Comparative analysis of mite genomes reveals positive selection for diet adaptation

Diet is a powerful evolutionary force for species adaptation and diversification. Acari is one of the most abundant clades of Arachnida, exhibiting diverse dietary types, while the underlying genetic adaptive mechanisms are not fully understood. Based on comparative analyses of 15 Acari genomes, we found genetic bases for three specialized diets. Herbivores experienced stronger selection pressure than other groups; the olfactory genes and gene families involving metabolizing toxins showed strong adaptive signals. Genes and gene families related to anticoagulation, detoxification, and haemoglobin digestion were found to be under strong selection pressure or significantly expanded in the blood-feeding species. Lipid metabolism genes have a faster evolutionary rate and been subjected to greater selection pressures in fat-feeding species; one positively selected site in the fatty-acid amide hydrolases 2 gene was identified. Our research provides a new perspective for the evolution of Acari and offers potential target loci for novel pesticide development.

Line 100. There was a long-time debate for the Acari is diphyletic or monophyletic. It would be good to discuss this based on figure 1.
Line 166 and 367. Authors should provide the source of RNAseq data for the four spider mite populations.
Line 167,168 and 172. How were the top 200 genes found based on Sixteen samples from four population of spider mite? Please the authors provide the method and the gene list. Which method did authors used for GO and KEGG enrichment for the expanded genes? Authors should clarify in the method part.
Line 174. How were the top 50 genes found based on Sixteen samples from four population of spider mite? Please the authors provide the method and the gene list.
Line 211. Please provide high resolution Supplementary Fig. 4. It is not readable. Authors should label the bootstrap numbers on the tree, and also provide gene alignment file and the tree construction method in the method part.
Line 332. For the KEGG and GO enrichment analysis, please authors explain why they selected the far related Drosophila melanogaster as background set. The Kobas used in this enrichment analysis supports a wide range of species, including mite and tick models species, such as Ixodes scapularis and Tetranychus urticae. The insects are innately very different from the mites and ticks in some pathways. Please authors explain how the gene enriched with the D. melanogaster gene sets could reflect the mite and tick biology? What is the p-value or q-value cutoff for Kobas result?
Reviewer #3 (Remarks to the Author): In this study, the authors compared genomes of 16 Arachnida species, and showed different patterns of diet adaptations. I have a few concerns: 1. The manuscript was mainly based on published genome data. Expression data of spider mites were also used. It is not clear why only expression data of herbivory species were used. It seems the authors are choosing data arbitrary. I think they should provide more detailed and scientific explanations about data selection. 2. The authors showed that Arachnida species with different diet showed different evolutionary patterns based on some regular analyses. These results are expectable. It is not clear what is the key scientific problem the authors want to solve based on their analyses. What is the purpose of doing these analyses as well as publishing a paper, and how these results will serve the purpose? In the abstract and the conclusions, the authors mentioned twice: "These different genetic bases provide a new perspective for the study of the evolution and diversification of this group, and offers potential drug targets for pest control." These statements are repetitive, and are very vague. What is the "new perspective" and how to offer potential drug targets? Details should be discussed at least. In addition, not all mite species analyzed in this study are pests. For example, Metaseiulus occidentalis is a predatory mite natural enemy. 3. There are many typos and formatting errors in manuscript (including the references). I think the authors should pay more attentions to fix these problems.

Response letter COMMSBIO-20-3585
Genomic implications in diet evolution: Comparative analysis of mite genomes reveals positive selection for diet adaptation Dear Reviewers, Many thanks for your important and helpful suggestions for our manuscript entitled "Genomic implications in diet evolution: Comparative analysis of mite genomes reveals positive selection for diet adaptation" . In the following responses, we have carefully addressed all the issues, and we have revised our manuscript accordingly. Our references to line numbers refer to the no markup view that we have uploaded as a 'Acari_Diet_maintext_changes_tracked.docx' file. All changes have been accepted in the clean revised manuscript uploaded as a 'Acari_Diet_supplementary_maintext_clean.docx' file. We hope you find that we have adequately addressed all of the suggestions and that our manuscript is now suitable for publication. Please let us know if you have any further questions or suggestions.
Our point-by-point responses are as follows:

Reviewer #1 (Remarks to the Author):
This manuscript studied the natural selection patterns of mites with different feeding habits through comparative genomics method. They identified the positively selected genes and gene family expansion related with herbivorous, blood-feeding, and fat-feeding habits. These potential candidate genes provide insights into the dietary evolution of mites under different selection pressures. I have some comments for improving the manuscript.
Thank you for your recognition! We appreciate your positive comments and we made corresponding revisions to your suggestions.
Comment (1) Line 98: how many genes were used for phylogenetic tree construction A total of 17,910 homologous genes were identified. Due to the different techniques of sequencing and assembly of the 16 genomes, critical filtration has been carried out in the species selection (lines 285-291) and reannotation was conducted for 6 genomes. However, the discrepancy of the amount of gene information cannot be avoided and the number of single-copy homologous genes annotated is limited. Finally, 65 single-copy homologous genes with 48,831 nucleotides were annotated by all 16 species and applied to construct the species tree.
Thank you for your suggestion. We have corrected it in line 101. Gene CES2 (ID: OG0000064) is an outlier in blood-feeding group shown in Fig1D with a total of 1,128 amino acid sites after deleting aligned gaps. 285 positively selected sites were found, among which 197 sites with a posterior probability of over 95% were detected. We haven't found any errors in its result of sequence alignment (See following figure 1). Thank you for your suggestion! We have added this reference no.62 in the revised manuscript. Comment (12) Line 262: "any expansion gene families"??
We are sorry for the confusing expression. We have revised it to "no common expanded gene family was found in this group" (see line 151).
We are sorry for the confusing expression. We have revised it to "expanded gene families" in line 150.
Comment (14) Lines 274-275: What do your mean about "Different levels of genetic convergence under the same diet was identified"?
We are sorry for the confusing expression. We have deleted the sentence. Thanks for raising this issue! To find the significant pathways and gene ontology, we set the cutoff of P value < 0.05 in Fisher's exact test and FDR-corrected P < 0.1 according to the thresholds recommended in the previous studies (Liu, Yao-Zhong et al., 2017;Hulsegge, I et al., 2017). The KEGG pathway and Gene ontology related to detoxification function mentioned in the lines 165-181 (the following table 1) were considered statistically significant according to the thresholds. Comment (17) Figure 3D: the figure is unclear.
Sorry for the unclear version. We have split Figure 3 into new Figure 3 and 4 to show the information of figures better.

Reviewer #2 (Remarks to the Author):
This study provides the genetic evidence of mites and ticks adaptation to different dietary habits based on the comparative analysis of genomes of 16 arachnids. Authors analyzed mites and ticks of adaptations to Herbivory, blood-sucking and Fat-feeding by using the methods such as gene family expansion and gene selection pressure.
We appreciate your positive comments and we have made corresponding revisions to your suggestions.
The finding of this working is interesting, but I got some comments as below: Comment (1) Line 89. The genome assembles in Table 1 used in this analysis should be cited clearly and separately, so that the other researcher can find correct paper resources to repeat this work.
Thank you for your suggestion! We have supplemented separate citations for each genome in the revised manuscript (Table 1). Researchers can also obtain the genome data through the GenBank Assembly Accession ID in Table 1. Comment (4) Line 100. There was a long-time debate for the Acari is diphyletic or monophyletic.
It would be good to discuss this based on figure 1.
Thank you for your suggestion! The debate about the monophyly of Acari has always been a hot topic. A recent study indicated that Acari constitutes a monophyletic group (Showed in the figure below) nested within a monophyletic Arachnida based on transcriptomic data from 95 species.
(Jesus Lozano-Fernandez, et al., 2019). However, the study published at the same year recovered the major mite lineages by using ultraconserved genomic elements (UCEs) and found mites to be non-monophyletic (Van Dam, 2019). Our data was insufficient to address this problem, unless we add the groups that can represent all the spiders of Arachnida, such as whip spiders, llamshade spiders, mygalomorph spiders, hooded tickspiders, Sun Spiders, and so on. However, it is not the main focus of our study. It is a good idea for our follow-up research. melanogaster was set as background set, 60 genes could be obtained from KEGG pathway while 147 genes could be obtained from GO annotation. When I. scapularis was set as background set, 53 genes could be obtained from KEGG pathway while 110 genes could be obtained from GO annotation. When T. urticae was set as background set, 57 genes could be obtained from KEGG pathway while no information was obtained from GO annotation. Among these genes obtained from KEGG annotation, more than 90 percent genes of I. scapularis and of T. urticae were annotated in D. melanogaster. Among the genes obtained from GO annotation, 88 percent genes of I. scapularis were annotated in D. melanogaster. Hence, we chose D. melanogaster as the background set to generate more information. We set a threshold of P value < 0.05 in Fisher's exact test and FDR-corrected P < 0.1 for KOBAS result.

