Characterization of the pigeon isthmo-tectal pathway by selective uptake and retrograde movement of radioactive compounds and by Golgi-like horseradish peroxidase labeling.

Horseradish peroxidase (HRP) injected into developing limb buds of Xenopus laevis tadpoles is carried by retrograde axonal transport to the somata of motoneurones in the ventral horn. Small injection of 10% HRP were found to remain well localised to specified sites in the limb bud. Two types of labelled cells were found: diffusely labelled and granular labelled. Diffusely labelled cells result from axonal damage in the presence of HRP. Granular labelled cells result only from uptake of HRP from the region of the axon endings. No gradular uptake was found from axon shafts. It is concluded that the distribution of granular labelled cells accurately reflects the region of the ventral horn projecting to the site of injection in the limb.     

Absence of postnatal death among motoneurones supplying the inferior gluteal nerve of the rat

Motoneurones innervating the caudal part of the gluteus maximus muscle of 0-2 day, 10-12 day and 2-3-month-old rats were labelled by a half-hour application of a solution of 30% horseradish peroxidase (HRP) and 2% lysophatidylcholine delivered by suction electrode to the cut inferior gluteal nerves. The numbers of motoneurones labelled 24-48 h later were not significantly different in the 3 age groups (mean = 58.75, 54.0, 57.5, respectively). When a simple 30% HRP solution was used in adult rats, the number of motoneurones labelled was significantly less (mean = 48.75). In contrast, application of 0.5 microliter of HRP in a pledget of gelfoam to either the cut or uncut inferior gluteal nerve of neonates labelled large numbers of motoneurones, presumably by diffusion into nearby muscles. It is concluded that no death of motoneurones innervating the gluteus maximus muscle occurs postnatally, and that spread of HRP to neighbouring muscles can give rise to spuriously high motoneurone counts in neonates, and that incomplete uptake or transport of HRP in adults can lead to incorrectly low counts.     

Compartmental and topographical distributions of axons in nerves to the amphibian (Bufo marinus) glutaeus muscle

The present work seeks to determine if axons to an amphibian muscle are segregated in nerve trunks between the spinal cord and muscle according to their primary nerve destination or their topographical projection in the muscle. The distribution of axons to different compartments and subcompartments of the amphibian (Bufo marinus) glutaeus muscle has been determined in transverse sections of spinal and limb nerves after retrogradely labelling the axons with horseradish peroxidase. Glutaeus axons were dispersed widely through spinal nerves 8 and 9 but loosely gathered together in one quadrant of the sciatic nerve after passing through the lumbar plexus. Glutaeus axons became tightly clustered to the exclusion of other axons along the length of the triceps femoris nerve after it divides from the sciatic nerve. Furthermore, axons destined for one of the two glutaeus primary nerve branches segregate from those of the other branch at the level of the triceps femoris nerve before the glutaeus nerve forms. On the other hand, motoneurones that subserve a primary branch are not segregated, but are found throughout the rostrocaudal extent of the glutaeus motoneurone pool. Injection of horseradish peroxidase under the epimysium of either the ventral or the dorsal surfaces of the glutaeus muscle labelled motoneurones preferentially in either the rostral or caudal part of the motoneurone pool, respectively. This confirms studies that have shown a topographical projection from the spinal motoneurone pool onto the glutaeus muscle. However, there was no segregation of dorsally projecting axons in the glutaeus and primary nerve branches. Thus, glutaeus axons segregate according to their muscle compartmental projections well before entering the muscle, but they show no organization in nerves with respect to their topographical projections within a compartment.     

Quantitative synaptology of feline motoneurones to external anal sphincter muscle

Motoneurones innervating the cat external anal sphincter muscle were labelled retrogradely following intramuscular injections with horseradish peroxidase (HRP). Labelled motoneurones were examined by correlative light and electron microscopy (LM and EM) with special regard to a qualitative and morphometric analysis of the axon terminals resident on the neuronal membrane. By LM, labelled motoneurones were (1) ipsilateral to the injections; (2) all in S1-S2; (3) found only in the superior dorsomedial region of Onuf's nucleus; and (4) exhibited a broad spectrum of diameters (25-72 micron, mean 47.4 +/- 11.3 micron). By EM, axon terminals on the neuronal membrane when classified according to size, vesicle shape, and synaptic complex ultrastructure conformed to the S-, F-, T-, M-, and C-type terminals previously described for cat lumbosacral motoneurones. C-terminals confirmed these sphincteric motoneurones to be skeletomotor. Pooled data from midnuclear sections through 15 random labelled motoneurones (20-64-micron diameter) revealed that S- and F-type terminals predominated, with numerically few M and C types. Notwithstanding their low frequency (0.3/100 micron membrane) C-terminals contributed 1% of the mean areal coverage by terminals, which implies a potentially larger synaptic influence relative to other terminal types. Linear relationships occurred between terminal frequency (or cover) and motoneurone diameter. While motoneurones greater than 40 micron in diameter exhibited all five terminal types, labelled motoneurones less than or equal to 30 micron generally possessed only S-, F-, and occasional T-type terminals, and in this respect resembled gamma motoneurones.     

