Asymmetric distribution of color-opponent response types across mouse visual cortex supports superior color vision in the sky

Reviewer #1 (Public Review):

Summary:

In this study, Franke et al. explore and characterize color response properties across primary visual cortex, revealing specific color opponent encoding strategies across the visual field. The authors use awake 2P imaging to define the spectral response properties of visual interneurons in layer 2/3. They find that opponent responses are more pronounced at photopic light levels, and that diversity in color opponent responses exists across the visual field, with green ON/ UV OFF responses more strongly represented in the upper visual field. This is argued to be relevant for the detection of certain features that are more salient when using chromatic space, possibly due to noise reduction. In the revised version, Franke et al. have addressed the potential pitfalls in the discussion, which is an important point for the non-expert reader. Thus, this study provides a solid characterization of the color properties of V1 and is a valuable addition to visual neuroscience research.

My remaining concerns are based more on the interpretation. I'm still not convinced by the statement "This type of color-opponency in the receptive field center of V1 neurons was not present in the receptive field center of retinal ganglion cells and, therefore, is likely computed by integrating center and surround information downstream of the retina." and I would suggest rewording it in the abstract.

As discussed previously and now nicely added to the discussion, it is difficult to make a direct comparison given the different stimulus types used to characterize the retina and V1 recordings and the different levels of adaptation in both tissues. I will leave this point to the discussion, which allows for a more nuanced description of the phenomenon. Why do I think this is important? In the introduction, the authors argue that "the discrepancy [of previous studies] may be due to differences in stimulus design or light levels." However, while different light levels can be tested in V1, this cannot be done properly in the retina with 2P experiments. To address this, one would have to examine color-opponency in RGC terminals in vivo, which is beyond the scope of this study. Addressing these latter points directly in the discussion would, in my opinion, only strengthen the study.

Reviewer #2 (Public Review):

Summary:

Franke et al. characterize the representation of color in the primary visual cortex of mice, highlighting how this changes across the visual field. Using calcium imaging in awake, head-fixed mice, they characterize the properties of V1 neurons (layer 2/3) using a large center-surround stimulation where green and ultra-violet colors were presented in random combinations. Clustering of responses revealed a set of functional cell-types based on their preference to different combinations of green and UV in their center and surround. These functional types were demonstrated to have different spatial distributions across V1, including one neuronal type (Green-ON/UV-OFF) that was much more prominent in the posterior V1 (i.e. upper visual field). Modelling work suggests that these neurons likely support the detection of predator-like objects in the sky.

Strengths:

The large-scale single-cell resolution imaging used in this work allows the authors to map the responses of individual neurons across large regions of the visual cortex. Combining this large dataset with clustering analysis enabled the authors to group V1 neurons into distinct functional cell types and demonstrate their relative distribution in the upper and lower visual fields. Modelling work demonstrated the different capacity of each functional type to detect objects in the sky, providing insight into the ethological relevance of color opponent neurons in V1.

Weaknesses:

While the study presents convincing evidence about the asymmetric distribution of color-opponent neurons in V1, the paper would greatly benefit from a more in-depth discussion of the caveats related to the conclusions drawn about their origin. This is particularly relevant regarding the conclusion drawn about the contribution of color opponent neurons in the retina. The mismatch between retinal color opponency and V1 color opponency could imply that this feature is not solely inherited from the retina, however, there are other plausible explanations that are not discussed here. Direct evidence for this statement remains weak.

In addition, the paper would benefit from adding explicit neuron counts or percentages to the quadrants of each of the density plots in Figures 2-5. The variance explained by the principal components does not capture the percentage of color opponent cells. Additionally, there appear to be some remaining errors in the figure legend and labels that have not been addressed (e.g. '??' in Fig 2 legend).

Overall, this study will be a valuable resource for researchers studying color vision, cortical processing, and the processing of ethologically relevant information. It provides a useful basis for future work on the origin of color opponency in V1 and its ethological relevance.

