Intensity versus identity coding in an olfactory system

Extracting the Behaviorally Relevant Stimulus: Unique Neural Representation of Farnesol, a Component of the Recruitment Pheromone of Bombus terrestris

To trigger innate behavior, sensory neural networks are pre-tuned to extract biologically relevant stimuli. Many male-female or insect-plant interactions depend on this phenomenon. Especially communication among individuals within social groups depends on innate behaviors. One example is the efficient recruitment of nest mates by successful bumblebee foragers. Returning foragers release a recruitment pheromone in the nest while they perform a ‘dance’ behavior to activate unemployed nest mates. A major component of this pheromone is the sesquiterpenoid farnesol. How farnesol is processed and perceived by the olfactory system, has not yet been identified. It is much likely that processing farnesol involves an innate mechanism for the extraction of relevant information to trigger a fast and reliable behavioral response. To test this hypothesis, we used population response analyses of 100 antennal lobe (AL) neurons recorded in alive bumblebee workers under repeated stimulation with four behaviorally different, but chemically related odorants (geraniol, citronellol, citronellal and farnesol). The analysis identified a unique neural representation of the recruitment pheromone component compared to the other odorants that are predominantly emitted by flowers. The farnesol induced population activity in the AL allowed a reliable separation of farnesol from all other chemically related odor stimuli we tested. We conclude that the farnesol induced population activity may reflect a predetermined representation within the AL-neural network allowing efficient and fast extraction of a behaviorally relevant stimulus. Furthermore, the results show that population response analyses of multiple single AL-units may provide a powerful tool to identify distinct representations of behaviorally relevant odors.

Endogenous sequential cortical activity evoked by visual stimuli

Although the functional properties of individual neurons in primary visual cortex have been studied intensely, little is known about how neuronal groups could encode changing visual stimuli using temporal activity patterns. To explore this, we used in vivo two-photon calcium imaging to record the activity of neuronal populations in primary visual cortex of awake mice in the presence and absence of visual stimulation. Multidimensional analysis of the network activity allowed us to identify neuronal ensembles defined as groups of cells firing in synchrony. These synchronous groups of neurons were themselves activated in sequential temporal patterns, which repeated at much higher proportions than chance and were triggered by specific visual stimuli such as natural visual scenes. Interestingly, sequential patterns were also present in recordings of spontaneous activity without any sensory stimulation and were accompanied by precise firing sequences at the single-cell level. Moreover, intrinsic dynamics could be used to predict the occurrence of future neuronal ensembles. Our data demonstrate that visual stimuli recruit similar sequential patterns to the ones observed spontaneously, consistent with the hypothesis that already existing Hebbian cell assemblies firing in predefined temporal sequences could be the microcircuit substrate that encodes visual percepts changing in time.

Single-trial dynamics of motor cortex and their applications to brain-machine interfaces

Increasing evidence suggests that neural population responses have their own internal drive, or dynamics, that describe how the neural population evolves through time. An important prediction of neural dynamical models is that previously observed neural activity is informative of noisy yet-to-be-observed activity on single-trials, and may thus have a denoising effect. To investigate this prediction, we built and characterized dynamical models of single-trial motor cortical activity. We find these models capture salient dynamical features of the neural population and are informative of future neural activity on single trials. To assess how neural dynamics may beneficially denoise single-trial neural activity, we incorporate neural dynamics into a brain–machine interface (BMI). In online experiments, we find that a neural dynamical BMI achieves substantially higher performance than its non-dynamical counterpart. These results provide evidence that neural dynamics beneficially inform the temporal evolution of neural activity on single trials and may directly impact the performance of BMIs.

In online experiments with monkeys the authors demonstrate, for the first time, that incorporating neural dynamics substantially improves brain–machine interface performance. This result is consistent with a framework hypothesizing that motor cortex is a dynamical machine that generates movement.

Stable odor recognition by a neuro-adaptive electronic nose

Sensitivity, selectivity and stability are decisive properties of sensors. In chemical gas sensors odor recognition can be severely compromised by poor signal stability, particularly in real life applications where the sensors are exposed to unpredictable sequences of odors under changing external conditions. Although olfactory receptor neurons in the nose face similar stimulus sequences under likewise changing conditions, odor recognition is very stable and odorants can be reliably identified independently from past odor perception. We postulate that appropriate pre-processing of the output signals of chemical sensors substantially contributes to the stability of odor recognition, in spite of marked sensor instabilities. To investigate this hypothesis, we use an adaptive, unsupervised neural network inspired by the glomerular input circuitry of the olfactory bulb. Essentially the model reduces the effect of the sensors’ instabilities by utilizing them via an adaptive multicompartment feed-forward inhibition. We collected and analyzed responses of a 4 × 4 gas sensor array to a number of volatile compounds applied over a period of 18 months, whereby every sensor was sampled episodically. The network conferred excellent stability to the compounds’ identification and was clearly superior over standard classifiers, even when one of the sensors exhibited random fluctuations or stopped working at all.

A model for non-monotonic intensity coding

Peripheral neurons of most sensory systems increase their response with increasing stimulus intensity. Behavioural responses, however, can be specific to some intermediate intensity level whose particular value might be innate or associatively learned. Learning such a preference requires an adjustable trans- formation from a monotonic stimulus representation at the sensory periphery to a non-monotonic representation for the motor command. How do neural systems accomplish this task? We tackle this general question focusing on odour-intensity learning in the fruit fly, whose first- and second-order olfactory neurons show monotonic stimulus-response curves. Nevertheless, flies form associative memories specific to particular trained odour intensities. Thus, downstream of the first two olfactory processing layers, odour intensity must be re-coded to enable intensity-specific associative learning. We present a minimal, feed-forward, three-layer circuit, which implements the required transformation by combining excitation, inhibition, and, as a decisive third element, homeostatic plasticity. Key features of this circuit motif are consistent with the known architecture and physiology of the fly olfactory system, whereas alternative mechanisms are either not composed of simple, scalable building blocks or not compatible with physiological observations. The simplicity of the circuit and the robustness of its function under parameter changes make this computational motif an attractive candidate for tuneable non-monotonic intensity coding.

On simplicity and complexity in the brave new world of large-scale neuroscience

Technological advances have dramatically expanded our ability to probe multi-neuronal dynamics and connectivity in the brain. However, our ability to extract a simple conceptual understanding from complex data is increasingly hampered by the lack of theoretically principled data analytic procedures, as well as theoretical frameworks for how circuit connectivity and dynamics can conspire to generate emergent behavioral and cognitive functions. We review and outline potential avenues for progress, including new theories of high dimensional data analysis, the need to analyze complex artificial networks, and methods for analyzing entire spaces of circuit models, rather than one model at a time. Such interplay between experiments, data analysis and theory will be indispensable in catalyzing conceptual advances in the age of large-scale neuroscience.

