Skip to main content
NCBI home page
Biophysical Journal logoLink to Biophysical Journal
. 2015 Oct 20;109(8):1676–1685. doi: 10.1016/j.bpj.2015.07.004

Nucleosome Core Particle Disassembly and Assembly Kinetics Studied Using Single-Molecule Fluorescence

Noa Plavner Hazan 1, Toma E Tomov 1, Roman Tsukanov 1, Miran Liber 1, Yaron Berger 1, Rula Masoud 1, Katalin Toth 2, Joerg Langowski 2, Eyal Nir 1,∗
PMCID: PMC4623893  PMID: 26488658

Abstract

The stability of the nucleosome core particle (NCP) is believed to play a major role in regulation of gene expression. To understand the mechanisms that influence NCP stability, we studied stability and dissociation and association kinetics under different histone protein (NCP) and NaCl concentrations using single-pair Förster resonance energy transfer and alternating laser excitation techniques. The method enables distinction between folded, unfolded, and intermediate NCP states and enables measurements at picomolar to nanomolar NCP concentrations where dissociation and association reactions can be directly observed. We reproduced the previously observed nonmonotonic dependence of NCP stability on NaCl concentration, and we suggest that this rather unexpected behavior is a result of interplay between repulsive and attractive forces within positively charged histones and between the histones and the negatively charged DNA. Higher NaCl concentrations decrease the attractive force between the histone proteins and the DNA but also stabilize H2A/H2B histone dimers, and possibly (H3/H4)2 tetramers. An intermediate state in which one DNA arm is unwrapped, previously observed at high NaCl concentrations, is also explained by this salt-induced stabilization. The strong dependence of NCP stability on ion and histone concentrations, and possibly on other charged macromolecules, may play a role in chromosomal morphology.

Introduction

Eukaryotic DNA is packed into nucleosomes (1) consisting of eight histone proteins wrapped by ∼1.75 turns of DNA. This structural arrangement restricts the accessibility of DNA to proteins; therefore, the nucleosome plays a major role in regulating important processes such as transcription, replication, and repair (2). Many structural studies use the nucleosome core particle (NCP), a complex of the histone octamer with 147 basepairs of DNA, as a standard model. Simple thermodynamics tell us that NCPs are more stable when histone and DNA concentrations are high (3, 4, 5, 6, 7, 8). High salt concentrations destabilize the NCP (3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18), partly due to the fact that counterions weaken the attraction between the positively charged histone proteins and the negatively charged DNA, resulting in partial or complete dissociation of the complex. The stability of H2A/H2B histone dimers and (H3/H4)2 tetramers in the absence of DNA, on the other hand, increases with an increase in salt concentration (19, 20, 21, 22). Gloss et al. suggested that this stabilization is a result of a salt-induced hydrophobic effect and screening of electrostatic repulsion between the positively charged histones (22). However, the influence of these stabilizing effects on salt-dependent stability of NCP is still unknown.

During salt-induced destabilization, intermediate NCP states are observed, and it has been suggested that dissociation and association are multistep processes (3, 9, 11, 12, 14, 16, 23, 24, 25, 26, 27, 28). For example, using single-molecule fluorescence, Gansen et al. (3) identified a stable intermediate state at NaCl concentrations >150 mM that they postulated to be an NCP with one unwrapped DNA arm. However, a mechanism that explains the presence of this intermediate in higher salt concentration and its absence in lower salt concentrations is still required. In dynamic studies using single-molecule fluorescence (29) and fluorescence correlation spectroscopy (FCS) (23), under equilibrium conditions, wrapping and unwrapping of the DNA arm was found to occur on a timescale of tens to hundreds of milliseconds. Other preliminary single-molecule fluorescence kinetic measurements revealed salt-induced NCP dissociation on timescales of hundreds of seconds (3, 14). Despite being the subject of much study, the detailed mechanisms of NCP assembly and disassembly, which influence the stability of the NCP complex, are not yet sufficiently understood. For example, it is unknown how counteracting ion-related effects and the total nucleosome concentration influence NCP stability and why intermediate states are stabilized at certain ion concentrations.

Single-molecule fluorescence approaches offer several advantages in the study of the structure and dynamics of the complex and often heterogeneous NCPs (3, 4, 5, 9, 13, 14, 15, 29, 30, 31, 32, 33). Single-pair Förster resonance energy transfer (FRET) provides high-resolution information about conformations and distribution of conformations. Alternating laser excitation (ALEX) techniques will, in addition, allow the removal of signal from an incompletely labeled sample, significantly improving the reliability of the data (34, 35, 36, 37). These methods can provide equilibrium data and nonequilibrium kinetic data at picomolar concentrations, which is essential due to the strong interactions (high affinity) between histones and DNA and between histone proteins.

Here, we present a comprehensive single-molecule fluorescence study of NCP stability and, within the time resolution allowed by the technique, a study of the kinetics of salt-induced NCP dissociation and association reactions. The dynamics of NCP formation and the influence of NCP concentration on the stability of the NCP were studied at different NaCl concentrations. Our results show that at close-to-physiological NaCl concentrations (≈150–300 mM), NCP stability is reduced, and that at higher NaCl concentrations, an intermediate NCP state is stabilized. We suggest that this behavior might be a result of salt-induced stabilization of H2A/H2B dimers and possibly (H3/H4)2 tetramers.

