Classes Of Retinal Ganglion Cells Research Articles

Retinal Ganglion Cell Subtypes and Their Vulnerability in Glaucoma.

The detection of selective retinal ganglion cell damage in glaucoma has been a long sought-after goal, not just for the development of clinical tests for the early detection of glaucoma but for the elucidation of potential mechanisms underlying retinal ganglion cell loss. Early reports of the selective vulnerability of larger retinal ganglion cells (RGCs) in human studies did not translate simply to the loss of a particular class of RGC but more likely reflected shrinkage and degeneration across all RGC classes. Subsequent studies of nonhuman primate (NHP) models of glaucoma indicated some selectivity with great damage to the magnocellular vs parvocellular pathways. More recently, rodent models of experimental glaucoma have highlighted a selective vulnerability of OFF-centered RGCs-particularly those with transient responses. Selectivity for OFF pathway damage is also seen as a trend in a rat model of glaucoma. These data support the concept that some RGCs are more vulnerable to the effects of glaucoma damage. This chapter covers some of the methods to elucidate RGC damage and the relevance of model selection to mimic human glaucoma rather than just RGC death.

Functional cell types in the mouse superior colliculus.

The superior colliculus (SC) represents a major visual processing station in the mammalian brain that receives input from many types of retinal ganglion cells (RGCs). How many parallel channels exist in the SC, and what information does each encode? Here, we recorded from mouse superficial SC neurons under a battery of visual stimuli including those used for classification of RGCs. An unsupervised clustering algorithm identified 24 functional types based on their visual responses. They fall into two groups: one that responds similarly to RGCs and another with more diverse and specialized stimulus selectivity. The second group is dominant at greater depths, consistent with a vertical progression of signal processing in the SC. Cells of the same functional type tend to cluster near each other in anatomical space. Compared to the retina, the visual representation in the SC has lower dimensionality, consistent with a sifting process along the visual pathway.

Neural and Müller glial adaptation of the retina to photoreceptor degeneration.

The majority of inherited retinal degenerative diseases and dry age-related macular degeneration are characterized by decay of the outer retina and photoreceptors, which leads to progressive loss of vision. The inner retina, including second- and third-order retinal neurons, also shows aberrant structural changes at all stages of degeneration. Müller glia, the major glial cells maintain retinal homeostasis, activating and rearranging immediately in response to photoreceptor stress. These phenomena are collectively known as retinal remodeling and are anatomically well described, but their impact on visual function is less well characterized. Retinal remodeling has traditionally been considered a detrimental chain of events that decreases visual function. However, emerging evidence from functional assays suggests that remodeling could also be a part of a survival mechanism wherein the inner retina responds plastically to outer retinal degeneration. The visual system´s first synapses between the photoreceptors and bipolar cells undergo rewiring and functionally compensate to maintain normal signal output to the brain. Distinct classes of retinal ganglion cells remain even after the massive loss of photoreceptors. Müller glia possess the regenerative potential for retinal recovery and possibly exert adaptive transcriptional changes in response to neuronal loss. These types of homeostatic changes could potentially explain the well-maintained visual function observed in patients with inherited retinal degenerative diseases who display prominent anatomic retinal pathology. This review will focus on our current understanding of retinal neuronal and Müller glial adaptation for the potential preservation of retinal activity during photoreceptor degeneration. Targeting retinal self-compensatory responses could help generate universal strategies to delay sensory disease progression.

A Role for Thalamic Projection GABAergic Neurons in Circadian Responses to Light.

