Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Electrophysiology wikipedia , lookup
Clinical neurochemistry wikipedia , lookup
Development of the nervous system wikipedia , lookup
Subventricular zone wikipedia , lookup
Neuroanatomy wikipedia , lookup
Neuropsychopharmacology wikipedia , lookup
Optogenetics wikipedia , lookup
Superior colliculus wikipedia , lookup
VS01CH12-Huberman Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. ANNUAL REVIEWS ARI 27 October 2015 22:15 Further Click here to view this article's online features: • Download figures as PPT slides • Navigate linked references • Download citations • Explore related articles • Search keywords Contributions of Retinal Ganglion Cells to Subcortical Visual Processing and Behaviors Onkar S. Dhande,∗ Benjamin K. Stafford, Jung-Hwan A. Lim, and Andrew D. Huberman Neurosciences Department, Neurobiology Section in the Division of Biological Sciences, and Department of Ophthalmology, University of California, San Diego, La Jolla, California 92093; email: [email protected], [email protected] Annu. Rev. Vis. Sci. 2015. 1:291–328 Keywords The Annual Review of Vision Science is online at vision.annualreviews.org retina, cell types, parallel pathways, vision, behavior This article’s doi: 10.1146/annurev-vision-082114-035502 Abstract c 2015 by Annual Reviews. Copyright All rights reserved ∗ Corresponding author Every aspect of visual perception and behavior is built from the neural activity of retinal ganglion cells (RGCs), the output neurons of the eye. Here, we review progress toward understanding the many types of RGCs that communicate visual signals to the brain, along with the subcortical brain regions that use those signals to build and respond to representations of the outside world. We emphasize recent progress in the use of mouse genetics, viral circuit tracing, and behavioral psychophysics to define and map the various RGCs and their associated networks. We also address questions about the homology of RGC types in mice and other species including nonhuman primates and humans. Finally, we propose a framework for understanding RGC typology and for highlighting the relationship between RGC type-specific circuitry and the processing stations in the brain that support and give rise to the perception of sight. 291 VS01CH12-Huberman ARI 27 October 2015 22:15 INTRODUCTION Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. For mammals, including humans, the eyes are the sole entry point for visual information to reach the brain. The retina—the thin sheet of light-sensing neural tissue at the back of the eye—is a true natural wonder; it is about as thick as a credit card and yet contains all the necessary machinery for converting light into a rich array of electrical signals that the brain can understand and unpack to create visual percepts (Azeredo da Silveira & Roska 2011). The brain then uses these signals to create internal representations of the location and qualities of visual features in the environment. The retina consists of five layers: three layers of cells and two layers of synaptic connections, called plexiform layers. The cell layer closest to the lens contains neurons called retinal ganglion cells (RGCs); RGC axons exit the eye to form the optic nerve, which connects with the brain (Figure 1a). The electrical signals carried by these RGC axons establish the crucial first link between the outside world and our internal perception of sight. A large amount of experimental attention has been dedicated to understanding the quality and type of visual signals encoded by the retina and communicated to the brain by RGCs. Indeed, progress in this area has been substantial, especially for the mouse. During the past decade, researchers have combined classic anatomical and electrophysiological tools with genetic tools to create a systematic and detailed exploration of both the number and type of RGCs that exist, as well as brain areas with which they communicate. Detailed maps of the polysynaptic circuits driven by specific types of mouse RGCs (e.g., Cruz-Martı́n et al. 2014), and causal tests of the specific roles they play in visual processing (e.g., Chen et al. 2011), are starting to reveal the organizational logic of the early visual pathway in this species. The hope and expectation is that the development and application of similar sets of tools in primate species will allow the field of visual neuroscience to develop an even stronger mechanistic understanding of human vision, as nonhuman primates have eyes and brains that are so similar to those of humans. In the following sections, we discuss the various criteria by which RGCs are categorized into distinct types and subtypes, as well as key tools and information that are still lacking for RGC typing. We also focus on the various subcortical structures to which RGCs connect, and we discuss how those structures process retinal information to influence different light-dependent perceptions and behaviors. −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−→ Figure 1 The retina is a layered structure consisting of five neural cell types. (a) Rod and cone photoreceptors in the outer nuclear layer signal the arrival of photons by modulating their glutamate release onto the dendrites of bipolar cells in the outer plexiform layer. Glutamate released by photoreceptors is further modulated via feedback and feedforward inhibition from horizontal cells. Bipolar cells, whose cell bodies reside in the inner nuclear layer, transmit input they receive from photoreceptors by releasing glutamate onto a complex network of ganglion cell and amacrine cell dendrites in the inner plexiform layer. Most amacrine cells are axonless and modulate the response of ganglion cells via feedback and feedforward mechanisms. Ganglion cells are the sole output neuron of the retina, and their axons form the optic nerve, which transmits information to visual processing centers in the brain. (b) The cells and circuits in the retina can be divided further on the basis of their functional output. This segregation occurs at the first synapse between photoreceptors and bipolar cells. Bipolar cells that transmit an On signal express what are called sign-inverting metabotropic glutamate receptors, which convert the excitatory glutamatergic signal from the photoreceptors into an inhibitory one. Those that transmit an Off signal express conventional ionotropic glutamate receptors and are excited by glutamate release from photoreceptors. This functional segregation also has an anatomical correlate: The axon terminals of On and Off bipolar cells stratify in different halves of the inner plexiform layer. Because ganglion cells typically inherit the functional polarity (On versus Off ) of the bipolar cells that innervate them, their function can be predicted on the basis of where in the inner plexiform layer their dendrites stratify. Abbreviation: mGluR6, metabotropic glutamate receptor. 292 Dhande et al. VS01CH12-Huberman ARI 27 October 2015 22:15 a Cone Layer Rod Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. Outer nuclear Outer plexiform Inner nuclear Bipolar cell Horizontal cell Inner plexiform Amacrine cell Ganglion cell Ganglion cell Optic nerve b On pathway Off pathway Sign-inverting mGluR6 synapse Conventional ionotropic synapse On bipolar cell Off bipolar cell Functional polarity Off On 1s 1s On ganglion cell Light-on response Off ganglion cell Light-off response www.annualreviews.org • RGC Contributions to Feature Processing 293 VS01CH12-Huberman ARI 27 October 2015 22:15 ORGANIZATIONAL LOGIC OF THE MAJOR RETINAL CELL TYPES AND CIRCUITS Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. Light Conversion and Signal Transduction in the Retina Processing of light information occurs first in the outer retina, where photoreceptors—rods and cones—absorb photons in their outer segments and, by using various opsin transduction cascades (reviewed in Korenbrot 2012), convert those photons into electrical signals that the rest of the nervous system can understand. Rods are highly sensitive to photons and operate under low-light, or scotopic, conditions, whereas cones absorb light only at brighter, or photopic, levels. Rods and cones synapse onto their postsynaptic partners, bipolar and horizontal cells. The details of the photoreceptor-to-target-cell communication are presented schematically in Figure 1 and are described in detail elsewhere (Demb & Singer 2015). For purposes of this review, it suffices to say that the presence of light in the visual field reduces glutamate release from rods and cones. This reduction in glutamate release effectively excites bipolar cells expressing metabotropic sign-inverting glutamate receptors (On bipolar cells) by removing inhibition and causes these cells to release glutamate from their terminals in the inner plexiform layer (IPL) (Figure 1b). In contrast, the presence of darkness triggers photoreceptors to increase glutamate release, thereby exciting bipolar cells that express conventional ionotropic glutamate receptors (Off bipolar cells) (Figure 1b). Horizontal cells, which are GABAergic and inhibitory, shape the flow of photoreceptor-to-bipolar-cell information through feedback and feedforward mechanisms (reviewed in Thoreson & Mangel 2012). As mentioned above, bipolar cells are excitatory glutamatergic neurons. They fall into two general categories: rod bipolar cells and cone bipolar cells, depending on their presynaptic inputs. In the mouse, there are at least 13 different types of bipolar cells that have been described morphologically, and each type stratifies its axon terminals at a distinct laminar depth in the inner retina (Figure 2a). Further, the response properties of bipolar cells differ among types, and bipolar cells impart these functional properties (e.g., On versus Off, sustained versus transient) upon the RGCs with which they communicate. In this way, bipolar cell terminals establish the first anatomical laminar segregation of parallel pathways for unique feature detection in the visual system. Indeed, this segregation dictates, in large part, the response properties of the different RGCs. As a general rule, RGCs stratify their dendrites at the same point in the IPL as the axon terminals of the bipolar cell type(s) from which they receive input. There is one known exception in the mouse, however: an Off stratifying ganglion cell that receives input from On bipolar cells via en passant synapses and, therefore, is functionally an On RGC (Figure 2b). Whether additional exceptions exist in the mouse or in other species remains to be seen. Lateral filtering of visual signals at the level of bipolar cell output is accomplished by a specific class of interneurons, the amacrine cells. More than three dozen types of amacrine cells have been −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−→ Figure 2 Origins of parallel pathways in the retina. (a) Bipolar cells can be divided into two broad types: rod and cone bipolar cells. There is only one type of rod bipolar cell ( green), an On type, that stratifies its axon terminals in the extreme vitreal portion of the inner plexiform layer (IPL). There are at least 13 types of bipolar cells. The axon terminals of cells that carry Off signals ( purple) stratify in the scleral half of the IPL, whereas those of cells that carry On signals (blue) stratify in the vitreal half. Figure adapted with permission from Euler et al. (2014). (b) There is at least one type of Off-stratifying ganglion cell that possesses an On light response (Hattar et al. 2002). This cell ( gray) receives On pathway input in the form of en passant synapses (red box) from specific type(s) of On bipolar cell(s) (blue) (Dumitrescu et al. 2009). (c) The extreme heterogeneity in the morphological scale of retinal cell types allows signals to be processed both locally and over long distances in the circuits that detect and process visual information. Wide-field amacrine cell adapted from Lin et al. (2004). Starburst amacrine cell adapted from Wei & Feller (2011). Alpha ganglion cell adapted from El-Danaf & Huberman (2015). 294 Dhande et al. VS01CH12-Huberman ARI 27 October 2015 22:15 identified; some types are wide-field and polyaxonal, allowing them to shape visual processing over considerable spans of the retina, whereas others are narrow-field and thus shape retinal circuit interactions more locally (Figure 2c). All but a few amacrine cells are inhibitory (Grimes et al. 2011, Lee et al. 2014, Krishnaswamy et al. 2015); they either release GABA or glycine (but not both). However, many amacrine types corelease neuromodulators such as acetylcholine (Ach) or Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. a Off bipolar cells 1 2 3a 3b On bipolar cells 4 5a 5b XBC 6 7 Rod bipolar cell 8 9 Off On b Conventional On bipolar cell En passant On bipolar cell En passant synapse Conventional On ganglion cell c Off-stratifying On ganglion cell Wide-field amacrine cell Alpha ganglion cell Starburst amacrine cell www.annualreviews.org • RGC Contributions to Feature Processing 295 VS01CH12-Huberman ARI 27 October 2015 22:15 dopamine (DA) along with GABA or glycine. For a thorough review of amacrine contributions to retinal processing see the accompanying review by Demb & Singer (2015). Ganglion Cells Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. At the risk of sounding RGC-centric, the ultimate purpose of all other retinal cell types and circuits is to pass along information about certain features of the visual world to specific types of RGCs so that they in turn can transmit that information to the brain. It is generally accepted that there are ∼20–30 different RGC types, and the study of how RGCs are classified into distinct types represents a useful framework for classifying cell types elsewhere in the mammalian central nervous system (Seung & Sümbül 2014, Sanes & Masland 2015). Indeed, RGCs offer the unique opportunity to study all of the parameters by which neurons have been traditionally classified: morphology, response to defined stimuli, presynaptic inputs, postsynaptic targets, and molecular expression profiles, among others. Before embarking on a detailed description of the known RGC types and subtypes, we briefly address how a given RGC is classified into a specific category and how, to some extent, this procedure is now standardized. How Are Retinal Ganglion Cell Types Defined? RGCs are classified by several parameters, none of which is definitive of a specific type on its own. One of the hallmark parameters is mosaicism; that is, the relatively uniform spacing of neuronal cell bodies of cells of a given type, and therefore function, in the plane of the retina (Figure 3a). Mosaicism makes intuitive sense because it ensures that visual information is encoded by the same number of cells of a given functional type at any point in the visual scene. Thus, this property is considered by many to be one of the most fundamental features for determining a unique RGC type (reviewed in Cook 1996, Field & Chichilnisky 2007), and remains a highly useful, and readily quantifiable metric for classifying RGC types (Cook 1996). Another classic way by which RGC types have been characterized is by their en face morphology, which includes two features: (a) soma size and (b) dendritic branching patterns in the IPL. These features work well to distinguish and categorize RGC types that tend to have very large versus very small dendritic arbors and somas (Figure 3b) (Rodieck et al. 1985). Indeed, physiologists have used soma size and shape to bias targeting of their recordings toward specific RGC types for decades. On its own, however, dendritic branching pattern is a somewhat precarious feature to rely on for purposes of RGC classification. RGCs that reside within the same mosaic, and that stratify at the same depth(s) within the IPL (and thus encode the same functional features), can have widely different en face morphologies (Bleckert et al. 2014; reviewed in Dhande & Huberman 2014b). These differences are especially prominent in species such as primates, in which high-acuity regions of the retina are used for detailed processing of specific regions of the visual scene (Watanabe & Rodieck 1989, Dacey & Peterson 1992; also see Stone 1983). The third major feature by which RGC types are categorized is their unique patterns of dendritic stratification (Figures 3c and 4) (also see Berson 2008, Sanes & Masland 2015). This property is observed as cell type–specific differences in the IPL depth at which they direct their dendritic arbors. The importance of this feature for RGC