The metabolome of Mexican cavefish shows a convergent signature highlighting sugar, antioxidant, and Ageing-Related metabolites
Curation statements for this article:-
Curated by eLife
Evaluation Summary:
Medley et al. study A. mexicanus, an extreme-adapted organism with important connections to human health. The authors test metabolic responses in this natural model of elevated blood glucose and extensive body fat deposits, conditions generally expected to predispose to higher risk for metabolic syndrome and higher frailty. The work is rigorous and will provide a reference for future studies aimed at dissecting the mechanistic basis underlying metabolic shifts in this uniquely attractive model. The authors also provide an open and accessible window into their data and analyses by sharing a Shiny app.
(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 agreed to share their name with the authors.)
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
Abstract
Insights from organisms, which have evolved natural strategies for promoting survivability under extreme environmental pressures, may help guide future research into novel approaches for enhancing human longevity. The cave-adapted Mexican tetra, Astyanax mexicanus , has attracted interest as a model system for metabolic resilience , a term we use to denote the property of maintaining health and longevity under conditions that would be highly deleterious in other organisms (Figure 1). Cave-dwelling populations of Mexican tetra exhibit elevated blood glucose, insulin resistance and hypertrophic visceral adipocytes compared to surface-dwelling counterparts. However, cavefish appear to avoid pathologies typically associated with these conditions, such as accumulation of advanced-glycation-end-products (AGEs) and chronic tissue inflammation. The metabolic strategies underlying the resilience properties of A. mexicanus cavefish, and how they relate to environmental challenges of the cave environment, are poorly understood. Here, we provide an untargeted metabolomics study of long- and short-term fasting in two A. mexicanus cave populations and one surface population. We find that, although the metabolome of cavefish bears many similarities with pathological conditions such as metabolic syndrome, cavefish also exhibit features not commonly associated with a pathological condition, and in some cases considered indicative of an overall robust metabolic condition. These include a reduction in cholesteryl esters and intermediates of protein glycation, and an increase in antioxidants and metabolites associated with hypoxia and longevity. This work suggests that certain metabolic features associated with human pathologies are either not intrinsically harmful, or can be counteracted by reciprocal adaptations. We provide a transparent pipeline for reproducing our analysis and a Shiny app for other researchers to explore and visualize our dataset.
Article activity feed
-
-
Author Response:
Reviewer #3 (Public Review):
In this study Medley and colleagues study the remarkable metabolic phenotypes of cave-adapted Mexican tetra - Astyanax mexicanus. Cave adapted fish populations have adapted several ways to cope with cave environments including lower metabolic rate, increased appetite, fat storage, and starvation resistance. Simultaneously though they are insulin resistant, hyperglycemic, and take in more calories. These fish also have a mutation in insulin receptor that in humans results in extremely deleterious metabolic consequences - however, cavefish do not appear to exhibit adverse effects.
To understand the adaptations that have led to these remarkable phenotypes the authors performed metabolomic profiling on two cave-adapted, and one non-cave adapted population under three different experimental …
Author Response:
Reviewer #3 (Public Review):
In this study Medley and colleagues study the remarkable metabolic phenotypes of cave-adapted Mexican tetra - Astyanax mexicanus. Cave adapted fish populations have adapted several ways to cope with cave environments including lower metabolic rate, increased appetite, fat storage, and starvation resistance. Simultaneously though they are insulin resistant, hyperglycemic, and take in more calories. These fish also have a mutation in insulin receptor that in humans results in extremely deleterious metabolic consequences - however, cavefish do not appear to exhibit adverse effects.
To understand the adaptations that have led to these remarkable phenotypes the authors performed metabolomic profiling on two cave-adapted, and one non-cave adapted population under three different experimental conditions. Overall, the experiment is really interesting and a wealth of data are generated from an important model system. However, I did not find the presented analyses of the data to be very convincing. While there are some interesting observations made, the individual results are often presented out of the context of the whole dataset - that is, it is hard to know how "significant" or important changes in any particular metabolite are when they are presented in isolation. There are a couple of places (Fig 4 and Fig 9), where a hypothesis is tested using the 'omics' data (see specific comments below) and I think focusing on these alongside presenting the data as a resource for the community would strengthen the manuscript.
Specific comments
The comment about "genetic ancestry" on page 5 is not correct I don't think - the shared homology between tissues would be better described as the "evolutionary conserved functions of individual tissues."
We are indebted to the reviewer for this comment! We find this a much clearer sentence and have changed the text accordingly
The increased similarity in metabolic profiles between cave-fish compared to surface fish presented in Fig 4 is very interesting but confusing in how it is presented. I think the text and the figure could be presented in a way that is more concise.
