A dynamic view of DNA structure within the nucleosome: Biological implications
Graphical abstract
Introduction
The nucleosome core particle (NCP) is one of the most notorious DNA-protein complexes, as the fundamental building block of packaged DNA in eukaryotic cells. X-ray structures showed that the DNA in NCP (DNANCP) wraps ~1.7 times around eight histone proteins (two copies of H2A, H2B, H3, and H4) to form a super-helical double helix, as described in numerous reviews (Cutter and Hayes, 2015, Kornberg and Lorch, 1999, Luger, 2001, Luger et al., 1997, Luger and Richmond, 1998, McGinty and Tan, 2015, Zhou et al., 2018). DNANCP of different sequences and lengths (145 to 147 bp) adapts to highly conserved histone binding motifs, regularly positioned at the surface of the histone structured domains. The DNANCP superhelical path and the associated curvature rely on two inter base-pair parameters, roll and slide (Olson and Zhurkin, 2011, Tolstorukov et al., 2007, Xu and Olson, 2010), complemented by twist and rise adjustments (Edayathumangalam et al., 2005, McGinty and Tan, 2015, Muthurajan et al., 2003, Ong et al., 2007, Tan and Davey, 2011). Along the DNA sequences, the values of these inter base-pair parameters and the groove dimensions follow more or less marked sinusoidal profiles, with oscillation period close to 10 base-pairs (bp) (Bishop, 2008, Olson and Zhurkin, 2011, Wu et al., 2010, Xu and Olson, 2010, Yang and Yan, 2011).
The DNA access and readability are often presented as closely related to the in vivo nucleosome positioning that results from complex processes involving a series of trans-acting factors (reviews: (Hughes and Rando, 2014, Lieleg et al., 2015, McGinty and Tan, 2016, Zhou et al., 2018)), but also the intrinsic DNA properties. However, disassembling DNANCP from the histones is not an absolute requirement for binding it and forming NCP-factor complexes. Even when complexed with the histones, DNANCP remains available enough to interact or even operate with chaperones, chromatin remodelers, enzymes or transcription factors (Fernandez Garcia et al., 2019, Kobayashi and Kurumizaka, 2019, Mayanagi et al., 2019, McGinty and Tan, 2016, Speranzini et al., 2016, Volokh et al., 2016, Zhou et al., 2018). The DNase I enzyme is an interesting example, since it has long been exploited in nucleosome studies (Klug and Lutter, 1981, Simpson and Stafford, 1983) and still is (Zhong et al., 2016). Essentially, the catalytic site of DNase I fills the DNA minor groove and cleaves the phosphodiester linkage (Lahm and Suck, 1991, Weston et al., 1992). Applied on NCP, this enzyme targets the wide minor grooves pointing outwards, opposite to the histone octamer, and produces a typical oscillatory cleavage profile (Zhong et al., 2016 and references herein). This property was cleverly exploited to detect transcription factors that bind to DNANCP and consequently disrupt the DNAse I periodic cleavage pattern (Zhong et al., 2016).
The ability of DNANCP to recognize proteins raises in particular the issue of its dynamical behavior. Indeed, it is now accepted that small, frequently subtle motions of free (unbound) DNA affect the assembly of nucleoprotein complexes (Battistini et al., 2019, Rohs et al., 2010). How the local (for instance roll or twist) or semi-local (groove dimensions, curvature) DNA sequence-dependent flexibility influences protein affinities was documented on various systems (Abe et al., 2015, Azad et al., 2018, Djuranovic et al., 2004, Djuranovic and Hartmann, 2005, Heddi et al., 2010a, Heddi et al., 2008, Koudelka and Carlson, 1992, Parvin et al., 1995, Tisné et al., 1999). As typical examples, the dynamics of NF-κB DNA targets is needed to transiently expose the specific base atom pattern recognized by the protein (Tisné et al., 1999, Wecker et al., 2002); also, the DNase I cleavage efficiency is increased by a malleable minor groove that favors the enzyme anchoring (Heddi et al., 2010a).
