Submit to this Journal Review for this Journal Propose a Special Issue

Article Menu

Share Help Cite Discuss in SciProfiles

Open AccessArticle

Peer-Review Record

Gene Expression Profiling of Trematomus bernacchii in Response to Thermal and Stabling Stress

Fishes 2022, 7(6), 387; https://doi.org/10.3390/fishes7060387

by Samuele Greco^1,†

, Anastasia Serena Gaetano^2,†

, Gael Furlanis¹, Francesca Capanni³, Chiara Manfrin¹

, Piero Giulio Giulianini¹

, Gianfranco Santovito⁴

, Paolo Edomi¹

, Alberto Pallavicini¹

and Marco Gerdol^1,*

Reviewer 1:

Huaping Zhu

Reviewer 2: Anonymous

Reviewer 3:

Silvia Portela-Bens

Reviewer 4:

Wen-Sheng Liu

Fishes 2022, 7(6), 387; https://doi.org/10.3390/fishes7060387

Submission received: 28 October 2022 / Revised: 3 December 2022 / Accepted: 8 December 2022 / Published: 13 December 2022

(This article belongs to the Special Issue Impacts of Anthropogenic Stressors on Fish Physiology)

Round 1

Reviewer 1 Report

The present study reports the gene expression profiling of Trematomus bernacchii in response to thermal and stabling stress. The result seems interesting, however some points remain to be revised or clarified.

For example:

1.Why did you choose the temperature (+1.5°C)? please add this information.

2.The fish of the stress group were transfer from -0.9°C water to +0.6°C water directly?

3.Line 102 "All the animals were euthanized with tricaine", Please add the concentration of tricaine.

4.what is the fish size of T. bernacchii in this study?

5. Why the three tissues (skeletal muscle, gills and brain) were chose to observe the response to a moderate heat stress, the other tissues were not included, such as liver?

6. The study analyzed the results from RNA-seq sequencing. It would be better if the differential expression gene and ratiocination were further verified.

Author Response

The present study reports the gene expression profiling of Trematomus bernacchii in response to thermal and stabling stress. The result seems interesting, however some points remain to be revised or clarified.

For example:

1.Why did you choose the temperature (+1.5°C)? please add this information.

Thank you for this remark. Since the very same request has been provided by reviewer #2, we realized that the reasons why this temperature increase was selected were not clear.

This temperature increase was selected to simulate a realistic environmental condition that could be eventually reached, and has indeed already been reached in some coastal regions of Antarctica. The introduction section has been improved to provide to clarify this concept.

2.The fish of the stress group were transfer from -0.9°C water to +0.6°C water directly?

This is correct. While we recognize that this sudden temperature change may have led to acute stress, temperature changes of similar magnitude are experienced (albeit for brief time periods) by Antarctic fish in their natural environment, as revealed by temperature data logger records collected during our expedition. However, the confinement of the fishes in the tanks did not allow the adoption of an avoidance behavior, which likely occurs in the natural environment, allowing the animals to move to areas with lower temperatures (e.g. by moving to higher depths, or to waters covered by the ice sheet.

3.Line 102 "All the animals were euthanized with tricaine", Please add the concentration of tricaine.

Thank you for pointing this out. This information has been added to the text in the materials and methods section.

4.what is the fish size of T. bernacchii in this study?

All the individuals analyzed in this study were adults. Information about the mean size plus standard deviation has been added in the materials and methods section.

Why the three tissues (skeletal muscle, gills and brain) were chose to observe the response to a moderate heat stress, the other tissues were not included, such as liver?

The original plan also included the sampling of liver and head kidney. Unfortunately, despite the placement of the samples in RNAlater, the quality of the RNA extracted from these two tissues was not sufficient to allow the preparation of sequencing libraries, preventing the inclusion of these samples in subsequent analyses. This is just one of the major limitations of carrying out experiments in Antarctica, since all samples need to be transported from the Antarctic base back to the mainland, where all subsequent phases of the analyses take place. In this particular case, the samples were transported by ship to Europe with a long travel. We can hypothesize that the high level of enzymatic activity (and in particular of RNase activity) typical of these tissues led to an excessive degradation of RNA that could not be avoided by the RNAlater preserving solution. We noticed the same issue for several samples obtained from the three other samples, even though the amount of degradation was still sufficient to allow the preparation of 3’-tag libraries.

