Choosing the Best Reference Genes for qPCR
Why Reference Genes Matter
In RT-qPCR, the Ct value of your gene of interest is meaningless on its own. It depends on how much total RNA you loaded, how efficient the reverse transcription was, and how much cDNA you pipetted into the reaction. Reference genes (also called housekeeping genes or endogenous controls) solve this problem by providing an internal standard: a gene whose expression should be constant across all your experimental conditions.
When you calculate DCt (Ct of GOI minus Ct of reference), you cancel out these sample-to-sample loading differences. But this only works if the reference gene truly is stable. If it changes even slightly between conditions, those changes propagate directly into your fold change results — and they do so silently, without any obvious error message.
Common Reference Genes
The following genes are the most frequently used internal controls in RT-qPCR. Each has strengths and limitations depending on your experimental system.
GAPDH — Glyceraldehyde-3-phosphate dehydrogenase
One of the most commonly used reference genes. Generally stable in many cell lines and tissues, but can be affected by hypoxia, diabetes, and high glucose conditions. Not ideal for metabolic studies.
ACTB — Beta-actin
Widely used, but expression can change with cell proliferation, mechanical stress, and some drug treatments. Often varies in cancer vs. normal tissue comparisons.
18S rRNA — 18S ribosomal RNA
Extremely abundant, so Ct values are typically very low (8-12). Useful when other reference genes are too variable, but its high abundance means it may not reflect mRNA population changes accurately. Not suitable for oligo-dT-primed cDNA synthesis.
HPRT1 — Hypoxanthine phosphoribosyltransferase 1
A good alternative reference gene with moderate expression levels. Generally stable across many tissue types, though it can vary in some cancers. Particularly useful in neuroscience and immunology research.
B2M — Beta-2-microglobulin
Stable in many contexts, but expression is strongly affected by interferon signaling and immune activation. Avoid in studies involving inflammation, viral infection, or immunotherapy.
Tissue-Specific Considerations
There is no universally stable reference gene. A gene that works perfectly in HeLa cells may be completely unsuitable in primary hepatocytes or adipose tissue. GAPDH, for example, is one of the most commonly cited reference genes in the literature, but it is directly involved in glycolysis and its expression changes significantly under metabolic stress, hypoxic conditions, and in diabetes models.
Before committing to a reference gene for a new experimental system, always perform a pilot validation. Run your candidate reference genes across all your experimental conditions and check whether their Ct values remain constant (within 1 Ct). If they do not, try alternative candidates or use the geometric mean of multiple reference genes to improve stability.
How to Validate Reference Gene Stability
Several algorithms have been developed to objectively rank candidate reference genes by stability. The three most widely used are:
geNorm
Developed by Vandesompele et al. (2002), geNorm calculates a stability measure (M value) based on the average pairwise variation between a candidate gene and all other candidates. Genes with the lowest M value are the most stable. geNorm also determines the optimal number of reference genes to use. An M value below 0.5 is considered stable for homogeneous samples; below 1.0 for heterogeneous tissues.
NormFinder
NormFinder (Andersen et al., 2004) uses a model-based approach that accounts for both intra-group and inter-group variation. It can identify genes that are stable overall but differ between experimental groups — a situation geNorm can miss. It also suggests optimal gene combinations.
BestKeeper
BestKeeper (Pfaffl et al., 2004) uses raw Ct values (rather than relative quantities) to calculate the standard deviation and coefficient of variation for each candidate. Genes with the lowest SD are ranked most stable. It is simpler to use but less powerful than geNorm or NormFinder for complex experimental designs.
Validate with VoilaPCR
VoilaPCR Plus includes built-in geNorm analysis. Upload your validation plate with multiple candidate reference genes, and VoilaPCR will calculate M values, rank your candidates by stability, and tell you the optimal number of reference genes to use. No more exporting to Excel or running standalone R scripts.