A tonic nicotinic brake controls spike timing in striatal spiny projection neurons

Article activity feed

1. 

   Version published to 10.7554/elife.75829 on eLife

   May 17, 2022
2. 

   Version published to 10.1101/2021.11.02.466870v4 on bioRxiv

   Apr 21, 2022
3. 

   eLife

   Apr 14, 2022

   Author Response

   Reviewer #1 (Public Review):

   This study provides data suggesting that tonic presynaptic a7 nicotinic receptor activity enhances corticostriatal input-mediated excitation of striatal medium spiny neurons; the data also suggest that tonic a4b2 nicotinic receptor activity on PV-fast spiking GABA interneurons inhibits striatal medium spiny neurons. These data advance our understanding about the complex cholinergic regulation of striatal neuronal circuits.

   The presented data are generally clean and high quality; but there are some problems that require the authors' attention.

   We thank the Reviewer for their insightful comments. We have addressed each point below with additional data and/or text. We believe these revisions have made the manuscript significantly stronger.

   1. In this study, ADP is a key parameter …

   Author Response

   Reviewer #1 (Public Review):

   This study provides data suggesting that tonic presynaptic a7 nicotinic receptor activity enhances corticostriatal input-mediated excitation of striatal medium spiny neurons; the data also suggest that tonic a4b2 nicotinic receptor activity on PV-fast spiking GABA interneurons inhibits striatal medium spiny neurons. These data advance our understanding about the complex cholinergic regulation of striatal neuronal circuits.

   The presented data are generally clean and high quality; but there are some problems that require the authors' attention.

   We thank the Reviewer for their insightful comments. We have addressed each point below with additional data and/or text. We believe these revisions have made the manuscript significantly stronger.

   1. In this study, ADP is a key parameter manipulated by several pharmacological treatments. But it is not clearly defined. The authors indicate EPSP and ADP are distinct by stating "LED pulse of increasing intensity generates excitatory postsynaptic potentials (EPSPs), or an AP followed by an after depolarization (ADP)." But the data (e.g. Fig. 1B) indicates that much of the ADP is probably EPSP. Please clarify. If much of the ADP is indeed EPSP, how are the data interpretation and the overall conclusion affected?

   We apologize for the oversight. The main focus of our study is on how tonic nAChR activation controls the timing of striatal output; our justification for including the ADP in our experimental analysis was simply corroborative, in that it represents an additional, easily measured parameter of the postsynaptic response to convergent cortical stimulation that 1) can be modulated by similar local inhibitory circuits that we show to mediate the effect of tonic nAChR activation and 2) is positioned (as opposed to EPSPs) to influence subsequent spiking, should the appropriate synaptic cues be present (which are deliberately omitted in our study). That said, under our experimental conditions EPSPs and ADPs were similar in both their kinetics and modulation by mecamylamine, suggesting that they represent mechanistically similar responses to cortical afferents. The defining difference (besides ADPs exhibiting larger amplitudes) is that they appear either in the absence of or following a spike. For these reasons we ultimately decided that reporting changes in both ADPs and EPSPs would be redundant, and limited our analyses to ADPs. Text has been added to the first paragraph of the results section to address these points.

   In Fig. 1F, ADP is absent. Why? Please clarify.

   Figure 1F shows an example of a SPN held at a mimicked ‘up-state’, achieved by injecting positive somatic current to produce a ‘resting’ membrane potential of -55-50mV. In this scenario, the ‘up-state’ membrane potential is higher than what would be reached during most ADPs evoked from Vrest, preventing the observation of ADPs in many trials. Text has been added to the end of the first paragraph in the results section to clarify this point.

   If ADP is distinct from EPSP here in MSNs, has it been reported in the literature, and how is it generated?

   Under our experimental conditions, we do not see any major differences between EPSPs and what we term ADPs (other than amplitude), at least in terms of kinetics and modulation by mecamylamine. That said, we have added text to the first paragraph of the results section that references previous work (Flores-Barrera et al.) describing suprathreshold depolarizations proceeding SPN spikes, which shaped our reasoning for including this measure in our study.

