Neuronal temperature perception induces specific defenses that enable C. elegans to cope with the enhanced reactivity of hydrogen peroxide at high temperature

  1. Francesco A Servello
  2. Rute Fernandes
  3. Matthias Eder
  4. Nathan Harris
  5. Olivier MF Martin
  6. Natasha Oswal
  7. Anders Lindberg
  8. Nohelly Derosiers
  9. Piali Sengupta
  10. Nicholas Stroustrup
  11. Javier Apfeld

  • Curated by eLife

    eLife logo

    eLife assessment

    The ability of organisms to cope with environmental stressors can be modified by their physiological conditions as well as life experience. Here, taking advantage of the tractability of the nematode C. elegans, the authors find that exposure to elevated temperatures enhances defenses against peroxides, agents whose toxicity is enhanced by temperature. The finding that a key thermosensory neuron is required for this phenomenon is an important advance in understanding the underlying mechanism; further, the authors' proposal that this is an "enhancer sensing" phenomenon is interesting and thought-provoking. The multidisciplinary approach and mechanistic detail revealed by this work will make it of interest to readers in the fields of sensory biology, signal transduction, and physiology.

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Hydrogen peroxide is the most common reactive chemical that organisms face on the microbial battlefield. The rate with which hydrogen peroxide damages biomolecules required for life increases with temperature, yet little is known about how organisms cope with this temperature-dependent threat. Here, we show that Caenorhabditis elegans nematodes use temperature information perceived by sensory neurons to cope with the temperature-dependent threat of hydrogen peroxide produced by the pathogenic bacterium Enterococcus faecium . These nematodes preemptively induce the expression of specific hydrogen peroxide defenses in response to perception of high temperature by a pair of sensory neurons. These neurons communicate temperature information to target tissues expressing those defenses via an insulin/IGF1 hormone. This is the first example of a multicellular organism inducing their defenses to a chemical when they sense an inherent enhancer of the reactivity of that chemical.

Article activity feed

  1. eLife

    eLife assessment

    The ability of organisms to cope with environmental stressors can be modified by their physiological conditions as well as life experience. Here, taking advantage of the tractability of the nematode C. elegans, the authors find that exposure to elevated temperatures enhances defenses against peroxides, agents whose toxicity is enhanced by temperature. The finding that a key thermosensory neuron is required for this phenomenon is an important advance in understanding the underlying mechanism; further, the authors' proposal that this is an "enhancer sensing" phenomenon is interesting and thought-provoking. The multidisciplinary approach and mechanistic detail revealed by this work will make it of interest to readers in the fields of sensory biology, signal transduction, and physiology.

  2. eLife

    Reviewer #1 (Public Review):

    Here, Servello et al explore the role of temperature and the temperature-sensing neuron AFD in promoting protection against peroxide damage. Unlike many other environmental threats, peroxide toxicity is expected to be temperature-dependent, since its chemical reactivity should be enhanced by higher temperatures. The authors convincingly and rigorously show that transient exposure to 25C, a condition of mild heat stress in C. elegans, activates animals' defenses against peroxides but potentially not other agents. Interestingly, this response requires the temperature-sensing AFD neurons, though whether temperature-dependent AFD activity is itself involved in this regulation is not explored. Further, the authors find that temperature regulates AFD's expression of the insulin ins-39 and provide evidence supporting the idea that repression of ins-39 at 25C contributes to enhanced peroxide defense. The authors use transcriptomic approaches to explore gene expression changes in animals in which AFD neurons are ablated, providing evidence that the FoxO-family transcription factor DAF-16 potentiates AFD signaling. However, because AFD ablation triggers effects broader than transient 25C exposure, the significance of these findings for temperature-dependent peroxide defense is somewhat unclear. Additionally, the possibility that DAF-16 (as well as another protective factor, SKN-1) function in parallel to temperature stress is consistent with many of the results shown but is not as thoroughly considered. Together, these studies identify a fascinating example of pre-emptive threat response triggered by the detection of a potentiator of that threat, a phenomenon they term "enhancer sensing." While some predictions of the specificity of this phenomenon remain untested, the paper provides intriguing insight into the potential mechanisms by which it may occur.

    Major issues:

    The dependence of the enhancer-sensing phenomenon on AFD leads the authors to conclude that the 25C stimulus is sensed by AFD itself, but this needs to be directly tested. To do this, they could ask whether tax-4 function is required in AFD, or use mutants in which AFD's thermosensory function is compromised.

    The enhancer-sensing model is fascinating, but as it stands it is somewhat oversold. The authors could tone down the writing, indicating that this model is suggested rather than shown. Alternatively, they could more carefully test some of its predictions - for example by exploring the response to other threats (e.g. some of the toxicants described in Fig. S5) at 20C and 25C in WT and AFD-ablated animals.

