Review Reports - Fish Hybridization Leads to Uncertainty Regarding Ciguatera Fish Poisoning Risk; Confirmation of Hybridization and Ciguatoxin Accumulation with Implications for Stakeholders

Round 1

Reviewer 1 Report

this is excellent work and an important reminder that hybridization of food fish can be a significant risk to consumers. This team had the needed expertise to approach the problem, resolve the issue and provide a template for future studies of toxin accumulation in hybrid species. Well done.

Author Response

Thank you for reviewing our manuscript, we greatly appreciate your time and efforts.

Kind regards

Reviewer 2 Report

ID: jmse-473445

The study by Loeffler et al., entitled "Methods to confirm hybridization and ciguatoxin accumulation between species of conflicting historical risk for ciguatera fish poisoning" focused on the contamination of fish with ciguatoxins, in the US Virgin Islands area. The harvesting of two fish not clearly identified by fishermen has revealed a hybridization between two species and raises an interesting question, as one parent is considered to be at high risk of ciguatera, while the other is not. The main carribean ciguatoxin (C-CTX-1) is found in one of the fish, using an analytical chemistry method (LC-MS / MS with a standard) and a cytolotoxicity method on neuroblastoma (Neuro2A). The presence of this toxin is well established, which raises an emerging problem related to phycotoxins: the questioning of fish "at risk" compared to "risk-free" fish in endemic areas.

This article offers interesting data that deserves to be published. I ask the authors about two remarks:

- This article discusses Ocyurus chrysurus "considered a 'safe' species for consumption over CFP" (L. 65-66) with references 9-11. Yet, the yellow-tail snapper (Ocyurus chrysurus) is listed among the ciguatoxic species in the western Atlantic of a 2001 study (Farstad DJ, Chow T, 2001. A brief case report and review of ciguatera poisoning Wilderness Environ Med 12: 263-269). The status of high or low probability of accumulating ciguatoxins needs to be further clarified. Perhaps the status of this fish has changed over time. The question deserves to be addressed.

- This article lacks a Figure 1 with two images: a photo of Ocyurus Chrysurus and a photo of Lutjanus apodus, so as to point out that these two fish, which do not belong to the same genus, are of the family Lutjanidae, and therefore not far from a phylogenetic point of view. If the authors have it, it would also be interesting to have an image of the hybrid fish in between.

Figure 3: please indicate the LC50 next to the curve ((a) and (b)). Please insert OV (Ouabain/Veratridin) in the figure caption.

Author Response

This article discusses Ocyurus chrysurus "considered a 'safe' species for consumption over CFP" (L. 65-66) with references 9-11. Yet, the yellow-tail snapper (Ocyurus chrysurus) is listed among the ciguatoxic species in the western Atlantic of a 2001 study (Farstad DJ, Chow T, 2001. A brief case report and review of ciguatera poisoning Wilderness Environ Med 12: 263-269). The status of high or low probability of accumulating ciguatoxins needs to be further clarified. Perhaps the status of this fish has changed over time. The question deserves to be addressed.

We have added a statement addressing this potential and the reference contained within the Farstad manuscript ‘Dammann 1969 survey of USVI fishing potential’.

We now include a Figure 1 with the suggested photos and include the museum voucher numbers for all pictured individuals.

Figure 3: please indicate the LC50 next to the curve ((a) and (b)). Please insert OV (Ouabain/Veratridin) in the figure caption.

This information has been added to figure 4 (former figure 3).

Reviewer 3 Report

This paper describes the identification, in US Virgin Islands, of two fish samples resulting in the hybridization between a local commercially prized species considered as ‘low risk of ciguatera’, Ocyurus chrysurus, and a regionally prohibited species considered as ‘medium risk of ciguatera’, Lutjanus apodus. The hybrid status of these two fish samples was confirmed through genetic analysis using COI mitochondrial gene (for the identification of maternal lineage) as well as nuclear gene S7 (for the identification of paternal lineage). In addition, the toxicity of these two hybrid samples was evaluated through N2a assay and LC-MS/MS analysis: one of the samples was found toxic (0.03 ppb C-CTX1 eq., confirmation of the presence of C-CTX1) whereas the other was found negative. Authors concluded that additional work is required to determine the risk this hybrid actually holds for consumers.

Broad comments

Overall, the topic of this study is really interesting. However, in my opinion, since only two fish were analyzed, there are not enough results to conclude on the toxic potential of this new fish hybrid species and toxicological analysis of several other samples are needed before publication.

