Voluntary activation of muscle in humans: does serotonergic neuromodulation matter?

Abstract Ionotropic inputs to motoneurones have the capacity to depolarise and hyperpolarise the motoneurone, whereas neuromodulatory inputs control the state of excitability of the motoneurone. Intracellular recordings of motoneurones from in vitro and in situ animal preparations have provided extraordinary insight into the mechanisms that underpin how neuromodulators regulate neuronal excitability. However, far fewer studies have attempted to translate the findings from cellular and molecular studies into a human model. In this review, we focus on the role that serotonin (5‐HT) plays in muscle activation in humans. 5‐HT is a potent regulator of neuronal firing rates, which can influence the force that can be generated by muscles during voluntary contractions. We firstly outline structural and functional characteristics of the serotonergic system, and then describe how motoneurone discharge can be facilitated and suppressed depending on the 5‐HT receptor subtype that is activated. We then provide a narrative on how 5‐HT effects can influence voluntary activation during muscle contractions in humans, and detail how 5‐HT may be a mediator of exercise‐induced fatigue that arises from the central nervous system.

; and line [l]) in an attempt to improve the quality of the manuscript: p2; l57: Ionotropic/neuromodulatory is not necessarily a class of input, rather they are inputs that activate classes of receptors. The inputs activate either ionotropic or metabotropic receptors and neuromodulatory inputs would predominantly activate those of the metabotropic class. Slight rewording for clarity would be helpful here. p5; l137: is this supposed to refer to rates of discharge? section ending on l142: The discussion of the role of 5ht on the dorsal cord seems a little light -one might consider a slight expansion of this topic here so that the (predominantly) inhibitory effects of 5ht on sensory transmission can be appreciated. p7; l207-209: It may be nice to put this magnitude of change into perspective for the naive reader. If a healthy young adult has near complete activation of the biceps (near 100%; see pre-fatigue %VA in Fig 3) then we would not expect that they could gain more than a percent or two as they cannot have >complete activation (theoretical, I know). Other muscles, with lower %VA, may show greater change. p8; l235: put "i.e. the motoneurone's hysteresis" in parenthesis? p8; l236-237: Although seldomly considered in the literature, the threshold of the PIC can vary and may be a key contributor to MN recruitment. Please refer to the discussion of Afsharipour et al 2021 Section ending l375: I suggest updating this section with reference to the the emerging findings from both CJ Heckman's lab and from the second author's lab (references below). These findings, albeit preliminary in nature, both support the notion that afferent feedback may dampen the effects of PICs in humans. Pearcey

31-Mar-2022
This review provides a comprehensive and balanced overview of the role of 5-HT on muscle activation. Insights from mammalian and from human studies are discussed. The authors do a commendable job of comparing the results from both animal and human studies and identify methodological challenges in testing hypotheses derived from animal work in human studies. The review will be very appreciated by the motor control community I only have minor comments: 1. I know that this review looks at voluntary muscle activation but I wonder if there are any insights about 5-HT perhaps modulating reflexive movements in humans and whether there could be a greater role in that type of movement based on the possibility that serotonergic systems are activated by novel or surprising stimuli 2. What about the role of 5-HT in mediating presynaptic inhibition of sensory afferents? Are there insights on this in humans, and does it relate to central fatigue? 3. If I can make a suggestion to add Bui et al. Please try to keep the diagram as simple as possible and without superfluous information that may distract from the main conclusion of the Review. Abstract Figures must be provided by authors no later than the revised manuscript stage and should be uploaded as a separate file during online submission labelled as File Type 'Abstract Figure'. Please ensure that you include the figure legend in the main article file. All Abstract Figures will be sent to a professional illustrator for redrawing and you may be asked to approve the redrawn figure before your paper is accepted. -Please upload separate high quality figure files via the submission form. -Author profile(s) must be uploaded via the submission form. Authors should submit a short biography (no more than 100 words for one author or 150 words in total for two authors) and a portrait photograph of the two leading authors on the paper. These should be uploaded, clearly labelled, with the manuscript submission. Any standard image format for the photograph is acceptable, but the resolution should be at least 300 dpi and preferably more. A group photograph of all authors is also acceptable, providing the biography for the whole group does not exceed 150 words. Author Contribution: Justin Kavanagh: Conception or design of the work; Acquisition or analysis or interpretation of data for the work; Drafting the work or revising it critically for important intellectual content; Final approval of the version to be published; Agreement to be accountable for all aspects of the work Janet Taylor: Conception or design of the work; Acquisition or analysis or interpretation of data for the work; Drafting the work or revising it critically for important intellectual content; Final approval of the version to be published; Agreement to be accountable for all aspects of the work

