Single-Particle cryo-EM as a Pipeline for Obtaining Atomic Resolution Structures of Druggable Targets in Preclinical Structure-Based Drug Design
- 533 Downloads
Single-particle cryo-electron microscopy (cryo-EM) and three-dimensional (3D) image processing have gained importance in the last few years to obtain atomic structures of drug targets. Obtaining atomic-resolution 3D structure better than ~2.5 Å is a standard approach in pharma companies to design and optimize therapeutic compounds against drug targets like proteins. Protein crystallography is the main technique in solving the structures of drug targets at atomic resolution. However, this technique requires protein crystals which in turn is a major bottleneck. It was not possible to obtain the structure of proteins better than 2.5 Å resolution by any other methods apart from protein crystallography until 2015. Recent advances in single-particle cryo-EM and 3D image processing have led to a resolution revolution in the field of structural biology that has led to high-resolution protein structures, thus breaking the cryo-EM resolution barriers to facilitate drug discovery. There are 24 structures solved by single-particle cryo-EM with resolution 2.5 Å or better in the EMDataBank (EMDB) till date. Among these, five cryo-EM 3D reconstructions of proteins in the EMDB have their associated coordinates deposited in Protein Data Bank (PDB), with bound inhibitor/ligand. Thus, for the first time, single-particle cryo-EM was included in the structure-based drug design (SBDD) pipeline for solving protein structures independently or where crystallography has failed to crystallize the protein. Further, this technique can be complementary and supplementary to protein crystallography field in solving 3D structures. Thus, single-particle cryo-EM can become a standard approach in pharmaceutical industry in the design, validation, and optimization of therapeutic compounds targeting therapeutically important protein molecules during preclinical drug discovery research. The present chapter will describe briefly the history and the principles of single-particle cryo-EM and 3D image processing to obtain atomic-resolution structure of proteins and their complex with their drug targets/ligands.
KeywordsSingle-particle cryo-EM Drug development Pharmacological targets Structural biology High resolution
Contrast Transfer Function
- DDD or DED
Direct Detection Device or Direct Electron Detector
Electron Microscopy Data Bank
Field Emission Gun
Fourier Shell Correlation
Multivariate Statistical Analysis
Protein Data Bank
Principle Component Analysis
Structure-Based Drug Design
Transmission Electron Microscopy
The importance of structural biology in understanding the principles of molecular function of proteins, the workforce of cellular world, underpins its use in health science and pharma industries. Classically, protein crystallography was ruling the world of structure-based drug design (SBDD). This was mainly due to the capability of protein crystallography to solve high (better than 1.8 Å), atomic (better than 1.2 Å), and ultra-high (better than 0.95 Å)-resolution 3D structures, which give information of protein drug molecular interaction at various levels. Particularly, the positions of hydrogen atoms were located in many atomic and ultra-high-resolution protein structures. There were no other methods that could rival the versatility of obtaining 3D atomic-level macromolecular structures with which crystallography could achieve. Of the 131,108 protein structures in PDB (as on June 15, 2018), 90% of structures among them were solved by X-ray crystallography technique and 8% by NMR technique. The remaining 2% of structures by large were solved by electron microscopy, electron crystallography, hybrid, and other methods, which include neutron diffraction, solution scattering, fiber diffraction. Clearly, the PDB data suggests that the protein crystallography technique dominates till date. However, the protein crystallography method comes with a proviso. That is, we need diffractable protein crystals of reasonable 10–100s of micron size, in order to obtain a high-resolution X-ray crystallography protein structure. Also, as the unit cell parameter of the protein crystals increase, the resolution of diffraction data drops as the cube of unit cell parameter . Moreover, many proteins, in particular membrane proteins and fibrous proteins, are recalcitrant to crystallization. An analysis of deposited protein structures in PDB by Kozma and co-workers in 2017  showed that the majority of the solved structures (97.6%) are globular proteins and only ~2.4% of them are membrane protein structures. This is primarily because obtaining good diffraction quality 3D crystals for membrane proteins is challenging. As a result, single-particle cryo-EM has gained popularity nowadays for solving membrane protein structures as well along with globular proteins. Also, in cases where single-particle cryo-EM cannot give high-resolution maps, protein crystallography and cryo-EM can be used as hybrid method to visualize macromolecular assemblies at pseudo-atomic resolution as described in Natesh  and references cited therein.
