Optimization of a cell surface vimentin binding peptoid to extract antagonist effect on lung cancer cells.

Targeting cytoskeletal proteins that are uniquely translocated to cancer cell surface may provide an alternative path for conventional drug discovery. Vimentin is such a cell surface-translocated cytoskeletal protein (CSV) found in non small cell lung cancer (NSCLC). We previously reported the identification of CSV-binding peptoid, named JM3A. While JM3A had no antagonist effect, here we used multiple strategies to optimize the binding of JM3A on CSV and extract the antagonistic effect. We first performed minimum pharmacophore identification studies using alanine/sarcosine scans. These studies revealed that residues 1-4 and 8 (from the C-terminus) are not important and those residues 5-7 are important for JM3A binding to CSV. We then found that our previous N-terminal benzophenone (BP)-coupled JM3A (JM3A-BP), which was used for pull-down and target identification studies, displayed 3-fold binding enhancement. The molecular docking studies indicated that the BP moiety binds to a new binding pocket on the vimentin coil 2 fragment, and further studies using 12 benzophenone-like moieties indicated that at least two phenyl groups are needed to occupy this new binding site. Interestingly, the binding was improved when non-important and bulky residues at the 4th and 8th positions were replaced with methyl groups (JM3A-4,8-BP). We next dimerized JM3A-4,8-BP to enhance the binding via the "avidity effect," using a central lysine linker to develop JM3A-4,8-BPD1 (EC = 300 nM). This showed 27- and 63-fold-improvement in binding over JM3A-4,8-BP and JM3A monomers, respectively. JM3A4,8BPD1 also displayed binding comparable to vimentin antibody. Finally, we observed an antagonist effect on H1299 NSCLC cell proliferation and viability from this most improved dimeric JM3A-4,8BPD1, which was not shown by the monomeric versions.