Behavioural correlates of combinatorial versus temporal features of odour codes

Most sensory stimuli evoke spiking responses that are distributed across neurons and are temporally structured. Whether the temporal structure of ensemble activity is modulated to facilitate different neural computations is not known. Here, we investigated this issue in the insect olfactory system. We found that an odourant can generate synchronous or asynchronous spiking activity across a neural ensemble in the antennal lobe circuit depending on its relative novelty with respect to a preceding stimulus. Regardless of variations in temporal spiking patterns, the activated combinations of neurons robustly represented stimulus identity. Consistent with this interpretation, locusts reliably recognized both solitary and sequential introductions of trained odourants in a quantitative behavioural assay. However, predictable behavioural responses across locusts were observed only to novel stimuli that evoked synchronized spiking patterns across neural ensembles. Hence, our results indicate that the combinatorial ensemble response encodes for stimulus identity, whereas the temporal structure of the ensemble response selectively emphasizes novel stimuli.

In the olfactory system, odourants typically evoke spiking responses in neurons that are both spatially and temporally structured. Here, the authors demonstrate that odour identity is encoded purely by the combinations of neurons activated and is insensitive to changes in temporal structure.

Trade-off between information format and capacity in the olfactory system

As information about the sensory environment passes between layers within the nervous system, the format of the information often changes. To examine how information format affects the capacity of neurons to represent stimuli, we measured the rate of information transmission in olfactory neurons in intact, awake locusts (Schistocerca americana) while pharmacologically manipulating patterns of correlated neuronal activity. Blocking the periodic inhibition underlying odor-elicited neural oscillatory synchronization increased information transmission rates. This suggests oscillatory synchrony, which serves other information processing roles, comes at a cost to the speed with which neurons can transmit information. Our results provide an example of a trade-off between benefits and costs in neural information processing.

Learning modifies odor mixture processing to improve detection of relevant components

Honey bees have a rich repertoire of olfactory learning behaviors, and they therefore are an excellent model to study plasticity in olfactory circuits. Recent behavioral, physiological, and molecular evidence suggested that the antennal lobe, the first relay of the olfactory system in insects and analog to the olfactory bulb in vertebrates, is involved in associative and nonassociative olfactory learning. Here we use calcium imaging to reveal how responses across antennal lobe projection neurons change after association of an input odor with appetitive reinforcement. After appetitive conditioning to 1-hexanol, the representation of an odor mixture containing 1-hexanol becomes more similar to this odor and less similar to the background odor acetophenone. We then apply computational modeling to investigate how changes in synaptic connectivity can account for the observed plasticity. Our study suggests that experience-dependent modulation of inhibitory interactions in the antennal lobe aids perception of salient odor components mixed with behaviorally irrelevant background odors.

Olfactory bulb coding of odors, mixtures and sniffs is a linear sum of odor time profiles

The olfactory system receives intermittent and fluctuating inputs arising from dispersion of odor plumes and active sampling by the animal. Previous work has suggested that the olfactory transduction machinery and excitatory-inhibitory olfactory bulb circuitry generate nonlinear population trajectories of neuronal activity that differ across odorants. Here we show that individual mitral/tufted (M/T) cells sum inputs linearly across odors and time. By decoupling odor sampling from respiration in anesthetized rats, we show that M/T cell responses to arbitrary odor waveforms and mixtures are well described by odor-specific impulse responses convolved with the odorant's temporal profile. The same impulse responses convolved with the respiratory airflow predict the classical respiration-locked firing of olfactory bulb neurons and several other reported response properties of M/T cells. These results show that the olfactory bulb linearly processes fluctuating odor inputs, thereby simplifying downstream decoding of stimulus identity and temporal dynamics.

Learning modulation of odor representations: new findings from Arc-indexed networks

We first review our understanding of odor representations in rodent olfactory bulb (OB) and anterior piriform cortex (APC). We then consider learning-induced representation changes. Finally we describe the perspective on network representations gained from examining Arc-indexed odor networks of awake rats. Arc-indexed networks are sparse and distributed, consistent with current views. However Arc provides representations of repeated odors. Arc-indexed repeated odor representations are quite variable. Sparse representations are assumed to be compact and reliable memory codes. Arc suggests this is not necessarily the case. The variability seen is consistent with electrophysiology in awake animals and may reflect top-down cortical modulation of context. Arc-indexing shows that distinct odors share larger than predicted neuron pools. These may be low-threshold neuronal subsets. Learning’s effect on Arc-indexed representations is to increase the stable or overlapping component of rewarded odor representations. This component can decrease for similar odors when their discrimination is rewarded. The learning effects seen are supported by electrophysiology, but mechanisms remain to be elucidated.

Computational models to understand decision making and pattern recognition in the insect brain

Odor stimuli reaching olfactory systems of mammals and insects are characterized by remarkable non-stationary and noisy time series. Their brains have evolved to discriminate subtle changes in odor mixtures and find meaningful variations in complex spatio-temporal patterns. Insects with small brains can effectively solve two computational tasks: identify the presence of an odor type and estimate the concentration levels of the odor. Understanding the learning and decision making processes in the insect brain can not only help us to uncover general principles of information processing in the brain, but it can also provide key insights to artificial chemical sensing. Both olfactory learning and memory are dominantly organized in the Antennal Lobe (AL) and the Mushroom Bodies (MBs). Current computational models yet fail to deliver an integrated picture of the joint computational roles of the AL and MBs. This review intends to provide an integrative overview of the computational literature analyzed in the context of the problem of classification (odor discrimination) and regression (odor concentration estimation), particularly identifying key computational ingredients necessary to solve pattern recognition.

Central processing in the mushroom bodies

The mushroom bodies in the insect brain serve as a central information processing area. Here, focusing mainly on olfaction, we discuss functionally related roles the mushroom bodies play in signal gain control, response sparsening, the separation of similar signals (decorrelation), and learning and memory. In sum, the mushroom bodies assemble and format a context-appropriate representation of the insect's world.