Materials and Methods

Preparation of the NCP

For a detailed description of the NCP preparation procedures, see Figs. S1–S4 in the Supporting Material. The NCP was reconstituted by salt dialysis of double-labeled DNA (147 bp Widom-601 sequence) and unlabeled recombinant Xenopus laevis histone octamers, generally according to published protocols (38). The DNA was labeled on thymine through a C6 linker at position +34 (relative to the dyad) with ATTO-550 and at position −45 with ATTO-647N (donor and acceptor, respectively). The labeling positions were chosen such that intact NCP, histone-free DNA, and the intermediate states showed different FRET values (see text). Xenopus laevis histones were expressed in Escherichia coli BL21 and purified at the University of Colorado Protein Facility according to established protocol (38).

Single-molecule fluorescence measurements

The diffusion-based single-molecule FRET (smFRET)/ALEX experiments were carried out on an in-house-built optical setup. For a detailed description of the technique, see Fig. S5 and previous publications (34, 35, 36, 37). The technique provides observation snapshots (also known as bursts) that capture the state of an NCP complex during the time it transits the confocal spot (1–5 ms duration, depending on the trajectory of the molecule and the focus size). Each burst is analyzed in terms of the proximity ratio, E (Eq. S1 in the Supporting Material, related to, or proportional to, the FRET efficiency), and the stoichiometry ratio, S (Eq. S2 in the Supporting Material) and the results are placed in a two-dimensional E/S histogram (see Fig. 1). The proximity ratio reports on the donor-acceptor distance, and the stoichiometry ratio on the donor/acceptor stoichiometry. Use of the ALEX technique enabled the removal of donor-only and acceptor-only populations that exist in small quantities in all histograms due to fluorophore bleaching and incomplete labeling. Only populations with the correct S values, centered on S = 0.5 (see Fig. 1, purple dotted rectangle), were selected to construct the E histograms and the kinetic profiles. All E histograms in this study are broader than the width expected from photon shot-noise statistics (34, 39), and the main peaks are accompanied by histogram tails or minor additional population(s). The broad peaks and the distribution of E values most likely indicate that the NCPs are not entirely homogeneous. In this work, however, analyses are focused on the main peaks; peak widths and tails were not interpreted further.

Figure 1.

Figure 1

Open in a new tab

smFRET/ALEX analyses of NCPs at three different NCP and NaCl concentrations, including a schematic of the NCP and an E/S histogram measured in each condition. (A) Measurement at 3 pM labeled NCP and 5 mM NaCl, with no unlabeled NCP present in the solution. Only an HF population, corresponding to intact NCP, was observed, and the NCP complex remained intact for the 2 h duration of the experiment. (B) Measurement at 3 pM labeled NCP at 80 mM NaCl with no unlabeled NCP present in the solution. The HF population decreased with time after the introduction of the NaCl, and an LF population, corresponding to free DNA, increased. (C) Measurement at 3 pM labeled NCP, 1 nM unlabeled NCP, and 750 mM NaCl. The HF population decreased with time and a low LF population increased. An MF population increased and then decreased with time. The MF population may correspond to NCP with one DNA arm unwrapped (3). To see this figure in color, go online.

Kinetic measurements and determination of rate constants

The NCP dissociation and association reactions were studied by monitoring the time evolution of the observed populations initiated by the increase in NaCl concentration. At time zero, the labeled NCP sample (stored at 3 μM [NCP] and 5 mM [NaCl]) was diluted to a concentration that allowed single-molecule observation (3 pM) in a buffer containing different concentrations of unlabeled NCP and NaCl (Supporting Material and Fig. S4). The resultant E histograms were fitted using two or three Gaussian functions, which describe the two or three different populations observed (Eq. S3 in the Supporting Material), and the relative sizes of these populations were determined as a function of time. To determine the transition rate constants, the kinetic profiles of the first-order reaction in Fig. 3 were fitted using an exponential function (Eq. S4 in the Supporting Material). Because the kinetic profiles at Figs. 4, 5, 6, and 7 are a result of complex two- and three-state dynamics, we fit the data using numerically simulated dynamic models (see the Supporting Material).

Figure 3.

Figure 3

Open in a new tab

Salt-induced NCP dissociation at 3 pM NCP concentration. (A) NCP dissociation kinetic profiles measured at different NaCl concentrations. At time zero, 3 pM NCP was introduced into buffered solution containing different NaCl concentrations. The HF population fractions, calculated every 300 s from smFRET/ALEX data, are plotted as a function of time. The solid lines are fits of the data using a first-order dissociation model (Eq. S4 in the Supporting Material). (B) Plot of the dissociation rate constants (kHF→LF) as a function of NaCl concentration. To see this figure in color, go online.

Figure 4.

Figure 4

Open in a new tab

Nucleosome dissociation and dependence of NCP stability on NCP concentration. (A) HF fraction as a function of time after addition to 80 mM NaCl at different NCP concentrations. (B) HF fraction as a function of time after addition to 300 mM NaCl in different NCP concentrations. The solid lines are fits to a two-state model. (C) Association rates calculated from the two-state fits. (D) Fraction of HF in equilibrium at different NCP concentrations. To see this figure in color, go online.

Figure 5.

Figure 5

Open in a new tab

Salt-induced NCP dissociation at 1 nM NCP concentration. (Left) HF fraction as a function of time after addition to the indicated NaCl concentration. (Right) Fraction of HF in equilibrium as a function of NaCl concentration. To see this figure in color, go online.

Figure 6.

Figure 6

Open in a new tab

Nucleosome dynamics in high salt concentration. (A) Fractions of HF, MF, and LF as a function of time after addition to the indicated NaCl concentration at 1 nM NCP. (B) Fractions of HF, MF, and LF as a function of time after addition to 900 mM NaCl at the indicated NCP concentrations. The lines are to guide the eye, and representative fitting of the kinetic curves is shown in Figs. 7A and S8. (C) The fractions of HF, MF, and LF at time 5000 s taken from the graphs in (A) and (B). Examples of fitting of the kinetic curves are shown in Figs. 7 and S8). To see this figure in color, go online.

