Wondering whether the use of DNA from saliva (or non-tissue-specific DNA in general) is generally considered to be acceptable for methylation analysis. I'm interested to look at this within the context of cognitive traits, learning abilities, etc.
There are large differences in distribution of DNA methylation and hydroxymethylation between different tissues in mammals, and even between cell-types in the same tissue.
I would strongly discourage such experiments unless there are really clearly established links between the tissue you are investigating and the tissue your are basing your research questions on.
The only benefit of using saliva would be if you're specifically looking for a saliva-based diagnostic. Like Erick says, there are large differences in DNA methylation between tissues. I expect the task of finding an association between cognitive traits and DNA meth in saliva will be considerable and positive results (if found) difficult to interpret. For example, a deficiency in folate might effect cognition and cause sufficient hypomethylation of some genes such that worse cognitive abilities correlate with the hypomethylation of gene X. You may infer that epigenetic regulation of gene X is biochemically linked with cognition somehow, when in fact, hypomethylation of gene X is just a proxy for poor nutrition.
http://www.ncbi.nlm.nih.gov/pubmed/12611557
It's definitely becoming a possibility, check out this article:
Schizophrenia brain methylation genes overlap with over 100 genes from a previous DNA methylation study of peripheral blood (http://www.nature.com/tp/journal/v4/n1/full/tp2013111a.html)
I used to be much more skeptical of comparing methylation across tissue because of tissue specific DMRs, but if an event altering methylation occurs in the developmental precursor cells, comparing across tissues could be effective for diagnosis.
Another advantage of saliva is that buccal cells and neurons are pretty closely developmentally related since both come from the ectoderm, so while they have their differences, they also have some similarities that I wouldn't expect to see across other cell types.
I tend to agree more with Rasmussen and Ross, in that DNA methylation is highly tissue and even cell-specific; so one should not be surprised by negative or no associations and even positive ones should be taken as chance or non-causal.
Regarding genetic 'imprinting' at the zygote or subsequent early stages, it could be a possibility, but it is well known that the entire genome does not take part in one disease such as Schizophrenia or any other; rather it is the specific tissues and organs that relate to expression problems. So even if the original methylation patterns were preserved in all cells (which i doubt) they would not be showing associations of any value, as the genetic activity in those non-relevant tissues would not take part in the 'functional' genotype expected in relevant tissues. However research is a world full of surprises..!
I also support the opinions give here.
Simply doing a study on saliva without any indications you might expect some differences there will be difficult.
Obviously non-invasive diagnostics would be the ideal solution. If you would eventually want to look for DNA methylation markers in saliva I would definately also look at specific tissues or cell types in the brain first to screen for differences there and then try to figure out whether this can be correlated to changes in blood/saliva. A stand-alone experiment in saliva without insights from the brain I feel would be extremely challenging.
Good luck
Maarten
NXT-Dx
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Biology
- Biomolecules
- Nucleic Acids
- DNA