Insect Mushroom Bodies Research Articles

Classifying chemical sensor data using GPU-accelerated bio-mimetic neuronal networks based on the insect olfactory system

Chemosensing “e-nose” technology has great potential applications in everyday life, ranging from drug detection to food quality assessment and even the diagnosis of illness. However, odour detection and classification remains a highly challenging domain, characterized by high dimensionality, unknown organization of the vast “odourant space” of volatile chemicals and complex, turbulent odour plumes. To compound these difficulties, current sensor technology continues to exhibit distinct shortcomings in speed, sensitivity, selectivity, recovery, and drift avoidance. In the research reported here we turn to a range of recent neuronal models [1-6] that were developed to describe the insect olfactory system, and have been shown to perform well across a range of classic, static classification tasks such as the MNIST handwritten digit set and the Sigma-Aldrich scent database. The insect olfactory system has been extensively studied and has been shown to be both fast and highly effective in complex natural conditions despite its limited size and complexity (when compared to the mammalian system) [2]. Insects such as moths, honey bees, locusts and fruit flies are capable of odour detection and classification tasks well beyond the abilities of current e-nose technology and machine learning algorithms [6]. We present results of applying an insect-inspired approach to the design of a learning spiking neural network that receives synchronized time series data from up to 12 metal-oxide based gas sensors, comprising an optimised [7] combination of classical doped tin oxide and novel zeolite-coated chromium titanium oxide sensors. We have collected sample data sets for classification tasks ranging from “easy” (single chemical identification presented under laboratory conditions) through to “hard” (identification of indicators of infectious diseases in breath samples taken from patients). To address slow sensor response we consider a range of transient-based processing before applying self-organisation techniques to most effectively locate “virtual receptors” (VR) in sensor space. We look to address decorrelation (separation in feature space) and the supervised association of responses with rewards by using correlates of the insect antennal lobe (AL) and mushroom body (MB) structures whilst applying reward-based spike-timing dependent plasticity mechanisms. Classification accuracy is compared with support vector machine (SVM) learning which we also look to match for speed through the use of GPU accelerated neural simulation [5] via the NVidia CUDA TM -based GeNN platform (http://sourceforge.net/projects/genn/).

The insect mushroom body, an experience-dependent recoding device

The insect mushroom body is a higher order integration center involved in cross-sensory integration and memory formation. The relatively large mushroom bodies of social Hymenoptera (e.g. bees) have been related to the demands of a social system and the neural processes required to allow the animal to navigate in an ever-changing environment. Here I review studies aiming to elucidate the neural processes that take place at the input and the output sites of the mushroom bodies and that underlie cross-sensory integration, associative learning, memory storage and retrieval. Highly processed sensory information is received at modality-specific compartments of the input site, the calyx. The large number of intrinsic neurons of the mushroom body receive multiple sensory inputs establishing combinations of processed sensory stimuli. A matrix-like memory structure characterizes the dendritic area of the intrinsic neurons allowing storage of rich combinations of sensory information. The rather small number of extrinsic neurons read out from multiple intrinsic neurons, thereby losing their sensory coding properties. The response properties of these neurons change according to the value of stimulus combinations experienced. It is concluded that the mushroom bodies transform the highly dimensional sensory coding space into a low dimensional coding space of value-based information. A model of such an experience-dependent recoding device is presented and compared with the available data.

Effect of GABAergic inhibition on odorant concentration coding in mushroom body intrinsic neurons of the honeybee

Kenyon cells, the intrinsic neurons of the insect mushroom body, have the intriguing property of responding in a sparse way to odorants. Sparse neuronal codes are often invariant to changes in stimulus intensity and duration, and sparse coding often depends on global inhibition. We tested if this is the case for honeybees' Kenyon cells, too, and used in vivo Ca²⁺ imaging to record their responses to different odorant concentrations. Kenyon cells responded not only to the onset of odorant stimuli (ON responses), but also to their termination (OFF responses). Both, ON and OFF responses increased with increasing odorant concentration. ON responses were phasic and invariant to the duration of odorant stimuli, while OFF responses increased with increasing odorant duration. Pharmacological blocking of GABA receptors in the brain revealed that ionotropic GABA(A) and metabotropic GABA(B) receptors attenuate Kenyon cells' ON responses without changing their OFF responses. Ionotropic GABA(A) receptors attenuated Kenyon cell ON responses more strongly than metabotropic GABA(B) receptors. However, the response dynamic, temporal resolution and paired-pulse depression did not depend on GABA(A) transmission. These data are discussed in the context of mechanisms leading to sparse coding in Kenyon cells.

