Semiquantitative PCR - I need some orientation, please (Nov/15/2006 )
Hi
I am thinking of trying to do semiquantitative RT-PCR. My plan is:
- Perform competitive PCR with the sequences I'm interested in and beta-actin from cDNA obtained from a human cell line.
- Then run 5ul of each PCR in a stained gel (all in the same gel). Use a software to measure the intensity of each band and normalize the data using the beta actin bands of each well respectively.
My questions are:
- It the band size of each band is different (beta actin would be always the same), could you let me know how the calculations should be done?
- How accurate is this methond?
- Could you tell me the advantages and disadvantages of this method vs Real Time PCR?
- What problems can it happen?
- Do you know any good chapter of a book or paper about this?
A lot of questions I know. All opinions, ideas and suggestions will be very wellcome.
Thanks a lot.
My suggestion is to perform RT-PCR for actin first for normalization of your samples’ concentration, then conduct RT-PCR for sequence of your interest using normalized templates.
For more accurate quantitation, you could clone your RT-PCR products and then diluted then into several concentration for PCR, then compare the PCR band intensity with PCR product of your sample.
The dynamic range of end point detection is small than real time PCR,
For refferece, see Ambion’s web site link
http://www.ambion.com/techlib/basics/rtpcr/index.html#7
Thanks a lot. It's a fantastic link, very very helpful.
REading that information, some new questions have come to my mind:
- I still don't know how the size of the band affects the intensity. Is it just a matter of dividing by the number of bp to normalize?
- If I want to determine the exponential range, do I need to run it in a polyacrylamide/ urea gel or can I do it in a agarose normal gel? I supposse it isn't so accurate but it will work well, isn't it?
- I want to compare the amount of five proteins (A,B,C,D and E) which form a heterocomplex. Could I use one of them to normalize the rest? Let's say A+B, A+C, A+D and A+E. The best part of this is that they will not reach the platoo so quick as housekeeping genes.
- Why can't I run the five pair of primers together with more amount of cDNA as a template? It is only because we cannot be sure if we are still in the exponential range?
I hope I make someone think. I always like questioning things.
Thanks in advance.
If I were you, I would use reference gene (such as rRNA) for normalization first, adjust all sample in to the same concentration before check examined genes’ expression.
If you hope compare the same gene expression in these samples, it is a little bit easy, just compare PCR product of the same cycles, which will give a roughly idea where they express high, for semiquanitive compare, you might dilute high sample with high expression into several level and compare band intensity with other samples and result in semi-quantitive data.
Since amplification efficiency affected by PCR products size (long size reduce the range of exponential range), primer annealing ability and others, so, it is not reliable to compare different gene by this way.
For compare the expression of different genes, it is better prepare a standard curve using know concentrated gene fragment (such as cloned PCR products), based on that, you could compare the “absolute copy” in your PCR reaction.
Thanks, rye. Yes, they are different genes.
My reference gene may be beta actin, am I right?
Because in the past other people used Real Time PCR with beta actin I know that 20 cycles is in the exponential range for beta actin. I assume that beta actin is better expressed in my cell line that the target proteins.
Having said this, can I run "relative PCR" with my samples? To say, protein A with beta actin for 20 cycles, protein B with beta actin for 20 cycles, etc and compare all of them?
Band sizes are in the same range. he difference between bands is only 80-100bp maximum (except for one which is 300bp larger). Do you think I can still do it.
Basically, what I would like to do is a easy and more important no time-consuming method to do that semiquantitative PCR approach in order to add some more data to our very soon paper. This is not a paper about gene expression, it's just additional information which can contribute to target a higher impact factor magazine.
In conclusion, is my approach feasible? Can it be used in research?
Thanks again and sorry for boring you with so many questions.
Come on, a fast and quicky reply. Anyone? Maybe the ever-helpful rye?
Ok,Let's make a deal. I promise I don't ask further question.
;-)
Usually people show RT-PCR results in this way:
Image of interest gene PCR products with different intensity showing which sample the interested gene is high expressed (strong band) or low expressed (weak or no band), some times with the band intensity value evaluated by gel image analysis system. In addition, there should be a picture of reference gene PCR products with identical intensity which showing the same amount of template have been added to PCR reaction for compare the abundance of interested gene. Usually PCR for reference gene are separated reaction but using the same amount of template as PCR for your interested gene.
Or might Just show values of relative or absolute data of expression level in different sample, which are based on estimation of standard curve.
I am not sure it is clear enough for your question, and I'd like hear any suggestions from other.
is it ok if I choose 36 cycles for my gene of interest and 30 cycles for housekeeping gene? Can I do a semiquantitative analysis?
I use different cycles because a 36 and 30 cycles bands are strong enough to see but did not reach the plateau. I use the same amount of template for both PCR reactions.
I use different cycles because a 36 and 30 cycles bands are strong enough to see but did not reach the plateau. I use the same amount of template for both PCR reactions.
This was perfectly okay before the advent of real-time PCR, and people still show such data, but I generally flinch at this.
housekeeping such as beta-actin may require even less than 30 cycles. Make sure that your conditions do not affect beta-actin message. For example if you are dealing with conditions that affect cytoskeletan, motility, it is likely that beta actin may have been affected. In that case, use additional loading controls such as lamin b1 and gapdh.
Plateau is not the only consideration, it is the logarithmically linear range of amplification, that starts getting skewed much before you reach plateau. Additionally, the slope of amplification for your gene and housekeeping gene primers may be quite different. So, unless there is a huge, dramatic difference (like all-or-none
, I would be skeptic of your results. 
..