Synaptic structure, distribution, and circuitry in the central ... - CiteSeerX

MICROSCOPY RESEARCH AND TECHNIQUE 56:210 –226 (2002)

Synaptic Structure, Distribution, and Circuitry in the Central Nervous System of the Locust and Related Insects ALAN HUGH DAVID WATSON1

AND

¨ RMANN2* FRIEDRICH-WILHELM SCHU

1

School of Biosciences, University of Wales, Cardiff, CF10 3US, Great Britain 2 Institute of Zoology and Anthropology, University of Go¨ttingen, Berliner Straße 28, D-37073 Go¨ttingen, Germany

KEY WORDS

synapses; ultrastructure; insect central nervous system; orthopterans

ABSTRACT The Orthopteran central nervous system has proved a fertile substrate for combined morphological and physiological studies of identified neurons. Electron microscopy reveals two major types of synaptic contacts between nerve fibres: chemical synapses (which predominate) and electrotonic (gap) junctions. The chemical synapses are characterized by a structural asymmetry between the pre- and postsynaptic electron dense paramembranous structures. The postsynaptic paramembranous density defines the extent of a synaptic contact that varies according to synaptic type and location in single identified neurons. Synaptic bars are the most prominent presynaptic element at both monadic and dyadic (divergent) synapses. These are associated with small electron lucent synaptic vesicles in neurons that are cholinergic or glutamatergic (round vesicles) or GABAergic (pleomorphic vesicles). Dense core vesicles of different sizes are indicative of the presence of peptide or amine transmitters. Synapses are mostly found on small-diameter neuropilar branches and the number of synaptic contacts constituting a single physiological synapse ranges from a few tens to several thousand depending on the neurones involved. Some principles of synaptic circuitry can be deduced from the analysis of highly ordered brain neuropiles. With the light microscope, synaptic location can be inferred from the distribution of the presynaptic protein synapsin I. In the ventral nerve cord, identified neurons that are components of circuits subserving known behaviours, have been studied using electrophysiology in combination with light and electron microscopy and immunocytochemistry of neuroactive compounds. This has allowed the synaptic distribution of the major classes of neurone in the ventral nerve cord to be analysed within a functional context. Microsc. Res. Tech. 56:210 –226, 2002. © 2002 Wiley-Liss, Inc. INTRODUCTION The brain and ventral nerve cord of locusts and other insect groups have provided important model systems for the analysis of the neural networks underlying both simple and complex behaviour. One of the key factors that has contributed to the interpretability of insect neuronal circuitry is that it incorporates many neurones that are uniquely identifiable in terms of their morphology and their physiological characteristics. This allows a synergistic approach to the analysis of circuitry, in which the physiological properties and synaptic distribution of single neurones can be viewed in the context of the behaviour that the circuit subserves. Among insects, the locust and other large Orthopterans are particularly well suited to this approach because of their large size and the accessibility of the nervous system. The Orthopteran nervous system also shares many properties with that of Diptera such as Drosophila, in which such studies are still largely impracticable but whose suitability for genetic manipulation has led to insights into the molecular biology of insect neurones and their synaptic equipment (Prokop, 1999). This review focuses on the morphology and distribution of synapses viewed in the context of synaptic circuitry and is based primarily on information derived from the locust nervous system. However, it also draws insights from studies of other insects. We provide a 2002 WILEY-LISS, INC. DOI 10.1002/jemt.10031

©

brief review of synaptic ultrastructure of both chemical and electrical synapses, and also some information on putative non-synaptic release of neurotransmitters from as yet unpublished data. The distribution and abundance of synaptic contacts on different classes of neurones are considered in some detail and we present examples of the complex synaptic microcircuitry in which certain individual physiologically characterised neurones are involved. The structural dynamics of synapses during pre- and postembryonic development, and during degeneration and regeneration, are outside the scope of this review. For an analysis of these topics, the reader is referred to the work of Budnik (1996), Davis and Goodman (1998), Meinertzhagen (1993), Meinertzhagen and Hu (1996), and Prokop (1999). SYNAPTIC ULTRASTRUCTURE Chemical Synapses The contacts between neurones that are interpreted as being the sites of chemical synapses typically share a number of structural features: synaptic vesicles as-

*Correspondence to: Prof. Friedrich-Wilhelm Schu¨rmann, Institute of Zoology and Anthropology, University of Go¨ttingen, Berliner Straße 28, D-37073 Go¨ttingen, Germany. E-mail: [email protected] Received 25 April 2001; accepted in revised form 3 July 2001 Contract grant sponsor: Deutsche Forschungsgemeinschaft SFB 406 Go¨ttingen; Contract grant sponsor: Wellcome Trust; Contract grant number: 052313/Z/97.

SYNAPSES IN THE CNS OF ORTHOPTERAN INSECTS

sociated with presynaptic electron dense membrane appositions, a widened synaptic cleft, and a sub-membranous electron-dense coating in the opposing postsynaptic elements (Figs. 1–3; see also Figs. 5,6). These basic components have been frequently observed in locusts and other insect forms using conventional electron microscopy (Boeckh et al., 1970; Burkhardt and Braitenberg, 1976; Osborne, 1966; Schu¨ rmann, 1980; Schu¨ rmann and Wechsler, 1969, 1970; Tyrer and Altman, 1976; Wood et al., 1977). The sub-membranous densities and the material within the synaptic cleft are best visualised in aldehyde fixed, non-osmicated tissue (Schu¨ rmann, 1980) in which the membranes themselves are left unstained (Fig. 1C). A single presynaptic terminal may lie opposite a single postsynaptic neurite (a monadic synapse) or more commonly, a pair of postsynaptic processes (a dyadic synapse) (e.g., Watson and Burrows, 1982). In insects, neuronal somata generally lie outside the neuropile and are heavily ensheathed in glial cells. However, synapse-like specialisations made by dopaminergic neurones are common on Kenyon cell somata in the mushroom bodies of bees (Blenau et al., 1999). Comparable studies on Orthopteran Keynon cells have yet to be carried out. While the postsynaptic element at central chemical synapses is almost universally another neurone, occasionally this can be demonstrated as a glial cell process (Fig. 1F) (Watson and Burrows, 1982). In insects, electron-lucent vesicles in the presynaptic terminal are typically associated with presynaptic bars or rows of electron-dense particles (Fig.1; see also Figs. 5,6). In locusts, the bar, which is by far the most abundant presynaptic structure, typically ranges from about half a micrometer to a few micrometers in length and, depending on fixation, a narrow plate may be seen running above it (Foelix and Choms, 1979; Watson and Burrows, 1982; Wood et al., 1977). This differs significantly from the presynaptic bars seen at synapses in the fly brain, at which the bar has a prominent broad plate giving it a “T”-shaped profile in cross-section (Burkhardt and Braitenberg, 1976; Fro¨ hlich and Meinertzhagen, 1983). A presynaptic grid of regularly arranged plated columns is found at many synapses in the bee brain (Schu¨ rmann, 1971). The functional significance of these different presynaptic figures remains obscure. At some dyadic synapses, two or even all three of the neurites involved may contain a presynaptic bar (Burrows et al., 1989; Watson and Pflu¨ ger, 1984). Though numerically uncommon, such forms are locally abundant in certain neuropiles (e.g., of the mushroom bodies) or at synapses involving certain identified neurones (Burrows et al., 1989; see below). As can be seen from contacts between the DCMD/LGMD neurones in the locust brain (Killmann et al., 1999), a single neuron may form both dyadic and monadic synapses with or without presynaptic structures. Freeze fracture studies of insect central nervous tissue have revealed clusters or rows of what appear to be vesicle attachment sites on either side of troughs that are interpreted as the presynaptic membrane of dyadic synapses (Fro¨ hlich, 1985; Tolbert and Hildebrand, 1981; Wood et al., 1977). Coated pits that may represent the sites of membrane retrieval have also been reported close to the active zone (Lane et al., 1983). The presynaptic membrane in the vicinity of the region of

211

synaptic contact may show an enhanced electron density compared to adjacent regions. However, this is less pronounced than that associated with the postsynaptic membrane (Killmann et al., 1999). At the postsynaptic membrane, the active zone is marked by a submembranous electron dense layer of variable extent (Figs. 1,2). A number of active zones in close proximity may occur along extended regions of membrane apposition between two neurons (Fig.1F) (Killmann et al., 1999), reminiscent of the distribution of synaptic loci at fly neuromuscular junctions (Feeny et al., 1998). The synaptic cleft has a width of 10 –20 nm, which varies depending on synaptic type (Fig. 2) (Killmann et al., 1999). The enhanced electron density of the material in the synaptic cleft and of the submembranous structures of the pre-and postsynaptic membranes is most probably a manifestation of the synaptic proteins associated with chemical transmission and synaptic stabilisation (Garner et al., 2000; Kneussel and Betz, 2000; Prokop, 1999). In the insect nervous system, the identities and functional roles of these synaptic proteins are now under intensive investigation, particularly at the glutamatergic neuromuscular synapses of Drosophila (Prokop 1999). Identification of the precise molecular structures of synaptic proteins at glutamatergic and non-glutamatergic synapses within the locust nervous system requires further molecular and pharmacological investigation. The synaptic vesicles that lie in closest proximity to the presynaptic densities in the locust central nervous system are generally small and electron lucent and fall into two categories: round with a typical diameter range of 35–50 nm, or pleomorphic with dimensions of 20 –30 nm. Pleomorphic vesicles are found almost exclusively in terminals that show evidence for the presence of GABA, while round agranular vesicles are found in terminals that are immunoreactive for glutamate (Watson, 1988; Fig. 3A,B). Neurones that appear to be cholinergic (Leitinger and Simmons, 2000; Lutz and Tyrer, 1988) also contain round agranular vesicles (Watson and England, 1991; Simmons and Littlewood, 1989; Watson et al., 1991, 1993). Synaptic terminals containing round agranular vesicles are generally more numerous than those containing pleomorphic vesicles and in ultrastructural immunocytochemical studies of identified interneurones or efferent neurones, about 20 – 40% of presynaptic terminals that contain electron-lucent vesicles exhibit GABA-like immunoreactivity. Following chemical fixation, electron lucent synaptic vesicles are only rarely seen to be undergoing exocytosis as shown for stimulated locust motorneurone-muscle synapses (Reinecke and Walther, 1981). Occasionally evidence of vesicle recycling in the form of clathrin-coated vesicles has been reported in insects (Killmann et al., 1999). Large dense core vesicles with a diameter of 60 –120 nm occur in all synaptic neuropiles of Orthopteran insects (Fig. 3C–G). These may be scattered among electron lucent vesicles but more often lie in distinct clusters in the vicinity of the presynaptic membrane. Dense core vesicles are particularly abundant in certain neuropile compartments within the brain (e.g., lateral protocerebrum, central body complex, mushroom bodies, and deutocerebral neuropiles) that may be richly endowed with neuropeptides and biogenic amines (Na¨ ssel, 1996; see Homberg, pages

212

¨ RMANN A.H.D. WATSON AND F.-W. SCHU

Fig. 1.

