A ‘double-edged’ role for type-5 metabotropic glutamate receptors in pain disclosed by light-sensitive drugs

Curation statements for this article:
  • Curated by eLife

    eLife logo

    eLife assessment

    In this interesting study, the authors have used light-sensitive mGlu5 negative allosteric modulators to determine the role of these receptors in a chronic pain model. These findings could be useful to the pain field, but the evidence supporting these claims is incomplete.

This article has been Reviewed by the following groups

Read the full article See related articles

Abstract

We used light-sensitive drugs to identify the brain region-specific role of mGlu5 metabotropic glutamate receptors in the control of pain. Optical activation of systemic JF-NP-26, a caged, normally inactive, negative allosteric modulator (NAM) of mGlu5 receptors, in cingulate, prelimbic, and infralimbic cortices and thalamus inhibited neuropathic pain hypersensitivity. Systemic treatment of alloswitch-1, an intrinsically active mGlu5 receptor NAM, caused analgesia, and the effect was reversed by light-induced drug inactivation in the prelimbic and infralimbic cortices, and thalamus. This demonstrates that mGlu5 receptor blockade in the medial prefrontal cortex and thalamus is both sufficient and necessary for the analgesic activity of mGlu5 receptor antagonists. Surprisingly, when the light was delivered in the basolateral amygdala, local activation of systemic JF-NP-26 reduced pain thresholds, whereas inactivation of alloswitch-1 enhanced analgesia. Electrophysiological analysis showed that alloswitch-1 increased excitatory synaptic responses in prelimbic pyramidal neurons evoked by stimulation of presumed BLA input, and decreased BLA-driven feedforward inhibition of amygdala output neurons. Both effects were reversed by optical silencing and reinstated by optical reactivation of alloswitch-1. These findings demonstrate for the first time that the action of mGlu5 receptors in the pain neuraxis is not homogenous, and suggest that blockade of mGlu5 receptors in the BLA may limit the overall analgesic activity of mGlu5 receptor antagonists. This could explain the suboptimal effect of mGlu5 NAMs on pain in human studies and validate photopharmacology as an important tool to determine ideal target sites for systemic drugs.

Article activity feed

  1. Author response:

    Reviewer #1 (Public Review):

    Metabotropic glutamate receptors (mGLuRs) play a key role in regulating neuronal activity and related behaviors. In different brain regions these receptors can be expressed presynaptically and postsynaptically in different classes of neurons. Therefore, it is difficult to predict the effects of systemically applied drugs that act on these receptors. Here, the authors harness the power of photopharmacology, applying modulators that can be activated or inactivated by light with spatial precision, to address this problem. Their stated goal is to determine the role of mGluRs in regulating pain behaviors, and the circuit mechanisms driving this regulation. Their findings suggest that mGluRs acting in medial prefrontal cortex and thalamus drive antinociception in animals with neuropathic pain, whereas these receptors drive pronociception when acting in the amygdala. Their circuit analysis suggests that, in the amygdala, mGluRs act by decreasing feedforward inhibition of the output neurons. These findings have the potential to affect the development of targeted treatment for pain and related disorders. The elegant photopharmacological approaches will likely inform future studies attempting to distinguish the action of neuroactive drugs in different brain regions.

    We thank the reviewer for the insightful evaluation of our study.

    Reducing the impact of these studies are several methodological, analytical, and interpretation issues.

    The authors report that "the effect of optical manipulations of photosensitive mGlu5 NAMs in individual brain regions in pain models has been studied before". It is, therefore, not immediately clear what is novel in the present study.

    We have clarified this in the following statement (page 3, lines 15‐17): “It remains to be determined if region‐specific actions play a role in the overall analgesic activity of mGlu5 receptor NAMs, considering that opposite actions have been reported”. The subsequent paragraph nicely explains the novelty of our approach, which is based on the combined use of a drug activated by light (JF‐NP‐26) and another drug inactivated by light (alloswitch‐1) to determine which region is sufficient and/or necessary for the analgesic effect of systemic mGlu5 receptor NAMs. In the Discussion (page 7) we state that “To the best of our knowledge, this is the first study to employ photopharmacological tools to compare and contrast distinct roles of mGlu5 receptors in different regions of the pain matrix”.

    The reliance only on reflexive measures of pain, especially in a study that examines the role of "affective and cognitive aspects of pain and pain modulation".