Odorant recognition using biological responses recorded in olfactory bulb of rats

In this study we applied pattern recognition (PR) techniques to extract odorant information from local field potential (LFP) signals recorded in the olfactory bulb (OB) of rats subjected to different odorant stimuli. We claim that LFP signals registered on the OB, the first stage of olfactory processing, are stimulus specific in animals with normal sensory experience, and that these patterns correspond to the neural substrate likely required for perceptual discrimination. Thus, these signals can be used as input to an artificial odorant classification system with great success. In this paper we have designed and compared the performance of several configurations of artificial olfaction systems (AOS) based on the combination of four feature extraction (FE) methods (Principal Component Analysis (PCA), Fisher Transformation (FT), Sammon NonLinear Map (NLM) and Wavelet Transform (WT)), and three PR techniques (Linear Discriminant Analysis (LDA), Multilayer Perceptron (MLP) and Support Vector Machine (SVM)), when four different stimuli are presented to rats. The best results were reached when PCA extraction followed by SVM as classifier were used, obtaining a classification accuracy of over 95% for all four stimuli.

Robust sequential working memory recall in heterogeneous cognitive networks

Psychiatric disorders are often caused by partial heterogeneous disinhibition in cognitive networks, controlling sequential and spatial working memory (SWM). Such dynamic connectivity changes suggest that the normal relationship between the neuronal components within the network deteriorates. As a result, competitive network dynamics is qualitatively altered. This dynamics defines the robust recall of the sequential information from memory and, thus, the SWM capacity. To understand pathological and non-pathological bifurcations of the sequential memory dynamics, here we investigate the model of recurrent inhibitory-excitatory networks with heterogeneous inhibition. We consider the ensemble of units with all-to-all inhibitory connections, in which the connection strengths are monotonically distributed at some interval. Based on computer experiments and studying the Lyapunov exponents, we observed and analyzed the new phenomenon—clustered sequential dynamics. The results are interpreted in the context of the winnerless competition principle. Accordingly, clustered sequential dynamics is represented in the phase space of the model by two weakly interacting quasi-attractors. One of them is similar to the sequential heteroclinic chain—the regular image of SWM, while the other is a quasi-chaotic attractor. Coexistence of these quasi-attractors means that the recall of the normal information sequence is intermittently interrupted by episodes with chaotic dynamics. We indicate potential dynamic ways for augmenting damaged working memory and other cognitive functions.

Dimensionality reduction for large-scale neural recordings

Most sensory, cognitive and motor functions depend on the interactions of many neurons. In recent years, there has been rapid development and increasing use of technologies for recording from large numbers of neurons, either sequentially or simultaneously. A key question is what scientific insight can be gained by studying a population of recorded neurons beyond studying each neuron individually. Here, we examine three important motivations for population studies: single-trial hypotheses requiring statistical power, hypotheses of population response structure and exploratory analyses of large data sets. Many recent studies have adopted dimensionality reduction to analyze these populations and to find features that are not apparent at the level of individual neurons. We describe the dimensionality reduction methods commonly applied to population activity and offer practical advice about selecting methods and interpreting their outputs. This review is intended for experimental and computational researchers who seek to understand the role dimensionality reduction has had and can have in systems neuroscience, and who seek to apply these methods to their own data.

Understanding odor information segregation in the olfactory bulb by means of mitral and tufted cells

Odor identification is one of the main tasks of the olfactory system. It is performed almost independently from the concentration of the odor providing a robust recognition. This capacity to ignore concentration information does not preclude the olfactory system from estimating concentration itself. Significant experimental evidence has indicated that the olfactory system is able to infer simultaneously odor identity and intensity. However, it is still unclear at what level or levels of the olfactory pathway this segregation of information occurs. In this work, we study whether this odor information segregation is performed at the input stage of the olfactory bulb: the glomerular layer. To this end, we built a detailed neural model of the glomerular layer based on its known anatomical connections and conducted two simulated odor experiments. In the first experiment, the model was exposed to an odor stimulus dataset composed of six different odorants, each one dosed at six different concentrations. In the second experiment, we conducted an odor morphing experiment where a sequence of binary mixtures going from one odor to another through intermediate mixtures was presented to the model. The results of the experiments were visualized using principal components analysis and analyzed with hierarchical clustering to unveil the structure of the high-dimensional output space. Additionally, Fisher's discriminant ratio and Pearson's correlation coefficient were used to quantify odor identity and odor concentration information respectively. Our results showed that the architecture of the glomerular layer was able to mediate the segregation of odor information obtaining output spiking sequences of the principal neurons, namely the mitral and external tufted cells, strongly correlated with odor identity and concentration, respectively. An important conclusion is also that the morphological difference between the principal neurons is not key to achieve odor information segregation.

Retronasal odor concentration coding in glomeruli of the rat olfactory bulb

The mammalian olfactory system processes odorants presented orthonasally (inhalation through the nose) and also retronasally (exhalation), enabling identification of both external as well as internal objects during food consumption. There are distinct differences between ortho- and retronasal air flow patterns, psychophysics, multimodal integration, and glomerular responses. Recent work indicates that rats can also detect odors retronasally, that rats can associate retronasal odors with tastes, and that their olfactory bulbs (OBs) can respond to retronasal odorants but differently than to orthonasal odors. To further characterize retronasal OB input activity patterns, experiments here focus on determining the effects of odor concentration on glomerular activity by monitoring calcium activity in the dorsal OB of rats using a dextran-conjugated calcium-sensitive dye in vivo. Results showed reliable concentration-response curves that differed between odorants, and recruitment of additional glomeruli, as odor concentration increased. We found evidence of different concentration-response functions between glomeruli, that in turn depended on odor. Further, the relation between dynamics and concentration differed remarkably among retronasal odorants. These dynamics are suggested to reduce the odor map ambiguity based on response amplitude. Elucidating the coding of retronasal odor intensity is fundamental to the understanding of feeding behavior and the neural basis of flavor. These data further establish and refine the rodent model of flavor neuroscience.

A temporal channel for information in sparse sensory coding

BACKGROUND

Sparse codes are found in nearly every sensory system, but the role of spike timing in sparse sensory coding is unclear. Here, we use the olfactory system of awake locusts to test whether the timing of spikes in Kenyon cells, a population of neurons that responds sparsely to odors, carries sensory information to and influences the responses of follower neurons.

RESULTS

We characterized two major classes of direct followers of Kenyon cells. With paired intracellular and field potential recordings made during odor presentations, we found that these followers contain information about odor identity in the temporal patterns of their spikes rather than in the spike rate, the spike phase, or the identities of the responsive neurons. Subtly manipulating the relative timing of Kenyon cell spikes with temporally and spatially structured microstimulation reliably altered the response patterns of the followers.

