Cholinergic and noradrenergic axonal activity contains a behavioral-state signal that is coordinated across the dorsal cortex

Article activity feed

1. 

   Version published to 10.7554/elife.81826 on eLife

   Apr 27, 2023
2. 

   Version published to 10.1101/2022.07.12.499792v2 on bioRxiv

   Mar 30, 2023
3. 

   eLife

   Mar 16, 2023

   Author Response

   Reviewer #1 (Public Review):

   Collins et al use mesoscopic two-photon imaging to simultaneously record activity from basal forebrain cholinergic or noradrenergic axons in several distant regions of the dorsal cortex during spontaneous behavior in head-fixed awake mice. They find that activity in axons from both neuromodulatory systems is closely correlated with measures of behavioral state, such as whisking, locomotion and face movements. While axons were globally correlated with these behavioral state-related metrics across the dorsal cortex, they also find evidence of behavioral state independent heterogenous signals.

   The use of simultaneous multiarea optical recordings across a large extent of dorsal cortex with single axon resolution for studying the coherence of neuromodulatory afferents across cortical areas is …

   Author Response

   Reviewer #1 (Public Review):

   Collins et al use mesoscopic two-photon imaging to simultaneously record activity from basal forebrain cholinergic or noradrenergic axons in several distant regions of the dorsal cortex during spontaneous behavior in head-fixed awake mice. They find that activity in axons from both neuromodulatory systems is closely correlated with measures of behavioral state, such as whisking, locomotion and face movements. While axons were globally correlated with these behavioral state-related metrics across the dorsal cortex, they also find evidence of behavioral state independent heterogenous signals.

   The use of simultaneous multiarea optical recordings across a large extent of dorsal cortex with single axon resolution for studying the coherence of neuromodulatory afferents across cortical areas is novel and addresses important questions regarding neuromodulation in the neocortex. The manuscript is clearly written, the data is well presented and, for the most part, carefully analyzed. Parts of the manuscript confirm previous results on the influence of behavioral state on norepinephrine and acetylcholine cortical afferents. However, the observation that these modulations are globally broadcasted to the dorsal cortex while behavioral state independent heterogenous signals are also present in these axons is novel and important for the field.

   While the evidence for a behavioral state driven global modulation of activity in both neuromodulatory systems is quite clear, I have concerns that the apparent heterogeneity in axonal responses might be driven by movement-induced artifacts. Moreover, even in the case that the heterogeneity in calcium activity across axons is confirmed, it might not be driven by differences in spiking activity across neuromodulatory axons as concluded, but by other mechanisms that are not explicitly discussed or considered.

   1. Motion artifacts are always a concern when imaging from small structures in behaving animals. This issue is addressed in the manuscript in Fig 2A-C by comparing axonal responses to "autofluorescent blebs that did not have calcium-dependent activity" (line 1011). Still, as calcium-dependent activity and motion artifacts can both be locked to behavioral variables the "bleb" selection criterion seems biased and flawed with a circular logic. "Blebs" presenting motion-induced changes in fluorescence that may pass as neural activity will be wrongly excluded when from the "bleb" control group using this criterion. This will result in an underestimation of the extent of the contamination of the GCaMP signals by movement-induced artifacts. This potential confound might generate apparent heterogeneity across axons and regions as some axons and some cortical areas might be more prone to movements artifacts than others.

   Thank you for the suggestion. We agree that motion artifacts are a reasonable concern. We rigorously addressed this concern by introducing non-calcium-dependent mCherry into cholinergic cortical axons and demonstrating that motion cannot explain our results (see Fig. 2F, Fig. 4H,L,P, Fig. 4 - figure supplement 1G, Video 3, and response above). These axons were chosen for analysis based solely on their ability to be imaged, in a manner identical to that of GCaMP6s containing axons.

   We agree that the observed evidence of heterogeneity is not as clear as the evidence of a common signal. We now carefully present our evidence. Heterogeneity may arise from variations in activity between single axons that is not explained by a common signal such as behavioral state. Heterogeneity could also be signaled by variations in correlated activity between axons. We now address these two possibilities in our manuscript. Our new analysis reveals that the correlated activity between axons is as expected for axons that are variably correlated to a common signal, such as behavioral state. Although we do find some evidence of correlation outside this common signal, we are not able to discern if this is related to imaging axon segments that are part of the same axon, or if it truly represents an independent signal. This is now stated in the text. On the other hand, strong variations in axonal activity from trial to trial that appear to be separate from the common signal is also prevalent. We now point out this variation as a possible source of heterogeneity. Since we do not know the source or meaning of this heterogeneous activity, we discuss only the possibility that it may hold behaviorally relevant information in these modulatory systems.

