Effect of the cancer-associated mutation H2BE76K on the structure and dynamics of nucleosome: a molecular dynamics study

The H2BE76K oncohistone mutation and its effect on nucleosome stability and dynamics has been studied through Molecular Dynamics (MD) simulations of both the wild-type and mutant H2BE76K forms. At the local level, the mutation results in a loss of hydrogen bonding interactions between residue H2BE76 and H4R92 in the wild-type nucleosome. In the mutant, hydrogen bonds cannot form and they are replaced by repulsive interactions. Such loss of interaction affects the local stability of the histone structure, leading, in the mutated nucleosome, to a distancing between the two involved histones. These local structural changes have long-range conformational and dynamic repercussions. Disruption of the H2B-H4 local interaction propagates along histone H4, inducing a conformational change in the H4 α 3-helix which acts as a structural mediator transmitting the perturbation toward the terminal region, enhancing the flexibility of the H4 N-terminal tail. This increased mobility reshapes DNA contacts, shifting from a single dominant interaction in the wild type to multiple dynamic interactions in the mutant, thereby providing a mechanistic link between a local structural perturbation and long-range nucleosome regulation.

The basic storage unit of eucaryot genome is the chromatin, whose fundamental repeating unit is the nucleosome (Horn and Peterson, 2002; Greer and Shi, 2012; Park and Luger, 2006). The nucleosome core particle consists of approximately 147 base pairs of DNA wrapped around an octamer of four core histone proteins (H3, H4, H2A, H2B) (see Figure 1). Ebrahimi et al. (2020), Sinden et al. (1998), McGinty and Tan (2015), Tolstorukov et al. (2007). Overall, the nucleosome plays a crucial role in gene expression regulation by controlling the accessibility of DNA to transcription factors and other regulatory proteins (Andrews and Luger, 2011). In this way it can determine which genes are expressed and at what levels. Given the regulatory role of chromatin for all DNA-templated processes, it is not surprising that genes encoding histone proteins are often mutated in tumor samples, (Espiritu et al., 2021; Bagert et al., 2021; Shi et al., 2017; Yuen and Knoepfler, 2013; Funato and Tabar 2018; Kanwal and Gupta, 2012; Espinoza Pereira et al., 2023; Lowe et al., 2019) and these mutations can have an impact on nucleosome stability, histone post-translational modification or chromatin dynamics (Arimura et al., 2018; Nacev et al., 2019; Wan and Chan, 2021; Li et al., 2023a; Buschbeck and Hake, 2017; Li et al., 2023b).

According to the analysis by Bennett and colleagues, the E76 polymorphism is found in cancer patients at a frequency 100 times higher than in the general population (Bennett et al., 2019). The E76 residue is embedded within the nucleosome and positioned at the dimer-tetramer interface, indicating that E76 mutations may contribute to tumor development by destabilizing the nucleosome structure. In fact, both in vitro studies (Nacev et al., 2019; Bennett et al., 2019; Arimura et al., 2018) and in vivo experiments Arimura et al., 2018; Kang et al., 2021) have confirmed that mutations at the E76 site weaken the interaction between H2B and H4. While the precise connection between nucleosome instability and changes in gene expression remains unclear, it has been shown that the presence of E76 mutations in various cell lines results in different phenotypic alterations and gene expression patterns as compared to their wild-type counterparts (Wan and Chan, 2021).

H2BE76 in the wild-type nucleosome forms a salt bridge with H4R92 (Luger and Mader, 1997) which is disrupted in the mutated nucleosome H2BE76K. Arimura et al. determined the crystal structure of the E76K nucleosome and compared it with that of the canonical nucleosome, showing that the side chain of the positively charged lysine causes electrostatic repulsion with H4R92, thus leading to a drastic conformational change in the α -helix of H4 (Arimura et al., 2018).

