Neuronal gene expression in two generations of the marine parasitic worm, Cryptocotyle lingua

Trematodes, or flukes, undergo intricate anatomical and behavioral transformations during their life cycle, yet the functional changes in their nervous system remain poorly understood. We investigated the molecular basis of nervous system function in Cryptocotyle lingua, a species of relevance for fisheries. Transcriptomic analysis revealed a streamlined molecular toolkit with the absence of key signaling pathways and ion channels. Notably, we observed the loss of nitric oxide synthase across the Platyhelminthes. Furthermore, we identified upregulated neuronal genes in dispersal larvae, including those involved in aminergic pathways, synaptic vesicle trafficking, TRPA channels, and surprisingly nitric oxide receptors. Using neuronal markers and in situ hybridization, we hypothesized their functional relevance to larval adaptations and host-finding strategies. Additionally, employing a behavior quantification toolkit, we assessed cercaria motility, facilitating further investigations into the behavior and physiology of parasitic flatworms. This study enhances our understanding of trematode neurobiology and provides insights for targeted antiparasitic strategies.

In the manuscript by Tolstenkov the authors describe the molecular workings of the nervous system in Cryptocotyle lingua.The findings show increased levels of certain neuronal genes in the dispersal larvae, particularly those linked to synaptic vesicle trafficking, TRPA channels, and nitric oxide receptors.The authors suggest that the findings have functional relevance to larval adaptations and host-finding strategies.The results and the related datasets shared by the authors improve knowledge on trematode neurobiology and will be of interest for the community.
Comments to the authors: 1) EdgeR is expected to yield better performance and greater statistical power compared to a t-test, so I do not understand why the authors decided to pursue this approach.Moreover, it remains unclear which analysis in the manuscript employs this test.
2) Authors should be more careful when extrapolating the absence of transcripts in the de novo assembly to the absence of the gene in the organism, as this method has biological and technical limitations to reveal all genes present in a species.
3) Specify the parameters used to classify a gene as coding or non-coding.4) How were mobile elements identified? 5) Line 88 change Fig 2D and E by Fig 1D and E 6) Line 117 -The deduction concerning homeobox genes cannot be inferred from Fig 2B .Given that the comprehensive list of differentially expressed genes isn't displayed, such a conclusion cannot be drawn from the provided data.Kindly consider adding a supplementary table that encompasses all DE genes along with their corresponding statistics.7) Line 150 -When you mention conducting differential expression analysis on specific genes, does this imply re-running the EdgeR analysis using raw counts solely for these genes, or does it refer to extracting the results only for this gene family from the overall table?8) Line 169 -looking at Fig 3B in some categories the percentage of regulated genes is more than 50% so this statement should be clarified.9) Line 181 -the is no figure 3C  Using transcriptional analysis and in situ hybridization, the investigators have made a comprehensive accounting for differences in molecular expression between two distinct developmental stages of Cryptocotyle lingua.There is a strong argument that we need to broaden the number of model systems with simple nervous systems for future study.In addition, there is a large gap in our understanding of how drastic life cycle changes are controlled at the genetic / molecular level as demonstrated in C. lingua.The paper for the most part presents the results clearly and I believe it is a comprehensive and important contribution in the establishment of C. lingua as a model system.
The results are technically sound, but I have some comments on the brief "behavioural" component of the paper.The abstract states that "Additionally, employing a behavior quantification toolkit, we assessed cercaria motility, facilitating further investigations in the behavior and physiology of parasitic flatworms."However the actual results on behavior are reported only as a few panels in the first figure and some panels show just a single trajectory from I assume a single animal.Thus the interpretation of the results is more subjective than objective, given the limited information we have from the panels and the methods sections.This needs to be addressed.Should a statement important enough to be included in the abstract be supported with such a limited amount of data?
Even though there is a reference, the "complexity" measure should be defined in the material and methods.Fig 1 IV.Using "min" and "max" and an arbitrary heat map, for complexity makes it impossible to judge the amount of variation in the behavior during the activity "bursts."It would be more clear if we could see the distribution of complexity measures.FIg 1 V.Only 3 bursts from a single path is shown here.How has this path been selected?Is this representative?What about other trajectories from other samples?
In the manuscript by Tolstenkov the authors describe the molecular workings of the nervous system in Cryptocotyle lingua.The findings show increased levels of certain neuronal genes in the dispersal larvae, particularly those linked to synaptic vesicle trafficking, TRPA channels, and nitric oxide receptors.The authors suggest that the findings have functional relevance to larval adaptations and host-finding strategies.The results and the related datasets shared by the authors improve knowledge on trematode neurobiology and will be of interest for the community.
Response: We appreciate the reviewer's positive feedback and acknowledgment of our work's contribution to flatworm neuroscience.Indeed, we believe that the comprehensive dataset on neuronal genes we have generated is of significant value for broader applications in the field.

