Gene expression interrogation, both at single-gene (classical gene expression analysis) and genome-wide (global transcriptome analysis) levels, are prominent fields of study. By providing quantitative measures of mRNA changes, it has the extraordinary potential of identifying the functional consequences of genetic variability and environmental influence. Accurate quantification of gene expression is providing great insight into the physiology and metabolic complexity of microbes and their consortia and is thus contributing to our understanding of both fundamental and applied issues of major interest.
Microarray transcript profiling is the most widely used technique to evaluate global gene expression in microbial systems. However, due to methodological uncertainties inherent to the technique, it is imperative to validate the expression of key genes by an alternative procedure and quantitative real-time PCR (qPCR) has become the method of choice. qPCR is accurate, exhibits a broad dynamic range, and is sensitive and reproducible [1–6]. However, when performing qPCR, several parameters need to be controlled in order to obtain accurate and reliable expression measurements; these include variations in the amounts of starting material between samples, RNA extraction efficiency, RNA integrity/quality, efficiency of cDNA synthesis, and differences in the overall transcriptional activity of the cells analyzed .
The most frequently used strategy to control for such variations is relative normalization, where the expression of a target gene is measured with respect to total RNA, rRNA or a stably expressed internal reference gene. The use of rRNA or other reference gene has the advantage that their expression permits normalization against the cumulative errors of the entire process . Ideally, the reference gene should be universally valid, expressed stably and at a similar level across all samples, cells, experimental treatments, and designs. Unfortunately, no such reference gene has been identified [9–12] and even widely used control genes have proven unsuitable in certain situations [6, 13, 14].
Housekeeping genes are usually chosen as reference genes both in eukaryotes  and prokaryotes  because they are assumed: a) to be essential, b) to be ubiquitous, c) not to be regulated or influenced by the experimental procedure and d) to be expressed at similar levels in different types of cells. However, it is becoming increasingly clear that some commonly chosen housekeeping genes vary considerably across cell types, with time, or due to experimental treatment . This may be explained partially by the fact that some housekeeping proteins participate in other functions as well as in basal cell metabolism [17–19].
Therefore, for an accurate comparison of mRNA transcription in different samples, either validated reference genes are required for normalization or new ones should be determined empirically for each experimental system and condition studied [20, 21]. Here, we evaluate a carefully selected set of reference genes for improving the interpretation of gene expression profiles in a model microorganism, Acidithiobacillus ferrooxidans. The implementation of high-throughput microarray analyses of A. ferrooxidans [22, 23] and PCR techniques for expression profiling [24–37], have greatly enhanced our understanding of the genetic and physiological potential of this bioleaching bacterium. However, suitable reference genes to evaluate gene expression have not previously been identified in A. ferrooxidans.
In order to address this lacuna, we screened high-density oligonucleotide array-based expression profiles available for A. ferrooxidans and identified a set of nine genes with minimal variability under different experimental setups. Quantitative real-time PCR was then used to determine the mRNA levels of these genes, comparing their transcription in five different experimental conditions. Finally, we evaluated the suitability of these and other previously reported reference genes to monitor the expression of four selected target genes from A. ferrooxidans grown with different energy sources. This study defines reference genes for normalization of gene expression for future research, sparing other researchers in this and related fields from cumbersome and time-consuming screenings for an ideal reference gene, provided that they verify the stability of these candidates under their conditions of study.