Dendro-dendritic cholinergic excitation controls dendritic spike initiation in retinal ganglion cells

The retina processes visual images to compute features such as the direction of image motion. Starburst amacrine cells (SACs), axonless feed-forward interneurons, are essential components of the retinal direction-selective circuitry. Recent work has highlighted that SAC-mediated dendro-dendritic inhibition controls the action potential output of direction-selective ganglion cells (DSGCs) by vetoing dendritic spike initiation. However, SACs co-release GABA and the excitatory neurotransmitter acetylcholine at dendritic sites. Here we use direct dendritic recordings to show that preferred direction light stimuli evoke SAC-mediated acetylcholine release, which powerfully controls the stimulus sensitivity, receptive field size and action potential output of ON-DSGCs by acting as an excitatory drive for the initiation of dendritic spikes. Consistent with this, paired recordings reveal that the activation of single ON-SACs drove dendritic spike generation, because of predominate cholinergic excitation received on the preferred side of ON-DSGCs. Thus, dendro-dendritic release of neurotransmitters from SACs bi-directionally gate dendritic spike initiation to control the directionally selective action potential output of retinal ganglion cells.

Specific comments -My major concern is focused on the results shown in Figures 1d-1h. The On DSGCs in these figures are aligned with their preferred directions (upward). Activating SBACs located in the upper side (labeled "Preferred subfield" in Figure 1d) produced a net depolarization in On DSGCs, while activation of most SBACs on the lower side (labeled "Preferred subfield" in Figure 1d) produced a hyperpolarization (Figure 1h). This is directly contradictory to all previous studies on On-Off DSGCs and On DSGCs. It is well-accepted that the strongest inhibitory inputs onto DSGCs come from SBAC dendrites that extend to the null direction of DSGCs. The authors need to address this surprising discrepancy in their manuscript.
-The Abstract and Introduction needs to be re-worked to include an updated description of the current literature on this topic. For example, the authors claim that "it remains unknown, however, how SBAC-mediated cholinergic excitation is integrated together with bipolar-cell-mediated glumatergic excitation and SBAC-mediated inhibition to control the action potential output of DSGCs". In fact, two studies published earlier this year ; Poleg-Polsky and Diamond, 2016) addressed exactly this question, but are omitted in this manuscript. Kittila and Massey, 1995 also needs to be included in the references.
-In Discussion, page 12, the third paragraph, the "preferred-side" and "null-side" defined in this paper is opposite to those defined in previous studies. Thus, this paragraph needs to be re-written to address the discrepancy mentioned above. Page 15, last paragraph: "Support for these electrophysiological data is provided by ultrastructural analysis, which shows a clear distinction amongst the distribution of different classes of dendro-dendritic contacts in the dendritic arbor of ON-OFF DSGCs 17. Morphologically identified classical dendro-dendritic synapses have been shown to be focussed in the null-dendritic subtree, whereas other classes of close SBAC dendro-dendritic appositions are uniformly distributed throughout the dendritic tree of DSGCs 17, suggesting a structural correlate for the dendritic-field dependent balance of GABAergic and cholinergic signalling." These sentences sound vague to me: what did the authors mean by "the different classes of dendro-dendritic contacts"? Which cell type does "the null-dendritic subtree" refer to? Reference #17 by Briggman and colleagues demonstrates an antiparallel relationship between the orientation of presynaptic SBAC dendrites and the preferred direction of On-Off DSGCs, which is not consistent with the wiring pattern shown in Figure 1 of this paper.

Reviewer #2 (Remarks to the Author):
In their study, Brombas et al. show that cholinergic input from starburst amacrine cells (SACs) gates the initiation of dendritic spikes in ON direction-selective ganglion cells (oDSGCs) in the rabbit retina. Direction-selective circuits in the retina have been extensively studied. Most attention has focused on the asymmetric inhibitory input from SACs to DSGCs, which is required for direction selectivity. By comparison, less is know about the excitatory cholinergic input from SACs and the dendritic integration of both inputs by DSGCs. The present work explores these topics and reveals that cholinergic input from single SACs in the preferred subfield is sufficient to elicits spikes in oDSGCs. These spikes are blocked by cholinergic antagonists. The authors also show that pre-and postsynaptic interference with cholinergic transmission reduces oDSGC responses to visual motion stimuli and shrinks their suprathreshold receptive field. They then use direct dendritic recordings from oDSGCs to demonstrate that the effects of cholinergic transmission on somatic spiking are secondary to effects on dendritic spiking. In particular, electrical stimulation of single SACs in the preferred subfield or stimulation with small spots of light elicit dendritic spikes, which propagate to the soma of oDSGCs. Blockers of cholinergic transmission block these dendritic spikes as well as their somatic counterparts. Finally, the authors use local manipulations of cholinergic transmission, light stimuli, and dendritic current injections to show that acetylcholine gates the initiation of dendritic spikes in oDSGCs.
Overall, this is a very interesting study from an excellent group. The technical quality of the experiments is outstanding. Results are clearly illustrated in the figures and explained well in the text. However, the authors need to include recent findings on cholinergic input to DSGCs in the introduction and discussion of their manuscript to put their findings in the appropriate context. In addition, they should further explore and note the influence of stimulus contrast on their results.
Specific comments: 1) The authors need to include in the introduction and discussion of their manuscript a recent study by Sethuramanujam et al (Neuron 2016), which revealed the importance of cholinergic transmission from SACs in driving ooDSGC spiking under naturalistic and low contrast stimulus conditions.
2) The presentation of the data as cumulative response distributions in Fig4 was counterintuitive to me. The authors should add panels that present response data of oDSGCS and ooDSGCs as a function of stimulus contrast.
3) The authors show in Fig5 that the suprathreshold receptive field size shrinks in the presence of cholinergic blockers. The strength of this effect is likely dependent on the stimulus contrast. The authors should explore how the RF size changes as a function of stimulus contrast. 4) Given the finding of Sethuramanujam et al that cholinergic input is more dominant at low contrast compared to high contrast the authors should mention the contrast of their stimuli throughout this study. 4) A minor point is that I would encourage the authors to replace SBAC with the more widely used abbreviation SAC for starburst amacrine cell.