typing is hard to overstate because it dictates the identities of the presynaptic cells—bipolars or amacrines—from which the RGCs are positioned to collect synaptic input (see, for example, Figure 2a). Indeed, new techniques focused on quantifying stratification position within the IPL are making it possible to detect differences amongst RGC types with ever greater precision (Manookin et al. 2008, Sümbül et al. 2014). Thus, above all else, dendritic stratification provides our first guess as to which visual features an RGC will respond best. 296 Dhande et al. VS01CH12-Huberman a ARI 27 October 2015 22:15 b Functional mosaics En face morphologies Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. Local object motion detectors Spot detectors Luminance detectors c Stratification patterns Layer ONL OPL INL IPL GCL Figure 3 Organizational principles for typing ganglion cells. (a) The functional output of a ganglion cell can be used as a classification criterion. The cell bodies of ganglion cells with homogenous functions tile the retina, forming semiuniform mosaics. (b) En face morphology, including dendritic field size and complexity, has also been used to classify different types of ganglion cells. Black cell adapted from Sümbül et al. (2014). Dark red cell adapted from Berson et al. (2010). Green and blue cells adapted from El-Danaf & Huberman (2015). (c) Stratification depth within the IPL can also be used to classify cells. The dendrites of cells with specific anatomical and functional properties stratify in highly stereotypical locations within the IPL. Abbreviations: GCL, ganglion cell layer; INL, inner nuclear layer, IPL, inner plexiform layer; ONL, outer nuclear layer; OPL, outer plexiform layer. www.annualreviews.org • RGC Contributions to Feature Processing 297 ARI 27 October 2015 22:15 In fact, modeling data support the idea that if one had to determine RGC type on the basis of only one morphological feature, the sine qua non feature is dendritic stratification (Sümbül et al. 2014). A similarly new and developing method for classifying RGCs into distinct types is their signature pattern of gene expression (Kay et al. 2011, Siegert et al. 2012). To date, the best examples of RGC typing using this method are the melanopsin-expressing intrinsically photosensitive RGCs (ipRGCs). As their name suggests, these RGCs can directly respond to light without the need for rod or cone input, and, therefore, act as a class of RGC photoreceptors (Berson et al. 2002; see below). The ipRGCs are a unique case among RGC types, however, because the expression of a single protein—melanopsin—determines the essence of the cell type classification. Indeed, this classification system is similar to how standard photoreceptors can be classified into distinct types (such as L- or S-cones) based on their unique expression of opsin photopigments (Nathans 1987), rather than on differences in, for example, morphology. Even melanopsin expression, however, has limitations when used for classifying RGCs. As discussed in greater detail below, there are numerous subtypes of ipRGCs which apparently express varying levels of melanopsin. Thus, melanopsin expression alone is insufficient for classifying the various ipRGC subtypes. Given the difficulty in finding individual genes that mark specific types of RGCs, the search for molecular markers has centered on developmental genes, such as transcription factors, that dictate stratification and connectivity (Xiang et al. 1996). A handful of such markers have been identified, but virtually all of them are required in combination to faithfully represent individual RGC types; there are very few bona fide individual markers for distinct RGC types (Table 1) (Badea et al. 2009, Mao et al. 2014, Sweeney et al. 2014). The adhesion molecule JamB is a notable exception, as it marks a very specific RGC type and, reportedly, does not mark any others (Kim et al. 2008, 2010). Other markers that have been used in the past to label specific populations of RGCs, such as the antibody SMI-32, which recognizes the nonphosphorylated form of heavy chain neurofilaments, have recently been shown to label more than a single type of RGC; SMI-32 is expressed at low levels in at least three or four types of RGCs (Lin et al. 2004, Coombs et al. 2006). In sum, many of these markers are selective for specific subtypes within the RGC population but are not exclusive to specific types of RGCs per se. In recent years, transgenic mice harboring fluorescent reporters under the control of specific promoter elements, as well as knockin mice with site-specific insertions of Cre and/or a fluorescent reporter, have been used to label and study specific subtypes of RGCs with great success. Although some of these mice come with the caveat that the transgenes may not recapitulate the endogenous activity of the promoters in their entirety (Kay et al. 2011), these systems have allowed researchers to address virtually all of the typing criteria described in the preceding sections (Huberman et al. 2008b, 2009; Kim et al. 2008; Yonehara et al. 2008; Ecker et al. 2010; Kay et al. 2011; Osterhout et al. 2011; Trenholm et al. 2011; Dhande et al. 2013). In total, these mice have been used to “genetically identify” 15 of the 20–30 RGC subtypes reported to exist in the mouse (Coombs et al. 2006, Völgyi et al. 2009). The morphology and stratification patterns of these RGCs and the transgenic mouse lines that mark them are shown in Figure 4. Indeed, a variety of genetically modified mice are now being used in dozens of laboratories around the world to generate detailed knowledge about RGC type–specific physiology, morphology, central projections, gene expression, and contributions to behavior. Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. VS01CH12-Huberman MAJOR TYPES OF RETINAL GANGLION CELLS As mentioned above, there are an estimated 20–30 morphologically and physiologically distinct RGC subtypes in the mouse and in primates. To parse these into distinct and well-defined groups, it is useful to divide them into several broad functional categories, such as motion-sensitive versus 298 Dhande et al. VS01CH12-Huberman ARI 27 October 2015 22:15 Table 1 Molecular markers and transgenic mouse lines used to study functionally distinct retinal ganglion cell subtypes Retinal ganglion cell subtype Directionselective Icon On–Off DSGC V T N Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. Unique marker(s) D On DSGC V Transgenic line(s) Posterior: CART, Mmp17 Anterior: NA Upward: CART, Col25a1, Cdh6 Downward: CART, Col25a1, Cdh6 Drd4-GFP TRHR-GFP Hoxd10-GFP Hb9-GFP TYW9 Cdh6-CreER BD(FSTL4/Spig1)CreER Huberman et al. (2009) Kay et al. (2011) Rivlin-Etzion et al. (2011) Trenholm et al. (2011) Dhande et al. (2013) Unknown FSTL4/Spig1-GFP Hoxd10-GFP Yonehara et al. (2008) Dhande et al. (2013) Jam-B JAM-B-CreER Kim et al. (2008) SCN-projecting Melanopsin+ Brn3b− Opn4-TauLacZ Opn4-GFP Opn4-Cre Opn4-CreER Hattar et al. (2002) Schmidt et al. (2008) Ecker et al. (2010) Chen et al. (2011) T N Reference(s) D Off DSGC V N Intrinsically photosensitive M1 T OPN-projecting Melanopsin+ Brn3b+ M2 Unknown Melanopsin+ Brn3b+ Opn4-TauLacZ Opn4-GFP Opn4-Cre Cdh3-GFP Schmidt et al. (2008) Ecker et al. (2010) Osterhout et al. (2011) Hattar et al. (2002) M3 Unknown Melanopsin+ Brn3b+ Opn4-GFP Schmidt et al. (2008) Melanopsin+ Brn3b+ SMI32+ Opn4-Cre Ecker et al. (2010) Melanopsin+ Brn3b+ Opn4-Cre Ecker et al. (2010) Unknown TYW3 Kim et al. (2010) Zhang et al. (2012) M4 M5 Object motion W3A W3B Unknown (Continued ) www.annualreviews.org • RGC Contributions to Feature Processing 299 VS01CH12-Huberman ARI 27 October 2015 22:15 Table 1 (Continued ) Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. Retinal ganglion cell subtype Icon Unique marker(s) Transgenic line(s) Reference(s) Center–surround (alpha) W7A W7B SMI-32 Osteopontin TYW7 CB2-GFP KCNG4-Cre Huberman et al. (2008) Kim et al. (2010) Duan et al. (2015) Loomingsensitive PV-5 Unknown PV-Cre Münch et al. (2009) Farrow et al. (2013) Abbreviations: DGSC, direction-selective ganglion cell; GFP, green fluorescent protein; NA, not available; OPN, olivary pretectal nucleus; SCN, suprachiasmatic nucleus. non-motion-sensitive or broad versus detailed spatial feature detectors. To clarify: we refer to an RGC type as a major category of, say, direction-selective RGCs, whereas our description of an RGC subtype resides within a particular categorization. For example, upward-sensitive On–Off direction-selective ganglion cells (DSGCs) represent one On–Off DSGC subtype, downwardsensitive On–Off DSGCs represent another subtype, and so on. Non-Motion-Sensitive Retinal Ganglion Cell Types Some of the earliest RGC types to be studied can be loosely described as spot detectors. These types of RGCs respond to increments or decrements of light in their receptive fields and, in some instances, possess an antagonistic surround. In the following subsections, we highlight several of these types of RGCs that have been well characterized in the past 50–60 years. Alpha retinal ganglion cells. One of the classic RGC types is the alpha cell. First described functionally by Kuffler (1953) in the cat retina, and coined the “alpha” cell by Boycott & Wässle (1974), these cells are monostratified and relatively large, both in terms of their somas and the diameters of their dendritic arbors, making them one of the easiest RGCs to target for filling and/or electrophysiological recordings. Alpha RGCs are center–surround spot detectors, of which there are four subtypes that stratify at different IPL depths in the mouse and that are named for the properties of their receptive field centers: On-transient, On-sustained, Off-transient, and Off-sustained (Pang et al. 2003, van Wyk et al. 2009, Stafford et al. 2010). Although the number of alpha cell subtypes varies across species, the receptive field properties of alpha RGCs are remarkably similar in cats, mice, and guinea pigs (Manookin et al. 2008, van Wyk et al. 2009), and at −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−→ Figure 4 Representative morphologies and inner plexiform layer (IPL) stratification depths of genetically labeled ganglion cells in the retinas of transgenic mouse lines. For each cell type, a representative en face morphology is shown, along with its stereotypical stratification depth within the IPL, normalized to two fiducial strata within the IPL; IPL strata (On and Off ChAT bands) are labeled by a choline acetyltransferase (ChAT) antibody, and their normalized positions are indicated by the dashed lines in the stratification panels. Solid gray boxes around cells indicate multiple cell types labeled in a single transgenic line. Known functions are listed below each cell type. M1 and M2 cells adapted from Berson et al. (2010). M3 cell adapted from Schmidt et al. (2014). M5 cell adapted from Hu et al. (2013). M4 cell adapted from Estevez et al. (2012). W3, W7A, W7B, and K cells adapted from Sümbül et al. (2014). Jam-B cell adapted from Kim et al. (2008). Hoxd10 and CB2 cells adapted from El-Danaf & Huberman (2015). Drd4/TRHR/W9/BDa cell based on unpublished data from the Huberman lab. Abbreviations: DS, direction-selective; t-Off, Off-transient; s-Off, Off-sustained; s-On, On-sustained; aOn–Off, anterior-tuned On–Off; pOn–Off, posterior-tuned On–Off; uOff, upward-tuned Off; uOn–Off, upward-tuned On–Off. 300 Dhande et al. VS01CH12-Huberman ARI 27 October 2015 22:15 Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. Opn4/Opn4-Cre M1 Luminance CB2 M2 Luminance W3 t-Off alpha Local motion Off On Off On M5 Drd4/TRHR/W9 BDa Off On pOn–Off DS uOn–Off DS M4 s-On alpha M3 Off On uOff DS Off On Hoxd10 aOn–Off DS Jam-B W7A/B s-Off Delta (W7B) On DS Off On t-Off (W7A) Off On Unknown Ka Kb Off On www.annualreviews.org • RGC Contributions to Feature Processing 301 VS01CH12-Huberman ARI 27 October 2015 22:15 least one of the subtypes is likely the center–surround RGC made famous by the classic studies by Kuffler (1953) and others in the cat retina (Barlow et al. 1957, Wiesel 1960, Rodieck & Stone 1965). Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. Local edge detectors. First described physiologically in rabbits by Levick (1967), cells belonging to this RGC type prefer stimuli restricted to their small receptive field centers and respond poorly to large spots, full-field stimuli, or extended gratings. Although these cells respond well to small light or dark edges, they exhibit no directional preference. Their circuit properties were later characterized in detail in the rabbit (van Wyk et al. 2006) and, very recently, were genetically identified in the mouse (Zhang et al. 2012). As in the rabbit, mouse local edge detectors are cells with very small dendritic fields and On–Off responses resulting from broad stratification of their dendrites within the central IPL, where they apparently receive mixed On and Off bipolar cell input. Orientation-selective retinal ganglion cells. Single-cell recordings in rabbit, cat, and primate retinas indicate that orientation-selective RGCs are present in these species (Levick & Thibos 1982, Bloomfield 1994, Passaglia et al. 2002, Venkataramani & Taylor 2010). Multisite electrode recordings (Zhao et al. 2013) suggest the presence of orientation-selective RGCs in the mouse retina as well, although such cells have yet to be genetically identified or studied in detail in the mouse at the level of morphology, circuitry, or contribution to central vision. Intrinsically photosensitive retinal ganglion cells. These cells represent a limited subset of ganglion cells (<2% of all RGCs) that express melanopsin, making them intrinsically photosensitive (Berson 2013). Physiologically, they display sustained responses to steady illumination and thereby encode global light intensity levels in the ambient environment (Zhao et al. 2014b). To date, five subtypes of ipRGCs (M1–M5) have been identified, all but one of which are monostratified, and each of which possesses a different level of intrinsic photosensitivity attributable to varying levels of melanopsin expression (Ecker et al. 2010, Hu et al. 2013). Further, some ipRGCs are actually subsets of other subtypes of RGCs. For example, the M4s are On-sustained alpha RGCs, although it is not known whether every On-sustained alpha RGC expresses melanopsin (Estevez et al. 2012). Nevertheless, because On-sustained alpha RGCs express melanopsin, it is clear that some ipRGCs can play multiple roles in visual processing and are not limited to encoding only ambient light intensity (Schmidt et al. 2014). Moreover, ipRGCs exist in both mice and primates, and some cell types have nearly identical physiology and morphology, with the primary difference being one of scale (Figure 5). Whether or not all five subtypes exist in every species in which these cells have been identified, however, remains unclear (Berson et al. 2002, Hannibal et al. 2004, Dacey et al. 2005). Motion-Sensitive Retinal Ganglion Cell Types Certain types of RGCs respond to moving stimuli in addition to responding to increments or decrements of light in their receptive fields. Some types of motion-sensitive RGCs respond to stimuli moving along a particular axis of the visual field, whereas other types respond to specific patterns of object motion. In the following subsections, we review some of the better characterized motion-sensitive RGCs. On–Off direction-selective retinal ganglion cells. On–Off DSGCs were first discovered by Barlow and coworkers in the early 1960s (Barlow & Hill 1963; reviewed in Vaney et al. 2012). As their name suggests, these cells respond to both light and dark edges moving along a particular 302 Dhande et al. VS01CH12-Huberman a ARI 27 October 2015 22:15 b Macaque Mouse Anti-Opn4 Anti-Opn4 200 μm 25 μm 20 mV 20 mV Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. 