EMDataBank (EMDB) entries having single-particle cryo-EM 3D reconstruction with bound ligands at 2.5 Å or better resolution and their corresponding PDB codes
EMDB entry ID (deposition date)
EMD-2984 (April 26, 2015)
E. coli beta-galactosidase (0.465 MDa)
Phenylethyl beta-D-thiogalactopyranoside (PETG)
EMD-3295 (January 12, 2016)
Homo sapiens p97/VCP Transitional endoplasmic reticulum ATPase (0.54 MDa)
UPCDC30245 (an allosteric inhibitor of VCP)
EMD-7025 (September 9, 2017)
Leishmania donovani 91s ribosome LSU
EMD-7770 (March 28, 2018)
E. coli beta-galactosidase (0.465 MDa)
EMD-7638 (March 27, 2018)
Enterovirus D68 (virus from Homo sapiens) vp1 (0.0330 MDa), vp3 (0.0272 MDa), vp2 (0.0276 MDa, vp4 (0.00734 MDa)
EMD-7599 (March 20, 2018)
Enterovirus D68 vp1 (0.0329 MDa), vp3 (0.0271 MDa), vp2 (0.0276 MDa), vp4 (0.00734 MDa)
No bound inhibitor
EMD-8194 (May 17, 2016)
Bos taurus Glutamate dehydrogenase (0.334 MDa, 0.0616 MDa)
No bound inhibitor
EMD-8762 (June 8, 2017)
Human rhinovirus B14 C5 antibody variable heavy domain (0.0120 MDa), C5 antibody variable light domain (0.0109 MDa), vp1 (0.0326 MDa), vp3 (0.0262 MDa), vp2 (0.0285 MDa), vp4 (0.00718 MDa)
No bound inhibitor
EMD-9012 (July 31, 2018)
Adeno-associated virus - 2 (3.9 MDa), empty virus from Homo sapiens VP1 (0.0820 MDa)
No bound inhibitor
2 The Single-Particle Cryo-EM at High Resolution
The single-particle cryo-EM method for high-resolution structure determination of proteins and protein complexes involves four major steps, viz. (i) the sample preparation, (ii) specimen preparation, (iii) data collection, and (iv) image processing and 3D reconstruction (i.e., structure determination, which includes model building and refinement of the protein/ligand coordinates in the EM map). Sample preparation involves protein purification either from the source or expressed recombinantly in a heterologous host system. The amount of sample required for cryo-EM is very less (~1 µM) in comparison with protein crystallography or NMR spectroscopy techniques, where typically ~200 µM sample is required.
For single-particle electron microscopy (EM), there are two main ways of specimen preparation: (a) negative stain specimen preparation and (b) solution-state “vitrification” for cryo-EM. The former is used for quick characterization of macromolecules and their complexes. However, this type of specimen preparation involves inherent drawbacks (e.g., artifacts and visualizing stain rather than actual protein), which limits the resolution of EM reconstruction map from 30 to 20 Å at its best. Single-particle cryo-EM, the focus of this chapter, on the other hand is synonymous to solution-state structure, and the specimen preparation does not induce artifacts over the protein sample being studied. The vitrified specimen preserves the resolution of the protein structure that is being studied.
