DriftRec: Adapting Diffusion Models to Blind JPEG Restoration.

In this work, we utilize the high-fidelity generation abilities of diffusion models to solve blind JPEG restoration at high compression levels. We propose an elegant modification of the forward stochastic differential equation of diffusion models to adapt them to this restoration task and name our method DriftRec. Comparing DriftRec against an L2 regression baseline with the same network architecture and state-of-the-art techniques for JPEG restoration, we show that our approach can escape the tendency of other methods to generate blurry images, and recovers the distribution of clean images significantly more faithfully. For this, only a dataset of clean/corrupted image pairs and no knowledge about the corruption operation is required, enabling wider applicability to other restoration tasks. In contrast to other conditional and unconditional diffusion models, we utilize the idea that the distributions of clean and corrupted images are much closer to each other than each is to the usual Gaussian prior of the reverse process in diffusion models. Our approach therefore requires only low levels of added noise and needs comparatively few sampling steps even without further optimizations. We show that DriftRec naturally generalizes to realistic and difficult scenarios such as unaligned double JPEG compression and blind restoration of JPEGs found online, without having encountered such examples during training.

Structural Elucidation and Antiviral Activity of Covalent Cathepsin L Inhibitors.

Emerging RNA viruses, including SARS-CoV-2, continue to be a major threat. Cell entry of SARS-CoV-2 particles via the endosomal pathway involves cysteine cathepsins. Due to ubiquitous expression, cathepsin L (CatL) is considered a promising drug target in the context of different viral and lysosome-related diseases. We characterized the anti-SARS-CoV-2 activity of a set of carbonyl- and succinyl epoxide-based inhibitors, which were previously identified as inhibitors of cathepsins or related cysteine proteases. Calpain inhibitor XII, MG-101, and CatL inhibitor IV possess antiviral activity in the very low nanomolar EC50 range in Vero E6 cells and inhibit CatL in the picomolar Ki range. We show a relevant off-target effect of CatL inhibition by the coronavirus main protease α-ketoamide inhibitor 13b. Crystal structures of CatL in complex with 14 compounds at resolutions better than 2 Å present a solid basis for structure-guided understanding and optimization of CatL inhibitors toward protease drug development.

Data reduction in protein serial crystallography.

Serial crystallography (SX) has become an established technique for protein structure determination, especially when dealing with small or radiation-sensitive crystals and investigating fast or irreversible protein dynamics. The advent of newly developed multi-megapixel X-ray area detectors, capable of capturing over 1000 images per second, has brought about substantial benefits. However, this advancement also entails a notable increase in the volume of collected data. Today, up to 2 PB of data per experiment could be easily obtained under efficient operating conditions. The combined costs associated with storing data from multiple experiments provide a compelling incentive to develop strategies that effectively reduce the amount of data stored on disk while maintaining the quality of scientific outcomes. Lossless data-compression methods are designed to preserve the information content of the data but often struggle to achieve a high compression ratio when applied to experimental data that contain noise. Conversely, lossy compression methods offer the potential to greatly reduce the data volume. Nonetheless, it is vital to thoroughly assess the impact of data quality and scientific outcomes when employing lossy compression, as it inherently involves discarding information. The evaluation of lossy compression effects on data requires proper data quality metrics. In our research, we assess various approaches for both lossless and lossy compression techniques applied to SX data, and equally importantly, we describe metrics suitable for evaluating SX data quality.

Observation of a single protein by ultrafast X-ray diffraction.

