Greiner et al . reply—in our paper1, we presented several lines of evidence that chaetocin functions as specific inhibitor of SU(VAR)3-9. We had initially concluded that the disulfide warhead was not essential for chaetocin's inhibitory capacity on the basis of our observation of substantial inhibition of Drosophila melanogaster SU(VAR)3-9 by chaetocin, even in the presence of high concentrations of DTT1. Cherblanc et al.2 now revisit this issue in the context of SUV39H1, the human ortholog of SU(VAR)3-9.

We note that Cherblanc et al.2 find a IC50 for hSUV39H1 similar to that observed in our initial experiments. However, the availability of new compounds has enabled them to directly test our initial hypothesis. Their use of 2, lacking chaetocin's disulfides, indeed suggests that, in contrast to our initial hypothesis, the disulfide bond within chaetocin has a substantial role. However, it is interesting that a structurally divergent ETP model compound (3) containing the disulfide bond has a 30-fold higher IC50. This clearly suggests that chaetocin does not solely inhibit Drosophila SU(VAR)3-9 activity through the presence of its disulfide bonds (which 3 has as well) but also direct binding to the enzyme.

There is an apparent discrepancy between our data and the data of Cherblanc et al.2 with regards to the mechanism of inhibition that may be due to differences in assay conditions: we stopped the methyltransferase reaction after 5 min (as opposed to 1.5 h for Cherblanc et al.2) and did not preincubate the enzyme with the inhibitor. By doing this, we excluded the effect that chaetocin's disulfide bonds might have on the enzyme after an extended incubation and could observe competitive binding of SAM to SU(VAR)3-9. On the basis of our findings and the new data of Cherblanc et al.2,we assume that chaetocin binds the enzyme competitively and subsequently reacts with critical residues, therefore acting as a suicide inhibitor. It will be interesting to see this tested in a structural analysis. This mechanism would argue against the suggestion from Cherblanc et al.2 that SU(VAR)3-9's susceptibility to the compound is solely based on the difference to reactive thiols, a hypothesis that can and should now be tested in a rigorous manner.

It is also worth noting that we were aware of the potential for nonspecific effects in our original work and had to perform a careful titration of chaetocin to prevent its global effect on cell physiology (we used SL2 Drosophila cells), establishing a narrow concentration window that enabled us to analyze the effects of chaetocin on histone modification separately from its induction of apoptosis. As we only detected changes in H3K9 methylation and not at any other methylation site either by MS or by western blotting, we concluded that chaetocin specifically targets the class of enzymes known to methylate this residue (for example, G9a and SU(VAR)3-9). We therefore think that chaetocin, at least in this narrow concentration window, can be used to modulate H3K9me in vivo and is thus a useful compound for further investigations.