   1. In Fig. 1F, the holding potential for mecamylamine is a few mV more negative than the control, but the spike latency is shorter under mecamylamine. This is hard to understand because membrane potential (current-injection-induced depolarization + EPSP) determines spike firing and latency. If the holding potential is the same, then it's easy to understand (larger EPSP under mycamylamine).

   Thanks for pointing this out! We agree that this might seem counter-intuitive in terms of Vrest and EPSP amplitude only. Given that mecamylamine reduces GABAergic inputs to SPNs, the reduction in spike latency in this case is consistent with a reduction of GABA receptor mediated shunting. We have added this point to the text in the 3rd paragraph of the results section, which we think strengthens our justification to look at GINs as the potential mediators of mecamylamine’s effect on spike latency.

   1. Data in Fig. 2D, E are weak. The spiking ability of whole-cell recorded neurons often declines over time (evidence: the AP duration for the red trace is longer); recovery/partial recovery from MLA is needed for the data to be reliable. Fig. 2E shows 8 cells: 6 had no response, 2 increased. Sample size needs to increase.

   We appreciate this comment. Our initial justification for this experiment was from previous reports that alpha-7 nAChRs reduce corticostriatal glutamate release probability. We have now added additional data (Figure 2 supplemental data) showing that blockade of tonically activated alpha-7 nAChRs with the more specific antagonist MLA was not sufficient to change corticostriatal synaptic strength or release probability. In parallel, as we began increasing the sample size of the experiment testing the effect of MLA on spike latency, we noticed that the effect size became smaller than what we initially reported, which was already modest. Given the modest effect size of MLA on spike latency (with no presynaptic mechanism to offer), we reason that it would likely have minimal impact compared to the larger effect of mecamylamine. For this reason, we have backed off our conclusion that TONIC activation of presynaptic alpha-7 nAChRs on corticostriatal axon terminals will have a meaningful physiological impact on SPN spike timing. Accordingly, we removed previous figure 2D/E, but supplemented Figure 2A/B/C with new data (figure 2 supplement) demonstrating the lack of effect of tonic nAChR activation on corticostriatal synapse release probability. The title of the manuscript has been altered to reflect this.

   1. Fig. 7: the data on DhbE increasing AP duration is not convincing: no effect in 4 neurons, increase in 4 other neurons, and decrease in other neurons. Data ismore important than p<0.05. How do you interpret DhbE increasing AP duration?

   Point taken. We shouldn’t let a statistical calculation dominate the interpretation of a mostly mixed population result. Furthermore, upon revisiting this figure we realized that the main points pertinent to our conclusions (mecamylamine hyperpolarizes PV-FSI Vrest) were obscured by data that were of limited relevance. We have re-focused this figure to highlight data that are directly pertinent to our interpretation. This included removing the AP duration data set in question, which does not add to or inform our conclusions. We have further strengthened our conclusion that PV-FSIs are a primary mediator of the effect of tonic nAChR activation on spike latency by adding new data showing that pharmacologically blocking cortical activation of PV-FSIs occludes the effect of mecamylamine (new figure 8, see comments to Reviewer 2).

   Fig. 7F shows AP duration for PV-FSI is around 1.75 ms (some are over 2 ms, recorded at 35 C). This is unusually long. Also, the AP rise time is around 1.4 ms, very long. 1.75 ms total rise time vs. 1.4 ms for just rise: they do not add up?

   Please see our response to the above point.

   Reviewer #2 (Public Review):

   This manuscript examines one aspect of how acetylcholine influences striatal microcircuit function. While striatal cholinergic interneurons are known to be engaged in key events and tasks related to the basal ganglia in vivo, and pharmacological studies indicate cholinergic signaling is complex and critical to striatal function, the mechanistic details by which acetylcholine regulates individual cell types within the striatum, as well as how these integrate to shape striatal output, remain largely unknown. This work thus addresses an important problem in the basal ganglia field, with likely relevance to both normal function and disease-related dysfunction. The authors used a brain slice preparation in which a large number of excitatory cortical inputs to the striatum are activated, and they could measure the resulting activation of striatal projection neurons (SPNs). Their primary finding was that in this preparation, blocking nicotinic acetylcholine signaling resulted in more rapid activation of SPNs. They then explored some of the potential mechanisms for this phenomenon, and conclude that in their preparation, cholinergic interneurons are engaged both tonically and phasically, resulting in recruitment of local GABAergic interneurons that provide feedforward inhibition onto SPNs. They show that one striatal GABAergic interneuron subclass, PV-FSI, are modestly excited by tonic nicotinic signaling, and suggest this may be one contributor to their primary finding.

