Electric shock causes a fleeing-like persistent behavioral response in the nematode Caenorhabditis elegans

Abstract Behavioral persistency reflects internal brain states, which are the foundations of multiple brain functions. However, experimental paradigms enabling genetic analyses of behavioral persistency and its associated brain functions have been limited. Here, we report novel persistent behavioral responses caused by electric stimuli in the nematode Caenorhabditis elegans. When the animals on bacterial food are stimulated by alternating current, their movement speed suddenly increases 2- to 3-fold, persisting for more than 1 minute even after a 5-second stimulation. Genetic analyses reveal that voltage-gated channels in the neurons are required for the response, possibly as the sensors, and neuropeptide signaling regulates the duration of the persistent response. Additional behavioral analyses implicate that the animal's response to electric shock is scalable and has a negative valence. These properties, along with persistence, have been recently regarded as essential features of emotion, suggesting that C. elegans response to electric shock may reflect a form of emotion, akin to fear.

Two experts in the field and myself have reviewed your manuscript.I am pleased to inform you that, with minor revisions, it is potentially suitable for publication in GENETICS.The reviewers have comments and concerns that need to be addressed in a revised manuscript.You can read their reviews at the end of this email.
It is most important that you address the following in your resubmission.The evidence for the behavioural response to AC being a primitive for fear is not strong enough yet.It is unclear whether the behavioural response has negative valence and whether it can be generalised (two of the characteristics extracted by Anderson and Adolphs).One way to address both of these is to test whether the AC treatment can negatively condition a stimulus (salt, benzaldehyde or butane would be good candidates) in a simple associative learning task.In addition, some knowledge of the neurons involved in sensing or executing the behavioural responses to AC would highly improve the manuscript.There are some suggestions about how to do this by the reviewers below.
When you resubmit, please let the editorial office know approximately how long you expect to need for revisions.
Upon resubmission, please include: 1.A clean version of your manuscript; 2. A marked version of your manuscript in which you highlight significant revisions carried out in response to the major points raised by the editor/reviewers (track changes is acceptable if preferred); 3. A detailed response to the editor's/reviewers' comments and to the concerns listed above.Please reference line numbers in this response to aid the editors.
Additionally, please ensure that your resubmission is formatted for GENETICS.https://academic.oup.com/genetics/pages/general-instructionsFollow this link to submit the revised manuscript: Link Not Available Sincerely,

Arantza Barrios Associate Editor GENETICS
Approved by: Oliver Hobert Senior Editor GENETICS --------------------------------------------------------------------Reviewer #1 (Comments for the Authors (Required)): Tee et al provide a fun study that explores the genetic basis for how C. elegans responds to AC electrical stimuli.They tested a wide range of parameters in WT and many different mutant backgrounds.Their most salient results are that certain AC stimuli cause the worm to maintain its speed during stimulation but speed up afterward, whereas strong stimuli tend to cause the worm to speed up only during stimulation.These responses appear to require calcium (egl-19 and unc-2) and potassium (slo-1) channels, which somewhat matches how other electrically sensitive animals sense electrical stimuli.The characteristic decline of response after stimuli appears to require an unspecified collection of neuropeptides because egl-3 mutants exhibited an unusually long response.The study is well conducted and controlled.The statistics appear appropriate.Major issues: 1.It would have been interesting to have determined which sensory neurons, if any, are required for these AC responses; however, they found that many sensory pathways appear to be dispensable for this response.This raises the possibility that multiple sensory pathways sense AC redundantly in parallel, and they missed this result because they didn't break the right combination of sensory neurons.This needs to be stated in the Discussion.Alternatively, sensory neurons might not even be required to respond.Indeed, the calcium and potassium channels that are required for response are expressed throughout the nervous system.I wonder if they could use cell-selective expression of RNAi to knock-down egl-19 in at least a large subset of sensory neurons to deduce whether sensory neurons are in fact required for this response?Or use RNAi by feeding in a worm strain that expresses sid-1 in sensory neurons?This would be very interesting.Some useful public strains include: TU3403 (touch neurons), XE1375 (GABA), XE1474 (dopamine), XE1581 (ACh), and XE1582 (glutamatergic).
