Related Experiment Videos
PCR-based calibration curves for studies of quantitative gene expression in human monocytes: development and
Reidun Øvstebø1, Kari Bente Foss Haug, Knut Lande
1The Research and Development Group, Department of Clinical Chemistry, Ullevål University Hospital, 0407 Oslo, Norway. reidun.ovstebo@ioks.uio.no
Background:
Quantitative reverse transcription-PCR (RT-PCR) used to detect small changes in specific mRNA concentrations is often associated with poor reproducibility. Thus, there is a need for stringent quality control in each step of the protocol.
Methods:
Real-time PCR-based calibration curves for a target gene, tissue factor (TF), and a reference gene, beta-actin, generated from PCR amplicons were evaluated by running cDNA controls. In addition, the reverse transcription step was evaluated by running mRNA controls. Amplification efficiencies of calibrators and targets were determined. Variances within and between runs were estimated, and power statistics were applied to determine the concentration differences that could reliably be detected.
Results:
Within- and between-run variations (CVs) of cDNA controls (TF and beta-actin), extrapolated from reproducible calibration curves (CVs of slopes, 4.3% and 2.7%, respectively) were 4-10% (within) and 15-38% (between) using both daily and "grand mean" calibration curves. CVs for the beta-actin mRNA controls were 12% (within) and 19-28% (between). Estimates of each step's contribution to the total variation were as follows: CV(RT-PCR), 28%; CV(PCR), 15%; CV(RT), 23% (difference between CV(RT-PCR) and CV(PCR)). PCR efficiencies were as follows: beta-actin calibrator/target, 1.96/1.95; TF calibrator/target, 1.95/1.93. Duplicate measurements could detect a twofold concentration difference (power, 0.8).
Conclusions:
Daily PCR calibration curves generated from PCR amplicons were reproducible, allowing the use of a grand mean calibration curve. The reverse transcription step contributes the most to the total variation. By determining a system's total variance, power analysis may be used to disclose differences that can be reliably detected at a specified power.