BioWAXS: experiment and interpretation

Conference: 2021: 71st ACA Annual Meeting
08/05/2021: 12:00 PM - 3:00 PM
Oral Session 


The molecules of life are in constant motion. To gain insight into dynamics on the atomic scale, it is increasingly common to apply wide-angle X-ray scattering (WAXS) to biomolecular solutions or condensed phases, especially in conjunction with time-resolved experiments. The WAXS signal is uniquely sensitive to dynamic structural fluctuations and hydration effects, both of which are key to a mechanistic understanding of function. However, the challenge of interpreting the WAXS signal has traditionally limited its biological applications. In recent years, the challenge has been addressed through a combination of new experimental techniques (such as temperature-jump and X-FEL) and sophisticated computational approaches (such as ensemble fitting, physics-based modeling, and machine learning). In this session, we aim to advance the field by highlighting recent experimental results and new methods for interpretation.


Revealing structural dynamics of proteins with time-resolved x-ray liquidography

Time-resolved x-ray scattering has now been fully established as a powerful method to investigate molecular structural dynamics in solution [1-10]. We have employed the technique to study structural dynamics of a wide variety of small molecules including diatomic molecules, haloalkanes, organometallic complexes over timescales from femtoseconds to milliseconds. Moreover, the technique has been applied to protein molecules to unveil structural dynamics of proteins in solution. One of the representative experimental methods commonly used for studying structural dynamics of proteins is time-resolved x-ray crystallography. Time-resolved x-ray crystallography can determine atomic-resolution structures of proteins, but has been applied to limited proteins due to the stringent prerequisites such as highly-ordered and radiation-resistant single crystals. The problem can be overcome by applying time-resolved x-ray diffraction directly to protein solutions rather than protein single crystals. To emphasize that structural information can be obtained from the liquid phase, this time-resolved x-ray solution scattering technique is named time-resolved x-ray liquidography (TRXL) in analogy to time-resolved x-ray crystallography where the structural information of reaction intermediates is obtained from the crystalline phase. We will present our recent results including the achievement of femtosecond TRXL using an x-ray free-electron laser. We investigated ultrafast structural dynamics of homodimeric hemoglobin (HbI), which is a model system for studying allosteric structural transition in proteins. From femtosecond TRXL data of HbI, we revealed detained ultrafast structural changes of HbI involving a coherent motion of the protein body and time-dependent change of electron density in the hydration shell.

[1] H. Ihee et al., Science, 309, 1223-1227 (2005).
[2] K. H. Kim et al., J. Am. Chem. Soc., 134, 7001-7008 (2012).
[3] T. W. Kim et al., J. Am. Chem. Soc., 134, 3145-3153 (2012).
[4] K. H. Kim et al., Nature, 518, 385-389 (2015).
[5] J. G. Kim et al., Acc. Chem. Res., 48, 2200-2208 (2015).
[6] J. G. Kim et al., Struct. Dyn., 3, 023601 (2016).
[7] H. Kim et al., J. Phys. Chem. B, 124, 1550-1556 (2020).
[8] T. W. Kim et al., Proc. Natl. Acad. Sci., 117, 14996-15005 (2020).
[9] J. G. Kim et al., Nature, 582, 520-524 (2020).
[10] M. Choi et al., Chem. Sci., DOI: 10.1039/D1SC01207J, Online publication: 2021-05-10. 

View Abstract 579


Hyotcherl Ihee, Dept of Chemistry, KAIST, & Inst for Basic Science Daejeon

Probing photoinduced protein function by time resolved X-ray scattering

Biological macromolecules are characterized by specific structural and dynamic features that are at the basis of their biological activity. Understanding macromolecular activity thus requires studying structural changes over time and on various time scales. Time-resolved X-ray scattering permits tracking macromolecular conformational changes along a photoinduced reaction pathway. I will show results from our studies on light-induced protein structural dynamics in model systems and in photoreceptors involved in cell signalling, photoprotection and photoinduced gene regulation. 

