Sensory pathways and motoneurons of the proboscis reflex in the suboesophageal ganglion of the honey bee

Pollen reinforces learning in honey bee pollen foragers but not in nectar foragers

Searching for reward motivates and drives behaviour. In honey bees Apis mellifera, specialized pollen foragers are attracted to and learn odours with pollen. However, the role of pollen as a reward remains poorly understood. Unlike nectar, pollen is not ingested during collection. We hypothesized that pollen (but not nectar) foragers could learn pollen by sole antennal or tarsal stimulation. Then, we tested how pairing of pollen (either hand- or bee-collected) and a neutral odour during a pre-conditioning affects performance of both pollen and nectar foragers during the classical conditioning of the proboscis extension response. Secondly, we tested whether nectar and pollen foragers perceive the simultaneous presentation of pollen (on the tarsi) and sugar (on the antennae) as a better reinforcement than sucrose alone. Finally, we searched for differences in learning of the pollen and nectar foragers when they were prevented from ingesting the reward during the conditioning. Differences in pollen-reinforced learning correlate with division of labour between pollen and nectar foragers. Results show that pollen foragers performed better than nectar foragers during the conditioning phase after being pre-conditioned with pollen. Pollen foragers also performed better than nectar foragers in both the acquisition and extinction phases of the conditioning, when reinforced with the dual reward. Consistently, pollen foragers showed improved abilities to learn cues reinforced without sugar ingestion. We discussed that differences in how pollen and nectar foragers respond to a cue associated with pollen greatly contribute to the physiological mechanism that underlies foraging specialization in the honeybee.

Mass Spectrometric Quantification of Arousal Associated Neurochemical Changes in Single Honey Bee Brains and Brain Regions

Honey bee foragers show a strong diurnal rhythm of foraging activity, and such behavioral changes are likely under the control of specific neuromodulators. To identify and quantify neuromodulators involved in regulating rest and arousal in honey bees, we established a mass spectrometric method for quantifying 14 different neurochemicals and precursor molecules. We measured forager type and brain region specific differences in amine levels from individual honey bee brains and brain regions. The observed differences in amine levels between resting and aroused foragers resemble findings in other species indicating a conserved molecular mechanism by glutamate and GABA in regulating arousal. Subesophageal ganglion specific changes in the histaminergic system and global increases in aspartate during arousal suggest a possible role of histamine and aspartate in feeding and arousal, respectively. More aminergic systems were significantly affected due to arousal in nectar foragers than in pollen foragers, implying that forager phenotypes differ not only in their food preference but also in their neuromodulatory signaling systems (brain states). Finally, we found that neurotransmitter precursors were better at distinguishing brain states in the central brain, while their end products correlated with arousal associated changes in sensory regions like the optic and antennal lobes.

Levels and activity of cyclic guanosine monophosphate-dependent protein kinase in nurse and forager honeybees

Age-dependent division of labour in honeybees was shown to be connected to sensory response thresholds. Foragers show a higher gustatory responsiveness than nurse bees. It is generally assumed that nutrition-related signalling pathways underlie this behavioural plasticity. Here, one important candidate gene is the foraging gene, which encodes a cyclic guanosine monophosphate-dependent protein kinase (PKG). Several roles of members of this enzyme family were analysed in vertebrates. They own functions in important processes such as growth, secretion and neuronal adaptation. Honeybee foraging messenger RNA expression is upregulated in the brain of foragers. In vivo activation of PKG can modulate gustatory responsiveness. We present for the first time PKG protein level and activity data in the context of social behaviour and feeding. Protein level was significantly higher in brains of foragers than in those of nurse bees, substantiating the role of PKG in behavioural plasticity. However, enzyme activity did not differ between behavioural roles. The mediation of feeding status appears independent of PKG signalling. Neither PKG content nor enzyme activity differed between starved and satiated individuals. We suggest that even though nutrition-related pathways are surely involved in controlling behavioural plasticity, which involves changes in PKG signalling, mediation of satiety itself is independent of PKG.

Structure and development of the subesophageal zone of the Drosophila brain. I. Segmental architecture, compartmentalization, and lineage anatomy

The subesophageal zone (SEZ) of the Drosophila brain houses the circuitry underlying feeding behavior and is involved in many other aspects of sensory processing and locomotor control. Formed by the merging of four neuromeres, the internal architecture of the SEZ can be best understood by identifying segmentally reiterated landmarks emerging in the embryo and larva, and following the gradual changes by which these landmarks become integrated into the mature SEZ during metamorphosis. In previous works, the system of longitudinal fibers (connectives) and transverse axons (commissures) has been used as a scaffold that provides internal landmarks for the neuromeres of the larval ventral nerve cord. We have extended the analysis of this scaffold to the SEZ and, in addition, reconstructed the tracts formed by lineages and nerves in relationship to the connectives and commissures. As a result, we establish reliable criteria that define boundaries between the four neuromeres (tritocerebrum, mandibular neuromere, maxillary neuromere, labial neuromere) of the SEZ at all stages of development. Fascicles and lineage tracts also demarcate seven columnar neuropil domains (ventromedial, ventro-lateral, centromedial, central, centrolateral, dorsomedial, dorsolateral) identifiable throughout development. These anatomical subdivisions, presented in the form of an atlas including confocal sections and 3D digital models for the larval, pupal and adult stage, allowed us to describe the morphogenetic changes shaping the adult SEZ. Finally, we mapped MARCM-labeled clones of all secondary lineages of the SEZ to the newly established neuropil subdivisions. Our work will facilitate future studies of function and comparative anatomy of the SEZ.

Structure and development of the subesophageal zone of the Drosophila brain. II. Sensory compartments

The subesophageal zone (SEZ) of the Drosophila brain processes mechanosensory and gustatory sensory input from sensilla located on the head, mouth cavity and trunk. Motor output from the SEZ directly controls the movements involved in feeding behavior. In an accompanying paper (Hartenstein et al., ), we analyzed the systems of fiber tracts and secondary lineages to establish reliable criteria for defining boundaries between the four neuromeres of the SEZ, as well as discrete longitudinal neuropil domains within each SEZ neuromere. Here we use this anatomical framework to systematically map the sensory projections entering the SEZ throughout development. Our findings show continuity between larval and adult sensory neuropils. Gustatory axons from internal and external taste sensilla of the larva and adult form two closely related sensory projections, (a) the anterior central sensory center located deep in the ventromedial neuropil of the tritocerebrum and mandibular neuromere, and (b) the anterior ventral sensory center (AVSC), occupying a superficial layer within the ventromedial tritocerebrum. Additional, presumed mechanosensory terminal axons entering via the labial nerve define the ventromedial sensory center (VMSC) in the maxilla and labium. Mechanosensory afferents of the massive array of chordotonal organs (Johnston's organ) of the adult antenna project into the centrolateral neuropil column of the anterior SEZ, creating the antenno-mechanosensory and motor center (AMMC). Dendritic projections of dye back-filled motor neurons extend throughout a ventral layer of the SEZ, overlapping widely with the AVSC and VMSC. Our findings elucidate fundamental structural aspects of the developing sensory systems in Drosophila.

