Lysosome-related organelles promote stress and immune responses in C. elegans

Lysosome-related organelles (LROs) play diverse roles and their dysfunction causes immunodeficiency. However, their primordial functions remain unclear. Here, we report that C. elegans LROs (gut granules) promote organismal defenses against various stresses. We find that toxic benzaldehyde exposure induces LRO autofluorescence, stimulates the expression of LRO-specific genes and enhances LRO transport capacity as well as increases tolerance to benzaldehyde, heat and oxidative stresses, while these responses are impaired in glo-1/Rab32 and pgp-2 ABC transporter LRO biogenesis mutants. Benzaldehyde upregulates glo-1- and pgp-2-dependent expression of heat shock, detoxification and antimicrobial effector genes, which requires daf-16/FOXO and/or pmk-1/p38MAPK. Finally, benzaldehyde preconditioning increases resistance against Pseudomonas aeruginosa PA14 in a glo-1- and pgp-2-dependent manner, and PA14 infection leads to the deposition of fluorescent metabolites in LROs and induction of LRO genes. Our study suggests that LROs may play a role in systemic responses to stresses and in pathogen resistance.

1. Fundamentally, the only data that the authors show to support their conclusion that LROs are involved in stress resistance programs is the following: glo mutants suppress autofluoresce and also blunt the BA-induced protection to various stress.However, these two observations could be "true, true, and unrelated."It seems equally plausible that BA, a toxic xenobiotic, activates stress response pathways (e.g., p38, SKN-1, ATFS-1), and gut granule autofluoresence changes to BA treatment as a secondary consequence, rather an a direct effect of LRO biology.The fact that BA-mediated protection is blunted in glo mutants could be secondary to non-specific pleiotropies in these mutant backgrounds, rather than specific changes in the LRO compartment.
2. A direct connection between the LRO and protective stress responses was not established in this paper.Thus, line 24 of the abstract, the core conclusion of the manuscroipt, "LROs called gut granules promote organismal defenses" is a gross overstatement of the author's findings.
3. Transcription analyses was performed by qRT-PCR on cherry-picked genes, making their biological significance impossible to interpret.Maybe BA treatment increases the transcription of half the genome?

Reviewer #1
In this paper the authors further expand their idea that the lysosome-related organelles (LROs) promote stress and immune defenses in C. elegans and they presented a convincing model for the LRO stress response.The article provides new insight into the mechanism of how the toxic insults (benzaldehyde, methyl-salycilate and P. aeruginosa PA14 infection) are able to induce a deposition of (auto)fluorescent metabolites and an increased burden of C. elegans intestinal LROs.Next it was also evidenced, that this in turn stimulates the early expression of LRO-specific membrane transporter and vesicular traffic genes (via yet unidentified factors), and thereby ultimately resulting in LRO adaptation.As the authors outlined, their novel findings illuminate a role for LROs in mounting systemic stress, detoxification and antimicrobial defenses.This process requires a co-operation between ancient and highly conserved processes such as compartmentalization in the endolysosomal system, an ABC membrane transporter and transcriptional stress and innate immune pathways.In my opinion, this is an excellent paper.This is of sufficient general interest to warrant publication in this journal.The paper is written clearly and the evidences justifies the conclusions.All experimental protocols are given in sufficient detail.We sincerely thank Reviewer #1 for taking the time to attentively review our manuscript, and for the positive comments and recommendation for publication.

Reviewer #2 Brief Summary
The work by Hajdu et al. addresses the function of gut granules in C. elegans.These lysosomerelated organelles have been implicated in a number of metabolic processes, including metal and toxin storage.This interesting work shows that gut granules function in protecting C. elegans from chemical toxins, heat and oxidative stress, and bacterial pathogens.Moreover, the work identifies transcriptional responses that require gut granules.Together, this work makes a compelling case for a role of gut granules in varied and important physiological processes.