We agree and based on similar comments from all reviewers have decided to remove the figure. We believe the trends shown in this figure are well-represented by the remaining figures in the manuscript.
On page 5 the authors comment that fructose and fructose phosphates tend to be upregulated in the brain, but that does seem to be the case in Figure 5?
We regret the confusion caused by this figure. Fructose levels are elevated in cave populations, but levels in the brain are much lower than liver. This figure was previously normalized per row (so that the tissue with the highest concentration tended to show up clearly but tissues with lower concentrations were barely visible). We have substituted a figure that is normalized per-tissue so that relative differences between populations can be discerned within a given tissue. Separately from this, we have removed the fructose comment as it was a speculative comparison with naked mole-rats.
Are any of the results in figure 5 significant?
At a technical level, our Bayesian approach cannot predict significance (because null hypothesis rejection is a frequentist concept), but Supplementary Table 2 shows which metabolites are most likely to be up- or down-regulated according to our statistical model.
Throughout the paper there does not seem to be multiple testing corrections?
While multiple testing correction is standard for a frequentist-based statistical model, in the case of a Bayesian statistical model such as ours, the choice of prior can help in reducing erroneous conclusions due to overestimation of effect size. We use the highly conservative default prior of the “bayesglm” function of the arm R package, which covers a range of ±5 on the logistic scale. We hope that this choice of prior will serve in lieu of multiple testing correction.
The entire section on Obesity and Inflammation-related metabolites refers the reader to supplementary data. It would be helpful to have some display items / tables for the reader to refer to here to interpret these results.
I'm not sure Fig 8 is significant after multiple testing correction.
Significance values in this figure are based on Bayesian GLM posteriors (they are not technically “significance” values in the frequentist sense), but we find them helpful in determining which metabolites have the most skewed posteriors (largest effect size). As above, the choice of prior should help eliminate erroneous conclusions.
I think a more robust approach is needed to compare the data from different organisms to the cavefish. Perhaps correlating the metabolites or projecting them into the PCA from these conditions? It's hard to know in the Obesity and Inflammation-related metabolites what to make of the similarities and differences between humans and cave fish. The observations are indeed intriguing, but, I can't tell how different / similar they are to expectation given the handful of examples presented.
We thank the reviewer for these suggestions, and we agree a more comprehensive approach is needed to draw comparisons between human and cavefish metabolic trends. However, as part of the overall “toning down” of the mechanistic language in the manuscript, we have decided to instead remove comparisons with other organisms such as humans and naked mole rats.
The comment about positive selection (page 10) seems a bit out of place - suggest being more circumspect, "perhaps a locus under selection."
Indeed, we have incorporated this suggestion into the manuscript.
The statistical analyses for the section on Resistance to Nutrient Deprivation are very clear and the explicit "omics" test of a hypothesis is well laid out. I wish previous analyses had taken a similar approach. However, that said I think a multiple testing correction might need to be applied in Fig 9 data.
We hope our choice of Bayesian prior will serve in lieu of multiple testing correction.
Fig S7 is quite interesting and seems well suited to the main text!
We thank the reviewer for this suggestion. With so many figures, we were uncertain which ones to include in the main text. We have moved this figure (ROS imaging) to the main text (it is now Fig 8).
A lot of redundant information is in the figures - they could be streamlined quite a bit. There also seems to be a too many figures, and they could potentially be combined.
We agree and note similar comments from all reviewers. We have removed Fig 4 and 8 of the original submission (Venn diagrams and blood lipoprotein).
The observed overlap between cavefish metabolic adaptations and those found in naked mole rats seems tenuous - certainly there are similarities and this should be pointed out, but it's hard to judge how significant / important these are.
We agree and as part of our revision we have removed comparisons with naked mole rats and humans.
-
Evaluation Summary:
Medley et al. study A. mexicanus, an extreme-adapted organism with important connections to human health. The authors test metabolic responses in this natural model of elevated blood glucose and extensive body fat deposits, conditions generally expected to predispose to higher risk for metabolic syndrome and higher frailty. The work is rigorous and will provide a reference for future studies aimed at dissecting the mechanistic basis underlying metabolic shifts in this uniquely attractive model. The authors also provide an open and accessible window into their data and analyses by sharing a Shiny app.
(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 agreed to share their name …
Evaluation Summary:
Medley et al. study A. mexicanus, an extreme-adapted organism with important connections to human health. The authors test metabolic responses in this natural model of elevated blood glucose and extensive body fat deposits, conditions generally expected to predispose to higher risk for metabolic syndrome and higher frailty. The work is rigorous and will provide a reference for future studies aimed at dissecting the mechanistic basis underlying metabolic shifts in this uniquely attractive model. The authors also provide an open and accessible window into their data and analyses by sharing a Shiny app.
