Cobalt mapping of the nervous system: how to avoid artifacts

A half century of experimental neuroanatomical tracing

Most of our current understanding of brain function and dysfunction has its firm base in what is so elegantly called the 'anatomical substrate', i.e. the anatomical, histological, and histochemical domains within the large knowledge envelope called 'neuroscience' that further includes physiological, pharmacological, neurochemical, behavioral, genetical and clinical domains. This review focuses mainly on the anatomical domain in neuroscience. To a large degree neuroanatomical tract-tracing methods have paved the way in this domain. Over the past few decades, a great number of neuroanatomical tracers have been added to the technical arsenal to fulfill almost any experimental demand. Despite this sophisticated arsenal, the decision which tracer is best suited for a given tracing experiment still represents a difficult choice. Although this review is obviously not intended to provide the last word in the tract-tracing field, we provide a survey of the available tracing methods including some of their roots. We further summarize our experience with neuroanatomical tracers, in an attempt to provide the novice user with some advice to help this person to select the most appropriate criteria to choose a tracer that best applies to a given experimental design.

Neurons projecting from the brain to the corpora allata in orthopteroid insects: anatomy and physiology

Retrograde and orthograde labeling of neurons projecting to the corpus allatum was performed in locust, grasshopper, cricket, and cockroach species in order to identify brain neurons that may be involved in the regulation of juvenile hormone production. In the acridid grasshopper Gomphocerus rufus L., and the locusts Locusta migratoria (R.&F.) and Schistocerca gregaria Forskal, the corpora allata are innervated by two morphologically distinguishable types of brain neurons. One group of 9-13 neurons (depending on species) with somata in the pars lateralis extend axons via the nervus corporis cardiaci 2 and nervus corporis allati 1 to the ipsilateral corpus allatum, whereas two cells in each pars lateralis have bilateral projections and innervate both glands. No direct connection between the pars intercerebralis and corpus allatum has been found. In contrast, neurons with paired axons innervating both glands are not present in Periplaneta americana (L.) and Gryllus bimaculatus de Geer. Instead, two cells in each pars lateralis project only to the gland contralateral to their somata. Electrophysiological experiments on acridid grasshoppers have confirmed the existence of a direct conduction pathway between the two glands via the paired axons of four cells that have been identified by neuroanatomy. These cells are not spontaneously active under experimental conditions. Ongoing discharges in the left and right nerves are unrelated, suggesting that the corpora allata receive independent neuronal inputs from the brain.

Structural homology of identified motoneurones in larval and adult stages of hemi- and holometabolous insects

The set of neurones innervating the dorsal longitudinal muscles was studied with cobalt and nickel backfills in: (1) larval and adult locusts (Schistocerca gregaria and Locusta migratoria), (2) the larval and adult beetle (Zophobas morio), and (3) various segments of these insect species. In all specimens 11 neurones were encountered, which can be subdivided into a group of 7 motoneurones that stem from the next anterior ganglion and 4 neurones located in the ganglion of the segment containing the muscles. The latter group comprises 2 contralateral and 2 medial somata, of which one is a dorsal unpaired median neurone. The results were analysed under different aspects. This neural set and the basic structure of the dendritic fields is similar in: (1) different segments (serial homology), (2) the larval stage and imago of the same species with or without a pronounced metamorphosis (ontogeny), and (3) the studied hemi- and holometabolous insects (phylogeny). Our results support the notion that the structure of these neurones is conserved irrespectively of changes in the periphery and strategy of postembryonic development.

Motor neurons of grasshopper metathoracic ganglion occur in stereotypic anatomical groups

Anatomical groups containing identified motor neurons of the main muscles of the legs and the wings are described in a segmental ganglion of the adult grasshopper. The groups occur reproducibly in ganglia of different individuals and are a simplifying and organizing feature of ganglionic morphology. The motor neurons within each group have cell bodies near each other in the cortex of the ganglion and primary neurites that enter the ganglionic core as a discrete bundle. The primary neurite bundles are distinctive in shape and position and have the same composition in every individual, despite variations in the positions of the cell bodies of the contributing motor neurons. The primary neurite bundle of a group is separate from those of other groups and separate from bundles of motor axons that exit or sensory axons that enter the ganglion. Each group of cell bodies in the cortex appears from light microscope examination to be held separately within a glial surround. Areas of glial cell cytoplasm may extend considerably beyond the boundaries of the neuronal cell bodies, to give shape and structural integrity to the cortex. Similarities between the morphology of the adult groups reported here and the descriptions by others of embryonic and larval nervous systems suggest to us that the motor neurons of each group are the progeny of a single neuroblast.

