Neural mechanisms for filtering self-generated sensory signals in cerebellum-like circuits

An internal model architecture for novelty detection: implications for cerebellar and collicular roles in sensory processing

The cerebellum is thought to implement internal models for sensory prediction, but details of the underlying circuitry are currently obscure. We therefore investigated a specific example of internal-model based sensory prediction, namely detection of whisker contacts during whisking. Inputs from the vibrissae in rats can be affected by signals generated by whisker movement, a phenomenon also observable in whisking robots. Robot novelty-detection can be improved by adaptive noise-cancellation, in which an adaptive filter learns a forward model of the whisker plant that allows the sensory effects of whisking to be predicted and thus subtracted from the noisy sensory input. However, the forward model only uses information from an efference copy of the whisking commands. Here we show that the addition of sensory information from the whiskers allows the adaptive filter to learn a more complex internal model that performs more robustly than the forward model, particularly when the whisking-induced interference has a periodic structure. We then propose a neural equivalent of the circuitry required for adaptive novelty-detection in the robot, in which the role of the adaptive filter is carried out by the cerebellum, with the comparison of its output (an estimate of the self-induced interference) and the original vibrissal signal occurring in the superior colliculus, a structure noted for its central role in novelty detection. This proposal makes a specific prediction concerning the whisker-related functions of a region in cerebellar cortical zone A₂ that in rats receives climbing fibre input from the superior colliculus (via the inferior olive). This region has not been observed in non-whisking animals such as cats and primates, and its functional role in vibrissal processing has hitherto remained mysterious. Further investigation of this system may throw light on how cerebellar-based internal models could be used in broader sensory, motor and cognitive contexts.

The neuroethology of electrocommunication: how signal background influences sensory encoding and behaviour in Apteronotus leptorhynchus

Weakly-electric fish are a well-established model system for neuroethological studies on communication and aggression. Sensory encoding of their electric communication signals, as well as behavioural responses to these signals, have been investigated in great detail under laboratory conditions. In the wave-type brown ghost knifefish, Apteronotus leptorhynchus, transient increases in the frequency of the generated electric field, called chirps, are particularly well-studied, since they can be readily evoked by stimulating a fish with artificial signals mimicking conspecifics. When two fish interact, both their quasi-sinusoidal electric fields (called electric organ discharge, EOD) superimpose, resulting in a beat, an amplitude modulation at the frequency difference between the two EODs. Although chirps themselves are highly stereotyped signals, the shape of the amplitude modulation resulting from a chirp superimposed on a beat background depends on a number of parameters, such as the beat frequency, modulation depth, and beat phase at which the chirp is emitted. Here we review the influence of these beat parameters on chirp encoding in the three primary stages of the electrosensory pathway: electroreceptor afferents, the hindbrain electrosensory lateral line lobe, and midbrain torus semicircularis. We then examine the role of these parameters, which represent specific features of various social contexts, on the behavioural responses of A. leptorhynchus. Some aspects of the behaviour may be explained by the coding properties of early sensory neurons to chirp stimuli. However, the complexity and diversity of behavioural responses to chirps in the context of different background parameters cannot be explained solely on the basis of the sensory responses and thus suggest that critical roles are played by higher processing stages.

From static electric images to electric flow: towards dynamic perceptual cues in active electroreception

Active electroreception is an ancestral trait found in many aquatic vertebrates and has evolved independently in two teleost lineages, the Gymnotiformes and the Mormyriformes. Unique to these so-called weakly electric fish is their ability to actively generate electrical currents in the water and sense the electrical properties of the environment. How natural behavior contributes to this sensory system has been of interest to neuroethologists since the pioneering works of Lissmann. Here we report on a mutual modeling and experimental study of the stimuli available during active electrolocation of Gnathonemus petersii (Mormyridae). We show the validity of the model (I) by demonstrating that localized spatial patterns of object induced modulations in the electric field (electric images) are comparable to experimentally mapped 2-dimensional electric images and (II) by replicating earlier key findings showing that a normalized metric of electric image width provides an unambiguous cue for distance estimation. We then show that electric images and the distance metric vary systematically when an object is moved along the trunk. These potential ambiguities with regard to localization lead us to a spatiotemporal analysis of electric images. We introduce a new temporal metric for distance estimation that is based on the normalized spatial properties of electrical images. Finally, based on a survey of exploratory behavior, we show how objects situated at the tail, a region previously neglected, cast global electric images that extend over the whole sensory epithelium of the animals.

Cellular and circuit properties supporting different sensory coding strategies in electric fish and other systems

Neural codes often seem tailored to the type of information they must carry. Here we contrast the encoding strategies for two different communication signals in electric fish and describe the underlying cellular and network properties that implement them. We compare an aggressive signal that needs to be quickly detected, to a courtship signal whose quality needs to be evaluated. The aggressive signal is encoded by synchronized bursts and a predictive feedback input is crucial in separating background noise from the communication signal. The courtship signal is accurately encoded through a heterogenous population response allowing the discrimination of signal differences. Most importantly we show that the same strategies are used in other systems arguing that they evolved similar solutions because they faced similar tasks.

