Modeling Ligands into Maps Derived from Electron Cryomicroscopy

Summary

This protocol introduces the tools available for modeling small-molecule ligands in cryoEM maps of macromolecules.

Abstract

Deciphering the protein-ligand interactions in a macromolecular complex is crucial for understanding the molecular mechanism, underlying biological processes, and drug development. In recent years, cryogenic sample electron microscopy (cryoEM) has emerged as a powerful technique to determine the structures of macromolecules and to investigate the mode of ligand binding at near-atomic resolution. Identifying and modeling non-protein molecules in cryoEM maps is often challenging due to anisotropic resolution across the molecule of interest and inherent noise in the data. In this article, the readers are introduced to various software and methods currently used for ligand identification, model building, and refinement of atomic coordinates using selected macromolecules. One of the simplest ways to identify the presence of a ligand, as illustrated with the enolase enzyme, is to subtract the two maps obtained with and without the ligand. The extra density of the ligand is likely to stand out in the difference map even at a higher threshold. There are instances, as shown in the case of metabotropic Glutamate receptor mGlu₅, when such simple difference maps cannot be generated. The recently introduced method of deriving the Fo-Fc omit map can serve as a tool for validating and demonstrating the presence of the ligand. Finally, using the well-studied β-galactosidase as an example, the effect of resolution on modeling the ligands and solvent molecules in cryoEM maps is analyzed, and an outlook on how cryoEM can be used in drug discovery is presented.

Introduction

Cells accomplish their functions by carrying out innumerable chemical reactions simultaneously and independently, each meticulously regulated to ensure their survival and adaptability in response to environmental cues. This is achieved by molecular recognition, which enables biomolecules, especially proteins, to form transient or stable complexes with other macromolecules as well as small molecules or ligands¹. Thus, protein-ligand interactions are fundamental to all processes in biology, which include the regulation of protein expression and activity, the recognition of substrates and cofactors by enzymes, as well as how cells perceive and relay signals^1,2. A better understanding of the kinetic, thermodynamic, and structural properties of the protein-ligand complex reveals the molecular basis of ligand interaction and also facilitates rational drug design by optimizing drug interaction and specificity. An economical and faster approach to studying protein-ligand interaction is to use molecular docking, which is a computational method that virtually screens a diverse range of small molecules and predicts the binding mode and affinity of these ligands to target proteins³. However, experimental evidence from high-resolution structures determined by X-ray Diffraction (XRD), Nuclear Magnetic Resonance (NMR), or electron cryomicroscopy (cryoEM) provides the essential proof for such predictions and aids in the development of newer and more effective activators or inhibitors for a given target. This article uses the abbreviation 'cryoEM' as the technique is commonly referred to. However, there is an ongoing debate on choosing the correct nomenclature, and recently, the term cryogenic-sample Electron Microscopy (cryoEM) has been proposed to indicate that the sample is at cryogenic temperature and imaged with electrons⁴. Similarly, the maps derived from cryoEM have been called electron potential, electrostatic potential, or Coulomb potential, and for simplicity, here we use cryoEM maps^5,6,7,8,9,10.

Although XRD has been the gold standard technique in high-resolution structure determination of protein-ligand complexes, post resolution-revolution¹¹ cryoEM has gained momentum, as indicated by the surge of Coulomb potential maps or cryoEM maps deposited in the Electron Microscopy Database (EMDB)^12,13 over the last few years¹⁴. Owing to advancements in sample preparation, imaging, and data processing methods, the number of Protein Data Bank (PDB)¹⁴ depositions employing cryoEM increased from 0.7% to 17% between 2010 and 2020, with approximately 50% of reported structures in 2020 being determined at 3.5 Å resolution or better^15,16. CryoEM has been rapidly adopted by the structural biology community, including the pharmaceutical industry, as it allows the study of flexible and non-crystalline biological macromolecules, especially membrane proteins and multi-protein complexes, at near-atomic resolution, overcoming the process of crystallization and obtaining well-diffracting crystals required for high-resolution structure determination by XRD.

Accurate modeling of the ligand in the cryoEM map is paramount, as it serves as a blueprint of the protein-ligand complex at a molecular level. There are several automated ligand-building tools used in X-ray crystallography that depend on the shape and topology of the ligand density in order to fit or build the ligand into the electron density^17,18,19,20. Nevertheless, if the resolution is lower than 3 Å, these approaches tend to produce less desirable outcomes because the topological features on which they depend for recognition and building become less defined. In many instances, these methods have proven ineffective in accurately modeling ligands into cryoEM maps, as these maps have been determined in the low-to-medium resolution range, typically between 3.5 Å-5 Ź⁷.