Morphology of antennal motoneurons in the brains of two crickets, Gryllus bimaculatus and Gryllus campestris

The morphology of the antennal muscles of two cricket species, Gryllus campestris and G. bimaculatus, and their innervation are described. The motoneurons innervating the five tentorio-scapal muscles M4 and M5 and the two scapo-pedicellar muscles M6 and M7 were stained with cobalt chloride introduced via the cut axonal endings in the muscle. The seven antennal muscles are innervated by a total of 17 excitatory motoneurons and one common inhibitory neuron. These neurons branch in the dorsal neuropil of the deuto- and tritocerebrum. No difference in the morphology of the motoneurons between the two species was evident. Two dorsal-unpaired-medial (DUM) neurons located in the suboesophageal ganglion also innervate the antennal muscles. Intracellular recordings of some motoneurons combined with Lucifer Yellow injections corroborated the motoneuron morphology obtained by cobalt backfilling from the muscles. The recordings showed that the motoneurons are either of the fast or the slow type.

Cytochemical applications of X-ray microanalysis

X-ray microanalysis (XRMA) has been applied to a wide variety of cytochemical problems, but the most valuable applications have been to the validation of cytochemical methods (by the qualitative or quantitative analysis of reaction products), and to the simultaneous localization of more than one substance, which cannot easily be achieved by using alternative methods. The latter applications involve stoichiometric studies (the quantitative relationships between reaction products and substrates), and distribution studies. Ultrastructural cytochemistry with XRMA is limited by the need to use high-brightness electron sources. Apart from the limited availability of such sources, they may cause unacceptable damage to the specimen. Preparation methods for cytochemistry using XRMA are reviewed; in principle these do not differ from those used for other cytochemical applications, but it is important not to introduce extraneous elements (from fixative, buffer, or embedding medium) into the specimen, where the additional X-ray peaks may interfere with the analysis. Quantification in XRMA of cytochemical preparations poses special problems, because the addition of the reaction product to the specimen alters the yield of continuum X rays, used for assessing the mass of the specimen, and also dilutes endogenous elements. However, measurement of ratios between characteristic elemental peaks is a useful method in X-ray microanalytical cytochemistry, and it is concluded that one of the most important attributes of XRMA for cytochemical purposes is the ease with which the substances of interest can be measured.

Histological characterization of neurones innervating functionally different muscles of Locusta

The innervation of four functionally different muscles (subalar, remotor 1, remotor 2, pleuroalar), all served by the same nerve branch, was studied in both winged segments of the locust, Locusta migratoria. Several anatomical techniques were applied: With the cobalt backfill and silver intensification technique four cell types (motoneurone, dorsal unpaired median neurone, common inhibitory neurone, and small median neurone) were demonstrated. Serial sections enabled the morphology of the motoneurones to be described in more detail and in respect to a possible functional organization of the arborizations. A differential staining technique allowed us to stain various neurones in different colours in the same preparation. With this technique the anatomy of both the "rostral" and the "caudal" subalar motoneurones could be described in parallel, thus avoiding errors in comparison due to possible individual variations from preparation to preparation. Axon counts in the peripheral nerve branch enabled us to compile a list of the total innervation for each muscle. Results from other orthopterans are integrated and whether differences in the dendritic fields might be of functional significance is discussed.

Synaptic connections between the hindwing stretch receptor and flight motor neurones in the locust revealed by double cobalt labelling for electron microscopy