The study analyzed the results from RNA-seq sequencing. It would be better if the differential expression gene and ratiocination were further verified.

Undoubtedly, verifying the consistency of gene expression trends with other approaches on additional biological replicates would provide further support to our findings. Nevertheless, all the scientific activities carried out in Antarctica within PNRA expeditions are tightly regulated and, for all animal experimentation, only a limited number of individuals can be sampled in order to minimize the impact of scientific activities on natural populations. Consequently, such validation experiments cannot be easily planned for this target species. Nevertheless, we would like to remark the fact that erythrocyte morphology data, that has been also collected within the frame of this experiment, have been recently published (see Rizzotti et al. 2022, https://doi.org/10.1016/j.jtherbio.2021.103139), providing support to our conclusions.

Reviewer 2 Report

In the present study, authors studied the thermal and stabling stress on the transcriptomic response in the emerald rockcod Trematomus bernacchii. They found that the brain was identified as the most susceptible tissue to heat stress and immune response, protein synthesis and folding, and energy metabolism-related genes were differentially expressed. While the gills also displayed smaller significant alterations. What’s more, stabling stress also had impact on gene expression profiles in the brain. The authors suggested that more attention should be dedicated to an improved design of the experiments carried out on Antarctic organism. Overall, this study is of interest, although somewhere is needed to be improved. Please find my suggestions below.

1. Line 42-49, these sentences were a little redundant. Please simply.

2. The authors should explain why they use +1.5°C as a moderate heat stress.

3. How did authors control the +0.6°C in the experiment?

4. Were the fishes fed in the experiment?

5. The qRT-PCR is necessary to verify the accuracy of RNA-seq.

6. It is concern that the mapping rate was too low (21.77%).

7. The quality control information including (Raw reads, Clean reads, Q20, Q30, mapping rate) should be list in a table.

8. The Figure S1 was not found.

9. The genes in the supplementary file should be described by full name.

10. The DEG numbers under stabling stress should be list in a table.

11. Fig. 7 should be described in the Result section.

12. Line 436, he genes?

13. The conclusion was too long. For example, line 545-552 might be unsuitable in the conclusion.

Author Response

Line 42-49, these sentences were a little redundant. Please simply.

We have carefully revised the text, improved the readability of the introduction and removing redundancy. Revised parts are shown in red.

The authors should explain why they use +1.5°C as a moderate heat stress.

Thank you for this remark. Since the very same request has been provided by reviewer #1, we realized that the reason why this temperature increase was selected were not clear. This temperature increase was selected to simulate a realistic environmental condition that has already been reached in the Antarctic Peninsula and may interest increasing coastal areas in the future. We improved the description of the rationale behind our study in the revised introduction.

How did authors control the +0.6°C in the experiment?

Temperatures of the tanks were monitored with a Tinytag Aquatic 2 datalogger. This information has been added in the manuscript.

Were the fishes fed in the experiment?

The specimens were fed with pieces of the Antarctic scallop Adamussium colbecki. This information has been added to the manuscript.

The qRT-PCR is necessary to verify the accuracy of RNA-seq.

In principle, we would agree that extending gene expression analyses to additional biological samples would provide additional support to the results of this study. However, the accuracy of RNA-seq is not, by itself, a cause of concern, as this methodology has been clearly demonstrated on multiple occasions not to require a systematic validation by qRT-PCR, also because the two techniques have a different dynamic range and may lead to discordant results due to alternative splicing (depending on PCR primer design, see Everaert et al. 2017). See for example Coenye, 2021 (“the data available suggest that RNA-seq methods and data analysis approaches are robust enough to not always require validation by qPCR and/or other approaches, although there are situations where this may be of added value”).

