designing primer for methylation - (Apr/23/2006 )

Hi ALL

I am new to this field & has lots of questions. One major question is: I have found that the gene of my interest has a CpG island in it. So when designing a primer for MSP, should I include the sequence upstream of the CpG island to design a primer. If so how much length should I include from the upstream sequence...

Does anyone have any experience in designing primers for methylation analysis using a light cycler.

Any advice will be appreciated

---

First thing is to check if someone has already done MSP on your gene of interest. if not, then for MSP you would need to design primers within the CpG island. MEthprimer is a good start but is not the most ideal program that designs good primers. I would suggest you try a couple of the primers selected with this.

Typical sizes are below 200bp in length, but methprimer will do this for you. As for lightcycler, it is very similar to the ABI RT machines so there is no problem on that front.

Nick

---

Hi,everyone.
I am still new to this field,though I have been starting the MSP about FHIT gene for months.As what you nick had said above,I design the primers with methprimer.I used the DNA of MOLT-4 cell line as the positive control,and the PBMC from healthy people as the negative.I can get both methylated and unmethylated band in the PBMC.I don't why.As we know ,usually the healthy people are unmethylated,how the mehhy-band appear?I exclude the contamination,starting with another PBMC,and it was still there.IS there something wrong with my primer.I design the primer with methprimer,it should have worked well like the others.Could someone help me ,please?Much thanks.
jim

---

Hi Jim,

I'm not familiar with the FHIT gene. But I'm not sure if you can assume that it is normally unmethylated. To my understanding many genes are normally methylated and if by chance they become unmethylated, they become overactive and cancerous. On the other hand a number of genes, especially tumor suppressor genes are unmethylated. If they happen to get methylated and therefore inactivated, then one may be more susceptible to cancerous processes.

Unless it's documented somewhere before, I don't know if you can assume the methylation status of a particular gene such as FHIT.

---

Hi,Purplefetus.Thank you for your reply.
Well,FHIT gene is a tumor suppressor gene.As what you have said above,tumor suppressor genes are normally unmethylated,in healthy people.I donot know why I often amplified the methy-bands in healthy people' PBMC DNA.I am sure that there is no contamination when extracting the DNA.How do you design your primers?By methprimer?Have you runinto the same promblem?
jim006

---

Jim,

are PBMC, peripherial blood monocytes? if so, you may well have in the normal samples, a subpopulation where your FHIT is methylated, however, there are not enough to give rise to the phenotype, but are sufficent for a PCR reaction to pick up, MSP is very sensitive is picking up 1/10th of the population of DNA that is methylated and much less than that also I suspect.

How many cycles are you performing for MSP? if over 30 I suggest you lower this significantly, as saturating the MSP with the number of cycles also results in false positives.

Good luck, nick.

---

Hi Nick,

Thnx for ur advice. As far as my genes are concerned, no one has done any metylation analysis on them. But I choose them based on some previous microarray data, in which they showed differential expression or inactivation. I am not sure if it is ok to do it this way....

We have a Roche Light cycler in our lab. Can I use the Bisulfite primers which I get using the methprimer to do a reaction in the light cycler using melting curve analysis to distinguish between methylated & unmethylated samples ? I read in one paper that methylated & unmethylated strands will have different melting temperatures. If I am currect, then the primers for bisulfite sequencing does'nt distinguish between methylated & unmethylated strands, so it is wise to use the same primers for light cycler ?

---

Sure that is how we are doing things also, look for expression changes and based on the assumption expression is regulated by DNA methylation, then go ahead and look for methylation differences!!



Certainly possible as the GC content of methylated and unmethylated templates would be different implying their melting point would also be different. I have not tried this myself, but I would say, yes it is possible to do so this way, however, you would need to have the appropriate controls run at the same time as well.

let me know how you go with this experiment!

Nick

---

hello,i am new too.so i have some questions and hope have answers from you .if the gene have not be done ,can i do it by msp?beacause i do not know if the C of CpG is methylation, and i do not know which CpG maybe methlation in unmomal tissue ,so how i design the primer? if i use the online designer,it gives me primers and the primer include 3 CpG ,if 2 of them methylation,the two couple primer is not useful,isthem?

---

Thanks,nick.

Yes,PBMC are peripherial blood monocytes.I run a nested MSP.Both cycles are 35.
I will try your advice,lower to 25 cycles in the second round.IS that OK?Thank you very much.

---