All strains were treated exactly the same to assess baseline transcriptional variation.
Growth protocol
Grown separately at 25C in defined media with glucose as sole C source. Grown to OD 0.400 (exponential phase) at 600nm for total RNA extraction.
Extracted molecule
total RNA
Extraction protocol
Total RNA was extracted using Ambions RiboPure bacterial RNA kit. mRNA was enriched using Ambions MICROBExpress kit.
Label
Alexa Flour-647
Label protocol
aRNA was synthesized using Ambions MessageAmp II kit. aRNA was indirectly labeled with Alexa Flour 555 and 647 dyes.
Hybridization protocol
Anti-sense RNA from strains OS223, OS195, OS185, and OS155 were used for the assay. Two different strains were simultaneously hybridized per slide in a competitive two-color scheme for each possible pairwise comparison excluding self. Each 60 µl hybridization reaction contained approximately 2 µg aRNA per strain, oppositely labeled with either Alexa-Flour 555 or Alexa-Flour 647 dyes. The aRNA solution was contained on the slide using 25 X 40 mm LifterSlips (Thermo Scientific, Waltham, MA). Slides were then statically incubated at 50⁰C for 18 hours in single-slide rubber hybridization chambers (CamLab, Cambridge, U.K.).
Scan protocol
Three biological replicates were used per strain comparison, allowing for 18 hybridizations to be performed, using a loop design format. In addition, dye swaps hybridizations were performed for all biological replicates for a total of 36 slide hybridizations. Slides were scanned at a 30-µm resolution using a Genepix 4200A scanner (Axon Instruments, Sunnyvale, CA).
Description
OS155 aRNA
Data processing
Overall expression was analyzed using the Genespring GX11 software (Agilent technologies, Santa Clara, CA), with foreground-background signal intensities imported separately per channel in a one color format. The data set including 36 slides, representing 72 samples, was normalized by applying a quantile algorithm, then Log2 transformed to a threshold of 1.0. Normalized signal intensities were filtered using an upper cutoff of 95% and lower cutoff of 10% per channel, keeping all entities falling within that range in at least half of the samples.
