Analysis of rod/cone gap junctions from the reconstruction of mouse photoreceptor terminals
Curation statements for this article:-
Curated by eLife
Evaluation Summary:
This paper uses a powerful combination of imaging techniques to provide a thorough view of the structure of the gap junction network connecting rod and cone photoreceptors in the mouse retina. The main conclusion - that rod-cone coupling is much more prevalent than rod-rod or cone-cone coupling - is well supported by the data although some results require qualification. The main concern in review centers around the importance of this result beyond the retina community.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 and Reviewer #2 agreed to share their names with the authors.)
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Neuroscience (eLife)
Abstract
Electrical coupling, mediated by gap junctions, contributes to signal averaging, synchronization, and noise reduction in neuronal circuits. In addition, gap junctions may also provide alternative neuronal pathways. However, because they are small and especially difficult to image, gap junctions are often ignored in large-scale 3D reconstructions. Here, we reconstruct gap junctions between photoreceptors in the mouse retina using serial blockface-scanning electron microscopy, focused ion beam-scanning electron microscopy, and confocal microscopy for the gap junction protein Cx36. An exuberant spray of fine telodendria extends from each cone pedicle (including blue cones) to contact 40–50 nearby rod spherules at sites of Cx36 labeling, with approximately 50 Cx36 clusters per cone pedicle and 2–3 per rod spherule. We were unable to detect rod/rod or cone/cone coupling. Thus, rod/cone coupling accounts for nearly all gap junctions between photoreceptors. We estimate a mean of 86 Cx36 channels per rod/cone pair, which may provide a maximum conductance of ~1200 pS, if all gap junction channels were open. This is comparable to the maximum conductance previously measured between rod/cone pairs in the presence of a dopamine antagonist to activate Cx36, suggesting that the open probability of gap junction channels can approach 100% under certain conditions.
Article activity feed
-
-
Author Response:
Reviewer #1 (Public Review):
This paper uses a combination of confocal and electron microscopy to localize gap junctions in the outer retina. Electrical coupling between photoreceptors is an important aspect of retinal function, and past work provides (often indirect) evidence for rod-rod, rod-cone and cone-cone coupling. The work described here indicates that rod-cone coupling dominates. The combination of techniques is quite convincing and very elegant. My concerns are primarily about the appeal of the work to non-retina readers. Some of these concerns could be mitigated by a more accessible presentation of some of the results. Suggestions along these lines, and a few other minor issues, follow.
Introduction:
The introduction is a bit retina-centric. I think more needs to be done to explain how each type of …
Author Response:
Reviewer #1 (Public Review):
This paper uses a combination of confocal and electron microscopy to localize gap junctions in the outer retina. Electrical coupling between photoreceptors is an important aspect of retinal function, and past work provides (often indirect) evidence for rod-rod, rod-cone and cone-cone coupling. The work described here indicates that rod-cone coupling dominates. The combination of techniques is quite convincing and very elegant. My concerns are primarily about the appeal of the work to non-retina readers. Some of these concerns could be mitigated by a more accessible presentation of some of the results. Suggestions along these lines, and a few other minor issues, follow.
Introduction:
The introduction is a bit retina-centric. I think more needs to be done to explain how each type of coupling (rod-rod, rod-cone, cone-cone) could impact retinal processing, and why it is important to resolve which are present or dominant. One issue that could get emphasized is the difference between gap junctions between like cell types (presumably involved in lateral spread of signals, averaging, etc) and between unlike cells (potentially providing an alternate path for signal flow - as in the secondary rod pathway).
We have included new text in the introduction to address this issue. We have tried to provide background material of a general nature and we have included some introductory text about different types of gap junctions, as requested. We thank reviewer 1 for this helpful suggestion.