Molecular characterization of tsetse's proboscis and its response to Trypanosoma congolense infection

Tsetse flies (Glossina spp.) transmit parasitic African trypanosomes (Trypanosoma spp.), including Trypanosoma congolense, which causes animal African trypanosomiasis (AAT). AAT detrimentally affects agricultural activities in sub-Saharan Africa and has negative impacts on the livelihood and nutrient availability for the affected communities. After tsetse ingests an infectious blood meal, T. congolense sequentially colonizes the fly’s gut and proboscis (PB) organs before being transmitted to new mammalian hosts during subsequent feedings. Despite the importance of PB in blood feeding and disease transmission, little is known about its molecular composition, function and response to trypanosome infection. To bridge this gap, we used RNA-seq analysis to determine its molecular characteristics and responses to trypanosome infection. By comparing the PB transcriptome to whole head and midgut transcriptomes, we identified 668 PB-enriched transcripts that encoded proteins associated with muscle tissue, organ development, chemosensation and chitin-cuticle structure development. Moreover, transcripts encoding putative mechanoreceptors that monitor blood flow during tsetse feeding and interact with trypanosomes were also expressed in the PB. Microscopic analysis of the PB revealed cellular structures associated with muscles and cells. Infection with T. congolense resulted in increased and decreased expression of 38 and 88 transcripts, respectively. Twelve of these differentially expressed transcripts were PB-enriched. Among the transcripts induced upon infection were those encoding putative proteins associated with cell division function(s), suggesting enhanced tissue renewal, while those suppressed were associated with metabolic processes, extracellular matrix and ATP-binding as well as immunity. These results suggest that PB is a muscular organ with chemosensory and mechanosensory capabilities. The mechanoreceptors may be point of PB-trypanosomes interactions. T. congolense infection resulted in reduced metabolic and immune capacity of the PB. The molecular knowledge on the composition and putative functions of PB forms the foundation to identify new targets to disrupt tsetse’s ability to feed and parasite transmission.

Tsetse flies are economically important insects responsible for transmitting African trypanosomes, which cause debilitating and fatal diseases in humans and animals in sub-Saharan Africa. In the tsetse vector, trypanosomes undergo complex developmental processes in the midgut, culminating with the generation of mammalian infective forms in the salivary glands for Trypanosoma brucei spp. and in the proboscis (PB) for Trypanosoma congolense and Trypanosoma vivax. Molecular studies on tsetse’s PB, and its interactions with trypanosomes, are limited. We used RNA-seq analysis to obtain molecular information on the putative products associated with tsetse’s PB and characterized PB responses to infection with T. congolense. Based on the predicted putative protein profile, the PB appears to be a muscular organ with mechanoreceptors and may have the capacity to sense and respond to chemical cues. Parasite infections of the PB lead to decreased expression of genes whose products are associated with metabolic and immune functions. These data provide insights into tsetse-trypanosome interactions in the PB organ and identify potential candidate targets that can be further explored to develop biotechnological strategies to reduce transmission of trypanosomes by tsetse flies.

A Tyrosine-Hydroxylase Characterization of Dopaminergic Neurons in the Honey Bee Brain

Dopamine (DA) plays a fundamental role in insect behavior as it acts both as a general modulator of behavior and as a value system in associative learning where it mediates the reinforcing properties of unconditioned stimuli (US). Here we aimed at characterizing the dopaminergic neurons in the central nervous system of the honey bee, an insect that serves as an established model for the study of learning and memory. We used tyrosine hydroxylase (TH) immunoreactivity (ir) to ensure that the neurons detected synthesize DA endogenously. We found three main dopaminergic clusters, C1-C3, which had been previously described; the C1 cluster is located in a small region adjacent to the esophagus (ES) and the antennal lobe (AL); the C2 cluster is situated above the C1 cluster, between the AL and the vertical lobe (VL) of the mushroom body (MB); the C3 cluster is located below the calyces (CA) of the MB. In addition, we found a novel dopaminergic cluster, C4, located above the dorsomedial border of the lobula, which innervates the visual neuropils of the bee brain. Additional smaller processes and clusters were found and are described. The profuse dopaminergic innervation of the entire bee brain and the specific connectivity of DA neurons, with visual, olfactory and gustatory circuits, provide a foundation for a deeper understanding of how these sensory modules are modulated by DA, and the DA-dependent value-based associations that occur during associative learning.

Worker brain development and colony organization in ants: Does division of labor influence neuroplasticity?

Brain compartment size allometries may adaptively reflect cognitive needs associated with behavioral development and ecology. Ants provide an informative system to study the relationship of neural architecture and development because worker tasks and sensory inputs may change with age. Additionally, tasks may be divided among morphologically and behaviorally differentiated worker groups (subcastes), reducing repertoire size through specialization and aligning brain structure with task-specific cognitive requirements. We hypothesized that division of labor may decrease developmental neuroplasticity in workers due to the apparently limited behavioral flexibility associated with task specialization. To test this hypothesis, we compared macroscopic and cellular neuroanatomy in two ant sister clades with striking contrasts in worker morphological differentiation and colony-level social organization: Oecophylla smaragdina, a socially complex species with large colonies and behaviorally distinct dimorphic workers, and Formica subsericea, a socially basic species with small colonies containing monomorphic workers. We quantified volumes of functionally distinct brain compartments in newly eclosed and mature workers and measured the effects of visual experience on synaptic complex (microglomeruli) organization in the mushroom bodies-regions of higher-order sensory integration-to determine the extent of experience-dependent neuroplasticity. We demonstrate that, contrary to our hypothesis, O. smaragdina workers have significant age-related volume increases and synaptic reorganization in the mushroom bodies, whereas F. subsericea workers have reduced age-related neuroplasticity. We also found no visual experience-dependent synaptic reorganization in either species. Our findings thus suggest that changes in the mushroom body with age are associated with division of labor, and therefore social complexity, in ants. © 2017 Wiley Periodicals, Inc. Develop Neurobiol 77: 1072-1085, 2017.

Sucrose Sensitivity of Honey Bees Is Differently Affected by Dietary Protein and a Neonicotinoid Pesticide

Over a decade, declines in honey bee colonies have raised worldwide concerns. Several potentially contributing factors have been investigated, e.g. parasites, diseases, and pesticides. Neonicotinoid pesticides have received much attention due to their intensive use in crop protection, and their adverse effects on many levels of honey bee physiology led the European Union to ban these compounds. Due to their neuronal target, a receptor expressed throughout the insect nervous system, studies have focused mainly on neuroscience and behaviour. Through the Geometric Framework of nutrition, we investigated effects of the neonicotinoid thiamethoxam on survival, food consumption and sucrose sensitivity of honey bees (Apis mellifera). Thiamethoxam did not affect protein and carbohydrate intake, but decreased responses to high concentrations of sucrose. Interestingly, when bees ate fixed unbalanced diets, dietary protein facilitated better sucrose detection. Both thiamethoxam and dietary protein influenced survival. These findings suggest that, in the presence of a pesticide and unbalanced food, honey bee health may be severely challenged. Consequences for foraging efficiency and colony activity, cornerstones of honey bee health, are also discussed.

