Motion detection: neuronal circuit meets theory.

Motion detection in fly vision has been investigated experimentally and theoretically for half of a century, yet mechanistic insights into the neuronal computation have only started to emerge. In a recent issue of Nature, two studies provide major insights into how motion direction is extracted from the image flow projected onto the retina.

RSK-Mediated Non-canonical Activation of EphA2 by Tamoxifen.

The long-term administration of tamoxifen to estrogen receptor α (ERα)-positive breast cancer patients is an established treatment that reduces mortality and recurrence. However, resistance to tamoxifen and an increased risk of endometrial cancer may occur; therefore, the mechanisms by which tamoxifen causes these adverse effects warrant further study. Tamoxifen has been shown to activate mitogen-activated protein kinase (MAPK) in an ERα-independent manner; therefore, we investigated its effects on the MAPK-mediated non-canonical activation of EphA2, a critical event regulating cell migration. Tamoxifen at slightly higher concentrations induced the rapid phosphorylation of EphA2 at Ser-897 via the MAPK/extracellular signal-regulated kinase (ERK) kinase (MEK)-ERK-ribosomal S6 kinases (RSK) pathway in HeLa cells. In addition, tamoxifen significantly enhanced the migration ability of ERα-negative MDA-MB-231 breast cancer cells in RSK- and EphA2-dependent manners. Phosphorylated EphA2 was internalized and re-localized to the plasma membrane, including lamellipodia, in an RSK-dependent manner. Collectively, the present results provide novel insights into the tumor-promoting activity of tamoxifen.

Spatially asymmetric reorganization of inhibition establishes a motion-sensitive circuit.

Spatial asymmetries in neural connectivity have an important role in creating basic building blocks of neuronal processing. A key circuit module of directionally selective (DS) retinal ganglion cells is a spatially asymmetric inhibitory input from starburst amacrine cells. It is not known how and when this circuit asymmetry is established during development. Here we photostimulate mouse starburst cells targeted with channelrhodopsin-2 (refs 6-8) while recording from a single genetically labelled type of DS cell. We follow the spatial distribution of synaptic strengths between starburst and DS cells during early postnatal development before these neurons can respond to a physiological light stimulus, and confirm connectivity by monosynaptically restricted trans-synaptic rabies viral tracing. We show that asymmetry develops rapidly over a 2-day period through an intermediate state in which random or symmetric synaptic connections have been established. The development of asymmetry involves the spatially selective reorganization of inhibitory synaptic inputs. Intriguingly, the spatial distribution of excitatory synaptic inputs from starburst cells is significantly more symmetric than that of the inhibitory inputs at the end of this developmental period. Our work demonstrates a rapid developmental switch from a symmetric to asymmetric input distribution for inhibition in the neural circuit of a principal cell.

SPIG1 negatively regulates BDNF maturation.

We previously identified SPARC-related protein-containing immunoglobulin domains 1 (SPIG1, also known as Follistatin-like protein 4) as one of the dorsal-retina-specific molecules expressed in the developing chick retina. We here demonstrated that the knockdown of SPIG1 in the retinal ganglion cells (RGCs) of developing chick embryos induced the robust ectopic branching of dorsal RGC axons and failed to form a tight terminal zone at the proper position on the tectum. The knockdown of SPIG1 in RGCs also led to enhanced axon branching in vitro. However, this was canceled by the addition of a neutralizing antibody against brain-derived neurotrophic factor (BDNF) to the culture medium. SPIG1 and BDNF were colocalized in vesicle-like structures in cells. SPIG1 bound with the proform of BDNF (proBDNF) but very weakly with mature BDNF in vitro. The expression and secretion of mature BDNF were significantly decreased when SPIG1 was exogenously expressed with BDNF in HEK293T or PC12 cells. The amount of mature BDNF proteins as well as the tyrosine phosphorylation level of the BDNF receptor, tropomyosin-related kinase B (TrkB), in the hippocampus were significantly higher in SPIG1-knockout mice than in wild-type mice. Here the spine density of CA1 pyramidal neurons was consistently increased. Together, these results suggest that SPIG1 negatively regulated BDNF maturation by binding to proBDNF, thereby suppressing axonal branching and spine formation.