This has been subsequently confirmed by a molecular dynamics (MD) simulation carried out by Bennett et al., in which the disruption of the H2B-H4 association in the H2BE76K mutated histone was cleary observed (Bennett et al., 2019). However, such MD simulation has been carried out for a short time-frame (approximately 20 nanoseconds) and on a system restricted to only the two protein chains involved (Bennett et al., 2019). On the other hand, to obtain accurate information about the mutation’s effect on nucleosome stability and dynamics, it is essential to study the entire mutated nucleosome through MD over extended simulation times (several microseconds), and to compare its structural and dynamic properties to those of the wild-type canonical counterpart. In this respect, the MD technique, which has been widely used in the last decades to investigate both biological and non-biological systems. Armeev et al. (2021); Migliorati and D’Angelo (2016), Li and Kono (2018), Migliorati et al. (2011), Shaytan et al. (2016), Sessa et al. (2015), Rajagopalan et al. (2017), Migliorati et al. (2012), Migliorati et al. (2019) represents an invaluable tool to shed light on differences and analogies between the wild-type and the H2BE76K nucleosomes. MD simulations of the H2BE76K mutated nucleosome have been recently carried out and on the basis of the MD results it was concluded that the H2B and H4 helices at the site of mutation have the highest magnitude of motion (Onyema et al., 2025). Moreover, analysis of the normal modes showed that the mutation increases the frequency of fluctations of histone H2B, while that of histone H4 is somewhat similar to the wild type (Onyema et al., 2025). In such study only the dynamics of the two histones directly involved in the mutation have been investigated, namely, H2B and H4 (Onyema et al., 2025). However, to fully rationalize the effect of the mutation and to understand its increased tumorigenic propensity, it is crucial to investigate the dynamics of the entire nucleosome rather than focusing solely on the two directly involved histones. The aim of this work is to study, at the atomistic level, the effects of the H2BE76K mutation through MD simulations, assessing its impact on nucleosome stability and DNA accessibility, and identifying mutation-induced structural and dynamic changes relevant to chromatin regulation and oncological diseases.

MD simulations of the wild-type and the mutated H2BE76K nuclesome were performed using the GPU-accelerated GROMACS software (Berendsen et al., 1995). Protein atoms were described using the AMBER99SB force field, while DNA was modeled using the AMBER99 nucleic acid parameters with parmbsc0 corrections to the backbone dihedral potentials (Wang et al., 2004). The initial structures of the wild-type and of the mutated H2BE76K nucleosome used in this study correspond to the crystal structures obtained by Arimura et al., available in the Protein Data Bank (PDB) with code 5Y0C and 5Y0D, respectively (Arimura et al., 2018). For both wild-type and mutant systems, three simulations were conducted, each starting from different initial velocities in order to explore a wider set of possible configurations, thereby reducing the risk of the results being influenced by specific initial conditions. In each simulation system, the initial structural model was solvated in a box of TIP3P water molecules with a minimum distance of 13.5 Å between the nucleosome atoms and the edges of the box. NaCl was added to the system up to a concentration of 0.1 M. The obtained simulated cubic boxes have a side length of 135 Å. The solvated systems were first energy minimized using steepest descent minimization for 10000 steps, gradually heated to 300 K throughout 1 ns using position restraints, and then equilibrated for a period of 1 ns. After that, the production simulations were carried out with a time step of 2 fs in the NPT ensemble with the temperature and pressure maintained at 300 K and 1 atm using the modified Berendsen thermostat (velocity-rescaling) Bussi et al. (2007) and Parrinello-Raman barostat, Parrinello and Rahman (1981) respectively. The final production period in the NVT ensemble simulation was 7 μ s, during which configurations were saved every 90 steps. The intermolecular interactions were calculated using a 1.2 nm cut-off for building neighbor lists, for van der Waals interactions, and for Coulomb interactions. Long-range electrostatic effects were treated using the PME (Particle Mesh Ewald) method Essmann et al. (1995) with a Fourier grid step set to 0.12 nm. All stretching vibrations related to bonds containing hydrogen atoms were constrained using the LINCS algorithm (Hess et al., 1997). To evaluate the influence of ionic strength on the obtained results, six additional simulations of the wild-type and E76K nucleosomes (three replicas per each system) have been performed for 0.5 μ s, at a NaCl concentration of 0.15 M, by using the same protocol described above. Principal component analysis (PCA) was performed for each nucleosome system using the concatenated trajectories of the three independent simulations. In addition, PCA was also carried out on each individual trajectory separately to assess the conformational space sampled across independent simulations.

The comparative analysis between the wild-type nucleosome and its H2BE76K mutated variant through MD allowed us to determine whether and how the mutation affects the structural stability and dynamics of the nucleosome, with potential implications for DNA accessibility regulation. Before proceeding with the analysis of the trajectories, we compared the B-factors calculated from the MD simulations for all C α atoms of both the wild-type and mutant nucleosomes with the experimental B-factors derived from the crystal structure reported by Arimura et al. (2018) (see Supplementary Figure S1 of the Supplementary Material). In both cases, the B-factors obtained from the simulations show qualitative agreement with the experimental values, supporting the reliability of the simulated dynamics and validating the structural fluctuations sampled during the MD simulations.

Conformational stability of wildtype and H2BE76K nucleosomes

The first analysis performed on the trajectories was the Root Mean Square Deviation (RMSD). This type of analysis provides a global view of the structural stability over time. From the obtained results (Supplementary Figure S2 of the Supplementary Material), it can be observed that both the wild-type and mutated nucleosomes maintain an RMSD value below 0.40 nm throughout the simulations. This finding indicates that the nucleosome as a whole remains structurally stable throughout the simulations and the introduction of the E76K mutation in histone H2B does not compromise the global integrity and fold of the nucleosome. Moreover, the RMSD results show that the wild-type and H2BE76K nucleosomes stabilize in a time span shorther than 500 ns in all the six MD simulations.

