PyMOL: Display the electrostatic potential with the APBS plugin

The Schrodinger version of PyMOL as well as the open source PyMOL version (I have version 1.8.x) come along with the APBS plugin installed. Plugins are extensions to PyMOL developed by various authors.

When calculating the electrostatic potential, use the "PyMOL + console" icon to start PyMOL on windows. Like this, you know where the APBS saves temporary files, which you need to check, as outlined in the following. Next load the structure, for which you want to display the electrostatic potential. In the example here this is a CryoEM complex structure of the Y2 receptor. We load it directly from the pdb via fetch 7yon. From the top menu choose: Menu: Plugins, APBS Electrostatics. A menu window open from which you start APBS.

It is important to specify the correct selection of atoms for which we calculate the surface. We want to calculate the electrostatic potential at the molecular surface for the Y2 receptor alone, i.e. in the absence of the peptide agonist and the G proteins bound to the receptor. The receptor is chain R. Therefore we extend the selection string to: "polymer and 7on and chain R". This will include all protein atoms of the Y2 receptor, but exclude the other chains as well as water molecules, lipids, cofactors, metal ions, etc. This is an important point. For other cases one may want to include metal ions or anions into the calculation. In this case they would need to be selected as well.

Next under "Advanced configuration" you activate the option "Don't delete temporary files". After that you hit the "Run" button and the calculation should start. If there is an error related to file access, read this.

If everything runs fine, you should see the molecular surface colored by the electrostatic potential in PyMOL graphic window. Before we analyse this further, we check the following. The electrostatic potential is calculated by continuum electrostatics algorithms and these depend on the atomic radii and partial charges. Full charges influence the electrostatic potential most strongly. It is therefore important, that the carboxylates, histidines, cysteines, metal ions etc. have the correct protonation state and charge. The pdb2pqr program, which does this assignment, will assign no charge to histidines (single protonated state) for example. In the PyMOL console window you will find the path of the directory where the APBS plugin stores temporary files. Open this path in the file explorer and open the file mol.pqr in a text editor. The file has the following format:

                                                        partial q    radius
ATOM      1  N   LEU    40     179.452 167.105 112.830  0.10100000   1.824
ATOM      2  CA  LEU    40     178.947 166.472 114.042  0.01040000   1.908
ATOM      3  C   LEU    40     179.477 167.179 115.285  0.61229998   1.908

This file has the same format as a pdb file, but the atomic partial charges and the radius are written to the occupancy and B-factor columns, respectively. You need to check this file if all metal ions are present, if any, and if protonable residues are present in the correct protonation state. It may be necessary to use programs to titrate such residues and feed this information into the APBS plugin of PyMOL. Meaningless or wrong results will be obtained, if charge assignment is in error or if atoms are missing.

The following figure shows the result of the calculation for the Y2R receptor.

Figure 1: Electrostatic potential of the Y2 receptor and interacting pancreatic peptide and G_α1 in the complex of pdb id 7yon. Note that the membrane has not been included in the calculation and is replaced by water in the continuum electrostatic calculation. Therefore the potential of the Y2 receptor at the membrane region is not physiological. Top views from the extracelluar and cytosolic sides indicate distinct negative electrostatic potential at the extracellular face and orthosteric peptide binding pocket and positive electrostatic potential at the cytosolic side. The potentials of the pancreatic peptide (mostly positive) and the G_α1 protein (negative potential at and around the C-terminal helix that inserts into the 7TM domain) are complementary to the binding interfaces of the Y2 receptor indicating the electrostatic interactions play a role in complex formation.

Note that the molecular surface and electrostatic potential of the pancreatic polypeptide in the image above is erroneous. When the APBS plugin was run, a warning was issued that the polypeptide chain has breaks and that the chain contains an unknown residue type named TYC. This is the C-terminal tyrosine residue that is amidated at its main-chain C-terminus. This residue has thus been omitted from the structure (see image above, you can see the tyrosine sticking out of the molecular surface). Furthermore, examination of the file mol.pqr shows that the penultimate arginine residue (R35) bears a C-terminal negatively charged carboxylate group. One would need to include the tyrosine with the amidation into the calculation with the correct charges, but this is not described here. This would likely render the potential of the peptide even more positive.

Back to PyMOL tutorial main page.