Modeling the insect mushroom bodies: Application to a delayed match-to-sample task

Despite their small brains, insects show advanced capabilities in learning and task solving. Flies, honeybees and ants are becoming a reference point in neuroscience and a main source of inspiration for autonomous robot design issues and control algorithms. In particular, honeybees demonstrate to be able to autonomously abstract complex associations and apply them in tasks involving different sensory modalities within the insect brain. Mushroom Bodies (MBs) are worthy of primary attention for understanding memory and learning functions in insects. In fact, even if their main role regards olfactory conditioning, they are involved in many behavioral achievements and learning capabilities, as has been shown in honeybees and flies. Owing to the many neurogenetic tools, the fruit fly Drosophila became a source of information for the neuroarchitecture and biochemistry of the MBs, although the MBs of flies are by far simpler in organization than their honeybee orthologs. Electrophysiological studies, in turn, became available on the MBs of locusts and honeybees. In this paper a novel bio-inspired neural architecture is presented, which represents a generalized insect MB with the basic features taken from fruit fly neuroanatomy. By mimicking a number of different MB functions and architecture, we can replace and improve formerly used artificial neural networks. The model is a multi-layer spiking neural network where key elements of the insect brain, the antennal lobes, the lateral horn region, the MBs, and their mutual interactions are modeled. In particular, the model is based on the role of parts of the MBs named MB-lobes, where interesting processing mechanisms arise on the basis of spatio-temporal pattern formation. The introduced network is able to model learning mechanisms like olfactory conditioning seen in honeybees and flies and was found able also to perform more complex and abstract associations, like the delayed matching-to-sample tasks known only from honeybees. A biological basis of the proposed model is presented together with a detailed description of the architecture. Simulation results and remarks on the biological counterpart are also reported to demonstrate the possible applications of the designed computational model. Such neural architecture, able to autonomously learn complex associations is envisaged to be a suitable basis for an immediate implementation within an robot control architecture.

Fine structural organization of the hemiellipsoid body of the land hermit crab, Coenobita clypeatus

Electron microscopical observations of the hemiellipsoid bodies of the land hermit crab Coenobita clypeatus resolve microglomerular synaptic complexes that are comparable to those observed in the calyces of insect mushroom bodies and which characterize olfactory inputs onto intrinsic neurons. In an adult hermit crab, intrinsic neurons and one class of efferent neurons originate from neuronal somata of globuli cells covering the hemiellipsoid bodies. Counts of their nucleoli show that about 120,000 globuli cells supply each hemiellipsoid body in an adult hermit crab. This number is comparable to the number of globuli cells supplying mushroom bodies of certain insects, such as honey bees and cockroaches. Counts of axons in tracts leading from the olfactory lobes to the hemiellipsoid bodies resolve 20,000 afferent axons, however, an order of magnitude greater than known for any insect. These afferent axons provide numerous swollen varicosities, each presynaptic to many small profiles, and thus comparable to the microglomeruli that characterize insect mushroom body calyces. Also, common to mushroom bodies and hemiellipsoid bodies are arrangements of intrinsic neurons, afferent neurons containing dense core vesicles, and systems of serial synaptic complexes that relate to postsynaptic profiles of efferent neurons. Together, the ultrastructural organization of the hemiellipsoid bodies of C. clypeatus supports the proposition that this center may share a common origin with the insect mushroom body despite obvious divergent evolution of overall shape.