   1. In the case that the heterogeneity is indeed due to differences in calcium activity, it might be not due to modularity in spiking activity within the LC or the BF as interpreted and discussed in the manuscript. As calcium signaling in axons not only relates to spiking activity but can also reflect presynaptic modulations, the observed heterogeneity might be due to local action of presynaptic modulators in a context of global identical broadcasted activity. The current dataset does not allow distinguishing which of the two different mechanisms underlies the observed signal heterogeneity.

   It is true that our data set is unable to determine whether presynaptic modulations contribute to any observed heterogeneity. We have adjusted our interpretation of heterogeneity throughout the manuscript and have specifically addressed this comment in the discussion by presenting the possibility that a global signal could be locally modulated.

   Reviewer #3 (Public Review):

   Acetylcholine and Norepinephrine are two of the most powerful neuromodulators in the CNS. Recently developments of new methods allow monitoring of the dynamic changes in the activity of these agents in the brain in vivo. Here the authors explore the relationship between the dynamic changes in behavioral states and those of ACh and NE in the cortex. Since neuromodulatory systems cover most of the cortical tissue, it is essential to be able to monitor the activity of these systems in many cortical areas simultaneously. This is a daunting task because the axons releasing NE and ACh are very thin. To my knowledge, this study is the first to use mesoscopic imaging over a wide range of the cortex at the single axon resolution in awake animals. They find that almost any observable change in behavioral state is accompanied by a transient change in the activity of cortical ACh and NE axonal segments. Whisking is significantly correlated with ACh and NE. The authors also explore the spatial pattern of activity of ACh and NE axons over the dorsal cortex and find that most of the dynamics is synchronous over a wide spatial scale. They look for deviation from this pattern (which I will discuss later). Lastly, the authors monitor the activity of cortical interneurons capable of releasing ACh.

   Comments:

   1. On a broad overview, I find the discussion of behavioral states, brain states, and neuromodulation states quite confusing. To begin with, I am not convinced by the statement that "brain states or behavioral states change on a moment-to-moment basis." I find that the division of brain activity into microstates (e.g., microarousal) is counterproductive. After all, at the extreme, going along this path, we might eventually have an extremely high dimensional space of all neuronal activity, and any change in any neuron would define a new brain state. Similarly, mice can walk without whisking, can whisk without walking, can walk and whisk, are all these different behavioral states? And if so, are they all associated with different brain states? And if so, are they all associated with different brain states? Most importantly, in the context of this manuscript, one would expect that different states (brain, behavior) would be associated with at least four potential states of the ACh x NE system (high ACh and High NE, High ACh and Low NE, etc.). However, the reported findings indicate that the two systems are highly synchronized (or at least correlated), and both transiently go on with any change from a passive state to an active state. Therefore, the manuscript describes a rather confined relationship of the neuromodulation systems with the rather rich potential of brain and behavioral states. Of course, this is only my viewpoint, and the authors are not obliged to accept it, but they should recognize that the viewpoint they take for granted is not shared by all and consider acknowledging it in the manuscript.

   We thank this reviewer for this thoughtful comment. While it is clear that animals do in fact exhibit distinct and clear brain and behavioral states (e.g. sleep, waking, grooming, still, walking, etc.), it is beyond the scope of the present manuscript to attempt to tackle this complex field - rather, we refer the reader to a recent review that we have published on this important topic (McCormick, Nestvogel, and He 2020). We agree that properly delineating brain and behavioral states is of great importance, as it could significantly impact experimental design and interpretation of results. Since all of the relevant substates that a mouse may exhibit have not yet been determined, we decided to use changes in whisking and walking behaviors to differentiate between distinct behavioral states owing to: 1) historical use of these measures in behavioral and neural states in head-fixed mice, 2) relative ease of measurement of these variables, 3) a clearly observable relationship with cholinergic and noradrenergic activity with these measures of behavior, and, arguably most importantly, 4) assumed relevance to the animal (Musall et al. 2019; Reimer et al. 2016; Salkoff et al. 2020; Stringer et al. 2019).