CONCLUSIONS

Our results show that even remarkably sparse spiking responses can provide information through stimulus-specific variations in timing on the order of tens to hundreds of milliseconds and that these variations can determine the responses of downstream neurons. These results establish the importance of spike timing in a sparse sensory code.

Precise spatial coding is preserved along the longitudinal hippocampal axis

Compared with the dorsal hippocampus, relatively few studies have characterized neuronal responses in the ventral hippocampus. In particular, it is unclear whether and how cells in the ventral region represent space and/or respond to contextual changes. We recorded from dorsal and ventral CA1 neurons in freely moving mice exposed to manipulations of visuospatial and olfactory contexts. We found that ventral cells respond to alterations of the visuospatial environment such as exposure to novel local cues, cue rotations, and contextual expansion in similar ways to dorsal cells, with the exception of cue rotations. Furthermore, we found that ventral cells responded to odors much more strongly than dorsal cells, particularly to odors of high valence. Similar to earlier studies recording from the ventral hippocampus in CA3, we also found increased scaling of place cell field size along the longitudinal hippocampal axis. Although the increase in place field size observed toward the ventral pole has previously been taken to suggest a decrease in spatial information coded by ventral place cells, we hypothesized that a change in spatial scaling could instead signal a shift in representational coding that preserves the resolution of spatial information. To explore this possibility, we examined population activity using principal component analysis (PCA) and neural location reconstruction techniques. Our results suggest that ventral populations encode a distributed representation of space, and that the resolution of spatial information at the population level is comparable to that of dorsal populations of similar size. Finally, through the use of neural network modeling, we suggest that the redundancy in spatial representation along the longitudinal hippocampal axis may allow the hippocampus to overcome the conflict between memory interference and generalization inherent in neural network memory. Our results indicate that ventral population activity is well suited for generalization across locations and contexts.

From molecule to mind: an integrative perspective on odor intensity

A fundamental problem in systems neuroscience is mapping the physical properties of a stimulus to perceptual characteristics. In vision, wavelength translates into color; in audition, frequency translates into pitch. Although odorant concentration is a key feature of olfactory stimuli, we do not know how concentration is translated into perceived intensity by the olfactory system. A variety of neural responses at several levels of processing have been reported to vary with odorant concentration, suggesting specific coding models. However, it remains unclear which, if any, of these phenomena underlie the perception of odor intensity. Here, we provide an overview of current models at different stages of olfactory processing, and identify promising avenues for future research.

Antennal lobe representations are optimized when olfactory stimuli are periodically structured to simulate natural wing beat effects

Animals use behaviors to actively sample the environment across a broad spectrum of sensory domains. These behaviors discretize the sensory experience into unique spatiotemporal moments, minimize sensory adaptation, and enhance perception. In olfaction, behaviors such as sniffing, antennal flicking, and wing beating all act to periodically expose olfactory epithelium. In mammals, it is thought that sniffing enhances neural representations; however, the effects of insect wing beating on representations remain unknown. To determine how well the antennal lobe (AL) produces odor dependent representations when wing beating effects are simulated, we used extracellular methods to record neural units and local field potentials (LFPs) from moth AL. We recorded responses to odors presented as prolonged continuous stimuli or periodically as 20 and 25 Hz pulse trains designed to simulate the oscillating effects of wing beating around the antennae during odor guided flight. Using spectral analyses, we show that ~25% of all recorded units were able to entrain to "pulsed stimuli"; this includes pulsed blanks, which elicited the strongest overall entrainment. The strength of entrainment to pulse train stimuli was dependent on molecular features of the odorants, odor concentration, and pulse train duration. Moreover, units showing pulse tracking responses were highly phase locked to LFPs during odor stimulation, indicating that unit-LFP phase relationships are stimulus-driven. Finally, a Euclidean distance-based population vector analysis established that AL odor representations are more robust, peak more quickly, and do not show adaptation when odors were presented at the natural wing beat frequency as opposed to prolonged continuous stimulation. These results suggest a general strategy for optimizing olfactory representations, which exploits the natural rhythmicity of wing beating by integrating mechanosensory and olfactory cues at the level of the AL.

Coding and transformations in the olfactory system

How is sensory information represented in the brain? A long-standing debate in neural coding is whether and how timing of spikes conveys information to downstream neurons. Although we know that neurons in the olfactory bulb (OB) exhibit rich temporal dynamics, the functional relevance of temporal coding remains hotly debated. Recent recording experiments in awake behaving animals have elucidated highly organized temporal structures of activity in the OB. In addition, the analysis of neural circuits in the piriform cortex (PC) demonstrated the importance of not only OB afferent inputs but also intrinsic PC neural circuits in shaping odor responses. Furthermore, new experiments involving stimulation of the OB with specific temporal patterns allowed for testing the relevance of temporal codes. Together, these studies suggest that the relative timing of neuronal activity in the OB conveys odor information and that neural circuits in the PC possess various mechanisms to decode temporal patterns of OB input.

Neural decoding with kernel-based metric learning

In studies of the nervous system, the choice of metric for the neural responses is a pivotal assumption. For instance, a well-suited distance metric enables us to gauge the similarity of neural responses to various stimuli and assess the variability of responses to a repeated stimulus-exploratory steps in understanding how the stimuli are encoded neurally. Here we introduce an approach where the metric is tuned for a particular neural decoding task. Neural spike train metrics have been used to quantify the information content carried by the timing of action potentials. While a number of metrics for individual neurons exist, a method to optimally combine single-neuron metrics into multineuron, or population-based, metrics is lacking. We pose the problem of optimizing multineuron metrics and other metrics using centered alignment, a kernel-based dependence measure. The approach is demonstrated on invasively recorded neural data consisting of both spike trains and local field potentials. The experimental paradigm consists of decoding the location of tactile stimulation on the forepaws of anesthetized rats. We show that the optimized metrics highlight the distinguishing dimensions of the neural response, significantly increase the decoding accuracy, and improve nonlinear dimensionality reduction methods for exploratory neural analysis.