(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 agreed to share their name with the authors.)
-
Reviewer #1 (Public Review):
The authors compare metabolic states in one surface and two cave populations of the Mexican tetra (Astyanax mexicanus). This specis has a great appeal in that the derived cave populations present traits associated with human metabolic syndrome, insulin resistance and would generally be expected to be associated with a series of fitness-challenging conditions. However, extant cave populations seem to be resilient to such metabolic state, and the authors here provide an account of metabolic states in liver, muscle and brain.
The authors are able to pinpoint convergent metabolic signatures in cave populations respect to one surface population. Specifically, sugar phosphate metabolism appears somehow convergent among the cave populations, muscle-specific ascorbic acid is also more aboundant among cave populations.
Reviewer #1 (Public Review):
The authors compare metabolic states in one surface and two cave populations of the Mexican tetra (Astyanax mexicanus). This specis has a great appeal in that the derived cave populations present traits associated with human metabolic syndrome, insulin resistance and would generally be expected to be associated with a series of fitness-challenging conditions. However, extant cave populations seem to be resilient to such metabolic state, and the authors here provide an account of metabolic states in liver, muscle and brain.
The authors are able to pinpoint convergent metabolic signatures in cave populations respect to one surface population. Specifically, sugar phosphate metabolism appears somehow convergent among the cave populations, muscle-specific ascorbic acid is also more aboundant among cave populations.
Rather than validating novel mechanistic insights into cavefish biology (e..g explaining their resilience to the poor nutrient conditions of the cave environment), this work points towards plausible associations between metabolic states in cave populations and cavefish biology.
What strikes most about this work is the laudable effor to share this important resource with the community, via the development of a Shiny app.
-
Reviewer #2 (Public Review):
The goal of this study is to provide a broad, untargeted study of primary metabolites and lipids of long- and short-term fasting in two A. mexicanus cave populations and one surface population. The authors have tried to identify the molecular signatures underline the low-nutrient adaptation in cave-dwelling subpopulations, and clarify the metabolic differences between morphotypes that might explain A. mexicanus longevity.
In the last few years A. mexicanus has been advanced as a model to investigate metabolic strategies for survival in harsh cave environment, which are not currently well-understood. This paper provide new insights about the evolutionary and physiological implications of these metabolic changes. The hypothesis behind this study is that A. mexicanus troglomorphic populations exhibit …
Reviewer #2 (Public Review):
The goal of this study is to provide a broad, untargeted study of primary metabolites and lipids of long- and short-term fasting in two A. mexicanus cave populations and one surface population. The authors have tried to identify the molecular signatures underline the low-nutrient adaptation in cave-dwelling subpopulations, and clarify the metabolic differences between morphotypes that might explain A. mexicanus longevity.
In the last few years A. mexicanus has been advanced as a model to investigate metabolic strategies for survival in harsh cave environment, which are not currently well-understood. This paper provide new insights about the evolutionary and physiological implications of these metabolic changes. The hypothesis behind this study is that A. mexicanus troglomorphic populations exhibit similarities with human obesity and diabetes mellitus, but also evolved "biological counterbalance" in order to offset the negative effects of this pathological phenotype.
Strengths:
This study provides for the first time a large, untargeted metabolomics and lipidomics study of A. mexicanus surface and cave morphs, in different tissues and feeding regimes. This study confirms that cavefish morphs have an altered sugar metabolism, and exhibit higher levels of sugars and sugar phosphates. Interesting, the opposite trend occurs for uronic acids. Additionally, the authors have also found important differences in ascorbate and glutathione (which have been demonstrated to be effective modulators of antioxidant systems), α-KGn (which has been implicated in longevity in C. elegans) nicotinamide (which is a precursor to NAD` synthesis and hence is related to oxidative metabolism). Moreover, they have also shown via ROS staining that Pachòn cavefish presents a lower level of superoxide radicals in the liver after 30 days of fasting. All these results correlate with the cavefish physiological state, and suggests that selection in cave environments favours resistance to oxidative stress. In conclusion, these findings confirm in a reasonable manner the authors' hypothesis, but at the same time raise new questions about the molecular mechanisms underlines the low-nutrient adaptation in cave-dwelling subpopulations.
This data is useful for the cavefish field, especially the availability of an interactive app.
Weaknesses:
The conclusions of this paper are based almost exclusively on metabolomics data. Mass spectrometry techniques, because of their sensitivity and selectivity, have become methods of choice to characterize the metabolome. However there are still many unsolved problems related to these techniques, such as the lack of well established and standardized methods or procedures, and the difficulties still met in the identification of the metabolites influenced by a given feeding regime. Therefore, metabolomics study should be always validated with alternative techniques (ex. transcriptomic analysis). In this study, the lack of a rigorous validation of the data represent a major issue. Validation would also allow the authors to be less speculative.