Single-particle cryo-EM technique has the capability to solve protein structures to better than 4 Å resolution nowadays. It is to be noted that, there is a consensus in the EM community that better then 4 Å depicts high-resolution structures, while, in the X-ray crystallography community, high resolution corresponds to better than 1.8 Å resolution, as described in the beginning of this chapter. Prior to the resolution revolution in the year 2015, most of the cryo-EM structures with resolution 4 Å or better were virus structures [20, 21, 22]. This was possible due to their large size and high symmetry (e.g., icosahedron symmetry). Most of these data were collected on photographic film (KODAK SO-163 FILM). However, the asymmetric particles (i.e., particles without higher-order symmetry) were limited to sub-nanometer (around 6–10 Å) resolution. Only 1/10th of the total number of structures in EMDB were with resolution 4 Å or better before the resolution revolution. This has significantly increased to 1/6th of the total number of single-particle cryo-EM structures in EMDB as on July 29, 2018, clearly indicating that, currently, there are more structures solved with resolution better than 4 Å in the database. These were possible due to the advancement in the hardware and software and the way the projection images are captured and processed during cryo-EM data collection and processing. Main steps involved in single-particle cryo-EM for obtaining high-resolution protein structure are presented in three subsections. First, we will begin with the details of the specimen preparation in Sect. 2.1, followed by data collection in Sect. 2.2, and finally image processing and 3D reconstruction in Sect. 2.3, respectively.
2.1 Specimen Preparation for Single-Particle Cryo-EM
2.2 Data Collection
Once the grid is on the stage of the TEM, the data is collected on a highly sensitive direct detection device (DDD) also called as direct electron detector (DED) under low electron dose (typically <15 e−/Å2). Low electron dose is necessary since high dose (>15–20 e−/Å2) will cause radiation damage. However, high dose ~1000 e−/Å2 is required for atomic-resolution reconstruction . This problem can be overcome by averaging similar looking particles as described in image processing Sect. 2.3 below. DDD is more sensitive (technically, this feature is called improved detective quantum efficiency (DQE)) and can detect lower doses more effectively with low noise as compared to the conventional photographic film or the CCD (charged coupled device) detectors. Data collection at the focus gives the best resolution, but however the phase contrast is lost in the image (i.e., you cannot clearly visualize the particles). In order to visualize the particles, the images are captured at a defocus that restores the phase contrast in the image, which enables us to visualize particles. Hence, data is collected at a range of defocus between ~4 µm (lower resolution) and ~1 µm (higher resolution). Modern-day advancements in hardware have led to the use of phase plates and energy filters that can restore contrast in the images collected closer to focus. Thus, preserving high-resolution information in the images and at the same time preserving the image phase/amplitude contract as a result alleviate the need for contrast transfer function (CTF) modulation correction at image processing stage.
For high-resolution structure determination, the data is collected on DED as movie frames, which is actually a dose fractionated image stack. The movie frames collected can be corrected for loss of resolution due to stage drifts, charging, and beam-induced motion. The individual movie frames or subset of movie frames in batches are then aligned with respect to each other in order to restore the high-resolution information . Relatively, high exposures up to 20 e−/Å2 can be used for movie mode while DEDs can also be used in electron counting mode where dose rate must be kept below 10 e−/pixel/s [26, 27]. Movie corrections are applied immediately on the micrographs after the data collection using programs like MotionCor2 , optical flow algorithm as implemented in Xmipp [29, 30], Unblur/Summovie [31, 32]. In addition, improved stability of specimen can be provided by the use of grids with graphene and gold support [25, 33, 34]. Hence, in the last six years there has been many breakthroughs in detector, imaging, and image processing technology that has led to high-resolution data collection for even smaller proteins like hemoglobin with mass 64 kDa using Volta phase plate (VPP) , thus leading to resolution revolution with structures determination to better than 2.5 Å. Another aspect of data collection is the automation. Not all proteins give homogenous samples for atomic-resolution reconstruction. The fact that proteins are dynamic leads to heterogeneity and underlies the need for large amount of data collection (in a hope to group particles into homogenous groups), which is tedious to be done manually. In recent years, many software packages have been developed to interface with the advanced electron microscopes for automatic data acquisition. Some examples of such software that can be used for fully automated data collection on a well-calibrated cryo-TEM are Leginon , SerialEM , UCSFImage4 , FEI-EPU, JEOL-JADAS , GATAN-Latitude S. Most of the software is used for automated data collection for both single-particle cryo-EM and electron tomography (ET) work. Some programs like Appion  extend the automated data collection through a pipeline from automated data collection all the way through automated particle picking to image processing (CTF estimation, classification, and 3D reconstruction).
