Sensory systems exist to provide an organism with information about the state of the environment that can be used to guide future actions and decisions. Remarkably, two conceptually simple yet general theorems from information theory can be used to evaluate the performance of any sensory system. One theorem states that there is a minimal amount of energy that an organism has to spend in order to capture a given amount of information about the environment. The second theorem states that the maximum rate with which the organism can acquire resources from the environment, relative to its competitors, is limited by the information this organism collects about the environment, also relative to its competitors. These two theorems provide a scaffold for formulating and testing general principles of sensory coding but leave unanswered many important practical questions of implementation in neural circuits. These implementation questions have guided thinking in entire subfields of sensory neuroscience, and include: What features in the sensory environment should be measured? Given that we make decisions on a variety of time scales, how should one solve trade-offs between making simpler measurements to guide minimal decisions vs. more elaborate sensory systems that have to overcome multiple delays between sensation and action. Once we agree on the types of features that are important to represent, how should they be represented? How should resources be allocated between different stages of processing, and where is the impact of noise most damaging? Finally, one should consider trade-offs between implementing a fixed strategy vs. an adaptive scheme that readjusts resources based on current needs. Where adaptation is considered, under what conditions does it become optimal to switch strategies? Research over the past 60 years has provided answers to almost all of these questions but primarily in early sensory systems. Joining these answers into a comprehensive framework is a challenge that will help us understand who we are and how we can make better use of limited natural resources.
Tatyana O. Sharpee
Justin D. Lieber and Sliman J. Bensmaia
The ability to identify tactile objects depends in part on the perception of their surface microstructure and material properties. Texture perception can, on a first approximation, be described by a number of nameable perceptual axes, such as rough/smooth, hard/soft, sticky/slippery, and warm/cool, which exist within a complex perceptual space. The perception of texture relies on two different neural streams of information: Coarser features, measured in millimeters, are primarily encoded by spatial patterns of activity across one population of tactile nerve fibers, while finer features, down to the micron level, are encoded by finely timed temporal patterns within two other populations of afferents. These two streams of information ascend the somatosensory neuraxis and are eventually combined and further elaborated in the cortex to yield a high-dimensional representation that accounts for our exquisite and stable perception of texture.
Kenway Louie and Paul W. Glimcher
A core question in systems and computational neuroscience is how the brain represents information. Identifying principles of information coding in neural circuits is critical to understanding brain organization and function in sensory, motor, and cognitive neuroscience. This provides a conceptual bridge between the underlying biophysical mechanisms and the ultimate behavioral goals of the organism. Central to this framework is the question of computation: what are the relevant representations of input and output, and what algorithms govern the input-output transformation? Remarkably, evidence suggests that certain canonical computations exist across different circuits, brain regions, and species. Such computations are implemented by different biophysical and network mechanisms, indicating that the unifying target of conservation is the algorithmic form of information processing rather than the specific biological implementation. A prime candidate to serve as a canonical computation is divisive normalization, which scales the activity of a given neuron by the activity of a larger neuronal pool. This nonlinear transformation introduces an intrinsic contextual modulation into information coding, such that the selective response of a neuron to features of the input is scaled by other input characteristics. This contextual modulation allows the normalization model to capture a wide array of neural and behavioral phenomena not captured by simpler linear models of information processing. The generality and flexibility of the normalization model arises from the normalization pool, which allows different inputs to directly drive and suppress a given neuron, effectively separating information that drives excitation and contextual modulation. Originally proposed to describe responses in early visual cortex, normalization has been widely documented in different brain regions, hierarchical levels, and modalities of sensory processing; furthermore, recent work shows that the normalization extends to cognitive processes such as attention, multisensory integration, and decision making. This ubiquity reinforces the canonical nature of the normalization computation and highlights the importance of an algorithmic framework in linking biological mechanism and behavior.
Quentin Gaudry and Jonathan Schenk
Olfactory systems are tasked with converting the chemical environment into electrical signals that the brain can use to optimize behaviors such as navigating towards resources, finding mates, or avoiding danger. Drosophila melanogaster has long served as a model system for several attributes of olfaction. Such features include sensory coding, development, and the attempt to link sensory perception to behavior. The strength of Drosophila as a model system for neurobiology lies in the myriad of genetic tools made available to the experimentalist, and equally importantly, the numerical reduction in cell numbers within the olfactory circuit. Modern techniques have recently made it possible to target nearly all cell types in the antennal lobe to directly monitor their physiological activity or to alter their expression of endogenous proteins or transgenes.
