Activation of β2-Adrenergic Receptors in Microglia Alleviates Neuropathic Hypersensitivity in Mice
Round 1
Reviewer 1 Report (Previous Reviewer 1)
All my previous concerns have been sufficiently addressed and, in my opinion, the current version is significantly improved.
Author Response
We thank the Reviewer for his/her positive feedback
Reviewer 2 Report (New Reviewer)
Damo et al. 2022 Cells
The study investigates effects of beta-2 agonist formoterol on neuropathic pain in mice. The authors argue that activation of beta 2 receptors on microglia is responsible for the reduction in pain scores which they have detected using their models. By using a microglia-delimited knock out of beta 2 AR, the authors demonstrate that the inhibition of allodynia is primarily dependent on the receptors expressed by microglia. The study uses conventional experimental methods in vitro and in vivo. These results are discussed in line with the known application of antidepressants in chronic pain and the authors argue for the development of b2 AR agonists for therapy of chronic pain.
While the paper is in general interesting and the topic is extremely important, an obvious question to the authors is, if beta 2 agonists are really effective in chronic pain, why do we now know it from clinics? In recent years we have seen numerous studies when all sorts of curious effects were shown in mice but then never materialise in human medicine. Beta 2 agonists have been in use for many decades. If they had any noticeable effect on chronic pain it would be reported and known, such an effect cannot be missed. The literature which authors cite does not prove the validity of their case, although we do know about the use of antidepressants or alpha 2 agonists in pain clinics. So, this reviewer is left with an impression that this is yet another paper which describes some effects one can observe in mice under some experimental conditions but their relevance to the problem of human pain is questionable.
There are some serious technical issues which have to be resolved for the paper to be considered further.
Main issues
Figure 1C,D
Images in Fig 1 C look identical. Why do the authors say that there were increases? What is to be compared with what and what do the numbers in panel D show, is there a statistical value in them?
Mechanical hypersensitivity is shown in AUC in Fig 1E. This is some kind of integral of frequency (line 377) but why? Usually von Frey test output is in units of pressure, grams… Why not here? I would like to just see mechanical threshold here. Altogether here and on other Figures changes are modest, one could wonder whether they would be relevant. How would nurofen affect them?
Line 427.
Since Formoterol-induced antinociceptive effects last for less than 6 hours after a single application, this prerequisite was fulfilled in our experiments (Supplementary 1B, C).
Graphs in Supplementary Figure 1 B and C show that there was only 1 time point when the effect was significant and it was different timing for response rate (1 hour) and response latency (3 hours). Therefore the effect of formoterol only appeared briefly and for some reason different parameters were affected differently. So the question is really, is this a big effect, something to really impress?
Immuno on Fig 4 and 5.
These images are unconvincing. Possibly because of low quality of the submitted figures, I can hardly see any green colour on these panels and arrows seem to be pointing at random locations. In fact the objects which can be seen look odd. P-JNK staining looks like dots of about 1 mkM or less, sometimes at the very edge of a cell, sometimes they look outside of the cell, some cells have 3 of them. What are these dots, are they nuclei of cells? Then why are they in such odd places, sometimes several in one cell? I am not convinced by any of these images. Images of these markers (p-JNK and p-38) available from other published studies look very different.
The authors must provide high quality of individual channels and controls to convince readers that this is real staining.
I am asking the editorial office to send this paper to an author who has experience with pJNK staining to carefully scrutinise high quality images with separate channels (green and red), not only overlays.
Otherwise these experiments have to be taken out altogether.
Figure 5D.
Astrocyte density is measured by GFAP staining. This happens quite often but is a mistake. Upregulation of GFAP makes cells brighter and creates a false impression of an increased cell density. To make this claim the authors must stain for S100 which shows cell somas and co-stain for DPI.
Additional issues
Line 50 Recently, antidepressant drugs, which enhance Antidepressants have been in use for treatment of chronic pain for many decades, this is by no means recent.
Line 92 Microglia are known to particularly express high levels of Gs-coupled β2-adre93
noceptors (β2-AR) and respond to the application of NA [9,23,24].