The first step in 3D structure determination of a protein-ligand complex by cryoEM involves either co-purifying the ligand with the protein (when the ligand has a high binding affinity to the protein) or incubating the protein solution with the ligand for a specific duration before grid preparation. Subsequently, a small sample volume is placed onto a plasma-cleaned holey TEM grid, followed by flash freezing in liquid ethane and ultimately imaging with a cryo-TEM. The 2D projection images from hundreds of thousands to millions of individual particles are averaged to reconstruct a 3-dimensional (3D) Coulomb potential map of the macromolecule. Identifying and modeling ligands and solvent molecules in these maps pose significant challenges in many cases due to the anisotropic resolution across the map (i.e., the resolution is not uniform across the macromolecule), flexibility in the region where the ligand is bound, and the noise in the data. Many of the modeling, refinement, and visualization tools that were developed for XRD are now being adapted for use in cryoEM for the same purposes^{18, 19, 20, 21}. In this article, an overview of various methods and software currently used to identify ligands, build models, and refine the coordinates derived from cryoEM is presented. A step-by-step protocol has been provided to illustrate the processes involved in modeling ligands using specific protein-ligand complexes with varying resolution and complexity.

The first step in modeling ligands in cryoEM maps includes the identification of the ligand density (non-protein) in the map. If ligand binding does not induce any conformational change in the protein, then calculating a simple difference map between the protein-ligand complex and the apo-protein essentially highlights the regions of extra density, suggesting the presence of the ligand. Such differences can be observed immediately, as it just requires two maps, and even intermediate maps during the process of 3D refinement can be used to check if the ligand is present. Additionally, if the resolution is high enough (<3.0 Å), then the difference map can also provide insights into the location of water molecules as well as ions interacting with the ligand and the protein residues.

In the absence of the apo-protein map, it is now possible to use Servalcat²², which is available as a standalone tool and has also been integrated into the CCP-EM software suite^23,24 as part of the Refmac refinement and in CCP4 8.0 release^25,26. Servalcat allows the calculation of an FSC weighted difference (Fo-Fc) map using the unsharpened half-maps and the apoprotein model as input. The Fo-Fc omit map represents the disparity between the experimental map (Fo) and the map derived from the model (Fc). In the absence of a ligand in the model, a positive density in a Fo-Fc map that overlaps with the experimental EM map typically suggests the presence of the ligand. The assumption here is that the protein chain is well-fitted in the map, and the remaining positive density indicates the location of the ligand. However, it is important to meticulously examine whether the positive density stems from modeling inaccuracies, such as the wrong rotamer of a protein side-chain.

The second step involves obtaining or creating a cartesian coordinate file of the ligand with well-defined geometry from the available chemical information. Standard ligands (for example, ATP and NADP⁺) that are already available in the CCP4 monomer library can be used for refinement by retrieving the coordinate and geometry files via their monomer accession code. However, for unknown or non-standard ligands, various tools are available to create the geometry files. Some of the examples include the eLBOW²⁷ - (electronic ligand builder and optimization workbench) in Phenix²⁸, Lidia - an in-built tool in Coot²⁹, JLigand/ACEDRG^30,31, CCP-EM^23,24, Ligprep³²-a module of Glide within the Schrodinger suite. The ligand coordinate file is then fitted in the density, guided by both the experimental cryoEM map and the difference map in Coot. This is followed by real-space refinement in Phenix²⁸ or reciprocal refinement in Refmac³³. A Linux workstation or a laptop equipped with a good graphics card and the above-mentioned software is required. Most of these programs are included in various suites. CCP-EM²⁴ and Phenix²⁸ are freely available to academic users and include a variety of tools that are used in this article, including Coot, Refmac5^33,34,35,36, Servalcat, phenix.real_space_refine, etc. Similarly, Chimera³⁷, and ChimeraX³⁸ provide free licenses to academic users.

