Ruler elements in chromatin remodelers set nucleosome array spacing and phasing

Abstract

Arrays of regularly spaced nucleosomes dominate chromatin and are often phased by alignment to reference sites like active promoters. How the distances between nucleosomes (spacing), and between phasing sites and nucleosomes are determined remains unclear, and specifically, how ATP dependent chromatin remodelers impact these features. Here, we used genome-wide reconstitution to probe how Saccharomyces cerevisiae ATP dependent remodelers generate phased arrays of regularly spaced nucleosomes. We find that remodelers bear a functional element named the 'ruler' that determines spacing and phasing in a remodeler-specific way. We use structure-based mutagenesis to identify and tune the ruler element residing in the Nhp10 and Arp8 modules of the INO80 remodeler complex. Generally, we propose that a remodeler ruler regulates nucleosome sliding direction bias in response to (epi)genetic information. This finally conceptualizes how remodeler-mediated nucleosome dynamics determine stable steady-state nucleosome positioning relative to other nucleosomes, DNA bound factors, DNA ends and DNA sequence elements.

The Active Rather than Intrinsic Mechanism of Nucleosome Depletion by Poly(dA:dT) Tracts

“…This allows direct comparison of occupancies across conditions. We compared on the one hand ODM-seq data for S. cerevisiae in vivo chromatin, averaged over five replicates, as well as for SGD reconstituted chromatin along the S. cerevisiae genome at two different nucleosome densities with on the other hand published MNase-seq data of in vivo chromatin and for two replicate sets of SGD chromatin at three different densities each [62]. As ODM-seq reports nucleosome occupancy over the whole nucleosome length, we plotted MNase-seq data as 147 bp extended dyads, too.…”

“…MNase-seq data were plotted as bp extended dyads. Nucleosome density is quantified as described in [17,62].…”

“…This argues that promoter NFRs over poly(dA:dT) tracts are actively maintained by these factors and we suggest that this is the main reason why these factors are essential. Other non-essential remodelers, like the INO80 or the ISW2 complex, also participate in shaping promoter NFRs as NFR borders are affected in their respective deletion mutants [70,74,80,81] and in in vitro reconstitutions [39,62,64]. Further, NFRs with or without poly(dA:dT) tracts are assembled into nucleosomes in the wake of replication and only later become nucleosome-free again [82][83][84].…”

“…This means that they are able to read out DNA sequence and other information, e.g., distance to barriers like GRFs, and process this information together with their own input into remodeler-specific nucleosome organization [39,64,98]. The remodeler-specific information was conceptualized as a "remodeler ruler" [62,99]. Especially for the INO80 complex, the modules that make up the ruler and are involved in sensing and relaying DNA sequence information, in particular DNA shape features [100], were recently defined in structural terms [62,64].…”

“…The remodeler-specific information was conceptualized as a "remodeler ruler" [62,99]. Especially for the INO80 complex, the modules that make up the ruler and are involved in sensing and relaying DNA sequence information, in particular DNA shape features [100], were recently defined in structural terms [62,64]. Also recently, the concept of intrinsic DNA bendability was revisited by a novel genome-scale cyclizibility assay for nucleosome-sized DNA fragments [101].…”