25 μm 50 μm On (10 s) c On (10 s) Mouse ipRGC scaled to primate Figure 5 Molecular markers label homogeneous populations of ganglion cells in primates and in the mouse. (a) Melanopsin-expressing ganglion cells in the macaque retina have large dendritic trees and sustained light responses. Example en face image and physiological trace adapted with permission from Dacey et al. (2005). Antibody staining based on unpublished data from the Huberman lab. (b) Melanopsin-expressing ganglion cells in the mouse have the same morphological and functional properties as those in the macaque retina, albeit on a much smaller scale. Example en face image adapted with permission from Berson et al. (2010). Antibody staining and physiological trace based on unpublished data from the Huberman lab. (c) Inset showing a mouse melanopsin ganglion cell in panel b drawn to scale relative to the macaque cell in panel a. Abbreviation: ipRGC, intrinsically photosensitive retinal ganglion cell. axis of the visual field [up, down, forward (nasal), or backward (temporal)], and they can respond to a fairly broad range of speeds (up to 60 deg/s) (Oyster et al. 1972, Weng et al. 2005, Sivyer et al. 2010). Moreover, their morphology predicts their functionality: They are bistratified and extend dendrites in both the On and Off layers of the IPL. The On–Off DSGCs, and the cells and circuits that endow these cells with their direction selectivity (reviewed in Vaney et al. 2012), have been a major focus of visual neuroscience and, indeed, stand as one of its great success stories in defining exactly how neurons perform discrete computations. Accessory optic system–projecting direction-selective ganglion cells. A different type of DSGC comprises those that project to accessory optic targets in the brainstem to control imagestabilizing eye movements (reviewed in Simpson 1984; Yonehara et al. 2008, 2009; Dhande et al. 2013). These accessory optic system (AOS)-projecting DSGCs historically have been called OnDSGCs because they were believed to consist of three subtypes of monostratified cells that respond to light edges moving along one of three axes of the visual field: upward, forward, or backward. This term warrants revision, however, as we now know from studies in mice that AOS-projecting RGCs include the three subtypes of monostratified On-DSGCs, along with a bistratified forwardsensitive On–Off DSGC subtype (Dhande et al. 2013) and a bistratified downward-sensitive On–Off DSGC subtype (Kay et al. 2011). In addition, many On-DSGCs extend a fraction of www.annualreviews.org • RGC Contributions to Feature Processing 303 VS01CH12-Huberman ARI 27 October 2015 22:15 Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. their dendritic arbor into the Off sublamina, although this fraction is apparently not sufficient to generate a functional Off response in these cells (Dhande et al. 2013) (see, e.g., Hoxd10-RGCs in Figure 4). The key distinction between the AOS-projecting DSGCs and the canonical On–Off DSGCs described above is that all AOS-projecting DSGCs are tuned for slow speeds. This tuning makes intuitive sense given the behavioral requirements of this system: The AOS offsets the slow drift motion of the head and eyes and allows for image stabilization on the retina. (see Dhande et al. 2013 and movies at the following URL: http://www.hubermanlab.com/movies.html). Off-type direction-selective ganglion cells. A monostratified, purely Off-DSGC type, called a J-RGC, was discovered recently in the mouse (Kim et al. 2008). These J-RGCs are named for their expression of the cell adhesion molecule JamB and are unusual among DSGCs for two reasons: (a) They respond best to upward motion of dark stimuli, and (b) their directional tuning properties are determined by their fanlike dendritic shape and by their orientation with respect to the retinal axes (Kim et al. 2008, 2014), rather than by assymetric inhibition from starburst amacrine cells, which generates direction selectivity in the canonical On–Off DSGCs and the AOS-projecting RGCs (Briggman et al. 2011, Yonehara et al. 2011). Object motion–sensitive retinal ganglion cells. A remarkable type of RGC comprises the object motion–sensitive (OMS) cells. These cells do not respond to motion in a specific direction; instead, they respond to differential motion. Broadly defined, differential motion occurs when a pattern within the receptive field of a given cell moves differently from a pattern in its surround. These cells are the “bug detectors” described by Lettvin and coworkers (1959) in their classic treatise “What the Frog’s Eye Tells the Frog’s Brain.” The circuitry underlying these cells was later worked out in both salamander and rabbit retinas, although it is not clear whether OMS cells represent a pure RGC type or include several types and/or subtypes (Ölveczky et al. 2003). For example, it is possible that the LED cells described above are a subtype of OMS RGCs. Nevertheless, the ethological significance of these cells is apparent and underscores the extent to which RGCs can inform the brain about stimuli far richer than just spots and moving bars. Looming-sensitive retinal ganglion cells. Other monostratified, Off RGCs in the mouse retina have been reported to encode looming objects, particularly dark, expanding objects overhead (Münch et al. 2009). Interestingly, the functional properties and stratification patterns of these RGCs strongly resemble those of Off-transient alpha RGCs, which have also been described in the mouse (Pang et al. 2003, Huberman et al. 2008b). However, whether or not they are a different set of RGCs altogether remains unclear. Specialized Retinal Ganglion Cell Types of the Primate The primate retina possesses several specializations that mouse, rat, and rabbit retinas lack. First and foremost, the primate retina has a fovea, that is, an area devoted to high-acuity vision. Second, within the fovea, RGC types that are unique to the primate help generate the high-acuity vision subserved by this region. Finally, RGCs of the primate retina exhibit robust color opponency. In the following subsections, we discuss some of these primate-specific RGC types and, where possible, attempt to draw parallels with RGC types of other species. Midget retinal ganglion cells in primates. Midget RGCs are the most numerous type in the primate retina. These cells are monostratified, consist of two subtypes—On and Off—depending 304 Dhande et al. VS01CH12-Huberman ARI 27 October 2015 22:15 Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. upon where they stratify their dendrites, have very small receptive fields, and carry chromatic information to the brain. Although comparisons have been drawn between the midget RGCs of primates and the X (sometimes called beta) RGCs of the cat retina, the primary relationship between these two groups is that, relatively speaking, each type represents the highest-acuity system in its species. As such, the midget RGC may be a primate specialization designed for highacuity red–green color-opponent vision (Watanabe & Rodieck 1989, Dacey & Petersen 1992). Small bistratified retinal ganglion cells. These cells have been documented in various primate species (marmoset, macaque), in which they carry yellow–blue opponent signals to the koniocellular layers of the dorsal lateral geniculate nucleus (dLGN) (Field et al. 2007, Szmajda et al. 2008; reviewed in Hendry & Reid 2000). They have been hypothesized to be homologous to the general group of so-called W (also called gamma) cells in cats. Given that the W or gamma type classification encompasses many different RGC types, however, this comparison is not exact (Berson 2008). For a review of the morphologies of the various non-parasol, non-midget, non-yellow–blue color-opponent RGCs in monkeys, the reader is referred to an article by Dacey (2004). Parasol cells. Parasol cells are monostratified RGCs with medium-sized dendritic trees and comprise two subtypes—On and Off—that vary depending on where they stratify their dendrites in the IPL. Some researchers have considered alpha RGCs to be homologous to the parasol cells of the primate retina (Dacey & Petersen 1992, Crook et al. 2008). Others, however, contest this notion, suggesting instead that the primate upsilon cell is the alpha homolog, and the parasols are a unique type altogether (Petrusca et al. 2007). The hope is that by extending the molecular techniques being developed to classify mouse RGC types into the primate retina, and combining them with our expanding knowledge of primate RGC physiology, will help resolve the parasol versus upsilon/alpha debate. The above list of RGC types is not intended to be exhaustive; surely additional types and subtypes of RGCs are still yet to be found. In the next section, we review progress in understanding where in the brain the different types of information encoded by RGCs are routed, and the ways in which the brain targets of RGC information streams use the information received. CENTRAL PROJECTIONS OF RETINAL GANGLION CELLS The retinofugal pathway contains ∼46 different target structures, each of which receives direct input from RGCs (Morin & Studholme 2014). Some of these RGC targets are well known, such as the dLGN, which relays visual information to the cortex for conscious perception of visual scenes. Other targets, such as the amygdala and habenula, are less well known as being retinorecipient nuclei, and their roles in visual processing remain poorly understood (Hattar et al. 2006, Morin & Studholme 2014). In the following subsections, we review several of the major parallel optic pathways and targets whose functions are generally understood and for which detailed knowledge about the type and role of RGC input is starting to emerge (Figure 6). For the sake of simplicity, we divide the optic pathway into three general categories: (a) subcortical visual circuits that govern the whole-animal physiological state, such as the retinofugal pathways governing entrainment of circadian rhythms; (b) circuits that control visually driven reflexive behaviors, such as the pupillary light reflex (PLR) and eye movements for stabilizing images on the retina; and (c) circuits for encoding complex visual features in the superior colliculus (SC) and dLGN. www.annualreviews.org • RGC Contributions to Feature Processing 305 VS01CH12-Huberman ARI 27 October 2015 22:15 SC mdPPN Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. NOT/DTN OPN Habenula dLGN IGL vLGN MTNd Zona incerta SCN MTNv Amygdala On–Off DSGCs Fast-tuned BD TRHR Drd4 Slow-tuned On DSGCs Hoxd10 Hoxd10/Spig-1 ipRGCs (M1-5) Opn4/Opn4-Cre/Cdh3 M1 M2 M3 M4 Off DSGCs Object motion Jam-B W3 t-Off Alpha s-Off CB2 W7A M5 Figure 6 Schematic showing projection patterns of genetically labeled ganglion cell types to subcortical visual targets for all transgenic mouse lines in which one or a small number of ganglion cells are labeled. For simplicity, axonal projections of melanopsin-expressing ganglion cells have been shown as a single population. Abbreviations: dLGN, dorsal lateral geniculate nucleus; DSGCs, direction-selective ganglion cells; IGL, intergeniculate leaflet; ipRGCs, intrinsically photosensitive retinal ganglion cells; mdPPN, medial division of the posterior pretectal nucleus; MTNd, dorsal medial terminal nucleus; MTNv, ventral medial terminal nucleus; NOT/DTN, nucleus of the optic tract/dorsal terminal nucleus; OPN, olivary pretectal nucleus; SC, superior colliculus; SCN, suprachiasmatic nucleus; s-Off, Off-sustained; t-Off, Off-transient; vLGN, ventral lateral geniculate nucleus. 306 Dhande et al. VS01CH12-Huberman ARI 27 October 2015 22:15 Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. Setting the Brain’s Clock: A Retinofugal Circuit Governing Entrainment of Endogenous Circadian Rhythms The suprachiasmatic nucleus (SCN) is the master circadian clock in the brain; it synchronizes endogenous rhythms to a 24-hour cycle and is coupled to the ambient light–dark cycle via RGC inputs (reviewed in Welsh et al. 2010, Lucas et al. 2012, Morin 2013, Legates et al. 2014). The SCN is the first retinorecipient target in the optic pathway. It consists of two densely packed, bilateral nuclei situated above the optic chiasm, each of which receives direct synaptic input via the retinohypothalmic tract. Just over a decade ago, it was discovered that the axons of melanopsinexpressing ipRGCs form the predominant projection pathway to the SCN (Gooley et al. 2001; Berson et al. 2002; Hattar et al. 2002, 2006) (Figure 7a). In early studies, multiple research groups took advantage of the fact that so-called intrinsic photosensitivity is a direct consequence of melanopsin expression in these RGCs and therefore generated mice that (a) express a marker (LacZ) from the melanopsin locus and (b) lack the melanopsin gene. Their results revealed the direct projection from melanopsin-expressing RGCs to the SCN and provided direct evidence that melanopsin signaling is involved in the photoentrainment of circadian rhythms (Panda et al. 2002, Ruby et al. 2002, Hattar et al. 2003). Although these studies demonstrated an important role for melanopsin signaling in entrainment to light–dark cycles, circadian photoentrainment was not completely abolished in melanopsin knockout mice. This finding suggests that although RGCs that express melanopsin may be essential for circadian photoentrainment, the same RGCs may also influence photoentrainment via signals arising from the classic rod–cone pathway. The alternative hypothesis is that RGCs other than the melanopsin-expressing RGCs are also important for entrainment. Support for the idea that ipRGCs are the major drivers in this system came in the form of three studies published in 2008 (Göz et al. 2008, Güler et al. 2008, Hatori et al. 2008), each of which independently demonstrated that specifically killing melanopsin-expressing RGCs resulted in a complete loss of circadian photoentrainment. These articles remain landmark studies because they are the first to directly link a specific RGC subtype (defined by its morphology, electrophysiological properties, connections with the brain, and gene expression) to a clear and defined subset of light-driven behaviors. Although the importance of ipRGCs in entraining the master clock in the SCN is now well established, several important questions remain unresolved. For example: How does each of the different ipRGC subtypes contribute to photoentrainment? The SCN receives input predominantly from M1 and M2 ipRGCs, and they appear to innervate distinct subregions of the SCN (Abrahamson & Moore 2001, Baver et al. 2008, Ecker et al. 2010). However, the functional significance of this specific innervation pattern remains unclear. In terms of the identities of the SCN-projecting ipRGCs versus non-SCN-projecting ipRGCs, Chen et al. (2011) elegantly showed that the M1 ipRGCs that lack the transcription factor Brn3b target the SCN, whereas Brn3b+ M1 ipRGCs target other regions of the visual system (see the subsection titled “The Pupillary Light Reflex: A Circuit for Ensuring that Appropriate Light Levels Reach the Brain” and Figure 7). Thus, it is clear that even genetically unique populations within ipRGC subtypes have important roles in driving these behaviors. The retino–ventral thalamic pathway helps modulate nonphotic circadian function. The retino-SCN projection pathway is indispensable for resetting the phase of the circadian pacemaker. However, retinal pathways that indirectly drive the SCN have also been implicated in entraining circadian rhythms, mainly to nonphotic cues such as locomotion activity and feeding times, among others (Mrosovsky 1995, Harrington 1997, Verwey & Amir 2009, Morin 2013). www.annualreviews.org • RGC Contributions to Feature Processing 307 VS01CH12-Huberman ARI 27 October 2015 22:15 a M1 (Brn3b–) M3 M2 M4 M5 Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. SCN dLGN IGL vLGN Nonphotic entrainment Photoentrainment b Non-M1 ipRGCs (OPN core–projecting) M1 (Brn3b+) ipRGCs (OPN shell–projecting) M2 M3 Pupillary light reflex M5 M4 OPN Edinger– Westphal nucleus Ciliary ganglion Figure 7 Reflexive behaviors controlled by intrinsically photosensitive retinal ganglion cells (ipRGCs). (a) Circadian entrainment is intimately linked to ipRGCs. Direct retinal input to the suprachiasmatic nucleus (SCN) from melanopsin-expressing (M1 and M2) ipRGCs is required for photoentrainment of circadian rhythms to a 24-hour light–dark cycle. Signals to the SCN via the ventral lateral geniculate nucleus (vLGN) and intergeniculate nucleus (IGL), thalamic structures that receive direct inputs from a mix of ipRGC subtypes (M1–M5), are involved in mediating nonphotic entrainment of circadian rhythms (e.g., entrainment to the timing of wheel running episodes). (b) The olivary pretectal nucleus (OPN) is an obligatory structure in the pupillary light reflex (PLR) pathway (ipRGCs → OPN → Edinger–Westphal nucleus → ciliary ganglion → intrinsic eye muscles → pupil constriction). M1 ipRGCs expressing the transcription factor Brn3b project to the shell of the OPN and have been genetically demonstrated to be essential for driving the PLR. The contributions of other ipRGC subtypes that innervate the OPN core to the PLR remain to be determined. Abbreviation: dLGN, dorsal lateral geniculate nucleus. 308 Dhande et al. Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. VS01CH12-Huberman ARI 27 October 2015 22:15 These indirect routes from the eye to the SCN reside in the thalamus. The major retinorecipient targets of the visual thalamic complex comprise three nuclei: the dLGN, the ventral LGN (vLGN), and the intergeniculate leaflet (IGL). Although the dLGN has received the most experimental attention (see the subsection titled “The Dorsal Lateral Geniculate Nucleus: The Primary Subcortical Target Feeding Visual Cortex”), it is noteworthy that in the mouse, the vLGN is comparable in size to the dLGN, and it also receives inputs from multiple functionally distinct RGC subtypes, including ipRGCs (Figures 6 and 7a). The IGL, also a fairly large target spanning millimeters of the thalamus, receives retinal inputs from ipRGCs (Hattar et al. 2006, Ecker et al. 2010) (Figures 6 and 7a). A subset of neurons in the IGL/vLGN send direct projections to the SCN, forming the geniculohypothalamic tract (Harrington 1997, Morin 2013) (Figure 7a). What is the function of IGL and vLGN inputs to the SCN? Brief periods of light exposure set at different times of day (a skeleton photoperiod) are sufficient for photoentrainment of the circadian cycle in nocturnal animals. Edelstein & Amir (1999) demonstrated that rats in which the IGL/vLGN was ablated failed to entrain to a skeleton photoperiod. The geniculohypothalamic tract also appears to be part of the neural circuit that allows nonvisual cues to associate with light–dark cycles in the natural environment. For example, Amir & Stewart (1996) demonstrated that circadian rhythms could be reset by a nonphotic stimulus (20-minute exposure to air puffs) in rats that had been trained to associate the air stimulus with a light cue that would normally reset their circadian clocks. They implicated the IGL/vLGN as a potential location for integration of this conditioned photic and nonphotic cue pairing on the basis of upregulation of the neural activity marker Fos in these thalamic targets. More recently, Lumsden and colleagues (Delogu et al. 2012) showed that disrupting the organization of postsynaptic cells in the IGL and vLGN during development results in abnormal circadian entrainment by photic and nonphotic cues. Interestingly, these mice also exhibited impaired negative masking (i.e., motor activity suppression by light exposure during the active dark phase of a nocturnal animal). Therefore the IGL and vLGN may represent part of a larger sensory integration network linking visual and nonvisual pathways for regulating endogenous biological rhythms and behaviors. The Pupillary Light Reflex: A Circuit for Ensuring that Appropriate Light Levels Reach the Brain The ambient luminance can vary drastically as we move from one location, such as a dimly lit room, to another, such as a sunny street. Adjusting the aperture of the pupil, and thereby adjusting the amount of light entering the eye, aids in optimizing image quality over such a wide range of light levels (Campbell & Gregory 1960). The visual system is equipped with customized machinery, built in large part from specific RGCs and their connections with the brain, that can change the aperture of the pupil in response to changes in luminance levels. This adjustment involves the subconscious, reflexive response of constricting or dilating the pupil, termed the pupillary light reflex (PLR). Initially, only conventional rod- and cone-dependent pathways were thought to drive the PLR. In 2001, however, this reflex was shown to persist in rodless, coneless mice (Lucas et al. 2001), and, subsequently, the role of melanopsin signaling in driving the PLR was revealed (Panda et al. 2003, Lucas et al. 2003). We now recognize that although rods and cones do contribute to the PLR, direct activation of ipRGCs also plays a major role in this behavior. Indeed, as we discuss below, the availability of molecular tools for marking and manipulating ipRGCs has led to significant advances in understanding their contributions to this important reflex. The olivary pretectal nucleus (OPN)—a midbrain retinorecipient target—is a critical hub in the central PLR pathway (Trejo & Cicerone 1984; Young & Lund 1994; Gamlin & Clarke 1995; Gamlin et al. 1995; Distler & Hoffmann 1989a,b; Gamlin 2006). For many decades, it www.annualreviews.org • RGC Contributions to Feature Processing 309 ARI 27 October 2015 22:15 was unclear which RGCs specifically drive the PLR, largely because the location of the OPN— nested between the dLGN and the SC and surrounded by many other small pretectal nuclei— made selectively targeting the OPN for retrograde tracer injections difficult. Genetic tagging of melanopsin-expressing RGCs with markers that also labeled their axons (Hattar et al. 2002, 2006; Ecker et al. 2010) revealed the crucial link between ipRGCs and the OPN and set the stage for determining the role played by these neurons as the sole driver of the PLR. The peripheral region, or shell, of the OPN receives inputs mainly from M1 ipRGCs, whereas the core region is innervated by other ipRGCs, such as M2s (Hattar et al. 2002, 2006; Baver et al. 2008; Ecker et al. 2010) (Figure 7b). Targeted removal of ipRGCs as a general group, which can be accomplished by toxin-based killing of these cells (Göz et al. 2008, Güler et al. 2008, Hatori et al. 2008) or by altering the expression of transcription factors that drive their development (Mao et al. 2014, Sweeney et al. 2014), abolishes the PLR. Deleting the transcription factor Eomes (also called Tbr2) leads to the loss of all ipRGCs, resulting in a loss of retinal projections to the OPN and a concomitant loss of light-induced pupillary constriction (Mao et al. 2014, Sweeney et al. 2014). What role do the different OPN-projecting ipRGC subtypes play in driving the PLR? As mentioned above, M1 ipRGC terminations are restricted to the OPN shell. Interestingly, neurons in the OPN shell are strongly activated by changes in light levels and project to the Edinger– Westphal nucleus (Prichard et al. 2002, Baver et al. 2008). Neurons in the Edinger–Westphal nucleus in turn send projections to the ciliary ganglion, which activates the intrinsic eye muscles, eventually leading to pupil constriction (Figure 7b). In a critical test of which RGCs drive the PLR, Hattar and colleagues (Chen et al. 2011) showed that the M1 ipRGCs that feed the shell of the OPN preferentially express the transcription factor Brn3b. They specifically killed these OPN inputs by crossing mice that conditionally express a toxin from the Brn3b locus with mice that express Cre recombinase from the melanopsin locus. This intersectional genetic ablation of OPN-projecting M1 ipRGCs resulted in an abrogated PLR (Chen et al. 2011). The contributions to the PLR of other non-M1 ipRGCs remain to be determined, as do those of the OPN core. Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. VS01CH12-Huberman A Reflex-Generating Circuit for Stabilizing Visual Scenes on the Retina Head and eye movements cause images to slip on the retina. Left unchecked, such retinal slip would lead to blurry perceptual representations of visual scenes. Thus, the eyes and brain generate compensatory head and eye movements in order to null this movement-generated effect. These gazestabilizing mechanisms are highly conserved and are found in all jawed species including frogs, fish, birds, and mammals (reviewed in Masseck & Hoffmann 2009, Distler & Hoffmann 2011). Two oculomotor reflexes, the optokinetic reflex (OKR) and the vestibuloocular reflex (VOR), collaborate to generate the slip-compensating eye and head movements necessary for gaze stabilization. The VOR can drive near perfect image-stabilizing eye movements in response to fast head movements because the biomechanics of the semicircular canals are ideal for detecting rapid head rotations. When the head moves at velocities too slow to engage the vestibular system, however, the OKR compensates for retinal slip. Thus, the visual and vestibular systems work in concert over a wide range of movement velocities to elicit the compensatory eye movements necessary to support accurate encoding of the visual scene (reviewed in Simpson 1984). What are the visual circuit components that drive the optokinetic reflex? The AOS, a group of functionally specialized RGCs and the brainstem targets to which they connect, has been implicated in generating the reflexive eye movements required for gaze stabilization in multiple species, including in mice (Yonehara et al. 2008, 2009; Dhande et al. 2013; reviewed in Simpson 1984, Giolli et al. 2006, Distler & Hoffmann 2011). Work done mainly in rabbits implicated On DSGCs as the main RGC subtype involved in generating gaze-stabilizing eye movements (Simpson 310 Dhande et al. ARI 27 October 2015 22:15 1984). These cells respond best to slow-moving visual stimuli, and their directional preferences closely match the axes along which retinal slip is generated during head rotations. In addition, retrograde labeling and electrophysiological recordings from the different terminal nuclei of the AOS, namely, the nucleus of the optic tract (NOT) and the dorsal, medial, and lateral terminal nuclei (DTN, MTN, and LTN, respectively), suggested that On DSGCs do indeed constitute the primary RGC inputs (Simpson 1984, Distler & Hoffmann 2011). Thus, a model emerged that the On DSGCs that encode forward motion and feed the NOT and DTN drive horizontal reflexive eye movements, and those that encode upward and downward motion and innervate the MTN drive vertical eye movements (Simpson 1984, Distler & Hoffmann 2011) (Figure 8). a On–Off DSGC (forward) On DSGC (forward) NOT–DTN On DSGC (downward) On DSGC (upward) MTNd MTNv b 20 Angle (º) Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. VS01CH12-Huberman 0 –20 0 5 10 15 Time (s) Figure 8 The accessory optic system (AOS) generates reflexive eye movements necessary for stable visual perception. (a) Schematic showing genetically identified direction-selective retinal ganglion cells (DSGCs) that innervate brainstem nuclei of the AOS, namely the nucleus of the optic tract (NOT), and the dorsal and medial terminal nuclei (DTN and MTN, respectively). The NOT–DTN receives synaptic inputs from both On and On–Off DSGCs encoding forward motion and drives horizontal eye movements. The dorsal and ventral MTN (MTNd and MTNv, respectively) receive inputs from On DSGCs encoding upward and downward motion, respectively, and are involved in generating vertical reflexive eye movements. For a detailed schematic of the fiber tracts that form the AOS, the reader is referred to Pak et al. (1987) and Dhande et al. (2013). (b) Gaze-stabilizing eye movements can be quantitatively measured in head-fixed, awake mice. Sample image (left) captured from a CCD camera used to record optokinetic nystagmus in response to horizontal or vertical moving grating stimuli (middle) and example eye movement trace (right). Figure adapted with permission from Dhande et al. (2013). www.annualreviews.org • RGC Contributions to Feature Processing 311 ARI 27 October 2015 22:15 Recent advances in viral, genetic, and behavioral tools reveal several striking similarities between the mouse and rabbit AOS. These advances are uncovering new facets of the functional organization of the mouse accessory optic system. Studies using mice with genetically marked RGCs have confirmed that On DSGCs also constitute the major retinal inputs to the mouse AOS (Yonehara et al. 2008, Dhande et al. 2013). Moreover, Yonehara and colleagues (2009) showed that neurons in the dorsal MTN of the mouse are strongly activated by upward motion and that neurons in the ventral MTN respond preferentially to downward-moving visual stimuli. This finding suggests that within the MTN, inputs from On DSGCs encoding upward motion are compartmentalized separately from those encoding downward motion (Figure 8). Two new studies (Kay et al. 2011, Dhande et al. 2013) discovered that the AOS also receives an appreciable number of inputs from On–Off DSGCs. The MTN receives inputs from On–Off DSGCs encoding downward motion, consistent with the role the MTN plays in driving vertical compensatory eye movements (Kay et al. 2011). In contrast, the NOT receives inputs from a population of On–Off DSGCs that are tuned to forward motion and prefer relatively slow-moving visual stimuli, making these neurons ideal for participating in the role played by the NOT in generating horizontal eye movements (Dhande et al. 2013) (Figure 8). Curiously, On–Off DSGCs that prefer downward motion may also innervate the NOT (Kay et al. 2011). Collectively, analyses of retinal projections to various AOS nuclei have begun to reveal the unexpected complexity of the inputs received by these nuclei and presumably used to drive reflexive eye movements. To date, researchers have not clearly established a causal relationship among specific AOSprojecting RGC types and image stabilization behavior. One study eliminated DS tuning of all direction-selective retinal signals by killing the presynaptic inputs to DSGCs, namely, the cholinergic starburst amacrine cells. Interestingly, killing the starburst amacrine cells also eliminated image-stabilizing eye movements (Yoshida et al. 2001). Silencing the On bipolar pathway has a similar effect on this set of behaviors (Sugita et al. 2013). Together, these findings bolster the correlation between On DSGCs and image-stabilizing eye movements. Both manipulations are somewhat indirect, however, because they also alter retinal signals to the SC and the cortex, both of which may provide feedback signals to the AOS network (reviewed in Gamlin 2006). Thus, the contribution of specific RGC subtypes or AOS target nuclei in generating reflexive, slip-compensating eye movements still remains unresolved. Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. VS01CH12-Huberman Integration of Diverse Visual Features in the Midbrain The SC, homologous to the optic tectum in birds and cold-blooded vertebrates, is a large, multisensory structure in the midbrain and has a topographic map of visual space in its superficial layers. Other sensory maps, such as somatosensory and auditory maps, reside below the visual input layers. The SC harbors circuitry that is crucial for both convergence of multiple optic pathways and for integration of afferent signals from multiple sensory modalities (reviewed in May 2006, Dhande & Huberman 2014a). A key organizational feature of the SC is that all of its sensory maps are integrated along the depth of the SC layers such that a sound occurring in a particular location in visual space will generate a head or eye movement to that location (reviewed in Knudsen et al. 1987, Sparks & Nelson 1987, Stein et al. 2014). The SC is also a key structure for initiating motor commands for head, body, and/or voluntary eye movements (e.g., saccades) with the goal of gaze orientation (reviewed in Sparks 2002, Gandhi & Katnani 2011). In addition, considerable work in primates has focused on the role of the SC in smooth pursuit and attention (reviewed in Wurtz & Albano 1980, Mysore & Knudsen 2011, Krauzlis et al. 2013). The SC (optic tectum) has also long been implicated in mediating innate defensive behaviors such as looming-evoked escape and freezing in multiple species (Ingle 1973, Sahibzada et al. 1986, Dean et al. 1988, Ellard & Goodale 312 Dhande et al. VS01CH12-Huberman a ARI 27 October 2015 22:15 b 300 Distance to nest (pixels) Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. 2° 20° 200 100 Flight 0 Nest 0 0.2 0.4 0.6 Time (s) Figure 9 An overhead threat elicits an innate flight response in mice. (a) Schematic showing how an expanding looming stimulus triggers an escape flight to the home nest in mice. (b) Example trace [adapted from Yilmaz & Meister (2013)] showing the flight trace (red ) to the nest ( green) evoked by a disc stimulus (black circles) expanding from 2 to 20◦ in diameter. The superior colliculus is thought to mediate this innate aversive behavior. 