Article activity feed

  1. eLife

    Author Response

    Reviewer #2 (Public Review):

    The manuscript presents an interesting study on a timely topic (hyperacusis). The study was carried out in awake animals using modern approaches in neurosciences (calcium imaging, optogenetic). The amount of data is impressive, the study is very ambitious, and overall its quality is indisputable. However, I have some general comments and questions on some concepts that are critical for the study, and also on the interpretation of the data, in particular the behavioral data.

    We appreciate Reviewer 2’s overall positive evaluation as well as their more specific critiques, which we address below.

    The first point I want to mention is the concept of 'homeostatic plasticity'. I am not sure we agree on its definition. My understanding of it is that the AVERAGE of central activity will remain constant around a set point value. In case of a reduction of sensory inputs (hearing loss), the neurons' sensitivity will be enhanced in such a way that the averaged activity will be preserved. So, neural hyperactivity after partial or sensory deprivation is not 'maladaptive': it is a collateral effect, 'the price to pay' for maintaining neural activity stable around a given value. In my opinion, this point is crucial. The authors should also mention and cite the model's paper from Schaette et al.

    “Homeostasis” is a term used widely in physiology to describe a negative feedback process in which an internal adjustment compensates for an external perturbation to return a given system (temperature, pH, etc.) to a set point. To the reviewer’s point, homeostatic processes – broadly defined – can work at many different biological scales including perhaps large, distributed systems like the example s/he gave of neurons throughout the central auditory pathway. By contrast, “homeostatic plasticity” is a mechanism studied by dozens of laboratories in hundreds of papers by which neurons (typically studied in cortical neurons) adjust their synaptic and intrinsic excitability to maintain their activity around a set point range. A key feature of homeostatic plasticity is that neurons “sense” deviations from their set point and initiate a compensatory process to offset this deviation. Up to this point, it seems that we are on the same page as the reviewer.

    The first point of possible disagreement lies in the interpretation of how excess neural activity relates to homeostatic plasticity. The reviewer mentioned modeling papers by Schaette and Kempter (2006, 2007, 2012) on the cochlear nucleus, which are also based on homeostatic plasticity and their work is now cited in the revised text (see line 71). The reviewer is correct that there is a difference in how the term is used and interpreted, but the difference is fairly subtle. Their work and our work propose that homeostatic plasticity processes are applied within a single neuron to offset the reduced afferent input that accompanies cochlear damage. As the reviewer recalled, they describe hyperactivity as a consequence of this compensation, as we do as well. The only difference is that they and the reviewer describe hyperactivity as the byproduct of the normal, successful implementation of homeostatic plasticity, which it unequivocally is not because – by definition – homeostatic plasticity is a stabilizing process that maintains activity at a predetermined set point range.

    The second point of disagreement lies in the reviewer’s statement that “neural hyperactivity after partial or sensory deprivation is not 'maladaptive': it is a collateral effect, 'the price to pay' for maintaining neural activity stable around a given value.” We disagree. Hyperactivity can be both a collateral and maladaptive effect. Hyperactivity and hypersynchrony are understood to be the basis of tinnitus, which is a maladaptive, disordered state. The reviewer’s comment implies that there is no alternative for compensating for sensory deprivation but to make cortical neurons hyperactive. We see no reason why this must be so. In fact, stabilization of activity rates after sensory deprivation has been demonstrated in hundreds of studies in the developing visual system. In the adult auditory system, activity in cortical neurons is initially depressed after injury before rebounding to exceed baseline levels (see Resnik Polley 2017 eLife, Asokan 2018 Nat Comm., Resnik Polley 2021 Neuron). It is not obligatory for cortical activity rates to pass through the set point range and continue into hyperactivity, nor is it obligatory for cortical activity rates to remain elevated above baseline many days after the injury. Additional evidence for this point comes from Figures 4, 6, and 8, which show that some cortical neurons actually do homeostatically regulate their activity back to baseline (i.e., show stable gain). This raises the intriguing question of why some neurons recover to their homeostatic activity set point while others do not. Figure 8 provides new insight into this question by showing that that their baseline response properties can account for 40% of the variability in gain stabilization after peripheral insult.

