Druyan, S., Department of Poultry Science, North Carolina State University, Raleigh, NC 27695-7608, United States, Institute of Animal Science, Agricultural Research Organization, Volcani Center, PO Box 6, Bet Dagan 50250, Israel
De Oliveira, J.E., Department of Poultry Science, North Carolina State University, Raleigh, NC 27695-7608, United States, Provimi Research and Innovation Centre, Lenneke Marelaan 2, Sint-Stevens Woluwe, B1932, Belgium
Ashwell, C.M., Department of Poultry Science, North Carolina State University, Raleigh, NC 27695-7608, United States

Recent studies using microarray technologies for the chicken have reported information regarding the effects of specific experimental treatments on gene expression levels often resulting in large gene lists and limitations on the statistical significance levels detected. In most cases, with these limitations, along with thresholds of ±2-fold differences in expression levels, that are used to create these gene lists, much of the biological information may have been overlooked. In this study, a focused 70-mer oligonucleotide microarray was developed to address the apparent limits of detection and issues with multiple testing resulting from the use of microarrays that include only a single spot (probe) for each gene. Gene expression was assayed across the development of the chicken embryonic heart from d 7 to 20 of incubation. When using a mixed-model approach and ANOVA with Bonferroni correction for multiple testing, including replicates within the focused array significantly increased the sensitivity with which differences could be detected across sample groups, as compared with single-spot data. By incorporating replication into the focused array, 50 genes were detected as being differentially expressed in the embryonic heart across the time points sampled. This compares with only 7 genes detected as being differentially expressed when a more typical, less statistically stringent single-spot analysis is conducted. Based on our observations, the use of focused microarrays allows for the thorough investigation of gene expression patterns, with detection of significant changes in gene expression of ±7%. This limit of detection is far superior to that of real-time PCR, which is able to detect significant changes in expression from ±33 to 55%, depending on the specific application. The ability to detect small differences in expression will allow investigators to identify subtle effects that have perhaps been overlooked in many prior assays, including single-spot arrays. Subtle shifts in gene expression are exactly those that occur during embryonic development, nutritional manipulation, and the initial stages of disease before clinical signs appear. ©2008 Poultry Science Association Inc.