Selectivity and sparseness in randomly connected balanced networks

Neurons in sensory cortex show stimulus selectivity and sparse population response, even in cases where no strong functionally specific structure in connectivity can be detected. This raises the question whether selectivity and sparseness can be generated and maintained in randomly connected networks. We consider a recurrent network of excitatory and inhibitory spiking neurons with random connectivity, driven by random projections from an input layer of stimulus selective neurons. In this architecture, the stimulus-to-stimulus and neuron-to-neuron modulation of total synaptic input is weak compared to the mean input. Surprisingly, we show that in the balanced state the network can still support high stimulus selectivity and sparse population response. In the balanced state, strong synapses amplify the variation in synaptic input and recurrent inhibition cancels the mean. Functional specificity in connectivity emerges due to the inhomogeneity caused by the generative statistical rule used to build the network. We further elucidate the mechanism behind and evaluate the effects of model parameters on population sparseness and stimulus selectivity. Network response to mixtures of stimuli is investigated. It is shown that a balanced state with unselective inhibition can be achieved with densely connected input to inhibitory population. Balanced networks exhibit the “paradoxical” effect: an increase in excitatory drive to inhibition leads to decreased inhibitory population firing rate. We compare and contrast selectivity and sparseness generated by the balanced network to randomly connected unbalanced networks. Finally, we discuss our results in light of experiments.

Spontaneous activity governs olfactory representations in spatially organized habenular microcircuits

The medial habenula relays information from the sensory areas via the interpeduncular nucleus to the periaqueductal gray that regulates animal behavior under stress conditions. Ablation of the dorsal habenula (dHb) in zebrafish, which is equivalent to the mammalian medial habenula, was shown to perturb experience-dependent fear. Therefore, understanding dHb function is important for understanding the neural basis of fear. In zebrafish, the dHb receives inputs from the mitral cells (MCs) of the olfactory bulb (OB), and odors can trigger distinct behaviors (e.g., feeding, courtship, alarm). However, it is unclear how the dHb processes olfactory information and how these computations relate to behavior. In this study, we demonstrate that the odor responses in the dHb are asymmetric and spatially organized despite the unorganized OB inputs. Moreover, we show that the spontaneous dHb activity is not random but structured into functionally and spatially organized clusters of neurons, which reflects the favored states of the dHb network. These dHb clusters are also preserved during odor stimulation and govern olfactory responses. Finally, we show that functional dHb clusters overlap with genetically defined dHb neurons, which regulate experience-dependent fear. Thus, we propose that the dHb is composed of functionally, spatially, and genetically distinct microcircuits that regulate different behavioral programs.

Neural circuits as computational dynamical systems

Many recent studies of neurons recorded from cortex reveal complex temporal dynamics. How such dynamics embody the computations that ultimately lead to behavior remains a mystery. Approaching this issue requires developing plausible hypotheses couched in terms of neural dynamics. A tool ideally suited to aid in this question is the recurrent neural network (RNN). RNNs straddle the fields of nonlinear dynamical systems and machine learning and have recently seen great advances in both theory and application. I summarize recent theoretical and technological advances and highlight an example of how RNNs helped to explain perplexing high-dimensional neurophysiological data in the prefrontal cortex.

Cooperative integration and representation underlying bilateral network of fly motion-sensitive neurons

How is binocular motion information integrated in the bilateral network of wide-field motion-sensitive neurons, called lobula plate tangential cells (LPTCs), in the visual system of flies? It is possible to construct an accurate model of this network because a complete picture of synaptic interactions has been experimentally identified. We investigated the cooperative behavior of the network of horizontal LPTCs underlying the integration of binocular motion information and the information representation in the bilateral LPTC network through numerical simulations on the network model. First, we qualitatively reproduced rotational motion-sensitive response of the H2 cell previously reported in vivo experiments and ascertained that it could be accounted for by the cooperative behavior of the bilateral network mainly via interhemispheric electrical coupling. We demonstrated that the response properties of single H1 and Hu cells, unlike H2 cells, are not influenced by motion stimuli in the contralateral visual hemi-field, but that the correlations between these cell activities are enhanced by the rotational motion stimulus. We next examined the whole population activity by performing principal component analysis (PCA) on the population activities of simulated LPTCs. We showed that the two orthogonal patterns of correlated population activities given by the first two principal components represent the rotational and translational motions, respectively, and similar to the H2 cell, rotational motion produces a stronger response in the network than does translational motion. Furthermore, we found that these population-coding properties are strongly influenced by the interhemispheric electrical coupling. Finally, to test the generality of our conclusions, we used a more simplified model and verified that the numerical results are not specific to the network model we constructed.

Searching for collective behavior in a large network of sensory neurons

Maximum entropy models are the least structured probability distributions that exactly reproduce a chosen set of statistics measured in an interacting network. Here we use this principle to construct probabilistic models which describe the correlated spiking activity of populations of up to 120 neurons in the salamander retina as it responds to natural movies. Already in groups as small as 10 neurons, interactions between spikes can no longer be regarded as small perturbations in an otherwise independent system; for 40 or more neurons pairwise interactions need to be supplemented by a global interaction that controls the distribution of synchrony in the population. Here we show that such "K-pairwise" models--being systematic extensions of the previously used pairwise Ising models--provide an excellent account of the data. We explore the properties of the neural vocabulary by: 1) estimating its entropy, which constrains the population's capacity to represent visual information; 2) classifying activity patterns into a small set of metastable collective modes; 3) showing that the neural codeword ensembles are extremely inhomogenous; 4) demonstrating that the state of individual neurons is highly predictable from the rest of the population, allowing the capacity for error correction.

Context-dependent computation by recurrent dynamics in prefrontal cortex

Prefrontal cortex is thought to have a fundamental role in flexible, context-dependent behaviour, but the exact nature of the computations underlying this role remains largely unknown. In particular, individual prefrontal neurons often generate remarkably complex responses that defy deep understanding of their contribution to behaviour. Here we study prefrontal cortex activity in macaque monkeys trained to flexibly select and integrate noisy sensory inputs towards a choice. We find that the observed complexity and functional roles of single neurons are readily understood in the framework of a dynamical process unfolding at the level of the population. The population dynamics can be reproduced by a trained recurrent neural network, which suggests a previously unknown mechanism for selection and integration of task-relevant inputs. This mechanism indicates that selection and integration are two aspects of a single dynamical process unfolding within the same prefrontal circuits, and potentially provides a novel, general framework for understanding context-dependent computations.

DataHigh: graphical user interface for visualizing and interacting with high-dimensional neural activity

OBJECTIVE

Analyzing and interpreting the activity of a heterogeneous population of neurons can be challenging, especially as the number of neurons, experimental trials, and experimental conditions increases. One approach is to extract a set of latent variables that succinctly captures the prominent co-fluctuation patterns across the neural population. A key problem is that the number of latent variables needed to adequately describe the population activity is often greater than 3, thereby preventing direct visualization of the latent space. By visualizing a small number of 2-d projections of the latent space or each latent variable individually, it is easy to miss salient features of the population activity.

