Bubble statistics and positioning in superhelically stressed DNA

TL;DR

This paper introduces a computational framework to analyze how superhelical stress influences DNA bubble formation, revealing increased bubble sizes and localization near transcription start sites, with implications for gene regulation.

Contribution

The study develops an efficient transfer-matrix method for genome-scale analysis of DNA denaturation under superhelical stress, combining the Zimm-Bragg and Benham models.

Findings

Superhelical DNA exhibits larger, more cooperative bubbles than unconstrained DNA.

Bubble localization is influenced by sequence disorder and is often near transcription start sites.

Genomic DNA has a higher probability of forming large bubbles under superhelical stress compared to random sequences.

Abstract

We present a general framework to study the thermodynamic denaturation of double-stranded DNA under superhelical stress. We report calculations of position- and size-dependent opening probabilities for bubbles along the sequence. Our results are obtained from transfer-matrix solutions of the Zimm-Bragg model for unconstrained DNA and of a self-consistent linearization of the Benham model for superhelical DNA. The numerical efficiency of our method allows for the analysis of entire genomes and of random sequences of corresponding length ($10^6-10^9$ base pairs). We show that, at physiological conditions, opening in superhelical DNA is strongly cooperative with average bubble sizes of $10^2-10^3$ base pairs (bp), and orders of magnitude higher than in unconstrained DNA. In heterogeneous sequences, the average degree of base-pair opening is self-averaging, while bubble localization and statistics are dominated by sequence disorder. Compared to random sequences with identical GC-content, genomic DNA has a significantly increased probability to open large bubbles under superhelical stress. These bubbles are frequently located directly upstream of transcription start sites.