Reviewer #3 (Remarks to the Author):
The manuscript "Dendro-dendritic cholinergic excitation drives dendritic spike initiation in retinal ganglion cells" by Brombas et al investigates the effect of cholinergic transmission in the starburst amacrine cell to direction selective ganglion cell circuit in the rabbit retina.
The study reports somatic and dendritic patch-clamp recordings of starburst amacrine cells and direction-selective ganglion cells which are an important tool for precisely measuring synaptic signal transmission and details of electrical signal modulation in neurons. The study is technically of high quality and can be of substantial interest. However, I have a major concern that has essentially prevented me from understanding the scope and impact of this study. I was not able to understand how the provided data can be interpreted in the context of the now rather well-established synaptic circuitry in this tissue, namely the number of synaptic contacts between SBACs and DSGCs. This has been precisely established at the electron microscopic level and shown to be dependent on the angle of the SBAC's dendrite and the preferred direction of the DSGC . It has been shown that the number of synaptic contacts is about 13 fold higher in the null-direction than in the reversed direction. Based on this circuitry it is not clear how the results presented can be interpreted. In the discussion the authors mention that one may have to consider all the non-synaptic appositions as sites of cholinergic release. However, this then creates a picture of a rather unspecific background cholinergic release in the retina, where the inhibitory signals are the decisive modulators of computationally relevant synaptic transmission. If so, the statements about cholinergic excitation "driving", "powerfully augmenting" etc stimulus sensitivity should be phrased more cautiously. Moreover, when the authors refer to "dendro-dendritic" release, it should be made very clear upfront that this is nonsynaptic (if I understand correctly). As a constructive suggestion I would strongly advise to start with a presentation of the model of the system that is being investigated which builds on the known synaptic circuitry, non-synaptic appositions etc. This would allow the reader to put the presented data into a meaningful context. As it is now, one is left confused. Secondly, I was not able to extract from the figure legends what exactly is shown in the figures. Clear statements about whether neurons were stimulated using the patch pipettes or using light stimuli are missing, rather the interpretations are presented. Together with the missing reference circuit, this rendered me unable to fully judge the presented data. Moreover, when stimulated using visual stimuli, somatic recording in a neuron like SBAC (which is expected to have substantial dendritic nonlinear mechanisms, which may not be visible at the soma) it is much less informative.
This lack of precise description has made it virtually impossible to follow the results and determine whether what is shown is an interesting new result or just a speculation based on the notion of dendritic excitability.
Again I would like to emphasize it is well possible that these results are important. They add on to the data existing for the effect of cholinergic modulation for on-off DSGCs. Presenting physiology, however, without precise inclusion of the known circuits is in my view below standard.

Reviewer #1:
We thank the reviewer for highlighting the relevance of our work, and pinpointing that our work is the first to demonstrate that SAC-mediated cholingeric excitation is locally integrated in the dendritic tree of DSGCs to control dendritic spike generation. In the revised manuscript we have substantially revised the text, and improved the presentation of figures to directly address the reviewers specific comments.