1988, Northmore et al. 1988, Westby et al. 1990; reviewed in Dean et al. 1989). As a result of the recent advances in genetic tools to visualize and manipulate neural circuits, along with the design of more sophisticated behavioral analyses, the dissection of SC-mediated behaviors is once again gaining attention (Yilmaz & Meister 2013, Liang et al. 2015, Shang et al. 2015, Wei et al. 2015) (Figure 9). The SC is one of the most distal retinorecipient targets. It is located on the dorsal surface of the midbrain and is partially covered by the visual cortex. RGCs from the contralateral eye terminate in the superficial layers of the SC to generate a continuous map of visual space. Axons from the ipsilateral eye terminate in the slightly deeper layers of the SC and arise from RGCs residing in the ventral retina, thereby establishing eye-specific visuotopic maps (reviewed in Assali et al. 2014). For many decades, researchers have taken advantage of the clear structure–function attributes of the SC, such as the precise organization of retinocollicular afferents, in order to gain insight into the molecular and activity-dependent mechanisms that sculpt topographic maps during development (reviewed in McLaughlin & O’leary 2005, Cang & Feldheim 2013, Kirkby et al. 2013, Ackman & Crair 2014, Triplett 2014). Recently, the advent of transgenic tools for labeling specific RGC subtypes and in vivo imaging techniques have allowed researchers to make significant gains in understanding retinocollicular circuit organization and function in the mouse. It is estimated that more than 90% of RGCs project to the SC in rodents (Linden & Perry 1983, Hofbauer & Dräger 1985). This massive input raises some key questions, such as the following: Are all the visual features relayed by the RGCs maintained as parallel maps within the SC? It has long been known that different RGCs terminate at different depths in the SC of the cat and that of the hamster (Bowling & Michael 1980, Berson 1988, Mooney & Rhoades 1990, Tamamaki et al. 1995), and retrograde labeling experiments hinted that a similar overall organization might also exist in the mouse (Hofbauer & Dräger 1985). Genetic labeling of RGCs that encode distinct visual features (for example, different axes of motion, local object motion, or luminance) has now provided detailed information on this issue by confirming that different RGC subtypes target distinct laminar depths www.annualreviews.org • RGC Contributions to Feature Processing 313 ARI 27 October 2015 22:15 in the retinorecipient SC, subdividing the SC into multiple distinct sublaminar maps (reviewed in Dhande & Huberman 2014a). RGC axons that respond to local object motion selectively terminate in the upper half of the retinorecipient SC (Kim et al. 2010). On–Off and Off DSGCs also terminate in this domain of the SC (Kim et al. 2008, 2010; Huberman et al. 2009; Kay et al. 2011; Rivlin-Etzion et al. 2011). In contrast, ipRGCs and alpha RGCs terminate preferentially in the lower half of the retinorecipient SC (Hattar et al. 2006, Huberman et al. 2008b, Brown et al. 2010, Ecker et al. 2010). Clearly there are at least four distinct feature maps based on the organization of retinal inputs, raising the issue of whether these maps are inherited and maintained as distinct channels by collicular networks or are combined to generate more sophisticated feature maps (see also Dhande & Huberman 2014a). Several studies have documented the presence of a morphologically diverse collection of excitatory and inhibitory neurons in the SC of rats and of mice (Langer & Lund 1974, May 2006, Isa & Hall 2009, Gale & Murphy 2014). The four types of collicular neurons that have received the most focus are (a) wide-field neurons, which have large dendritic trees that extend at oblique angles to the SC surface; (b) horizontal cells, which are putative inhibitory projection neurons; (c) narrow-field cells, which have vertical, cylindrically organized dendrites; and (d ) stellate neurons, which have small, nonoriented dendritic fields (reviewed in May 2006). Gale & Murphy (2014) recently demonstrated that these four classes of collicular neurons can be separated not only based on morphometric parameters, but also on the basis of their visual response properties and distinct projection patterns to downstream targets. So the question now becomes: Do different SC neurons receive inputs from distinct RGC subtypes? Although the answer to this question has yet to be resolved conclusively, recent studies have made significant progress in extending our knowledge of the functional properties of different subtypes of collicular neurons (Gale & Murphy 2014), and, from that, inferences can now be made about the preservation and/or transformation of retinal signals arriving within the SC. Ursula Dräger and colleagues (Dräger 1975; Dräger & Hubel 1975a, 1975b, 1976) were pioneers in describing the functional cartography of sensory modalities in the mouse SC. Only within the past few years, however, have researchers again begun to systematically probe the visual properties of collicular neurons, this time using newly developed in vivo imaging and electrophysiological tools in conjunction with genetic and classical anatomical tracing methods. From this effort, it is becoming clear that neurons in the retinorecipient layers of the SC in mice display a wide range of visual response properties, such as direction-independent motion selectivity; looming sensitivity; orientation and direction selectivity; and tuning for contrast, speed, and size (Wang et al. 2009, 2010; Feinberg & Meister 2014; Gale & Murphy 2014; Ahmadlou & Heimel 2015). Key issues to resolve in the coming years will be (a) the precise architecture of the local SC circuits by which these properties arise (e.g., the integration of specific retinal afferents onto specific SC target cells) and (b) how each component of these circuits contributes to specific behaviors such as looming-evoked escape. We predict that genetic drivers for specific RGC and SC neuron types will play a central role in assisting efforts to resolve these issues. Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. VS01CH12-Huberman What is the circuit diagram for generating feature selectivity in the superior colliculus? Neurons in layer 5 of the visual cortex send topographically organized, direct glutamatergic inputs to the SC. The mouse visual cortex also encodes visual features similar to those encoded in the SC, such as looming, direction, and orientation, among others. (Wang et al. 2010, Feinberg & Meister 2014, Zhao et al. 2014a, Ahmadlou & Heimel 2015). Ablation or silencing of corticocollicular inputs showed that this pathway modulates certain receptive field and tuning properties of SC neurons, such as their gain, but these manipulations of the corticocollicular pathway do not appear to result in a complete loss of feature selectivity in SC neurons (Wang et al. 2010, Feinberg & 314 Dhande et al. Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. VS01CH12-Huberman ARI 27 October 2015 22:15 Meister 2014, Zhao et al. 2014a, Ahmadlou & Heimel 2015). These results contrast with those in species such as cats and hamsters, in which certain response features of collicular neurons, such as direction selectivity, critically depend on cortical inputs to the SC (Wickelgren & Sterling 1969, Rosenquist & Palmer 1971, Chalupa & Rhoades 1977). Although current studies suggest that the visual features encoded by SC neurons in mice arise largely from cortex-independent mechanisms, more explicit experiments to test whether or not cortical inputs can function as drivers that impart feature selectivity to collicular neurons in mice remain to be performed. Another possibility is that properties such as orientation and direction selectivity arise from local and interlaminar circuits in the SC, as is the case for some tectal neurons in zebrafish (reviewed in Gebhardt et al. 2013, Dhande & Huberman 2014a). Finally, the parsimonious explanation for the diversity of visual features encoded by collicular neurons is that these properties are directly inherited from the many different types of RGCs that project to the SC. As we discussed above, different types of RGCs are capable of encoding these specialized visual features, and several of these RGC subtypes (direction-selective, speed-tuned, motion-selective) send projections to the SC. Recent advances in combining synaptic tracing techniques with in vivo two-photon imaging (Rancz et al. 2011, Yonehara et al. 2013, Wertz et al. 2015) have made this hypothesis readily testable. The Dorsal Lateral Geniculate Nucleus: The Primary Subcortical Target Feeding Visual Cortex The dLGN is a prominent thalamic relay station that, together with the SC, forms the majority of the optic pathway in all mammals. The dLGN has received much attention from vision researchers working in a variety of species because it is the gateway for visual information to reach the primary visual cortex (V1) and thus the starting point of higher-order visual representations and perception (see Usrey & Alitto 2015). The dLGN receives inputs from several different RGC subtypes, and all of these inputs appear to be organized in a retinotopic manner (Pfeiffenberger et al. 2006, Piscopo et al. 2013). In addition, binocular inputs are segregated on the basis of the eye of origin within the dLGN (Rakic 1976, Jaubert-Miazza et al. 2005), a feature that has served as a valuable model for understanding how both spontaneous activity during development and visual experience shape neural circuits in a wide variety of species (reviewed in Katz & Shatz 1996, Huberman et al. 2008a, Hong & Chen 2011, Kirkby et al. 2013). The image of the six-layered primate dLGN is iconic in neuroscience. Rodents, however, lack overt cytoarchitectural lamination in their dLGN (Reese 1988). The recent discovery of genetic tools that label the axons of functionally defined RGC subtypes confirmed that RGCs projecting to the mouse dLGN are organized into discrete visual channels that have a laminar-specific overall organization (Kim et al. 2008, 2010; Huberman et al. 2008b, 2009; Ecker et al. 2010; Kay et al. 2011; Rivlin-Etzion et al. 2011; reviewed in Dhande & Huberman, 2014a). Broadly speaking, the mouse dLGN can be divided into a core region and a shell region, each of which contains a complete retinotopic map and receives input from functionally distinct populations of RGCs (Figure 10). The core compartment of the dLGN receives input from alpha-like RGCs, which are not directionally tuned and which respond best to center–surround stimuli (Huberman et al. 2008b), as well as from ipRGCs (Brown et al. 2010, Ecker et al. 2010). In contrast, the shell of the dLGN, receives input from multiple direction-selective RGC subtypes, each of which is tuned to one of the four cardinal directions (Kim et al. 2008, 2010; Huberman et al. 2009; Kay et al. 2011; Rivlin-Etzion et al. 2011). In recent studies of the postsynaptic organization of dLGN neurons, Guido and colleagues (Krahe et al. 2011, Seabrook et al. 2013, Bickford et al. 2015) demonstrated that the mouse dLGN www.annualreviews.org • RGC Contributions to Feature Processing 315 ARI 27 October 2015 22:15 contains at least four distinct types of neurons. Morphologically, these neurons resemble the X, Y, and W relay neurons described in cats and primates, as well as the local inhibitory neurons. Interestingly, some dLGN neuron subtypes exhibit shell versus core distribution biases. For example, W-like cells are found predominantly in the shell (where DSGC inputs terminate), whereas Y-like cells reside in the core. X-like cells and local inhibitory neurons are found dispersed across both the shell and the core. Neurons in the mouse dLGN respond to a variety of visual features, such as direction of motion, stimulus orientation, and contrast modulation (Marshel et al. 2012, Piscopo et al. 2013), but how the visual properties of X, Y, W, and inhibitory cells are determined by their specific RGC inputs remains unclear (Figure 10). RGC receptive field properties no doubt influence the shape of dLGN cell responses, but the convergence of RGC projections (1–6 RGCs synapse onto each dLGN neuron) (Chen & Regehr 2000, Jaubert-Miazza et al. 2005, AcunaGoycolea et al. 2008, Seabrook et al. 2013) offers ample opportunity for both combining different visual feature properties (e.g., opposing directionally tuned inputs) onto individual dLGN neurons and, by virtue of connections with inhibitory dLGN cells, possibly sharpening those features (see, e.g., Levick et al. 1969). Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. VS01CH12-Huberman Challenging the textbook model of cortical orientation and direction selectivity? Since the discovery of orientation-selective neurons in the cat and primate visual cortex by Hubel & Wiesel in the 1960s (Hubel & Wiesel 1962, 1968), along with the subsequent discovery of orientation columns and pinwheel organization in the cortex (Blasdel & Salama 1986, Ts’o et al. 1990, Bonhoeffer & Grinvald 1991), orientation selectivity has been a major focus of research in visual neuroscience. Traditionally, studies addressing the emergence and underlying mechanisms of orientation selectively have used rabbits, cats, ferrets, or primates as their model of choice. However, all of the recent advances in genetic and optical imaging tools for recording neural activity in rodents have made mouse V1 a premier model system for studies exploring the organization and mechanistic basis of orientation selectivity. Mouse V1 is rich with neurons that respond strongly to bars of light of a particular orientation (Dräger 1975, Mangini & Pearlman 1980, Métin et al. 1988, Niell & Stryker 2008), but unlike the columnar and pinwheel organization of orientationselective cortical neurons described in higher mammals, orientation-selective neurons in mouse (and rat) V1 are organized in a salt-and-pepper fashion (Girman et al. 1999, Ohki et al. 2005, Sohya et al. 2007, Kerlin et al. 2010, Bonin et al. 2011). The long-standing model for cortical orientation selectivity is that untuned center–surround inputs from thalamic relay cells are summed by each cortical neuron to generate elongated cortical receptive fields that are strongly activated by oriented bars of light. Although this method may be the dominant one for generating orientation-selective units in the cortex, recent studies confirm that orientation-selective (axial) and direction-selective neurons are certainly present in the dLGN of both rodents and primates. Marshel et al. (2012) studied populations of neurons residing in the dorsal tip of the dLGN and, using in vivo two-photon calcium imaging, identified dLGN neurons that were tuned for motion in the anterior and posterior directions, as well as neurons that were tuned to axial motion along the horizontal (anterior–posterior) axis. Subsequent studies used recording electrodes that allow for a more thorough investigation of the entire dLGN to establish the existence of mouse dLGN neurons that encode directional and axial motion along both the vertical (upward–downward) and horizontal axes (Piscopo et al. 2013, Scholl et al. 2013, Zhao et al. 2013). Recently Cruz-Martı́n and colleagues (2014) discovered that direction- and orientation-tuned inputs from the dLGN do indeed feed superficial layers 1 and 2 of the mouse visual cortex (Figure 10). Furthermore, Cheong et al. (2013) found that in marmosets, a small portion of neurons in the koniocellular layers of the dLGN are strongly orientation selective (on par with V1 units), and, recently, Ling et al. (2015) discovered orientation-selective responses in 316 Dhande et al. VS01CH12-Huberman ARI 27 October 2015 22:15 Cortical neuron properties V1 1 2 Directionselective Orientationselective Loomingselective Suppressedby-contrast • Métin et al. 1988 • Niell & Stryker 2008 • Dräger 1975 • Mangini & Pearlman 1980 • Métin et al. 1988 • Niell & Stryker 2008 • Zhao et al. 2014 • Niell & Stryker 2010 3 4 Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. 