The thalamus is an important hub for sensory information and participates in sensory perception, regulation of attention, arousal and sleep. These functions are executed primarily by glutamatergic thalamocortical neurons that extend axons to the cortex and initiate cortico-thalamocortical connectional loops. However, the thalamus also contains projection GABAergic neurons that do not extend axons toward the cortex. Here, we have harnessed recent insight into the development of the intergeniculate leaflet (IGL) and the ventral lateral geniculate nucleus (LGv) to specifically target and manipulate thalamic projection GABAergic neurons in female and male mice. Our results show that thalamic GABAergic neurons of the IGL and LGv receive retinal input from diverse classes of retinal ganglion cells (RGCs) but not from the M1 intrinsically photosensitive retinal ganglion cell (ipRGC) type. We describe the synergistic role of the photoreceptor melanopsin and the thalamic neurons of the IGL/LGv in circadian entrainment to dim light. We identify a requirement for the thalamic IGL/LGv neurons in the rapid changes in vigilance states associated with circadian light transitions.SIGNIFICANCE STATEMENT The intergeniculate leaflet (IGL) and ventral lateral geniculate nucleus (LGv) are part of the extended circadian system and mediate some nonimage-forming visual functions. Here, we show that each of these structures has a thalamic (dorsal) as well as prethalamic (ventral) developmental origin. We map the retinal input to thalamus-derived cells in the IGL/LGv complex and discover that while RGC input is dominant, this is not likely to originate from M1ipRGCs. We implicate thalamic cells in the IGL/LGv in vigilance state transitions at circadian light changes and in overt behavioral entrainment to dim light, the latter exacerbated by concomitant loss of melanopsin expression.

Light-evoked dendritic spikes in sustained but not transient rabbit retinal ganglion cells

Dendritic computations have a central role in neuronal function, but it is unknown how cell-class heterogeneity of dendritic electrical excitability shapes physiologically engaged neuronal and circuit computations. To address this, we examined dendritic integration in closely related classes of retinal ganglion cells (GCs) using simultaneous somato-dendritic electrical recording techniques in a functionally intact circuit. Simultaneous recordings revealed sustained OFF-GCs generated powerful dendritic spikes in response to visual input that drove action potential firing. In contrast, the dendrites of transient OFF-GCs were passive and did not generate dendritic spikes. Dendritic spike generation allowed sustained, but not transient, OFF-GCs to signal into action potential output the local motion of visual stimuli to produce a continuous wave of action potential firing in adjacent cells as images moved across the retina. Conversely, this representation was highly fragmented in transient OFF-GCs. Thus, a heterogeneity of dendritic excitability defines the computations executed by classes of GCs.

A novel approach to the functional classification of retinal ganglion cells.

Retinal neurons are remarkedly diverse based on structure, function and genetic identity. Classifying these cells is a challenging task, requiring multimodal methodology. Here, we introduce a novel approach for retinal ganglion cell (RGC) classification, based on pharmacogenetics combined with immunohistochemistry and large-scale retinal electrophysiology. Our novel strategy allows grouping of cells sharing gene expression and understanding how these cell classes respond to basic and complex visual scenes. Our approach consists of several consecutive steps. First, the spike firing frequency is increased in RGCs co-expressing a certain gene (Scnn1a or Grik4) using excitatory DREADDs (designer receptors exclusively activated by designer drugs) in order to single out activity originating specifically from these cells. Their spike location is then combined with post hoc immunostaining, to unequivocally characterize their anatomical and functional features. We grouped these isolated RGCs into multiple clusters based on spike train similarities. Using this novel approach, we were able to extend the pre-existing list of Grik4-expressing RGC types to a total of eight and, for the first time, we provide a phenotypical description of 13 Scnn1a-expressing RGCs. The insights and methods gained here can guide not only RGC classification but neuronal classification challenges in other brain regions as well.

Differential susceptibility of retinal ganglion cell subtypes against neurodegenerative diseases.

Retinal ganglion cells (RGCs) are essential to propagate external visual information from the retina to the brain. Death of RGCs is speculated to be closely correlated with blinding retinal diseases, such as glaucoma and traumatic optic neuropathy (TON). Emerging innovative technologies have helped refine and standardize the classification of RGCs; at present, they are classified into more than 40 subpopulations in mammals. These RGC subtypes are identified by a combination of anatomical morphologies, electrophysiological functions, and genetic profiles. Increasing evidence suggests that neurodegenerative diseases do not collectively affect the RGCs. In fact, which RGC subtype exhibits the strongest or weakest susceptibility is hotly debated. Although a consensus has not yet been reached, it is certain that assorted RGCs display differential susceptibility against irreversible degeneration. Interestingly, a single RGC subtype can exhibit various vulnerabilities to optic nerve damage in diverse injury models. Thus, elucidating how susceptible RGC subtypes are to various injuries can protect vulnerable RGCs from damage and improve the possibility of preventing and treating visual impairment caused by neurodegenerative diseases. In this review, we summarize in detail the progress and status quo of research on the type-specific susceptibility of RGCs and point out current limitations and the possible directions for future research in this field.