Division of labour in honey bees: age- and task-related changes in the expression of octopamine receptor genes

The honey bee (Apis mellifera L.) has developed into an important ethological model organism for social behaviour and behavioural plasticity. Bees perform a complex age-dependent division of labour with the most pronounced behavioural differences occurring between in-hive bees and foragers. Whereas nurse bees, for example, stay inside the hive and provide the larvae with food, foragers leave the hive to collect pollen and nectar for the entire colony. The biogenic amine octopamine appears to play a major role in division of labour but the molecular mechanisms involved are unknown. We here investigated the role of two characterized octopamine receptors in honey bee division of labour. AmOctαR1 codes for a Ca(2+) -linked octopamine receptor. AmOctβR3/4 codes for a cyclic adenosine monophosphate-coupled octopamine receptor. Messenger RNA expression of AmOctαR1 in different brain neuropils correlates with social task, whereas expression of AmOctβR3/4 changes with age rather than with social role per se. Our results for the first time link the regulatory role of octopamine in division of labour to specific receptors and brain regions. They are an important step forward in our understanding of complex behavioural organization in social groups.

Octopamine-like immunoreactive neurons in the brain and subesophageal ganglion of the parasitic wasps Nasonia vitripennis and N. giraulti

Octopamine is an important neuromodulator in the insect nervous system, influencing memory formation, sensory perception and motor control. In this study, we compare the distribution of octopamine-like immunoreactive neurons in two parasitic wasp species of the Nasonia genus, N. vitripennis and N. giraulti. These two species were previously described as differing in their learning and memory formation, which raised the question as to whether morphological differences in octopaminergic neurons underpinned these variations. Immunohistochemistry in combination with confocal laser scanning microscopy was used to reveal and compare the somata and major projections of the octopaminergic neurons in these wasps. The brains of both species showed similar staining patterns, with six different neuron clusters being identified in the brain and five different clusters in the subesophageal ganglion. Of those clusters found in the subesophageal ganglion, three contained unpaired neurons, whereas the other three consisted in paired neurons. The overall pattern of octopaminergic neurons in both species was similar, with no differences in the numbers or projections of the ventral unpaired median (VUM) neurons, which are known to be involved in memory formation in insects. In one other cluster in the brain, located in-between the optic lobe and the antennal lobe, we detected more neurons in N. vitripennis compared with N. giraulti. Combining our results with findings made previously in other Hymenopteran species, we discuss possible functions and some of the ultimate factors influencing the evolution of the octopaminergic system in the insect brain.

Central projections of gustatory receptor neurons in the medial and the lateral sensilla styloconica of Helicoverpa armigera larvae

Food selection behavior of lepidopteran larvae is predominantly governed by the activation of taste neurons present in two sensilla styloconica located on the galea of the maxilla. In this study, we present the ultrastructure of the sensilla styloconica and the central projection pattern of their associated receptor neurons in larvae of the heliothine moth, Helicoverpa armigera. By means of light microscopy and scanning electron microscopy, the previous findings of two morphologically fairly similar sensilla comprising a socketed conic tip inserted into a large peg were confirmed. However, the peg size of the medial sensillum was found to be significantly bigger than that of the lateral sensillum. The sensory neurons derived from each sensillum styloconicum were mapped separately using anterograde staining experiments combined with confocal laser-scanning microscopy. For determining the afferents' target regions relative to each other, we reconstructed the labeled axons and placed them into a common reference framework. The sensory axons from both sensilla projected via the ipsilateral maxillary nerve to the suboesophageal ganglion and further through the ipsilateral circumoesophageal connective to the brain. In the suboesophageal ganglion, the sensory projections targeted two areas of the ipsilateral maxillary neuropil, one located in the ventrolateral neuromere and the other adjacent to the neuromere midline. In the brain, the axon terminals targeted the dorso-anterior area of the ipsilateral tritocerebrum. As confirmed by the three-dimensional reconstructions, the target regions of the neural projections originating from each of the two sensilla styloconica were identical.

Cognition with few neurons: higher-order learning in insects

Insects possess miniature brains but exhibit a sophisticated behavioral repertoire. Recent studies have reported the existence of unsuspected cognitive capabilities in various insect species that go beyond the traditionally studied framework of simple associative learning. Here, I focus on capabilities such as attentional modulation and concept learning and discuss their mechanistic bases. I analyze whether these behaviors, which appear particularly complex, can be explained on the basis of elemental associative learning and specific neural circuitries or, by contrast, require an explanatory level that goes beyond simple associative links. In doing this, I highlight experimental challenges and suggest future directions for investigating the neurobiology of higher-order learning in insects, with the goal of uncovering the basic neural architectures underlying cognitive processing.

Serotonin depresses feeding behaviour in ants

Feeding behaviour is a complex functional system that relies on external signals and the physiological state of the animal. This is also the case in ants as they vary their feeding behaviour according to food characteristics, environmental conditions and - as they are social insects - to the colony's requirements. The biogenic amine serotonin (5-HT) was shown to be involved in the control and modulation of many actions and processes related to feeding in both vertebrates and invertebrates. In this study, we investigated whether 5-HT affects nectar feeding in ants by analysing its effect on the sucking-pump activity. Furthermore, we studied 5-HT association with tissues and neuronal ganglia involved in feeding regulation. Our results show that 5-HT promotes a dose-dependent depression of sucrose feeding in Camponotus mus ants. Orally administered 5-HT diminished the intake rate by mainly decreasing the volume of solution taken per pump contraction, without modifying the sucrose acceptance threshold. Immunohistochemical studies all along the alimentary canal revealed 5-HT-like immunoreactive processes on the foregut (oesophagus, crop and proventriculus), while the midgut and hindgut lacked 5-HT innervation. Although the frontal and suboesophageal ganglia contained 5-HT immunoreactive cell bodies, serotonergic innervation in the sucking-pump muscles was absent. The results are discussed in the frame of a role of 5-HT in feeding control in ants.

Distribution of the octopamine receptor AmOA1 in the honey bee brain

Octopamine plays an important role in many behaviors in invertebrates. It acts via binding to G protein coupled receptors located on the plasma membrane of responsive cells. Several distinct subtypes of octopamine receptors have been found in invertebrates, yet little is known about the expression pattern of these different receptor subtypes and how each subtype may contribute to different behaviors. One honey bee (Apis mellifera) octopamine receptor, AmOA1, was recently cloned and characterized. Here we continue to characterize the AmOA1 receptor by investigating its distribution in the honey bee brain. We used two independent antibodies produced against two distinct peptides in the carboxyl-terminus to study the distribution of the AmOA1 receptor in the honey bee brain. We found that both anti-AmOA1 antibodies revealed labeling of cell body clusters throughout the brain and within the following brain neuropils: the antennal lobes; the calyces, pedunculus, vertical (alpha, gamma) and medial (beta) lobes of the mushroom body; the optic lobes; the subesophageal ganglion; and the central complex. Double immunofluorescence staining using anti-GABA and anti-AmOA1 receptor antibodies revealed that a population of inhibitory GABAergic local interneurons in the antennal lobes express the AmOA1 receptor in the cell bodies, axons and their endings in the glomeruli. In the mushroom bodies, AmOA1 receptors are expressed in a subpopulation of inhibitory GABAergic feedback neurons that ends in the visual (outer half of basal ring and collar regions) and olfactory (lip and inner basal ring region) calyx neuropils, as well as in the collar and lip zones of the vertical and medial lobes. The data suggest that one effect of octopamine via AmOA1 in the antennal lobe and mushroom body is to modulate inhibitory neurons.