    The role of ins-39 remains somewhat speculative. Fig 4F shows that ins-39 mutants have a reduced induction of peroxide defense, but it seems that this could be the result of a ceiling effect. The authors' model predicts that overexpression of ins-39, particularly at 25C, should sensitize animals to peroxide damage, a prediction that should be tested directly. Further, the authors seem to assume that AFD is the relevant site of ins-39 function, but this needs to be better supported.

    Most of the daf-16 and skn-1 experiments are carried out in AFD-ablated animals, making the relevance of these findings for the 25C-dependent induction of peroxide defense somewhat unclear. As the authors show, AFD ablation causes much more extensive changes than transient 25C exposure, clearly seen in slope of the line in 3C. Further, unlike 25C exposure, AFD ablation is a chronic and non-physiological state. It would be useful for the authors to be cautious in their interpretation of these findings and to be clearer about how strongly they can connect them to the "enhancer sensing" phenomenon. Along these lines, the potentiation idea could be toned down a bit. Much of the data is consistent with parallel function for daf-16 (and skn-1) - for example, Fig 5C indicates additive effects of daf-16 and 25C exposure; 6C shows that AFD ablation still has a clear effect on peroxide sensitivity in the absence of both daf-16 and skn-1; and Fig S8a shows that much of the transcriptional response to AFD ablation (along PC1) is intact in daf-16 animals.

  3. eLife

    Reviewer #2 (Public Review):

    In this study, Servello and the colleagues characterize how a temperature sensing neuron AFD regulates increased resistance to hydrogen peroxide in worms cultivated at a higher temperature. They show that loss of AFD and the insulin-like peptide INS-39 produced by AFD increase H2O2 resistance similarly as high temperature growth. To understand the molecular basis, they use mRNA-seq and analysis of gene expression at the whole-genome scale and transgenic lines to show that AFD ablation and high cultivation temperature generate overlapping changes in gene expression via the function of the FOXO transcription factor DAF-16 in the intestine.

    This study is built on their previous work that established C. elegans as a model to study mechanisms for sensing and resistance of H2O2, an important environmental chemical threat for living organisms. Here, the authors uncover the neuronal and molecular basis for H2O2 resistance induced by high cultivation temperature. The authors use multiple approaches, including genetics, transgenics, whole-genome gene expression analysis, to characterize "enhancer sensing" that they discovered in this study. The experiments are well designed with appropriate controls. The data analysis is comprehensive and revealing. The findings are novel and explain a common and interesting phenomenon. The new understanding generated in this study will appeal to the readers in the fields of sensory biology, signaling transduction and physiology. The implications or conclusions of a few results presented here could be further discussed or clarified in the context of several previous studies.

  4. eLife

    Reviewer #3 (Public Review):

    This paper offers novel mechanistic insights into how pre-exposure to warm temperature increases the resistance of C. elegans to peroxides, which are more toxic at warmer temperature. The temperature range tested in this study lies within the animal's living conditions and is much lower than that of heat shock. Therefore, this study expands our understanding of how past thermosensory experience shapes physiological fitness under chemical stress. The paper is technically sound with most experiments or analyses carried out rigorously, and therefore the conclusions are solid. However, it challenges our current understanding of the role of the C. elegans thermosensory system in coping with stress. The traditional view is that the AFD thermosensory neuron is activated upon sensing temperature rise, and that temperature sensation through AFD positively regulates systemic heat shock response and promotes longevity in C. elegans. Thus, it is quite unexpected that AFD ablation activates DAF-16 and improves peroxide resistance. It also appears counterintuitive that genes upregulated at 25 degrees overlap extensively with those upregulated by AFD ablation at 20 degrees. I feel that it is premature to coin the term "enhancer sensing" for such a phenomenon, as their work does not rule out the possibility that AFD ablation increases resistance to other stresses that are independent of temperature regarding their toxicity or magnitude of hazard. Additional work is necessary to clarify these issues.

    1. Whether the role of AFD in inhibiting peroxide resistance is related to AFD activity needs further clarification. AFD activity depends on the animal's thermosensory experience. As animals in this study are maintained at 20 degrees unless indicated specifically, the AFD displays activities starting around 17 degrees and peaks around 20 degrees. Under such condition, the AFD displays little or no activity to thermal stimuli around 15 degrees. It will be important to test whether cultivation of animals at 20 degrees improves peroxide resistance at 15 degrees, compared to 15 degrees-cultivation/15 degrees peroxide testing. The authors should also test whether AFD ablation further improves survival under peroxides at 15 degrees for animals grown at 20 degrees, whose AFD should show little or no activities at 15 degrees.