Figure 7.

Figure 7

Open in a new tab

Nucleosome dynamics at high salt concentration. (A) Typical reaction profiles demonstrating the decrease of the HF population, the increase followed by a decrease of the MF intermediate population, and the increase of the LF population. More examples are given in Fig. S8. (B and C) Summaries of the calculated rate constants measured at 1 nM NCP (B) and at 900 mM NaCl (C). To see this figure in color, go online.

Calculating standard deviation

The error bars for the fraction of high FRET population (see Figs. 2, 4 D, and 5, right) and for the rate constants (see Figs. 3 B, 4 C, and 7, B and C) are the standard deviations calculated from three independent experiments measured under identical conditions.

Figure 2.

Figure 2

Open in a new tab

The fraction of HF population measured 900–1200 s after introduction of the indicated concentration of NaCl and at 1 nM unlabeled NCP concentration. The increase in NCP stability from 150 to 650 mM NaCl is a result of salt-induced stabilization of the dimers and possibly tetramers (see main text).

Results and Discussion

General observations

For clarity, we introduce first our general observations, followed by detailed analysis at different ranges of salt concentration. The data from single-pair FRET with ALEX are displayed in Fig. 1 as two-dimensional histograms of the stoichiometry (S) versus the proximity ratio (E) measured at three NaCl concentrations and two unlabeled NCP concentrations (for more information about the method, see Materials and Methods, the Supporting Material, and Fig. S5, and for more representative E/S histograms, see Fig. S6). Single-molecule fluorescence requires a low concentration of labeled molecules. To achieve higher concentrations of NCP, the concentration of labeled NCP was maintained at 3 pM, and different concentrations of unlabeled NCP were used (see the Supporting Material and Fig. S4). Depending on conditions, two or three peaks can be distinguished in the FRET histograms. High-FRET values (HF, E = ∼0.65) were assigned as intact NCP, low-FRET values (LF, E = ∼0.22) as free DNA, and medium-FRET values (MF, E = ∼0.40) as an intermediate state in which one DNA arm might be partially unwrapped and/or a histone dimer dissociated (3). These results, and the kinetic data detailed below, demonstrate that at different NaCl and NCP concentration regions, NCPs behave in significantly different ways. At 5 mM NaCl and in the absence of unlabeled NCP (Fig. 1 A), the E histogram consists of a predominantly HF population and only a very minor LF population. Under these low salt conditions the NCP remained stable for >2 h. In contrast, at 80 mM NaCl, the HF population decreased with time after the introduction of NaCl, and the LF population increased (Fig. 1 B), indicating that salt induced dissociation of the NCP. No intermediate MF population was observed in the range 80–300 mM NaCl concentration at any NCP concentration (Fig. S6). In high NaCl concentrations and in the presence of elevated unlabeled NCP (i.e., 750 mM NaCl and 1 nM NCP (Fig. 1 C)), an MF population was observed in addition to the HF and LF populations (3). The HF population decreased and the LF population increased with time after the introduction of NaCl, and the MF population first increased and then decreased over time. In such high salt concentrations, low NCP concentrations (several picomolars) resulted in almost complete dissociation. Three populations were observed previously by Gansen et al. (40), but due to the different labeling positions in that study, compared to the study presented here, their HF population corresponded to the intermediate state and their MF population to the folded NCP.

Previous studies have shown that an increase in the concentration of NaCl reduces NCP stability (3, 4, 5, 6, 7, 9, 10, 11, 12, 13, 14, 15, 16). Here, using NaCl concentrations from 5 mM to 1200 mM, we observed that the dependence of NCP stability on NaCl concentration was nonmonotonic. As demonstrated in Fig. 2, the fraction of the intact HF population measured 900–1200 s after an increase of NaCl concentration to the indicated value in the presence of 1 nM unlabeled NCP was lower at near-physiological conditions (80–400 mM NaCl) than at lower or higher salt concentrations.

A similar behavior has been observed in previous studies of salt-induced destabilization of nucleosomes on double-labeled DNA, but with the labels in different positions (40). In that study, a nonmonotonic behavior of the overall FRET signal at intermediate NaCl concentrations was observed that was similar to the dip in the HF population observed here. At higher salt concentrations, similar to where we observe the MF peak here, Gansen et al. observed a third peak at higher FRET than the main peak. Both these peaks could indicate the same state, with the DNA partially unwrapped or shifted, but with an increase in FRET in the previous study (40) and a decrease here, due to the different dye positions.

Salt-induced dissociation of NCP at very low NCP concentrations

At very low NCP concentration (i.e., no unlabeled NCP added) the salt-induced dissociation is practically irreversible. The 3 pM concentration of labeled NCP used in these measurements is low enough to avoid reassembly of NCP from its components (DNA, dimers, and tetramers), enabling direct observation of the dissociation kinetics. Only two peaks, HF and LF, were observed in the E histograms under these conditions (Fig. 1 A). The fraction of HF population as a function of time is plotted for different NaCl concentrations in Fig. 3 A.

Since the NCP concentration was very low and the data showed that no reassembly reaction was taking place (given enough time and salt concentration, all the HF population disappears), we assumed a first-order dissociation reaction (HF → LF). Knowing the initial HF fraction (∼0.85), we characterized the decay by a forced-fit single exponential (Eq. S4 in the Supporting Material). This resulted in a characteristic relaxation time at each salt concentration and from that, an apparent first-order rate constant. Although the data deviated from strict monoexponential behavior, their quality did not allow fitting a more complicated model.

