High precision FRET studies reveal reversible transitions in nucleosomes between microseconds and minutes

Abstract

Nucleosomes play a dual role in compacting the genome and regulating the access to DNA. To unravel the underlying mechanism, we study fluorescently labeled mononucleosomes by multi-parameter FRET measurements and characterize their structural and dynamic heterogeneity upon NaCl-induced destabilization. Species-selective fluorescence lifetime analysis and dynamic photon distribution analysis reveal intermediates during nucleosome opening and lead to a coherent structural and kinetic model. In dynamic octasomes and hexasomes the interface between the H2A-H2B dimers and the (H3-H4)₂ tetramer opens asymmetrically by an angle of ≈20° on a 50 and 15 µs time scale, respectively. This is followed by a slower stepwise release of the dimers coupled with DNA unwrapping. A mutation (H2A-R81A) at the interface between H2A and H3 facilitates initial opening, confirming the central role of the dimer:tetramer interface for nucleosome stability. Partially opened states such as those described here might serve as convenient nucleation sites for DNA-recognizing proteins.

Fig. 1

Overview on FRET dye positions in the nucleosome constructs. The names of the constructs assign the position and side of the fluorescent labels: first the donor (green circle) and second the acceptor (red circle). In the cartoon representations, the H2A-H2B dimers are shown in orange, the (H3-H4)₂ tetramer in turquoise, DNA in gray. To account for the asymmetry, we denote by α the left side of the DNA sequence (forward (-) strand) with base pairs counted in negative numbers from the fragment center (position 0). The other side is called β, with base pairs counted in positive numbers ( (+) strand). The dye labeling positions on the DNA are given as relative base shifts to the middle of the sequence: Dy acceptor close to dyad axis at positions −15 and + 15, E donor labeling position at the end of the DNA strand at positions −85 and + 85, I internal labeling position on the DNA + 41 (donor) and −53 (acceptor), H2B donor at position 112 of H2B using the H2B-T112C mutant, mut the H2A-R81A mutation is represented by a black triangle on the heterodimer

Fig. 2

MFD-PIE analysis of H2A-H2B dimer release. The nucleosome construct H2B-Dy_α labeled with Alexa488 on each H2A-H2B dimer and Cy5 on the DNA was used. a Crystal structure (3LZ1) with accessible fluorophore volumes for H2B-Dy_α nucleosomes (green and red clouds for Alexa488 and Cy5, respectively) that describe the spatial distributions of the dyes. Histones are shown in blue (H3), pale green (H4), ruby (H2A), and yellow (H2B). b FRET efficiency versus stoichiometry, S, for 20 pM H2B-Dy_α and 980 pM unlabeled nucleosomes at 150 mM NaCl. Subspecies are identified according to their position in the 2D distribution. c Relative population of donor-only, D_αA, D_βA, and D_αD_βA for 20 pM H2B-Dy_α and 980 pM unlabeled nucleosomes as a function of NaCl concentration with fit to a sigmoid function (Supplementary Note 5, Supplementary Equation 36, global fit with x_{Donor only} = 1−x_DαDβA−x_DαA−x_DβA). D_βA nucleosomes prevail at significantly higher NaCl concentration than D_αA nucleosomes, indicating that H2A-H2B heterodimers are predominantly evicted from the α-side (see details in text)

Fig. 3

Asymmetric unwrapping of DNA ends. Results of DNA unwrapping in the nucleosome constructs E_αDy_β and E_βDy_α (see Methods and Supplementary Note 5) studied by ensemble FRET studies. The dyes were placed at the DNA ends and near the dyad axis to report on unwrapping of each side through loss of FRET. For visualization, data were normalized to the maximal value of the proximity ratio at 100 mM NaCl and approximated by a sigmoidal function (Supplementary Note 5, Supplementary Equation 36) to obtain the c_1/2 salt concentration. The β-side was significantly more resilient against salt-induced unwrapping than the α-side. Error bars are standard errors from three replicates