Reviewer #3 (Public Review):

This paper studies chromatic coding in mouse primary visual cortex. Calcium responses of a large collection of cells are measured in response to a simple spot stimulus. These responses are used to estimate chromatic tuning properties - specifically sensitivity to UV and green stimuli presented in a large central spot or a larger still surrounding region. Cells are divided based on their responses to these stimuli into luminance or chromatic sensitive groups. The results are interesting and many aspects of the experiments and conclusions are well done; several technical concerns, however, limit the support for several main conclusions,

Limitations of stimulus choice
The paper relies on responses to a large (37.5 degree diameter) modulated spot and surround region. This spot is considerably larger than the receptive fields of both V1 cells and retinal ganglion cells (it is twice the area of the average V1 receptive field). As a result, the spot itself is very likely to strongly activate both center and surround mechanisms, and responses of cells are likely to depend on where the receptive fields are located within the spot (and, e.g., how much of the true neural surround samples the center spot vs the surround region). Most importantly, the surrounds of most of the recorded cells will be strongly activated by the central spot. This brings into question statements in the paper about selective activation of center and surround (e.g. page 2, right column). This in turn raises questions about several subsequent analyses that rely on selective center and surround activation.

Comparison with retina
A key conclusion of the paper is that the chromatic tuning in V1 is not inherited from retinal ganglion cells. This conclusion comes from comparing chromatic tuning in a previously-collected data set from retina with the present results. But the retina recordings were made using a considerably smaller spot, and hence it is not clear that the comparison made in the paper is accurate. For example, the stimulus used for the V1 experiments almost certainly strongly stimulates both center and surround of retinal ganglion cells. The text focuses on color opponency in the receptive field centers of retinal ganglion cells, but center-surround opponency seems at least as relevant for such large spots. This issue needs to be described more clearly and earlier in the paper.

Limitations associated with ETA analysis
One of the reviewers in the previous round of reviews raised the concern that the ETA analysis may not accurately capture responses of cells with nonlinear receptive field properties such as On/Off cells. This possibility and whether it is a concern should be discussed.

Discrimination performance poor
Discriminability of color or luminance is used as a measure of population coding. The discrimination performance appears to be quite poor - with 500-1000 neurons needed to reliably distinguish light from dark or green from UV. Intuitively I would expect that a single cell would provide such discrimination. Is this intuition wrong? If not, how do we interpret the discrimination analyses?

Author response:

The following is the authors’ response to the original reviews.

We thank the reviewers for their thoughtful comments. We were pleased that they thought our study was "well crafted and written", "important", and that it provides a "valuable resource for researchers studying color vision". They also expressed several constructive criticisms, concerning – among other things – the lack of details regarding experimental procedures and analysis, the challenge in relating retinal data to cortical recordings, and consistency of results across animals. In response to the reviewers’ comments and following their suggestions, we performed additional analyses, and substantially revised the paper:

We added a section in the Discussion about "Limitations of the stimulus paradigm". In addition, we added a new Suppl. Figure that illustrates the effect of deconvolution of calcium traces on our results and clarified in the text why we use deconvolved signals for all analyses. The new Suppl. Figure also shows an additional analysis with a more conservative threshold of neuron exclusion.

We now clarify how retinal signaling relates to our cortical results and rewrote the text to be more conservative regarding our conclusions.

In addition, we added a new Suppl. Figure showing the key analyses from Figures 2 and 4 separately for each animal. We now mention consistency across animals in the Results section and clearly state which analyses were performed an data pooled across animals.

We are positive that these additions address the issues raised by the reviewers. Please find our point-by-point replies to all comments below.

eLife assessment

Franke et al. explore and characterize the color response properties in the mouse primary visual cortex, revealing specific color opponent encoding strategies across the visual field. The data is solid; however, the evidence supporting some conclusions and details about some procedures are incomplete. In its current form, the paper makes a useful contribution to how color is coded in mouse V1. Significance would be enhanced with some additional analyses and resolution of some technical issues.

We thank the reviewers for appreciating our manuscript and their thoughtful comments.