   Strengths of the study include the focus on cholinergic signaling across multiple striatal cell types, careful and clearly displayed slice electrophysiology, good writing, and a methodical approach to pharmacology.

   Weaknesses include reliance on the Thy1-ChR2 line to activate excitatory cortical inputs to the striatum (this line may be less specific to cortical pyramidal neurons than a specific Cre recombinase mouse line used with Cre-dependent ChR2, and thus have unintended influences on the results), and despite a strong start, a fairly weak mechanistic exploration of what GABAergic neuron subclasses might contribute to their original phenomenon.

   We thank the Reviewer for their thoughtful and constructive comments. The Reviewer identified two weakness of our study, as presented. The first weakness was our reliance on a transgenic mouse line (Thy1-ChR2) to activate cortical inputs to the striatum. Specifically, how a potential lack of specificity/ectopic expression of ChR2 in non-glutamatergic cortical neurons may impact our interpretation of the data. The second is that we did not make an effort to identify the specific subclass(es) of GINs that contribute to the phenomenon we describe. We have addressed both of these comments with new experiments, which we will describe individually below.

   1. Specificity of corticostriatal afferent activation in Thy1-ChR2 mice. As the Reviewer keenly points out, although Thy1-ChR2 mice are often used as a tool to specifically activate excitatory corticostriatal nerve terminals with optogenetic stimuli, there is concern that ChR2 expression is not exclusively limited to glutamatergic cortical neurons. If present, direct optogenetic activation of non-cortical striatal afferents would influence our results and impact our interpretation. We have addressed this issue experimentally by adding two new types of experiments (and related text, pages 7-8).

   We have added new data using immunohistochemical staining to survey for ectopic expression of ChR2 in the cortex. Staining for GAD, to broadly identify GABAergic neurons, displayed no overlap with ChR2-expressing cortical neurons in Thy1-ChR2 mice. Since a population of GABAergic somatostatin-expressing cortical neurons (particularly in the auditory cortex), have been shown to directly innervate the striatum (Rock et al., 2016), we also show that we found no evidence for somatostatin-ChR2 colocalization in our mice. Furthermore, we report no evidence for somatic expression of ChR2 in the striatum. We do report somatic expression of ChR2 in a population of globus pallidus soma, and add text to describe the above data (figure 3 supplement ) as well as published data identifying ChR2 in axons of the substantia nigra. Together, these data suggest that cortical expression of ChR2 is limited to non-GABAergic neurons, though do not eliminate the possibility of a direct monosynaptic GABAergic input to the striatum form non-cortical (and extrastriatal) brain regions. We describe newly added experimental data below to address this possibility.

   We have added new data to directly test if the optogenetic stimulation protocol used in this study induces a monosynaptic GABAergic current in SPNs (figure 3 supplement). We report that an optogenetically-evoked monosynaptic GABAergic current is indeed detected in SPNs, though it is unlikely to affect our results or interpretations for two reasons. First, based on the newly added histological data, the source of this GABAergic current is non-cortical and extrastriatal. Second, and more importantly, this input is insensitive to mecamylamine (new data, figure 3 supplement) and as such would not be modulated by the key manipulations presented in this study. Finally, experiments described below – instructed by a suggestion made by Reviewer 2 (see below) – show that blocking glutamatergic synaptic activation of a class of striatal GINs eliminates the effect of mecamylamine on SPN spike latency, ruling out the involvement of a monosynaptic GABAergic input in mediating the phenomenon.