Specific comments (Reviewers comments in italics)
1. My major concern is focused on the results shown in Figures 1d-1h. The On DSGCs in these figures are aligned with their preferred directions (upward). Activating SBACs located in the upper side (labeled "Preferred subfield" in Figure 1d) produced a net depolarization in On DSGCs, while activation of most SBACs on the lower side (labeled "Preferred subfield" in Figure 1d) produced a hyperpolarization (Figure 1h). This is directly contradictory to all previous studies on On-Off DSGCs and On DSGCs. It is well-accepted that the strongest inhibitory inputs onto DSGCs come from SBAC dendrites that extend to the null direction of DSGCs. The authors need to address this surprising discrepancy in their manuscript.
We thank the reviewer for highlighting this important point. We have clarified in the text of the revised manuscript the design and interpretation of these data, and have revised Figure 2 (formally Figure 1 panels d-h) to better illustrate our findings. We believe that our poor data illustration contributed to a misunderstanding. Previous work has highlighted the direction of light movement (see for example 1,2 ) to delineate SACs located on the preferred-and null-side of the dendritic arbor of ON-OFF-DSGCs. In our analysis we aligned morphologies by the preferred direction vector of action potential output, derived from polar plot analysis (Figure 2 and Supplementary Figure 2), which is customary in the field (see for example red lines fit to polar plots in 1 ). We, however, failed to adequately highlight that this vector is 180 degrees rotated to the direction of light bar movement. To highlight this we have added schematic light bars indicating the direction of light bar movement to the appropriate Figure  (Figure 2e), reworked the legend of Figure 2, and modified the appropriate text in the Results section (page 5 to 6 of the revised manuscript). We believe that our findings from ON-SAC to ON-DSGCs connections are in line with those previously reported for ON-OFF-DSGCs (page 6 of the revised manuscript). The results section now reads (page 5 to 6): We next enquired if the synaptic impact of ON-SACs aligned with the directional tuning of postsynaptic ON-DSGCs (Fig. 2). To do this we mapped the action potential output of ON-DSGCs generated by light bars moved across their receptive fields and made paired recordings from ON-SACs positioned close to the edge of the dendritic arbor first activated by light stimuli moving in a preferred direction, the preferred dendritic subfield, or ON-SACs positioned close to the edge of the dendritic arbor first activated by null direction light stimuli, the null dendritic subfield (Fig. 2a,b inset and Fig.  2c,d, inset, Supplementary Fig. 2; light bar size= 100 by 300-400 µm, moved in one of 12 directions at 0.24 mm / s; direction selectivity index (DSI)= 0.93 ± 0.03; SAC-DSGC soma separation= 270 ± 14 mm; n= 17). When the somata of ON-SACs were positioned close to the edge of the dendritic tree of ON-DSGCs first activated by preferred direction light stimuli, and the sites of close dendro-dendritic SAC-DSGC apposition restricted to the preferred dendritic subfield, net excitatory responses were generated ( Fig. 2a-b,e; PSP amplitude= 2.02 ± 0.37 mV, integral= 31.4 ± 4.3 mV.s; n= 6 pairs). In contrast, when the somata and sites of dendro-dendritic apposition were focussed on the null subfield, responses were on average inhibitory, but demonstrated a wide range (Fig. 2c-e; PSP amplitude= -0.59 ± 0.8 mV; integral= -33.7 ± 21.0 mV.s; n= 11 pairs; significantly different from preferred subfield: P= 0.035, T= 2.31; integral: P= 0.041, T= 2.24). Under current-clamp recording conditions, therefore the balance of SAC-mediated cholinergic excitation outweighs that of inhibition in the preferred dendritic subfield of ON-DSGCs, while inhibition outweighs excitation in the null dendritic subfield (Fig. 2e). This finding parallels the targeted dendritic impact of SAC-mediated excitation and inhibition reported for ON-OFF DSGCs 1-4 , and so reveals stereo-typed dendro-dendritic circuitry in the ON-and OFFsublamina of the inner plexiform layer.
2. The Abstract and Introduction needs to be re-worked to include an updated description of the current literature on this topic. For example, the authors claim that "it remains unknown, however, how SBAC-mediated cholinergic excitation is integrated together with bipolar-cell-mediated glumatergic excitation and SBAC-mediated inhibition to control the action potential output of DSGCs".
In fact, two studies published earlier this year Poleg-Polsky and Diamond, 2016) addressed exactly this question, but are omitted in this manuscript. Kittila and Massey, 1995 also needs to be included in the references.
We thank the reviewer for highlighting this important point. We agree that the recent work of Sethuramanujam et al and Poleg-Polsky & Diamond has shed light on the important contribution of cholinergic excitation as a gating mechanism of the voltage-dependent NMDA receptor-mediated conductances -activated by glutamate release from bipolar-cells. We have completely rewritten the Introduction to highlight this point and added careful comment in the Discussion section of our revised manuscript. In the Introductory section of the revised manuscript we indicate (page 3): SACs are, however, not simply feed-forward inhibitory interneurons, as ultrastructural and functional evidence indicates that both GABA and acetylcholine (ACh) are co-released by SAC to drive postsynaptic inhibition and excitation [1][2][3][4][5][6][7][8][9][10][11][12][13] . The physiological role of the co-released neurotransmitter ACh is however less well understood. Previous studies, using somatic recording techniques, have revealed that SAC-mediated excitation does not directly contribute to directional selectivity, but acts non-specifically to enhance the action potential output of DSGCs 12,14 . Consistent with this, during development the dendritic subfield-dependent strengthening of dendro-dendritic GABAergic inhibition occurs in parallel with the refinement of the symmetry of cholinergic signalling in DSGCs 4 . SACmediated cholinergic signalling has however been demonstrated to powerfully control the action potential output of DSGCs in response to time-varying visual stimuli, and act as an essential excitatory signal under mesopic, dim light, conditions 15,16 . Light-evoked feed-forward SAC-mediated cholinergic excitation may therefore function to provide local dendritic depolarization that acts to gate bipolar-cellmediated glutamatergic excitation of DSGCs 16 , which is largely mediated NMDA receptors 17 .
Furthermore, in the Discussion section we highlight this work (pages 15-16): We suggest that SAC-mediated cholinergic excitation acts as a dendritic subunit-specific gate of bipolar-cell signalling, that when activated by moving light bars powerfully controls the action potential output of DSGCs, to define the stimulus sensitivity and receptive field area over which the computation of direction selectivity is executed (Fig. 10). An idea supportive of previous observations made from ON-OFF-DSGCs, which have revealed that cholinergic excitation acts to control bipolarcell mediated glutamatergic excitation by supplying membrane depolarization to relieve the voltagedependent block of NMDA receptor-mediated PSPs 16 . Our findings however extend this idea to demonstrate that these excitatory inputs act in concert to drive the initiation of terminal dendritic spikes, and thus launch a cascade of dendritic spike generation which ultimately results in the driving of action potential output, providing a substrate for the localized dendritic processing of light stimuli 18 (Fig. 10).
We apologize for the omission of reference to , which has now been appropriately referenced in the revised manuscript.
3. In Discussion, page 12, the third paragraph, the "preferred-side" and "null-side" defined in this paper is opposite to those defined in previous studies. Thus, this paragraph needs to be re-written to address the discrepancy mentioned above.
Please see response to specific comment #1.
4. Page 15, last paragraph: "Support for these electrophysiological data is provided by ultrastructural analysis...." These sentences sound vague to me: what did the authors mean by "the different classes of dendro-dendritic contacts"? Which cell type does "the null-dendritic subtree" refer to? Reference #17 by Briggman and colleagues demonstrates an antiparallel relationship between the orientation of presynaptic SBAC dendrites and the preferred direction of On-Off DSGCs, which is not consistent with the wiring pattern shown in Figure 1 of this paper.
We thank the reviewer for raising this important issue, we have completely revised our discussion of previous ultrastuctural work. Careful analysis of the onset time of ON-SAC-evoked cholinergic and GABAergic postsynaptic potentials revealed a significant difference in onset time (revised Figure 1), that coupled with previous work 16,19 suggests that cholinergic signalling may occur in a local paracrine fashion in the IPL. We have revised our Discussion section to reflect this new analysis to form a more coherent interpretation of our data (page 17, revised manuscript): A structural substrate for the dual excitatory and inhibitory roles of SACs in the retinal circuitry is well documented 1-10, 12, 13 . The dendrites of ON-and OFF-SACs form dense choline acetyl transferase and vesicular ACh transporter delineated fascicles in the inner plexiform layer, the passage of which are closely followed by the dendrites of DSGCs [20][21][22] . Notably this mapping is most refined for small terminal DSGC dendrites, the predominant site of SAC-DSGC synaptic connectivity 23 . Ultrastructural analysis of murine ON-OFF DSGCs has, however, shown that the distribution of dendro-dendritic synapses is asymmetrical, with a ~13-fold greater density established by SACs positioned on the functional determined null dendritic subfield of ON-OFF-DSGCs 19 . Our results demonstrate in the simplified circuitry of the ON-sublamina of the inner plexiform layer that the postsynaptic impact of ON-SACs is position dependent, with cholinergic excitation dominating in the preferred, and GABAergic inhibition in the null-dendritic subtree, consistent with previous results from ON-OFF DSGCs 1-4, 24 . As our paired recordings, and previous results, have revealed the obligatory co-release of ACh and GABA from SACs, these data suggest that ACh may be released from, or have postsynaptic impact at, sites other than dendro-dendritic synapses, a finding that is supported by the relatively long time to onset of cholinergic PSPs. Furthermore, direct dendritic recordings revealed that light stimuli, and direct activation of SACs drove localized nAChR-dependent dendritic excitation, which was integrated locally in the dendritic arbor to control dendritic spike initiation. Our findings are therefore consistent with a model in which ACh operates in a localized paracrine fashion in the IPL 19 , at sites constrained by the co-fasciculation of SAC and DSGC dendrites. We therefore conclude that the excitatory and inhibitory neurotransmitters released from the dendrites of feed-forward interneurons in the output layer of the retina are integrated together with direct bipolar-cell-mediated glutamatergic excitation to drive and control the generation of dendritic spikes in DSGCs, the initiation of which determines the light stimulus sensitivity, receptive field size and direction selectivity of action potential output (Fig. 10).

Reviewer #2:
We thank the reviewer for their supportive comments indicating that "Overall, this is a very interesting study from an excellent group. The technical quality of the experiments is outstanding. Results are clearly illustrated in the figures and explained well in the text".
In the revised manuscript we have added a large body of new experimental data, substantially revised the text, and improved the presentation of figures to directly address the reviewers specific concerns.

