RNA expression data from Halobacterium NRC-1 in response to oxic/anoxic transitions, initial experiment

To gain a comprehensive systems-level understanding of cellular phenotypes, it is critical to characterize the relationship between the dynamic transcriptome and proteome during environmental perturbations. Previous comparisons have shown a lack of correlation between mRNA and protein level measurements suggesting a predominant role for post-transcriptional regulation in mediating cellular environmental responses. To investigate the extent of post-transcriptional regulation, we have analyzed transcriptome and proteome level changes over a 13-hour 28-point time course during transitions between oxic and anoxic physiologies of Halobacterium. Integrated computational analyses of these data show that temporally shifting mRNA and protein profiles relative to one another significantly increases the mRNA/protein correlation. Although time lags for unrelated genes vary widely, we observe similar temporal lags between the transcription and translation of functionally related genes. In contrast, no significant temporal separation was observed within the transcript profiles. Taken together, these data suggest that while there is indeed a direct correlation between many corresponding changes at mRNA and protein levels, translational delay may be the predominant mechanism for the temporal regulation of protein abundance during physiological oxic/anoxic transitions in Halobacterium. The approach and algorithms delineated in this study provide a framework for incorporating the temporal dimension of information processing across many different layers of gene regulation. Keywords: time course Halobacterium sp. NRC-1 (ATCC700922) was routinely grown in complex medium (CM; 250g/L NaCl, 20g/L MgSO4.7H2O, 3g/L sodium citrate, 2g/L KCl, 10g/L peptone) at 37ºC under full-spectrum white light. For turbidostat experiments, starter cultures of NRC-1 were inoculated into 2L of CM in a 3.0 L vessel (5-10% inoculum) and grown to mid-logarithmic phase (OD600 ~ 0.5) in batch mode in a BioFlo100 modular bench top fermentor (New Brunswick Scientific, Edison, NJ) at 300 – 500 rpm, pH 7.0. Prior to each experiment, an oxygen sensor (model InPro 6000, Mettler Toledo, Columbus, OH) was calibrated to 100% oxygen at 1200rpm and sparging with 3.2 VVM of air. These conditions were approximately equivalent to oxygen saturation in CM medium, which is 1.6 mg/L (~5uM). Once the culture reached mid-logarithmic phase, the oxygen level was rapidly decreased to 0.8-4.8% within 10 minutes (achieved by turning off airflow and reducing agitation to 250 rpm), and allowed to equilibrate for 12 h prior to the start of sampling. The flow rate was zero during low oxygen conditions. The oxygen concentration in the culture was then rapidly increased to approximately 90-95% (achieved by agitating at 1200 rpm and sparging the medium with 3.2 VVM of air) and sampling commenced and continued at the times indicated in the Sample records. The culture was subsequently kept in active growth at a flow rate of approximately 2.25 ml/min (specific growth rate ~0.1/hr) for 6hr. Cultures were then rapidly shifted (within 10 minutes) back to low oxygen for 6hr, and then finally back to high oxygen for 30 min. During these perturbations, all other parameters were kept constant (pH 7.2-7.3, 37ºC, ambient light, O.D.600~0.5-6.5). Each culture sample was split in half, one half being used for RNA extraction and the other for protein preparation. Samples were removed from the turbidostat according to the times indicated in the titles of the Sample records.