Functional independence of endogenous μ- and δ-opioid receptors co-expressed in cholinergic interneurons

  1. Seksiri Arttamangkul
  2. Emily J Platt
  3. James Carroll
  4. David Farrens

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This study is significant because it addresses a debated question in the field about whether different opioid receptor subtypes expressed in the same cells must function as a unit or function independently. Here the authors show that while mu and delta opioid receptors each signals in a similar manner in response to specific treatments, their interactions are largely independent of one another in modulating the firing and regulation by desensitization and internalization mechanisms in striatal cholinergic interneurons.

    (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 #3 agreed to share their name with the authors.)

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Class A G-protein-coupled receptors (GPCRs) normally function as monomers, although evidence from heterologous expression systems suggests that they may sometimes form homodimers and/or heterodimers. This study aims to evaluate possible functional interplay of endogenous µ- and δ-opioid receptors (MORs and DORs) in mouse neurons. Detecting GPCR dimers in native tissues, however, has been challenging. Previously, MORs and DORs co-expressed in transfected cells have been reported to form heterodimers, and their possible co-localization in neurons has been studied in knock-in mice expressing genetically engineered receptors fused to fluorescent proteins. Here, we find that single cholinergic neurons in the mouse striatum endogenously express both MORs and DORs. The receptors on neurons from live brain slices were fluorescently labeled with a ligand-directed labeling reagent, NAI-A594. The selective activation of MORs and DORs, with DAMGO (µ-agonist) and deltorphin (δ-agonist) inhibited spontaneous firing in all cells examined. In the continued presence of agonist, the firing rate returned to baseline as the result of receptor desensitization with the application of deltorphin but was less observed with the application of DAMGO. In addition, agonist-induced internalization of DORs but not MORs was detected. When MORs and DORs were activated simultaneously with [Met 5 ]-enkephalin, desensitization of MORs was facilitated but internalization was not increased. Together, these results indicate that while MORs and DORs are expressed in single striatal cholinergic interneurons, the two receptors function independently.

Article activity feed

  1. eLife

    Author Response:

    Reviewer #1 (Public Review):

    This paper addresses an important question in the opioid field, whether the mu opioid receptor (MOR) and the delta opioid receptor (DOR) are likely to occur as independent receptors or whether their signaling is coupled and could be the result of interactions. The authors take advantage of a fluorescent label, NAI-A594, which binds to both receptors, to do live imaging experiments in the cholinergic neurons of striatum and test how the responses to a selective agonist to one receptor affects subsequent responses to an agonist of the other. They use receptor internalization and electrophysiological recordings to gauge the likelihood that the two receptors are independently expressed or act as a unit. The work is carefully done, and the authors conclude that the two receptors act independently in these neurons; the data support the idea that at all of the receptors are not necessarily linked. However, the work cannot exclude that some of the receptors also act together. One issue is that opioid receptors and GPCRs generally can produce distinct effects: recruitment of the beta-arrestin pathway promotes desensitization and receptor internalization, while signaling via Galpha or beta-gamma produces other signaling events. While the agonists used in the present study likely target both pathways in the MOR and DOR, it remains possible that in a heterodimer, signaling might instead be biased. Moreover, the similar downstream signaling pathways make it more difficult to untangle the possible interactions between the two receptor types.

    We have included discussion about biased signaling of ME. We agree that our work has not ruled out the possible transient formation of a subset of heterodimers as there could not be determined by approaches used in this study. However, at the macroscopic level and time scale, our results do not support the possibility of a stable dimer. It is not known if the lifetime of heterodimer association will be long enough for activation process and causing biased signaling. In basal conditions, no evidence of MOR-MOR homodimers was found by single molecule analyses (Moller et al., 2020; Asher et al., 2021). MOR-MOR-homodimers could be induced by DAMGO with a lifetime of ~460 milliseconds indicating the fast decay of dimers (Moller et al., 2020).

