A genome-wide map of CTCF multivalency redefines the CTCF code

The “CTCF code” hypothesis posits that CTCF pleotropic functions are driven by recognition of diverse DNA sequences through combinatorial use of its 11 zinc fingers (ZFs). This model however is supported by in vitro binding studies of a limited number of sequences. To directly test CTCF multivalency in vivo we here define ZF binding requirements at ~50,000 genomic sites in primary lymphocytes. We find that CTCF reads sequence diversity through ZF clustering. ZFs4-7 anchor CTCF to ~80% of targets containing the 20bp core motif. Non-conserved flanking sequences are recognized by ZFs1-2 and ZFs8- 11 clusters, which also stabilize CTCF broadly. Alternatively, CTCF employ ZFs9-11 to associate with a second phylogenetically-conserved upstream motif at ~15% of its sites. Individually, ZFs increase overall binding affinity and chromatin residence time. Unexpectedly, we also uncover a conserved downstream DNA motif that destabilizes CTCF occupancy. CTCF thus associates with a wide array of DNA modules via combinatorial clustering of its 11 ZFs. Overall design: ChIP-Seq of wt and mutant CTCF in activated mouse B cells. For the comparison of wt and mutant CTCF binding profiles, CTCF fused to a biotag was overexpressed in the presence of the endogenous wt CTCF copies together with BirA. The exongenous CTCF was then precipitated using streptavidin beads. For the comparison of wt CTCF binding in Mus musculus C57BL/6 and Mus spretus, endogenous CTCF was precipitated with an anti-CTCF antibody. All these experiments were carried out in triplicate. In addition, several overexpressed mutants were subjected to exonuclease treatment prior to sequencing (ChIP-Exo).