A Researcher’s work is never easy but it is even harder when relative data are to be interpreted. This is especially true for Real-Time PCR. It is one of the most accurate ways to evaluate gene expression. However, despite it being such a powerful technique, it still carries many pitfalls which can lead a scientist to the wrong conclusion. Often a new user does not have thorough sample/RNA preparation, equipment or knowledge. So what are the considerations and aspects that the researcher should pay attention to?
Typically, raw data directly reflects biological events, but it is harder to explain. While normalized data can emphasize/exaggerate the specific interest points it also hides a lot of facts, which can lead to a wrong conclusion. My colleague Isobel Maciver wrote an excellent blog emphasizing the standardization guidelines of quantification principles of quantitative-real time PCR experiments (Bustin et al., 2009; Bustin et al., 2010). She ends her blog by raising awareness for the continuous evolution of guidelines and together with authors of the paper calls for an active input from researchers. This is true for any step described there but even more so for normalization. For example, normalization can simplify the data and minimize the manipulation variations. But at the same time, any form of normalization is data manipulation, which can lead you to a wrong conclusion either because you miss the true biological difference or you describe a non existent one.
Although the relative quantification data can be normalized in a many ways (Pfaffl, 2006), a comparison to a fully validated reference gene/s is the most common.
“It cannot be emphasized enough that the choice of housekeeping genes or lineage specific genes is critical….It is vital to develop universal, artificial, stable, internal standard materials, that can be added prior to the RNA preparation, to monitor the efficiency of RT as well as the kinetic PCR respectively. Usually more than one gene should be tested in a pair-wise correlation analysis and a summary reference gene index obtained …” Pfaffl (2006).
Indeed, Taylor et al., (2011) has shown that even when following the guidelines for standardization of qPCR experiments, it is rather a state of art to choose the right controls and interpret results. In his study of MCM7 in tumor versus normal tissue, Taylor found that normalization with commonly used GAPDH and 18S reference genes can lead to inconclusive and even opposite results. These two genes were changing their expression in opposite directions in tumor samples. An increase in expression of MCM7 (when normalized using a combination of HPRT1/TBP) gave stable values for both tissue types when used with good-quality samples.
This is only one example that shows we cannot offer a precize normalization suggestions without comprehensive understanding of what we are doing, because by doing so we are making conclusions that we should not be making. The key element is to know the details of experiment design and testing goals. Similarly, when researcher asks how to express/publish the data, there is no definite answer. Whether a researcher should present the data in percentage of control (of what control?) or present the raw data that is highly dependent on the research’s focus.
Taken altogether it is the researcher who will decide how to interpret data and who needs to be critical about the decision.
Bustin, S.A., et al. (2009). The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin Chem 55(4), 611 – 622.
Bustin, S.A. et al. (2010). MIQE précis: Practical implementation of minimum standard guidelines for fluorescence-based quantitative real-time PCR experiments. BMC Molecular Biology, 11, 74.
Pfaffl, M.W. (2006). Relative Quantification.– In: Real-time PCR (Hrsg. Dorak T.), International University Line, La Jolla, CA, USA , 63-82.
Taylor, S.A. (2011). MIQE Case Study — Effect of RNA Sample Quality and Reference Gene Stability on Gene Expression Data. Tech Note 6245