The obtained rate constants (kHF → LF) are summarized in Fig. 3 B. The dissociation rates increased monotonically with the salt concentration, reaching the technical time-resolution limit of ∼40 s at ∼150 mM NaCl. These results are in agreement with previous data (3) and can be explained by the reduced electrostatic attraction between the positively charged histone proteins and the negatively charged DNA. Thus, a higher concentration of ions results in faster dissociation. At this low concentration of histones, the reassembly reaction does not occur, and eventually all NCPs dissociate. In addition to the histone-free DNA, the LF population may also include a tetrasome structure, in which the (H3-H4)2 tetramer is still bound to DNA but without significantly bending the DNA.

Dissociation and association of NCP in the presence of unlabeled NCP

Next, we studied the influence of NCP concentration on the stability of NCP within the physiologically relevant range of NaCl concentrations (150 mM NaCl). Labeled NCP (3 pM) was introduced into a solution containing different concentrations of unlabeled NCP and either 80 mM or 300 mM NaCl, and the change in relative populations was monitored over time. In this salt range, only two peaks, HF and LF, were observed in the E histograms at all NCP concentrations (Figs. 1 B and S6). The plot of the fraction of HF versus time shows a rapid decrease of the HF population and an increase in the LF population to a constant value that was different for different NaCl and NCP concentrations (Fig. 4). The simplest model to explain these observations is a two-state model with a forward and a backward reaction (HF⇄LF), and the data were fitted accordingly (for details of the analysis, see Materials and Methods and the Supporting Material). The reactions were somewhat too fast for the disassembly and assembly rate constants (kHF → LF and kHF ← LF, respectively) to be independently and accurately determined; however, the ratio between the rates (HF(t = ∞) = kHF ← LF/(kHF → LF + kHF ← LF)), which is equal to the fraction of HF in equilibrium, could be determined reliably. In the case of the 80 mM NaCl experiment, the disassembly rate constants were fixed to the value determined in the absence of unlabeled NCP (kHF → LF = 5 × 10−3s−1; see Fig. 3 B), and only the reassembly rates were fitted (Fig. 4, A and C). When both the association and the dissociation rate constants were fitted, the dissociation rates appeared to be roughly constant, whereas the association rates increased with increasing NCP concentration (Fig. S7). In the case of 300 mM NaCl, the dissociation rates in the absence of NCP are faster than the experimental time resolution, and therefore, only equilibrium HF concentrations were calculated (Fig. 4, B and D). The results indicate that both the association rates and the HF population in equilibrium increased roughly linearly with the increase in NCP concentration, and the equilibrium dissociation constant (Kd) was calculated to be 0.25 for 80 mM NaCl and 4 nM for 300 mM NaCl.

In this range of salt concentrations the disassembly process, which is a unimolecular reaction, does not appear to depend on the concentration of NCP, although we cannot rule out some minor (undetectable) dependence. Assuming that unfolded NCP contributes histone dimers, and possibly tetramers (unbound to DNA), to the solution, the disassembly process appears not to depend on the concentration of dimers and tetramers either, and it only accelerates with increasing NaCl concentration. On the other hand, the reassembly reaction is NCP-concentration-dependent, because the pool of unlabeled NCPs supplies histone dimers and tetramers, shifting the equilibrium toward intact NCP.

NCP stability at intermediate to high NaCl concentrations

We then studied the stability of NCP at NaCl concentrations in the range 300–700 mM. Although a high NaCl concentration is not directly physiologically relevant, investigation of a high salt concentration is a common tool for exploring possible pathways for nucleosome disassembly. It may help explain intermediates and may also allow back-extrapolation to physiological conditions. In fact, the intermediate state discovered in Böhm et al. (9) could be estimated to occur at a fraction of ∼1% in physiological conditions. It may also explain the in vitro salt-induced NCP reconstitution (38). Fig. 5 shows the kinetic profiles of salt-induced NCP dissociation measured for 1 nM NCP at different NaCl concentrations. Under these conditions, the reactions were too fast to allow independent determination of the disassembly and assembly reaction rate constants. Therefore, only the equilibrium HF was determined. In the presence of 1 nM unlabeled NCP, from ∼250 mM NaCl and up to ∼650 mM NaCl, increases in NaCl concentration increase the NCP stability, in agreement with the data reported in our earlier work (40).

Ions may contribute to the stability of NCP in several ways. It was shown that increased concentration of ions increases the stability of H2A/H2B histone dimers and (H3/H4)2 tetramers (19, 20, 21, 22). For isolated H2A-H2B dimers, it was suggested that this stabilization results from both salt-induced hydrophobic effects (via the Hofmeister effect and preferential hydration) and screening of electrostatic repulsion between the positively charged histones (22). Increased dimer and tetramer stability may increase the stability of NCP by two different mechanisms. First, it might increase the concentration of intact dimers and tetramers not bound to DNA in the solution, which would increase the NCP association rate, as demonstrated in Fig. 4. Second, stabilized dimers and tetramers may better maintain the specific conformation that enables strong binding with DNA, which would promote refolding. For example, when one arm of the NCP is unwrapped, it is more likely that this arm will rewrap when the H2A/H2B dimer adjacent to the arm (whether it is bound to the unwrapped arm or comes from the solution and binds to the arm) maintains a conformation that encourages refolding. The time resolution of our experimental technique does not allow us to determine unequivocally whether the increase in stability is a result of a decrease in dissociation rates or an increase in association rates or both; therefore, currently we cannot determine which of these processes contribute, and to what degree, to the salt-induced NCP stabilization in the 250–650 NaCl concentration range.

