Retinal Circuits Research Articles

Spatial regulation of amacrine cell genesis by Semaphorin 3f

PurposeThe axonal projections of retinal ganglion cells (RGCs) of the eye are topographically organized so that spatial information from visual images is preserved. This retinotopic organization is established during development by secreted morphogens that pattern domains of transcription factor expression within naso-temporal and dorso-ventral quadrants of the embryonic eye. Poorly understood are the downstream signaling molecules that generate the topographically organized retinal cells and circuits. The secreted signaling molecule Semaphorin 3fa (Sema3fa) belongs to the Sema family of molecules that provide positional information to developing cells. Here, we test a role for Sema3fa in cell genesis of the temporal zebrafish retina. MethodsWe compare retinal cell genesis in wild type and sema3fa CRISPR zebrafish mutants by in situ hybridization and immunohistochemistry. ResultsWe find that mRNAs for sema3fa and known receptors, neuropilin2b (nrp2b) and plexina1a (plxna1a), are expressed by progenitors of the temporal, but not nasal zebrafish embryonic retina. In the sema3faca304/ca304 embryo, initially the domains of expression for atoh7 and neurod4, transcription factors necessary for the specification of RGCs and amacrine cells, respectively, are disrupted. Yet, post-embryonically only amacrine cells of the temporal retina are reduced in numbers, with both GABAergic and glycinergic subtypes affected. ConclusionsThese data suggest that Sema3fa acts early on embryonic temporal progenitors to control in a spatially-dependent manner the production of amacrine cells, possibly to allow the establishment of neural circuits with domain-specific functions. We propose that spatially restricted extrinsic signals in the neural retina control cell genesis in a domain-dependent manner.

Robust cone-mediated signaling persists late into rod photoreceptor degeneration.

Rod photoreceptor degeneration causes deterioration in the morphology and physiology of cone photoreceptors along with changes in retinal circuits. These changes could diminish visual signaling at cone-mediated light levels, thereby limiting the efficacy of treatments such as gene therapy for rescuing normal, cone-mediated vision. However, the impact of progressive rod death on cone-mediated signaling remains unclear. To investigate the fidelity of retinal ganglion cell (RGC) signaling throughout disease progression, we used a mouse model of rod degeneration (Cngb1neo/neo). Despite clear deterioration of cone morphology with rod death, cone-mediated signaling among RGCs remained surprisingly robust: spatiotemporal receptive fields changed little and the mutual information between stimuli and spiking responses was relatively constant. This relative stability held until nearly all rods had died and cones had completely lost well-formed outer segments. Interestingly, RGC information rates were higher and more stable for natural movies than checkerboard noise as degeneration progressed. The main change in RGC responses with photoreceptor degeneration was a decrease in response gain. These results suggest that gene therapies for rod degenerative diseases are likely to prolong cone-mediated vision even if there are changes to cone morphology and density.

Suppression without inhibition: how retinal computation contributes to saccadic suppression

Visual perception remains stable across saccadic eye movements, despite the concurrent strongly disruptive visual flow. This stability is partially associated with a reduction in visual sensitivity, known as saccadic suppression, which already starts in the retina with reduced ganglion cell sensitivity. However, the retinal circuit mechanisms giving rise to such suppression remain unknown. Here, we describe these mechanisms using electrophysiology in mouse, pig, and macaque retina, 2-photon calcium imaging, computational modeling, and human psychophysics. We find that sequential stimuli, like those that naturally occur during saccades, trigger three independent suppressive mechanisms in the retina. The main mechanism is triggered by contrast-reversing sequential stimuli and originates within the receptive field center of ganglion cells. It does not involve inhibition or other known suppressive mechanisms like saturation or adaptation. Instead, it relies on temporal filtering of the inherently slow response of cone photoreceptors coupled with downstream nonlinearities. Two further mechanisms of suppression are present predominantly in ON ganglion cells and originate in the receptive field surround, highlighting another disparity between ON and OFF ganglion cells. The mechanisms uncovered here likely play a role in shaping the retinal output following eye movements and other natural viewing conditions where sequential stimulation is ubiquitous.

Divergent outer retinal circuits drive image and non-image visual behaviors.