The authors have performed a comparative metabolomics analysis between A. mexicanus surface and cave morphs in three different tissues and feeding regimes. The large amounts of data with a high number of variables greatly complicates the understanding of the data. This issue emerges for example in figure 2: the authors show a complex system of Venn diagrams across different morphs, tissues and feeding conditions. This figure is confusing and does not provide a clear understanding of the result.
Another limitation of this study is the fish age. As the authors stated, the analysis was based on juvenile, pre-sexually mature fish. This early developmental stage does not take into account age-related hormonal variations, which may have an impact on the fish metabolism and homeostasis. Therefore, further studies are needed to confirm the results across different developmental stages.
-
Reviewer #3 (Public Review):
In this study Medley and colleagues study the remarkable metabolic phenotypes of cave-adapted Mexican tetra - Astyanax mexicanus. Cave adapted fish populations have adapted several ways to cope with cave environments including lower metabolic rate, increased appetite, fat storage, and starvation resistance. Simultaneously though they are insulin resistant, hyperglycemic, and take in more calories. These fish also have a mutation in insulin receptor that in humans results in extremely deleterious metabolic consequences - however, cavefish do not appear to exhibit adverse effects.
To understand the adaptations that have led to these remarkable phenotypes the authors performed metabolomic profiling on two cave-adapted, and one non-cave adapted population under three different experimental conditions. Overall, …
Reviewer #3 (Public Review):
In this study Medley and colleagues study the remarkable metabolic phenotypes of cave-adapted Mexican tetra - Astyanax mexicanus. Cave adapted fish populations have adapted several ways to cope with cave environments including lower metabolic rate, increased appetite, fat storage, and starvation resistance. Simultaneously though they are insulin resistant, hyperglycemic, and take in more calories. These fish also have a mutation in insulin receptor that in humans results in extremely deleterious metabolic consequences - however, cavefish do not appear to exhibit adverse effects.
To understand the adaptations that have led to these remarkable phenotypes the authors performed metabolomic profiling on two cave-adapted, and one non-cave adapted population under three different experimental conditions. Overall, the experiment is really interesting and a wealth of data are generated from an important model system. However, I did not find the presented analyses of the data to be very convincing. While there are some interesting observations made, the individual results are often presented out of the context of the whole dataset - that is, it is hard to know how "significant" or important changes in any particular metabolite are when they are presented in isolation. There are a couple of places (Fig 4 and Fig 9), where a hypothesis is tested using the 'omics' data (see specific comments below) and I think focusing on these alongside presenting the data as a resource for the community would strengthen the manuscript.
Specific comments
The comment about "genetic ancestry" on page 5 is not correct I don't think - the shared homology between tissues would be better described as the "evolutionary conserved functions of individual tissues."
The increased similarity in metabolic profiles between cave-fish compared to surface fish presented in Fig 4 is very interesting but confusing in how it is presented. I think the text and the figure could be presented in a way that is more concise.
On page 5 the authors comment that fructose and fructose phosphates tend to be upregulated in the brain, but that does seem to be the case in Figure 5?
Are any of the results in figure 5 significant?
Throughout the paper there does not seem to be multiple testing corrections?
The entire section on Obesity and Inflammation-related metabolites refers the reader to supplementary data. It would be helpful to have some display items / tables for the reader to refer to here to interpret these results.
I'm not sure Fig 8 is significant after multiple testing correction.
I think a more robust approach is needed to compare the data from different organisms to the cavefish. Perhaps correlating the metabolites or projecting them into the PCA from these conditions? It's hard to know in the Obesity and Inflammation-related metabolites what to make of the similarities and differences between humans and cave fish. The observations are indeed intriguing, but, I can't tell how different / similar they are to expectation given the handful of examples presented.
The comment about positive selection (page 10) seems a bit out of place - suggest being more circumspect, "perhaps a locus under selection."
The statistical analyses for the section on Resistance to Nutrient Deprivation are very clear and the explicit "omics" test of a hypothesis is well laid out. I wish previous analyses had taken a similar approach. However, that said I think a multiple testing correction might need to be applied in Fig 9 data.
Fig S7 is quite interesting and seems well suited to the main text!
A lot of redundant information is in the figures - they could be streamlined quite a bit. There also seems to be a too many figures, and they could potentially be combined.
The observed overlap between cavefish metabolic adaptations and those found in naked mole rats seems tenuous - certainly there are similarities and this should be pointed out, but it's hard to judge how significant / important these are.
-
-
-