Distribution of calcitonin gene-related peptide-like immunoreactivity in the nucleus ambiguus of the cat

The distribution of calcitonin gene-related peptide (CGRP) in the cat nucleus ambiguus was examined by means of a combination of horseradish peroxidase (HRP) tracing and immunohistochemical techniques. Vagal motoneurones in the nucleus ambiguus were identified by applying HRP to either the thoracic vagus or the superior laryngeal nerve or the cervical vagus. Motoneurones in the nucleus ambiguus labelled with HRP from the thoracic vagus did not contain CGRP-like immunoreactivity although CGRP-like immunoreactive cells were present in this nucleus on the same sections. In contrast, a large proportion of the motoneurones labelled from the superior laryngeal nerve and a smaller proportion of cells labelled from the cervical vagus did contain CGRP-like immunoreactivity. It is concluded that CGRP-like immunoreactivity is absent from vagal preganglionic motoneurones projecting to structures in the thorax and abdomen but is present in vagal motoneurones projecting to striated muscle of the larynx and pharynx.     

Dibutyryl cyclic guanine monophosphate inhibits natural but not limb amputation-induced motoneurone death in the chick embryo

The number and distribution of motoneurones in the lateral motor columns of normal chick embryos have been compared with embryos treated daily with dibutyryl cyclic guanine monophosphate and subject to unilateral limb bud amputation. Treatment with cyclic nucleotide rescued about 17% of the motoneurones compared with controls by Stage 37, but did not prevent amputation-induced motoneurone death. The role of cyclic nucleotides in motoneurone survival is therefore permissive, rather than instructive.     

Chromatophore motoneurons in the brain of the squid, Lolliguncula brevis: a HRP study

The location of the motoneuron somata controllingg activity of the chromatophore muslces was studied in the squid Lollinguncula brevis. Retrograde transport of horseradish peroxidase from injection sites in the skin or in the mantle muscle established that the chromatophore motoneurons are situated in the subesophageal mass of the brain while at least some of the mantle muscle motoneurones are in the stellate ganglia. Motoneurons to chromatophores in the mantle have their somata in the posterior subesophageal mass, mainly in the chromatophore or fin lobes. Motoneurons to chromatophores in the head are located in the anterior pedal lobes and those to the chromatophores in the arms project mainly from the anterior chromatophore lobes. However, some neurons in the posterior chromatophore lobes project to the head or arm regions. A few cells in both the anterior and posterior chromatophore lobes project contralaterally. Somata in other lobes of the subesophageal mass are also labelled by injections in the skin or in the mantle muscle. Evidence presented here suggests that some of the neurons labelled outside the chromatophore lobes are chromatophore motoneurons.     

Further indications for enhancement of retrograde transneuronal transport of WGA-HRP by synaptic activity

Factors affecting the retrograde transneuronal transport of wheat germ agglutinin conjugated with horseradish peroxidase (WGA-HRP), from spinal motoneurones to interneurones, have been studied in the cat. To this end, the location of transneuronally labelled interneurones was compared in animals which were awake or remained anaesthetized after WGA-HRP had been injected into the semitendinosus or the medial gastrocnemius nerve. In the anaesthetized animals motor axons of the injected nerves were stimulated selectively, to activate only Renshaw cells, or together with group I afferents, to activate also laminae V–VI interneurones with input from these muscles. The interneurones labelled in this study were distributed in different spinal cord regions than the interneurones labelled in preparations in which group I afferents of antagonist muscles were stimulated, as described in a previous study. The reported observations extend evidence of Harrison et al. that the retrograde transneuronal transport of WGA-HRP is facilitated by synaptic activity.     