Running Title: Serotonin and voluntary muscle activation
Dual Publication: Figure 1B, 1C, and 1D: Our schematic of the serotonergic system and motoneurones is supported by cellular data published in Cotel, Exley, Cragg, and Perrier JF, Proceedings of the National Academy of Sciences USA, 2013, 110(12):4774-9. This does not constitute dual publication as we use this representative image only to illustrate the main theme of the Topical review. Namely, that prior to our human work, cellular studies had revealed the dual role that serotonin plays in modulating motoneurone activity. The source publication is included in the figure caption.  Benoit-Marand, 2017). It appears that depleting brain 5-HT in rats has the effect of reducing non-invasive brain stimulation or motor practice can influence excitatory synaptic plasticity.

111
In particular, there is evidence to suggest enhanced long-term potentiation-like plasticity from    the depolarising current for motoneurone firing so that less synaptic input is required.

240
The paired motor unit technique uses the firing rate of low threshold motor units, that already 241 have PICs fully active, as estimates of synaptic input to other motor units in the same pool.

242
The measure that is calculated is known as ∆F (difference in frequency of firing of the lower 243 threshold motor unit at recruitment and derecruitment of the higher threshold motor unit).  HT release (Orssatto et al., 2021b). Reductions in motoneurone PIC may also provide a 256 mechanism for loss of force production from skeletal muscle. In younger populations, there is 257 evidence that ∆F reduces up to 25% for soleus motoneurones following passive stretching of  Competing interests and funding 406 The authors declare that no competing interests exist for this work and no funding was 407 received to perform this work.  Pharmacology 164, S115-S135.      This manuscript has been reviewed by 3 expert reviewers. They unanimously enjoyed reading this review, noted its scholarship qualities and ultimately recommend publication. Several minor suggestions and clarifications are proposed and they appear to be expert and reasonable. As such, the authors should have no difficulties in addressing them in a timely manner.
We thank the Reviewing Editor for the positive comments regarding our manuscript. We have addressed each referee in turn and have highlighted in our manuscript where amendments have been made.

REFEREE #1
Kavanagh and Taylor provide a timely review with the goal of highlighting and summarizing recent work in humans that describe roles and underlying mechanisms for 5HT and target receptors in the voluntary activation of muscle in humans. To accomplish this end, they place their human work in the context of multiple animal models that describe how serotonin modulates intrinsic properties and ion channels of motoneurons. The review draws the conclusions that studying 5HT control of muscle activation in humans might not be as straight forward as initially thought and propose/discuss several potential challenges. Because of the cross-human-animal comparison, this review will be of interest and useful to those studying neural control of muscle in humans and likewise, will be of interest to those studying mechanisms governing motoneuron excitability in animal models. This review is well written and was enjoyable to read. That being said, I have several comments that I think should be addressed. It is my hope that these comments will strengthen the review.
We thank Referee #1 for these positive comments. The referee has provided exceptional comments for us to consider, and we have incorporated nearly all suggestions in our revised manuscript. We would, however, like to preface our responses by pointing out that the majority of recommendations were for the animal and cellular sections of the manuscript. We had to carefully consider how much additional information to include in these sections so our Topical Review retains its focus on human muscle activation. We have included line numbers in our revised manuscript.