Cone-cone coupling:
It would be helpful to put the conclusions about rod-cone and cone-cone coupling together. The paragraph starting on line 585 is a bit confusing that way. It starts by summarizing evidence that blue cones are not coupled with red/green cones. But then (in mouse) all the cones are coupled to rods, so that specific exclusion of blue cones seems unlikely to hold. You come back to this a bit later in the discussion, and there indicate that there appears to be weak cone-cone coupling. Merging the text in those two locations might help. It might also help to make the (seemingly clear) prediction that blue and green cone signals in mouse will get mixed.
Thank you for pointing out that this section is not clear. It seems two different points are muddled: 1) Blue cones do not make gap junctions with other cones, perhaps to minimize spectral mixing: the evidence from primate and ground squirrel suggests that blue cones are not coupled to red/green cones or green cones. 2) In contrast, we find no evidence of color selectivity in rod/cone coupling: green cones and blue cones are both coupled to all nearby rods. Thus, rod signals can be injected into the downstream pathways of both blue and green cones.
We have rewritten the text and separated these points into separate paragraphs for clarity, as below.
Revised Text:
Blue cone pedicles are also coupled to rods.
In the cone networks of primate and ground squirrel retina, there is good evidence that blue cones are not coupled to neighboring red/green (primate) or green cones (ground squirrel) (Hornstein et al., 2004; Li and DeVries, 2004; O’Brien et al., 2012). In the primate retina, the telodendria of blue cones are few in number and too short to reach the neighboring red/green cones (O’Brien et al., 2012). Thus, blue cones appear to be electrically separated from other cones in these two species, perhaps to maintain spectral discrimination (Hsu et al., 2000). In the mouse retina, although the blue cones were identified by Behrens et al., (2016), we were unable to find any cone to cone gap junctions, regardless of color (see below).
In contrast to the selective connections between cones in some species, rods were coupled to both blue and green cones indiscriminately in the mouse retina (present work) and in primate retina (O’Brien et al., 2012). Blue cones, identified in confocal work by the presence of S-cone opsin, and in SBF-SEM by their connections with blue cone bipolar cells (Behrens et al., 2016; Nadal-Nicolás et al., 2020), and green cones both made telodendrial contacts at Cx36 clusters with all nearby rod spherules (Fig. 4). Thus, we find no evidence for color specificity in rod/cone coupling. In fact, a single rod spherule may be coupled to both blue and green cones (Fig. 5, supplement 5). Therefore, rod signals can pass via the secondary rod pathway into both blue and green cones and their downstream pathways. Considering blue cone circuits specifically, rod input to blue cone bipolar cells and downstream circuits is predicted via the secondary rod pathway, in addition to the previously reported primary rod pathway inputs from AII amacrine cells to blue cone bipolar cells (Field et al., 2009; Whitaker et al., 2021).
Relation to other circuits:
Are there implications of the present results for gap junctional coupling in other circuits that could be emphasized? Things like the open probability how strongly it can be modulated seem like points of general interest - but I don't have enough expertise to know if those are established facts on other systems. Some of that is touched on in the Discussion, but quite briefly.
In an effort to keep the discussion short, we have perhaps been too abrupt. We have added text to the discussion to include some general issues concerning gap junctions.
Location of Cx36:
Can you speculate on why Cx36 is generally located at the mouth of the synaptic opening in the rod spherule? This was a very clear result, but it was unclear (at least to me) if it was important.
This is an interesting topic and we have expanded the discussion to consider potential functions and mechanisms.
Added to discussion:
The position of rod/cone gap junctions, at the base of the rod spherule, close to the opening of the post-synaptic cavity, appears to be systematic in that the vast majority of rod/cone gap junctions occur at this site. We may speculate that gap junctions are localized with some of the same scaffolding proteins that occur at the rod synaptic terminal, but the functional significance of this repeated motif is unknown. In mutant mouse lines, where Cx36 has been deleted from either rods or cones, cone telodendria are still present and they still reach out to contact nearby rod spherules in the absence of rod/cone gap junctions. Therefore, the specificity of synaptic connections is not determined or maintained by the presence of Cx36 gap junctions.
