A simple connectivity scheme for sparse coding in an olfactory system

Spatial constraints and cell surface molecule depletion structure a randomly connected learning circuit

The brain can represent almost limitless objects to "categorize an unlabeled world" (Edelman, 1989). This feat is supported by expansion layer circuit architectures, in which neurons carrying information about discrete sensory channels make combinatorial connections onto much larger postsynaptic populations. Combinatorial connections in expansion layers are modeled as randomized sets. The extent to which randomized wiring exists in vivo is debated, and how combinatorial connectivity patterns are generated during development is not understood. Non-deterministic wiring algorithms could program such connectivity using minimal genomic information. Here, we investigate anatomic and transcriptional patterns and perturb partner availability to ask how Kenyon cells, the expansion layer neurons of the insect mushroom body, obtain combinatorial input from olfactory projection neurons. Olfactory projection neurons form their presynaptic outputs in an orderly, predictable, and biased fashion. We find that Kenyon cells accept spatially co-located but molecularly heterogeneous inputs from this orderly map, and ask how Kenyon cell surface molecule expression impacts partner choice. Cell surface immunoglobulins are broadly depleted in Kenyon cells, and we propose that this allows them to form connections with molecularly heterogeneous partners. This model can explain how developmentally identical neurons acquire diverse wiring identities.

Interpretable deep learning for deconvolutional analysis of neural signals

The widespread adoption of deep learning to build models that capture the dynamics of neural populations is typically based on "black-box" approaches that lack an interpretable link between neural activity and function. Here, we propose to apply algorithm unrolling, a method for interpretable deep learning, to design the architecture of sparse deconvolutional neural networks and obtain a direct interpretation of network weights in relation to stimulus-driven single-neuron activity through a generative model. We characterize our method, referred to as deconvolutional unrolled neural learning (DUNL), and show its versatility by applying it to deconvolve single-trial local signals across multiple brain areas and recording modalities. To exemplify use cases of our decomposition method, we uncover multiplexed salience and reward prediction error signals from midbrain dopamine neurons in an unbiased manner, perform simultaneous event detection and characterization in somatosensory thalamus recordings, and characterize the responses of neurons in the piriform cortex. Our work leverages the advances in interpretable deep learning to gain a mechanistic understanding of neural dynamics.

Neuroscience: Hacking development to understand sensory discrimination

Fine sensory discrimination abilities are enabled by specific neural circuit architectures. A new study reveals how manipulating particular network parameters in the fly's memory centre, the mushroom body, alters sensory coding and discrimination.

Input density tunes Kenyon cell sensory responses in the Drosophila mushroom body

The ability to discriminate sensory stimuli with overlapping features is thought to arise in brain structures called expansion layers, where neurons carrying information about sensory features make combinatorial connections onto a much larger set of cells. For 50 years, expansion coding has been a prime topic of theoretical neuroscience, which seeks to explain how quantitative parameters of the expansion circuit influence sensory sensitivity, discrimination, and generalization. Here, we investigate the developmental events that produce the quantitative parameters of the arthropod expansion layer, called the mushroom body. Using Drosophila melanogaster as a model, we employ genetic and chemical tools to engineer changes to circuit development. These allow us to produce living animals with hypothesis-driven variations on natural expansion layer wiring parameters. We then test the functional and behavioral consequences. By altering the number of expansion layer neurons (Kenyon cells) and their dendritic complexity, we find that input density, but not cell number, tunes neuronal odor selectivity. Simple odor discrimination behavior is maintained when the Kenyon cell number is reduced and augmented by Kenyon cell number expansion. Animals with increased input density to each Kenyon cell show increased overlap in Kenyon cell odor responses and become worse at odor discrimination tasks.

Connectomics of the Octopus vulgaris vertical lobe provides insight into conserved and novel principles of a memory acquisition network

Here, we present the first analysis of the connectome of a small volume of the Octopus vulgaris vertical lobe (VL), a brain structure mediating the acquisition of long-term memory in this behaviorally advanced mollusk. Serial section electron microscopy revealed new types of interneurons, cellular components of extensive modulatory systems, and multiple synaptic motifs. The sensory input to the VL is conveyed via~1.8 × 106 axons that sparsely innervate two parallel and interconnected feedforward networks formed by the two types of amacrine interneurons (AM), simple AMs (SAMs) and complex AMs (CAMs). SAMs make up 89.3% of the~25 × 106VL cells, each receiving a synaptic input from only a single input neuron on its non-bifurcating primary neurite, suggesting that each input neuron is represented in only~12 ± 3.4SAMs. This synaptic site is likely a 'memory site' as it is endowed with LTP. The CAMs, a newly described AM type, comprise 1.6% of the VL cells. Their bifurcating neurites integrate multiple inputs from the input axons and SAMs. While the SAM network appears to feedforward sparse 'memorizable' sensory representations to the VL output layer, the CAMs appear to monitor global activity and feedforward a balancing inhibition for 'sharpening' the stimulus-specific VL output. While sharing morphological and wiring features with circuits supporting associative learning in other animals, the VL has evolved a unique circuit that enables associative learning based on feedforward information flow.

Gain modulation and odor concentration invariance in early olfactory networks

The broad receptive field of the olfactory receptors constitutes the basis of a combinatorial code that allows animals to detect and discriminate many more odorants than the actual number of receptor types that they express. One drawback is that high odor concentrations recruit lower affinity receptors which can lead to the perception of qualitatively different odors. Here we addressed the contribution that signal-processing in the antennal lobe makes to reduce concentration dependence in odor representation. By means of calcium imaging and pharmacological approach we describe the contribution that GABA receptors play in terms of the amplitude and temporal profiles of the signals that convey odor information from the antennal lobes to higher brain centers. We found that GABA reduces the amplitude of odor elicited signals and the number of glomeruli that are recruited in an odor-concentration-dependent manner. Blocking GABA receptors decreases the correlation among glomerular activity patterns elicited by different concentrations of the same odor. In addition, we built a realistic mathematical model of the antennal lobe that was used to test the viability of the proposed mechanisms and to evaluate the processing properties of the AL network under conditions that cannot be achieved in physiology experiments. Interestingly, even though based on a rather simple topology and cell interactions solely mediated by GABAergic lateral inhibitions, the AL model reproduced key features of the AL response upon different odor concentrations and provides plausible solutions for concentration invariant recognition of odors by artificial sensors.

Characterizing neural coding performance for populations of sensory neurons: comparing a weighted spike distance metrics to other analytical methods

The identity of sensory stimuli is encoded in the spatio-temporal patterns of responses of the encoding neural population. For stimuli to be discriminated reliably, differences in population responses must be accurately decoded by downstream networks. Several methods to compare patterns of responses have been used by neurophysiologists to characterize the accuracy of the sensory responses studied. Among the most widely used analyses, we note methods based on Euclidean distances or on spike metric distances. Methods based on artificial neural networks and machine learning that recognize and/or classify specific input patterns have also gained popularity. Here, we first compare these three strategies using datasets from three different model systems: the moth olfactory system, the electrosensory system of gymnotids, and leaky-integrate-and-fire (LIF) model responses. We show that the input-weighting procedure inherent to artificial neural networks allows the efficient extraction of information relevant to stimulus discrimination. To combine the convenience of methods such as spike metric distances but leverage the advantages of weighting the inputs, we propose a measure based on geometric distances where each dimension is weighted proportionally to how informative it is. We show that the result of this Weighted Euclidian Distance (WED) analysis performs as well or better than the artificial neural network we tested and outperforms the more traditional spike distance metrics. We applied information theoretic analysis to LIF responses and compared their encoding accuracy with the discrimination accuracy quantified through this WED analysis. We show a high degree of correlation between discrimination accuracy and information content, and that our weighting procedure allowed the efficient use of information present to perform the discrimination task. We argue that our proposed measure provides the flexibility and ease of use sought by neurophysiologists while providing a more powerful way to extract relevant information than more traditional methods.