Neuronal organization of the hemiellipsoid body of the land hermit crab, Coenobita clypeatus: Correspondence with the mushroom body ground pattern

Malacostracan crustaceans and dicondylic insects possess large second-order olfactory neuropils called, respectively, hemiellipsoid bodies and mushroom bodies. Because these centers look very different in the two groups of arthropods, it has been debated whether these second-order sensory neuropils are homologous or whether they have evolved independently. Here we describe the results of neuroanatomical observations and experiments that resolve the neuronal organization of the hemiellipsoid body in the terrestrial Caribbean hermit crab, Coenobita clypeatus, and compare this organization with the mushroom body of an insect, the cockroach Periplaneta americana. Comparisons of the morphology, ultrastructure, and immunoreactivity of the hemiellipsoid body of C. clypeatus and the mushroom body of the cockroach P. americana reveal in both a layered motif provided by rectilinear arrangements of extrinsic and intrinsic neurons as well as a microglomerular organization. Furthermore, antibodies raised against DC0, the major catalytic subunit of protein kinase A, specifically label both the crustacean hemiellipsoid bodies and insect mushroom bodies. In crustaceans lacking eyestalks, where the entire brain is contained within the head, this antibody selectively labels hemiellipsoid bodies, the superior part of which approximates a mushroom body's calyx in having large numbers of microglomeruli. We propose that these multiple correspondences indicate homology of the crustacean hemiellipsoid body and insect mushroom body and discuss the implications of this with respect to the phylogenetic history of arthropods. We conclude that crustaceans, insects, and other groups of arthropods share an ancestral neuronal ground pattern that is specific to their second-order olfactory centers.

Brain development in an insect with extensive maternal care, the tsetse fly Glossina morsitans (Diptera: Glossinidae)

Brain development in an insect with extensive maternal care, the tsetse fly Glossina morsitans (Diptera: Glossinidae)

A Putative Vesicular Transporter Expressed in Drosophila Mushroom Bodies that Mediates Sexual Behavior May Define a Neurotransmitter System

Vesicular transporters are required for the storage ofall classical and amino acid neurotransmitters in synaptic vesicles. Some neurons lack known vesicular transporters, suggesting additional neurotransmitter systems remain unidentified. Insect mushroom bodies (MBs) are critical for several behaviors, including learning, but the neurotransmitters released by the intrinsic Kenyon cells (KCs) remain unknown. Likewise, KCs do not express a known vesicular transporter. We report the identification of a novel Drosophila gene portabella (prt) that is structurally similar to known vesicular transporters. Both larval and adult brains express PRT in the KCs of the MBs. Additional PRT cells project to the central complex and optic ganglia. prt mutation causes an olfactory learning deficit and an unusual defect in the male's position during copulation that is rescued by expression in KCs. Because prt is expressed in neurons that lack other known vesicular transporters or neurotransmitters, it may define a previously unknown neurotransmitter system responsible for sexual behavior and a component of olfactory learning.

Are mushroom bodies cerebellum-like structures?

The mushroom bodies are distinctive neuropils in the protocerebral brain segments of many protostomes. A defining feature of mushroom bodies is their intrinsic neurons, masses of cytoplasm-poor globuli cells that form a system of lobes with their densely-packed, parallel-projecting axon-like processes. In insects, the role of the mushroom bodies in olfactory processing and associative learning and memory has been studied in depth, but several lines of evidence suggest that the function of these higher brain centers cannot be restricted to these roles. The present account considers whether insight into an underlying function of mushroom bodies may be provided by cerebellum-like structures in vertebrates, which are similarly defined by the presence of masses of tiny granule cells that emit thin parallel fibers forming a dense molecular layer. In vertebrates, the shared neuroarchitecture of cerebellum-like structures has been suggested to underlie a common functional role as adaptive filters for the removal of predictable sensory elements, such as those arising from reafference, from the total sensory input. Cerebellum-like structures include the vertebrate cerebellum, the electrosensory lateral line lobe, dorsal and medial octavolateral nuclei of fish, and the dorsal cochlear nucleus of mammals. The many architectural and physiological features that the insect mushroom bodies share with cerebellum-like structures suggest that it might be fruitful to consider mushroom body function in light of a possible role as adaptive sensory filters. The present account thus presents a detailed comparison of the insect mushroom bodies with vertebrate cerebellum-like structures.

