Antibody Humanization: CDR Grafting Panel

Graft the CDR loops of a murine antibody onto a human framework, with mutations around the join points of the loops to improve the geometry.

To open this panel, click the Tasks button and browse to Biologics → Humanization - CDR Grafting.

  • Using
  • Features
  • Additional Resources

Using the Antibody Humanization: CDR Grafting Panel

Before you can use this panel, you must have a structure for the antibody that you want to humanize. If you do not have a structure, you should first build a structure using the Antibody Structure Prediction Panel. You can also obtain a structure from the PDB, or import one from a file.

The Vernier zone, the canonical structural residues, and the interface residue are all grafted together with the CDRs by default., i.e. their residue identity and structures are preserved from the mouse structure. A modified CDR definition based on Kabat numbering is used by default: the H1 loop is extended by 5 residues on the left, and the H3 loop is extended by 2 residues on the left.

You can use sequences such as human germline sequences instead of structures for the framework templates. In this case, the mouse structure is mutated to the chosen template sequence, and any insertions or deletions are rebuilt.

Our recommended strategy is to start the CDR grafting with the default settings, which generally leads to a stable humanized structure by the inclusion of the three special regions mentioned above. Then you can decrease the number of residues in the grafting, to gradually reduce the number of mouse residues in the humanized model. Options for the three regions as well as customized regions and the CDR loop definitions are available in the CDR Grafting - Advanced Options Dialog Box.

If the humanized model is not stable, or you simply want to improve its stability, back mutations (i.e. mutating the residues in the humanized model back to the original mouse model or to some other residue) are often required. When you perform a back mutation, the change in protein stability, assessed with the same computational model as in the residue mutation workflow, is returned with the results. A negative value means that the mutated protein is more stable.

After you have performed the humanization, you may want to refine the structure of the antibody further to relieve strain. For example, you might want to do a side-chain refinement on the framework to relieve minor clashes with the CDR loops.

The color scheme used for the resultant model is as follows:

Antibody colors
Color Region
FR
CDR
VZ
CSR
VZ+CSR
IR

To run CDR grafting from the command line, you can use the following command. Run the command with -h for more information.

$SCHRODINGER/run -FROM bioluminate cdr_graft.py

For information on command options, see cdr_graft.py Command Help.

Antibody Humanization: CDR Grafting Panel Features

 

Numbering scheme option menu

Choose the numbering scheme to use, from Chothia, Kabat, IMGT, EnhancedChothia, AHo. The default is Kabat.

Import antibody structure to humanize buttons

Click one of these buttons to import the structure of the antibody that you want to humanize into the workflow.

  • From Workspace—Use the structure that is displayed in the Workspace.
  • PDB ID—Use the structure from the specified PDB ID. Opens the Enter PDB ID dialog box, in which you can enter the PDB ID of the structure. The structure is retrieved from a local copy of the PDB if it is available, or from the RCSB web site, depending on the preference set for PDB retrieval.
  • Browse—Open a file browser in which you can navigate to the desired location and select the file that contains the antibody.

Note: When using the CDR Grafting Panel to humanize an antibody, the light chain human % and heavy chain human % properties will be associated with the complete antibody entry in addition to the individual heavy and light chain entries the Project Table.

Specify replacement framework section
Use framework from option menu

Choose the source of the structures for the framework region from this option menu:

  • Structural database—Search the antibody databases for the best-fitting human antibody frameworks. The human antibodies are aligned to the query antibody to find the best alignment of the framework regions to the query framework regions where the CDR loops join. A Load button is shown to the right when you choose this option, which you click to perform the import.

  • Framework structure—Read the framework structure from a file. The structure does not have to be an entire antibody, provided it contains the framework region. A Browse button is shown to the right when you choose this option, which you click to locate the file.

  • Framework sequences—Read the sequences for the framework region from a file. A Browse button is shown to the right when you choose this option, which you click to locate the files for the heavy and light chains, and also for the J region if you use human germline sequences.

  • Default human germline sequences—Read the sequences for the framework region from a default human germline compilation obtained from the Antibody Society's Germline Database Working Group.

  • A Load button is shown to the right when you choose this option, which you click to load the sequences.

For the last two choices (i.e. for sequence import), the framework table is divided into two, one for the heavy region and one for the light region, so you can choose templates for these regions independently. A homology model is built for the framework region using the default options, as represented in the Antibody Structure Prediction Panel.

Load button

Load framework templates from the specified databases or sequences from the default human germline. Only present when you choose Structural database or Default human germline from the Use framework from option menu.

Choose Databases button

Choose the databases to search for human antibodies. These databases are the ones that are searched for human frameworks that are the best candidates for grafting the CDR loops of the input antibody onto. Opens the Choose Database dialog box, in which you can add databases to the list, remove them from the list, and mark them as active. All databases that you select as active are searched when you run a search for human frameworks. Only present when you choose Structural database from the Use framework from option menu.

Browse button

Click this button to locate the file containing the framework structure or the files containing the framework sequences. Only present when you choose Framework structure or Framework sequences from the Use framework from option menu.

Weight Options button

Specify weights for the stem geometry score and the framework similarity score for the calculation of a weighted score. Opens the Weights dialog box. The weights are specified separately for templates from crystal structures and templates from homology models. Only present when you choose Structural database or Framework structures from the Use framework from option menu.

