Parallel functional subnetworks embedded in the macaque face patch system

During normal vision, our eyes provide the brain with a continuous stream of useful information about the world. How visually specialized areas of the cortex, such as face-selective patches, operate under natural modes of behavior is poorly understood. Here we report that, during the free viewing of movies, cohorts of face-selective neurons in the macaque cortex fractionate into distributed and parallel subnetworks that carry distinct information. We classified neurons into functional groups on the basis of their movie-driven coupling with functional magnetic resonance imaging time courses across the brain. Neurons from each group were distributed across multiple face patches but intermixed locally with other groups at each recording site. These findings challenge prevailing views about functional segregation in the cortex and underscore the importance of naturalistic paradigms for cognitive neuroscience.

Dynamic Suppression of Average Facial Structure Shapes Neural Tuning in Three Macaque Face Patches

The visual perception of identity in humans and other primates is thought to draw upon cortical areas specialized for the analysis of facial structure. A prominent theory of face recognition holds that the brain computes and stores average facial structure, which it then uses to efficiently determine individual identity, though the neural mechanisms underlying this process are controversial. Here, we demonstrate that the dynamic suppression of average facial structure plays a prominent role in the responses of neurons in three fMRI-defined face patches of the macaque. Using photorealistic face stimuli that systematically varied in identity level according to a psychophysically based face space, we found that single units in the AF, AM, and ML face patches exhibited robust tuning around average facial structure. This tuning emerged after the initial excitatory response to the face and was expressed as the selective suppression of sustained responses to low-identity faces. The coincidence of this suppression with increased spike timing synchrony across the population suggests a mechanism of active inhibition underlying this effect. Control experiments confirmed that the diminished responses to low-identity faces were not due to short-term adaptation processes. We propose that the brain's neural suppression of average facial structure facilitates recognition by promoting the extraction of distinctive facial characteristics and suppressing redundant or irrelevant responses across the population.

Functional Subpopulations of Neurons in a Macaque Face Patch Revealed by Single-Unit fMRI Mapping

Neurons within fMRI-defined face patches of the macaque brain exhibit shared categorical responses to flashed images but diverge in their responses under more natural viewing conditions. Here we investigate functional diversity among neurons in the anterior fundus (AF) face patch, combining whole-brain fMRI with longitudinal single-unit recordings in a local population (<1 mm³). For each cell, we computed a whole-brain correlation map based on its shared time course with voxels throughout the brain during naturalistic movie viewing. Based on this mapping, neighboring neurons showed markedly different affiliation with distant visually responsive areas and fell coarsely into subpopulations. Of these, only one subpopulation (∼16% of neurons) yielded similar correlation maps to the local fMRI signal. The results employ the readout of large-scale fMRI networks and, by indicating multiple functional domains within a single voxel, present a new view of functional diversity within a local neural population.

One month in the life of a neuron: longitudinal single-unit electrophysiology in the monkey visual system

Conventional recording methods generally preclude following the activity of the same neurons in awake animals across days. This limits our ability to systematically investigate the principles of neuronal specialization, or to study phenomena that evolve over multiple days such as experience-dependent plasticity. To redress this shortcoming, we developed a drivable, chronically implanted microwire recording preparation that allowed us to follow visual responses in inferotemporal (IT) cortex in awake behaving monkeys across multiple days, and in many cases across months. The microwire bundle and other implanted components were MRI compatible and thus permitted in the same animals both functional imaging and long-term recording from multiple neurons in deep structures within a region the approximate size of one voxel (<1 mm). The distinct patterns of stimulus selectivity observed in IT neurons, together with stable features in spike waveforms and interspike interval distributions, allowed us to track individual neurons across weeks and sometimes months. The long-term consistency of visual responses shown here permits large-scale mappings of neuronal properties using massive image libraries presented over the course of days. We demonstrate this possibility by screening the visual responses of single neurons to a set of 10,000 stimuli.

All in the timing: Priming, repetition suppression, and synchrony

Abstract The terms "priming" and "repetition suppression" are commonly used to refer to phenomena occurring on time scales that can differ by several orders of magnitude, ranging from seconds to days or even years. The models discussed by Gotts et al. provide a thought-provoking theoretical framework for relating neuronal and behavioral plasticity. I argue that whereas both the sharpening and the Bayesian models may mediate the gradual acquisition of perceptual expertise, they are unlikely to account for more rapid behavioral changes. The synchrony model, however, could potentially operate within the timing constraints imposed by the fastest forms of repetition priming.