    Reviewer #2 (Public Review):

    Strengths:

    1. The authors use a nice combination of pharmacology, in situ hybridization, and fluorescence analysis to demonstrate that MORs and DORs are both expressed on ChIs.
    1. The authors specifically analyze expression patterns based on sex or dorsal vs ventral striatal zones.
    1. The authors use extracellular recordings of ChIs to demonstrate that their firing patterns are sensitive to MOR or DOR stimulation.
    1. The authors use pharmacology and extracellular recordings to demonstrate that MOR and DOR mediated inhibition is relatively independent of each other.
    1. The authors use live cell imaging and pharmacology to examine whether selective agonist administration results in receptor internalization.

    Weaknesses:

    1. It remains unclear why Met-enk pretreatment results in MOR desensitization.

    We agree this is an interesting question, and is something we are pursuing.

    Reviewer #3 (Public Review):

    This is an important study that addresses a standing question regarding whether different types of opioid receptors expressed in the same cell signal independently of one another or operate as functional units (heterodimers). This study specifically explores this question by investigating the co-expression of mu and delta opioids receptors (MORs and DORs respectively) in cholinergic interneurons of the striatum. It has been known for some time that these neurons express both MORs and DORs, but functional interactions between them in these neurons has not been explored. The study uses a variety of methodologies to investigate these interactions and the experiments were generally well designed to test this hypothesis. The results are quite striking, suggesting that this study will be of high impact to help show that while these receptors may co-exist within cells, that they do not necessarily have to act in concert with one another, which is especially relevant in deciphering opioid signaling in neural function and neurotransmission. There are some concerns related to how the data were analyzed, raising some questions about how to interpret the data and whether certain conclusions are warranted, but these do not detract too heavily from a very interesting study.

    Strengths:

    1. The use of the NAI compounds in combination with receptor-specific antagonists is a nice way to measure receptor expression and internalization, especially across the whole of the striatum.
    1. Performing measures in both male and female tissue is a strength.
    1. There was a nice comparison made between receptor-specific ligands and the endogenous opioid peptide met-enkephalin.
    1. The experiments were generally thorough with proper controls and used a variety of methodologies to address the hypothesis.
    1. Even without detailed statistical reporting (see below) many of the findings appear to be robust.

    Weaknesses:

    1. The manuscript lacks detailed statistical analyses, mostly relying on descriptions of the data as interpreted by the authors. In most places there are no indications of which statistical analyses were utilized and what the outcomes of those analyses were. In the few places where analyses were indicated, it is not clear that the appropriate tests were used (e.g. using an unpaired t-test when an ANOVA, and possibly a repeated measures ANOVA should have been used). The description of the statistical analyses utilized in the paper also conflicts with what is actually used in most places. It is difficult to evaluate the data without this information. Figure 3 uses standard deviation as a measure of variability, but other figures use standard error and there is no explanation for why this is the case. In many cases there are statements regarding data being different than baseline or 100%, but these are not supported by any statistical measures as being truly different.

    The statistical analyses have been added. The SD for the labeling data was used to demonstrate the distribution of fluorescence of each labeling condition, which also showed variability depending on location of neurons in the striatum. There were no hypothetical values for comparing in this data set, therefore SE was not used.

    1. The results section makes claims about the kinetics of desensitization as a result of met-enkephalin treatment (referring to Figure 5E), but there is no indication that a time by treatment factor was significantly different. The authors cannot make claims about the rate of desensitization without an actual assessment of rate. Relatedly, the authors do not fully discuss that while MORs and DORs have different degrees of desensitization at the times they measure, the two receptors may have similar maximal extents of desensitization, just at different time scales. Figure 5D has the implication that MORs are beginning to desensitize, just at a slower rate than DORs. Essentially, the authors are trying to have it both ways: ignoring rates in most cases and implicating rates in one case without actually testing them.

    The difference of desensitization curves as the result of ME treatments was analyzed by two-way ANOVA and the p values of time as treatment factor were included in the results. Regarding to the second concern, although it is possible that the two receptors may have similar maximal extents of desensitization if application of agonists is prolonged, this is not the purpose of our study that reports the different regulation processes of MORs and DORs endogenously co-expressed in a single neuron. It is clear that at during 5-minute application of agonists, desensitization of MOR occurs slower and with lesser degree than that of DOR.

