Cryo-EM Analyses Permit Visualization of Structural Polymorphism of Biological Macromolecules

Abstract

The functions of biological macromolecules are often associated with conformational malleability of the structures. This phenomenon of chemically identical molecules with different structures is coined structural polymorphism. Conventionally, structural polymorphism is observed directly by structural determination at the density map level from X-ray crystal diffraction. Although crystallography approach can report the conformation of a macromolecule with the position of each atom accurately defined in it, the exploration of structural polymorphism and interpreting biological function in terms of crystal structures is largely constrained by the crystal packing. An alternative approach to studying the macromolecule of interest in solution is thus desirable. With the advancement of instrumentation and computational methods for image analysis and reconstruction, cryo-electron microscope (cryo-EM) has been transformed to be able to produce "in solution" structures of macromolecules routinely with resolutions comparable to crystallography but without the need of crystals. Since the sample preparation of single-particle cryo-EM allows for all forms co-existing in solution to be simultaneously frozen, the image data contain rich information as to structural polymorphism. The ensemble of structure information can be subsequently disentangled through three-dimensional (3D) classification analyses. In this review, we highlight important examples of protein structural polymorphism in relation to allostery, subunit cooperativity and function plasticity recently revealed by cryo-EM analyses, and review recent developments in 3D classification algorithms including neural network/deep learning approaches that would enable cryo-EM analyese in this regard. Finally, we brief the frontier of cryo-EM structure determination of RNA molecules where resolving the structural polymorphism is at dawn.

Keywords: classification; conformation alanalysis; cryo-TEM; crystal; heterogeneity; solution; strucural analysis; x-ray.

FIGURE 1

Snapshots of TRPV cryo-EM structures (A) four TRPV structures superimposed: apo followed by three different conformations bound with DkTx toxin (EMD-23136, PDB: 7L2P; EMD-23143; PDB: 7L2S; EMD-23141; PDB: 7L2T; EMD-23141; PDB: 7L2U) (Zhang et al., 2021). (B) a cartoon of energy landscape for illustrating the four different structures in (A).

FIGURE 2

Single particle cryo-EM processing and milestones (A) Single particle processing. There are in general a total of 10 steps in the processing where the red box encloses the pre-processing steps, which can be executed in an on-the-fly manner. At the bottom, the post-processing steps encased by the green box are used to improve the overall resolution and local map quality as well. Bayesian polishing on RELION takes care of the dose weighting to compensate the frame-dependent radiation damage effect. Once a high-quality map is obtained, the variation of defocus and higher order aberration at per particle level can be further estimated and corrected. (B) Cryo-EM milestones. The boxes on the left of the timeline highlight key advancement of hardware and software where “AF image shift” stands for “aberration-free image shift.” The boxes on the right of the timeline indicate significant cryo-EM structures and events including recent rapid structure determination of COVID-19 proteins, achievement of atomic resolution with apo-ferritin, and the Nobel Physiology or Medicine in 2021 that recognizes the discoveries of TRPV and Piezo channels as heat and pressure sensors respectively where the structures were solely obtained by cryo-EM.

FIGURE 3

Single particle cryo-EM reveals co-exiting conformations of haemoglobin. (A) Visualization of haemoglobin with the aid of Volta phase plate (see the milestones in Figure 2B). (This figure is adopted from Figure 1A of Ref 41, an open access article distributed under the terms of the Creative Commons CC BY license. https://creativecommons.org/) (B) Modeling the cryo-EM map with known X-ray structures suggests the co-existing conformations. (This figure is adopted from Figure 1E of Ref 41, an open access article distributed under the terms of the Creative Commons CC BY license. https://creativecommons.org/). (C) Visualization of haemoglobin in thin ice without phase plate. This figure is adopted from Figure 1A of Ref 42. (D) Resolving the cryo-EM map with the usage of K-way 3D classification on RELION indicates the co-existing conformations. (This figure is adopted from Figure 1E of Ref 42, an open access article distributed under the terms of the Creative Commons CC BY license. https://creativecommons.org/).

FIGURE 4

Nanodisc system facilitates cryo-EM analyses of membrane proteins in native environment. (A) Reconstitution of a nanodisc-membrane protein complex: lipids enclosed by membrane scaffold proteins (MSP) to form a nanodisc are assembled with a membrane protein to form a nanodisc-membrane protein complex. (This figure is made by modifying nanodisc cartoons available from BioRender with permission under a purchased license.) (B) Cryo-EM structure of a connexin protein in nanodisc. (C) Revelation of lipid-protein interactions (B and C) are adopted from Figures 1, 3 in Denisov and Sligar (2016) with permission from Creative Commons license).

FIGURE 5

Cryo-EM analyses of GroEL, a multi-subunit protein machinery. (A) Segmentation of a high-resolution cryo-EM reconstruction. This allows extraction of individual subunit for further analysis. (B) Local Resolution map of GroEl. It shows the apical domain appears to be more dynamic as compared to equatorial domain. (C) Focused 3D classification on individual subunit. It reveals three major conformers (This set of figures is adopted from Roh et al (2017) with permission granted by NAS).

FIGURE 6

Data visualization of structural variability resolved via PCA analyses. (A) t-SNE plot of the results from classical PCA (Penczek et al., 2011). (B) t-SNE plot of the results from two-stage PCA (2SDR) (Chung et al., 2020). As stated in Chung et al., 2020, We followed the procedure in Penczek et al., 2011 to generate a dataset containing 9,453 simulated cryo-EM particle images projected from five 70S ribosome conformations with minor differences resulting from combinations of the absence or presence of tRNA (transfer RNA) and EF-G (elongation factor G). We then resampled these particle images to generate 11,000 3D volumes (density maps) on 75 × 75 × 75 voxels. Next, we solved the eigenvolumes using PCA (Penczek et al., 2011). or 2SDR (Chung et al., 2020). and compared the performance of these two methods using the factorial coordinates defined in Penczek et al., 2011. The t-SNE plots and k-means with 5 classes on their factorial coordinates indicate that the eigenvolumes solved by 2SDR have clearly resolved the structural variability (This set of figures is adopted from Chung et al., 2020 with permission given by International Press.).