5 6 dLGN relay neuron properties ~15,000 axons dLGN Directionselective Orientationselective Non-directionselective ~50,000 axons Ganglion cell properties Retina Direction-selective On–Off DSGCs Non-direction-selective t-Off alpha Figure 10 Visual features encoded by the retinogeniculocortical pathway in mice. Cells at each level of this pathway have been shown to possess a variety of functional properties, some of which overlap. For example, direction-selective ganglion cells (DSGCs) send signals to cells in the dorsal lateral geniculate nucleus (dLGN) that, in turn, exhibit direction-selective signals in their axon terminals in the visual cortex (V1). In contrast, other properties (e.g., looming selectivity) have so far only been described in cells in V1. It is not yet clear whether these cells inherit these properties from the geniculocortical projecting cells that innervate them or from complex local circuits within V1. Given the high level of convergence within the retinogeniculocortical pathway, however, properties unique to V1 neurons likely arise from a combination of the properties of the geniculocortical projecting neurons that innervate them, as well as from modification resulting from local circuits within the cortex. the human LGN using fMRI. Therefore, in light of the growing body of work demonstrating the presence of orientation-selective responses in the LGN, the textbook framework for cortical orientation selectivity may need revision. What is the source of orientation selectivity in the dLGN? One possibility is that corticogeniculate projections impose this feature on relay neurons. This possibility is unlikely, however, because cortical inactivation does not impair the orientation selectivity of geniculate neurons (Scholl et al. 2013, Zhao et al. 2013). Another possibility is that combining DSGC inputs with opposing direction preferences generates orientation-selective units in the dLGN. During their survey of the www.annualreviews.org • RGC Contributions to Feature Processing 317 ARI 27 October 2015 22:15 visual response properties of dLGN neurons, Piscopo et al. (2013) found an enriched presence of direction- and orientation-selective units in the dorsolateral shell of the LGN, the same region in which different subtypes of DSGCs are known to terminate (Huberman et al. 2009, Kim et al. 2010, Rivlin-Etzion et al. 2011). In addition, Cruz-Martı́n et al. (2014) used transsynaptic rabies tracing techniques to show that relay neurons in the shell do in fact receive synaptic connections from DSGC axon terminals (Figure 10). Collectively, this work suggests that direction-selective responses in the dLGN are a direct consequence of their presynaptic partners, DSGCs. A third possibility is that orientation selectivity in the dLGN arises from inputs from the SC onto the dLGN. Several collicular neurons are orientation selective (see below), and, in rodents, collicular projections are known to terminate exclusively in the shell region of the dLGN (Reese 1984, Harting et al. 1991, Grubb & Thompson 2004, Bickford et al. 2015). Finally, orientation-selective cells in the thalamus could also arise as a direct consequence of orientation selectivity in the retina. Orientation-selective ganglion cells are found in rabbits and cats, and they are likely present in primates as well (Levick & Thibos 1982, Bloomfield 1994, Passaglia et al. 2002). Zhao et al. (2014a) used retinal multielectrode array recordings to show that orientation-selective cells are also present in the mouse retina. However, targeted recordings of orientation-selective ganglion cells have not yet been performed, nor have their dendritic morphology and projections to central visual targets been revealed. Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. VS01CH12-Huberman Isoorientation domains in the superior colliculus: a bit of primate V1 in the mouse superior colliculus. A significant fraction of neurons in the superficial retinorecipient layers of the SC display strong responses to oriented visual stimuli (Wang et al. 2009, 2010; Feinberg & Meister 2014; Gale & Murphy 2014). Feinberg & Meister (2014) recently developed a surgical preparation that allows for visualization of the SC without compromising the overlying cortex. By exposing the SC in this manner, they were able to map the population activity of collicular neurons using genetically encoded calcium sensors (GCaMP6) in awake, head-fixed mice. In doing so, they made a remarkable discovery: Orientation-selective neurons in the mouse SC have a columnar organization similar to that in cats and primates. Ahmadlou & Heimel (2015) used a combination of extracellular electrode recordings and in vivo calcium imaging techniques, and they not only came to the same conclusion but also found that these orientation maps are organized in a concentric manner. Interestingly, both their study and the one by Feinberg & Meister found that not all orientations are represented at every retinotopic location. Collicular orientation selectivity in mice appears to be independent of cortical inputs (Wang et al. 2010, Feinberg & Meister 2014, Ahmadlou & Heimel 2015). Thus, orientation selectivity in the SC is likely to be generated by combining inputs from direction-selective or orientation-selective RGC afferents and/or is an emergent property of intracollicular circuits. As mentioned above, the classic model for orientation selectivity posits that selectivity for stimulus orientation in areas such as the visual cortex arises from the specific organization pattern of non-orientation-tuned thalamic inputs. This model may not hold true for some species, however, given that orientation selectivity is now reported at every stage of the retinogeniculocortical pathway and that orientation columns have been discovered in the mouse SC. It will be exciting to empirically test how the selectivity present in higher visual areas for stimulus features such as orientation and direction relates to selectivity for the same features encoded by the retina. ipRGCs: new players in the retinogeniculocortical pathway. An emerging theme in vision science is the contribution of ipRGCs to pattern vision. The recent development of transgenic mice with genetically tagged ipRGCs revealed that the axons of ipRGCs project to the core region of the dLGN (Hattar et al. 2006, Brown et al. 2010, Ecker et al. 2010). Brown and colleagues 318 Dhande et al. Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. VS01CH12-Huberman ARI 27 October 2015 22:15 (2010) assessed the visual response properties of dLGN and cortical neurons in wild-type mice and in mice lacking either melanopsin or rod and/or cone photoreceptors. Surprisingly, they found that melanopsin-mediated light responses can drive nearly 40% of LGN neurons and can thereby also drive neurons in V1 (Brown et al. 2010). Moreover, in a more recent study, they showed that melanopsin signaling can influence the spatial and temporal frequency tuning of neurons in the LGN and can impact the speed preference of LGN neurons in response to moving gratings (Allen et al. 2014). How does the ipRGC pathway influence pattern vision and behavior? Hattar and colleagues demonstrated that mice with altered cone and rod function but intact melanopsin signaling could still discriminate high-contrast spatial patterns, although the altered animals had to perform more perceptual trials than wild-type animals (Ecker et al. 2010). Studies have also shown that some On alpha-like RGCs express melanopsin (M4 ipRGCs), are intrinsically photosensitive, and project to the core compartment of the dLGN (Estevez et al. 2012) and, very recently, melanopsin signaling and M4 ipRGCs in particular were implicated as being essential for normal contrast detection (Schmidt et al. 2014). We note that the presence of ipRGC input to the dLGN is not specific to rodents; ipRGCs have also been shown to project to the dLGN in macaque monkeys (Dacey et al. 2005). Melanopsinexpressing RGCs are also present in the human retina (Hannibal et al. 2004, Dacey et al. 2005), but their central projection patterns have not yet been charted. Nevertheless, these studies collectively build a strong case that ipRGCs are not just part of primordial visual pathways that entrain circadian rhythms and drive the PLR. Rather, they appear to play several critical roles in supporting more traditionally studied aspects of vision, including visual perception itself. DISCUSSION In this article, we have attempted to convey the rich set of signals that RGCs encode, along with the brain circuits that unpack and use that information. The field of visual neuroscience has clearly entered a particularly exciting phase in which classic techniques such as dye labeling and single unit electrophysiology are being joined by modern genetic tools for marking, monitoring, and manipulating the activity of specific retinal output pathways. These tools in turn enable scientists to define the precise contributions of these pathways to vision. Given that the past decade has brought genetic identification of ∼75% of all mouse RGC subtypes, the next major effort will be to apply the same logic to identification of specific types of visual neurons within the brain. One can also imagine that in the next decade, there will be a significant expansion in the effort to bring molecular tools to the study and control of retinal and brain neurons in nonhuman primates and in humans. The motivation for this is to understand how we process visual information and to improve quality of life by better diagnosing and treating diseases that compromise vision. DISCLOSURE STATEMENT The authors are not aware of any affiliations, memberships, funding, or financial holdings that might be perceived as affecting the objectivity of this review. ACKNOWLEDGMENTS We thank Drs. David Berson, Rana El-Danaf and Tania Seabrook for helpful comments on earlier versions of this manuscript. A.D.H. is supported by the National Institutes of Health and National Eye Institute (grants: RO1 EY022157 and U01 U01NS090562), the Glaucoma Research Foundation Catalyst for a Cure Initiative, the E. Matilda Ziegler Foundation, the Pew Charitable www.annualreviews.org • RGC Contributions to Feature Processing 319 VS01CH12-Huberman ARI 27 October 2015 22:15 Trusts, and the McKnight Foundation. O.S.D was supported in part by the Knights Templar Postdoctoral Fellowship and Grant. LITERATURE CITED Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. Abrahamson EE, Moore RY. 2001. Suprachiasmatic nucleus in the mouse: retinal innervation, intrinsic organization and efferent projections. Brain Res. 916(1–2):172–91 Ackman JB, Crair MC. 2014. Role of emergent neural activity in visual map development. Curr. Opin. Neurobiol. 24(1):166–75 Acuna-Goycolea C, Brenowitz SD, Regehr WG. 2008. Active dendritic conductances dynamically regulate GABA release from thalamic interneurons. Neuron 57(3):420–31 Ahmadlou M, Heimel JA. 2015. Preference for concentric orientations in the mouse superior colliculus. Nat. Commun. 6:6773 Allen AE, Storchi R, Martial FP, Petersen RS, Montemurro MA, et al. 2014. Melanopsin-driven light adaptation in mouse vision. Curr. Biol. 24(21):2481–90 Amir S, Stewart J. 1996. Resetting of the circadian clock by a conditioned stimulus. Nature 379(6565):542–45 Assali A, Gaspar P, Rebsam A. 2014. Activity dependent mechanisms of visual map formation—from retinal waves to molecular regulators. Semin. Cell Dev. Biol. 35:136–46 Azeredo da Silveira R, Roska B. 2011. Cell types, circuits, computation. Curr. Opin. Neurobiol. 21:664–71 Badea TC, Cahill H, Ecker J, Hattar S, Nathans J. 2009. Distinct roles of transcription factors Brn3a and Brn3b in controlling the development, morphology, and function of retinal ganglion cells. Neuron 61(6):852–64 Barlow HB, Fitzhugh R, Kuffler SW. 1957. Change of organization in the receptive fields of the cat’s retina during dark adaptation. J. Physiol. 137(3):327–37 Barlow HB, Hill RM. 1963. Selective sensitivity to direction of movement in ganglion cells of the rabbit retina. Science 139(3553):412–14 Baver SB, Pickard GE, Sollars PJ, Pickard GE. 2008. Two types of melanopsin retinal ganglion cell differentially innervate the hypothalamic suprachiasmatic nucleus and the olivary pretectal nucleus. Eur. J. Neurosci. 27:1763–70 Berson DM. 1988. Retinal and cortical inputs to cat superior colliculus: composition, convergence and laminar specificity. Prog. Brain Res. 75:17–26 Berson DM. 2008. Retinal ganglion cell types and their central projections. In The Senses: A Comprehensive Reference, Vol. 1, ed. AI Basbaum, A Kaneko, GM Shepherd, G Westheimer, pp. 491–520. San Diego: Academic Berson DM. 2013. Intrinsically photosensitive retinal ganglion cells. In The New Visual Neurosciences, ed. JS Werner, LM Chalupa, pp. 183–196. Cambridge, MA: MIT Press Berson DM, Castrucci AM, Provencio I. 2010. Morphology and mosaics of melanopsin-expressing retinal ganglion cell types in mice. J. Comp. Neurol. 518(13):2405–22 Berson DM, Dunn FA, Takao M. 2002. Phototransduction by retinal ganglion cells that set the circadian clock. Science 295(5557):1070–73 Bickford ME, Zhou N, Krahe TE, Govindaiah G, Guido W. 2015. Retinal and tectal “driver-like” inputs converge in the shell of the mouse dorsal lateral geniculate nucleus. J. Neurosci. 35:10523–34 Blasdel GG, Salama G. 1986. Voltage-sensitive dyes reveal a modular organization in monkey striate cortex. Nature 321(6070):579–85 Bleckert A, Schwartz GW, Turner MH, Rieke F, Wong ROL. 2014. Visual space is represented by nonmatching topographies of distinct mouse retinal ganglion cell types. Curr. Biol. 24(3):310–15 Bloomfield SA. 1994. Orientation-sensitive amacrine and ganglion cells in the rabbit retina. J. Neurophysiol. 71(5):1672–91 Bonhoeffer T, Grinvald A. 1991. Iso-orientation domains in cat visual cortex are arranged in pinwheel-like patterns. Nature 353(6343):429–31 Bonin V, Histed MH, Yurgenson S, Reid RC. 2011. Local diversity and fine-scale organization of receptive fields in mouse visual cortex. J. Neurosci. 31(50):18506–21 320 Dhande et al. Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. VS01CH12-Huberman ARI 27 October 2015 22:15 Bowling DB, Michael CR. 1980. Projection patterns of single physiologically characterized optic tract fibres in cat. Nature 286(5776):899–902 Boycott BB, Wässle H. 1974. The morphological types of ganglion cells of the domestic cat’s retina. J. Physiol. 240:397–419 Briggman KL, Helmstaedter M, Denk W. 2011. Wiring specificity in the direction-selectivity circuit of the retina. Nature 471(7337):183–88 Brown TM, Gias C, Hatori M, Keding SR, Semo M, et al. 2010. Melanopsin contributions to irradiance coding in the thalamo-cortical visual system. PLOS Biol. 8(12):e1000558 Campbell FW, Gregory AH. 1960. Effect of size of pupil on visual acuity. Nature 187:1121–23 Cang J, Feldheim DA. 2013. Developmental mechanisms of topographic map formation and alignment. Annu. Rev. Neurosci. 36:51–77 Chalupa LM, Rhoades RW. 1977. Responses of visual, somatosensory, and auditory neurones in the golden hamster’s superior colliculus. J. Physiol. 270(3):595–626 Chen C, Regehr WG. 2000. Developmental remodeling of the retinogeniculate synapse. Neuron 28(3):955–66 Chen S-K, Badea TC, Hattar S. 2011. Photoentrainment and pupillary light reflex are mediated by distinct populations of ipRGCs. Nature 476:92–95 Cheong SK, Tailby C, Solomon SG, Martin PR. 2013. Cortical-like receptive fields in the lateral geniculate nucleus of marmoset monkeys. J. Neurosci. 33(16):6864–76 Cook JE. 1996. Spatial properties of retinal mosaics: an empirical evaluation of some existing measures. Vis. Neurosci. 13(1):15–30 Coombs J, van der List D, Wang G-Y, Chalupa LM. 2006. Morphological properties of mouse retinal ganglion cells. Neuroscience 140(1):123–36 Crook JD, Peterson BB, Packer OS, Robinson FR, Troy JB, Dacey DM. 2008. Y-cell receptive field and collicular projection of parasol ganglion cells in macaque monkey retina. J. Neurosci. 28(44):11277–91 Cruz-Martı́n A, El-Danaf RN, Osakada F, Sriram B, Dhande OS, et al. 2014. A dedicated circuit links direction-selective retinal ganglion cells to the primary visual cortex. Nature 507(7492):358–61 Dacey D. 2004. Origins of perception: retinal ganglion cell diversity and the creation of parallel visual pathways. In The Cognitive Neurosciences, ed. MS Gazzaniga, pp. 281–301. Cambridge, MA: MIT Press Dacey DM, Liao H-W, Peterson BB, Robinson FR, Smith VC, et al. 2005. Melanopsin-expressing ganglion cells in primate retina signal colour and irradiance and project to the LGN. Nature 433(7027):749–54 Dacey DM, Petersen MR. 1992. Dendritic field size and morphology of midget and parasol ganglion cells of the human retina. PNAS 89(20):9666–70 Dean P, Mitchell IJ, Redgrave P. 1988. Responses resembling defensive behaviour produced by microinjection of glutamate into superior colliculus of rats. Neuroscience 24:501–10 Dean P, Redgrave P, Westby GW. 1989. Event or emergency? Two response systems in the mammalian superior colliculus. Trends Neurosci. 12:137–47 Delogu A, Sellers K, Zagoraiou L, Bocianowska-Zbrog A, Mandal S, et al. 2012. Subcortical visual shell nuclei targeted by ipRGCs develop from a Sox14+-GABAergic progenitor and require Sox14 to regulate daily activity rhythms. Neuron 75(4):648–62 Demb J, Singer J. 2015. Functional circuitry of the retina. Annu. Rev. Vis. Sci. 1:263–89 Dhande OS, Estevez ME, Quattrochi LE, El-Danaf RN, Nguyen PL, et al. 2013. Genetic dissection of retinal inputs to brainstem nuclei controlling image stabilization. J. Neurosci. 33(45):17797–813 Dhande OS, Huberman AD. 2014a. Retinal ganglion cell maps in the brain: implications for visual processing. Curr. Opin. Neurobiol. 