Emerging computational motifs: Lessons from the retina.

The retinal neuronal circuit is the first stage of visual processing in the central nervous system. The efforts of scientists over the last few decades indicate that the retina is not merely an array of photosensitive cells, but also a processor that performs various computations. Within a thickness of only ∼200 µm, the retina consists of diverse forms of neuronal circuits, each of which encodes different visual features. Since the discovery of direction-selective cells by Horace Barlow and Richard Hill, the mechanisms that generate direction selectivity in the retina have remained a fascinating research topic. This review provides an overview of recent advances in our understanding of direction-selectivity circuits. Beyond the conventional wisdom of direction selectivity, emerging findings indicate that the retina utilizes complicated and sophisticated mechanisms in which excitatory and inhibitory pathways are involved in the efficient encoding of motion information. As will become evident, the discovery of computational motifs in the retina facilitates an understanding of how sensory systems establish feature selectivity.

Recombinase-independent AAV for anterograde transsynaptic tracing.

Viral transsynaptic labeling has become indispensable for investigating the functional connectivity of neural circuits in the mammalian brain. Adeno-associated virus serotype 1 (AAV1) allows for anterograde transneuronal labeling and manipulation of postsynaptic neurons. However, it is limited to delivering an AAV1 expressing a recombinase which relies on using transgenic animals or genetic access to postsynaptic neurons. We reasoned that a strong expression level could overcome this limitation. To this end, we used a self-complementary AAV of serotype 1 (scAAV1) under a strong promoter (CAG). We demonstrated the anterograde transneuronal efficiency of scAAV1 by delivering a fluorescent marker in mouse retina-superior colliculus and thalamic-amygdala pathways in a recombinase-independent manner in the mouse brain. In addition to investigating neuronal connectivity, anterograde transsynaptic AAVs with a strong promoter may be suitable for functional mapping and imaging.

Subcortico-amygdala pathway processes innate and learned threats.

Behavioral flexibility and timely reactions to salient stimuli are essential for survival. The subcortical thalamic-basolateral amygdala (BLA) pathway serves as a shortcut for salient stimuli ensuring rapid processing. Here, we show that BLA neuronal and thalamic axonal activity in mice mirror the defensive behavior evoked by an innate visual threat as well as an auditory learned threat. Importantly, perturbing this pathway compromises defensive responses to both forms of threats, in that animals fail to switch from exploratory to defensive behavior. Despite the shared pathway between the two forms of threat processing, we observed noticeable differences. Blocking ß-adrenergic receptors impairs the defensive response to the innate but not the learned threats. This reduced defensive response, surprisingly, is reflected in the suppression of the activity exclusively in the BLA as the thalamic input response remains intact. Our side-by-side examination highlights the similarities and differences between innate and learned threat-processing, thus providing new fundamental insights.

Therapeutic Neuromodulation toward a Critical State May Serve as a General Treatment Strategy.

Brain disease has become one of this century's biggest health challenges, urging the development of novel, more effective treatments. To this end, neuromodulation represents an excellent method to modulate the activity of distinct neuronal regions to alleviate disease. Recently, the medical indications for neuromodulation therapy have expanded through the adoption of the idea that neurological disorders emerge from deficits in systems-level structures, such as brain waves and neural topology. Connections between neuronal regions are thought to fluidly form and dissolve again based on the patterns by which neuronal populations synchronize. Akin to a fire that may spread or die out, the brain's activity may similarly hyper-synchronize and ignite, such as seizures, or dwindle out and go stale, as in a state of coma. Remarkably, however, the healthy brain remains hedged in between these extremes in a critical state around which neuronal activity maneuvers local and global operational modes. While it has been suggested that perturbations of this criticality could underlie neuropathologies, such as vegetative states, epilepsy, and schizophrenia, a major translational impact is yet to be made. In this hypothesis article, we dissect recent computational findings demonstrating that a neural network's short- and long-range connections have distinct and tractable roles in sustaining the critical regime. While short-range connections shape the dynamics of neuronal activity, long-range connections determine the scope of the neuronal processes. Thus, to facilitate translational progress, we introduce topological and dynamical system concepts within the framework of criticality and discuss the implications and possibilities for therapeutic neuromodulation guided by topological decompositions.

Spatiotemporal properties of glutamate input support direction selectivity in the dendrites of retinal starburst amacrine cells.

The asymmetric summation of kinetically distinct glutamate inputs across the dendrites of retinal 'starburst' amacrine cells is one of the several mechanisms that have been proposed to underlie their direction-selective properties, but experimentally verifying input kinetics has been a challenge. Here, we used two-photon glutamate sensor (iGluSnFR) imaging to directly measure the input kinetics across individual starburst dendrites. We found that signals measured from proximal dendrites were relatively sustained compared to those measured from distal dendrites. These differences were observed across a range of stimulus sizes and appeared to be shaped mainly by excitatory rather than inhibitory network interactions. Temporal deconvolution analysis suggests that the steady-state vesicle release rate was ~3 times larger at proximal sites compared to distal sites. Using a connectomics-inspired computational model, we demonstrate that input kinetics play an important role in shaping direction selectivity at low stimulus velocities. Taken together, these results provide direct support for the 'space-time wiring' model for direction selectivity.

EyeLoop: An Open-Source System for High-Speed, Closed-Loop Eye-Tracking.

Eye-trackers are widely used to study nervous system dynamics and neuropathology. Despite this broad utility, eye-tracking remains expensive, hardware-intensive, and proprietary, limiting its use to high-resource facilities. It also does not easily allow for real-time analysis and closed-loop design to link eye movements to neural activity. To address these issues, we developed an open-source eye-tracker - EyeLoop - that uses a highly efficient vectorized pupil detection method to provide uninterrupted tracking and fast online analysis with high accuracy on par with popular eye tracking modules, such as DeepLabCut. This Python-based software easily integrates custom functions using code modules, tracks a multitude of eyes, including in rodents, humans, and non-human primates, and operates at more than 1,000 frames per second on consumer-grade hardware. In this paper, we demonstrate EyeLoop's utility in an open-loop experiment and in biomedical disease identification, two common applications of eye-tracking. With a remarkably low cost and minimum setup steps, EyeLoop makes high-speed eye-tracking widely accessible.

Short- and Long-Range Connections Differentially Modulate the Dynamics and State of Small-World Networks.

The human brain contains billions of neurons that flexibly interconnect to support local and global computational spans. As neuronal activity propagates through the neural medium, it approaches a critical state hedged between ordered and disordered system regimes. Recent work demonstrates that this criticality coincides with the small-world topology, a network arrangement that accommodates both local (subcritical) and global (supercritical) system properties. On one hand, operating near criticality is thought to offer several neurocomputational advantages, e.g., high-dynamic range, efficient information capacity, and information transfer fidelity. On the other hand, aberrations from the critical state have been linked to diverse pathologies of the brain, such as post-traumatic epileptiform seizures and disorders of consciousness. Modulation of brain activity, through neuromodulation, presents an attractive mode of treatment to alleviate such neurological disorders, but a tractable neural framework is needed to facilitate clinical progress. Using a variation on the generative small-world model of Watts and Strogatz and Kuramoto's model of coupled oscillators, we show that the topological and dynamical properties of the small-world network are divided into two functional domains based on the range of connectivity, and that these domains play distinct roles in shaping the behavior of the critical state. We demonstrate that short-range network connections shape the dynamics of the system, e.g., its volatility and metastability, whereas long-range connections drive the system state, e.g., a seizure. Together, these findings lend support to combinatorial neuromodulation approaches that synergistically normalize the system dynamic while mobilizing the system state.

Binocular integration of retinal motion information underlies optic flow processing by the cortex.

