Since more than 20 years, there has been a tremendous effort both on packing proteins in crystals and on improving synchrotrons so as to make high performance protein crystallography a routine task. This has allowed to resolve the structure of thousands of proteins, regardless of them being soluble or embedded in membranes. Since a few years, a new revolution is on the way with the advent of X-ray free electron lasers (FELs), capable of producing the most intense and shortest X-ray pulses on earth. Facilities such as the Linac Coherent Light Source (LCLS, Stanford) or the European FEL (XFEL in Hamburg) offer users the possibility to conduct time-resolved X-ray experiments with variable energies (diffraction, scattering, XANES), in a laser-pump/X-ray-probe operation, with down to ≈30 fs time resolution.
Such pump-probe experiments have been applied in recent years to photo-sensitive proteins, allowing to observe photo-initiated motion of the protein-forming molecules, the light-absorbing chromophore and amino acids, with atomic resolution (1.5 Å). Rhodopsin, the photo-sensor of vision, has not been subjected to an FEL yet, but its cousin, bacterio-rhodopsin (bR) was. The latter is used by archaebacteria in a primitive form of photo-synthesis to generate ATP. bR thus stands as an example of the large family of retinal proteins, used for diverse functions by microbacteria. It is known that the chromophore retinal undergoes a trans-cis isomerisation upon absorption of light, and standard lab-based laser spectroscopy has shown in the late 1980’s[1] that this is accomplished in half a picosecond or less (5x 10-13 s) !
In the past months, two femtosecond X-ray diffraction experiments reported astounding details of the photo-initiated motion of retinal, and of the amino acids and water molecules forming the retinal binding pocket: within less than 200 fs, when retinal is in an excited state, it slightly stretches and twists, hydrogen-bonds are disrupted, and one water molecule “disappears” [1,2]. The theoretically predicted aborted bicycle-pedal mechanism for isomerisation, a special form of synchronised rotation about two C-C bond axes of the retinal backbone, was confirmed. In the second paper [2], an international consortium led by I. Schlichting (Max-Planck-Institute Heidelberg), including S. Haacke at IPCMS, reports coherent oscillations both of the retinal as well as of some key amino acids, flanking retinal. The paper highlights the electronic and vibrational coupling of retinal and its environment, which will be crucial to account for fully in the next generation of extended quantum chemistry simulations, and which is probably key to understand the catalytic effect the protein exerts on the photo-isomerisation reaction.