SYNAPSES IN THE CNS OF ORTHOPTERAN INSECTS

213

Fig. 2. Quantification of the morphological characters of two distinct structural types of synapses between unidentified presynaptic elements and the purely postsynaptic dendrite of a cobaltstained DCMD plurisegmental interneuron in the brain of the locust; from densitometry of film negatives at high magnifications. Horizontal bars: range of variation; vertical bars and rectangles, mean ⫾ standard deviation; n, number of scanning lines. Note the different horizontal scales. The two types of synapses significantly differ with respect to the thickness of subsynaptic membrane appositions, but not with respect to the cleft width or thickness of the membranes. There is a significant negative correlation between the cleft width and the thickness of subsynaptic material only in Type I synapses (from Killmann et al., 1999).

189 –209, in this issue). Such regions have been tentatively proposed as centres of neuromodulation (Schu¨ rmann et al., 1991). An abundance of dense core vesicles is also typical of neurosecretory efferent neurones con-

Fig. 1. A–G: Electron micrographs of synaptic contacts in the cricket (A–E) and locust (F,G) nervous system. Sites of chemical synapses are defined by intracellular electron dense membrane appositions of variable extent, and synaptic clefts. Presynaptic profiles containing electron-lucent synaptic vesicle that are generally clustered around bars or bar-like figures (arrowheads). At these sites two or more postsynaptic profiles generally lie in appostition to the presynaptic membrane (dyadic and polyadic synapses) (A–C,G). More rarely, only a single process may lie in a postsynaptic postion (monadic synapse) (D,E). Extended divergent synaptic contacts in a presynaptic bouton may involve numerous postsynaptic elements as seen here in mushroom body calyx. A: A large presynaptic bouton with divergent connectivity shows extended active zones (upper part of panel). B: A typical synaptic contact showing the presynaptic bar (arrowhead) separated by a synaptic gap from welldefined membrane densities in the two postsynaptic processes (asterisks). C: Proteinaceous material in the synaptic gap and within the pre(arrowhead) and postsynaptic (asterisks) processes can be selectively stained by certain histochemical stains (aldehyde-bismuth impregnation). D,E: Synaptic convergence of many presynaptic fibres onto a postsynaptic profile (asterisk), seen here in mushroom body ␣-lobe, is more rarely encountered. A mixture of abundant small electron-lucent synaptic vesicles and rarer electron-dense vesicles is seen in the vicinity of the synaptic contacts (arrowheads). At some synaptic appositions, two presynaptic process containing synaptic bars lie opposite the single postsynaptic element. F: A bar in a neurite belonging to the Fast Extensor Tibiae motorneuron (FETI) neurite marks a putative monadic synapse (arrowhead) on a glial cell. G: A cobalt-sulphide impregnated descending contralateral movement detector (DCMD) neurone lies postsynaptic to the lateral giant movement detector (LGMD) neurone. This micrograph shows several monadic synaptic contacts (arrowheads) between the two neurones, which serial section reconstruction shows to be without presynaptic bars. Two similar synapses involving two presynaptic neurites of unknown origin are also shown (arrows). Scale bars ⫽ 1 ␮m (A, D); 0.1 ␮m (B,C); 0.2 ␮m (E–G).

taining these classes of transmitters. Peptides can often be demonstrated to be co-localised with the fastacting transmitters associated with electron lucent vesicles. For example, the peptide proctolin is present in some Orthopteran glutamatergic motorneurones (Adams and O’Shea, 1983; Bartos et al 1994; Keshisnian and O’Shea, 1985). In the Orthoptera, biogenic amines such as 5-HT (Elekes and Hustert, 1988; Peters et al., 1987; Peters and Tyrer, 1987), dopamine (Elekes et al., 1987), octopamine (Schu¨ rmann et al., 1995), and histamine (Ho¨ rner et al., 1996) are found in terminals that make conventional synapses and that contain both dense core vesicles and electron lucent vesicles (Fig. 3C–E). Similar observations have been made in other insect groups (Na¨ ssel, 1996). The pre-embedding immunoctyochemical methods that have been most commonly used to investigate the ultrastructural distribution of these neurotransmitters do not allow a clear identification of the vesicle populations with which they are associated. The post-embedding immunogold technique has revealed that serotonin is associated with large dense core vesicles in neurosecretory terminals of Rhodnius (Miksys and Orchard, 1994). Immunocytochemical evidence for the presence of neuropeptides such as adipokinetic hormones, FMRFamid, allostatin, and proctolin within dense core vesicles has been obtained from studies in the locust (Diederen et al., 1987, pages 227–236, in this volume) and the cockroach (Ude and Agricola, 1995; Ude and Eckert, 1988). In some classes of axon terminals within the nervous system of the locusts and other Orthopterans (Fig. 1D,E), dense core vesicles appear closely associated with synapse-like active zones (Mancini and Frontali, 1970; Schu¨ rmann, 1972). Morphological evidence for

Fig. 3. A–H: Synaptic electron-lucent and dense core vesicles. A,B: Electron-lucent vesicles immunoreactive for glutamate are round (A) while those immunoreactive for GABA are pleomorphic (B), revealed here in the locust prothoracic ganglion with the postembedding immunogold technique. C–E: In synaptic neuropiles, neuronal profiles immunoreactive for biogenic amines (asterisks) are here revealed using pre-embedding peroxidase immunocytochemistry in the cricket prothoracic ganglion. Processes immunoreactive for octopamine(OA), dopamine- (DA), and 5-HT contain clusters of dense core vesicles. In addition, electron-lucent synaptic vesicles associated with

conventional dyadic synaptic sites (arrows) may be encountered. F,G: Non-synaptic release of dense core vesicles is commonly observed after potassium stimulation and tannic acid infusion followed by chemical fixation. Membrane fusion of dense core vesicles (arrows) is here seen accompanied by a release of vesicle contents into the extracellular space in cricket brain protocerbral neuropile. H: Omegafigures representing dense core vesicle fusion with the cell membrane are only rarely encountered after conventional fixation of stimulated or unstimulated tissue (Cricket mushroom body ␣-lobe). Scale bars ⫽ 0.2 ␮m (A,B,H); 1 ␮m (C–E); 0.5 ␮m (F,G).

SYNAPSES IN THE CNS OF ORTHOPTERAN INSECTS

release of the vesicles at such sites has been found in studies of other invertebrates (Gu¨ nther and Schu¨ rmann, 1973). Putative Sites of Non-Synaptic Chemical Transmission It has been proposed that large diameter dense core vesicles release their contents mainly at sites other than identifiable synaptic active zones (Buma and Roubos, 1986; Golding, 1994; Schu¨ rmann et al., 1991). Following conventional aldehyde fixation, however, exocytotic omega profiles have rarely been seen at extrasynaptic sites in neuropiles where such vesicles are common. Following stimulation of brain tissue with potassium ringer solution containing tannic acid, a dramatic increase in the incidence of non-synaptic omega profiles in these neuropile areas is seen with conventional fixation (Fig. 3F,G). The dimensions of the omega profiles correspond to dense core vesicles that lie free in the cytoplasm (Kerwien and Schu¨ rmann 1990). This suggests a paracrine form of release of neuroactive compounds (volume transmission) in insects, in addition to focal synaptic transmission associated with small agranular vesicles (Vizi and Labos, 1991). Volume transmission has been proposed as a feature the central nervous systems of both vertebrates and invertebrates and is considered to be an early evolutionary development (Golding, 1994). Despite its probable importance, the contribution and functional significance of non-synaptic release of neuroactive compounds from vesicles are largely unknown. Electrical Synapses In the fly central nervous system, both physiological and ultrastructural evidence has been obtained for electrotonic coupling between identified neurones (Bacon and Strausfeld, 1986; Strausfeld and Bassemir, 1983; Tanouye and Wyman, 1980). Mixed synapses with both the ultrastructural features of chemical and electrotonic connections have also been described (Blagburn et al., 1999). Though less common, electrotonic connections are present between neurones (Siegler, 1982) and between glial cells in the Orthoptera and in other insect groups. Ultrastructural studies of a connection between two visual interneurones in the locust brain (the LGMD and DCMD) have demonstrated that while it is based on thousands of chemical synapses, there are in addition regions of gap-junction like contacts between the two neurones. At these sites, which are typically about 0.1 nm in length, the membranes of the two neurones lie only 2 nm apart (Killmann and Schu¨ rmann, 1985). The synapse between the two neurones is a rectifying connection with a high safety factor, but whether its electrophysiological properties justify the conclusion that it has a mixed chemical-electrical nature, is still a matter of debate (Rind, 1984; Rowell and O’Shea, 1980). Dye coupling between adult and developing insect neurones has been reported for other insects groups (Jacobs et al., 2000; Phelan et al., 1996; Strausfeld and Bassemir, 1983) and has also been observed in the cricket and locust nervous system (Schu¨ rmann, personal observation). In Drosophila, studies of electrical connections involving giant fibres have lead to the discovery of a new family of gap junction proteins called

