RNA-Seq experiment

Cells were grown in 200 ml cultures in 1 litre flasks under white light as above until they reached mid to late log phase (20–40 million cells per ml). They were then transferred to constant monochromatic red, green, or blue light. After 72 h, cells were centrifuged at 5000 g, pooled into 1 ml cold phosphate-buffered saline, then spun again at maximum speed in a microfuge. The supernatant was discarded, and the cell pellet frozen in liquid nitrogen. For RNA extraction, 1 ml of TRIzol reagent was added to the frozen samples before thawing at room temperature. Two glass beads were added before shaking for 3 min using a Tissue Lyser (Qiagen) at maximum speed. Chloroform (200 μl) was added and mixed by shaking for 15 s before incubation at room temperature for 3 min and centrifugation at 12 000 g at 4 °C for 15 min. The aqueous phase was transferred to a fresh tube, combined with 0.5 ml isopropanol and 5 μl of glycogen (20 mg ml⁻¹), incubated for 10 min at room temperature, then spun down at 12 000 g at 4 °C for 15 min. The supernatant was removed, and the pellet washed twice with 1 ml 70% (v/v) ethanol before resuspension in 50 μl of RNase-free water. Samples were treated with RNase-free DNase (Sigma-Aldrich) according to the manufacturer’s instructions, before RNA purification using the Spectrum Plant Total RNA Kit (Sigma-Aldrich). RNA quality was verified using a Bioanalyser before preparation of RNA-Seq libraries using the Illumina Tru-Seq RNA library preparation kit, which enriches for mRNAs using Oligo-dT beads to capture polyA tails before cDNA synthesis. This method does not capture chloroplast RNA. One hundred base pair paired-end sequencing was carried out on an Illumina HiSeq at the Oxford Genomics Centre.

RNA-Seq data were analysed within the Cyverse Discovery Environment (https://www.cyverse.org). Contaminating adapter sequences and poor quality sequences were removed using Trimmomatic v 0.36.0 (Bolger et al., 2014). The Fastqc tool version 0.2 was then used to produce fastq files (Andrews, 2015), and TopHat version 2 to map reads to reference genomes using Bowtie 2 (Langmead and Salzberg, 2012; Kim et al., 2013). Samples that gave low numbers of reads were removed from further analyses. CuffDiff version 2.2.1a was used to calculate differential expression values for each gene in each pair of samples (Trapnell et al., 2013). Read counts were normalized by transcript length and by the total number of fragments, using the geometric fragments per kilobase of transcript per million fragments mapped (FPKM) method. Differentially expressed genes were identified based on corrected P-values (Q-values) and a false discovery rate (FDR) less than 0.05. This gave pairwise comparisons in the form of expression levels sorted by log2-fold change between each pair of light conditions. Gene expression patterns were visualized using the CummeRbund package (Goff et al., 2014), and heatmaps were refined using the Pheatmap package (Kolde, 2019). Reference genomes used were the ORCAE OTTH0595 v2 genome (Derelle et al., 2006; Palenik et al., 2007; Blanc-Mathieu et al., 2014) and the RCC809 v2 genome (Grigoriev et al., 2012). Reference genome sequence (fasta) files and annotation (gff3) files for each Ostreococcus ecotype were sourced from the Online Resource for Community Annotation of Eukaryotes (ORCAE) database (http://bioinformatics.psb.ugent.be/orcae/).