In addition, the presentation of toxicological analysis (material and methods as well as results) has to be improved, especially for N2a. Indeed, no information about the protocol used in this study are provided while some data are crucial to evaluate the quality of the results reported. It is essential to provide at least LOD and LOQ values, as well as EC₅₀ of the C-CTX1 standard used to calibrate the assay. Furthermore, the presentation of N2a results is not satisfying (see specific comments below).

Specific comments

L2-4: the title is unclear, according to me ‘ciguatoxin accumulation’ is not in the right place and should be placed at the end of the sentence. Furthermore, methods proposed in the present study are not new and were previously well described so, in my opinion, the title is ambiguous and should be entirely revised.

L19 and L60: ‘several fish that appeared to be hybrids’ vs. ‘two fish that appeared to be the result of hybridization’. It is unclear for me if several specimens of two potential hybrid species have been observed by fishermen or if only two hybrid samples have been observed and caught? Thus, I don’t know if these hybrid fish are uncommon or frequently observed.

L24: What do you mean by ‘composite cytotoxicity’?

L28: What about the comparison with the C-CTX1 levels found in the ‘prohibited’ parent?

L31-32 and 124-125: ‘provides a methodology for future studies into novel CFP vectors’ & ‘provide methodology that can be used to determine the risks these hybrids pose for causing CFP’. Like the title, these sentences let think that a new methodology is proposed in this study, however, in my opinion, it is not the case.

L47-49: You could cite one or two references at the end of this sentence.

L56-59: Why do you cite only the study by Tosteton (1985) whereas more recent data are available? Indeed, prevalence estimations reported by [6] are very high and more recent studies (Tester et al. 2009 & 2010 for example) have reported more realistic data for this region.

L59-62: It could be interesting to provide pictures of the two parent and two hybrid fish, even if you place them in supplemental information.

L133: please define N2a since it is its first apparition in the text.

L139: two fish samples appears to be not enough. Furthermore, why don’t you have analyzed at the same time some fish samples of the two parent species?

L149 & 155: please specify the origin of the C-CTX1 standard used to calibrate your toxicological analysis.

L149-153: You have to indicate at least the LOD and LOQ of your assay. It would be appreciated to briefly described, as for the other sections, the N2a assay protocol (final volume in wells, concentrations of ouabain/veratridine, time of incubation, volume of samples in wells, dilution ranges, etc…).

L155-171: What are the LOD and LOQ ?

L200: please specify the codes for the 2 hybrids samples (i.e. FDA321 and H2) in order to help the lector to find these two samples in the Figure 1.

L207-208: ‘The results of the S7 dataset strongly suggest that the two individuals are indeed hybrids and that the paternal species was O. chrysurus’. Why do you say ‘strongly suggest’ and not ‘confirm’? Can’t you be sure of the hybridization with these results? In this case, in the discussion section, you should propose additional analysis to firmly confirm these preliminary results.

L271: ‘Data regarding CTX levels for L. apodus are currently lacking’. Why don’t you have also analyzed L. apodus samples to provide these lacking data and thus be able to compare the toxicity of your hydrid samples with the ‘real toxicity’ of the parent fish species considered as ‘medium risk of ciguatera’?

L273: ‘one contained a 0.03 ppb C-CTX-1 Eq. concentration, with the presence of C-CTX-1 confirmed by LC-MS/MS’. Why don’t you have quantify levels of C-CTX1 in the hybrid sample using LC-MS/MS results?

L274-275: ‘the second hybrid fish (FDA321/USNM444947) contained no detectable levels of CTXs’. Again, what are the LOD of your N2a assay and LC-MS/MS analysis?

L278 (Figure 3): Usually, the x axis is expressed in mass/volume (e.g. pg/mL). Here, as there is no information about the final volume in wells, it is impossible for the lector to deduce the EC₅₀ of the C-CTX1 standard in order to compare it with the values reported in other studies. You have to change the x axis of your figure or to indicate in the text the EC₅₀ of the C-CTX1 (expressed in pg/mL for example). Furthermore, what do you mean by ‘C-CTX-1 reference standard applied at 5 pg’? To obtain this sigmoidal response, you have tested several concentrations of your standard (dilution range), haven’t you? Same comments for the hybrid fish sample.

L308 (Figure 4): I don’t understand why no quantification of C-CTX1 in the hybrid sample is provided. The lector can’t calculate it because no information is available about the concentration of the sample injected but the authors should them be able to give a quantification.

L315-316: ‘Efforts at obtaining additional individuals to fully assess the CFP risk for this hybrid species were unsuccessful’. For which reason? You have to better explain this point because it is one of the major ‘lack’ of your study. Is it because these two hybrid fish were never observed again? But, in this case, does this hybrid species really represent a risk for consumers?