SUMMARYImage- and non-image-forming vision are essential for animal behavior. Here we use genetically modified mouse lines to examine retinal circuits driving image- and non-image-functions. We describe the outer retinal circuits underlying the pupillary light response (PLR) and circadian photoentrainment, two non-image-forming behaviors. Rods and cones signal light increments and decrements through the ON and OFF pathways, respectively. We find that the OFF pathway drives image-forming vision but cannot drive circadian photoentrainment or the PLR. Cone light responses drive image formation but fail to drive the PLR. At photopic levels, rods use the primary and secondary rod pathways to drive the PLR, whereas at the scotopic and mesopic levels, rods use the primary pathway to drive the PLR, and the secondary pathway is insufficient. Circuit dynamics allow rod ON pathways to drive two non-image-forming behaviors across a wide range of light intensities, whereas the OFF pathway is potentially restricted to image formation.

Characterization of retinal function and structure in the MPTP murine model of Parkinson’s disease

In addition to well characterized motor symptoms, visual disturbances are increasingly recognized as an early manifestation in Parkinson’s disease (PD). A better understanding of the mechanisms underlying these changes would facilitate the development of vision tests which can be used as preclinical biomarkers to support the development of novel therapeutics for PD. This study aims to characterize the retinal phenotype of a mouse model of dopaminergic dysfunction and to examine whether these changes are reversible with levodopa treatment. We use a 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) mouse model of PD to characterize the neurotoxic effects of MPTP on in vivo retinal function (electroretinography, ERG), retinal structure (optical coherence tomography, OCT) and retinal dopaminergic cell number (tyrosine hydroxylase immunohistochemistry, IHC) at two time points (21 and 45 days) post MPTP model induction. We also investigate the effect of levodopa (L-DOPA) as a proof-of-principle chronic intervention against MPTP-induced changes in the retina. We show that MPTP decreases dopaminergic amacrine cell number (9%, p < 0.05) and that a component of the ERG that involves these cells, in particular oscillatory potential (OP) peak timing, was significantly delayed at Day 45 (7–13%, p < 0.01). This functional deficit was paralleled by outer plexiform layer (OPL) thinning (p < 0.05). L-DOPA treatment ameliorated oscillatory potential deficits (7–13%, p < 0.001) in MPTP animals. Our data suggest that the MPTP toxin slows the timing of inner retinal feedback circuits related to retinal dopaminergic pathways which mirrors findings from humans with PD. It also indicates that the MPTP model causes structural thinning of the outer retinal layer on OCT imaging that is not ameliorated with L-DOPA treatment. Together, these non-invasive measures serve as effective biomarkers for PD diagnosis as well as for quantifying the effect of therapy.

Trans-Seq maps a selective mammalian retinotectal synapse instructed by Nephronectin.

The mouse visual system serves as an accessible model to understand mammalian circuit wiring. Despite rich knowledge in retinal circuits, the long-range connectivity map from distinct retinal ganglion cell (RGC) types to diverse brain neuron types remains unknown. In this study, we developed an integrated approach, called Trans-Seq, to map RGCs to superior collicular (SC) circuits. Trans-Seq combines a fluorescent anterograde trans-synaptic tracer, consisting of codon-optimized wheat germ agglutinin fused to mCherry, with single-cell RNA sequencing. We used Trans-Seq to classify SC neuron types innervated by genetically defined RGC types and predicted a neuronal pair from αRGCs to Nephronectin-positive wide-field neurons (NPWFs). We validated this connection using genetic labeling, electrophysiology and retrograde tracing. We then used transcriptomic data from Trans-Seq to identify Nephronectin as a determinant for selective synaptic choice from αRGC to NPWFs via binding to Integrin α8β1. The Trans-Seq approach can be broadly applied for post-synaptic circuit discovery from genetically defined pre-synaptic neurons.

Cone-Driven Retinal Responses Are Shaped by Rod But Not Cone HCN1.