Fig. 4

Analysis of MFD smFRET data. Results of I_βI_α-wt nucleosome construct studied at 150 mM NaCl labeled with Alexa488 (donor) and Alexa594 (acceptor). a Crystal structure 3LZ1 and accessible fluorophore volumes for donor (green) and acceptor (red). Histones are shown in blue (H3), pale green (H4), ruby (H2A) and yellow (H2B). b, c FRET efficiency (E) versus donor lifetime (〈τ_D〉_F) for 20 pM and 2 nM (20 pM labeled and 1.98 nM unlabeled) nucleosomes. Color coded horizontal lines correspond to FRET efficiency levels of different species. The E - 〈τ_D〉_F relations are presented by black line for static FRET species (Supplementary Methods, Supplementary Equation 40), violet line - for dynamic interconversion between HF (〈τ_D〉_F(HF) = 1.3 ns) and MF* (〈τ_D〉_F(MF*) = 2.65 ns) states and by yellow hypothetical dynamic FRET line connecting MF* and MF. d dynPDA of data shown in b. Experimental FRET efficiency histogram for 2 ms TW and fitted distributions for a model with two interconverting species (MF* and HF) and three static components as Gaussian-distributed distances (MF, LF, and donor-only). The quality of the fit is demonstrated by weighted residuals in the upper panel. The right panel shows the model with the intrinsic distribution of 〈R_DA〉_E for the limit of large photon counts (shot noise free). e Relaxation times $t_{\mathrm{R}}^{\mathrm{dyn}}={\left(k_{{\mathrm{MF}}^{*}\to \mathrm{HF}}+k_{\mathrm{HF}\to {\mathrm{MF}}^{*}}\right)}^{-1}$ and mean FRET efficiency 〈E〉_dyn=x_MFE^MF+x_HFE^HF of dynamic species as a function of nucleosome concentration, calculated from dynPDA fits. x_MF(HF) were obtained from rate constants in f whereas E^MF(HF) were calculated from mean distances of interconverting states (〈R_DA^MF〉_E = 61.2 ± 1.5 Å and 〈R_DA^HF〉_E = 46.7 ± 0.8 Å). f Kinetic rate constants for the MF*⇄HF transition as a function of the free H2A-H2B concentration. Free H2A-H2B is estimated from the total nucleosome concentration and the different species fractions as obtained by PDA. The decrease of k_MF*→HF with increasing free heterodimer concentration implies the existence of two different MF* species. Error bars are standard errors from at least three different measurements. Lines are a weighted global fit of a four-state kinetic model to the data (see Supplementary Equations 16, 17)

Fig. 5

Kinetic parameters of I_βI_α-wt nucleosomes as function of NaCl concentration. The dependence of fractions of the four FRET species on the NaCl concentration were determined from single molecule measurements of 20 pM I_βI_α-wt nucleosomes by dynPDA. The transitions are described by weighted fits to a sigmoidal function (Supplementary Note 5, Supplementary Equation 36). a Salt-dependence of the static fractions x_i of i = LF, MF, MF*, and HF. Intact nucleosomes (MF) are dominant at low salt concentrations but decrease rapidly above 600 mM NaCl in favor of LF and MF*. The fraction of nucleosomes that was found in a dynamic transition between MF* and HF peaks around 800–900 mM NaCl. Global fit with x_MF* = (1 – x_LF – x_MF)/(1 + K_cl/op) and x_HF = (1 – x_LF – x_MF)* K_cl/op/(1 + K_cl/op). The salt-dependent equilibrium constant K_cl/op was obtained from the fit in c. c_1/2 (MF) = 725 ± 25 mM, c_1/2 (LF) = 957 ± 41 mM. b Salt-dependent kinetic rates for the MF* ⇄ HF transition. Transitions from MF* to HF are suppressed at higher NaCl concentrations, while the back reaction from HF to MF* is promoted by increasing salt. Global fit with c_1/2 = 635 ± 60 mM. c Equilibrium constant K_cl/op as obtained from the rates in b: $K_{\mathrm{cl}∕\mathrm{op}}=k_{{\mathrm{MF}}^{*}\to \mathrm{HF}}∕k_{\mathrm{HF}\to {\mathrm{MF}}^{*}}\mathrm{=}{x}_{\mathrm{HF}}∕x_{{\mathrm{MF}}^{*}}$. c_1/2(K_cl/op) = 629 ± 34 mM. Error bars are standard errors from at least three independent measurements