Neural architecture of the primary gustatory center of Drosophila melanogaster visualized with GAL4 and LexA enhancer-trap systems

Gustatory information is essential for animals to select edible foods and avoid poisons. Whereas mammals detect tastants with their taste receptor cells, which convey gustatory signals to the brain indirectly via the taste sensory neurons, insect gustatory receptor neurons (GRNs) send their axons directly to the primary gustatory center in the suboesophageal ganglion (SOG). In spite of this relatively simple architecture, the precise structure of the insect primary gustatory center has not been revealed in enough detail. To obtain comprehensive anatomical knowledge about this brain area, we screened the Drosophila melanogaster GAL4 enhancer-trap strains that visualize specific subsets of the gustatory neurons as well as putative mechanosensory neurons associated with the taste pegs. Terminals of these neurons form three branches in the SOG. To map the positions of their arborization areas precisely, we screened newly established LexA::VP16 enhancer-trap strains and obtained a driver line that labels a large subset of peripheral sensory neurons. By double-labeling specific and landmark neurons with GAL4 and LexA strains, we were able to distinguish 11 zones in the primary gustatory center, among which 5 zones were identified newly in this study. Arborization areas of various known GRNs on the labellum, oesophagus, and legs were also mapped in this framework. The putative mechanosensory neurons terminate exclusively in three zones of these areas, supporting the notion of segregated primary centers that are specialized for chemosensory and mechanosensory signals associated with gustatory sensation.

Parallel reinforcement pathways for conditioned food aversions in the honeybee

Avoiding toxins in food is as important as obtaining nutrition. Conditioned food aversions have been studied in animals as diverse as nematodes and humans [1, 2], but the neural signaling mechanisms underlying this form of learning have been difficult to pinpoint. Honeybees quickly learn to associate floral cues with food [3], a trait that makes them an excellent model organism for studying the neural mechanisms of learning and memory. Here we show that honeybees not only detect toxins but can also learn to associate odors with both the taste of toxins and the postingestive consequences of consuming them. We found that two distinct monoaminergic pathways mediate learned food aversions in the honeybee. As for other insect species conditioned with salt or electric shock reinforcers [4-7], learned avoidances of odors paired with bad-tasting toxins are mediated by dopamine. Our experiments are the first to identify a second, postingestive pathway for learned olfactory aversions that involves serotonin. This second pathway may represent an ancient mechanism for food aversion learning conserved across animal lineages.

Biochemical properties of heterologously expressed and native adenylyl cyclases from the honeybee brain (Apis mellifera L.)

Cyclic AMP is an important intracellular signaling molecule participating e.g. in sensory signal transduction, cardiac myocyte regulation, learning and memory. The formation of cAMP is catalyzed by adenylyl cyclases. A variety of factors can modulate the properties of these enzymes and lead to dynamic changes of the intracellular cAMP concentration. Here we determined the tissue distribution of a recently cloned adenylyl cyclase (AmAC3) in honeybee brain. The protein is present in all neuropils. Intensive immunoreactivity was found in parts of the proto- and deutocerebrum and in the suboesophageal ganglion. Biochemical and pharmacological properties of AmAC3 and of native adenylyl cyclases in subregions of the honeybee brain were examined. Values for half-maximal activation with NKH477 were in the low micromolar range with 10.2 μM for AmAC3 and 3.6-8.1 μM for native enzymes. Biosynthesis of cAMP was specifically blocked by P-site inhibitors. Adenylyl cyclases in antennal lobes and AmAC3 share the inhibitory profile with 2',5'dd3'ATP>3'AMP>2'deoxyadenosine. In addition to P-site inhibitors AmAC3 activity was impaired by Ca(2+)/calmodulin. The results suggest that AmAC3 is a likely candidate to fulfill an integrative role in sensory, motor and higher-order information processing in the honeybee brain.

Central gustatory neurons integrate taste quality information from four appendages in the moth Heliothis virescens

Discrimination between edible and noxious food, crucial for animal survival, is based on separate gustatory receptors for phagostimulants and deterrents. In the moth Heliothis virescens, gustatory receptor neurons (GRNs) tuned to phagostimulants like sucrose and deterrents like quinine, respectively, have indicated a labeled line mechanism for mediating appetitive and aversive information to the CNS. In the present study, we have investigated the central gustatory neurons (CGNs) in this moth as an approach to understand how gustatory information is coded in the CNS. Intracellular recordings from CGNs in the suboesophageal ganglion (SOG) combined with fluorescent staining revealed a large diversity of CGN types responding to sucrose, quinine, water, and mechanosensory stimuli applied to the antennae, the proboscis, and the right tarsus. The CGNs responded with varying tuning breadth to tastants applied to more than one appendage. This integration of information across stimuli and appendages, contradict a simple labeled line mechanism in the CNS for coding identity and location of taste stimuli. Instead the distinct pattern of activity found in an ensemble of CGNs, suggests a population coding mechanism. Staining revealed that the majority of the CGNs were confined locally within the SOG/tritocerebrum, whereas others projected to the deutocerebrum, protocerebrum, frontal ganglion, and thoracic ganglia. Some CGNs were reconstructed and registered into the H. virescens standard brain atlas, showing dendritic overlap with the previously described GRN projections. In general, the physiology and morphology of the CGNs suggested multifunctional properties, where a single CGN might belong to several networks executing different functions.

Vibration-processing interneurons in the honeybee brain

The afferents of the Johnston's organ (JO) in the honeybee brain send their axons to three distinct areas, the dorsal lobe, the dorsal subesophageal ganglion (DL-dSEG), and the posterior protocerebral lobe (PPL), suggesting that vibratory signals detected by the JO are processed differentially in these primary sensory centers. The morphological and physiological characteristics of interneurons arborizing in these areas were studied by intracellular recording and staining. DL-Int-1 and DL-Int-2 have dense arborizations in the DL-dSEG and respond to vibratory stimulation applied to the JO in either tonic excitatory, on-off-phasic excitatory, or tonic inhibitory patterns. PPL-D-1 has dense arborizations in the PPL, sends axons into the ventral nerve cord (VNC), and responds to vibratory stimulation and olfactory stimulation simultaneously applied to the antennae in long-lasting excitatory pattern. These results show that there are at least two parallel pathways for vibration processing through the DL-dSEG and the PPL. In this study, Honeybee Standard Brain was used as the common reference, and the morphology of two types of interneurons (DL-Int-1 and DL-Int-2) and JO afferents was merged into the standard brain based on the boundary of several neuropiles, greatly supporting the understanding of the spatial relationship between these identified neurons and JO afferents. The visualization of the region where the JO afferents are closely appositioned to these DL interneurons demonstrated the difference in putative synaptic regions between the JO afferents and these DL interneurons (DL-Int-1 and DL-Int-2) in the DL. The neural circuits related to the vibration-processing interneurons are discussed.

Effect of fipronil on side-specific antennal tactile learning in the honeybee

In the honeybee, the conditioning of the proboscis extension response using tactile antennal stimulations is well suited for studying the side-specificity of learning including the possible bilateral transfer of memory traces in the brain, and the role of inhibitory networks. A tactile stimulus was presented to one antenna in association with a sucrose reward to the proboscis. The other antenna was either not stimulated (A+/0 training), stimulated with a non-reinforced tactile stimulus B (A+/B- training) or stimulated with B reinforced with sucrose to the proboscis (A+/B+ training). Memory tests performed 3 and 24h after training showed in all situations that a tactile stimulus learnt on one side was only retrieved ipsilaterally, indicating no bilateral transfer of information. In all these groups, we investigated the effect of the phenylpyrazole insecticide fipronil by applying a sublethal dose (0.5 ng/bee) on the thorax 15 min before training. This treatment decreased acquisition success and the subsequent memory performances were lowered but the distribution of responses to the tactile stimuli between sides was not affected. These results underline the role of the inhibitory networks targeted by fipronil on tactile learning and memory processes.

