Unique fluorophores in the dimeric archaeal histones hMfB and hPyA1 reveal the impact of nonnative structure in a monomeric kinetic intermediate

Abstract

Homodimeric archaeal histones and heterodimeric eukaryotic histones share a conserved structure but fold through different kinetic mechanisms, with a correlation between faster folding/association rates and the population of kinetic intermediates. Wild-type hMfB (from Methanothermus fervidus) has no intrinsic fluorophores; Met35, which is Tyr in hyperthermophilic archaeal histones such as hPyA1 (from Pyrococcus strain GB-3A), was mutated to Tyr and Trp. Two Tyr-to-Trp mutants of hPyA1 were also characterized. All fluorophores were introduced into the long, central alpha-helix of the histone fold. Far-UV circular dichroism (CD) indicated that the fluorophores did not significantly alter the helical content of the histones. The equilibrium unfolding transitions of the histone variants were two-state, reversible processes, with DeltaG degrees (H2O) values within 1 kcal/mol of the wild-type dimers. The hPyA1 Trp variants fold by two-state kinetic mechanisms like wild-type hPyA1, but with increased folding and unfolding rates, suggesting that the mutated residues (Tyr-32 and Tyr-36) contribute to transition state structure. Like wild-type hMfB, M35Y and M35W hMfB fold by a three-state mechanism, with a stopped-flow CD burst-phase monomeric intermediate. The M35 mutants populate monomeric intermediates with increased secondary structure and stability but exhibit decreased folding rates; this suggests that nonnative interactions occur from burial of the hydrophobic Tyr and Trp residues in this kinetic intermediate. These results implicate the long central helix as a key component of the structure in the kinetic monomeric intermediates of hMfB as well as the dimerization transition state in the folding of hPyA1.

Figure 1.

Ribbon diagram of the archaeal histone homodimer hMfB. The two monomers are shaded light and dark. The side chain of Met35 is shown in CPK representation; the structurally equivalent position is Tyr36 in hPyA1. The Cα of the residue equivalent to Tyr32 in hPyA1 is indicated by a single sphere. The structure was determined by NMR (1BFM.pdb) (Starich et al. 1996), and the figure rendered with PyMOL (DeLano Scientific).

Figure 2.

Spectroscopic properties of the hMfB and hPyA1 fluorescent variants. (A) Far-UV CD of hMfB dimers in 0 M GdmCl: wild-type (⋄, dotted line), M35Y (○), and M35W (□). (B) Trp FL spectra in 0 M (spectra with symbols) and 4 M GdmCl (spectra without symbols): hMfB M35W (black solid lines, ■); hPyA1 Y32W (dark gray lines, ▲); hPyA1 Y36W (light gray lines, ▼). Conditions: 5 μM monomer for CD and 2 μM monomer for FL; 200 mM KCl, 20 mM KPi, pH 7.2, 25°C.

Figure 3.

Representative F_app plots for the fluorescent histone variants. (Open symbols) FL data; (closed symbols) far-UV CD. Wild-type data is from Topping and Gloss (2004). (Solid lines) Global fits of multiple titrations to a two-state dimer unfolding mechanism. (A) hMfB at 5 μM monomer: wild-type data (dotted line), M35Y (circles), M35W (squares). (B) hPyA1 at 3 μM monomer: wild type (dotted line), Y32W (circles), Y36W (diamonds). The range of GdmCl concentrations in the titrations generally spanned 0–5 M, but the transition region has been expanded for clarity. Conditions: 200 mM KCl, 20 mM KPi, pH 7.2, 25°C.

Figure 4.

GdmCl dependence of the folding and unfolding rates. For all panels: (cyan ●) hMfB M35Y, (blue ■) hMfB M35W, (red ▲) hPyA1 Y32W, (green ▼) Y36W; (solid lines) global fits for the mutants, (dashed lines) global fit for wild-type hMfB, (dotted lines) global fit for wild-type hPyA1 (Topping and Gloss 2004). (A, B) Data points represent the results of semi-local fits of multiple FL and CD traces at varied monomer concentrations at a given [GdmCl]. (A) Folding kinetics for hMfB variants. (B) Folding kinetics for hPyA1 variants. (C) Unfolding kinetics. Data points represent the results of semi-local fits of multiple CD and FL traces at 2 μM for Y32W and 5 μM monomer for all other proteins. As described in the text, there is very little protein concentration dependence for the unfolding rates. Conditions: 200 mM KCl, 20 mM KPi, pH 7.2, 25°C.

Figure 5.

Burst-phase analysis for wild-type and hMfB mutants. The observed SF–CD amplitude is shown as a percent of the amplitude expected based on the native and unfolded baselines determined from equilibrium experiments. Lines are drawn to guide the eye and do not represent fits of the data. (Diamonds, dashed line) Wild type, (squares, solid line) M35W, (circles, dotted line) M35Y. (Inset) Representative M35W SF–CD trace at 0.7 M GdmCl; (arrows) expected ellipticities for the folded and unfolded species; (solid line) semi-local fit of the data as described in the text. Conditions: 5 μM monomer, 200 mM KCl, 20 mM KPi, pH 7.2, 25°C.

Figure 6.

Reaction coordinate diagram for the folding of the wild type and M35W hMfB to illustrate the changes in the ΔΔG equilibrium and kinetic values. The free energies of the unfolded states were set at zero for both histones. The free energy for the dimerization transition state was estimated from the Kramers formalism, with a second-order diffusion-limited pre-exponential factor of 3 × 10⁹ M⁻¹s⁻¹ (Gloss and Matthews 1998). Because the folding of the monomeric intermediate occurs in the stopped-flow dead time, its rate of formation and thus its ΔG^‡ _fold could not be estimated; thus, it was arbitrarily set to half that of the second-order reaction. The reaction profiles are: wild type (dotted line), M35W (solid line). The changes in ΔG for the BP monomer stability and the dimerization transition state are denoted by the δ values. (Arrows) Changes associated with the ΔΔG^‡ values.