Signal integration of converging neural circuits is poorly understood. One example is in the retina where the integration of rod and cone signaling is responsible for the large dynamic range of vision. The relative contribution of rods versus cones is dictated by a complex function involving background light intensity and stimulus temporal frequency. One understudied mechanism involved in coordinating rod and cone signaling onto the shared retinal circuit is the hyperpolarization activated current (Ih) mediated by hyperpolarization-activated cyclic nucleotide-gated 1 (HCN1) channels expressed in rods and cones. Ih opposes membrane hyperpolarization driven by activation of the phototransduction cascade and modulates the strength and kinetics of the photoreceptor voltage response. We examined conditional knock-out (KO) of HCN1 from mouse rods using electroretinography (ERG). In the absence of HCN1, rod responses are prolonged in dim light which altered the response to slow modulation of light intensity both at the level of retinal signaling and behavior. Under brighter intensities, cone-driven signaling was suppressed. To our surprise, conditional KO of HCN1 from mouse cones had no effect on cone-mediated signaling. We propose that Ih is dispensable in cones because of the high level of temporal control of cone phototransduction. Thus, HCN1 is required for cone-driven retinal signaling only indirectly by modulating the voltage response of rods to limit their output.SIGNIFICANCE STATEMENT Hyperpolarization gated hyperpolarization-activated cyclic nucleotide-gated 1 (HCN1) channels carry a feedback current that helps to reset light-activated photoreceptors. Using conditional HCN1 knock-out (KO) mice we show that ablating HCN1 from rods allows rods to signal in bright light when they are normally shut down. Instead of enhancing vision this results in suppressing cone signaling. Conversely, ablating HCN1 from cones was of no consequence. This work provides novel insights into the integration of rod and cone signaling in the retina and challenges our assumptions about the role of HCN1 in cones.

ERG and Behavioral CFF in Light-Damaged Albino Rats.

The full-field ERG is useful for index rod- or cone-mediated retinal function in rodent models of retinal degeneration. However, the relationship between the ERG response amplitudes and visually guided behavior, such as flicker detection, is not well understood. A comparison of ERG to behavioral responses in a light-damage model of retinal degeneration allows us to better understand the functional implications of electrophysiological changes. Flicker-ERG and behavioral responses to flicker were used to determine critical flicker frequency (CFF) under scotopic and photopic conditions before and up to 90 d after a 10-day period of low-intensity light damage. Dark- and light-adapted ERG flash responses were significantly reduced after light damage. The a-wave was permanently reduced, while the b-wave amplitude recovered over three weeks after light damage. There was a small, but significant dip in scotopic ERG CFF. Photopic behavioral CFF was slightly lower following light damage. The recovery of the b-wave amplitude and flicker sensitivity demonstrates the plasticity of retinal circuits following photopic injury.

Feature Detection by Retinal Ganglion Cells.

Retinal circuits transform the pixel representation of photoreceptors into the feature representations of ganglion cells, whose axons transmit these representations to the brain. Functional, morphological, and transcriptomic surveys have identified more than 40 retinal ganglion cell (RGC) types in mice. RGCs extract features of varying complexity; some simply signal local differences in brightness (i.e., luminance contrast), whereas others detect specific motion trajectories. To understand the retina, we need to know how retinal circuits give rise to the diverse RGC feature representations. A catalog of the RGC feature set, in turn, is fundamental to understanding visual processing in the brain. Anterograde tracing indicates that RGCs innervate more than 50 areas in the mouse brain. Current maps connecting RGC types to brain areas are rudimentary, as is our understanding of how retinal signals are transformed downstream to guide behavior. In this article, I review the feature selectivities of mouse RGCs, how they arise, and how they are utilized downstream. Not only is knowledge of the behavioral purpose of RGC signals critical for understanding the retinal contributions to vision; it can also guide us to the most relevant areas of visual feature space.

Strip1 regulates retinal ganglion cell survival by suppressing Jun-mediated apoptosis to promote retinal neural circuit formation.