24(1):133–42 Dhande OS, Huberman AD. 2014b. Visual circuits: mouse retina no longer a level playing field. Curr. Biol. 24(4):R155–56 Distler C, Hoffmann KP. 1989a. The pupillary light reflex in normal and innate microstrabismic cats, I: behavior and receptive-field analysis in the nucleus praetectalis olivaris. Vis. Neurosci. 3(2):127–38 Distler C, Hoffmann KP. 1989b. The pupillary light reflex in normal and innate microstrabismic cats, II: retinal and cortical input to the nucleus praetectalis olivaris. Vis. Neurosci. 3(2):139–53 Distler C, Hoffmann K. 2011. The optokinetic reflex. In The Oxford Handbook of Eye Movements, ed. SP Liversedge, ID Gilchrist, S Everling, pp. 65–83. New York: Oxford Univ. Press www.annualreviews.org • RGC Contributions to Feature Processing 321 ARI 27 October 2015 22:15 Dräger UC. 1975. Receptive fields of single cells and topography in mouse visual cortex. J. Comp. Neurol. 160(3):269–90 Dräger UC, Hubel DH. 1975a. Physiology of visual cells in mouse superior colliculus and correlation with somatosensory and auditory input. Nature 253(5488):203–4 Dräger UC, Hubel DH. 1975b. Responses to visual stimulation and relationship between visual, auditory, and somatosensory inputs in mouse superior colliculus. J. Neurophysiol. 38(3):690–713 Dräger UC, Hubel DH. 1976. Topography of visual and somatosensory projections to mouse superior colliculus. J. Neurophysiol. 39(1):91–101 Duan X, Qiao M, Bei F, Kim I-J, He Z, Sanes JR. 2015. Subtype-specific regeneration of retinal ganglion cells following axotomy: effects of osteopontin and mTOR signaling. Neuron 85:1244–56 Dumitrescu ON, Pucci FG, Wong KY, Berson DM. 2009. Ectopic retinal ON bipolar cell synapses in the OFF inner plexiform layer: contacts with dopaminergic amacrine cells and melanopsin ganglion cells. J. Comp. Neurol. 517(2):226–44 Ecker JL, Dumitrescu ON, Wong KY, Alam NM, Chen S-K, et al. 2010. Melanopsin-expressing retinal ganglion-cell photoreceptors: cellular diversity and role in pattern vision. Neuron 67(1):49–60 Edelstein K, Amir S. 1999. The role of the intergeniculate leaflet in entrainment of circadian rhythms to a skeleton photoperiod. J. Neurosci. 19(1):372–80 El-Danaf RN, Huberman AD. 2015. Characteristic patterns of dendritic remodeling in early-stage glaucoma: evidence from genetically identified retinal ganglion cell types. J. Neurosci. 35(6):2329–43 Ellard CG, Goodale MA. 1988. A functional analysis of the collicular output pathways: a dissociation of deficits following lesions of the dorsal tegmental decussation and the ipsilateral collicular efferent bundle in the Mongolian gerbil. Exp. Brain Res. 71(2):307–19 Estevez ME, Fogerson PM, Ilardi MC, Borghuis BG, Chan E, et al. 2012. Form and function of the M4 cell, an intrinsically photosensitive retinal ganglion cell type contributing to geniculocortical vision. J. Neurosci. 32(39):13608–20 Euler T, Haverkamp S, Schubert T, Baden T. 2014. Retinal bipolar cells: elementary building blocks of vision. Nat. Rev. Neurosci. 15(8):507–19 Farrow K, Teixeira M, Szikra T, Viney TJ, Balint K, et al. 2013. Ambient illumination toggles a neuronal circuit switch in the retina and visual perception at cone threshold. Neuron 78:325–38 Field GD, Sher A, Gauthier JL, Greschner M, Shlens J, et al. 2007. Spatial properties and functional organization of small bistratified ganglion cells in primate retina. J. Neurosci. 27(48):13261–72 Field GD, Chichilnisky EJ. 2007. Information processing in the primate retina: circuitry and coding. Annu. Rev. Neurosci. 30:1–30 Feinberg EH, Meister M. 2014. Orientation columns in the mouse superior colliculus. Nature 519(7542):229– 32 Gale SD, Murphy GJ. 2014. Distinct representation and distribution of visual information by specific cell types in mouse superficial superior colliculus. J. Neurosci. 34(40):13458–71 Gamlin PD, Clarke RJ. 1995. The pupillary light reflex pathway of the primate. J. Am. Optom. Assoc. 66(7):415– 18 Gamlin PD, Zhang H, Clarke RJ. 1995. Luminance neurons in the pretectal olivary nucleus mediate the pupillary light reflex in the rhesus monkey. Exp. Brain Res. 106(1):169–76 Gamlin PDR. 2006. The pretectum: connections and oculomotor-related roles. Prog. Brain Res. 151:379–405 Gandhi NJ, Katnani HA. 2011. Motor functions of the superior colliculus. Annu. Rev. Neurosci. 34:205–31 Gebhardt C, Baier H, Del Bene F. 2013. Direction selectivity in the visual system of the zebrafish larva. Front. Neural Circuits 7:111 Giolli RA, Blanks RHI, Lui F. 2006. The accessory optic system: basic organization with an update on connectivity, neurochemistry, and function. Prog. Brain Res. 151:407–440 Girman SV, Sauvé Y, Lund RD. 1999. Receptive field properties of single neurons in rat primary visual cortex. J. Neurophysiol. 82(1):301–11 Grimes WN, Seal RP, Oesch N, Edwards RH, Diamond JS. 2011. Genetic targeting and physiological features of VGLUT3+ amacrine cells. Vis. Neurosci. 28(5):381–92 Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. VS01CH12-Huberman 322 Dhande et al. Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. VS01CH12-Huberman ARI 27 October 2015 22:15 Grubb MS, Thompson ID. 2004. Biochemical and anatomical subdivision of the dorsal lateral geniculate nucleus in normal mice and in mice lacking the β2 subunit of the nicotinic acetylcholine receptor. Vision Res. 44(28):3365–76 Gooley JJ, Lu J, Chou TC, Scammell TE, Saper CB. 2001. Melanopsin in cells of origin of the retinohypothalamic tract. Nat. Neurosci. 4(12):1165 Göz D, Studholme K, Lappi DA, Rollag MD, Provencio I, Morin LP. 2008. Targeted destruction of photosensitive retinal ganglion cells with a saporin conjugate alters the effects of light on mouse circadian rhythms. PLOS ONE 3(9):e3153 Güler AD, Ecker JL, Lall GS, Haq S, Altimus CM, et al. 2008. Melanopsin cells are the principal conduits for rod-cone input to non-image-forming vision. Nature 453:102–5 Hannibal J, Hindersson P, Ostergaard J, Georg B, Heegaard S, et al. 2004. Melanopsin is expressed in PACAP-containing retinal ganglion cells of the human retinohypothalamic tract. Investig. Ophthalmol. Vis. Sci. 45(11):4202–9 Harrington ME. 1997. The ventral lateral geniculate nucleus and the intergeniculate leaflet: interrelated structures in the visual and circadian systems. Neurosci. Biobehav. Rev. 21(5):705–27 Harting JK, Huerta MF, Hashikawa T, van Lieshout DP. 1991. Projection of the mammalian superior colliculus upon the dorsal lateral geniculate nucleus: organization of tectogeniculate pathways in nineteen species. J. Comp. Neurol. 304(2):275–306 Hatori M, Le H, Vollmers C, Keding SR, Tanaka N, et al. 2008. Inducible ablation of melanopsin-expressing retinal ganglion cells reveals their central role in non-image forming visual responses. PLOS ONE 3(6):e2451 Hattar S, Liao HW, Takao M, Berson DM, Yau KW. 2002. Melanopsin-containing retinal ganglion cells: architecture, projections, and intrinsic photosensitivity. Science 295(5557):1065–70 Hattar S, Lucas RJ, Mrosovsky N, Thompson S, Douglas RH, et al. 2003. Melanopsin and rod–cone photoreceptive systems account for all major accessory visual functions in mice. Nature 424:76–81 Hattar S, Kumar M, Park A, Tong P, Tung J, et al. 2006. Central projections of melanopsin-expressing retinal ganglion cells in the mouse. J. Comp. Neurol. 497(3):326–49 Hendry SH, Reid RC. 2000. The koniocellular pathway in primate vision. Annu. Rev. Neurosci. 23:127–53 Hofbauer A, Dräger UC. 1985. Depth segregation of retinal ganglion cells projecting to mouse superior colliculus. J. Comp. Neurol. 234(4):465–74 Hong YK, Chen C. 2011. Wiring and rewiring of the retinogeniculate synapse. Curr. Opin. Neurobiol. 21(2):228–37 Hu C, Hill DD, Wong KY. 2013. Intrinsic physiological properties of the five types of mouse ganglion-cell photoreceptors. J. Neurophysiol. 109:1876–89 Hubel DH, Wiesel TN. 1962. Receptive fields, binocular interaction and functional architecture in the cat’s visual cortex. J. Physiol. 160:106–54 Hubel DH, Wiesel TN. 1968. Receptive fields and functional architecture of monkey striate cortex. J. Physiol. 195(1):215–43 Huberman AD, Feller MB, Chapman B. 2008a. Mechanisms underlying development of visual maps and receptive fields. Annu. Rev. Neurosci. 31:479–509 Huberman AD, Manu M, Koch SM, Susman MW, Lutz AB, et al. 2008b. Architecture and activity-mediated refinement of axonal projections from a mosaic of genetically identified retinal ganglion cells. Neuron 59(3):425–38 Huberman AD, Wei W, Elstrott J, Stafford BK, Feller MB, Barres BA. 2009. Genetic identification of an OnOff direction-selective retinal ganglion cell subtype reveals a layer-specific subcortical map of posterior motion. Neuron 62(3):327–34 Ingle D. 1973. Two visual systems in the frog. Science 181(4104):1053–55 Isa T, Hall WC. 2009. Exploring the superior colliculus in vitro. J. Neurophysiol. 102(5):2581–93 Jaubert-Miazza L, Green E, Lo F-S, Bui K, Mills J, Guido W. 2005. Structural and functional composition of the developing retinogeniculate pathway in the mouse. Vis. Neurosci. 22(5):661–76 Katz LC, Shatz CJ. 1996. Synaptic activity and the construction of cortical circuits. Science 274(5290):1133–38 www.annualreviews.org • RGC Contributions to Feature Processing 323 ARI 27 October 2015 22:15 Kay JN, De la Huerta I, Kim I-J, Zhang Y, Yamagata M, et al. 2011. Retinal ganglion cells with distinct directional preferences differ in molecular identity, structure, and central projections. J. Neurosci. 31(21):7753–62 Kerlin AM, Andermann ML, Berezovskii VK, Reid RC. 2010. Broadly tuned response properties of diverse inhibitory neuron subtypes in mouse visual cortex. Neuron 67(5):858–71 Kim I-J, Zhang Y, Meister M, Sanes JR. 2010. Laminar restriction of retinal ganglion cell dendrites and axons: subtype-specific developmental patterns revealed with transgenic markers. J. Neurosci. 30(4):1452–62 Kim I-J, Zhang Y, Yamagata M, Meister M, Sanes JR. 2008. Molecular identification of a retinal cell type that responds to upward motion. Nature 452(7186):478–82 Kim JS, Greene MJ, Zlateski A, Lee K, Richardson M, et al. 2014. Space–time wiring specificity supports direction selectivity in the retina. Nature 509(7500):331–36 Kirkby LA, Sack GS, Firl A, Feller MB. 2013. A role for correlated spontaneous activity in the assembly of neural circuits. Neuron 80(5):1129–44 Knudsen EI, du Lac S, Esterly SD. 1987. Computational maps in the brain. Annu. Rev. Neurosci. 10:41–65 Korenbrot JI. 2012. Speed, sensitivity, and stability of the light response in rod and cone photoreceptors: facts and models. Prog. Retin. Eye Res. 31(5):442–66 Krahe TE, El-Danaf RN, Dilger EK, Henderson SC, Guido W. 2011. Morphologically distinct classes of relay cells exhibit regional preferences in the dorsal lateral geniculate nucleus of the mouse. J. Neurosci. 31(48):17437–48 Krauzlis RJ, Lovejoy LP, Zénon A. 2013. Superior colliculus and visual spatial attention. Annu. Rev. Neurosci. 36:165–82 Krishnaswamy A, Yamagata M, Duan X, Hong YK, Sanes JR. 2015. Sidekick 2 directs formation of a retinal circuit that detects differential motion. Nature. 524:466–70 Kuffler SW. 1953. Discharge patterns and functional organization of mammalian retina. J. Neurophysiol. 16(1):37–68 Langer TP, Lund RD. 1974. The upper layers of the superior colliculus of the rat: a Golgi study. J. Comp. Neurol. 158(4):418–35 Lee S, Chen L, Chen M, Ye M, Seal RP, Zhou ZJ. 2014. An unconventional glutamatergic circuit in the retina formed by vGluT3 amacrine cells. Neuron 84(4):708–15 LeGates TA, Fernandez DC, Hattar S. 2014. Light as a central modulator of circadian rhythms, sleep and affect. Nat. Rev. Neurosci. 15(7):443–54 Lettvin JY, Maturana HR, McCulloch WS, Pitts WH. 1959. What the frog’s eye tells the frog’s brain. Proc. IRE 47(11):1940–51 Leventhal AG, Rodieck RW, Dreher B. 1981. Retinal ganglion cell classes in the Old World monkey: morphology and central projections. Science 213(4512):1139–42 Levick WR. 1967. Receptive fields and trigger features of ganglion cells in the visual streak of the rabbit’s retina. J. Physiol. 188(3):285–307 Levick WR, Oyster CW, Takahashi E. 1969. Rabbit lateral geniculate nucleus: sharpener of directional information. Science 165(3894):712–14 Levick WR, Thibos LN. 1982. Analysis of orientation bias in cat retina. J. Physiol. 329:243–61 Liang F, Xiong XR, Zingg B, Ji X-Y, Zhang LI, Tao HW. 2015. Sensory cortical control of a visually induced arrest behavior via corticotectal projections. Neuron 86(3):755–67 Lin B, Wang SW, Masland RH. 2004. Retinal ganglion cell type, size, and spacing can be specified independent of homotypic dendritic contacts. Neuron 43(4):475–85 Linden R, Perry VH. 1983. Massive retinotectal projection in rats. Brain Res. 272(1):145–49 Ling S, Pratte MS, Tong F. 2015. Attention alters orientation processing in the human lateral geniculate nucleus. Nat. Neurosci. 18:496–98 Lucas RJ, Douglas RH, Foster RG. 2001. Characterization of an ocular photopigment capable of driving pupillary constriction in mice. Nat. Neurosci. 4(6):621–26 Lucas RJ, Hattar S, Takao M, Berson DM, Foster RG, Yau K-W. 2003. Diminished pupillary light reflex at high irradiances in melanopsin-knockout mice. Science 299:245–47 Lucas RJ, Lall GS, Allen AE, Brown TM. 2012. How rod, cone, and melanopsin photoreceptors come together to enlighten the mammalian circadian clock. Prog. Brain Res. 199:1–18 Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. VS01CH12-Huberman 324 Dhande et al. Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. VS01CH12-Huberman ARI 27 October 2015 22:15 Mangini NJ, Pearlman AL. 1980. Laminar distribution of receptive field properties in the primary visual cortex of the mouse. J. Comp. Neurol. 193(1):203–22 Manookin MB, Beaudoin DL, Ernst ZR, Flagel LJ, Demb JB. 2008. Disinhibition combines with excitation to extend the operating range of the OFF visual pathway in daylight. J. Neurosci. 28(16):4136–50 Mao C-A, Li H, Zhang Z, Kiyama T, Panda S, et al. 2014. T-box transcription regulator Tbr2 is essential for the formation and maintenance of Opn4/melanopsin-expressing intrinsically photosensitive retinal ganglion cells. J. Neurosci. 34(39):13083–95 Marshel JH, Kaye AP, Nauhaus I, Callaway EM. 2012. Anterior-posterior direction opponency in the superficial mouse lateral geniculate nucleus. Neuron 76(4):713–20 Masseck OA, Hoffmann K-P. 2009. Comparative neurobiology of the optokinetic reflex. Ann. N. Y. Acad. Sci. 1164:430–39 May PJ. 2006. The mammalian superior colliculus: laminar structure and connections. Prog. Brain Res. 151:321–78 McLaughlin T, O’Leary DDM. 2005. Molecular gradients and development of retinotopic maps. Annu. Rev. Neurosci. 28:327–55 Métin C, Godement P, Imbert M. 1988. The primary visual cortex in the mouse: receptive field properties and functional organization. Exp. Brain Res. 69(3):594–612 Mooney RD, Rhoades RW. 1990. Relationships between physiological and morphological properties of retinocollicular axons in the hamster. J. Neurosci. 10(9):3164–77 Morin LP. 2013. Neuroanatomy of the extended circadian rhythm system. Exp. Neurol. 243:4–20 Morin LP, Studholme KM. 2014. Retinofugal projections in the mouse. J. Comp. Neurol. 522(16):3733–53 Mrosovsky N. 1995. A non-photic gateway to the circadian clock of hamsters. Ciba Found. Symp. 183:154–67; discussion pp. 167–74 Münch TA, da Silveira RA, Siegert S, Viney TJ, Awatramani GB, Roska B. 2009. Approach sensitivity in the retina processed by a multifunctional neural circuit. Nat. Neurosci. 12(10):1308–16 Mysore SP, Knudsen EI. 2011. The role of a midbrain network in competitive stimulus selection. Curr. Opin. Neurobiol. 21(4):653–60 Nathans J. 1987. Molecular biology of visual pigments. Annu. Rev. Neurosci. 10:163–94 Niell CM, Stryker MP. 2008. Highly selective receptive fields in mouse visual cortex. J. Neurosci. 28(30):7520–36 Niell CM, Stryker MP. 2010. Modulation of visual responses by behavioral state in mouse visual cortex. Neuron 65(4):472–79 Northmore DP, Levine ES, Schneider GE. 1988. Behavior evoked by electrical stimulation of the hamster superior colliculus. Exp. Brain Res. 73(3):595–605 Ohki K, Chung S, Ch’ng YH, Kara P, Reid RC. 2005. Functional imaging with cellular resolution reveals precise micro-architecture in visual cortex. Nature 433(7026):597–603 Ölveczky BP, Baccus SA, Meister M. 2003. Segregation of object and background motion in the retina. Nature 423(6938):401–8 Osterhout JA, Josten N, Yamada J, Pan F, Wu S, et al. 2011. Cadherin-6 mediates axon-target matching in a non-image-forming visual circuit. Neuron 71(4):632–39 Oyster CW, Takahashi E, Collewijn H. 1972. Direction-selective retinal ganglion cells and control of optokinetic nystagmus in the rabbit. Vision Res. 12(2):183–93 Pak MW, Giolli RA, Pinto LH, Mangini NJ, Gregory KM, Vanable JW Jr. 1987. Retinopretectal and accessory optic projections of normal mice and the OKN-defective mutant mice beige, beige-J, and pearl. J. Comp. Neurol. 