Normalization Methods for qPCR

Real-time PCR has been used for gene expression analysis for over a decade.1,2 High quality reagents, techniques, and assay design have increased the accuracy of quantifying nucleic acids, making quantitative PCR (qPCR) a very powerful tool. Many factors can contribute to variability in qPCR, making the results difficult to reproduce between experiments. Not only can the quantity and quality of extracted RNA vary between samples, but the reversetranscription efficiency can also vary due to sample concentration, RNA integrity, the reagents used, and the presence of contaminants. Because RNA quantity can vary greatly between samples, accurate analysis requires normalization. This article describes two methods of sample
normalization for gene expression analysis: normalizing to input RNA and normalizing to reference genes.

Normalization to Input RNA

Normalization to input RNA is typically done by measuring the absorbance at 260 nm (A260), but cannot determine RNA integrity, because intact RNA and degraded RNA absorb light equally. Therefore, a secondary analysis, typically on a formaldehyde agarose gel, is required to assay for degradation. Microfluidic electrophoresis on the Experion™ system is faster and requires much less RNA for this assessment. The expression of four genes: annexin A3, CAPZB, cofilin, and destrin, was monitored in HeLa cells subjected to different treatments (Figure 1). Two sets of samples were transfected with small interfering RNAs (siRNAs, Act8 and Act9) targeting different regions of the b-actin gene. Control transfections included treatment with an siRNA targeting Green Fluorescent Protein (GFP) as a nonspecific control, transfection reagent, or buffer only. Normalizing to the amount of input RNA showed that annexin A3 expression increased approximately 2- to 5-fold depending on the treatment. In addition, CAPZB, cofilin, and destrin varied from slightly below baseline to a 50% increase in expression level.

INSERT Figure 1 here
Caption: Fig 1. Expression analysis of four genes normalized to initial amount of input RNA.

Normalization to Reference Genes

Because normalizing to input RNA amount cannot compensate for variations in reverse transcription efficiency, researchers often normalize to a reference gene. This accounts for variations in reverse-transcription efficiency because the reference gene is reverse-transcribed along with the gene of interest. Housekeeping genes such as b-actin, tubulin, GAPDH, and 18S ribosomal RNA are often chosen, with the assumption that their expression is constitutively high and that given treatments will have little effect on this expression. Reanalysis of the data in Figure 1 shows a 6.3-
fold increase in annexin A3 expression when normalized to 18S RNA (Figure 2: Act8), compared to a 4.9-fold increase when normalized to input RNA amount.

INSERT Figure 2 here
Caption: Fig 2. Expression analysis of four genes normalized to 18S ribosomal RNA.

While the ideal reference gene does not vary as a function of treatment or condition, it is often difficult to identify genes that meet this criterion.3,4 A more accurate strategy for normalization has been proposed by Vandesompele and colleagues.5 Instead of normalizing to a single reference gene, multiple genes that display minimal variation are selected. After determining the geometric mean of variation for these genes, the expression levels for gene(s) of interest are normalized to the geometric mean.

INSERT Figure 3 here
Caption: Fig 3. Expression analysis of four genes normalized to the geometric mean of three reference genes.

Figure 3 shows the expression levels of the four genes, calculated using the multiple reference gene normalization strategy. This analysis reveals that treatment with either siRNAs against b-actin increased annexin A3 expression levels about 5.5-fold,
and CABZB, cofilin, and destrin expression increased about 2-fold.

Summary

Proper normalization is essential for obtaining accurate gene expression data. There are many strategies for normalization, and with proper controls and replicates, all can be valid. The most comprehensive strategy uses a normalization factor calculated from the geometric mean of multiple reference genes. To simplify data analysis, iQ™5 and MyiQ™ real-time PCR detection systems come with analysis software that permits normalization to a standardized input amount, a single reference gene, or the geometric mean of multiple reference genes. Additionally, the software can take individual assay efficiencies into consideration, as well as combine multiple data sets to generate a complete gene expression study.

References

1. Heid CA et al., Real time quantitative PCR, Genome Res 6, 986–994 (1996)
2. Higuchi R et al., Kinetic PCR analysis: real-time monitoring of DNA amplification reactions, Biotechnology 11, 1026–1030 (1993)
3. Glare EM et al., b-Actin and GAPDH housekeeping gene expression in asthmatic airways is variable and not suitable for normalising mRNA levels, Thorax 57, 765–770 (2002)
4. Schmittgen TD and Zakrajsek BA, Effect of experimental treatment on housekeeping gene expression: validation by real-time, quantitative RT-PCR, J Biochem Biophys Methods 46, 69–81 (2000)
5. Vandesompele J et al., Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes, Genome Biol 3, RESEARCH0034, Epub Jun 18 (2002)
   

Bio-Rad’s real-time thermal cyclers are licensed real-time thermal cyclers under Applera’s United States Patent No. 6,814,934 B1 for use in research and for all other fields except the fields of human diagnostics and veterinary diagnostics.