2.3 Image Processing and Three-Dimensional Reconstruction
After CTF correction, the images are normalized to set the mean density of the particles to zero and same standard deviation . The particles are then manually or auto-picked into boxes of 1.5–2.5x, the size of the largest axis of the particle using suitable software. A guide for choosing the right box size is given at the online documentation http://blake.bcm.edu/emanwiki/EMAN2/BoxSize. Number of softwares are available for manual and automatic picking of particles and subsequent image processing. Examples of such programs are FindEM  (only for automated particle picking), EMAN (e2boxer.py) [49, 50], IMAGIC , Ximdisp  (only for interactive display, analyses, and particle picking; now a part of CCP-EM package [53, 54]), Xmipp , RELION-autopick , cryoSPARC , APPLE Picker  (completely automatic particle picking, a part of ASPIRE Suite ), gEMpicker  (only for template-based particle picking), SIGNATURE  (only for particle picking and data analysis), etc. Most of the auto-picking software employ initial manual picking routine (except APPLE picker), where a couple of thousands of particles are manually picked from a subset of available micrographs and use the best class averages generated from them (having as many different representative orientations) as templates to auto-pick particles from rest of the micrographs. This is the preferred method. Alternatively, the auto-picking programs can use low-pass-filtered EM maps as templates for particle picking (less preferred, but useful in protein drug complex where you have the apo-protein structure already). Using maps from PDB (Protein Data Bank) coordinates, as reference model is not preferred at this stage in order to avoid “Einstein-from-noise” effect , i.e., to avoid any 2D model bias. CTF corrections can also be performed on picked particle images as compared to whole micrographs in some software, e.g., EMAN .
After particle picking, the next stage is to get the 3D reconstruction of the biological macromolecules using the different but identifiable 2D projections of particles. The first 3D reconstruction from a 2D projection was carried out on negative stained tail of bacteriophage T4 by De Rosier and Klug . However, the 2D projections of particle images cutout from the motion-corrected micrographs have still low signal-to-noise ratio (SNR) due to low electron dose data collection as described in data collection section. Hence, in order to improve the SNR of the particles, many identical looking particle images are aligned and summed (clustering) thus effectively increasing the SNR and dose without increasing the damage . There are three main advantages of reference-free (unsupervised) 2D classification: (i) to select few 2D classes from which we can make starting 3D map, which can be projected as references for refinement. (ii) We can identify the fraction of bad classes (which may contain artifacts, invalid particles, or simply empty), and thus, those images with anomalies can be deleted from the data set in the beginning itself. (iii) It also helps in identifying the conformational and compositional variability in the data set . Two-dimensional (2D) and 3D classifications are carried out by using various statistical analysis software suite IMAGIC , Spider , EMAN , RELION-3 , FREALIGN , Appion , cryoSPARC , ASPIRE Suite , Xmipp , SPHIRE (sphire.mpg.de), etc., or a combination of more than one of these suites. Several of these software packages are integrated into one processing framework, for example, as in Scipion . An exhaustive list of EM software programs is available at EMDataBank (EMDB, http://www.emdatabank.org/emsoftware.html).
Spider [63, 67] and IMAGIC  were among the first programs to be developed for single-particle reconstruction in the year 1996 followed by FREALIGN  in the year 1998 and other program suites followed. The clustering of similar particle images was first introduced by van Heel and Frank  in the year 1981 using multivariate statistical analysis. Clustering in the currently available programs uses one of the following methods: multivariate statistical analysis (MSA)/principle component analysis (PCA), hierarchical clustering, k-means clustering, and the maximum-likelihood methods  or by recently proposed empirical Bayesian approach . Currently, the EMAN2  and RELION-3  are among the popular program that do reference-free 2D class-averaging (references are generated from within the data set) and 3D reconstruction. EMAN2 uses iterative MSA-based reference-free 2D classification. The latest one, the RELION, uses empirical Bayesian likelihood approach for 2D classification .