258:435–46 Panda S, Provencio I, Tu DC, Pires SS, Rollag MD, et al. 2003. Melanopsin is required for non-image-forming photic responses in blind mice. Science 301(5632):525–27 Panda S, Sato TK, Castrucci AM, Rollag MD, DeGrip WJ, et al. 2002. Melanopsin (Opn4) requirement for normal light-induced circadian phase shifting. Science 298(5601):2213–16 Pang J-J, Gao F, Wu SM. 2003. Light-evoked excitatory and inhibitory synaptic inputs to ON and OFF alpha ganglion cells in the mouse retina. J. Neurosci. 23(14):6063–73 Passaglia CL, Troy JB, Rüttiger L, Lee BB. 2002. Orientation sensitivity of ganglion cells in primate retina. Vision Res. 42(6):683–94 www.annualreviews.org • RGC Contributions to Feature Processing 325 ARI 27 October 2015 22:15 Petrusca D, Grivich MI, Sher A, Field GD, Gauthier JL, et al. 2007. Identification and characterization of a Y-like primate retinal ganglion cell type. J. Neurosci. 27(41):11019–27 Pfeiffenberger C, Yamada J, Feldheim DA. 2006. Ephrin-As and patterned retinal activity act together in the development of topographic maps in the primary visual system. J. Neurosci. 26(50):12873–84 Piscopo DM, El-Danaf RN, Huberman AD, Niell CM. 2013. Diverse visual features encoded in mouse lateral geniculate nucleus. J. Neurosci. 33(11):4642–56 Prichard JR, Stoffel RT, Quimby DL, Obermeyer WH, Benca RM, Behan M. 2002. Fos immunoreactivity in rat subcortical visual shell in response to illuminance changes. Neuroscience 114(3):781–93 Rakic P. 1976. Prenatal genesis of connections subserving ocular dominance in the rhesus monkey. Nature 261(5560):467–71 Rancz EA, Franks KM, Schwarz MK, Pichler B, Schaefer AT, Margrie TW. 2011. Transfection via whole-cell recording in vivo: bridging single-cell physiology, genetics and connectomics. Nat. Neurosci. 14(4):527–32 Reese BE. 1984. The projection from the superior colliculus to the dorsal lateral geniculate nucleus in the rat. Brain Res. 305(1):162–68 Reese BE. 1988. “Hidden lamination” in the dorsal lateral geniculate nucleus: the functional organization of this thalamic region in the rat. Brain Res. 472(2):119–37 Rivlin-Etzion M, Zhou K, Wei W, Elstrott J, Nguyen PL, et al. 2011. Transgenic mice reveal unexpected diversity of On-Off direction-selective retinal ganglion cell subtypes and brain structures involved in motion processing. J. Neurosci. 31(24):8760–69 Rodieck RW. 1979. Visual pathways. Annu. Rev. Neurosci. 2:193–225 Rodieck RW, Binmoeller KF, Dineen J. 1985. Parasol and midget ganglion cells of the human retina. J. Comp. Neurol. 233(1):115–32 Rodieck RW, Stone J. 1965. Analysis of receptive fields of cat retinal ganglion cells. J. Neurophysiol. 28:832–49 Rosenquist AC, Palmer LA. 1971. Visual receptive field properties of cells of the superior colliculus after cortical lesions in the cat. Exp. Neurol. 33(3):629–52 Ruby NF, Brennan TJ, Xie X, Cao V, Franken P, et al. 2002. Role of melanopsin in circadian responses to light. Science 298(5601):2211–13 Sahibzada N, Dean P, Redgrave P. 1986. Movements resembling orientation or avoidance elicited by electrical stimulation of the superior colliculus in rats. J. Neurosci. 6(3):723–33 Sanes JR, Masland RH. 2015. The types of retinal ganglion cells: current status and implications for neuronal classification. Annu. Rev. Neurosci. 38:221–46 Schmidt TM, Alam NM, Chen S, Kofuji P, Li W, et al. 2014. A role for melanopsin in alpha retinal ganglion cells and contrast detection. Neuron 82(4):781–88 Schmidt TM, Taniguchi K, Kofuji P. 2008. Intrinsic and extrinsic light responses in melanopsin-expressing ganglion cells during mouse development. J. Neurophysiol. 100:371–84 Scholl B, Tan AYY, Corey J, Priebe NJ. 2013. Emergence of orientation selectivity in the mammalian visual pathway. J. Neurosci. 33(26):10616–24 Seabrook TA, Krahe TE, Govindaiah G, Guido W. 2013. Interneurons in the mouse visual thalamus maintain a high degree of retinal convergence throughout postnatal development. Neural Dev. 8:24 Seung HS, Sümbül U. 2014. Neuronal cell types and connectivity: lessons from the retina. Neuron 83(6):1262–72 Shang C, Liu Z, Chen Z, Shi Y, Wang Q, et al. 2015. A parvalbumin-positive excitatory visual pathway to trigger fear responses in mice. Science 348(6242):1472–77 Siegert S, Cabuy E, Scherf BG, Kohler H, Panda S, et al. 2012. Transcriptional code and disease map for adult retinal cell types. Nat. Neurosci. 15(3):487–95, S1–2 Simpson JI. 1984. The accessory optic system. Annu. Rev. Neurosci. 7:13–41 Sivyer B, van Wyk M, Vaney DI, Taylor WR. 2010. Synaptic inputs and timing underlying the velocity tuning of direction-selective ganglion cells in rabbit retina. J. Physiol. 588:3243–53 Sohya K, Kameyama K, Yanagawa Y, Obata K, Tsumoto T. 2007. GABAergic neurons are less selective to stimulus orientation than excitatory neurons in layer II/III of visual cortex, as revealed by in vivo functional Ca2+ imaging in transgenic mice. J. Neurosci. 27(8):2145–49 Sparks DL. 2002. The brainstem control of saccadic eye movements. Nat. Rev. Neurosci. 3(12):952–64 Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. VS01CH12-Huberman 326 Dhande et al. Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. VS01CH12-Huberman ARI 27 October 2015 22:15 Sparks DL, Nelson IS. 1987. Sensory and motor maps in the mammalian superior colliculus. Trends Neurosci. 10(8):312–17 Stafford BK, Kupershtok M, Demb JB. 2010. Cell type–specific differences in NMDA receptor contributions to mouse retinal ganglion cell responses. Presented at Fed. Am. Soc. Exp. Biol. Conf. Retin. Neurobiol. Vis. Process. Saxtons River, VT, Jul. 11–16 Stein BE, Stanford TR, Rowland BA. 2014. Development of multisensory integration from the perspective of the individual neuron. Nat. Rev. Neurosci. 15(8):520–35 Stone J. 1983. Parallel Processing in the Visual System. Boston, MA: Springer Sugita Y, Miura K, Araki F, Furukawa T, Kawano K. 2013. Contributions of retinal direction-selective ganglion cells to optokinetic responses in mice. Eur. J. Neurosci. 38(6):2823–31 Sümbül U, Song S, McCulloch K, Becker M, Lin B, et al. 2014. A genetic and computational approach to structurally classify neuronal types. Nat. Commun. 5:3512 Sweeney NT, Tierney H, Feldheim DA. 2014. Tbr2 is required to generate a neural circuit mediating the pupillary light reflex. J. Neurosci. 34(16):5447–53 Szmajda BA, Grünert U, Martin PR. 2008. Retinal ganglion cell inputs to the koniocellular pathway. J. Comp. Neurol. 510(3):251–68 Tamamaki N, Uhlrich DJ, Sherman SM. 1995. Morphology of physiologically identified retinal X and Y axons in the cat’s thalamus and midbrain as revealed by intraaxonal injection of biocytin. J. Comp. Neurol. 354(4):583–607 Thoreson WB, Mangel SC. 2012. Lateral interactions in the outer retina. Prog. Retin. Eye Res. 31(5):407–41 Trejo LJ, Cicerone CM. 1984. Cells in the pretectal olivary nucleus are in the pathway for the direct light reflex of the pupil in the rat. Brain Res. 300(1):49–62 Trenholm S, Johnson K, Li X, Smith RG, Awatramani GB. 2011. Parallel mechanisms encode direction in the retina. Neuron 71(4):683–94 Triplett JW. 2014. Molecular guidance of retinotopic map development in the midbrain. Curr. Opin. Neurobiol. 24(1):7–12 Ts’o DY, Frostig RD, Lieke EE, Grinvald A. 1990. Functional organization of primate visual cortex revealed by high resolution optical imaging. Science 249(4967):417–20 Usrey WM, Alitto HJ. 2015. Visual functions of the thalamus. Annu. Rev. Vis. Sci. 1:351–71 van Wyk M, Taylor WR, Vaney D. 2006. Local edge detectors: a substrate for fine spatial vision at low temporal frequencies in rabbit retina. J. Neurosci. 26(51):13250–63 van Wyk M, Wässle H, Taylor WR. 2009. Receptive field properties of ON- and OFF-ganglion cells in the mouse retina. Vis. Neurosci. 26(3):297–308 Vaney DI, Sivyer B, Taylor WR. 2012. Direction selectivity in the retina: symmetry and asymmetry in structure and function. Nat. Rev. Neurosci. 13(3):194–208 Venkataramani S, Taylor WR. 2010. Orientation selectivity in rabbit retinal ganglion cells is mediated by presynaptic inhibition. J. Neurosci. 30(46):15664–76 Verwey M, Amir S. 2009. Food-entrainable circadian oscillators in the brain. Eur. J. Neurosci. 30(9):1650–57 Völgyi B, Chheda S, Bloomfield SA. 2009. Tracer coupling patterns of the ganglion cell subtypes in the mouse retina. J. Comp. Neurol. 512(5):664–87 Wang L, Rangarajan KV, Lawhn-Heath CA, Sarnaik R, Wang B-S, et al. 2009. Direction-specific disruption of subcortical visual behavior and receptive fields in mice lacking the β2 subunit of nicotinic acetylcholine receptor. J. Neurosci. 29(41):12909–18 Wang L, Sarnaik R, Rangarajan K, Liu X, Cang J. 2010. Visual receptive field properties of neurons in the superficial superior colliculus of the mouse. J. Neurosci. 30(49):16573–84 Watanabe M, Rodieck RW. 1989. Parasol and midget ganglion cells of the primate retina. J. Comp. Neurol. 289(3):434–54 Wiesel TN. 1960. Receptive fields of ganglion cells in the cat’s retina. J. Physiol. 153:583–94 Wei P, Liu N, Zhang Z, Liu X, Tang Y, et al. 2015. Processing of visually evoked innate fear by a non-canonical thalamic pathway. Nat. Commun. 6:6756 Wei W, Feller MB. 2011. Organization and development of direction-selective circuits in the retina. Trends Neurosci. 34(12):638–45 www.annualreviews.org • RGC Contributions to Feature Processing 327 ARI 27 October 2015 22:15 Welsh DK, Takahashi JS, Kay SA. 2010. Suprachiasmatic nucleus: cell autonomy and network properties. Annu. Rev. Physiol. 72:551–77 Weng S, Sun W, He S. 2005. Identification of ON–OFF direction-selective ganglion cells in the mouse retina. J. Physiol. 562:915–23 Wertz A, Trenholm S, Yonehara K, Hillier D, Raics Z, et al. 2015. Single-cell–initiated monosynaptic tracing reveals layer-specific cortical network modules. Science 349(6243):70–74 Westby GW, Keay KA, Redgrave P, Dean P, Bannister M. 1990. Output pathways from the rat superior colliculus mediating approach and avoidance have different sensory properties. Exp. Brain Res. 81(3):626– 38 Wickelgren BG, Sterling P. 1969. Influence of visual cortex on receptive fields in the superior colliculus of the cat. J. Neurophysiol. 32(1):16–23 Witkovsky P. 2004. Dopamine and retinal function. Doc. Ophthalmol. 108(1):17–40 Wurtz RH, Albano JE. 1980. Visual-motor function of the primate superior colliculus. Annu. Rev. Neurosci. 3:189–226 Xiang M, Zhou H, Nathans J. 1996. Molecular biology of retinal ganglion cells. PNAS 93(2):596–601 Yilmaz M, Meister M. 2013. Rapid innate defensive responses of mice to looming visual stimuli. Curr. Biol. 23(20):2011–15 Yonehara K, Balint K, Noda M, Nagel G, Bamberg E, Botond R. 2011. Spatially asymmetric reorganization of inhibition establishes a motion-sensitive circuit. Nature 469:407–410 Yonehara K, Farrow K, Ghanem A, Hillier D, Balint K, et al. 2013. The first stage of cardinal direction selectivity is localized to the dendrites of retinal ganglion cells. Neuron 79(6):1078–85 Yonehara K, Ishikane H, Sakuta H, Shintani T, Nakamura-Yonehara K, et al. 2009. Identification of retinal ganglion cells and their projections involved in central transmission of information about upward and downward image motion. PLOS ONE 4(1):e4320 Yonehara K, Shintani T, Suzuki R, Sakuta H, Takeuchi Y, et al. 2008. Expression of SPIG1 reveals development of a retinal ganglion cell subtype projecting to the medial terminal nucleus in the mouse. PLOS ONE 3(2):e1533 Yoshida K, Watanabe D, Ishikane H, Tachibana M, Pastan I, Nakanishi S. 2001. A key role of starburst amacrine cells in originating retinal directional selectivity and optokinetic eye movement. Neuron 30(3):771–80 Young MJ, Lund RD. 1994. The anatomical substrates subserving the pupillary light reflex in rats: origin of the consensual pupillary response. Neuroscience 62(2):481–96 Zhang Y, Kim I-J, Sanes JR, Meister M. 2012. The most numerous ganglion cell type of the mouse retina is a selective feature detector. PNAS 109(36):E2391–98 Zhao X, Chen H, Liu X, Cang J. 2013. Orientation-selective responses in the mouse lateral geniculate nucleus. J. Neurosci. 33(31):12751–63 Zhao X, Liu M, Cang J. 2014a. Visual cortex modulates the magnitude but not the selectivity of loomingevoked responses in the superior colliculus of awake mice. Neuron 84(1):202–13 Zhao X, Stafford BK, Godin AL, King WM, Wong KY. 2014b. Photoresponse diversity among the five types of intrinsically photosensitive retinal ganglion cells. J. Physiol. 592:1619–36 Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. VS01CH12-Huberman 328 Dhande et al. VS01-FrontMatter ARI 30 October 2015 19:57 Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. Annual Review of Vision Science Contents Volume 1, 2015 An autobiographical article by Horace Barlow is available online at www.annualreviews.org/r/horacebarlow. Image Formation in the Living Human Eye Pablo Artal p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 Adaptive Optics Ophthalmoscopy Austin Roorda and Jacque L. Duncan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p19 Imaging Glaucoma Donald C. Hood p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p51 What Does Genetics Tell Us About Age-Related Macular Degeneration? Felix Grassmann, Thomas Ach, Caroline Brandl, Iris M. Heid, and Bernhard H.F. Weber p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p73 Mitochondrial Genetics and Optic Neuropathy Janey L. Wiggs p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p97 Zebrafish Models of Retinal Disease Brian A. Link and Ross F. Collery p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 125 Angiogenesis and Eye Disease Yoshihiko Usui, Peter D. Westenskow, Salome Murinello, Michael I. Dorrell, Leah Scheppke, Felicitas Bucher, Susumu Sakimoto, Liliana P. Paris, Edith Aguilar, and Martin Friedlander p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 155 Optogenetic Approaches to Restoring Vision Zhuo-Hua Pan, Qi Lu, Anding Bi, Alexander M. Dizhoor, and Gary W. Abrams p p p p 185 The Determination of Rod and Cone Photoreceptor Fate Constance L. Cepko p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 211 Ribbon Synapses and Visual Processing in the Retina Leon Lagnado and Frank Schmitz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 235 Functional Circuitry of the Retina Jonathan B. Demb and Joshua H. Singer p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 263 vii VS01-FrontMatter ARI 30 October 2015 19:57 Contributions of Retinal Ganglion Cells to Subcortical Visual Processing and Behaviors Onkar S. Dhande, Benjamin K. Stafford, Jung-Hwan A. Lim, and Andrew D. Huberman p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 291 Annu. Rev. Vis. Sci. 2015.1:291-328. Downloaded from www.annualreviews.org Access provided by Stanford University - Main Campus - Robert Crown Law Library on 09/08/16. For personal use only. Organization of the Central Visual Pathways Following Field Defects Arising from Congenital, Inherited, and Acquired Eye Disease Antony B. Morland p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 329 Visual Functions of the Thalamus W. Martin Usrey and Henry J. Alitto p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 351 Neuronal Mechanisms of Visual Attention John Maunsell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 373 A Revised Neural Framework for Face Processing Brad Duchaine and Galit Yovel p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 393 Deep Neural Networks: A New Framework for Modeling Biological Vision and Brain Information Processing Nikolaus Kriegeskorte p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 417 Visual Guidance of Smooth Pursuit Eye Movements Stephen G. Lisberger p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 447 Visuomotor Functions in the Frontal Lobe Jeffrey D. Schall p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 469 Control and Functions of Fixational Eye Movements Michele Rucci and Martina Poletti p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 499 Color and the Cone Mosaic David H. Brainard p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 519 Visual Adaptation Michael A. Webster p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 547 Development of Three-Dimensional Perception in Human Infants Anthony M. Norcia and Holly E. Gerhard p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 569 Errata An online log of corrections to Annual Review of Vision Science articles may be found at http://www.annualreviews.org/errata/vision viii Contents