Real-time PCR, Volume VIII

Table of Contents
Letter from the Editor
Index of Experts
Q1: How do you determine your sample integrity or quality?
Q2: What guides your choice of fluorophore and quencher for a particular experiment?
Q3: What controls do you typically include, and why?
Q4: What normalization strategy do you follow?
Q5: What do you find to be acceptable levels of intra- and inter-assay variation?
Q6: Do you usually quantify the absolute or relative amount of the PCR product? What strategy do you use to do that?
List of Resources

Letter from the Editor

In this latest issue of Genome Technology's technical guide series, we delve into the intricacies of running real-time PCR experiments. It's our eighth foray into this topic and this time around, we asked our experts how they determine their sample quality, what sorts of controls they deem necessary, and what level of variation they find acceptable, among other topics.

Our experts took to the task with good humor and say that their controls, of course, often depend on the experiment at had. Generally, though, they employ both biological and technical replicates to ensure accuracy, but some favor biological replicates a bit more. Throughout this issue, they also share other tidbits on how to perform quality real-time PCR experiments. As always, be sure to check out the resource pages for related papers, websites, and upcoming conferences.

Ciara Curtin

Index of Experts

Many thanks to our tireless experts for taking the time to contribute to this technical guide, which would not be possible without them.

Willard Freeman
Director, Functional Genomics Facility,
Hershey Center for Applied Research
Pennsylvania State College of Medicine

Allison Gillaspy
Associate Director, Laboratory for Genomics and Bioinformatics
The University of Oklahoma
Health Sciences Center

Jim Huggett
Senior Research Fellow
Windeyer Institute of Medical Research
University College London

Q1: How do you determine your sample integrity or quality?

For our studies, we want to determine the RNA concentration, presence of any organic contaminants which interfere with subsequent reactions, and any sample degradation. Most studies have a large number of samples and we start by determining the RNA concentration spectrophotometrically, with a NanoDrop instrument. This also helps us initially determine if there is organic contamination of the sample through the 260 nm to 230 nm ratio. Next, we use fairly standard lab-on-a-chip methods, like the Bioanalyzer, to determine RNA quality using the RNA integrity number. Depending on the sample concentration from the spectrophotometry we can determine the appropriate Bioanalyzer chip (Nano or Pico) needed. While the Bioanalyzer does return a concentration value, we prefer the values from the spectrophotometer. For less routine protocols such as samples derived from laser capture microdissected tissue, we have to use some alternate techniques like fluorescence-based assays to quantify the RNA concentration.
— Willard Freeman

We run all RNA samples on the Bioanalyzer using the RNA nano kit to check for integrity prior to reverse transcription. If the Bioanalyzer shows signs of degradation or an abnormal ratio of ribosomal peaks then we do not proceed to the reverse transcription stage. We also determine if there is residual chromosomal DNA by using the Bioanalyzer and by always setting up an RT negative control for each sample.
— Allison Gillaspy

This is largely dependent on the experiment; for measuring RNA from cell culture experiments, using epithelial cells for example, we would expect good ribosomal bands with minimal degradation. However, this can vary with cell and tissue type (or age — like when using formalin-fixed or archeological samples). The important thing is that sample integrity or quality is defined. Another example of this is looking for pathogen DNA in the soluble fraction of fresh urine as a diagnostic. Here, quality is much more difficult to define and can be poor; however, this does not preclude us from using these extracts as long as we discuss this problem.
— Jim Huggett

Q2: What guides your choice of fluorophore and quencher for a particular experiment?

In our gene expression (RT-PCR) work we try to utilize pre-designed primer probe sets whenever possible, depending on the gene and the species being examined. Our standard workflow uses 5' exonuclease assays with a FAM reporter and a minor groove binder quencher since these provide more specificity than intercalating dyes such as SYBR Green.

Additionally, many of these probes are exon spanning and will therefore not anneal to any amplified genomic DNA. Generally, we avoid multiplex reactions because the time and amount of reagents needed to optimize these is greater than just running only one primer/probe per well. When necessary we use intercalating dyes, such as in chromatin immunoprecipitation-PCR (ChIP-qPCR).
— Willard Freeman

Some of our facility users prefer the TaqMan reagents if there are validated primers available for their specific experiments or if they are interested in only a few genes. However, we typically use SYBR Green because it is cost-efficient and works well for our purposes, specifically when we are examining several different genes for one experiment. We also have several projects with historical data using SYBR Green so continued use makes it possible for us to more easily compare the results.
— Allison Gillaspy

Price and a good delta Rn. We tend to use black hole quenchers as they provide a low background and are reasonably priced. We do still use SYBR Green for many of our applications.
— Jim Huggett

Q3: What controls do you typically include, and why?

