Probe Masking of Arabidopsis Microarrays to Detect Differential Gene Expression in Ovules of Sexual and Apomictic Boechera (Brassicaceae).

Microarrays are powerful tools for gene expression profiling, but their use is generally restricted to sequenced and well-known genomes.  Although microarrays have been used to detect gene expression in related species, reliable results have been limited to genes that exhibit sequence homology with most (preferably all) of the probes of the probe set of each respective gene.  For example, there are 11 probes per gene (probe set) on the Affymetrix <em>Arabidopsis</em> (ATH1) GeneChip^R.  Recently, various methods have been suggested to deal with polymorphisms between array probes and target organism DNA, thus increasing transcriptome coverage for cross-species array analyses.  One such method involves a probe-selection strategy that masks one or more probes per probe set, that are not homologous with the target species gene.&nbsp; We report herein the use of this strategy to profile gene expression during multiple stages of ovule development for sexual and apomictic species of <em>Boechera</em> (Brassicaceae).  Twenty ATH1 arrays were used to assess gene expression in ovules of three species &ndash; two apomicts and one sexual &ndash; at four stages of ovule development.  Probe intensities were masked at a range of thresholds (0 to 1000) using the Xspecies Perl script (Nottingham <em>Arabidopsis</em> Stock Center), and resulting custom Chip Description Files were used to test for differential gene expression at a BH-FDR &lt; 0.05.  qRT-PCR was used to evaluate relative expression levels of our probe masking approach.  Probe masking generally increased <em>i</em>) the detection of differentially expressed genes by up to five-fold, <em>ii</em>) the level of fold-change observed for previously detected differentially expressed genes, and <em>iii</em>) the number of overrepresented gene ontology categories that were enriched above that predicted by random distribution.  Results of this study are being used to determine candidate genes for the control of apomixis, and suggest a novel theory of apomixis control by epigenetic regulation.