Specific comments (Reviewers comments in italics)
1. The authors need to include in the introduction and discussion of their manuscript a recent study by Sethuramanujam et al (Neuron 2016), which revealed the importance of cholinergic transmission from SACs in driving ooDSGC spiking under naturalistic and low contrast stimulus conditions.
We thank the reviewer for highlighting this point we have revised the Introduction and Discussion sections of the manuscript to fully describe the recent Sethuramanujam paper. In the Introductory section of the revised manuscript we indicate (page 3): SACs are, however, not simply feed-forward inhibitory interneurons, as ultrastructural and functional evidence indicates that both GABA and acetylcholine (ACh) are co-released by SAC to drive postsynaptic inhibition and excitation [1][2][3][4][5][6][7][8][9][10][11][12][13] . The physiological role of the co-released neurotransmitter ACh is however less well understood. Previous studies, using somatic recording techniques, have revealed that SAC-mediated excitation does not directly contribute to directional selectivity, but acts non-specifically to enhance the action potential output of DSGCs 12,14 . Consistent with this, during development the dendritic subfield-dependent strengthening of dendro-dendritic GABAergic inhibition occurs in parallel with the refinement of the symmetry of cholinergic signalling in DSGCs 4 . SACmediated cholinergic signalling has however been demonstrated to powerfully control the action potential output of DSGCs in response to time-varying visual stimuli, and act as an essential excitatory signal under mesopic, low contrast, conditions 15,16 . Light-evoked feed-forward SAC-mediated cholinergic excitation may therefore function to provide local dendritic depolarization that acts to gate bipolar-cell-mediated glutamatergic excitation of DSGCs 16 , which is largely mediated NMDA receptors 17 .
Furthermore, in the Discussion section we highlight this work (pages 15-16): We suggest that SAC-mediated cholinergic excitation acts as a dendritic subunit-specific gate of bipolar-cell signalling, that when activated by moving light bars powerfully controls the action potential output of DSGCs, to define the stimulus sensitivity and receptive field area over which the computation of direction selectivity is executed (Fig. 10). An idea supportive of previous observations made from ON-OFF-DSGCs, which have revealed that cholinergic excitation acts to control bipolarcell mediated glutamatergic excitation by supplying membrane depolarization to relieve the voltagedependent block of NMDA receptor-mediated PSPs 16 . Our findings however extend this idea to demonstrate that these excitatory inputs act in concert to drive the initiation of terminal dendritic spikes, and thus launch a cascade of dendritic spike generation which ultimately results in the driving of action potential output, providing a substrate for the localized dendritic processing of light stimuli 18 (Fig. 10).

The presentation of the data as cumulative response distributions in Fig 4 was counterintuitive to me. The authors should add panels that present response data of oDSGCS and ooDSGCs as a function of stimulus contrast.
We have revised the presentation of this material. In the revised manuscript (revised To examine if cholinergic excitation influenced the responsiveness of ON-DSGCs to light stimuli across a wide stimulus range, we systematically varied the intensity of light stimuli and their speed of motion to generate stimulus-response relationship under control conditions and in the presence of a nAChR antagonist ( Fig. 5a-b; intensity: 10 to 100% above background; speed: 0.04 to 0.9 mm / s). Under control conditions the directional selective output responses of ON-DSGCs emerged at the light stimulus threshold for action potential firing and were maintained across a 50-fold preferred direction firing range (Fig. 5c). The antagonism of nAChRs increased the threshold light intensity, constrained the speed of stimuli that generated action potential output, and attenuated action potential firing throughout the range of light stimuli ( Fig. 5a-b). Notably, this compression of dynamic range was not accompanied by a degradation of the fidelity of the computation of direction selectivity ( Fig. 5c; DSI: control= 0.946 ± 0.014; Hex= 0.990 ± 0.004; n= 70 trials, n= 9 cells). To further explore the light range over which SAC-mediated cholinergic signalling controlled the responsiveness of ON-DSGCs, we made recordings in preparations adapted to mesopic light conditions (background illumination and stimulus intensity reduced to 0.06 of the standard photopic levels). Under these dim light conditions antagonism of nAChRs dramatically reduced the action potential output of ON-DSGCs evoked by preferred and null direction light bars (Supplementary Fig. 6; preferred direction: control= 15.1 ± 2.3 Hz; nAChR antagonist= 0.4 ± 0.1 Hz; P< 0.0001, T= 6.314; n= 11; null direction: control= 1.0 ± 0.3 Hz; nAChR antagonist= 0.04 ± 0.02 Hz; P= 0.0027, T= 3.961). Together these data show that cholinergic excitation acts to control the light stimulus-sensitivity and the magnitude of ON-DSGC output responses.

The authors show in Fig5 that the suprathreshold receptive field size shrinks in the presence of cholinergic blockers. The strength of this effect is likely dependent on the stimulus contrast. The authors should explore how the RF size changes as a function of stimulus contrast.
We thank the reviewer for this insight. We have undertaken a major series of experiments to address this concern. In our new data summarized in Figure 6 we have analysed the impact of cholinergic excitation on the relationship between dendritic field size and supra-threshold receptive field size across a broad range of light stimuli. In addition to the standard photopic 100% stimulus contrast condition reported in the original submission, we have used photopic 50% contrast stimuli, as well as mesopic, dim light stimuli (6% of the intensity of photopic stimuli). As presented in revised Figure 6 and in the Results section (page 9), blockade of nAChRs profoundly disrupted the wide field receptive field properties of ON-DSGCs to a similar degree under photopic (both 100% and 50% stimulus contrast) and mesopic light conditions (compare graphs in Fig. 6d-e), suggesting that SAC-mediated cholinergic signalling functions to powerful drive dendritic excitation across a broad physiological range of stimuli. We have revised our presentation of this material, now showing the displacement between first action potential generation and the edge of the preferred dendritic tree as averaged data and cumulative probability distributions for each trial. In addition we present new Supplementary material detailing the reduction of light responses evoked by mesopic light stimuli (Supplementary Figure 6). These changes are detailed in the revised manuscript (page 9): To gain insight into how SAC-mediated cholinergic excitation is integrated in postsynaptic ganglion cells we first determined if cholinergic signalling controlled the receptive field structure of ON-DSGCs. Under control conditions we found a tight spatial relationship between the size of the suprathreshold receptive field and the dendritic field size of ON-DSGCs across a wide-range of light conditions, consistent with previous reports 25 ( Fig. 6; receptive field size= 923 ± 24 µm; dendritic field size= 851 ± 23 µm, r= 0.574; P< 0.0001, n= 66 cells). Notably, when preferred direction light bars were swept across the receptive field a close relationship was found between the time at which light stimuli activated the retinal circuitry which innervated the edge of the preferred dendritic subfield of ON-DSGCs and the time of occurrence of the first light-evoked action potential, which could be converted into a spatial relationship (Fig. 6a-e; control: displacement= 19.8 ± 3.6 µm; n= 642 trials, n= 66 cells). The pharmacological blockade of nAChRs reduced action potential firing and delayed its onset as preferred direction light bars were swept across the receptive field, resulting in a disruption of the relationship between dendritic field edge and action potential generation, and a dramatic constriction of supra-threshold receptive field size (Fig. 6a-e; control= 923 ± 24 µm; nAChR antagonist= 471 ± 40 µm; P< 0.0001; T= 11.27; n= 66). Notably the blockade of SAC-mediated cholinergic excitation disrupted the wide field receptive field properties of ON-DSGCs to a similar degree under photopic and mesopic light conditions (compare graphs in Fig. 6d-e), suggesting that SAC-mediated cholinergic signalling functions to powerful drive dendritic excitation across a broad physiological range of stimuli.

Given the finding of Sethuramanujam et al that cholinergic input is more dominant at low contrast compared to high contrast the authors should mention the contrast of their stimuli throughout this study.
We thank the reviewer for highlighting this point, we have indicated stimulus intensity relative to background, as well as the adaptation state of the retina, throughout the revised manuscript.

