Functional cell types in the mouse superior colliculus

Curation statements for this article:
  • Curated by eLife

    eLife logo

    eLife assessment

    This paper will be of importance to visual neuroscientists, in particular those interested in the functional organization of subcortical visual pathways. The work provides evidence for a much greater diversity of functional cell types in the mouse superior colliculus than previously suggested, and that the functional organization of cell types in the superior colliculus is distinct from that of the retina. These results are based on an impressive data set. However, the conclusions require additional support.

This article has been Reviewed by the following groups

Read the full article See related articles

Abstract

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.

Article activity feed

  1. Author Response

    Reviewer #2 (Public Review):

    This paper has collected an impressive data set of the visual response properties of neurons in the visual layers of the mouse superior colliculus. There are 3 main findings of the study. First, the authors identify 24 functional classes of neurons based on the clustering of each neuron's visual response properties. Second, unlike in the retina where each cell type is regularly spaced, functional classes in the superior colliculus appear to cluster near each other. Third, visual representation has a lower dimensionality in the superior colliculus compared to the retina. The dataset has the potential to support the conclusions of the paper, but further analysis is required to make the claims convincing.

    Strengths:

    The main strength of the paper is its impressive dataset of more than 5000 neurons from the visual layers of the superior colliculus. This data set includes recordings from both an interesting set of genetically labelled classes of cells and from a reasonably large portion of the superior colliculus. This dataset offers the opportunity to support the major claims of the paper. This includes i) the identification of 24 functional classes of neurons, ii) the intriguing possibility that functional classes form local patches within the superior colliculus and iii) that the representation of visual information in the superior colliculus has a lower dimensionality compared to the retina.

    Weaknesses:

    The weakness of the paper is that its main claims are not adequately supported by the presented data or analysis. First, support for the existence of 24 functional classes is not clear enough. Our major concern is that it is not clear that each class of neurons was distributed across different mice. Are certain cell types overrepresented in individual animals, or do you find examples of each cell type in most animals?

    The new Supplementary Figure 7G shows how individual mice contribute to the functional types for all neurons. Further, the new Supplementary Figure 12 shows the receptive field locations derived from recordings in each of the animals.

    In addition, it should be made explicit how the responses of each genetically labeled class of neurons are distributed among the 24 functional clusters.

    We have added a new Figure 5D to show this.

    Second, the analysis of the spatial clustering of functional cell types is not complete. Do the same functional clusters sample the same retinotopic locations in different mice? How are clusters of the functional type distributed in visual space?

    Please see our point-by-point responses below to the concerns.

    Third, the lower dimensionality of representation in the superior colliculus may be the result of selective projections of retinal ganglion cells, not all retinal ganglion cell types project to the superior colliculus. Please estimate the dimensionality of the visual representation of those retinal ganglion cell types that projects to the superior colliculus.

    Certainly part of the dimensionality reduction may come from the incomplete retino-geniculate projection; we have added discussion on this topic.

  2. eLife assessment

    This paper will be of importance to visual neuroscientists, in particular those interested in the functional organization of subcortical visual pathways. The work provides evidence for a much greater diversity of functional cell types in the mouse superior colliculus than previously suggested, and that the functional organization of cell types in the superior colliculus is distinct from that of the retina. These results are based on an impressive data set. However, the conclusions require additional support.

  3. Reviewer #1 (Public Review):

    In this work, Li & Meister provide an extensive description of functional cell types within the posterior-medial part of the mouse superior colliculus, corresponding to the upper lateral visual field. Presenting a battery of visual stimuli to head-fixed wild-type mouse lines, they use calcium imaging and subsequent clustering of functional responses to identify 24 functional cell types. Besides the comprehensive sampling of SC cell types, a major strength of the manuscript in my view is the direct comparison with the previously published in vitro RGC data. Overall, the manuscript and the associated data promise to be a valuable resource to experimentalists and computational neuroscientists comparing visual processing across the major processing stages of the mouse visual system. However, in the current form the manuscript still comes with some limitations. In my view, these are related to some parts, where statistical justifications for the conclusions are still missing, where the findings should be more strongly embedded into the current and past literature, and where more efforts should be made to relate the findings in the different mouse lines to those for the overall population.

  4. Reviewer #2 (Public Review):

    This paper has collected an impressive data set of the visual response properties of neurons in the visual layers of the mouse superior colliculus. There are 3 main findings of the study. First, the authors identify 24 functional classes of neurons based on the clustering of each neuron's visual response properties. Second, unlike in the retina where each cell type is regularly spaced, functional classes in the superior colliculus appear to cluster near each other. Third, visual representation has a lower dimensionality in the superior colliculus compared to the retina. The dataset has the potential to support the conclusions of the paper, but further analysis is required to make the claims convincing.

    Strengths:

    The main strength of the paper is its impressive dataset of more than 5000 neurons from the visual layers of the superior colliculus. This data set includes recordings from both an interesting set of genetically labelled classes of cells and from a reasonably large portion of the superior colliculus. This dataset offers the opportunity to support the major claims of the paper. This includes i) the identification of 24 functional classes of neurons, ii) the intriguing possibility that functional classes form local patches within the superior colliculus and iii) that the representation of visual information in the superior colliculus has a lower dimensionality compared to the retina.

    Weaknesses:

    The weakness of the paper is that its main claims are not adequately supported by the presented data or analysis. First, support for the existence of 24 functional classes is not clear enough. Our major concern is that it is not clear that each class of neurons was distributed across different mice. Are certain cell types overrepresented in individual animals, or do you find examples of each cell type in most animals? In addition, it should be made explicit how the responses of each genetically labeled class of neurons are distributed among the 24 functional clusters. Second, the analysis of the spatial clustering of functional cell types is not complete. Do the same functional clusters sample the same retinotopic locations in different mice? How are clusters of the functional type distributed in visual space? Third, the lower dimensionality of representation in the superior colliculus may be the result of selective projections of retinal ganglion cells, not all retinal ganglion cell types project to the superior colliculus. Please estimate the dimensionality of the visual representation of those retinal ganglion cell types that projects to the superior colliculus.

  5. Reviewer #3 (Public Review):

    How parallel retinal outputs are processed in recipient visual areas is largely unknown. The present paper tackles this issue in the mouse superior colliculus - a key target of retinal outputs. Calcium signals of SC neurons were measured in response to a set of stimuli known to differentiate retinal ganglion cells. The resulting responses were then clustered to identify distinct cell types. These measurements were repeated in several transgenic mice with specific subsets of SC neurons labeled. The experiments and analysis generally support the conclusions well. There are several places, however, where the work could be presented more clearly.