Authors need to back up this statement with the data from recent transcriptomic studies from cells other but cultured.
Line 337 Next, we ascertained the purity of microglia extraction through quantitative RT-PCR (qPCR) analysis, which re vealed enrichment of microglial genes
It is not unusual to put forward such data as argument for purity. However without knowing how well the primers for Aqp4, Syt1 and MBP work, this comparison is meaningless. In addition, different mRNA are expressed in cells at different levels and if you compare a more abundant RNA with less abundant species you will always get "specificity". Can the authors assure that their primers for all markers work equally well?
Line 743-4 A few studies have reported beneficial effects of β2-AR agonists, such as terbutaline,
744 on diverse types of inflammatory pain and more recently, also pain of neuropathic origin
745 [39,40].
This is a questionable statement. The key issue is that these drugs are in clinical use for very long time. If they had any noticeable analgesic effect in chronic pain we would have known it for a long time. Such effects do not go unnoticed. However we have no such data from patients. This makes one wonder, whether these observations have clinical relevance at all.
Line 793-794 This also has implications for the current therapeutic applications of SNRIs, such as
Duloxetine, in neuropathic pain…
One cannot draw direct parallels between NA uptake blockers and the agonists of beta2 receptors. There many more potential mechanisms in case of antidepressants. After all amitriptyline which by no means is selective to NA is one of the most frequently used drugs for chronic pain from this group.
Author Response
Please see the attachment
Author Response File: 
Author Response.pdf
Round 2
Reviewer 2 Report (New Reviewer)
Thanks for revising the manuscript. The main issue, however, persists, if beta 2 agonists had such an effect in humans we would have known it for decades.
So the clinical relevance of this research is unclear.
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
The manuscript uses different approaches to establish the role of microglial β2-adrenergic receptors in contributing to neuropathic pain in SNI mice.
The paper is very reach in results and figures and the findings are novel and interesting. Yet, the overall organization of the paper is quite confusing and it is difficult for the reader to follow the rational of the experiments.
major concerns:
1- the reader is continuously forced to jump from one topic to another, or from one protocol to another. Different time points are used in different experiments displayed in different figures or insupplementary materials, and it is very hard to take track of them and follow the rationale of the work. Initially, the effect of SNI and formoterol on microglia is shown at 3 days after injury (fig 1-3), then microglia is investigated in transgenic mice at 6 and 21 days after injury (fig4-5). Then, authors report some experiments on astrocytes at 6 and 21 days after injury in WT animals (fig.6) or both astrocyte and microglia at day 21 ( Fig 7). Than again microglia only after 21 days (figS9). Animal behavior is sparsely reported in main figures and supplementary figures.
I would definitely suggest to revise the organization of the manuscript to make it more readable. here some suggestions: the results should start with a full description of pain behavior (mechanical and cold sensitivity ) at 3, 6 and 21 days after injury along with the effect of formoterol. Then, a detailed description of microglia activation, including p-p38 and JNK expression and the relative effects formoterol at 3, 6 and 21 day after injury should be reported. then astrocytic activation (+JNK) at days 3, 6 and 21 after injury should be reported in a subsequent and separate section. Finally, I would report the data on transgenic mice to dissect more in depth the specific role of β2-adrenergic receptors in microglia only.
2- There is an overall abuse of supplementary figures. Main results (such as JNK expression or the effect of SNI on microglia morphology) should be in main figures.
3- A main issue in all the paper is the relative low numerosity of experimental groups (typically 3-5 animals per group). This is fine if the variance is small, but becomes an issue when the variance increases. I have estimated that the test power in most of the experiments is about 10%, which means that in case of negative results it might be difficult to discriminate a true negative from a false negative (type II error). This is a problem for the interpretation, for instance, of FIG.S3B or FIG.S8B, where an effect can be apparently seen in the graph, but that turned out to be negative to the test. In such situations, I recommend that the claim that there is no effect should be, at least, mitigated.
