Read on Twitter
*Is CTCF a master weaver of all genomes containing CTCF?*
*What is the molecular basis of gene insulation by CTCF?*
CTCF is required to form only ~10% of all contact domain boundaries in Drosophila. CTCF stably associates with a regulatory cofactor, Cp190.
Recap in threads
*What is the molecular basis of gene insulation by CTCF?*
CTCF is required to form only ~10% of all contact domain boundaries in Drosophila. CTCF stably associates with a regulatory cofactor, Cp190.
Recap in threads
1/8 Neurons are the only cell type in which CTCF is critically required for fly viability. (Movie: CTCF0 mutants never hatch from the pupal case, but they are brought back to life by expressing CTCF only in neurons.) So all experiments were done on CTCF0 central nervous systems.
2/8 @erezaterez made Hi-C possible on CTCF0 brains.
CTCF was bound at 8% of all contact domain boundaries.
Boundary defects in CTCF0 mutants were specifically observed at former CTCF peaks.
CTCF was bound at 8% of all contact domain boundaries.
Boundary defects in CTCF0 mutants were specifically observed at former CTCF peaks.
3/8 Fly and human CTCF N-termini are highly (!) diverged except for 10 amino acids that bind to cohesin in human cells (Li et al. 2019) and in flies (see gel). But consistent with Rowley et al. 2017, we found that CTCF-dependent boundaries lack clear directionality.
4/8 In CTCF[0] mutant central nervous systems, some genes near CTCF-dependent boundaries are misexpressed in various patterns – below, in a pattern consistent with loss of insulation from putative neuronal enhancers.
![4/8 In CTCF[0] mutant central nervous systems, some genes near CTCF-dependent boundaries are misexpressed in various patterns – below, in a pattern consistent with loss of insulation from putative neuronal enhancers.](https://pbs.twimg.com/media/EuDP9psWgAoVM4d.jpg "4/8 In CTCF[0] mutant central nervous systems, some genes near CTCF-dependent boundaries are misexpressed in various patterns – below, in a pattern consistent with loss of insulation from putative neuronal enhancers.")
5/8 We developed a quantitative insulator reporter assay and found that CTCF does not directly repress or activate transcription. Instead, CTCF insulates a promoter from an enhancer in a manner that scales with CTCF occupancy on DNA. (A-N = different CTCF-bound loci).
6/8 We found that CTCF stably associates with and recruits the boundary-associated factor Cp190 to CTCF-occupied boundaries. Partial boundary retention at some CTCF sites correlated with residual, CTCF-independent Cp190 binding in CTCF0 mutants.
7/8 We found that certain genes near CTCF/Cp190-bound boundaries are co-regulated by CTCF and Cp190. In Cp190 mutants, CTCF is still recruited to these sites – suggesting that Cp190 is required for CTCF function independently of CTCF binding to DNA.
8/8 In short, there is a profound difference in CTCF-requirement for genome folding in flies and vertebrates - in which a large fraction of boundaries are CTCF-dependent.
Fine collaboration with @erezaterez and all authors. https://www.nature.com/articles/s41467-021-21366-2
Fine collaboration with @erezaterez and all authors. https://www.nature.com/articles/s41467-021-21366-2