Protocol

1. Modeling phosphoenolpyruvate (PEP) in enolase from Mycobacterium tuberculosis

1. Identification of ligand density in the cryoEM map of PEP-enolase complex
   1. Download the unsharpened half-maps (emd_30988_additional_1.map and emd_30988_additional_2.map) of apo-enolase from the additional data in EMDB (see Table of Materials).
   2. Open ChimeraX (see Table of Materials). Open the half-maps of apo-enolase by clicking on Open from the toolbar and selecting the file names. Type vop add #1 #2 in the command line to combine both the half-maps to obtain the apo-enolase unsharpened map.
      NOTE: #1 and #2 denote the two half maps for the apo-enolase enzyme, as mentioned in step 1.1.1.
   3. Rename the combined map (ChimeraX map ID: #3) by typing rename #3 Apo-enolase-unsharpened-map. Color the map gray from the model panel. The map indicates that enolase is octameric in the solution with D4 symmetry.
   4. Repeat steps 1.1.1-1.1.3 using the following half-maps emd_30989_additional_1.map (map ID: 4) and emd_30989_additional_2.map (map ID:5) to obtain PEP-enolase-unsharpened-map (map ID: #6).
   5. In order to compute a difference map, first fit the two maps (PEP-enolase and apo-enolase unsharpened maps from step 1.1.3 and step 1.1.4) onto each other by clicking on Fit (map in map) in the Map panel (toolbar) in ChimeraX.
      NOTE: Normalization is not performed between the two maps. In some cases, normalization may be required to bring the maps to the same scale.
   6. Subtract the PEP-enolase map from the apo-enolase map by typing vop subtract #6 #3 in the command line.
      NOTE: #6 = PEP-enolase unsharpened map, #3 = Apo-enolase unsharpened map.
   7. Color the subtracted map (map ID:7) in green, denoting positive density (as per the convention in X-ray crystallography)³⁹.
   8. Download the coordinates of M. tuberculosis octameric enolase (PDB: 7e4x) from the Protein Data Bank (PDB), click on open from the toolbar, and select the filename.
   9. Rename the model 7e4x.pdb by typing rename #8 Apo_enolase.pdb in the command line. #8 denotes the Apo_enolase.pdb in ChimeraX.
   10. Select the model from the model panel. Click on Right Mouse from the toolbar. Click on Move molecule to position the model in proximity to the PEP-enolase map (#6) and align the model with respect to the map. Fit this model into the PEP-enolase-unsharpened-map by typing fit #8 in #6 in the command line. Here, the map is numbered 6, and the model is numbered 8.
       NOTE: In some cases, the local fit in ChimeraX may not be enough to align the model. In these cases, an additional step of global fitting (DockinMap, see Table of Materials) may be performed in Phenix.
   11. Save the fitted model by typing save Apo-enolase.pdb #8 in the command line.
2. Modeling and refinement of PEP in the B-factor sharpened cryoEM map of PEP-enolase complex
   1. Open "coot" (see Table of Materials) from the terminal by typing ./Coot &.
   2. Display "Apo-enolase.pdb" by clicking on File > Open Coordinates Apo-enolase.pdb. The coordinate file is displayed by bonds (color by atom).
   3. Download the PEP-bound Enolase sharpened map, i.e., emd_30989.map from EMDB. Display the same by clicking on File > Open Map > emd_30989.map . Set the threshold to 7.00 σ by scrolling the middle mouse button.
   4. Find unmodeled blobs by clicking on Validate > Unmodeled blobs > Find Blobs.
   5. Locate the unmodeled ligand density near the active site residues, Ser 42, Lys-386 and Arg-364 of the enolase model.
   6. Get the PEP monomer model file from the Coot monomer library by clicking on File > Get Monomer and entering PEP as the 3 Letter code.
   7. Move the PEP molecule to the density using the Rotate/Translate- Zone/Chain/Molecule option in the sidebar menu. Use real-space refinement in Coot to fit the PEP molecule in the density by clicking on Real Space Refine Zone in the sidebar menu. Merge the fitted ligand with the Apo-enolase.pdb by using the merge molecule option in the edit tab. Save the model as PEP-Enolase.pdb.