The vestibular system: multimodal integration and encoding of self-motion for motor control

Understanding how sensory pathways transmit information under natural conditions remains a major goal in neuroscience. The vestibular system plays a vital role in everyday life, contributing to a wide range of functions from reflexes to the highest levels of voluntary behavior. Recent experiments establishing that vestibular (self-motion) processing is inherently multimodal also provide insight into a set of interrelated questions. What neural code is used to represent sensory information in vestibular pathways? How do the interactions between the organism and the environment shape encoding? How is self-motion information processing adjusted to meet the needs of specific tasks? This review highlights progress that has recently been made towards understanding how the brain encodes and processes self-motion to ensure accurate motor control.

Tactile sensing in specialized predators - from behavior to the brain

A number of predators depend heavily on tactile cues for pursuing and capturing prey. Here I describe and discuss the whiskers of carnivorous grasshopper mice and shrews, the sensory rays of the star-nosed mole, and the tactile appendages of the tentacled snake. These diverse sensors are accompanied by remarkable corresponding specializations in the central nervous system. But understanding their function and the significance of the central nervous system correlates requires the careful documentation of behavior inherent to a neuroethological approach.

Sustained firing of cartwheel cells in the dorsal cochlear nucleus evokes endocannabinoid release and retrograde suppression of parallel fiber synapses

Neurons in many brain regions release endocannabinoids from their dendrites that act as retrograde signals to transiently suppress neurotransmitter release from presynaptic terminals. Little is known, however, about the physiological mechanisms of short-term endocannabinoid-mediated plasticity under physiological conditions. Here we investigate calcium-dependent endocannabinoid release from cartwheel cells (CWCs) of the mouse dorsal cochlear nucleus (DCN) in the auditory brainstem that provide feedforward inhibition onto DCN principal neurons. We report that sustained action potential firing by CWCs evokes endocannabinoid release in response to submicromolar elevation of dendritic calcium that transiently suppresses their parallel fiber (PF) inputs by >70%. Basal spontaneous CWC firing rates are insufficient to evoke tonic suppression of PF synapses. However, elevating CWC firing rates by stimulating PFs triggers the release of endocannabinoids and heterosynaptic suppression of PF inputs. Spike-evoked suppression by endocannabinoids selectively suppresses excitatory synapses, but glycinergic/GABAergic inputs onto CWCs are not affected. Our findings demonstrate a mechanism of transient plasticity mediated by endocannabinoids that heterosynaptically suppresses subsets of excitatory presynaptic inputs to CWCs that regulates feedforward inhibition of DCN principal neurons and may influence the output of the DCN.

Preparing for the unpredictable: adaptive feedback enhances the response to unexpected communication signals

To interact with the environment efficiently, the nervous system must generate expectations about redundant sensory signals and detect unexpected ones. Neural circuits can, for example, compare a prediction of the sensory signal that was generated by the nervous system with the incoming sensory input, to generate a response selective to novel stimuli. In the first-order electrosensory neurons of a gymnotiform electric fish, a negative image of low-frequency redundant communication signals is subtracted from the neural response via feedback, allowing unpredictable signals to be extracted. Here we show that the cancelling feedback not only suppresses the predictable signal but also actively enhances the response to the unpredictable communication signal. A transient mismatch between the predictive feedback and incoming sensory input causes both to be positive: the soma is suddenly depolarized by the unpredictable input, whereas the neuron's apical dendrites remain depolarized by the lagging cancelling feedback. The apical dendrites allow the backpropagation of somatic spikes. We show that backpropagation is enhanced when the dendrites are depolarized, causing the unpredictable excitatory input to evoke spike bursts. As a consequence, the feedback driven by a predictable low-frequency signal not only suppresses the response to a redundant stimulus but also induces a bursting response triggered by unpredictable communication signals.

Single-neuron recordings from unanesthetized mouse dorsal cochlear nucleus

Because of the availability of disease and genetic models, the mouse has become a valuable species for auditory neuroscience that will facilitate long-term goals of understanding neuronal mechanisms underlying the perception and processing of sounds. The goal of this study was to define the basic sound-evoked response properties of single neurons in the mouse dorsal cochlear nucleus (DCN). Neurons producing complex spikes were distinguished as cartwheel cells (CWCs), and other neurons were classified according to the response map scheme previously developed in DCN. Similar to observations in other rodent species, neurons of the mouse DCN exhibit relatively little sound-driven inhibition. As a result, type III was the most commonly observed response. Our findings are generally consistent with the model of DCN function that has been developed in the cat and the gerbil, suggesting that this in vivo mouse preparation will be a useful tool for future studies of auditory physiology.