    1. The authors conclude that MORs do not internalize, whereas DORs do, but their time course does not align with previous experiments, involving a very long treatment followed by a long washout period. The treatment differences could play a role in their differential outcomes (MOR recycling v. DOR recycling). The authors should address this disparity either experimentally or discuss it as a limitation.

    The time course of previous internalization and the desensitization were different. We did a new set of experiments in that internalization was studied with 5-minute application of agonists as it was used in desensitization experiments. There was no detectable internalization of either MOR or DOR at this 5-minute time point. We added these data in the results (Figure 6C), and discussed that desensitization and internalization were separable.

    1. The staining of DORs (as inferred by CTAP treatment) in Fig 1Bc does not match the pattern of DOR expression in the literature, appearing like there is no DOR anywhere besides the most dorsolateral region of the striatum. This also conflicts with their data in Figure 3. This is curious and should be addressed/discussed. The species differences between figures could play a role in this or it could be experimental methods.

    The result showed that there was no patch-like structure that indicated the staining of MOR. The image was taken from a macroscope with low resolution. The low fluorescence signal was difficult to acquire. In figure 3, we used 2-photon microscope to determine each stained neuron and thus a high-quality image was obtained.

    1. The authors used a variety of pharmacological agents and curiously failed to discuss instances where some of the agents didn't produce expected results. For example, morphine only partially decreased firing, which was surprising, but also wasn't discussed. CTAP and naloxone did not fully reverse the effects of DAMGO (Figure 5C), but this was glossed over.

    Morphine is known to be a partial agonist and thus our finding is not a surprising result based on several significant literatures (reviewed in Birdsong and Williams, Mol Pharmacol 2020, 98, 401-409). CTAP and naloxone reverse the action of DAMGO. The time course of blockade varied from cell to cell most likely as a result of the location of the cell within the brain slice. Cells deeper in the slice will be more slowly affected by both agonists and antagonists.

    1. It is curious that the authors found heterologous desensitization with met-enkephalin treatment, but did not explicitly test this with their receptor-specific ligands. This relates to a larger concern, and one that is lightly touched upon in the discussion: the indication that depending on the signaling pathway (G protein v. arrestin) there could be different outcomes for receptor function and regulation (i.e. biased signaling). It would be important for the authors to discuss this given that some of the pharmacological treatments they employ have different biases in their signaling which could affect their measured outcomes.

    We include a discussion on biased signaling of each receptor with ME. We also discuss the non-biased signaling of arrestin and G protein by DAMGO and deltorphin at MORs and DORs, respectively.

    1. Experiments were performed on tissue from both male and female mice, but the proportion of each sex used in each experiment was not clear, aside from Figure 3 and its accompanying supplemental figure. While overall expression may not differ between sex, sex differences could account for variability in functional data and the sexes used should be indicated in each experiment or at least discussed as a limitation of the study.

    We pooled the data from male and female mice. Reported numbers of male and female mice used were now shown in Table 1. We discussed the limitation of this finding and did not investigate potential sex difference. Each data point was the result from recording of one neuron from a single slice.

    1. The use of MOR knockout mice is a good control, but there are no details provided of how cholinergic interneurons were identified in these mice.

    We included in the figure legend (Supplemental Figure 3) that ChIs in MORKO were identified by morphology of neurons being larger than other neurons nearby. The staining of ChIs from MORKO was compared to the staining of ChAT-GFP using the same protocol and analysis. Finding the large cells and confirming with green fluorescence of ChAT-GFP helped in identifying and assigning ChIs without GFP.

    1. The description of the methods used to calculate desensitization (lines 236-240) did not seem to match what was actually performed and the methods did not clarify this. It is difficult to evaluate the data when it is not clear how the data were obtained.

    We have now corrected the calculation of the desensitization that is described in the results.

    1. The descriptions of MOR desensitization was muddled. It was described as having persistent inhibition (i.e. implied lack of desensitization), but the Table and Figures indicated that MORs do desensitize, just not to the extent that DORs do.

    We have now changed the description of MOR desensitization to be clearer that MORs also desensitized, but at a much lesser degree when compared to DOR.

    1. The authors cite literature that assessed cholinergic interneuron function in dorsal and ventral striatum and their staining data show expression of opioid receptors in both dorsal and ventral striatum, but they chose to focus on cholinergic interneurons in the ventral striatum. The authors should provide a clear rationale in the results section where this decision was made.