NCP dynamics in high NaCl concentration and with the presence of an intermediate state

At high NaCl concentrations, a third population, the intermediate MF population, is observed (Fig. 1 C) (3). The HF, MF, and LF population fractions were measured as a function of time at 1 nM unlabeled NCP over the NaCl concentration range from 760 to 1200 mM (Fig. 6 A; for E/S histograms, see Fig. S6). We also evaluated these fractions as a function of time at 900 mM NaCl and over a range of NCP concentrations from 3 pM to 3 nM (Fig. 6 B). For clarity, the fractions of HF, MF, and LF measured 5000 s after the initiation of the salt-induced dissociation are also presented (Fig. 6 C). In this NaCl concentration range, the fraction of HF continuously decreases and that of LF increases. The fraction of the MF state has a maximum at ∼900 mM NaCl.

Analysis of the time evolution of the populations indicates the following trends: in all cases, the fractions of HF decrease and the fractions of LF increase with time. Under most conditions, the fraction of MF first increased and then either decreased (Figs. 7 A and S8) or plateaued. As was observed for lower NaCl concentrations (Figs. 4 and 5), in the presence of sufficient NCP concentration, not all HF (or MF) disappeared with time. A minimal model that can describe these observations adequately includes HF and LF states that interconvert via the intermediate MF state (HF⇄MF⇄LF). The kinetic data were analyzed using this model (see Fig. 7 A, for example) and the four obtained rate constants are summarized in Fig. 7, B and C.

The data show conclusively that the dissociation rates (kHF → MF and kMF → LF) increased with the increase of NaCl concentration, and, interestingly, the association rates (kHF → MF and kMF → LF) also increased, although less dramatically. The results for the different NCP concentrations are somewhat less clear. The transition rate from HF to MF did not significantly change as the NCP concentration increased, and it somewhat decreased for the transition from MF to LF. The association rates from LF to MF and from MF to HF increased with the increase in NCP concentration above 1 nM. We hypothesize that as the higher NaCl concentration weakens the DNA-histone interaction, it also increases the transition rate from the folded NCP to the intermediate state and from the intermediate state to the unfolded state. The increase in the association rate with the increase in NaCl concentration (Fig. 7 B) may be a result of the increased concentration of free dimers, and possibly tetramers, due to salt-induced stabilization of dimers and tetramers not bound to DNA, as discussed above. In addition, it is possible that the increased stability of the dimers accelerates refolding in cases where the dimers do not dissociate from the DNA open arm in the transition state. In a similar fashion, higher concentrations of NCP, which increase the concentration of available dimers and tetramers, accelerate refolding of partially unfolded and possibly fully unfolded states. The MF state is observed only at higher NaCl concentrations (>700 mM), and it can be assigned as an NCP with one DNA arm opened, with or without the adjacent dimer (MF and MF∗, respectively; see NCP schematics in Fig. 1 C) (3).

The decrease in the MF→LF transition rate with increasing NCP concentration may necessitate extension of the minimal model to include another fast concentration-dependent preequilibrium in the MF state (MF⇄MF∗) in which conversion to the LF state occurs (mainly) from the MF∗ state. Experiments to explore this phenomenon in more detail are currently underway.

To explain the presence of the intermediate state at high salt concentration, we propose that the salt stabilizes the histone-histone interactions (dimers, tetramers, and dimer-tetramer interactions). This may help to maintain the specific dimer and tetramer conformations that enable strong interaction with the DNA and reduce the dimer-tetramer repulsion in the open-arm structure, stabilizing the complex. In lower NaCl concentrations (<500 mM NaCl; see Fig. S6), the dimers and tetramers are less stable and more repulsive, destabilizing the open-arm complex.

By labeling the H2A or H2B histone proteins it should be possible to determine the presence or absence of one or two H2A/H2B dimers in the intermediate state using the ALEX technique. However, our preliminary data were not conclusive (data not shown), and currently, we cannot determine under which conditions the H2A/H2B dimer remains attached to the open arm (see NCP schematics in Fig. 1 C).

Summary of the dependence of NCP stability on NCP concentration

As was demonstrated in prior studies (3, 4, 5, 6, 7), NCP is stabilized by increasing its concentration. We suggest that this stabilization is mainly a result of an increased concentration of H2A/H2B dimers (and possibly tetramers) not bound to DNA, which results in faster refolding. Our results permit a rough estimate of the dissociation constant (half of the population is in the LF state), which is ∼4 nM for 300 mM salt and ∼0.25 nM for 80 mM salt.

Summary of the dependence of NCP stability on NaCl concentration

In light of our results, we propose that even in the absence of free histone proteins in the solution, at a very low salt concentration (<5 mM), the NCP is very stable. The attractive forces between the negatively charged DNA and the positively charged histone proteins, together with the specific DNA protein interactions, are sufficient to hold the NCP components together and maintain the folded conformation. However, even a low concentration of NaCl is sufficient to reduce the attractive forces such that the NCP quickly unfolds. It is important to remember, however, that this is the case for the stable Widom-601 sequence used in this study, and that other sequences are less stable (40).

