sample type: AX2 cells treated with E. Coli B/r as control. Total RNA isolated at 1h time point. strain: AX2
Extracted molecule
total RNA
Extraction protocol
RNA was isolated using TRIzol
Label
Cy5
Label protocol
Fairplay indirect labelling kit (Stratagene)
Hybridization protocol
Cy3 and Cy5 labelled targets were mixed, ethanol precipitated and dissolved in 65 µl of hybridisation buffer (Noegel et al., 1985) with 500 µg/ml Fish sperm DNA (Roche, Mannheim, Germany) and 2 µM Oligo dA 18-mer. The hybridisation mix was heated to 80 °C for 10 min, applied to the microarray under a cover-slip and incubated in a hybridisation chamber (Corning, New York, USA) for 15 hours at 37°C. Post-hybridisation washes were performed twice with 2x SSC, 0.1% SDS and once with 0.1x SSC, 0.1% SDS for 5 min each, five times with 0.1x SSC and once with with 0.01x SSC for 5 sec each and dried by centrifugation at 235xg for 5 min. See also Farbrother et al., Cell. Microbiol., 3:438-456, 2006.
Scan protocol
Signal detection was performed with the ScanArray 4000XL confocal laser scanner (PerkinElmer Life Sciences, Wellesley, USA). Images for Cy3 and Cy5 were obtained, spots were detected and quantified with ScanArray Express v2.2 (PerkinElmer Life Sciences), then manually inspected and if necessary corrected. See also Farbrother et al., Cell. Microbiol., 3:438-456, 2006.
Description
DNA microarrays cotainining 5,423 non-redundant ESTs from the D. discoideum cDNA project (Urushihara et al., 2004), partial sequences of 450 selected genes, probes of the SpotReport-10 Array Validation System (Stratagene, La Jolla, USA) and appropriate positive and negative controls were produced in-house using standard protocols (see also: http://www.uni-koeln.de/medfak/biochemie/transcriptomics/ production.e.shtml). Briefly, probes were amplified by two 100µl PCR reactions in 96-well microtiter plates using standard primers (M13F2: 5’-GTAAAACGACGGCCAGTG-3’, M13R2: 5’-C6-Aminolink-ACCATGATTACGCCAAGC-3’) for the cloned ESTs and gene-specific primers for the selected genes and the positive controls. After analysis by agarose gel electrophoresis (10µl of each reaction) the two reactions were combined, precipitated, solubilised in 11µl 25% DMSO, 1µl separated again by agarose gel electrophoresis and evaluated (Farbrother et al., 2002). Probes were transferred with the proteineer dp pipetting robot (Bruker, Germany) into 384-well microarray plates with cylindrical wells (Abgene, Hamburg, Germany) and printed onto UltraGaps slides (Corning, New York, USA) at 18°C and 45% humidity with a BioRobotics MicroGrid 600 microarray robot (Genomic Solutions, Huntingdon, UK) using MicroSpot 2500 microarraying pins. Each probe was spotted in duplicate, positive and negative controls were spotted 6, 24, 192 or 286 times and the SpotReport controls 96 times. The microarray is composed of three grids with 16 subarrays each and 14,620 spots in total. The complete microarray dataset is available at GEO, http://www.ncbi.nlm.nih.gov/geo; accession number GPL1972.
Data processing
For each individual comparison four microarrays were hybridised and analysed. Two imagepairs were produced per microarray slide, one with high laser intensity so that signals for most probes and also some saturated signals were obtained and a second one with lower laser intensity so that none of the signals was saturated. This way the dynamic range of the measurement was expanded. To handle the import and export of microarray data to different analysis programs an Excel Add-In, Array tools (http://www.uni-koeln.de/med-fak/biochemie/ transcriptomics/tools-array.e.shtml), was programmed in Visual Basic. Upon import of two data files of the same microarray scanned with different laser powers the saturated spots of the high laser power scan were replaced by non-saturated spots from the low laser power scan. In addition the import also performed data filtering by flagging SpotReport controls, negative controls, empty spots, spots where only spotting solution was printed and spots whose intensities were below or equal to zero as "Bad". Fluorescence ratios were normalised by LOWESS-norma-lisation using R1.6.2 (BioConductor, http://www.bioconductor.org/). The normalized M values (M = log2 (Intensityexperiment/Intensitycontrol) were transferred into a new worksheet for significance analysis of microarrays (SAM) analysis (Tusher et al., 2001). At the transfer all probes spotted in higher replicates were reduced to double spots by averaging. Furthermore all probes that were flagged "Found" in less than half of the spots, were excluded from SAM analysis. Differentially regulated genes were identified using the one-class SAM method calculating 1000 permutations. The SAM program not only identifies the differentially regulated genes, but also predicts the number of false positives. These are the genes that are falsely reported as differentially expressed. This feature was used in all microarray experiments to set the significance level such, that the 90th percentile of the false discovery rate (FDR) was minimal.
See also Farbrother et al., Cell. Microbiol., 3:438-456, 2006.