4- Following point 3, I would mitigate any strong claim that there are relevant sex differences in this study. In fact, the few sex differences detected are minor and are typically associated to negative results that are not very convincing (see above). Just as an example, the authors rule out an increase int expression of microglial JNK in females 3 days after injury based on the ambiguous results shown in figure S3B (not that the mean of the black column is even higher in females than in males); Indeed, in contradiction, few figures later (S7B) they show a significant increase of microglial JNK in females at day 6.
5- I recommend the authors to expand the discussion by including some considerations on sex. In particular, they should emphasize the overall lack of sex difference in this study (rather than emphasizing the few, ambiguous differences). I would also suggest removing the sentence “in a sex-dependent manner” from line 21 of the abstract, as it is not clear which sex-dependency should be noticed in the manuscript
Minor points:
1- Line 47 introduction, “which also exert inhibition via Gi-coupled opioid receptors”, please add a reference
2- Line 61 introduction, “Thus, it has been proposed drugs that reinforce this descending system can be more efficient against neuropathic pain”. Please, rephrase: English needs revision and the sentence seems referring to the facilitatory descending system rather than to the inhibitory descending system.
3- Line 282 methods, authors stated that they use ANOVA, but actually they use two-way ANOVA in the text. Please specify in methods. If a two ways ANOVA is used, the significance of the interaction between variables (e.g. spinal cord side vs treatment) should be reported.
4- Figure 3A: what is the blue? Is it DAPI? In case, please add DAPI protocol in the method section.
5- Line 295 results, not clear what is the meaning of “first” in this sentence.
6- Line 431 results, please add a sentence to state that the deletion of the receptor in microglia does not significantly affect the course of allodynia/hyperalgesia following SNI
7- Line 493: the sentence is unclear. What do the authors mean for neuropathic pain being “fully established” after 3 weeks? What do the authors mean saying that microglia activation is not reliable in females?
8- Line 496 results: here the authors introduce data on astrocytes. They should explain HERE why they suddenly decided to investigate astrocytes, which is instead mentioned later at line 560. Also, the quantification of astrocytic activation induced by SNI at the different time points after injury should be reported before mentioning the effect of formoterol.
Reviewer 2 Report
Damo and Simonetti investigated the contribution of β2-adrenergic receptor activation in microglia to antinociception in neuropathic mice. They conducted interesting experiments and attempted to analyse gender differences in the antinociceptive effects of formoterol, a β2-AR agonist. They also used a conditional KO for microglial β2-ARs. They also highlighted the possible usage of β2-AR agonists to treat pain. Although they performed different complementary investigations, there are some major concerns about n numbers, animal age, time-points and sample collection which need to be clarified, before the manuscript is accepted for publication. Comments are as follows:
1) I appreciate the commitment of the authors with the 3R´s in regards of reducing animal numbers. However, most of the data was gathered from n = 3-4 for each gender used. An n equal to n does not allow to draw conclusions in regards of statistical analyses and more importantly, in regards of experimental reproducibility. I mean, in how many different occasions were the experiments performed? They should be performed at least in two different occasions.
2) Age varied according the analysis. They authors should justify why animals of different ages were employed for different methods as age can influence pain perception.
3) Treatment schedules must be made clearer as well as experiment time-points. For example, the authors state they treated animals with formoterol at days 3, 6 and 21, or 21 following neuropathy induction. They then proceeded to measure nociceptive responses at 1, 3, 6, 12, and 24 h (mechanical nociception), and at 1, 3, 8, 14, and 26 h to measure responses to cold after Formoterol injection. The 3h time-point is missing in the mechanical nociception graph and the 1h time-point is missing in the cold plate test graph (Figure S1).
4) CPP test was stated to be performed at days 4 and on day 32 after the operation (session 2.4.1) and then at days 8 and 36 (session 3.8) and days 5 and 32 (legend to figure 8). Which one is the correct combination of time-points? Why they do differ to the measurements of mechanical and cold nociception?
5) For some experiments, L3 – L4 spinal segments were used, for others, L3 -L5. Why different segments were used?
6) Please indicate the total numbers of male and female mice used in the study in the appropriate session.