We assume that the NCP structure at NaCl concentrations above the physiological range (300 mM and above) is similar to the earlier-postulated partially open state found using FRET (9). In the range from 300 to 700 mM NaCl and in the presence of NCP (1 nM), this state is stabilized as the NaCl concentration increases, an effect not explained before. We propose that this is a result of salt-induced stabilization of the dimers and tetramers, which may contribute to the stability of NCP. The higher salt concentration may increase the fraction of correctly folded free dimers and tetramers (22), which would increase the association rate. Furthermore, it stabilizes dimers and tetramers in the NCP complex, which decreases the dissociation rate. In high NaCl concentrations, an intermediate state is observed, similar to the one described in our earlier work (40). We propose that this state is stabilized by salt-induced stabilization of the dimers, and possibly tetramers, whereas DNA-histone and dimer-tetramer interactions are sufficiently reduced to open the nucleosome globally, leading to a state with one DNA arm open (3), or the butterfly state proposed in a previous study (9). In addition, it is likely that because of the salt-induced stabilization of the dimers and tetramers, these complexes maintain their morphology and remain intact to a greater extent to the DNA in this salt concentration range than at lower salt.

Under conditions studied here (for example, 1 nM NCP concentration (Fig. 2)), and for the Widom-601 sequence, the negative effect of the salt on NCP stability predominates around the physiological salt concentration over the positive effect of salt-induced stabilization of the dimers and tetramers. For that reason, the NCP stability is lower in the range 150–350 mM NaCl than at higher NaCl concentrations. At 1 nM NCP concentration, salt-induced stabilization of dimers and tetramers does not have a dominant effect until NaCl concentration is >350 mM. As was shown before, however, higher NCP concentrations may supersede this effect, as the concentration of dimers and tetramers not yet bound to DNA is sufficient to prevent NCP dissociation at NaCl concentrations <500 mM (6, 11, 13, 16).

Conclusions

Use of single-molecule fluorescence methods enabled us to distinguish among three different NCP states and to measure the transition rates between these states. We confirm our previous observation of nonmonotonic dependence of NCP stability on NaCl concentration (40). The dependence of the stability of the three states and the transition rates on NaCl and NCP concentrations was carefully studied. We proposed a mechanism to explain this nonmonotonic behavior, as well as the presence of an intermediate open-arm conformation state. We suggest that NCP stability is influenced by an interplay of repulsive and attractive forces between positively charged histones and between the histones and the negatively charged DNA, an effect that was previously unnoticed or rarely discussed. We showed that the concentration of NCP influences its stability, most likely due to increased concentration of free histone dimers and tetramers. We show that at physiological salt concentrations, NCP stability strongly depended on NaCl and NCP concentrations.

What do these salt-induced dissociation experiments tell us about the function of the nucleosome at physiological salt levels? We can presume that other external forces acting on chromatin, such as topological stress, molecular machines such as chromatin-remodeling factors, or other chromatin-associated proteins, will change the mechanical and electrostatic environment of the nucleosome. For instance, it is probable that modification of the histone tails acts in part through the charge neutralization of lysines (41, 42, 43), and it has been proposed that such changes may induce internal destabilizing transitions in the nucleosome core (44). The salt-induced destabilization study presented here suggests possible pathways for nucleosome opening under external forces and opens the way for further research into the forces that hold the nucleosome subunits together. Partially opened states, such as those characterized here, may then act as nucleation sites for proteins that need to access DNA. Furthermore, this suggests the possibility of mechanisms where changes in ions or other charged molecules, and in histone concentrations, may signal chromatin morphological changes during the cell cycle or cell differentiation.

Author Contributions

N.P.H. conducted most of the sample preparation, experiments, and data analysis, prepared the figures, and assisted with preparation of the manuscript. T.E.T. and R.T. assisted with data analysis, the optical setup alignment, and the manuscript. T.E.T., R.T., M.L., Y.B., and R.M. assisted with DNA labeling and with data collection. K.T. and J.L. provided valuable suggestions for sample preparation and assisted in manuscript preparation. E.N. supervised the research and prepared the manuscript.

Acknowledgments

We thank Ashraf Brik and Raz Zarivach for their support with high-performance liquid chromatography.

T.E.T. and Y.B were supported by a Negev Fellowship, M.L. by a Darom Fellowship, and E.N. by an Alon Fellowship.

Editor: Elizabeth Rhoades.

Footnotes

Rula Masoud’s present address is Department of Biomedical Engineering, Technion, Israel Institute of Technology, 32000 Haifa, Israel.

Supporting Materials and Methods, Supporting Results, eight figures, and one table are available at http://www.biophysj.org/biophysj/supplemental/S0006-3495(15)00675-X.

Supporting Material

Document S1. Supporting Materials and Methods, Supporting Results, eight figures, and one table
mmc1.pdf (1.9MB, pdf)
Document S2. Article plus Supporting Material
mmc2.pdf (3.4MB, pdf)

