methylation in patient samples - (May/23/2006 )

Dear Methylation experts,
Hi there! I have a rather a very basic question regarding bisulfite conversion.Is unconverted sequence always supposed to be due to methylation. Can DNA quality from patient samples with various kinds of cancer also govern the appearance of methylated band even though it is much fainter than the strong unmethtylated band?The pbl control and universal methylated DNA are given bisulfite treatment as per the Frommer protocol and give predicted bands. However how to be sure that the faint methylated bands visible in patients are indeed for real and not because of poor quality/ tight 3D structure or incomplete denaturation. We follow Frommer et al 1992 as psoted in the MD anderson methyaltion page which involves 10 minute denaturstion in 2M NaoH before strating the bisulfite treatment.Thanks in advance for all queries . It will help us a lot in interpreting our results.
S.S

incomplete conversion could lead to what you are seeing, and indeed protein contaminants can give rise to incomplete conversion, see Warnecke et al, 2002 with regard to bisulfite conversion artefacts. I haven' got the complete reference for you, but a pubmed searh will reveal it.

Nick