2. I cannot tell from your experiments if the heightened speed during or after shock reflects "fear" or other types of emotion.Do any of the AC stimuli serve as effective unconditioned cues for associative learning?This can be easily tested.Without testing this, I find it hard to know if the worms are "excited", "happy", or "fearful" when maintaining their speed during AC or moving fast afterwards.Why not test whether worms form a negative association with innately attractive NaCl in the agar substrate with a chemotaxis assay after training.Then you might have clear results.There are other useful ways to test for negative association too.Alternatively, I would be happy if they just toned down the idea of fear without more experimental evidence.
Minor issues: Line 67. "express" probably should be rewritten as "reflect" Line 68. ...the response if not mediated by any single well-known... (you need to say this because they may be mediated by more than one in redundantly in parallel) Line 73. ...indicate that the response of C. elegans to electric shock... Line 83.Please add info here and elsewhere in Results and Discussion from Chrisman et al., 2016.They got distinct results from Gabel et al. 2007 Line 100.Was there any mutant that speeds up *during* this stimuli?Lines 109-113.This is confusing as written.I'd recommend taking out this out here and elaborate more carefully in the Discussion.Figure 2A and B. The think lines are supposed to reflect individual measurements and not SEM.Then why does absolute speed measurements appear to drop below zero?This doesn't make sense.Figure 2C.I'm confused why you needed to have worms start on the food.Why not have the worms start anywhere.The presence of the food seems to complicate interpretation of results because you do not clarify in most cases when worms are on or off food. Figure 4A.I'd recommend adding sublabels to help explain the six results you have within this single panel.For instance, the bottom one where you stimulate for 600 seconds is interesting.You speculate that the worms may be experiencing exhaustion.To determine this, you might try using laser blue light to stimulate the worms.This may dishabituate the worms and cause them to speed up again, or not. Figure 5.You could try backing up your egl-19 results with L-type calcium channel blocker that effects C. elegans.This would test if you get the same result even if you transiently block function of the channel rather than lack EGL-19 channel the entire life of the animal.
Figure 5 E and F.You are measuring different time windows in these two panels.To help clarify that, I'd recommend adding text above these graphs to distinguish During or After AC stimulation.Figure 6.I seem to remember that null mutations in mec-10 can sense touch because with it missing, MEC-4 can form homomeric channels.The mec-10 hypomorphic mutants cannot sense touch.Is mec-10(tm1552) defective in touch or not? Figure 6G.Was this peak speed?When?Maybe explain with text above the graph?Discussion: Clarify which neuropeptides are effected or not effected by egl-3 mutation.Line 296: I thought according to Fig 5F you did see a positive result for mec-4.Line 308: citation is missing Line 343: I'm interested in seeing how the duration and strength of the electric shock here compares with those used in fear conditioning paradigms in other experimental animal models and humans.
Reviewer #2 (Comments for the Authors (Required)): The present study by Tee et al., demonstrates that C. elegans exhibits a persistent speed increase in response to AC electric stimulation thus reflecting a fear-like state in the animal that has not been reported previously.The authors have nicely demonstrated that food does not influence the emotional state of behavior through well designed experiments.This study also identifies certain genes that are required to elicit this response.While I believe this study is well presented and may evoke strong interest in the field, I do have some minor concerns regarding certain experiments and would also recommend one key experiment be added before its publication.
1.In Figure 1 supplemental video 3 and 4, the animals shown when electrically stimulated do not show any movement as if a state of shock.However, the authors claim this is not the case for every plate.I would suggest showing a representative video and explain why animals show such varied response.Is the voltage or frequency not uniform in the plate?
2. The authors should address why would DC and AC current have totally different response in the animal.Can the animal sense difference between DC and AC currents as one would expect the animal to just respond to electric stimulation?3. I would suggest moving Figure 1 that shows the set up to supplemental figure and moving Supplemental Figure 2 that highlights on and off response to the main figures.4. Since fleeing response is mostly forward locomotion and bends, it wasn't clear in the list of genes analyzed, if the genes involved in this locomotory behavior were tested.5. To gain insight into how persistent behavior is generated, it is important to demonstrate the key neurons that influence behavior states in the animal.To this effect, the authors have identified a few genes, but do not try to uncover the neurons that might be affected.It wouldn't be hard to express calcium sensor in neurons (for example the motor neurons) and image these during before, during and after electric stimulation to identify which neurons specifically respond during fleeing.