Reviewer #2 (Public Review):
Previous studies demonstrate that modulation of gap junctional coupling in the outer plexiform layer of the mouse retina regulates the balance between sensitivity and resolution. The authors use optical and electron microscopy to structurally characterize this coupling. They find that gap junctional coupling in mouse OPL is produced by a dense meshwork of cone photoreceptor telodendrions that selectively innervate the rim surrounding the synaptic openings of rod photoreceptor spherules. The density of this coupling network is such that each cone is coupled to dozens of rods and each rod is coupled to multiple cones. Rod/rod and cone/cone gap junctions were not detected.
The combination of antibody labeling, reconstruction of the photoreceptor terminal network, and ultrastructural analysis provides a remarkably clear view of the gap junctional connectivity that constitutes the first stage of visual processing. A few results are only weakly supported due to sample size or technical limitations. However, the overall conclusions are well supported and the data is presented with unusual transparency. The map of the network organization of photoreceptor coupling generated here is an important contribution to visual science.
Optical imaging:
The quality of the confocal imaging is high and the images of the Cx36 distribution relative to rod spherules is convincing. There does seem to be a significant amount of processing in the images and a lack of background signal in antibody images. Whether this processing is due to the airy scan software or additional filtering and thresholding, it can be difficult to judge the distribution of signal in several images.
In general, there was no filtering or processing of any confocal images, except for adjusting brightness and contrast. However, we may have been over-zealous in reducing the background. Therefore, we have adjusted Figures 1 and 2 to include more background as requested, to enable the reader to better judge the specificity of the immunolabeling. In addition, we have prepared supplementary figures to show the individual channels with background, as well as the combined images, to be absolutely clear and transparent. Finally, for each confocal image, the confocal series from which it was derived has been archived and is publicly accessible.
Former Figure 1D, now Fig. 2D is an exception because it shows a 3D projection of the colocalization between a single EGFP labeled cone pedicle and Cx36. We have revised this figure, providing new 2D optical sections to show how the image was prepared, in addition to revising the final 3D projection, labeling it as a 3D projection with colocalized Cx36.
Electron microscopy:
The authors perform annotations on two previously acquired volume EM datasets. The first serial blockface EM dataset is relatively low resolution and lacks ultrastructural labeling but is used effectively to reconstruct the terminal morphology and points of contacts between photoreceptors. The second EM data set uses FIB SEM to obtain smaller voxel sizes from tissue stained in such a way that the darkened membranes of putative gap junctions are distinct from surrounding membrane. Most measures of gap junction number come from the ultrastructure free dataset. In isolation, counting of gap junctions in this type of image volume could be unreliable. However, comparing the putative gap junctions in this dataset to the morphology and distribution of Cx36 antibody clusters in the confocal imaging and the darkened plaques in the FIB SEM images greatly increases confidence that the network description of rod/cone gap junctional coupling is accurate.
Quantification:
Most quantification is presented with an unusually high degree of transparency, with scatterplots showing all data points, data source files showing the animals that data came from, and standard deviations being supplied in descriptive statistics. There are a few places where Ns are difficult to determine or the analysis is not quite clear. For several results, claims are made when the sample size is too small to be sufficiently confident. The reconstruction of 5 blue cones suggests that, overall, blue cones are not radically different from other cones in their terminal morphology or gap junctional coupling to rod spherules. Claims that the blue cones are identical to other cones in most measures or that their telodendrions are smaller, but not statistically smaller are not well supported by the sampling. Similarly, the fact that the 6 nearby cones closely analyzed for cone/cone gap junctions yield no junctions, strongly suggests that vast majority of gap junctions are cone/rod gap junctions. However, the sample is too small to argue that there could not be infrequent, atypical, or region-specific cone/cone gap junctions.
We have addressed the issues of blue cones and cone/cone coupling to soften our conclusions and explicitly point out the small numbers.