215

innexins, that are structurally distinct from the vertebrate connexins (Phelan et al., 1998). Studies of embryonic gap junctions in the nervous system of Schistocerca have revealed two novel members of the innexin family, G-Inx (1) and G-Inx (2), in neuroblasts and glia that are orthologous to Drosophila innexins (Ganfornina et al., 1999). G-Inx (1) appears to be expressed predominantly in glial elements of the abdominal ganglia. The distribution and dynamic properties of electrical synapses and their contribution to circuitry of the adult locust central nervous system, however, have yet to be determined. SYNAPTIC CIRCUITRY Brain Neuropiles and Neurons In some prominent regions of the insect brain (e.g., the mushroom bodies, central body complex, and antennal and optic lobes) long known from numerous histological studies, the neuropile is organised into columns, strata, and glomeruli with repeated functional units and with a high degree of obvious geometrical order, which considerably aided analysis of their synaptic circuitry (Leise, 1991; Strausfeld, 1976). Surrounding neuropiles display a structural organisation that is more comparable to that of the ventral cord. Differences in the architecture of these neuropiles can be demonstrated by immunocyto-chemical studies using antibodies against Drosophila synapsin I-protein (Klagges et al., 1996) in insect species and in other invertebrates (Fabian-Fine et al., 1999; Schu¨ rmann, unpublished observations). The distribution of Drosophila synapsin I-like immunoreactivity is correlated with clustered synaptic vesicles at synaptic sites and is, therefore, useful in pinpointing synaptic loci in some neuropiles at the light microscopical level (Fig. 4). In neuropiles appearing less geometrically ordered, but that are nevertheless highly structured (the “diffuse neuropiles” of the ventral nerve cord and brain), high levels of synapsin I-like immunoreactivity are quite uniformly distributed. This is in contrast to the neuropiles looking more geometrically structured where the labelling is concentrated in regions that other studies (Schu¨ rmann, 1987) have revealed to contain high densities of synaptic vesicles and synaptic sites (Fig. 4D,E). Within discrete regions of neuropile containing repeated neuronal units, unusually structured synaptic complexes that comprise elements of microcircuitry (divergent, convergent, serial, reciprocal synaptic coupling) are found in both the vertebrate and invertebrate nervous systems (Figs. 1A–E, 4E, 5, 6). The concept of microcircuitry within the insect nervous system is well established (Pearson, 1979) and has received support by the finding of co-distribution of transmitters and neuromodulators in discrete neuropile compartments (see Homberg, pages 189 –209, in this issue). Examples of synaptic complexes in Orthopterans studied using electron microscopy, with or without the aid of intracellular neuronal labelling techniques, are found in the mushroom bodies (Iwasaki et al., 1999; Leitch and Laurent, 1996; Mancini and Frontali, 1970; Schu¨ rmann, 1974, 1987; Strausfeld and Li, 1999; Trujillo-Cenoz and Melamed, 1962), and in the antennal lobe glomeruli (Boeckh et al., 1970, 1989; Distler, 1990; Distler and Boeckh, 1996). Studies of the optic lobes of

Fig. 4. A–D: Synapsin I-C3 immunocytochemical staining differs in distribution and intensity in different brain neuropilar compartments of the locust (A) and cricket brain (B–D) seen in laser confocal micrographs (A,D) and in conventional fluorescence micrographs (B,C) of vibratome sections. The distribution of immunoreactivity corresponds to vesicle clusters at synaptic sites. A: Note differential staining of strata in the loptic lobe neuropiles of the lamina (LA), medulla (ME), and lobula (LO) and immuno-negative parts in the chiasmata (asterisks) and lines perpendicular to strata in the medulla. B,C: Mushroom body compartments of the calyces (CA), pedun-

cle (PE), and ␣-lobe (␣) and the central body part (CB) appear less intensely stained than surrounding neuropiles (asterisk); ocellar neuropiles (triangles) and bridge (BR) are strongly stained. Fluorescence in the neurolemma (star) is non-specific. (Fluorescence micrograph of a horizontal vibratome section through the cricket brain.) D: The pattern of stained boutons in the mushroom body calyx (asterisk) of the cricket is seen in this laser confocal micrograph. E: Similar presynaptic boutons (asterisks) in the same calycal neuropile are seen in the electron micrograph. Scale bars ⫽ 100 ␮m (A); 50 ␮m (B,C); 10 ␮m (D); 1 ␮m (E).

SYNAPSES IN THE CNS OF ORTHOPTERAN INSECTS

Fig. 5.

See figure legend on next page.

217

218

¨ RMANN A.H.D. WATSON AND F.-W. SCHU

Orthopterans are relatively rare (O’Carrol et al., 1992). A recent paper describes a special convergent wiring of cholinergic visual interneurons in the locust brain (Rind and Leitinger, 2000; Rind, to appear in Part II of this special issue). In comparison, by far the most intensely investigated insect brain neuropiles are the optic lobes of flies (see Meinertzhagen, 1993). A deeper understanding of the synaptic circuitry within less geometrically ordered neuropiles is dependent on intracellular staining methods that reveal the structure of single or small numbers of physiologically or morphologically characterised neurones. When carried out in combination with immuno-electron microscopy of neuroactive compounds, this has proved a powerful tool for the structural analysis of synaptic interactions. While the small number of neurones revealed in any one preparation may be seen as a disadvantage, this can be outweighed by the possibility of correlating the physiological properties of a neurone with a knowledge of the nature and distribution of its synaptic interactions. While such methods have been quite widely applied to neurons in the locust ventral nerve cord (see below), only a few ultrastructural studies have applied these labour intensive techniques to neurones in the brain of the locust and other Orthopterans. Furthermore, the studies so far carried out either aim to describe the nature, distribution, and numbers of synaptic connections in small regions of a neurone’s arborisation or to understand the synaptic wiring of several units in a functional context. No inspection of a single neuron at the ultrastructural level has yet delivered a complete analysis of the range and numbers of input- and output synapses for identified neurons. However, the demonstration of the entire synaptic compliment of single neurons may soon become possible through the refinement of methods of confocal imaging used in combination with immunocytochemistry. The total number of synapses of single identified neurones has so far only been estimated from selected samples of serial sections or from serial ultrathin sections of restricted parts of their central arborisations. One neurone for which this has been done is the large visual intersegmental interneurone, the descending contralateral movement detector (DCMD). The arborisation of this neurone within

Fig. 5. A: Synaptic interaction sites of the metathoracic fast extensor tibiae motorneurone. B: A FETI neurite, labelled with horseradish peroxidase, makes a monadic input synapse (arrowhead, main panel and inset) onto an unlabelled neurite that in turn makes a dyadic synapse (arrow) . One postsynaptic element of the dyadic synapse is the FETI neurite itself, while the other is a thin sheet of glia. The FETI neurite also receives a monadic input (arrow) from another unlabelled neurite. C: Both the FETI and a flexor tibiae motorneurone (asterisk) have been labelled by intracellular injection with horseradish peroxidase. The FETI makes a synapse (arrowhead) onto the flexor tibiae neurone. At the same site, an unlabelled neurite makes a dyadic synapse (arrowhead) onto both the FETI and the flexor motorneurone. D: I–V: Five sections through a complex set of synapses (arrows) linking the FETI and flexor tibiae motorneurone. The stars mark monadic or dyadic synapses from the FETI onto the flexor, and the triangles dyadic synapses at which both the FETI and flexor are postsynaptic. The FETI and flexor are also linked by serial synapses through neurite A and C. There are feedback circuits from the FETI onto itself through neurites A, C, and D. Neurite A also synapses with C, and C with D. Scale bars ⫽ 50 ␮m (A), 0.5 ␮m (B,C), 1 ␮m (D).

the locust brain receives only input synapses that are of two morphologically distinct types and are most abundant on very small diameter branches (Killmann et al., 1999). A total of 8,500 input synapses of different types has been estimated for the whole dendritic tree, among which at least 2,250 contacts arise from the presynaptic lateral giant movement detector (LGMD) neuron. For other identified neurones, estimated total numbers of synaptic contacts vary widely (Killmann et al. 1999; Littlewood and Simmons, 1992; Simmons and Littlewood, 1989). Thus, estimates of the numbers of synaptic contacts that comprise a single physiological connection between a pair of neurones also vary considerably depending on the neurones concerned. For some neurones, the number of contacts constituting a synapse may be as low as a few tens (Meinertzhagen, 1989) but more often this lies in the range of several hundreds if not thousands of synaptic contacts (Burrows et al., 1989; Simmons and Littlewood, 1989; Watson and Burrows, 1983). Synaptic Distribution and Circuitry of Single Identified Neurones of the Ventral Nerve Cord In the locust ventral nerve cord, representatives of each of the major classes of neurone present have been examined ultrastructurally following intracellular staining with horseradish peroxidase. This has allowed the structure and distribution of their synapses to be described and related to their general physiological properties, and in some cases to their role in the circuitry to which they belong. Comparisons can also be made between neurones within each class and between neurones in different classes. Sensory neurones. The synaptic relationships of a broad range of locust sensory afferent neurones have been investigated. Some supply receptors on the body surface that have a purely exteroceptive role (e.g., prosternal filiform and cercal hairs; Blagburn and Beadle, 1984; Blagburn et al., 1985; Watson, 1990; Watson and Pflu¨ ger, 1994), while other surface receptors have a largely proprioceptive role (campaniform sensilla and hair plates; Watson and England, 1991; Watson et al., 1991). The sensory terminals of proprioceptive chordotonal organs that monitor joint movement and position (Altman and Tyrer, 1980; Watson et al., 1993) and of auditory afferent terminals in the cricket (Hardt and Watson, 1999) have also been examined. The central terminals of sensory afferent neurones project somatotopically to modality specific regions of sensory neuropile (Newland, 1991). Most sensory neurones appear to be cholinergic and their terminals, containing round agranular vesicles, make predominantly dyadic output synapses, though there is some evidence that a small proportion may be serotonergic (Lutz and Tyrer, 1988; Parker and Newland, 1995). The central terminals of all of the afferent classes examined receive synaptic input from neuropilar processes that are immunoreactive for GABA (Hardt and Watson, 1999; Watson, 1992; Watson and Pflu¨ ger, 1994). Many also receive input from glutamate-immunoreactive processes though the proportion of these inputs varies widely (between 5–50%) depending on the type of afferent neurone (Hardt and Watson, 1999; Watson et al., 1991; Watson and Pflu¨ ger, 1984). The release of GABA results in an increase in chloride conductance that induces a presyn-

SYNAPSES IN THE CNS OF ORTHOPTERAN INSECTS

Fig. 6.