Advanced Options button

Set options for additional regions to graft and for customizing the CDR loop region. Opens the CDR Grafting - Advanced Options Dialog Box.

Framework table

This table lists the templates that were found or imported for the framework regions, along with sequence identity and stem geometry scores. The table has the following columns when you choose Structural Database or Framework Structure from the Use framework from option menu. The similarity is the number of matching residues divided by the total number of residues, where "matching" means that the two residues have a positive score in the BLOSUM62 matrix. The identity is the number of identical residues divided by the total number of residues.

Heavy PDB ID of the template used for the heavy framework region
Light PDB ID of the template used for the light framework region
Light Chain Type Type of chain in the light framework region (Kappa or Lambda)
Weighted Geom+Sim Weighted combination of the stem geometry fitness score and the sequence similarity score. The rows in the table are ordered by this score. The weights can be specified by clicking Weight Options and specifying the weights in the Weights dialog box.
Composite Score Average of the Heavy Sim. and Light Sim. scores. This score is used as the sequence similarity score in the calculation of the weighted score.
Heavy Fr. Sim. Sequence similarity of the framework region for the heavy chain.
Heavy Fr. Iden. Sequence identity of the framework region for the heavy chain.
Light Fr. Sim. Sequence similarity of the framework region for the light chain.
Light Fr. Iden. Sequence identity of the framework region for the light chain.
CDR Stem Geom Fitness score for the geometry of the CDR loop stem residues. The stem residues are the residues in the framework region that are directly attached to the CDR loops. The geometry score is related to the RMSD between the native and the grafted antibody for the C-N distance, the C-N-C angle, and the C-C-N-C dihedral across the bond between the stem residue and the first (or last) loop residue.
PDB Resolution Resolution of the PDB structure, if available; otherwise experimental method, if available.
Structure Source of the structure used for the framework regions, which can either be a crystal structure or a homology model. The scoring depends on the structure source.

If you choose Framework Sequences or Default human germline from the Use framework from option menu, two tables are presented, one for the heavy chain and one for the light chain. The tables have the following columns:

Heavy| Light PDB ID of the template used for the heavy/light framework region
Light Chain Type Type of chain in the light framework region (Kappa or Lambda). Only in the light chain table.
Heavy|Light Fr. Sim. Sequence similarity of the framework region for the heavy/light chain. The similarity is the number of matching residues divided by the total number of residues, where "matching" means that the two residues have a positive score in the BLOSUM62 matrix.
Heavy|Light Fr. Iden. Sequence identity of the framework region for the heavy/light chain. The identity is the number of identical residues divided by the total number of residues.
Heavy Chain - Closest Germline The heavy chain germline gene exhibiting the greatest sequence identity to the query.
Replace Framework button

Replace the query framework with the framework selected in the table. When the resulting model is build, a sequence viewer panel opens, showing the sequences of the original light and heavy variable chains and the chains of the grafted model, annotated with disulfide bridges, SSA, and CDR regions. This sequence viewer allows you to compare the original and the replacement framework. The framework replacement is done without adjusting the loops in the CDR regions.

Define back mutation in framework region section

If you want to mutate selected residues in the framework region to relieve clashes with the grafted CDR region, you can do so in this section.

Show residues with CDR/framework vdW clash option

Select this option to list residues in the table in which there are clashes between the CDR region and the framework region, based on van der Waals interactions.

Show residues within N of CDR side chain option and box

Select this option to list residues in the table that are within the specified distance of a CDR side chain.

Mutations table

This table lists the mutations that match the criteria specified by the Show residues with CDR/framework vdW clash and Show residues within N of CDR side chain options. The framework residue is shown, along with the distance to the CDR side chain and a check for CDR clashes.

The Mutations column lists the residue to mutate to, and allows you to select this residue, by clicking in the column and choosing from a menu. In addition to the standard residues, this menu has an item for the corresponding query residue and an item for no mutation. Setting the mutation to No mutation disables mutation of this residue. Residues from the default human germline have text indicating how many germlines the residue is found in. The tooltip for these residues lists the germline titles.

The Δ Stability column gives the change in stability on mutation of the residue. A negative value means that the mutation is beneficial to protein stability. The stability calculation is calculated with the same methods as for residue scanning (see residue scanning, affinity maturation, humanize, antibody, stability, prediction, surface complementarity, substitution, optimization, mutant, variant, delta prime energy, delta SASA, delta pka, delta affinity, delta hydropathy, delta total rotatable bonds, delta stability, delta vdW surface complimentarity, delta prime energy).

Clear Mutations button

Clear the Mutations column for the residues that are selected in the table. This button allows you to clear the mutation for multiple table rows, so that the mutation is not performed.

Mutate to Query button

Set the mutation in the Mutations column to the query residue. This button allows you to set the mutation for multiple table rows.

Pick residue option

Pick residues in the Workspace to select them in the table.

Perform button

Select an option to perform Single mutations or Simultaneous mutations. If Single mutations is selected, the Perform Back Mutations button becomes available, which performs back mutations on the residues shown in the table that have a valid mutation shown in the Mutations column. If Simultaneous mutations is selected, the Send Mutations to Residue Scanning button becomes available. Clicking this button opens the MM-GBSA Residue Scanning Panel, which performs the mutations. After all the mutations are performed, the mutated residues and residues within 4 Å are refined, and the change in protein stability due to the mutation is calculated.

Reset button

Clear all data from the panel and reset all settings to their default values.

Tutorials

Related Topics