Of note, the manuscript was not submitted with line numbers but was generated by
Page 2 Line 38-40: the statement saying 5HT can have profound effects on force generation during muscle contractions is a bit of an overstatement as the review highlights recent work that shows a very modest change MVC following manipulation of the serotonergic system. While I can agree that 5HT has a potent effect on firing rates, this doesn't seem to translate to huge changes in maximal muscle force -which is interesting and discussed in the review. I think this discrepancy should either be highlighted in the abstract to provide context for the title of the review. Alternatively, I suggest toning back this statement.
To keep the abstract succinct, we have decided to tone back this statement. We feel that drawing attention to this particular discrepancy may detract from the many other comparisons that are made throughout the review. Our statement now reads, "Serotonin (5-HT) is a potent regulator of neuronal firing rates which can influence the force that can be generated by muscles during voluntary contractions.
Line 73: Why has serotonergic modulation been so extensively studied? While I can agree that it is a potent modulatory system in the control of spinal circuits, it is far from the only one. I think that this would be worth highlighting and would be useful to provide reviews that describe roles for other neuromodulators in the control of spinal circuits/motoneuron excitability (eg. Cholinergic, dopaminergic, noradrenergic, etc.).
We agree that several neuromodulator systems are involved in the regulation of motor circuits in the CNS. We have included the following information in the revised introduction with a view of remaining focussed on the serotonergic system, but also acknowledging that other neurotransmitter systems in the CNS contribute to muscle activation: "There are several neuromodulator systems that regulate the excitability of cortical and spinal motor circuits, and many excellent reviews have been written that describe the effects of each system on motor activity (e.g. Line 88: While the effects of ionotropic neurotransmitters may be terminated by removal of the ligand, the effects of neuromodulators, such as 5HT on neuronal function may not necessarily be terminated by reuptake of 5HT given that activation of GPCRs can elicit long lasting changes in ion channel function.
We agree that this is an important consideration that should be included in the manuscript. We have now modified our text to state: "It is often suggested that the termination of 5-HT effects at the synapse occurs with reuptake into pre-synaptic terminals or glial cells via monoamine transporters. However, serotonergic effects on neuronal function may not necessarily end with reuptake of 5-HT, as activation of G protein-coupled receptors can elicit long lasting changes in ion channel function (Pavlos & Friedman, 2017)".

Line 125: it is not clear what is meant by evoked responses. Would this be using TMS in humans or in animal models using other approaches?
This was a poorly worded sentence, and we have changed the text to read, "Most of the work that has examined 5-HT effects on motor cortical activity has employed magnetic or electrical stimulation techniques to explore cortical plasticity, or neuroimaging techniques to map changes in regional activity due to changes in neurotransmitter concentration." Line 137: worth mentioning is work from Brian Noga's Lab that used voltammetry to show increases in 5HT that occurs on the time scale of seconds during MLR-evoked fictive locomotion in cats (https://doi.org/10.3389/fncir.2017.00059). Interestingly, 5HT levels begin to increase prior to the onset of fictive locomotion which is supportive of a role for 5HT in setting the network tone for locomotion.
We thank Reviewer 1 for highlighting this work, as it has provided a valuable piece of information for our manuscript. Although the reviewer has suggested that this study could be included in our discussion of contraction intensity, we believe that it is best suited several sentences earlier where the onset of 5-HT release complements our anatomical descriptions of the 5 Line 146-150: the three mechanisms described would collectively increase neuronal excitability. I think it is worthwhile using precise language to highlight this here. Eg change modulate on Line 150 to increase.
We have changed the term 'modulate' to 'increase' in the revised manuscript.

Line 152: It is unclear what is exactly meant by 'inhibition of hyperpolarization'. It is not clear how reducing spike threshold would inhibit hyperpolarization. Lines 152-156 should be rethought and reworded for clarity.
We have modified the text in the former line 152 to read, "5-HT can also increase discharge rate via mechanisms that reduce the amplitude of slow and medium afterhyperpolarisation phases that follow the action potential ( Similarly, we have reworded the sentences that follow this statement to improve clarity in our discussion, "Although 5-HT can reduce hyperpolarisation by decreasing the threshold for generating a Na+ based action potentials (Fedirchuk & Dai, 2004), 5-HT also generates a Ca+ dependent plateau potential by reducing a K+ current responsible for slow afterhyperpolarisation (Hounsgaard & Kiehn, 1989).
Line 160: the DLF is not a specific analogue of the raphe-spinal pathway, but does contain descending serotonergic fibres, in addition to many other. This might be misleading as written.
We have changed our description of the dorsolateral funiculus to, "DLF, contains descending serotonergic fibres".