The radius of gyration (R g) of DNA is a useful parameter to characterize the size and the degree of compactness of nucleosome and it has been computed as a function of time for both wild type and mutated nucleosomes. The results of this analysis are reported in Supplementary Figure S3 of the Supplementary Material. Note that the time reported in the graphs represents an aggregate time of 21 μ s, derived from the sum of three simulations of 7 μ s each, a sufficiently long period to obtain a robust assessment of system properties. The average DNA R g value is identical for both the wild-type and mutated nucleosomes, corresponding to R g ̄ = 4.52 ± 0.01 nm in both systems. The steadiness of such parameter over the time and the absence of significant differences between wild-type and mutant implies that the biological system’s stability does not undergo major variations, confirming that the nucleosome does not experience significant structural relaxation or alterations that could compromise its integrity.

Local effects of the mutation

In order to examine the local effects of the mutation, the number of hydrogen bonds between the two residues that, in the wild-type, are expected to stably interact, namely, the glutamate E76 of histone H2B and the arginine R92 of histone H4, have been calculated. Figure 2A presents the distribution, expressed as percentage, of the number of hydrogen bonds between residues E76 and R92, calculated over the entire aggregated simulation of 21 μ s, obtained by combining the three independent trajectories. Note that hydrogen bonds were identified using a geometric definition by considering a donor-acceptor distance ≤ 0.35 nm and a hydrogen-donor-acceptor angle ≤ 30 °.

The results show that in the wild-type nucleosome, the number of hydrogen bonds between E76 and R92 fluctuates mainly between 1 and 3, with a dominant configuration in which two hydrogen bonds are formed between the two residues. Figures 2B–D illustrate examples of MD snapshots in which the two residues form 2, 1 and 3 hydrogen bonds, respectively. However, it is important to emphasize that these configurations are dynamic and continuously vary over time. Conversely, in the mutated nucleosome, the substitution of glutamic acid with lysine at position 76 prevents the formation of stabilizing interactions with R92 of hystone H4, as both residues are positively charged and therefore subject to electrostatic repulsion (see as an example the MD snapshot shown in Figure 2E). This loss of interaction could have implications for the local stability of the histone structure. The analysis of the minimum distances between residues 76 of H2B and H4R92 allows one to evaluate whether and how these residues interact with each other and to observe any changes in their behavior following the mutation. As a consequence of the hydrogen bond formation between H2B E76 and H4 R92 in the wild-type nucleosome, the minimum distance between the two residues remains below 2.0 Å for most of the time (see Figure 2F). Note that the distances shown in Figure 2F are not heavy-atom distances but were computed considering any pair of atoms belonging to the two interacting residues, including hydrogen atoms. In the case of the mutant instead, the absence of a favorable molecular interaction between the two residues leads to a significantly larger minimum distance compared to the wild-type, with considerably wider fluctuations, ranging between 2.0 and 6.0 Å (Figure 2G). Figure 3 shows the region surrounding residue 76 of histone H2B, obtained by superimposing the mutant and wild-type nucleosome structures within the energy minima basins identified through PCA of the MD simulations (vide infra). The presence of a repulsive interaction between lysine (K) at position 76 of H2B and arginine (R92) of histone H4 induces a conformational change in the H4 α 3-helix, in agreement with the crystal structure results reported by Arimura et al. (2018).

To provide direct evidence of the lack of favorable electrostatic interaction between residue 76 (H2B) and R92 (H4) in the mutant, we have calculated the per-residue electrostatic interaction energy between such two residues both in the wild-type and E76K systems. Our results show a significant difference in the local interaction environment: the wild-type system exhibits a strong attractive interaction with an average energy of −130 kJ/mol, whereas the E76K mutant shows a negligible/weakly repulsive interaction of +1.02 *10 -5 kJ/mol. To provide a dynamic view of this interaction, we have also shown the time evolution of the short-range electrostatic interaction energy between H2B-E76/K76 and H4-R92 in a 2-microsecond representative window of our simulations (see Supplementary Figure S4 of the Supplementary Material). This quantitative evidence confirms that the glutamate-to-lysine substitution disrupts the favorable electrostatic interaction between H2B and H4, replacing it with a weakly repulsive state. It should be noted that the repulsive energy value in the E76K mutant remains near-zero because the two positively charged residues undergo a conformational rearrangement to avoid unfavorable contact. This is further supported by our minimum distance analysis between residue 76 (H2B) and R92 (H4) throughout the MD trajectory, which shows that these residues tend to remain further apart in the mutant compared to the wild-type, to minimize electrostatic repulsion.