The source of the respiratory drive to nasolabialis motoneurones in the rabbit; a HRP study

Lower brainstem projections to the motoneurones of the nasolabialis muscles, which show rhythmic respiratory-phased activity were studied in the rabbit using the horseradish peroxidase (HRP) technique. The nasolabial motoneurone pool was first identified by the retrograde transport of HRP injected intramuscularly, and by antidromic stimulation and microelectrode recording techniques. The results from subsequent iontophoretic injection of HRP into the lateral division of the facial nucleus (the nasolabial pool) produced significant ipsilateral labelling in the nucleus ambiguus-retroambigualis (NA-NRA) complex. Labelled cells, predominantly ipsilateral, were also consistently observed in the parvocellular reticular nucleus. Smaller numbers of labelled cells were identified in the ventral, dorsal and gigantocellular nuclei of the reticular formation on both sides of the medulla. A large proportion of HRP-labelled cells of the NRA was located in the caudal medulla where the presence of propriobulbar and bulbospinal respiratory neurones has been well documented. These results suggest that some neurones of the NA-NRA complex may serve as upper respiratory motoneurones to the nasolabial musculature.     

Input conductance, axonal conduction velocity and cell size among hindlimb motoneurones of the cat

Input conductance and axonal conduction velocity were measured for hindlimb motoneurones that were anatomically labelled by substances injected through the intracellular microelectrode (Procion dyes, horseradish peroxidase). We confirmed that there is a good correlation between the axonal conduction velocity of a hindlimb motoneurone and the size of its cell body. Furthermore, we confirmed that the power relation between neuronal input conductance and axonal conduction velocity has an exponent of about 3–4. If large motoneurones were simply scaled-up versions of the smaller ones, this exponent should be have been between 1.5 and 2.0. We showed that the unexpectedly high input conductance of fast-axoned motoneurones, compared to that of the more slow-axoned ones, was not due to a corresponding disproportion between the axonal conduction velocity and the size of the cell body. Neither could it be explained by differences between large and small cells with respect to the relative sizes and numbers of dendritic stems. The unexpectedly high input conductance of large cells seems likely to be largely caused by a lower average value for the specific membrane resistance among these cells than among the smaller ones. Hitherto unknown differences in dendritic architecture between large and smaller cells might conceivably be of some importance as well. Our results are consistent with the view that, in muscle contractions evoked by the central nervous system, thin-axoned motoneurones might be recruited more easily than more thick-axoned ones even if all the cells were activated by the same density of equipotent synapses.     

Almost all ganglion cells in the rabbit retina project to the superior colliculus

After a localized injection of horseradish peroxidase into the superior colliculus of the rabbit almost all of the ganglion cells in the monocular portion of the contralateral retina were retrogradely labelled with the enzyme. Of the few ganglion cells which were not labelled, most had large cell bodies.     

The retinohypothalamic tract in the cat: retinal ganglion cell morphology and pattern of projection

The pattern of retinal projection to the hypothalamus and the morphological properties of the retinal ganglion cells that comprise the retinohypothalamic tract have been examined in the cat. Intraocular injections of horseradish peroxidase revealed a dense retinal projection to the ventral suprachiasmatic nucleus; however, lighter projections were seen in the dorsal suprachiasmatic nucleus, and in hypothalamic regions both dorsal and lateral to the suprachiasmatic nucleus. Intrasuprachiasmatic nucleus injections of horseradish peroxidase retrogradely labelled retinal ganglion cells that were small to medium in soma size. The labelled ganglion cells exhibited long thin dendrites that were sparsely branched. The labelled retinal ganglion cells exhibited a significant change in soma size associated with retinal eccentricity. The morphological characteristics of the ganglion cells that project to the suprachiasmatic nucleus are similar to those of gamma cells.     

Transneuronal transport of wheat germ agglutinin-conjugated horseradish peroxidase into trigeminal interneurones of the rat

Intramuscular injections of either horseradish peroxidase (HRP) or wheat germ agglutinin-conjugated horseradish peroxidase (WGA-HRP) were made into the masseter muscle of rats. Both tracers labeled primary sensory neurones in the V mesencephalic nucleus, motoneurones in the V motor nucleus, and some motoneurones in the facial motor nucleus. WGA-HRP labeled additional neurones in the V main sensory nucleus and the rostral pole of the V nucleus oralis. These were classed as interneurones because they lay in areas outside those known to contain either first-order afferent or motoneurone somata. We argue that these were labeled by retrograde transport of tracer because they lay close to the V motor nucleus, and from some of them processes could be followed into the region of the V motor nucleus.     