In the vertebrate retina, an interplay between retinal ganglion cells (RGCs), amacrine (AC), and bipolar (BP) cells establishes a synaptic layer called the inner plexiform layer (IPL). This circuit conveys signals from photoreceptors to visual centers in the brain. However, the molecular mechanisms involved in its development remain poorly understood. Striatin-interacting protein 1 (Strip1) is a core component of the striatin-interacting phosphatases and kinases (STRIPAK) complex, and it has shown emerging roles in embryonic morphogenesis. Here, we uncover the importance of Strip1 in inner retina development. Using zebrafish, we show that loss of Strip1 causes defects in IPL formation. In strip1 mutants, RGCs undergo dramatic cell death shortly after birth. AC and BP cells subsequently invade the degenerating RGC layer, leading to a disorganized IPL. Mechanistically, zebrafish Strip1 interacts with its STRIPAK partner, Striatin 3 (Strn3), and both show overlapping functions in RGC survival. Furthermore, loss of Strip1 or Strn3 leads to activation of the proapoptotic marker, Jun, within RGCs, and Jun knockdown rescues RGC survival in strip1 mutants. In addition to its function in RGC maintenance, Strip1 is required for RGC dendritic patterning, which likely contributes to proper IPL formation. Taken together, we propose that a series of Strip1-mediated regulatory events coordinates inner retinal circuit formation by maintaining RGCs during development, which ensures proper positioning and neurite patterning of inner retinal neurons.

Connexin45 colocalization patterns in the plexiform layers of the developing mouse retina.

Chemical and electrical synapses (gap junctions) are widely prevalent in the nervous system. Gap junctions emerge long before chemical synapses, allowing communication between developing cells, and are thought to be involved in establishing neural circuits. Mounting evidence indicates that these two modalities of synaptic transmission closely interact during retinal development and that such interactions play a critical role in synaptogenesis and circuit formation during the perinatal period. In vertebrates, gap junctions consist of two connexins which in turn are made up of six connexins (Cx). To what extent Cx45 and Cx36, the most abundant connexins in the retina, are involved in synaptogenesis and retinal circuit formation is not known. The here presented immunohistochemical study used stainings of Cx45, Cx36 and Synaptophysin in the outer and inner (IPL) plexiform layers of postnatal day 8-16 mice retinas to shed light on the role of connexins during critical neuronal developmental processes. Cx45 and Cx36 expressions in both plexiform layers of the mouse retina increased till eye opening and dropped afterwards. The percentage of heterotypic Cx45/Cx36 gap junctions is also higher before the critical event of eye opening. Finally, Cx45 is closely located and/or colocalized with Synaptophysin also shortly before eye opening in the IPL of the mouse retina. All findings point towards a pivotal role for Cx45 during postnatal synaptogenesis in the mouse retina. However, a more functional study is needed to determine the role of Cx45 during synaptogenesis and circuit formation.

Emergence of Direction-Selective Retinal Cell Types in Task-Optimized Deep Learning Models.

Convolutional neural networks (CNNs), a class of deep learning models, have experienced recent success in modeling sensory cortices and retinal circuits through optimizing performance on machine learning tasks, otherwise known as task optimization. Previous research has shown task-optimized CNNs to be capable of providing explanations as to why the retina efficiently encodes natural stimuli and how certain retinal cell types are involved in efficient encoding. In our work, we sought to use task-optimized CNNs as a means of explaining computational mechanisms responsible for motion-selective retinal circuits. We designed a biologically constrained CNN and optimized its performance on a motion-classification task. We drew inspiration from psychophysics, deep learning, and systems neuroscience literature to develop a toolbox of methods to reverse engineer the computational mechanisms learned in our model. Through reverse engineering our model, we proposed a computational mechanism in which direction-selective ganglion cells and starburst amacrine cells, both experimentally observed retinal cell types, emerge in our model to discriminate among moving stimuli. This emergence suggests that direction-selective circuits in the retina are ecologically designed to robustly discriminate among moving stimuli. Our results and methods also provide a framework for how to build more interpretable deep learning models and how to understand them.

Müller Glia in Retinal Development: From Specification to Circuit Integration.

Müller glia of the retina share many features with astroglia located throughout the brain including maintenance of homeostasis, modulation of neurotransmitter spillover, and robust response to injury. Here we present the molecular factors and signaling events that govern Müller glial specification, patterning, and differentiation. Next, we discuss the various roles of Müller glia in retinal development, which include maintaining retinal organization and integrity as well as promoting neuronal survival, synaptogenesis, and phagocytosis of debris. Finally, we review the mechanisms by which Müller glia integrate into retinal circuits and actively participate in neuronal signaling during development.

