The time revolution in macromolecular crystallography

Abstract

Macromolecular crystallography has historically provided the atomic structures of proteins fundamental to cellular functions. However, the advent of cryo-electron microscopy for structure determination of large and increasingly smaller and flexible proteins signaled a paradigm shift in structural biology. The extensive structural and sequence data from crystallography and advanced sequencing techniques have been pivotal for training computational models for accurate structure prediction, unveiling the general fold of most proteins. Here, we present a perspective on the rise of time-resolved crystallography as the new frontier of macromolecular structure determination. We trace the evolution from the pioneering time-resolved crystallography methods to modern serial crystallography, highlighting the synergy between rapid detection technologies and state-of-the-art x-ray sources. These innovations are redefining our exploration of protein dynamics, with high-resolution crystallography uniquely positioned to elucidate rapid dynamic processes at ambient temperatures, thus deepening our understanding of protein functionality. We propose that the integration of dynamic structural data with machine learning advancements will unlock predictive capabilities for protein kinetics, revolutionizing dynamics like macromolecular crystallography revolutionized structural biology.

FIG. 1.

Evolution and milestones in time-resolved crystallography. Graph shows Google Scholar results per year using the term “time-resolved crystallography.” Key historical developments are annotated on the timeline, including from left to right: the publication of the first Laue diffraction pattern from protein crystals (1984) [picture from Fig. 2(b) reprinted with permission from Moffat et al., Science 223(4643), 1423–1425 (1984). Copyright 1984 AAAS], the inauguration of the Swiss Light Source (SLS) (2001) (Picture: SLS), and the Linac Coherent Light Source (LCLS) tunnel (2009) upon first lasing (picture: Brad Plummer/SLAC National Accelerator Laboratory). The backdrop displays all time-resolved structures archived in the Protein Data Bank (PDB). The inception of time-resolved crystallography coincides with the advent of Laue crystallography's rapid data acquisition capabilities. Its initial growth phase was bolstered by the launch of third-generation synchrotrons between 1994 (ESRF) and 2016 (SESAME). Time-resolved crystallography grew further with the emergence of x-ray free-electron lasers (XFELs) between 2009 and 2018, further bolstered by serial crystallography techniques. The field is expected to gain additional momentum over the next decade, propelled by advancements in high-speed detectors, fourth-generation synchrotron sources, and novel computational methods, transcending its niche origins to broaden its impact and research output.

FIG. 2.

Growth trends in structural biology. (a) The total count of x-ray crystallography structures released annually by the Protein Data Bank (PDB) reached its peak in 2020 (red line). Starting in 1999, the release of cryo-structures (blue line) surpassed the number of “dynamic” structures here defined as temperatures above the glass transition (≥180 K) (green line). (b) The total number of PDB entries released annually continues to increase (red line), indicating an ongoing productivity increase in structural biology. (c) This upward trajectory is driven by cryo-EM (green line), which eclipsed the number of unique (molecules with the exact same sequence are counted just once) x-ray crystallography entries released by the PDB (red line) in 2022. (d) Anticipating this shift, the percentage of publications featuring cryo-EM structures (green line) outpaced those based on x-ray crystallography (blue line) in three high-impact journals (total number of publications shown as red line).

FIG. 3.

Rise in single-shot serial crystallography and high-resolution structural studies. (a) There is a noticeable uptick in “dynamic” structures, determined at temperatures over 180 K since 2018 (red line). Single-shot serial crystallography (ssSX) is becoming more prominent, making up 30% of these dynamic entries in the PDB by 2023 (blue line). (b) At the same time, high-resolution structures are claiming a larger share in the PDB. The red line shows the number of structures with resolutions higher than 1.5 Å, and the blue line indicates their percentage of total submissions. This trend highlights the crucial role of crystallography as a high-resolution method for in-depth structural insights.

FIG. 4.

Variants of time-resolved single-shot serial crystallography (ssSX) via a high-viscosity jet. (a) The pump-probe method activates crystals (depicted as purple blocks) with a laser pulse (gray circle), with the jet (yellow) extrusion allowing collection from non-illuminated crystals (white blocks). Data acquisition follows with an x-ray pulse for a single diffraction pattern at a set time delay, then a dark pattern once the jet moves the activated zone past the laser's effective diameter (at least 1.5 times the 1/e² size of the pump laser). This mode is suitable when the probe pulse dose exceeds the damage threshold, a scenario well-characterized at XFELs but requiring rigorous assessment at fourth-generation synchrotrons, especially when utilizing multi-layer monochromators that deliver high doses in short durations. (b) The pump-scan technique continuously probes the illuminated region, compiling diffraction patterns over a duration until the pumped area is displaced by jet extrusion, thus capturing an entire time-series rather than a fixed delay. This approach is feasible when the dose per frame remains within the radiation damage threshold of about 100 kGy per frame.

FIG. 5.

Overview of triggering mechanisms and corresponding biological processes in time-resolved crystallography. (a) Classification of triggering mechanisms by the nature of the initiating stimuli, with the corresponding time scales increasing from left to right. (b) The time-scales of biologically relevant processes and their alignment with the triggering mechanisms depicted in (a), illustrating the timescale coverage for capturing dynamic biological events.