Simple synaptic modulations implement diverse novelty computations.

Detecting novelty is ethologically useful for an organism's survival. Recent experiments characterize how different types of novelty over timescales from seconds to weeks are reflected in the activity of excitatory and inhibitory neuron types. Here, we introduce a learning mechanism, familiarity-modulated synapses (FMSs), consisting of multiplicative modulations dependent on presynaptic or pre/postsynaptic neuron activity. With FMSs, network responses that encode novelty emerge under unsupervised continual learning and minimal connectivity constraints. Implementing FMSs within an experimentally constrained model of a visual cortical circuit, we demonstrate the generalizability of FMSs by simultaneously fitting absolute, contextual, and omission novelty effects. Our model also reproduces functional diversity within cell subpopulations, leading to experimentally testable predictions about connectivity and synaptic dynamics that can produce both population-level novelty responses and heterogeneous individual neuron signals. Altogether, our findings demonstrate how simple plasticity mechanisms within a cortical circuit structure can produce qualitatively distinct and complex novelty responses.

Adaptation supports short-term memory in a visual change detection task.

The maintenance of short-term memories is critical for survival in a dynamically changing world. Previous studies suggest that this memory can be stored in the form of persistent neural activity or using a synaptic mechanism, such as with short-term plasticity. Here, we compare the predictions of these two mechanisms to neural and behavioral measurements in a visual change detection task. Mice were trained to respond to changes in a repeated sequence of natural images while neural activity was recorded using two-photon calcium imaging. We also trained two types of artificial neural networks on the same change detection task as the mice. Following fixed pre-processing using a pretrained convolutional neural network, either a recurrent neural network (RNN) or a feedforward neural network with short-term synaptic depression (STPNet) was trained to the same level of performance as the mice. While both networks are able to learn the task, the STPNet model contains units whose activity are more similar to the in vivo data and produces errors which are more similar to the mice. When images are omitted, an unexpected perturbation which was absent during training, mice often do not respond to the omission but are more likely to respond to the subsequent image. Unlike the RNN model, STPNet produces a similar pattern of behavior. These results suggest that simple neural adaptation mechanisms may serve as an important bottom-up memory signal in this task, which can be used by downstream areas in the decision-making process.

Characterization of Learning, Motivation, and Visual Perception in Five Transgenic Mouse Lines Expressing GCaMP in Distinct Cell Populations.

To study the mechanisms of perception and cognition, neural measurements must be made during behavior. A goal of the Allen Brain Observatory is to map the activity of distinct cortical cell classes underlying visual and behavioral processing. Here we describe standardized methodology for training head-fixed mice on a visual change detection task, and we use our paradigm to characterize learning and behavior of five GCaMP6-expressing transgenic lines. We used automated training procedures to facilitate comparisons across mice. Training times varied, but most transgenic mice learned the behavioral task. Motivation levels also varied across mice. To compare mice in similar motivational states we subdivided sessions into over-, under-, and optimally motivated periods. When motivated, the pattern of perceptual decisions were highly correlated across transgenic lines, although overall performance (d-prime) was lower in one line labeling somatostatin inhibitory cells. These results provide important context for using these mice to map neural activity underlying perception and behavior.

Higher-Order Thalamic Circuits Channel Parallel Streams of Visual Information in Mice.

Higher-order thalamic nuclei, such as the visual pulvinar, play essential roles in cortical function by connecting functionally related cortical and subcortical brain regions. A coherent framework describing pulvinar function remains elusive because of its anatomical complexity and involvement in diverse cognitive processes. We combined large-scale anatomical circuit mapping with high-density electrophysiological recordings to dissect a homolog of the pulvinar in mice, the lateral posterior thalamic nucleus (LP). We define three broad LP subregions based on correspondence between connectivity and functional properties. These subregions form corticothalamic loops biased toward ventral or dorsal stream cortical areas and contain separate representations of visual space. Silencing the visual cortex or superior colliculus revealed that they drive visual tuning properties in separate LP subregions. Thus, by specifying the driving input sources, functional properties, and downstream targets of LP circuits, our data provide a roadmap for understanding the mechanisms of higher-order thalamic function in vision.

Automated identification of mouse visual areas with intrinsic signal imaging.

Intrinsic signal optical imaging (ISI) is a rapid and noninvasive method for observing brain activity in vivo over a large area of the cortex. Here we describe our protocol for mapping retinotopy to identify mouse visual cortical areas using ISI. First, surgery is performed to attach a head frame to the mouse skull (∼1 h). The next day, intrinsic activity across the visual cortex is recorded during the presentation of a full-field drifting bar in the horizontal and vertical directions (∼2 h). Horizontal and vertical retinotopic maps are generated by analyzing the response of each pixel during the period of the stimulus. Last, an algorithm uses these retinotopic maps to compute the visual field sign and coverage, and automatically construct visual borders without human input. Compared with conventional retinotopic mapping with episodic presentation of adjacent stimuli, a continuous, periodic stimulus is more resistant to biological artifacts. Furthermore, unlike manual hand-drawn approaches, we present a method for automatically segmenting visual areas, even in the small mouse cortex. This relatively simple procedure and accompanying open-source code can be implemented with minimal surgical and computational experience, and is useful to any laboratory wishing to target visual cortical areas in this increasingly valuable model system.

Topography and areal organization of mouse visual cortex.

To guide future experiments aimed at understanding the mouse visual system, it is essential that we have a solid handle on the global topography of visual cortical areas. Ideally, the method used to measure cortical topography is objective, robust, and simple enough to guide subsequent targeting of visual areas in each subject. We developed an automated method that uses retinotopic maps of mouse visual cortex obtained with intrinsic signal imaging (Schuett et al., 2002; Kalatsky and Stryker, 2003; Marshel et al., 2011) and applies an algorithm to automatically identify cortical regions that satisfy a set of quantifiable criteria for what constitutes a visual area. This approach facilitated detailed parcellation of mouse visual cortex, delineating nine known areas (primary visual cortex, lateromedial area, anterolateral area, rostrolateral area, anteromedial area, posteromedial area, laterointermediate area, posterior area, and postrhinal area), and revealing two additional areas that have not been previously described as visuotopically mapped in mice (laterolateral anterior area and medial area). Using the topographic maps and defined area boundaries from each animal, we characterized several features of map organization, including variability in area position, area size, visual field coverage, and cortical magnification. We demonstrate that higher areas in mice often have representations that are incomplete or biased toward particular regions of visual space, suggestive of specializations for processing specific types of information about the environment. This work provides a comprehensive description of mouse visuotopic organization and describes essential tools for accurate functional localization of visual areas.

Functional specialization of seven mouse visual cortical areas.

To establish the mouse as a genetically tractable model for high-order visual processing, we characterized fine-scale retinotopic organization of visual cortex and determined functional specialization of layer 2/3 neuronal populations in seven retinotopically identified areas. Each area contains a distinct visuotopic representation and encodes a unique combination of spatiotemporal features. Areas LM, AL, RL, and AM prefer up to three times faster temporal frequencies and significantly lower spatial frequencies than V1, while V1 and PM prefer high spatial and low temporal frequencies. LI prefers both high spatial and temporal frequencies. All extrastriate areas except LI increase orientation selectivity compared to V1, and three areas are significantly more direction selective (AL, RL, and AM). Specific combinations of spatiotemporal representations further distinguish areas. These results reveal that mouse higher visual areas are functionally distinct, and separate groups of areas may be specialized for motion-related versus pattern-related computations, perhaps forming pathways analogous to dorsal and ventral streams in other species.