Hacking brain development to test models of sensory coding

Animals can discriminate myriad sensory stimuli but can also generalize from learned experience. You can probably distinguish the favorite teas of your colleagues while still recognizing that all tea pales in comparison to coffee. Tradeoffs between detection, discrimination, and generalization are inherent at every layer of sensory processing. During development, specific quantitative parameters are wired into perceptual circuits and set the playing field on which plasticity mechanisms play out. A primary goal of systems neuroscience is to understand how material properties of a circuit define the logical operations-computations--that it makes, and what good these computations are for survival. A cardinal method in biology-and the mechanism of evolution--is to change a unit or variable within a system and ask how this affects organismal function. Here, we make use of our knowledge of developmental wiring mechanisms to modify hard-wired circuit parameters in the Drosophila melanogaster mushroom body and assess the functional and behavioral consequences. By altering the number of expansion layer neurons (Kenyon cells) and their dendritic complexity, we find that input number, but not cell number, tunes odor selectivity. Simple odor discrimination performance is maintained when Kenyon cell number is reduced and augmented by Kenyon cell expansion.

Harnessing insect olfactory neural circuits for detecting and discriminating human cancers

There is overwhelming evidence that presence of cancer alters cellular metabolic processes, and these changes are manifested in emitted volatile organic compound (VOC) compositions of cancer cells. Here, we take a novel forward engineering approach by developing an insect olfactory neural circuit-based VOC sensor for cancer detection. We obtained oral cancer cell culture VOC-evoked extracellular neural responses from in vivo insect (locust) antennal lobe neurons. We employed biological neural computations of the antennal lobe circuitry for generating spatiotemporal neuronal response templates corresponding to each cell culture VOC mixture, and employed these neuronal templates to distinguish oral cancer cell lines (SAS, Ca9-22, and HSC-3) vs. a non-cancer cell line (HaCaT). Our results demonstrate that three different human oral cancers can be robustly distinguished from each other and from a non-cancer oral cell line. By using high-dimensional population neuronal response analysis and leave-one-trial-out methodology, our approach yielded high classification success for each cell line tested. Our analyses achieved 76-100% success in identifying cell lines by using the population neural response (n = 194) collected for the entire duration of the cell culture study. We also demonstrate this cancer detection technique can distinguish between different types of oral cancers and non-cancer at different time-matched points of growth. This brain-based cancer detection approach is fast as it can differentiate between VOC mixtures within 250 ms of stimulus onset. Our brain-based cancer detection system comprises a novel VOC sensing methodology that incorporates entire biological chemosensory arrays, biological signal transduction, and neuronal computations in a form of a forward-engineered technology for cancer VOC detection.

Mechanism of carbachol-induced 40 Hz gamma oscillations and the effects of NMDA activation on oscillatory dynamics in a model of the CA3 subfield of the hippocampus

Gamma oscillations are a prominent feature of various neural systems, including the CA3 subfield of the hippocampus. In CA3, in vitro carbachol application induces ∼40 Hz gamma oscillations in the network of glutamatergic excitatory pyramidal neurons (PNs) and local GABAergic inhibitory neurons (INs). Activation of NMDA receptors within CA3 leads to an increase in the frequency of carbachol-induced oscillations to ∼60 Hz, a broadening of the distribution of individual oscillation cycle frequencies, and a decrease in the time lag between PN and IN spike bursts. In this work, we develop a biophysical integrate-and-fire model of the CA3 subfield, we show that the dynamics of our model are in concordance with physiological observations, and we provide computational support for the hypothesis that the 'E-I' mechanism is responsible for the emergence of ∼40 Hz gamma oscillations in the absence of NMDA activation. We then incorporate NMDA receptors into our CA3 model, and we show that our model exhibits the increase in gamma oscillation frequency, broadening of the cycle frequency distribution, and decrease in the time lag between PN and IN spike bursts observed experimentally. Remarkably, we find an inverse relationship in our model between the net NMDA current delivered to PNs and INs in an oscillation cycle and cycle frequency. Furthermore, we find a disparate effect of NMDA receptors on PNs versus INs - we show that NMDA receptors on INs tend to increase oscillation frequency, while NMDA receptors on PNs tend to slightly decrease or not affect oscillation frequency. We find that these observations can be explained if NMDA activity above a threshold level causes a shift in the mechanism underlying gamma oscillations; in the absence of NMDA receptors, the 'E-I' mechanism is primarily responsible for the generation of gamma oscillations (at 40 Hz), while when NMDA receptors are active, the mechanism of gamma oscillations shifts to the 'I-I' mechanism, and we argue that within the 'I-I' regime (which displays a higher baseline oscillation frequency of ∼60 Hz), slight changes in the level of NMDA activity are inversely related to cycle frequency.

Optogenetic and thermogenetic manipulation of defined neural circuits and behaviors in Drosophila

Neural network dynamics underlying flexible animal behaviors remain elusive. The fruit fly Drosophila melanogaster is considered an excellent model in behavioral neuroscience because of its simple neuroanatomical architecture and the availability of various genetic methods. Moreover, Drosophila larvae's transparent body allows investigators to use optical methods on freely moving animals, broadening research directions. Activating or inhibiting well-defined events in excitable cells with a fine temporal resolution using optogenetics and thermogenetics led to the association of functions of defined neural populations with specific behavioral outputs such as the induction of associative memory. Furthermore, combining optogenetics and thermogenetics with state-of-the-art approaches, including connectome mapping and machine learning-based behavioral quantification, might provide a complete view of the experience- and time-dependent variations of behavioral responses. These methodologies allow further understanding of the functional connections between neural circuits and behaviors such as chemosensory, motivational, courtship, and feeding behaviors and sleep, learning, and memory.

The anterior paired lateral neuron normalizes odour-evoked activity in the Drosophila mushroom body calyx

To identify and memorize discrete but similar environmental inputs, the brain needs to distinguish between subtle differences of activity patterns in defined neuronal populations. The Kenyon cells (KCs) of the Drosophila adult mushroom body (MB) respond sparsely to complex olfactory input, a property that is thought to support stimuli discrimination in the MB. To understand how this property emerges, we investigated the role of the inhibitory anterior paired lateral (APL) neuron in the input circuit of the MB, the calyx. Within the calyx, presynaptic boutons of projection neurons (PNs) form large synaptic microglomeruli (MGs) with dendrites of postsynaptic KCs. Combining electron microscopy (EM) data analysis and in vivo calcium imaging, we show that APL, via inhibitory and reciprocal synapses targeting both PN boutons and KC dendrites, normalizes odour-evoked representations in MGs of the calyx. APL response scales with the PN input strength and is regionalized around PN input distribution. Our data indicate that the formation of a sparse code by the KCs requires APL-driven normalization of their MG postsynaptic responses. This work provides experimental insights on how inhibition shapes sensory information representation in a higher brain centre, thereby supporting stimuli discrimination and allowing for efficient associative memory formation.