5.
A minor point is that I would encourage the authors to replace SBAC with the more widely used abbreviation SAC for starburst amacrine cell.
We have used the abbreviation SAC for Starburst amacrine cell throughout the revised manuscript.

Reviewer #3:
We thank the reviewer for indicating that "The study is technically of high quality and can be of substantial interest".
We apologize to the reviewer for the poor description of our work in the original submission. We have undertaken a body of additional experimental work, and have rewritten large sections of the manuscript and clarified our illustrated materials to directly address the reviewers concerns, stated as "However, I have a major concern that has essentially prevented me from understanding the scope and impact of this study".
Specific comments (Reviewers comments in italics) 1. I was not able to understand how the provided data can be interpreted in the context of the now rather well-established synaptic circuitry in this tissue, namely the number of synaptic contacts between SBACs and DSGCs. This has been precisely established at the electron microscopic level and shown to be dependent on the angle of the SBAC's dendrite and the preferred direction of the DSGC  We thank the reviewer for highlighting these important issues. We have thoroughly revised the manuscript to directly address these concerns.
First we have rewritten the Introduction section of the manuscript to better describe the known SAC-DSGC circuitry in the inner plexiform layer of the retina, highlighting previous ultrastrutural work. These issues are summarised on pages 2 to 3 of the Introduction: The direction-selective action potential output of DSGCs is believed to be computed by the integration of a directionally un-tuned excitatory input, predominately mediated by glutamate release from bipolar cells 26-28 , with a directionally tuned inhibitory synaptic input generated by the dendritic release of GABA from axonless feed-forward interneurons, termed starburst amacrine cells (SACs) 3, 19, 29-31 . SACs represent essential components of the direction-selective circuitry of the retina 32 . Both ON-and OFF-SACs are distributed, in a mosaic throughout the retina 33 , and possess a unique radial dendritic morphology, in which each dendrite operates in electrical isolation 29, 34, 35 . Functionally, two-photon calcium imaging has revealed that SAC dendritic calcium responses are preferentially generated when light stimuli move from the soma toward terminal dendritic sites 29 , a direction-selective calcium signal that is thought to generate the dendritic release of neurotransmitters from terminal dendritic synaptic output zones 12,23,36 . The directional tuning of DSGCs is disrupted by the pharmacological antagonism of GABA A receptors 11,12 , evincing a prominent role of SAC-mediated synaptic inhibition. Consistent with this, electrophysiological and high-resolution morphological studies have demonstrated that the SAC-mediated inhibitory synaptic control of null direction light responses is mediated by a greater GABA A receptor-mediated synaptic conductance, synapse number, and distribution of dendro-dendritic synapses in the null dendritic subfield of DSGCs 1,3,4,19,24 . SACs are, however, not simply feedforward inhibitory interneurons, as ultrastructural and functional evidence indicates that both GABA and acetylcholine (ACh) are co-released by SAC to drive postsynaptic inhibition and excitation [1][2][3][4][5][6][7][8][9][10][11][12][13] . The physiological role of the co-released neurotransmitter ACh is however less well understood. Previous studies, using somatic recording techniques, have revealed that SAC-mediated excitation does not directly contribute to directional selectivity, but acts non-specifically to enhance the action potential output of DSGCs 12,14 . Consistent with this, during development the dendritic subfielddependent strengthening of dendro-dendritic GABAergic inhibition occurs in parallel with the refinement of the symmetry of cholinergic signalling in DSGCs 4 . SAC-mediated cholinergic signalling has however been demonstrated to powerfully control the action potential output of DSGCs in response to time-varying visual stimuli, and act as an essential excitatory signal under mesopic, low contrast, conditions 15,16 . Light-evoked feed-forward SAC-mediated cholinergic excitation may therefore function to provide local dendritic depolarization that acts to gate bipolar-cell-mediated glutamatergic excitation of DSGCs 16 , which is largely mediated NMDA receptors 17 .
Second, we have undertaken detailed analysis of the time course of ON-SAC evoked excitatory and inhibitory postsynaptic potentials, revealing that the time to onset of ON-SAC evoked cholinergic EPSPs is longer than those of GABAergic IPSPs. We interpret these findings in line with previous literature to indicate that cholinergic signalling occurs in a localised paracrine fashion in the inner plexiform layer. We have specifically addressed this point in Figure 1 of the revised manuscript, and describe these results in the Results section (pages 4-5): Pharmacologically isolated cholinergic excitatory and GABAergic inhibitory PSPs exhibited a characteristic difference in their time to onset following SAC activation, with the onset of GABAergic inhibitory PSPs leading that of cholinergic excitatory PSPs (Fig. 1f-h, Supplementary Fig. 1a-b; IPSP onset time= 6.6 ± 0.9 ms; EPSP onset time= 10.3 ± 1.3 ms, P= 0.0432; T= 2.295; onset time measured at 5% of PSP amplitude; IPSP rise time= 8.9 ± 1.9 ms; EPSP rise time= 5.8 ± 0.5 ms; rise time measured between 10-90% of PSP amplitude; n= 14). This difference could not be accounted for by a presynaptic mechanism, such as temporal jitter in the engagement of dendritic transmitter release, as the distribution of onset latencies around the mean was similar for pharmacologically isolated excitatory and inhibitory PSPs (Supplementary Fig. 1c). The onset time of SAC-evoked PSPs could however be influenced by the time course of transmitter diffusion from SAC dendritic sites. Consistent with this previous studies have suggested cholinergic signalling may occur in a paracrine fashion in the retina, and so may not be reliant on structural determined dendro-dendritic synapses 16,19 , an idea consistent with the observed slower onset time of cholinergic excitation. Functionally, the different onset of dendro-dendritic excitation and inhibition would be predicted to lead to the generation of biphasic compound PSPs, if both GABA and ACh are co-released from a single ON-SAC. Indeed when pharmacologically isolated excitatory and inhibitory components were arithmetically summed, a clear biphasic compound PSP was generated (Fig. 1h). Consistent with this, SAC-evoked PSPs recorded under control conditions frequently exhibited a biphasic waveform, that was clearly evident in single trials and when consecutive responses were digitally averaged ( Fig. 1d & h, respectively). Together, these data reveal that ACh and GABA are co-released from single ON-SACs.
Third, we have rewritten the Discussion section of the revised manuscript to indicate how our new data can be integrated together with previous electrophysiological and structural findings. The close of the Discussions section now reads (page 17): A structural substrate for the dual excitatory and inhibitory roles of SACs in the retinal circuitry is well documented 1-10, 12, 13 . The dendrites of ON-and OFF-SACs form dense choline acetyl transferase and vesicular ACh transporter delineated fascicles in the inner plexiform layer, the passage of which are closely followed by the dendrites of DSGCs [20][21][22] . Notably this mapping is most refined for small terminal DSGC dendrites, the predominant site of SAC-DSGC synaptic connectivity 23 . Ultrastructural analysis of murine ON-OFF DSGCs has, however, shown that the distribution of dendro-dendritic synapses is asymmetrical, with a ~13-fold greater density established by SACs positioned on the functional determined null dendritic subfield of ON-OFF-DSGCs 19 . Our results demonstrate in the simplified circuitry of the ON-sublamina of the inner plexiform layer that the postsynaptic impact of ON-SACs is position dependent, with cholinergic excitation dominating in the preferred, and GABAergic inhibition in the null-dendritic subtree, consistent with previous results from ON-OFF-DSGCs 1-4, 24 . As our paired recordings, and previous results, have revealed the obligatory co-release of ACh and GABA from SACs, these data suggest that ACh may be released from, or have postsynaptic impact at, sites other than dendro-dendritic synapses, a finding that is supported by the relatively long time to onset of cholinergic PSPs. Furthermore, direct dendritic recordings revealed that light stimuli, and direct activation of SACs drove localized nAChR-dependent dendritic excitation, which was integrated locally in the dendritic arbor to control dendritic spike initiation. Our findings are therefore consistent with a model in which ACh operates in a localized paracrine fashion in the IPL 19 , at sites constrained by the co-fasciculation of SAC and DSGC dendrites. We therefore conclude that the excitatory and inhibitory neurotransmitters released from the dendrites of feed-forward interneurons in the output layer of the retina are integrated together with direct bipolar-cell-mediated glutamatergic excitation to drive and control the generation of dendritic spikes in DSGCs, the initiation of which determines the light stimulus sensitivity, receptive field size and direction selectivity of action potential output (Fig. 10).