Synaptic interactions between sensory and motor neurones in the locust flight system have been investigated by using intracellular labelling with cobalt and nickel for electron microscopy. Simultaneous axonal filling of two neurones with different concentrations of metal ions produces differential labelling, so that contacts between them in the central nervous system can be recognized. We have investigated the connectivity of the hindwing stretch receptor neurone (SR) with a direct hindwing depressor motor neurone (MN 127) known from physiological experiments to receive monosynaptic input from the SR, and an indirect hindwing depressor motor neurone (MN 112/1), for which no monosynaptic connection with the SR has been reported. We have found no direct synapses between the SR and MN 112/1, although some of their branches lie close together in the neuropile. We have, however, found some evidence for polysynaptic connections between them. There are many synapses of conventional dyadic morphology from both the lateral and mediolateral branches of the SR to MN 127; the medial branch was not examined. Those from the lateral branch contact the motor neurone on branches close to the neuropilar segment, while those from the mediolateral branch contact long, thin distal twigs. We estimate that there are about 600 anatomical synapses between these two neurones. Our results suggest that a large number of widely distributed anatomical synapses constitute the physiological synaptic connection between the SR and MN 127. The dyadic arrangement of these synapses provides an anatomical correlate for the physiologically established divergence of SR outputs onto interneurones and motor neurones.

A comparative study of neck muscle motor neurons in a cricket and a locust

The gross morphology of the neck muscles of a cricket (Gryllus campestris) and their innervation are described and compared with a locust (Schistocerca gregaria). The motor neurons innervating the neck muscles were stained in crickets and locusts with cobalt chloride introduced via the nerve endings in the muscle. The two species show overall similarities, not only in position of the neck motor neurons in suboesophageal, prothoracic, and mesothoracic ganglia but also in motor neuron morphology. However, muscle 60 in the cricket is innervated by a unique motor neuron with its axon in prothoracic nerve 3, instead of sharing motor neurons in suboesophageal nerve 8 and mesothoracic nerve 1 with muscle 59, as in locust. Muscle 62 has the same attachments and innervation with similar motor neurons in cricket and locust but a different mechanical function in the two species. The findings are discussed with respect to possible segmental homologies and to the origins of the muscles as either dorso-ventral or longitudinal. As several muscles share the same motor neurons, we suggest that neck muscle function be described in terms of "behavioural units of action."

Differential backfilling of interneuron populations based upon axon projections in a lobster abdominal ganglion

Intersegmental interneurons in the lobster second abdominal ganglion were differentially stained by simultaneously backfilling their axons in the anterior and posterior hemiconnectives with nickel and cobalt ions. When precipitated with rubeanic acid the nickel and cobalt formed different colors in the cell bodies depending upon the location of the axon(s) in the connectives. In the most completely stained preparations 55 ascending, 56 descending, and 25 bidirectional interneurons were observed. Soma diameters ranged from 15-120 micron. One of the somata was located in the connective. The differential staining technique has advantages over conventional backfilling techniques for examining the morphological relationships of populations of neurons having different axon projections.

A cobalt study of medullary sensory projections from lateral line nerves, associated cutaneous nerves, and the VIIIth nerve in adult Xenopus

The medullary projections of the anterior lateral line nerve, dorsal branch (Alln.d), the posterior lateral line nerve, dorsal branch (PLLn.d), associated cutaneous nerves, and the VIIIth nerve in Xenopus laevis have been delineated by axonal infusion of cobalt chloride and silver intensification. The peripheral innervation of the posterior lateral line sense organs has also been traced. From wholemount and sectioned preparations, we describe three central projections, extending the length of the ipsilateral medulla but occupying distinct zones: lateral line afferents dorsomedially, stato-acoustic dorsolaterally, and cutaneous ventrolaterally. Arborizations of ALLn.d and PLLn.d afferents are morphologically similar, intermingling throughout the lateral line lobe. Each divides into ascending and descending limbs bearing collaterals, which terminate in the lateral line neuropile and nucleus. Evidence is presented for directional and positional mapping in the branching of individual PLLn.d afferents and for topography in the ALLn.d projection. Second-order neurones have been identified by transneuronal staining and their axons traced into the contralateral torus semicircularis. The morphology of efferent neurones is also described. Rostral branches of PLLn.d also contain cutaneous afferents which run through the medulla into the spinal cord, similar to the nerve V (cutaneous) projection. In nerve VIII preparations, the projection to the compact cochlear nucleus and the massive vestibular projection are identified. Cutaneous and vestibular but not lateral line afferents extend into the cerebellum. The separation of VIIIth nerve and lateral line afferents in Xenopus medulla is considered as evidence against the validity of the acousticolateralis concept. Information processing in the lateral line lobe is discussed in relation to connectivity patterns between first- and second-order neurones.