We recognize that the limited number of biological samples analyzed in this study (n=3) may not be optimal to appropriately address the impact of inter-individual variability of response in T. bernacchii. All the scientific activities carried out in Antarctica within PNRA expeditions are tightly regulated and, for all animal experimentation, only a limited number of individuals can be sampled in order to minimize the impact of scientific activities on natural populations. Due to these limitations, we could not sample more individuals than those analyzed in the present study.

It is concern that the mapping rate was too low (21.77%).

We applied very stringent criteria to assess the quality of the datasets used in this study, removing a few samples that did not meet these thresholds, appearing as outliers. The few samples that displayed a particularly low mapping rate (in a single case, 21.77%) still retained a sufficiently high number of reads (4.32e6 reads, ~940k reads mapped) to allow a reliable calculation of gene expression levels, in line with indications from the original QuantSeq protocol publication (see Moll, P., Ante, M., Seitz, A. et al. QuantSeq 3′ mRNA sequencing for RNA quantification. Nat Methods 11, i–iii (2014). https://doi.org/10.1038/nmeth.f.376). In fact, the expression profile obtained from this sample was consistent with those obtained from the two other biological replicates.

Hence, we do not believe that low mapping rates were, per se, an issue for this type of RNA-seq protocol, which is often used for the analysis of low input/FFPE samples. We interpret the high variation in read mapping rates among samples as the result of the significant variability of the quality of extracted RNAs, which was often, as mentioned in a previous comment, not optimal, leading to the complete removal of all liver and head kidney samples from further analysis. Under such circumstances, other classes of abundant RNAs (such as non-depleted rRNA, mitochondrial mRNAs, etc.) can make up for a significant fraction of the fragments included in RNA-seq libraries. Luckily, the 3’-tag library preparation approach used in this case limited the impact of RNA degradation on the accuracy of gene expression estimates, as this method is not impacted by the strong 3’end bias of read mapping that is typically observed from classical RNA-seq libraries prepared using a poly-A selection strategy in samples with highly degraded RNA.

The quality control information including (Raw reads, Clean reads, Q20, Q30, mapping rate) should be list in a table.

This information is now provided, for all sequencing libraries, in the supplementary information.

The Figure S1 was not found.

Thank you for pointing this out. This issue with the supplementary files has been fixed.

The genes in the supplementary file should be described by full name.

This information was added to the supplementary information.

The DEG numbers under stabling stress should be list in a table.

We added a simple additional table listing the number of DEGs observed in response to stabling stress, together with the number of associated clusters of coregulated genes.

Fig. 7 should be described in the Result section.

Thank you for the suggestion, we moved the figure and the reporting of this data to the results section, keeping the discussion part in the proper section

Line 436, he genes?

Thank you. This was corrected to “The genes”

The conclusion was too long. For example, line 545-552 might be unsuitable in the conclusion.

Thank you for this suggestion. Following a similar comment provided by reviewer #4, we have completely rewritten the conclusions section.

Reviewer 3 Report

Experimental design is not correct for this type of experiment. First, acclimate period of the specimens is too short to leave that stress due to confinement get down. Each specie has his own time to acclimate to new conditions and also its depends on the change that you produce in its routine compared with its wild environment. This should keep in mind when you design an experiment and in this case it seems that have not been consider. Then I have not been able to see the duplicate of the experimental tanks, that you need to make your experiment reliable, robust and statistically significant. It means, you should do the experiment with a duplicate of your groups.

Finally, in the text you have quite clear that the acclimation period has masked your result about the objective of this paper, maybe you should reconsider this objective and raise what your results want to say again.

Author Response

With all due respect, we would like to remark that we are not working with zebrafish in a fully equipped laboratory on mainland, but rather on an Antarctic species in an Antarctic base.