APPROACH

To address this limitation, we developed a Matlab graphical user interface (called DataHigh) that allows the user to quickly and smoothly navigate through a continuum of different 2-d projections of the latent space. We also implemented a suite of additional visualization tools (including playing out population activity timecourses as a movie and displaying summary statistics, such as covariance ellipses and average timecourses) and an optional tool for performing dimensionality reduction.

MAIN RESULTS

To demonstrate the utility and versatility of DataHigh, we used it to analyze single-trial spike count and single-trial timecourse population activity recorded using a multi-electrode array, as well as trial-averaged population activity recorded using single electrodes.

SIGNIFICANCE

DataHigh was developed to fulfil a need for visualization in exploratory neural data analysis, which can provide intuition that is critical for building scientific hypotheses and models of population activity.

Olfactory memories are intensity specific in larval Drosophila

Learning can rely on stimulus quality, stimulus intensity, or a combination of these. Regarding olfaction, the coding of odour quality is often proposed to be combinatorial along the olfactory pathway, and working hypotheses are available concerning short-term associative memory trace formation of odour quality. However, it is less clear how odour intensity is coded, and whether olfactory memory traces include information about the intensity of the learnt odour. Using odour-sugar associative conditioning in larval Drosophila, we first describe the dose-effect curves of learnability across odour intensities for four different odours (n-amyl acetate, 3-octanol, 1-octen-3-ol and benzaldehyde). We then chose odour intensities such that larvae were trained at an intermediate odour intensity, but were tested for retention with either that trained intermediate odour intensity, or with respectively higher or lower intensities. We observed a specificity of retention for the trained intensity for all four odours used. This adds to the appreciation of the richness in 'content' of olfactory short-term memory traces, even in a system as simple as larval Drosophila, and to define the demands on computational models of associative olfactory memory trace formation. We suggest two kinds of circuit architecture that have the potential to accommodate intensity learning, and discuss how they may be implemented in the insect brain.

Theoretical foundations of the sound analog membrane potential that underlies coincidence detection in the barn owl

A wide variety of neurons encode temporal information via phase-locked spikes. In the avian auditory brainstem, neurons in the cochlear nucleus magnocellularis (NM) send phase-locked synaptic inputs to coincidence detector neurons in the nucleus laminaris (NL) that mediate sound localization. Previous modeling studies suggested that converging phase-locked synaptic inputs may give rise to a periodic oscillation in the membrane potential of their target neuron. Recent physiological recordings in vivo revealed that owl NL neurons changed their spike rates almost linearly with the amplitude of this oscillatory potential. The oscillatory potential was termed the sound analog potential, because of its resemblance to the waveform of the stimulus tone. The amplitude of the sound analog potential recorded in NL varied systematically with the interaural time difference (ITD), which is one of the most important cues for sound localization. In order to investigate the mechanisms underlying ITD computation in the NM-NL circuit, we provide detailed theoretical descriptions of how phase-locked inputs form oscillating membrane potentials. We derive analytical expressions that relate presynaptic, synaptic, and postsynaptic factors to the signal and noise components of the oscillation in both the synaptic conductance and the membrane potential. Numerical simulations demonstrate the validity of the theoretical formulations for the entire frequency ranges tested (1–8 kHz) and potential effects of higher harmonics on NL neurons with low best frequencies (<2 kHz).

Encoding of mixtures in a simple olfactory system

Natural odors are usually mixtures; yet, humans and animals can experience them as unitary percepts. Olfaction also enables stimulus categorization and generalization. We studied how these computations are performed with the responses of 168 locust antennal lobe projection neurons (PNs) to varying mixtures of two monomolecular odors, and of 174 PNs and 209 mushroom body Kenyon cells (KCs) to mixtures of up to eight monomolecular odors. Single-PN responses showed strong hypoadditivity and population trajectories clustered by odor concentration and mixture similarity. KC responses were much sparser on average than those of PNs and often signaled the presence of single components in mixtures. Linear classifiers could read out the responses of both populations in single time bins to perform odor identification, categorization, and generalization. Our results suggest that odor representations in the mushroom body may result from competing optimization constraints to facilitate memorization (sparseness) while enabling identification, classification, and generalization.

A spatiotemporal coding mechanism for background-invariant odor recognition

Sensory stimuli evoke neural activity that evolves over time. What features of these spatiotemporal responses allow the robust encoding of stimulus identity in a multistimulus environment? Here we examined this issue in the locust (Schistocerca americana) olfactory system. We found that sensory responses evoked by an odorant (foreground) varied when presented atop or after an ongoing stimulus (background). These inconsistent sensory inputs triggered dynamic reorganization of ensemble activity in the downstream antennal lobe. As a result, partial pattern matches between neural representations encoding the same foreground stimulus across conditions were achieved. The degree and segments of response overlaps varied; however, any overlap observed was sufficient to drive background-independent responses in the downstream neural population. Notably, recognition performance of locusts in behavioral assays correlated well with our physiological findings. Hence, our results reveal how background-independent recognition of odors can be achieved using spatiotemporal patterns of neural activity.

Reliable sex and strain discrimination in the mouse vomeronasal organ and accessory olfactory bulb

Animals modulate their courtship and territorial behaviors in response to olfactory cues produced by other animals. In rodents, detecting these cues is the primary role of the accessory olfactory system (AOS). We sought to systematically investigate the natural stimulus coding logic and robustness in neurons of the first two stages of accessory olfactory processing, the vomeronasal organ (VNO) and accessory olfactory bulb (AOB). We show that firing rate responses of just a few well-chosen mouse VNO or AOB neurons can be used to reliably encode both sex and strain of other mice from cues contained in urine. Additionally, we show that this population code can generalize to new concentrations of stimuli and appears to represent stimulus identity in terms of diverging paths in coding space. Together, the results indicate that firing rate code on the temporal order of seconds is sufficient for accurate classification of pheromonal patterns at different concentrations and may be used by AOS neural circuitry to discriminate among naturally occurring urine stimuli.