   1. Identification of the key GIN subclass that mediates the phenomenon. Our initial manuscript included data demonstrating the feasibility of PV-FSIs in participating in the phenomenon we described, but we agree with the Reviewer that we stopped well short of identifying the class of GINs that are actually involved. We have added two new data sets to the manuscript that now corroborate both the involvement and necessity of PV-FSIs in mediating this phenomenon. First, we have added data showing that striatal SOM+ interneurons respond to mecamylamine differently than PV-FSIs do: while mecamylamine hyperpolarizes PV-FSIs, it depolarizes the average membrane potential of SOM+ interneurons and has no effect on their spontaneous firing frequency, making them unlikely candidates to mediate the phenomenon we describe. Second, we have added data showing that pharmacologically preventing cortical activation of PV-FSIs both mimics and occludes the effect of mecamylamine on spike latency and ADP amplitude (new figure 8). This data also rules out the involvement of certain other classes of GINs, such as PLTS interneurons, as the pharmacological manipulation we performed (blockade of calcium-permeable GluA2-lacking AMPA receptors) does not affect their response to cortical inputs (Gittis et al., 2010).

   Reviewer #3 (Public Review):

   The manuscript by Matityahu et al., investigated the role of tonic activation of AChRs on the spike timing of striatal spiny projection neurons (SPNs) in acute striatal slices. By selectively activation of corticostrialal projections using optogenetic tools (ChR2), they find that pharmacological blockade of presynaptic α7 nAChRs delays SPN spikes, whereas blockade of α4β2 nAChRs on GABAergic interneurons advances SPN spikes. The work is carefully done with proper control experiments, and the main conclusions are mostly well supported by data.

   Although they only constitute ~1% of the total striatal neurons in rodents and humans, cholinergic interneurons (ChINs) are gatekeepers of striatal circuitry because of their extensively arborized axons and varicosities which tonically release ACh. Whereas the role of muscarinic AChRs (mAChRs) in modulating striatal output has been well established, the role of nAChRs (especially the tonic activation) remains to be elucidated. The study is solid and the results are new and convincing. The data suggest that tonic activation of nAChRs may place a "brake" on SPN activity, and the lift of this brake during pauses of ChIN firing in response to salient stimuli may be critical for striatal information processing and learning. The findings from this study will enhance our understanding of the role of tonic nAChR activation in controlling SPNs and striatal output.

   We thank the reviewer for their careful reading of our manuscript and for their kind words and helpful suggestions.

   Unjustified Conclusions and Suggestions:

   1. The change of the SPN spike timing by AChR modulation is on a few milliseconds time scale. To make the current study more significant, the authors should design and perform additional experiments to demonstrate the functional consequence in controlling striatal output and learning. For example, will activation or blockade of nAChRs have effects on striatal STDP?

   We too would be thrilled to see the results of such experiments. Unfortunately our early attempts to perform such tests (e.g., crossing Thy1-ChR2 mice with ChAT-Cre mice to selectively express halorhodopsin in CINs, and combine cortical excitation with silencing of CINs) have been plagued by technical challenges, and would require time and resources that we feel are pragmatically beyond the scope of this study. That said, we’ve included new text (particularly, page 15) discussing how our results may fit with a newly published study on the role of CINs in corticostriatal LTP (Reynolds et al., 2022).

   1. Modulation of striatal circuitry is complex. The addition of a diagram illustrating the hypothesis and key results would help.

   Excellent suggestion. We have added a summary diagram, which is now figure 9.

   Read the original source
4. 

   eLife

   Jan 19, 2022

   Evaluation Summary:

   Matityahu et al investigate the influence of nicotinic acetylcholine receptor signaling on striatal microcircuit function through a combination of slice electrophysiology, optogenetics, and pharmacology. They find that nicotinic signaling delays spiking of striatal projection neurons in response to excitatory input, likely through the tonic release of acetylcholine by cholinergic interneurons onto local GABAergic interneurons and their influence on striatal projection neurons. Understanding how acetylcholine shapes striatal circuits is important, as this neurotransmitter is implicated in multiple movement disorders as well as other basal ganglia-related diseases.

   (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested …

   Evaluation Summary:

   Matityahu et al investigate the influence of nicotinic acetylcholine receptor signaling on striatal microcircuit function through a combination of slice electrophysiology, optogenetics, and pharmacology. They find that nicotinic signaling delays spiking of striatal projection neurons in response to excitatory input, likely through the tonic release of acetylcholine by cholinergic interneurons onto local GABAergic interneurons and their influence on striatal projection neurons. Understanding how acetylcholine shapes striatal circuits is important, as this neurotransmitter is implicated in multiple movement disorders as well as other basal ganglia-related diseases.

   (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 agreed to share their name with the authors.)

