Abstract
Microarrays are widely used to study genome-wide gene expression changes in different conditions most notably disease, growth, or to investigate the effects of drugs on entire genomes. While the number and gene probe sequences to investigate individual gene expression changes differs amongst manufactures, the design for all of the probes is biased towards the 3' region. With the advent of exon arrays, transcripts of any known or predicted exon can be investigated to facilitate the study of genome-wide alternative splicing events. Thus, the use of exon arrays provides unprecedented opportunities in gene expression studies. However, it remains a major challenge to directly compare gene expression data derived from oligonucleotide to exon arrays. In the present study, genome-wide expression profiling of Laser Micro-dissected Pressure Catapulted (LMPC) samples of c-Raf mouse lung adenocarcinoma, dysplasia, unaltered transgenic and non-transgenic tissues was performed using the Affymetrix GeneChip Mouse Genome 430 2.0 Array and whole genome Mouse Exon 1.0 ST Array. Based on individual group comparisons 52 to 83% of regulated genes were similar in direction, but fold changes of regulated genes disagreed when data amongst the two platforms were compared. Furthermore, for 27 regulated genes opposite direction of gene expression was observed when the two platforms were compared pointing to the need to assess alternative splicing events at the 3' end. Taken collectively, exon arrays can be performed even with laser microdissected samples but fold change gene expression changes differ considerably between 3'IVT array and exon arrays with alternative splicing events contributing to apparent differences in gene expression changes.