Linearly additive shape and color signals in monkey inferotemporal cortex

How does the brain represent a red circle? One possibility is that there is a specialized and possibly time-consuming process whereby the attributes of shape and color, carried by separate populations of neurons in low-order visual cortex, are bound together into a unitary neural representation. Another possibility is that neurons in high-order visual cortex are selective, by virtue of their bottom-up input from low-order visual areas, for particular conjunctions of shape and color. A third possibility is that they simply sum shape and color signals linearly. We tested these ideas by measuring the responses of inferotemporal cortex neurons to sets of stimuli in which two attributes-shape and color-varied independently. We find that a few neurons exhibit conjunction selectivity but that in most neurons the influences of shape and color sum linearly. Contrary to the idea of conjunction coding, few neurons respond selectively to a particular combination of shape and color. Contrary to the idea that binding requires time, conjunction signals, when present, occur as early as feature signals. We argue that neither conjunction selectivity nor a specialized feature binding process is necessary for the effective representation of shape-color combinations.

Repetition suppression in monkey inferotemporal cortex: relation to behavioral priming

In tasks requiring judgments about visual stimuli, humans exhibit repetition priming, responding with increased speed when a stimulus is repeated. Repetition priming might depend on repetition suppression, a phenomenon first observed in monkey inferotemporal cortex (IT) whereby, when a stimulus is repeated, the strength of the neuronal visual response is reduced. If the reduction resulted in sharpening of the cortical representation of the stimulus, and did not just scale it down, then speeded processing might result. To explore the relation between repetition priming and repetition suppression, we monitored neuronal activity in IT while monkeys performed a symmetry decision task. We found 1) that monkeys exhibit repetition priming, 2) that IT neurons simultaneously exhibit repetition suppression, 3) that repetition priming and repetition suppression do not vary in a significantly correlated fashion across trials, and 4) that repetition suppression scales down the representation of the stimulus without sharpening it. We conclude that repetition suppression accompanies repetition priming but is unlikely to be its cause.

Short- and long-term plasticity of the perforant path synapse in hippocampal area CA3

The direct perforant path (PP) projection to CA3 is a major source of cortical input to the hippocampal region, yet relatively little is known about the basic properties of physiology and plasticity in this pathway. We tested whether PP long-term potentiation (LTP) in CA3 possesses the Hebbian property of associativity; i.e., whether the firing of fibers of different orders can induce PP LTP. We stimulated PP with weak trains of high-frequency stimulation (HFS), which by itself was below the threshold for LTP induction. The identical HFS was effective in inducing LTP when the mossy fiber pathway (MF) was activated simultaneously, thus demonstrating associative plasticity between the two pathways. We also demonstrated associative LTP between PP and recurrent collateral fibers (RC). PP LTP was blocked by the N-methyl-D-aspartate receptor (NMDAR) antagonist 2-amino-5-phosphonovaleric acid in both the associative and homosynaptic induction conditions. Neither MF nor RC fiber HFS alone resulted in permanent changes in PP field excitatory postsynaptic potential (fEPSP) amplitude. However, HFS delivered to either MF or RC alone led to transient heterosynaptic depression of the PP fEPSP. Our results support the conceptual framework that regards CA3 as an autoassociative memory network in which efficient retrieval of previously stored activity patterns is mediated by associative plasticity of the PP synapse.

Giant miniature EPSCs at the hippocampal mossy fiber to CA3 pyramidal cell synapse are monoquantal

The mechanisms generating giant miniature excitatory postsynaptic currents (mEPSCs) were investigated at the hippocampal mossy fiber (MF) to CA3 pyramidal cell synapse in vitro. These giant mEPSCs have peak amplitudes as large as 1,700 pA (13.6 nS) with a mean maximal mEPSC amplitude of 366 +/- 20 pA (mean +/- SD; 5 nS; n = 25 cells). This is compared with maximal mEPSC amplitudes of <100 pA typically observed at other cortical synapses. We tested the hypothesis that giant mEPSCs are due to synchronized release of multiple vesicles across the release sites of single MF boutons by directly inducing vesicular release using secretagogues. If giant mEPSCs result from simultaneous multivesicular release, then secretagogues should increase the frequency of small mEPSCs selectively. We found that hypertonic sucrose and spermine increased the frequency of both small and giant mEPSCs. The peptide toxin secretagogues alpha-latrotoxin and pardaxin failed to increase the frequency of giant mEPSCs, but the possible lack of tissue penetration of the toxins make these results equivocal. Because a multiquantal release mechanism is likely to be mediated by a spontaneous increase in presynaptic calcium concentration, a monoquantal mechanism is further supported by results that giant mEPSCs were not affected by manipulations of extracellular or intracellular calcium concentrations. In addition, reducing the temperature of the bath to 15 degrees C failed to desynchronize the rising phases of giant mEPSCs. Together these data suggest that the giant mEPSCs are generated via a monovesicular mechanism. Three-dimensional analysis through serial electron microscopy of the MF boutons revealed large clear vesicles (50 to 160 nm diam) docked presynaptically at the MF synapse in sufficient numbers to account for the amplitude and frequency of giant mEPSCs recorded electrophysiologically. It is concluded that release of the contents of a single large clear vesicle generates giant mEPSCs at the MF to CA3 pyramidal cell synapse.