Structural compactness of the nucleosome

To assess the impact of the H2B E76K mutation on the structural compactness of histones in the nucleosome core, an analysis of the distances between the centers of mass of single histones, of the dimers and of the tetramers has been carried out (Figure 4). The distribution of the distances between the centers of mass of the two histones direcly involved in the mutation, namely, H2B and H4, show that in the mutant a slight increase in the separation between the two histones is observed (Figure 5A). The most probable distance in the wild-type and E76K nucleosome is indeed 3.12 and 3.15 nm, respectively. This effect is likely due to the loss, in the mutant, of the hydrogen bonds formed between residue E76 of histone H2B and residue R92 of histone H4. In order to determine if the mutation affects also the relative positioning of the dimers within the nucleosome core, we have calculated the distribution of the distances between the center of masses of the two dimers (H3-H4) and of the two dimers (H2A-H2B), shown in Figures 5B,C, respectively. The results show that, in the H2B E76K mutant, the dimers tend to move slightly closer together. In particular, the distributions for the (H3-H4) dimers present a peak at 3.64 nm and 3.68 nm in the mutated and wild-type nucleosome, respectively, while in the case of (H2A-H2B) the maxima of the distributions are found at 3.66 nm and 3.72 nm. On the other hand, when considering the entire octameric histone core, the calculated distance between the centers of mass of the two tetramers, (H2A-H2B-H2A-H2B) and (H3-H4-H3-H4), is larger in the mutant, with the most probable value at 1.97 nm compared to 1.92 nm of the wild-type nucleosome (Figure 5D). The distribution of the distances between the centers of mass have been also calculated separately for each of the three individual trajectories and the results of this analysis are shown in Supplementary Figures S5, S6 of the Supplementary Material for the wild-type and E76K system, respectively. Supplementary Table S1 of the Supplementary Material lists the average distances obtained from the individual distributions, the total averages and the standard deviations. The analysis of individual trajectories confirms the consistency of our findings, as the resulting mean distances are statistically distinct, further supporting the observed trends. These results suggest that, while the dimers tend to reduce their internal distances, the overall organization of the tetramers appears to be less compact in the mutated variant. This could reflect an altered structural dynamics of the histone core, potentially affecting the overall stability of the nucleosome.

Long-range effects of the mutation on nucleosome dynamics

Analysis of Root Mean Square Fluctuations

The Root Mean Square Fluctuations (RMSF) measure how much each residue in a protein fluctuates relative to a reference position, which corresponds here to its average. Residues that fluctuate significantly from their average position indicate greater mobility during the simulation. This behavior is typically observed in more flexible regions, such as loops or protein termini. Conversely, residues with low fluctuation indicate more stable regions, commonly found in structurally rigid areas such as helices or beta-sheets. RMSF is thus a valuable measure to identify the most mobile and flexible regions of a protein throughout a simulation.

From the results shown in Figure 6, it emerges that histone chains within the mutated nucleosome exhibit larger flexibility compared to those in the wild-type nucleosome. This trend is evident across all four histone types forming the nucleosome core, although with specific differences for each histone. Notably, an increase in flexibility is observed in the protein termini, regions that are typically more dynamic and less structured. For H3 and, to a great extent, for H4, there is a significant increase of the fluctuations at the N-terminal ends of the mutated nucleosome as compared to the wild-type counterpart, suggesting that the H2BE76K mutation indirectly affects the dynamics of these histones. In the case of H2A, increased flexibility is particularly evident around residues 100–110. Lastly, for H2B, while the overall trend is similar between the wild-type and mutant nucleosome, a larger fluctuation is still observed near the residue where the mutation occurs.

The increased flexibility observed in the mutant nucleosome may result in a less rigid and more dynamic structure, potentially impacting nucleosome stability and chromatin regulation. This observation supports the hypothesis that the H2BE76K mutation may enhance nucleosome plasticity, making it more susceptible to remodeling processes and interactions with other protein factors.

The ends of the histones are the regions of the proteins showing the highest mobility and they are also the regions that interact with DNA. A visual representation of this behavior is provided in Figure 7, which depicts the regions of the different histone proteins that form contacts with DNA. As concerns H4, the residues near the N-terminus make contacts with DNA and are also highly mobile, while for H2A the same occurs for the C-terminal region. H3 is, in general, the histone that establishes the highest number of contacts with DNA, wedging itself between the DNA wraps. Due to its central position in the nucleosome core, H3 plays also a crucial role in the stability of the nucleosome structure (see Figure 1).