Comments to the authors:
1) EdgeR is expected to yield better performance and greater statistical power compared to a t-test, so I do not understand why the authors decided to pursue this approach.Moreover, it remains unclear which analysis in the manuscript employs this test.
We acknowledge the reviewer's concern regarding the use of the T-test.In response, we expanded our EdgeR analysis to not only encompass the general characterization of the transcriptome (as originally done) but also a targeted analysis of the specific genes described.This subset comprises 74 homeoboxes and 226 characterized neuronal genes, detailed in the updated supplementary table 1-9.While this revised analysis refined our list of DEGs, the overall conclusions of the paper remain unchanged.We've incorporated these changes in the text and updated Figure 3, as well as Supplement figures 2 and 3 (Please note that we added a Supplementary figure 1 in response to the comments of the Reviewer 2 that caused a shift of Supplementary figures numbers).

Revisions Made:
Line 444 (Please note that the line numbers are corresponded to the revised manuscript): We've clarified our approach in Methods: "To ensure that biologically significant differences were not overlooked, we conducted a detailed EdgeR analysis specifically on a subset of 300 verified neuronal genes, which includes the mentioned homeobox domain-containing proteins (see Supplementary Table 1).For this analysis, genes with an FDR <= 0.001 and FC> [1] were considered as differentially expressed."Please note that adding a Supplementary Table 1 shifted the Supplement Table numbers.
Line 98: Additional details have been provided for context: "The primary biological function of the redia stage is the production of dispersal larvae, the cercariae.Accordingly, of the 74 development regulatory Homeobox genes, only six were upregulated in the cercariae.In contrast, 27 Homeobox genes showed specific upregulation during the redia stage.This suggests a potential homeobox signature for C. lingua embryonic development (refer to Supplementary Table 1)."Furthermore, to align with these revisions, information on differentially expressed neuronal genes in the Result section, all heatmap figures in the Supplementary materials have been updated to reflect the newly refined estimates of DEG significance.
In Figure 3 we updated the proportion of upregulated genes.Supplement Figure 2. Supplement Figure 3.
Please also see updated Supplementary Tables 1 -9.
2) Authors should be more careful when extrapolating the absence of transcripts in the de novo assembly to the absence of the gene in the organism, as this method has biological and technical limitations to reveal all genes present in a species.
We completely agree with your point.Not all genes can be expressed at the developmental stages we studied.Therefore, all cases of expected but missing genes were confirmed by blast searches in the genomes of the other 25 trematode species (WormBase Parasite).This was stated directly in the Results for the case of

and has now been clarified in Methods (line 472). "To validate the absence of anticipated genes, we conducted BLASTp and tBLASTn searches within the genomes of other Platyhelminthes and Nematoda species on WormBase Parasite v. WBPS14, contingent on the specific context (be it species, class, or phylum)."
3) Specify the parameters used to classify a gene as coding or non-coding.

Thank you for highlighting this aspect. As previously outlined in the Methods section (line 449), we've provided additional clarity on our methodology:
"The coding sequences were identified using the TransDecoder tool (default criteria) with either BLASTp (Swissprot) and PFAM hits as ORF retention criteria.Transcripts encoding peptides shorter than 100 amino acids were excluded from the analysis."

4) How were mobile elements identified?
We appreciate the attention to this detail.Indeed, the prevalence of expressed transposons and other mobile elements is often overlooked in protostomian transcriptome studies.Although this topic is beyond the primary focus of our current paper, we felt it was necessary to address it briefly.We are in the process of conducting a detailed analysis of mobile elements within the C. lingua transcriptome and aim to present our findings in a subsequent publication.
To identify these elements in the current study, we relied on several criteria.These included the presence of Pfam domains characteristic of various families of reverse transcriptases, transposases, and integrases.Furthermore, a significant similarity based on blastp searches (with a threshold of 1.0E<10) to prototype transposons (e.g., PiggyBac,Tigger,Pogo) and retrovirus-like hits was also used to classify certain proteins as derivatives of transposable elements.
We have now added clarity in the manuscript (line 93): "Our C. lingua transcriptome analysis revealed a high number of mobile elements such as LINE-1, PiggyBac, Tigger, Pogo transposable elements, and gag-pol polyproteins.However, for the scope of this study, these elements were excluded from our primary analyses."Given that the comprehensive list of differentially expressed genes isn't displayed, such a conclusion cannot be drawn from the provided data.Kindly consider adding a supplementary table that encompasses all DE genes along with their corresponding statistics.

