during this period; control embryos had an average of 22,500 lumbar motoneurons on day 5.5, and 13,500 on day 10, whereas treated embryos had approximately 21,000 cells on day 10. Curare, alpha-CTX, alpha-bungarotoxin (alpha-BTX) and botulinum toxin were all about equally effective in preventing cell death. Similar treatment begun after day 12, however, had no effect on cell number. If even a partial immobilization was continued after day 10 (in embryos totally immobilized earlier) most of the excess neurons were maintained, in some cases right up to hatching, at which time the embryos died due to respiratory failure. In contrast, when administration of the immobilizing agents was stopped, allowing the embryos' motility to return to control levels, the excess neurons underwent a delayed cell death and total cell number fell to below control levels by days 16--18. Limb muscles from embryos with excess motoneurons exhibited relatively normal differentiation and had acetylcholinesterase (AChE) stained endplates which were innervated. Following curare treatment the two wing muscles, anterior and posterior latissimus dorsi, were found to have an increased number of AChE-stained endplates, whereas the only leg muscle examined quantitatively--the ischioflexorius (IFL)--did not; the IFL, did, however, have a markedly reduced variance in endplate distance, as well as other apparent differences suggesting an altered pattern of innervation. Our findings imply that the number of motoneurons undergoing natural cell death is closely related to muscle activity. Thus, functional interactions at the developing neuromuscular junction seem to be critical in controlling cell death. If a retrograde trophic factor is involved its action is somehow related to muscle activity.