Locomotion creates various patterns of optic flow on the retina, which provide the observer with information about their movement relative to the environment. However, it is unclear how these optic flow patterns are encoded by the cortex. Here, we use two-photon calcium imaging in awake mice to systematically map monocular and binocular responses to horizontal motion in four areas of the visual cortex. We find that neurons selective to translational or rotational optic flow are abundant in higher visual areas, whereas neurons suppressed by binocular motion are more common in the primary visual cortex. Disruption of retinal direction selectivity in Frmd7 mutant mice reduces the number of translation-selective neurons in the primary visual cortex and translation- and rotation-selective neurons as well as binocular direction-selective neurons in the rostrolateral and anterior visual cortex, blurring the functional distinction between primary and higher visual areas. Thus, optic flow representations in specific areas of the visual cortex rely on binocular integration of motion information from the retina.

Direction selectivity in retinal bipolar cell axon terminals.

The ability to encode the direction of image motion is fundamental to our sense of vision. Direction selectivity along the four cardinal directions is thought to originate in direction-selective ganglion cells (DSGCs) because of directionally tuned GABAergic suppression by starburst cells. Here, by utilizing two-photon glutamate imaging to measure synaptic release, we reveal that direction selectivity along all four directions arises earlier than expected at bipolar cell outputs. Individual bipolar cells contained four distinct populations of axon terminal boutons with different preferred directions. We further show that this bouton-specific tuning relies on cholinergic excitation from starburst cells and GABAergic inhibition from wide-field amacrine cells. DSGCs received both tuned directionally aligned inputs and untuned inputs from among heterogeneously tuned glutamatergic bouton populations. Thus, directional tuning in the excitatory visual pathway is incrementally refined at the bipolar cell axon terminals and their recipient DSGC dendrites by two different neurotransmitters co-released from starburst cells.

FLRT3 Marks Direction-Selective Retinal Ganglion Cells That Project to the Medial Terminal Nucleus.

The mammalian retina extracts a multitude of diverse features from the visual scene such as color, contrast, and direction of motion. These features are transmitted separately to the brain by more than 40 different retinal ganglion cell (RGC) subtypes. However, so far only a few genetic markers exist to fully characterize the different RGC subtypes. Here, we present a novel genetic Flrt3-CreERT2 knock-in mouse that labels a small subpopulation of RGCs. Using single-cell injection of fluorescent dyes in Flrt3 positive RGCs, we distinguished four morphological RGC subtypes. Anterograde tracings using a fluorescent Cre-dependent Adeno-associated virus (AAV) revealed that a subgroup of Flrt3 positive RGCs specifically project to the medial terminal nucleus (MTN), which is part of the accessory optic system (AOS) and is essential in driving reflex eye movements for retinal image stabilization. Functional characterization using ex vivo patch-clamp recordings showed that the MTN-projecting Flrt3 RGCs preferentially respond to downward motion in an ON-fashion. These neurons distribute in a regular pattern and most of them are bistratified at the level of the ON and OFF bands of cholinergic starburst amacrine cells where they express the known ON-OFF direction-selective RGC marker CART. Together, our results indicate that MTN-projecting Flrt3 RGCs represent a new functionally homogeneous AOS projecting direction-selective RGC subpopulation.

Rapid multi-directed cholinergic transmission in the central nervous system.

In many parts of the central nervous system, including the retina, it is unclear whether cholinergic transmission is mediated by rapid, point-to-point synaptic mechanisms, or slower, broad-scale 'non-synaptic' mechanisms. Here, we characterized the ultrastructural features of cholinergic connections between direction-selective starburst amacrine cells and downstream ganglion cells in an existing serial electron microscopy data set, as well as their functional properties using electrophysiology and two-photon acetylcholine (ACh) imaging. Correlative results demonstrate that a 'tripartite' structure facilitates a 'multi-directed' form of transmission, in which ACh released from a single vesicle rapidly (~1 ms) co-activates receptors expressed in multiple neurons located within ~1 µm of the release site. Cholinergic signals are direction-selective at a local, but not global scale, and facilitate the transfer of information from starburst to ganglion cell dendrites. These results suggest a distinct operational framework for cholinergic signaling that bears the hallmarks of synaptic and non-synaptic forms of transmission.

Contributions of Retinal Direction Selectivity to Central Visual Processing.