Referee 1 (Remarks to the Author):

Summary:

In this study, Franke et al. explore and characterize the color response properties across the primary visual cortex, revealing specific color opponent encoding strategies across the visual field. The authors use awake-behaving 2P imaging to define the spectral response properties of visual interneurons in Layer 2/3. They find that opponent responses are more prominent at photopic light levels, and diversity in color opponent responses exists across the visual science, with green ON/ UV OFF responses being stronger represented in the upper visual field. This is argued to be relevant for detecting certain features that are more salient when the chromatic space is used, possibly due to noise reductions.

Strengths:

The work is well crafted and written and provides a thorough characterization that reveals an uncharacterized diversity of visual properties in V1. I find this characterization important because it reveals how strongly chromatic information can modulate the response properties in V1. In the upper visual field, 25% of the cells differentially relay chromatic information, and one may wonder how this information will be integrated and subsequently used to aid vision beyond the detection of color per see. I personally like the last paragraph of the discussion that highlights this fact.

We thank the reviewer for appreciating our manuscript.

Weaknesses: One major point highlighted in this paper is the fact that Green ON/UV OFF responses are not generated in the retina. But glancing through the literature, I saw this is not necessarily true. Fig 1. of Joesch and Meister, a paper cited, shows this can be the case. Thus, I would not emphasize that this wasn’t present in the retina. This is a minor point, but even if the retina could not generate these signals, I would be surprised if the diversity of responses would only arise through feed-forward excitation, given the intricacies of cortical connectivity. Thus, I would argue that the argument holds for most of the responses seen in V1; they need to be further processed by cortical circuitries.

We thank the reviewer for this comment. When analyzing available data from the retina using a similar center-surround color flicker stimulus (Szatko et al. 2020), we found that Green On/UV Off color opponency is very rare in the RF center of retinal ganglion cells (Suppl. Fig. 5). This suggests that center Green On/UV Off color opponency in V1 neurons is not inherited by the RF center of retinal neurons. However, we agree with the reviewer that retinal neurons might still contribute to V1 color opponency, for example by being center-surround color opponent (e.g. Joesch et al. 2016 and Szatko et al. 2020). We rephrased the text to acknowledge this fact.

This takes me to my second point, defining center and surround. The center spot is 37.5 deg of visual angle, more than 1 mm of the retinal surface. That means that all retinal cells, at least half and most likely all of their surrounds will also be activated. Although 37.5 deg is roughly the receptive field size previously determined for V1 neurons, the one-to-one comparison with retinal recording, particularly with their center/surround properties, is difficult. This should be discussed. I assume that the authors tried a similar approach with sparse or dense checker white noise stimuli. If so, it would be interesting if there were better ways of defining the properties of V1 neurons on their complex/simple receptive field properties to define how much of their responses are due to an activation of the true "center" or a coactivation of the surround. Interestingly, at least some of the cells (Fig. 1d, cells 2 and 5) don’t have a surround. Could it be that in these cases, the "center" and "surround" are being excited together? How different would the overall statistics change if one used a full-filed flicker stimulus instead of a center/surround stimulus? How stable are the results if the center/surround flicker stimulus is shifted? These results won’t change the fact that chromatic coding is present in the VC and that there are clear differences depending on their position, but it might change the interpretation. Thus, I would encourage you to test these differences and discuss them.

Thanks for this comment. We agree with the reviewer that a one-to-one comparison of retina and V1 data is challenging, due to differences in both RF and stimulus size. We rephrased the Results text to clarify this point and now also mention it in the Discussion.

To be able to record from many V1 neurons simultaneously, we used a stimulus size of 37.5 degree visual angle in diameter, which is slightly larger than center RFs of single V1 neurons. As the reviewer mentions, the disadvantage of this approach is that the stimulus is only roughly centered on the neurons’ center RFs. To reduce the impact of potential stimulus misalignment on our results, we used the following steps:

For each recording, we positioned the monitor such that the mean RF across all neurons lies within the center of the stimulus field of view.

We confirmed that this procedure results in good stimulus alignment for the large majority of recorded neurons within individual recording fields by using a sparse noise stimulus (Suppl. Fig. 1a-c). Specifically, we found that for 83% of tested neurons, more than two thirds of their center RF, determined by the sparse noise stimulus, overlapped with the center spot of the color noise stimulus.