    2. The importance of the thermosensory function of AFD should be verified. In the current study, the tax-4 mutation was used to infer AFD activity, but tax-4 is expressed in sensory neurons other than AFD. In addition to AFD, AWC can sense temperature and it also expresses tax-4. Therefore, influence on AFD from other tax-4-expressing neurons cannot be excluded. On the other hand, ablation of AFD removes all AFD functions, including those that are constitutive and temperature-independent. Therefore, the authors should test the gcy-18 gcy-8 gcy-23 triple mutant, in which the AFD neurons are fully differentiated but completely insensitive to thermal stimuli. These three thermosensor genes are exclusively expressed in AFD. Compared to the tax-4 mutant that is broadly defective in multiple sensory modalities, this triple gcy mutant shows defects specifically in thermosensation. They should see whether results obtained from the AFD ablated animals could be reproduced by experiments using the gcy-18 gcy-8 gcy-23 triple mutant. The authors are also recommended to investigate ins-39 expression in AFD and profile gene expression patterns in the gcy-18 gcy-8 gcy-23 triple mutant.

    3. The literature suggests that AFD promotes longevity likely in part through daf-16 (Chen at al., 2016) or independent of daf-16 (Lee & Kenyon, 2009). Whatever it is, various studies show that activation of AFD and daf-16 promote a normal lifespan at higher temperature, and AFD ablation shortens lifespan at either 20 or 25 degrees. Therefore, the finding that DAF-16-upregulated genes overlap extensively with those upregulated by AFD ablation is quite unexpected (Figure 5B). The authors should perform further gene ontology (GO) analysis to identify subsets of genes co-regulated by DAF-16 and AFD ablation, whether these genes are reported to be involved in longevity regulation, immunity, stress response, etc.

    4. I feel that "enhancer sensing" is an overstatement, or at least a premature term that is not sufficiently supported without further investigations. The authors should explore whether AFD ablation or pre-exposure to warm temperature specifically enhances resistance to a stressor the toxicity of which is increased at higher temperature, but does not affect the resistance to other temperature-insensitive threats.

  5. eLife

    Author Response:

    Reviewer #1 (Public Review):

    Here, Servello et al explore the role of temperature and the temperature-sensing neuron AFD in promoting protection against peroxide damage. Unlike many other environmental threats, peroxide toxicity is expected to be temperature-dependent, since its chemical reactivity should be enhanced by higher temperatures. The authors convincingly and rigorously show that transient exposure to 25C, a condition of mild heat stress in C. elegans, activates animals' defenses against peroxides but potentially not other agents. Interestingly, this response requires the temperature-sensing AFD neurons, though whether temperature-dependent AFD activity is itself involved in this regulation is not explored. Further, the authors find that temperature regulates AFD's expression of the insulin ins-39 and provide evidence supporting the idea that repression of ins-39 at 25C contributes to enhanced peroxide defense. The authors use transcriptomic approaches to explore gene expression changes in animals in which AFD neurons are ablated, providing evidence that the FoxO-family transcription factor DAF-16 potentiates AFD signaling. However, because AFD ablation triggers effects broader than transient 25C exposure, the significance of these findings for temperature-dependent peroxide defense is somewhat unclear. Additionally, the possibility that DAF-16 (as well as another protective factor, SKN-1) function in parallel to temperature stress is consistent with many of the results shown but is not as thoroughly considered. Together, these studies identify a fascinating example of pre-emptive threat response triggered by the detection of a potentiator of that threat, a phenomenon they term "enhancer sensing." While some predictions of the specificity of this phenomenon remain untested, the paper provides intriguing insight into the potential mechanisms by which it may occur.

    Major issues:

    The dependence of the enhancer-sensing phenomenon on AFD leads the authors to conclude that the 25C stimulus is sensed by AFD itself, but this needs to be directly tested. To do this, they could ask whether tax-4 function is required in AFD, or use mutants in which AFD's thermosensory function is compromised.

    We thank the reviewer for suggesting these experiments. As requested, we determined whether previously identified mechanisms for temperature perception by the AFD neurons were required for the temperature-dependent regulation of peroxide resistance using gcy-18 gcy-8 gcy-23 triple mutants and the respective single mutants. The findings from the new experiments lead us to conclude that temperature perception by AFD via the GCY-8, GCY-18, and GCY-23 receptor guanylate cyclases, which are exclusively expressed in the AFD neurons, contributes to the temperature-dependent regulation of peroxide resistance in C. elegans. These experiments are detailed in the following new paragraph in the results section:

    “Last, we determined whether previously identified mechanisms for temperature perception by the AFD neurons were required for the temperature-dependent regulation of peroxide resistance. The AFD neurons sense temperature using receptor guanylate cyclases, which catalyze cGMP production, leading to the opening of TAX-4 channels (Goodman and Sengupta, 2019). Three receptor guanylate cyclases are expressed exclusively in AFD neurons: GCY-8, GCY-18, and GCY-23 (Inada et al., 2006; Yu et al., 1997) and are thought to act as temperature sensors (Takeishi et al., 2016). Triple mutants lacking gcy-8, gcy-18, and gcy-23 function are behaviorally atactic on thermal gradients and fail to display changes in intracellular calcium or thermoreceptor current in the AFD neurons in response to temperature changes (Inada et al., 2006; Ramot et al., 2008; Takeishi et al., 2016; Wang et al., 2013; Wasserman et al., 2011). We found that when grown and assayed at 20°C, gcy-23(oy150) gcy-8(oy44) gcy-18(nj38) triple null mutants survived 43% longer in the presence of tBuOOH than wild-type controls (Figure 3J). In contrast, at 25°C, the gcy-23 gcy-8 gcy-18 triple mutants showed a 12% decrease in peroxide resistance relative to wild-type controls (Figure 3K). Therefore, the three AFD-specific receptor guanylate cyclases influenced the temperature dependence of peroxide resistance, lowering peroxide resistance at 20°C and slightly increasing it at 25°C. At 20°C, the gcy-8(oy44), gcy-18(nj38), and gcy-23(oy150) single mutants increased peroxide resistance by 10%, 51%, and 21%, respectively, relative to wild-type controls (Figure 3L). Therefore, each of the three AFD-specific receptor guanylate cyclases regulates peroxide resistance. We conclude that temperature perception by AFD via GCY-8, GCY-18, and GCY-23 enables C. elegans to lower their peroxide resistance at the lower cultivation temperature.”

    The enhancer-sensing model is fascinating, but as it stands it is somewhat oversold. The authors could tone down the writing, indicating that this model is suggested rather than shown. Alternatively, they could more carefully test some of its predictions - for example by exploring the response to other threats (e.g. some of the toxicants described in Fig. S5) at 20C and 25C in WT and AFD-ablated animals.

    We edited the manuscript and expanded the manuscript’s discussion to address these concerns as well as similar concerns from reviewer #3. In the paper we show that the regulation of the induction of H2O2 defenses in C. elegans is coupled to the perception of temperature (an inherent enhancer of the reactivity of H2O2). To understand the significance of this finding in an evolutionary context, and to explain why such a regulatory system would evolve, we introduced in the discussion a new conceptual framework, “enhancer sensing,” and devoted a section of the discussion to demonstrating that the phenomenon that we observed could not be adequately explained by existing frameworks used to understand the evolutionary origins of the regulatory systems for defense responses.

    We now realize that we did not sufficiently and clearly explain the scope for the criterion for establishing a phenomenon represents enhancer sensing, leading to incorrect predictions by reviewer’s 1 and 3 about (a) whether what we observed in C. elegans is an instance of enhancer sensing (or more proof is needed) and (b) what the enhancer sensing model for the coupling of temperature perception to H2O2 defense would predict about how temperature and the AFD neurons would affect resilience to other chemicals. We regret failing to adequately explain the model’s scope and predictions and believe that we have now explicitly addressed the scope of what constitutes enhancer sensing and the predictions of the model. In particular, we previously did not spell out (a) the distinction between the enhancer sensing strategy and the mechanistic implementation of that strategy; and, importantly, (b) we did not discuss what the enhancer sensing strategy coupling temperature perception to H2O2 defense in C. elegans predicted (and did not predict) about whether a similar strategy would be expected to be used by C. elegans to deal with other temperature-dependent threats. We now address these issues in two new paragraphs in the discussion that read:

    “We show here that C. elegans uses an enhancer sensing strategy that couples H2O2 defense to the perception of high temperature. We expect this strategy’s output (the level of H2O2 defense) to provide the nematodes with an evolutionarily optimal strategy across ecologically relevant inputs (cultivation temperatures) (Kussell and Leibler, 2005; Maynard Smith, 1982; Wolf et al., 2005). This strategy is implemented at the organismic level through the division of labor between the AFD neurons, which sense and broadcast temperature information, and the intestine, which responds to that information by providing H2O2 defense (Figure 9D). Ascertaining that C. elegans relies on this enhancer sensing strategy does not depend on the temperature information broadcast by AFD exclusively regulating defense responses to temperature-dependent threats, because the regulation of defenses towards temperature-insensitive threats could affect defenses towards temperature-dependent threats; for example, suppressing defenses towards a temperature-insensitive threat would be beneficial if those defenses interfered with H2O2 defense or depleted energy resources contributing to H2O2 defense.

    As with any sensing strategy, enhancer sensing strategies are more likely to evolve when sensing is informative and responding is beneficial. In their natural habitat, C. elegans encounter many environmental chemicals that, like H2O2, are inherently more reactive at higher temperatures. It will be interesting to determine the extent to which C. elegans uses enhancer sensing strategies coupling temperature perception to the induction of defenses towards those chemicals, and whether those strategies rely on temperature perception and broadcasting by the AFD neurons. We expect that sensing strategies regulating defense towards those chemicals would be more likely to evolve when those chemicals are common, reactive, and cause consequential damage.”