The idea of using ultrashort X-ray pulses to obtain images of single proteins frozen in time has fascinated and inspired many. It was one of the arguments for building X-ray free-electron lasers. According to theory, the extremely intense pulses provide sufficient signal to dispense with using crystals as an amplifier, and the ultrashort pulse duration permits capturing the diffraction data before the sample inevitably explodes. This was first demonstrated on biological samples a decade ago on the giant mimivirus. Since then, a large collaboration has been pushing the limit of the smallest sample that can be imaged. The ability to capture snapshots on the timescale of atomic vibrations, while keeping the sample at room temperature, may allow probing the entire conformational phase space of macromolecules. Here we show the first observation of an X-ray diffraction pattern from a single protein, that of Escherichia coli GroEL which at 14 nm in diameter is the smallest biological sample ever imaged by X-rays, and demonstrate that the concept of diffraction before destruction extends to single proteins. From the pattern, it is possible to determine the approximate orientation of the protein. Our experiment demonstrates the feasibility of ultrafast imaging of single proteins, opening the way to single-molecule time-resolved studies on the femtosecond timescale.

Time-resolved crystallography captures light-driven DNA repair.

Photolyase is an enzyme that uses light to catalyze DNA repair. To capture the reaction intermediates involved in the enzyme's catalytic cycle, we conducted a time-resolved crystallography experiment. We found that photolyase traps the excited state of the active cofactor, flavin adenine dinucleotide (FAD), in a highly bent geometry. This excited state performs electron transfer to damaged DNA, inducing repair. We show that the repair reaction, which involves the lysis of two covalent bonds, occurs through a single-bond intermediate. The transformation of the substrate into product crowds the active site and disrupts hydrogen bonds with the enzyme, resulting in stepwise product release, with the 3' thymine ejected first, followed by the 5' base.

Structure of the Lysinibacillus sphaericus Tpp49Aa1 pesticidal protein elucidated from natural crystals using MHz-SFX.

The Lysinibacillus sphaericus proteins Tpp49Aa1 and Cry48Aa1 can together act as a toxin toward the mosquito Culex quinquefasciatus and have potential use in biocontrol. Given that proteins with sequence homology to the individual proteins can have activity alone against other insect species, the structure of Tpp49Aa1 was solved in order to understand this protein more fully and inform the design of improved biopesticides. Tpp49Aa1 is naturally expressed as a crystalline inclusion within the host bacterium, and MHz serial femtosecond crystallography using the novel nanofocus option at an X-ray free electron laser allowed rapid and high-quality data collection to determine the structure of Tpp49Aa1 at 1.62 Å resolution. This revealed the packing of Tpp49Aa1 within these natural nanocrystals as a homodimer with a large intermolecular interface. Complementary experiments conducted at varied pH also enabled investigation of the early structural events leading up to the dissolution of natural Tpp49Aa1 crystals-a crucial step in its mechanism of action. To better understand the cooperation between the two proteins, assays were performed on a range of different mosquito cell lines using both individual proteins and mixtures of the two. Finally, bioassays demonstrated Tpp49Aa1/Cry48Aa1 susceptibility of Anopheles stephensi, Aedes albopictus, and Culex tarsalis larvae-substantially increasing the potential use of this binary toxin in mosquito control.

Calpeptin is a potent cathepsin inhibitor and drug candidate for SARS-CoV-2 infections.

Several drug screening campaigns identified Calpeptin as a drug candidate against SARS-CoV-2. Initially reported to target the viral main protease (Mpro), its moderate activity in Mpro inhibition assays hints at a second target. Indeed, we show that Calpeptin is an extremely potent cysteine cathepsin inhibitor, a finding additionally supported by X-ray crystallography. Cell infection assays proved Calpeptin's efficacy against SARS-CoV-2. Treatment of SARS-CoV-2-infected Golden Syrian hamsters with sulfonated Calpeptin at a dose of 1 mg/kg body weight reduces the viral load in the trachea. Despite a higher risk of side effects, an intrinsic advantage in targeting host proteins is their mutational stability in contrast to highly mutable viral targets. Here we show that the inhibition of cathepsins, a protein family of the host organism, by calpeptin is a promising approach for the treatment of SARS-CoV-2 and potentially other viral infections.

Form factor determination of biological molecules with X-ray free electron laser small-angle scattering (XFEL-SAS).