In addition to histone analysis, the RMSF of DNA was also evaluated, measuring the average fluctuation of each nucleotide relative to its mean position during the simulation. Comparing RMSF values between the wild-type and mutant DNA sequences reveals an increase in flexibility in the mutated condition compared to the wild-type, which is particularly significant at the DNA end (Figure 8). The enhanced mobility of DNA may be a consequence of the reduced stability of the mutant nucleosome, leading to a more dynamic structure. This phenomenon could have important implications for chromatin regulation, potentially facilitating easier access to transcription factors and other protein complexes involved in gene regulation.

RMSF profiles for both histones and DNA were also calculated independently for each trajectory (see Supplementary Figures S7, S8 for the wild-type and E76K nucleosome, respectively). These individual analyses confirmed the trends observed in the concatenated trajectories, demonstrating the consistency of the fluctuations across all replicates.

Principal component analysis

PCA of the atomic positional fluctuations covariance matrix (Amadei et al., 1993) is a very useful technique to shed light on biologically relevant movements in the study of protein systems. Note that throughout the text, PCA always refers to the PCA of the atomic positional fluctuations covariance matrix. The PCA results for the wild-type and mutant H2BE76K nucleosomes are shown in Supplementary Figure S9 of the Supplementary Material, where the projections of the first two eigenvectors in a two-dimensional space is shown. Note that the autovectors were obtained by analyzing the entire 21-microsecond trajectories, derived from concatenating the three individual trajectories.

The obtained distributions indicate that an extensive conformational sampling has been achieved during the simulations, with the systems exploring a wide range of configurations in terms of eigenvectors. These results can be used to construct multidimensional histograms, representing the distribution of these movements in terms of energy. Such histograms provide insights into the system thermodynamic properties. By applying a Boltzmann transformation to these multidimensional histograms, Gibbs free energy landscapes can be generated (Figure 9). The wild-type energy profile exhibits a single well-defined minimum, located on the right side of the graph, with an energy difference of 12.8 kJ/mol between the most stable and less stable states. This indicates that the system tends to remain confined in an energetically favorable configuration, suggesting high structural stability. Conversely, the H2BE76K mutation significantly increases the dispersion of energy minima, causing the system to explore a broader range of conformational states. The presence of multiple minima suggests a larger variety of accessible states, implying that the mutant nucleosome has a more complex and less confined dynamic, resulting in a more disordered system which explores multiple conformations with nearly equal probabilities.

In addition, PCA was independently performed on each individual trajectory to evaluate the conformational space sampled across the three replicate simulations. The results obtained from these separate analyses (see Supplementary Figure S10 of the Supplementary Material) are consistent with those derived from the concatenated trajectory, confirming that the mutant nucleosome exhibits a more complex and less confined dynamical behavior compared to the wild-type system. The Gibbs free-energy landscapes reconstructed from the individual trajectories (Supplementary Figure S11 of the Supplementary Material) allowed us to estimate the uncertainty associated with the free-energy minima. For the wild-type nucleosome, the variability among replicas corresponds to an uncertainty of approximately 0.4 kJ/mol. For the mutant system, a standard deviation of 1.2 kJ/mol was obtained, reflecting a larger variability in the sampled conformational ensemble.

It is important to note that the absolute values of the free-energy minima differ when comparing individual trajectories with the concatenated dataset. This behavior is expected, as the free-energy landscape computed from concatenated trajectories reflects the combined probability distribution of all sampled configurations rather than the average of the individual replica estimates. Consequently, the concatenated analysis provides a more statistically robust representation of the dominant conformational basins, while the replica-based analyses quantify the sampling variability.

To further validate the interpretation of the conformational landscape and address the dependence of free energy projections on the choice of collective variables, an additional analysis was performed by projecting the trajectories onto the second and third principal components obtained from the PCA (see Supplementary Figure S12 of the Supplementary Material). The corresponding two-dimensional free energy landscape was therefore computed (see Supplementary Figure S13 of the Supplementary Material) and compared with that derived from the first two principal components. The analysis revealed a consistent behavior between the two projections: the wild-type nucleosome remains confined within a limited energetic basin, indicative of high structural stability, whereas the H2BE76K mutant exhibits a broader distribution of energy minima, exploring a wider conformational space and displaying a more heterogeneous and less confined dynamic behavior. The Gibbs free energy analysis thus confirms that the mutation alters the nucleosome’s conformational behavior, promoting a larger structural flexibility and a lower stability compared to the wild-type form.