Invited Session IV: Top-down vs. bottom-up approaches to computational modeling of vision: Retinal encoding of natural images

Current predictive models of natural scene encoding by retinal ganglion cells perform poorly, especially when tested on stimuli that they are not trained on. This failure is likely to originate at least in part because the functional architecture (e.g. locations of linear and nonlinear components) of these models does not match that of real retinal circuits. We take an empirical approach to simplify the structure of natural scenes. We find that reducing natural movies to 16 linearly integrated regions describes ~80% of the structure of parasol RGC spike responses. We test our understanding by using these simplified stimuli to create high-dimensional metamers that recapitulate the spike response of full-field naturalistic movies. Finally, we identify the retinal computations that convert natural images in 16-dimensional space into 1-dimensional spike outputs. This work provides a description of spatial integration that generalizes well across images and that corresponds closely to known anatomical and functional aspects of retinal circuits.\.

Contributed Session II: In-vivo classification of human cone photoreceptors reveals crystalline S-cone sub-mosaics in the central retina

The topography of S-cones in the macula sets the neural constraints for coding the short-wavelength spectrum of color vision. We find that S-cones tile the central human retina with a non-random crystalline arrangement. This finding departs from previous studies, likely due to limited sampling. In 2 subjects we classified cones using Adaptive Optics Line-scan OCT and a bleaching stimulus of 660±10 nm. 8 ROIs per subject were classified at 1.5° and ~4° eccentricity across the 4 meridians. Numbers of total and S- cones per ROI spanned 541-3545 (mean: 1823) and 38-171 (mean: 99), respectively. We measured S-cone spacing in each ROI using the established method of Density Recovery Profile (DRP). To compare with random arrangement, we generated 1000 Monte Carlo (MC) simulations of each ROI such that its cone locations were maintained but locations of S-cones within it were randomized. We then measured the radius in units of inter-cone distance for which S-cone density was significantly lower than MC distributions, finding low density in a 1-cone radius in 13/16 ROIs (8/8 at 1.5°, p ≤ .037. 5/8 at 3.5°-4.5°, p ≤ .002), and up to a 2-cone radius in 12/16 ROIs (7/8 at 1.5°, p ≤ .002. 5/8 at 3.5°-4.5° p ≤ .003). Further experiments will include additional human subjects, ROIs at higher eccentricities, and classification using a short-wavelength bleach. Together, these findings have important implications for retinal development and color coding retinal circuits.

Inhibition, but not excitation, recovers from partial cone loss with greater spatiotemporal integration, synapse density, and frequency.

SUMMARYNeural circuits function in the face of changing inputs, either caused by normal variation in stimuli or by cell death. To maintain their ability to perform essential computations with partial inputs, neural circuits make modifications. Here, we study the retinal circuit’s responses to changes in light stimuli or in photoreceptor inputs by inducing partial cone death in the mature mouse retina. Can the retina withstand or recover from input loss? We find that the excitatory pathways exhibit functional loss commensurate with cone death and with some aspects predicted by partial light stimulation. However, inhibitory pathways recover functionally from lost input by increasing spatiotemporal integration in a way that is not recapitulated by partially stimulating the control retina. Anatomically, inhibitory synapses are upregulated on secondary bipolar cells and output ganglion cells. These findings demonstrate the greater capacity for inhibition, compared with excitation, to modify spatiotemporal processing with fewer cone inputs.

The Intrinsic Blue Light Responses of Avian Müller Glial Cells Imply Calcium Release from Internal Stores.