   Our manuscript seeks to simply relate the activity of cholinergic and noradrenergic axons across the dorsal surface of the cortex in comparison to these commonly used measures of spontaneous behavior in head-fixed mice to discern to what relative degree there are common, global signals in these two modulatory systems and how they relate to changes in the measured behaviors. Somewhat surprisingly, previous studies have found that neural activity throughout the dorsal cortex of mice is strongly related to movements of the face and body as well as behavioral arousal (Stringer et al. 2019; Musall et al. 2019; Salkoff et al. 2020). Here we determine to what degree these commonly used measures of “state” are already reflected in the GCaMP6s activity of cholinergic and noradrenergic axons (and local cortical interneurons).

   We agree with the interpretation that our results suggest a confined relationship between spontaneous cholinergic and noradrenergic activity in the cortex within the spontaneous behaviors that we observe. We, by no means, mean to suggest that this confined relationship is the only relationship cholinergic and noradrenergic systems exhibit to each other or to behavior. It seems very likely that in the wide variety of behavior exhibited by freely moving mice in their lifetime, there are times in which the activity of cholinergic and noradrenergic systems exhibit a radically different relationship to each other and to behavior. We simply cannot know this without experimental examination. We now mention this possibility in the discussion and give a few appropriate references.

   1. Most of the manuscript (bar one case) reports nearly identical dynamics of ACh and NE. Is that a principle? What makes these systems behave so similarly? Why have two systems that act nearly the same? Still, if there is a difference, it is the time scale of the ACh compared to the NE. Can the authors explain this difference or speculate what drives it?

   Perhaps one of the most striking findings in recent years from examination of mouse brain activity is the prominence and prevalence of a general signal in nearly all neural systems that relates to movement and arousal of the animal (Stringer et al. 2019; Salkoff et al. 2020). Here we report that this signal is also strongly present within the cholinergic and noradrenergic systems. Perhaps this is unsurprising, since everywhere one looks, one finds this global signal. However, we feel that understanding the presence and nature of this large signal is critical to deciphering behavior-related signals in these systems in the future. We discuss this point in the discussion. The one difference we did find is in the more transient nature of NE axonal activity versus both behavior and cholinergic axon activity. We now speculate on this difference in the discussion.

   1. Whisker activity explains most strongly the neuromodulators dynamics, but pupil dilation almost does not (in contrast to many previous reports including reports of the same authors). If I am not mistaken, this was nearly ignored in the presentation of the results and the discussion section. Could the author elaborate more on what is the reason for this discrepancy?

   We apologize for the misleading presentation of our results. In Fig. 3C and D it is clear that pupil diameter is highly coherent with both cholinergic and noradrenergic axon activity, as published previously. In the present study, this coherence peaks at 0.4 to 0.5 for both. In our previous study (Reimer et al. 2016), the cholinergic activity also peaked in coherence at low frequencies at around 0.4 to 0.5 (Reimer et al., Fig. 1H) while the noradrenergic activity coherence peaked at 0.6 to 0.7. The present study was not optimized for pupil diameter examination, since we kept the light levels as low as possible (resulting in low dynamic range of pupil dilations since they were nearly always enlarged to near maximum) in order to increase the S/N of cortical axon activity. We now mention these similarities and differences and caveats in the manuscript. An additional important point is that the kinetics of pupil diameter changes are slow in comparison to whisker movements, reducing the ability of pupil dilation to accurately track changes in axonal activity at frequencies greater than approximately 0.2 Hz (Fig. 2 - figure supplement 2). This is now mentioned in the text.

   1. I find the question of homogenous vs. heterogenous signaling of both the ACh and NE systems quite important. It is one thing if the two systems just broadcast "one bit" information to the whole brain or if there are neuromodulation signals that are confined in space and are uncorrelated with the global signal. However, the way the analysis of this question is presented in the manuscript is very difficult to follow, and eventually, the take-home message is unclear. The discussion section indicates that the results support that beyond a global synchronized signal, there is a significant amount of heterogeneous activity. I think this question could benefit from further analysis. I suggest trying to demonstrate more specific examples of axonal ROIs where their activity is decorrelated with the global signal, test how consistent this property is (for those ROIs), and find a behavioral parameter that it predicts.

   Also, in the discussion part, I am missing a discussion of the potential mechanism that allows this heterogeneity. On the one hand, an area may receive NE/ACh innervation from different BF/LC neurons, which are not completely synchronized. But those neurons also innervate other areas, so what is the expected eventual pattern? Also, do the results support neuromodulation control by local interneuron circuits targeting the axons (as is the case with dopaminergic axons in the Basal Ganglia)?