Line 167: which neuron?
We have changed the term 'neuron' to 'motoneurone' in the revised manuscript. Line 172: by 'increased hyperpolarisation' do you mean that the resting membrane potential was hyperpolarized? If so, this should be reworded for clarity.
The reviewer is correct. We have amended our terminology to indicate that 5-HT injected in the vicinity of the motoneurone caused hyperpolarisation of the resting membrane.
Line 174: while the work was conducted in JF Perrier's lab, I think it is important to acknowledge and the credit should be drawn to the first author, F Cotel, who likely performed the majority of the work. Please rephrase.
We have changed this sentence to read, "An underlying mechanism of 5-HT-mediated inhibition of motoneurone activity was clarified in a series of elegant experiments by Cotel and Perrier".
Line 189: An interesting observation that I think is worth mention, that 5HT fibres have been reported at the AIS of rodent motoneurons (Deardorff, Romer, and Fyffe, 2021), which is in contrast to that has been suggested by Cotel et al. It is possible that this is a species difference. Alternatively, it might be due to differences in motoneuron types, which have not really been explored. Regardless, it still provides a means for direct compartmental modulation of channels that are expressed at the AIS, which differ from those expressed in the dendrites.
We agree with the referee that this is an important consideration. We have amended our text to read, "However, it is worth noting that these findings are only reported for adult turtle motoneurones and may not be applicable for other species. Indeed, a recent study have identified 5-HT boutons present on the AIS of rodent motoneurones (Deardorff et al., 2021), and compartmentalisation of 5-HT receptor subtypes on human motoneurones is yet to be detailed".
Line 224: Indeed, previous work has demonstrated contributions of CaV and NaV channels to PIC and PIC-mediated firing properties, and historically it was thought that PICs were relatively straight forward. However more recent work from the Brocard group has suggested that it might be a bit more complex than initially thought as other ion channels can contribute to PIC and PIC-mediated intrinsic properties. Examples include activation of TRPM5 (Bos et al., 2021) and inactivation of KV1.2 (Bos et al., 2018). Further, M-type potassium currents (mediated by KCNQ channels) influence measures of PIC and their effect on intrinsic properties and oppose NaV1.6-mediated inward currents in excitatory interneurons. While this direct interaction has not been shown in motoneurons, it is worth noting given that motoneurons do express KCNQ channels in somatodendritic and AIS compartments alongside NaV1.6 channels (Verneuril et al., 2020). Whilst historically it was thought that PICs were straightforward, it is becoming more apparent that this is not the case as multiple channels contribute to and even oppose their actions in parallel. A similar argument could be made for modulation of 'PICchannels' as changes in PIC-mediated intrinsic properties could be mediated through changes in some of these other channels. I highly recommend checking these works out and think that these points are worth highlighting in this review.
We have included the sentence below to highlight evidence of more complex mechanisms of PIC-like non-linearities in motoneurone behaviour.
"However, recent evidence shows that non-linearities in motoneurone discharge may have more complex mechanisms including activation of TRPm5 and inactivation of Kv1. Line 249: I would suggest softening this statement from PICs mediate serotonin's effect on motor performance, to PICs may partially contribute to serotonin's effect on motor performance. As previously highlighted, many other ion channels can be influenced by serotonin.
We agree that this statement was too strong, so we now suggest that "PICs may be partially responsible for 5-HT mediated increases in voluntary activation". These comments have provided a valuable inclusion for our manuscript. We have amended our text to the following: "These findings provide support that PIC activation in humans is linked to voluntary drive and hence 5-HT release. The difference between gastrocnemius and soleus also highlights that PICs, and their influence on motoneuron intrinsic properties, differ between muscles and motoneuron subtypes ( We have rewritten this statement to improve the clarity of the text: "However, it is known that PICs associated with gastrocnemius medialis increase from 10% to 20% MVC, and PICs associated with soleus increase from 10% to 30% MVC, during slow ramped plantarflexions (Orssatto et al., 2021b). These findings provide support that PIC activation in humans is linked to voluntary drive and hence 5-HT release." Line 256: Not to mention that changes in the expression of the PIC channels or those that oppose their actions could also contribute.
We are not sure what the reviewer is suggesting here. Acute changes in Delta F and maximal voluntary force are seen after 5 muscle stretches of ~1 min each and then recover over ~10 min. We have not amended the manuscript based on this comment.
Line 313: suggest toning back language from '...in the spinal cord was due to...' to '...in the spinal cord could be due to...' as although plausible, this is purely speculative.
We have changed the phrase 'in the spinal cord was due to' to 'in the spinal cord could be due to' in the revised manuscript Line 348: Modulated 'by' 5HT.
We have changed the phrase 'modulated with 5-HT' to 'modulated by 5-HT'.
Line 369-373: Maybe not such a recent idea with the citations provided approaching 15 years ago. Suggest rephrasing.
We have changed the phrase 'A recent proposal' to 'A reasonable proposal'.