Estimate of open channels:
The authors estimate that 89% of gap junction channels are open during times of maximum rod/cone coupling and point out that this number is surprisingly high relative to previous estimates. However, this estimate appears to be subject to many significant potential errors. The estimate combines previous freeze fracture studies of the density of gap junctions from various species and various parts of the retina the measurements of the length and width of the gap junctions in the current study. Differences in tissue processing, density variation within and between systems, reconstruction error, and variation and error in the inputs to the model could all contribute to an underestimate of the total number of channels linking mouse rods and cones. Moreover, without an accounting of these issues, the real error bars on the range of possible open channels would seem to include both surprising and less surprising estimates of open gap junction fractions.
This is a major issue. In short, for the calculations of open probability, we have estimated the cumulative errors, added these numbers to the text and attached an appendix showing the statistical analysis. We have also added a section to the discussion to address the possible sources of error enumerated by reviewer 2.
Reviewer #3 (Public Review):
In the presented work, Ishibashi and colleagues combine immunohistochemistry, analysis of a publicly available large scale 3D EM dataset and smaller but more detailed newly acquired EM datasets to qualitatively and quantitatively study gap junctions of mouse rod and cone axon terminals. The existence of rod-to-cone gap junctions has been known before, but the use of larger 3D EM data allows to determine an average number of contacts as well as an estimate of the strength of gap junctions. This as well as the (very likely) exclusion of direct cone-to-cone coupling in the mouse as opposed to some other mammals are the main contributions of this paper and one more puzzle piece of the big picture of mouse retinal connectivity. However, while the findings are a valuable addition towards a complete picture of the connectivity in the mouse retina, the novelty of the findings is limited to the number of contacts per photoreceptor and gap junction sizes.
In my opinion, while the authors present a thorough analysis of their data, the manuscript in its current state has stylistic flaws on the motivational side. To me, abstract and introduction lack a motivation or stronger statement of relevance for this analysis. Similarly, while each individual analysis is discussed one by one, I'm missing a broader discussion of the implications of the findings for the field and possible directions for future research to highlight relevance for a broader readership.
Thank you for the positive comments. We have rewritten and added material to the Abstract, Introduction and Discussion in an attempt to explain the reasoning for this study and to explain the findings to a broader audience.
-
Evaluation Summary:
This paper uses a powerful combination of imaging techniques to provide a thorough view of the structure of the gap junction network connecting rod and cone photoreceptors in the mouse retina. The main conclusion - that rod-cone coupling is much more prevalent than rod-rod or cone-cone coupling - is well supported by the data although some results require qualification. The main concern in review centers around the importance of this result beyond the retina community.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 and Reviewer #2 agreed to share their names with the authors.)
-
Reviewer #1 (Public Review):
This paper uses a combination of confocal and electron microscopy to localize gap junctions in the outer retina. Electrical coupling between photoreceptors is an important aspect of retinal function, and past work provides (often indirect) evidence for rod-rod, rod-cone and cone-cone coupling. The work described here indicates that rod-cone coupling dominates. The combination of techniques is quite convincing and very elegant. My concerns are primarily about the appeal of the work to non-retina readers. Some of these concerns could be mitigated by a more accessible presentation of some of the results. Suggestions along these lines, and a few other minor issues, follow.
Introduction:
The introduction is a bit retina-centric. I think more needs to be done to explain how each type of coupling (rod-rod, …
Reviewer #1 (Public Review):
This paper uses a combination of confocal and electron microscopy to localize gap junctions in the outer retina. Electrical coupling between photoreceptors is an important aspect of retinal function, and past work provides (often indirect) evidence for rod-rod, rod-cone and cone-cone coupling. The work described here indicates that rod-cone coupling dominates. The combination of techniques is quite convincing and very elegant. My concerns are primarily about the appeal of the work to non-retina readers. Some of these concerns could be mitigated by a more accessible presentation of some of the results. Suggestions along these lines, and a few other minor issues, follow.