References

  • 1.Olins A.L., Olins D.E. Spheroid chromatin units (v bodies) Science. 1974;183:330–332. doi: 10.1126/science.183.4122.330. [DOI] [PubMed] [Google Scholar]
  • 2.Workman J.L., Kingston R.E. Alteration of nucleosome structure as a mechanism of transcriptional regulation. Annu. Rev. Biochem. 1998;67:545–579. doi: 10.1146/annurev.biochem.67.1.545. [DOI] [PubMed] [Google Scholar]
  • 3.Gansen A., Valeri A., Seidel C.A. Nucleosome disassembly intermediates characterized by single-molecule FRET. Proc. Natl. Acad. Sci. USA. 2009;106:15308–15313. doi: 10.1073/pnas.0903005106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Gansen A., Hauger F., Langowski J. Single-pair fluorescence resonance energy transfer of nucleosomes in free diffusion: optimizing stability and resolution of subpopulations. Anal. Biochem. 2007;368:193–204. doi: 10.1016/j.ab.2007.04.047. [DOI] [PubMed] [Google Scholar]
  • 5.Gansen A., Tóth K., Langowski J. Structural variability of nucleosomes detected by single-pair Förster resonance energy transfer: histone acetylation, sequence variation, and salt effects. J. Phys. Chem. B. 2009;113:2604–2613. doi: 10.1021/jp7114737. [DOI] [PubMed] [Google Scholar]
  • 6.Hoch D.A., Stratton J.J., Gloss L.M. Protein-protein Förster resonance energy transfer analysis of nucleosome core particles containing H2A and H2A.Z. J. Mol. Biol. 2007;371:971–988. doi: 10.1016/j.jmb.2007.05.075. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Gottesfeld J.M., Luger K. Energetics and affinity of the histone octamer for defined DNA sequences. Biochemistry. 2001;40:10927–10933. doi: 10.1021/bi0109966. [DOI] [PubMed] [Google Scholar]
  • 8.Thåström A., Gottesfeld J.M., Widom J. Histone-DNA binding free energy cannot be measured in dilution-driven dissociation experiments. Biochemistry. 2004;43:736–741. doi: 10.1021/bi0302043. [DOI] [PubMed] [Google Scholar]
  • 9.Böhm V., Hieb A.R., Langowski J. Nucleosome accessibility governed by the dimer/tetramer interface. Nucleic Acids Res. 2011;39:3093–3102. doi: 10.1093/nar/gkq1279. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Kelbauskas L., Chan N., Lohr D. Sequence-dependent variations associated with H2A/H2B depletion of nucleosomes. Biophys. J. 2008;94:147–158. doi: 10.1529/biophysj.107.111906. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Oohara I., Wada A. Spectroscopic studies on histone-DNA interactions. II. Three transitions in nucleosomes resolved by salt-titration. J. Mol. Biol. 1987;196:399–411. doi: 10.1016/0022-2836(87)90700-5. [DOI] [PubMed] [Google Scholar]
  • 12.Li G., Widom J. Nucleosomes facilitate their own invasion. Nat. Struct. Mol. Biol. 2004;11:763–769. doi: 10.1038/nsmb801. [DOI] [PubMed] [Google Scholar]
  • 13.Bönisch C., Schneider K., Hake S.B. H2A.Z.2.2 is an alternatively spliced histone H2A.Z variant that causes severe nucleosome destabilization. Nucleic Acids Res. 2012;40:5951–5964. doi: 10.1093/nar/gks267. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Koopmans W.J., Buning R., van Noort J. spFRET using alternating excitation and FCS reveals progressive DNA unwrapping in nucleosomes. Biophys. J. 2009;97:195–204. doi: 10.1016/j.bpj.2009.04.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Lovullo D., Daniel D., Woodbury N.W. A fluorescence resonance energy transfer-based probe to monitor nucleosome structure. Anal. Biochem. 2005;341:165–172. doi: 10.1016/j.ab.2005.03.022. [DOI] [PubMed] [Google Scholar]
  • 16.Park Y.J., Dyer P.N., Luger K. A new fluorescence resonance energy transfer approach demonstrates that the histone variant H2AZ stabilizes the histone octamer within the nucleosome. J. Biol. Chem. 2004;279:24274–24282. doi: 10.1074/jbc.M313152200. [DOI] [PubMed] [Google Scholar]
  • 17.Yager T.D., McMurray C.T., van Holde K.E. Salt-induced release of DNA from nucleosome core particles. Biochemistry. 1989;28:2271–2281. doi: 10.1021/bi00431a045. [DOI] [PubMed] [Google Scholar]
  • 18.Yager T.D., van Holde K.E. Dynamics and equilibria of nucleosomes at elevated ionic strength. J. Biol. Chem. 1984;259:4212–4222. [PubMed] [Google Scholar]
  • 19.Karantza V., Baxevanis A.D., Moudrianakis E.N. Thermodynamic studies of the core histones: ionic strength and pH dependence of H2A-H2B dimer stability. Biochemistry. 1995;34:5988–5996. doi: 10.1021/bi00017a028. [DOI] [PubMed] [Google Scholar]
  • 20.Karantza V., Freire E., Moudrianakis E.N. Thermodynamic studies of the core histones: pH and ionic strength effects on the stability of the (H3-H4)/(H3-H4)2 system. Biochemistry. 1996;35:2037–2046. doi: 10.1021/bi9518858. [DOI] [PubMed] [Google Scholar]
  • 21.Karantza V., Freire E., Moudrianakis E.N. Thermodynamic studies of the core histones: stability of the octamer subunits is not altered by removal of their terminal domains. Biochemistry. 2001;40:13114–13123. doi: 10.1021/bi0110140. [DOI] [PubMed] [Google Scholar]
  • 22.Gloss L.M., Placek B.J. The effect of salts on the stability of the H2A-H2B histone dimer. Biochemistry. 2002;41:14951–14959. doi: 10.1021/bi026282s. [DOI] [PubMed] [Google Scholar]
  • 23.Li G., Levitus M., Widom J. Rapid spontaneous accessibility of nucleosomal DNA. Nat. Struct. Mol. Biol. 2005;12:46–53. doi: 10.1038/nsmb869. [DOI] [PubMed] [Google Scholar]