    The rationale has been added and focuses on functional interaction between MOR and DOR in the ventral striatum. The distribution of receptors measured with NAI-594 suggest a comparable expression of MOR and DOR in this area. This is a key point as it allows possible functional studies using cells with similar expressions of the two receptors.

  2. Version published to 10.7554/elife.69740 on eLife
  3. eLife

    Evaluation Summary:

    This study is significant because it addresses a debated question in the field about whether different opioid receptor subtypes expressed in the same cells must function as a unit or function independently. Here the authors show that while mu and delta opioid receptors each signals in a similar manner in response to specific treatments, their interactions are largely independent of one another in modulating the firing and regulation by desensitization and internalization mechanisms in striatal cholinergic interneurons.

    (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 #3 agreed to share their name with the authors.)

  4. eLife

    Reviewer #1 (Public Review):

    This paper addresses an important question in the opioid field, whether the mu opioid receptor (MOR) and the delta opioid receptor (DOR) are likely to occur as independent receptors or whether their signaling is coupled and could be the result of interactions. The authors take advantage of a fluorescent label, NAI-A594, which binds to both receptors, to do live imaging experiments in the cholinergic neurons of striatum and test how the responses to a selective agonist to one receptor affects subsequent responses to an agonist of the other. They use receptor internalization and electrophysiological recordings to gauge the likelihood that the two receptors are independently expressed or act as a unit. The work is carefully done, and the authors conclude that the two receptors act independently in these neurons; the data support the idea that at all of the receptors are not necessarily linked. However, the work cannot exclude that some of the receptors also act together. One issue is that opioid receptors and GPCRs generally can produce distinct effects: recruitment of the beta-arrestin pathway promotes desensitization and receptor internalization, while signaling via Galpha or beta-gamma produces other signaling events. While the agonists used in the present study likely target both pathways in the MOR and DOR, it remains possible that in a heterodimer, signaling might instead be biased. Moreover, the similar downstream signaling pathways make it more difficult to untangle the possible interactions between the two receptor types.

  5. eLife

    Reviewer #2 (Public Review):

    Strengths:

    1. The authors use a nice combination of pharmacology, in situ hybridization, and fluorescence analysis to demonstrate that MORs and DORs are both expressed on ChIs.

    2. The authors specifically analyze expression patterns based on sex or dorsal vs ventral striatal zones.

    3. The authors use extracellular recordings of ChIs to demonstrate that their firing patterns are sensitive to MOR or DOR stimulation.

    4. The authors use pharmacology and extracellular recordings to demonstrate that MOR and DOR mediated inhibition is relatively independent of each other.

    5. The authors use live cell imaging and pharmacology to examine whether selective agonist administration results in receptor internalization.

    Weaknesses:

    1. It remains unclear why Met-enk pretreatment results in MOR desensitization.
  6. eLife

    Reviewer #3 (Public Review):

    This is an important study that addresses a standing question regarding whether different types of opioid receptors expressed in the same cell signal independently of one another or operate as functional units (heterodimers). This study specifically explores this question by investigating the co-expression of mu and delta opioids receptors (MORs and DORs respectively) in cholinergic interneurons of the striatum. It has been known for some time that these neurons express both MORs and DORs, but functional interactions between them in these neurons has not been explored. The study uses a variety of methodologies to investigate these interactions and the experiments were generally well designed to test this hypothesis. The results are quite striking, suggesting that this study will be of high impact to help show that while these receptors may co-exist within cells, that they do not necessarily have to act in concert with one another, which is especially relevant in deciphering opioid signaling in neural function and neurotransmission. There are some concerns related to how the data were analyzed, raising some questions about how to interpret the data and whether certain conclusions are warranted, but these do not detract too heavily from a very interesting study.

    Strengths:

    1. The use of the NAI compounds in combination with receptor-specific antagonists is a nice way to measure receptor expression and internalization, especially across the whole of the striatum.

