qRT-PCR (quantitative reverse transcription-polymerase chain reaction) is now the gold standard technique for mRNA detection and quantification, sensitive enough to enable quantification of RNA from a single cell.
The reverse transcription (RT) step is the main source of variability in a qRT-PCR experiment, so an optimal reverse transcription is essential for a reliable and successful qRT-PCR assay.
The reason for this is that the total RNA template can contain inhibitors such as buffer salts, fatty acids, alcohol, phenol etc, left over from the extraction process. This results in reduced RT and PCR reaction efficiencies, and generate unreliable quantification results.
Choosing the right RNA isolation or RNA clean-up kits from the many available can help, but a more sure-fire way to get great qRT-PCR results is to perform the reaction in two steps; RT, followed by qPCR.
Here are a few advantages that two-step RT-PCR reactions has over one-step qRT-PCR.
I. Reduced Primer Dimer Formation:
Problem: Primer dimer (PD) formation can be detected in qRT-PCR reactions using SYBR® green, although not Taqman®. PD formation during RT-PCR are a big problem because if the reaction resources are being used up producing a load of PD,it will result in reduced yield or inaccurate estimation of the specific product.
Primers that can form a strong 3′ duplex will self-hybridize readily at lower temperatures like, say42-50C, which is the temperature that most RT reactions are carried out at.
So primer-dimer can occur right there in the RT reaction. Even just a small amount of PD here is bad news in one-step qPCR because it will be efficiently amplified in the by DNA polymerase.
The reverse transcriptase itself can also participate to some extent in PD amplification, because reverse transcriptase exhibits DNA-dependent polymerase activity so will act on DNA templates and RNA: DNA hybrids.
Fix: By taking the RT-PCR reaction from a one-step to a two-step reaction, carry-over of accumulated PD from RT reaction can be minimized. This is done by simply diluting the cDNA from first strand synthesis before using it as a template for the qPCR reaction.
So the probability of non-specific product formation from PD as a result of accumulation from the RT reaction can be drastically reduced. But to fully eliminate PD formation you must also optimise the qPCR conditions.
Problem: Variation between two different RT reactions can complicate assay interpretation greatly.
To try and account for this, an internal control gene, such as a housekeeping gene, is amplified along with the target gene when doing one-step RT-PCR. This is called relative quantification. The target gene is quantified based on the relative expression level of the housekeeping gene.
For one-step RT-PCR, the control reaction for the housekeeping gene can either be carried out in a separate reaction (monoplex) or the same (duplex) reaction.
With the monoplex reactions, the target and control genes will be amplified from different pools of cDNA template. Since different RT reactions can have different efficiencies, performing control in a separate reaction can complicate comparisons between target RNA levels. Extensive optimization of primer balance is often required to obtain equal amplification efficiency for the two targets to allow them to be compared.
Fix: Two-step, monoplex reactions may be a way to fix this problem. In the first step of the two-step assay a common pool of cDNA with non-specific primers (such as oligo dT, random hexamers, octamers, nonamers or decamers) is generated.
This is then followed by separate qPCR assays performed with aliquots from the same cDNA pool. Variability in cDNA levels is overcome since the template for qPCR comes from the same pool. Also, the complement of genes is the same in each separate reaction and there is no preferential amplification of one target over the other.
III. Redundant Controls:
Problem: As I mentioned in II, housekeeping gene amplification is often used as a way to estimate the quantity of the target gene. When many targets are assayed by one-step RT-PCR, multiple housekeeping gene amplifications may be required for each target.
Fix: With a two-step RT-PCR protocol, the need to perform multiple reactions for housekeeping genes can be eliminated. Only control, housekeeping gene reaction is needed.
You can just use the common cDNA pool produced after RT for the detection of control and multiple targets. This serves as a good control against sample to sample variation and efficiency of the RT.
Two-step RT-PCR allow you to be flexible with the amount of reverse transcriptase you add to the reaction. For example, the amount of reverse transcriptase used for the first strand cDNA synthesis can be increased to give better results.
In this case, it is then important to limit the amount of RT product transferred to the qPCR reaction since Taq activity can be affected by RT. Clean up after RT reaction or dilution of the RT product is needed to avoid adding no more than 10% of the total qPCR reaction.
Most one-step RT-PCR reagents come with optimized buffers to work with both RT and PCR. This limits the choice of RT and PCR enzymes that you can try. In such cases, there is no other option than a 2-step RT-PCR.
V. Intra-assay variation
A common belief is that one-step RT-PCR gives reduced experimental variation because, since all of the enzymatic steps occur in the same tube under controlled cycling conditions and there is template handling.
Intra-assay variation is therefore an important issue to consider with 2-step RT-PCR. However, Spelman and colleagues  show that two-step RT-PCR is reproducible and shows good correlation between assays. Please take some time to look through that paper. It also discusses the advantages that I list above.
My final two cents: In any assay, not just qRT-PCR, is advantageous to have as many breaks as possible built in. I like to call these breaks “quality-control breaks”. They allow you to take stock and examine your results to see if things are still on track, and makes troubleshooting far easier. Even just for this reason, I am sold on two-step, rather than one-step, qRT-PCR.
1. Spelman, F et al, Analytical Biochemistry, 303, 95-98 (2002).
2. Sigma Aldrich Technical Guide to quantitative PCR: http://www.sigmaaldrich.com/life-science/molecular-biology/pcr/quantitative-pcr/qpcr-technical-guide.html