For analysis, we excluded neurons without a significant center STA, which may be caused by misalignment of the stimulus.

Together, we believe these points strongly suggest that the center spot and the surround annulus of the noise stimulus predominantly drive center (i.e. classical RF) and surround (i.e. extraclassical RF), respectively, of the recorded V1 neurons. This is further supported by the fact that color response types identified using an automated clustering method were robust across mice (Suppl. Fig. 6c), indicating consistent stimulus centering.

Nevertheless, we cannot exclude that the stimulus was misaligned for a subset of the recorded neurons used for analysis. We agree with the reviewer that such misalignment might have contributed to cells not having surround STAs, due to simultaneous activation of antagonistic center and surround RF components by the surround stimulus. While a full-field stimulus would get rid of the misalignment problem, it would not allow to study color tuning in center and surround RF components separately. Instead, one could compare the results of our approach with an approach that centers the stimulus on individual neurons. However, we believe that performing these additional experiments is out of the scope of the current study.

To acknowledge the experimental limitations of our study and the concerns brought up by the reviewer, we now explicitly mention the steps we perform to reduce the effects of stimulus misalignment in the Results section and discuss the problem of stimulus alignment in the Discussion. We believe these changes will help the reader to interpret our results.

Referee 2 (Remarks to the Author):

Summary:

Franke et al. characterize the representation of color in the primary visual cortex of mice and how it changes across the visual field, with a particular focus on how this may influence the ability to detect aerial predators. Using calcium imaging in awake, head-fixed mice, they characterize the properties of V1 neurons (layer 2/3) using a large center-surround stimulation where green and ultra-violet were presented in random combinations. Using a clustering approach, a set of functional cell-types were identified based on their preference to different combinations of green and UV in their center and surround. These functional types were demonstrated to have varying spatial distributions in V1, including one neuronal type (Green-ON/UV-OFF) that was much more prominent in the posterior V1 (i.e. upper visual field). Modelling work suggests that these neurons likely support the detection of predator-like objects in the sky.

Strengths:

The large-scale single-cell resolution imaging used in this work allows the authors to map the responses of individual neurons across large regions of the visual cortex. Combining this large dataset with clustering analysis enabled the authors to group V1 neurons into distinct functional cell types and demonstrate their relative distribution in the upper and lower visual fields. Modelling work demonstrated the different capacity of each functional type to detect objects in the sky, providing insight into the ethological relevance of color opponent neurons in V1.

We thank the reviewer for appreciating our manuscript.

Weaknesses:

While the study presents solid evidence a few weaknesses exist, including the size of the dataset, clarity regarding details of data included in each step of the analysis and discussion of caveats of the work. The results presented here are based on recordings of 3 mice. While the number of neurons recorded is reasonably large (n > 3000) an analysis that tests for consistency across animals is missing. Related to this, it is unclear how many neurons at each stage of the analysis come from the 3 different mice (except for Suppl. Fig 4).

Thank you for this comment. We apologize that the original manuscript did not clearly indicate the consistency of our results across animals. We have revised the manuscript in the following ways:

We have added an additional Suppl. Figure, which shows the variability of the data within and across animals (Suppl. Fig. 4). Specifically, we show the distribution of color and luminance selectivity for (i) center and surround components of V1 RFs and (ii) for upper and lower visual field. This data is used for all analyses shown in Figures 2-4. The figure legend of this figure also states the number of neurons per animal.

We now clearly state in the Results section that all analyses in the main figures were performed by pooling data across animals, and refer to the Suppl. Figures for consistency across animals.

We believe these changes help the reader to interpret our results.

Finally, the paper would greatly benefit from a more in depth discussion of the caveats related to the conclusion drawn at each stage of the analysis. This is particularly relevant regarding the caveats related to using spike triggered averages to assess the response preferences of ON-OFF neurons, and the conclusions drawn about the contribution of retinal color opponency.

Thanks. We substantially revised the text to discuss caveats and limitations of the approach. For example, we added a section into the Discussion called "Limitations of the stimulus paradigm". In addition, we clarified how retinal signals relate to cortical ones and phrased our conclusions more conservatively.