Non-synaptic interactions between olfactory receptor neurons, a possible key feature of odor processing in flies

When flies explore their environment, they encounter odors in complex, highly intermittent plumes. To navigate a plume and, for example, find food, they must solve several challenges, including reliably identifying mixtures of odorants and their intensities, and discriminating odorant mixtures emanating from a single source from odorants emitted from separate sources and just mixing in the air. Lateral inhibition in the antennal lobe is commonly understood to help solving these challenges. With a computational model of the Drosophila olfactory system, we analyze the utility of an alternative mechanism for solving them: Non-synaptic ("ephaptic") interactions (NSIs) between olfactory receptor neurons that are stereotypically co-housed in the same sensilla. We find that NSIs improve mixture ratio detection and plume structure sensing and do so more efficiently than the traditionally considered mechanism of lateral inhibition in the antennal lobe. The best performance is achieved when both mechanisms work in synergy. However, we also found that NSIs decrease the dynamic range of co-housed ORNs, especially when they have similar sensitivity to an odorant. These results shed light, from a functional perspective, on the role of NSIs, which are normally avoided between neurons, for instance by myelination.

A model of lateral interactions as the origin of multiwhisker receptive fields in rat barrel cortex

While cells within barrel cortex respond primarily to deflections of their principal whisker (PW), they also exhibit responses to non-principal, or adjacent, whiskers (AWs), albeit responses with diminished amplitudes and longer latencies. The origin of multiwhisker receptive fields of barrel cells remains a point of controversy within the experimental literature, with three contending possibilities: (i) barrel cells inherit their AW responses from the AW responses of thalamocortical (TC) cells within their aligned barreloid; (ii) the axons of TC cells within a barreloid ramify to innervate multiple barrels, rather than only terminating within their aligned barrel; (iii) lateral intracortical transmission between barrels conveys AW responsivity to barrel cells. In this work, we develop a detailed, biologically plausible model of multiple barrels in order to examine possibility (iii); in order to isolate the dynamics that possibility (iii) entails, we incorporate lateral connections between barrels while assuming that TC cells respond only to their PW and that TC cell axons are confined to their home barrel. We show that our model is capable of capturing a broad swath of experimental observations on multiwhisker receptive field dynamics within barrels, and we compare and contrast the dynamics of this model with model dynamics from prior work in which employ a similar general modeling strategy to examine possibility (i).

Changes in Brain Functional Network Connectivity in Adult Moyamoya Diseases

Moyamoya disease (MMD) is a cerebrovascular disease that is characterized by progressive stenosis or occlusion of the internal carotid arteries and its main branches, which leads to the formation of abnormal small collateral vessels. However, little is known about how these special vascular structures affect cortical network connectivity and brain function. By applying EEG analysis and graphic network analyses undergoing EEG recording of subjects with eyes-closed (EC) and eyes-open (EO) resting states, and working memory (WM) tasks, we examined the brain network features of hemorrhagic (H_MMD) and ischemic MMD (I_MMD) brains. For the first time, we observed that I_MMD had the much lower alpha-blocking rate during EO state than healthy controls while H_MMD exhibited the relatively low EEG activity rate across all the behavior states. Further, I_MMD showed strong network connections in the alpha-wave band in frontal and parietal regions during EO and WM states. EEG frequency and network topological maps during both resting and WM states indicated that the left frontal lobe and left parietal lobe in H_MMD patients and the right parietal lobe and temporal lobe in I_MMD patients have clear differences compared with controls, which provides a new insight to understand distinct electrophysiological features of MMD. However, due to the small sample size of recruited patient subjects, the result conclusion may be limited.

Supplementary information

The online version contains supplementary material available at (10.1007/s11571-021-09666-1).

Olfactory encoding within the insect antennal lobe: The emergence and role of higher order temporal correlations in the dynamics of antennal lobe spiking activity

In this review, we focus on the antennal lobe (AL) of three insect species - the fruit fly, sphinx moth, and locust. We first review the experimentally elucidated anatomy and physiology of the early olfactory system of each species; empirical studies of AL activity, however, often focus on assessing firing rates (averaged over time scales of about 100 ms), and hence the AL odor code is often analyzed in terms of a temporally evolving vector of firing rates. However, such a perspective necessarily misses the possibility of higher order temporal correlations in spiking activity within a single cell and across multiple cells over shorter time scales (of about 10 ms). Hence, we then review our prior theoretical work, where we constructed biophysically detailed, species-specific AL models within the fly, moth, and locust, finding that in each case higher order temporal correlations in spiking naturally emerge from model dynamics (i.e., without a prioriincorporation of elements designed to produce correlated activity). We therefore use our theoretical work to argue the perspective that temporal correlations in spiking over short time scales, which have received little experimental attention to-date, may provide valuable coding dimensions (complementing the coding dimensions provided by the vector of firing rates) that nature has exploited in the encoding of odors within the AL. We further argue that, if the AL does indeed utilize temporally correlated activity to represent odor information, such an odor code could be naturally and easily deciphered within the Mushroom Body.

The connectome of the adult Drosophila mushroom body provides insights into function

Making inferences about the computations performed by neuronal circuits from synapse-level connectivity maps is an emerging opportunity in neuroscience. The mushroom body (MB) is well positioned for developing and testing such an approach due to its conserved neuronal architecture, recently completed dense connectome, and extensive prior experimental studies of its roles in learning, memory, and activity regulation. Here, we identify new components of the MB circuit in Drosophila, including extensive visual input and MB output neurons (MBONs) with direct connections to descending neurons. We find unexpected structure in sensory inputs, in the transfer of information about different sensory modalities to MBONs, and in the modulation of that transfer by dopaminergic neurons (DANs). We provide insights into the circuitry used to integrate MB outputs, connectivity between the MB and the central complex and inputs to DANs, including feedback from MBONs. Our results provide a foundation for further theoretical and experimental work.