Response characteristics of vibration-sensitive interneurons related to Johnston's organ in the honeybee, Apis mellifera

Honeybees detect airborne vibration by means of Johnston's organ (JO), located in the pedicel of each antenna. In this study we identified two types of vibration-sensitive interneurons with arborizations in the primary sensory area of the JO, namely, the dorsal lobe-interneuron 1 (DL-Int-1) and dorsal lobe-interneuron 2 (DL-Int-2) using intracellular recordings combined with intracellular staining. For visualizing overlapping areas between the JO sensory terminals and the branches of these identified interneurons, the three-dimensional images of the individual neurons were registered into the standard atlas of the honeybee brain (Brandt et al. [2005] J Comp Neurol 492:1-19). Both DL-Int-1 and DL-Int-2 overlapped with the central terminal area of receptor neurons of the JO in the DL. For DL-Int-1 an on-off phasic excitation was elicited by vibrational stimuli applied to the JO when the spontaneous spike frequency was low, whereas tonic inhibition was induced when it was high. Moreover, current injection into a DL-Int-1 led to changes of the response pattern from on-off phasic excitation to tonic inhibition, in response to the vibratory stimulation. Although the vibration usually induced on-off phasic excitation in DL-Int-1, vibration applied immediately after odor stimulation induced tonic inhibition in it. DL-Int-2 responded to vibration stimuli applied to the JO by a tonic burst and were most sensitive to 265 Hz vibration, which is coincident with the strongest frequency of airborne vibrations arising during the waggle dance. These results suggest that DL-Int-1 and DL-Int-2 are related to coding of the duration of the vibration as sensed by the JO.

Tritocerebral tract input to the insect mushroom bodies

Insect mushroom bodies, best known for their role in olfactory processing, also receive sensory input from other modalities. In crickets and grasshoppers, a tritocerebral tract containing afferents from palp mechanosensory and gustatory centers innervates the accessory calyx. The accessory calyx is uniquely composed of Class III Kenyon cells, and was shown by immunohistochemistry to be present sporadically across several insect orders. Neuronal tracers applied to the source of tritocerebral tract axons in several species of insects demonstrated that tritocerebral tract innervation of the mushroom bodies targeted the accessory calyx when present, the primary calyces when an accessory calyx was not present, or both. These results suggest that tritocerebral tract input to the mushroom bodies is likely ubiquitous, reflecting the importance of gustation for insect behavior. The scattered phylogenetic distribution of Class III Kenyon cells is also proposed to represent an example of generative homology, in which the developmental program for forming a structure is retained in all members of a lineage, but the program is not "run" in all branches.

Does an insect's unconditioned response to sucrose reveal expectations of reward?

We asked whether and how a sequence of a honeybee's experience with different reward magnitudes changes its subsequent unconditioned proboscis extension response (PER) to sucrose stimulation of the antennae, 24 hours after training, in the absence of reward, and under otherwise similar circumstances. We found that the bees that had experienced an increasing reward schedule extended their probosces earlier and during longer periods in comparison to bees that had experienced either decreasing or constant reward schedules, and that these effects at a later time depend upon the activation of memories formed on the basis of a specific property of the experienced reward, namely, that its magnitude increased over time. An anticipatory response to reward is typically thought of as being rooted in a subject's expectations of reward. Therefore our results make us wonder to what extent a long-term ‘anticipatory’ adjustment of a honeybee's PER is based upon an expectation of reward. Further experiments will aim to elucidate the neural substrates underlying reward anticipation in harnessed honeybees.

Involvement of NO-synthase and nicotinic receptors in learning in the honey bee

Restrained worker honey bees (Apis mellifera) are one of the main models for the comparative study of learning and memory processes. Bees easily learn to associate a sucrose reward to antennal tactile scanning of a small metal plate (associative learning). Their proboscis extension response can also be habituated through repeated sucrose stimulations (non-associative learning). We studied the role of nitric oxide synthase and nicotinic acetylcholine receptors in these two forms of learning. The nicotinic antagonist MLA or the nitric oxide synthase inhibitor l-NAME impaired the formation of tactile associative long-term memory that specifically occurs during multiple-trial training; however these drugs had no effect on single-trial training. None of the drugs affected retrieval processes. These pharmacological results are consistent with data previously obtained with olfactory conditioning and indicate that MLA-sensitive nicotinic receptors and NO-synthase are specifically involved in long-term memory. MLA and l-NAME both reduced the number of trials required for habituation to occur. This result suggests that a reduction of cholinergic nicotinic neurotransmission promotes PER habituation in the honey bee.

Topographic organization of sensory afferents of Johnston's organ in the honeybee brain

Johnston's organ (JO) in insects is a multicellular mechanosensory organ stimulated by movement of the distal part of the antenna. In honeybees JO is thought to be a primary sensor detecting air-particle movements caused by the waggling dance of conspecifics. In this study projection patterns of JO afferents within the brain were investigated. About 720 somata, distributed around the periphery of the second segment of the antenna (pedicel), were divided into three subgroups based on their soma location: an anterior group, a ventral group, and a dorsal group. These groups sent axons to different branches (N2 to N4) diverged from the antennal nerve. Dye injection into individual nerve branches revealed that all three groups of afferents, having fine collaterals in the dorsal lobe, sent axons broadly through tracts T6I, T6II, and T6III to terminate ipsilaterally in the medial posterior protocerebral lobe, the dorsal region of the subesophageal ganglion, and the central posterior protocerebral lobe, respectively. Within these termination fields only axon terminals running in T6I were characterized by thick processes with large varicosities. Differential staining using fluorescent dyes revealed that the axon terminals of the three groups were spatially segregated, especially in T6I, showing some degree of somatotopy. This spatial segregation was not observed in axon terminals running in other tracts. Our results show that projection patterns of JO afferents in the honeybee brain fundamentally resemble those in the dipteran brain. The possible roles of extensive termination fields of JO afferents in parallel processings of mechanosensory signals are discussed.

Central gustatory projections and side-specificity of operant antennal muscle conditioning in the honeybee

Gustatory stimuli to the antennae, especially sucrose, are important for bees and are employed in learning paradigms as unconditioned stimulus. The present study identified primary antennal gustatory projections in the bee brain and determined the impact of stimulation of the antennal tip on antennal muscle activity and its plasticity. Central projections of antennal taste hairs contained axons of two morphologies projecting into the dorsal lobe, which is also the antennal motor centre. Putative mechanosensory axons arborised in a dorso-lateral area. Putative gustatory axons projected to a ventro-medial area. Bees scan gustatory and mechanical stimuli with their antennae using variable strategies but sensory input to the motor system has not been investigated in detail. Mechanical, gustatory, and electrical stimulation of the ipsilateral antennal tip were found to evoke short-latency responses in an antennal muscle, the fast flagellum flexor. Contralateral gustatory stimulation induced smaller responses with longer latency. The activity of the fast flagellum flexor was conditioned operantly by pairing high muscle activity with ipsilateral antennal sucrose stimulation. A proboscis reward was unnecessary for learning. With contralateral antennal sucrose stimulation, conditioning was unsuccessful. Thus, muscle activity induced by gustatory stimulation was important for learning success and conditioning was side-specific.

