Drug Design Lab Exercise

In this exercise you are going to develop a small virtual combinatorial library of huprines (potential treatment for Alzheimers Disease) to test for inhibition of acetylcholinesterase (AChE) via virtual screening.

1)  The basis for your scaffold will be huprine-A itself.  To obtain the huprine molecule, first load in the huprine + AChE crystal structure:
  • [SYBYL]  File  >>  Read... 
  • [Read File]  Select "1E66.pdb"  >>  {OK}
  • [Center the Molecule]  Select "CENTER_VIEW"  >>  {OK}

2) Remove everything except the huprine ligand:
  • [SYBYL]  Build/Edit  >>  Delete  >>  Substructure...
  • [Substructure Expression]  {Substructures...}  >>  select "A/HUX803"  >>  {OK}
  • [Substructure Expression]  {Invert}  >>  {OK}

3) You now have the huprine molecule on screen.  Note that it has no hydrogen atoms, but don't worry about this for the time being.  There are, however, some errors in the atomic valences that must be corrected, in that the proper bonding should be

Huprine-x molecule

 

 

 

 

 

 

Figure 1.  huprine-x molecule (with all hydrogens) showing the proper bonding structure and the three combinatorial chemistry attachment points (R1-R3)

4) To fix the structure:
  • [SYBYL]  Build/Edit  >>  Modify  >>  Atom... 
  • [Option]  select "ONLY_TYPE" (this will change atom types but not positions)
  • >>  {OK}
  • [SYBYL (on screen)]  click on the each atom of the quinoline moiety which is
  • Shown as a haphazard mis of single and double bonds, but should be all aromatic.
  • [Option]  if the first atom highlighted for changing is the N (i.e., the blue atom), modify the value for 'Selection' to "N.ar"  (or if it was a C, use "C.ar")  >>  {OK}  >>  Note that the atom highlighted on screen has moved from the first atom you selected to the next.  Continue making the appropriate modifications and clicking {OK} until you've cycled through all ten atoms.
5) Note that the bonding in the aromatic ring has been fixed.  The molecule as shown, however, is huprine-x, and what we want as out basic scaffold is the dechlorinated analog huprine-a.  To make this change
  • [SYBYL] Build/Edit >> Modify >> Atom...
  • [Option] select "TYPE" >> {OK}
  • [SYBYL (on screen)] click on the Cl atom (green)
  • [Option] change the value for 'Selection' to "H" >> {OK}
     
6) Now add hydrogens:
  • [SYBYL]  Build/Edit  >>  Add  >>  Hydrogens
     
7) Label the atoms which you are going to combinatorially modify (circled in Fig. 1) as follows:  
  • [SYBYL]  View  >>  Label  >>  Atom ID...
  • [Atom Expression]  select "M1:" (or whichever molecular area contains your ​hup from steps 1-6)
  • [SYBYL (on screen)]  locate on screen the three H's circled in Fig. 1 and click on each.
  • [Atom Expression]  {OK]
  • The three H's should now be labeled (probably 7, 16 and 20 if your structure is correct).
  • [SYBYL] Build/Edit >> Name Molecule...
  • [Name Molecule] enter "hup" for the Molecule name >> {OK}

8) Finally, name your molecule:
  • [SYBYL] Build/Edit >> Name Molecule...
  • [Name Molecule] enter "hup" for the Molecule name >> {OK}

9) You are now ready to use CombiLibMaker to design a combinatorial library. First initialize a new system:
  • [SYBYL] Tools >> CombiLibMaker >> Core + Sidechains...
  • [SYBYL] Core + Sidechains >> Session >> Create New...
  • [String] Enter a name for your session, e.g., "hups" >> {OK}

10) Specify your backbone structure:
  • [SYBYL] Core + Sidechains >> Define Backbones...
  • This will bring up a small unnamed command widget. On it, click:
  • [ ] Add Backbone
  • [Option] select "03_Molecular_Area" >> {OK}
  • [Option] select "Completely_Provided:_3D_Coordinates" >> {OK}
  • [Atom Expression] select "M1:" (or whichever molecular area contains your hup from steps 1-6)
  • [SYBYL (on screen)] click on the three H's you labeled in step 7
  • [Atom Expression] {OK}
  • [Yes or No] {No} (they should not be equivalent attachment points)
  • [Yes or No] {No} (you will not be defining additional backbones)
  • [ ] Close