Nigrostriatal projections in the rat as demonstrated by retrograde transport of horseradish peroxidase. I. Projections to the rostral striatum

The distribution of nigral neurons projecting to the rostral part of the striatum was studied in 12 rats using the horseradish peroxidase or the wheatgerm lectin bound horseradish peroxidase labelling techniques. Labelled neurons localized in the substantia nigra pars compacta (SNc) were demonstrated throughout the anteroposterior extent of the nucleus. Most of the labelled neurons were localized in the medial half of the SNc, predominantly in its basal part. Labelled neurons localized in the substantia nigra pars reticulata (SNr) predominated in the caudal half of the nucleus where they were found in its medial, central and lateral parts. A quantitative evaluation of the shape and size of the labelled neurons showed no statistically significant differences between the SNc and SNr as to the shape of the labelled perikarya. In contrast, nigrostriatal neurons in the SNc were found to have larger perikarya than their counterparts in the SNr.     

Development of motor innervation of the chick following dorsal-ventral limb bud rotations

To test mechanisms which motoneurons may use to grow to their appropriate targets, I rotated the limb around the dorsal-ventral axis prior to motoneuron outgrowth. The positions of motoneurons in the spinal cord innervating individual muscles and muscle masses were then determined using retrograde horseradish peroxidase uptake. Motoneurons innervated their appropriate muscles after dorsal-ventral limb rotation, before and after motoneuron death. Thus, cell death does not serve to remove errors in matching between motor nuclei and their corresponding muscles after dorsoventral rotation of the limb. Motoneurons must be specified for a peripheral target prior to outgrowth, and they grow to that target relatively directly. Axons compensated for the limb rotation by first collecting into groups in a position appropriate for the normal limb orientation, then shifting dorsal-ventral position within the plexus and proximal nerve trunk. Based on these results it is hypothesized that axons destined to innervate dorsal or ventral musculature might use chemospecific cues during growth to maintain appropriate positions within the nerve with respect to limb orientation.     

Transplantation of embryonic neurones to replace missing spinal motoneurones

Loss of spinal motoneurones results in severe functional impairment. The most successful way to replace missing motoneurones is the use of embryonic postmitotic motoneurone grafts. It has been shown that grafted motoneurones survive, differentiate and integrate into the host cord. If grafted motoneurones are provided with a suitable conduit for axonal regeneration (e.g. a reimplanted ventral root) the grafted cells are able to grow their axons along the whole length of the peripheral nerves to reach muscles in the limb and restore function. Grafted motoneurones show excellent survival in motoneurone-depleted adult host cords, but the developing spinal cord appears to be an unfavourable environment for these cells. The long term survival and maturation of the grafted neurones are dependent on the availability of a nerve conduit and one or more target muscles, no matter whether these are ectopic nerve-muscle implants or limb muscles in their original place. Thus, grafted and host motoneurones induce functional recovery of the denervated limb muscles when their axons regenerate into an avulsed and reimplanted ventral root. On the other hand, motoneurone-enriched embryonic grafts placed into a hemisection cavity in the cervical spinal cord induce axonal regeneration from great numbers of host motoneurones, possibly by the bridging effect of the grafts. In this case the regenerating host motoneurones reinnervate their original target muscles while the graft provides few axons for the reinnervation of muscles. These results suggest that reconstruction of the injured spinal cord with embryonic motoneurone-enriched spinal cord graft is a feasible method to improve severe functional motor deficits.     

The source of the respiratory drive to nasolabialis motoneurons in the rabbit; a HRP study

Lower brainstem projections to the motoneurons of the nasolabialis muscles, which show rhythmic respiratory-phased activity were studied in the rabbit using the horseradish peroxidase (HRP) technique.The nasolabial motoneurone pool was first identified by the retrograde transport of HRP injected intramuscularly, and by antidromic stimulation and microelectrode recording techniques. The results from subsequent iontophoretic injection of HRP into the lateral division of the facial nucleus (the nasolabial pool) produced significant ipsilateral labelling in the nucleus ambiguus-retroambigualis (NA-NRA) complex. Labelled cells, predominantly ipsilateral, were also consistently observed in the parvocellular reticular nucleus. Smaller numbers of labelled cells were identified in the ventral, dorsal and gigantocellular nuclei of the reticular formation on both sides of the medulla. A large proportion of HRP-labelled cells of the NRA was located in the caudal medulla where the presence of propriobulbar and bulbospinal respiratory neurones has been well documented.These results suggest that some neurones of the NA-NRA complex may serve as upper respiratory motoneurones to the nasolabial musculature.     