A subpopulation of mushroom body intrinsic neurons is generated by protocerebral neuroblasts in the tobacco hornworm moth, Manduca sexta (Sphingidae, Lepidoptera)

Subpopulations of Kenyon cells, the intrinsic neurons of the insect mushroom bodies, are typically sequentially generated by dedicated neuroblasts that begin proliferating during embryogenesis. When present, Class III Kenyon cells are thought to be the first born population of neurons by virtue of the location of their cell somata, farthest from the position of the mushroom body neuroblasts. In the adult tobacco hornworm moth Manduca sexta, the axons of Class III Kenyon cells form a separate Y tract and dorsal and ventral lobelet; surprisingly, these distinctive structures are absent from the larval Manduca mushroom bodies. BrdU labeling and immunohistochemical staining reveal that Class III Kenyon cells are in fact born in the mid-larval through adult stages. The peripheral position of their cell bodies is due to their genesis from two previously undescribed protocerebral neuroblasts distinct from the mushroom body neuroblasts that generate the other Kenyon cell types. These findings challenge the notion that all Kenyon cells are produced solely by the mushroom body neuroblasts, and may explain why Class III Kenyon cells are found sporadically across the insects, suggesting that when present, they may arise through de novo recruitment of neuroblasts outside of the mushroom bodies. In addition, lifelong neurogenesis by both the Class III neuroblasts and the mushroom body neuroblasts was observed, raising the possibility that adult neurogenesis may play a role in mushroom body function in Manduca.

DrosophilaLarvae Establish Appetitive Olfactory Memories via Mushroom Body Neurons of Embryonic Origin

Insect mushroom bodies are required for diverse behavioral functions, including odor learning and memory. Using the numerically simple olfactory pathway of the Drosophila melanogaster larva, we provide evidence that the formation of appetitive olfactory associations relies on embryonic-born intrinsic mushroom body neurons (Kenyon cells). The participation of larval-born Kenyon cells, i.e., neurons that become gradually integrated in the developing mushroom body during larval life, in this task is unlikely. These data provide important insights into how a small set of identified Kenyon cells can store and integrate olfactory information in a developing brain. To investigate possible functional subdivisions of the larval mushroom body, we anatomically disentangle its input and output neurons at the single-cell level. Based on this approach, we define 10 subdomains of the larval mushroom body that may be implicated in mediating specific interactions between the olfactory pathway, modulatory neurons, and neuronal output.

Modulatory action of acetylcholine on the Na+-dependent action potentials in Kenyon cells isolated from the mushroom body of the cricket brain

Kenyon cells, intrinsic neurons of the insect mushroom body, have been assumed to be a site of conditioning stimulus (CS) and unconditioned stimulus (US) association in olfactory learning and memory. Acetylcholine (ACh) has been implicated to be a neurotransmitter mediating CS reception in Kenyon cells, causing rapid membrane depolarization via nicotinic ACh receptors. However, the long-term effects of ACh on the membrane excitability of Kenyon cells are not fully understood. In this study, we examined the effects of ACh on Na+ dependent action potentials (Na+ spikes) elicited by depolarizing current injection and on net membrane currents under the voltage clamp condition in Kenyon cells isolated from the mushroom body of the cricket Gryllus bimaculatus. Current-clamp studies using amphotericin B perforated-patch recordings showed that freshly dispersed cricket Kenyon cells could produce repetitive Na+ spikes in response to prolonged depolarizing current injection. Bath application of ACh increased both the instantaneous frequency and the amplitudes of Na+ spikes. This excitatory action of ACh on Kenyon cells is attenuated by the pre-treatment of the cells with the muscarinic receptor antagonists, atropine and scopolamine, but not by the nicotinic receptor antagonist mecamylamine. Voltage-clamp studies further showed that bath application of ACh caused an increase in net inward currents that are sensitive to TTX, whereas outward currents were decreased by this treatment. These results indicate that in order to mediate CS, ACh may modulate the firing properties of Na+ spikes of Kenyon cells through muscarinic receptor activation, thus increasing Na conductance and decreasing K conductance.

Long-Term Memory Leads to Synaptic Reorganization in the Mushroom Bodies: A Memory Trace in the Insect Brain?

The insect mushroom bodies (MBs) are paired brain centers which, like the mammalian hippocampus, have a prominent function in learning and memory. Despite convergent evidence for their crucial role in the formation and storage of associative memories, little is known about the mechanisms underlying such storage. In mammals and other species, the consolidation of stable memories is accompanied by structural plasticity involving variations in synapse number and/or size. Here, we address the question of whether the formation of olfactory long-term memory (LTM) could be associated with changes in the synaptic architecture of the MB networks. For this, we took advantage of the modular architecture of the honeybee MB neuropil, where synaptic contacts between olfactory input and MB neurons are segregated into discrete units (microglomeruli) which can be easily visualized and counted. We show that the density in microglomeruli increases as a specific olfactory LTM is formed, while the volume of the neuropil remains constant. Such variation is reproducible and is clearly correlated with memory consolidation, as it requires gene transcription. Thus stable structural synaptic rearrangements, including the growth of new synapses, seem to be a common property of insect and mammalian brain networks involved in the storage of stable memory traces.