Tee et al. present a novel experimental paradigm for persistent behavior in C. elegans induced by transient electric shocks with AC stimulation.This behavior is modulated by voltage, duration and frequency of the electric stimulus, showing an ON response at a lower voltage (30 V) and an OFF response at higher voltage (75 V).They demonstrate that the behavior shows persistency, that is, it lasts for up to several minutes after a short stimulus; scalability -meaning that it varies in duration with different intensity of the stimulus; and that it has negative valence, where the response to the presentation of the electric shock overrides foraging behavior.The authors also show that the avoidance response depends on the EGL-19 and UNC-2 voltage-gated calcium channels, as well as the SLO-1 BK channel -but is not dependent on monoamine signaling.Instead, its persistency is modulated by neuropeptide signaling.Interestingly, these avoidance responses to AC stimuli appears to be different from electrosensory behavior to DC stimuli published by the Samuel lab in 2007 -for example, DC responses are dependent on five amphid neurons, whereas mutants that eliminate amphid function still display the AC responses.
Persistent behavior and concomitant persistency in brain states represent the different brain states found in humans and other animals, such as motivation or emotion.The behavioral paradigm presented entails three of four hallmarks of emotion that were proposed by Anderson and Adolphs, who defined these characteristics specifically with the study of animal models of emotion in mind.Because of the importance of persistent behavioral states for an understanding of the function of the brain, it is highly valuable that this study establishes a new paradigm in the C. elegans genetic model.Electric shocks are extensively used in the study of associative learning in other models, especially rodents, and a comparable C. elegans paradigm can be useful for the study of conserved principles of cognition across species.The establishment of the paradigm opens up the possibility to use electric shock as an unconditioned stimulus in experiments on associative learning in C. elegans.In this regard, I note that emotional processing e.g. in the amygdala is modulated by monoamine signaling, especially dopamine and serotonin.It may be possible to uncover a role of these modulators in AC electrosensation in future studies.
Since the study explicitly interpret their finding in the light of the concept of four hallmarks of emotion, it would significantly benefit this study -and strengthen the argument made by the authors -to also address the fourth hallmark, generalization of emotional states.This could be done for example by testing if responses to a sensory stimulus of a different modality, applied shortly afterwards, are altered by a previously applied electric shock -such as a blue light stimulus.A different way of address the question of generalization would be to test if the emotional response demonstrates pleiotropy: that is, if there are consequences of an emotion state that "fan out" to a multitude of effects in response to the sensory stimulation, such as vegetative effects.Examples in C. elegans would be to measure if pharyngeal pumping or defecation frequency also change after a short electric shock.
Overall, the experiments have been carefully executed and appropriately interpreted.
Reviewer #3 (Comments for the Authors (Required)): Importantly, the authors address two caveats -whether the increase in speed is due to food leaving, induced by the electric stimulus; and whether it is due to an increase in temperature.To further address this caveat, it would be desirable to measure the temperature on the surface of the agar plate with the electric stimulus regime applied here.Thirdly, the OFF behavior shown after 75 V stimuli, where no or little increase in speed occurs during the stimulation, could simply be an inhibition of locomotion by interference with neural communication due to the externally applied strong electric field.The authors provide several arguments however that this is not the case -for instance, that a reduction of salt to decrease the current does not change the OFF behavior.Instead, because it occurs at a particular high voltage, it could be an example of a "freeze" response to a strong shock; this can be seen in C. elegans as well in response to particularly harsh touch.
In the discussion, I miss a discussion of how this paradigm is actually also relevant for arousal, which is a concept closely related to emotion.Indeed, the hallmarks of persistency, scalability and valence would seem to apply to arousal states as well.