    We note that our ability to predict survival to other toxicants, such as those that trigger specific gene-expression responses that are AFD-dependent but are unaffected between 20C and 25C (as proposed by the reviewer), is limited not only by our lack of knowledge about the specific mechanisms that protect worms from those toxicants, but also by our lack of knowledge about whether defense towards hydrogen peroxide interferes (or synergizes) with defense towards each of those toxicants and whether defense towards those toxicants interferes (or synergizes) with H2O2 defense. We therefore think that those experiments would be better addressed in future studies.

    The role of ins-39 remains somewhat speculative. Fig 4F shows that ins-39 mutants have a reduced induction of peroxide defense, but it seems that this could be the result of a ceiling effect. The authors' model predicts that overexpression of ins-39, particularly at 25C, should sensitize animals to peroxide damage, a prediction that should be tested directly. Further, the authors seem to assume that AFD is the relevant site of ins-39 function, but this needs to be better supported.

    As requested by all three reviewers, we determined whether ins-39 gene expression in AFD was sufficient to lower peroxide resistance by restoring ins-39(+) gene expression only in the AFD neurons using the AFD-specific gcy-8 promoter. As predicted by the reviewer, these worms were more sensitive to peroxide than wild-type worms. The findings from this experiment lead us to conclude that expression of ins-39 in the AFD neurons was sufficient to regulate the nematode’s peroxide resistance. The new section reads:

    “Next, we determined whether the INS-39 signal from AFD regulated the nematode’s peroxide resistance. The tm6467 null mutation in ins-39 deletes 520 bases, removing almost all the ins-39 coding sequence (Figure 5A), and inserts in that location 142-bases identical to an intervening sequence located between ins-39 and its adjacent gene. In nematodes grown and assayed at 20°C, ins-39(tm6467) increased peroxide resistance by 26% relative to wild-type controls (Figure 5F). To determine whether ins-39 gene expression in AFD was sufficient to lower peroxide resistance, we restored ins-39(+) expression only in the AFD neurons using the AFD-specific gcy-8 promoter (Inada et al., 2006; Yu et al., 1997) in ins-39(tm6467) mutants. Expression of ins-39(+) only in AFD eliminated the increase in peroxide resistance of ins-39(tm6467) mutants (Figure 5F). Notably, the peroxide resistance of the two independent transgenic lines was 28% and 30% lower than that of wild-type controls, likely due to overexpression of the gene beyond wild-type levels. We conclude that the gene dose-dependent expression of ins-39 in the AFD neurons regulated the nematode’s peroxide resistance.”

    The temperature-shift experiments in figure 5G (formerly 4F) indicated that the effect on peroxide resistance at 20C of growth at 25C and of the ins-39 mutation were non additive. We interpreted this epistatic interaction to be due to action in a common pathway. It is possible that while growth at 25C increases the subsequent peroxide resistance at 20C, it could limit the nematodes’ subsequent peroxide resistance at 20C (beyond those peroxide-resistance increasing effects) when in combination with another intervention, even if those interventions acted via parallel mechanisms—a ceiling effect, as proposed by the reviewer. We favor the alternative interpretation, that the mechanisms act sequentially, because of our findings that ins-39 gene expression within AFD was lower at 25C than at 20C, leading us to propose the sequential model in figure 5H (formerly 4G).

    Most of the daf-16 and skn-1 experiments are carried out in AFD-ablated animals, making the relevance of these findings for the 25C-dependent induction of peroxide defense somewhat unclear. As the authors show, AFD ablation causes much more extensive changes than transient 25C exposure, clearly seen in slope of the line in 3C. Further, unlike 25C exposure, AFD ablation is a chronic and non-physiological state. It would be useful for the authors to be cautious in their interpretation of these findings and to be clearer about how strongly they can connect them to the "enhancer sensing" phenomenon. Along these lines, the potentiation idea could be toned down a bit. Much of the data is consistent with parallel function for daf-16 (and skn-1) - for example, Fig 5C indicates additive effects of daf-16 and 25C exposure; 6C shows that AFD ablation still has a clear effect on peroxide sensitivity in the absence of both daf-16 and skn-1; and Fig S8a shows that much of the transcriptional response to AFD ablation (along PC1) is intact in daf-16 animals.

    We have made several adjustments in the text to address these concerns. As the reviewer noted, the experiments with skn-1 were performed only in AFD ablated worms. We have renamed the section heading to “SKN-1/NRF and DAF-16/FOXO collaborate to increase the nematodes’ peroxide resistance in response to AFD ablation” to make that clear.