Innervation pattern of suboesophageal ventral unpaired median neurones in the honeybee brain

In honeybees (Apis mellifera), the biogenic amine octopamine has been shown to play a role in associative and non-associative learning and in the division of labour in the hive. Immunohistochemical studies indicate that the ventral unpaired median (VUM) neurones in the suboesophageal ganglion (SOG) are putatively octopaminergic and therefore might be involved in the octopaminergic modulation of behaviour. In contrast to our knowledge about the behavioural effects of octopamine, only one neurone (VUMmx1) has been related to a behavioural effect (the reward function during olfactory learning). In this study, we have investigated suboesophageal VUM neurones with fluorescent dye-tracing techniques and intracellular recordings combined with intracellular staining. Ten different VUM neurones have been found including six VUM neurones innervating neuropile regions of the brain and the SOG exclusively (central VUM neurones) and four VUM neurones with axons in peripheral nerves (peripheral VUM neurones). The central VUM neurones innervate the antennal lobes, the protocerebral lobes (including the lateral horn) and the mushroom body calyces. Of these, a novel mandibular VUM neurone, VUMmd1, exhibits the same branching pattern in the brain as VUMmx1 and responds to sucrose and odours in a similar way. The peripheral VUM neurones innervate the antennal and the mandibular nerves. In addition, we describe one labial unpaired median neurone with a dorsal cell body, DUMlb1. The possible homology between the honeybee VUM neurones and the unpaired median neurones in other insects is discussed.

Two closely located areas in the suboesophageal ganglion and the tritocerebrum receive projections of gustatory receptor neurons located on the antennae and the proboscis in the moth Heliothis virescens

Sucrose stimulation of gustatory receptor neurons on the antennae, the tarsi, and the mouthparts elicits the proboscis extension reflex in many insect species, including lepidopterans. The sensory pathways involved in this reflex have only partly been investigated, and in hymenopterans only. The present paper concerns the pathways of the gustatory receptor neurons on the antennae and on the proboscis involved in the proboscis extension reflex in the moth Heliothis virescens (Lepidoptera; Noctuidae). Fluorescent dyes were applied to the contact chemosensilla, sensilla chaetica on the antennae, and sensilla styloconica on the proboscis, permitting tracing of the axons of the gustatory receptor neurons in the central nervous system. The stained axons showed projections from the two appendages in two closely located but distinct areas in the suboesophageal ganglion (SOG)/tritocerebrum. The projections of the antennal gustatory receptor neurons were located posterior-laterally to those from the proboscis. Electrophysiological recordings from the receptor neurons in s. chaetica during mechanical and chemical stimulation were performed, showing responses of one mechanosensory and of several gustatory receptor neurons. Separate neurons showed excitatory responses to sucrose and sinigrin. The effect of these two tastants on the proboscis extension reflex was tested by repeated stimulations with solutions of the two compounds. Whereas sucrose elicited extension in 100% of the individuals in all repetitions, sinigrin elicited extension in fewer individuals, a number that decreased with repeated stimulation.

A confined taste area in a lepidopteran brain

Knowledge about the neuronal pathways of the taste system is interesting both for studying taste coding and appetitive learning of odours. We here present the morphology of the sensilla styloconica on the proboscis of the moth Heliothis virescens and the projections of the associated receptor neurones in the central nervous system. The morphology of the sensilla was studied by light microscopy and by scanning- and transmission electron microscopy. Each sensillum contains three or four sensory neurones; one mechanosensory and two or three chemosensory. The receptor neurones were stained with neurobiotin tracer combined with avidin-fluorescein conjugate, and the projections were viewed in a confocal laser-scanning microscope. The stained axons entered the suboesophageal ganglion via the maxillary nerves and were divided into two categories based on their projection pattern. Category one projected exclusively ipsilaterally in the dorsal suboesophageal ganglion/tritocerebrum and category two projected bilaterally and more ventrally in the suboesophageal ganglion confined to the anterior surface of the neuropil. The bilateral projecting neurones had one additional branch terminating ipsilaterally in the dorsal suboesophageal ganglion/tritocerebrum. A possible segregation of the two categories of projections as taste and mechanosensory is discussed.

Insect odor and taste receptors

Insect odor and taste receptors are highly sensitive detectors of food, mates, and oviposition sites. Following the identification of the first insect odor and taste receptors in Drosophila melanogaster, these receptors were identified in a number of other insects, including the malaria vector mosquito Anopheles gambiae; the silk moth, Bombyx mori; and the tobacco budworm, Heliothis virescens. The chemical specificities of many of the D. melanogaster receptors, as well as a few of the A. gambiae and B. mori receptors, have now been determined either by analysis of deletion mutants or by ectopic expression in in vivo or heterologous expression systems. Here we discuss recent advances in our understanding of the molecular and cellular basis of odor and taste coding in insects.

Different physiological properties in a pool of mandibular closer motor neurons in a caterpillar, Bombyx mori

To determine whether a large pool of mandibular closer motor neurons in an insect consists of different functional groups, the physiological properties of the motor neurons were investigated by intracellular recording techniques. The mandibular closer muscle of a caterpillar, Bombyx mori, was innervated by at least 11 motor neurons with large somata and 2-3 neurons with small somata. The large 11 motor neurons could be divided into slow and fast types on the basis of thresholds of neurons and frequency of soma spikes by current injections. The fast motor neurons usually evoked relatively larger muscle potentials, whereas the slow neurons evoked smaller potentials. The slow neurons spiked while the mandible was held in a fully closed position, whereas the fast neurons did not. The slow neurons did not spike while the mandible was free to move. These results suggest different physiological properties among closer motor neurons, supporting the hypothesis that the closer motor neurons in a large neuron pool can be divided into different functional groups.

A new ascending sensory tract to the calyces of the honeybee mushroom body, the subesophageal-calycal tract

The mushroom bodies of the honeybee are important neuropils for learning and memory. Therefore, knowledge about their input and output connections is essential to understanding how these neuropils function. A newly described input tract to the mushroom body is presented here, which is called the subesophageal-calycal tract (SCT) and connects the subesophageal ganglion with the calyces of the mushroom bodies. The neuronal somata of the SCT neurons lie in one cluster between the lobula of the optic lobe and a neuropil area that is formed from the fusion of the tritocerebrum and the subesophageal ganglion. Within the subesophageal ganglion, the dendritic fibers of SCT neurons overlap with terminals of sensory neurons from the proboscis. Therefore, we conclude that the SCT neurons might process gustatory and mechanosensory information from the proboscis. Individual SCT neurons receive unilateral input within the subesophageal ganglion and may connect to either the ipsilateral or the contralateral mushroom body. On their way to the mushroom bodies, the SCT neuron axons meet the roots of the antennocerebralis tracts (ACTs) and from this point follow the same path as the median ACT neurons for a short distance. Within the calyces, the SCT neurons innervate two separate areas, a small area within the dorsal collar just below the lip and a part of the basal ring. Double-labeling experiments show that the projections of the SCT neurons do not overlap with the projections of the olfactory projection neurons and visual projection neurons from the dorsal medulla. The possible function of the SCT neurons and the relation of the SCT to known input tracts of the mushroom bodies in other insects are discussed.

