Evolutionary developmental biology studies of non-model organisms are often limited by the absence of a sequenced genome, and thus tend to focus on small numbers of candidate genes cloned individually using degenerate PCR. To facilitate larger-scale gene discovery and move beyond the candidate gene approach, we have used next-generation sequencing to generate transcriptomes from the ovaries and embryos of multiple emerging model arthropod organisms, and de novo assemble and annotate them. In each of the transcriptomes from the milkweed bug Oncopeltus fasciatus, the cricket Gryllus bimaculatus, the amphipod Parhyale hawaiensis and the spider Parasteatoda tepidariorum, we identified over 10,000 genes, including members of all conserved metazoan signaling pathways and genes implicated in many important developmental processes. We have created a searchable database called ASGARD to provide public access to these annotated transcriptomes, their associated raw data, and bioinformatic tools for analyzing and annotating such de novotranscriptomes.
We are also interested in improving the tools available for functional genetic analysis in non-traditional model organisms. Maternal RNA interference is frequently used in non-model insects to determine the function of genes in embryonic development. However, we find that while this technique is often effective at knocking down maternally supplied transcripts, the effects on zygotic transcripts can be highly variable. We have therefore optimized embryonic injection protocols that enable us to bypass maternal requirements by performing zygotic knockdowns, and to assess requirements for gene functions at different stages of embryogenesis.We have also collaborated with Prashant Sharma (American Museum of Natural History) to develop the first laboratory model for developmental studies in a daddy longlegs species, and have generated protocols for in situ hybridization and RNAi in this animal. Building on the recently reported success of transgenesis and targeted genome editing in Gryllus, we are currently working on CRISPR-based knockout and knock-in strategies, as well as tissue-specific reporter constructs, for use in the cricket. In collaboration with Quaid Morris (University of Toronto), we are working towards generating predictive bioinformatics tools to discover gene networks involved in developmental processes of interest. If successful, these techniques should be broadly applicable to a wide range of study organisms.