    In contrast, the peroxide resistance experiments with daf-16 were done also in worms grown at 25C and then shifted to 20C during the peroxide resistance assay. The connection of daf-16 with the temperature dependent regulation of peroxide resistance was established in temperature shifts experiments in daf-16 single mutants (Figure 6C, formerly 5C) and in transgenic worms rescuing the daf-16 mutant only in the intestine (Figure 6F). In the revised text we make it clearer that the effect of the daf-16 mutation is bigger when the nematodes are shifted from 25C to 20C: “The daf-16(mu86) null mutation decreased peroxide resistance in nematodes grown at 25°C and assayed at 20°C by 35%, a greater extent than the 21% reduction in peroxide resistance induced by that mutation in nematodes grown and assayed at 20°C (Figure 6C).”

    As the reviewer noted, daf-16 and skn-1 have a role in peroxide resistance when the AFD neurons are not ablated (albeit a smaller one than when those neurons are ablated). We have made several changes and additions to the text to make that explicit. Most notably, the revised last paragraph of the SKN-1 section now reads: “We propose that when nematodes are cultured at 20°C, the AFD neurons promote signaling by the DAF-2/insulin/IGF1 receptor in target tissues, which subsequently lowers the nematode’s peroxide resistance by repressing transcriptional activation by SKN-1/NRF and DAF-16/FOXO. However, this repression is not complete, because both daf-16(mu86) and skn-1(RNAi) lowered peroxide resistance at 20°C when the AFD neurons were present. It is also likely that DAF-16 and SKN-1 are not the only factors that contribute to peroxide resistance in AFD-ablated nematodes at 20°C, because AFD ablation increased peroxide resistance in daf-16(mu86); skn-1(RNAi) nematodes, albeit to a lesser extent than in daf-16(+) or skn-1(+) backgrounds.”

    The potentiation idea was specific to the effects of DAF-16 on gene expression. As the reviewer noted, much of the transcriptional response to AFD ablation is intact (albeit reduced in magnitude) in AFD-ablated daf-16 mutants, leading to a shift in the PC1 score for the mutant. At the level of the expression of individual genes, we quantified those effects in Figure 8G (formerly 7D). When we did the RNAseq experiments we had expected that lack of daf-16 would eliminate either all the changes in gene expression induced by AFD ablation or eliminate those changes for a subset of genes. Instead, what we found was much more subtle, and unexpected: the size of the gene expression change induced by AFD ablation was reduced by the daf-16 mutation, and that reduction was systematic. Specifically, we found that the bigger the change in gene expression induced by AFD ablation, the bigger the effect of daf-16 in the AFD ablated animals (that is, potentiation), leading to a change in the slope in the regression line in Figure 8G. We revised the paper to ensure we only used the word potentiation in this context (gene expression), even though formally DAF-16 also potentiated the effects of AFD ablation (and temperature shift from 25C to 20C) on peroxide resistance.

    Reviewer #3 (Public Review):

    This paper offers novel mechanistic insights into how pre-exposure to warm temperature increases the resistance of C. elegans to peroxides, which are more toxic at warmer temperature. The temperature range tested in this study lies within the animal's living conditions and is much lower than that of heat shock. Therefore, this study expands our understanding of how past thermosensory experience shapes physiological fitness under chemical stress. The paper is technically sound with most experiments or analyses carried out rigorously, and therefore the conclusions are solid. However, it challenges our current understanding of the role of the C. elegans thermosensory system in coping with stress. The traditional view is that the AFD thermosensory neuron is activated upon sensing temperature rise, and that temperature sensation through AFD positively regulates systemic heat shock response and promotes longevity in C. elegans. Thus, it is quite unexpected that AFD ablation activates DAF-16 and improves peroxide resistance. It also appears counterintuitive that genes upregulated at 25 degrees overlap extensively with those upregulated by AFD ablation at 20 degrees. I feel that it is premature to coin the term "enhancer sensing" for such a phenomenon, as their work does not rule out the possibility that AFD ablation increases resistance to other stresses that are independent of temperature regarding their toxicity or magnitude of hazard. Additional work is necessary to clarify these issues.

    1. Whether the role of AFD in inhibiting peroxide resistance is related to AFD activity needs further clarification. AFD activity depends on the animal's thermosensory experience. As animals in this study are maintained at 20 degrees unless indicated specifically, the AFD displays activities starting around 17 degrees and peaks around 20 degrees. Under such condition, the AFD displays little or no activity to thermal stimuli around 15 degrees. It will be important to test whether cultivation of animals at 20 degrees improves peroxide resistance at 15 degrees, compared to 15 degrees-cultivation/15 degrees peroxide testing. The authors should also test whether AFD ablation further improves survival under peroxides at 15 degrees for animals grown at 20 degrees, whose AFD should show little or no activities at 15 degrees.