    2. Performing measures in both male and female tissue is a strength.

    3. There was a nice comparison made between receptor-specific ligands and the endogenous opioid peptide met-enkephalin.

    4. The experiments were generally thorough with proper controls and used a variety of methodologies to address the hypothesis.

    5. Even without detailed statistical reporting (see below) many of the findings appear to be robust.

    Weaknesses:

    1. The manuscript lacks detailed statistical analyses, mostly relying on descriptions of the data as interpreted by the authors. In most places there are no indications of which statistical analyses were utilized and what the outcomes of those analyses were. In the few places where analyses were indicated, it is not clear that the appropriate tests were used (e.g. using an unpaired t-test when an ANOVA, and possibly a repeated measures ANOVA should have been used). The description of the statistical analyses utilized in the paper also conflicts with what is actually used in most places. It is difficult to evaluate the data without this information. Figure 3 uses standard deviation as a measure of variability, but other figures use standard error and there is no explanation for why this is the case. In many cases there are statements regarding data being different than baseline or 100%, but these are not supported by any statistical measures as being truly different.

    2. The results section makes claims about the kinetics of desensitization as a result of met-enkephalin treatment (referring to Figure 5E), but there is no indication that a time by treatment factor was significantly different. The authors cannot make claims about the rate of desensitization without an actual assessment of rate. Relatedly, the authors do not fully discuss that while MORs and DORs have different degrees of desensitization at the times they measure, the two receptors may have similar maximal extents of desensitization, just at different time scales. Figure 5D has the implication that MORs are beginning to desensitize, just at a slower rate than DORs. Essentially, the authors are trying to have it both ways: ignoring rates in most cases and implicating rates in one case without actually testing them.

    3. The authors conclude that MORs do not internalize, whereas DORs do, but their time course does not align with previous experiments, involving a very long treatment followed by a long washout period. The treatment differences could play a role in their differential outcomes (MOR recycling v. DOR recycling). The authors should address this disparity either experimentally or discuss it as a limitation.

    4. The staining of DORs (as inferred by CTAP treatment) in Fig 1Bc does not match the pattern of DOR expression in the literature, appearing like there is no DOR anywhere besides the most dorsolateral region of the striatum. This also conflicts with their data in Figure 3. This is curious and should be addressed/discussed. The species differences between figures could play a role in this or it could be experimental methods.

    5. The authors used a variety of pharmacological agents and curiously failed to discuss instances where some of the agents didn't produce expected results. For example, morphine only partially decreased firing, which was surprising, but also wasn't discussed. CTAP and naloxone did not fully reverse the effects of DAMGO (Figure 5C), but this was glossed over.

    6. It is curious that the authors found heterologous desensitization with met-enkephalin treatment, but did not explicitly test this with their receptor-specific ligands. This relates to a larger concern, and one that is lightly touched upon in the discussion: the indication that depending on the signaling pathway (G protein v. arrestin) there could be different outcomes for receptor function and regulation (i.e. biased signaling). It would be important for the authors to discuss this given that some of the pharmacological treatments they employ have different biases in their signaling which could affect their measured outcomes.

    7. Experiments were performed on tissue from both male and female mice, but the proportion of each sex used in each experiment was not clear, aside from Figure 3 and its accompanying supplemental figure. While overall expression may not differ between sex, sex differences could account for variability in functional data and the sexes used should be indicated in each experiment or at least discussed as a limitation of the study.

    8. The use of MOR knockout mice is a good control, but there are no details provided of how cholinergic interneurons were identified in these mice.

    9. The description of the methods used to calculate desensitization (lines 236-240) did not seem to match what was actually performed and the methods did not clarify this. It is difficult to evaluate the data when it is not clear how the data were obtained.

    10. The descriptions of MOR desensitization was muddled. It was described as having persistent inhibition (i.e. implied lack of desensitization), but the Table and Figures indicated that MORs do desensitize, just not to the extent that DORs do.

    11. The authors cite literature that assessed cholinergic interneuron function in dorsal and ventral striatum and their staining data show expression of opioid receptors in both dorsal and ventral striatum, but they chose to focus on cholinergic interneurons in the ventral striatum. The authors should provide a clear rationale in the results section where this decision was made.

  7. Version published to 10.1101/2021.04.23.441175v1 on bioRxiv