Designing for Shape

ACS MEETING NEWS

Natural products are classic inspirations for, and indeed sources of, therapeutic agents, but they are scarce and often painstaking to synthesize. A more practical way to capture their bioactivity is through simple metal complexes.

That was the gist of a talk that Eric Meggers, an assistant professor of chemistry at the University of Pennsylvania, gave last month at the American Chemical Society's national meeting in Washington, D.C., in a session sponsored by the Division of Organic Chemistry.

Meggers showed that simple, rigid organometallic molecules designed to mimic a natural product's shape not only can replicate the unique protein-binding activity of the natural product but also can be more effective than the inspiration itself. This approach to designing bioactivity, which he refers to as "morphing natural products into simple metal complexes," already has yielded easily accessible compounds with unprecedented selectivity and affinity for key proteins. The compounds may find use as probes of biological processes and even as therapeutic agents.

A natural product's bioactivity often is due to a defined shape complementing the pocket of a protein, Meggers explained. Rigid, defined shapes, he showed, are more easily achieved by complexes assembled around a central metal core than by plain organic compounds.

"It is not simple to make an organic molecule with a defined shape that is not complicated," Meggers said. One carbon atom bonded to four different groups yields only two isomers, but an octahedral metal center with six different ligands offers 30 stereoisomers. That means a metal center is a more sophisticated platform on which to build a three-dimensional structure.

Meggers focuses on molecules that bind to one target only. This one-to-one selectivity, he believes, can be accomplished by molecules that are rigid enough that they cannot adapt their shape to anything but the intended target. His group easily achieves such rigidity around a metal core through chelating or multidentate ligands.

"Our compounds are really rigid," Meggers said. "We did not realize how useful that [rigidity] might be until we found how selective the compounds are."

Several ruthenium complexes--designed by graduate students Douglas S. Williams, G. Ekin Atilla, and Howard Bregman to mimic the shape of the alkaloid staurosporine--have turned out to be highly potent and selective inhibitors of protein kinases. One compound binds the protein kinase Pim-1 with picomolar affinity, Meggers reported. "It does not bind to anything else, even though the active sites of protein kinases are very similar." The only explanation, he said, is that the compound's defined, rigid shape exquisitely matches that of the active site.

Crystal structures of the inhibitor bound to human Pim-1--obtained in collaboration with Stefan Knapp at Oxford University--show that the inhibitor closely mimics the binding mode of staurosporine. Furthermore, Meggers said, the crystal structure reveals that ruthenium is not involved in the binding of the complex in the protein pocket. "It is buried in the middle, acting like a glue to keep the organic ligands together," he explained. This use of metals as scaffolds, rather than as reactive centers, is not widely practiced, he said.

The caveat in Meggers' approach is that the metal complex should be chemically stable. Unlike the covalent bonds in plain organic compounds, the coordinate covalent bonds in metal complexes can dissociate, Meggers explained. With certain metals such as ruthenium, though, the coordinate covalent bonds are often completely kinetically inert--that is, the activation energy for dissociation is high.

Various tests have convinced Meggers that the metal complexes designed in his lab are chemically stable and are not any more toxic than plain organic compounds. "We've tested in mammalian cells, frog embryos, and zebrafish," he explained. "These are all fine, except when we hit intended targets," in which case, effects associated with inhibition of that target appear.

Meggers' collaborator Peter S. Klein, at Penn's Department of Medicine, tested the effect of an inhibitor of glycogen synthase kinase 3 on frog embryos. GSK-3 plays a role in what is known as the Wnt signaling pathway, which is involved in the formation of a single body axis in frog embryos, Meggers explained. A second body axis develops when GSK-3 is inhibited. When the GSK-3-inhibiting ruthenium complex was injected into a developing frog embryo, the embryo developed into a two-headed tadpole.

That experiment, Meggers noted, verifies that the metal complex works in a biological environment with no obvious effect on cells or organisms other than what is expected. This point is important because metal-containing compounds for therapeutics are perceived to be high risk. Although Meggers is not necessarily looking for therapeutic agents, he believes metal complexes with defined shapes have therapeutic potential.

"There is no reason not to think of using our metal compounds to treat life-threatening disease," Meggers told C&EN. "Take cancer therapy. One of the most widely used anticancer drugs is cisplatin, a nasty and toxic metal compound. It has two chloride leaving groups, and they react with everything. Our compounds are not that nasty, because they have no leaving groups."

Furthermore, Meggers continued, many kinases are involved in cancer. Earlier this year, Penn filed a patent on the strategy to inhibit protein kinases with metal compounds.

Meggers also pointed out that the strategy of morphing natural products into metal complexes is validated by its inclusion in Penn's Center for Molecular Discovery, part of the National Institutes of Health's initiative to support discovery of molecular probes. At the moment, his group participates by taking initial hits from the screening of organic-compound libraries and improving their properties by transforming them into rigid metal complexes.

The Meggers lab will be looking at other complicated natural products, such as paclitaxel, that might be mimicked by metal complexes. "We think we can make derivatives that are at least equally effective but much simpler," he told C&EN. "That would be enormously exciting."