Next step is to get the 3D reconstruction from selected good class averages. High-resolution 3D reconstructions require an initial 3D model that can be iteratively refined to obtain the best possible resolution for the data set. The first starting 3D model is obtained using experimental methods or by finding the relative orientations of 2D projection averages (and hence the particles) by computational methods. Assigning orientations by programs involves finding the location and Euler angles of the particles in the boxed region. The earliest one among them was the popular angular reconstitution method  by Marin van Heel, which uses real-space implementation of “common lines” principle to get relative orientations of the class averages as implemented in the program IMAGIC . Thus, the Euler angles assigned 2D class averages can be used to get the starting 3D model. This method does not require reference for assigning relative orientation, while another program Spider by Joachim Frank and co-workers uses projection matching and cross-correlation approach [63, 71]. This method requires a starting 3D model which is generated from ab initio random conical tilt method  from EM images taken at a pair of know angles. Most of the present-day programs generate the starting 3D model by using statistical approach and comparison with back-projections to assign the Euler angles to a subset of manually selected good class averages. For example, EMAN2 uses a Monte Carlo method, RELION uses Bayesian methods, and VIPER  a module in SPHIRE suite (http://sphire.mpg.de/) uses a stochastic hill-climbing algorithm. Iterative rounds of projection matching with the references generated from starting 3D model (called as 3D projection matching procedure) followed by subsequent 3D reconstruction (using various algorithms) are used until the resolution of the reconstruction during subsequent refinement cycles does not further improve. This will lead to the final 3D reconstruction with the best possible resolution.
3 Resolution, Model Building, and Validation
Resolution estimation of the EM maps is still subjective, with differences among various groups still not settled . Resolution of 3D EM map is calculated from a plot of Fourier shell correlation (FSC)  as a function of spatial frequency (the resolution estimation of 3D reconstructions in Fig. 4 is shown in Fig. 5). FSC is the cross-correlation (CC) calculated between two 3D reconstruction maps, where each map is calculated from half the data images. The resolution that is reported in publication essentially as a single number is the value of maximum spatial frequency up to which the EM map is reliable. The identification of resolution is subjective as it is arbitrary what one considers as reliable. The procedure for resolution assessment is described in detail by Penczek . There are several suggestions for identifying the cutoff: (i) the 3-sigma criteria where the spectral SNR (SSNR) = 0 in which case FSC = 0; (ii) point at which power of signal is equal to the power of noise, i.e., SSNR = 1 or FSC = 0.33; (iii) the classic midpoint of FSC curve, i.e., FSC = 0.5  where SSNR = 2, which means signal dominates noise; and finally (iv) point where FSC = 0.143, derived by Rosenthal and Henderson  in comparison with X-ray crystallography. Hence, which cutoff is chosen is a matter of present-day debate. Recently, in order to reduce further any possible reference bias, “gold-standard FSC” was suggested with FSC calculated between two completely independent refinements and 3D reconstruction .
There are other computational ways to improve the resolution nominally without improving the image alignment, e.g., masking/threshold flattening. In any case, the resolution estimations have their own limitations and hence reported EM resolution should be treated as only broad guideline rather than a definitive number and cannot be used as validation. Nonetheless, it is an important parameter to be reported with each EM map deposition at the EMDB. Resolution anisotropy is common in cryo-EM structures, and it is a common practice to document it as color ramping from low to high resolution on the cryo-EM 3D reconstruction map using programs ResMap  and blocres . The results can be visualized independently or with chimera (e.g., blocres with Local FSC plug-in for chimera).