11) Specify your sidechains:
  • [SYBYL] Core + Sidechains >> Define Sidechains...
  • [ ] Add Sidechain
  • [Option] select "05_User_Molecular_Database"
  • [Database Selection] select "sidechains.mdb" >> {Open}
  • [Option] select "Loop_through_all_molecules" >> {OK}
  • [Molecular Area] select the lowest numbered molecular area that is currently labeled "" >> {OK}
  • If you don't see a molecule on your screen at this point, is must be hiding somewhere off screen. On your SYBYL screen, scan out by pulling down on your mouse at the same time as simultaneously depressing the middle and right mouse buttons.
  • Once you see it, translate it closer to the center of the screen and scan back (i.e., push back with the mouse while depressing the middle and right buttons) until the molecule is a normal size. Once you have the molecule in a convenient place, look for an atom that is colored magenta -- this will be the H atom that gets specified for substitution.
  • [Yes or No] {Yes} (you do want to add this to the list)
  • [SYBYL (on screen)] click on the H-atom that will be substituted: in most of the side chains, this will be an H that is just off an aldehyde carbonyl; it had been colored magenta to make it easier to recognize but when you clicked "Yes" in the previous step it got recolored cyan.
  • [Atom Expression] {OK}
  • Continue repeating the previous 5 substeps as the database cycles through each sidechain molecule. When it's done, you should be left with a spreadsheet of sidechains plus an instructions widget labeled [ ].
  • On the latter: [ ] click {Close}

12) You will now place some restrictions on the products that can be achieved:
  • [SYBYL] Core + Sidechains >> Define Restrictions >> Backbone...
  • [Setting Restrictions] read the dialog and click {OK}
  • On Attachment Point 1 (column ATTACH_1) it has proven difficult to place any of the substituent groups except methyl and ethyl, so we're going to change the cells for H, F and Cl in column 1 from Y (yes) to N (no).
  • [BACKBONE_1_RESTRICTIONS] click on the cell corresponding to row 'CL' and column '1:ATTACH_1' >> in the text field, type "N"
  • [keyboard] hit the enter key
  • [BACKBONE_1_RESTRICTIONS] repeat the above process for the F and H rows in column 1:ATTACH_1
  • [ ] (small side widget) click {Close}

13) Generate the library:
  • [SYBYL] Core + Sidechains >> Generate Library...
  • [Options for Library Generation] change the Stereochemistry Output Options from "Preserve_All" to "Preserve_None" >> change the Product Naming from {By_Numbers} to {By_Names} >> change the Run Mode from {batch} to {interactive} >> {OK}
  • [Yes or No] {Yes} (you wish to continue)

14) The initial library generation just created 'SMILES' strings which are simply expressions that define atomic connectivity. You are now going to need to convert these strings into 3D molecule structures via Concord:
  • [SYBYL] Tools >> CONCORD Standalone...
  • [XConcord 5.1.2 Control Window] for the Name of the compound input file, click {Modify...}
  • [Compound source file] Double click on "hups.clm" directory >> scroll down the Files list and select "hups.smi" >> {OK}
  • [XConcord 5.1.2 Control Window] for the Format of the input file, click {Modify...}
  • [Compound source file format] select "SMILES" >> {OK}
  • [XConcord 5.1.2 Control Window] for the Name(s) and format(s) of the output file(s) click {Modify...}
  • [Output formats and files] deselect "SYBYL" >> select "MDL SD"
  • [Concord output file -- MDL SD format] change the name to "hups_3D.sdf" >> {OK}
  • [Output formats and files] {OK}
  • [XConcord 5.1.2 Control Window] click {Perform}
  • The format conversion should run in a minute or so, and statistics will be reported to the green log window. Once it's completed, kill the session via:
  • [XConcord 5.1.2 Control Window] {Session} >> Exit
  • [Warning] click {Yes} (it is okay to exit)

15) In order to examine your structures, load them into a molecular spreadsheet as follows:
  • [SYBYL] File >> Convert MACCS File >> To Spreadsheet...
  • [ANY_FILE] select "hups_3D.sdf" >> {OK}
  • [STRING] {OK}
  • [IDENTIFIER] give it a name such as "hups" >> {OK}
  • [Yes or No] {No} (do not include property data}
  • You can now view individual structures by selecting rows on the spreadsheet and displaying them on the main screen:
  • [HUPS] click on a row name >> {Show RowSel}
  • NOTE: once again, it's possible that your structures will appear off screen and you may need to scan out and then rotate/translate the molecule into a favorable viewing position.
 
16) Note that you do a lot of analysis on your ligands within the molecular spreadsheet. For example, you can compute numerous physicochemical properties such as ClogP:
  • [HUPS] click on the header box for column 1 (should have a "1" in it) >> click {AutoFill}
  • [New colmn type] select "CLOGP" >> {OK}
  • [IDENTIFIER] {OK}
  • [IDENTIFIER] {OK}
 
17) Similarly, you can try other Lipinski parameters, such as molecular weight:
  • [HUPS] click on the header box for column 3 >> {AutoFill}
  • [New colmn type] select "MOL_WEIGHT" >> {OK}
  • [IDENTIFIER] {OK}
 
18) H-bond acceptors:
  • [HUPS] click on the header box for column 4 >> {AutoFill}
  • [New colmn type] select "SLN_PROPERTY" >> {OK}
  • [Property to Calculate] select "Acceptor" >> {OK}
 
19) H-bond donors:
  • [HUPS] click on the header box for column 5 >> {AutoFill}
  • [New colmn type] select "SLN_PROPERTY" >> {OK}
  • [Property to Calculate] select "Donor" >> {OK}
 
20) Rotatable bonds:
  • [HUPS] click on the header box for column 6 >> {AutoFill}
  • [New colmn type] select "SLN_PROPERTY" >> {OK}
  • [Property to Calculate] select "RotBonds" >> {OK}
  • and note that there are numerous other properties that can be computer.
 