Impressions of the work
Overall, the data is of high quality, is well controlled, and provides convincing support for the major conclusions of the work.The graphical representation and statistical analysis of the work is a strength.The authors show a robust understanding of the literature on C. elegans LROs, stress, and pathogens.The authors present a significant amount of data, all of which are central to their questions and conclusions.The work is highly novel, being the first to analyze the regulation of genes involved in LRO formation and function.The work will be of broad interest to the C. elegans community as well as those that study the GLO-1 related Rab32/38 and ABC transporters in other organisms, making it a good fit for publication in Communications Biology.Specific comments (numbered) 1.The authors might consider supporting their discussion of the evolutionary origins of LROs by mentioning acidocalcisomes in unicellular organisms, which share functions and biogenesis factors with LROs.We sincerely thank Reviewer #2 for the time to evaluate our manuscript, for finding it novel, relevant and sound.We are grateful for the useful comments to improve our study.We supplemented the Introduction with the inclusion of acidocalcisomes as evolutionary predecessors of LROs (line 56 of the revised MS).
2. On line 43 the authors use malanine when they likely are referring to melanin.We thankfully corrected the typo (line 48 of the revised MS).
3. On lines 56-57 the wording regarding the function of Rab32/38 and BLOC-3 is unclear.We are grateful to Reviewer #2 for drawing our attention on the confounding wording.We rephrased the sentence (line 61 of the revised MS).

4.
In the experiments presented in Figure 1, were identical fluorescence signal acquisition settings used in control vs BA or N2 vs glo-1(-) conditions?If yes, this should be described in the methods or figure legend.In addition, it would be helpful to define CTCF units and describe how they were calculated in the methods.We appreciate the comment.Yes, we employed identical acquisition settings for all conditions.Both this and the definition and calculation of CTCF were described in the revised Methods (lines 856-861 of the revised MS).1B, there is some concern that the addition of Nile Red could increase the green fluorescence, as it can fluoresce in the green.A Control/non-BA condition is not provided to assess this issue.The colocalization of the BA induced fluorescence might be better assed by localizing gut granules with its red autofluorescence.We are indebted to Reviewer #2 for pointing out the lack of a control non-BA condition in Fig. 1B.We included this in the revised Fig. 1B, where we did not detect green fluorescence signal from Nile Red (please compare Figure 1A Control and 1B Nile Red upper left panel).We note that we 6.The data in 1B nicely show colocalization that can only be seen when I really zoom in, for readers it would be helpful to greatly zoom in on one section of the intestine so that individual organelles can be more easily resolved.We thank Reviewer #2 for the important comment.We revised Figure 1B and included higher magnification micrographs and insets to visualize individual gut granules.

Please include scale bars in the insets in Fig 1C.
We included the scale bars in the insets in Figure 1C.8.In Fig 2E it looks like there was not a significant reduction in lawn avoidance of BApreconditioned glo-1(-) compared to naïve glo-1(-), yet the authors state on line 183 that there was a partial reduction.This conclusion does not appear to be supported by the data and needs to be changed or addressed more clearly.We thank for the comment and revised the conclusion of Figure 2E accordingly (line 197 of the revised MS). 9.The authors need to consider and possibly experimentally test the possibility that the behavioral effects that are seen in glo-1(-) mutants could result from LRO independent processes.For example, necrotic cell death of neurons is altered by glo-1(-) https://doi.org/10.1083/jcb.200511103and glo-1 expression has not been analyzed for neuronal expression, leaving open the possibility for a role for this Rab in neurons.We thank for this note, cited the suggested reference and included the possibility that the behavioural change might be attributed to a neuronal function of glo-1 (lines 200-202 of the revised MS).10.The suggestion that pre-exposure to BA causes increased Nile Red staining due to increased transport /capacity is only weakly supported by the data (lines 200-203).It could easily be that gut granules accumulate more hydrophobic materials stained by Nile Red in response to BA (due to defects in transport).Assays of fluid phase endocytosis (Rhod-BSA or Rhod-dextran) and organelle diameter would more directly assess changes in transport/capacity.We sincerely thank Reviewer #2 for pointing out weakly supported conclusions regarding a more efficient transport or capacity of LROs.Based on the proposal we monitored the kinetics of Nile Red release from the LROs and found that BA-treated animals exhibited a rapid release of Nile Red from their LROs which became identical to control worms within 30-60 min (revised Fig. 3B).This excludes an unspecific accumulation and transport defect and suggest an increased transport capacity, which is consistent with the transcriptional upregulation of LRO membrane transporters pgp-2 and K09C4.5 (revised Figure 3D).We revised the corresponding paragraph accordingly (lines 222-226 of the revised MS).
Although the determination of organelle diameter would provide an important characteristic, we did not observe a robust increase in their size by BA treatment, only a significantly increased red and green fluorescence.The difficulty with these data is that they do not provide direct estimates on organelle size or number, only the quantity of their content.Therefore, we removed the misleading term increased 'capacity' which might refer to morphological enlargement and only use increased 'transport capacity' or 'functional capacity' of LROs from the revised manuscript (see for instance the title of the chapter at lines 211-212 of the revised MS).
11.If the investigators have done the analysis, it would be interesting to report the effects of stress, pre-treatment with BA, and loss of gut granules in glo-3(-) on glo-1 mRNA levels, given its key role in LRO biogenesis.We thank the important suggestion.We tested the effect of BA exposure and different stress regulator mutations on glo-1 mRNA levels and found no significant change.These results were included in revised Supplementary Fig. S3E and lines 234-237 of the revised MS.Intriguingly, PA14 infection elevated glo-1 mRNA level, which was absent in pgp-2 mutants, suggesting that pathogen stress causes a demand which requires the upregulation of glo-1 to support LRO biogenesis and a regulatory role of pgp-2 (see revised Fig. 8H and lines 533 and 536-537 of the revised MS).
12. The images in Fig 7I do not enable the reader to assess the patterns and localization of fluorescent materials.Larger high resolution images are necessary for the authors to support their claims.We thank Reviewer #2 for the important comment.We included higher magnification micrographs to visualize the fluorescent patterns in revised Figure 8I.
We thank Reviewer #2 again for the constructive comments, and we hope that our reply and the revised manuscript will be recommended for publication in Communications Biology.