   Read the original source
5. 

   eLife

   Jan 19, 2022

   Reviewer #1 (Public Review):

   This study provides data suggesting that tonic presynaptic a7 nicotinic receptor activity enhances corticostriatal input-mediated excitation of striatal medium spiny neurons; the data also suggest that tonic a4b2 nicotinic receptor activity on PV-fast spiking GABA interneurons inhibits striatal medium spiny neurons. These data advance our understanding about the complex cholinergic regulation of striatal neuronal circuits.

   The presented data are generally clean and high quality; but there are some problems that require the authors' attention.

   1. In this study, ADP is a key parameter manipulated by several pharmacological treatments. But it is not clearly defined. The authors indicate EPSP and ADP are distinct by stating "LED pulse of increasing intensity generates excitatory postsynaptic potentials (EPSPs), …

   Reviewer #1 (Public Review):

   This study provides data suggesting that tonic presynaptic a7 nicotinic receptor activity enhances corticostriatal input-mediated excitation of striatal medium spiny neurons; the data also suggest that tonic a4b2 nicotinic receptor activity on PV-fast spiking GABA interneurons inhibits striatal medium spiny neurons. These data advance our understanding about the complex cholinergic regulation of striatal neuronal circuits.

   The presented data are generally clean and high quality; but there are some problems that require the authors' attention.

   1. In this study, ADP is a key parameter manipulated by several pharmacological treatments. But it is not clearly defined. The authors indicate EPSP and ADP are distinct by stating "LED pulse of increasing intensity generates excitatory postsynaptic potentials (EPSPs), or an AP followed by an after depolarization (ADP)." But the data (e.g. Fig. 1B) indicates that much of the ADP is probably EPSP. Please clarify.

   If much of the ADP is indeed EPSP, how are the data interpretation and the overall conclusion affected?

   In Fig. 1F, ADP is absent. Why? Please clarify.

   If ADP is distinct from EPSP here in MSNs, has it been reported in the literature, and how is it generated?

   2. In Fig. 1F, the holding potential for mecamylamine is a few mV more negative than the control, but the spike latency is shorter under mecamylamine. This is hard to understand because membrane potential (current-injection-induced depolarization + EPSP) determines spike firing and latency. If the holding potential is the same, then it's easy to understand (larger EPSP under mycamylamine).

   3. Data in Fig. 2D, E are weak. The spiking ability of whole-cell recorded neurons often declines over time (evidence: the AP duration for the red trace is longer); recovery/partial recovery from MLA is needed for the data to be reliable.
   Fig. 2E shows 8 cells: 6 had no response, 2 increased. Sample size needs to increase.

   4. Fig. 7: the data on DhbE increasing AP duration is not convincing: no effect in 4 neurons, increase in 4 other neurons, and decrease in other neurons. Data ismore important than p<0.05. How do you interpret DhbE increasing AP duration?

   Fig. 7F shows AP duration for PV-FSI is around 1.75 ms (some are over 2 ms, recorded at 35 C). This is unusually long. Also, the AP rise time is around 1.4 ms, very long. 1.75 ms total rise time vs. 1.4 ms for just rise: they do not add up?

   Read the original source
6. 

   eLife

   Jan 19, 2022

   Reviewer #2 (Public Review):

   This manuscript examines one aspect of how acetylcholine influences striatal microcircuit function. While striatal cholinergic interneurons are known to be engaged in key events and tasks related to the basal ganglia in vivo, and pharmacological studies indicate cholinergic signaling is complex and critical to striatal function, the mechanistic details by which acetylcholine regulates individual cell types within the striatum, as well as how these integrate to shape striatal output, remain largely unknown. This work thus addresses an important problem in the basal ganglia field, with likely relevance to both normal function and disease-related dysfunction. The authors used a brain slice preparation in which a large number of excitatory cortical inputs to the striatum are activated, and they could measure the …

   Reviewer #2 (Public Review):