However, to which extent are structural fluctuations, potentially functionally relevant, preserved, damped or even suppressed in the case of a bound, and therefore constrained, DNA? For such a question, DNANCP is a paradigm: on one hand, at least a residual malleability should be preserved to ensure the structural fit with external partners; on the other hand, DNANCP seems robustly constrained since its super helix path is maintained by an especially dense interaction network involving both the histone structured domains and specific N-terminal regions along the two DNA strands, as recently observed on a nucleosome simulated in solution (Elbahnsi et al., 2018). Despite this, studies suggest that at least some DNANCP properties are reminiscent of those of free DNA. Indeed, free and DNANCP profiles are qualitatively parallel for i) roll and twist (Chua et al., 2012, Richmond and Davey, 2003), ii) roll, twist and phosphate group conformers (Heddi et al., 2010b), iii) roll, twist, and slide (Marathe and Bansal, 2011, Wu et al., 2010), or iv) roll, twist and groove width (Xu and Olson, 2010). This is probably indicative of coordinated motions at the dinucleotide level as observed in free DNA (Dans et al., 2019, Dršata et al., 2013, Heddi et al., 2010b, Heddi et al., 2006, Imeddourene et al., 2016). Concerning the dispersion of DNANCP helical parameter values, a direct way to characterize the DNANCP malleability, studies based on X-ray (Dlakić et al., 2005) or simulations (Roccatano et al., 2007, Sun et al., 2019) mentioned that the ranges of thermal motion of DNANCP and free DNA were similar.
The scarcity of insights on DNANCP dynamics reflects the notorious difficulty to decipher dynamical properties experimentally, in particular DNA flexibility, a situation further aggravated in the case of nucleoprotein complex as large as NCP. While potentially influenced by biases such as intermolecular NCP contacts and crystal packing (Harp et al., 2000, McGinty and Tan, 2015, Tsunaka et al., 2005), examining X-ray structures is of course helpful, but the essentially static character of solid state models limits the information about dynamics. Solution or solid-state NMR is potentially of great interest to approach DNANCP dynamics but remains impracticable because the combination of the DNANCP molecular size and the limited dispersion of chemical shifts causes uninterpretable broadened and overlapping signals (van Emmerik and van Ingen, 2019). In this situation, molecular simulations are a promising alternative, especially considering the improvements of DNA force fields (Ben Imeddourene et al., 2015, Dans et al., 2017, Hart et al., 2012, Zgarbová et al., 2017, Zgarbová et al., 2013). Two among these force fields, Parmbsc0εςOLI (Zgarbová et al., 2013) and CHARMM36 (Hart et al., 2012), were tested on a series of simulations of 1 μs each on four free DNA dodecamers related to the 601 sequence (Ben Imeddourene et al., 2015). We recall that the 601 sequence, also called Widom sequence, is widely used for positioning nucleosomes in in vitro and in vivo experiments because of its high affinity for the histone octamer (Thåström et al., 2004). The MDs of dodecamers yielded output in reasonable agreement with NMR-inferred data, despite being intentionally carried out without experimental structural restraints, and in spite of strong differences in the conception and parametrization of the two underlying force fields. CHARMM36 in particular provided a satisfactory representation of the backbone dihedral transitions and their modulation by the dinucleotide sequence, features that had long been defective in prior simulations.
Given the force field progress, we decided to carry out MD simulations of a nucleosome formed with the 601 sequence in explicit solvent using the CHARMM36 force field (Hart et al., 2012), for a total duration of more than 1 μs. Our aim in this previous work was to describe the DNA/histone interface (Elbahnsi et al., 2018) in conjunction with VLDM, a Voronoi tessellation-based method analyzing the topology of interacting elements without any empirical or subjective adjustment (Elbahnsi et al., 2018, Esque et al., 2013, Retureau et al., 2019). Here, we exploited the same simulations to gain a direct view of the behavior of DNANCP in solution and highlight the main features of the DNANCP dynamics. To obtain a comparison between free and histone bound DNA, we also draw on MDs of free dodecamer segments that together cover 39 base pairs of the 5′ half of 601 sequence (Fig. 1). As implied above, these dodecamer simulations were initially performed for force field tests (Ben Imeddourene et al., 2015), with protocol and force-field consistent with the MDs of NCP.