Thank you for the constructive feedback. We recognize the importance of providing comprehensive data to support our conclusions regarding homeobox genes. In response, we have incorporated a new table detailing Homeobox domain-containing proteins as Supplementary table 1. This table, along with others, now includes results from the specific
EdgeR analysis accompanied by the relevant statistics.Please refer to our response to comment 1 for a detailed explanation and the associated text corrections.
7) Line 150 -When you mention conducting differential expression analysis on specific genes, does this imply re-running the EdgeR analysis using raw counts solely for these genes, or does it refer to extracting the results only for this gene family from the overall table ?For the 300 genes specifically characterized in the paper, we implied re-running the EdgeR analysis using raw counts solely for these genes.Please refer to our response to comment 1 for a detailed explanation and the associated text corrections.
8) Line 169 -looking at Fig 3B in some categories the percentage of regulated genes is more than 50% so this statement should be clarified.
Corrected consistent with the new DEG results after specific subset of genes EdgeR analysis (see also updated Figure 3B and Supplemental Table 2).Corrected (complete list of the enzymes can be found in the Supplemental information).
10) What statistical measure does the e-value presented in the supplementary tables correspond to?Is it the blast e-value?This is blastp e-value.Now it is clarified in all supplementary tables.11) Supplementary Figure 5 -comments in column H are not in English.Table 6 (former 5).12) Line 225 -Supplementary Figure 5 does not show expression levels.

Corrected in the Supplement
We could not identify Supplementary Figure 5. Regarding the Supplement Tables and the Supplement Figures 1 and 2, they were updated.Reviewer #2 (Remarks to the Author): Using transcriptional analysis and in situ hybridization, the investigators have made a comprehensive accounting for differences in molecular expression between two distinct developmental stages of Cryptocotyle lingua.There is a strong argument that we need to broaden the number of model systems with simple nervous systems for future study.In addition, there is a large gap in our understanding of how drastic life cycle changes are controlled at the genetic / molecular level as demonstrated in C. lingua.The paper for the most part presents the results clearly and I believe it is a comprehensive and important contribution in the establishment of C. lingua as a model system.
The results are technically sound, but I have some comments on the brief "behavioural" component of the paper.The abstract states that "Additionally, employing a behavior quantification toolkit, we assessed cercaria motility, facilitating further investigations in the behavior and physiology of parasitic flatworms."However the actual results on behavior are reported only as a few panels in the first figure and some panels show just a single trajectory from I assume a single animal.Thus the interpretation of the results is more subjective than objective, given the limited information we have from the panels and the methods sections.This needs to be addressed.Should a statement important enough to be included in the abstract be supported with such a limited amount of data?
We deeply appreciate the reviewer's constructive feedback and positive remarks on the our efforts towards establishing Cryptocotyle lingua as a model system and presenting a few toolkits for the community.
We recognize the presented behavioral data might have been too brief, not fully representing its significance as indicated in the abstract.This was a result of our aim for conciseness of the text which we agreed was not too successful.
To balance this inequality we have introduced Supplementary Figure 1, which encompasses the distribution of speed, complexity, activity timings, and extra trajectories for the sampled animals.
From the batch derived from several infected mollusks, 10 random individual larvae were recorded, with one trajectory being unsuccessful.Our primary goal was to illustrate the toolkit's capabilities rather than provide an exhaustive behavioral analysis, which was not the main scope of the manuscript, explaining the limited sample size.
(probably this refers to Fig 3A III) 10) What statistical measure does the e-value presented in the supplementary tables correspond to?Is it the blast e-value?11) Supplementary Figure 5 -comments in column H are not in English.12) Line 225 -Supplementary Figure 5 does not show expression levels.13) Line 248 -change to S Figure 2B 14) Line 253 -change to S Figure 2C 15) Line 261 and 269 -change to S Figure 2D 16) "S." is a more standard abbreviation for the Schistosoma genus than "Sch."Reviewer #2 (Remarks to the Author): 5) Line 88 change Fig 2D and E by Fig 1D and E Corrected.6) Line 117 -The deduction concerning homeobox genes cannot be inferred from Fig 2B.
9) Line 181 -the is no figure 3C (probably this refers to Fig 3A III) 13) Line 248 -change to S Figure2BAddressed.14) Line 253 -change to S Figure 2C Corrected.15) Line 261 and 269 -change to S Figure 2D Addressed.16) "S." is a more standard abbreviation for the Schistosoma genus than "Sch."Addressed.