Global motions of nucleosome

Nucleosome dynamics, analyzed through the first principal component (PC1) obtained via PCA, reveal significant differences between the wild-type and H2BE76K mutant structures. In order to visualize the regions of the nucleosome that are involved in the global motion associated to the first principal mode obtained via PCA, it is very useful to show the dynamics of the nucleosome explained by such eigenvector in the two systems (see Figure 10). In the wild-type nucleosome, the dominant motion is primarily localized at the DNA entry-exit regions and involves the terminal DNA segments together with the C-terminal tail of histone H2A, indicating a concerted fluctuation between DNA breathing and H2A tail dynamics. The remaining regions of the nucleosome core show comparatively limited displacement along this principal mode. Conversely, the E76K mutant exhibits a substantially more delocalized motion along PC1. The dominant mode involves multiple distinct regions of the nucleosome, including the C-terminal regions of both H2A chains as well as the N-terminal tails of histones H3 and H4. This redistribution of collective dynamics suggests that the mutation promotes a broader propagation of structural fluctuations across the nucleosome particle, consistent with increased dynamical coupling between histone tails and DNA. These results are consistent with the broader and more rugged free energy landscape of the H2BE76K mutant, supporting a scenario in which the mutation enhances conformational heterogeneity and promotes more delocalized nucleosome dynamics. It should be highlighted that DNA dynamics can depend on the choice of nucleic acid force field. In the present work, DNA was described using the AMBER99+parmbsc0 parametrization, which has been extensively validated for nucleosomal DNA simulations. Although newer parametrizations such as OL15 may provide improved quantitative descriptions of DNA flexibility, comparative studies indicate that parmbsc0 preserves the overall structural and dynamical properties of DNA. Consequently, the relative trends in end-DNA mobility reported here are expected to be robust with respect to the specific AMBER-family DNA force field used.

It is important to note that the percentage of nucleosome motion associated with the first, second, and third eigenvectors amounts to 33%, 12%, and 4%, respectively, in the wild-type system, whereas it is 23%, 15%, and 11% in the mutant. Therefore, the observation that the terminal region of the DNA explores a larger conformational space in the wild type - as evident from the superposition of configurations, where the sampled spatial region appears broader - does not imply that the DNA is less mobile in the mutant. Indeed, in the mutant system the first eigenvector contributes less to the total fluctuations compared to the wild type, indicating that the motion is distributed over multiple modes rather than being dominated by a single collective motion. When the overall dynamics are considered, the terminal DNA regions are in fact more mobile in the mutant nucleosome, as demonstrated by the RMSD analysis.

It is interesting now to assess whether the increased DNA end mobility observed in the mutant also corresponds to a greater degree of DNA unwrapping. To address this point, we calculate nucleosome opening angles and DNA end-octamer distances throughout the concatenated trajectories, that are shown in Supplementary Figure S14 of the Supplementary Material. Distances were defined as the separation between the center of mass of the histone octamer and the center of mass of the last five DNA residues, while opening angles were calculated by using the three following points: the center of mass of the nine DNA residues surrounding the dyad, the center of mass of the histone octamer, and the center of mass of the last five DNA residues. The results of this analysis show that the average unwrapping in the E76K nucleosome remains comparable to the wild type system. Instead, the mutant displays bimodal distributions of opening angles and DNA end-octamer distances, indicating that it samples multiple conformational states. This suggests that the mutation primarily modulates nucleosome dynamics and conformational heterogeneity rather than promoting a persistently more unwrapped state.

Mutation-induced modulation of H4-DNA interactions

It is particularly interesting to extract nucleosome conformations corresponding to the free-energy minima in order to determine whether these basins correspond to specific structural states. Figure 11A shows the superposition of the two conformations obtained in this way for the wild-type and the mutant systems. The structures appear similar and largely superimposable, both as concerns the histone octamer and the two DNA strands. The most significant difference between the minimum conformations concerns the position of the N-terminal tail of histone H4, which extends in different directions in the two systems (see the red arrows in the figure). This result is especially intriguing in light of the fact that histone H4 is directly involved in the interaction affected by the mutation, namely, the disruption of the interaction between residue 76 of H2B and arginine 94 of histone H4.

To understand how this local interaction change propagates along histone H4 and leads to such a drastic rearrangement of the terminal region of the protein, we superimposed this region of the nucleosome (Figure 11B). The comparison shows that the local mutation induces a conformational change in the H4 α 3 -helix, as previously noted, while not significantly altering the conformation of the remaining regions of histone H4, which remain largely superimposable in the two systems. In contrast, the N-terminal tail is strongly affected and adopts markedly different orientations.