Line 375: This paragraph ends kind of open. Is there a possibility to capitalise on new genetic tools in rodents that allow for identification and manipulation of defined interneuron subtypes to advance these ideas further? Are there new methods/technologies/techniques in humans that might provide some insight?
We agree that this paragraph finished quite open-endedly. However, we believe that the simple explanation for the open-ended paragraph was its lack of conclusion. In particular, the final sentence did not emphasis the take home message from the paragraph. We have changed the final sentence in this paragraph to read, "Thus, it is possible that a 5-HT mechanism involving inhibitory spinal circuitry can regulate the amplitude of agonist and antagonist muscle contractions".
Line 377: I'm not entirely sure that this question was answered in this conclusion paragraph. Indeed, challenges are highlighted, but I think it would be worthwhile to outright statement that might provide more of an answer to this question.
We have changed this paragraph to provide more direction for the reader: "Animal experiments provide clear evidence that 5-HT is a potent modulator of spinal circuits and motoneurone output. However, the effects of 5-HT on voluntary muscle activity in humans are less clear. Effects on motor performance during whole-body exercise are inconsistent. This is not to say that serotonergic neuromodulation does not matter for humans, but instead highlights the challenges associated with studying how a complex neuromodulatory system acts during muscle contractions. Controlled experiments using single-joint, single muscle, contraction protocols have found that maximal force can be changed by altering 5-HT activity in the CNS. The differences observed in voluntary activation are small but are present despite an otherwise intact system and the actions of other neuromodulators, including noradrenaline. This suggests that serotonin has a non-redundant role in maximal voluntary contractions but still begs the question of exactly how important it is in typical motor tasks. Nonetheless, 5-HTrelated changes in muscle activation typically emerge with strong contractions for both the unfatigued and fatigued motor system. Thus, it appears that the magnitude of descending drive to the muscle may be aligned with the level of 5-HT neuromodulation in humans. Indeed, we are beginning to reveal evidence where 5-HT-effects may be scaled to the intensity of muscle activation in humans (Goodlich et al., 2022; Henderson et al., 2022)".
Line 378: neuromodulation of what? This sentence is rather vague as 5HT is a neuromodulator and it can be inferred that it would contribute to neuromodulation. I would suggest stating that 5HT is a potent modulator of spinal circuits and motoneuron output.
As suggested, we have changed this sentence to, "Animal experiments provide clear evidence that 5-HT is a potent modulator of spinal circuits and motoneurone output".
Line 393-396: this sentence is a little misleading as it implies that pharmacological manipulation of neurotransmitters in the CNS is unique to humans. However, I think that the main point is detailed in the sentence that follows. Perhaps consider rephrasing these two sentences.
This was a poorly worded sentence where we should not have used the word 'unique'. We have removed this term and stated in the revised manuscript: Participant safety is an additional challenge in human experiments that use pharmacology to manipulate neurotransmitter activity. Human studies must operate within a window of safe drug administration, and typically use therapeutic doses of 5-HT modulating medications. Thus, very little is known about how dosage effects influence 5-HT activity in humans. We thank Referee #3 for this suggestion. Please note that the intent of the figure was to provide an overview of the 5-HT system and not to provide specific details across species. If we were to modify the figure itself to highlight where all of the data was derived from, it would compromise the layout and clarity of its content. Instead, we have highlighted for caption 1A that the subsequent information is derived from the mammalian spinal cord, and for caption 1B that the subsequent information is derived from preparations of the adult turtle spinal cord. The source reference for the turtle study is listed at the bottom of the caption.