  • 24.Claudet C., Angelov D., Bednar J. Histone octamer instability under single molecule experiment conditions. J. Biol. Chem. 2005;280:19958–19965. doi: 10.1074/jbc.M500121200. [DOI] [PubMed] [Google Scholar]
  • 25.White C.L., Luger K. Defined structural changes occur in a nucleosome upon Amt1 transcription factor binding. J. Mol. Biol. 2004;342:1391–1402. doi: 10.1016/j.jmb.2004.07.080. [DOI] [PubMed] [Google Scholar]
  • 26.Andrews A.J., Chen X., Luger K. The histone chaperone Nap1 promotes nucleosome assembly by eliminating nonnucleosomal histone DNA interactions. Mol. Cell. 2010;37:834–842. doi: 10.1016/j.molcel.2010.01.037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Akey C.W., Luger K. Histone chaperones and nucleosome assembly. Curr. Opin. Struct. Biol. 2003;13:6–14. doi: 10.1016/s0959-440x(03)00002-2. [DOI] [PubMed] [Google Scholar]
  • 28.Zlatanova J., Bishop T.C., van Holde K. The nucleosome family: dynamic and growing. Structure. 2009;17:160–171. doi: 10.1016/j.str.2008.12.016. [DOI] [PubMed] [Google Scholar]
  • 29.Tomschik M., Zheng H., Leuba S.H. Fast, long-range, reversible conformational fluctuations in nucleosomes revealed by single-pair fluorescence resonance energy transfer. Proc. Natl. Acad. Sci. USA. 2005;102:3278–3283. doi: 10.1073/pnas.0500189102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Koopmans W.J., Brehm A., van Noort J. Single-pair FRET microscopy reveals mononucleosome dynamics. J. Fluoresc. 2007;17:785–795. doi: 10.1007/s10895-007-0218-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Tomschik M., van Holde K., Zlatanova J. Nucleosome dynamics as studied by single-pair fluorescence resonance energy transfer: a reevaluation. J. Fluoresc. 2009;19:53–62. doi: 10.1007/s10895-008-0379-1. [DOI] [PubMed] [Google Scholar]
  • 32.Koopmans W.J., Schmidt T., van Noort J. Nucleosome immobilization strategies for single-pair FRET microscopy. ChemPhysChem. 2008;9:2002–2009. doi: 10.1002/cphc.200800370. [DOI] [PubMed] [Google Scholar]
  • 33.Luo Y., North J.A., Poirier M.G. Single molecule fluorescence methodologies for investigating transcription factor binding kinetics to nucleosomes and DNA. Methods. 2014;70:108–118. doi: 10.1016/j.ymeth.2014.09.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Tsukanov R., Tomov T.E., Nir E. Conformational dynamics of DNA hairpins at millisecond resolution obtained from analysis of single-molecule FRET histograms. J. Phys. Chem. B. 2013;117:16105–16109. doi: 10.1021/jp411280n. [DOI] [PubMed] [Google Scholar]
  • 35.Tsukanov R., Tomov T.E., Nir E. Detailed study of DNA hairpin dynamics using single-molecule fluorescence assisted by DNA origami. J. Phys. Chem. B. 2013;117:11932–11942. doi: 10.1021/jp4059214. [DOI] [PubMed] [Google Scholar]
  • 36.Tomov T.E., Tsukanov R., Nir E. Disentangling subpopulations in single-molecule FRET and ALEX experiments with photon distribution analysis. Biophys. J. 2012;102:1163–1173. doi: 10.1016/j.bpj.2011.11.4025. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Masoud R., Tsukanov R., Nir E. Studying the structural dynamics of bipedal DNA motors with single-molecule fluorescence spectroscopy. ACS Nano. 2012;6:6272–6283. doi: 10.1021/nn301709n. [DOI] [PubMed] [Google Scholar]
  • 38.Dyer P.N., Edayathumangalam R.S., Luger K. Reconstitution of nucleosome core particles from recombinant histones and DNA. Methods Enzymol. 2004;375:23–44. doi: 10.1016/s0076-6879(03)75002-2. [DOI] [PubMed] [Google Scholar]
  • 39.Nir E., Michalet X., Weiss S. Shot-noise limited single-molecule FRET histograms: comparison between theory and experiments. J. Phys. Chem. B. 2006;110:22103–22124. doi: 10.1021/jp063483n. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Gansen A., Hieb A.R., Langowski J. Closing the gap between single molecule and bulk FRET analysis of nucleosomes. PLoS One. 2013;8:e57018. doi: 10.1371/journal.pone.0057018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Potoyan D.A., Papoian G.A. Regulation of the H4 tail binding and folding landscapes via Lys-16 acetylation. Proc. Natl. Acad. Sci. USA. 2012;109:17857–17862. doi: 10.1073/pnas.1201805109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Tóth K., Bühm V., Langowski J. Histone- and DNA sequence-dependent stability of nucleosomes studied by single-pair FRET. Cytometry A. 2013;83:839–846. doi: 10.1002/cyto.a.22320. [DOI] [PubMed] [Google Scholar]
  • 43.Erler J., Zhang R., Langowski J. The role of histone tails in the nucleosome: a computational study. Biophys. J. 2014;107:2911–2922. doi: 10.1016/j.bpj.2014.10.065. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Biswas M., Voltz K., Langowski J. Role of histone tails in structural stability of the nucleosome. PLOS Comput. Biol. 2011;7:e1002279. doi: 10.1371/journal.pcbi.1002279. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Document S1. Supporting Materials and Methods, Supporting Results, eight figures, and one table
mmc1.pdf (1.9MB, pdf)
Document S2. Article plus Supporting Material
mmc2.pdf (3.4MB, pdf)

Articles from Biophysical Journal are provided here courtesy of The Biophysical Society

RESOURCES