Medicinal chemists design molecules before they build them.  Believe me, it is a joy to make the molecule that you imagined and wanted.

My mentor at Purdue University, Prof. Mark Cushman, devised a way to design hybrid molecules.1 We have referred to it as the “platform hop” strategy.1 A piece of one molecule is transported, in the imaginary sense, to a different second one.  The sequence of graphics below illustrates the process.

First, you get X-ray crystal structure #1.2 It contains the drug target, which in this case is Topoisomerase IB-DNA covalent complex (gray), with a bound ligand (purple).  This ligand is a known inhibitor of the target and it displays weak in vitro anticancer activity.


Next, you get X-ray crystal structure #2.3 It contains the same target, but with a different bound ligand (SA315F; green).  This structurally distinct compound is another known inhibitor of the drug target.  Note the protuberance on the left side of the ligand.  That is beta glucose.


Then, the target structures are superposed, using the Topoisomerase IB backbone alpha carbons as guides.


In our case, we were interested in whether the glucose of the green molecule could be transferred to the purple scaffold—and where exactly it could be attached.4 By removing the target structure and changing the viewing angle, you can see that the glucose connects to the rest of the green ligand at a nitrogen (blue vertex indicated by arrow), and that this nitrogen aligns well in three-dimensional space with the nitrogen on the purple ligand.


We designed 20 molecules, all similar to the purple one, and with a sugar moiety connected at the nitrogen.  Eleven of the molecules bear the aromatic ring substituents that decorate Phase I anticancer drugs indotecan and indimitecan.  The activities of the new molecules varied widely.  One of the new analogues performed comparably to camptothecin in the target inhibition assay and displayed potent (49 ± 3 nM) cancer cell growth inhibitory activity.4


This strategy is imperfect, like all computational drug-design strategies.  There are many confounding factors, including differences in solvation and induced fit.  But “platform hopping” can be tried out with different targets and ligands.



1. J. Med. Chem. 2011, 54, 4937–4953.

2. PDB code: 1SC7 (DOI: 10.2210/pdb1sc7/pdb)

3. PDB code: 1SEU (DOI: 10.2210/pdb1seu/pdb)

4. J. Med. Chem. 2014, 57, 1495–1512.

Software credits: Pymol and Sybyl

Note to drug designers:  the details for our process are described in the experimental section of each paper.