Protocol Online logo
Top : Forum Archives: : DNA Methylation, Histone and Chromatin Study

why i am not luky to find any potential tumor suppressor by epigenetic study - (Mar/13/2007 )

we just finish genome wide profile of DNA methylation of liver cancer,
by pyrosequence we identified dozens tumor specific hypermethylated gene ( we screen 30 patient sample, and more than 10 has hypermethylation in tumor but no hypermethylation in normal and proximal tissue)
by RT-PCR, we further found several gene is both hypermethylated and silenced in tumor cell line; now, the candidate is only 3. they are all novel genes, 1 contain high mobility domain, one cantian ?, i forget, that domain is also releated to transcritpion factor.

then we did functional assay, construct expression vector and transfect them into tumor cell line which lack this gene expression, we are expecting some inhibitory effect. but we are disapointed to find: no growth inhibiton; no apoptosis induction after transfection,

what should we do? we screen 6000 gene at begining, then at last we got 0. perhaps epigenetics silening is not important mechanism to silence tumor suprresor, is there some better way to identify tumor suppressor gene?

should i give up these gnes? they are just incidently hypermethylated and silenced, or by-stander effect, but they themselves do not have any important function?


A tumor cell lines may contain additional mutations that obscure the importance of your genes being methylated. In many cases tumor suppressor methylation is an early event in tumorigenesis that may increase the change of additional mutations. Once these mutations may occurred, the methylation may no longer be required.


Congratulations on the amount and the quality of the work you have done. That is the strategy many people want to pursue including myself to find some TSGs. But a fundamental question regarding DNA methylation is unanswered: Why are genes methylated in cancer? Without a clear answer to this question, any efforts trying to identify TSGs or biomarkers by methylation profiling are likely to end up with nothing.