See figure legend on next page.

219

220

¨ RMANN A.H.D. WATSON AND F.-W. SCHU

aptic depolarisation of the afferent terminal. This inhibits transmitter release by shunting the currents induced by the action potential as it invades the terminal (Burrows and Laurent, 1993; Hue and Callec, 1983; Watson, 1992). The effects of the glutamatergic inputs are unknown. The only known central glutamatergic synapse in the locust nervous system is one between the fast extensor tibiae and flexor tibiae motorneurones (Parker, 1994; Sombati and Hoyle, 1984). Here glutamate induces a large cation-mediated depolarisation. In the crustacean central nervous system, however, central glutamatergic synapses often induce an increase in chloride conductance (Clarac and Cattaert, 1996; Pearlstein et al., 1998). The input synapses received by the locust sensory afferent terminals are dyadic and where only a single afferent has been labelled, it contributes only one process to the postsynaptic dyad. Where several afferents from a single hair plate were labelled in the same preparation, however, many synapses were seen at which both postsynaptic elements were afferent terminals (Watson et al., 1991). Only two types of hair plate afferent have been identified on physiological criteria, and this configuration would be consistent with many or all of the afferents being gated as a sensory unit. Sensory neurones may also be paired postsynaptically with other classes of neurone. For example, synapses at which a sensory afferent terminal and a sensory interneurone with which it makes synaptic contact elsewhere, form the two postsynaptic elements at a dyadic synapse have been observed (Watson and Pflu¨ ger, unpublished data). Efferent Neurones. Locust muscle is supplied by inhibitory and excitatory motorneurones that induce inhibitory or excitatory postsynaptic potentials at the muscle membrane, and dorsal unpaired median (DUM) modulatory neurones that alter the rate of contraction but have no effect on membrane potential. Each of these classes of neurone has been studied in the electron microscope. The inhibitory neurones are GABAergic and there are three pairs in each thoracic ganglion, each neurone supplying many muscles (Watson, 1986). The common inhibitor neurone of the metathoracic ganglion supplies 13 muscles in the proximal part of the leg (Hale and Burrows, 1985) but its central

Fig. 6. A–E: Synapses of a local spiking interneurone from the metathoracic ganglion of the locust that has been labelled by intracellular injection with horseradish peroxidase. A,B: Reconstructions from serial sections through ventral (A) and dorsal (B) neurites. These neurones have two distinct fields of branches, one ventral and one dorsal, linked by a short axon like neurite. The ventral branches are in contact with the terminals of sensory neurones supplying the hind leg. Fine ventral branches are predominantly sites of synaptic input (arrowheads), though larger diameter ventral branches also make output synapses (circles). A typical varicosity from a dorsal branch is shown in B. Though this makes several output synapses (arrowheads), it also receives some synaptic input (circles). C: Section through the varicosity in B, in which two output synapses (arrowheads) are visible. D: A fine ventral branch receives an input synapse (arrowhead). E: The larger diameter ventral branch makes a divergent output synapse (arrowhead). F,G: A neurite from a non-spiking local interneurone in the metathoracic ganglion of the locust, labelled by intracellular injection with horseradish peroxidase. Inset (F): Several output synapses made from the region marked by the box in F. G: A higher magnification image of an output synapse from the same neurone shown in F. Scale bars ⫽ 2 ␮m (F), 0.5 ␮m (inset in F, G).

branches contain no synaptic vesicles and it makes no central output synapses (Watson et al., 1985). It receives abundant dyadic input synapses to which it contributes one of the pair of postsynaptic neurites. The excitatory motor neurones, which are glutamatergic (Bicker et al., 1988; Emson et al., 1974; Watson, 1988; Watson and Seymour-Laurent, 1993), fall into two groups on the basis of their central synaptic connections. Some, such as the slow extensor tibiae and flexor tibiae motorneurones, receive many dyadic input synapses, but make no central output synapses (Watson and Burrows, 1982). Others, such as the tergosternal flight motorneurones and fast extensor tibiae motorneurone (FETi), also make numerous central output synapses (Burrows et al., 1989; Watson, 1984). In most neurones, tracing their neurites over distances of 10 –20 ␮m reveals only simple, one-way synaptic connections with neighbouring neurites (e.g., tergosternal flight motorneurones). The FETi, however, is unique among the arthropod neurones so far studied by serial section reconstruction, in the complexity of the local circuitry in which it participates (Burrows et al., 1989; Watson and Burrows, 1982). Reconstructions of short segments of neurite, or even single ultrathin sections, reveal a complex circuitry involving direct, indirect, and reciprocal connections with other neurones (Fig. 5). The FETi has a dense arborisation of very fine neurites mostly arising directly from the primary neurite, which is quite unlike the sparser more tree-like branching patterns of other motor neurones or interneurones. This may be a consequence of the need to make such complex connections. A considerable proportion of the synapses received by the FETi are monadic (i.e., it is the only postsynaptic element), and this may also reflect the difficulty of establishing and weighting the synapses required for such intricate circuitry within confined volumes of neuropile. At some of the dyadic synapses to which the FETi contributes the presynaptic component, one of the other neurites involved also contains a presynaptic bar (Fig. 5). The possible significance of these arrangements only becomes apparent if some of the participating neurones can be identified. For example, during the motor programme driving the locust jump, two antagonistic muscles (an extensor and flexor of the tibia) contract at the same time. This is, in part, brought about by a direct synaptic connection between the single fast extensor motor neurone and a number of flexor tibiae motorneurones (Burrows et al., 1989; Hoyle and Burrows, 1973). Direct synaptic connections between the fast extensor and a flexor tibiae motor neurones can be demonstrated when both are labelled by intracellular dye injection (Burrows et al., 1989). Figure 5 shows an example of one of these synapses at which the second, unlabelled postsynaptic process also contains a presynaptic bar. Even where only a conventional dyadic synapse exists between them, a common input onto both occurs within 1 ␮m in about 85% of the synapses (Fig. 5), thus repeating the pattern in a less rigorous configuration. The identity of the unlabelled third element is unknown but several candidates can be suggested on the basis of known physiological properties of the extensor/flexor connection. An action potential in the extensor tibiae motorneurone causes an excitatory postsynaptic potential or an increase in fir-

SYNAPSES IN THE CNS OF ORTHOPTERAN INSECTS

ing frequency of the flexor motorneurones (Burrows et al., 1989). In addition, a common drive onto both the extensor and the flexors motorneurones is present that could (though it need not) arise from the unlabelled presynaptic element of the triad. Two identified sources of common input onto the fast extensor and flexor motorneurones are known to be active during the preparatory phase of the jump motor programme. The first comes from a descending mesothoracic interneurone (Pearson and Robertson, 1981) and the second from sensory neurones supplying two campaniform sensilla (cuticular strain detectors) on the proximal tibia of the hind leg (Burrows and Pflu¨ ger, 1988). Another possible role for the unlabelled presynaptic element would be one of modulation of the connection between the extensor and the flexor motorneurone. The static position of the tibia at the time of an action potential in the extensor motorneurone can have a marked effect on the amplitude of the excitatory postsynaptic potential generated in the flexor motorneurone (Heitler and Burrows, 1977a,b). This effect probably arises from sensory input from the femoral chordotonal organ, which monitors the movement and position of the femoro-tibial joint. Because changes in the membrane potential and spike amplitude of the fast extensor motorneurone induced by alterations in joint angle are very small, these could only produce the observed postsynaptic effect if the synapses mediating them were located near the synaptic connection between the fast extensor motorneurone and flexor motorneurones (Jellema et al., 1997). The observed arrangement of synapses would precisely fulfill this condition. The intricate interconnections of the fast extensor motorneurone revealed by physiological studies, therefore, suggest not just one, but at least three possible roles for the synaptic configurations observed. The third class of efferent neurone studied in the thoracic ganglia of the locust are octopaminergic dorsal unpaired median (DUM) neurones. At the neuromuscular junction, the axon terminals of these neurones contain numerous large granular vesicles (Hoyle et al., 1974) but though such vesicles are present in small numbers in the bifurcating primary neurites, they are absent from the fine neuropilar branches within the neuropile of the thoracic and abdominal ganglia (Pflu¨ ger and Watson, 1995; Watson, 1984). Input synapses received by the DUM neurones are generally dyadic. In preparations where more than one DUM neurone has been labelled, dyadic synapses to which two different DUM neurones each contribute one of the postsynaptic elements can be found (Watson, 1984). Recordings from pairs of DUM neurones in the same ganglion often show a high proportion of common postsynaptic potentials (Duch et al., 1999; Hoyle and Dagan, 1978). Structures that have features of output synapses are only very rarely seen on the fine neuropilar branches of DUM neurones and are associated predominantly with agranular vesicles. Though there is some evidence that octopamine may have central effects, the virtual absence of vesicles, granular or agranular, in these branches suggests that even nonsynaptic release is unlikely in the population examined. In the primary neurites, well-developed presynaptic bars are present. However, these are associated with few, if any, vesicles and the postsynaptic pro-

