0.5% isobutonal for B1,B2,B3,D1,D2,D3 was added at the beginning of cultivation.
Growth protocol
Glucose minimal medium (plus 0.5% isobutanol for B1,B2,B3,D1,D2,D3), 30°C with 150 to 200 rpm shaking.
Extracted molecule
total RNA
Extraction protocol
Total RNA was extracted from early log phase cultures using RNeasy kit (Qiagen). 100 μL RNAase free water was used for final elution; 90 μL aliquot was immediately ethanol precipitated and stored at -80°C. Then, mRNA was enriched from 10μg of total RNA using the MICROBExpressTM Bacterial mRNA Purification Kit (#1905, Ambion) by the removal of 16S and 23S ribosomal RNAs.
Label
Cy3
Label protocol
The enriched mRNA (200ng) was converted to cRNA containing aminoallyl-UTP with Message Amp II -Bacteria Prokaryotic RNA amplification kit (#1790, Ambion) following the manufacturer’s instructions. Amino-allyl modified cRNA (45μg) was coupled with amine reactive fluorescent dye (Alexa Fluor-555, #32756, Invitrogen) in a 10μl reaction following the manufacturer's instructions. After fluorescent dye coupling, unincorporated dye was removed with RNEasy Mini Columns (Qiagen) following the manufacturer’s instructions. Labeled cRNA was eluted from RNeasy Mini Columns with RNAse/DNAse free water. The extent of dye incorporation was determined by the Microarray function on a NanoDrop 1000 spectrophotometer (Thermo Scientific). The dye incorporation calculations were performed as described by Invitrogen/Molecular Probes for Alexa Fluor dye products. All samples had dye incorporation between 35 to 43 bases per dye. Fluorescent dye coupled cRNA (20μg) was fragmented by exposure to zinc sulphate (5mM final concentration in a 60μl reaction) at 75°C for 10min and the reaction was stopped by the addition of 500mM EDTA to a final concentration of 20mM. The extent of fragmentation was visualized with a 2100 Bioanalyzer (Agilent Technologies) using a RNA Nano Chip (Agilent Technologies, #5067-1511). Samples with a mean fragment size of 100-200 nucleotides were qualified for hybridization
Hybridization protocol
Each labeled and fragmented cRNA sample was hybridized individually to one custom 80K microarray by dynamic hybridization as follows. Twenty micrograms of sample was added to hybridization solution (600μl final volume) and incubated at 65°C for 5min and then placed on ice. Hybridization solution contained 6X SSPE (1M NaCl, 6.7mM EDTA, 40mM NaH2PO4 and 27.3mM Na2HPO4), 0.01 μg/μl acetylated BSA (#R3961, Promega), 0.01% Tween-20 (Sigma, #P9416), 10% deionized formamide (#P9037, Sigma). A large volume of hybridization solution master mix was prepared from which aliquots were removed to prepare each sample. Hybridizations were performed using a hybridization gasket slide (#G2534-60003, Agilent Technologies) and a Microarray Hybridization Chamber assembly (#G2534A, Agilent Technologies) following the manufacturer’s instructions except that 585μl of hybridization solution was used per gasket slide and both the gasket slide and microarray slide were preheated to 65°C prior to Hybridization Chamber assembly. The final Hybridization Chamber assembly was incubated at 50°C for 20hrs while rotating at ~5rpm (to assure free movement of the mixing bubble) in a hybridization oven (#G2545A, Agilent Technologies).
Scan protocol
Slides were scanned using an Axon 4000B scanner (Molecular Devices) at 5 micron resolution and 100% laser power. The PMT gain in the 532nm channel was adjusted to appreciate the full dynamic range (0 – 65,000) of signal intensity such that only a few pixels were saturated in a few spots. A signal intensity value for each probe on the array was extracted from the scanned image using Axon GenePixPro 6.1 software (version 6.1.0.4, Molecular Devices). Fixed diameter (35 microns) circular feature indicators were placed over the centre of each spot (probe) and median pixel intensity was calculated for each feature.
Description
mRNA enriched from total RNA by removal of 16S and 23S ribosomal RNAs. On the array, there are 12078 probes targeting 4280 CDSs of this strain.
Data processing
First, background fluorescence intensities of individual spots were subtracted directly from foreground intensities for background adjustment. Then variance-stabilizing normalization (vsn) and quantile methods without weight were applied sequentially for normalization with software R (2.11.0, “vsn” and “aroma.light” package). For each strain/isobutanol condition, spots with acceptable signals in at least two out of three biological replicates were chosen for further analysis, which resulted in 4-6 replicates for each probe. Two or three probes for each gene were included on the array. For simplicity, we chose the probe with the smallest overall variation across replicates (measured by the sum of standard deviations across 12-18 technical/biological replicates over all the four strain/isobutanol conditions) to represent the gene in further analysis. For each chosen probe, the median of the technical replicates on each array was calculated and the log2-transformed value was used to represent the expression level of the corresponding gene for each sample.
