RNA sequencing in Saccharomyces cerevisiae with abrogated binding of Nab3 to PIC2

Whole-transcriptome sequencing allowed us to compare transcriptomic differences between a Saccharomyces cerevisiae BY4741 parental strain (PIC2-GFP) and two derived mutants lacking Nab3 RNA-binding sites in PIC2. Previous experiments had established that disrupting Nab3 binding to PIC2 while maintaining Nrd1 binding to the same transcript led to severe growth defects, larger cell size, delayed cell cycle and resistance to oxidative stress. Following characterization of a PIC2-overexpressing mutant we established that these phenotypes were not related to Pic2 overabundance. Therefore, the purpose of this experiment was to determine (i) whether the transcriptomic signature of the mutant lacking Nab3 RNA-binding sites in PIC2 mutant could account for the observed phenotypes and (ii) if the transcripts that were differentially expressed in such mutant were targeted by Nab3 and Nrd1. Overall design: Three biological replicates of RNA-sequencing experiments performed in a parental Saccharomyces cerevisiae strain (PIC2-GFP) and two derived mutants: mut-Nab3-BS and mut-NNS-BS. While mut-Nab3-BS lacks Nab3 RNA-binding sites in PIC2, the other strain has no Nab3 or Nrd1 RNA-binding sites in PIC2 (mut-NNS-BS). Cells were grown in 25 mL-cultures of SC-Ura containing 1% (w/v) raffinose and harvested at by centrifugation (4000 rpm, 5 minutes) at an OD600 of 0.5. Pellets were then snap frozen by submersion in liquid nitrogen and stored at -80ºC. On the day of the experiments, cells were thawed on ice and denatured by the addition of 200 µL of guanidinium thiocyanate (GTC) acid phenol mix. Cell lysis was performed by adding 400 µL of glass beads to each sample and vortexing all tubes at full speed for five 1-minute intervals separated by four 1-minute incubations on ice. An additional 1.5 mL of GTC phenol mix was added to the samples before incubating them at 65ºC for 10 minutes. Samples were then placed on ice for another 10 min. 800 µL of sodium acetate mix (3.3 mL 3 M NaOAc pH5.2, 0.2 mL 0.5M EDTA pH 8, 1mL 1M Tris-HCl pH 8, water to 100 mL) and 1.5 mL of chloroform were pipetted into each extract before vortexing at maximum speed for 5 seconds. Lysates were centrifuged at 4ºC (10000 rpm, 30 minutes) and the resulting aqueous phase was transferred into a new tube containing an equivalent volume of Phenol:Chloroform:Isoamylalcohol mix. The mixtures were vortexed vigorously and centrifuged at 10000 rpm for 5 minutes. Once more, the aqueous phase of each sample was mixed with an equal volume of Chloroform:Isoamylalcohol and vortexed at maximum speed before centrifugation at 10000 rpm for 5 minutes. Finally, the aqueous phase of each sample was relocated to tubes containing 3 volumes of 96% ethanol. The mix was vortexed vigorously and preserved at -80°C for at least 30 minutes or at -20°C overnight. Following ethanol precipitation, samples were spun at 1000 rpm for 30 minutes at 4ºC. After discarding the resulting supernatants, 70% ethanol was pipetted into each tube. After centrifuging samples at 10000 rpm for 5 minutes at 4ºC, ethanol was removed, and pellets were left to air-dry for approximately 5 minutes before resuspension in DEPC water. Finally, RNA concentrations were measured in the Qubit™ 4 fluorometer (Thermo Fisher Scientific) using the Qubit™ RNA HS assay kit (Thermo Fisher Scientific). RNA extracts were treated with RQ1 RNase-Free DNase as per manufacturer's guidelines. The quality of the RNA samples was then inspected in an Agilent 2100 bioanalyzer (Agilent Technologies) using an RNA 6000 pico assay (Agilent Technologies) according to the manufacturer's instructions. Samples with RNA integrity number (RIN) scores greater than 7 were submitted to Novogene for paired-end sequencing of 150 bp-reads in a NovaSeq 6000 system (Illumina). Before sequencing, samples underwent a rRNA depletion step and libraries were generated using the TruSeq preparation protocol (Illumina). Raw sequencing outputs were returned as FASTQ files. Hence, after conversion into FASTA format, files were processed using the paired-end (PE) version of the pyCRAC pipeline (Webb et al., 2014). This package demultiplexes barcoded reads and removes adapter sequences applying the previously developed Flexbar tool (Dodt et al., 2012). The resulting reads undergo two quality control steps. Firstly, sequences are filtered based on their length and nucleotide quality values (i.e., PHRED scores). Secondly, since PCR artifacts will display identical random prefixes, they can easily be pinpointed and excluded from downstream analysis (collapsed). The remaining unique reads were subsequently aligned to the corresponding regions of the Saccharomyces_cerevisiae.R64-1-1.75 reference genome with the Novoalign tool (version 2.07). The pyReadCounters function of the pyCRAC package (version 1.5.1; Webb et al., 2014) was used to compute the number of reads that map to genomic features. The resulting tables for all the samples were merged into a single TXT file that was analysed by the DESeq2 R package (Love et al., 2014). The pipeline corrects for size variability by computing a scaling factor for each sample. To obtain such normalisation factor, the software calculates the ratio of the raw count of a given gene in one of the samples to the geometric mean of the counts for that same gene across the compared samples. The median of all the ratios calculated within a sample is defined as the scaling factor for that replicate. Having adjusted the raw reads, the package then performs a logarithmic transformation of the data and calculates a fold change and an FDR for each entry. Statistical significance was assessed for entries which exceeded or were lower than log2 fold change o 1 and -1 respectively and had an FDR value of 0.05 or less.