With the booming medium- and high-resolution cryo-EM 3D structures, it is necessary to have consistency between crystallography and cryo-EM terms currently used for defining what is an atomic- or high-resolution structure. While it is very common to use the term “atomic resolution” for cryo-EM resolutions better than 3.5 Å, the crystallography definition of the term “atomic resolution” means the resolution is 1.2 Å or better  and ultra-high resolution means 0.95 Å or better ( and references cited therein). Similarly, 1.8 Å or better is called high resolution , 3.0 Å or better up to 1.9 Å is treated as medium resolution while low resolution is between 4 and 3.1 Å. Resolution below 4 Å is considered as poor resolution in protein crystallography. While the method of estimation of resolution is quite different between crystallography and cryo-EM techniques, the conventions for using the terms should be consistent, irrespective of the method. Hence, the author would like to suggest that it is necessary for the cryo-EM field to maintain consistency in the future, while using the terms ultra-high, atomic, high, medium, and low resolution.
3.2 Model Building
If the resolution of the 3D reconstruction (i.e., the electron potential map) is sufficiently high, e.g., better than 3 or 4 Å, it is often possible to build ab initio atomic model and do refinement with the EM map using known chemical constraints/restraints. If X-ray crystallography coordinates of the segment or its homologues are available, one can rigid body fit the segment coordinates into the cryo-EM map using programs like UCSF Chimera . Where the resolution of the 3D reconstruction map is limited to worse than 4 Å, combining crystallography and cryo-EM as a hybrid method is a powerful tool to obtain a pseudo-atomic model(s). Iterative rounds of model building using programs like Coot , O  and refinement using programs like Coot, refmac , or PHENIX real-space-refinement  are carried out. De novo backbone tracing and model building can be carried out using programs like Pathwalking and Gorgon ; it can also build macromolecular assemblies at non-atomic resolution . When the cryo-EM map shows variation by domain movements or flexibility to the available protein coordinates, programs like FlexEM [91, 92] and MDFF  with its graphical user interface VMD  can be used to flexibly fit the coordinates in the EM map. The fitted model and EM map can be visualized in programs like PyMOL /Chimera  to generate publication quality figures.
Validation in cryo-EM reconstruction is important to avoid errors in particle alignment, reference bias, over-fitting of atomic coordinates, and over-estimation of resolution. Validation tools for cryo-EM similar to the free R value (Rfree) in X-ray crystallography  have been introduced in 2003 by Joachim and co-workers  and recently by Chen and co-workers . There are some general guidelines [80, 99] suggested, and they are actively evolving. As described in the resolution paragraph of this section, the reporting of resolution with one number cannot be used as validation; however, it is an important parameter to be reported during EMDB deposition. Further, FSC may fail when the particles are significantly misaligned. So one has to estimate the resolution properly  and use the reported single number resolution with caution. It is suggested that gold-standard FSC provides a realistic estimate of the true signal , and this will lead ultimately to a better map. In recent days, reporting local resolution has also formed a common practice in publication and thesis [81, 82]. Also, the local resolution will be helpful in avoiding over-interpretation of poor regions in the cryo-EM map. If the 3D map is of sufficient resolution (better than 4 Å), it can resolve the secondary structural features. A good validation would be especially if you can see a right-handed alpha helix or even the side chain residues, especially the bulky residues like tryptophan, phenylalanine, or tyrosine in the high-resolution cryo-EM map. Even the comparison of the new EM structure with the available EM structure will be one way of validating the newly reconstructed 3D EM map . Further, programs TEMPy  and refmac  can be used to assess the validity of the fitted coordinates to the EM map. In Coot  program, one can use the Ramachandran plot (Validate →Ramachandran plot) and geometrical quality (Validate →Geometry analysis) to validate the quality of the refined model. One more way to validate is to compare the 3D reconstruction results from different techniques, e.g., projection matching and the angular reconstitution. For low-resolution maps (worse than 4 or 10 Å), measure of confidence can be provided by a priori random conical tilt experiments .
