I think you need to do the mutation in both strands,no?. If not you will need to use one primer with the mutation and one without, to allow amplification of both strainds but you need to be carefull with that because that will produce a mismach in the DNA and the E.coli MismachRepair System will try to repair it. You could try with two primers and transform the DNA in to a MRS defective strain. But that mismach will be unestable also in eukaryotic cells. And you will need to sequence after each experiment to be sure that you still have the mutation.

Hi Domingo,

Thanks for your message. I want to test effects of non-CpG based promoter methyation(i.e. meC(A/T) which are usually not pallindromic) on the gene regulation. Thus methylation on only one strand. I have tried using both single (with methylation group) and double primer (+other primer without any mutation), however, I had no success until now. Any suggestions on PCR conditions would be helpful.

meanwhile, you mentioned about MRS defective strain. I suppose you mean, dcm- strains?

Why do you need a mutation in only one strand? Surely that will be corrected by your cells, so you end up with lots of cells without the mutation, and some cells with the mutation in both strands? I understand the methylation is needed in only one strand. I understand the sequence is non-palindromic - so if you introduce the required mutation in the antisense strand, and the matching complementary mutation in the sense strand, that would seem more efficient (say, 5-GAGATT-3 > 5-GACATT-3 on the antisense strand, so 5-AATCTC-3 > 5-AATGTC-3 on the sense strand). That way, you could use the same, or a tweaked, protocol you used earlier for mutagenesis of two strands, What am I missing?

Hi Marjolein,

I need methylated base on only one strand because of so called non-CpG based DNA regulation. Dnmt3a have been shown to methylate sometimes only one strand.

My preliminary data suggests similar single strand methylation only on antisense strand (by bisulfite seq). Thus I want to validate if methylated base only on antisense strand can affect promoter activity.

I understand that, but methylation on only one strand does not mean you cannot introduce the matching (complementary) mutation on the other strand, does it? To mutate a T to a C (for example) so that you get methylation of that C, you would introduce an A -> G mutation on the complementary strand; this would not lead to methylation on that second strand, would it?

I think you can make your desired mutation in one strand, you just can not amplify it in bacterial. you need to find a system to amplify this. or find a way to do in vitro methylation on only one strand.

How are you assaying for your desired product after the PCR mutagenesis?

I think you try the following PCR condition i.e. thermal cycling: (90C 1 min 55C 1 min then 55C 15 min) x 24 55C 15min. Primers should be 40pg/ul - 3ul; Plasmid DNA should be around 100 ng/ul - 1 ul (you may safely increase it to 3ul and see). 1 ul of Pfu enzyme (check the activity and purity of this enzyme. If required open a new vial). 1ul of 10mM dNTP should be sufficient. Rest should be as per your second protocol (i.e the buffers etc). While running the electrophoresis check the temperature of the buffers ( make sure that these do not heat up to a large extent). If it does then you may need to chill these by adding fresh chilled buffers. Try these and let us know your outcome. I believe it should work .

Hi Roshan,

Since you are talking about a luciferase gene after the promoter I assume that you're trying to do in vivo assays.

As suggested previously, DNA methylation in only one strand can be make with proper oligonucleotides that contain the desire modification, but if you clone the PCR product in a living cell, you might end up either without any methylation or with methylation in both strands.

I also assume that you have a TF to check with this promoter. So instead of going directly to in vivo, I'd make a test in vitro by EMSA with the methylated and not methylated version, and methylated in one strand or the other. Directly from the PCR product (you can use SYBR Green as marker for the DNA shift in the gel).

In the case you want to treat whole cells and see different luciferase luminescence depending on the treatment and the methylation status of the promoter I would suggest a different strategy. I would cotransform cells with whatever candidate (DNA-glycosilase, demethylases, etc) that would modify that promoter making it either methylated and not methylated. Thus, you could see how important the methylation of that promoter is based on luciferase transcription.

With more info I might be able to help better ...

Good luck!

Wich is the site you want to methylate? Because If you want to perform a non CpG methylation It shuoldn´t bother you to have the right complemetary base in the other DNA strain. As long as your meC is not next to any G, there will be no complementary Cytosine to methylate on the other strain.

Consider that a missmatch in any cloning step will be lost as soon as the bvacteriar replicates the plasmid. The same will probably happen if you use your missmatched product to transform cells in a stable way.

May be you could explain wich is the sequence you want to mutate and in what system you want introduce it?