Celebrating with Julia Schaepe, William Greenleaf, and all other authors!

@stockholm-uni.bsky.social, @scilifelab.se, @vetenskapsradet.bsky.social, @strategiska.bsky.social
(6/6)

🧬We do not need to invoke highly cooperative biological state changes, such as phase separation or condensation, to describe 𝘪𝘯 𝘷𝘪𝘷𝘰 transcription factor binding in our observed contexts. In sum, we can understand chromatin states 𝘪𝘯 𝘷𝘪𝘷𝘰 with “classical” and interpretable biochemistry. (5/6)

🧬 A thermodynamic model with 𝘪𝘯 𝘷𝘪𝘵𝘳𝘰-derived affinities predicts 𝘪𝘯 𝘷𝘪𝘷𝘰 chromatin states in human cells across KLF motif grammars. Our models predict chromatin states with few physical and interpretable free parameters, in contrast to highly parameterized “black box” deep learning models. (4/6)

🧬 Motif flanking bases that KLF1 does not make direct contacts with in the crystal structure drive ~40-fold affinity variation. Distal flanking sequences helical turns away (≥10.5 bp) from the core motif tune affinity ~2-fold, in a manner consistent with sequence-dependent search of the motif (3/6)

🧬 Differences in DNA binding affinity are driven by differences in motif recognition. This is the same binding mechanism that we previously found for the bacterial TF l𝘢𝘤 repressor, which is structurally dissimilar to KLF1, implying a universal mechanism across domains of life and TF types. (2/6)