While we are fully aware of the relevance of the acclimation period for the interpretation of gene expression results in fish, it would be simply impossible to apply longer acclimation periods at the Mario Zucchelli Italian station in Antarctica. All scientific activities in the base (and in any other base in Antarctica) take place within a rather short time period, which lasts approximately 3 months and a half. Due to the size of the base, not all research teams can be present at the base for the entire period. They are therefore granted shorter visiting times of approximately 40 days each, and all activities have to be carried out within this timeframe. Under such circumstances, longer acclimation periods would clearly not be feasible.

Due to the aforementioned limitations, the overwhelming majority of previously published experimental protocols carried out in T. bernacchii and other notothenioid fishes have used acclimation periods equal to or shorter than the 11 days used in our experiment, with the standard acclimation period being one week. In absence of any support whatsoever in scientific literature for the use of longer acclimation times in the target species, we do not believe that the criticism provided by the referee with this respect is reasonable.

See below a list of previously published studies, together with the length of the acclimation period used, for your reference:

-Sleadd & Buckley 2012 (https://doi.org/10.1007/s00300-012-1262-8): 48 hours

-Illuminati et al. 2008 (https://doi.org/10.1080/02772240902902349): 2/3 days

-Enzor et al. 2017 (https://doi.org/10.1093/conphys/cox019): one week

-Canapa et al. 2007 (https://doi.org/10.1016/j.chemosphere.2006.07.026): one week

-Bakiu et al. 2022 (https://doi.org/10.3390/ijms232112799): one week

-Benedetti et al. 2007 (https://doi.org/10.3390/ijms232112799): one week

-Zucchi et al. 2010 (https://doi.org/10.1016/j.envpol.2010.04.012): one week

-Vasadia et al. 2019 (https://doi.org/10.1016/j.margen.2019.100698): one week

-Di Bello et al. 2007 (https://doi.org/10.1016/j.aquatox.2007.05.010): one week

-Truzzi et al. 2018 (https://doi.org/10.1016/j.marenvres.2018.03.017) 10 days

-Giuliani et al. 2021 (https://doi.org/10.3390/antiox10030410): 10 days

With all the aforementioned considerations in mind, we respectfully disagree with the reviewer about the fact that “acclimation time was not considered for the design of the experiment”. We specifically chose the acclimation period based on the overwhelming consensus of scientific literature and on the experience of personnel who had previously worked on this species in previous Antarctic expeditions.

Moreover, we would like to remind to the reviewer that naïve samples wild (fishes dissected immediately after catch) were explicitly included in the experimental design with the precise aim to provide a snapshot of baseline expression profiles, something that is rarely done even in studies conducted in more favorable environments and with better equipment.

Please note that the differences observed between the naive samples and T0 (i.e. the first samples analyzed at the end of the acclimation period) did not highlight significant differences in terms of gene expression in any of the three studied tissues. This is clearly visible in Figure 5 (for brain) and Figure 8 (for gills), since the vast majority of the genes altered in response to stabling was up- or downregulated at later time points, with the only exception being represented by the few DEGs included in gill cluster 4. When we refer to “stabling stress” we are implying something much more complex than simple acclimation to laboratory conditions: as reported in the discussion, some components of stabling stress may be linked to the movement to smaller experimental tanks, the inability to adopt an avoidance behavior, the change in fish density per tank (all these situations would have clearly not been amended by the use of longer acclimation times), or even long-term effects of the confinement in laboratory conditions.

The reviewer also complained about the lack of a second tank to provide a technical replicate. In an ideal experiment, carried out in laboratories located in the mainland, this would have been an obvious choice. However, this was not possible in Antarctica: the space at the base was limited and, as mentioned above, facilities need to be shared by multiple research teams. Consequently, we only had access to two tanks at that time (i.e. one used for the control temperature, the other for the increased temperature treatment). Please note that the very same limitations have been described on multiple occasions for other previously published manuscripts that had to deal with the same limitations. See for example Huth & Place, 2016, where this issue has been explicitly addressed (https://doi.org/10.1186/s12864-016-2454-3). Previous experiments have, in any case, demonstrated that the tank effect was very limited in notothenioid fish, as reported in Enzor and Place 2014 (https://doi.org/10.1242%2Fjeb.108431) and Enzor et al. 2013 (https://doi.org/10.1016/j.cbpa.2012.07.016).

