I hope somebody is able to give me suggestions.
After treating a 50:50 mix of fully methylated (FM) and non-methylated (NM) commercial DNA (zymosearch) I amplified the locus I want to study. Once excluded any PCR bias - i.e. by digesting the PCR product with a restriction enzyme that cut only the FM "allele" and confirming an almost equal amount of the two products - I proceed with TA cloning (pGEM + T4 DNA ligase Promega). The ligation product has been used for transforming DH-5a cells. When we checked colonies (nearly 20) to confirm that the 50:50 proportion has been mainteined, we observed instead a 70:30 (FM:NM). This situation has been already reported (Warnecke et al. / Methods 27 (2002) 101–107) as possible.
Do any of you experienced a similar situation and how have you fixed it? Changing maybe ligase or cells?
I thank you in advance
I'm cloning bisulfite treated DNA, in particular a 50:50 mix of fully methylated and non-methylated commercial DNA (zymosearch)
Hi Gianluca
I hope that I have correctly understood
Summarizing, you did a 50:50 mixture of methylated and unmethylated DNA, then you did a PCR for methylated DNA (I assume you used oligonucleotides for the BSP method), from there you cloned and sequenced, am I okay?
Through these methods it would not be possible to observe post-cloning a 50:50 methylation ratio, because there are three steps in cloning that make the whole process random. First, despite confirming that there are the same amount of methylated and unmethylated DNA amplicons, it is random that the ligation in the vector is also 50:50; second, it is also random that competent cells capture half of vectors with methylated and unmethylated DNA; and thirdly, the selection of the colonies is also random, there is nothing to ensure that you choose 10 colonies that having the vector with methylated DNA and 10 colonies that having the vector with unmethylated DNA.
Thus, one question comes to mind, is not it clear to me why you using the BSP method to measure the global methylation of a gene? If you use a mix of methylated and unmethylated DNA from Zymo as a control, would not it be better to find for the relationship through High Resolution Melting?
Dear Pedro, you correctly assume I'm doing a Cloning-based BSP.
I do not perform HRM because I have not access to such an instrument at the moment, so I've tried to reach the same result using an alternative approach.
Concerning instead your observation that it would not be possible to observe post-cloning a 50:50 methylation ratio, I disagree and this because, as you correctly say, there all the steps in the cloning process are random. Just to explain what I intend let me make an example. As you know, when we flip a coin (assuming there is no major difference between the two sides) we expect in half of cases for heads, and 1/2 for tails and this is particularly true when we did a very large number of trial flips. This is exactly what happen in the cloning experiments when no bias occur. In the ligation phase every vector can ligate either the fully methylated allele or the non methylated one, and if this occurs a large number of time (and this exactly what happens) in half of cases we got one insert (FM) and in the other half the other one (NM). This will be exactly the same for the transformation process in which every bacterial cell can capture either one or the other vector. Being the number of transformed cells very high we expect that in half of cases a cell will inglobate the FM construct and in the other 50% of cases the NM one. The only process that could in theory be a bias is the selection of colonies. Indeed we picked up just a limited number (about 30). But repeating the experiments twice I got the same results.
That's why I think there is a cloning bias
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts