Transcriptome-Wide Comparisons and Virulence Gene Polymorphisms of Host-Associated Genotypes of the Cnidarian Parasite Ceratonova shasta in Salmonids

Abstract

Ceratonova shasta is an important myxozoan pathogen affecting the health of salmonid fishes in the Pacific Northwest of North America. Ceratonova shasta exists as a complex of host-specific genotypes, some with low to moderate virulence, and one that causes a profound, lethal infection in susceptible hosts. High throughput sequencing methods are powerful tools for discovering the genetic basis of these host/virulence differences, but deep sequencing of myxozoans has been challenging due to extremely fast molecular evolution of this group, yielding strongly divergent sequences that are difficult to identify, and unavoidable host contamination. We designed and optimized different bioinformatic pipelines to address these challenges. We obtained a unique set of comprehensive, host-free myxozoan RNA-seq data from C. shasta genotypes of varying virulence from different salmonid hosts. Analyses of transcriptome-wide genetic distances and maximum likelihood multigene phylogenies elucidated the evolutionary relationship between lineages and demonstrated the limited resolution of the established Internal Transcribed Spacer marker for C. shasta genotype identification, as this marker fails to differentiate between biologically distinct genotype II lineages from coho salmon and rainbow trout. We further analyzed the data sets based on polymorphisms in two gene groups related to virulence: cell migration and proteolytic enzymes including their inhibitors. The developed single-nucleotide polymorphism-calling pipeline identified polymorphisms between genotypes and demonstrated that variations in both motility and protease genes were associated with different levels of virulence of C. shasta in its salmonid hosts. The prospective use of proteolytic enzymes as promising candidates for targeted interventions against myxozoans in aquaculture is discussed. We developed host-free transcriptomes of a myxozoan model organism from strains that exhibited different degrees of virulence, as a unique source of data that will foster functional gene analyses and serve as a base for the development of potential therapeutics for efficient control of these parasites.

Keywords: Myxozoa; SNPs; aquaculture; cell migration/motility; proteases.

Fig. 1

(A) Rainbow trout infected with Ceratonova shasta genotype IIR, showing swollen abdomen due to accumulation of ascitic fluid in the visceral cavity. (B) Ascitic fluid rich in C. shasta presporogonic, sporogonic, and spore stages.

Fig. 2

The workflow developed during this study for host contaminant filtration, assembly, and annotation. We filtered out host contamination at both the read level before assembly, using read mappings to reference genomes of the parasite Ceratonova shasta and host Oncorhynchus mykiss, and at the contig level after assembly, using BlastN against the same reference genomes. We produced two versions of the reference transcriptome: 1) C. shasta only, which is the more conservative assembly from only those reads that mapped to the parasite genome and 2) C. shasta + NHP, which was assembled from both the reads that mapped to the C. shasta genome and those that mapped to NHP (neither host nor parasite).

Fig. 3

Results of first stage filtering: the percentages of reads that best matched our draft Ceratonova shasta genome, rainbow trout genome, or NHP (neither host nor parasite) for each of the five C. shasta genotype transcriptome libraries.

Fig. 4

Results of second stage filtering: the percentages and size distributions of assembled contigs that best matched to the Ceratonova shasta genome, rainbow trout genome, or NHP (neither host nor parasite) for the two versions of the reference transcriptome: (A) C. shasta only and (B) C. shasta + NHP.

Fig. 5

Maximum likelihood phylogenetic trees. (A) Phylogenomic tree showing Ceratonova shasta genotypes relationships. Alignment based on 51/78 genes from Yahalomi et al. 2020. Polypodium hydriforme was set as the outgroup. (B, C) SNPs-based trees of transcriptomes of C. shasta genotypes using different subset of genes: (B) all genes (N = 22,755) and (C) only genes present across all ten transcriptomes pairwise comparisons (N = 593). Values at nodes indicate bootstrap support. LKR: Lower Klamath River; UKR: Upper Klamath River; WR: Willamette River.

Fig. 6

Relative conservation of motility and protease/inhibitor gene sets based on pairwise genetic difference comparisons and on SNPs-based phylogenetic analyses of Ceratonova shasta genotypes. (A) Chart shows relative frequency of conserved genes in pairwise comparisons between data sets, subtracting relative frequencies of all conserved genes. Negative values indicate that the subset of genes is less conserved than average. Positive values indicate that the genes are more conserved than average; (B, C) SNP-based ML trees of transcriptomes of C. shasta genotypes using strictly curated data sets of (B) cell migration genes (N = 164) and (C) proteases and inhibitor genes (N = 41). Values at nodes indicate bootstrap support. LKR: Lower Klamath River; UKR: Upper Klamath River; WR: Willamette River.