   Our results clearly demonstrate a robust global signal that is common across cholinergic and noradrenergic axons which is related to behavioral state. We have less strong, but still present, evidence for a heterogeneous signal in addition to this global signal. This evidence is based largely upon the large variation in activities in different axon segments during behavioral events that appear similar. This result suggests that the axon segments we monitored do not all act as if they are members of the same axon. We now discuss the strong evidence for the global signal present in our data, and leave open the possibility of a heterogeneous signal whose mechanisms and importance remains to be determined.

   1. The axonal signal seems to be very similar across the cortex. I am not sure this is technically possible, but given that NE axons are thin and non-myelinated and taking advantage of the mesoscopic scale, could the author find any clue for the propagation of the signal on the rostral to caudal axis?

   We were unable to detect propagation across the cortical sheet and believe this is beyond the scope of the present study.

   1. While the section about local VCIN is consistent with the story, it is somehow a sidetrack and ends the manuscript on the wrong note. I leave it to the authors to decide but recommend them to reconsider if and where to include it. Unfortunately, the figure attached was on a very poor resolution, and I could not look into the details, so I am afraid that I could not review this section properly.

   We believe this adds to the manuscript and therefore have decided to include this data.

   Read the original source
4. 

   eLife

   Mar 14, 2023

   eLife assessment

   This study uses behavioral monitoring and cutting-edge calcium imaging approaches to track the activity of cholinergic and noradrenergic axons in cortex of head-fixed mice, and correlate activity with behavioral state. While the evidence that behaviorally related signals are broadly broadcasted to the dorsal cortex is clear from the data, the conclusion that there is also heterogeneity across axons and areas is of less certain significance and might be undermined by methodological artifacts.

   Read the original source
5. 

   eLife

   Mar 14, 2023

   Reviewer #1 (Public Review):

   Collins et al use mesoscopic two-photon imaging to simultaneously record activity from basal forebrain cholinergic or noradrenergic axons in several distant regions of the dorsal cortex during spontaneous behavior in head-fixed awake mice. They find that activity in axons from both neuromodulatory systems is closely correlated with measures of behavioral state, such as whisking, locomotion and face movements. While axons were globally correlated with these behavioral state-related metrics across the dorsal cortex, they also find evidence of behavioral state independent heterogenous signals.

   The use of simultaneous multiarea optical recordings across a large extent of dorsal cortex with single axon resolution for studying the coherence of neuromodulatory afferents across cortical areas is novel and addresses …

   Reviewer #1 (Public Review):

   Collins et al use mesoscopic two-photon imaging to simultaneously record activity from basal forebrain cholinergic or noradrenergic axons in several distant regions of the dorsal cortex during spontaneous behavior in head-fixed awake mice. They find that activity in axons from both neuromodulatory systems is closely correlated with measures of behavioral state, such as whisking, locomotion and face movements. While axons were globally correlated with these behavioral state-related metrics across the dorsal cortex, they also find evidence of behavioral state independent heterogenous signals.

   The use of simultaneous multiarea optical recordings across a large extent of dorsal cortex with single axon resolution for studying the coherence of neuromodulatory afferents across cortical areas is novel and addresses important questions regarding neuromodulation in the neocortex. The manuscript is clearly written, the data is well presented and, for the most part, carefully analyzed. Parts of the manuscript confirm previous results on the influence of behavioral state on norepinephrine and acetylcholine cortical afferents. However, the observation that these modulations are globally broadcasted to the dorsal cortex while behavioral state independent hetetogenous signals are also present in these axons is novel and important for the field.

   While the evidence for a behavioral state driven global modulation of activity in both neuromodulatory systems is quite clear, I have concerns that the apparent heterogeneity in axonal responses might be driven by movement-induced artifacts. Moreover, even in the case that the heterogeneity in calcium activity across axons is confirmed, it might not be driven by differences in spiking activity across neuromodulatory axons as concluded, but by other mechanisms that are not explicitly discussed or considered.