The brain monitors the sensory environment via signals from the sensory periphery, such as the olfactory epithelium, the inner ear, and the retina. Understanding how sensory stimuli are processed throughout the sensory hierarchy, and how this relates to behavior, is a central outstanding question in the field of neuroscience. The processing of visual motion in mice offers unique opportunities for addressing these questions thanks to a rich literature on the anatomical and physiological properties of motion-sensitive neurons across the visual system, paired with recent developments of cutting-edge genetic and imaging approaches. A visual scene typically contains motion originating from either moving objects or optic flow caused by self-generated movements. Neurons encoding the direction of visual motion are said to be 'direction-selective'. It was historically believed the circuits responsible for creating direction selectivity de novo exist within the visual cortex. Yet, in mice, direction-selective responses can be found already in the retina, suggesting in this model organism visual motion analysis starts at the earliest stage of the visual hierarchy. This minireview presents emerging literature demonstrating how retinal direction-selective cells make causal contributions to central visual motion processing and visually guided behaviors in mice, and their potential clinical relevance, and outlines experiments for testing remaining questions. Research in this field will undoubtedly continue to advance our understanding of the basic principles of the visual system and how sensory neurons extract fundamental features of the world.

A segregated cortical stream for retinal direction selectivity.

Visual features extracted by retinal circuits are streamed into higher visual areas (HVAs) after being processed along the visual hierarchy. However, how specialized neuronal representations of HVAs are built, based on retinal output channels, remained unclear. Here, we addressed this question by determining the effects of genetically disrupting retinal direction selectivity on motion-evoked responses in visual stages from the retina to HVAs in mice. Direction-selective (DS) cells in the rostrolateral (RL) area that prefer higher temporal frequencies, and that change direction tuning bias as the temporal frequency of a stimulus increases, are selectively reduced upon retinal manipulation. DS cells in the primary visual cortex projecting to area RL, but not to the posteromedial area, were similarly affected. Therefore, the specific connectivity of cortico-cortical projection neurons routes feedforward signaling originating from retinal DS cells preferentially to area RL. We thus identify a cortical processing stream for motion computed in the retina.

Spatiotemporally Asymmetric Excitation Supports Mammalian Retinal Motion Sensitivity.

The detection of visual motion is a fundamental function of the visual system. How motion speed and direction are computed together at the cellular level, however, remains largely unknown. Here, we suggest a circuit mechanism by which excitatory inputs to direction-selective ganglion cells in the mouse retina become sensitive to the motion speed and direction of image motion. Electrophysiological, imaging, and connectomic analyses provide evidence that the dendrites of ON direction-selective cells receive spatially offset and asymmetrically filtered glutamatergic inputs along motion-preference axis from asymmetrically wired bipolar and amacrine cell types with distinct release dynamics. A computational model shows that, with this spatiotemporal structure, the input amplitude becomes sensitive to speed and direction by a preferred direction enhancement mechanism. Our results highlight the role of an excitatory mechanism in retinal motion computation by which feature selectivity emerges from non-selective inputs.

Adaptive disinhibitory gating by VIP interneurons permits associative learning.

Learning drives behavioral adaptations necessary for survival. While plasticity of excitatory projection neurons during associative learning has been extensively studied, little is known about the contributions of local interneurons. Using fear conditioning as a model for associative learning, we found that behaviorally relevant, salient stimuli cause learning by tapping into a local microcircuit consisting of precisely connected subtypes of inhibitory interneurons. By employing deep-brain calcium imaging and optogenetics, we demonstrate that vasoactive intestinal peptide (VIP)-expressing interneurons in the basolateral amygdala are activated by aversive events and provide a mandatory disinhibitory signal for associative learning. Notably, VIP interneuron responses during learning are strongly modulated by expectations. Our findings indicate that VIP interneurons are a central component of a dynamic circuit motif that mediates adaptive disinhibitory gating to specifically learn about unexpected, salient events, thereby ensuring appropriate behavioral adaptations.

Visual Circuits: Division of Labor Revealed.

Each mosaic of retinal ganglion cells is thought to extract the same visual feature across mouse retina, but a new study shows that ganglion cells of the same type actually show different light response depending on retinal location.