Abstract
CONSPECTUS: A series of diseases, ranging from cholera via travelers' diarrhea to hamburger disease, are caused by bacterially produced toxic proteins. In particular, a toxic protein unit is brought into the host cell upon binding to specific membrane-bound oligosaccharides on the host cell membrane. For example, the protein that causes cholera, cholera toxin (CT), has five identical, symmetrically placed binding pockets (B proteins), on top of which the toxic A protein resides. A promising strategy to counteract the devastating biological effects of this AB, protein involves the development of inhibitors that can act as mimics of membrane-bound GM1 molecules, i.e., that can bind CT strongly and selectively. To reach this goal, two features are essential: First of all, the inhibitor should display oligosaccharides that resemble as much as possible the naturally occurring cell-surface pentasaccharide onto which CT normally binds, the so-called GM1 sugar (the oligosaccharide part of which is then labeled GMlos). Second, the inhibitor should be able to bind CT via multivalent interactions so as to bind CT as strongly as possible to allow for a real competition with the cell-membrane-bound GM1 molecules.
In this Account, we present elements of the path that leads to strong CT inhibition by outlining the roles of multivalency and the development and use of GM1 mimics. First, multivalency effects were investigated using "sugar-coated" platforms, ranging from dendritic structures with up to eight oligosaccharides to platforms that mimicked the fivefold symmetry of CT itself. The latter goal was reached either via synthetic scaffolds like corannulene or calix[5]arene or via the development of a neolectin CT mimic that itself carries five GMlos groups. Second, the effect of the nature of the oligosaccharide appended to this platform was investigated via the use of oligosaccharides of increasing complexity, from galactose and lactose to the tetrasaccharide GM2os and eventually to GMlos itself. The combination of these threads gives rise to a series of inhibitors that can strongly bind CT, with IC50 values below 100 pM, and in some cases can even bind one-on-one.