Free-electron lasers (FEL) are revolutionizing X-ray-based structural biology methods. While protein crystallography is already routinely performed at FELs, Small Angle X-ray Scattering (SAXS) studies of biological macromolecules are not as prevalent. SAXS allows the study of the shape and overall structure of proteins and nucleic acids in solution, in a quasi-native environment. In solution, chemical and biophysical parameters that have an influence on the structure and dynamics of molecules can be varied and their effect on conformational changes can be monitored in time-resolved XFEL and SAXS experiments. We report here the collection of scattering form factors of proteins in solution using FEL X-rays. The form factors correspond to the scattering signal of the protein ensemble alone; the scattering contributions from the solvent and the instrument are separately measured and accurately subtracted. The experiment was done using a liquid jet for sample delivery. These results pave the way for time-resolved studies and measurements from dilute samples, capitalizing on the intense and short FEL X-ray pulses.

Heterogeneity in M. tuberculosis ß-lactamase inhibition by Sulbactam.

For decades, researchers have elucidated essential enzymatic functions on the atomic length scale by tracing atomic positions in real-time. Our work builds on possibilities unleashed by mix-and-inject serial crystallography (MISC) at X-ray free electron laser facilities. In this approach, enzymatic reactions are triggered by mixing substrate or ligand solutions with enzyme microcrystals. Here, we report in atomic detail (between 2.2 and 2.7 Å resolution) by room-temperature, time-resolved crystallography with millisecond time-resolution (with timepoints between 3 ms and 700 ms) how the Mycobacterium tuberculosis enzyme BlaC is inhibited by sulbactam (SUB). Our results reveal ligand binding heterogeneity, ligand gating, cooperativity, induced fit, and conformational selection all from the same set of MISC data, detailing how SUB approaches the catalytic clefts and binds to the enzyme noncovalently before reacting to a trans-enamine. This was made possible in part by the application of singular value decomposition to the MISC data using a program that remains functional even if unit cell parameters change up to 3 Å during the reaction.

Online dynamic flat-field correction for MHz microscopy data at European XFEL.

The high pulse intensity and repetition rate of the European X-ray Free-Electron Laser (EuXFEL) provide superior temporal resolution compared with other X-ray sources. In combination with MHz X-ray microscopy techniques, it offers a unique opportunity to achieve superior contrast and spatial resolution in applications demanding high temporal resolution. In both live visualization and offline data analysis for microscopy experiments, baseline normalization is essential for further processing steps such as phase retrieval and modal decomposition. In addition, access to normalized projections during data acquisition can play an important role in decision-making and improve the quality of the data. However, the stochastic nature of X-ray free-electron laser sources hinders the use of standard flat-field normalization methods during MHz X-ray microscopy experiments. Here, an online (i.e. near real-time) dynamic flat-field correction method based on principal component analysis of dynamically evolving flat-field images is presented. The method is used for the normalization of individual X-ray projections and has been implemented as a near real-time analysis tool at the Single Particles, Clusters, and Biomolecules and Serial Femtosecond Crystallography (SPB/SFX) instrument of EuXFEL.

Ab initio spatial phase retrieval via intensity triple correlations.

Second-order intensity correlations from incoherent emitters can reveal the Fourier transform modulus of their spatial distribution, but retrieving the phase to enable completely general Fourier inversion to real space remains challenging. Phase retrieval via the third-order intensity correlations has relied on special emitter configurations which simplified an unaddressed sign problem in the computation. Without a complete treatment of this sign problem, the general case of retrieving the Fourier phase from a truly arbitrary configuration of emitters is not possible. In this paper, a general method for ab initio phase retrieval via the intensity triple correlations is described. Simulations demonstrate accurate phase retrieval for clusters of incoherent emitters which could be applied to imaging stars or fluorescent atoms and molecules. With this work, it is now finally tractable to perform Fourier inversion directly and reconstruct images of arbitrary arrays of independent emitters via far-field intensity correlations alone.

Hard X-ray stereographic microscopy for single-shot differential phase imaging.