The authors provide solid evidence to support an asymmetric distribution of color opponent cells in V1 and a reduced color contrast representation in lower light levels. Some statements would benefit from more direct evidence such as the integration of upstream visual signals for color opponency in V1.

Based on the comments from Reviewer 1, we have rephrased the statements regarding the integration of upstream visual signals for color opponency in V1. We think these revisions increase the clarity of the results and help the reader with interpretation.

Overall, this study will be a valuable resource for researchers studying color vision, cortical processing, and the processing of ethologically relevant information. It provides a useful basis for future work on the origin of color opponency in V1 and its ethological relevance.

Thanks! We thank the reviewer again for the helpful comments.

Referee 3 (Remarks to the Author):

This paper studies chromatic coding in mouse primary visual cortex. Calcium responses of a large collection of cells are measured in response to a simple spot stimulus. These responses are used to estimate chromatic tuning properties - specifically sensitivity to UV and green stimuli presented in a large central spot or a larger still surrounding region. Cells are divided based on their responses to these stimuli into luminance or chromatic sensitive groups. Several technical concerns limit how clearly the data support the conclusions. If these issues can be fixed, the paper would make a valuable contribution to how color is coded in mouse V1.

We thank the reviewer for the helpful comments.

Analysis: The central tool used to analyze the data is a "spike triggered average" of the responses to randomly varying stimuli. There are several steps in this analysis that are not documented, and hence evaluating how well it works is difficult. Central to this is that the paper does not measure spikes. Instead, measured calcium traces are converted to estimated spike rates, which are then used to estimate STAs. There are no raw calcium traces shown, and the approach to estimate spike rates is not described in any detail. Confirming the accuracy of these steps is essential for a reader to be able to evaluate the paper. Further, it is not clear why the linear filters connecting the recorded calcium traces and the stimulus cannot be estimated directly, without the intermediate step of estimating spike rates.

Thank you for this comment. We have used the genetically encoded calcium sensor GCaMP6s in our recordings. This sensor is a very sensitive GCaMP6 variant, but also one with slow kinetics. To remove the effect of the slow sensor kinetics from recorded calcium responses, the recorded traces are commonly deconvolved with the impulse function of the sensor to obtain the deconvolved calcium traces. We now include this reasoning in the Results section. To illustrate the effect of the deconvolution, we added a new Suppl. Figure (Suppl. Fig. 2) showing raw calcium and deconvolved traces, and the STAs estimated from both types of traces. This illustrates that the results regarding neuronal color preferences are consistent across raw and deconvolved calcium traces.

We agree with the reviewer that the term STA might be confusing. We have replaced it with the term "even-triggered-average" (ETA). In addition, we have replaced the phrase "estimated spike rate" with "deconvolved calcium trace" throughout the manuscript because the unit of the deconvolved traces is not interpretable, like spike rate would be (spikes per second). In the revised version, we now clarify in the Methods section that we estimate the ETAs based on deconvolved calcium traces, which is correlated with and an approximation for spike rate.

A further issue about the STAs is that the inclusion criterion (correlation of predicted vs measured responses of 0.25) is pretty forgiving. It would be helpful to see a distribution of those correlation values, and some control analyses to check whether the STA is providing a sufficiently accurate measure to support the results (e.g. do the central results hold for the cells with the highest correlations).

We thank the reviewer for this comment. To exclude noisy neurons from analysis, we used the following procedure:

For each of the four stimulus conditions (center and surround for green and UV stimuli), kernel quality was measured by comparing the variance of the STA with the variance of the baseline, defined as the first 500 ms of the STA. Only cells with at least 10-times more variance of the kernel compared to baseline for UV or green center STA were considered for further analysis.

We have added the distribution of quality values to a new Suppl. Figure (Suppl. Fig. 2d,e). We now also show the percentage of neurons above threshold, given different quality thresholds. Finally, we have repeated the analysis shown in Figure 2 for a much more conservative threshold, including only the top 25% of neurons (Suppl. Fig. 2e,f). We now mention this new analysis in the Methods and Results section.