PTEN Expression Regulates Gap Junction Connectivity in the Retina.

Manipulation of the phosphatase and tensin homolog (PTEN) pathway has been suggested as a therapeutic approach to treat or prevent vision loss due to retinal disease. In this study, we investigated the effects of deleting one copy of Pten in a well-characterized class of retinal ganglion cells called α-ganglion cells in the mouse retina. In Pten+/– retinas, α-ganglion cells did not exhibit major changes in their dendritic structure, although most cells developed a few, unusual loop-forming dendrites. By contrast, α-ganglion cells exhibited a significant decrease in heterologous and homologous gap junction mediated cell coupling with other retinal ganglion and amacrine cells. Additionally, the majority of OFF α-ganglion cells (12/18 cells) formed novel coupling to displaced amacrine cells. The number of connexin36 puncta, the predominant connexin that mediates gap junction communication at electrical synapses, was decreased by at least 50% on OFF α-ganglion cells. Reduced and incorrect gap junction connectivity of α-ganglion cells will affect their functional properties and alter visual image processing in the retina. The anomalous connectivity of retinal ganglion cells would potentially limit future therapeutic approaches involving manipulation of the Pten pathway for treating ganglion cell degeneration in diseases like glaucoma, traumatic brain injury, Parkinson’s, and Alzheimer’s diseases.

Spike-triggered Clustering for Retinal Ganglion Cell Classification.

Retinal ganglion cells (RGCs), the retina’s output neurons, encode visual information through spiking. The RGC receptive field (RF) represents the basic unit of visual information processing in the retina. RFs are commonly estimated using the spike-triggered average (STA), which is the average of the stimulus patterns to which a given RGC is sensitive. Whereas STA, based on the concept of the average, is simple and intuitive, it leaves more complex structures in the RFs undetected. Alternatively, spike-triggered covariance (STC) analysis provides information on second-order RF statistics. However, STC is computationally cumbersome and difficult to interpret. Thus, the objective of this study was to propose and validate a new computational method, called spike-triggered clustering (STCL), specific for multimodal RFs. Specifically, RFs were fit with a Gaussian mixture model, which provides the means and covariances of multiple RF clusters. The proposed method recovered bipolar stimulus patterns in the RFs of ON-OFF cells, while the STA identified only ON and OFF RGCs, and the remaining RGCs were labeled as unknown types. In contrast, our new STCL analysis distinguished ON-OFF RGCs from the ON, OFF, and unknown RGC types classified by STA. Thus, the proposed method enables us to include ON-OFF RGCs prior to retinal information analysis.

Characterizing Retinal Ganglion Cell Responses to Electrical Stimulation Using Generalized Linear Models.

The ability to preferentially stimulate different retinal pathways is an important area of research for improving visual prosthetics. Recent work has shown that different classes of retinal ganglion cells (RGCs) have distinct linear electrical input filters for low-amplitude white noise stimulation. The aim of this study is to provide a statistical framework for characterizing how RGCs respond to white-noise electrical stimulation. We used a nested family of Generalized Linear Models (GLMs) to partition neural responses into different components—progressively adding covariates to the GLM which captured non-stationarity in neural activity, a linear dependence on the stimulus, and any remaining non-linear interactions. We found that each of these components resulted in increased model performance, but that even the non-linear model left a substantial fraction of neural variability unexplained. The broad goal of this paper is to provide a much-needed theoretical framework to objectively quantify stimulus paradigms in terms of the types of neural responses that they elicit (linear vs. non-linear vs. stimulus-independent variability). In turn, this aids the prosthetic community in the search for optimal stimulus parameters that avoid indiscriminate retinal activation and adaptation caused by excessively large stimulus pulses, and avoid low fidelity responses (low signal-to-noise ratio) caused by excessively weak stimulus pulses.