221

cesses is most commonly a glial cell (Watson, 1984). The physiological significance of these is unknown. Most efferent neurones, regardless of type, make few if any central output synapses. However, those that do may be involved in very complex local circuitry. The extensor tibiae muscle of the locust hind leg is supplied by only two excitatory motorneurones, the FETi whose synaptic properties have just been described, and the slow extensor tibiae motorneurone (SETi) that makes no central output synapses. The FETi is active only in the explosive tibial extension of jumping and kicking, and the SETi only in walking and other slow movements. This distinctiveness in terms of behaviour and circuitry emphasises the uniqueness of individual locust neurones even among those that belong to the same class and contact the same postsynaptic structure. Local interneurones. Local interneurones (Fig. 6) have all of their neurites confined within a single ganglion or neuromere or even a particular region within a ganglion. Functionally, they fall into two broad categories, those that produce action potentials and those that do not. Most spiking local interneurones so far studied in the Orthopteran nervous have one set of branches in a sensory neuropile where they receive direct input from sensory afferent neurones (Burrows, 1996). A short unbranching, axon-like neurite may link this field with a second field of branches that lie in a different region of the neuropile and that are predominantly sites of synaptic output. The branches of the input field are often of small diameter and relatively smooth, while those of the output field may be more varicose than those of the input field (Watson and Burrows, 1985). It has been suggested that on the basis of branch morphology alone, the polarity of the neurone and the distribution of input and output synapses can be inferred (Ro¨ mer and Marquart, 1984; Strausfeld, 1976; Tyrer and Altman, 1974). However, as will be seen in this and other sections of this paper, ultrastructural studies demonstrate that morphology alone is an unreliable guide to synapse distribution. For example, in a population of metathoracic spiking local interneurones receiving sensory input from the leg, about a third of the synapses on the dorsal field of branches that were contacted by sensory afferent neurones were output synapses (Watson and Burrows, 1985). These were particularly concentrated on the larger dorsal neurites, suggesting a degree of compartmentalisation even within a single field of branches. In cricket Omega neurones, however, the input synapses on the branches that receive contacts from auditory afferents outnumber outputs by 10:1 (Watson and Hardt, 1996). In both cases, the identity of the postsynaptic neurones is unknown. The output fields of these two types of interneurone also receive presynaptic input synapses. For the locust metathoracic spiking interneurones, these were outnumbered by the outputs by 4.5:1 and for the Omega neurone by 10:1. Nevertheless, small numbers of input synapses could effectively control the overall output of these neurones. The synaptic morphology and distribution of nonspiking local interneurones (Fig. 6G) has been studied in both the locust and crayfish ventral nerve cord and the locust brain (Kondoh and Hisada, 1986a,b; Littlewood and Simmons, 1992; Simmons and Littlewood,

222

¨ RMANN A.H.D. WATSON AND F.-W. SCHU

1989; Watson and Burrows, 1988). Many populations of non-spiking local interneurones appear to have only a single arborisation within the neuropile, and on these branches input and output synapses may be intermingled though inputs are more abundant on smaller diameter branches and outputs on larger ones (Kondoh and Hisada, 1986b). A similar pattern is seen in some local non-spiking interneurones in the locust, while in others, small diameter branches make predominantly output synapses (Watson and Burrows, 1988). One population of crayfish non-spiking interneurones has two distinct arborisations, one of which is composed of smooth branches that received predominantly input synapses, the other being more varicose with intermingled inputs and outputs (Kondo and Hisada, 1986a). There are, therefore, a variety of patterns of synaptic distribution possible on the branches of these small neurones. Because these neurones function without action potentials, a crucial factor for the way in which they operate is how efficiently potentials can spread through the arborisation. The crayfish neurone with separate input and output arborisations presents one extreme form of organisation. Here the two arborisations are linked by a very large diameter and presumably a low resistance neurite to ensure the efficient spread of potentials between the two fields. In neurones where input and output synapses are intermingled, however, an output synapse will be most strongly affected by the inputs nearest to it. If input synapses from single or functionally equivalent presynaptic and postsynaptic neurones are clustered on particular branches, the interneurone might be functionally compartmentalised. The compartments would not be exclusive, rather inputs from different sources would constitute a series of overlapping spheres of influence having strong effects on nearby output synapses and progressively weaker influences on more distant output sites. This scenario has not been specifically tested, but some of the known physiology of non-spiking interneurones would lend themselves to such an interpretation (Burrows,1996; Laurent and Burrows, 1989). Non-spiking neurones can be demonstrated to release neurotransmitter tonically at membrane potentials at which they are typically encountered (Burrows and Siegler, 1978). In the ventral nerve cord of the locust (Watson and Burrows, 1988) and the crayfish (Kondoh and Hisada, 1986a,b), the morphology of the output synapses made by non-spiking interneurones resembles that of synapses seen in spiking interneurones. They are typically dyadic, with a presynaptic bar surrounded by populations of vesicles that are not noticeably more abundant than those in spiking neurones. A different situation prevails, however, in a population of locust brain ocellar L-neurones, which each make both tonic excitatory output synapses onto other L-neurones and phasic inhibitory reciprocal output synapses with each other (Simmons and Littlewood, 1989; Littlewood and Simmons, 1992). Each tonic synapse is composed of some hundreds of dyadic associations at which a presynaptic bar is surrounded by small agranular vesicles. These synapses are characterised by numerous coated omega profiles that may represent sites of membrane re-uptake. By contrast, though the phasic synapses are in many ways similar, the number of dyadic associations and the number of omega profiles

at each is lower, perhaps reflecting lower levels of activity. Intersegmental Interneurones. Intersegmental neurones are spiking neurones whose axons link one ganglion or neuromere to another. They may ascend or descend between two adjacent ganglia or run for considerable distances along the nerve cord arborizing in several ganglia and within the brain (Killmann et al., 1999; see Heinrich, to appear in Part II of this special issue). At the light microscope level, examination of the arborisations of these neurones may reveal some fields that are composed of relatively smooth fine branches and others that are composed of varicose branches. These have often been interpreted as being sites of input and output, respectively, but as with the local interneurones, ultrastructural examination usually reveals a more complex pattern that suggests that the neurone may have a variety of functions or modes of operation in different ganglia. Most studies of the synaptic distribution of neurones of this type have examined the arborisations in a single ganglion only, usually the one that contains the cell body of the neurone. One such study is of a locust ascending mesothoracic interneurone that plays a role in flight initiation (Watson and Burrows, 1983). This neurone contains predominantly round agranular vesicles and is probably glutamatergic (Watson and Seymour-Laurent, 1993) and in the mesothoracic ganglion, input and output synapses are present in approximately equal numbers and are intermingled on all branches. In the cricket prothoracic ganglion, the synaptic distribution of two ascending interneurones, AN1 and AN2, both of which carry auditory information to the brain, have been compared (Hardt and Watson, 1994). The two neurones have quite similar branching patterns in the prothoracic ganglion, but whereas AN1 receives only input synapses, AN2 also makes outputs. Both neurones have many fine branches in the ventral medial auditory neuropile where they receive monosynaptic input from auditory afferents. However, the outputs from AN2 are made only from major branches in this neuropile that are of considerably larger diameter than the equivalent ones in AN1. This pattern of synaptic distribution bears some similarity to that of the input field of the local spiking interneurones though the separation between the fine branches as sites of input and the large diameter ones as sites of output is more extreme in AN2. An intersegmental auditory interneurone in the locust mesothoracic ganglion shows some branches that are input only, while others have both inputs and outputs, with the latter predominating (Peters et al., 1986). The input branches are smooth while those with output are varicose, though as the authors point out, varicosities may be more reliable indicators of clustered mitochondria than synapse sites. AN1 is not the only intersegmental interneurone that has been found to receive only input in its ganglion or region of origin. The same is true of dendrites of the DCMD neurone that lie within the locust brain (Killmann et al., 1999) and of a population of descending intersegmental interneurones that coordinate local locomotor reflexes involving the mesoand metathoracic legs (Watson and Laurent, 1990). This latter population contains both GABAergic and non-GABAergic neurones, which receive direct excita-

SYNAPSES IN THE CNS OF ORTHOPTERAN INSECTS

tory input from leg mechanoreceptors in the mesothoracic ganglion and inhibitory inputs from local mesothoracic interneurones. The most complete study so far of synaptic distribution on an Orthopteran intersegmental interneurone has been carried out on the neurone A4I1 (Watson and Pflu¨ ger, 1989). This neurone receives its major input from fields of wind sensitive hairs on the prosternum and head of the locust that terminate in the ventral neuropile of the prothoracic ganglion. The cell body, however, lies in the fourth abdominal ganglion and its axon projects to the brain, arborizing in all of the ganglia in between. Action potentials can be initiated from both the prothoracic and the fourth abdominal ganglia. In the thoracic ganglia, the neurone has fine varicose ventral neurites and much thicker dorsal neurites. In the prothoracic ganglion, the ventral branches receive only input synapses and the dorsal branches make only output. A similar trend is seen in the meso- and metathoracic ganglia, though here some outputs are also made from the ventral branches and some inputs received by the dorsal ones. In the fourth abdominal ganglion, the neurone receives synaptic input only and rostrally running action potentials can be elicited from this ganglion following wind stimulation of the cerci. The effect of these would, however, normally be swamped by the greater sensitivity of the prosternal sensory input that sends action potentials both rostrally and caudally. Physiological studies so far have not revealed the role of the abundant output synapses on the dorsal prothoracic branches though the neurone is known to make direct connections onto mesothoracic flight motorneurones and indirect connections with metathoracic flight steering motorneurones. CONCLUSIONS In insects, synaptic ultrastructure and synaptic circuitry have been most intensely studied in two groups, the Diptera and the Orthoptera. The fruitfly Drosophila, has proved an important model organism for investigations of synapse structure due to its suitability for combined genetic, molecular biological and morphological approaches. Its major disadvantage is the limited accessibility of its neuropiles for combined structural and electrophysiological studies on neuronal connectivity though this has proved possible in other Dipteran species. On the other hand, the studies of identified neurons of locusts and other Orthopterans have delivered by far the most detailed knowledge of physiology in relation to their roles in simple and complex behaviours (Burrows, 1996). The molecular components of the Orthopteran synaptic apparatus have, however, not yet been investigated in detail. The study of Orthopteran synapses is likely to profit directly from the knowledge that has been acquired of the molecules associated with Drosophila synapses and from antibodies raised against them. Further investigations of the functional morphology and role of mature synapses of identified locust neurons should be directed towards three important problems: (1) Understanding the subcellular topography of the synaptic complexes, (2) Investigating the principles of synaptic microcircuitry, (3) Unravelling the pharmacology of identified central synaptic interactions, which has so far received scant attention.