Functional and comparative assessements of the octopus learning and memory system

The octopus and its close relatives the cuttlefish and squid are the most advanced of the invertebrates, possessing the largest brains both in weight and cell numbers. Here I review recent studies of the neurophysiological properties of the vertical lobe system (VL) in the cephalopod brain, a system already thought to be dedicated to learning and memory. Summarizing from the point of view of comparative evolution, I relate these results to other systems where anatomical and electrophysiological data are available, mainly the insect mushroom bodies and the mammalian hippocampus. The emerging results suggest that a convergent evolutionary process has resulted in similar neural organization and activity-dependent long-term synaptic plasticity in all these learning and memory systems, even though the invertebrate systems conserve their typical anatomical and electrophysiological features. And finally, functional inferences based on the comparison with the insect mushroom.

Synaptic organization in the adult Drosophila mushroom body calyx

Insect mushroom bodies are critical for olfactory associative learning. We have carried out an extensive quantitative description of the synaptic organization of the calyx of adult Drosophila melanogaster, the main olfactory input region of the mushroom body. By using high-resolution confocal microscopy, electron microscopy-based three-dimensional reconstructions, and genetic labeling of the neuronal populations contributing to the calyx, we resolved the precise connections between large cholinergic boutons of antennal lobe projection neurons and the dendrites of Kenyon cells, the mushroom body intrinsic neurons. Throughout the calyx, these elements constitute synaptic complexes called microglomeruli. By single-cell labeling, we show that each Kenyon cell's claw-like dendritic specialization is highly enriched in filamentous actin, suggesting that this might be a site of plastic reorganization. In fact, Lim kinase (LimK) overexpression in the Kenyon cells modifies the shape of the microglomeruli. Confocal and electron microscopy indicate that each Kenyon cell claw enwraps a single bouton of a projection neuron. Each bouton is contacted by a number of such claw-like specializations as well as profiles of gamma-aminobutyric acid-positive neurons. The dendrites of distinct populations of Kenyon cells involved in different types of memory are partially segregated within the calyx and contribute to different subsets of microglomeruli. Our analysis suggests, though, that projection neuron boutons can contact more than one type of Kenyon cell. These findings represent an important basis for the functional analysis of the olfactory pathway, including the formation of associative olfactory memories.

Fast and Robust Learning by Reinforcement Signals: Explorations in the Insect Brain

We propose a model for pattern recognition in the insect brain. Departing from a well-known body of knowledge about the insect brain, we investigate which of the potentially present features may be useful to learn input patterns rapidly and in a stable manner. The plasticity underlying pattern recognition is situated in the insect mushroom bodies and requires an error signal to associate the stimulus with a proper response. As a proof of concept, we used our model insect brain to classify the well-known MNIST database of handwritten digits, a popular benchmark for classifiers. We show that the structural organization of the insect brain appears to be suitable for both fast learning of new stimuli and reasonable performance in stationary conditions. Furthermore, it is extremely robust to damage to the brain structures involved in sensory processing. Finally, we suggest that spatiotemporal dynamics can improve the level of confidence in a classification decision. The proposed approach allows testing the effect of hypothesized mechanisms rather than speculating on their benefit for system performance or confidence in its responses.

