Bio-orthogonal profiling of protein methylation

Protein methylation is an important biological process e.g. histone lysine methylation is involved in both gene activation (on histone 3 and Lys 4, 36 and 79) and silencing (H3 Lys 9 or 27 and H4 Lys 20). But how can other protein methyltransferases (PMTs) be investigated? Genetic approaches are always useful, but can have limitations particularly if your protein isn’t very common, forms complexes with other proteins (which would be disrupted by its absence and cause phenotypic changes beyond protein methylation) or if your process results in a non-viable cell. Chemical approaches would be extremely useful and complimentary, but as always problems with selectivity have to be overcome; a particular problem with PMTs as SAM (S-Adenosyl methionine) is a particularly prolific enzyme co-factor.

What to do then? Islam et. al. have developed a rather elegant solution that could allow general examination of a variety of PMTs, by creating mutants of proteins that can accept a synthetic azido-SAM donor. The enzyme then tags its target as before but instead of a methyl an azide containing group is left behind, this can subsequently be reacted with tags (like biotin) which contain strained alkynes in a bio-orthogonal manner. 

This type of mutation-induced-selectivity approach to small molecules is becoming increasingly common and has already been used by other groups (particularly Shokat’s work to use mutation of Kinases to allowing the use of highly selective small molecule inhibitors). Islam’s work uses Shokats approach, but in a subtly different way. By ensuring the selectivity of their SAM analogue in donating azides, bio-orthogonal chemistry can then be used to reliably unpick the targets and potentially the use of a variety of markers such as fluorescent compounds, potentially of different colours; biotin, etc. Maybe this could mean that (as is discussed in the paper) the “methylome” of the protein can be potentially obtained?  Indeed the authors report that new targets for their modified enzymes (EuHMT1 and 2).

Hopefully this type of approach can be used more generally for other methyl transferases. The acid test in the long run will be if more than these two enzymes can be mutated to selectively accept only the azido SAM. Also if the bio-orthoganal approaches can be expanded as well such as in-situ labelling (like Bertozzi has demonstrated in Zebra fish embryos) it could be an even more powerful technique, although maybe I am bit naive in thinking that bit is possible!