In vivo targeting of de novo DNA methylation by histone modifications in yeast and mouse

In this case, we would like to ask you to revise your manuscript swiftly by addressing the comments below and return your revision within four weeks of time (the sooner the better), in order for us to further consider publication.

1) The authors have to provide a clear rationale for their selection of DNMT3b only (instead of DNMT3a or both) to study de novo DNA methylation, since DNMT3a has been shown to have stronger enzymatic activity than DNMT3b in (Baubec et al. 2015), and DNMT3a (but not DNMT3b) is frequently mutated in tumors.

We expressed both murine de novo DNA methyltransferases in yeast, but we weren’t able to detect any expression both at RNA and protein levels of the murine DNMT3a in our transformed yeast strains. Moreover, whole genome bisulfite sequencing data of the DNMT3a strains do not show any level of DNA methylation. We do not understand the reason for why DNMT3a expression did not work, while DNMT3b did.

2) The authors need to discuss any differences of their work in comparison with Baubec et al. 2015.

Discussion of similarities and differences have been included in the Introduction and Discussion sections of the manuscript, as illustrated below.

Introduction, fifth paragraph: “Recently it has been reported that the PWWP domain is important in specifying the localization of DNMT3b in mouse embryonic stem cells (Baubec et al., 2015).”

Discussion, first paragraph: “It has been recently shown (Baubec et al., 2015) that in embryonic stem cells the PWWP domain is responsible for the targeting of DNMT3b […] and during a period of biologically important de novo DNA methylation (germ cells).”

Discussion, second paragraph: “In our yeast system we detected an anti-correlation between transcript levels and DNA methylation […] into the gene-body, shaping their intragenic DNA methylation distribution.”

3) The final section on mouse male germline DNA methylation is the least developed, but this should not prove difficult to improve. Many results and conclusions in the section “Correspondence between H3K36me3 and early DNA methylation in mammalian cells” are primarily supported by browser captures (Figure 6B–D) rather than quantitative plots. This is not appropriate. Furthermore, there is a paucity of summary information on the WGBS experiments. An intelligent reader will refer to this in supplementary data in order to judge the reliability of the data.

Figure 6C was modified: the previous genome browser snapshot (which was similar to that of 6D) was substituted with a quantitative plot showing that regions with high H3K36me3 levels at E13.5 are associated with the de novo DNA methylation increase occurring between E13.5 and E16.5. Data from WGBS experiments in mice (Pastor et al., 2014 and Seisenberger et al., 2012) were summarized in Supplementary file 1 panel G and Supplementary file 2 panel C, in the same manner as the yeast DNA methylation data. Also, Figures 6E and 6F are quantitative metaplots over all protein coding genes, which show global correlation of DNA methylation and H3K36me3 (6F) and anti-correlation with H3K4me3 (6E).