Though cryo-EM can handle heterogeneous particles, we need homogenous particles, which are equally dispersed in the vitrified ice in order to achieve atomic resolution. Ideally speaking, all data sets are heterogeneous! The question is how much one is willing to tolerate . Further, during cryo-EM specimen preparation, non-physiological structural heterogeneity is often introduced . While structural heterogeneity is a problem to obtain high resolution, it also provides a unique opportunity to study the conformational flexibility/dynamics of the macromolecular assemblies. Ideally, homogenous samples have to be biochemically standardized and prepared before the vitrification process. However, this is not possible with all protein samples due to the inherent protein flexibility which is necessary for its function, for example, rotation of 30S subunit of ribosome  or rotational states in case of eukaryotic V-ATPase . In such cases, the heterogeneous sample data images can be classified computationally to classes containing homogenous particles (an example of such classification can be seen in Fig. 4).
Three main techniques are currently in use to identify and sort the macromolecular structural conformational variability or heterogeneity . The first approach depends on classifying the 2D images based on the eigen images/eigenvectors [74, 108, 109, 110, 111] without any starting model. First, classify using MSA to obtain orientation classes, and then, the major variation among the picked particles in each orientation classes can be identified in the low-order eigen images by MSA, and using these, information particles can be classified into homogenous classes, leading to preliminary 3D reconstruction from a class containing majority of homogenous particles as shown in Fig. 3. The preliminary 3D reconstructions can be projected as references for competitive alignment. Further, the quality of 3D reconstruction can be iteratively improved until the eigen images show no major variations within the class and the particles stabilize from jumping to another class during competitive projection matching. In this manner, three class reconstructions were obtained as shown in Fig. 4. The second method depends on detection in 2D variations using starting model . The third method also needs initial starting model and uses a statistical approach to obtain 3D classification. In this case, large number of 3D maps are calculated from randomly selected subset of particles (with previously assigned orientation based on initial 3D map). Determination of the 3D variance can be used to assess the heterogeneity, and estimation of covariance enables one to carry out the 3D classification according to variable regions. Alternatively, the molecular states can be separated using maximum likelihood classification [104, 112] or by the latest multi-body refinement method .
5 Single-Particle Cryo-EM Applications in SBDD
6 Conclusions and Future Prospective
Recent advances in cryo-EM have enabled us to use single-particle cryo-EM as a method of choice to resolve solution-state 3D structures of proteins and protein complexes at atomic resolution, thus breaking the cryo-EM resolution barrier to facilitate SBDD . In recent years, many pharmaceutical companies like Bayer, Merck Research Laboratories, Sonafi, AstraZeneca, Regeneron Pharmaceuticals, NovAliX, Genentech etc. have realized the importance of this method and started hiring experts in single-particle cryo-EM to get involved in their SBDD pipeline. Table 1 lists the protein structures with bound ligands solved by single particle cryo-EM at resolution 2.5 Å or better; i.e., five of the structures have bound inhibitors/glycans, which underscore the importance of single-particle cryo-EM in SBDD. Apart from these, there are many more structures with bound ligands in the EMDB at resolutions below 2.5 Å. The future of this technique will be in obtaining the sub-nanometer resolution and perhaps atomic-resolution structures of proteins and protein complexes in vivo. This methodology called the cellular tomography, although not the scope of this chapter, is a promising future technology for atomic-resolution structures of proteins and protein complexes in its native environment “the cell.” With the advent of phase plates, energy filters, and automation in cryo-EM data collection, promising efforts are being made to achieve that goal and the realization of that goal may not be far away which would in turn potentially further accelerate the SBDD program.
RN was supported by Ramalingaswamy Fellowship from DBT. RN would like to thank his laboratory members and colleagues for their constant support, valuable scientific and technical discussion. Last but not least, RN would like to thank IISER-TVM past Director Prof. ED Jemmis and present Director Prof. V. Ramakrishnan for their unstinted support.