    A third point of disagreement related to the reviewer’s statement that “My understanding of it is that the AVERAGE of central activity will remain constant around a set point value. In case of a reduction of sensory inputs (hearing loss), the neurons' sensitivity will be enhanced in such a way that the averaged activity will be preserved”. We agree that homeostatic plasticity processes are influenced by activity propagating through distributed neural networks. However, the biological implementation of the process is programmed into individual neurons. The activity set point is neuron-specific, the error signal that encodes a deviation from the set point is neuron-specific, and the transcriptional/translational changes deployed to stabilize the activity rate are neuron-specific. As an analogy, home climate control systems work autonomously for each house, because the sensors (thermostat) and actuators (heating/cooling) are sensitive to fluctuations in that home, not across other houses in the town. The heating and cooling systems for each house in town may be driven by a distributed, common source (e.g., a hot day) but the mechanisms that bring the ambient temperature back to the set point for each house are autonomous and reflect the particular thermostat programming for each house. The widely studied homeostatic plasticity mechanisms mentioned in our manuscript (e.g., excitatory synaptic scaling) are not sensitive to and do not target the averaged neural activity among millions of neurons distributed throughout the sensory neuroaxis.

    As a final point on this statement, there is no demonstration that we are aware of that average central activity remains constant after a reduction of sensory inputs. This would require recording from many neurons across multiple stages of the sensory pathway in a single animal to show that the increased gain at later stages in the system exactly offsets the reduced responsiveness at earlier stages of the system. So, the reviewer’s definition of homeostatic plasticity is based on a general supposition about a distributed process that has never been empirically demonstrated whereas the definition we use is consistent with the mechanisms and terminology used throughout the neuroscience literature (albeit often incorrectly in the hearing loss literature).

    The second point is that a lot is built on the behavioral procedure and d'. I am not convinced by the behavioral procedure (and the d') is a convincing measurement of loudness (and therefore loudness hyperacusis). So, in my opinion, the title may be changed and more importantly the entire spirit of the paper should be modified.

    The reviewer’s critique as well as comments from other reviewers helped us realize that we had used the terms “hyperacusis” and “loudness” imprecisely. We think that is part of the confusion. What we have studied here is auditory hypersensitivity after sensorineural hearing loss, which may or may not be a model of why persons with hyperacusis can exhibit loudness hypersensitivity.

    Once “hyperacusis” and “loudness” have been stripped away from the behavior, we contend that we have a behavioral assay for auditory hypersensitivity, which is the main point of our study. To be clear, the behavioral readout most commonly employed in the animal literature to model hyperacusis is reaction time, which has a less direct relationship to hypersensitivity than does d’. D-prime is widely used as the sensitivity index in detection behaviors. The main advantage of d’ is that it controls for differences in response bias either between subjects or after noise exposure. We used the d’ metric to show that mice can more reliably detect tone levels near their sensation threshold and can more reliably detect direct stimulation of thalamocortical projection neurons after acoustic trauma. These observations provide the framework for all of the neural measurements that follow.

    On the balance, the reviewer was correct that our imprecise use of hyperacusis and loudness was confusing and contradictory. The terms “hyperacusis” and “loudness” now only appear in the manuscript to describe other published findings or to describe what our study does not address. This resulted in several small text changes throughout the manuscript as well as a direct statement about the relationship between our work, loudness, and hyperacusis on Pg. 14, Lns 448-466.

    “While the findings presented here support an association between sensorineural peripheral injury, excess cortical gain, and behavioral hypersensitivity, they should not be interpreted as providing strong evidence for these factors in clinical conditions such as tinnitus or hyperacusis. Our data have nothing to say about tinnitus one way or the other, simply because we never studied a behavior that would indicate phantom sound perception. If anything, one might expect that mice experiencing a chronic phantom sound corresponding in frequency to the region of steeply sloping hearing loss would instead exhibit an increase in false alarms on high-frequency detection blocks after acoustic trauma, but this was not something we observed. Hyperacusis describes a spectrum of aversive auditory qualities including increased perceived loudness of moderate intensity sounds, a decrease in loudness tolerance, discomfort, pain, and even fear of sounds (Pienkowski et al., 2014a). The affective components of hyperacusis are more challenging to index in animals, particularly using head-fixed behaviors, though progress is being made with active avoidance paradigms in freely moving animals (Manohar et al., 2017). Our noise-induced high-frequency sensorineural hearing loss and Go-NoGo operant detection behavior were not designed to model hyperacusis. Hearing loss is not strongly associated with hyperacusis, where many individuals have normal hearing or have a pattern of mild hearing loss that does not correspond to the frequency dependence of their auditory sensitivity (Sheldrake et al., 2015). While the excess central gain and behavioral hypersensitivity we describe here may be related to the sensory component of hyperacusis, this connection is tentative because it was elicited by acoustic trauma and because the detection behavior provides a measure of stimulus salience, but not the perceptual quality of loudness, per se.”