    The reviewer raises an interesting point about the relation between the mechanisms that determine AFD activity in response to temperature and those that enable AFD to regulate peroxide resistance. In the revised manuscript we tested whether known mechanisms enabling AFD to sense changes in temperature acutely (receptor guanylate cyclases GCY-8, GCY-18, and GCY-23) played a role in the temperature dependence of peroxide resistance. We found that they did, as detailed in our response to reviewer #1’s point 1.

    As noted by reviewer #2 in their point 1, and in our reply to that comment (and in a new discussion paragraph in the revised manuscript), the relationship between the known mechanisms the acutely regulate the activity of AFD in response to temperature and the mechanisms by which constant cultivation temperature regulates gene expression in AFD (and therefore the expression of peroxide resistance regulating signals like INS-39) is not well understood. Therefore, it is difficult to predict which temperatures will cause induction of peroxide defenses via AFD-dependent mechanisms, or via other mechanisms. While we agree with the reviewer that it will be interesting to characterize the extent to which other cultivation temperatures besides 25C lead to increased peroxide resistance at lower temperatures (including the proposed shifts from 20C to 15C), we think that those questions will be better addressed in future studies.

    2. The importance of the thermosensory function of AFD should be verified. In the current study, the tax-4 mutation was used to infer AFD activity, but tax-4 is expressed in sensory neurons other than AFD. In addition to AFD, AWC can sense temperature and it also expresses tax-4. Therefore, influence on AFD from other tax-4-expressing neurons cannot be excluded. On the other hand, ablation of AFD removes all AFD functions, including those that are constitutive and temperature-independent. Therefore, the authors should test the gcy-18 gcy-8 gcy-23 triple mutant, in which the AFD neurons are fully differentiated but completely insensitive to thermal stimuli. These three thermosensor genes are exclusively expressed in AFD. Compared to the tax-4 mutant that is broadly defective in multiple sensory modalities, this triple gcy mutant shows defects specifically in thermosensation. They should see whether results obtained from the AFD ablated animals could be reproduced by experiments using the gcy-18 gcy-8 gcy-23 triple mutant. The authors are also recommended to investigate ins-39 expression in AFD and profile gene expression patterns in the gcy-18 gcy-8 gcy-23 triple mutant.

    We thank the reviewer for this suggestion. We have performed the requested experiments, as detailed in our response to reviewer #1’s point 1. Briefly, we determined found that gcy-18 gcy-8 gcy-23 triple mutants increased peroxide resistance at 20C but not at 25C, and found that the respective gcy single mutants affected peroxide resistance at 20C. In light of these findings, we concluded that temperature perception by AFD via GCY-8, GCY-18, and GCY-23 enables C. elegans to lower their peroxide defenses at the lower cultivation temperature.

    3. The literature suggests that AFD promotes longevity likely in part through daf-16 (Chen at al., 2016) or independent of daf-16 (Lee & Kenyon, 2009). Whatever it is, various studies show that activation of AFD and daf-16 promote a normal lifespan at higher temperature, and AFD ablation shortens lifespan at either 20 or 25 degrees. Therefore, the finding that DAF-16-upregulated genes overlap extensively with those upregulated by AFD ablation is quite unexpected (Figure 5B). The authors should perform further gene ontology (GO) analysis to identify subsets of genes co-regulated by DAF-16 and AFD ablation, whether these genes are reported to be involved in longevity regulation, immunity, stress response, etc.

    We thank the reviewer for this interesting comment about the complex mechanisms by which AFD regulates longevity. We note that AFD also has additional temperature-dependent roles in lifespan regulation, as Murphy et al. 2003 found that RNAi of gcy-18 increased lifespan in wild-type worms at 20C but not at 25C. Therefore, AFD-specific interventions can also be lifespan extending at 20C.

    We performed WormCat analysis, which is similar to gene ontology, in Figure 8-figure supplement 2 (formerly Figure S8G), which we described in the results section: “we found that the extent to which AFD ablation affected the average expression of sets of genes with related functions (Higgins et al., 2022; Holdorf et al., 2020) was systematically lower in daf-16(mu86) mutants than in daf-16(+) nematodes (_R_2 = 86%, slope = 0.67, P < 0.0001, Figure 8—figure supplement 2).” Visual inspection of the plot and the very high coefficient of determination of 86% indicate that the size of the effect of AFD ablation on gene expression was systematically smaller when the contribution of DAF-16 to gene expression was removed.

    In the revised manuscript we also moved the three panels quantifying the expression of DAF-16 targets and daf-16-regulated genes from the supplement to the main figure. One of those panels (Figure 8F) shows that genes upregulated by daf-16(+) in daf-2 mutants were disproportionally affected by lack of daf-16 in AFD-ablated worms, as we described in the results section: “In addition, in AFD ablated nematodes, lack of daf-16 lowered the expression of genes upregulated in a daf-16-dependent manner in daf-2(-) mutants (Murphy et al., 2003) to a greater degree than in unablated nematodes (Figure 8F).”