Sequential sparsing by successive adapting neural populations

In the principal cells of the insect mushroom body, the Kenyon cells (KC), olfactory information is represented by a spatially and temporally sparse code. Each odor stimulus will activate only a small portion of neurons and each stimulus leads to only a short phasic response following stimulus onset irrespective of the actual duration of a constant stimulus. The mechanisms responsible for the sparse code in the KCs are yet unresolved. Here, we explore the role of the neuron-intrinsic mechanism of spike-frequency adaptation (SFA) in producing temporally sparse responses to sensory stimulation in higher processing stages. Our single neuron model is defined through a conductance-based integrate-and-fire neuron with spike-frequency adaptation [1]. We study a fully connected feed-forward network architecture in coarse analogy to the insect olfactory pathway. A first layer of ten neurons represents the projection neurons (PNs) of the antenna lobe. All PNs receive a step-like input from the olfactory receptor neurons, which was realized by independent Poisson processes. The second layer represents 100 KCs which converge onto ten neurons in the output layer which represents the population of mushroom body extrinsic neurons (ENs). Our simulation result matches with the experimental observations. In particular, intracellular recordings of PNs show a clear phasic-tonic response that outlasts the stimulus [2] while extracellular recordings from KCs in the locust express sharp transient responses [3]. We conclude that the neuron-intrinsic mechanism is can explain a progressive temporal response sparsening in the insect olfactory system. Further experimental work is needed to test this hypothesis empirically. [1] Muller et. al., Neural Comput, 19(11):2958-3010, 2007. [2] Assisi et. al., Nat Neurosci, 10(9):1176-1184, 2007. [3] Krofczik et. al. Front. Comput. Neurosci., 2(9), 2009.

Intrinsic Membrane Properties and Inhibitory Synaptic Input of Kenyon Cells as Mechanisms for Sparse Coding?

The insect mushroom bodies (MBs) are multimodal signal processing centers and are essential for olfactory learning. Electrophysiological recordings from the MBs' principal component neurons, the Kenyon cells (KCs), showed a sparse representation of olfactory signals. It has been proposed that the intrinsic and synaptic properties of the KC circuitry combine to reduce the firing of action potentials and to generate relatively brief windows for synaptic integration in the KCs, thus causing them to operate as coincidence detectors. To better understand the ionic mechanisms that mediate the KC intrinsic firing properties, we used whole cell patch-clamp recordings from KCs in the adult, intact brain of Periplaneta americana to analyze voltage- and/or Ca2+-dependent inward (ICa, INa) and outward currents [IA, IK(V), IK,ST, IO(Ca)]. In general the currents had properties similar to those of currents in other insect neurons. Certain functional parameters of ICa and IO(Ca), however, had unusually high values, allowing them to assist sparse coding. ICa had a low-activation threshold and a very high current density compared with those of ICa in other insect neurons. Together these parameters make ICa suitable for boosting and sharpening the excitatory postsynaptic potentials as reported in previous studies. IO(Ca) also had a large current density and a very depolarized activation threshold. In combination, the large ICa and IO(Ca) are likely to mediate the strong spike frequency adaptation. These intrinsic properties of the KCs are likely to be supported by their tonic, inhibitory synaptic input, which was revealed by specific GABA antagonists and which contributes significantly to the hyperpolarized membrane potential at rest.

Pavlov's moth: olfactory learning and spike timing–dependent plasticity

Spike-timing dependent plasticity is a favored synaptic mechanism for learning. However, a surprising new study by Ito and colleagues in the insect mushroom body suggests that it cannot account for a paradigmatic form of learning.

Metamorphosis and adult development of the mushroom bodies of the red flour beetle, Tribolium castaneum

The insect mushroom bodies play important roles in a number of higher processing functions such as sensory integration, higher level olfactory processing, and spatial and associative learning and memory. These functions have been established through studies in a handful of tractable model systems, of which only the fruit fly Drosophila melanogaster has been readily amenable to genetic manipulations. The red flour beetle Tribolium castaneum has a sequenced genome and has been subject to the development of molecular tools for the ready manipulation of gene expression; however, little is known about the development and organization of the mushroom bodies of this insect. The present account bridges this gap by demonstrating that the organization of the Tribolium mushroom bodies is strikingly like that of the fruit fly, with the significant exception that the timeline of neurogenesis is shifted so that the last population of Kenyon cells is born entirely after adult eclosion. Tribolium Kenyon cells are generated by two large neuroblasts per hemisphere and segregate into an early-born delta lobe subpopulation followed by clear homologs of the Drosophila gamma, alpha'/beta' and alpha/beta lobe subpopulations, with the larval-born cohorts undergoing dendritic reorganization during metamorphosis. BrdU labeling and immunohistochemical staining also reveal that a proportion of individual Tribolium have variable numbers of mushroom body neuroblasts. If heritable, this variation represents a unique opportunity for further studies of the genetic control of brain region size through the control of neuroblast number and cell cycle dynamics.