Dissecting acid stress response in Rhizobium tropici CIAT 899

Rhizobium tropici CIAT899 is a nodule-forming α-proteobacterium displaying intrinsic resistance to several abiotic stress conditions such as low soil pH and high temperatures, which are common in tropical environments. It is a good competitor for Phaseolus vulgaris (common bean) nodule occupancy at low pH values, however little is known about the genetic or physiological basis of acid tolerance about gene expression under acidic conditions. To identify genes responding to pH stress we studied the transcriptomes of cells grown under different pH conditions. RNA was extracted from cells grown for several generations in minimal medium at 6.8 or 4.5 (adapted cells). In addition, we acid-shocked cells pre-grown at pH 6.8 for 45 minutes at pH 4.5. Transcriptomes were determined by RNA-Seq. From a total of 6289 protein-coding genes, 383 were found to be differentially expressed under acidic conditions versus control, among which 351 were induced and 32 repressed; only 11 genes were induced upon acid shock. The acid stress response of R. tropici CIAT899 is versatile: we found genes encoding response regulators and membrane transporters, but also enzymes involved in amino acid and carbohydrate metabolism and proton extrusion. Our findings enhance our understanding of the core genes that are important during the acid stress response in R. tropici. Bacterial strains and culture conditions. Rhizobium tropici CIAT899 was routinely grown in TY medium (Beringer, 1974) or the minimal medium described by Kingsley and Bohlool (1992), adjusted to pH 6.8 (MM- buffered to pH 6.8 with 20 mM N-(2-Hydroxyethyl)piperazine-N′-(2-ethanesulfonic acid) Hepes)) or to pH 4.5 (MAM- buffered to pH 4.5 with 25 mM Homopiperazine-N,N’-bis-2[ethanesulfonic] acid [Homopipes]; Research organics, Cleveland, OH, U.S.A.) at 30° C. MM and MAM were solidified with 0.8% gelrite (Carl Roth GmbH, Karlsruhe, Germany). Antibiotics were added at the following final concentrations (µg/mL): kanamycin (Km) 50; carbenicillin (Cb) 100; tetracycline (Tc) 10 and nalidixic acid (Nal) 20, if not indicated otherwise. RNA isolation. R. tropici CIAT899 was pre-grown in TY medium, then cultured in MM (control pH 6.8) (sample AR) or MAM (pH 4.5) (sample BR) until an OD60 of 0.6-0.7. For acid shock treatment cells were cultured in MM medium to mid-log phase, then washed with MAM and incubated another 45 minutes in MAM (pH 4.5) (sample CR). Cultures were harvested by centrifugation and the pellets were immediately frozen in liquid nitrogen. Total RNA was prepared using the RNeasy mini kit (QIAGEN, Hildesheim, Germany) with some modifications. Pellets were resuspended in RLT buffer, with lysozyme (20 mg per mL; QIAGEN, Hildesheim, Germany) containing Zirconium Oxide Beads (0.5 mm, Next Advance). Cells were disrupted using the Bullet Blender Tissue Homogenizer (Next-Advance) in impact-resistant 2 mL tubes. Genomic DNA was digested with RNase free DNase (QIAGEN, Hildesheim, Germany) for 20 minutes at room temperature. Final RNA concentrations were determined using a NanoDrop (Thermo Scientific). The typical OD260 to OD280 ratio of RNA samples was approximately 2.0. The integrity of RNA samples was verified using a TapeStation 2200 instrument (Agilent Technologies) and the RNA integrity number (RINe) was determined. RNA-seq and data analysis. RNA-Seq libraries were prepared using the TruSeq RNA sample Prep kit (Illumina) and sequenced (150 nt per read) by HiSeq 2500 instrument (Illumina) at the Beijing Genomics Institute (BGI, China). For the analysis of RNA-Seq data, Bowtie2 was used to align raw reads to the R. tropici CIAT 899 genome (Genbank entry U00096.2) and samtools was used to obtain BAM files. Differentially expressed genes (DEGs) were obtained via NOISeq 2.14.1 Bioconductor package using local fit and betaPrior parameter set to False. NOISeq implements differential expression analysis based on the Negative Binomial distribution. A false discovery rate (FDR) threshold of 0.95 was set for DEG calling. Sample clustering and principal component analyses were performed upon variance stabilizing transformation of expression data (NOISeq package). Transcripts were called as differentially expressed when the FDR-Log2FC adjusted p values were below 0.05 and fold-changes over 2 (Tarasona et al., 2011, Tarasona et al., 2015).