L322-324: ‘A CTX risk profile for hybrid species has not yet been reported. Therefore, methods applicable for the determination of CFP risk profiles of a hybrid between two species with differing historical risk is of ecological and toxicological importance’. The methods are exactly the same that for parent species.

L357-358: ‘This study was an opportunity to observe how the crossing of two species is reflected in CTX bioaccumulation, compared with the normal historical CTX content of the parent species’. No data are available for CTXs contents found in L. apodus, so no comparison is possible with the second parent species. In addition, you have analyzed only two hybrid samples, one found slightly toxic and the other found no toxic so how could you conclude about the reflection of the crossing of two species in the CTX bioaccumulation? You have obtained contrasting results so, as you say in the rest of your discussion, you have to analyze more samples to conclude. Unfortunately, in my opinion, you have to do this before to be able to publish because the present results are too preliminary.

L366-368: ‘This can be described by observing the maximum CTX toxin content in O. chrysurus throughout the St. Thomas region (including individuals in excess of 1 kg), which was lower than the concentration measured in the hybrid’. You have only one value for the hybrid sample. If you had analyzed several samples, maybe the mean toxin content would be similar to the one found for the 17 samples of O. chrysurus.

L370-371: ‘this hybrid demonstrated the ability to seek and consume, or accumulate, higher levels of CTXs than its lower risk parental species’. Is the difference between 0.02 and 0.03 ppb really significantly different? You can’t conclude because you haven’t analyze enough hybrid samples. Maybe the value of 0.03 ppb would be the maximum levels found in hybrid samples?

Author Response

We have amended the title to clear up this confusion.

We have adjusted the language on L60 to include the word ‘caught’. This should indicate to the reader that the observed individuals were the same captured individuals. Since this study is the first to describe this type of hybrid the status of its frequency is unknown. However, informal communications with recreational fishers have indicated that they have seen or are now becoming aware of this particular hybrid (but have offered no samples). Other hybrids between snapper species are being caught in the territory as well.

L24: What do you mean by ‘composite cytotoxicity’?

Sometimes several ciguatoxin congeners are present in a specimen which can attribute to the toxicity of the fish. The N2a assay can’t distinguish between the different congeners hence it generates the results as total toxicity i.e. composite cytotoxicity.

L28: What about the comparison with the C-CTX1 levels found in the ‘prohibited’ parent?

Because this species is prohibited and not regularly sold it was not a target of our investigation, instead we relied on the regional government prohibitory actions and a history of poisoning to determine its status. CTX contaminated fish can be sporadic, within and among species, within and among time, and locations. This species has been implicated in a number of CFP outbreaks, therefore given the guidance level (with a built in 10x safety factor) that these fish have the potential to exceed this guidance level of 0.1 ppb CCTX1.

The method for identifying hybrids is novel and this is the first description of this method. It can be used to identify novel CFP vectors (other hybrid fish) previously undescribed for identification and thus eventually risk assessment.

L47-49: You could cite one or two references at the end of this sentence.

We have added two references to support this statement.

In addition to the Tosteson 1995, we have also included the suggested reference as well as an additional reference for a survey conducted in St. Thomas. The Tester rates in fig. 1 for St. Thomas do not include the most heavily affected islands for CFP (STT and STJ); St. Croix has a low incidence rate. However, that being said it still gives a nice overview of the region and includes many relevant studies and serves a value to the reader. It is difficult to obtain accurate incidence rate data, and most all underreport, so we’ve made references to these papers and the range of incidence rates reported.

L59-62: It could be interesting to provide pictures of the two parent and two hybrid fish, even if you place them in supplemental information.

We have added the photos of both parent species and the hybrid: now Figure 1.

L133: please define N2a since it is its first apparition in the text.

This was originally defined in the abstract but we have added the acronym definition here as well for clarity.

L139: two fish samples appears to be not enough. Furthermore, why don’t you have analyzed at the same time some fish samples of the two parent species?

The low sample size is insufficient for a full risk assessment; therefore this study serves as a proof of concept to address several questions. This study proves that these types of hybrids can contain CTXs and that they can exceed the levels so far observed in the ‘safe’ parental species. We have previously conducted a survey into the toxin content of the lower risk species that is commercially available and found no levels exceeding the guidance for C CTX 1. The higher risk species is restricted on the market and only selective fish from ‘safe areas’ are supplied with the intention that they do not contain CTXs. Therefore, we would be repeating an already completed study into the lower risk species, and if we tested more of the higher risk species we could not answer any novel questions, as there is high variability within species, among locations, and among time points.

