Ultrafast folding of alpha3D: a de novo designed three-helix bundle protein

Abstract

Here, we describe the folding/unfolding kinetics of alpha3D, a small designed three-helix bundle. Both IR temperature jump and ultrafast fluorescence mixing methods reveal a single-exponential process consistent with a minimal folding time of 3.2 +/- 1.2 micros (at approximately 50 degrees C), indicating that a protein can fold on the 1- to 5-micros time scale. Furthermore, the single-exponential nature of the relaxation indicates that the prefactor for transition state (TS)-folding models is probably >or=1 (micros)-1 for a protein of this size and topology. Molecular dynamics simulations and IR spectroscopy provide a molecular rationale for the rapid, single-exponential folding of this protein. alpha3D shows a significant bias toward local helical structure in the thermally denatured state. The molecular dynamics-simulated TS ensemble is highly heterogeneous and dynamic, allowing access to the TS via multiple pathways.

Fig. 1.

(a) IR spectra of α₃DinD₂O (50 mM phosphate buffer, pH* 2.2 uncorrected pH reading in D₂O) were collected every 7°C, from 2.5 to 86.5°C. Representative IR spectra in the amide I′ region (○) are shown. These spectra (total 13) were modeled (solid lines) globally by 6 G with the following parameters: ν₁ = 1,587.5 cm^–1, Δν₁ = 26.8 cm^–1; ν₂ = 1,606.6 cm^–1, Δν₂ = 18.6 cm^–1; ν₃ = 1,628.6 cm^–1, Δν₃ = 25.7 cm^–1; ν₄ 1644.2 cm^–1, Δν₄ 23.4 cm^–1; ν₅ = 1660.1 cm^–1, Δν₅ = 32.8 cm^–1; and ν₆ = 1,705.7 cm^–1, Δν₆ = 49.0 cm^–1; where ν is band position and Δν is full width at half maximum. The bands that make up the fit for the 9.5°C spectrum are shown. The bands at 1,705 and 1,587 cm^–1 are due to protonated and deprotonated carboxylates, respectively, and the band at 1,606 cm^–1 arises from amino acid side chains. The remaining three bands are due to amide COs. (b) Difference IR spectra generated by subtracting the spectrum collected at 2.5°C from the spectra in a. Arrows indicate the direction of changes when temperature is increased. (c) Relative band areas of the three amide bands as a function of temperature.

Fig. 2.

T-jump-induced relaxation kinetics of α₃D measured by time-resolved IR spectroscopy at 1,631 and 1,665 cm^–1 as well as different final temperatures. The fast phase rose instantaneously and was not resolvable with our instrument setup, whereas the slow phase was modeled by a single-exponential function. The relaxation time constants corresponding to different final temperatures are τ (45.4°C) = 2.8 μs, τ (70.0°C) = 2.6 μs, τ (81.0°C) = 1.2 μs, and τ (72.3°C) = 2.2 μs. The T-jump amplitude was ≈10°C for each case, and the final temperature is indicated.

Fig. 3.

Unfolding kinetics of α₃D measured by continuous-flow fluorescence. (a) Representative continuous-flow fluorescence traces, including unfolding experiments at 12, 31, and 38°C and a final urea concentration of 6.5 M. Solid lines represent single-exponential fits. (b) Semilogarithmic plot of the rate constant of unfolding vs. urea concentration; representative data at three temperatures are shown. The rates in the absence of urea, obtained by linear regression (lines), are included in Fig. 4b.

Fig. 4.

(a) Arrhenius plots of the observed (•), folding (○), and unfolding (▵) rate constants. Lines are fits to the Eyring equation, i.e., ln(k) = ln(D) – ΔG^≠/RT, where D is a constant and ΔG^≠ is the free energy of activation that is a function of temperature, as described by the following function: ΔG^≠ = ΔH^≠(T_m) + ΔC_p^≠ (T – T_m) – T [ΔS^≠(T_m) +ΔC_p^≠ ln(T/T_m)], where T_m is 73.2°C. Global fitting of the folding and unfolding kinetics yields the following thermodynamic parameters of activation when D = 10⁶ s^–1 is used: (i) for folding, ΔH^≠(T_m) = –6.1 kcal·mol^–1, ΔS^≠(T_m) = –20.5 cal·mol^–1·K^–1, ΔC_p^≠ = –255 cal·mol^–1·K^–1; (ii) for unfolding, ΔH^≠(T_m) = 29.8 kcal·mol^–1, ΔS^≠(T_m) = 83.1 cal·mol^–1·K^–1, ΔC_p^≠ = 392 cal·mol^–1·K^–1. (b) Same as a but with an extended x axis. For comparison, unfolding rate constants (□) measured in mixing experiments are also shown. These data show that both folding and unfolding rates exhibit nonlinear temperature dependence. The folding time at 25°C is predicted to be ≈4.8 μs according to the fit in a, and the unfolding time at 100°C is predicted to be ≈100 ns. The times required to reach the TS in two independent simulations at 100°C are shown as ▪.

Fig. 5.

Representative structures from the MD simulations. (a) Loss of helical and tertiary structure as a function of time for two of the simulations. The regions of the native helix are red (HI), green (HII), and blue (HIII). Thickened ribbons denote helical structure, as determined by using the dssp algorithm (58). (b) A representative structure from each TS ensemble, identified by using a conformational clustering procedure (32), is displayed along with a side chain contact pay with native contacts shown above the diagonal and nonnative below. The map shows the average population of the contacts over a 5-ps TS ensemble, moving from blue to green to red for the most to least populated contacts.