All samples are always run in triplicate with no-template and no-enzyme control. It is important to run properly powered qPCR studies like any good biological experiment. For each group in a study, we use anywhere from n=3 to n=12 independent samples to assess the biological variance and have appropriate power to determine statistically significant changes. Each sample is run in technical triplicates to provide tight data through elimination of drop-out wells. The no-template and no-enzyme reactions give us the quality control for both the assay performance and any genomic DNA contamination. All of our studies are performed in 384-well formats which gives us the number of wells we need for biological and technical replicates as well as reducing reagents costs through small reaction volumes and numbers of plates run.
— Willard Freeman

Every sample is set up in duplicate along with positive and negative controls (in duplicate) for each run. We repeat this two more times to equal three technical replicates for each sample.
— Allison Gillaspy

Depending on the experiment, we tend to use multiple negative controls spread throughout the reactions (e.g. three are inputted before any DNA is aliquoted, three after the standard curve, and typically four after the samples) allowing us to track ontamination. Usually when contamination does occur it is sporadic, low-level, affecting only one negative-control replicate. Positive controls are used with our diagnostic PCRs — these are usually samples we know to be positive and we also target internal sequences like human DNA. We do not always replicate the sample PCR reactions favoring biological replicates where possible. We do, however, run our standard curves as either duplicate or triplicate reactions for each concentration.
— Jim Huggett

Q4: What normalization strategy do you follow?

For the vast majority of our studies, we use a single endogenous control that we have determined to be stable for that experimental protocol. The levels of the endogenous control are determined by an absolute quantitation experiment. While there is a great deal of discussion in the literature regarding the use of multiple endogenous controls, this is usually necessary in experiments where cells are undergoing extensivetranscriptomic alterations. Our studies are centered around neuro­behavioral experiments which examine modest changes in neuronal gene expression and there are stable endogenous controls.

Also, all of our samples are quantified at the RNA and cDNA steps to ensure equivalent sample loading.

The challenge comes in studies that attempt to compare different tissues, or cells undergoing differentiation or dedifferentiation. These studies require careful examination of the endogenous control(s) to ensure that these gene(s) do not change in expression and provide a stable normalization factor.
— Willard Freeman

After reverse transcription, all cDNA samples are treated as though the recovery is equal to the input (per guidelines from AB). From there, all samples are diluted to a final concentration of either 1 ng/ul or 2 ng/ul before proceeding with the real-time experiments.
— Allison Gillaspy

For gene expression work, we use either a single validated reference gene for greater than five-fold measurements or multiple validated reference genes (using Jo Vandesompele's geNorm) for finer measurements.
— Jim Huggett

Q5: What do you find to be acceptable levels of intra- and inter-assay variation??

From our tests, technical replicates should generally produce a CV of less than 10 percent. Electronic pipettes and robotics can reduce this value to below 5 percent. Variability between samples is dependent on the gene being examined and the nature of the samples. The variability between samples is best reduced by good experimental practices upstream of the qPCR studies (e.g. animal handling, dissection, RNA isolation, etc). There is always some biological variability between samples and that is why designing the experiment with the appropriate sample size to determine statistical significance is so important.
— Willard Freeman

A difference of 1 Ct (about 0.4 Ct standard error) wouldn’t be that much cause for alarm for us because of the number of technical replicates that we do. If there is a huge difference it is normally easy to see that it may have been due to a pipetting error or something else during the set-up because of the replicates too. If there is a lot of variation between the different replicates, as opposed to just one of the replicates being odd (more than 1 Ct difference), then those samples would definitely be repeated.
— Allison Gillaspy

We would expect inter-assay variation to be less than 20 percent CV when measuring more than 100 copies per reaction.
— Jim Huggett

Q6: Do you usually quantify the absolute or relative amount of the PCR product? What strategy do you use to do that?

With the exception of determining endogenous control suitability, we do not perform absolute quantitation. We find that very little additional biological insight is gained through absolute quantitation as the number of cells in the samples is difficult to impossible to determine. Relative quantitation allows us to determine fold changes with experimental treatments and we use standard 2^(-delta delta Ct) methods with one of the samples from the control group as the calibrator. Relative quantities are the primary information of interest in our neuro­behavioral and biomarker studies; the great deal of extra effort required for absolute quantitation does not really provide additional useful information in our case.
— Willard Freeman

We usually look at relative quantitation based on an internal control, but we've also done some standard curves. The strategy for relative quantitation is the delta delta Ct method that is used by many groups. Our normalization strategy for absolute quantitation is a typical standard curve that is created for us by the Applied Biosystems software after running the samples using the 7500 Fast system.
— Allison Gillaspy

Not usually, although we have done this using the absolute normalized ­fluorescence reading at the end of the reaction. Alternatively, PicoGreen can be used for an accurate measurement.
— Jim Huggett

List of Resources

Publications

Bengtsson M, Hemberg M, Rorsman P, Ståhlberg A. (2008). Quantification of mRNA in single cells and modelling of RT-qPCR induced noise. BMC Molecular Biology. 17(9):63.