    4. I feel that "enhancer sensing" is an overstatement, or at least a premature term that is not sufficiently supported without further investigations. The authors should explore whether AFD ablation or pre-exposure to warm temperature specifically enhances resistance to a stressor the toxicity of which is increased at higher temperature, but does not affect the resistance to other temperature-insensitive threats.

    We edited the manuscript and expanded the manuscript’s discussion to address these concerns as well as similar concerns from reviewer #1. For clarity, we repeat much of our response to reviewer #1’s point 2 here, with the last paragraph of this response specific to this reviewer’s comment.

    In the paper we show that in C. elegans the regulation of the induction of H2O2 defenses is coupled to the perception of temperature (an inherent enhancer of the reactivity of H2O2). To understand the significance of this finding in an evolutionary context, and to explain why such a regulatory system would evolve, we introduced in the discussion a new conceptual framework, “enhancer sensing,” and devoted a section of the discussion to demonstrating that the phenomenon that we observed could not be adequately explained by existing frameworks used to understand the evolutionary origins of the regulatory systems for defense responses.

    We now realize that we did not sufficiently and clearly explain the scope for the criterion for establishing a phenomenon represents enhancer sensing, leading to incorrect predictions by reviewer’s 1 and 3 about (a) whether what we observed in C. elegans is an instance of enhancer sensing (or more proof is needed) and (b) what the enhancer sensing model for the coupling of temperature perception to H2O2 defense would predict about how temperature and the AFD neurons would affect resilience to other chemicals. We regret failing to adequately explain the model’s scope and predictions and believe that we have now explicitly addressed the scope of what constitutes enhancer sensing and the predictions of the model. In particular, we previously did not spell out (a) the distinction between the enhancer sensing strategy and the mechanistic implementation of that strategy; and, importantly, (b) we did not discuss what the enhancer sensing strategy coupling temperature perception to H2O2 defense in C. elegans predicted (and did not predict) about whether a similar strategy would be expected to be used by C. elegans to deal with other temperature-dependent threats. We now address these issues in two new paragraphs in the discussion that read:

    “We show here that C. elegans uses an enhancer sensing strategy that couples H2O2 defense to the perception of high temperature. We expect this strategy’s output (the level of H2O2 defense) to provide the nematodes with an evolutionarily optimal strategy across ecologically relevant inputs (cultivation temperatures) (Kussell and Leibler, 2005; Maynard Smith, 1982; Wolf et al., 2005). This strategy is implemented at the organismic level through the division of labor between the AFD neurons, which sense and broadcast temperature information, and the intestine, which responds to that information by providing H2O2 defense (Figure 9D). Ascertaining that C. elegans relies on this enhancer sensing strategy does not depend on the temperature information broadcast by AFD exclusively regulating defense responses to temperature-dependent threats, because the regulation of defense towards temperature-insensitive threats could affect defenses towards temperature-dependent threats; for example, suppressing defenses towards a temperature-insensitive threat would be beneficial if those defenses interfered with H2O2 defense or depleted energy resources contributing to H2O2 defense.

    As with any sensing strategy, enhancer sensing strategies are more likely to evolve when sensing is informative and responding is beneficial. In their natural habitat, C. elegans encounter many environmental chemicals that, like H2O2, are inherently more reactive at higher temperatures. It will be interesting to determine the extent to which C. elegans uses enhancer sensing strategies coupling temperature perception to the induction of defenses towards those chemicals, and whether those strategies rely on temperature perception and broadcasting by the AFD neurons. We expect that sensing strategies regulating defense towards those chemicals would be more likely to evolve when those chemicals are common, reactive, and cause consequential damage.”

    We note, in the first of the new discussion paragraphs, that the existence of an enhancer sensing strategy is not contingent on whether the AFD neurons (that implement the temperature sensing and temperature-information broadcasting functions regulating peroxide defenses) also do not regulate defense responses to temperature-insensitive threats. For example, it may be beneficial to an animal facing high concentrations of environmental peroxides to suppress defense against a temperature-insensitive threat when those defenses are detrimental towards defense towards hydrogen peroxide. This could occur, for example, because there is an energetic trade off when mounting multiple defense responses, or because specific defenses towards temperature-insensitive threats interfere with peroxide defense. As we noted in our response to reviewer #1’s point 2, our ability to predict survival to threats other than H2O2 (including temperature-independent threats) is limited not only by our lack of knowledge about the specific mechanisms that protect worms from those threats, but also by our inability to predict the extent to which defenses towards different threats operate independently, constructively, or destructively with those that provide hydrogen peroxide defense. We therefore think that those experiments would be better addressed in future studies.

  6. Version published to 10.7554/elife.78941 on eLife
  7. Version published to 10.1101/2022.03.21.485202v1 on bioRxiv