Structural and Functional Plasticity in the Regenerating Olfactory System of the Migratory Locust

Regeneration after injury is accompanied by transient and lasting changes in the neuroarchitecture of the nervous system and, thus, a form of structural plasticity. In this review, we introduce the olfactory pathway of a particular insect as a convenient model to visualize neural regeneration at an anatomical level and study functional recovery at an electrophysiological level. The olfactory pathway of the locust (Locusta migratoria) is characterized by a multiglomerular innervation of the antennal lobe by olfactory receptor neurons. These olfactory afferents were axotomized by crushing the base of the antenna. The resulting degeneration and regeneration in the antennal lobe could be quantified by size measurements, dye labeling, and immunofluorescence staining of cell surface proteins implicated in axonal guidance during development. Within 3 days post lesion, the antennal lobe volume was reduced by 30% and from then onward regained size back to normal by 2 weeks post injury. The majority of regenerating olfactory receptor axons reinnervated the glomeruli of the antennal lobe. A few regenerating axons project erroneously into the mushroom body on a pathway that is normally chosen by second-order projection neurons. Based on intracellular responses of antennal lobe output neurons to odor stimulation, regenerated fibers establish functional synapses again. Following complete absence after nerve crush, responses to odor stimuli return to control level within 10–14 days. On average, regeneration of afferents, and re-established synaptic connections appear faster in younger fifth instar nymphs than in adults. The initial degeneration of olfactory receptor axons has a trans-synaptic effect on a second order brain center, leading to a transient size reduction of the mushroom body calyx. Odor-evoked oscillating field potentials, absent after nerve crush, were restored in the calyx, indicative of regenerative processes in the network architecture. We conclude that axonal regeneration in the locust olfactory system appears to be possible, precise, and fast, opening an avenue for future mechanistic studies. As a perspective of biomedical importance, the current evidence for nitric oxide/cGMP signaling as positive regulator of axon regeneration in connectives of the ventral nerve cord is considered in light of particular regeneration studies in vertebrate central nervous systems.

A spiking neural program for sensorimotor control during foraging in flying insects

Foraging is a vital behavioral task for living organisms. Behavioral strategies and abstract mathematical models thereof have been described in detail for various species. To explore the link between underlying neural circuits and computational principles, we present how a biologically detailed neural circuit model of the insect mushroom body implements sensory processing, learning, and motor control. We focus on cast and surge strategies employed by flying insects when foraging within turbulent odor plumes. Using a spike-based plasticity rule, the model rapidly learns to associate individual olfactory sensory cues paired with food in a classical conditioning paradigm. We show that, without retraining, the system dynamically recalls memories to detect relevant cues in complex sensory scenes. Accumulation of this sensory evidence on short time scales generates cast-and-surge motor commands. Our generic systems approach predicts that population sparseness facilitates learning, while temporal sparseness is required for dynamic memory recall and precise behavioral control. Our work successfully combines biological computational principles with spike-based machine learning. It shows how knowledge transfer from static to arbitrary complex dynamic conditions can be achieved by foraging insects and may serve as inspiration for agent-based machine learning.

Non-synaptic interactions between olfactory receptor neurons, a possible key feature of odor processing in flies.. [DOI: 10.1101/2020.07.23.217216]

When flies explore their environment, they encounter odors in complex, highly intermittent plumes. To navigate a plume and, for example, find food, they must solve several challenges, including reliably identifying mixtures of odorants and their intensities, and discriminating odorant mixtures emanating from a single source from odorants emitted from separate sources and just mixing in the air. Lateral inhibition in the antennal lobe is commonly understood to help solving these challenges. With a computational model of the Drosophila olfactory system, we analyze the utility of an alternative mechanism for solving them: Non-synaptic (“ephaptic”) interactions (NSIs) between olfactory receptor neurons that are stereotypically co-housed in the same sensilla.

We found that NSIs improve mixture ratio detection and plume structure sensing and they do so more efficiently than the traditionally considered mechanism of lateral inhibition in the antennal lobe. However, we also found that NSIs decrease the dynamic range of co-housed ORNs, especially when they have similar sensitivity to an odorant. These results shed light, from a functional perspective, on the role of NSIs, which are normally avoided between neurons, for instance by myelination.

Author summary

Myelin is important to isolate neurons and avoid disruptive electrical interference between them; it can be found in almost any neural assembly. However, there are a few exceptions to this rule and it remains unclear why. One particularly interesting case is the electrical interaction between olfactory sensory neurons co-housed in the sensilla of insects. Here, we created a computational model of the early stages of the Drosophila olfactory system and observed that the electrical interference between olfactory receptor neurons can be a useful trait that can help flies, and other insects, to navigate the complex plumes of odorants in their natural environment.

With the model we were able to shed new light on the trade-off of adopting this mechanism: We found that the non-synaptic interactions (NSIs) improve both the identification of the concentration ratio in mixtures of odorants and the discrimination of odorant mixtures emanating from a single source from odorants emitted from separate sources – both highly advantageous. However, they also decrease the dynamic range of the olfactory sensory neurons – a clear disadvantage.

Effect of Circuit Structure on Odor Representation in the Insect Olfactory System

In neuroscience, the structure of a circuit has often been used to intuit function—an inversion of Louis Kahn’s famous dictum, “Form follows function” (Kristan and Katz, 2006). However, different brain networks may use different network architectures to solve the same problem. The olfactory circuits of two insects, the locust, Schistocerca americana, and the fruit fly, Drosophila melanogaster, serve the same function—to identify and discriminate odors. The neural circuitry that achieves this shows marked structural differences. Projection neurons (PNs) in the antennal lobe innervate Kenyon cells (KCs) of the mushroom body. In locust, each KC receives inputs from ∼50% of PNs, a scheme that maximizes the difference between inputs to any two of ∼50,000 KCs. In contrast, in Drosophila, this number is only 5% and appears suboptimal. Using a computational model of the olfactory system, we show that the activity of KCs is sufficiently high-dimensional that it can separate similar odors regardless of the divergence of PN–KC connections. However, when temporal patterning encodes odor attributes, dense connectivity outperforms sparse connections. Increased separability comes at the cost of reliability. The disadvantage of sparse connectivity can be mitigated by incorporating other aspects of circuit architecture seen in Drosophila. Our simulations predict that Drosophila and locust circuits lie at different ends of a continuum where the Drosophila gives up on the ability to resolve similar odors to generalize across varying environments, while the locust separates odor representations but risks misclassifying noisy variants of the same odor.

Circuit and Cellular Mechanisms Facilitate the Transformation from Dense to Sparse Coding in the Insect Olfactory System

Transformations between sensory representations are shaped by neural mechanisms at the cellular and the circuit level. In the insect olfactory system, the encoding of odor information undergoes a transition from a dense spatiotemporal population code in the antennal lobe to a sparse code in the mushroom body. However, the exact mechanisms shaping odor representations and their role in sensory processing are incompletely identified. Here, we investigate the transformation from dense to sparse odor representations in a spiking model of the insect olfactory system, focusing on two ubiquitous neural mechanisms: spike frequency adaptation at the cellular level and lateral inhibition at the circuit level. We find that cellular adaptation is essential for sparse representations in time (temporal sparseness), while lateral inhibition regulates sparseness in the neuronal space (population sparseness). The interplay of both mechanisms shapes spatiotemporal odor representations, which are optimized for the discrimination of odors during stimulus onset and offset. Response pattern correlation across different stimuli showed a nonmonotonic dependence on the strength of lateral inhibition with an optimum at intermediate levels, which is explained by two counteracting mechanisms. In addition, we find that odor identity is stored on a prolonged timescale in the adaptation levels but not in the spiking activity of the principal cells of the mushroom body, providing a testable hypothesis for the location of the so-called odor trace.