Introduction:
The introduction is a bit retina-centric. I think more needs to be done to explain how each type of coupling (rod-rod, rod-cone, cone-cone) could impact retinal processing, and why it is important to resolve which are present or dominant. One issue that could get emphasized is the difference between gap junctions between like cell types (presumably involved in lateral spread of signals, averaging, etc) and between unlike cells (potentially providing an alternate path for signal flow - as in the secondary rod pathway).
Cone-cone coupling:
It would be helpful to put the conclusions about rod-cone and cone-cone coupling together. The paragraph starting on line 585 is a bit confusing that way. It starts by summarizing evidence that blue cones are not coupled with red/green cones. But then (in mouse) all the cones are coupled to rods, so that specific exclusion of blue cones seems unlikely to hold. You come back to this a bit later in the discussion, and there indicate that there appears to be weak cone-cone coupling. Merging the text in those two locations might help. It might also help to make the (seemingly clear) prediction that blue and green cone signals in mouse will get mixed.
Relation to other circuits:
Are there implications of the present results for gap junctional coupling in other circuits that could be emphasized? Things like the open probability how strongly it can be modulated seem like points of general interest - but I don't have enough expertise to know if those are established facts on other systems. Some of that is touched on in the Discussion, but quite briefly.
Location of Cx36:
Can you speculate on why Cx36 is generally located at the mouth of the synaptic opening in the rod spherule? This was a very clear result, but it was unclear (at least to me) if it was important.
-
Reviewer #2 (Public Review):
Previous studies demonstrate that modulation of gap junctional coupling in the outer plexiform layer of the mouse retina regulates the balance between sensitivity and resolution. The authors use optical and electron microscopy to structurally characterize this coupling. They find that gap junctional coupling in mouse OPL is produced by a dense meshwork of cone photoreceptor telodendrions that selectively innervate the rim surrounding the synaptic openings of rod photoreceptor spherules. The density of this coupling network is such that each cone is coupled to dozens of rods and each rod is coupled to multiple cones. Rod/rod and cone/cone gap junctions were not detected.
The combination of antibody labeling, reconstruction of the photoreceptor terminal network, and ultrastructural analysis provides a …
Reviewer #2 (Public Review):
Previous studies demonstrate that modulation of gap junctional coupling in the outer plexiform layer of the mouse retina regulates the balance between sensitivity and resolution. The authors use optical and electron microscopy to structurally characterize this coupling. They find that gap junctional coupling in mouse OPL is produced by a dense meshwork of cone photoreceptor telodendrions that selectively innervate the rim surrounding the synaptic openings of rod photoreceptor spherules. The density of this coupling network is such that each cone is coupled to dozens of rods and each rod is coupled to multiple cones. Rod/rod and cone/cone gap junctions were not detected.
The combination of antibody labeling, reconstruction of the photoreceptor terminal network, and ultrastructural analysis provides a remarkably clear view of the gap junctional connectivity that constitutes the first stage of visual processing. A few results are only weakly supported due to sample size or technical limitations. However, the overall conclusions are well supported and the data is presented with unusual transparency. The map of the network organization of photoreceptor coupling generated here is an important contribution to visual science.
Optical imaging:
The quality of the confocal imaging is high and the images of the Cx36 distribution relative to rod spherules is convincing. There does seem to be a significant amount of processing in the images and a lack of background signal in antibody images. Whether this processing is due to the airy scan software or additional filtering and thresholding, it can be difficult to judge the distribution of signal in several images.