Several analyses presented here address DNA backbone motions, for the following reasons. Backbone motions in free B-DNA consist in coordinated conformational transitions of ε and ζ dihedral angles between two states called BI and BII. BI and BII populations on oligonucleotides are extrapolated from 31P NMR chemical shifts (Heddi et al., 2006, Tian et al., 2009), relatively easily measured (Gorenstein, 1992). NMR-based studies have provided a general framework to understand the BI ↔ BII equilibrium and its structural effect on the double helix (Heddi et al., 2010b, Heddi et al., 2006, Imeddourene et al., 2016, Isaacs and Spielmann, 2001, Schwieters and Clore, 2007, Tian et al., 2009). Combining experimental data and modelling established that, in NpN·NpN steps (N for any nucleotide), motions of the two facing phosphodiester linkages are concomitant with variations of the relative positions of two successive bases (Dans et al., 2019, Dršata et al., 2013, Heddi et al., 2010b, Heddi et al., 2006, Imeddourene et al., 2016). This reciprocal dependence involves the BI/BII states, the slide, roll, twist (Ben Imeddourene et al., 2015, Heddi et al., 2010b, Imeddourene et al., 2016) and, to a lesser extent, the rise, tilt and shift (Dršata et al., 2013). The BI or BII density per 4–5 bp segments and the groove dimensions are also coupled (Oguey et al., 2010, Xu et al., 2014). Another key point is that NMR data ascertained that the BI/BII propensity is highly sensitive to the DNA sequence (Heddi et al., 2010b, Heddi et al., 2006, Schwieters and Clore, 2007, Tian et al., 2009, Xu et al., 2014). Owing to the structural couplings, the BI/BII populations of each dinucleotide reflect the conformational landscape that this dinucleotide explores. So, in free DNA, ApA/G·T/CpT and ApT/C·A/GpT have only access to a restricted conformational region characterized by predominant BI conformer, negative slide, null or positive roll and low or moderate twist. On the other hand, GpG·CpC, CpG·CpG, GpC·GpC and CpA·TpG are more flexible, with BI ↔ BII oscillations associated to the broadening of slide, roll and twist ranges towards positive, negative and high values, respectively. In sum, the phosphate group motion is an effective reporter of the sequence-dependent local deformability of DNA, and its intramolecular energetics. This is the reason why the backbone states deserve such attention.
This study presents extensive analyzes from careful simulations carried out on DNANCP, and relevant free oligomers, with a state-of-the-art protocol and validated energy model. It documents multiple aspects of DNANCP dynamics including atomic fluctuations, motions of backbone, variability of inter base pair parameters and couplings involving helical descriptors. The similarities or differences between DNANCP and free DNA are systematically surveyed. Thanks to the exhaustive dataset collected on the DNA-histone interface, we also examine the relation between the DNANCP dynamics and the contacts with the histones. Then, the role of the different dinucleotide intrinsic properties on the propensity of a DNA sequence to assemble into a nucleosome is considered in the light of the above properties of DNANCP. Taken together, those observations provide a clearer picture of nucleosomal DNA when considering its biological functions.
Section snippets
Materials and methods
The details of our Molecular Dynamics (MD) setup were presented in two previous papers, one on the nucleosome (Elbahnsi et al., 2018) and the other on the free DNA dodecamers (Ben Imeddourene et al., 2015). We therefore only summarize here the main aspects of the protocols.
Overview of the studied trajectories
Four nucleosome systems (SYS1, SYS1-bis, SYS2 and SYS2-bis) were considered, which differed initially by the conformations of the histone unstructured domains, called tails (see Fig. S1, Table S1 and Materials and Methods). As explained in Introduction, the four simulations were recently used to provide a detailed but focused description of the DNA-histone interface in solution (Elbahnsi et al., 2018). In this previous work, the integrity and stability of the nucleosome was checked following
Conclusions
The widespread depictions of NCPs have popularized the view of a packed DNA appearing quite different from its free state, because of its super-helical path wrapped around the histone structured domains. Such tight bending around the histones could suggest a regime for DNANCP with dramatically altered properties compared to those of free DNA, including highly restricted dynamics. The DNA-histone interface seems a priori to corroborate this view, considering the dense interaction network that
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
The simulations were carried out on the GENCI-CEA platform.
References (99)
- et al.
Deconvolving the recognition of DNA shape from sequence
Cell
(2015) - et al.
How B-DNA dynamics decipher sequence-selective protein recognition
J. Mol. Biol.
(2019) Geometry of the nucleosomal DNA superhelix
Biophys. J.
(2008)- et al.
The conformation of DNA and protein within chromatin subunits
FEBS Lett.
(1977) - et al.
A brief review of nucleosome structure
FEBS Lett.
(2015) - et al.
Solvent mediated interactions in the structure of the nucleosome core particle at 1.9 a resolution
J. Mol. Biol.
(2002) - et al.
Molecular dynamics studies on free and bound targets of the bovine papillomavirus type I e2 protein: the protein binding effect on DNA and the recognition mechanism
Biophys. J.
(2005) - et al.
The role of DNA structure and dynamics in the recognition of bovine papillomavirus E2 protein target sequences
J. Mol. Biol.
(2004) - et al.
Nucleosomes in solution exist as a mixture of twist-defect states
J. Mol. Biol.
(2005) - et al.
