A representative double-tiled microarray.
Like fast food, researchers like their data super sized. Microarrays are limited in size and scope by the number of oligonucleotide probes (features) that can fit on a slide. Using several microarrays for one experiment is expensive and increases data variability. Now, Wheelan et al. double the number of features in a microarray by stacking oligonucleotides on top of each other in a recent article in Nature Methods.
The authors called their stacked feature arrays 'double-tiled' microarrays. Each 60-nucleotide oligonucleotide (60-mer) feature contained two unrelated 30-mers joined end to end. The 'inner' 30-mer was closest and the 'outer' 30-mer was farthest from the slide.
To cover the entire Saccharomyces cerevisiae genome, the authors selected 80,897 30-mers spaced 123 nucleotides apart. The authors arranged the 30-mers in sequence order from left to right (horizontally) on the slide. This series of 30-mers became the inner series. Once it was complete, the remaining 30-mers became the outer series and were placed in sequence order from top to bottom (vertically) on the slide. Therefore, each hybridized gene lit up a line of horizontal or vertical features. In all, the yeast microarray contained 44,000 60-mer features.
The array included positive control 30-mers in both the inner and outer series complementary to the retrotransposon Ty1. The authors hybridized the double-tiled microarray with labeled plasmid DNA containing yeast sequences, including Ty1, and found horizontal and vertical lines of fluorescence. The horizontal and vertical Ty1 series had similar signal intensities, suggesting that sequences hybridized to the inner and outer series with equal efficiency.
Two-color labeling differentiated two samples on one double-tiled array. The authors grew one yeast culture in galactose and labeled its RNA with the red fluorophore Cy5. They grew another yeast culture in glucose and labeled its RNA with the green fluorophore Cy3. Most horizontal and vertical lines on the resulting hybridized array were yellow, indicating that most genes were present in both galactose- and glucose-grown yeast. However, two red lines spanned features complementary to genes encoding galactokinase and galactose permease, which are both necessary for galactose metabolism.
The authors found little difference between double-tiled and conventional 'single-tiled' microarrays. They made single-tiled microarrays with 44,000 60-mer features and hybridized them with the labeled RNA from the previous experiment. Variation between replicate microarrays was similarly low for both double- and single-tiled microarrays. Rank-ordered lists of gene signal intensity were similar for the double- and single-tiled arrays, suggesting that they produced similar results despite using entirely different probes.
The authors believe that double-tiled microarrays will work as well for larger genomes as they did for yeast. To accommodate larger genomes, greater genomic space might be necessary between 30-mer oligonucleotides. Alternatively, triple-tiled microarrays, with outermost 30-mers arranged diagonally on the slide, might expand the data available from a single microarray even further.