Polarella genomics: understanding the evolutionary transition to algal symbiosis and cold adaptation

Abstract

Dinoflagellates are one of the most ubiquitous and diverse groups of phytoplankton in the marine ecosystem, playing a crucial role in the nutrient cycle of the world oceans. Some lineages of dinoflagellates form symbiotic relationships with corals and coral reef animals, while other are free-living, including those that cause harmful algal blooms, or inhabit brine channels in polar sea ice. To better understand the molecular mechanisms that underpin the diversification of these species, an approach to genome-data generation that incorporates a diverse array of species is required. However, due to the immense size of free-living dinoflagellate nuclear genomes (up to 250 Gbp) little data has been generated. Their genomes also possess highly idiosyncratic features, e.g. non-canonical splice sites. In this thesis, I sequenced and analysed the genomes of two isolates of Polarella glacialis, a psychrophilic free-living dinoflagellate species that is sister to the symbiotic lineage Symbiodiniaceae (the coral reef symbionts). These genomes represent the first draft assemblies from free-living (and psychrophilic) dinoflagellates, and provide unparalleled insights into the genomes of these organisms.In Chapter 1, I establish the significance, hypotheses and aims of my thesis research. In Chapter 2, I present a detailed literature review, in which I set out the current state of dinoflagellate genomic research (including an overview of dinoflagellate genome data already published) and research into cold adaptation. I cover the current state of genome and transcriptome sequencing, including the challenges that must be considered when designing a sequencing experiment. I also cover the challenges facing the ab initio prediction of genes in non-model eukaryotic organisms, highlighting the methods previously applied to dinoflagellates.In Chapter 3 (the first research chapter), I detail the development and application of a comprehensive gene-prediction workflow tailored for the idiosyncratic features in dinoflagellate genomes. This workflow, incorporating both ab initio and evidence-based gene-prediction strategies, is the most comprehensive yet developed for dinoflagellates. I adopt this workflow in a comprehensive analysis of genomes from two Symbiodiniaceae species, one of which is the first isolate sequenced from the Great Barrier Reef. Through comparative genomics using two other Symbiodiniaceae genomes, my results reveal 2460 nuclear gene families showing evidence of positive selection, including genes involved in photosynthesis, transmembrane ion transport, synthesis and modification of amino acids and glycoproteins, and stress response. The results also reveal an extensive set of genes for meiosis and response to light stress.In Chapter 4 (the second research chapter), I present my analysis of all available dinoflagellate transcriptomes, in which I identify genes and functions that are conserved across all dinoflagellate lineages, or are lineage-specific. I demonstrate the abundance of dark genes (that have no annotated function) in dinoflagellates, and show that they are potentially a result of lineage-specific adaptation. I also identify distinctive sequence features (e.g. protein domains) enriched among cold-adapted dinoflagellate species. One such domain, DUF3494, has been extensively characterised as being ice-binding.In Chapter 5 (the third research chapter), I present the assembled and annotated genomes of the two P. glacialis isolates, generated using both short- and long-read sequence data. These genomes are diploid (~3 Gbp), and encode a large number of repetitive elements, with a high proportion of simple and complex repeats. They also showed a high proportion of tandem repeated single-exon genes, which appears to facilitate the expansion of gene families related to photosynthetic functions, particularly rhodopsin and chlorophyll-binding proteins. The two isolates share high sequence similarity but possess genomes of differing sizes, likely a result of uneven repeat expansion. I found the ice-binding protein domain DUF3494 identified in Chapter 4 to be encoded in tandem repeated single-exon blocks, similar to other functionally important gene families.This research represents the first comprehensive genome-scale analysis of any free-living psychrophilic dinoflagellate, and the first comparative genomic analysis within a single species. The data and established analytical workflows generated from this research represent a foundational reference for comparative analyses of dinoflagellate genomes, encompassing the coral reef symbionts of Symbiodiniaceae and bloom-forming species.