    A lot is derived/interpreted from the results, but I believe there is a lot of over-interpretation. I would suggest the authors be more cautious and moderate in their speculations and conclusions. I would reconfigure the manuscript, and simplify it.

    We believe that the changes mentioned above and in the response to their specific comments below reduce over-interpretation and simplify the manuscript.

    As an example of a change made to moderate the conclusions from our work, we added the following to Pg. 14, Lns 442-447

    “Further, while the perceptual salience (Figure 2) and neural decoding of spared, 8kHz tones (Figure 5) were both enhanced after high-frequency sensorineural hearing loss, these measurements were not performed in the same animals (and therefore not at the same time). Definitive proof that increased cortical gain is the neural substrate for auditory hypersensitivity after hearing loss would require concurrent monitoring and manipulations of cortical activity, which would be an important goal for future experiments.”

    Reviewer #3 (Public Review):

    The study uses a mouse animal model of sensorineural hearing loss after sound overexposure at high frequencies that mimics ageing sensorineural hearing loss in humans. Those mice present behavioural hypersensitivity to mid-frequency tones stimuli that can be recreated with optogenetic stimulation of thalamocortical terminals in the auditory cortex. Calcium chronic imaging in pyramidal neurons in layers 2-3 of the auditory cortex shows reorganization of the tonotopic maps and changes in sound intensity coding in line with the loudness hypersensitivity showed behaviourally. After an initial state of neural diffuse hyperactivity and high correlation between cells in the auditory cortex, changes concentrate in the deafferented high-frequency edge by day 3, especially when using mid-frequency tones as sound stimuli. Those neurons can show homeostatic gain control or non-homeostatic excess gain depending on their previous baseline spontaneous activity, suggesting a specific set of cortical neurons prompt to develop hyperactivity following acoustic trauma.

    This study is excellent in the combination of techniques, especially behaviour and calcium chronic imaging. Neural hyperactivity, increase in synchrony, and reorganization of the tonotopic maps in the auditory cortex following peripheral insult in the cochlea has been shown in seminal papers by Jos Eggermont or Dexter Irvine among others, although intensity level changes are a new addition. More importantly, the authors show data that suggest a close association between loudness hypersensitivity perception and an excess of cortical gain after cochlear sensorineural damage, which is the main message of the study.

    The problem is that not all the high-frequency sensorineural hearing loss in humans present hyperacusis and/or tinnitus as co-morbidities, in the same manner that not all animal models of sensorineural hearing loss present combined tinnitus and/or hyperacusis. In fact, among different studies on the topic, there is a consensus that about 2/3rds or 70% of animals with hearing loss develop tinnitus too, but not all of them. A similar scenario may happen with hearing loss and hyperacusis. Therefore, we need to ask whether all the animals in this study develop hyperacusis and tinnitus with the hearing loss or not, and if not, what are the differences in the neural activity between the cases that presented only hearing loss and the cases that presented hearing loss and hyperacusis and/or tinnitus. It could be possible that the proportion of cells showing non-homeostatic excess gain were higher in those cases where tinnitus and hyperacusis were combined with hearing loss.

    We thank the reviewer for her/his careful reading of the original manuscript and many helpful suggestions and critiques that have been addressed in the revision. Both Reviewer 2 and Reviewer 3 understood that we were presenting our high-frequency sensorineural hearing loss manipulation as a way to model the clinical phenomenon of hyperacusis. This was not our intent, and we regret the wording of the original manuscript communicated this point. In fact, the clinical literature shows that hyperacusis does not have a strong association with hearing loss and moreover our behavioral and neural outcome measures were not designed to index the core phenotype of hyperacusis (a spectrum of sound-evoked distress, disproportionate scaling of loudness with sound level, and sound-evoked pain). Our study addresses the neural and behavioral signatures of auditory hypersensitivity, which is an “upstream” condition that may (or may not) be related to the presentation of clinical phenomena like hyperacusis and tinnitus.