As a constructive suggestion I would strongly advise to start with a presentation of the model of the system that is being investigated which builds on the known synaptic circuitry, non-synaptic appositions etc. This would allow the reader to put the presented data into a meaningful context. As it is now, one is left confused.
We thank the reviewer for this suggestion, we have included a cartoon of the basic retinal circuitry activated by light stimuli in the revised manuscript (Figure 3a, inset). We have produced a schematic figure that summarizes our findings and interpretations, which we use to guide the reader through the Discussion section of the manuscript (Figure 10). Furthermore, we have modified the final paragraph of the Discussion section as noted in response to specific comment #1.

I was not able to extract from the figure legends what exactly is shown in the figures.
Clear statements about whether neurons were stimulated using the patch pipettes or using light stimuli are missing, rather the interpretations are presented. Together with the missing reference circuit, this rendered me unable to fully judge the presented data.
We have revised the presentation of the Figures throughout the revised manuscript, and rewritten the figure legends to more clearly indicate the illustrated material.

Moreover, when stimulated using visual stimuli, somatic recording in a neuron like SBAC (which is expected to have substantial dendritic nonlinear mechanisms, which may not be visible at the soma) it is much less informative.
We apologize to the reviewer for the lack of clarity in our original submission. All recordings from SACs were made from the somata. We did not directly investigate dendritic mechanisms of SACs. All presented simultaneous somato-dendritic recordings were made from ON-DSGCs. In some cases, this was complemented by presynaptic somatic recordings from SACs to demonstrate that ACh release from ON-SACs evokes localized dendritic EPSPs, which were capable of directly driving dendritic spike generation in postsynaptic ON-DSGCs. We have rewritten the Results section to clarify this issue (pages 16 to 17): To explore the determinants of this form of sub-cellular integration, we tested if cholinergic excitation provided by a single SAC was capable of driving the initiation of dendritic spikes in ON-DSGCs. To do this we made simultaneous somato-dendritic recordings from ON-DSGCs and a third somatic recording from a presynaptic ON-SAC, in the presence of the GABA A receptor antagonist GABAzine ( Fig. 7e-g; n= 7). Under these conditions the activation of a single SAC led to the generation of nAChR-mediated dendritic excitatory PSPs, which were crowned by the firing of dendritic spikes, in recordings where dendro-dendritic SAC-DSGC appositions were focussed in the recorded dendritic arbor of the ON-DSGCs at loci distal to the site of dendritic recording ( Fig. 7e-g; dendritic recordings 145 ± 5 mm from soma; average site of SAC dendro-dendritic appositions from DSGC soma= 437 ± 26 mm; n= 5). Simultaneous somatic recording revealed that each dendritic spike preceded and drove action potential firing (dendritic spike to action potential delay= 0.40 ± 0.05 ms). In contrast, when dendritic recordings were made from the dendritic subfield contralateral to the site of predominate SAC innervation, SAC-evoked cholinergic excitation drove action potential firing, which was first recorded somatically and subsequently back-propagated to the dendritic recording site, consistent with cholinergic excitation of the contralateral dendritic tree (Supplementary Fig. 8). Taken together these data directly demonstrate that dendro-dendritic cholinergic excitation is capable of driving the generation of dendritic spikes in ON-DSGCs.

This lack of precise description has made it virtually impossible to follow the results and determine whether what is shown is an interesting new result or just a speculation based on the notion of dendritic excitability. Again I would like to emphasize it is well possible that these results are important. They add on to the data existing for the effect of cholinergic modulation for on-off DSGCs. Presenting physiology, however, without precise inclusion of the known circuits is in my view below standard.
We thank the reviewer for this important comment. We have thoroughly revised the Introductory, Results and Discussion sections of the revised manuscript and carefully revised the main and supplementary figures to carefully explain and illustrate our findings. We hope that the new presentation, and Discussion of the local paracrine release of ACh helps to bridge the gap between our physiological findings and previous high-resolution circuit mapping experiments. Indeed, previous high-resolution circuit mapping studies have found it difficult to explain the role of SAC-mediated cholinergic signalling in the output layer of the retina, see discussion in 19 . We hope that our new analysis and more thorough discussion better explains the role of SAC-mediated cholinergic signalling in the retina, by demonstrating how local integration within the dendritic tree of ON-DSGCs controls this class of RGC action potential output. To avoid confusion between our study and previous work, we have removed all functional data concerning the role of cholinergic signalling in controlling the output properties of rabbit ON-OFF-DSGCs in the revised manuscript. We hope that our revised presentation and discussion of material, which is summarised in a new schematic diagram (Figure 10), place our new findings into a logical and understandable framework.

