Expression profiling by high throughput sequencing
Beginning in 2013, sea stars throughout the Eastern North Pacific were decimated by wasting disease, also known as ‘asteroid idiopathic wasting syndrome’ (AIWS) due to its elusive etiology. The geographic extent and taxonomic scale of AIWS meant events leading up to the outbreak were heterogeneous, multifaceted, and oftentimes unobserved; progression from morbidity to death was rapid, leaving few tell-tale symptoms. Here we take a forensic genomic approach to discover candidate genes that may help explain sea star wasting syndrome. We report the first genome and annotation for P. ochraceus, along with differential gene expression (DGE) analyses in four size classes, three tissue types, and in symptomatic and asymptomatic individuals. We integrate nucleotide polymorphisms associated with survivors of the wasting disease outbreak, DGE associated with temperature treatments in P. ochraceus, and DGE associated with wasting in another asteroid Pycnopodia helianthoides. In P. ochraceus, we find DGE across all tissues, among size classes, and between asymptomatic and symptomatic individuals; the strongest wasting-associated DGE signal is in pyloric caecum. We also find previously identified outlier loci co-occur with differentially expressed genes. In cross-species comparisons of symptomatic and asymptomatic individuals, consistent responses distinguish genes associated with invertebrate innate immunity and chemical defense, consistent with context-dependent stress responses, defensive apoptosis, and tissue degradation. Our analyses thus highlight genomic constituents that may link suspected environmental drivers (elevated temperature) with intrinsic differences among individuals (age/size, alleles associated with susceptibility) that elicit organismal responses (e.g. coelomocyte proliferation) and manifest as sea star wasting mass mortality.
Tissue biopsies - of tube feet, pyloric caecum, and dermis - were taken from 17 individuals: 15 collected at McClures Beach and two from Lifeboat House in Point Reyes National Seashore (SI Appendix, Table S1) on the 19th and 21st August 2017, respectively, and preserved in RNAlater® in the field. A DGE analysis was first conducted to evaluate differential expression between tissue types. Subsequent analyses were conducted separately for each tissue type to assess differential expression between size classes and between symptomatic and asymptomatic individuals. Each size class (i.e. small, medium, large, or x-large) was compared to the mean of all other size classes. An additional analysis was conducted comparing just the small individuals directly to the x-large individuals to reveal potential DGE that might be masked by variation in intermediate size classes. Genes were filtered at a minimum threshold of mean 1 count per million (CPM). Significance was assessed at a false discovery rate (FDR) < 0.01 (unless otherwise noted).