Methylation detection

Alex Dobrovic adobrovic at
Thu Oct 23 02:30:07 EST 1997

You can't just use any primers after bisulfite treatment as it modifies the
Cs in the primer. Primer design is accordingly quite complex especially as
the 2 strands of the DNA are no longer complementary after modification.

>"Dr. Duncan Clark" <duncan at> writes: > In article
><6154gl$hgb$1 at>, Paul Rohde
>> <P.Rohde at> writes
>> >Sorry for my ignorant responce, and I don't know if it is applicable
>> >to you, but there is a chemical method that transforms the methylated
>> >base from one type of base to another.
>> >
>> >So you PCR half your sample and sequence it.
>> >On the other sample, you treat it, PCR and sequence, then look
>> >for the base change differences (transformed bases).
>> Good idea, the only snag is that all PCR products are always
>> unmethylated unless you give the reaction methylated nucleotides. You
>> therefore automatically lose whatever methylation you started from.
>Hi Duncan!
>I don't want to be in a position where I have to defend something I said
>about a subject area that isn't mine.  I just heard it at a seminar
>a fair while ago.  But still, you missed the point!!
>Yes, its true that the PCR product is unmethylated, but I didn't say to
>measure the methylation on the PCR product!!  Read what I said again.
>Thankfully, Aimee answered me -- I wasn't to help her, but the little she
>fed back to me will help me now to re-explain to you.  Again:
>A)  You have you DNA sample that is methylated somehow.
>B)  Take half of this sample and treat it by bisulphate modifications.
>    This transforms any methylated C to T.
>Ci)  PCR this modified sample for sequencing
>Cii) On the other half of the DNA sample that hasn't been treated
>     by bisulphate modifications, just PCR this for sequencing.
>D)  Compare the two sequences.  The "untreated" sequence will be the
>    true DNA sequence.  The "treated" sequence could show some differences
>    ie some T now exist that have replaced a C compared to the "untreated"
>    sequence.  These Ts show the Cs that were methylated.  Hey Presto!
>Duncan, you must of read my posting way too late in the night. ;-)
>If any body knows more about specifically methylation detection in
>A/T rich genomes (I'm not sure what this implies really) please post
>something because Aimee sounded pretty desperate!
>> Duncan
>> --
>> The problem with being on the cutting edge is that you occasionally get
>> sliced from time to time....
>> Duncan Clark
>> DNAmp Ltd.
>> TEl/FAX 01252376288

Alexander Dobrovic, Ph.D.
Chief Scientist
Department of Haematology-Oncology
The Queen Elizabeth Hospital
Woodville, SA 5011, Australia

Affiliate Senior Lecturer
Department of Medicine
University of Adelaide

Tel 61-8-8222 6884
Fax 61-8-8222 6046

More information about the Methods mailing list