223

Electron microscopy will remain indispensable for synaptic studies, as it is still the only direct means of unequivocally demonstrating sites of synaptic contact and for localising neurotransmitters and other synapse-associated molecules to the synaptic site. The knowledge of synaptic distribution that is required for the detailed modelling of neuronal interactions is dependent on laborious quantitative electron microscopy. These will most effectively be directed at identified Orthopteran neurones about whose role in central circuitry a wealth of information is already available. Quantitative computer-aided electron microscopy has demonstrated the existence of two types of synapse in a locust neuron, similar to type 1 and type 2 synapses in vertebrates, which would otherwise be hard to distinguish. If the physiological significance of these differences can be determined, this approach may greatly strengthen the power of ultrastructural analysis. Studies of intracellularly labelled, physiologically characterised neurones have revealed that though dyadic synapses are the most prevalent type of contact between insect neurones, a single neuron may express a mixture of dyadic, monadic, bar-containing, or bar-free synaptic active zones and of subsynaptic densities of variable extent. The functional significance of bars and of the structural variability seen at synapses is still not understood, though some of the implications of dyadic synaptic configurations can be derived from investigations of connections between pairs of identified neurons. Whereas small electron lucent vesicles in synaptic terminals are assumed to be released from active zone, the sites and consequences of dense core vesicle release remain unclear even though dense core vesicles are found in the majority of nerve fibre profiles. Neuropeptides outnumber classical transmitters by far (see Homberg, pages 189 –209, in this issue) and their distribution is concentrated in some brain areas, among them the central body complex. However, in Orthopteran central nervous systems, as well as in other insects and invertebrates, the connectivity and functions of peptidergic interneurons remain largely uninvestigated. These neurons deserve intensified study, as they contribute significantly to local networks. In addition, despite extensive immunocytochemical studies of the Orthopteran nervous system using antibodies raised against most classical transmitters and many insect peptides, the nature of the neurotransmitters used by many populations of interneurones remains completely unknown and until this information is available, a complete understanding of the principles underlying insect neuronal circuitry will not be possible. Despite the attention that has been devoted to investigations of the Orthopteran central nervous systems, our knowledge of their neural circuits is still fragmentary. Numerous light microscopical studies have provided a useful framework for the understanding of the different neuropilar compartments of the insect brain, revealing the major classes of nerve cells present and many common underlying architectural principles. Inevitably, these studies have been directed at circuitry that is particularly amenable for study and consequently much remains to be discovered. For the fly brain, a considerable amount of valuable data has been

¨ RMANN A.H.D. WATSON AND F.-W. SCHU

224

obtained from studies of the optic lobe neuropiles (Meinertzhagen 1989; Strausfeld, 1976), while in Orthopteran brains the antennal lobes and mushroom bodies have been fertile ground for studies of olfactory information processing, learning, and memory. A deeper knowledge of synaptic structure and principles of circuitry, and their structural and functional dynamics, will be gained by further intense studies from these highly organised neuropiles, whereas the investigations of single identified neurons of all types are indispensable for a more profound understanding the design principles of circuitry expressed in less structurally organised brain neuropiles and the ventral cord. ACKNOWLEDGMENTS Thanks are due to Drs. W. Buchner and A. Hofbauer/ Wu¨ rzburg, for kindly providing antibodies against synapsin I, and Dr. H. Agricola/Jena, for antibodies against octopamine. We thank Ms. M. Knierim and M. Winkler/Go¨ ttingen for skillful technical assistance, Dr. H. Gras/Go¨ ttingen for computer help, and Ms. I. Frambach/Go¨ ttingen for synapsin-ir pictures. This work was supported by DFG-SFB 406 Go¨ ttingen (to F.W.S.) and Wellcome Trust grant 052313/Z/97 (to A.H.D.W). REFERENCES Adams ME, O’Shea M. 1983. Peptide cotransmitter at a neuromuscular junction. Science 221:286 –289. Altman JS, Shaw MK, Tyrer NM. 1980. Input synapses on to a locust sensory neurone revealed by cobalt-electron microscopy. Brain Res 189:245–250. Bacon JP, Strausfeld NJ. 1986. The Dipteran “giant fibre“ pathway: neurons and signals. J Comp Physiol 158:529 –548. Bartos M, Allgauer C, Eckert M, Honegger HW. 1994. The antennal motor system of crickets: proctolin in slow and fast motoneurons as revealed by double labelling. Eur J Neurosci 6:825– 836. Bicker G, Schafer S, Ottersen OP, Storm-Mathisen J. 1988. Glutamate-like immunoreactivity in identified neuronal populations of insect nervous systems. J Neurosci 8:2108 –2122. Blagburn JM, Beadle DJ. 1984. Synapses between an identified giant interneurone and a filiform hair sensory neurone in the terminal ganglion of first instar cockroaches (Periplaneta americana L.). J Exp Biol 113:477– 481. Blagburn JM, Beadle DJ, Sattelle DB. 1985. Development of synapses between identified sensory neurones and giant interneurones in the cockroach Periplaneta americana. J Embryol Exp Morphol 86:227– 246. Blagburn JM, Alexopoulos H, Davis JA, Bacon JP. 1999. Null mutation in shaking-B eliminates electrical, but not chemical, synapses in the Drosophila giant fibre system: a structural study. J Comp Neurol 404:449 – 458. Blenau W, Schmidt M, Schu¨ rmann F-W. 1999. Neurons with dopamine-like immunoreactivity target mushroom body Kenyon cell somata in the brain of some hymenopteran insects. Int J Insect Morphol Embryol 28:203–210. Boeckh J, Sandri C, Akert K. 1970. Sensorische Einga¨ nge und synaptische Verbindungen im Zentralnervensystem von Insekten. Z Zellforsch 103:429 – 446 Boeckh J, Ernst KD, Selsam. 1989. Double labelling reveals monosynaptic connection between antennal receptor cells and identified local interneurons of the deutocerebrum in the American cockroach. Zool J Anat 119:303–312. Budnik V. 1996. Synapse maturation and structural plasticity at Drosophila neuromuscular junctions. Curr Opin Neurobiol 6:558 – 867. Buma P, Roubos EW. 1986. Ultrastructural demonstration of nonsynaptic release sites in the central nervous system of the snail Lymnaea stagnalis, the insect Periplaneta americana, and the rat. Neuroscience 17:867– 879. Burkhardt W, Braitenberg V. 1976. Some peculiar synaptic complexes in the first visual ganglion of the fly, Musca domestica. Cell Tissue Res 173:287–308. Burrows M. 1996. The neurobiology of an insect brain. New York: Oxford University Press.

Burrows M, Laurent G. 1993. Synaptic potentials in the central terminals of locust proprioceptive afferents generated by other afferents from the same sense organ. J Neurosci 13:808 – 819. Burrows M, Pflu¨ ger HJ. 1988. Positive feedback loops from proproceptors invovled in leg movements of the locust. J Comp Physiol [A] 163:437– 446. Burrows M, Siegler MV. 1978. Graded synaptic transmission between local interneurones and motor neurones in the metathoracic ganglion of the locust. J Physiol (Lond) 285:231–255. Burrows M, Watson AHD, Brunn DE. 1989. Physiological and ultrastructural characterization of a central synaptic connection between identified motor neurons in the locust. Eur J Neurosci 1:111– 126. Clarac F, Cattaert D. 1996. Invertebrate presynaptic inhibition and motor control. Exp Brain Res 112:163–180. Davis GW, Goodman CS. 1998. Gentic analysis of synaptic development and plasticity: homeostatic rgeulation of synaptic efficacy. Curr Opin Neurobiol 8:149 –156. Diederen JHB, Maas HA, Pel HJ, Schooneveld H, Jansen WF, Vullings HG. 1987. Co-localization of the adipokinetic hormones I and II in the same glandular cells in the same secretory granules of corpus cardiacum of Locusta migratoria and Schistocerca gregeria. Cell Tissue Res 249:379 –389. Distler P. 1990. Synaptic connections of dopamine-immunoreactive neurons in the antennal lobes of Periplaneta americana. Colocalization with GABA-like immunoreactivity. Histochemistry 93:617– 626. Distler P, Boeckh J. 1996. Synaptic connection between olfactory receptor cells and uniglomerular projection neurons in the antennal lobe of the American cockroach, Periplaneta americana. J Comp Neurol 370:35– 46. Duch C, Mentel T, Pflu¨ ger HJ. 1999. Distribution and activation of different types of octopaminergic DUM neurons in the locust. J Comp Neurol 403:119 –134. Elekes K, Hustert R. 1988. The efferent innervation of the genital chamber by an identified serotonergic neuron in the female cricket Acheta domestica. Cell Tissue Res 252:449 – 457. Elekes K, Hustert R, Geffard M. 1987. Serotonin-immunoreactive and dopamine-immunoreactive neurones in the terminal ganglion of the cricket, Acheta domesticus. Cell Tissue Res 250:167–180. Emson PC, Burrows M, Fonnum F. 1974. Levels of glutamate decarboxylase, choline acetyltransferase and acetylcholinesterase in identified motorneurons of the locust. J Neurobiol 5:33– 42. Fabian-Fine R, Volknandt W, Seyfarth E-A. 1999. Peripheral synapses at identifiable mechanosensory neurons in the spider Cupiennius salei: synapsin-like immunoreactivity. Cell Tissue Res 295: 13–19 Feeny CJ, Karmanithi S, Pearce J, Govind CK, Atwood HL. 1998. Motor nerve terminals on abdominal muscles in larval flesh flies, Sarcophaga bullata: comparisons with Drosophila. J Comp Neurol 402:197–209. Foelix RF, Choms A. 1979. Fine structure of a spider joint receptor and associated synapses. Eur J Cell Biol 19:149 –159. Fro¨ hlich A. 1985. Freeze-fracture study of an invertebrate multiplecontact synapse: the fly photoreceptor tetrad. J Comp Neurol 241: 311–326. Fro¨ hlich A, Meinertzhagen IA. 1983. Quantitative features of synapse formation in the fly’s visual system. I. The presynaptic photoreceptor terminal. J Neurosci 3:2336 –2349. Ganfornina MD, Sanchez D, Herrera M, Bastiani J. 1999. Developmental expression and molecular characterization of two gap junction channel proteins expressed during embryogenesis in the grasshopper Schistocerca americana. Dev Genet 24:137–150. Garner CC, Kindler S, Gundelfinger ED.2000. Molecular determinants of presynaptic active zones. Curr Opin Neurobiol 10:321–341. Golding DW. 1994. A pattern confirmed and refined- synaptic, nonsynaptic and parasynaptic exocytosis. BioEssays 16:503–508. Gu¨ nther J, Schu¨ rmann F-W. 1973. Zur Feinstruktur des dorsalen Riesenfasersystems im Bauchmark des Regenwurms. II. Synaptische Beziehungen der proximalen Riesenfaserkollateralen. Z Zellforsch 139:369 –396. Hale JP, Burrows M. 1985. Innervation patterns of inhibitory motor neurones in the thorax of the locust. J Exp Biol 117:401– 413. Hardt M, Watson AHD. 1994. Distribution of synapses on two ascending interneurones carrying frequency-specific information in the auditory system of the cricket: evidence for GABAergic inputs. J Comp Neurol 345:481– 495. Hardt M, Watson AHD. 1999. Distribution of input and output synapses on the central branches of bushcricket and cricket auditory afferent neurones: immunocytochemical evidence for GABA and

SYNAPSES IN THE CNS OF ORTHOPTERAN INSECTS glutamate in different populations of presynaptic boutons. J Comp Neurol 403:281–294. Heitler WJ, Burrows M. 1977a. The locust jump. I. The motor programme. J Exp Biol 66:203–219. Heitler WJ, Burrows M. 1977b. The locust jump. II. Neural circuits of the motor programme. J Exp Biol 66:221–241. Ho¨ rner M, Helle J, Schu¨ rmann F-W. 1996. The distribution of immunoreactive neurons in the ventral nerve cord of the cricket Gryllus bimaculatus. Cell Tissue Res 286:393– 405. Hoyle G, Burrrows M. 1973. Neural mechanisms underlying behavior in the locust Schistocerca gregaria. Iphysiology of identified motorneurons in the metathoracic ganglion. J Neurobiol 4:3– 41. Hoyle G, Dagan D. 1978. Physiological characteristics and reflex activation of DUM (octopaminergic) neurons of locust metathoracic ganglion. J Neurobiol 9:59 –79. Hoyle G, Dagan D, Moberley B, Colquhoun W. 1974. Dorsal unpaired median insect neurones make neurosecretory endings on skeletal muscle. J Exp Zool 187:159 –165. Hue B, Callec JJ. 1983. Presynaptic inhibition in the cercal-afferent giant-interneurone synapses of the cockroach, Periplaneta americana. J Insect Physiol 291:741–748. Jacobs K, Todman MG, Allen MJ, Davies JA, Bacon JP. 2000. Synaptogenesis in the giant fibre system of Drosophila: interaction of the giant fibre and ist major motorneuronal target. Development 127:5203–5212. Jellema T, Tait DS, Heitler WJ. 1997. The concerted activity in parallel proprioceptive pathways controls the initiation of co-activation in the locust kick motor programme. Eur J Neurosci 9:55– 64. Iwasaki M, Mizunami M, Nishikawa M, Itoh T, Tominaga Y. 1999. Ultrastructural analyses of modular subunits in the mushroom bodies of the cockroach. J Electron Microsc 48: 55– 62. Kerwien P, Schu¨ rmann FW. 1990. Nichtsynaptische Exocytose im Hirn und in den Corpora cardiaca von Gryllus bimaculatus: Eine Tannin-Studie. In: Elsner N, Roth G, editors. Brain-perceptioncognition. Proc 18th Go¨ ttingen Neurobiology Conf. Stuttgart: Thieme, p 312. Keshisnian H, O’Shea M. 1985. The distribution of a peptide neurotransmitter in the postembryonic grasshopper central nervous system. J Neurosci 5:992–1004. Killmann F. Schu¨ rmann F-W. 1985. Both electrical and chemical transmission between the ‘lobula giant movement detector’ and the ‘descending contralateral movement detctor’ neurons of locusts are supported by electron microscopy. J Neurocytol 14: 637– 652 Killmann F, Gras H, Schu¨ rmann F-W. 1999. Types, numbers and distribution of synapses on the dendritic tree of an identified visual interneuron in the brain of the locust. Cell Tissue Res 296:645– 665. Klagges BER, Heimbeck G, Godenschwege TG, Hofbauer A, Pflugfelder GO, Reifengerste R, Reisch D, Schaupp M, Buchner S, Buchner E. 1996. Invertebrate Synapsins: A single gene codes for several isoforms in Drosophila. J Neurosci 15:3154 –3165. Kneussel M, Betz H. 2000. Receptors, gephyrin and gephyrin-associated proteins: novel insights into the assembly of inhibitory postsynaptic membrane specializations. J Physiol 525:1–9. Kondoh Y, Hisada M. 1986a. Distribution and ultrastructure of synapses on a premotor local nonspiking interneuron of the crayfish. J Comp Neurol 254:259 –270. Kondoh Y, Hisada M. 1986b. Regional specialization in synaptic input and output in an identified local nonspiking interneuron of the crayfish revealed by light and electron microscopy. J Comp Neurol 251:334 –348. Lane NJ, Sattelle DB, Hufnagel LA. 1983. Pre- and post-synaptic structures in insect CNS: intramembranous features and sites of alpha-bungarotoxin binding. Tissue Cell 15:921–937. Laurent G, Burrows M. 1989. Intersegmental interneurons can control the gain of reflexes in adjacent segments of the locust by their action on nonspiking local interneurons. J Neurosci 9:3030 –3039. Leise EM. 1991. Evolutionary trends in invertebrate ganglionic structure. Neurosciences 2:369 –377. Leitch B, Laurent G. 1996. GABAergic synapses in the antennal lobe and mushroom boy of the locust olfactory system. J Comp Neurol 372:487–514. Leitinger G, Simmons PJ. 2000. Cytochemical evidence that acetylcholine is a neurotransmitter of neurons that make excitatory and inhibitory outputs in the locust ocellar visual system. J Comp Neurol 416:345–355 Littlewood PM, Simmons PJ. 1992. Distribution and structure of identified tonic and phasic synapses between L-neurones in the locust ocellar tract. J Comp Neurol 325:493–513. Lutz EM, Tyrer NM. 1988. Immunohistochemical localization of serotonin and choline acetyltransferase in sensory neurones of the locust. J Comp Neurol 267:335–342.

225

Mancini G, Frontali N. 1970. On the ultrstructural localization of catecholamines in the ␤-lobes (corpora pedunculata) of Periplaneta americana. Z Zellforsch 103:341–350. Meinertzhagen IA. 1989. Fly photoreceptor synapses: their development, evolution, and plasticity. J Neurobiol 20:276 –294. Meinertzhagen IA. 1993. The synaptic population of the fly’s optic neuropil and their dynamic regulation: parallels with the vertebrate retina. Prog Retinal Res 12:13–39. Meinertzhagen IA, Hu X. 1996. Evidence for site selectionduring synaptogenesis: The surface distribution of synaptic sites in photoreceptor terminals of the flies Musca and Drosophila. Cell Mol Neurobiol 16:677– 678. Miksys S, Orchard I. 1994. Immunogold labelling of serotonin-like and FMRFamide-like immunoreactive material in neurohaemal areas on abdominal nerves of Rhodnius prolixus. Cell Tissue Res 278:145–151. Na¨ ssel DR. 1996 Neuropeptides, amines and amino acids in an elementary insect ganglion: functional and chemical anatomy of the unfused abdominal ganglion. Prog Neurobiol 48:325– 420. Newland PL. 1991. Morphology and somatotopic organisation of the central projections of afferents from tactile hairs on the hind leg of the locust. J Comp Neurol 312:493–508. O’Carroll DC, Osorio D, James AC, Bush. 1992. Local feedback mediated via amacrine cells in the insect optic lobe. J Comp Physiol 171:447– 455. Osborne MP. 1966. The fine structure of synapses and tight junctions in the central nervous system of the blowfly larva. J Insect Physiol 12:1503–1512 Parker D. 1994. Glutamatergic transmission between antagonistic motor neurons in the locust. J Comp Physiol [A] 119:267–283. Parker D, Newland PL. 1995. Cholinergic synaptic transmission between proprioceptive afferents and a hind leg motor neuron in the locust. J Neurophysiol 73:586 –594. Pearlstein E, Watson AH, Bevengut M, Cattaert D. 1998. Inhibitory connections between antagonistic motor neurones of the crayfish walking legs. J Comp Neurol 399:241–254. Pearson KG. 1979. Local neurons and local interactions in the nervous system of invertebrates. In: Schmitt FO, Worden FG, editors. The neurosciences. Fourth study program. Cambridge, MA: MIT Press, p 145–157. Pearson KG, Robertson RM. 1981. Interneurons co-activating hindleg flexor and extensor motorneurons in the locust. J Comp Physiol 94:391– 400. Peters BH, Butler SV, Tyrer NM. 1987. Morphology, ultrastructure and synapse distribution of putative serotonergic salivary neurons in the locust. Neuroscience 23:705–719. Peters BH, Ro¨ mer H, Marquart V. 1986. Spatial segregation of synaptic inputs and outputs in a locust auditory interneurone. J Comp Neurol 254:34 –50. Peters BH, Tyrer NM. 1987. Electron microscopy of serotonin-immunoreactive neuron branches and terminals in the locust central nervous system. Neuroscience 23:333–341. Pflu¨ ger HJ, Watson AHD. 1995. GABA and glutamate-like immunoreactivity at synapses received by dorsal unpaired median neurones in the abdominal nerve cord of the locust. Cell Tissue Res 280:325– 333. Phelan P, Nakagawa M, Wilkin MB, Moffat KG, Okane CJ, Davis JA, Bacon JP. 1996. Mutations in shaking-B prevent electriical synapse formation in the Drosophila giant fibre system. J Neurosci 16:1101– 1113. Phelan P, Bacon JP, Davies JA, Stebbings LA, Todman MG, Avery L, Baines RA, Barnes TM, Ford CC, Hekimi S, Lee R, Shaw JE, Starich TA, Curtin KD, Sun YA, Wyman RJ. 1998. Innexins: a family of invertebrate gap-junction proteins. Trends Genet 14:348 – 349. Prokop A. 1999. Integrating bits and pieces: synapse structure and formation in Drosophila embryos. Cell Tissue Res 297:169 –186. Reinecke M, Walther C. 1981. Ultrastructural changes with high activity and subsequent recovery at locust motor nerve terminals. A stereological analysis. Neuroscience 6:489 –503. Rind FC. 1984. A chemical synapse between two motion detecting neurones in the locust brain. J Exp Biol 110:143–167. Rind FC, Leitinger G. 2000. Immunocytochemical evidence that collision sensing neurons in the locust visual system contain acetylcholine. J Comp Neurol 423:389 – 401. Ro¨ mer H, Marquart V. 1984. Morphology and physiology of auditory interneurons in the metathoracic ganglion of the locust. J Comp Physiol [A] 155. Rowell CHF, O’Shea M. 1980. Modualtion of transmission of an electrical system in the locust movement detector system. J Comp Physiol 137:233–241.