In light of these observations, we calculated the contacts between the DNA and the terminal region of histone H4, defined by 7 residues: HIS18, ARG19, LYS20, VAL21, LEU22, ARG23, ASP24. Contacts were defined based on the distance between heavy atoms, excluding hydrogen atoms, specifically considering two heavy atoms in contact when their distance is less than 4.50 Å. The resulting contact distributions are very different in the two systems (Supplementary Figure S15 of the Supplementary Material). In the wild-type, the probability distribution exhibits a maximum at approximately 14 contacts, whereas the mutant displays a bimodal distribution with a maximum probability around 38 contacts. Analysis of individual simulations (Supplementary Figure S16 of the Supplementary Material) yielded an average number of contacts of 25.1 ± 4.0 for the mutant and 16.8 ± 3.8 for the wild type. These results indicate that the terminal region of histone H4 tends to form a larger number of interactions with DNA in the mutant as compared to the wild type. To further characterize this behavior, we computed the contact map between DNA and the terminal region of histone H4 (Figure 11C). Contacts were defined based on distances between heavy atoms (excluding hydrogens), with two atoms considered in contact when separated by less than 4.50 Å. In the wild-type system, the most stable interaction occurs between H4 HIS18 and bp196 (located at SHL -2). Conversely, the mutant displays a more dynamic contact profile, characterized by multiple interactions rather than a single dominant one: ARG19-bp207, LEU22-bp207, LEU22-bp206 (all located at SHL-1), HIS18-bp86, ARG19-bp86 and ARG23-bp88 (located in the same region but belonging to the other DNA strand). These results highlight two major findings. First, the two N-terminal tails of histone H4 interact with different DNA segments in the wild-type and mutant systems. To illustrate how this is possible, Figure 11D shows once again the superposition of the histone core within the two nucleosomes, considering the conformations extracted from the minimum basin identified by the PCA analysis. The comparison clearly shows that the two tails of histone H4, pointing in markedly different directions, can establish interactions with distinct regions of DNA.

The second important result is that, in the wild-type system, only a single dominant contact is observed, whereas the mutant interacts with multiple residues along the DNA. This broader variety of contacts reflects the dynamic behavior of the histone tail. Indeed, analysis of the motion associated with the first eigenvector showed that, in the mutant, the H4 tail spans a wide region of conformational space, while in the wild type it remains comparatively restricted and does not display significant motion. Note that the significantly higher mobility of the histone H4 tail in the mutant nucleosome compared to the wild-type was also observed by the RMSF analysis. By exploring this larger conformational space, the mutant tail dynamically establishes interactions with DNA via different residues over time.

From a functional perspective, this enhanced conformational flexibility may represent a key feature of the oncohistone behavior. Notably, these findings suggest that a mutation disrupting a local interaction can propagate its effects over long distances along the protein structure, ultimately increasing the flexibility of the H4 terminal region and reshaping its interaction landscape with DNA. Such long-range allosteric effects provide a possible mechanistic link between a localized structural perturbation and global chromatin regulation changes associated with oncogenic mutations.

Our findings of increased nucleosomal flexibility could provide a structural basis for the transcriptional dysregulation observed in H2BE76K-associated cancers. Specifically, Kang et al. identified ADAM19 as a direct target of this mutation, showing that H2BE76K promotes ADAM19 upregulation by facilitating faster RNA polymerase II (Pol II) progression along the gene body (Kang et al., 2021). The enhanced DNA end dynamics observed in our MD simulations suggests that the mutation increases nucleosome breathing without significantly altering the average degree of DNA unwrapping. Such increased dynamical fluctuations may transiently reduce the physical barrier normally imposed by the nucleosome, creating short-lived windows of accessibility for the transcriptional machinery. This dynamic accessibility may also favor the binding of pioneer transcription factors, such as FoxA1 or GATA factors, which possess the unique ability to target DNA sequences wrapped around the nucleosome (Cirillo et al., 2002). Rather than promoting a persistently destabilized nucleosome, the H2BE76K mutation appears to reshape nucleosome dynamics by enhancing conformational heterogeneity, thereby facilitating regulatory factor engagement. Unlike other oncohistones such as H3K27M, which primarily perturb the epigenetic landscape by affecting histone modifications, H2BE76K appears to function as a ’structural oncohistone’ that drives oncogenesis through direct physical alteration of nucleosome dynamics.

