Post-transcriptional regulation of redox homeostasis by the small RNA SHOxi in haloarchaea

Oxidative stress responsive small non-coding RNAs (sRNAs) have been reported in the model archaeon, Haloferax volcanii, but targets and mechanisms of actions have not been elucidated. While haloarchaea are highly resistant to oxidative stress, a comprehensive understanding of the mechanisms regulating this remarkable response is lacking. Here, using a combination of high throughput and reverse molecular genetic approaches we elucidated the functional role of the most up-regulated intergenic sRNA during oxidative stress in H. volcanii, aptly named Small RNA in Haloferax Oxidative Stress (SHOxi). We demonstrated that SHOxi is a functional non-coding RNA that plays gene regulatory roles in the oxidative stress response of an extremophilic archaeon. We found that SHOxi likely regulates redox homeostasis during oxidative stress by the post-transcriptional destabilization of malic enzyme mRNA. The decrease in the NAD+/NADH ratio resulting from the direct RNA-RNA interaction between SHOxi and its trans-target is instrumental in the survival of H. volcanii. The regulatory effects of SHOxi provides evidence that the fine-tuning of metabolic cofactors could be a core strategy to mitigate damage from oxidative stress and confer resistance. This study is the first to establish the regulatory effects of sRNAs on mRNAs during the oxidative stress response in Archaea. H. volcanii auxotrophic strain H53 (Δpyre2, ΔtrpA) and H98 (Δpyre2, ΔthyH) were used for all experiments. Culturing in liquid and solid media was done in rich medium (Hv-YPC) or selection medium (Hv-Cab), at 42°C and with shaking at 220 rpm (Amerix Gyromax 737) (Dyall-Smith, 2009). Uracil, tryptophan, thymidine, and hypoxanthine were added to a final concentration of 50 µg/mL, each. H. volcanii liquid cultures were exposed to H2O2 as previously described (Gelsinger & DiRuggiero, 2018b). In brief, cultures were grown in 160 mL of Hv-YPC or Hv-Cab under optimal conditions to an OD of 0.4 (mid exponential phase). 2 mM H2O2 was directly added to the cultures followed by an hour incubation at 42 °C with shaking at 220 rpm. Cultures were then rapidly cooled down, centrifuged at 5,000 x g for 5 minutes and the pellets resuspended in 18% sea water. The cell suspensions were then transferred to a 1 mL tube and centrifuged at 6,000 x g for 3 minutes, the pellets were flash frozen and stored at -80 °C until ready for RNA extraction. Total RNA was extracted using the Zymo Quick-RNA Miniprep kit with the following modifications: H. volcanii liquid culture is slimy and viscous thus to increase cellular lysis a 23 G needle and syringe were used to break down the cell pellet after addition of RNA lysis buffer to the frozen pellets to insure complete cell lysis. Total RNA was then extracted following the standard kit protocol. Total RNA was DNase I (NEB) treated (37 °C for 2 hours) as previously described (ref). Total RNA was then rRNA-depleted using the Ribo-zero Bacteria kit (Illumina). Strand-specific libraries were prepared using the SMART-seq Ultralow RNA input kit (Takara), insert sizes checked with the Bioanalyzer RNA pico kit (Agilent), and either paired-end sequenced (2 x 150 bp) or single-end sequenced (100 bp) on the Illumina HiSeq 2500 platform at the Johns Hopkins University Genetic Resources Core Facility (GRCF). We used a read count-based differential expression analysis to identify putative targets of SHOxi that were differentially expressed during oxidative stress and in ΔSHOxi. The program featureCounts was used to rapidly count reads that map to the NCBI H. volcanii annotation. featureCounts was run with strand-specific options on, paired-end mode on or off, multi-mapping off. The read counts were then used in the R differential expression software package DESeq2 (ref). Briefly, read counts were converted into a data matrix and normalized by sequencing depth and geometric mean. Differential expression was calculated by finding the difference in read counts between the SHOxi knockout oxidative stress state to the normalized read counts from the wild-type oxidative stress normalized read counts. The differentially expressed mRNAs were filtered based on the statistical parameter of False Discovery Rate (FDR) under 5%. In addition, only mRNAs with converse differential expression levels (FDR < 5%) in our previous wild type no challenge/oxi stress differential expression comparison (Gelsinger & DiRuggiero, 2018b) were labeled as specific putative targets of SHOxi.