   1. Motion artifacts are always a concern when imaging from small structures in behaving animals. This issue is addressed in the manuscript in Fig 2A-C by comparing axonal responses to "autofluorescent blebs that did not have calcium-dependent activity" (line 1011). Still, as calcium-dependent activity and motion artifacts can both be locked to behavioral variables the "bleb" selection criterion seems biased and flawed with a circular logic. "Blebs" presenting motion-induced changes in fluorescence that may pass as neural activity will be wrongly excluded when from the "bleb" control group using this criterion. This will result in an underestimation of the extent of the contamination of the GCaMP signals by movement-induced artifacts. This potential confound might generate apparent heterogeneity across axons and regions as some axons and some cortical areas might be more prone to movements artifacts than others.

   2. In the case that the heterogeneity is indeed due to differences in calcium activity, it might be not due to modularity in spiking activity within the LC or the BF as interpreted and discussed in the manuscript. As calcium signaling in axons not only relates to spiking activity but can also reflect presynaptic modulations, the observed heterogeneity might be due to local action of presynaptic modulators in a context of global identical broadcasted activity. The current dataset does not allow distinguishing which of the two different mechanisms underlies to the observed signal heterogeneity.

   Read the original source
6. 

   eLife

   Mar 14, 2023

   Reviewer #2 (Public Review):

   This study uses behavioral monitoring and cutting-edge calcium imaging approaches to track the activity of cholinergic and noradrenergic axons in cortex of head-fixed mice, and correlate activity with behavioral state. The data confirm that much of this activity is dependent on behavioral state, and in particular is strongly correlated with arousal of the animal and is highly coordinated across axons. They also show that a small fraction of axonal activity is heterogenous, and does not seem to be dependent on global behavioral state. They describe additional details of this activity, such as that whisking activity is the best predictor of cholinergic and noradrenergic axon activity, and that noradrenergic activity is more transient during bouts of arousal (whisking) than cholinergic activity. Altogether this …

   Reviewer #2 (Public Review):

   This study uses behavioral monitoring and cutting-edge calcium imaging approaches to track the activity of cholinergic and noradrenergic axons in cortex of head-fixed mice, and correlate activity with behavioral state. The data confirm that much of this activity is dependent on behavioral state, and in particular is strongly correlated with arousal of the animal and is highly coordinated across axons. They also show that a small fraction of axonal activity is heterogenous, and does not seem to be dependent on global behavioral state. They describe additional details of this activity, such as that whisking activity is the best predictor of cholinergic and noradrenergic axon activity, and that noradrenergic activity is more transient during bouts of arousal (whisking) than cholinergic activity. Altogether this manuscript is generally very thorough analytically, most of the data appear technically sound, and the presentation is largely clear. However, the significance of the findings - exactly how much they enhance what is already known - is less clear.

   The main advanced novelty of the approach is the use of mesoscale imaging, giving them the ability to analyze the degree to which neuromodulatory cholinergic and noradrenergic signals are uniform across cortex, or might be correlated with distinct behavioral states or events. They attempt to get at this in Figure 4, by determining how much of their detected signal from cholinergic and noradrenergic axon activity comes from a 'common signal' versus how much of the signal is residual once the common signal is subtracted, so presumably reflects a unique influence. This analysis and the reasoning behind it is very hard to follow, and it is not clear to us that these residual signals are truly meaningful (i.e. not coming just from some source of noise). The authors try to get at this meaning in Figure 4K by plotting partial minus ordinary correlations in different arousal states, but it is not clear to us what exactly this difference means, considering the ordinary correlation itself is different in those comparisons as well. The fact that there is a bigger difference between partial and ordinary correlations during whisking than in other states does not give us real information about where the partial correlation is from.

   Read the original source
7. 

   eLife

   Mar 14, 2023

   Reviewer #3 (Public Review):

   Acetylcholine and Norepinephrine are two of the most powerful neuromodulators in the CNS. Recently developments of new methods allow monitoring of the dynamic changes in the activity of these agents in the brain in vivo. Here the authors explore the relationship between the dynamic changes in behavioral states and those of ACh and NE in the cortex. Since neuromodulatory systems cover most of the cortical tissue, it is essential to be able to monitor the activity of these systems in many cortical areas simultaneously. This is a daunting task because the axons releasing NE and ACh are very thin. To my knowledge, this study is the first to use mesoscopic imaging over a wide range of the cortex at the single axon resolution in awake animals. They find that almost any observable change in behavioral state is …

   Reviewer #3 (Public Review):

