The implications of using an inappropriate reference gene for real-time reverse transcription PCR data normalization.

By K. Dheda, J. Huggett, J. Chang, L. Kim, Stephen A. Bustin, M. Johnson, G. Rook, Alimuddin Zumla. Analytical Biochemistry, 2005

670 citations recorded by the source corpus.

hasbecome a routine technique for gene expression analysis,allowing accurate high-throughput RNA quantiWcationover a wide dynamic range at a relatively low cost.Although the technique has become well establishedover the past few years, data normalization remainsproblematic and is subject to frequent criticism. Normal-ization is essential to control for experimental errorbetween samples that can be introduced at a number ofstages throughout the procedure. Several di Verent strate-gies can be applied to the normalization of the expres-sion of an RNA of interest (for a review, see [1]). Themost common strategy is to use a reference gene as aninternal standard that is assumed to remain constantbetween experimental groups. It is now clear, however,that in most experimental situations the use of conven-tional reference genes, such as glyceraldehyde-3-phos-phate dehydrogenase (GAPDH) and -actin, isinappropriate due to their variability [1,2]. If the refer-ence gene chosen has a large expression error, the noiseof the assay is increased and detection of small changesbecomes unfeasible. Far worse, if expression of the refer-ence gene is altered by the experimental condition or thedisease under study, the results obtained may be entirelyincorrect. Therefore, it is essential to validate potentialreference genes to establish whether they are appropriatefor a speciWc experimental purpose. We have shown pre-viously that human acidic ribosomal protein (HuPO)was the least variable reference gene investigated in ourmodel of pulmonary tuberculosis (PTB) [3].Despite the fact that many validation exercises havebeen undertaken conWrming the variability of convention-ally used reference genes [1], it has been argued that over-all study conclusions remain unchanged because thisvariability is often similar in the control and interventiongroups [4]. Furthermore, reports that address the net e Vectof reference gene variability on experimental conclusionsare sparse and do not provide the Wnal proof that refer-ence gene variability may have a variable inXuence onstudy conclusions. To address the impact of a validationexercise on intergroup comparisons, we normalized diVer-ent sets of data with diVering reference genes, includingconventionally chosen genes. This report demonstratesthat experimental results are highly dependent on the ref-erence gene chosen. Moreover, whether statistically sig-niWcant diVerences between study groups were present orabsent depended on which reference gene was used.In the study, 21 patients with proven PTB and 21matched healthy volunteers were recruited. E Vector T-cellIFN- ELISPOT responses to TB-speciWc antigens wereused to conWrm that control subjects were not latentlyinfected with

Subfields: Molecular Biology, Surgery

View Paper: https://www.semanticscholar.org/paper/0a4bab41cbe912c24ddadeed15822e8978438437

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

Citations in the corpus

Quantification of mRNA using real-time RT-PCR (2006, background): “The inevitable consequence of inappropriate normalization is the acquisition of biologically irrelevant dat”
An optimized grapevine RNA isolation procedure and statistical determination of reference genes for real-time RT-PCR during berry development (2006, positive): “It has been shown that real-time RT-PCR results are highly dependent on the reference genes chosen [8], which supports putting considerable effort into validating gene(s) chosen for normalization prior to extensive experimentation.”