Structural features of transcription factors associating with nucleosome binding
Mol. Cell
(2019)
31P NMR of DNA
Methods Enzymol.
Sequence-dependent DNA flexibility mediates DNase I cleavage
J. Mol. Biol.
Importance of accurate DNA structures in solution: the Jun-Fos model
J. Mol. Biol.
NMR evidence for mechanical coupling of phosphate B(I)-B(II) transitions with deoxyribose conformational exchange in DNA
J. Mol. Biol.
Structural transition of the nucleosome during chromatin remodeling and transcription
Curr. Opin. Struct. Biol.
Twenty-five years of the nucleosome, fundamental particle of the eukaryote chromosome
Cell
Fast empirical pKa prediction by Ewald summation
J. Mol. Graph. Model.
DNase I-induced DNA conformation. 2 A structure of a DNase I-octamer complex
J. Mol. Biol.
DNA binding within the nucleosome core
Curr. Opin. Struct. Biol.
Satisfying hydrogen bonding potential in proteins
J. Mol. Biol.
Recognition of the nucleosome by chromatin factors and enzymes
Curr. Opin. Struct. Biol.
Working the kinks out of nucleosomal DNA
Curr. Opin. Struct. Biol.
DNA stretching and extreme kinking in the nucleosome core
J. Mol. Biol.
A new fluorescence resonance energy transfer approach demonstrates that the histone variant H2AZ stabilizes the histone octamer within the nucleosome
J. Biol. Chem.
Nucleosome structural studies
Curr. Opin. Struct. Biol.
Nucleosomal locations of dominant DNA sequence motifs for histone-DNA interactions and nucleosome positioning
J. Mol. Biol.
How NF-kappaB can be attracted by its cognate DNA
J. Mol. Biol.
A Novel ‘Roll-and-Slide’ mechanism of DNA folding in chromatin. Implications for nucleosome positioning
J. Mol. Biol.
Unspinning chromatin: revealing the dynamic nucleosome landscape by NMR
Prog. Nucl. Magn. Reson. Spectrosc.
X-ray structure of the DNase I-d(GGTATACC)2 complex at 2.3 A resolution
J. Mol. Biol.
Structural insight into the sequence dependence of nucleosome positioning
Struct. Lond. Engl.
Experimental maps of DNA structure at nucleotide resolution distinguish intrinsic from protein-induced DNA deformations
Nucl. Acids Res.
Finite representation of an infinite bulk system - solvent boundary potential for computer-simulations
J. Chem. Phys.
Simulations meet experiment to reveal new insights into DNA intrinsic mechanics
PLoS Comput. Biol.
CHARMM: the biomolecular simulation program
J. Comput. Chem.
The mechanics behind DNA sequence-dependent properties of the nucleosome
Nucl. Acids Res.
The static and dynamic structural heterogeneities of B-DNA: extending Calladine-Dickerson rules
Nucl. Acids Res
How accurate are accurate force-fields for B-DNA?
Nucl. Acids Res.
Particle Mesh Ewald - an N.log(n) Method for Ewald Sums in Large Systems
J. Chem. Phys.
Conformational characteristics and correlations in crystal structures of nucleic acid oligonucleotides: evidence for sub-states
J. Biomol. Struct. Dyn.
DNA bendability and nucleosome positioning in transcriptional regulation
Structure, stiffness and substates of the dickerson-drew dodecamer
J. Chem. Theory Comput.
Holding the nucleosome together: a quantitative description of the DNA-histone interface in solution
J. Chem. Theory Comput.
VLDP web server: a powerful geometric tool for analysing protein structures in their environment
Nucl. Acids Res.
1H NMR investigation of the conformational states of DNA in nucleosome core particles
Nucl. Acids Res.
Asymmetries in the nucleosome core particle at 2.5 A resolution
Acta Crystallogr. D Biol. Crystallogr.
Optimization of the CHARMM additive force field for DNA: improved treatment of the BI/BII conformational equilibrium
J. Chem. Theory Comput.
High-resolution biophysical analysis of the dynamics of nucleosome formation
Sci. Rep.
Quantification of DNA BI/BII backbone states in solution. Implications for DNA overall structure and recognition
J. Am. Chem. Soc.
Cited by (4)
Nucleosome assembly and disassembly pathways in vitro
2022, PLoS ONEDeciphering the mechanical properties of B-DNA duplex
2022, Wiley Interdisciplinary Reviews: Computational Molecular ScienceDNA based service data security in cloud computing environment
2021, CITISIA 2021 - IEEE Conference on Innovative Technologies in Intelligent System and Industrial Application, Proceedings