Feedback inhibition and its control in an insect olfactory circuit

Inhibitory neurons play critical roles in regulating and shaping olfactory responses in vertebrates and invertebrates. In insects, these roles are performed by relatively few neurons, which can be interrogated efficiently, revealing fundamental principles of olfactory coding. Here, with electrophysiological recordings from the locust and a large-scale biophysical model, we analyzed the properties and functions of GGN, a unique giant GABAergic neuron that plays a central role in structuring olfactory codes in the locust mushroom body. Our simulations suggest that depolarizing GGN at its input branch can globally inhibit KCs several hundred microns away. Our in vivorecordings show that GGN responds to odors with complex temporal patterns of depolarization and hyperpolarization that can vary with odors and across animals, leading our model to predict the existence of a yet-undiscovered olfactory pathway. Our analysis reveals basic new features of GGN and the olfactory network surrounding it.

Optimality of sparse olfactory representations is not affected by network plasticity

The neural representation of a stimulus is repeatedly transformed as it moves from the sensory periphery to deeper layers of the nervous system. Sparsening transformations are thought to increase the separation between similar representations, encode stimuli with great specificity, maximize storage capacity of associative memories, and provide an energy efficient instantiation of information in neural circuits. In the insect olfactory system, odors are initially represented in the periphery as a combinatorial code with relatively simple temporal dynamics. Subsequently, in the antennal lobe this representation is transformed into a dense and complex spatiotemporal activity pattern. Next, in the mushroom body Kenyon cells (KCs), the representation is dramatically sparsened. Finally, in mushroom body output neurons (MBONs), the representation takes on a new dense spatiotemporal format. Here, we develop a computational model to simulate this chain of olfactory processing from the receptor neurons to MBONs. We demonstrate that representations of similar odorants are maximally separated, measured by the distance between the corresponding MBON activity vectors, when KC responses are sparse. Sparseness is maintained across variations in odor concentration by adjusting the feedback inhibition that KCs receive from an inhibitory neuron, the Giant GABAergic neuron. Different odor concentrations require different strength and timing of feedback inhibition for optimal processing. Importantly, as observed in vivo, the KC–MBON synapse is highly plastic, and, therefore, changes in synaptic strength after learning can change the balance of excitation and inhibition, potentially leading to changes in the distance between MBON activity vectors of two odorants for the same level of KC population sparseness. Thus, what is an optimal degree of sparseness before odor learning, could be rendered sub–optimal post learning. Here, we show, however, that synaptic weight changes caused by spike timing dependent plasticity increase the distance between the odor representations from the perspective of MBONs. A level of sparseness that was optimal before learning remains optimal post-learning.

Kenyon cells (KCs) of the mushroom body represent odors as a sparse code. When viewed from the perspective of follower neurons, mushroom body output neurons (MBONs) reveal an optimal level of coding sparseness that maximally separates the representations of odors. However, the KC–MBON synapse is highly plastic and may be potentiated or depressed by odor–driven experience that could, in turn, disrupt the optimality formed by pre–synaptic circuits. Contrary to this expectation, we show that synaptic plasticity based on spike timing of pre- and postsynaptic neurons improves the ability of the system to distinguish between the representations of similar odors while preserving the optimality determined by pre–synaptic circuits.

Molecular characterization and distribution of the voltage-gated sodium channel, Para, in the brain of the grasshopper and vinegar fly

Voltage-gated sodium (NaV) channels, encoded by the gene para, play a critical role in the rapid processing and propagation of visual information related to collision avoidance behaviors. We investigated their localization by immunostaining the optic lobes and central brain of the grasshopper Schistocerca americana and the vinegar fly Drosophila melanogaster with an antibody that recognizes the channel peptide domain responsible for fast inactivation gating. NaV channels were detected at high density at all stages of development. In the optic lobe, they revealed stereotypically repeating fascicles consistent with the regular structure of the eye. In the central brain, major axonal tracts were strongly labeled, particularly in the grasshopper olfactory system. We used the NaV channel sequence of Drosophila to identify an ortholog in the transcriptome of Schistocerca. The grasshopper, vinegar fly, and human NaV channels exhibit a high degree of conservation at gating and ion selectivity domains. Comparison with three species evolutionarily close to Schistocerca identified splice variants of Para and their relation to those of Drosophila. The anatomical distribution of NaV channels molecularly analogous to those of humans in grasshoppers and vinegar flies provides a substrate for rapid signal propagation and visual processing in the context of visually-guided collision avoidance.

Separate But Interactive Parallel Olfactory Processing Streams Governed by Different Types of GABAergic Feedback Neurons in the Mushroom Body of a Basal Insect

The basic organization of the olfactory system has been the subject of extensive studies in vertebrates and invertebrates. In many animals, GABA-ergic neurons inhibit spike activities of higher-order olfactory neurons and help sparsening of their odor representations. In the cockroach, two different types of GABA-immunoreactive interneurons (calyceal giants [CGs]) mainly project to the base and lip regions of the calyces (input areas) of the mushroom body (MB), a second-order olfactory center. The base and lip regions receive axon terminals of two different types of projection neurons, which receive synapses from different classes of olfactory sensory neurons (OSNs), and receive dendrites of different classes of Kenyon cells, MB intrinsic neurons. We performed intracellular recordings from pairs of CGs and MB output neurons (MBONs) of male American cockroaches, the latter receiving synapses from Kenyon cells, and we found that a CG receives excitatory synapses from an MBON and that odor responses of the MBON are changed by current injection into the CG. Such feedback effects, however, were often weak or absent in pairs of neurons that belong to different streams, suggesting parallel organization of the recurrent pathways, although interactions between different streams were also evident. Cross-covariance analysis of the spike activities of CGs and MBONs suggested that odor stimulation produces synchronized spike activities in MBONs and then in CGs. We suggest that there are separate but interactive parallel streams to process odors detected by different OSNs throughout the olfactory processing system in cockroaches.SIGNIFICANCE STATEMENT Organizational principles of the olfactory system have been the subject of extensive studies. In cockroaches, signals from olfactory sensory neurons (OSNs) in two different classes of sensilla are sent to two different classes of projection neurons, which terminate in different areas of the mushroom body (MB), each area having dendrites of different classes of MB intrinsic neurons (Kenyon cells) and terminations of different classes of GABAergic neurons. Physiological and morphological assessments derived from simultaneous intracellular recordings/stainings from GABAergic neurons and MB output neurons suggested that GABAergic neurons play feedback roles and that odors detected by OSNs are processed in separate but interactive processing streams throughout the central olfactory system.

