Genomic DNA was extracted using Qiagen Plant Mini Kit as per manufacturer guidelines
Label
Biotin/Steptavidin
Label protocol
The preparation of targets in all cases involved the double digestion of 0.5 μg of gDNA with AluI and HaeIII, and purification using Qiaquick® PCR Purification Kit (Qiagen Inc.). Biotin-11-dUTP was then incorporated into restriction digested gDNA fragments using the Biotin DecaLabel™ DNA Labeling Kit (Fermentas, ON, Canada) following the manufacturer’s guidelines. However, the incubation time was increased to 20 h, the reaction stopped with 1 µl 0.5M EDTA, pH 8.0 and labelled gDNA fragments were purified using Qiaquick® PCR Purification Kit.
Hybridization protocol
The SDA slides were pre-hybridized for 45 min at 42°C in a pre-warmed solution containing 5X standard saline citrate (SSC), 0.1% sodium dodecyl sulphate (SDS), 1% bovine serum albumin (BSA) and 25% formamide. The slides were rinsed twice with sterile MilliQ water and immediately dried with an air gun. The biotin-labelled targets (dried to 16 μL) were added to 17.5 μL of fresh 2X Hybridization buffer (250 μL of formamide, 250 μL of 10X SSC, 10 μL of 10% SDS), 0.5 μL of 5 μg/μL Human Cot1 DNA (Sigma-Aldrich, St Louis, MO), 0.5 μL of 10 mg/mL Poly A (Sigma-Aldrich) and 0.5 μL of 10 mg/mL salmon sperm DNA (Sigma-Aldrich). The mixture was denatured at 100°C for 2 min and immediately applied onto the array under a 22 × 25-mm lifter slip (Grale Scientific, Victoria, Australia). The slides were then placed in waterproof, humidified hybridization chambers (Corning Incorporated Life Sciences) and incubated overnight in a 42°C water bath. Following hybridization, the slides were washed twice for 5 min in 500 mL Wash buffer 1 (1X SSC with 0.1% SDS), once for 5 min in 500 mL Wash buffer 3 (0.1X SSC with 0.1% SDS), and once for 5 min in 500 mL Wash buffer 4 (0.1X SSC). Subsequently the slides were transferred to 500 mL of 6X SSPE-T buffer (0.9 M NaCl, 0.06 M NaH2PO4.H2O, 0.006 M EDTA, 0.005% Triton X-100, pH 7.4) without allowing them to dry. The biotinylated DNA targets bound on the array were then labelled with fluorescent FluoroLink™ streptavidin-labelled Cy3 dye (Amersham Pharmacia, UK) using a biotin–streptavidin system. Briefly, 200 μL of a Detection solution (0.5 µL of 0.8 µg/µL streptavidin-labelled Cy3, 0.8 µl of 25 µg/µL BSA, made to 200 µL with 6X SSPE-T) was applied directly onto the array surface and a 22 × 25-mm lifter slip was placed over it to evenly distribute the solution on the array. The slides were placed in hybridization chambers, wrapped in aluminum foil and incubated at 37oC for 1 h in the dark. Finally, the slides were washed thrice in 6X SSPE-T for 5 min and rinsed with sterile MilliQ water before being dried with an air gun. All hybridizations were performed with six technical replicates (corresponding to six sub-arrays) and two biological replicates, resulting in 12 data points for each array feature.
Scan protocol
Slides were scanned with a ScanArray Gx (PerkinElmer Life and Analytical Sciences, Downers Grove, IL) microarray scanner in conjunction with the supplied software. The slides were scanned with a resolution of 5 µm at 532 nm (Cy3, green laser) and at 55% photomultiplicator (PMT) gain whilst keeping background noise low. The scanned array was quantified using PerkinElmer ScanArray Express software v 2.0. The program individually quantified the signal intensity at each probe and normalized the data using the adaptive circle and LOWESS functions.
Data processing
Probes which did not hybridize were automatically flagged by the scanning software and labelled as ‘bad’. The values of all Bad features (signal to noise ratio (SNR) value of less than 5 in more than half of the technical replications) have been converted to Zero . Manual flagging was used to remove spots displaying inconsistent hybridization such as ‘donut’ spots. ‘Good’ probes were accepted as having a mean ‘signal to noise ratio’ (SNR) value of greater than 5 in more than half of the technical replications. The signal intensities and flag values of the two biological replicates were compared and average signal intensities were calculated for only those features that were flagged ‘Good’ in both the replicates. The values of features that had a ‘Bad’ flag in either or both the replicates were converted to zero. Data analysis included subtracting the background from median signal intensity for each feature, log2 transformation and combining technical replicates by taking average.