    The main endpoint of the study was not to examine the cognitive and affective aspects of pain, although some of the regions examined are involved in these aspects of pain besides the regulation of sensory aspects (pain thresholds). However, we followed the kind suggestion and measured depression‐like and risk‐taking (anxiety‐like) behaviors in mice. To optimize the number of mice and be still consistent with the number of mice approved by the regulatory agency we used the following groups of mice for the evaluation of risk‐taking behavior with the light‐dark box: (i) sham‐operated mice treated with vehicle; (ii) CCI mice treated with vehicle; (iii) CCI mice treated with JF‐NP‐26 without light activation; and (iv) CCI mice treated with JF‐NP‐26 and irradiated with activating light (the test cannot be performed in the same mice before and after light activation to avoid habituation); depression‐like behavior with the tail suspension test was performed in two separate groups of mice: (i) CCI mice treated with JF‐NP‐26 with no light; and (ii) CCI mice treated with JF‐NP‐26 and light activation. All mice had been implanted with optic fibers in the basolateral amygdala.

    Data are shown in the new Supplementary Fig. S4 and reported in the Results section (page 5) as follows: “Knowing that mGlu5 receptors in the BLA shape susceptibility to stress and fear in rodents (35, 36), we also measured depression‐like and risk‐taking behavior after light‐induced activation of JF‐NP26 in the BLA of neuropathic mice. Light‐induced activation of JF‐NP‐26 decreased risk‐taking hence increased anxiety‐like behavior in CCI mice as shown by the decreased number of entries into, and reduced time spent in, the light compartment of the light‐dark box (Fig. S4a‐c). Depression‐like behavior assessed with the tail‐suspension test was unchanged in CCI mice after light‐induced irradiation of JF‐NP‐26 in the BLA (Fig. S4d).”

    The inclusion of only males is unfortunate because of known, significant sex differences in neuronal circuits driving pain conditions, in both preclinical models (including form work by the authors) and in clinical populations.

    We are aware that there are important sex differences in the pain neuraxis, but this study was not about sex differences. The goal was to evaluate any region‐specific actions of systemically administered compounds (mGlu5 NAMs) and the contribution and requirement of specific brain regions to the observed drug effects, using photopharmacology and drugs activated or inactivated/reactivated by light. This analysis would have been less straightforward in female mice given for example that it is known that mGlu5 receptors interact with estrogen receptors. This aspect could be addressed in a future project. The present study provides the basis for comparative studies in females.

    The elegant slice experiments (especially Fig. 3) were designed to probe circuit mechanisms through which mGluRs act in different brain regions. These experiments also provide a control to assess whether the photopharmacological compounds act as advertised. Surprisingly, the effect size produced by these compounds on neuronal activity are rather small (and, at times, seems driven by outliers). How this small effect affects the interpretation of the behavioral findings is not clear.

    These small effect sizes should also be considered when interpreting the circuit actions studied here.

    We greatly appreciate your insightful comments and constructive feedback on our findings. The mean effect sizes observed in certain experiments are quite small, but effects or changes were very consistent. And we illustrate this now by including lines to connect individual data points for the same neuron in the modified Figure 3 (f, g, n, o) to show consistent changes observed in the EPSC and IPSC graphs. We would like to add that is not quite clear how neuronal effects translate into behavioral consequence, how much of a change in individual neurons or in a population of neurons or change of a certain magnitude is sufficient and required. These are all interesting questions, but the results of our behavioral and electrophysiological data match quite nicely, including differential or opposing drug effects.

    Some of the sample sizes are as small as n=3. Without an a priori power analysis, it is difficult to assess the validity of the analyses.

    The authors present intriguing data on changes in InsP levels in some (but not all) animals after injury, but not in sham animals. They also report an increase in the expression of mGLuRs expression in some, but not all brain regions. These findings are not discussed. It is not clear how these selective changes in mGluR expression and activity might affect the interpretation of the photopharmacological results.

    We performed new experiments to increase sample size in PI experiments in the infralimbic and prelimbic cortices where the n was low. Now the data are more solid. New statistical values are reported in the legend of Fig. 1. We also added a discussion of the signaling data (page 9) as follows:

    “We found that mGlu5 receptor‐mediated PI hydrolysis was significantly amplified in all subregions of the contralateral mPFC and in the contralateral amygdala after induction of neuropathic pain whereas mGlu5 receptor protein levels were significantly increased only in the contralateral infralimbic cortex of neuropathic mice. This suggests that, at least in the anterior cingulate cortex, prelimbic cortex, and basolateral amygdala, mGlu5 receptors become hyperactive after induction of pain. It remains to be determined if this is mediated by an enhanced coupling of mGlu5 receptors to Gq/11 proteins, increased expression of phospholipase‐C or other mechanisms. Interestingly, mGlu5 receptor signaling was down‐regulated in the thalamus of neuropathic mice, but mGlu5 blockade in the thalamus still had antinociceptive effects (see below). Downregulation of mGlu5 receptor signaling in the thalamus might represent a compensatory mechanism aimed at mitigating pain in neuropathic mice.”

    The behavioral data seem to represent discrete, and not continuous variables. The statistical tests applied are likely inappropriate for these analyses.