L149 & 155: please specify the origin of the C-CTX1 standard used to calibrate your toxicological analysis.

We have added the C CTX1 standard information in section 2.1

L149-153: You have to indicate at least the LOD and LOQ of your assay. t would be appreciated to briefly described, as for the other sections, the N2a assay protocol (final volume in wells, concentrations of ouabain/veratridine, time of incubation, volume of samples in wells, dilution ranges, etc…).

We have added the LOD and LOQ of the assay and added additional information for the N2a assay protocol to section 2.2.

L155-171: What are the LOD and LOQ ?

We have added the following information to the section in question: ‘The limit of detection (estimated at LC30) for the assay using O. chrysurus was determined to be 0.002 ng Caribbean ciguatoxin-1 eq/g. The limit of quantification (where a full dose–response curve was evident without matrix interference) for spiked negative tissue controls 0.005 ng C-CTX-1 eq./g. The maximum negative sample tissue equivalent dose used for the hybrid was 870 mg TE/mL.’

L200: please specify the codes for the 2 hybrids samples (i.e. FDA321 and H2) in order to help the lector to find these two samples in the Figure 1.

We have added the suggested information. We have also included the museum voucher numbers for the samples in Figure 1, also shown in Figure 2.

We have adjusted the language to ‘confirm’ that the specimen is a hybrid.

The ‘medium risk species’ is restricted from sale at the market and has been linked to CFP events, which we have documented. A survey of fish from various regions to establish a range of toxicity was beyond the scope of this study. We limited the scope just to focus on the hybrid and to determine if it was indeed a hybrid, if it could accumulate or contain CTXs, and if so at what level? Our goal was to answer these questions. Conceivably if we set out to re-determine the risk posed by L. apodus, we would be facing a large regional and spatial study. Even with that evidence we could not change the status of the fish.

Due to the limited amount of CTX standard available, the CTX analysis is carried out in a two-tiered analysis, where the N2a assay estimates the concentration of the toxin while the LC-MS unambiguously confirms the presence of the biomarker (C-CTX-1). The amount of standard needed to perform the N2a assay is more that 4-fold lower in comparison the LC-MS/MS; therefore, the FDA is currently not doing quantification of CTX by LC-MS/MS until the generation of more standard. Furthermore, the FDA Guidance level for C-CTXs is reported as composite toxicity in C-CTX-1 equivalence. The LC-MS method for Caribbean ciguatoxin is set to analyze C-CTX-1 only.

L274-275: ‘the second hybrid fish (FDA321/USNM444947) contained no detectable levels of CTXs’. Again, what are the LOD of your N2a assay and LC-MS/MS analysis?

The LOD and LOQ for this assay and these matrices for the N2a have been added. It is known that the N2a assay is more sensitive than the LC-MS/MS. Currently, the FDA is in the processes of a SLV for C-CTX-1 by LC-MS/MS and the LOD and LOQ are being determined. In the present study, determining the LOD and LOQ for LC-MS/MS is out of the scope since the estimation of the toxins is performed by N2a.

We have changed the x axis and methods section to include this information.

Part 1: This question is answered in L273.Part 2: The injection volume was added to section 2.3. (‘and 10 uL injections were made using the LC’s autosampler’)

We have added an additional statement further clarifying this point. Hybrids in nature are very rare, and to come across them in a directed fishing effort is difficult. These hybrids were opportunistic samples, therefore a sustained effort to find additional hybrids for testing was not feasible. Currently from this data the acute risk for CFP form these hybrids is not supported, but it does provide evidence for their ability to accumulate toxins, and for a potential consumer to be exposed to CTXs through consumption.

A hybrid in a CFP endemic region has not been tested for CTXs (or any hybrid to our knowledge). The methods for determining a hybrid have not yet been reported. What was of interest to us and the fishermen was investigating the toxin content of a hybrid between species with long held toxin understandings.

We have adjusted the language of the sentence. This study serves as a proof of concept and provides methodology to address these questions when new hybrids from endemic regions present themselves.

That is true, but averages in toxin content from a large sample of fish are not a primary concern with respect to CFP outbreaks and poisoning events. Therefore our focus was on the maximum amount recorded for the low risk species and how it compared to what we observed in the hybrid. This information provides a new high mark for the species, serving as the maximum potential so far observed.

The difference between 0,02 and 0,03 ppb might not be significant, which is why we did not make a statement of a significance difference between the maximums observed between the two species. But the highest maximal observed for the lower risk species is lower than the current presentation for the highest maximal observed for the hybrid. Beyond the number to number comparison we note that this hybrid came from a micro region that has been shown to produce lower toxin individuals, and it is a smaller size, both contributing factors to a lower CTX content. If these hybrids are similar to the marketable fish reported in the referenced survey, we could expect higher toxin contents in larger individuals from a more toxic location.

