Selection of normalizer genes in conducting relative gene expression analysis of embryos.

Abstract

In relative gene expression analysis, a reference gene for sample normalization is required for determining target expression changes among experimental treatment groups. Since some developmental toxicants secondarily cause general growth retardation and/or other general biological changes, commonly used housekeeping genes may not serve as accurate normalizers. We conducted real-time polymerase chain reaction (PCR) with normalization to calculate relative target transcriptional change, using housekeeping and structure-specific expression genes as normalizers. Relative levels of Hoxb1 expression were measured in cultured rodent embryos at 24 hr post retinoic acid (RA) administration. Transcriptional response was also evaluated using two novel compounds that produced posterior axial and growth defects in rat whole-embryo culture. Embryos treated with these compounds were evaluated for general biological processes, and their respective biological states were considered in the context of the relative gene expression change calculated with the housekeeping normalizers. Normalized RA-induced Hoxb1 expression demonstrated that only some reference genes accurately quantitated the expected 1.5- to 2-fold increase in Hoxb1 expression. Evaluation of the test compounds demonstrated that only normalization with the spatially-restricted hindbrain gene, Krox-20, calculated significant expression decreases of T-gene, a gene known to be functionally relevant in posterior axial development. Reduction in T-gene expression was confirmed qualitatively by whole-mount in situ hybridization. Prudent reference gene selection is important in evaluating relative gene expression in embryos. An experimental control design is proposed to facilitate the identification of normalizing genes that will accurately calculate relative gene expression change in treated embryos.