      NOTE: One can also use ligand> Jiggle-Fit Ligand option to fit the ligand as well.
   8. To add ligands to the rest of the monomers in the coordinate file, click on Calculate > NCS Tools > NCS ligands. A separate window named Find NCS-Related Ligands will appear. For the Protein with NCS option, select the coordinate file Apo-enolase.pdb and Chain ID A as NCS Master Chain.
      1. For the molecule containing ligand, select the Apo-enolase.pdb as the coordinate file and Chain ID, J, and Residue Number 1 to 1. Click on Find Candidate Positions. A window named Fitted Ligands will appear with a list of candidate positions. Evaluate the fitting by clicking on the individual candidate ligands and analyze the fit visually.
         NOTE: In Servalcat/Refmac, only a monomer model can be provided, and the symmetry used in reconstruction can be given to obtain an expanded model.
   9. Additional density for the solvent molecules was also observed near the ligand. High-resolution crystal structures of enolase from various homologs suggest the presence of two Mg²⁺ ions^40,41, which probably stabilizes the negative charge of the reaction intermediate. Model two Mg²⁺ ions in the active site density by clicking on place atom at the pointer and selecting MG from the Pointer Atom type list.
      NOTE: The bond geometry and distance may be used as a guide to select the metal ion. In the case of Mg²⁺ bound to oxygen atoms in protein residues, the bond distance varies between 2.1 Å -2.4 Å, with an octahedral geometry. However, in the case of low-resolution maps, the coordination sphere is often incomplete, and the distance may vary as well because of limitations in the resolution.
   10. Repeat step 1.2.8 in order to add the Mg²⁺ ions in the symmetry-related monomers.
   11. Save the model by clicking on File > Save coordinates > Select Filename > and typing Enolase+PEP+Mg.pdb.
   12. Open Phenix GUI (see Table of Materials). Run a real space refinement job with the Enolase+PEP+Mg.pdb and the emd_30989.map as the input model and the map, respectively, using default parameters. This step is done iteratively with Coot to achieve good map-model fit and geometry.
       NOTE: The unsharpened difference map is used to demonstrate the presence of ligand, such as in a figure. For modeling purposes, the B-factor sharpened map has been used and is recommended. The B-factor sharpened map can also be used to calculate the difference map for demonstration purposes as opposed to the unsharpened map. However, if the resolution is lower in the region where the ligand is bound, then the single B-factor used for sharpening the entire map may not clearly reveal the ligand density.
3. Visualization and figure generation of the modeled ligand
   1. Open the refined Enolase+PEP+Mg model from the final Phenix refinement job in PyMOL⁴² (see Table of Materials) by clicking on File > Open and selecting the file name.
   2. Load the emd-30989.map (PEP-bound sharpened map) by clicking on File > Open and selecting the filename.
   3. Rename the map as PEP-sharpened by clicking on actions > rename against the emd_30989 object and typing PEP-sharpened.
      NOTE: In most cases, for example enolase, the maps when opened in pymol are normalized as the normalize map option is checked by default in pymol. As cryoEM maps are centered on a larger box, the normalisation will include everything in the box. Thus, normalisation can be turned off before opening in pymol.
   4. Select the ligands by clicking on display > sequence.
   5. Type isomesh mesh_ligands, PEP-sharpened, 6.0, sele, carve=3.0 in the command line and press enter.
   6. Color the mesh_ligand blue by clicking on the last box, C, next to the mesh_ligand object.
   7. Display the active site residues, Ser-42, Asp-241, Glu-283, Asp-310, Arg-364 and Lys-386 interacting with the PEP and Mg²⁺ in stick and sphere representation, respectively, and the enolase enzyme in cartoon representation.
   8. Ray trace by typing ray 3600, 3600 to generate publication quality images and save as a png file by clicking on File > save and choosing PEP.png as the file name.