Reviewer #3
Hajdu and colleagues present a manuscript where they attempt to assign a function to lysosomalrelated organelles in immune and stress response in C. elegans.The primary tool employed by these investigators in this study is benzaldehyde, a toxic xenobiotic that increases gut granule autofluorescence.Glo-1 and Glo-3 abrogated the induction of gut granules by BA, as expected.They authors show that pre-exposure to BA provides protection from a subsequent challenge with BA, and show that BA induces the expression of a wide array of stress and immune response genes, albiet subtly (2-8 fold expression for most).In addition, BA can provide some protection form heat stress and paraquat challenge.However, the core conclusion of this paper is that unifying mechanism by which BA promotes stress resistance is through the expansion of lysomal-related organelles.However, the data supporting this conclusion is only correlative, and multiple alternate explanations exist that could explain their findings, which were not evaluated by the authors.My opinion, which I expand further below, is that the authors have grossly over-interpreted their findings.
1. Fundamentally, the only data that the authors show to support their conclusion that LROs are involved in stress resistance programs is the following: glo mutants suppress autofluoresce and also blunt the BA-induced protection to various stress.However, these two observations could be "true, true, and unrelated."It seems equally plausible that BA, a toxic xenobiotic, activates stress response pathways (e.g., p38, SKN-1, ATFS-1), and gut granule autofluoresence changes to BA treatment as a secondary consequence, rather an a direct effect of LRO biology.The fact that BA-mediated protection is blunted in glo mutants could be secondary to non-specific pleiotropies in these mutant backgrounds, rather than specific changes in the LRO compartment.
We sincerely thank Reviewer #3 for the time to evaluate our manuscript.We are very grateful for pointing out the missing pieces of evidence and the discrepancy between our results and the proposed model.
Based on the Reviewer's suggestion, we investigated the BA-induced autofluorescence in stress pathway regulator mutants shown or suspected to be involved in the BA-induced stress response.
We observed that daf-16, hsf-1, pmk-1 and skn-1, but not atfs-1 and kgb-1, were required for autofluorescence (revised Fig. 3E and Supplementary Fig. 3D; lines 275-285 and 617-620 of the revised MS).The involvement of other regulator(s) cannot be excluded.These results suggest that the deposition of autofluorescent BA-related material in LROs is a consequence of the activation of several stress pathways, which represent a regulated defensive response against BA toxicity.We also tested the requirement of these regulators in the BA-induced expression of LRO-associated genes and found that daf-16 (for pgp-2), hsf-1 (for pgp-2) and pmk-1 (for glo-3) were, but skn-1 was not, required for the expression of LRO mRNAs.Moreover, the expression of K09C4.5 was affected by none of the above factors, suggesting the involvement of other unknown regulator(s) (revised Fig. 3F-H and Supplementary Fig. 3E,F; lines 286-297 and 640-642 of the revised MS).These findings suggest that the transcriptional upregulation of LRO biogenetic and transport genes upon BA exposure relies on specific, partially identified regulators and appears as an organellar adaptive response to xenobiotic BA stress.We observed that BA-preconditioning induced protection towards various stresses was compromised in LRO mutants.This raised the question whether LRO mutants might exhibit nonspecific pleiotropic effects, i.e. sickness.However, we found that LRO mutants exhibit identical basal physiological tolerance to BA and paraquat and only slightly reduced thermotolerance and pathogen resistance, compared to wild-type (Figs.4, 6 and 8).Further, although both pmk-1 and atfs-1 mutants have lower basal paraquat tolerance, only pmk-1 failed to respond to BA preconditioning (revised Supplementary Fig. 7E and lines 492-493 of the revised MS).These findings suggest that young adult LRO mutants appear to have quasi-normal physiology, but their response to a preconditioning mild stress is impaired.Moreover, LRO