Transient Protein Softening during the Working Cycle of a Molecular Machine

Proper functioning of proteins usually requires a certain internal flexibility provided by stochastic structural fluctuations on the picosecond time scale. In contrast with conventional steady-state experiments, we report on a novel type of (laser-neutron) pump-probe experiment combining in situ activation of protein function with a time-dependent test of protein dynamics using quasielastic neutron scattering. A “transient protein softening” is shown to occur during the photocycle of bacteriorhodopsin as a direct proof for the functional significance of protein flexibility.

Figure 1

Absorbance changes of BR at 412 nm at room temperature as a function of time after laser excitation at 532 nm and $t=0$. The actinic laser pulse energies are—from top to bottom—70, 36, 26, 18, 10, 8, 3.3, and $1.4\mathrm{mJ}/{\mathrm{cm}}^2$. Data were obtained directly at PM samples used for light-excited QENS experiments. Diamonds and squares indicate the arrival times of the neutron pulses at the sample relative to the laser pulse in two different measurement modes, i.e., experiments without (B) and with time selection (C), respectively. Here, the letters B and C correspond to the subscripts of the fit functions to the respective QENS spectra shown in Fig. 2. The inset shows the superposition of the structures of a BR monomer in the ground state ${\mathrm{BR}}_{568}$ (purple) and in the $M$ intermediate (yellow), respectively, according to Sass et al. [16].

Figure 2

QENS spectra of wild-type PM hydrated at 98% r.h. ${\mathrm{D}}_2\mathrm{O}$ measured with NEAT using an incident neutron wavelength of 5.1 Å and an elastic energy resolution of $93\mu \mathrm{eV}$ at room temperature (295 K). The energy transfer axis is composed of a linear (left) and a logarithmic part (right side). Frame I shows a comparison of the PM QENS spectra obtained in the dark (black squares) and selectively during the presence of the $M$ intermediate (red squares), the full lines are fits of the data. Labels A, B, and C correspond to the fit functions $S_A(Q,\omega )$, $S_B(Q,\omega )$, and $S_C(Q,\omega )$, respectively. In Frame II the fit function $S_{\mathrm{total}}(Q,\omega )$ of the light-induced PM QENS spectrum obtained during the $M$ intermediate is composed of two components: (a) a fraction of 80% of the BR molecules remaining in the dark [0.8 $S_A(Q,\omega )$] and (b) an experimentally determined fraction of 20% of the BR molecules in the $M$ intermediate after absorption of a light quantum [0.2 $S_M(Q,\omega )$]. The line shape of the vibrational contribution to the QENS spectra is labeled with “DHO.”