2. Modeling of ligands in metabotropic Glutamate receptor mGlu₅

1. Identifying and modeling the ligand density in the cryoEM map of agonist and antagonist-bound mGlu₅
   1. Download the two half-maps along with their corresponding coordinate files for each of the agonist bound (EMD-31536, 7fd8.pdb) and the antagonist bound (EMD-31537,7fd9.pdb) mGlu₅ complexes.
   2. Open ChimeraX
      1. Open the coordinate file, 7fd8.pdb by clicking open and selecting 7fd8.pdb.
      2. Type delete ligand in the command line and press Enter.
      3. Similarly, use the command delete/B to delete the chain B.
         NOTE: Ligands and chains can be deleted by using the drop-down menu at the top using select and actions > atoms/bonds > delete.
      4. Save the unliganded pdb by typing the following in the command line: save 7fd8_noligand_chainA.pdb #1. #1 denotes the model ID.
      5. Repeat the steps 2.1.2.1-2.1.2.4 for 7fd9.pdb.
   3. Open CCP-EM. Create a new project named mGlu₅ and specify the project directory and username.
      1. Open Relion from the options on the left panel.
         1. Go to the Mask Creation tab.
         2. Provide one of the half-maps for EMD-31536 as the input.
            NOTE: Relion accepts maps with .mrc as a suffix, and the EMD maps have .map suffix. The map files can be opened in ChimeraX for examination and saved in the .mrc format.
         3. In the mask tab, provide pixel size as 0.89 Å and initial binarization threshold as 0.007.
         4. Run the job with the alias 31536 (or any other name that is easy to follow).
         5. Similarly, create a mask for the EMD - 31537 half map.
      2. Open Refmac Servalcat program from the options in the left panel in CCPEM.
         1. Provide the name as mGlu₅_agonist.
         2. Import the 7fd8_noligand_chainA .pdb file as described in the model section 2.1.2.4. Provide the location of the two half maps and the corresponding mask created in Relion.
         3. Specify the resolution as 3.8 Å.
         4. Go to refinement settings > Strict Symmetry and mention C2 as the Relion point group symmetry.
         5. Start the program by pressing the start button at the top.
   4. Once the job is finished, open the result in Coot (option available at the top in the result section). This will display a diffmap.mtz in Coot by default, where the colors red and green indicate negative and positive densities, respectively.
      1. Go to Display manager > properties of the difference map labeled DELFWT PHDELWT and change the contour level to 4.0 (absolute value).
         NOTE: The choice of the threshold depends on the map and resolution.
      2. Hide the map labeled FWT PHWT and the molecule labeled refined.pdb.
      3. Move the map around using the mouse to visualize the positive density (colored green) and bring the larger blob to the centre.
      4. Go to File > Get monomer, type QUS, and hit enter.
      5. Go to Calculate > Modeling > Rigid Body Fit molecule and double-click on the QUS monomer to adjust the molecule to fit in the Fo-Fc density.
      6. Use the merge molecule option in the edit tab to add the QUS molecule to the coordinate file, refined.pdb.
      7. Repeat steps 2.1.4.3-2.1.4.6 to fit the ligand with monomer code NAG and CHS in the smaller blob in the extracellular and at the top of the transmembrane domain, respectively.
      8. Save the coordinate as refined_ligands.pdb.
         NOTE: The resolution of the transmembrane (TM) domain is lower, and thus, it is not discussed here.
2. Refinement of the mGlu₅ liganded structure
   1. Open CCPEM and clone the previous refinement job (step 2.1.3.2) and run it with the refined_ligands.pdb and sharpened map (emd. 31536.mrc) as input, using default parameters and C2 symmetry.
      NOTE: If the program doesn't recognize the ligand, then the CIF files need to be provided. These CIF files can be downloaded from the PDB or generated using various programs (see step 3.1.1 for details).
3. Visualization and figure generation in PyMOL
   1. Open the refined_expanded.pdb model from the latest Refmac refinement job in PyMOL.
   2. Go to File > Open and browse to the Refmac job directory to load the diffmap_normalised_fofc.mrc (difference map from Servalcat job in step 2.1.3.2) file.
      NOTE: If the .mrc extension files are not visible while browsing, then change the file type to all files and browse.
   3. Select the ligands using display > sequence.
   4. Go to Actions button and rename the selection as ligands.
   5. Type the following in the command line and hit enter: isomesh mesh_ligands, diffmap_normalised_fofc, 6.0, ligands, carve=2.5.
      NOTE: The mesh width can be adjusted. Here, 0.2 mesh width is used, and this is set by using the following command: set mesh_width=0.2.
   6. Right-click on one of the monomer and go to chain > color to specify the color of each monomer. Color the ligand and mesh using the last box labeled c in the ligand and mesh_ligands objects. The mGlu₅ receptor dimer is shown in cartoon representation, and the monomers are colored in teal and wheat. Ligands are shown in stick, and the mesh contouring the ligands is colored green.
   7. Use the actions > orient/ center/zoom option against the objects and adjust manually to visualize the mesh and clearly. Select the nearby residues that could be interacting with the ligand, for example, Tyr64, Trp100 for QUS, and as required, display their side chain or main chain or both as stick representation.
   8. Ray trace to generate high-resolution images and save them as a png file.
      NOTE: The above steps are repeated for the antagonist bound map EMD-31537 and the corresponding coordinate file PDB-7fd9.

3. Modeling the inhibitor, deoxygalacto-nojirimycin (DGN), and solvent molecules in a high-resolution sharpened map of β-galactosidase