REFEREE #2
It was with pleasure that I read the work submitted by Kavanagh and Taylor entitled "Voluntary activation of muscle in humans: does serotonergic neuromodulation matter?" For many years, the field has been left wondering if PICs that are facilitated by 5HT are actually relevant to human function. This topical review starts to address this, and other, aspects of 5HT neuromodulation during normal and fatigued motor output. It is with great certainty that I can say that this article will be highly read and cited. Great work! Below, I have provided some specific comments (with page ; and line [l]) in an attempt to improve the quality of the manuscript.
We thank Referee #2 for these kind comments. We believe that our Topical Review has significantly improved with the recommendations that the referee has made. p2; l57: Ionotropic/neuromodulatory is not necessarily a class of input, rather they are inputs that activate classes of receptors. The inputs activate either ionotropic or metabotropic receptors and neuromodulatory inputs would predominantly activate those of the metabotropic class. Slight rewording for clarity would be helpful here.
We have modified this sentence to read, "Synaptic inputs to motoneurones will typically activate two classes of receptors: ionotropic and neuromodulatory". p5; l137: is this supposed to refer to rates of discharge?
The referee is correct. We have modified the sentence to clarify that we are referring to discharge rate, "Incremental increases in treadmill walking speed correspond to incremental increases in the discharge rate of single raphe-spinal fibres".
Section ending on l142: The discussion of the role of 5ht on the dorsal cord seems a little light -one might consider a slight expansion of this topic here so that the (predominantly) inhibitory effects of 5ht on sensory transmission can be appreciated.
We have added additional information to this paragraph. In doing so, we have been cautions with the amount of information that we present. In particular, the role that 5-HT plays in regulating afferent and interneuron excitability is extraordinarily complex, and to do justice to the topic would require a review by itself. We have included the following information in the revised manuscript:  , 1993). Adding further complexity to our understanding of how 5-HT regulates sensory neurone in the spinal cord is that 5-HT effects may be exerted presynaptically and post-synaptically. In particular, actions of muscle spindle (Ia fibres) and tendon organ (II fibres) afferents on spinal interneurons by 5-HT is dependent on the type of the afferent that is activated and the functional type of the interneuron it is connected to (Jankowska et al., 2000). Thus, it is difficult to predict how motor function may be influenced by the actions of 5-HT on afferent neurons and interneurons".
p7; l207-209: It may be nice to put this magnitude of change into perspective for the naive reader. If a healthy young adult has near complete activation of the biceps (near 100%; see pre-fatigue %VA in Fig 3) then we would not expect that they could gain more than a percent or two as they cannot have >complete activation (theoretical, I know). Other muscles, with lower %VA, may show greater change.
We agree that more detail could have been provided to give our voluntary activation data context for a naïve audience. Hence, we have modified the text as follows: A finding of only small drug-related increases in voluntary activation is perhaps not surprising, as increases in motoneurone excitation during near-maximal contraction intensities produce only small changes in interpolated twitch amplitude (Herbert & Gandevia, 1999). Hence, voluntary activation of the biceps (that is calculated from interpolated twitches) may be as high as 98% or 99% in healthy individuals, and any intervention that may increase activation is limited by a very close ceiling (i.e. 100% voluntary activation).
We have placed these words in parenthesis in the revised manuscript.
p8; l236-237: Although seldomly considered in the literature, the threshold of the PIC can vary and may be a key contributor to MN recruitment. Please refer to the discussion of Afsharipour et al 2021