Round 2

Reviewer 3 Report

Thank you for having taken into consideration most of my comments.

Other minor corrections/comments:

L3-4: Confirmation of hybridization and ciguatoxins accumulation.

L176: You should precise the final concentrations of ouabain and veratridine applied to neuroblastoma cells.

L317: There are no standard deviations for the EC50 values (and thus for the toxicity data given L305 and 309). Haven't you tested the C-CTX1 standard and the fish samples in triplicate?

L318: Is the starting dose of 434.8 mg TE/mL or 870 mg TE/mL (i.e. maximal concentration of ww fish tissue tested)?

L319: It would be more informative to indicate final concentrations of ouabain and veratridine.

L374-377: The same issue would be plausible for O. chrysurus since a low concentration of CTXs (i.e. 0.019 ng/g) has been detected in one sample.

L395: Yes, it is important to test some L. apodus samples to have a "positive control" and a toxicity reference to compare results obtained for the low toxic hybrid sample with levels of CTXs contained in its other parent considered as "medium risk of CFP".

Author Response

Dear reviewer,

We greatly appreciate your time and efforts. Your suggestions, comments, and insight have helped improve this manuscript.

Please find below a point by point response to your comments.

Kind regards

L3-4: Confirmation of hybridization and ciguatoxins accumulation.

We have added the suggested language.

L176: You should precise the final concentrations of ouabain and veratridine applied to neuroblastoma cells.

The concentrations are included in the referred to method (30), but for clarity we have added the amount here on L179.

L317: There are no standard deviations for the EC50 values (and thus for the toxicity data given L305 and 309). Haven't you tested the C-CTX1 standard and the fish samples in triplicate?

Each sample and CTX-1 standard is serially diluted (8×) and 10 ul from each dilution is added in triplicate (see standard error bars in figure 4a and b). The mean sample dose (avg of these three readings) where a 50% reduction in cell viability is recorded for each the standard and the test sample and then compared (Standard EC50 (avg of triplicate)/hybrid EC50 (avg of triplicate)). This calculation is discussed in detail in the reference for the method (30).

L318: Is the starting dose of 434.8 mg TE/mL or 870 mg TE/mL (i.e. maximal concentration of ww fish tissue tested)?

L316: ‘while the second hybrid fish (FDA321/USNM444947) contained no detectable levels of CTXs at the maximal concentration of wet weight fish tissue tested (870 mg TE/mL).’

870mgTE/mL was the maxium tested for the negative sample, to show that we tested beyond the positive fish into the range below the LOD of 0.001ppb before declaring the second hybrid ‘negative’. The 434,8mgTE/mL was the optimal dose amount used for the positive fish.

L319: It would be more informative to indicate final concentrations of ouabain and veratridine.

We have added this information on L179.

L374-377: The same issue would be plausible for O. chrysurus since a low concentration of CTXs (i.e. 0.019 ng/g) has been detected in one sample.

Yes, you are correct; and this issue was discussed in reference 11. From our experience investigating CTXs in endemic areas most organisms involved in the food web where CTXs are present usually can contain some detectable level of CTXs, given the right conditions. However, these levels are of concern for low dose exposure (a difficult problem epidemiologically). The guidance level of 0.1ppb for C-CTX-1 is set with a 10X safety factor with the intention that a person will not become ill when consuming a fish with this concentration.

The positive control used in this study is the C-CTX-1 standard, this in addition to molecular confirmation by LC-MSMS gives us confidence in our confirmation of the presence of C-CTX-1 in the hybrid sample. We used the C-CTX-1 standard in both the assay as well as by LC-MSMS to compare with the hybrid sample (figures 4 and 5). C-CTX-1 is what the guidance level is based on for consumer safety in the Caribbean, therefore it is the most relavent and appropriate control target to use. The threshold for samples implicated in outbreaks is in excess of 0.1 ppb. C-CTX-1 eq., a guidance level established by the FDA after analyzing numerous outbreak samples responsible for illnesses, as discussed in reference number 5. The ‘medium risk’ species has been implicated in 6,8% of recent outbreaks (between 2011-2016yr) in Guadeloupe and is restricted from commercial sale by the government due to its established risk profile as reported in reference numbers 6, 15, and 16. Therefore, given the evidence from hospital reports and government intervention we agree with their assessment and legislation enacted regulating the sale of the fish and that this is a medium risk fish for CFP; as it has been implicated in ciguatera fish poisoning events that resulted in hospitalization.