    The behavioral values reported here represent measurements of force (g) required to elicit a reflex (i.e., reflex thresholds) and can be considered continuous variables. The statistical tests used for the behavioral experiments included either t‐test to determine if the difference between two groups was statistically significant or One‐Way ANOVA (repeated measures when appropriate) to determine if there were any statistically significant differences between the means of three or more groups. This form of analysis for the outcome measures in this study is well‐established in the literature.

    The authors assume (and state in the abstract) that they can selectively stimulate BLA afferents to the neocortex. This is technically highly unlikely.

    We appreciate the reviewer's insightful comment regarding the technical challenges associated with the selective stimulation of BLA afferents to the neocortex. We are aware that the electrical stimulation does not allow the exclusive stimulation of a specific pathway, though BLA afferents form the major component of afferent fibers running in the layer IV of the infralimbic cortex on their way to targets in layer II/III and layer V or infra‐ and pre‐limbic cortices.

    Our previous work (Kiritoshi et al., 2016) compared directly electrical and optogenetic stimulation in the mPFC, and found that they match, suggesting that electrical stimulation provides a reliable means to activate BLA input in the mPFC. We acknowledge the technical limitations of selective BLA activation with electrical stimulation, though we are confident that our approach allowed the investigation of mGlu5 manipulations in the BLA‐mPFC circuitry. We have modified the abstract to read as follows: “Electrophysiological analysis showed that alloswitch‐1 increased excitatory synaptic responses in prelimbic pyramidal neurons evoked by stimulation of presumed BLA input, and decreased BLA‐driven feedforward inhibition of amygdala output neurons”.

    The results from the experiment on rostroventral medulla (RVM) neurons are less than convincing because only a "trend" towards decreased excitation is reported. As above, without consideration of effect size, it is hard to appreciate the significance of these findings. The absence of a demonstration of a classical ON Cell firing pattern is also unfortunate.

    We appreciate this observation. Based on the Reviewer’s suggestion, we report below the effect size of optical modulation in the prelimbic cortex on RVM activity, according to Cohen’s d calculation from ttests (now shown in the Table 1). This information is also included in Results (page 6).

    Moreover, in this study we classified ON‐ or OFF‐cells based on their firing patterns relative to nocifensive withdrawal responses (H.L. Fields and M.M. Heinricher 1985). As ON‐cells with high basal firing can be easily misclassified as NEUTRAL‐cells (N.M. Barbaro, M.M. Heinricher, H.L. Fields, 1986), potential NEUTRAL‐cells with continuous spontaneous activity were verified by giving a brief bolus of anesthetic to the point that the withdrawal reflex was abolished. Indeed, firing of spontaneously active ON‐cells slows or stops with this manipulation, which unmasks reflex‐related responses. This is now reported and explained in Methods (page 14).

  2. eLife assessment

    In this interesting study, the authors have used light-sensitive mGlu5 negative allosteric modulators to determine the role of these receptors in a chronic pain model. These findings could be useful to the pain field, but the evidence supporting these claims is incomplete.

  3. Reviewer #1 (Public Review):

    Metabotropic glutamate receptors (mGLuRs) play a key role in regulating neuronal activity and related behaviors. In different brain regions these receptors can be expressed presynaptically and postsynaptically in different classes of neurons. Therefore, it is difficult to predict the effects of systemically applied drugs that act on these receptors. Here, the authors harness the power of photopharmacology, applying modulators that can be activated or inactivated by light with spatial precision, to address this problem. Their stated goal is to determine the role of mGluRs in regulating pain behaviors, and the circuit mechanisms driving this regulation. Their findings suggest that mGluRs acting in medial prefrontal cortex and thalamus drive antinociception in animals with neuropathic pain, whereas these receptors drive pronociception when acting in the amygdala. Their circuit analysis suggests that, in the amygdala, mGluRs act by decreasing feedforward inhibition of the output neurons. These findings have the potential to affect the development of targeted treatment for pain and related disorders. The elegant photopharmacological approaches will likely inform future studies attempting to distinguish the action of neuroactive drugs in different brain regions.

    Reducing the impact of these studies are several methodological, analytical, and interpretation issues.

