I'm setting up a real time PCR protocol to assess copy number variation of some mRNAs in the eel. I'm trying to use ultramers synthesized by IDT as standards for the absolute quantification, but I had really bad results! It seems as if the reaction is "too" efficient, since the mean Ct for my lowest standard (10^1 copies) is about 16, and the mean Ct for my highest standard (10^8 copies) is about 9. My primer concentration is 0.2 uM and the reaction volume 10 uL. Can someone give me some help?
You might have a problem with background (trend) correction. I have no experience with ultramers. I used PCR-product (nested primers), purified and quantified by OD260. This worked well. If you are going to try this, take care with contaminations. Purify and quantify the "standard" in another lab/building and take only very diluted samples from this standard to the place where you prepare the qPCRs.
If your 10^8 (copies/rxn) standard appears at ~9, then, doing the math, 1 copy should appear at a Cq of about 35.6 at 100% efficiency. The only thing that is unreasonable, like you said, is that your 10-copy reaction appears at a Cq of about 16. Where (at 100% efficiency) about 790,000-copy reaction should appear at ~16 in your scenario.
So it seems likely that your dilution series may be off (but by a factor of 79,000)... yikes.
Be sure the ultramer has been fully solubilized before serial diluting? Or else it does seem possible that major contamination of the work-space may have happened somehow (as Jochen alludes to above).
Also - another troubling note... if you are using SYBR Green qPCR, I have sometimes seen standard curves that flatten unreasonably - no response to serial dilution, and I have never been able to explain that, given that the product (melt curve) was correct and no primer dimers were present. I think it was associated with the particular SYBR Green mastermix chosen.
The only other explanation is that you have inadvertently out-done Linus Pauling's early theory of a triple helix and have discovered the 79,000-fold helix. ;]
Take care to use proper buffers for dilution of your calibration concentrations, e.g. 10 mM Tris pH 7.5 to 8.0, 50 mM KCl. Unbuffered dilutions will not be stable, and salt-free buffers may produce unexpected deviations (by clumping?).
Beside this, I follow Jochen's suspicion that the PCR may be unspecific. Have you testes dilution series of your eel DNA? An agarose gel (also of the qPCR) will tell you if the products are specific.
A general risk with a (highly concentrated) synthetic standard is cross-contamination through smear or aerosole (pipet)!
thank you all for the advices!! I will try to perform additional tests with my ultramers. The (other) unreasonable stuff is that if I use a standard made by cloning my PCR product into a plasmid vector, the protocol works well! My standard curves are good to be used for abs quantification: PCR efficiency around 96% over a 10^8 - 10^2 copies. This should say I don't have contaminations, and I was quite sure about this since I always work with filter tips and different aliquots of water, master mix, and primers.
I don't know if IDT ultramer preps commonly contain truncated forms of the intended oligo or not (if not extensively purified). Perhaps, if they do contain truncates, this may confound the qPCR in the way that you are experiencing when you try to use it. (Not sure)
Up front, I would say that if your PCR product/linearized plasmid vector works well -then run with that. But, realize that different contexts of your target will often give you different amplification efficiencies (using the same primers) during qPCR.
I.e., your linearized plasmid will amplify at a different efficiency than your biological mRNA/cDNA template does. And, if/when you notice this, be sure to reconcile the two universes mathematically...
One way to do this is by the attached file.
In short: the equation in the attached file takes the y-intercept of the absolute template dilution curve (the Cq at which 1 copy would appear) and expresses it in the language of what the experimental/biological sample's y-intercept would be (for 1 copy). It does this by acknowledging the exponential amplification efficiency for the absolute standard ("Eamp"a) and the exponential amplification efficiency for the bioloigal sample template dilution series ("Eamp"s). Both kinds of standard curve must be run to obtain these values.
Once this transformation is performed, i.e., once you know where 1 copy of target would appear for biologically-derived template, you can estimate the # of target copies in all of your biological samples based on what you learned from your absolute standard curve..
Assuming no qPCR inhibition (or proof of no inhibition using the SPUD assay or other means, etc.) this approach helps you obtain reasonably good estimates of target copy number in your biological samples (as long as you absolutely trust the quantification method used to estimate copies of your original absolute template/uL with 100% confidence).
The take home point here would be that, absolute standards (for a certain target), if amplifying at a different efficiency than the same target does in your biological samples, are not yet a faithful direct indication of copy number in your biological samples until you use the attached mathematical transformation to 'express one Cq scale in the language of the other Cq scale' by taking the different apparent exponential amplification efficiencies into account.
- Badges
- Science topic