Robust encoding of stimulus identity and concentration in the accessory olfactory system

Sensory systems represent stimulus identity and intensity, but in the neural periphery these two variables are typically intertwined. Moreover, stable detection may be complicated by environmental uncertainty; stimulus properties can differ over time and circumstance in ways that are not necessarily biologically relevant. We explored these issues in the context of the mouse accessory olfactory system, which specializes in detection of chemical social cues and infers myriad aspects of the identity and physiological state of conspecifics from complex mixtures, such as urine. Using mixtures of sulfated steroids, key constituents of urine, we found that spiking responses of individual vomeronasal sensory neurons encode both individual compounds and mixtures in a manner consistent with a simple model of receptor-ligand interactions. Although typical neurons did not accurately encode concentration over a large dynamic range, from population activity it was possible to reliably estimate the log-concentration of pure compounds over several orders of magnitude. For binary mixtures, simple models failed to accurately segment the individual components, largely because of the prevalence of neurons responsive to both components. By accounting for such overlaps during model tuning, we show that, from neuronal firing, one can accurately estimate log-concentration of both components, even when tested across widely varying concentrations. With this foundation, the difference of logarithms, log A - log B = log A/B, provides a natural mechanism to accurately estimate concentration ratios. Thus, we show that a biophysically plausible circuit model can reconstruct concentration ratios from observed neuronal firing, representing a powerful mechanism to separate stimulus identity from absolute concentration.

Principal component analysis of odor coding at the level of third-order olfactory neurons in Drosophila

Olfactory information in Drosophila is conveyed by projection neurons from olfactory sensory neurons to Kenyon cells (KCs) in the mushroom body (MB). A subset of KCs responds to a given odor molecule, and the combination of these KCs represents a part of the neuronal olfactory code. KCs are also thought to function as coincidence detectors for memory formation, associating odor information with a coincident punishment or reward stimulus. Associative conditioning has been shown to modify KC output. This plasticity occurs in the vertical lobes of MBs containing α/α' branches of KCs, which is shown by measuring the average Ca(2+) levels in the branch of each lobe. We devised a method to quantitatively describe the population activity patterns recorded from axons of >1000 KCs at the α/α' branches using two-photon Ca(2+) imaging. Principal component analysis of the population activity patterns clearly differentiated the responses to distinct odors.

Sparse and background-invariant coding of vocalizations in auditory scenes

Vocal communicators such as humans and songbirds readily recognize individual vocalizations, even in distracting auditory environments. This perceptual ability is likely subserved by auditory neurons whose spiking responses to individual vocalizations are minimally affected by background sounds. However, auditory neurons that produce background-invariant responses to vocalizations in auditory scenes have not been found. Here, we describe a population of neurons in the zebra finch auditory cortex that represent vocalizations with a sparse code and that maintain their vocalization-like firing patterns in levels of background sound that permit behavioral recognition. These same neurons decrease or stop spiking in levels of background sound that preclude behavioral recognition. In contrast, upstream neurons represent vocalizations with dense and background-corrupted responses. We provide experimental evidence suggesting that sparse coding is mediated by feedforward suppression. Finally, we show through simulations that feedforward inhibition can transform a dense representation of vocalizations into a sparse and background-invariant representation.

Mushroom body miscellanea: transgenic Drosophila strains expressing anatomical and physiological sensor proteins in Kenyon cells

The fruit fly Drosophila melanogaster represents a key model organism for analyzing how neuronal circuits regulate behavior. The mushroom body in the central brain is a particularly prominent brain region that has been intensely studied in several insect species and been implicated in a variety of behaviors, e.g., associative learning, locomotor activity, and sleep. Drosophila melanogaster offers the advantage that transgenes can be easily expressed in neuronal subpopulations, e.g., in intrinsic mushroom body neurons (Kenyon cells). A number of transgenes has been described and engineered to visualize the anatomy of neurons, to monitor physiological parameters of neuronal activity, and to manipulate neuronal function artificially. To target the expression of these transgenes selectively to specific neurons several sophisticated bi- or even multipartite transcription systems have been invented. However, the number of transgenes that can be combined in the genome of an individual fly is limited in practice. To facilitate the analysis of the mushroom body we provide a compilation of transgenic fruit flies that express transgenes under direct control of the Kenyon-cell specific promoter, mb247. The transgenes expressed are fluorescence reporters to analyze neuroanatomical aspects of the mushroom body, proteins to restrict ectopic gene expression to mushroom bodies, or fluorescent sensors to monitor physiological parameters of neuronal activity of Kenyon cells. Some of the transgenic animals compiled here have been published already, whereas others are novel and characterized here for the first time. Overall, the collection of transgenic flies expressing sensor and reporter genes in Kenyon cells facilitates combinations with binary transcription systems and might, ultimately, advance the physiological analysis of mushroom body function.

Local interneurons and projection neurons in the antennal lobe from a spiking point of view

Local computation in microcircuits is an essential feature of distributed information processing in vertebrate and invertebrate brains. The insect antennal lobe represents a spatially confined local network that processes high-dimensional and redundant peripheral input to compute an efficient odor code. Social insects can rely on a particularly rich olfactory receptor repertoire, and they exhibit complex odor-guided behaviors. This corresponds with a high anatomical complexity of their antennal lobe network. In the honeybee, a large number of glomeruli that receive sensory input are interconnected by a dense network of local interneurons (LNs). Uniglomerular projection neurons (PNs) integrate sensory and recurrent local network input into an efficient spatio-temporal odor code. To investigate the specific computational roles of LNs and PNs, we measured several features of sub- and suprathreshold single-cell responses to in vivo odor stimulation. Using a semisupervised cluster analysis, we identified a combination of five characteristic features as sufficient to separate LNs and PNs from each other, independent of the applied odor-stimuli. The two clusters differed significantly in all these five features. PNs showed a higher spontaneous subthreshold activation, assumed higher peak response rates and a more regular spiking pattern. LNs reacted considerably faster to the onset of a stimulus, and their responses were more reliable across stimulus repetitions. We discuss possible mechanisms that can explain our results, and we interpret cell-type-specific characteristics with respect to their functional relevance.

Imaging a population code for odor identity in the Drosophila mushroom body

The brain represents sensory information in the coordinated activity of neuronal ensembles. Although the microcircuits underlying olfactory processing are well characterized in Drosophila, no studies to date have examined the encoding of odor identity by populations of neurons and related it to the odor specificity of olfactory behavior. Here we used two-photon Ca(2+) imaging to record odor-evoked responses from >100 neurons simultaneously in the Drosophila mushroom body (MB). For the first time, we demonstrate quantitatively that MB population responses contain substantial information on odor identity. Using a series of increasingly similar odor blends, we identified conditions in which odor discrimination is difficult behaviorally. We found that MB ensemble responses accounted well for olfactory acuity in this task. Kenyon cell ensembles with as few as 25 cells were sufficient to match behavioral discrimination accuracy. Using a generalization task, we demonstrated that the MB population code could predict the flies' responses to novel odors. The degree to which flies generalized a learned aversive association to unfamiliar test odors depended upon the relative similarity between the odors' evoked MB activity patterns. Discrimination and generalization place different demands on the animal, yet the flies' choices in these tasks were reliably predicted based on the amount of overlap between MB activity patterns. Therefore, these different behaviors can be understood in the context of a single physiological framework.

