Determining the structure of a target with its ligand is key in guiding the development of new drug molecules. Cryo-electron microscopy (cryo-EM) is transforming drug discovery by helping researchers obtain structural insights into a broad range of otherwise intractable biological targets. Because samples are flash frozen at cryogenic temperatures (below -175 °C), proteins can be examined in their near-native state. Additionally, as there are no upper size limits for samples, complete, functional molecular machines can be imaged in 3D, including one or several ligands, revealing biologically relevant mechanisms. Understanding such mechanisms will facilitate the development of new therapeutics.
Cryo-EM structural insight is useful for drug discovery
Structural biology has been key in optimizing new leads. X-ray diffraction (XRD) and nuclear magnetic resonance (NMR) were the two most widely used techniques for the determination of macromolecular structures at high resolution, but both have limitations relating to particle size, the hydrophobic nature of macromolecules, and the amount of sample available. With cryo-EM, researchers don’t have these restrictions and can therefore image targets that have not yet been solved. Cryo-EM is directly applicable to fragment-based drug discovery (FBDD), as it can bring a suitable level of detail to molecular interactions between fragments and their targets. For example, the structures of three fragment-size ligands in complex with β-galactosidase were obtained by cryo-EM at resolutions greater than 2.3Å.1 In general, a resolution of 3Å or better is required to determine the binding mode of a fragment-sized ligand structure, and such resolution can be routinely achieved using cryo-EM. Crystals of small parts of commercially interesting targets bound to therapeutic antibody fragments can be analyzed by XRD, but because cryo-EM does not require crystallization and there are no limitations in terms of size, a whole target bound to fragment antibodies (Fabs) can be imaged. For example, the antibody rituximab, known to treat certain autoimmune diseases and types of cancer, has been imaged recently with its target CD20 by cryo-EM.2 A structure-based model of the CD20-rituximab complex (figure 1) has been devised, and new observations made based on the structure of the entire complex revealed its mechanism, which was unclear until now. Deciphering the binding mode of the antibody to the receptor is essential to discover and develop more efficient antibody-based drugs.
Figure 1: CD20: rituximab complex by cryo-EM. The structural details revealed the interactions between CD20, a B cell membrane protein dimer, and two Fabs (heavy chain in purple and light chain in pink) in a glyco-diosgenin (GDN) micelle.
Credit: Model adapted from2 by Hans Raaijmakers.
Cryo-EM is a unique tool for lead discovery
Since a cryo-EM 3D molecular structure is based on a single particle analysis (SPA) of the macromolecules in solution, crystallization is not required. Therefore, the particle being studied does not need to be in a specific buffer and can be advantageously analyzed in native conditions, resulting in biologically relevant structures. Hence, the structure of native complexes can be imaged in 3D. Among the most difficult targets are G protein-coupled receptors (GPCRs). They represent the largest receptor family in the human genome and are implicated in numerous diseases, with more than 30% of marketed drugs targeting this receptor family. In this context, tremendous efforts were made over decades to determine GPCR structures using XRD and NMR, but only truncated models of more stable complexes could be obtained. In 2014, the cryo-EM “resolution revolution” marked a turning point3 where cryo-EM joined XRD and NMR to form a complementary trio of techniques. Researchers were prompt to adopt cryo-EM and solved outstanding structures of GPCRs (class B GPCR–G-protein complex4; active, complexed, human CGRP receptor bound to Gs-protein5; adenosine A2A receptor coupled to an engineered heterotrimeric G protein6; biased agonist-bound human GLP-1 receptor–Gs complex7). Recently, the structure of a GPCR-G complex stabilized using an antibody fragment has been elucidated.8 The ability of the mAb16 antibody to bind both the Gα and Gβγ subunits, stabilizing the GPCR-G complex, even in the presence of GTPγS, is unique. This approach could be applied to other G-proteins with minimal effort, facilitating structural studies of GPCR/G-protein complexes.
The scope of cryo-EM is not limited to small membrane protein complexes. Cryo-EM also allows researchers to solve structures of whole viruses and dissect how they interact with therapeutics. As various conformations of structures can be extracted in silico from the vitrified sample, cryo-EM can reveal multiple functional states. Recent health outbreaks invite us to look at some examples where structural variations between different conformations could explain the mechanism of infection of cells by a virus. Over the past decade, about one-third of the diagnosed cases of the viral respiratory infection caused by the Middle East Respiratory Syndrome coronavirus (MERS-CoV) were lethal. Using cryo-EM structural insights, a potential mechanism has been suggested for the initiation of the virus–cell membrane fusion.9 This generalizable structure-based approach keeps the large transmembrane spike protein (responsible for fusion) of the virus in an antigenically optimal conformation so that it can be targeted by the immune system. Based on this finding, strongly neutralizing antibodies against MERS-CoV are currently being developed.