mutants have slightly shorter lifespan and weakened protection against proteotoxicity (Kumar et al Austin Aging Res 2017).We note that stress response mutants (hsf-1, daf-16, skn-1, pmk-1) exhibit similar phenotypes, such as shorter lifespan and sensitivity to stresses, including infection.Finally, instead of a compensatory upregulation of other stress responses as a result of sickness, LRO mutations prevented the BAand PA14-induced activation of specific, but not all transcriptional stress responses (Figs. 5 through  8).Although a mild sickness, i.e. a sensitivity to stress is evident, all the regulated deposition of toxic metabolites, the induction of LRO genes and the expression of other LRO-dependent specific stress responses suggest that LROs regulate organismal stress defenses.We agree with the Reviewer that the deposition of autofluorescent material in gut granules represent a metabolic process, while the LRO-dependent transcriptional activation of stress and antimicrobial responses might indicate a signaling event, hence, they might be independent processes.Also, the extent of contribution of these processes in the organismal defense and their connection to each other is unclear (see lines 671-673 of the revised MS).
2. A direct connection between the LRO and protective stress responses was not established in this paper.Thus, line 24 of the abstract, the core conclusion of the manuscroipt, "LROs called gut granules promote organismal defenses" is a gross overstatement of the author's findings.We are thankful for the important note.We studied the role of the respective regulators of the BAinduced, LRO-dependent stress and antimicrobial gene expression and found a requirement for daf-16 (for hsp-16.2and cyp-35B) and pmk-1 (for hsp-70, irg-5 and clec-60) and nhr-86 (for irg-5), but not hsf-1 and skn-1.Moreover, the extended nuclear translocation of DAF-16::GFP with a simultaneous inability to activate its target genes in LRO mutants provides evidence that LROderived signals are indispensable for the BA-induced expression of stress and immune response genes (Fig. 7 and Supplementary Fig. 7; lines 455-492 and 687-708 of the revised MS).These findings confirm the original conclusion that LROs promote specific organismal defenses and extend it with the demonstration of the extensive cooperation of LROs with cellular stress and immune pathways.We updated the manuscript and the model (revised Fig. 9 and its legend) to reflect a more precise interpretation of the findings.
3. Transcription analyses was performed by qRT-PCR on cherry-picked genes, making their biological significance impossible to interpret.Maybe BA treatment increases the transcription of half the genome?
We highly appreciate the relevant question.It may well be that BA and other toxic agents, such as those of PA14 induce a more comprehensive stress response, which, together with the underlying LRO-dependent mechanisms are subject of a future system-wide research project.In the current study we analyzed the expression of carefully selected marker genes of specific stress responses validated in a plethora of previous studies.Further, besides the induction of specific stress pathways, the biological significance of our findings lies in the induction of adaptive LRO responses (exemplified by pgp-2 and glo-3 expression and increased Nile Red transport) as well as the elicitation of organismal defenses to various stresses, including infection (see lines 659-669 of the revised MS).
We are indebted to Reviewer #3 again for the constructive critics and for the thoughtful comments.We hope our reply and the revised manuscript adequately addresses the concerns raised and the study will be recommended for publication in Communications Biology.I note that the version of the manuscript supplied to me did not have line numbers that matched the authors description in the rebuttal letter making it difficult to assess changes to the manuscript.
Most of my original concerns have been addressed in this improved manuscript.However, I do still have some minor comments and concerns that relate to major conclusions made by the authors.
