Total RNA Quality Control An aliquot of RNA was centrifuged and the pellet was washed with 80% ethanol. Following centrifugation, the supernatant was carefully aspirated, the residual ethanol was removed by brief drying (37oC) and the pellet was suspended in RNase free water. Total RNA was quantified with the use of a NanoDrop-1000 spectrophotometer. In addition, the integrity of the total RNA was analyzed (Agilent 2100 Bioanalyzer). Enrichment of mRNA Ten micrograms of total RNA was converted to enriched mRNA using Applied Biosystem’s (Ambion) MicrobExpressTm Kit (AM1905). The manufacturer’s protocol was followed. Enriched mRNA was analyzed by the Agilent 2100 Bioanalyzer. The expected yield of enriched mRNA from ten micrograms of total RNA is typically 1.0 - 2.5?g. Production of cRNA Enriched mRNA (200ng) was converted to cRNA using Applied Biosystem’s (Ambion) MessageAmp™ II-Bacteria RNA Amplification Kit (AM1790). Amino-allyl-UTP was incorporated into the cRNA during the IVT reaction (aa-cRNA). The manufacturer’s protocol was followed. The expected yield from 200ng is about 100?g. Fluorescent Labeling of cRNA Alexa Fluor 555 (Invitrogen A32756) was coupled to cRNA following the manufacturer’s instructions. Unincorporated dye was removed using RNeasy Mini Columns (Qiagen 74104). The Qiagen’s quick clean-up protocol was followed. cRNA Fragmentation cRNA coupled to fluorescent dye was fragmented using heat and Zn++ cations. The expected mean fragment length is 100 -125 nucleotides. Fragmented cRNA is considered sufficiently fragmented for use on MYcroarrays if the mean fragment length less than 200 nucleotides. (Parameters: Amount of nucleic acid labeled = 10, Amplification = RNA polymerases, Mass unit = Micro gram)
Bacteria were grown in LB broth after a 1:100 seeding from overnight cultures. The bacteria were allowed to propagate (shaking at 37C) until the OD600 reading of 0.6 prior to RNA extraction. (Parameters: time unit = hours, temperature unit = C, media = LB Broth)
Fold change for Test vs Control hybs was calculated, and log trasnformed.
Normalization To adjust for differences in sample behavior from array to array, a scale factor was created to equalize signal across all arrays. Only probes that met the following criteria were used to calculate the scale factor: only Kp spots were used; spots must have less than 10% saturated pixels; spots could not be manually flagged “bad”, median signal had to be more than 1.5 fold background; and all 6 arrays had to have the spot qualify as a normalization spot. There were 16,200 qualifying spots. The average signal for these qualifying spots, called normalization features (NF), was calculated for each array. A scale factor (SF) for each array was calculated as follows:  SF = ? (median pixel signal of NF)array i / ? (median pixel signal of NF)Brightest array An adjusted F532 median signal for each probe was calculated by multiplying the background adjusted median F532 signal for each feature by the scale factor for each array (see equation 3). Scaling is a minimalistic normalization approach. There are numerous normalization techniques for single color microarray analysis. MYcroarray provides a tab delimited raw data file output (.gpr) files so that if the customer desires a more fancy or sophisticated normalization approach, the raw data can be used to for this. The final algorithm to calculate scaled and background corrected signal was:  Adjusted Signal = (Max[median signal – bkg, 0.1] x SF) + C Where bkg (background) was the 2nd percentile darkest spot in the distribution of Kp spots, SF was the scale factor (see discussion above and equation 2) and C was a constant (C = 25). The constant shifts the entire background corrected and scaled signal distribution to the right, which helps avoid confoundment of low signal values with residual noise in the system. Trimmed Mean. A trimmed mean was used to generate the final signal value for the 4 identical probe replicates that survey each gene on each array. The data used for the trimmed mean calculation was the adjusted median pixel signal. The trimmed mean was calculated by discarding the maximum and minimum signal of each series of 4 probe replicates and averaging the remaining four signal values. If a gene had less than four “good” probe replicates, than the average of the remaining spots was used to estimate signal for each probe series. Present Call. A present call is an estimation of whether or not the transcript for genei was detected. The present call has three components. First component is at the probe level. The raw median pixel signal for each probe surveying genei must be at least 1.5 fold greater than the global background signal for a probe-level present call. Second component is at the probe series level. For genei, at least 3 out of 4 probe replicates must have a probe level present call. Finally, a gene level decision is made within each condition where 2 out of 3 biologic replicates had to have a probe series level present call for a gene to be considered present.
Single color Hybridizations. The customer sent 3 biological replicates of each experimental condition; control (C) and test (P). Fifteen micrograms of each fluorescent target was hybridized separately to one array (n=6). The experimental design is shown in Table 1. Hybridization was performed for 22hrs at 50oC in a proprietary hybridization buffer. Washing and Scanning. The arrays were washed as follows. Our wash stock solution is 20X SSPE (3M NaCl, 20mM EDTA, 118.2mM NaH2PO4 and 81.8mM Na2HPO4). Each array was removed from the hybridization cassette in while submerged in 1X SSPE (24C), transferred to fresh 1X SSPE (24C) and stored briefly while the rest of the arrays were liberated from their cassettes (estimated storage time was <2min). All arrays were washed at the same time. Wash# [SSPE] Temp Time 1 1X 24C 3min 2 1X 24C 3min 3 1X 50C 5min 4 1X 24C 3min 5 0.5X 24C 30sec All washes were performed in baths while the wash solution was gently circulated by stir bar. Following the final wash, arrays were spun dry in a microarray centrifuge. (Parameters: Chamber type = OTHER: Agilent Two-Gasket Slides , Quantity of label target used = 10, Mass unit = Micro gram, time = 20, Tiny time unit = hours, Volume = 255, Volume unit = Micro litre, temperature = 50)
Cell pellets were harvested at OD600 = 0.6 and RNA was extracted using the RNAeasy Extraction Kit (Qiagen, Hilden, Germany). All extractions were performed in biological triplicates. RNA was measured with Bioanalyzer 2100 (Agilent) to confirm RNA integrity level above 9. (Parameters: Extracted product = total_RNA, Amplification = none)