Microsecond time-resolved X-ray scattering by utilizing MHz repetition rate at second-generation XFELs

Sebastian Westenhoff, Patrick Konold, Leonardo Monrroy, Alfredo Bellisario, Diogo de Matos Filipe, Patrick W. Adams, Roberto Álvarez, Richard Bean, Johan Bielecki, Szabolcs Bódizs, Gabriel Ducroq, Helmut Grubmueller, Richard A. Kirian, Marco Kloos, Jayanat Koliyadu, F. Koua, Taru Larkiala, Romain Letrun, Fredrik Lindsten, Michael Maihöfer, Andrew V. Martin, Petra Mésźaros, Jennifer Mutisya, Amke Nimmrich, Kenta Okamoto, Adam Round, Tokushi Sato, Joana Valério, Daniel Westphal, August Wolter, Tej Varma Yenupuri, Tong You, Filipe R. N. C. MaiaResearch Square 2023 · DOI

TL;DR

Researchers have unlocked a new way to film proteins in motion by leveraging the ultra-fast, megahertz (MHz) pulse rates of second-generation X-ray Free Electron Lasers (XFELs). By achieving an unprecedented noise floor of 0.001%, the team successfully captured the microsecond-scale unfolding of a light-sensing protein, revealing a "hidden" intermediate state that had previously eluded experimental detection.

What Problem Does This Paper Address?

Understanding how proteins change shape is fundamental to biology and medicine, but capturing these "structural perturbations" in real-time is notoriously difficult. Traditional "pump-probe" methods at X-ray facilities are often slow and sample-hungry: you trigger a reaction (the pump), hit it with one X-ray pulse (the probe) at a specific delay, and repeat the process thousands of times to build a movie.

The challenge is that many biological processes happen on the microsecond timescale—too fast for standard synchrotrons but often overlooked by femtosecond-focused XFEL studies. Furthermore, the "noise" in these experiments often drowns out the subtle signals of moving atoms. This paper addresses the need for a high-efficiency, low-noise method to observe sub-millisecond structural dynamics in near-native (liquid) environments.

Methods

The authors utilized the European XFEL (EuXFEL), which is unique because it delivers X-ray pulses in rapid-fire "trains" at MHz frequencies.

  • TR-WAXS: They used Time-Resolved Wide-Angle X-ray Scattering, a technique that looks at how X-rays bounce off proteins in solution to deduce their shape.
  • The "Single-Trigger" Strategy: Instead of one probe pulse per trigger, they fired a single laser pulse to start the protein reaction and then used a series of 175 X-ray pulses (spaced just 1.77 microseconds apart) to record a continuous time-series.
  • Massive Averaging: By repeating this "train" sequence and averaging the data, they suppressed experimental noise to a staggering 0.001%—at least an order of magnitude better than previous attempts.
  • Structural Modeling: To interpret the data, they used AlphaFold with "dropout enabled inference" to generate 20,000 possible structural variations of the protein, then used a search algorithm to find which models best fit the experimental X-ray scattering patterns.

Datasets

  • Experimental Data: The raw X-ray scattering data for the AsLOV2 protein is hosted in the EuXFEL repository (DOI: 10.22003/XFEL.EU-DATA-003046-00).
  • Structural Models: A library of 20,000 AlphaFold-predicted structures was used for the fitting process.
  • Benchmark: The study used the well-characterized AsLOV2 (a light-oxygen-voltage domain from oats) as a benchmark system to prove the method's efficacy.

Key Insights and Ideas

  • The "Hidden" Intermediate (State B): The most significant discovery was the identification of a previously unrecognized intermediate state. While researchers knew the protein formed a chemical bond (State A) and eventually unfolded its helix (State C), the data revealed a distinct "State B" that occurs after the bond forms but before the major unfolding. This state had been predicted by computer simulations but never clearly seen in the lab.
  • Biphasic Unfolding: The results prove that the unfolding of the J$\alpha$ helix is not a simple "on/off" switch but a two-step (biphasic) process occurring within 300 microseconds.
  • Noise is the Real Enemy: The paper demonstrates that the primary advantage of MHz XFELs isn't just speed—it's the ability to average out systematic detector noise and jet fluctuations to see signals that were previously invisible.

Key Results

  • Temporal Resolution: Achieved a time resolution of 1.77 $\mu$s, determined by the 564 kHz repetition rate of the XFEL.
  • Spatial Resolution: Captured oscillations in the scattering data out to 4.2 Å, allowing for high-resolution structural insights.
  • Precision: The noise floor of 0.001% allowed the team to resolve State A and State B, which involve very small structural changes compared to the final unfolded state.
  • Structural Confirmation: The best-fitting models showed that not only does the J$\alpha$ helix unfold, but the N-terminal A'$\alpha$ helix also unfolds in most structures, providing a more complete picture of the protein's signaling mechanism.

Limitations and Future Work

  • Jet Stability: About 30% of the collected data had to be discarded because fluctuations in the liquid sample jet distorted the scattering patterns. Improving the stability of these "micro-jets" is a priority for future experiments.
  • Triggering Mechanisms: This study used a laser to trigger a light-sensitive protein. To make this method "widely applicable," future work will need to integrate rapid-mixing technology (microfluidics) to trigger non-photosensitive reactions, such as enzyme catalysis or RNA folding.
  • Data Complexity: The massive amount of data generated by MHz XFELs requires significant computational power for processing and modeling, suggesting a growing role for AI and automated pipelines in structural biology.

Summary updated

Abstract

<title>Abstract</title> Detecting microsecond structural perturbations in biomolecules has wide relevance in biology, chemistry, and medicine. Here, we show how MHz repetition rates at X-ray free electron lasers (XFELs) can be used to produce microsecond time-series of protein scattering with exceptionally low noise levels of 0.001%. We demonstrate the approach by deriving new mechanistic insight into Jα helix unfolding of a Light-Oxygen-Voltage (LOV) photosensory domain. This time-resolved acquisition strategy is easy to implement and widely applicable for direct observation of structural dynamics of many biochemical processes.
BibTeX
@inproceedings{westenhoff2023microsecond,
  author = {Sebastian Westenhoff and Patrick Konold and Leonardo Monrroy and Alfredo Bellisario and Diogo de Matos Filipe and Patrick W. Adams and Roberto Álvarez and Richard Bean and Johan Bielecki and Szabolcs Bódizs and Gabriel Ducroq and Helmut Grubmueller and Richard A. Kirian and Marco Kloos and Jayanat Koliyadu and F. Koua and Taru Larkiala and Romain Letrun and Fredrik Lindsten and Michael Maihöfer and Andrew V. Martin and Petra Mésźaros and Jennifer Mutisya and Amke Nimmrich and Kenta Okamoto and Adam Round and Tokushi Sato and Joana Valério and Daniel Westphal and August Wolter and Tej Varma Yenupuri and Tong You and Filipe R. N. C. Maia},
  title = {{Microsecond time-resolved X-ray scattering by utilizing MHz repetition rate at second-generation XFELs}},
  booktitle = {Research Square},
  year = {2023},
  doi = {10.21203/rs.3.rs-3638353/v1},
}

Loading related content...

Figure