Specific points:
-At least to demonstrate a causative role of egl-19 in AC responses, a transgenic rescue of the gene should be tested if it restores the avoidance behavior.Such a genomic rescue strain has been published by Gao and Zhen (10.1073/pnas.1012346108).Notably, in this study they used the strain to generate mosaic animals that express egl-19 only in neurons but not muscles.Data from such a neuron-specific rescue would provide further support for the hypothesis that EGL-19 may act as part of an electrosensory mechanism in neurons.
-Methods section: A description of how the stripes of food were seeded is not provided.
-Statistics: The authors are performing the Kruskal-Wallis test on repeated measures for figures 2 and 3 (the ones with speed measurements before, during and after electric stimulationassuming that it is the same animals from which three measurements are taken at different time points based on the description), which is not appropriate for these data because Kruskal-Wallis requires the samples to be independent.A Friedman test (also non-parametric, but designed for repeated measures/related samples) would be adequate instead.For the other figures (5 and 6), the Kruskal-Wallis tests used are appropriate, as here they are used to compare between different genotypes.-It is not mentioned if any pre-processing of the data was done to check for the shape of their distributions or the variance sizes.Speed data often show normal distributions, which would allow to employ a parametric test instead.
In conclusion, I support acceptance of this study for publication, once these points have been considered I am pleased to accept your manuscript entitled "Electric shock causes a fleeing-like persistent behavioral response in the nematode Caenorhabditis elegans" for publication in GENETICS, pending minor revision (as suggested by reviewer 1).
All reviewers and myself empathise with the authors' struggle to address empirically some of the concerns that were initially raised and we are satisfied with your responses.Please submit your revision along with a brief description of how you modified the manuscript in response to the reviewers' concerns and suggestions (which can be viewed at the bottom of this email).Mainly, the revisions should include the couple of citations that reviewer # 1 highlights.I expect you should be able to submit a revised manuscript within 30 days.A suitably revised manuscript will be acceptable for publication; I don't expect to send it out for review.
When revising the ms., please make an effort to shorten it, because that almost always improves a manuscript.We urge authors to heed the advice of Strunk and White: "omit needless words"1.Follow this link to submit the revised manuscript: Link Not Available Thank you for submitting this story to Genetics.

Sincerely, Arantza Barrios Associate Editor GENETICS
Approved by: Oliver Hobert Senior Editor GENETICS ----------------------------------------------------------------------------Reviewer comments: Reviewer #1 (Comments for the Authors (Required)): Tee et al. provide an update on their fun study that examines how C. elegans responds to electric shock.They have worked hard to solidify their finding that egl-19 is required for the response.They also attempted to check if shock conferred a negative association in plasticity paradigm.I'm overall satisfied and happy with their response.I have a few comments below.I totally respect the Iino's lab expertise on associative learning with NaCl as a cue.However, I sense that Tee et al may have performed suboptimal follow-up experiments to test if electric shock might serve as am unconditioned negative (fearful) stimulus.Chemotaxis assays can be set up to test naïve *attraction* to moderate concentrations of salt (e.g.Ward, 1973), but the authors instead show avoidance of higher concentrations of salt (Figure 1 in response).For attraction, the chemotaxis plate usually contains a background concentration of zero with a salt gradient.In this condition, the worms robustly perform positive chemotaxis.If worms are starved on NGM plates containing normal salt concentration, they will perform chemotaxis with a negative or neutral (0) index in these conditions, signifying that they associated the salt with a negative experience (Iino's JEB paper).The authors claim that they tested a full range of parameters, but I'm not sure if the authors actually performed this experiment and tested whether shock could substitute for starvation as a negative stimulus.As written, it is hard to know if by "spots" they mean small microliter volume drops that would quickly diffuse, resulting in the reported concentration of salt at the two locations (e.g.I doubt there was zero salt).-Nevertheless, I agree with the authors that simply toning down language on "fear" and "emotion" is appropriate without offering further positive results.So I'm fine with their revised paper.I appreciate Tee et al's struggle to test cellular loci that underlie the behavioral responses to shock.I agree that RNAi for behavioral phenotypes can be tricky, and as such negative results need to be carefully considered.Their new positive results