Validation and application of a high fidelity mRNA linear amplification procedure for profiling gene expression

Abstract

The need for microgram quantities of RNA for microarray experiments has hindered application of this novel technology in cell types/tissue samples with limited abundance of RNA. In this study, potential application of T7-based linear RNA amplification was investigated for use in gene expression profiling experiments where starting material is limited. Yield and integrity of amplified antisense RNA (aaRNA), microarray hybridization intensities, and fidelity of differential gene expression detected were determined for arrays generated for unamplified versus amplified RNA from the same homogenous starting pools. Total RNA was extracted from bovine spleen and fetal ovary, serially diluted to concentrations ranging from 2 μg to 500 pg and amplified. Quality and quantity of total input RNA and aaRNA were assessed by spectrophotometry, gel electrophoresis and bioanalyzer. In experiment 1, we determined the optimal amounts of aaRNA generated from 20, 40, 200 ng and 2 μg input total RNA for use in cDNA synthesis, labeling and array hybridization that would yield robust and consistent hybridization signals on a bovine oocyte cDNA microarray. In experiment 2, comparison of microarray hybridization intensities and fidelity of differential gene expression between aaRNA generated from 2, 20 and 40 ng input total RNA versus unamplified RNA (uRNA) were conducted. The hybridization intensities for each of the 7000 spots per slide for microarrays conducted using aaRNA versus uRNA were highly correlated (2 ng = 0.84, 20 ng = 0.88, 40 ng = 0.90; P < 0.01). The false positive rate was low and similar (4.0% versus 4.4%) for arrays done with uRNA and aaRNA. Ninety-seven ESTs were detected as differentially expressed in the fetal ovary versus spleen at >1.5- or <0.5-fold using uRNA ( P < 0.05). However, the number of genes detected in arrays using aaRNA was approximately 1.5–2.5 times greater than with uRNA. Approximately, 65–70% of differentially expressed genes were common between uRNA and aaRNA arrays. Relative fold-expression (Cy3/Cy5 ratios) for 25 overlapping abundant genes was comparable for uRNA versus aaRNA arrays with 2 and 20 ng total RNA as input. Results demonstrate that T7-based linear amplification of small amounts of input RNA and use of aaRNA in microarray experiments retains fidelity of detection of differential gene expression that is relatively comparable to experiments done with uRNA and provides a potentially viable approach to facilitate gene expression profiling using limited amounts of starting material.