1. Identification of the ligand, DGN, using the half-maps from cryoEM
   1. Preparation of unknown Ligand dictionary for modeling [Deoxygalacto-nojirimycin(DGN)]
      NOTE: The Deoxygalacto-nojirimycin dictionary file is included in the CCP4 monomer library as DGJ (3 letter ligand ID). Nevertheless, here it is considered as a new and unknown ligand to illustrate the process of generating dictionary files for unknown ligands, DGN is used as the 3 letter code.
      1. Open the compound summary for deoxygalacto-nojirimycin (DGN) in the PubChem webpage and copy the smiles string for the compound deoxygalacto-nojirimycin.
      2. Open Phenix > Ligands > eLBOW.
      3. Specify the smiles string as input.
      4. Specify appropriate ligand building optimization method. Here, simple optimization is used.
      5. Paste the chemical smiles string in the ligand definition section.
      6. Provide a job title, output file prefix, and ligand ID appropriately and press run.
         NOTE: The output of the job contains a coordinate file, DGN.pdb, and a dictionary file, DGN.cif file. Jligand²⁹ can also be used to make the cif file.
   2. Download the β-galactosidase coordinate file (6tsh.pdb) from the PDB and remove the coordinates for the protein chains (B, C, D), ligand, DGN, and other solvent molecules by using a text editor or delete chain/zone option in Coot. Save the coordinate file as apo-betaGal_chainA.pdb.
      NOTE: ChimeraX or Pymol can also be used to remove the ligand/solvent molecule.
   3. Download the half-maps (emd_10563_half_map_1.map and emd_10563_half_map_2.map) from additional data of the EMD-10563 entry from EMDB.
   4. Open CCPEM and use Relion to create the mask for the map at a threshold of 0.01 (as described in steps 2.1.3.1.1-2.1.3.1.4).
   5. Run Servalcat (within the CCPEM suite as described in step 2) with the half-maps obtained in step 3.1.3 and the apo-model in step 3.1.2 and D2 symmetry.
      NOTE: The model has to be aligned to one of the half-maps or full maps for locating the ligand density. This can be done by fitting the model into the map using ChimeraX and then saving the coordinates relative to the half-map. In this example, half-maps and the primary map in EMDB have different box sizes/grids, and the model needs to be placed in the half-map.
   6. Open Coot.
      1. Open DGN.cif for the ligand dictionary as generated in step 3.1.1 with Phenix eLBOW. This is done by going to File > Import CIF dictionary and browsing the eLBOW job directory.
      2. Open the protein and ligand coordinate files, apo-betaGal_chainA.pdb from step 3.1.2 and DGN.pdb from step 3.1.6 respectively.
      3. Auto-open the diffmap.mtz file from the Servalcat job (step 3.1.5) and visualize the map labeled DELFWT PHDELWT in Coot. Contour the map at a threshold of ~ 4 absolute value.
         NOTE: It is important to check the map statistics by typing header map.mrc or using tools in Chimera. All the necessary information on the pixel size, box size, minimum, mean, and maximum density, and rms deviation from mean density can be obtained. It is crucial to check the pixel size and box size of the cryoEM maps. The 3D map represents the protein-ligand map volume in a grid of 3D voxels. In each voxel, the electron potential value or the probability of finding the electron at that location is stored. When a certain threshold is chosen, then the voxels, which have a density lower than the specified value, are set to zero. This helps to minimize the experimental noise in the map and visualize the map features better⁴³. Thus, the map should be contoured at a threshold where the map features are easily visible and the noise in the map is minimum.
      4. Go to Ligand > Find Ligands. In the new window that appears, select ligand, the difference map, and the apo-betaGal_chainA.pdb model. Leave the other options at their default settings and click Find Ligands to start the search.
      5. A list of top hits showing potential ligand density is displayed with a copy placed in each of the densities. Manually check all the hits where the ligand has been placed. Keep the most probable hits and remove the ambiguous ones.
         NOTE: The difference in map density at a high threshold clearly suggests the presence of the ligand in the active site.
2. Modeling the ligand DGN and solvent molecules
   1. Perform a real space refinement in Coot to fit the ligand (DGN.pdb) in the difference density map, and then merge the ligand coordinates with the apo-betaGal_chainA.pdb and save it as betaGal_chainA+ligand.pdb.
      NOTE: The ligands that have been placed in regions of other chains can be removed and are not merged while saving the merged coordinate file. The ligands in other chains can be generated by NCS Ligands as shown in example 1, step 1.2.8, or in Refmac/Servalcat using symmetry expansion.
   2. Run another Servalcat job as above (step 3.1.5) with betaGal_chainA+ligand.pdb as input model and the half-maps (from step 3.1.3) with D2 symmetry.
   3. Difference density (from the Servalcat job at step 3.2.2) surrounding the modeled ligand suggests the presence of the solvent molecules coordinating with the ligand molecule. The central density close to the ligand appears to be too big for water molecules.
   4. Based on previous biochemical and structural data indicating the presence of Mg²⁺ at that position, the Mg²⁺ ion is modeled here by clicking the place atom option and specifying the atom as MG.