Innervation and structure of extraocular muscles in the monkey in comparison to those of the cat

Motoneurones that innervate the medial rectus, lateral rectus, and accessory lateral rectus muscles in the monkey have been identified and localized by retrograde transport of horseradish peroxidase. Medial rectus motoneurones were located within both dorsal and ventral regions of the oculomotor nucleus, with a differential distribution along the rostral-caudal axis of the nucleus. Lateral rectus motoneurones were located predominantly within the abducens nucleus, and were distributed throughout the rostral-caudal extent of the nucleus. Motoneurones that innervate the accessory lateral rectus muscle comprised a group of large cells located approximately 0.5 mm ventral to the rostral protion of the abducens nucleus, corresponding to the ventral abducens nucleus of Tsuchida ('06). The ventral subgroup of abducens motoneurones, which innervate both the lateral rectus and accessory lateral rectus muscles, thus do not occupy a brain stem location similar to the cat accessory abducens nucleus, whose motoneurones innervate the retractor bulbi muscle, to which the accessory lateral rectus muscle presumably is homologous. A few accessory lateral rectus motoneurones also were located within the abducens nucleus, overlapping the distribution of lateral rectus motoneurones. Electron microscope examination of the lateral rectus muscle revealed the presence of three morphological types of singly innervated muscle fibers and two morphological types of multiply innervated muscle fibers that exhibited a differential distribution within the orbital, intermediate, and global regions of the muscle. The accessory lateral rectus muscle resembled the global portion of the lateral rectus muscle in containing two morphological types of singly innervated fibers and one type of multiply innervated fiber. These findings indicate that the central differences in the brainstem locations of motoneurones that innervate the cat retractor bulbi and monkey accessory lateral rectus muscles are correlated with peripheral differences not only in the morphology, but also possibly in the mechanical roles, of the muscles they innervate. The accessory lateral rectus muscle thus appears to have evolved both structurally and functionally towards more of a role in patterned eye movement. Furthermore, with the phylogenetic regression of the retractor bulbi muscle, the various types of eye movement with which this muscle is associated in lower vertebrates may be assumed by the other extraocular muscles in higher mammals, including humans.     

Glycine-immunoreactive terminals in the rat trigeminal motor nucleus: light- and electron-microscopic analysis of their relationships with motoneurones and with GABA-immunoreactive terminals

Post-embedding immunolabelling methods were applied to semi-thin and ultrathin resin sections to examine the relationships between glycine- and γ-aminobutyric acid (GABA)-immunoreactive terminals on trigeminal motoneurones, which were identified by the retrograde transport of horseradish peroxidase injected into the jaw-closer muscles. Serial sections were cut through boutons and alternate sections were incubated with antibodies to glycine and GABA. Light-microscopic analysis of semi-thin sections revealed a similar pattern of glycine and GABA-immunoreactive boutons along the motoneurone soma and proximal dendrites, and of immunoreactive cell bodies in the parvocellular reticular and peritrigeminal areas surrounding the motor nucleus. Immunoreactive synaptic terminals on motoneurones were identified on serial ultrathin sections at electron-microscopic level using a quantitative immunogold method. Three populations of immunolabelled boutons were recognized: boutons immunoreactive for glycine alone (32%), boutons immunoreactive for GABA alone (22%), and boutons showing co-existence of glycine and GABA immunoreactivities (46%). Terminals which were immunoreactive for glycine only contained a higher proportion of flattened synaptic vesicles than those which were immunoreactive for GABA only, which contained predominantly spherical vesicles. Terminals which exhibited both immunoreactivities contained a mixture of vesicle types. All three classes of terminal formed axo-dendritic and axo-somatic contacts onto retrogradely labelled motoneurones. A relatively high proportion (25%) of boutons that were immunoreactive for both transmitters formed synapses on somatic spines. However, only GABA-immunoreactive boutons formed the presynaptic elements at axo-axonic contacts: none of these were found to contain glycine immunoreactivity. These data provide ultrastructural evidence for the role of glycine and GABA as inhibitory neurotransmitters at synapses onto jaw-closer motoneurones, but suggest that presynaptic control of transmission at excitatory (glutamatergic) synapses on motoneurones involves GABAergic, but not glycinergic inhibition.