The retina of vertebrates is responsible for capturing light through visual (cones and rods) and non-visual photoreceptors (intrinsically photosensitive retinal ganglion cells and horizontal cells) triggering a number of essential activities associated to image- and non-image forming functions (photic entrainment of daily rhythms, pupillary light reflexes, pineal melatonin inhibition, among others). Although the retina contains diverse types of neuronal based-photoreceptors cells, originally classified as ciliary- or rhabdomeric-like types, in recent years, it has been shown that the major glial cell type of the retina, the Müller glial cells (MC), express blue photopigments as Opn3 (encephalopsin) and Opn5 (neuropsin) and display light responses associated to intracellular Ca2 + mobilization. These findings strongly propose MC as novel retinal photodetectors (Rios et al., 2019). Herein, we further investigated the intrinsic light responses of primary cultures of MC from embryonic chicken retinas specially focused on Ca2 + mobilization by fluorescence imaging and the identity of the internal Ca2 + stores responsible for blue light responses. Results clearly demonstrated that light responses were specific to blue light of long time exposure, and that the main Ca2 + reservoir to trigger downstream responses came from intracellular stores localized in the endoplasmic reticulum These observations bring more complexity to the intrinsic photosensitivity of retinal cells, particularly with regard to the detection of light in the blue range of visible spectra, and add novel functions to glial cells cooperating with other photoreceptors to detect and integrate ambient light in the retinal circuit and participate in cell to cell communication. Summary statement: Non-neuronal cells in the vertebrate retina, Muller glial cells, express non-canonical photopigments and sense blue light causing calcium release from intracellular stores strongly suggesting a novel intrinsic photosensitivity and new regulatory events mediating light-driven processes with yet unknown physiological implications.

The roles of microglia in neural remodeling during retinal degeneration.

Retina remodeling is a consequence of many retinal degenerative diseases that are characterized by progressive photoreceptor death. Retina remodeling involves a series of complex pathological processes, consisting of photoreceptor degeneration and death, as well as retinal cell reprogramming and "rewiring". This rewiring alters retinal neural circuits that are centered on synaptic connections and lead to widespread death of retinal cells. Retinal remodeling, especially inner retinal remodeling, is the major factor that limits the effectiveness of various treatment strategies, including cell therapy; thus, it is important to elucidate the mechanisms involved in retinal remodeling during retinal degeneration. Microglia are the dominant immune cells in the retina. Microglia monitor the retinal microenvironment, are activated following retinal injury or degeneration, have powerful phagocytosis capabilities, and play a critical role in synaptic pruning during central neural system development. Analogously, microglia have been found to participate in the clearance of synaptic elements in a complement-dependent manner in the classic retinitis pigmentosa (RP) model, Royal College of Surgeons (RCS) rats, and retard the formation of ectopic neuritogenesis and the deterioration of visual function during retinal degeneration. Since previous research on microglia has rarely concentrated on synaptic remodeling during retinal degeneration, summarizing the microglial mechanisms involved in retinal remodeling is necessary in order to design compounds targeting microglia and retinal remodeling that might be promising therapeutic strategies for treating retinal degeneration.

Photoreceptors for Entrainment of Circadian Rhythms

In mammals, the intrinsic circadian clock regulates various physiological rhythms, including sleep-wake cycles. These circadian rhythms can be entrained to daily 24-hour light-dark cycles by virtue of photoreception within the retina. The photo-entrainment of circadian rhythms is predominantly mediated by intrinsically photosensitive retinal ganglion cells (ipRGCs). In this article, we review the mechanisms of action of ipRGC photoreception, the retinal circuit involving ipRGCs, and the heterogeneity of ipRGCs in mice.

The cell adhesion molecule Sdk1 shapes assembly of a retinal circuit that detects localized edges.

Nearly 50 different mouse retinal ganglion cell (RGC) types sample the visual scene for distinct features. RGC feature selectivity arises from their synapses with a specific subset of amacrine (AC) and bipolar cell (BC) types, but how RGC dendrites arborize and collect input from these specific subsets remains poorly understood. Here we examine the hypothesis that RGCs employ molecular recognition systems to meet this challenge. By combining calcium imaging and type-specific histological stains, we define a family of circuits that express the recognition molecule Sidekick-1 (Sdk1), which include a novel RGC type (S1-RGC) that responds to local edges. Genetic and physiological studies revealed that Sdk1 loss selectively disrupts S1-RGC visual responses, which result from a loss of excitatory and inhibitory inputs and selective dendritic deficits on this neuron. We conclude that Sdk1 shapes dendrite growth and wiring to help S1-RGCs become feature selective.