The quantization error in a Self-Organizing Map as a contrast and colour specific indicator of single-pixel change in large random patterns

The quantization error in a fixed-size Self-Organizing Map (SOM) with unsupervised winner-take-all learning has previously been used successfully to detect, in minimal computation time, highly meaningful changes across images in medical time series and in time series of satellite images. Here, the functional properties of the quantization error in SOM are explored further to show that the metric is capable of reliably discriminating between the finest differences in local contrast intensities and contrast signs. While this capability of the QE is akin to functional characteristics of a specific class of retinal ganglion cells (the so-called Y-cells) in the visual systems of the primate and the cat, the sensitivity of the QE surpasses the capacity limits of human visual detection. Here, the quantization error in the SOM is found to reliably signal changes in contrast or colour when contrast information is removed from or added to the image, but not when the amount and relative weight of contrast information is constant and only the local spatial position of contrast elements in the pattern changes. While the RGB Mean reflects coarser changes in colour or contrast well enough, the SOM-QE is shown to outperform the RGB Mean in the detection of single-pixel changes in images with up to five million pixels. This could have important implications in the context of unsupervised image learning and computational building block approaches to large sets of image data (big data), including deep learning blocks, and automatic detection of contrast change at the nanoscale in Transmission or Scanning Electron Micrographs (TEM, SEM), or at the subpixel level in multispectral and hyper-spectral imaging data.

The quantization error in a Self-Organizing Map as a contrast and colour specific indicator of single-pixel change in large random patterns

The quantization error in a fixed-size Self-Organizing Map (SOM) with unsupervised winner-take-all learning has previously been used successfully to detect, in minimal computation time, highly meaningful changes across images in medical time series and in time series of satellite images. Here, the functional properties of the quantization error in SOM are explored further to show that the metric is capable of reliably discriminating between the finest differences in local contrast intensities and contrast signs. While this capability of the QE is akin to functional characteristics of a specific class of retinal ganglion cells (the so-called Y-cells) in the visual systems of the primate and the cat, the sensitivity of the QE surpasses the capacity limits of human visual detection. Here, the quantization error in the SOM is found to reliably signal changes in contrast or colour when contrast information is removed from or added to the image, but not when the amount and relative weight of contrast information is constant and only the local spatial position of contrast elements in the pattern changes. While the RGB Mean reflects coarser changes in colour or contrast well enough, the SOM-QE is shown to outperform the RGB Mean in the detection of single-pixel changes in images with up to five million pixels. This could have important implications in the context of unsupervised image learning and computational building block approaches to large sets of image data (big data), including deep learning blocks, and automatic detection of contrast change at the nanoscale in Transmission or Scanning Electron Micrographs (TEM, SEM), or at the subpixel level in multispectral and hyper-spectral imaging data.

How Does Light Regulate Mood and Behavioral State?

The idea that light affects mood and behavioral state is not new. However, not much is known about the particular mechanisms and circuits involved. To fully understand these, we need to know what properties of light are important for mediating changes in mood as well as what photoreceptors and pathways are responsible. Increasing evidence from both human and animal studies imply that a specialized class of retinal ganglion cells, intrinsically photosensitive retinal ganglion cells (ipRGCs), plays an important role in the light-regulated effects on mood and behavioral state, which is in line with their well-established roles in other non-visual responses (pupillary light reflex and circadian photoentrainment). This paper reviews our current understanding on the mechanisms and paths by which the light information modulates behavioral state and mood.

Non-parametric Physiological Classification of Retinal Ganglion Cells in the Mouse Retina.

Retinal ganglion cells, the sole output neurons of the retina, exhibit surprising diversity. A recent study reported over 30 distinct types in the mouse retina, indicating that the processing of visual information is highly parallelised in the brain. The advent of high density multi-electrode arrays now enables recording from many hundreds to thousands of neurons from a single retina. Here we describe a method for the automatic classification of large-scale retinal recordings using a simple stimulus paradigm and a spike train distance measure as a clustering metric. We evaluate our approach using synthetic spike trains, and demonstrate that major known cell types are identified in high-density recording sessions from the mouse retina with around 1,000 retinal ganglion cells. A comparison across different retinas reveals substantial variability between preparations, suggesting pooling data across retinas should be approached with caution. As a parameter-free method, our approach is broadly applicable for cellular physiological classification in all sensory modalities.

Heterogeneity of retinogeniculate axon arbors.

The retinogeniculate synapse transmits information from retinal ganglion cells (RGC) in the eye to thalamocortical relay neurons in the visual thalamus, the dorsal lateral geniculate nucleus (dLGN). Studies in mice have identified genetic markers for distinct classes of RGCs encoding different features of the visual space, facilitating the dissection of RGC subtype-specific physiology and anatomy. In this study, we examine the morphological properties of axon arbors of the BD-RGC class of ON-OFF direction selective cells that, by definition, exhibit a stereotypic dendritic arbor and termination pattern in the retina. We find that axon arbors from the same class of RGCs exhibit variations in their structure based on their target region of the dLGN. Our findings suggest that target regions may influence the morphologic and synaptic properties of their afferent inputs.