Reviewer #1 (Remarks to the Author):
This manuscript highlights the importance of cholinergic inputs from SACs in initiating local dendritic spikes in DSGCs. The combination of dendritic and somatic recordings from DSGCs and stimulation of SACs is technically elegant and impressive, and provides direct evidence of ACh-dependent dendritic spike generation in DSGCs. The authors have added new experiments and changed the text to address the points I have raised. However, I still have several concerns, mainly about the conceptual framework related to this study, and interpretation of the data.
1. The authors describe the SAC-DSGC connectivity pattern in terms of "preferred or null dendritic subfield of DSGCs". This is not consistent with the current models of direction selectivity. In the current models, at least for On-Off DSGCs, the asymmetric wiring between SACs and DSGCs arises from dendritic branches of SACs, not those of DSGCs. Indeed, the dendritic field of a DSGC is considered rather uniform along the preferred-null axis, and the entire dendritic tree receives inhibitory inputs from SAC dendrites oriented in the null direction of the DSGC. Therefore, the use of "preferred or null dendritic subfield of DSGCs" throughout the manuscript (e.g. in pages 2, 5, 14) is inappropriate. pairs to address this discrepancy. 5. Page 6 last line, it is unclear to me how "stereo-typed dendro-dendritic circuitry in the On-and Off-sublamina of the inner plexiform layer" is supported, because the asymmetric wiring between Off-SACs and DSGCs has not been directly demonstrated. 6. Fig. 5c, the authors showed that Hex reduces direction index, a commonly used measure of directional tuning in DSGCs, and thus concluded in the text that Hex does not affect "the fidelity of the computation of direction selectivity". However, I think a more careful interpretation of direction index is necessary. For example, if there is zero or very few spikes in the null direction in the control condition, the null direction firing cannot be reduced further by Hex, while the preferred direction firing is greatly reduced (as shown in Fig. 5c). In this scenario, the direction index is not changed, but the difference between null and preferred direction firing rates is dramatically altered, which could indeed be interpreted as a reduction in the fidelity of the computation of direction selectivity.
Minor points: Fig 1c and 1e? 8. Fig. 2b and d, the polar plots should be rotated 180 degrees so that the vector sums of the spiking responses point to the same preferred direction shown in Fig 2a and 2c.

Reviewer #2 (Remarks to the Author):
The authors have satisfactorily addressed my previous comments. Congratulations on a very nice study

Reviewer #3 (Remarks to the Author):
The authors have taken the concerns of the other reviewers and myself very seriously and have very successfully revised the manuscript. The description of the circuit context and the significance of the authors' findings has much improved. Also the figures are now clearly understandable, allowing a proper assessment of the presented data. As such this is a very successful revision and I can recommend publication without hesitation.
We thank the Reviewers of our manuscript for their comments. We note that Reviewer #2 and Reviewer #3 were entirely satisfied with our revised manuscript, indicating: Reviewer #2: "The authors have satisfactorily addressed my previous comments. Congratulations on a very nice study." Reviewer #3: "The authors have taken the concerns of the other reviewers and myself very seriously and have very successfully revised the manuscript. The description of the circuit context and the significance of the authors' findings has much improved. Also the figures are now clearly understandable, allowing a proper assessment of the presented data. As such this is a very successful revision and I can recommend publication without hesitation." We thank Reviewer #1 for their further interest in our work. Please find below our point-by-point reply.

Reviewer #1 Major concerns:
1. The authors describe the SAC-DSGC connectivity pattern in terms of "preferred or null dendritic subfield of DSGCs". This is not consistent with the current models of direction selectivity. In the current models, at least for On-Off DSGCs, the asymmetric wiring between SACs and DSGCs arises from dendritic branches of SACs, not those of DSGCs. Indeed, the dendritic field of a DSGC is considered rather uniform along the preferred-null axis, and the entire dendritic tree receives inhibitory inputs from SAC dendrites oriented in the null direction of the DSGC. Therefore, the use of "preferred or null dendritic subfield of DSGCs" throughout the manuscript (e.g. in pages 2,5,14) is inappropriate.
Previous work detailing the impact of SAC-mediated synaptic input to ON-OFF DSGCs have differentiated the preferred and null-sides of the dendritic arbor of ON-OFF DSGCs using an approach similar to the one illustrated and described in our manuscript [1][2][3][4] . We therefore consider the approach we have adopted to describe the dendritic field of ON-DSGCs to be standard in the field, and as so will allow a direct comparison of our work with that previously reported for ON-OFF DSGCs 1-4 .  We thank the reviewer for highlight the involvement of the SAC-mediated cholinergic signalling in the direction selective circuitry of the retina. We acknowledge that previous studies have demonstrated a variable contribution of cholinergic signalling to the light-evoked excitation of DSGCs. We also acknowledge that there exists controversy concerning the role of cholinergic signalling in the computation of direction selectivity. As this body of literature motivated our study, we have simplified the Introductory section of the revised manuscript to state as clearly, and concisely as possible existing work (page 3): The physiological role of the co-released neurotransmitter ACh is however less well understood, and controversy remains on the contribution of this feed-forward excitatory signal to the generation of light-evoked DSGC action potential output, and its role in the computation of direction selectivity 1, 2, 4-13 . We believe that these two points are of the upmost significance as we have previously experimentally demonstrated that somatic voltage-clamp techniques profoundly distorts the amplitude, reversal potential, kinetics and ability to separate temporally overlapping excitatory and inhibitory of dendritically generated synaptic input in central neurons (Williams and Mitchell, Nature Neurosci. 2008) 14 . Such direct experimental analysis of the distortions imposed by somatic voltage-clamp recording techniques have been confirmed in DSGCs by computer simulations (Poleg-Polsky & Diamond, Plos One, 2011) 15 .
We note that both these studies have demonstrated that the somatic voltage clamp does not universally control voltage in the dendritic tree, allowing escape potentials to be generated at the dendritic site of synaptic activation. This limitation of the somatic voltage clamp recording technique has a profound impact when both excitatory and inhibitory synaptic inputs are co-generated 14,15 -as is the case in the recordings of Lee et al. 2010 1 . We note that as Lee et al. 2010 1 employed somatic voltage clamp techniques, and used the control of the somatic holding potential to "isolate" SACevoked excitatory and inhibitory synaptic currents, considerable errors must be considered in their latency analysis, as at the dendritic site of synaptic activation the local membrane potential is not voltage controlled and so free to be charged by excitatory and inhibitory synaptic input, an effect that will distort latency measurement of "isolated" components. We therefore respectfully indicate: i) previous measurement of the latency of temporally overlapping SAC-evoked IPSCs and EPSCs recorded by Lee et al. 2010 1 are subject to considerable uncertainty, and ii) somatic voltage-clamp techniques do not represent an adequate tool to address this problem.
We contend that our findings illustrated in Figure 1 and Supplementary Figure 1 using current clamp recordings of pharmacologically isolated SAC-evoked EPSPs and IPSPs represents the most accurate method available to undertake latency analysis. We also note that we have demonstrated that a clear separation of latency is stable across many trials (Figure 1 panel f) and cannot be accounted for by other processes (Supplementary Fig. 1). Furthermore, such latency analysis yielded a prediction concerning the waveform of PSPs evoked under control conditions. Analysis of control SAC-evoked PSPs supported this prediction (Figure 1, panel h).
Evidence for paracrine ACh signalling in the literature: The reviewer indicates that: "In contrast, no evidence of paracrine SAC-DSGC signaling has been shown in the literature." We contend that there is substantive evidence for a local-paracrine action of ACh in the direction selective circuitry of the retina, to which our direct observations add. We note that discussion of this issue has been addressed in the most recent paper to explore SAC function in the retina, published in late 2016 in the journal Neuron 13 , a reference that the reviewer cites at other points in their review.
We note that these authors indicate that SAC signalling in the preferred subfield of ON-OFF-DSGCs to be paracrine (see Fig. 1B of 13 ). We further note that the Discussion section of this paper indicates (reproduced from 13 Ford et al., 2012;Schmidt et al., 1987). Paracrine transmission would allow DSGCs to pool cholinergic signals arising from many dendrites orientated in different directions. In contrast to ACh, the clearance of GABA from the synaptic cleft relies on strong uptake mechanisms that confine its action to the synapse. In this way, co-release of ACh and GABA by SACs could lead to distinct spatiotemporal patterns of activity at the level of the DSGC." Reproduced from 13 .
Indeed Ford et al. 2012 has provided direct evidential support for a role of local paracrine SAC-mediated cholinergic signalling in the developing retina 16 . When taken together with our new findings, which provide direct evidence in support of a local-paracrine action of ACh, we with respect strongly disagree with the reviewers point that "..no evidence of paracrine SAC-DSGC signaling has been shown in the literature." But concur with  that new tools are required to directly address this issue, we have therefore stated in the revised discussion section of the manuscript (page 25): Our findings are therefore consistent with a model in which ACh operates in a localized paracrine fashion in the IPL 13, 17 16 , at sites constrained by the cofasciculation of SAC and DSGC dendrites. The development of new tools for the direct visualization of ACh release and diffusion are, however, required to definitively address this issue.
We further note that our latency analysis, and detailed description of the structure of the IPL was called for by Reviewer # 3 in the first round of review, who indicated that they were entirely happy with the revised manuscript.