The characterisation of fast phenomena at the microscopic scale is required for the understanding of catastrophic responses of materials to loads and shocks, the processing of materials by optical or mechanical means, the processes involved in many key technologies such as additive manufacturing and microfluidics, and the mixing of fuels in combustion. Such processes are usually stochastic in nature and occur within the opaque interior volumes of materials or samples, with complex dynamics that evolve in all three dimensions at speeds exceeding many meters per second. There is therefore a need for the ability to record three-dimensional X-ray movies of irreversible processes with resolutions of micrometers and frame rates of microseconds. Here we demonstrate a method to achieve this by recording a stereo phase-contrast image pair in a single exposure. The two images are combined computationally to reconstruct a 3D model of the object. The method is extendable to more than two simultaneous views. When combined with megahertz pulse trains of X-ray free-electron lasers (XFELs) it will be possible to create movies able to resolve 3D trajectories with velocities of kilometers per second.

Fourth-generation light sources.

New fourth-generation synchrotron radiation facilities bring large gains in X-ray source brightness, but also challenges in making full use of their potentials. Some of these challenges have been faced at X-ray free-electron laser facilities.

Imaging via Correlation of X-Ray Fluorescence Photons.

We demonstrate that x-ray fluorescence emission, which cannot maintain a stationary interference pattern, can be used to obtain images of structures by recording photon-photon correlations in the manner of the stellar intensity interferometry of Hanbury Brown and Twiss. This is achieved utilizing femtosecond-duration pulses of a hard x-ray free-electron laser to generate the emission in exposures comparable to the coherence time of the fluorescence. Iterative phasing of the photon correlation map generated a model-free real-space image of the structure of the emitters. Since fluorescence can dominate coherent scattering, this may enable imaging uncrystallised macromolecules.

Dose-efficient scanning Compton X-ray microscopy.

The highest resolution of images of soft matter and biological materials is ultimately limited by modification of the structure, induced by the necessarily high energy of short-wavelength radiation. Imaging the inelastically scattered X-rays at a photon energy of 60 keV (0.02 nm wavelength) offers greater signal per energy transferred to the sample than coherent-scattering techniques such as phase-contrast microscopy and projection holography. We present images of dried, unstained, and unfixed biological objects obtained by scanning Compton X-ray microscopy, at a resolution of about 70 nm. This microscope was realised using novel wedged multilayer Laue lenses that were fabricated to sub-ångström precision, a new wavefront measurement scheme for hard X rays, and efficient pixel-array detectors. The doses required to form these images were as little as 0.02% of the tolerable dose and 0.05% of that needed for phase-contrast imaging at similar resolution using 17 keV photon energy. The images obtained provide a quantitative map of the projected mass density in the sample, as confirmed by imaging a silicon wedge. Based on these results, we find that it should be possible to obtain radiation damage-free images of biological samples at a resolution below 10 nm.

Crystal diffraction prediction and partiality estimation using Gaussian basis functions.

The recent diversification of macromolecular crystallographic experiments including the use of pink beams, convergent electron diffraction and serial snapshot crystallography has shown the limitations of using the Laue equations for diffraction prediction. This article gives a computationally efficient way of calculating approximate crystal diffraction patterns given varying distributions of the incoming beam, crystal shapes and other potentially hidden parameters. This approach models each pixel of a diffraction pattern and improves data processing of integrated peak intensities by enabling the correction of partially recorded reflections. The fundamental idea is to express the distributions as weighted sums of Gaussian functions. The approach is demonstrated on serial femtosecond crystallography data sets, showing a significant decrease in the required number of patterns to refine a structure to a given error.

JINXED: just in time crystallization for easy structure determination of biological macromolecules.