    - The authors report that "the effect of optical manipulations of photosensitive mGlu5 NAMs in individual brain regions in pain models has been studied before". It is, therefore, not immediately clear what is novel in the present study.
    - The reliance only on reflexive measures of pain, especially in a study that examines the role of "affective and cognitive aspects of pain and pain modulation".
    - The inclusion of only males is unfortunate because of known, significant sex differences in neuronal circuits driving pain conditions, in both preclinical models (including form work by the authors) and in clinical populations.
    - The elegant slice experiments (especially Fig. 3) were designed to probe circuit mechanisms through which mGluRs act in different brain regions. These experiments also provide a control to assess whether the photopharmacological compounds act as advertised. Surprisingly, the effect size produced by these compounds on neuronal activity are rather small (and, at times, seems driven by outliers). How this small effect affects the interpretation of the behavioral findings is not clear.
    - These small effect sizes should also be considered when interpreting the circuit actions studied here.
    - Some of the sample sizes are as small as n=3. Without an a priori power analysis, it is difficult to assess the validity of the analyses.
    - The authors present intriguing data on changes in InsP levels in some (but not all) animals after injury, but not in sham animals. They also report an increase in the expression of mGLuRs expression in some, but not all brain regions. These findings are not discussed. It is not clear how these selective changes in mGluR expression and activity might affect the interpretation of the photopharmacological results.
    - The behavioral data seem to represent discrete, and not continuous variables. The statistical tests applied are likely inappropriate for these analyses.
    - The authors assume (and state in the abstract) that they can selectively stimulate BLA afferents to the neocortex. This is technically highly unlikely.
    - The results from the experiment on rostroventral medulla (RVM) neurons are less than convincing because only a "trend" towards decreased excitation is reported. As above, without consideration of effect size, it is hard to appreciate the significance of these findings. The absence of a demonstration of a classical ON Cell firing pattern is also unfortunate.

  4. Reviewer #2 (Public Review):

    In this study, Notartomaso et al. used optical activation of systemic JF-NP-26, a caged, baseline inactive, negative allosteric modulator (NAM) of mGlu5 receptors, in cingulate, prelimbic and infralimbic cortices, thalamus, and BLA to investigate the roles of these receptors in various brain regions in pain processing. They found that alloswitch-1, an intrinsically active mGlu5 receptor NAM, caused analgesia, but this analgesic effect was reversed by light-induced drug inactivation in the prelimbic and infralimbic cortices, and thalamus. In contrast, these pharmacological effects were reversed in the BLA. They further found that alloswitch-1 increased excitatory synaptic responses in prelimbic pyramidal neurons evoked by stimulation of BLA input, and decreased feedforward inhibition of amygdala output neurons by BLA. They thus concluded that mGlu5 receptors had differential effects in distinct brain regions. mGlu5 receptors are important receptors in pain processing, and their regional specificity has not been studied in detail. Further, this is an interesting study regarding the use of optical activation of pro-drugs, and the findings are timely. The combination of in vivo pharmacology, biochemistry, and slice EP provides complementary results.

  5. Reviewer #3 (Public Review):

    In this manuscript, Notartomaso, Antenucci et al. use two different light-sensitive metabotropic glutamate receptor negative allosteric modulators (NAMs) to determine how mGlu5 receptor signaling in distinct brain regions of mice influences mechanical sensitivity in chronic constriction injury (CCI) model of neuropathic pain. This is an extension of their previous work using photocaged mGlu5 negative allosteric modulators and incorporates a systemically active NAM that can be locally photoswitched off and on with violet and green light, respectively. The authors found that NAM signaling in the thalamus and prefrontal cortical regions consistently reduced mechanical hypersensitivity. However, they observed divergent effects on these measures in the basolateral amygdala. The authors attempted to solve the discrepancy in behavioral measurements between mGlu5 signaling in the basolateral amygdala by determining how NAMs modulate synaptic transmission or in vivo firing and found that these effects were projection-dependent.

    Strengths:

    This study demonstrates the importance of local signaling by mGlu5 receptors across multiple pain-processing circuits in the brain, and the use of optical activation and inactivation strategies is very intriguing.

    Weaknesses:

    One major limitation is the lack of sham surgery groups and vehicle/light-only controls in behavior and physiology experiments, though the authors did test mechanical sensitivity in the contralateral paw. The reliance on a single behavioral measure in these groups is also problematic. Many of these brain regions are known to influence distinct aspects of somatosensory processing or other behaviors entirely, which may be interpreted as hypersensitivity (e.g. fear or anxiety-like behaviors in the basolateral amygdala). Details on the light intensities used is also absent, and it is important to test whether violet light had any unintended effects on these cells/mice.

    While the effort to provide some mechanistic understanding using slice physiology and in vivo recordings is appreciated, they ignore any effects that these NAMs have directly on the excitability of the recorded output neuron. In the models, mGlu5 is proposed to exist on some upstream inhibitory (mPFC) or excitatory (BLA) projection, but no evidence of a direct effect on these synaptic inputs was observed. Given the widespread distribution of mGlu5 in these brain regions, the proposed model seems unlikely. Perhaps CCI induces changes in functional coupling of mGlu5 in different cell types, and this could be revealed with appropriate control experiments.

    Overall the broad profiling taken here across multiple brain regions lacks controls and some cohesion, making it challenging to conclude how mGlu5 signaling is acutely impacting these circuits and/or specific cell types.