Analysis of feedforward mechanisms of multiwhisker receptive field generation in a model of the rat barrel cortex

There is substantial anatomical segregation in the organization of the rodent barrel system - each whisker on the mystacial pad sends input to TC cells within a dedicated thalamic barreloid, which in turn innervates a corresponding cortical barrel, and RS cells within a barrel respond primarily to deflections of the corresponding whisker at the beginning of the dedicated transmission line (the principal whisker, PW). However, it is also well-established that barrel cells exhibit multiwhisker receptive fields (RFs), and display lower amplitude, longer latency responses to deflections of non-PWs (or adjacent whiskers, AWs). There is considerable controversy regarding the origin of such multiwhisker RFs; three possibilities include: (i) TC cells within a barreloid respond to multiple whiskers, and barrel RS cells simply inherit multiwhisker responses from their aligned barreloid; (ii) TC cells respond only to the PW, but individual barreloids innervate multiple barrels; (iii) multiwhisker responses of barrel cells arise from lateral corticocortical (barrel-to-barrel) synaptic transmission. Ablation studies attempting to pinpoint the source of RS cell AW responses are often contradictory (though experimental work tends to suggest possibilities (i) or (iii) to be most plausible), and hence it is important to carefully evaluate these hypotheses in terms of available physiological data on barreloid and barrel response dynamics. In this work, I employ a biologically detailed model of the rat barrel cortex to evaluate possibility (i), and I show that, within the model, hypothesis (i) is capable of explaining a broad range of the available physiological data on responses to single (PW or AW) deflections and paired whisker deflections (AW deflection followed by PW deflection), as well as the dependence of such responses on the angular direction of whisker deflection. In particular, the model shows that barrel RS cells can exhibit AW direction tuning despite the fact that barreloid to barrel wiring has no systematic dependence on the AW direction preference of TC cells. Future modeling work will examine the other possibilities for the generation of multiwhisker RS cell RFs, and compare and contrast the different possible mechanisms within the context of available experimental data.

Neuroscience: Intelligence in the Honeybee Mushroom Body

Intelligence, in most people's conception, involves combining pieces of evidence to reach non-obvious conclusions. A recent theoretical study shows that intelligence-like brain functions can emerge from simple neural circuits, in this case the honeybee mushroom body.

Re-evaluating Circuit Mechanisms Underlying Pattern Separation

When animals interact with complex environments, their neural circuits must separate overlapping patterns of activity that represent sensory and motor information. Pattern separation is thought to be a key function of several brain regions, including the cerebellar cortex, insect mushroom body, and dentate gyrus. However, recent findings have questioned long-held ideas on how these circuits perform this fundamental computation. Here, we re-evaluate the functional and structural mechanisms underlying pattern separation. We argue that the dimensionality of the space available for population codes representing sensory and motor information provides a common framework for understanding pattern separation. We then discuss how these three circuits use different strategies to separate activity patterns and facilitate associative learning in the presence of trial-to-trial variability.

Biological Mechanisms for Learning: A Computational Model of Olfactory Learning in the Manduca sexta Moth, With Applications to Neural Nets

The insect olfactory system, which includes the antennal lobe (AL), mushroom body (MB), and ancillary structures, is a relatively simple neural system capable of learning. Its structural features, which are widespread in biological neural systems, process olfactory stimuli through a cascade of networks where large dimension shifts occur from stage to stage and where sparsity and randomness play a critical role in coding. Learning is partly enabled by a neuromodulatory reward mechanism of octopamine stimulation of the AL, whose increased activity induces synaptic weight updates in the MB through Hebbian plasticity. Enforced sparsity in the MB focuses Hebbian growth on neurons that are the most important for the representation of the learned odor. Based upon current biophysical knowledge, we have constructed an end-to-end computational firing-rate model of the Manduca sexta moth olfactory system which includes the interaction of the AL and MB under octopamine stimulation. Our model is able to robustly learn new odors, and neural firing rates in our simulations match the statistical features of in vivo firing rate data. From a biological perspective, the model provides a valuable tool for examining the role of neuromodulators, like octopamine, in learning, and gives insight into critical interactions between sparsity, Hebbian growth, and stimulation during learning. Our simulations also inform predictions about structural details of the olfactory system that are not currently well-characterized. From a machine learning perspective, the model yields bio-inspired mechanisms that are potentially useful in constructing neural nets for rapid learning from very few samples. These mechanisms include high-noise layers, sparse layers as noise filters, and a biologically-plausible optimization method to train the network based on octopamine stimulation, sparse layers, and Hebbian growth.

Neuronal Response Latencies Encode First Odor Identity Information across Subjects

Odorants are coded in the primary olfactory processing centers by spatially and temporally distributed patterns of glomerular activity. Whereas the spatial distribution of odorant-induced responses is known to be conserved across individuals, the universality of its temporal structure is still debated. Via fast two-photon calcium imaging, we analyzed the early phase of neuronal responses in the form of the activity onset latencies in the antennal lobe projection neurons of honeybee foragers. We show that each odorant evokes a stimulus-specific response latency pattern across the glomerular coding space. Moreover, we investigate these early response features for the first time across animals, revealing that the order of glomerular firing onsets is conserved across individuals and allows them to reliably predict odorant identity, but not concentration. These results suggest that the neuronal response latencies provide the first available code for fast odor identification.SIGNIFICANCE STATEMENT Here, we studied early temporal coding in the primary olfactory processing centers of the honeybee brain by fast imaging of glomerular responses to different odorants across glomeruli and across individuals. Regarding the elusive role of rapid response dynamics in olfactory coding, we were able to clarify the following aspects: (1) the rank of glomerular activation is conserved across individuals, (2) its stimulus prediction accuracy is equal to that of the response amplitude code, and (3) it contains complementary information. Our findings suggest a substantial role of response latencies in odor identification, anticipating the static response amplitude code.

Dynamic contrast enhancement and flexible odor codes

Sensory stimuli evoke spiking activities patterned across neurons and time that are hypothesized to encode information about their identity. Since the same stimulus can be encountered in a multitude of ways, how stable or flexible are these stimulus-evoked responses? Here we examine this issue in the locust olfactory system. In the antennal lobe, we find that both spatial and temporal features of odor-evoked responses vary in a stimulus-history dependent manner. The response variations are not random, but allow the antennal lobe circuit to enhance the uniqueness of the current stimulus. Nevertheless, information about the odorant identity is conf ounded due to this contrast enhancement computation. Notably, predictions from a linear logical classifier (OR-of-ANDs) that can decode information distributed in flexible subsets of neurons match results from behavioral experiments. In sum, our results suggest that a trade-off between stability and flexibility in sensory coding can be achieved using a simple computational logic.

Sensory stimuli are encountered in multiple ways necessitating a flexible and adaptive neural population code for identification. Here, the authors show that the dynamics of odor coding in the locust antennal lobe varies with stimulus context so as to enhance the target stimulus representation.

Effects of Adaptation on Discrimination of Whisker Deflection Velocity and Angular Direction in a Model of the Barrel Cortex

Two important stimulus features represented within the rodent barrel cortex are velocity and angular direction of whisker deflection. Each cortical barrel receives information from thalamocortical (TC) cells that relay information from a single whisker, and TC input is decoded by barrel regular-spiking (RS) cells through a feedforward inhibitory architecture (with inhibition delivered by cortical fast-spiking or FS cells). TC cells encode deflection velocity through population synchrony, while deflection direction is encoded through the distribution of spike counts across the TC population. Barrel RS cells encode both deflection direction and velocity with spike rate, and are divided into functional domains by direction preference. Following repetitive whisker stimulation, system adaptation causes a weakening of synaptic inputs to RS cells and diminishes RS cell spike responses, though evidence suggests that stimulus discrimination may improve following adaptation. In this work, I construct a model of the TC, FS, and RS cells comprising a single barrel system—the model incorporates realistic synaptic connectivity and dynamics and simulates both angular direction (through the spatial pattern of TC activation) and velocity (through synchrony of the TC population spikes) of a deflection of the primary whisker, and I use the model to examine direction and velocity selectivity of barrel RS cells before and after adaptation. I find that velocity and direction selectivity of individual RS cells (measured over multiple trials) sharpens following adaptation, but stimulus discrimination using a simple linear classifier by the RS population response during a single trial (a more biologically meaningful measure than single cell discrimination over multiple trials) exhibits strikingly different behavior—velocity discrimination is similar both before and after adaptation, while direction classification improves substantially following adaptation. This is the first model, to my knowledge, that simulates both whisker deflection velocity and angular direction and examines the ability of the RS population response to pinpoint both stimulus features within the context of adaptation.