Differences between morphological and electrophysiological retinal ganglion cell classes.

Retinal prostheses work by delivering electrical pulses to the surviving retinal neurons. A pattern of electrical stimulation can generate a perception of vision in blind patients. To improve efficacy of retinal implants, it is important to understand how different classes of retinal neurons respond to electrical stimulation and if a classification can be made based on the electrophysiological properties of neurons. We use previously recorded patch clamp data from retinal ganglion cells classified into morphological classes (A,B,C, D) and functional types (ON, OFF, ON-OFF). We use a machine learning technique to separate data based on the recorded electrophysiological parameters. Results show that the clusters discovered using the machine learning technique do not correspond to the morphological or functional classes used by neuroscientists.

A Fine-Scale Functional Logic to Convergence from Retina to Thalamus

Numerous well-defined classes of retinal ganglion cells innervate the thalamus to guide image-forming vision, yet the rules governing their convergence and divergence remain unknown. Using two-photon calcium imaging in awake mouse thalamus, we observed a functional arrangement of retinal ganglion cell axonal boutons in which coarse-scale retinotopic ordering gives way to fine-scale organization based on shared preferences for other visual features. Specifically, at the ∼6μm scale, clusters of boutons from different axons often showed similar preferences for either one or multiple features, including axis and direction of motion, spatial frequency, and changes in luminance. Conversely, individual axons could "de-multiplex" information channels by participating in multiple, functionally distinct bouton clusters. Finally, ultrastructural analyses demonstrated that retinal axonal boutons in a local cluster often target the same dendritic domain. These data suggest that functionally specific convergence and divergence of retinal axons may impart diverse, robust, and often novel feature selectivity to visual thalamus.

Automated Identification of Axonal Action Potentials in Microelectrode Array Recordings

Event Abstract Back to Event Automated Identification of Axonal Action Potentials in high-density CMOS Microelectrode Array Recordings David Hoffmann1, 2*, Miriam Reh1, 2 and Günther Zeck1 1 Natural and Medical Sciences Institute, Neurophysics, Germany 2 Graduate Training Centre of Neuroscience, International Max Planck Research School, Germany When recording or modulating activity in neural tissue, such as the retina, it is of great advantage to obtain online information about cell location, about their axons and putative synaptically connected partners [1]. In this work we present and evaluate an algorithm for automated classification of retinal ganglion cells and identification of their axons based on recordings with high-density CMOS-based MEAs (CMOS MEA 5000, MultiChannelSystems MCS GmbH) from ex vivo mouse retina. The algorithm, which is based on a convolutional neural network (CNN) computes spike triggered average (STAs) electrical image and performs a classification of the electrical images during the experiment. This may enable the user to selectively stimulate a cell at a specified position. Material and Methods Spike sorting and computation of STAs was done using the software CMOS-MEA-TOOLs, which is based on an cICA algorithm [2]. This software visualizes the STAs as short videos, which are used here for producing a ground truth data set. A total of 1309 neurons were hand labelled and used to train the CNN classifier. Input to the CNN are feature maps of the STAs, namely a Boolean indicating threshold crossing for each electrode of the MEA, the mean maximum cross-correlation and the minimal voltage values per electrode. For a single electrode i the mean maximum cross-correlation is calculated by computing the maximum cross-correlation of this electrode with all 8 adjacent electrodes and calculating the mean over these 8 values. The algorithm was implemented in python and the resulting package, called axonFinder, comes with methods for semi-automatic labelling, to increase the amount of available training data. With more training data the classification accuracy can be expected to increase further. Results The mean accuracy of the CNN over all data sets (n=15) is 0.76 ± 0.1 (mean ± standard deviation), which is somewhat lower than human accuracy 0.84 ± 0.09. Human accuracy was determined by one human expert classifying the cells based on the input feature maps of the CNN, each visualized as image. For 6 out of 15 data sets the CNN and the human accuracy is indistinguishable. We found that the true positive rate (classifying a cell correctly as RGC comprising an axon) is very similar between human (0.74 ± 0.15) and CNN (0.74 ± 0.16). Differences are observed for true negative rate. The lower accuracy of the CNN can partially be explained by occasional failure of the optimization, which is a result of the small training and, by chance, ill-chosen early stopping data set. Some of these problems can be solved using N-fold cross-validation. Indeed, we found that usage of N-fold cross-validation for training increases the accuracy of the CNN to 0.8 ± 0.07. Conclusion Here we demonstrated how axonal action potentials are automatically identified in CMOS MEA recordings. The method can be used offline for unequivocal separation of ganglion cells from possible displaced amacrine cells and online to select electrical stimulus positions. Conflict of Interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Acknowledgements This work is partially funded by the German Ministry for education and Research (BMBF, FKZ: 031L0059A). DH acknowledges support by the Graduate Training Centre Neuroscience, Tübingen. We thank Larissa Höfling and Florian Jetter for sharing their recorded data.