Macromolecular crystallography is a well established method in the field of structural biology and has led to the majority of known protein structures to date. After focusing on static structures, the method is now under development towards the investigation of protein dynamics through time-resolved methods. These experiments often require multiple handling steps of the sensitive protein crystals, e.g. for ligand-soaking and cryo-protection. These handling steps can cause significant crystal damage, and hence reduce data quality. Furthermore, in time-resolved experiments based on serial crystallography, which use micrometre-sized crystals for short diffusion times of ligands, certain crystal morphologies with small solvent channels can prevent sufficient ligand diffusion. Described here is a method that combines protein crystallization and data collection in a novel one-step process. Corresponding experiments were successfully performed as a proof-of-principle using hen egg-white lysozyme and crystallization times of only a few seconds. This method, called JINXED (Just IN time Crystallization for Easy structure Determination), promises high-quality data due to the avoidance of crystal handling and has the potential to enable time-resolved experiments with crystals containing small solvent channels by adding potential ligands to the crystallization buffer, simulating traditional co-crystallization approaches.

Decreasing ultrafast x-ray pulse durations with saturable absorption and resonant transitions.

Saturable absorption is a nonlinear effect where a material's ability to absorb light is frustrated due to a high influx of photons and the creation of electron vacancies. Experimentally induced saturable absorption in copper revealed a reduction in the temporal duration of transmitted x-ray laser pulses, but a detailed account of changes in opacity and emergence of resonances is still missing. In this computational work, we employ nonlocal thermodynamic equilibrium plasma simulations to study the interaction of femtosecond x rays and copper. Following the onset of frustrated absorption, we find that a K-M resonant transition occurring at highly charged states turns copper opaque again. The changes in absorption generate a transient transparent window responsible for the shortened transmission signal. We also propose using fluorescence induced by the incident beam as an alternative source to achieve shorter x-ray pulses. Intense femtosecond x rays are valuable to probe the structure and dynamics of biological samples or to reach extreme states of matter. Shortened pulses could be relevant for emerging imaging techniques.

An experimental study of liquid micro-jets produced with a gas dynamic virtual nozzle under the influence of an electric field.

The results of an experimental study of micro-jets produced with a gas dynamic virtual nozzle (GDVN) under the influence of an electric field are provided and discussed for the first time. The experimental study is performed with a 50% volume mixture of water and ethanol, and nitrogen focusing gas. The liquid sample and gas Reynolds numbers range from 0.09-5.4 and 0-190, respectively. The external electrode was positioned 400-500 µm downstream of the nozzle tip and an effect of electric potential between the electrode and the sample liquid from 0-7 kV was investigated. The jetting parametric space is examined as a function of operating gas and liquid flow rates, outlet chamber pressure, and an external electric field. The experimentally observed jet diameter, length and velocity ranged from 1-25 µm, 50-500 µm and 0.5-10 m/s, respectively. The jetting shape snapshots were processed automatically using purposely developed computer vision software. The velocity of the jet was calculated from the measured jet diameter and the sample flow rate. It is found that micro-jets accelerate in the direction of the applied electric field in the downstream direction at a constant acceleration as opposed to the standard GDVNs. New jetting modes were observed, where either the focusing gas or the electric forces dominate, encouraging further theoretical and numerical studies towards optimized system design. The study shows the potential to unlock a new generation of low background sample delivery for serial diffraction measurements of weakly scattering objects.

Speckle contrast of interfering fluorescence X-rays.

With the development of X-ray free-electron lasers (XFELs), producing pulses of femtosecond durations comparable with the coherence times of X-ray fluorescence, it has become possible to observe intensity-intensity correlations due to the interference of emission from independent atoms. This has been used to compare durations of X-ray pulses and to measure the size of a focusedX-ray beam, for example. Here it is shown that it is also possible to observe the interference of fluorescence photons through the measurement of the speckle contrast of angle-resolved fluorescence patterns. Speckle contrast is often used as a measure of the degree of coherence of the incident beam or the fluctuations of the illuminated sample as determined from X-ray diffraction patterns formed by elastic scattering, rather than from fluorescence patterns as addressed here. Commonly used approaches to estimate speckle contrast were found to suffer when applied to XFEL-generated fluorescence patterns due to low photon counts and a significant variation of the excitation pulse energy from shot to shot. A new method to reliably estimate speckle contrast under such conditions, using a weighting scheme, is introduced. The method is demonstrated by comparing the speckle contrast of fluorescence observed with pulses of 3 fs to 15 fs duration.