Systematic comparison of the fidelity of aRNA, mRNA and T-RNA on gene expression profiling using cDNA microarray

Abstract

In cDNA microarray technology, there are three main reverse transcription based RNA labeling methods, using total RNA (T-RNA), mRNA, and amplified antisense RNA (aRNA), respectively. However, despite the common use of the three types of RNAs, limited data are available regarding their differences and concordances. In this report, we compared the three methods through two sets of self-comparison experiments using the same RNA sample in all cases. Within each method, duplicate hybridizations are highly reproducible with low biases, which are randomly produced. When combining different RNAs within a single array, correlation coefficients between the two channels are rather low, while the discrepancies are persistent. Furthermore, the fidelity of aRNA and mRNA microarrays in the expression profile study shows no significant difference with standard T-RNA based labeling methods. These results suggest that some RNA abundance are selectively changed during aRNA amplification/mRNA purification processes, but it will not affect the gene expression ratio of the two samples if the same type RNA are used. Therefore all three types of RNAs can be used in expression profiling analysis as long as the test and reference samples are generated by identical method within single study.

Introduction

Glass cDNA microarray technology has resulted in fruitful applications in the biomedicine field since its introduction. Parallel qualification of large numbers of RNA transcripts using microarray technology promises to provide detailed insight into cellular processes involved in the regulation of gene expression. This should allow new understanding of signaling networks that operate in the cell and of the molecular basis and classification of diseases (Alon et al., 1999, Golub et al., 1999, Sgroi et al., 1999 Alizadeh et al., 2000, Perou et al., 2000, Ross et al., 2000, Dhanasekaran et al., 2001).

However, a main problem with gene expression profiling by microarray is that 10–100 μg of total RNA or 2–5 μg mRNA may be required to determine an expression profile, which makes analyzing of small biopsies extremely difficult, especially in the case of fine needle aspiration or laser capture microdissection samples. To widen the capability of microarray technology, it is crucial to develop effective and high fidelity RNA amplification system to detect gene expression with limited amount of sample inputs. Such protocols for cDNA microarray experiments amplify RNA either in an in vitro transcription based linear way (Van Gelder et al., 1990, Eberwine et al., 1992, Wang et al., 2000), or a PCR based exponential method (Iscove et al., 2002), or a combination of a linear and exponential amplification (Puskas et al., 2002b). The T7 RNA polymerase based linear amplification protocol is the most widely used among the available amplification techniques. Using this method, the starting amount of total RNA can be lowered to 2 ng–10 μg (Baugh et al., 2001), and a few investigators have reported that this method is reliable enough to detect differentially expressed genes with microarray (Ono et al., 2000, Pabon et al., 2001, Hu et al., 2002, Puskas et al., 2002a).

Despite the common use of total RNA (T-RNA), mRNA, and amplified antisense RNA (aRNA) in gene expression profiling analysis, there has been lack of systematic comparison of the differences and concordances among them. In this report, we compare the three methods by two sets of self-comparison experiments using the same RNA sample in all cases. Moreover, the fidelity and reliability of the three methods in gene expression profiling study is compared with the same pair of RNA samples.