   This manuscript examines one aspect of how acetylcholine influences striatal microcircuit function. While striatal cholinergic interneurons are known to be engaged in key events and tasks related to the basal ganglia in vivo, and pharmacological studies indicate cholinergic signaling is complex and critical to striatal function, the mechanistic details by which acetylcholine regulates individual cell types within the striatum, as well as how these integrate to shape striatal output, remain largely unknown. This work thus addresses an important problem in the basal ganglia field, with likely relevance to both normal function and disease-related dysfunction. The authors used a brain slice preparation in which a large number of excitatory cortical inputs to the striatum are activated, and they could measure the resulting activation of striatal projection neurons (SPNs). Their primary finding was that in this preparation, blocking nicotinic acetylcholine signaling resulted in more rapid activation of SPNs. They then explored some of the potential mechanisms for this phenomenon, and conclude that in their preparation, cholinergic interneurons are engaged both tonically and phasically, resulting in recruitment of local GABAergic interneurons that provide feedforward inhibition onto SPNs. They show that one striatal GABAergic interneuron subclass, PV-FSI, are modestly excited by tonic nicotinic signaling, and suggest this may be one contributor to their primary finding.

   Strengths of the study include the focus on cholinergic signaling across multiple striatal cell types, careful and clearly displayed slice electrophysiology, good writing, and a methodical approach to pharmacology.

   Weaknesses include reliance on the Thy1-ChR2 line to activate excitatory cortical inputs to the striatum (this line may be less specific to cortical pyramidal neurons than a specific Cre recombinase mouse line used with Cre-dependent ChR2, and thus have unintended influences on the results), and despite a strong start, a fairly weak mechanistic exploration of what GABAergic neuron subclasses might contribute to their original phenomenon.

   Read the original source
7. 

   eLife

   Jan 19, 2022

   Reviewer #3 (Public Review):

   The manuscript by Matityahu et al., investigated the role of tonic activation of AChRs on the spike timing of striatal spiny projection neurons (SPNs) in acute striatal slices. By selectively activation of corticostrialal projections using optogenetic tools (ChR2), they find that pharmacological blockade of presynaptic α7 nAChRs delays SPN spikes, whereas blockade of α4β2 nAChRs on GABAergic interneurons advances SPN spikes. The work is carefully done with proper control experiments, and the main conclusions are mostly well supported by data.

   Although they only constitute ~1% of the total striatal neurons in rodents and humans, cholinergic interneurons (ChINs) are gatekeepers of striatal circuitry because of their extensively arborized axons and varicosities which tonically release ACh. Whereas the role of …

   Reviewer #3 (Public Review):

   The manuscript by Matityahu et al., investigated the role of tonic activation of AChRs on the spike timing of striatal spiny projection neurons (SPNs) in acute striatal slices. By selectively activation of corticostrialal projections using optogenetic tools (ChR2), they find that pharmacological blockade of presynaptic α7 nAChRs delays SPN spikes, whereas blockade of α4β2 nAChRs on GABAergic interneurons advances SPN spikes. The work is carefully done with proper control experiments, and the main conclusions are mostly well supported by data.

   Although they only constitute ~1% of the total striatal neurons in rodents and humans, cholinergic interneurons (ChINs) are gatekeepers of striatal circuitry because of their extensively arborized axons and varicosities which tonically release ACh. Whereas the role of muscarinic AChRs (mAChRs) in modulating striatal output has been well established, the role of nAChRs (especially the tonic activation) remains to be elucidated. The study is solid and the results are new and convincing. The data suggest that tonic activation of nAChRs may place a "brake" on SPN activity, and the lift of this brake during pauses of ChIN firing in response to salient stimuli may be critical for striatal information processing and learning. The findings from this study will enhance our understanding of the role of tonic nAChR activation in controlling SPNs and striatal output.

   Unjustified Conclusions and Suggestions:

   1. The change of the SPN spike timing by AChR modulation is on a few milliseconds time scale. To make the current study more significant, the authors should design and perform additional experiments to demonstrate the functional consequence in controlling striatal output and learning. For example, will activation or blockade of nAChRs have effects on striatal STDP?
   2. Modulation of striatal circuitry is complex. The addition of a diagram illustrating the hypothesis and key results would help.

   Read the original source
8. 

   Version published to 10.1101/2021.11.02.466870v3 on bioRxiv

   Dec 6, 2021
9. 

   Version published to 10.1101/2021.11.02.466870v2 on bioRxiv

   Nov 24, 2021
10. 

    Version published to 10.1101/2021.11.02.466870v1 on bioRxiv

    Nov 3, 2021