I am trying to figure out how to demonstrate that I cannot induce expression of a specific gene in a cell line. However, the gene is not expressed at baseline, so baseline Cts are very high. This makes it appear that the gene is inducible with treatment if I use the 2^(-ddCt) method, even when the treated samples also have a high Ct. For example I will have the change in Ct from 34 to 33 or from 40 to 34 despite consistent actin Cts. Basically, the expression is going from nothing to nothing. I know the primers are good because they are able to detect the gene in other cell types with more reasonable Cts (lower 20s). What is the best way to show this data so that it doesn't appear as though expression of the gene is being induced in any sort of meaningful way?
Ct-values above 35 cycles or so should not be considered as measurements. These are "non-detects". Report the numbers of non-detects. Do not use arbitrarily large Ct values to provide some "quantity" or amount of expression.
I agree with Jochen's statements. Also, did you do triplicates by any chance? I have used liquid handling Epmotion to setup PCRs for CFX96, I can tell you that even with liquid handling for Cts above 35, your PCR triplicates will not be tight. What can we do when we see Cts for a sample like this; 34,35.5,36. If you want to see how reliable the Cts are after 35, just find a sample that is around 35-40 and do a few replicates. That will give you an idea of the variation. Remember 1-2 cycle difference is 2-4 fold change difference.
Honestly, it's great that you're already aware that Ct values of 35+ represent "not really there": this is something many people do not really appreciate (and like you say, converting things to 2^ddCt completely masks this).
PCR is an incredibly sensitive technique, with a potential lower limit of detection (LLOD) of "one target molecule", which is usually Ct ~35 or so.
You can't _quantify_ that easily, though, so the LLOQ is usually substantially higher (10+ molecules), and even then it's sketchy.
And of course, you need to put all this into biological context, which you are (this is good).
If your gene is barely detectable (which it appears to be), then just saying "not meaningfully expressed" should be fine. If you want to be really specific, then "not meaningfully expressed (Cq > 34)" would also be fine, and would have the benefit of conveying to informed readers that yes, your gene is not meaningfully expressed, and to less-informed readers that Cq > 34 corresponds to 'not meaningfully expressed'.
qPCR is a wonderful quantitative technique, but just because it spits out numbers, you don't have to use them. You are allowed to look at the numbers and conclude "these are consistent with no meaningful expression".
One thing to John's post, not particularily to the problem as stated here, but possibly relevant to others:
If mixtures of different cell tyes (e.g. organ or tissue homogenates) or cell states (adhering, migrating, proliferationg, binding to a specific partner cell,...) and the target gene is expressed specifically in one of the cell types or in one of the cell states, and if this cell type or state is rare, then the average expression may be not quantifiable, but the few cells that do express the target gene may nevertheless express it in biologically meaningful amounts.
So just take care: it's not always the case that "I was unable to measure / quantify gene expression" means "there is no meaningful expression". It depends a bit on several details of your experiment.
- Badges
- Science topic
- Similar topics
- Filing