   5. Model water molecules in the difference density in the active site that indicate the presence of solvent molecule by clicking on the place atom at pointer and selecting water.
      NOTE: Coot also has the option to pick water molecules automatically. Here, since the focus is only on the active site, water molecules are added manually.
   6. Merge the coordinates for Mg²⁺ and water with the betaGal+ligand.pdb file and save it as betaGal+ligand+solvent_chainA.pdb.
   7. With the betaGal+ligand+solvent_chainA.pdb, perform Refmac refinement in Servalcat with D2 symmetry.
      NOTE: The output difference map from this job can serve as a means to validate whether the ligand has been accurately modeled, as the difference map should show minimal residual positive or negative density.
3. Visualization and figure generation with PyMOL
   NOTE: The steps outlined below provide a few examples of making figures using models and maps that can be used for illustrating the presence of ligands and solvents.
   1. Open the refined_expanded.pdb from the last Servalcat job (step 3.2.6) with the ligand and solvent modeled.
   2. Select different chains separately to color them magenta, yellow, green, and teal, as described in example 2.
   3. Go to Display > Sequence, make separate selections for the water molecules, magnesium, and DGN, and rename the objects as water, MG, and DGN, respectively.
   4. Display MG and water molecules as spheres by selecting the objects, right-clicking, and show > sphere. Similarly, display the ligand in stick representation and color the ligands and solvents appropriately. In this case, the ligand has been colored yellow by a heteroatom, the magnesium ion is colored purple, and water molecules are colored red.
   5. Select DGN, MG, and water molecules. Right-click and select actions > copy to object and name it as ligands.
   6. Open the diffmap_normalised_fo.mrc from the Servalcat job in step 3.2.6 and rename as ligand_fo.mrc. Type the following command: isomesh mesh_ligands_fo, ligand_fo, 3, ligands, carve=2.
      NOTE: The carve option of 2 Å is applied around the atoms, and depending on the resolution and quality of the maps, values of 3-5 can be used. Furthermore, when opening cryoEM maps in PyMOL, it is advisable to set normalize_ccp4_maps to off.
   7. Use the actions > orient/ zoom options and adjust manually to visualize the ligands and nearby interacting residues (wherever applicable) as shown in step 2.3.7.
   8. Ray trace these scenes to save high-resolution images using the command ray 3600, 3600.
   9. Save the scene as .png file.
      NOTE: The following steps are optional and outline the process to visualize (1) the difference density (Fo-Fc) for the unmodeled ligand, DGN, (2) the density (Fo) of the modeled ligand, DGN as well as the difference density (Fo-Fc) for unmodeled solvents and lastly for the (3) the density (Fo) for the modeled ligand, DGN and the solvent molecules in the active site.
   10. To demonstrate the presence of the ligand, DGN, and surrounding solvent molecules using the difference map, open diffmap_normalised_fofc.mrc from the Servalcat job in step 3.1.5. Rename the map file as unliganded_fofc.mrc by clicking actions > rename next to the map object. Type the following in the command line: isomesh mesh_ligands_fofc, unliganded_fofc, ligands, level=6, carve=2.
   11. To demonstrate the density of the modeled ligand, DGN, and the surrounding unmodeled solvent density, open the diffmap_normalised_fo.mrc as well as diffmap_normalised_fofc.mrc from the servalcat job in step 3.2.2. Rename the diffmap_normalised_fo.mrc as DGN_fo and the diffmap_normalised_fofc.mrc as DGN_unmodelled_solvents_fofc.
   12. Select MG and water from Display > Sequence, right-click and select actions > copy to object, and name them as solvents.
   13. Type the following in the command line: isomesh mesh_DGN_fo, DGN_fo, DGN, level=3, carve=2; isomesh mesh_solvent_fofc, DGN_unmodelled_solvents_fofc, solvents, level=6, carve=2.
       NOTE: The mesh_DGN_fo will show the density of the ligand, and the mesh_solvent_fofc will show the omit density for the MG and water.
   14. Finally, to visualize the modeled ligands as well as the surrounding solvent molecules in the active site, open the diffmap_normalised_fo.mrc from the Servalcat job in step 3.2.6 and rename it as ligand_fo.mrc.
   15. Type the following in the command line: isomesh mesh_ligand_fo, ligand_fo, selection=ligands, level=3, carve=2.
       NOTE: Here, we used diffmap_normalised_fo.mrc, but the final sharpened map can be used to show the density of ligands and the surrounding residues. It is a good practice to mention the type of map used, B-factor sharpening values in the figure legend.
   16. The mesh width and sphere size can be set by typing the following commands: set mesh_width=0.2 and set sphere_scale=0.25 to make the sphere size smaller.
   17. Color the fo mesh blue and the fofc mesh green.
   18. Use the actions > orient/ zoom options and adjust manually to visualize the ligands and nearby interacting residues (wherever applicable), as shown in step 2.3.7.
   19. Ray trace these scenes to save high-resolution images using the command ray 3600, 3600.
   20. Save the individual scenes as .png files.

