to BioTechniques free email alert service to receive content updates.
Flaws in Gene Expression Studies?

Melissa Lee Phillips

Do standard normalization procedures hide true changes in gene expression?

Current analyses of global gene expression may suffer from a pervasive flaw, according to a new study. The authors report that different cell types can produce vastly different quantities of total RNA – an effect that is masked by experimental procedure in standard expression studies.

In a paper published in Cell on October 26, research scientist Tony Lee from the Massachusetts Institute of Technology and colleagues reported that a recently developed type of experimental control provides accurate gene expression quantification and should be introduced in current expression studies (1).

Studies of global gene expression are one of the most popular methods in biomedical studies. By measuring levels of RNA found in different cells or tissues, these analyses give quantitative information about gene expression changes between different cell types or disease states.

Most gene expression analyses rely on the assumption that all cells produce about the same amount of total RNA, but this assumption was called into question recently by Lee’s group. In a previous experiment, they found that tumor cells that express the oncogene c-Myc show transcriptional amplification: those cells produce about two or three times more total RNA than normal cells (2).

In traditional analyses of gene expression, such as DNA microarrays, researchers isolate and purify about the same amount of RNA from each cell type they're studying. But normalizing RNA amounts between cell types hide cases in which one cell type is actually expressing several times more total RNA.

Standard methods "only work if you know that the same amount of RNA is being produced per cell," said Lee. This is not the case in cells that express c-Myc. Not only is total gene expression amplified in these cells, but the researchers also found that expression of certain genes is amplified more than expression of others.

To fix this problem, Lee and colleagues used synthetic RNA molecules called spiked-in standards as controls. Because these standards are introduced by the experimenters, their levels are not affected by total RNA expression levels in the cells. Spiked-in standards have only been commercially available a short time, so most research groups have not had an opportunity to use them.

By performing standard gene expression analysis of healthy cells and cells expressing c-Myc, the authors found that some genes were upregulated while others were downregulated in the cancerous cells. But using spiked-in RNA standards, the researchers found that more than 90% of genes were actually upregulated in the c-Myc cells. "It was surprising to us how big the effect was," said Lee.

The authors don't know yet if these results will apply to other, non-cancerous cell types "but there are plenty of situations where we could imagine bulk shifts in in transcription," said Lee. "This may be relevant for a lot of research."


1. Lovén, J., D.A. Orlando, A.A. Sigova, C.Y. Lin, P.B. Rahl, C.B. Burge, D.L. Levens, T.I. Lee, and R.A. Young. 2012. Revisiting global gene expression analysis. Cell 151(3):476-482.

2. Lin, C.Y., J. Lovén, P.B. Rahl, R.M. Paranal, C.B. Burge, J.E. Bradner, T.I. Lee, and R.A. Young. 2012. Transcriptional amplification in tumor cells with elevated c-Myc. Cell 151(1):56–67.

Keywords:  gene expression