View Abstract 643


Giorgio Schiro, IBS - CNRS Grenoble

Coffee Break

Instrumentation and applications of simultaneous SAXS/WAXS at the LiX beamline

At the Life Science X-ray Scattering (LiX) beamline at NSLS2, we have developed instrumentation to collect scattering data that covers an extended range of scattering vectors (q). This is accomplished using a custom C-shaped wide-angle detector along side the small-angle detector, providing a combined q-range of ~0.006-3.2 Å-1 in routine data collection, with reasonable coverage of azimuthal angles. I will describe the technical details in data processing to produce the combined SAXS/WAXS data. I will then discuss applications of the wide-angle data in user research, for both solution scattering and characterization of structures in biological tissues. 

View Abstract 534


Lin Yang, Brookhaven National Laboratory Upton, NY 

Classification of tissue variations in X-ray scanning microdiffraction from thin sections of human brain

β-amyloid plaques and neurofibrillary tangles (NFTs) are major hallmarks of pathology in Alzheimer's disease (AD). The plaques and NFTs exhibit wide structural variations in different regions of human brain. With the use of small (SAXS) and wide (WAXS) angle scattering from histological sections of AD human brain tissue, the underlying fibrillar shape, structure, and types of tissue lesions can be analyzed, detected, and classified over the different parts of individual and multiple human brains from AD subjects. Brain tissue from AD subjects is sectioned and scanned with the 5µm diameter X-ray beam to generate thousands of patterns arrayed across a region of interest. These data are used to generate maps of scattering attributes that reflect the variation of tissue structure and the spatial distribution of amyloid plaques and NFTs. However, the low intensity scattering from tissue lesions can be obscured by substrate, air, and mica scattering leading to low signal to noise ratio in the resulting diffraction patterns. The variation of tissue structure on multiple length scales combined with the polymorphism of NFT and amyloid aggregates presents significant challenges to structural characterization. We are addressing these challenges using state-of-art signal detection and classification methods including principal component analysis (PCA), support vector machine (SVM), and artificial neural network (ANN). The spatial distribution of structural attributes of β-amyloid and NFTs as mapped with scanning X-ray microdiffraction may lead to a better understanding of molecular basis of disease progression in AD.

This work is supported by NIH grant number R21AG068972 

View Abstract 610


Abdullah Al Bashit, Northeastern University Medford, MA 

Additional Author(s)

Prakash Nepal, Dept. of Bioengineering Boston, MA 
Lee Makowski, Northeastern University Boston, MA 

Decoding hidden structural information in solution wide-angle X-ray scattering (WAXS): from ensemble modeling to machine learning

Solution X-ray scattering is a powerful method for determining in vitro macromolecular structures. Small-angle X-ray scattering (SAXS) has been used to investigate global shape and size of a myriad of molecules at 10s-100
Å resolution. Its extension, wide-angle X-ray scattering (WAXS), sharpens the resolution of solution scattering to below 10 Å, enhancing information about macromolecular structures. These WAXS profiles, although informative, can be challenging to interpret, both as a result of contributions of the solvent, and the unknown structures related to the finer features of the solute. We applied WAXS to study the salt dependent structures of a short (12-base pair long) RNA duplexes. With a structural pool containing a variety of conformations, we leveraged ensemble modeling to understand the RNA's solution conformations [1]. Furthermore, we applied trained supervised machine learning (ML) models, mapping WAXS profiles to structural features, to interpret the noisy, yet information rich experimental WAXS profiles [2]. These well-trained ML models faithfully captured the WAXS fingerprints of RNAs. Importantly, these models identified specific scattering angles, or regions of scattering angles, which reflect (and can be used to predict) distinct structural parameters (WAXS feature importance). Therefore, we demonstrated that the ensemble and ML methods can provide novel frameworks for extracting highly relevant structural information from solution experiments on biological macromolecules.