4. Effect of resolution on ligand modeling in β-galactosidase

1. Open Terminal and go to the directory containing the two half-maps.
2. Open the two half-maps (step 3.1.3) in .map format in ChimeraX, visualize them and save them as emd10563_half1_1_unfil.mrc and emd10563_half2_1_unfil.mrc.
3. The mask created in step 3.1.4 can be used as an input.
4. Run a PostProcessing job in Relion using half-maps and masks from steps 4.2-4.3, with the default parameters.
5. Repeat step 4.4, now with a Skip FSC-weighing enabled and specifying an ad-hoc low-pass filter of 3 Å.
6. Repeat the step 4.4 with an ad-hoc low-pass filter of 3.5 Å.
7. Open the maps (postprocess.mrc) from steps 4.4-4.6, filter at different resolutions, and the model coordinate, 6tsh.pdb (aligned to the half-maps in ChimeraX) file from step 3.1.2 in PyMOL. Rename the different maps as 3.0, 3.5, and 2.3 as defined by their resolutions.
   NOTE: One can confirm the low-pass filtering of the maps by visualizing them in ChimeraX or Coot.
8. Visualise the effect of resolution on the density of ligand and solvent molecules by creating a mesh of 2 Å around the ligand and solvent molecules at a threshold of 6σ as described in step 3.3.6 with maps filtered to different resolutions.

Results

Example 1
The enzyme enolase from M. tuberculosis catalyzes the penultimate step of glycolysis and converts 2-phosphoglycerate to phosphoenolpyruvate (PEP), which is an essential intermediate for several metabolic pathways^44,45. CryoEM data for the apo-enolase and PEP-bound enolase samples were collected at the same pixel size of 1.07 Å, and image processing was performed with Relion 3.1^46,

Discussion

The improvements in microscope hardware and software have resulted in an increase in the number of cryoEM structures in recent years. Although the highest resolution achieved at the moment in single particle cryoEM is 1.2 Å^57,58,59, the majority of the structures are being determined around 3-4 Å resolution. Modeling ligands in medium to low-resolution maps can be tricky and often fraught with ambiguity. Given the wide...

Materials

Name	Company	Catalog Number	Comments
CCP4-8.0	Consortium of several institutes	https://www.ccp4.ac.uk	Free for academic users and includes Coot and list of tools developed for X-ray crystallography
CCP-EM	Consortium of several institutes	https://www.ccpem.ac.uk/download.php	Free for academic users and includes Coot, Relion and many others
Coot	Paul Emsley, LMB, Cambridge	https://www2.mrc-lmb.cam.ac.uk/personal/pemsley/coot/	General software for model building but also available with other suites described above
DockinMap (Phenix)	Consortium of several institutes	https://phenix-online.org/documentation/reference/dock_in_map.html	Software inside the Phenix suite for docking model into cryoEM maps
Electron Microscopy Data Bank	Consortium of several institutes	https://www.ebi.ac.uk/emdb/	Public Repository for Electron Microscopy maps
Falcon	Thermo Fisher Scientific	https://assets.thermofisher.com/TFS-Assets/MSD/Technical-Notes/Falcon-3EC-Datasheet.pdf	Commercial, camera from Thermo Fisher
Phenix	Consortium of several institutes	https://phenix-online.org/download	Free for academic users and includes Coot
Protein Data Bank	Consortium of several institutes	https://rcsb.org	Public database of macromolecular structures
Pymol	Schrodinger	https://pymol.org/2/	Molecular viusalization tool. Educational version is free but comes with limitation. The full version can be obtained with a small fee.
Relion	MRC-LMB, Cambridge	https://relion.readthedocs.io/en/release-4.0/Installation.html	Software for cryoEM image processing, also available with CCP-EM
Titan Krios	Thermo Fisher Scientific	https://www.thermofisher.com/in/en/home/electron-microscopy/products/transmission-electron-microscopes/krios-g4-cryo-tem.html?cid=msd_ls_xbu_xmkt_tem-krios_285811_gl_pso_gaw_tpne1c& gad_source=1&gclid=CjwKCAiA-P-rBhBEEiwAQEXhHyw5c8MKThmdA AkZesWC4FYQSwIQRk ZApkj08MfYG040DtiiuL8 RihoCebEQAvD_BwE	Commercial, cryoTEM from Thermo Fisher
UCSF Chimera	UCSF, USA	https://www.cgl.ucsf.edu/chimera/download.html	General purpose software for display, analysis and more
UCSF Chimera X	UCSF, USA	https://www.cgl.ucsf.edu/chimerax/	General purpose software for display, analysis and more