Calcium in Kenyon Cell Somata as a Substrate for an Olfactory Sensory Memory in Drosophila

Animals can form associations between temporally separated stimuli. To do so, the nervous system has to retain a neural representation of the first stimulus until the second stimulus appears. The neural substrate of such sensory stimulus memories is unknown. Here, we search for a sensory odor memory in the insect olfactory system and characterize odorant-evoked Ca2+ activity at three consecutive layers of the olfactory system in Drosophila: in olfactory receptor neurons (ORNs) and projection neurons (PNs) in the antennal lobe, and in Kenyon cells (KCs) in the mushroom body. We show that the post-stimulus responses in ORN axons, PN dendrites, PN somata, and KC dendrites are odor-specific, but they are not predictive of the chemical identity of past olfactory stimuli. However, the post-stimulus responses in KC somata carry information about the identity of previous olfactory stimuli. These findings show that the Ca2+ dynamics in KC somata could encode a sensory memory of odorant identity and thus might serve as a basis for associations between temporally separated stimuli.

A neural network model for familiarity and context learning during honeybee foraging flights

How complex is the memory structure that honeybees use to navigate? Recently, an insect-inspired parsimonious spiking neural network model was proposed that enabled simulated ground-moving agents to follow learned routes. We adapted this model to flying insects and evaluate the route following performance in three different worlds with gradually decreasing object density. In addition, we propose an extension to the model to enable the model to associate sensory input with a behavioral context, such as foraging or homing. The spiking neural network model makes use of a sparse stimulus representation in the mushroom body and reward-based synaptic plasticity at its output synapses. In our experiments, simulated bees were able to navigate correctly even when panoramic cues were missing. The context extension we propose enabled agents to successfully discriminate partly overlapping routes. The structure of the visual environment, however, crucially determines the success rate. We find that the model fails more often in visually rich environments due to the overlap of features represented by the Kenyon cell layer. Reducing the landmark density improves the agents route following performance. In very sparse environments, we find that extended landmarks, such as roads or field edges, may help the agent stay on its route, but often act as strong distractors yielding poor route following performance. We conclude that the presented model is valid for simple route following tasks and may represent one component of insect navigation. Additional components might still be necessary for guidance and action selection while navigating along different memorized routes in complex natural environments.

Spiking and Excitatory/Inhibitory Input Dynamics of Barrel Cells in Response to Whisker Deflections of Varying Velocity and Angular Direction

The spiking of barrel regular-spiking (RS) cells is tuned for both whisker deflection direction and velocity. Velocity tuning arises due to thalamocortical (TC) synchrony (but not spike quantity) varying with deflection velocity, coupled with feedforward inhibition, while direction selectivity is not fully understood, though may be due partly to direction tuning of TC spiking. Data show that as deflection direction deviates from the preferred direction of an RS cell, excitatory input to the RS cell diminishes minimally, but temporally shifts to coincide with the time-lagged inhibitory input. This work constructs a realistic large-scale model of a barrel; model RS cells exhibit velocity and direction selectivity due to TC input dynamics, with the experimentally observed sharpening of direction tuning with decreasing velocity. The model puts forth the novel proposal that RS→RS synapses can naturally and simply account for the unexplained direction dependence of RS cell inputs - as deflection direction deviates from the preferred direction of an RS cell, and TC input declines, RS→RS synaptic transmission buffers the decline in total excitatory input and causes a shift in timing of the excitatory input peak from the peak in TC input to the delayed peak in RS input. The model also provides several experimentally testable predictions on the velocity dependence of RS cell inputs. This model is the first, to my knowledge, to study the interaction of direction and velocity and propose physiological mechanisms for the stimulus dependence in the timing and amplitude of RS cell inputs.

Optimal Degrees of Synaptic Connectivity

Synaptic connectivity varies widely across neuronal types. Cerebellar granule cells receive five orders of magnitude fewer inputs than the Purkinje cells they innervate, and cerebellum-like circuits, including the insect mushroom body, also exhibit large divergences in connectivity. In contrast, the number of inputs per neuron in cerebral cortex is more uniform and large. We investigate how the dimension of a representation formed by a population of neurons depends on how many inputs each neuron receives and what this implies for learning associations. Our theory predicts that the dimensions of the cerebellar granule-cell and Drosophila Kenyon-cell representations are maximized at degrees of synaptic connectivity that match those observed anatomically, showing that sparse connectivity is sometimes superior to dense connectivity. When input synapses are subject to supervised plasticity, however, dense wiring becomes advantageous, suggesting that the type of plasticity exhibited by a set of synapses is a major determinant of connection density.

A Simple Computational Model of the Bee Mushroom Body Can Explain Seemingly Complex Forms of Olfactory Learning and Memory

Honeybees are models for studying how animals with relatively small brains accomplish complex cognition, displaying seemingly advanced (or "non-elemental") learning phenomena involving multiple conditioned stimuli. These include "peak shift" [1-4]-where animals not only respond to entrained stimuli, but respond even more strongly to similar ones that are farther away from non-rewarding stimuli. Bees also display negative and positive patterning discrimination [5], responding in opposite ways to mixtures of two odors than to individual odors. Since Pavlov, it has often been assumed that such phenomena are more complex than simple associate learning. We present a model of connections between olfactory sensory input and bees' mushroom bodies [6], incorporating empirically determined properties of mushroom body circuitry (random connectivity [7], sparse coding [8], and synaptic plasticity [9, 10]). We chose not to optimize the model's parameters to replicate specific behavioral phenomena, because we were interested in the emergent cognitive capacities that would pop out of a network constructed solely based on empirical neuroscientific information and plausible assumptions for unknown parameters. We demonstrate that the circuitry mediating "simple" associative learning can also replicate the various non-elemental forms of learning mentioned above and can effectively multi-task by replicating a range of different learning feats. We found that PN-KC synaptic plasticity is crucial in controlling the generalization-discrimination trade-off-it facilitates peak shift and hinders patterning discrimination-and that PN-to-KC connection number can affect this trade-off. These findings question the notion that forms of learning that have been regarded as "higher order" are computationally more complex than "simple" associative learning.