Bustin SA. (2005). Real-time, fluorescence-based quantitative PCR: a snapshot of current procedures and preferences. Expert Review of Molecular Diagnostics. 5(4):493-8.

Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M, Mueller R, Nolan T, Pfaffl MW, Shipley GL, Vandesompele J, Wittwer CT. (2009). The MIQE Guidelines: Minimum Information for Publication of Quantitative Real-Time PCR Experiments. Clinical Chemistry. 55:4. Online Early.

Bustin SA, Benes V, Nolan T, Pfaffl MW. (2005). Quantitative real-time RT-PCR—a perspective. Journal of Molecular Endocrinology. 34(3):597-601.

Fleige S, Pfaffl MW. (2006). RNA integrity and the effect on the real-time qRT-PCR performance. Molecular Aspects of Medicine. 27(2-3):126-39.

Fleige S, Walf V, Huch S, Prgomet C, Sehm J, Pfaffl MW. (2006). Comparison of relative mRNA quantification models and the impact of RNA integrity in quantitative real-time RT-PCR. Biotechnol Letters. 28(19):1601-13.

Kubista M, Andrade JM, Bengtsson M, Forootan A, Jonák J, Lind K, Sindelka R, Sjöback R, Sjögreen B, Strömbom L, Ståhlberg A, Zoric N. (2006). The real-time polymerase chain reaction. Molecular Aspects of Medicine. 27(2-3):95-125.

Lefever S, Vandesompele J, Speleman F, Pattyn F. (2008). RTPrimerDB: the portal for real-time PCR primers and probes. Nucleic Acid Research. 37 (Database issue):D942-5.

Lind K, Ståhlberg A, Zoric N, Kubista M. (2006). Combining sequence-specific probes and DNA binding dyes in real-time PCR for specific nucleic acid quantification and melting curve analysis. Biotechniques. 40(3):315-9.

Nolan T, Hands RE, Bustin SA. (2006). Quantification of mRNA using real-time RT-PCR. Nature Protocols. 1(3):1559-82.

Palais R, Wittwer CT. (2009). Mathematical algorithms for high-resolution DNA melting analysis. Methods in Enzymology. 454:323-43.

Willems E, Leyns L, Vandesompele. (2008). Standardization of real-time PCR gene expression data from independent biological replicates. Journal Analytical Biochemistry. 379(1):127-9.

Books

A-Z of Quantitative PCR. Ed. by Stephen Bustin. (July 2004). International University Line; ISBN 978-0963681782.

PCR: Methods Express. Ed. by Simon Hughes, Adrian Moody. (May 2007). Scion Publishing Ltd. ISBN 9781904842293.

Rapid Cycle Real-Time PCR-Methods and Applications. Ed. by Carl Wittwer, Meinhard Hahn, Karen Kaul. (February 2004). Springer; ISBN 9783540206293.

Real Time PCR. By Tevfik Dorak. (June 2006). Taylor & Francis; ISBN 9780415377348.

Real-Time PCR: Current Technology and Applications. Ed. by Julie Logan, Kirstin Edwards, Nick Saunders. (January 2009). Caister Academic Press; ISBN 9781904455394.

Real-Time PCR in Practice. By Jochen Wilhelm. (March 2009). Vch Verlagsgesellschaft Mbh; ISBN 978-3527316878.

Websites

BatchPrimer3
http://probes.pw.usda.gov/batchprimer3/index.html

GeneFisher2
http://bibiserv.techfak.uni-bielefeld.de/genefisher2/

GenomeTester 1.3
http://bioinfo.ut.ee/genometester/

geNorm
http://medgen.ugent.be/~jvdesomp/genorm/

MethBLAST
http://medgen.ugent.be/methBLAST/

MethPrimer
http://medgen.ugent.be/methprimerdb/

MFEprimer
http://biocompute.bmi.ac.cn/MFEprimer/

Multi-Objective Multiplex PCR Design
http://genomics14.bu.edu:8080/MuPlex/MuPlex.html

qPrimerDepot
http://primerdepot.nci.nih.gov/

OligoCalc
http://www.basic.northwestern.edu/biotools/oligocalc.html

PriFi
http://cgi-www.daimi.au.dk/cgi-chili/PriFi/main?config.x=101&config.y=30

PRIMEGENS
http://compbio.ornl.gov/structure/primegens/

Primer3Plus
http://www.bioinformatics.nl/cgi-bin/primer3plus/primer3plus.cgi

SNPbox
http://www.snpbox.org/

UCSC In Silico PCR
http://genome.ucsc.edu/cgi-bin/hgPcr

Conferences

qPCR Course-Core Module
TATAA BioCenter/Technical University of Munich
Freising-Weihenstephan, Germany
June 15-17, 2009

Advances in qPCR
Select Biosciences
Berlin, Germany
September 17-18, 2009

qPCR Congress 2009
Oxford Global Conferences
London, England
November 26-27, 2009

3rd qPCR Symposium USA
TATAA BioCenter
San Francisco, California
November 9-12, 2009