Parasite manipulation of host phenotypes inferred from transcriptional analyses in a trematode-amphipod system

Abstract

Manipulation of host phenotypes by parasites is hypothesized to be an adaptive strategy enhancing parasite transmission across hosts and generations. Characterizing the molecular mechanisms of manipulation is important to advance our understanding of host-parasite coevolution. The trematode (Levinseniella byrdi) is known to alter the colour and behaviour of its amphipod host (Orchestia grillus) presumably increasing predation of amphipods which enhances trematode transmission through its life cycle. We sampled 24 infected and 24 uninfected amphipods from a salt marsh in Massachusetts to perform differential gene expression analysis. In addition, we constructed novel genomic tools for O. grillus including a de novo genome and transcriptome. We discovered that trematode infection results in upregulation of amphipod transcripts associated with pigmentation and detection of external stimuli, and downregulation of multiple amphipod transcripts implicated in invertebrate immune responses, such as vacuolar ATPase genes. We hypothesize that suppression of immune genes and the altered expression of genes associated with coloration and behaviour may allow the trematode to persist in the amphipod and engage in further biochemical manipulation that promotes transmission. The genomic tools and transcriptomic analyses reported provide new opportunities to discover how parasites alter diverse pathways underlying host phenotypic changes in natural populations.

Keywords: Orchestia grillus; amphipod; differential expression; ecological genomics; host-parasite co-evolution; infection response; parasite manipulation; population genetics; trematode.

Fig. 1.. Field sampling and bioinformatic analyses of the amphipod-trematode system.

A) Samples were collected from The Plum Island Estuary - Long Term Ecological Reserve in Ipswich, Massachusetts USA (red box; see Methods for details). B) The life cycle of the trematode L. byrdi through its three hosts: eggs are laid by avian hosts, ingested by the first intermediate host Hydrobiid snails, which pass the larval stages to the second intermediate host, the amphipod O. grillus, which are then preyed upon by avian hosts. C, D) Amphipods infected by L. byrdi change color from light grey or brown to orange and move into more exposed areas of the marsh substrate, which may increase rates of predation (B, C, D from (Johnson & Heard, 2017)). E) The experimental protocol for obtaining genomic and transcriptomic data, and the bioinformatic pipelines used in the data analyses.

Fig. 2.. PCA and differential expression.

A) Principal Components Analysis (PCA) built using expression data analyzed in DESeq2 for all transcripts detected in the 24 infected and 24 uninfected amphipods. There is no clear genome-wide transcriptional differentiation of infected (blue) an uninfected (red) amphipods. B) Volcano plot of differential expression with fold change on the x-axis and FDR-corrected significance on the y-axis. The sign of the values on the x-axis is based on a contrast of (infected – uninfected), thus down-regulated transcripts with negative fold change tend to occur in infected amphipods. C) PCA built using the top 100 differentially expressed transcripts between infected vs. uninfected amphipods, showing a clear differentiation across PC1 for these transcripts.

Fig. 3.. Differential expression of gene functional categories.

A) Venn diagram of the significantly differentially expressed transcripts from the three different analysis pipelines edgeR, EBseq, and DEseq2. There are 610 transcripts shared by at least two of the three pipelines, and 100’s of transcripts unique to each pipeline. B) Gene Ontology (GO) terms associated with significant up- or down-regulated transcripts among the 610 FDR-corrected differentially expressed genes shared by at least two of the three analysis pipelines. The GO categories identified using DESeq2 and edgeR are similar, while those from EBseq differ somewhat (see Table S2 for details). GO annotations were based on homology of amphipod transcripts with Drosophila melanogaster gene sequence based on BLAST values of < 10⁻¹⁰ (see methods). The notable categories are upregulation of ‘detection of external stimuli’, ‘eye pigmentation’ and ‘pigment biosynthetic pathway’, and down regulation of general metabolic processes and ‘ion membrane transport’. These are associated with phenotypic effects of infection in adult amphipods.

Fig. 4.. Infected amphipods show downregulation in V-type proton ATPases.

The heatmap plot shows down- or up-regulated transcripts across various subunits of V-type proton ATPases. Colors above columns indicate infection status. Colors on rows indicate different types of subunits, as noted in the legend at right. Parasite-induced down regulation of host V-ATPases has been demonstrated in other systems as playing a role in the inhibition of host immune responses against parasite infection (Robinson et al., 2012), implying a common role for these proteins in the mechanisms of host-parasite interactions.