Odor representations in the olfactory bulb evolve after the first breath and persist as an odor afterimage

Rodents can discriminate odors in one breath, and mammalian olfaction research has thus focused on the first breath. However, sensory representations dynamically change during and after stimuli. To investigate these dynamics, we recorded spike trains from the olfactory bulb of awake, head-fixed mice and found that some mitral cells' odor representations changed following the first breath and others continued after odor cessation. Population analysis revealed that these postodor responses contained odor- and concentration-specific information--an odor afterimage. Using calcium imaging, we found that most olfactory glomerular activity was restricted to the odor presentation, implying that the afterimage is not primarily peripheral. The odor afterimage was not dependent on odorant physicochemical properties. To artificially induce aftereffects, we photostimulated mitral cells using channelrhodopsin and recorded centrally maintained persistent activity. The strength and persistence of the afterimage was dependent on the duration of both artificial and natural stimulation. In summary, we show that the odor representation evolves after the first breath and that there is a centrally maintained odor afterimage, similar to other sensory systems. These dynamics may help identify novel odorants in complex environments.

DataHigh: graphical user interface for visualizing and interacting with high-dimensional neural activity

The activity of tens to hundreds of neurons can be succinctly summarized by a smaller number of latent variables extracted using dimensionality reduction methods. These latent variables define a reduced-dimensional space in which we can study how population activity varies over time, across trials, and across experimental conditions. Ideally, we would like to visualize the population activity directly in the reduced-dimensional space, whose optimal dimensionality (as determined from the data) is typically greater than 3. However, direct plotting can only provide a 2D or 3D view. To address this limitation, we developed a Matlab graphical user interface (GUI) that allows the user to quickly navigate through a continuum of different 2D projections of the reduced-dimensional space. To demonstrate the utility and versatility of this GUI, we applied it to visualize population activity recorded in premotor and motor cortices during reaching tasks. Examples include single-trial population activity recorded using a multi-electrode array, as well as trial-averaged population activity recorded sequentially using single electrodes. Because any single 2D projection may provide a misleading impression of the data, being able to see a large number of 2D projections is critical for intuition-and hypothesis-building during exploratory data analysis. The GUI includes a suite of additional interactive tools, including playing out population activity timecourses as a movie and displaying summary statistics, such as covariance ellipses and average timecourses. The use of visualization tools like the GUI developed here, in tandem with dimensionality reduction methods, has the potential to further our understanding of neural population activity.

Gain control network conditions in early sensory coding

Gain control is essential for the proper function of any sensory system. However, the precise mechanisms for achieving effective gain control in the brain are unknown. Based on our understanding of the existence and strength of connections in the insect olfactory system, we analyze the conditions that lead to controlled gain in a randomly connected network of excitatory and inhibitory neurons. We consider two scenarios for the variation of input into the system. In the first case, the intensity of the sensory input controls the input currents to a fixed proportion of neurons of the excitatory and inhibitory populations. In the second case, increasing intensity of the sensory stimulus will both, recruit an increasing number of neurons that receive input and change the input current that they receive. Using a mean field approximation for the network activity we derive relationships between the parameters of the network that ensure that the overall level of activity of the excitatory population remains unchanged for increasing intensity of the external stimulation. We find that, first, the main parameters that regulate network gain are the probabilities of connections from the inhibitory population to the excitatory population and of the connections within the inhibitory population. Second, we show that strict gain control is not achievable in a random network in the second case, when the input recruits an increasing number of neurons. Finally, we confirm that the gain control conditions derived from the mean field approximation are valid in simulations of firing rate models and Hodgkin-Huxley conductance based models.

Rapid processing of chemosensor transients in a neuromorphic implementation of the insect macroglomerular complex

We present a biologically-constrained neuromorphic spiking model of the insect antennal lobe macroglomerular complex that encodes concentration ratios of chemical components existing within a blend, implemented using a set of programmable logic neuronal modeling cores. Depending upon the level of inhibition and symmetry in its inhibitory connections, the model exhibits two dynamical regimes: fixed point attractor (winner-takes-all type), and limit cycle attractor (winnerless competition type) dynamics. We show that, when driven by chemosensor input in real-time, the dynamical trajectories of the model's projection neuron population activity accurately encode the concentration ratios of binary odor mixtures in both dynamical regimes. By deploying spike timing-dependent plasticity in a subset of the synapses in the model, we demonstrate that a Hebbian-like associative learning rule is able to organize weights into a stable configuration after exposure to a randomized training set comprising a variety of input ratios. Examining the resulting local interneuron weights in the model shows that each inhibitory neuron competes to represent possible ratios across the population, forming a ratiometric representation via mutual inhibition. After training the resulting dynamical trajectories of the projection neuron population activity show amplification and better separation in their response to inputs of different ratios. Finally, we demonstrate that by using limit cycle attractor dynamics, it is possible to recover and classify blend ratio information from the early transient phases of chemosensor responses in real-time more rapidly and accurately compared to a nearest-neighbor classifier applied to the normalized chemosensor data. Our results demonstrate the potential of biologically-constrained neuromorphic spiking models in achieving rapid and efficient classification of early phase chemosensor array transients with execution times well beyond biological timescales.

Parallel neural pathways mediate CO2 avoidance responses in Drosophila

Different stimulus intensities elicit distinct perceptions, implying that input signals are either conveyed through an overlapping but distinct subpopulation of sensory neurons or channeled into divergent brain circuits according to intensity. In Drosophila, carbon dioxide (CO2) is detected by a single type of olfactory sensory neuron, but information is conveyed to higher brain centers through second-order projection neurons (PNs). Two distinct pathways, PN(v)-1 and PN(v)-2, are necessary and sufficient for avoidance responses to low and high CO2 concentrations, respectively. Whereas low concentrations activate PN(v)-1, high concentrations activate both PN(v)s and GABAergic PN(v)-3, which may inhibit PN(v)-1 pathway-mediated avoidance behavior. Channeling a sensory input into distinct neural pathways allows the perception of an odor to be further modulated by both stimulus intensity and context.