Page 6 last line, it is unclear to me how "stereo-typed dendro-dendritic circuitry in the On-and Off-sublamina of the inner plexiform layer" is supported, because the asymmetric wiring between Off-SACs and DSGCs has not been directly demonstrated.
Previous work has highlighted the direction-selectivity of OFF responses in DSGCs, and ultra structural data has highlighted asymmetric wring of OFF-SACs (see Figure  4 of Briggman et al. 2011) 17 . Fig. 5c, the authors showed that Hex reduces direction index, a commonly used measure of directional tuning in DSGCs, and thus concluded in the text that Hex does not affect "the fidelity of the computation of direction selectivity". However, I think a more careful interpretation of direction index is necessary. For example, if there is zero or very few spikes in the null direction in the control condition, the null direction firing cannot be reduced further by Hex, while the preferred direction firing is greatly reduced (as shown in Fig. 5c). In this scenario, the direction index is not changed, but the difference between null and preferred direction firing rates is dramatically altered, which could indeed be interpreted as a reduction in the fidelity of the computation of direction selectivity.

6.
We thank the reviewer for highlighting this issue. We entirely agree that it is necessary to document both the control of the firing rate and the computation of direction selectivity. That is why our results contains analysis of both parameters, which are illustrated together graphically (Fig. 5, panel c). We note that the calculation of direction selective indices is standard in the field, but also note that we are the first to produce a coherent analytical and graphical representation of direction selective indices and firing rate to illustrate the role of cholinergic signalling across a wide-range of light stimuli.
Minor points: Fig 1c and 1e? The indicated cumulative probability distributions represent analysis of the amplitude of each SAC-evoked synaptic potential recorded under the indicated conditions, the indicated number of paired recordings is documented. We note the synaptic potentials that reach the threshold for initiation of action potentials (Fig. 1, panels c, e, and  legend). We have clarified this by amending the legend of Fig. 1 to indicate (page 25):

What do individual data points represent in
(c) Cumulative probability distributions of the amplitude of each SAC-evoked PSP recorded under the indicated conditions, from the indicated number of paired recordings. Fig. 2b and d, the polar plots should be rotated 180 degrees so that the vector sums of the spiking responses point to the same preferred direction shown in Fig 2a and 2c.

8.
With respect we disagree. The polar plots represent analysis of light-evoked action potential output, and are standardly presented in the orientation presented. We have previously clarified the direction of light movement in Figure 2, panel e.

Reviewer #1
We thank the reviewer for indicating "In general, I have no concerns about the data in this study".
We are entirely in agreement with the Reviewer that "However, it is important that the results of this manuscript are placed in the correct context of current understanding of retinal direction selectivity", 1. I am not sure if the authors understand that the reason that they see an asymmetry of SAC impact in the so-call "null and preferred DSGC dendritic subfield" is because the presynaptic SACs they recorded from overlapped with the DSGC only with one side of the SAC dendrites. It is not because the overlap occurs on the one side or the other of DSGC dendritic field. For example, a SAC whose soma is very close to the DSGC soma may have its dendritic tree covering both the null and preferred subfields of the DSGC dendrites. But the inhibitory synapses only occur from the SAC dendrites pointing to the null direction of the DSGC, which is now in the preferred subfield (see the sketch below on the right). So the DSGC dendritic field is homogenous. They cannot be divided into null and preferred subfields.  Therefore, the authors need to rename these terms. For example, the last sentence in page 2 can be changed to " ……dendritic synapses from SACs on the null side of DSGCs." In page 5, last paragraph, and Figure 2: change "the preferred subfield" and "the null subfield" to "the preferred side" and "the null side".
We thank the reviewer for this comment. We have used the terms preferred side and null side throughout the revised manuscript. We thank the reviewer for their detailed comments about previous evidence concerning paracrine-like transmission between SACs and postsynaptic DSGCs. We agree that previous data is only suggestive of the role of paracrine-like neurotransmission. We note however that the points raised by the reviewer do not preclude our interpretation, specifically we provide positive evidence for vesicular ACh release in our work, and demonstrate the activation of ionotropic postsynaptic AChRs. We note that the calcium sensitivity of release does not suggest that paracrine-like transmission is not operational, in this regard previous work, for example Lee et al. presynaptic ACh release was depleted, preferred and null direction light responses were severely attenuated, and SAC-DSGC excitatory synaptic transmission selectively depressed (Fig. 4e-h; voltage integral of median filtered light responses: preferred direction: control= 14.5 ± 3.4 mV.s; vesamicol= -3.3 ± 1.5 mV.s; P= 0.0008, T= 9.05; null direction: control= 4.9 ± 3.2 mV.s; vesamicol= -11.1 ± 1.9 mV.s; P= 0.0004, T= 11.11; n= 5; excitatory PSP amplitude: control= 3.0 ± 0.6 mV; vesamicol= 0.9 ± 0.2 mV; P= 0.004, T= 4.47; n= 7). Taken together these data reveal that ACh release from ON-SACs powerfully controls the physiological responsiveness of ON-DSGCs through the activation of postsynaptic nAChRs, in a manner consistent with a local paracrine form of neurotransmission.
In addition we have amended the Discussion section (Page 17) to indicate: As our paired recordings, and previous results, have revealed the obligatory co-release of ACh and GABA from SACs, these data suggest that ACh may be released from, or have postsynaptic impact at, sites other than dendrodendritic synapses, a finding that is supported by the relatively long time to onset of cholinergic PSPs, and the effects of manipulating ACh hydrolysis 3,4 .