The actual bisulfite treatment - protocol and troubleshooting (Nov/18/2004 )

Hey there Ellen,

All is not lost, if there is a will there will always be a way .

Okay so you should choose a plasmid that contains HpaII/MspI sites (ie: CCGG). To check if your SssI methylase experiment worked, digest the treated plasmid with HpaII. HpaII is methylation sensitive and therefore you will not get digestion. You can also digest the same plasmid with MspI which is methylation insensitive and you would expect full digestion of all sites within the plasmid. Of course if you perform the digests with untreated plasmid (isolated from DAM- bacteria) then it will digest with both HpaII and MspI. (Of course you can use other methylation sensitive isoschizimers if you choose).

You do not have to perform sequencing of your PCR amplicon post bisulfite treatment. A neat way of checking if your positive control has been fully converted is to digest your amplicon with restriction enzymes. Again like above, if you were to digest your amplicon with MspI you will get no digestion if your bisulfite conversion has worked because all CCGG sites will be converted to either TCGA (if methylated) or TTAA (unmethylated), So that is enough to say that your bisulifte has worked efficently, To check if your methylase worked assuming that it did, all CCGG sites would be converted to TCGA after bisulfite and PCR amplification. So if you were to digest with TaqI, these sites will be cleaved. (This is of course cheaper than to directly sequence).

Hope this helps you, I am sure your supervisor will be impressed with the way you are attacking the problem!

Nick

---

You do not have to perform sequencing of your PCR amplicon post bisulfite treatment. A neat way of checking if your positive control has been fully converted is to digest your amplicon with restriction enzymes. Again like above, if you were to digest your amplicon with MspI you will get no digestion if your bisulfite conversion has worked because all CCGG sites will be converted to either TCGA (if methylated) or TTAA (unmethylated), So that is enough to say that your bisulifte has worked efficently, To check if your methylase worked assuming that it did, all CCGG sites would be converted to TCGA after bisulfite and PCR amplification. So if you were to digest with TaqI, these sites will be cleaved. (This is of course cheaper than to directly sequence).


Hi nick,

I am new to both bioforum and BSP. I read about it here. I find that suggestions given by u, pcrman etc are really nice. i had a question while going through this suggestion given by u to Ellen. CCGG sequences if methylated, after modification will become UCGG and if unmethylated will become UUGG and after PCR TCGG and TTGG. But u mentioned TCGA and TTAA. Please correct me where i am wrong.
Thankyou

---

Dear Nidhi,

you are correct in saying within a MspI site (CCGG) if methylated will become UCGG and unmethylated will become UUGG AFTER bisulfite treatment.

However, after PCR all U's are replaced with T's. Therefore, methylated will become TCGG and unmethylated will become TTGG after PCR amplification, you would think.

BUT!!!! you also have to take into account the complementary strand of the MspI site:

5' CCGG 3'
3' GGCC 5'

post bisulfite (if methylated):

5' UCGG 3'
3' GGCU 5'


post bisulfite (if unmethylated):

5' UUGG 3'
3' GGUU 5'

Notice that the strands are now mismatched? therefore you need a strand specific amplification and I normally amplify the top strand so the site now becomes :

post bisulfite (if methylated):

5' TCGA 3'
3' AGCT 5'


post bisulfite (if unmethylated):

5' TTAA3'
3' AATT 5'

Green represents all Thymines replacing uracil after PCR and the blue denotes bases complimenting the thymines after PCR.
So the amplification "corrects" the mismatches because you are only amplifying only one strand.....otherwise if you perform a normal PCR to both strands I hope you can see that you will get gobbledegook!

Nick

---

Hi! Nick,

I am new member of bioforum.I was really interested in this discussion on bisulphite treatment..I am still not able to comprehend how CCGG sequences if methylated, after modification and pcr will become TCGA (methylated) and TTAA (unmethylated)..I completely agree with you that one has to do strand specific amplification and normal pcr amplification won't work... it will be really nice if you could explain it further..

Thankyou
smith

---

smith,

Lets say the primers you design are for the top strand,

because a strand specific PCR is performed with strand specific PCR primers only one of the primer pairs will anneal to the templates available.

In this case it will be the reverse primer annealing to the top strand and the polymerase will extend from this.

So the initial rounds of PCR are a linear amplification whereby the reverse primer is extended. Once the nascent "bottom" strand is produced from this extension, the site for the forward primer binding is now generated and the reaction will go into it's exponential phase.

It is during the process, the methylated HpaII site becomes a TaqI site.

If only I had a picture to show you this!

Nick

---

Apologies to all,

smith has pointed out a little mistake I had made.

for the unmethylated HpaII site it goes from CCGG to UUGG and thus after pcr amplification to TTGG so for the unmethylated sites it will look like so

TTGG
AACC

thanks for pointing this out smith

---

HI!
I am back with some more questions..

what are the mechanisms or signals that determine the methylation pattern?

Say where methylation has to occur in a particular cell type..

hey nick thanks for ur immediate reply for my question ...

Smith

---

Hi Smith,

>>what are the mechanisms or signals that determine the methylation pattern? Say where methylation has to occur in a particular cell type..
We know very little why methylation occurs to some genes, sequences and in some cells but not others. So far no consensus sequences that can direct methylation have been discoved and confirmed.

---

A problem with the bead protocol:
Hydroquinone is soluble at only 7g / 100ml, while the protocol which is described in this series of postings proposes making a solution of 110 mg/ml, which is above the solubility limit of hydroquinone. When the solutions are mixed, the hydroquinone will dissolve, but you have to mix them together with the solids if this is to work.

---

Thanks a lot, labtechie!

Your protocol is EXCELLENT!!!

I just followed it and got a product from three different tissues simultaneously. Actually in one of the tissues it was possible to see a very weak band after the first round (40 cycles).

Previously I mostly used original protocol (Clark et al.,1994) but results were very unpredictable. Then I tried to follow Olek et al., but failed. Somehow this last protocol in your interpretation (very detailed) resolved all the problems.

I would like to include a special reference on your protocol in a line with Olek et al.


mustamakkara

---