Identification and localization of the nicotinic acetylcholine receptor alpha3 mRNA in the brain of the honeybee, Apis mellifera

The nicotinic acetylcholine receptors are ligand-gated ion channels responsible for rapid neurotransmission and are target sites for pesticides in insects. In the honeybee Apis mellifera, pharmacological and electrophysiological studies have shown that different nicotinic acetylcholine receptor subtypes may exist in the brain. Here, we have identified a honeybee cDNA that encodes a 537 amino acid protein with features typical of nicotinic acetylcholine receptor alpha subunit, and sequence homology to human alpha3. In situ hybridization on cryosections shows that the Apisalpha3 mRNA is differently expressed in larvae and adult. In larvae, Apisalpha3 mRNA expression is restricted to the suboesophageal ganglia. In adult, it is further expressed in the optic lobes, the dorsal lobes, the antennal lobes and the calyces of mushroom bodies. Together our results suggest that Apisalpha3 shows a controlled expression pattern during development.

Motor control of the mandible closer muscle in ants

Despite their simple design, ant mandible movements cover a wide range of forces, velocities and amplitudes. The mandible is controlled by the mandible closer muscle, which is composed of two functionally distinct subpopulations of muscle fiber types: fast fibers (short sarcomeres) and slow ones (long sarcomeres). The entire muscle is controlled by 10-12 motor neurons, 4-5 of which exclusively supply fast muscle fibers. Slow muscle fibers comprise a posterior and an antero-lateral group, each of which is controlled by 1-2 motor neurons. In addition, 3-4 motor neurons control all muscle fibers together. Simultaneous recordings of muscle activity and mandible movement reveal that fast movements require rapid contractions of fast muscle fibers. Slow and subtle movements result from the activation of slow muscle fibers. Forceful movements are generated by simultaneous co-activation of all muscle fiber types. Retrograde tracing shows that most dendritic arborizations of the different sets of motor neurons share the same neuropil in the subesophageal ganglion. In addition, fast motor neurons and neurons supplying the lateral group of slow closer muscle fibers each invade specific parts of the neuropil that is not shared by the other motor neuron groups. Some bilateral overlap between the dendrites of left and right motor neurons exists, particularly in fast motor neurons. The results explain how a single muscle is able to control the different movement parameters required for the proper function of ant mandibles.

Mandible movements in ants

Ants use their mandibles for almost any task, including prey-catching, fighting, leaf-cutting, brood care and communication. The key to the versatility of mandible functions is the mandible closer muscle. In ants, this muscle is generally composed of distinct muscle fiber types that differ in morphology and contractile properties. Fast contracting fibers have short sarcomeres (2-3 microm) and attach directly to the closer apodeme, that conveys the muscle power to the mandible joint. Slow but forceful contracting fibers have long sarcomeres (5-6 microm) and attach to the apodeme either directly or via thin thread-like filaments. Volume proportions of the fiber types are species-specific and correlate with feeding habits. Two biomechanical models explain why species that rely on fast mandible strikes, such as predatory ants, have elongated head capsules that accommodate long muscle fibers directly attached to the apodeme at small angles, whereas species that depend on forceful movements, like leaf-cutting ants, have broader heads and many filament-attached fibers. Trap-jaw ants feature highly specialized catapult mechanisms. Their mandible closing is known as one of the fastest movements in the animal kingdom. The relatively large number of motor neurons that control the mandible closer reflects the importance of this muscle for the behavior of ants as well as other insects.

Peripheral and central structures involved in insect gustation

Studies in insect gustation have a long history in general physiology, particularly with work on fly labellar and tarsal sensilla and in the general field of insect-plant interactions, where work on immature Lepidoptera and chrysomelid beetles has been prominent. Much more emphasis has been placed on the physiological characteristics of the sensory cells than on the central cellular mechanisms of taste processing. This is due to the fairly direct access for physiological experimentation presented by many taste sensilla and to the obvious importance of tastants in insect feeding and oviposition behaviour. In some of the insect models used for gustatory studies, advances have been made in understanding the basic morphology of the central neuropils involved in the first stages of taste processing. There is much less known about the physiology of interneurons involved. In this review, we concentrate on four insect models (Manduca sexta, Drosophila melanogaster, Neobellieria bullata (and other large flies), and Apis mellifera) to summarize morphological knowledge of peripheral and central aspects of insect gustation. Our views of current interpretations of available data are discussed and some important areas for future research are highlighted.

The control of mandible movements in the ant Odontomachus

Ants use their mandibles to manipulate many different objects including food, brood and nestmates. Different tasks require the modification of mandibular force and speed. Besides normal mandible movements the trap-jaw ant Odontomachus features a particularly fast mandible reflex during which both mandibles close synchronously within 3 ms. The mandibular muscles that govern mandible performance are controlled by four opener and eight closer motor neurons. During slow mandible movements different motor units can be activated successively, and fine tuning is assisted by co-activation of the antagonistic muscles. Fast and powerful movements are generated by the additional activation of two particular motor units which also contribute to the mandible strike. The trap-jaw reflex is triggered by a fast trigger muscle which is derived from the mandible closer. Intracellular recording reveals that trigger motor neurons can generate regular as well as particularly large postsynaptic potentials, which might be passively propagated over the short distance to the trigger muscle. The trigger motor neurons are dye-coupled and receive input from both sides of the body without delay, which ensures the synchronous release of both mandibles.

Behavioural pharmacology of dopamine, serotonin and putative aminergic ligands in the mushroom bodies of the honeybee (Apis mellifera)

The proboscis extension response (PER) which can be elicited in bees by stimulating one antenna with water vapour, was used to quantify the effects of dopamine, serotonin (5-HT) and putative receptor ligands in the mushroom body of the bee. The drugs were microinjected into the alpha-lobe of the mushroom body in one brain hemisphere. Injection of dopamine reduces the water vapour-elicited PER significantly. The effects of dopamine are limited to the treated side. Injection of 5-HT has similar effects to dopamine. The effects of 5-HT are apparent on the treated and partly also on the contralateral side. Significant effects for dopamine on the treated side were found when the concentration in the injected drop was 10(-7) M. For 5-HT significant effects on the treated side were apparent for concentrations of 10(-8) M. Putative dopamine and 5-HT receptor ligands were injected alone and coinjected with the amines. Two ligands with dopamine-antagonistic effects were found: buspirone > spiperone, while lisuride, sulpiride, chlorpromazine, SCH 23390, butaclamol and haloperidol had no dopamine-antagonistic effects. All tested putative 5-HT receptor ligands had significant 5-HT-antagonistic effects: butaclamol > methysergide > lisuride > cyproheptadine > SCH 23390. Good correlations between the behavioural data and in vitro radioligand binding studies were found for 5-HT receptor ligands, while there exist only partial correlations for dopamine receptor ligands.

The neural basis of associative reward learning in honeybees

Appetitive learning of food-predicting stimuli, an essential part of foraging behavior in honeybees, follows the rules of associative learning. In the learning of odors as reward-predicting stimuli, an individual neuron, one of a small group of large ascending neurons that serve principal brain neuropiles, mediates the reward and has experience-dependent response properties. This implies that this neuron functions as an integral part of associative memory, might underlie more complex features of learning, and could participate in the implementation of learning rules. Moreover, its structural properties suggest that it organizes the interaction of functionally different neural nets during learning and experience-dependent behavior.

A neuroanatomical map of the suboesophageal and prothoracic ganglia of the honey bee ( Apis mellifera )

The basic organization of the prothoracic and suboesophageal ganglia of the honey bee is described from transverse, horizontal, and sagittal serial sections by using a variety of staining methods. Drawings of sections demonstrate neuropils, tracts, commissures, clusters of somata and other conspicuous landmarks. These provide the elements of a map of the ganglion in which specific neurons can be related to their structural context. The prothoracic ganglion contains nine longitudinal tracts in each hemiganglion. Six main dorsal and four ventral commissures link the two halves of the ganglion together. Other conspicuous structures are the ventral association centre and a T-tract: a ring tract is not strongly developed. The suboesophageal ganglion (SOG) is a fusion of the mandibular, maxillary, and labial neuromeres, the latter clearly showing the same basic organization as that found in the prothoracic ganglion. The more anterior mandibular and maxillary neuromeres successively decrease in volume compared with the posterior labial neuromere. Whereas the ventral parts of the neuromeres are demarcated from each other by prominent midline tracts, the dorsal part of each neuromere is confluent with the next. This may be, in part, due to the tilt of about 90° in the anterior-posterior axis (neuraxis) between the suboesophageal ganglion and the brain. Presumably, in morphogenesis this results in a compression of the dorsal parts of the suboesophageal ganglion. Eight longitudinal tracts run through each half of the ganglion. The number of main dorsal commissures declines from six in the labial to five in the maxillary and three in the mandibular neuromere. There are five ventral commissures in the maxillary, and four in both the mandibular and labial neuromeres. The organization of the suboesophageal and pro­thoracic ganglia in the honey bee is compared with those found in the cockroach, locust and cricket. The data support the hypothesis that insect ganglia have a common morphological building plan.

Morphology of neuroglia in the antennal lobes and mushroom bodies of the brain of the honeybee

We investigated the distribution and anatomical organization of glial cells in the antennal lobes and mushroom bodies of the honeybee. Reconstructions from serial sections, prepared according to the ethyl gallate method, revealed the entire morphology of glial cells in neuropiles, tracts, and the soma rind. The distribution of the glial cell bodies in the neuropiles was derived from the staining of cell nuclei with a fluorescent dye. There are glial cells of different shape in the soma rind which are wrapped around the neuronal cell bodies of the antennal lobes and the Kenyon cells of the mushroom bodies. Glial cells surround neuropilar areas such as the external and lateral sides of the glomeruli of the antennal lobes. Whereas we could not detect glia in the glomerular neuropile, glial cells with long processes are located in the core of the antennal lobe. Extensions of these glial cells also invade tracts containing the olfactory projection neurons. A layer of glial cells separates the mushroom body neuropile from the surrounding protocerebral neuropile. The neuropile of the mushroom bodies is clearly compartmented by glial cells. There is a high density of astrocyte-like glia in a column of the pedunculus which can be followed to the ventral part of the alpha-lobe. A network of mushroom body intrinsic glial cells separates the alpha-lobe from the beta-lobe and the pedunculus. This anatomical description of glial cell types in olfactory information processing pathways of an insect brain provides a framework for further physiological studies of neuroglia in dissociated cell culture.

Anatomy of the antennal motoneurons in the brain of the honeybee (Apis mellifera)

This paper describes the morphology and location of the cerebral motoneurons that control the movement of the antennae in the honeybee. The position of each antenna is controlled by two muscle systems; the basal segment (scape) is moved by four muscles within the head capsule, and two muscles within the scape control the distal segments (flagellum) of the antenna. The motor system of the scape is controlled by nine motoneurons, and that of the flagellum by six motoneurons. All of these motoneurons share the dorsal lobe as a common projection area where their dendritic fields overlap extensively. These motoneurons do not have contralateral projections. The cell bodies of the antennal motoneurons are located in the soma layer lateral to the dorsal lobe. The somata for each muscle system are arranged in three clusters; two clusters are located in a region of the cortex dorsal to the dorsal lobe and one cluster is located in the cortex ventral to the dorsal lobe. In the cortex dorsal to the dorsal lobe, one cluster of each muscle system shares the same region. Altogether five groups of cell bodies can be distinguished. Double labeling of the motoneurons and presumptive mechanosensory primary antennal afferents with fluorescent dyes has shown that there is an extensive overlap of axonal projections of antennal mechanosensory afferents with dendritic fields of antennal motoneurons.

Food-induced arousal and nonassociative learning in honeybees: dependence of sensitization on the application site and duration of food stimulation

Stimulus-induced arousal (sensitization) of a component of appetitive behavior in honeybees, the proboscis extension reflex (PER), was used to investigate different aspects of nonassociative memory. The sensitizing stimulus (sucrose solution) was applied to one antenna, as a compound to antenna and proboscis, and to the proboscis. Stimulus duration was either 1 or 3 s. Sensitization was evaluated by monitoring PER toward an odor before (pretest) and after (test) application of the sensitizing stimulus. All responses were quantified by recording from muscle M17 which represents the motor program of PER. Data were analyzed by determining (1) the response probability to the odor and (2) the response strength by determining the number of M17-spikes and the percentage of licking bees per trial. The analysis of the response probability led to two main results: the proportion of animals responding to the test odor depended on stimulus site, and, dependent on stimulus site, a longer application of the sensitizing stimulus resulted in different sensitization rates. The strength of the sensitized response, however, did not correspond to the probability, with which it was elicited, but rather to the strength of the response to the sensitizing stimulus itself. Furthermore, the three groups were not equally affected by the short and long stimulation. The analysis of the proportion of animals licking during test confirmed the data obtained using the number of muscle spikes as a measure of response strength. These results suggest an internal evaluation of the sensitizing stimulus depending on its quality and intensity. The differential effects after antennal and proboscis stimulation may be realized via an arousal system which has two independent functions, a permissive one modulating response probability and one modulating response strength. The permissive function of arousal may be regulated via an intervening inhibitory system whose activation critically depends on the functional significance of the arousing stimulus. The content of this short-term form of memory may be interpreted as an expectation for food which is regulated according to experienced consequences.

Central projections of axons from taste hairs on the labellum and tarsi of the blowfly,Phormia regina Meigen

Taste hairs are located on the labellum and tarsi of blowflies. These multimodal hairs consist of four functionally distinct chemoreceptors and a mechanoreceptor. By staining selected multimodal hairs, we sought to identify the central projection patterns of multiple and single axons from those hairs. On each side of the labellum there are 11 "largest" hairs (LH). The neurons associated with the anteriormost (LH-1), posteriormost (LH-11), and one lateral (LH-6) hair on the labellum were stained selectively with cobaltous sulfide. The overall projection pattern in the central nervous system (CNS) for axons from LH-1 and LH-11 is similar and differs markedly from axons from LH-6. At least three individual axon-projection patterns were determined for each labellar hair filled, indicating a partial functional organization for axons from multimodal hairs. One identified axon, the dorsalmost axon, has terminal arborizations that do not differ with the location of its associated hair. Another axon, thicker than the others, projects to a region that is distinct from the four thin axons. Within this region the arborizations of the thick axons occupy different areas depending on the location of their associated hair. Neurons from the largest hairs on the distalmost tarsomere (D5) of each leg were also stained and consisted of one thick and four thin axons. All axons except one thin axon from tarsal D5 hairs terminate in their respective leg neuromeres. The remaining thin axon projects to the suboesophageal ganglion ipsilateral to the hair filled and terminates in the same region as a branch of the labellar dorsalmost axon. These data suggest that axonal arbors from multimodal hairs have a limited functional and somatotopic organization in the blowfly CNS.