Thank you for the kind words! We definitely overlooked that association in humans. Cool that CDCA7/CDCA7L variants are also associated with increased expression and CG methylation. Mechanistically, I'm curious to see how the CG-dominated function of plant CDCA7 will compare in animals!

Thank you Olivier!

Thank you James

This paper coming out today is a beautiful way for me to close up the #EMBOmobilegenome in Heidelberg. Great crowd, great talks, great things ahead! Shoutout to the organizers for putting this together 👏

#TEworldwide #TEsky
#Epigenetics #plantscience

and the link to the paper:
nature.com/articles/s41...

Finally I want to thank all co-authors who contributed, the support at the GMI & VBCF, and reviewers who improved the paper.

doi.org/10.7554/eLif...
www.science.org/doi/10.1126/...
doi.org/10.1093/nar/...
doi.org/10.1126/scia...
doi.org/10.1126/scia...
www.biorxiv.org/content/10.1...

I believe our study brings important insights into CDCA7’s relationship to DDM1 / HELLS and should be relevant in the context of recent papers on CDCA7 (links below).

This model fits with two published observations:
- CDCA7 zinc-finger domain binds hemi-methylated CG sites (which are created at replication forks) and prefers nucleosomal DNA over linker DNA.
- While most eukaryotes have HELLS / DDM1, lineages that lost DNA methylation also lost CDCA7.

To conclude, we propose a model of epigenetic inheritance where DDM1 functions are temporally separated during DNA replication, and CDCA7 predominantly promotes mCG maintenance, with minimal contribution to other DDM1 functions.

The worldwide distribution of CDCA7α alleles suggests it may contribute to local adaptation. We found these alleles correlate with an altitudinal cline in the mountains of Spain and are associated with changes in seed size, an adaptive trait for plants.

In natural populations, a 26-bp indel in the CDCA7α promoter acts as a rheostat, tuning its expression. This polymorphism affects the level of CG methylation and transposon silencing across the entire genome.

Quantitatively, CDCA7 disproportionately impacts mCG, indicating the predominant function of CDCA7 is to promote DDM1-dependent mCG. An important detail here is that we used ratios of mutant / WT to objectively compare methylation contexts.

We also mapped the physical interaction, showing the N-terminal domain of CDCA7α binds directly to the H2A.W-binding domain of DDM1.

The absence of CDCA7 has a weaker impact than DDM1 mutations, showing DDM1 can partially act without CDCA7. Conversely, we provide lots of genetic evidence that CDCA7 function exclusively through the DDM1 pathway.

We show that two A. thaliana genes, CDCA7α and CDCA7ß, redundantly contribute to all known DDM1 functions in heterochromatin maintenance. cdca7α/ß mutants phenocopy ddm1 mutants, affecting DNA methylation, histone variants and PTMs.

The Funabiki lab @HironoriFunabi1 has shown that Xenopus CDCA7 is a cofactor of HELLS / LSH (DDM1 in plants). While HELLS / DDM1 maintains both DNA methylation and heterochromatic histone variants, the function of CDCA7 remains unclear.
doi.org/10.1073/pnas...

Using CG methylation (mCG) of heterochromatic transposons as a trait for GWAS, we identified a major genetic association to an uncharacterized gene: a plant ortholog of vertebrate CDCA7.