1. Chen, Y.L. and Pollack, L., Salt dependence of A-form RNA duplexes: structures and implications, J. Phys. Chem. B 2019, 123, 46, 9773–9785
2. Chen, Y.L. and Pollack, L., Machine learning deciphers structural features of RNA duplexes measured with solution X-ray scattering, IUCrJ 2020, 7, 5 

View Abstract 636


Yen-Lin Chen, Cornell University Boston, MA 

Additional Author

Lois Pollack, Cornell University Ithaca, NY 

Visualization of biomolecular structures by WAXS and MD

Solution wide-angle X-ray scattering (WAXS) has exhibited its promising potential for characterizing conformational states of biomolecular structures with increasing accuracy at near-native conditions. WAXS probes structural information about biomolecules and their variations. However, the extraction of this hidden information is nontrivial and the accurate interpretation of the solution scattering data is prevented by the low resolution of the data. Here, we integrate WAXS with all-atom molecular dynamics (MD) simulations to investigate the solution structure of macromolecules, including DNA and RNA. This WAXS-MD strategy achieved excellent agreement between measured and simulated WAXS profiles and allows insights into the structural dynamics on the atomic scale which is inaccessible in experiment. From computer simulations and machine learning based analysis, we built correlation maps to visualize the relationship between well-defined features in the scattering profiles and real space characteristics of macromolecules. Notably, our analysis reveals that double stranded RNA (dsRNA) displays a marked insitu structural bending induced by G-tract-specific (i.e., GGG sequence dependence) cations binding, and tandem uracils (UUU) of RNA triple helices perform unique function in stabilizing their tertiary structure. Though demonstrated for nucleic acids duplex and triplex, our general approach can be applied to solve flexible single stranded RNA (ssRNA) in solution and can extend to study dynamic systems containing both protein and RNA partners. 

View Abstract 634


Weiwei He, New York University New York, NY 

Additional Author(s)

Yen-Lin Chen, Cornell University Boston, MA 
Serdal Kirmizialtin, New York University Abu Dhabi Abu Dhabi
Lois Pollack, Cornell University Ithaca, NY 

Probing structural dynamics of biomolecules in solution at high resolution via time-resolved small- and wide-angle x-ray scattering

To understand how biomolecules execute their designed function, it is crucial to know not only their native structures, but also how those structures evolve over time. The time-dependent x-ray scattering signature unique to a biomolecule is accompanied by scattering from its hydration shell, the surrounding buffer, the sample container, and anything else in the path between the x-ray source and the beamstop. When the sample undergoes a laser-triggered temperature jump, the scattering signature from both the biomolecule and its surrounding solvent can change. To isolate the biomolecule contribution, it is quite helpful to characterize the SAXS/WAXS pattern over a broad range of q and temperatures. To that end, we have helped develop on the BioCARS beamline at the APS the infrastructure needed to probe structural dynamics of biomolecules in solution at high resolution in both space and time by acquiring time-resolved small- and wide-angle x-ray scattering patterns (SAXS/WAXS) following either photo activation of a chromophore or a laser-induced temperature jump. This infrastructure includes home-built instrumentation that allows us to rapidly scan and precisely control the sample temperature over a range spanning -20 - 120 ˚C, and probe time-resolved changes in the scattering signature following a temperature jump with time resolution down to the <10 ns duration of the laser pulse. The time- and temperature-dependent scattering patterns recorded with this setup can be reproduced with reasonably high fidelity with a model involving linear combinations of temperature-independent biomolecule scattering patterns and temperature-dependent scattering from the hydration shell and surrounding solvent. The biomolecule contribution to the scattering, which is acquired out to 5.6 Å-1, contains structural information down to chemical bond distances, and is sensitive to secondary structure. An additional benefit of acquiring scattering data over a broad range of temperatures is that it allows precise characterization of relative populations of species in equilibrium as well as thermodynamic properties of folding/unfolding, substrate binding, or oligomerization. 

View Abstract 635


Philip Anfinrud, National Institutes of Health Bethesda, MD 

Additional Author(s)

Hyun Sun Cho, National Institutes of Health Bethesda, MD 
Friedrich Schotte, National Institutes of Health Bethesda, MD 
Valentyn Stadnytskyi, National Institutes of Health Bethesda, MD