I am trying to generate a five-data point standard curve from a 10X dilution series using an initial template cDNA of concentration 2100ng/ul. I use the Roche Light Cycler software 4.05 and MCF-7 breast cancer cell line-derived cDNA to construct a standard curve for the GAPDH gene. This is my 10th attempt and irrespective of my endless trials to vary each component of my RT-PCR cocktail, to maintain consistency throughout my experiments and to minimize contamination, I still get a non-straight standard curve with the first two data points forming a horizontal line before the slope can start to decrease. And with reference to the amplification curve, it looks like the two data points have got the same cross point, i.e. they start amplifying at the same time. I don't understand how this is possible because I did a serial dilution meaning the amount of cDNA in each RT-PCR mixture is not the same. I'm also very careful with my pipetting however I struggle to generate amplification curves that are equally spaced and sometimes my replicates scatter. Please help.
only two of your cDNA dilutions are diverting try to avoid them n go for the others....Lets see what happens...
Have you made a gel to see that the correct product is amplified in these two wells? (additions to melting curves, because melt curves do not always show distinct melting peaks for the wrong product).
What about negative controls (-RT and NTC)? Could it be a contamination?
This "broken line thing" is a Roche "feature". They think it's easier to ignore inappropriate background correction and instead adopt the calibration curve to the (erroneous) ct values, hoping the errors are systematically similar for standrad and unknowns.
@Warren Albertin please see my attached melting curves, they don't seem to have any anomalies such as asymmetrical peaks, shoulders nor unusually chunky peaks suggesting my amplification curves are not composite of more than one product, unless if i am misinterpreting the data. Would you say otherwise?
the metling curves are hard to interpret. They are very wide and may represent a convolution of several different curves. It is also difficult to judge if all curves show the same Tm. Run a gel. And check if you have contaminations.
@Jochen Wilhelm I only made a gel for the starting template cDNA and not for each dilution, are you saying i should try run a gel for the two dilutions? Also, i do try to minimize contamination, what i didn't try is negative controls, may i ask what do you mean by -RT and NTC?
@Pankaj Bagul the thing is I am trying to construct a 5-point standard curve, avoiding the two would demean that, but on the other hand it could help me see whether the problem is with my last two dilutions or the machine. I'll get back to you once I've tried that, thanks.
@Christoph Blossey thing is the software generates the curve automatically and even if i do it manually there are unequally spaced amplification curves to prove me wrong.
@Jochen Wilhelm Let me re-run the gel and see what i get, I'll get back to you.
Pontsho, yes, please make a gel of your dilutions to confirm that a) your melting curves are from the correct product and b) that there is only one product. The negative controls are required to check if you have contaminations. Btw, you should alwayys incluse negative controls. -RT is "minus revers transcrition": everything is done as for all other samples, except that the RT enzyme is added. A later PCR product indicates that you have DNA contamination that gets amplifiable product. NTC means "no-template control", here you add water instead of cDNA to see if you have some contaminations in the reaction components itself.
Just a small note here: too much cDNA in the reaction will inhibit the qPCR, so your 1st two points are exhibiting template inhibition of the Taq enzyme (therefore similar Cq values). Your real curve starts after that (where the slope starts to respond to dilutions; slope starts to decrease). Hopefully you're seeing somewhere near 3.3219 Cq difference between your serial 1:10 dilutions at the points after inhibition lets up. Avoid working in the inhibited range. With abundant ref. genes (such as GA3PDH) this is especially a concern. Even much much worse is 18S ribosomal.
Jack, when I am interpreting the plot correctly, then the problem are the highest dilutions. This still seems to be a problem of primer-dimer formation or contaminations. But you are right: too high cDNA conc can inhibit the PCR.
Hi Jochen, I see that I didn't look at the picture attached. I was initially responding to Pontsho Moela's original post here where she stated: "I still get a non-straight standard curve with the first two data points forming a horizontal line before the slope can start to decrease. " So I thought she was referring to her first two (most-concentrated) standards. My mistake, so, yes - I totally agree with you that her dilute points (where primer dimers and/or other non-specific amplifications can influence Cq) has flattened out the curve on the dilute end - causing the Cq values to not respond to dilution correctly due to the extra contaminating signal present in that region (whether primer dimer and/or non-specific and/or contaminating DNA or all 3) ... but, then, seeing the melt curve seems to contraindicate all of this. So it's back to Jochen's statement (which I now believe is closest to the truth):
"This "broken line thing" is a Roche "feature". They think it's easier to ignore inappropriate background correction and instead adopt the calibration curve to the (erroneous) ct values, hoping the errors are systematically similar for standrad and unknowns." I'm with Jochen on this.
Thank you so much everyone for your honest and insightful opinions. Jack and Jochen I'm in the process of re-doing my normal PCR so i can run an agarose gel, if i don't get distinct bands I'll extract RNA again and see where it takes me. I'll keep in contact with you on my findings.
Pontsho, please run the real-time PCR, NOT a normal PCR and check the products from the real-time PCR on a gel. You must have all the "real-time PCR conditions" (buffer & cycling times) to see what happens in the real-time PCR.
Oh okay...but I'm still using the old light cycler machine that uses capillary tubes and not a well plate. How do I transfer the RT-PCR product from the capillary tubes? Its impossible with a pipette.
Put the open capillaries upside-down into an Eppendorff tube and centrifuge gently (1000 rpm, not much more). This works perfectly fine.
Please take care of contaminations! Do these things *NOT* in the lab where you pipet the PCR reactions!!! The problem is less severe if you use an UNG mix, but I would anyway be careful.
Maybe you could use a different dilution factor in order to get enough points on your standard curve, and avoid the two 'problematic' concentrations?
- Badges
- Science topic