- 9.Danev R, Tegunov D, Baumeister W (2017) Using the Volta phase plate with defocus for cryo-EM single particle analysis. Elife 6:1–9Google Scholar
- 10.Banerjee S, Bartesaghi A, Merk A, Rao P, Bulfer SL, Yan Y, Green N, Mroczkowski B, Neitz RJ, Wipf P, Falconieri V, Deshaies RJ, Milne JL, Huryn D, Arkin M, Subramaniam S (2016) A resolution cryo-EM structure of human p97 and mechanism of allosteric inhibition. Science 351:871–875PubMedCrossRefGoogle Scholar
- 58.Singer A et al. (2010) Algorithms for Single Particle Reconstruction (ASPIRE), http://spr.math.princeton.edu/
- 64.Zivanov J, Nakane T, Scheres S (2018) A Bayesian approach to beam-induced motion correction in cryo-EM single-particle analysis, bioRxivGoogle Scholar
- 66.de la Rosa-Trevin JM, Quintana A, Del Cano L, Zaldivar A, Foche I, Gutierrez J, Gomez-Blanco J, Burguet-Castell J, Cuenca-Alba J, Abrishami V, Vargas J, Oton J, Sharov G, Vilas JL, Navas J, Conesa P, Kazemi M, Marabini R, Sorzano CO, Carazo JM (2016) Scipion: a software framework toward integration, reproducibility and validation in 3D electron microscopy. J Struct Biol 195:93–99PubMedCrossRefGoogle Scholar
- 73.Penczek PA (2014).http://sparx-em.org/sparxwiki/sxviper
- 75.Bartesaghi A, Aguerrebere C, Falconieri V, Banerjee S, Earl LA, Zhu X, Grigorieff N, Milne JLS, Sapiro G, Wu X, Subramaniam S (2018) Atomic resolution cryo-EM structure of beta-galactosidase. Structure 26(848–856):e3Google Scholar
- 77.Harauz G, van Heel M (1986) Exact filters for general geometry three-dimensional reconstruction. Optik 73:146–156Google Scholar
- 80.Henderson R, Sali A, Baker ML, Carragher B, Devkota B, Downing KH, Egelman EH, Feng Z, Frank J, Grigorieff N, Jiang W, Ludtke SJ, Medalia O, Penczek PA, Rosenthal PB, Rossmann MG, Schmid MF, Schroder GF, Steven AC, Stokes DL, Westbrook JD, Wriggers W, Yang H, Young J, Berman HM, Chiu W, Kleywegt GJ, Lawson CL (2012) Outcome of the first electron microscopy validation task force meeting. Structure 20:205–214PubMedCrossRefGoogle Scholar
- 83.Natesh R, Manikandan K, Bhanumoorthy P, Viswamitra MA, Ramakumar S (2003) Thermostable xylanase from Thermoascus aurantiacus at ultrahigh resolution (0.89 A) at 100 K and atomic resolution (1.11 A) at 293 K refined anisotropically to small-molecule accuracy. Acta Crystallogr D Biol Crystallogr 59:105–117PubMedCrossRefGoogle Scholar
- 94.Humphrey W, Dalke A, Schulten K (1996) VMD: visual molecular dynamics. J Mol Graph 14(33–8):27–28Google Scholar
- 95.Schrodinger LLC (2015) The PyMOL molecular graphics system, Version 1.8 inGoogle Scholar
- 98.Chen S, McMullan G, Faruqi AR, Murshudov GN, Short JM, Scheres SH, Henderson R (2013) High-resolution noise substitution to measure overfitting and validate resolution in 3D structure determination by single particle electron cryomicroscopy. Ultramicroscopy 135:24–35PubMedPubMedCentralCrossRefGoogle Scholar
- 103.Henderson R, Chen S, Chen JZ, Grigorieff N, Passmore LA, Ciccarelli L, Rubinstein JL, Crowther RA, Stewart PL, Rosenthal PB (2011) Tilt-pair analysis of images from a range of different specimens in single-particle electron cryomicroscopy. J Mol Biol 413:1028–1046PubMedPubMedCentralCrossRefGoogle Scholar