Reviewer #3 (Remarks to the Author):
In this study, the authors compared genomes of 16 Arachnida species, and showed different patterns of diet adaptations. I have a few concerns: Comment (1) The manuscript was mainly based on published genome data. Expression data of spider mites were also used. It is not clear why only expression data of herbivory species were used.
It seems the authors are choosing data arbitrary. I think they should provide more detailed and scientific explanations about data selection.
Thanks for raising this issue! Our study aimed to explore the dietary adaptions of mites based on released genomes. We have supplemented the details of data selection in the revised manuscript (lines 285-291). "(1) All Arachnida species in the database (before 2020.06) were searched as candidates; (2) Genome with the best completeness score was selected as the representative if there were two or more genomes for one species; (3) Species with genome completeness less than 80% were eliminated; (4) Gene prediction was conducted for the genomes lack of gene annotation information; (5) Contigs less than 1kb were excluded from the whole analysis". Finally, we chose one tick and fourteen mites, and the velvet spider as the background from 26 candidate species (see the following Table 1).
To see if the results of postive selection and gene expansion have been reflected in gene expression, we supplied transcriptome analysis for the different dietary groups. Sampling of mites is a big challenge, so we conducted database search for mite transcriptome of lipid-feeding, blood-feeding and plant-feeding species. The samples taken from whole blood or whole tissue transcriptome from natural individuals were included. After the database search, two samples of blood-feeding were insufficient to do biological repetition and excluded; four samples of lipid-feeding were included.
Only "Ribosome" was statistically significant with correction Fisher test P < 0.05 and FDR corrected P < 0.1 (see the following Table 2). Finally, we have shown details for gene expression of  Sorry for the confusing expression. We have changed the vague sentence "Based on comparative analyses of 15 Acari genomes, we found genetic bases for three specialized diets. Herbivores experienced stronger selection pressure than other groups; the olfactory genes and gene families involving metabolizing toxins showed strong adaptive signals. Genes and gene families related to anticoagulation, detoxification, and haemoglobin digestion were found to be under strong selection pressure or significantly expanded in the blood-feeding species. Lipid metabolism genes have a faster evolutionary rate and been subjected to greater selection pressures in fat-feeding species; one positively selected site in the fatty-acid amide hydrolases 2 gene was identified. Our research provides a new perspective for the evolution of Acari and offers potential target loci for novel pesticide development.". More information has been discussed in the specific dietary section in the revised manuscript.
Our study of evolutionary adaptation for different diets was not only based on pests but based on the mites with specialized dietary styles. However, the genetic adaptation of three specialized diets was implied and could consequently help with inhabitation of some pests such as spider mites, honey bee mites and so on. We could take efficient measures to weak the abilities for finding (olfaction), preparing (detoxification) and digesting (metabolism), which are found in the current study. For example, we have detected positive selection sites in HSP genes involving in olfactory pathways, which offered several drug targets for control the spider mites. Similarly, we have found that the evolutionary speed of lipid metabolism is significantly accelerated in honey bee mites, especially of arachidonic acid lipid metabolism pathway (Figure 6a and 6c). We could design drugs targeting at the candidate genes of arachidonic acid lipid metabolism such as FAAH2 to inhibit or kill honey bee mites in the following research.
Comment (3) There are many typos and formatting errors in manuscript (including the references).
I think the authors should pay more attentions to fix these problems.
Sorry for these mistakes. We have carefully modified our manuscript. And the revised manuscript was edited for proper English language, grammar, punctuation, spelling, and overall style by one or more of the highly qualified native English speaking editors at SNAS (verification code 30F9-7501-1FD4-89B3-2576).
Finally, thank you again for your great suggestions. We hope our manuscript can be of great interest to researchers of related fields now.