Limitations of stimulus choice: The paper relies on responses to a large (37.5 degree diameter) modulated spot and surrounding region. This spot is considerably larger than the receptive fields of both V1 cells and retinal ganglion cells. As a result, the spot itself is very likely to strongly activate both center and surround mechanisms, and responses of cells are likely to depend on where the receptive fields are located within the spot (and, e.g., how much of the true neural surround samples the center spot vs the surround region). The impact of these issues on the conclusions is considered briefly at the start of the results but needs to be evaluated in considerably more detail. This is particularly true for retinal ganglion cells given the size of their receptive fields (see also next point).

We agree with the reviewer that the centering of the stimulus is critical and apologize if this point was not discussed sufficiently. To be able to record from many V1 neurons simultaneously, we used a stimulus size of 37.5 degree visual angle in diameter, which is slightly larger than center RFs of single V1 neurons. As the reviewer mentions, the disadvantage of this approach is that the stimulus is only roughly centered on the neurons’ center RFs. To reduce the impact of potential stimulus misalignment on our results, we have used different experimental and analysis steps and controls (see also second comment of Reviewer 1):

For each recording, we positioned the monitor such that the mean RF across all neurons lies within the center of the stimulus field of view.

We confirmed that this procedure results in good stimulus alignment for the large majority of recorded neurons within individual recording fields by using a sparse noise stimulus (Suppl. Fig. 1a-c). Specifically, we found that for 83% of tested neurons, more than two thirds of their center RF, determined by the sparse noise stimulus, overlapped with the center spot of the color noise stimulus.

For analysis, we excluded neurons without a significant center STA, which may be caused by misalignment of the stimulus.

We now mention those clearly in the Results section and added the limitations of our approach to the Discussion section.

Comparison with retina: A key conclusion of the paper is that the chromatic tuning in V1 is not inherited from retinal ganglion cells. This conclusion comes from comparing chromatic tuning in a previously-collected data set from retina with the present results. But the retina recordings were made using a considerably smaller spot, and hence it is not clear that the comparison made in the paper is accurate. This issue may be handled by the analysis presented in the paper, but if so it needs to be described more clearly. The paper from which the retina data is taken argues that rod-cone chromatic opponency originates largely in the outer retina. This mechanism would be expected to be shared across retinal outputs. Thus it is not clear how the Green-On/UV-Off vs Green-Off/UV-On asymmetry could originate. This should be discussed.

We agree with the reviewer that a one-to-one comparison of retina and V1 data is challenging, due to differences in both RF and stimulus size. We rephrased the Results text to clarify this point and now also mention it in the Discussion.

When analyzing available data from the retina using a similar center-surround color flicker stimulus (Szatko et al. 2020), we found that Green On/UV Off color opponency is very rare in the RF center of retinal ganglion cells (Suppl. Fig. 5). This suggests that center Green On/UV Off color opponency in V1 neurons is not inherited by the RF center of retinal neurons. However, we agree with the reviewer that retinal neurons might still contribute to V1 color opponency, for example by being center-surround color opponent (e.g. Joesch et al. 2016 and Szatko et al. 2020). We rephrased the text to acknowledge this fact.

Residual chromatic cells at low mesopic light levels The presence of chromatically tuned cells at the lowest light level probed is surprising. The authors describe these conditions as rod-dominated, in which case chromatic tuning should not be possible. This again is discussed only briefly. It either reflects the presence of an unexpected pathway that amplifies weak cone signals under low mesopic conditions such that they can create spectral opponency or something amiss in the calibrations or analysis. Data collected at still lower light levels would help resolve this.

Thank you for this comment. We call the lowest light level "low mesopic" and "rod-dominated" because the spectral contrast of V1 center responses in posterior recording fields is green-shifted for this light level (Fig. 3a). This is only expected if responses in the UV-cone dominant ventral retina are predominantly driven by rod photoreceptors. We now explain this rationale in the Results section. In addition, we mention in the Discussion that future studies are required to test whether cone signals need to be amplified for low light levels. While we agree with the reviewer that it would be exciting to use even lower light levels during recordings, we believe this is out of the scope of the current study due to the technical challenges involved in achieving scotopic stimulation.