Oscillatory integration windows in neurons

Oscillatory synchrony among neurons occurs in many species and brain areas, and has been proposed to help neural circuits process information. One hypothesis states that oscillatory input creates cyclic integration windows: specific times in each oscillatory cycle when postsynaptic neurons become especially responsive to inputs. With paired local field potential (LFP) and intracellular recordings and controlled stimulus manipulations we directly test this idea in the locust olfactory system. We find that inputs arriving in Kenyon cells (KCs) sum most effectively in a preferred window of the oscillation cycle. With a computational model, we show that the non-uniform structure of noise in the membrane potential helps mediate this process. Further experiments performed in vivo demonstrate that integration windows can form in the absence of inhibition and at a broad range of oscillation frequencies. Our results reveal how a fundamental coincidence-detection mechanism in a neural circuit functions to decode temporally organized spiking.

Complete identification of four giant interneurons supplying mushroom body calyces in the cockroach Periplaneta americana

Global inhibition is a fundamental physiological mechanism that has been proposed to shape odor representation in higher-order olfactory centers. A pair of mushroom bodies (MBs) in insect brains, an analog of the mammalian olfactory cortex, are implicated in multisensory integration and associative memory formation. With the use of single/multiple intracellular recording and staining in the cockroach Periplaneta americana, we succeeded in unambiguous identification of four tightly bundled GABA-immunoreactive giant interneurons that are presumably involved in global inhibitory control of the MB. These neurons, including three spiking neurons and one nonspiking neuron, possess dendrites in termination fields of MB output neurons and send axon terminals back to MB input sites, calyces, suggesting feedback roles onto the MB. The largest spiking neuron innervates almost exclusively the basal region of calyces, while the two smaller spiking neurons and the second-largest nonspiking neuron innervate more profusely the peripheral (lip) region of the calyces than the basal region. This subdivision corresponds well to the calycal zonation made by axon terminals of two populations of uniglomerular projection neurons with dendrites in distinct glomerular groups in the antennal lobe. The four giant neurons exhibited excitatory responses to every odor tested in a neuron-specific fashion, and two of the neurons also exhibited inhibitory responses in some recording sessions. Our results suggest that two parallel olfactory inputs to the MB undergo different forms of inhibitory control by the giant neurons, which may, in turn, be involved in different aspects of odor discrimination, plasticity, and state-dependent gain control. J. Comp. Neurol. 525:204-230, 2017. © 2016 Wiley Periodicals, Inc.

A Robust Feedforward Model of the Olfactory System

Most natural odors have sparse molecular composition. This makes the principles of compressed sensing potentially relevant to the structure of the olfactory code. Yet, the largely feedforward organization of the olfactory system precludes reconstruction using standard compressed sensing algorithms. To resolve this problem, recent theoretical work has shown that signal reconstruction could take place as a result of a low dimensional dynamical system converging to one of its attractor states. However, the dynamical aspects of optimization slowed down odor recognition and were also found to be susceptible to noise. Here we describe a feedforward model of the olfactory system that achieves both strong compression and fast reconstruction that is also robust to noise. A key feature of the proposed model is a specific relationship between how odors are represented at the glomeruli stage, which corresponds to a compression, and the connections from glomeruli to third-order neurons (neurons in the olfactory cortex of vertebrates or Kenyon cells in the mushroom body of insects), which in the model corresponds to reconstruction. We show that should this specific relationship hold true, the reconstruction will be both fast and robust to noise, and in particular to the false activation of glomeruli. The predicted connectivity rate from glomeruli to third-order neurons can be tested experimentally.

Spatiotemporal Coding of Individual Chemicals by the Gustatory System

Four of the five major sensory systems (vision, olfaction, somatosensation, and audition) are thought to use different but partially overlapping sets of neurons to form unique representations of vast numbers of stimuli. The only exception is gustation, which is thought to represent only small numbers of basic taste categories. However, using new methods for delivering tastant chemicals and making electrophysiological recordings from the tractable gustatory system of the moth Manduca sexta, we found chemical-specific information is as follows: (1) initially encoded in the population of gustatory receptor neurons as broadly distributed spatiotemporal patterns of activity; (2) dramatically integrated and temporally transformed as it propagates to monosynaptically connected second-order neurons; and (3) observed in tastant-specific behavior. Our results are consistent with an emerging view of the gustatory system: rather than constructing basic taste categories, it uses a spatiotemporal population code to generate unique neural representations of individual tastant chemicals.

SIGNIFICANCE STATEMENT

Our results provide a new view of taste processing. Using a new, relatively simple model system and a new set of techniques to deliver taste stimuli and to examine gustatory receptor neurons and their immediate followers, we found no evidence for labeled line connectivity, or basic taste categories such as sweet, salty, bitter, and sour. Rather, individual tastant chemicals are represented as patterns of spiking activity distributed across populations of receptor neurons. These representations are transformed substantially as multiple types of receptor neurons converge upon follower neurons, leading to a combinatorial coding format that uniquely, rapidly, and efficiently represents individual taste chemicals. Finally, we found that the information content of these neurons can drive tastant-specific behavior.

Neural Encoding of Odors during Active Sampling and in Turbulent Plumes

Sensory inputs are often fluctuating and intermittent, yet animals reliably utilize them to direct behavior. Here we ask how natural stimulus fluctuations influence the dynamic neural encoding of odors. Using the locust olfactory system, we isolated two main causes of odor intermittency: chaotic odor plumes and active sampling behaviors. Despite their irregularity, chaotic odor plumes still drove dynamic neural response features including the synchronization, temporal patterning, and short-term plasticity of spiking in projection neurons, enabling classifier-based stimulus identification and activating downstream decoders (Kenyon cells). Locusts can also impose odor intermittency through active sampling movements with their unrestrained antennae. Odors triggered immediate, spatially targeted antennal scanning that, paradoxically, weakened individual neural responses. However, these frequent but weaker responses were highly informative about stimulus location. Thus, not only are odor-elicited dynamic neural responses compatible with natural stimulus fluctuations and important for stimulus identification, but locusts actively increase intermittency, possibly to improve stimulus localization.

It takes two-coincidence coding within the dual olfactory pathway of the honeybee

To rapidly process biologically relevant stimuli, sensory systems have developed a broad variety of coding mechanisms like parallel processing and coincidence detection. Parallel processing (e.g., in the visual system), increases both computational capacity and processing speed by simultaneously coding different aspects of the same stimulus. Coincidence detection is an efficient way to integrate information from different sources. Coincidence has been shown to promote associative learning and memory or stimulus feature detection (e.g., in auditory delay lines). Within the dual olfactory pathway of the honeybee both of these mechanisms might be implemented by uniglomerular projection neurons (PNs) that transfer information from the primary olfactory centers, the antennal lobe (AL), to a multimodal integration center, the mushroom body (MB). PNs from anatomically distinct tracts respond to the same stimulus space, but have different physiological properties, characteristics that are prerequisites for parallel processing of different stimulus aspects. However, the PN pathways also display mirror-imaged like anatomical trajectories that resemble neuronal coincidence detectors as known from auditory delay lines. To investigate temporal processing of olfactory information, we recorded PN odor responses simultaneously from both tracts and measured coincident activity of PNs within and between tracts. Our results show that coincidence levels are different within each of the two tracts. Coincidence also occurs between tracts, but to a minor extent compared to coincidence within tracts. Taken together our findings support the relevance of spike timing in coding of olfactory information (temporal code).