Questions: Time-Resolved Structural Methods

Question 1

What is the 'diffraction-before-destruction' principle at an XFEL, and why does it enable room-temperature crystallography?

• A. The crystal is protected from radiation by a special coating
• B. XFEL pulses are so intense (10^12 photons in ~10-50 femtoseconds) that the crystal diffracts before the radiation damage cascade destroys it. The pulse is shorter than the time needed for atomic motion in response to ionization (~100 femtoseconds), so the diffraction pattern captures the undamaged structure. This eliminates the need for cryogenic cooling (which standard crystallography uses to slow radiation damage), enabling data collection at room temperature where proteins occupy their physiologically relevant conformational ensemble
• C. The XFEL uses low-energy X-rays that do not damage the crystal
• D. Each crystal is imaged many times before it is destroyed

Answer: (B)

In conventional crystallography at synchrotrons, X-ray exposure progressively damages the crystal (breaking disulfide bonds, decarboxylating glutamates, reducing metal centers). Cryogenic cooling (100 K) slows this damage, allowing data collection but trapping the protein in a cryo-artifact conformational state. XFEL pulses are ~10^9 times brighter than synchrotron beams but last only 10-50 femtoseconds. The crystal is completely vaporized by each pulse, but the diffraction pattern is recorded before the atoms have time to move — the Coulomb explosion happens after the diffraction is complete. Each crystal gives one diffraction pattern (one orientation), so data from thousands of randomly oriented microcrystals are merged to reconstruct the full dataset (serial crystallography).

Question 2

Time-resolved crystallography at an XFEL can capture any biological process at atomic resolution, regardless of the timescale.

• T. True
• F. False

Answer: False

XFEL time-resolved crystallography works best for processes that can be synchronously triggered across the crystal. Light-activated processes (photoreceptors, photosynthetic reaction centers, light-driven ion pumps) are ideal because a laser pulse simultaneously triggers all molecules in the crystal within femtoseconds. For processes triggered by substrate binding (enzymatic catalysis), the limitation is diffusion time — substrate must diffuse into the crystal, which takes milliseconds to seconds depending on crystal size and substrate concentration. This sets a lower limit on the time resolution for diffusion-initiated experiments. Additionally, the process must be fast enough relative to the crystal lattice tolerance — large conformational changes may crack the crystal or disrupt the lattice. Mix-and-inject serial crystallography (MISC) with microcrystals (reducing diffusion path) pushes the time limit to ~milliseconds, while photocaged substrates (released by a light pulse) can synchronize non-light-driven processes on faster timescales.

Question 3

How does time-resolved cryo-EM differ from time-resolved crystallography in its approach to capturing structural intermediates?

Model answer: Time-resolved cryo-EM captures intermediates by rapidly mixing a macromolecule with its substrate or trigger, allowing the reaction to proceed for a defined time, and then rapidly freezing the sample by plunge-freezing into liquid ethane to trap the intermediate state in vitreous ice. The time resolution is limited by the mixing time (~milliseconds using microfluidic rapid-mixing devices) and the freezing time (~1 millisecond for vitrification). Unlike crystallography, cryo-EM works with molecules in solution (no crystal lattice constraints), so large conformational changes that would destroy a crystal are readily captured. Additionally, cryo-EM's ability to computationally classify particles into different conformational states means that a heterogeneous mixture of intermediates can be sorted after the fact — each particle is assigned to a conformational class, and 3D reconstructions of each class reveal the structural ensemble present at that time point.

Time-resolved cryo-EM has been applied to ribosome translocation (capturing pre- and post-translocation states after GTP hydrolysis), to GroEL-GroES chaperonin cycling (trapping ATP-driven conformational changes), and to spliceosomes (resolving multiple catalytic intermediates). The combination of millisecond time resolution and the ability to handle conformational heterogeneity makes it complementary to XFEL approaches, which offer femtosecond-to-millisecond resolution but require crystal-compatible motions.

Question 4

What structural insights about enzyme catalysis have been uniquely revealed by time-resolved serial crystallography that could not be obtained from static structures?

Model answer: Time-resolved serial crystallography has revealed transient catalytic intermediates — structural states that exist for femtoseconds to milliseconds during the catalytic cycle and cannot be trapped by conventional methods. Key examples include: (1) in bacteriorhodopsin (a light-driven proton pump), XFEL time series captured the complete photocycle at near-atomic resolution, showing the retinal isomerization, water wire rearrangement, and protein conformational changes that move protons across the membrane — individual intermediates lasting picoseconds were resolved; (2) in fluorescent proteins, the chromophore excited state and early photochemistry were captured before thermal relaxation; (3) in cytochrome c oxidase, intermediates in the oxygen reduction mechanism were trapped. These experiments revealed that enzyme active sites undergo coordinated sub-angstrom motions during catalysis — correlated movements of residues, water molecules, and substrates that facilitate bond breaking and formation — details invisible in static endpoint structures.

The bacteriorhodopsin work by Nango et al. (2016) and Nogly et al. (2018) is the landmark demonstration: 13 time points from femtoseconds to milliseconds produced an atomic-resolution movie of proton pumping, revealing water molecule rearrangements and hydrogen bond switches that mechanistic models had predicted but never been observed. This represents the fulfillment of structural biology's longstanding goal of watching molecular machines in action.