Protocol for competitive RT-PCR

For quantifying mRNA, we use a competitive RT-PCR protocol with internal standard RNAs. These are added in a defined quantity to the RNA sample prior to the RT reaction. The resulting standard cDNA is coamplified with the same primers as the endogenous target sequence. Its PCR product is approximately 50 nucleotides smaller. This method allows measurement of small differences (as low as factor 2) in mRNA amount between RNA samples. RNA standards have the big advantage that also the variation of the RT effiency is irrelevant, as well as the variation of the PCR effiency.

Making internal standard RNAs


The sequence to be amplified should best span an intron so that gDNA contamination does not play a crucial role. Otherwise, the RNA has to be treated with DNase I (RNase-free) and the success of this treatment has to be controlled by a PCR without prior RT. It is, however, always recommendable to treat the RNA samples with DNase so that no genomic DNA competes for the PCR components.
To make a standard, first a PCR with a conventional downstream primer and a modified upstream primer (40 nucleotides in length) is performed according to Celi et al. (1993). cDNA is used as template for the PCRs. The PCR products are isolated from an 1.5% agarose gel and cloned into a pGEM 3Z vector that containes a T7 RNA promoter sequence. The in vitro transcription of the cloned fragments is performed using T7 RNA polymerase (e.g. Gibco BRL). The internal standard RNA is then treated with RNase-free DNase I (e.g. Gibco BRL) to remove the plasmid DNA (success also checked by PCR without prior RT) and finally quantitated by measurement of the optical density at 260 nm and stored at -70°C.

The main problem with RNA standards is their instability. We found that especially thawing and refreezing damages them. Therefore it is best to store the standards in small aliquots in different dilutions and discard them if thawed too often. Also, this problem means that no absolute amounts can be measured, because there is no way of knowing how much standard is already degraded in the aliquot used. However, this method is very reliable and accurate for comparison of different samples if the same standard aliquots are used for measuring their mRNA amount.

Quantitation of mRNA


For the quantitation of the mRNA of one RNA preparation 4 RT reactions are prepared with 1 µg total RNA each and different amounts of standard RNA (if more than one mRNA is to be quantitated, the different standards can be mixed together). We use the SuperScriptPreamplification System from Gibco BRL for our RTs. For the first measurements of a mRNA it is best to add standard amounts which differ by factor 10 (i.e.: 100 fg; 1 pg; 10 pg; 100 pg; 1 ng) to determine the range in which the transcript amount is to be found. If that is known, factor 2-2.5 between the standard amounts gives more accurate results (i.e.: 25 pg; 50 pg; 100 pg; 250 pg).

In the following PCR 3-5 µl of cDNA, 1.5 units Taq DNA polymerase (Pharmacia), 200 µM of each dNTP, 250 nM of each primer and 1/10 volume of a 10x PCR standard buffer (15 mM MgCl2; 100 mM Tris/HCl, pH 8.3; 500 mM KCl) are added to a total volume of 50 µl. The PCR is run in the thermal cycler GeneAmp 9600 (Perkin Elmer). The PCR products are then separated on a 1.5% agarose gel, stained with ethidium bromide, SYBR-Green or SYBR-Gold (Molecular Probes, higher sensitivity, but also more light sensitive) and scanned by a CCD camera. The amount of cDNA used for a PCR, the number of cycles and the nucleic acid stain used depend on how abundant the transcript is that is being measured. If heteroduplices appear, one possibilty is to run less PCR cycles. If they still cannot be avoided, Eferl et al. addressed this problem extensively (Technical Tips Online).
Figure 1: Measurement of two RNA samples. The lower bands are the standard bands, the upper bands are the transcript bands. Lane 1-4: sample 3; lane 5-8: sample 4; lane 1+4: 250 pg standard; lane 2+5: 100 pg standard; lane 3+6: 50 pg standard; lane 4+8: 25 pg standard. The rectangles represent objects set with the ImageQuant software.

The band intensities are then analyzed by ImageQuant software (Molecular Dynamics) as follows: place rectangular objects over each band with a little background (which is set to local) and start „Volume Report". The ratios of transcript to standard band intensity (volume transcript/volume standard) are then fixed in a double logarithmic graphical representation (Harvard Graphics, Excel or Microcal Origin) in which the amount of standard mRNA equal to the amount of transcript RNA can be read at the intersection of the line with the y-value 1. Ideally, the 4 measured ratios should yield a straight line. However, because of variances due to pipetting, dilution etc, the ratios often amount to a slight zig-zag. In this case, a regression line is calculated or drawn.Figure 2: Graphical representation of the ratios measured above
In this example, sample 3 contains 55 pg of the mRNA in question (in our case the NF1 mRNA) and sample 4 86 pg. As said before, these absolute values are not reliable. In this case, we set one value (e.g. the control or the value under one condition) to 100% (in this case sample 4) and simply relate the other value to this. So the conclusion drawn from this measurement is that sample 3 (RNA from cells of a NF1 patient) contains only 64% of the mRNA amount as sample 4 (RNA from cells of a healthy control).

(Protocols of the Lab of Dieter Kaufmann, Department of Human Genetics, University of Ulm, Germany)