William B. Kristan Jr.
New techniques for recording the activity of many neurons simultaneously have given insights into how neuronal circuits make the decision to perform one of many possible behaviors. A long-standing hypothesis for how behavioral choices are made in any animal is that “command neurons” are responsible for selecting individual behaviors, and that these same neurons inhibit the command neurons that elicit other behaviors. In fact, this mechanism has turned out to be just one of several ways that such decision-making is accomplished. In particular, for some behavioral choices, the circuits appear to overlap, sometimes extensively, to perform two or more behaviors. Making decisions using such “multifunctional neurons” has been proposed for large neural networks, but this strategy appears to be used in relatively small nervous systems, too. These simpler nervous systems can serve as useful test systems to test hypotheses about how the dynamics of networks of neurons can be used to select among different behaviors, similar to the mechanisms used by leeches deciding to swim, shorten, crawl, or feed.
Mathew H. Evans, Michaela S.E. Loft, Dario Campagner, and Rasmus S. Petersen
Whiskers (vibrissae) are prominent on the snout of many mammals, both terrestrial and aquatic. The defining feature of whiskers is that they are rooted in large follicles with dense sensory innervation, surrounded by doughnut-shaped blood sinuses. Some species, including rats and mice, have elaborate muscular control of their whiskers and explore their environment by making rhythmic back-and-forth “whisking” movements. Whisking movements are purposefully modulated according to specific behavioral goals (“active sensing”). The basic whisking rhythm is controlled by a premotor complex in the intermediate reticular formation. Primary whisker neurons (PWNs), with cell bodies in the trigeminal ganglion, innervate several classes of mechanoreceptive nerve endings in the whisker follicle. Mechanotransduction involving Piezo2 ion channels establishes the fundamental physical signals that the whiskers communicate to the brain. PWN spikes are triggered by mechanical forces associated with both the whisking motion itself and whisker-object contact. Whisking is associated with inertial and muscle contraction forces that drive PWN activity. Whisker-object contact causes whiskers to bend, and PWN activity is driven primarily by the associated rotatory force (“bending moment”). Sensory signals from the PWNs are routed to many parts of the hindbrain, midbrain, and forebrain. Parallel ascending pathways transmit information about whisker forces to sensorimotor cortex. At each brainstem, thalamic, and cortical level of these pathways, there are one or more maps of the whisker array, consisting of cell clusters (“barrels” in the primary somatosensory cortex) whose spatial arrangement precisely mirrors that of the whiskers on the snout. However, the overall architecture of the whisker-responsive regions of the brain system is best characterized by multilevel sensory-motor feedback loops. Its intriguing biology, in combination with advantageous properties as a model sensory system, has made the whisker system the platform for seminal insights into brain function.
Angel Ariel Caputi
American gymnotiformes and African mormyriformes have evolved an active sensory system using a self-generated electric field as a carrier of signals. Objects polarized by the discharge of a specialized electric organ project their images on the skin where electroreceptors tuned to the time course of the self-generated field transduce local signals carrying information about impedance, shape, size, and location of objects, as well as electrocommunication messages, and encode them as primary afferents trains of spikes. This system is articulated with other cutaneous systems (passive electroreception and mechanoception) as well as proprioception informing the shape of the fish’s body. Primary afferents project on the electrosensory lobe where electrosensory signals are compared with expectation signals resulting from the integration of recent past electrosensory, other sensory, and, in the case of mormyriformes, electro- and skeleton-motor corollary discharges. This ensemble of signals converges on the apical dendrites of the principal cells where a working memory of the recent past, and therefore predictable input, is continuously built up and updated as a pattern of synaptic weights. The efferent neurons of the electrosensory lobe also project to the torus and indirectly to other brainstem nuclei that implement automatic electro- and skeleton-motor behaviors. Finally, the torus projects via the preglomerular nucleus to the telencephalon where cognitive functions, including “electroperception” of shape-, size- and impedance-related features of objects, recognition of conspecifics, perception based decisions, learning, and abstraction, are organized.