    The reviewer mentions a litmus test for animal models of tinnitus, in which the utility of an animal model for tinnitus would be evaluated in part based on whether a controlled insult only produced a behavioral change suggestive of a chronic phantom percept in a fraction of animals. That may be so, but our study is clearly not modeling tinnitus and we make no claims to this effect in the original or revised manuscript. The Reviewer then goes on to say that “a similar scenario may happen with hearing loss and hyperacusis”. “May” is the operative word here because the association between sensorineural hearing loss and the clinical presentation hyperacusis is quite weak overall in human subjects but no study (that we are aware of) has attempted to document the probabilistic appearance of hyperacusis before and after acoustic trauma. So, we really don’t know whether hyperacusis has a probabilistic appearance like tinnitus or is more deterministic like cochlear threshold shift. But, again, the main point is that our experiments make no direct claim about hyperacusis one way or the other, which we now clarify and discuss throughout the revised text, as detailed below.

    We do contend that our experiments allow us to study auditory hypersensitivity, though again there is no precedent or consensus in the literature for expecting auditory hypersensitivity to present probabilistically or deterministically across mice after a controlled insult. Regardless, we agree with the reviewer that it is a very good idea to provide the individual animal data to the reader. We added new panels to Figure 2C to show that an increase in the 8kHz d’ slope after noise exposure (i.e., a change > 1) was observed in 7/7 mice that underwent acoustic trauma but 1/6 mice in the sham exposure group, suggesting a deterministic, binary behavioral effect found in every mouse with noise-induced high-frequency sensorineural damage. On the other hand, within the acoustic trauma cohort, 3 mice showed marked increases in the d’ growth slope (> 2) while 4 showed more subtle changes, suggesting a more graded or probabilistic effect. By providing the individual animal data as per the Reviewer’s request, the reader can now make a more informed determination about the reliability of auditory hypersensitivity within the acoustic trauma cohort.

    Regarding the relationship between the peripheral/cortical/perceptual auditory hypersensitivity we report here and the clinical conditions of tinnitus and hyperacusis, we revised the text such that the word “hyperacusis” only appears in the context of other publications and have added the following text (Pg. 14, Lns 448-466).

    “While the findings presented here support an association between sensorineural peripheral injury, excess cortical gain, and behavioral hypersensitivity, they should not be interpreted as providing strong evidence for these factors in clinical conditions such as tinnitus or hyperacusis. Our data have nothing to say about tinnitus one way or the other, simply because we never studied a behavior that would indicate phantom sound perception. If anything, one might expect that mice experiencing a chronic phantom sound corresponding in frequency to the region of steeply sloping hearing loss would instead exhibit an increase in false alarms on high-frequency detection blocks after acoustic trauma, but this was not something we observed. Hyperacusis describes a spectrum of aversive auditory qualities including increased perceived loudness of moderate intensity sounds, a decrease in loudness tolerance, discomfort, pain, and even fear of sounds (Pienkowski et al., 2014a). The affective components of hyperacusis are more challenging to index in animals, particularly using head-fixed behaviors, though progress is being made with active avoidance paradigms in freely moving animals (Manohar et al., 2017). Our noise-induced high-frequency sensorineural hearing loss and Go-NoGo operant detection behavior were not designed to model hyperacusis. Hearing loss is not strongly associated with hyperacusis, where many individuals have normal hearing or have a pattern of mild hearing loss that does not correspond to the frequency dependence of their auditory sensitivity (Sheldrake et al., 2015). While the excess central gain and behavioral hypersensitivity we describe here may be related to the sensory component of hyperacusis, this connection is tentative because it was elicited by acoustic trauma and because the detection behavior provides a measure of stimulus salience, but not the perceptual quality of loudness, per se.”

  2. eLife

    Evaluation Summary:

    This study uses a mouse model of hyperacusis to further explore the hypothesis that this condition may be mediated by cortical hyperactivity. The authors here provide interesting optogenetic and calcium imaging experiments that reinforce this hypothesis and refine our understanding of the related plastic changes that are involved.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1, Reviewer #2 and Reviewer #3 agreed to share their names with the authors.)

