X-ray free-electron lasers (XFELs) have revolutionized ultrafast atomic-level studies. In structural biology, serial femtosecond crystallography (SFX) has advanced the field, but the requirement for crystals is a limitation. Cryo-electron microscopy (cryo-EM) offers high-resolution single-molecule imaging, but is limited to millisecond timescales. Femtosecond X-ray diffractive imaging (FXI) offers the potential to overcome these limitations, providing sub-picosecond time resolution and enabling the observation of a wider conformational landscape due to higher sample temperatures. The chaperonin GroEL, a large, well-studied protein complex, was chosen as an ideal candidate for this study due to its size, availability, and distinctive shape. Despite progress in FXI, no single-particle studies had previously been achieved. This work aims to demonstrate the principle of diffraction before destruction for single proteins, paving the way for ultrafast single-protein studies.

The introduction extensively reviews existing techniques such as SFX and cryo-EM, highlighting their strengths and limitations in studying ultrafast phenomena and single-molecule structures. It emphasizes the need for a technique that overcomes the limitations of both, such as the need for crystals in SFX and the millisecond timescale resolution of cryo-EM. The review positions FXI as a promising technique to address these limitations, and justifies the selection of GroEL as a suitable model protein due to its size, availability, and previous extensive study using other methods.

The experiment was conducted at the European XFEL (EuXFEL) facility using the Small Quantum Systems (SQS) instrument. GroEL particles were aerosolized using electrospray ionization, neutralized, and focused into a thin stream using an aerodynamic lens. The stream was intersected with femtosecond soft X-ray pulses (1200 eV, 6.5 mJ average pulse energy). Diffraction data were collected using a pnCCD detector. The experimental setup included measures to minimize background noise, such as apertures before and after the interaction region. Two types of background noise were identified: fluorescence from the injection gas and elastic scattering of X-rays by beamline elements. Data analysis involved identifying 'hits' (pulses intersecting a particle) from the background by considering the photon counts in each diffraction pattern. A hit detection threshold of 19,000 photons was employed to separate the signal from background noise. The size of the particles was estimated from the diffraction patterns, showing a peak at 15nm consistent with the expected size of GroEL. To verify that the diffraction patterns indeed originated from GroEL, a template matching scheme was employed, comparing the experimental diffraction pattern with 4.3 million simulated diffraction patterns considering various orientations and positions of the GroEL molecule in the beam. This comprehensive approach handled the uncertainties in orientation, center position and the presence of background noise in the experimental data.

The study successfully obtained and analyzed an X-ray diffraction pattern from a single GroEL protein molecule. The size of the particles exhibiting the diffraction signal was determined to be approximately 15 nm, matching the expected size of GroEL. The analysis showed that the diffraction pattern deviated from spherical symmetry, consistent with GroEL's barrel-shaped structure. A comparison of the experimental diffraction pattern with simulated patterns confirmed that the obtained data originated from a GroEL particle. The number of photons in the successful diffraction pattern was significantly higher than the average background (19,000 vs. ~17,600), indicating a clear signal from a single protein. The success rate of obtaining a diffraction pattern from a single protein was low (around 1% of the collected data), which was mainly attributed to the limitations of the detection system. Despite this, the work successfully demonstrated the feasibility of detecting a diffraction pattern from a single protein, representing the smallest biological sample ever imaged using X-rays, and extending the ‘diffraction before destruction’ principle to single biomolecules.

The successful observation of a diffraction pattern from a single GroEL protein validates the feasibility of FXI for single-molecule studies. The results demonstrate that it is possible to obtain structural information from single proteins using ultrafast X-ray pulses. While the success rate was low, this is mainly due to the detector limitations, suggesting that improvements in detector technology and data acquisition methods could significantly improve the efficiency of this technique. This work establishes FXI as a promising tool for time-resolved studies of single protein molecules with femtosecond time resolution, opening the path to observe dynamic processes in individual biomolecules at unprecedented detail.

This study represents a landmark achievement in single-molecule imaging, successfully capturing the X-ray diffraction pattern of a single GroEL protein for the first time. The results validate the theoretical concept of diffraction before destruction at the single-protein level. While challenges remain regarding data acquisition efficiency, improvements in detector technology are expected to overcome these limitations, opening up exciting possibilities for time-resolved structural studies of biological macromolecules at the femtosecond timescale.

The main limitation of the study was the low success rate in obtaining diffraction patterns from single GroEL proteins. This was primarily attributed to the limited capabilities of the CCD-based detector, which could not operate at the MHz repetition rate of the XFEL pulses, resulting in a significant reduction in data acquisition efficiency. Further development of high-speed detectors is needed to increase the efficiency of data acquisition. The analysis also relied on the comparison of the single experimental diffraction pattern with simulated patterns. More data is needed for a more statistically robust and complete structural characterization.