21) Once you're done verifying that the structures make sense, kill the spreadsheet:
 
22) You're now ready to set up the FlexX docking run:
  • [HUPS] File >> Close
  • [Yes or No] {No} (you don't need to save the table)
  • [SYBYL] Tools >> FlexX Suite >> Dock Multiple Ligands...
  • [FlexX - Run Multiple Molecules] for the Receptor Definition File (RDF),
  • specify a file name (e.g., AChE) >> click {Define...}
  • [FlexX - Create RDF File] for the PDB file, click on the {...}
  • [PDB_FILE] select "1E66.pdb" >> {OK}
  • [FlexX - Create RDF File] for the Active-Site File, click {Create...}
  • [STRING] enter "AChE_site" >> {OK}
  • You are going to define the active site to be the area around the originally co-crystallized huprine molecule as follows:
  • [FlexX - Create Active Site] scroll down the residue list all the way to the bottom and select "HUX803" >> {OK}
  • [FlexX - Create RDF File] click {OK}
  • [FlexX - Run Multiple Molecules] change the Ligands file structure from {SLN File} to {SD File} >> click on the nearby {...}
  • [MDL_SD_FILE] select "hups_3D.sdf" >> {OK}
  • [FlexX - Run Multiple Molecules] about 2/3 of the way down the widget (on the right) click on the {FlexX Details...}
  • [FlexX - FlexX Details] change the Num. Answers from "30" to "10" >> {OK}
  • [FlexX - Run Multiple Molecules] towards the bottom, click on the "Run FlexX in Batch" button >> specify a job name (e.g., "hups_AChE") >> {OK}
  • This job will take quite a long time to finish (~30 mins) however as soon as the message "NetBatch is processing job hups for batch execution." appears in your bottom console window (long blue rectangle) it has been submitted to the background which means that you can log off and return later without disrupting the job.
 
Note 22.1: the long job execution time and the limited number of FlexX licenses could mean that someone else's job in the class prevents yours from starting. If so, then you may try to schedule it to run later:
  • [SYBYL] Tools >> FlexX Suite >> Dock Multiple Ligands...
  • [FlexX - Run Multiple Molecules] click on {Netbatch Options...} (bottom right)
  • [Netbatch Options] click on the "Hold For Later" button >> change the time to some point at least 30 minutes later >> {OK}
 
Note 22.2: If you are simply unable to get the job started due to license conflicts, then you may copy a complete set of results already completed, by typing the following command into the SYBYL bottom console window:
  • [console] dcl cp -r /usr/people4/ghl/training/mdcmlab/hups_AChE .
  • [keyboard] hit the 'Enter' key
 
23) Once your job has completed, you may examine the results of your docking simulations:
  • [SYBYL] Tools >> FlexX Suite >> Browse Results...
  • [FlexX Answer Browser] for the jobname, click the {...} button
  • [DIRECTORY_FILE] select "hups_AChE" >> {OK}
 
24) Entries for the top scoring conformer for each successfully docked ligand are now given in the answer browser, as well as its docking score (approximately equal to its binding free energy in kcal/mol). You may see the structure of each ligand as follows:
  • [FlexX Answer Browser] click on the ligand's entry in the 'Filename' list >> {Display Ligand}
 
25) You can display as many ligands as you want, but note that it might get a bit messy looking unless you use the {Undisplay Ligand} ligand button to hide a few of the structures.

26) Also note that you can load the receptor model onto the screen as follows:

  • [FlexX Answer Browser] near the top, click on the 'Active Site' button.

27) Finally note that you can load the entire protein model onto the screen as follows:

28) [FlexX Answer Browser] near the top, click on the 'Protein' button. The originally co-crystallized huprine-x will appear on screen as well, which can be a useful comparison with your own ligands, but note that it and all other heteroatoms were omitted from your receptor model (so you weren't trying to dock into a receptor that already had an inhitor).

Questions:

1) How many of the 50 molecules in this study perfectly comply with the Lipinski rules of 5?

2) Which of the molecules gives you the best docking score? You many specify the molecule by listing the substituents for each of the R1, R2 and R3 substitution points.

3) Examine how the best scoring molecule in question 2) binds to the AChE receptor. List at least two important ligand-receptor interactions that play a role in the ligand's successful binding. Note that you may wish to label specific residues on screen via:

  • View >> Label >> Substructure...
  • [SYBYL (on screen)] click on the substructures of interest
  • [Substructure Expression] click {OK}

 

 


CONTACT
David Johnson
emaildkjohnson@ku.edu
Access to more than 40 computational chemistry software programs and databases
High-performance computational tools accelerate drug discovery and minimize costs
Analyze complex, multidimensional data sets to rapidly generate biological insights
KU Today