Normalizing RT-qPCR data: are we getting the right answers? An appraisal of normalization approaches and internal reference genes from a case study in the Finfish Lates calcarifer

Abstract

Commonly used normalization approaches for quantitative reverse transcription polymerase chain reaction analyses include (a) input nucleic acids standardization (ΔC q method), (b) normalizing target gene transcript abundance against a single internal reference gene (ΔΔC q method), and (c) geometric averaging of multiple reference gene abundance using the geNorm software. We compared these three approaches to examine expression of a negative muscle growth regulator gene, myostatin-I (mstn-I), in the finfish Lates calcarifer, following 4 weeks of nutritional fasting. Interestingly, these three different approaches led to widely divergent data interpretations. When ΔC q and subsequently ΔΔC q with α-tub as the reference gene were applied to mstn-I expression data, an ∼threefold upregulation of this gene was observed in fasted compared to fed fish. In contrast, mstn-I appeared unchanged when data was normalized against the geometric average of the two apparently most “stable” reference genes (elongation factor-1 α (ef1-α) and rpl8) selected from a panel comprising seven commonly utilized candidate reference genes (18S, cat-D, ef1-α, rpl8, gapdh, ubq, and α-tub). The geNorm software erroneously suggested that ef1-α, rpl8, and ubq were the most “stable” reference genes, whereas ΔC q analysis revealed these genes simply to exhibit similar patterns of regulation in response to fasting. The ΔC q approach showed that α-tub was the least variable in its expression level between fasted and fed fish after 4 weeks. The present study also highlights the importance of validating internal references for each time point under investigation when applying ΔΔC q and suggests that the more cost-effective ΔC q normalization approach, if carefully applied, may in fact produce the most biologically valid results.