Invasion history shapes host transcriptomic response to a body-snatching parasite

Abstract

By shuffling biogeographical distributions, biological invasions can both disrupt long-standing associations between hosts and parasites and establish new ones. This creates natural experiments with which to study the ecology and evolution of host-parasite interactions. In estuaries of the Gulf of Mexico, the white-fingered mud crab (Rhithropanopeus harrisii) is infected by a native parasitic barnacle, Loxothylacus panopaei (Rhizocephala), which manipulates host physiology and behaviour. In the 1960s, L. panopaei was introduced to the Chesapeake Bay and has since expanded along the southeastern Atlantic coast, while host populations in the northeast have so far been spared. We use this system to test the host's transcriptomic response to parasitic infection and investigate how this response varies with the parasite's invasion history, comparing populations representing (i) long-term sympatry between host and parasite, (ii) new associations where the parasite has invaded during the last 60 years and (iii) naïve hosts without prior exposure. A comparison of parasitized and control crabs revealed a core response, with widespread downregulation of transcripts involved in immunity and moulting. The transcriptional response differed between hosts from the parasite's native range and where it is absent, consistent with previous observations of increased susceptibility in populations lacking exposure to the parasite. Crabs from the parasite's introduced range, where prevalence is highest, displayed the most dissimilar response, possibly reflecting immune priming. These results provide molecular evidence for parasitic manipulation of host phenotype and the role of gene regulation in mediating host-parasite interactions.

Keywords: Rhizocephala; biological invasions; gene expression; host adaptation; host-parasite interactions; immune priming; parasitic manipulation.

FIGURE 1

Map of Loxothylacus panopaei invasion history and survey data from Tepolt, Darling, et al. (2020). Pie charts represent parasite prevalence in Rhithropanopeus harrisii at each site in summer 2015. *At AP, no parasitized R. harrisii were found, but the presence of ER clade L. panopaei was confirmed by infections in co-occurring Eurypanopeus depressus hosts

FIGURE 2

(a) Hierarchical clustering of Molecular Function GO terms significantly enriched among upregulated (red) and downregulated (blue) transcripts between infected and control crabs, as determined by a Mann–Whitney U test. (b) Curated subset of highly significant transcripts involved in innate immunity, ecdysis and neurotransmission, corresponding to GO term categories shown in (a). Error bars represent 95% confidence intervals. Note log₁₀ y-axis scale

FIGURE 3

Number of differentially expressed transcripts in response to Loxothylacus panopaei infection by host sex, across 1000 iterations of random downsampling. ***p << .001, numbers indicate Cohen’s d effect size

FIGURE 4

(a) Number of differentially expressed transcripts in response to L. panopaei infection according to parasite status in the host range, across 5000 iterations of random downsampling. *** = p << 0.001, numbers indicate Cohen’s d effect size. Some datapoints for the absent range not shown due to truncation of y-axis limit for display purposes. (b) Most significant transcript in likelihood ratio test of range:infection status interaction. Error bars represent 95% confidence intervals. Significant within-range t-test result indicated by *** (p < 0.001). (c) Composite response of the 848 transcripts exhibiting a significant range:infection status interaction with p < 0.01. Each line represents a single transcript, with y-axis indicating the mean expression level within each group relative to the overall mean, in units of standard deviation.