Effect of ionic strength on nucleosome dynamics

To evaluate the influence of ionic strength on the obtained results, we performed six additional MD simulations of the studied systems, consisting of three replicas of the wild-type nucleosome and three replicas of the E76K mutant, at a NaCl concentration of 0.15 M. The results obtained at this higher salt concentration confirm the trends previously observed at the lower ionic strength (0.10 M). Across most regions of the histone core, the mutant exhibits larger fluctuations compared to the wild type, indicating an overall increased mobility (see Supplementary Figure S17 of the Supplementary Material). The most pronounced mobility in the mutant is observed for residue HIS18 located in the N-terminal tail of histone H4, in agreement with the behavior detected at the lower salt concentration. Consistently with the results at 0.10 M NaCl, the N-terminal tail of the wild-type nucleosome displays markedly reduced mobility. PCA further supports these findings, showing conformational dynamics consistent with those observed at the lower ionic strength. In particular, the projections of the first two eigenvectors in a two-dimensional space (see Supplementary Figure S18 of the Supplementary Material) have been used also in this case to build Gibbs free energy landscapes (Supplementary Figure S19 of the Supplementary Material). In agreement with our findings at 0.10 M NaCl, simulations performed at 0.15 M further confirm that the H2BE76K mutation significantly increases the dispersion of energy minima within the conformational landscape. The mutant system explores a broader range of conformational states compared to the wild type, as evidenced by the presence of multiple energy minima. Indeed, this distribution shows a larger variety of accessible configurations, suggesting that the mutant nucleosome exhibits a more complex and less confined dynamic behavior.

In this work we have investigated the H2BE76K histone mutation and its effect on nucleosome stability and dynamics through MD simulations of both the wild-type and mutant H2BE76K forms. Among the recently identified histone mutations associated with tumor phenomena, special attention has been devoted to the H2BE76K mutation, since this mutation is located in a crucial region for the interaction between the H2A-H2B dimer and the H3-H4 tetramer. At the local level, the mutation results in a loss of hydrogen bonding interactions that are formed between residue H2BE76 and H4R92 in the wild-type nucleosome. Conversely, in the mutant hydrogen bonds cannot form but are replaced by repulsive interactions. Indeed, the substitution of glutamic acid with lysine at position 76 prevents the formation of stabilizing interactions with R92, as both residues are positively charged and thus subject to electrostatic repulsion. This loss of interaction affects the local stability of the histone structure, leading, in the mutated nucleosome, to a distancing between the two residues in histones H2B and H4 and to significant fluctuations in their distances. These local structural changes also have long-range repercussions: for the first time a distance analysis between histone center of mass has been carried out showing that the distances between the mass centers of the histones involved in the mutation increase, and the two histone tetramers also separate in the mutated nucleosome as compared to the wild-type form. These structural changes also impact nucleosome dynamics, particularly at the histone and DNA ends, which shows enhanced mobility in the mutant. In contrast to previous studies that focused exclusively on the dynamics of the histones directly involved in the mutation, Onyema et al. by carrying out a PCA analysis on the entire nucleosome it emerged that the H2BE76K nucleosome is able to explore a wider ensemble of conformational states as compared to the wild-type system. In the wild-type nucleosome, the principal mode is consistent with a localized DNA breathing dynamics, while the E76K mutant displays a markedly more delocalized motion encompassing multiple structural elements, including the C-terminal regions of both H2A chains and the N-terminal tails of histones H3 and H4. This redistribution of collective dynamics indicates that the mutation enhances long-range dynamical coupling within the nucleosome, promoting the spread of structural fluctuations from a local perturbation site to distal histone tails and DNA regions. Our results demonstrate that disruption of the local interaction between H2B residue 76 and histone H4 does not simply produce a localized structural perturbation but propagates along histone H4, ultimately reshaping nucleosome dynamics. Conformations extracted from the free-energy minima reveal that the overall nucleosome architecture remains largely preserved between wild-type and mutant systems, while the N-terminal tail of histone H4 adopts markedly different orientations. Structural comparison shows that the mutation induces a conformational rearrangement in the H4 α 3-helix, which acts as a structural mediator transmitting the perturbation toward the terminal region. Consistent with this propagation mechanism, the H4 N-terminal tail displays substantially altered interactions with DNA in the mutant nucleosome. Contact analysis reveals an increased number and diversity of DNA contacts, accompanied by a shift from a single dominant interaction in the wild type to multiple transient interactions in the mutant. The terminal histone H4 tail thus assumes a pivotal role in the structural and dynamic changes induced by the mutation, transitioning from a rigid state in the wild type to being the most dynamic component of the nucleosome in the mutant. To the best of our knowledge, this is the first time that such role of the H4 N-terminal tail has been highlighted in the H2BE76 mutant, both in terms of its dynamics and its specific interactions with the DNA. In the previous simulations carried out by Onyema et al. (2025) the H2B and H4 helices at the site of mutation were found to show the highest magnitude of motion and not the terminal tail, as obtained in the present investigation. As a consequence of its altered dynamical landscape, the H4 N-terminal tail engages distinct DNA regions in the mutant compared to the wild-type nucleosome, thus reshaping histone-DNA interaction patterns and potentially modulating chromatin accessibility and regulatory signaling, in line with the functional impact expected for an oncohistone. This study represent a further step in the understanding of the molecular mechanisms associated to histone mutations which can be fundamental to disentangle the epigenetic phenomena underlying many diseases, opening interesting future perspectives in both diagnostic and therapeutic fields.