226

¨ RMANN A.H.D. WATSON AND F.-W. SCHU

Schu¨ rmann F-W. 1971. Synaptic contacts of association fibers in the brain of the bee. Brain Res 26:169 –176. ¨ ber die Struktur der Pilzko¨ rper des InsekSchu¨ rmann F-W. 1972. U tenhirns. II. Synaptische Schaltungen im Alpha-Lobus des Heimchens Acheta domesticus L. Z Zellforsch 127:240 –257. Schu¨ rmann FW. 1974. Bemerkungen zur Funktion der Corpora pedunculata im Gehirn der Insekten aus morphologischer Sicht. Exp Brain Res 19:406 – 432 Schu¨ rmann F-W. 1980. Methods for special staining of synaptic sites. In: Strausfeld NJ, Miller TA, editors. Neuroanatomical techniques. Berlin: Springer, p 231–264. Schu¨ rmann F-W. 1987. The architecture of the mushroom bodies and related neuropils in the insect brain. In: Gupta PA, editor. Arthropod brain: evolution, development, structure, and functions. New York: Wiley. p 231–264. Schu¨ rmann F-W, Wechsler W. 1969. Elektronenmikroskopische Untersuchung am Antennallobus des Deutocerebrum der Wanderheuschrecke Locusta migratoria. Z Zellforsch 95:223–248. Schu¨ rmann F-W, Wechsler W. 1970. Synapsen im Antennenhu¨ gel von Locusta migratoria. Z Zellforsch 108:563–581. Schu¨ rmann F-W, Sandeman R, Sandeman D. 1991. Dense core vesicles and non-synaptic exocytosis in the central body of the crayfish brain. Cell Tissue Res. 265:493–501. Schu¨ rmann F-W, Ho¨ rner M, Spo¨ rhase-Eichmann U. 1995. The diverse layout of amine-containing systems in the ventral cord of an insect. Acta Biol Hung 46:485– 490. Siegler MV. 1982. Electrical coupling between supernumerary motor neurones in the locust. J Exp Biol 101:105–119. Simmons PJ, Littlewood PMH. 1989. Structure of a tonically transmitting synapse between identified interneurones in the locust brain. J Comp Neurol 283:129 –142 Sombati S, Hoyle G. 1984. Glutamatergic central nervous transmission in locusts. J Neurobiol 15:507–516. Strausfeld NJ. 1976. An atlas of an insect brain. Berlin: Springer Verlag. Strausfeld NJ, Bassemir UK. 1983. Cobalt-coupled neurons of a giant fibre system in Diptera. J Neurocytol 12:971–991. Strausfeld NJ, Li Y. 1999 Representation of the calyces in the medial and vertical lobes of cockroach mushroom bodies: J Comp Neurol 409:626 – 646. Tanouye MA, Wyman RJ. 1980. Motor output of giant fibre in Drosophila. J Neurophysiol 44:405– 421. Tolbert LP, Hildebrand JG. 1981. Organization and synaptic ultrastructure of glomeruli in the antennal lobes of the moth Manduca sexta: a study using thin sections and freeze fracture. Proc R Soc Lond B Biol Sci 213:279 –301. Trujillo-Cenoz O, Melamed J. 1962. Electron microscope observations on the calyces of the insect brain. J Ultrastruct Res 7:389 –398. Tyrer NM, Altman JS. 1974. Motor and sensory flight neurones in a locust demonstrated using cobalt chloride. J Comp Neurol 157:117– 138. Tyrer NM, Altman JS. 1976. Spatial distribution of synapses onto thoracic motor neurones in locusts. Cell Tissue Res 166:389 –398. Ude J, Agricola H. 1995. FMRFamide-like and allatostatin-like immunoreactivity in the lateral heart nerve of Periplaneta americana: colocalization at the electron- microscopic level. Cell Tissue Res 282:69 – 80. Vizi ES, Labos E. 1991. Non-synaptic Interactions at presynaptic level. Progr Neurobiol 37:145–163. Watson AHD. 1984. The dorsal unpaired median neurons of the locust metathoracic ganglion: neuronal structure and diversity, and synapse distribution. J Neurocytol 13:303–327.

Watson AHD. 1986. The distribution of GABA-like immunoreactivity in the thoracic nervous system of the locust Schistocerca gregaria. Cell Tissue Res 246:331–341. Watson AHD. 1988. Antibodies against GABA and glutamate label neurons with morphologically distinct synaptic vesicles in the locust central nervous system. Neuroscience 26:33– 44. Watson AHD. 1990. Ultrastructural evidence for GABAergic input onto cercal afferents in the locust (Locusta migratoria). J Exp Biol 148:509 –515. Watson AHD. 1992. Presynaptic modulation of sensory afferents in the invertebrate and vertebrate nervous system. Comp Biochem Physiol Comp Physiol 103:227–339. Watson AHD, Burrows M. 1982. The ultrastructure of identified locust motor neurones and their synaptic relationships. J Comp Neurol 205:383–397. Watson AHD, Burrows M. 1983. The morphology, ultrastructure, and distribution of synapses on an intersegmental interneuron of the locust. J Comp Neurol 214:154 –169. Watson AHD, Burrows M. 1985. The distribution of synapses on the two fields of neurites of spiking local interneurones in the locust. J Comp Neurol 240:219 –232. Watson AHD, Burrows M. 1988. Distribution and morphology of synapses on nonspiking local interneurones in the thoracic nervous system of the locust. J Comp Neurol 272:605– 616. Watson AHD, England RCD. 1991. The distribution of and interactions between GABA-immunoreactive and non-immunoreactive processes presynaptic to afferents from campaniform sensilla on the trochanter of the locust leg. Cell Tissue Res 266:331–341. Watson AHD, Hardt M. 1996. Distribution of synapses on two local auditory interneurones, ON1 and ON2, in the prothoracic ganglion of the cricket: relationships with GABA-immunoreactive neurones. Cell Tissue Res 283:231–246. Watson AHD, Laurent G. 1990. GABA-like immunoreactivity in a population of locust intersegmental interneurones and their inputs. J Comp Neurol 302:761–767. Watson AHD, Pflu¨ ger HJ. 1984. The ultrastructure of prosternal sensory hair afferents within the locust central nervous system. Neuroscience 11:269 –279. Watson AHD, Pflu¨ ger HJ. 1989. Regional specialisation for synaptic input and output on a locust intersegmental interneurone with multiple spike-initiating zones. J Comp Neurol 279:515–527. Watson AHD, Pflu¨ ger HJ. 1994. Distribution of input synapses from processes exhibiting GABA- or glutamate-like immunoreactivity onto terminals of prosternal filiform afferents in the locust. J Comp Neurol 343:617– 629. Watson AHD, Seymour-Laurent KJ. 1993. The distribution of glutamate-like immunoreactivity in the thoracic and abdominal ganglia of the locust (Schistocerca gregaria). Cell Tissue Res 273:557–570. Watson AHD, Burrows M, Hale JP. 1985. The morphology and ultrastructure of common inhibitory motor neurones in the thorax of the locust. J Comp Neurol 239:341–359. Watson AHD, Storm-Mathisen J, Ottersen OP. 1991. GABA and glutamate-like immunoreactivity in processes presynaptic to afferents from hair plates on the proximal joints of the locust leg. J Neurocytol 20:796 – 809. Watson AHD, Burrows M, Leitch B. 1993. GABA-immunoreactivity in processes presynaptic to the terminals of afferents from a locust leg proprioceptor. J Neurocytol 22:547–557. Wood MR, Pfenninger KH, Cohen MJ. 1977. Two types of presynaptic configurations in insect central synapses: an ultrastructural analysis. Brain Res 130:25– 45.