   Acetylcholine and Norepinephrine are two of the most powerful neuromodulators in the CNS. Recently developments of new methods allow monitoring of the dynamic changes in the activity of these agents in the brain in vivo. Here the authors explore the relationship between the dynamic changes in behavioral states and those of ACh and NE in the cortex. Since neuromodulatory systems cover most of the cortical tissue, it is essential to be able to monitor the activity of these systems in many cortical areas simultaneously. This is a daunting task because the axons releasing NE and ACh are very thin. To my knowledge, this study is the first to use mesoscopic imaging over a wide range of the cortex at the single axon resolution in awake animals. They find that almost any observable change in behavioral state is accompanied by a transient change in the activity of cortical ACh and NE axonal segments. Whisking is significantly correlated with ACh and NE. The authors also explore the spatial pattern of activity of ACh and NE axons over the dorsal cortex and find that most of the dynamics is synchronous over a wide spatial scale. They look for deviation from this pattern (which I will discuss later). Lastly, the authors monitor the activity of cortical interneurons capable of releasing ACh.

   Comments:
   1. On a broad overview, I find the discussion of behavioral states, brain states, and neuromodulation states quite confusing. To begin with, I am not convinced by the statement that "brain states or behavioral states change on a moment-to-moment basis." I find that the division of brain activity into microstates (e.g., microarousal) is counterproductive. After all, at the extreme, going along this path, we might eventually have an extremely high dimensional space of all neuronal activity, and any change in any neuron would define a new brain state. Similarly, mice can walk without whisking, can whisk without walking, can walk and whisk, are all these different behavioral states? And if so, are they all associated with different brain states? Most importantly, in the context of this manuscript, one would expect that different states (brain, behavior) would be associated with at least four potential states of the ACh x NE system (high ACh and High NE, High ACh and Low NE, etc.). However, the reported findings indicate that the two systems are highly synchronized (or at least correlated), and both transiently go on with any change from a passive state to an active state. Therefore, the manuscript describes a rather confined relationship of the neuromodulation systems with the rather rich potential of brain and behavioral states. Of course, this is only my viewpoint, and the authors are not obliged to accept it, but they should recognize that the viewpoint they take for granted is not shared by all and consider acknowledging it in the manuscript.
   2. Most of the manuscript (bar one case) reports nearly identical dynamics of ACh and NE. Is that a principle? What makes these systems behave so similarly? Why have two systems that act nearly the same? Still, if there is a difference, it is the time scale of the ACh compared to the NE. Can the authors explain this difference or speculate what drives it?
   3. Whisker activity explains most strongly the neuromodulators dynamics, but pupil dilation almost does not (in contrast to many previous reports including reports of the same authors). If I am not mistaken, this was nearly ignored in the presentation of the results and the discussion section. Could the author elaborate more on what is the reason for this discrepancy?
   4. I find the question of homogenous vs. heterogenous signaling of both the ACh and NE systems quite important. It is one thing if the two systems just broadcast "one bit" information to the whole brain or if there are neuromodulation signals that are confined in space and are uncorrelated with the global signal. However, the way the analysis of this question is presented in the manuscript is very difficult to follow, and eventually, the take-home message is unclear. The discussion section indicates that the results support that beyond a global synchronized signal, there is a significant amount of heterogeneous activity. I think this question could benefit from further analysis. I suggest trying to demonstrate more specific examples of axonal ROIs where their activity is decorrelated with the global signal, test how consistent this property is (for those ROIs), and find a behavioral parameter that it predicts. Also, in the discussion part, I am missing a discussion of the potential mechanism that allows this heterogeneity. On the one hand, an area may receive NE/ACh innervation from different BF/LC neurons, which are not completely synchronized. But those neurons also innervate other areas, so what is the expected eventual pattern? Also, do the results support neuromodulation control by local interneuron circuits targeting the axons (as is the case with dopaminergic axons in the Basal Ganglia)?
   5. The axonal signal seems to be very similar across the cortex. I am not sure this is technically possible, but given that NE axons are thin and non-myelinated and taking advantage of the mesoscopic scale, could the author find any clue for the propagation of the signal on the rostral to caudal axis?
   6. While the section about local VCIN is consistent with the story, it is somehow a sidetrack and ends the manuscript on the wrong note. I leave it to the authors to decide but recommend them to reconsider if and where to include it. Unfortunately, the figure attached was on a very poor resolution, and I could not look into the details, so I am afraid that I could not review this section properly.

   Read the original source
8. 

   Version published to 10.1101/2022.07.12.499792v1 on bioRxiv

   Jul 13, 2022