In summary, albeit certainly not meeting the standards that would have been used to carry out an experiment in model fish species in a European laboratory, we strongly believe that this was the best feasible experimental design to address the main biological question of our work in the facilities available at the Mario Zucchelli station. Far from not taking into account the issues linked with stabling, we feel like our experimental design was much more complex than what is routinely used for similar studies in Antarctica. Consequently, we respectfully (but strongly) disagree with the criticism provided by the reviewer and we kindly invite him/her to reconsider his/her suggestions based on the body of scientific literature available concerning fish experimental research in Antarctica.

Despite these disagreements with the reviewer, we agree about the fact that such limitations have to be clarified to the reader and we have consequently added a new subsection to specifically address this limitations in the materials and methods section. The added text reads as follows:

“The experimental phases of this study were conducted at the Mario Zucchelli Station at Terranova Bay, Antarctica and were therefore limited by the available hardware at the facilities and by the access time granted to each research team. All the available time and instrumentation (including tanks and related hardware) were used for this experiment, which could only rely on the use of a single tank for each of the two experimental conditions, without the possibility of including technical replicates. Nevertheless, as previously reported in other studies carried out on the same species, this factor is unlikely to have significantly impacted the outcome of the study. and long acclimation times. We also recognize that, in light of the results of our study, the selected acclimation time (i.e. 11 days) might have had a significant impact on the response of T. bernacchii to stabling. Nevertheless, this experimental design was planned based on the consensus of the scientific literature available on the subject at the time, which indicated an acclimation time between 7 and 10 days as appropriate for the target species and other notothenioid fishes.”

Once again, we respectfully disagree with the interpretation of the reviewer. The acclimation period had very little effect on gene expression profiles themselves, as revealed by the near-complete lack of DEGs in the comparison between the naive and T0 samples (please see Figures 5 and 8). Stabling stress (intended as a combination between the confinement in small experimental tank, the restricted possibility of movement, the variation of fish density, etc.) did on the other hand have a significant effect on gene expression trends. Nearly half of the results and discussion sections are dedicated to the reporting and interpretation of the effects of stabling stress, which is also clearly highlighted as one of the key points of this work in the title of the manuscript itself, so we believe that this topic has been already covered quite extensively in the present version of the manuscript.

Finally, please note that we have already discussed quite comprehensively the possibility that stabling stress might mask the effects of the heat treatment in a previously published paper, entitled “Morphological analysis of erythrocytes of an Antarctic teleost under heat stress: Bias of the stabling effect” (https://doi.org/10.1016/j.jtherbio.2021.103139).

Nevertheless, as suggested by reviewers #2 and #4, the conclusions section has been rewritten and now includes further mention of the fact that these results should be taken with a grain of salt.

Reviewer 4 Report

In this study, the emerald rockcod Trematomus bernacchii was used as a model species to investigate the effects of a 20-day long exposure to a +1.5°C increase in the brain, gills and skeletal muscle, using a RNA-sequencing approach. The results clearly identified the brain as the most susceptible tissue to heat stress, with evidence of a time-dependent response dominated by an alteration of immune response, protein synthesis and folding, and energy metabolism-related genes. While the gills displayed smaller but still significant alterations, the skeletal muscle was completely unaffected by the experimental conditions. The stabling conditions also had an important impact on gene expression profiles in the brain, suggesting the presence of significant alterations of the fish nervous system, possibly due to the confinement to tanks with limited water volume and of the restricted possibility of movement. Besides providing novel insights in the molecular mechanisms underlying thermal stress in notothenioids, these findings suggest that more attention should be dedicated to an improved design of the experiments carried out on antarctic organism, due to their extreme susceptibility to the slightest environmental alterations. However, there are still some problems that need to be corrected in the article.

