A key step in small molecule drug development is lead optimization, where a lead compound against a target protein is optimized for potency and selectivity, among other parameters. Sometimes diversity-oriented syntheses are sufficient to generate a lead with good pharmacokinetic and pharmacodynamic properties to obtain a drug candidate. However, direct visualization of how a compound binds to a protein target can greatly expedite lead optimization through structure-based drug design (SBDD) and provide stronger cases for mechanisms-of-action in patent and regulatory filings.
Here, we explore how cryo-EM overcomes bottlenecks and supports effective and successful drug discovery programs.
The power of the resolution revolution
Over the past decade, the “resolution revolution” has brought cryogenic transmission electron microscopy (cryo-EM) to the forefront of structural biology. Technological advances now allow for the routine structural determination of macromolecular complexes as small as 100 kDa (and even smaller, under specialized conditions). Cryo-EM is now the preferred method of structure determination for many classes of proteins that are not amenable to crystallography.
Cryo-EM requires a lower sample concentration (often by an order of magnitude), is more tolerant of local structural flexibility, is suitable for structure determination of proteins in lipid-detergent micelles and nanodiscs and can image proteins in the MegaDalton size range. For these reasons, cryo-EM has become a key method in SBDD for many high-priority classes of pharmaceutical targets, including G-protein coupled receptors (GPCRs), ion channels, transient receptor potential (TRP) channels, viral spike proteins, and multi-subunit complexes.
Structure-grade protein for use in cryo-EM
The first, and often the hardest, step towards obtaining a high-resolution cryo-EM structure is obtaining pure and monodisperse proteins. Purity refers to the ability to separate the protein of interest from other cellular contaminants. Monodispersity refers to the presence of a single oligomeric state, free of aggregates and other undesirable species. Each protein production workflow is unique and to some extent must be developed de novo. However, best practices have emerged for many classes of proteins that streamline expression and characterization. Below we will illustrate how NanoImaging Services’s protein science related-services can help you obtain “cryo-EM quality” samples.
Construct design and expression optimization
The first step towards obtaining a structure-grade purified protein is the design of one or more constructs optimized for structure determination. Genetically encoded affinity tags, such as FLAG-, Strep-, and His-tags are appended onto the N- and/or C-termini of the protein of interest to facilitate purification using affinity chromatography. His-tags are often preferred for crystallography due to the affinity resin’s lower cost and ease of elution. However, cryo-EM requires less sample material than X-ray crystallography, enabling the use of more expensive affinity tag resins such as those that bind Strep and FLAG tags. Poor protein expression is often addressed through the incorporation of one or more solubility tags, such as SUMO, MBP, and GFP, or by truncating regions of predicted disorder from the protein. These larger tags are often removed from proteins destined for crystallography, which lengthens the purification procedure and introduces further contaminants that must be purified away. These tags may be maintained for proteins targeted for cryo-EM, although they may ultimately induce undesirable preferred orientation issues and thus require removal.
In addition to the protein construct itself, selecting a proper expression system is critical to obtaining protein suitable for cryo-EM. For smaller and simpler proteins, bacteria and yeast might prove suitable expression hosts. However, many human proteins require efficient transcription or translation machinery, chaperones, trafficking and insertion into appropriate membranes, or specific post-translational modifications that require heterologous expression in more complex organisms. Therefore, many human protein classes such as membrane and multidomain proteins must be over-expressed in either mammalian or insect cell cultures. Compared to bacterial over-expression, utilizing mammalian or insect cell cultures is more expensive and time-consuming, and is best done with the help of scientists with specific expertise in construct design, cell type, and purification from complex systems.
Large-scale expression and purification
With an expression construct and affinity workflow validated, scale-up can begin. Compared to crystallography, cryo-EM requires considerably less protein, and the nature of the scale-up will be informed by the small-scale expressions and literature precedents. Many times, a small molecule ligand, co-factor, or co-expressed protein can stabilize the proteins to inhibit misfolding or aggregation.
The goals of any protein purification scale-up for structural biology are twofold: to obtain 1) pure and 2) monodisperse protein in large enough amount for structural workflows. Protein purification can take many different paths, but for structural work the final step in any purification should be size-exclusion chromatography (SEC). Unfolded, misfolded, and unstable proteins can oligomerize or aggregate. SEC can separate aggregates, desired, and undesired oligomers from within mixtures to yield a pure and monodisperse sample suitable for cryo-EM.
Higher purity is always preferred for any method in structural biology, but an advantage of cryo-EM over crystallography is a greater tolerance to sample contaminants. For particularly low expressing proteins where multiple rounds of chromatography would substantially reduce the yield below the quantities required for cryo-EM, it can be possible to use less pure (i.e., >80%) samples and overcome the limitations with longer data collections and careful particle classifications.
Biophysical characterization and optimization
Once the protein has been obtained in a pure and monodisperse form, the sample may appear ready for cryo-EM. However, there are also additional biophysical tools that can help characterize the protein in ways that can specifically impact how cryo-EM is performed.
SEC is a good preparative method for obtaining large quantities of monodisperse protein. However, the method is relatively low-resolution, so analytical-scale SEC is often coupled with multi-angle light scattering (MALS) to more accurately determine molecular weight and particle size in solution. Differential scanning fluorimetry (DSF) can be helpful to identify additives or ligands that stabilize the protein of interest. Results from these methods, along with post-translational modification analysis using mass spectrometry, can help the scientists at NanoImaging Services assess the likelihood of eventually obtaining a high-resolution cryo-EM structure.
Measurement of the affinity of the target of interest for specific ligands or compounds via a variety of biophysical methods including surface plasmon resonance (SPR), biolayer interferometry (BLI), and DSF can also greatly help in the design of appropriate experiments for high resolution structure determination. These measurements can inform, for example, with respect to the sample solubility, stability, and propensity to oligomerize, and provide a further tool for the selection of the sample best suited for structural work.
Reducing timelines for more efficient drug discovery
Structural biology adheres to the concept of “garbage in, garbage out” and cryo-EM is no exception. The rate-limiting step to obtaining a good cryo-EM structure is often generating structure-grade, well-characterized protein. NanoImaging Services has a protein scientist with over 15 years of experience in designing gene-to-protein-to-structure workflows for both academic and industry targets. Dr. Tara Davis is available for consultation on best practices, generation of target-specific proposals that can be presented to protein production CROs for evaluation and quote generation, and real-time troubleshooting of protein expression, production, and purification data as it is received from CROs or from internal protein purification resources. Together, Dr. Davis and the scientists at NanoImaging Services will work with your team throughout the process of obtaining pure and monodisperse protein for cryo-EM enabled drug design, helping to overcome the protein bottleneck, and increase your chances of successful structure determination.
- (1) https://www.pnas.org/content/118/41/e2115001118
- (2) Zhang, K., Zheludev, I.N., Hagey, R.J. et al. Cryo-EM and antisense targeting of the 28-kDa frameshift stimulation element from the SARS-CoV-2 RNA genome. Nat Struct Mol Biol28, 747–754 (2021). https://doi.org/10.1038/s41594-021-00653-y
- (3) https://www.sciencedirect.com/science/article/abs/pii/S0960894X20306351?via%3Dihub