Back

Mechanistic dissection of the CHD4 enhancers reveals cooperative functions among the homotypic ZNF410 clustered motifs

Xu, S.; Ren, R.; Peng, C.; Lu, H.; Ly, L. C.; Crossley, M.; Xu, B.; Cheng, X.; Blobel, G. A.; Lan, X.

2024-02-25 molecular biology
10.1101/2024.02.24.581901 bioRxiv
Show abstract

Transcription factors often regulate numerous target genes. However, ZNF410 controls only a single gene, CHD4, in human erythroid cells by its highly restricted chromatin occupancy to the CHD4 locus via two clusters of ZNF410 binding motifs. Here, we uncover that ZNF410 controls chromatin accessibility and activity of the two CHD4 enhancer regions. Combining CRISPR/Cas9 genomic deletion with CRISPRi approaches, we demonstrate that both enhancer regions additively contribute to CHD4 gene expression. Mutations of the di-adenine nucleotides within the ZNF410 binding motif fully disrupt ZNF410-DNA interaction. Luciferase assays, ChIP-seq, and ATAC-seq studies reveal that the homotypic clustered motifs within the CHD4 enhancers are recognized by ZNF410 in a collaborative fashion. Especially, the three ZNF410 motifs located in the 3 end of the distal enhancer act as "switch motifs" to control the chromatin accessibility and activity of the whole distal enhancer region. Together, our findings expose a complex functional hierarchy of motifs, where clustered motifs bound by the same transcription factor cooperate to fine-tune the expression of target gene. HighlightsO_LIZNF410 maintains the accessibility of the CHD4 enhancer regions C_LIO_LIThe two enhancer elements additively modulate CHD4 gene expression C_LIO_LIThe dinucleotide "AA" is the key bases within ZNF410 binding motif C_LIO_LIZNF410 binds to homotypic clustered motifs in a cooperative manner C_LI

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.