1. Prior comment 3. On lines 56-57 the wording regarding the function of Rab32/38 and BLOC-3 is unclear.Response: We rephrased the sentence (line 61 of the revised MS).The description of the relationship between Rab32/38 and BLOC-3 is not correct.BLOC-3 acts as the GEF to activate Rab32/38 and the description would be clearer if there were a description of Rab32/38 function in LRO formation: see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4970331/ 2. In Fig 1 the conclusion that induced green autofluorescence localizes to gut granules is not convincing.Without a red channel image in the 1A control/BA it could be that autofluorescence increases in the red channel as well so what is perceived as Nile Red in gut granules is really autofluorescence accumulating somewhere else.The data in Supp Fig 1A suggests that BA does not lead to strong red fluorescence, however the green signal is so weak and the objectives/image acquisition settings are so different in this image it is difficult to be convinced of the authors conclusion that BA does not lead to a significant increase in red autofluorescence signal.While the glo-1(-) result in Fig 1C is consistent with gut granules being the location of the induced autofluorescence it is not enough on its own to make a strong conclusion about the location of autofluorescent material in BA conditions.3. Scale bars should be included in the insets in Fig 1B and 1C. 4. In Fig 3A,B, E the experiment leading to the data is unclear.Is it the length of time after being exposed to Nile Red or the period of time after feeding ceased that is refereed to in the figures? 5.The authors response to my original comment #10 is not convincing.They present experiments in Fig 3A,B, E, which I admit are not very clear to me, that do not directly address anything about transport to and from LROs, they just measure the brightness of red fluorescence, which may not even be from Nile Red (see comment 2).If the signal comes from Nile Red it is unclear how Nile Red gets there or what it stains so conclusions relating to transport or not justified.It is important to keep in mind that Nile Red can move between organelles and accumulates in hydrophobic environments.Without stronger data, the authors should consider removing the conclusion that BA elicits increased functional capacity of C. elegans gut granules (line 537).6.The authors clearly show an increase in autofluorescence upon culturing with BA.However, it is unclear that this is deposition vs chemical changes in pre-existing materials to cause enhanced fluorescence.I appreciate the authors careful attention to the concerns I raised in the initial review of my paper, and I think the data in this paper present an advance.However, before this paper is formally accepted, I request that the authors use more careful language when describing their results, particularly in the abstract.This is very important in my opinion, because in places, their text does not accurately present their findings, and includes several phrases that are over-interpretations of their findings.
1. Line 29-30.Please delete "consistent with effector-triggered immunity" from the abstract.The authors did not test this, identify bacterial mutant effectors that deplete LROs or show that this is a form of ETI.
2. Line 24.Please be more specific in the sentence "C.elegans LROs promote organismal protection."Such as, "Here we report that RNAi of genes that reduce LRO number leads to enhance susceptibility to multiple stresses, including the toxic xenobiotic benzaldyde, heat and oxidative stress and pathogen infection" 3. Line 29 Rather than the phrase "involving daf-6 and pmk-1" I ask that the authors state exactly what they showed.Something like "Upregulation of glo-1 dependent genes by benzaldehyde was reduced in daf-16 and pmk-1 dependent mutants".This is important because the authors did not confirm that these pathways were activated (e.g., with western blots for active p38, showing that these mutants are required for their phenotype, genetic epistasis experiments, etc.).

7 .
It is unclear how the data in Supp Fig 1 indicate that the BA response is independent of AA-derived and of glycated proteins (104-105).What would the authors expect to see if flu-1 were involved in the creation of the fluorescence and how does the flu-1(-) experiment address glycated proteins?Reviewer #3 (Remarks to the Author):