Recently, the SARS-CoV-2 coronavirus (which causes the COVID-19 disease) was found to display an affinity for the human angiotensin-converting enzyme 2 receptor (ACE2), similar to the SARS-CoV that spread globally in 2002, allowing the fusion of the virus with the target cell, and its infection.10 Rapidly, near atomic structures of the SARS-CoV-2 spike glycoprotein trimer were solved using cryo-EM (figure 2), providing unmatched insights about conserved and accessible epitopes to support the development of vaccines and inhibitors of viral entry. Cryo-EM results show that the SARS-CoV-2 spike protein displays a range of conformations, from closed to open, and support hypotheses that the partially open states of the glycoprotein trimers are found in highly pathogenic human coronaviruses. This is different from the human coronaviruses associated with common colds, where largely closed glycoprotein trimers are found, and presents insights that were not visualized by any technique other than cryo-EM. Understanding the consequences of the variability of the 3D scaffold of a target will bring new theories about the ligand binding mechanism, potentially leading to the discovery of new drug molecules.
Figure 2: 3.2Å cryo-EM structure of the SARS-CoV-2 S glycoprotein in open conformation made publicly available.
Credit: Model adapted from 10 by Hans Raaijmakers.
Cryo-EM and lead optimization
By providing structural insights, cryo-EM is crucial in supporting both lead discovery and lead optimization, as the reproducibility, quality and throughput of cryo-EM is adequate to support fragment screening.1 Methods to streamline the rapid preparation of high-quality samples, resulting in a high-resolution map by cryo-EM were set up, exemplified by the 3.1Å structure of 2.6 MDa yeast fatty acid synthase (FAS) carried out within a day.11 This included the rapid production of the purified yeast FAS, the cryo-EM specimen vitrification for high-resolution data collection, and the data collection itself. This productivity with cryo-EM was inconceivable just a few years ago. Sample purification and preparation is different for each macromolecule: this step is particularly important for understanding the molecular mechanisms underlying the pharmacological actions of ligands in drug discovery. By focusing on the careful preparation of the sample, rather than the conditions required for its imaging, researchers managed to preserve the FAS phospho-pantetheine transferase domain, known to otherwise denature during crystallization.11 With this set up, the researcher can focus on issues upstream to the structural analysis including cloning, expression, purification and the strict monitoring of protein integrity.
Cryo-EM’s ability to meet the requirements for fragment screening in terms of both resolution and throughput has been demonstrated using the oncology target PKM2.1 Structure-based fragment screening of 68 ligands with PKM2 has been performed within the timeframe usually required for such experiments. Furthermore, to improve the throughput, the use of fragment “cocktails” can be successfully achieved using cryo-EM.1 Finally, with powerful detectors and new optical modes featured on modern cryo-EM microscopes, more than 400 images can be acquired per hour, and acquiring a full dataset can take less than two hours. Hence, at optimum use and capacity, a cryo-electron microscope could complete a 400-fragment screen in less than a month.
In conclusion, SPA by cryo-EM delivers resolutions suitable for lead discovery, and its throughput is high enough to support lead optimization. It produces observation that researchers did not have access to before, leading to expedited discovery of therapies and vaccines, helping make the world a healthier and safer place.
1. Saur M, J.Hartshorn M, Dong J, et al. Fragment-based drug discovery using cryo-EM. Drug Discovery Today. 2019:1-6. doi:doi.org/10.1016/j.drudis.2019.12.006.
2. Rougé L, Chiang N, Steffek M, et al. Structure of CD20 in complex with the therapeutic monoclonal antibody rituximab. Science. 2020.
3. Kühlbrandt W. The Resolution Revolution. Science. 2014;343:1443-1444.
4. Liang Y-L, Khoshouei M, Radjainia M, et al. Phase-plate cryo-EM structure of a class B GPCR–G-protein complex. Nature. 2017;546:118-123.
5. Liang Y, Khoshouei M, Deganutti G, et al. Cryo-EM structure of the active, Gs-protein complexed, human CGRP receptor. Nature. 2018;492:492-497.
6. García-Nafría J, Lee Y, Bai X, Carpenter B, Tate CG. Cryo-EM structure of the adenosine A2A receptor coupled to an engineered heterotrimeric G protein. eLIFE. 2018;7:1-19.
7. Liang Y-L, Khoshouei M, Glukhova A, et al. Phase-plate cryo-EM structure of a biased agonist-bound human GLP-1 receptor–Gs complex. Nature. 2018;555:121-125.
8. Maeda S, Koehl A, Matile H, et al. Development of an antibody fragment that stabilizes GPCR/G-protein complexes. Nature communications. 2018;9:3712:1-9. doi:doi:10.1038/s41467-018-06002-w.
9. Pallesen J, Wang N, Corbett KS, et al. Immunogenicity and structures of a rationally designed prefusion MERS-CoV spike antigen. PNAS. 2017:7348-7357. doi:doi:10.1073/pnas.1707304114.
10. Walls AC, Park Y-J, Tortorici MA, Wall A, McGuire AT, Veesler D. Structure, Function, and Antigenicity of the SARS-CoV-2 Spike Glycoprotein. Cell. 2020;180:1-12.
11. Joppe M, D’Imprima E, Salustros N, et al. The resolution revolution in cryoEM requires high-quality sample preparation: a rapid pipeline to a high-resolution map of yeast fatty acid synthase. IUCrJ. 2020;7:1-8. doi:doi:10.1107/S2052252519017366.