Major comment

It reached 1.5°C higher than pre-industrial at the end of this decade, but this warming is unstable in the Antarctic, and the warming process is slow, which may also lead to some adaptive evolution of Trematomus bernacchii. Therefore, the need to study the effect of moderate warming (1.5°C) on emerald rockfish needs to be further clarified.

Line 85: In the present work, why skeletal muscle, gills and brain were used as objects of study, the rest of the tissues similar to the liver are also important organs.

Line 95: Wild emerald rockfish were caught and acclimated for 11 days before the experiments were performed, which may affect the representativeness of the RNA-seq results of the heat stress experiment.

To further illustrate the effects of prolonged moderate warming on the skeletal muscle, gill and brain tissues of emerald rockfish, it might be better to add pathological section experiments of these tissues.

In the results section, images should not be displayed as separate results, which can lead to confusion and reduce the readability of the article.

Line 356: The results corresponding to Figure 7 should be described in the Results section and should not appear in the Discussion section.

Line 436: “he” should be revised to “The”.

The font size in Figures 2, 3 and 5 should be adjusted, some of the fonts overlap and are not conducive to reading.

The article is not very readable and the author needs to make extensive revisions to the introduction and discussion sections to make them more readable.

Line 515: The conclusion section is too unreadable, perhaps a shorter description would be better.

Author Response

Major comment

While the average temperature in Antarctica is rising at a slower rate than on other continents, this is not true for all regions (particularly the Antarctic Peninsula, which is about +3°C warmer than pre-industrial levels). As the reviewer correctly points out, the change is also unstable, with the greatest warming occurring in the summer and extreme freezing events occurring in the winter. As a result, the length of periods without ice covering coastal waters increases, allowing direct solar radiation into the water and causing warming of the water column. The temperature we used for the control tank, -0.9°C, was based on the seawater temperature measured close to the Mario Zucchelli station, where the fish were caught, which is already warmer than the yearly average Antarctic ocean temperature of around -1.7 °C.

Such events are brief, but the temperature increase is significant enough to affect marine life. Seasonal occurrences, as well as the relatively short time frames of direct warming from the sun, do not allow for adaptive evolution.

We further expanded the reasons behind the choice of experimental temperature, also following a similar remark made by reviewer #1. The introduction has been updated, with a reinforced statement about our temperature of choice.

Line 85: In the present work, why skeletal muscle, gills and brain were used as objects of study, the rest of the tissues similar to the liver are also important organs.

We most definitely agree with the reviewer. The original plan also included the sampling of liver and head kidney. Unfortunately, despite the placement of the samples in RNAlater, the quality of the RNA extracted from these two tissues was not sufficient to allow the preparation of sequencing libraries for at least 3 biological replicates per experimental time point, preventing the inclusion of these samples in subsequent analyses. This is just one of the major limitations of carrying out experiments in Antarctica, since all samples need to be transported from the Antarctic base back to the mainland, where all subsequent phases of the analyses take place. In this particular case, the samples were transported by ship to Europe with a long travel. We can hypothesize that the high level of enzymatic activity (and in particular of RNase activity) typical of these tissues led to an excessive degradation of RNA that could not be avoided by the RNAlater preserving solution. We noticed the same issue for several samples obtained from the three other samples, even though the amount of degradation was still sufficient to allow the preparation of 3’-tag libraries.

This is certainly possible. Unfortunately, carrying out similar experiments in an Antarctic base has had some important limitations, which among the others include the impossibility of planning longer acclimation periods. All scientific activities carried out at the Mario Zucchelli base (and in any other base in Antarctica) take place within a rather short time period, which lasts approximately 3 months and a half. Due to the size of the base, not all research teams can be present at the base for the entire period. They are therefore granted shorter visiting times of approximately 40 days each, and all activities have to be carried out within this timeframe. Under such circumstances, longer acclimation periods would be clearly not feasible.