Electron microscopy:
The authors perform annotations on two previously acquired volume EM datasets. The first serial blockface EM dataset is relatively low resolution and lacks ultrastructural labeling but is used effectively to reconstruct the terminal morphology and points of contacts between photoreceptors. The second EM data set uses FIB SEM to obtain smaller voxel sizes from tissue stained in such a way that the darkened membranes of putative gap junctions are distinct from surrounding membrane. Most measures of gap junction number come from the ultrastructure free dataset. In isolation, counting of gap junctions in this type of image volume could be unreliable. However, comparing the putative gap junctions in this dataset to the morphology and distribution of Cx36 antibody clusters in the confocal imaging and the darkened plaques in the FIB SEM images greatly increases confidence that the network description of rod/cone gap junctional coupling is accurate.
Quantification:
Most quantification is presented with an unusually high degree of transparency, with scatterplots showing all data points, data source files showing the animals that data came from, and standard deviations being supplied in descriptive statistics. There are a few places where Ns are difficult to determine or the analysis is not quite clear.
For several results, claims are made when the sample size is too small to be sufficiently confident. The reconstruction of 5 blue cones suggests that, overall, blue cones are not radically different from other cones in their terminal morphology or gap junctional coupling to rod spherules. Claims that the blue cones are identical to other cones in most measures or that their telodendrions are smaller, but not statistically smaller are not well supported by the sampling. Similarly, the fact that the 6 nearby cones closely analyzed for cone/cone gap junctions yield no junctions, strongly suggests that vast majority of gap junctions are cone/rod gap junctions. However, the sample is too small to argue that there could not be infrequent, atypical, or region-specific cone/cone gap junctions.Estimate of open channels:
The authors estimate that 89% of gap junction channels are open during times of maximum rod/cone coupling and point out that this number is surprisingly high relative to previous estimates. However, this estimate appears to be subject to many significant potential errors. The estimate combines previous freeze fracture studies of the density of gap junctions from various species and various parts of the retina the measurements of the length and width of the gap junctions in the current study. Differences in tissue processing, density variation within and between systems, reconstruction error, and variation and error in the inputs to the model could all contribute to an underestimate of the total number of channels linking mouse rods and cones. Moreover, without an accounting of these issues, the real error bars on the range of possible open channels would seem to include both surprising and less surprising estimates of open gap junction fractions.
-
Reviewer #3 (Public Review):
In the presented work, Ishibashi and colleagues combine immunohistochemistry, analysis of a publicly available large scale 3D EM dataset and smaller but more detailed newly acquired EM datasets to qualitatively and quantitatively study gap junctions of mouse rod and cone axon terminals. The existence of rod-to-cone gap junctions has been known before, but the use of larger 3D EM data allows to determine an average number of contacts as well as an estimate of the strength of gap junctions. This as well as the (very likely) exclusion of direct cone-to-cone coupling in the mouse as opposed to some other mammals are the main contributions of this paper and one more puzzle piece of the big picture of mouse retinal connectivity. However, while the findings are a valuable addition towards a complete picture of the …
Reviewer #3 (Public Review):
In the presented work, Ishibashi and colleagues combine immunohistochemistry, analysis of a publicly available large scale 3D EM dataset and smaller but more detailed newly acquired EM datasets to qualitatively and quantitatively study gap junctions of mouse rod and cone axon terminals. The existence of rod-to-cone gap junctions has been known before, but the use of larger 3D EM data allows to determine an average number of contacts as well as an estimate of the strength of gap junctions. This as well as the (very likely) exclusion of direct cone-to-cone coupling in the mouse as opposed to some other mammals are the main contributions of this paper and one more puzzle piece of the big picture of mouse retinal connectivity. However, while the findings are a valuable addition towards a complete picture of the connectivity in the mouse retina, the novelty of the findings is limited to the number of contacts per photoreceptor and gap junction sizes.
In my opinion, while the authors present a thorough analysis of their data, the manuscript in its current state has stylistic flaws on the motivational side. To me, abstract and introduction lack a motivation or stronger statement of relevance for this analysis. Similarly, while each individual analysis is discussed one by one, I'm missing a broader discussion of the implications of the findings for the field and possible directions for future research to highlight relevance for a broader readership.
-
-