Shared and distinct retinal input to the mouse superior colliculus and dorsal lateral geniculate nucleus.

The mammalian retina conveys the vast majority of information about visual stimuli to two brain regions: the dorsal lateral geniculate nucleus (dLGN) and the superior colliculus (SC). The degree to which retinal ganglion cells (RGCs) send similar or distinct information to the two areas remains unclear despite the important constraints that different patterns of RGC input place on downstream visual processing. To resolve this ambiguity, we injected a glycoprotein-deficient rabies virus coding for the expression of a fluorescent protein into the dLGN or SC; rabies virus labeled a smaller fraction of RGCs than lipophilic dyes such as DiI but, crucially, did not label RGC axons of passage. Approximately 80% of the RGCs infected by rabies virus injected into the dLGN were colabeled with DiI injected into the SC, suggesting that many dLGN-projecting RGCs also project to the SC. However, functional characterization of RGCs revealed that the SC receives input from several classes of RGCs that largely avoid the dLGN, in particular RGCs in which 1) sustained changes in light intensity elicit transient changes in firing rate and/or 2) a small range of stimulus sizes or temporal fluctuations in light intensity elicit robust activity. Taken together, our results illustrate several unexpected asymmetries in the information that the mouse retina conveys to two major downstream targets and suggest that differences in the output of dLGN and SC neurons reflect, at least in part, differences in the functional properties of RGCs that innervate the SC but not the dLGN.

Retinofugal Projections from Melanopsin-Expressing Retinal Ganglion Cells Revealed by Intraocular Injections of Cre-Dependent Virus.

To understand visual functions mediated by intrinsically photosensitive melanopsin-expressing retinal ganglion cells (mRGCs), it is important to elucidate axonal projections from these cells into the brain. Initial studies reported that melanopsin is expressed only in retinal ganglion cells within the eye. However, recent studies in Opn4-Cre mice revealed Cre-mediated marker expression in multiple brain areas. These discoveries complicate the use of melanopsin-driven genetic labeling techniques to identify retinofugal projections specifically from mRGCs. To restrict labeling to mRGCs, we developed a recombinant adeno-associated virus (AAV) carrying a Cre-dependent reporter (human placental alkaline phosphatase) that was injected into the vitreous of Opn4-Cre mouse eyes. The labeling observed in the brain of these mice was necessarily restricted specifically to retinofugal projections from mRGCs in the injected eye. We found that mRGCs innervate multiple nuclei in the basal forebrain, hypothalamus, amygdala, thalamus and midbrain. Midline structures tended to be bilaterally innervated, whereas the lateral structures received mostly contralateral innervation. As validation of our approach, we found projection patterns largely corresponded with previously published results; however, we have also identified a few novel targets. Our discovery of projections to the central amygdala suggests a possible direct neural pathway for aversive responses to light in neonates. In addition, projections to the accessory optic system suggest that mRGCs play a direct role in visual tracking, responses that were previously attributed to other classes of retinal ganglion cells. Moreover, projections to the zona incerta raise the possibility that mRGCs could regulate visceral and sensory functions. However, additional studies are needed to investigate the actual photosensitivity of mRGCs that project to the different brain areas. Also, there is a concern of "overlabeling" with very sensitive reporters that uncover low levels of expression. Light-evoked signaling from these cells must be shown to be of sufficient sensitivity to elicit physiologically relevant responses.