Fig. 6

Effect of the point mutation R81A in H2A on nucleosome dynamics and disassembly. a Crystal structure (3LZ1) with the H2A-R81A mutation (shown in a space filling magenta representation) that modifies the charge at the dimer:tetramer interface. Histones are shown in blue (H3), pale green (H4), ruby (H2A) and yellow (H2B). b Ensemble FRET analysis of salt-induced disassembly of I_βI_α-wt and I_βI_α-mut nucleosomes. Nucleosomes carrying mutated H2A are more susceptible to elevated ionic strength than their non-mutated counterpart; c_1/2(I_βI_α-mut) is about 165 mM lower than c_1/2(I_βI_α-wt). Average values from 6 independent measurements were: c_1/2(I_βI_α-wt) = 783 ± 5 mM and c_1/2(I_βI_α-mut) = 618 ± 7 mM. c MFD-smFRET analysis of 100 pM I_βI_α-wt nucleosomes at 450 mM NaCl. The majority of nucleosomes are found in MF; all three subspecies (LF, MF and dynF) remained stable over time. d MFD-smFRET analysis of 100 pM I_βI_α-mut nucleosomes at 450 mM NaCl. Over time, a significant portion of intact nucleosomes (MF) transit into the dynamic FRET state. The fraction of fully open nucleosomes (LF), however, did not change over time. All colored lines in panels c and d match the lines in Fig. 4a, c. Error bars are standard errors from three replicates

Fig. 7

A geometric model for nucleosome dynamics. All details on the parameterization of the intact nucleosome and definitions of relevant parameters and the computer code are given in Supplementary Note 3. The (H3-H4)₂ tetramer is shown in turquoise, while H2A-H2B heterodimers are shown in orange (α-side) and dark yellow (β-side). The interdye distances 〈R_DA^MF*〉_E are computed for the following three starting structures (opening angle θ = 0°): the intact nucleosome and both possible hexasomes with an open α-side and β-side, respectively. We compute the geometry of potentially dynamic nucleosome species as a function of the opening angle θ (between the H2A-H2B dimer and the (H3-H4)₂ tetramer) for five scenarios dimer:tetramer opening: (i, ii) intact nucleosomes (octasomes), where opening occurs either only on the α-side (orange line) or only on the β-side (wine dashed line), (iii) intact nucleosomes (octasomes) with simultaneous opening on both sides with equal angles (cyan line) and (iv, v) hexasomes that are missing the dimer at the α-side (black line) or at the β-side (gray line) with additional opening on the other so far closed side. The blue and red ribbons represent the ranges of experimentally determined average interdye distances for MF* (〈R_DA^MF*〉_E = (57 ± 1) Å) and HF (〈R_DA^HF〉_E = (46 ± 1) Å), respectively. Our model predicts an simultaneous and equal opening by ≈20° within a hexasome or an octasome, while rather unrealistic angles greater than 70° would be required for nucleosome species with only one open dimer:tetramer interface

Fig. 8

Full kinetic scheme for nucleosome disassembly. At 150 mM NaCl concentration nucleosomes disassembly proceeds through the sequential loss of H2A-H2B heterodimers as presented here with the corresponding relaxation times. The two heterodimers are shown in orange (α-side) and dark yellow (β-side), while the tetramer is shown in turquoise. We resolve five steps in our smFRET experiments: (I) Disassembly of the intact nucleosome ${\mathrm{O}}_{\mathrm{st}}^{\mathrm{MF}}$ is predominantly initiated from the weaker binding α-side, where opening of the first dimer:tetramer interface precedes dimer eviction ${\mathrm{O}}_{\mathrm{cl}}^{\mathrm{MF}*}$. (II–IV) The second dimer can reversibly detach from the tetramer once the first dimer:tetramer interface is broken ${\mathrm{O}}_{\mathrm{op}}^{\mathrm{HF}}$. The rate constant for the second dimer opening, however, is smaller when the first dimer is still present, probably due to stabilizing interactions between the dimers. Once the first dimer is gone ${\mathrm{H}}_{\mathrm{cl}}^{\mathrm{MF}*}$, reversible detachment ${\mathrm{H}}_{\mathrm{op}}^{\mathrm{HF}}$ and subsequent loss of the remaining dimer proceed at a faster rate. (V) The resulting tetrasome can adopt variable DNA geometries with most conformations leading to very low or no FRET (T^LF/NF). Note that at higher salt concentration the DNA can fully dissociate from the remaining tetramer (DNA^NF). All corresponding kinetic and thermodynamic parameters are compiled in Table 1