  3. eLife

    Reviewer #1 (Public Review):

    The authors of this manuscript aimed at demonstrating the hypothesis that hyperacusis is triggered by increased sensitivity in mid-range frequency following high-frequency cochlear trauma. The study combines a large variety of careful physiological and behavioral measurements that converge toward the above-mentioned interpretation, which was proposed in an earlier report. This will likely boost the development of hyperacusis mouse models which is beneficial for future treatments.

  4. eLife

    Reviewer #2 (Public Review):

    The manuscript presents an interesting study on a timely topic (hyperacusis). The study was carried out in awake animals using modern approaches in neurosciences (calcium imaging, optogenetic). The amount of data is impressive, the study is very ambitious, and overall its quality is indisputable. However, I have some general comments and questions on some concepts that are critical for the study, and also on the interpretation of the data, in particular the behavioral data.

    The first point I want to mention is the concept of 'homeostatic plasticity'. I am not sure we agree on its definition. My understanding of it is that the AVERAGE of central activity will remain constant around a set point value. In case of a reduction of sensory inputs (hearing loss), the neurons' sensitivity will be enhanced in such a way that the averaged activity will be preserved. So, neural hyperactivity after partial or sensory deprivation is not 'maladaptive': it is a collateral effect, 'the price to pay' for maintaining neural activity stable around a given value. In my opinion, this point is crucial. The authors should also mention and cite the model's paper from Schaette et al.

    The second point is that a lot is built on the behavioral procedure and d'. I am not convinced by the behavioral procedure (and the d') is a convincing measurement of loudness (and therefore loudness hyperacusis). So, in my opinion, the title may be changed and more importantly the entire spirit of the paper should be modified.

    A lot is derived/interpreted from the results, but I believe there is a lot of over-interpretation. I would suggest the authors be more cautious and moderate in their speculations and conclusions.

    I would reconfigure the manuscript, and simplify it.

  5. eLife

    Reviewer #3 (Public Review):

    The study uses a mouse animal model of sensorineural hearing loss after sound overexposure at high frequencies that mimics ageing sensorineural hearing loss in humans. Those mice present behavioural hypersensitivity to mid-frequency tones stimuli that can be recreated with optogenetic stimulation of thalamocortical terminals in the auditory cortex. Calcium chronic imaging in pyramidal neurons in layers 2-3 of the auditory cortex shows reorganization of the tonotopic maps and changes in sound intensity coding in line with the loudness hypersensitivity showed behaviourally. After an initial state of neural diffuse hyperactivity and high correlation between cells in the auditory cortex, changes concentrate in the deafferented high-frequency edge by day 3, especially when using mid-frequency tones as sound stimuli. Those neurons can show homeostatic gain control or non-homeostatic excess gain depending on their previous baseline spontaneous activity, suggesting a specific set of cortical neurons prompt to develop hyperactivity following acoustic trauma.

    This study is excellent in the combination of techniques, especially behaviour and calcium chronic imaging. Neural hyperactivity, increase in synchrony, and reorganization of the tonotopic maps in the auditory cortex following peripheral insult in the cochlea has been shown in seminal papers by Jos Eggermont or Dexter Irvine among others, although intensity level changes are a new addition. More importantly, the authors show data that suggest a close association between loudness hypersensitivity perception and an excess of cortical gain after cochlear sensorineural damage, which is the main message of the study.

    The problem is that not all the high-frequency sensorineural hearing loss in humans present hyperacusis and/or tinnitus as co-morbidities, in the same manner that not all animal models of sensorineural hearing loss present combined tinnitus and/or hyperacusis. In fact, among different studies on the topic, there is a consensus that about 2/3rds or 70% of animals with hearing loss develop tinnitus too, but not all of them. A similar scenario may happen with hearing loss and hyperacusis. Therefore, we need to ask whether all the animals in this study develop hyperacusis and tinnitus with the hearing loss or not, and if not, what are the differences in the neural activity between the cases that presented only hearing loss and the cases that presented hearing loss and hyperacusis and/or tinnitus. It could be possible that the proportion of cells showing non-homeostatic excess gain were higher in those cases where tinnitus and hyperacusis were combined with hearing loss.

  10. Version 1 published on bioRxiv