Exploring chemical space with organometallics makes complete sense; carbon-based molecules can only form linear, trigonal planar, or tetrahedral geometries, so why not explore elements that are pentavalent or hexavalent and can thus form unique bioactive scaffolds? In a recent lecture, Meggers pointed out that an asymmetric tetrahedral carbon can form 2 stereoisomers, but an octahedral center with six substituents can form 30 different stereoisomers. (For those non-believers, Meggers had actually drawn out all 30 different stereoisomers on a slide).
In their search for a octahedral carbon-substitute, Meggers and coworkers concentrated on ruthenium because of its low cost, low toxicity (in the II and III oxidation states), high stability, and synthetic tractability. Using the ATP-competitive protein kinase inhibitor staurosporine as the basis for a ruthenium ligand, a small library of complexes was synthesized (100 members total, with 12 different ligands overall). Screening against several kinases revealed that several of the ruthenium based inhibitors were quite potent, with IC50 values in the nanomolar range. What I find most amazing is the fact that the staurosporine-based pyridocarbazole ligand is 19,000 times less potent than the ruthenium complex containing the same ligand. Further application of this strategy has led to the discovery of highly selective protein kinase inhibitors (for Pim-1, GSK-3, MSK-1), some with picomolar binding constants. A crystal structure confirmed the initial hypothesis that the ruthenium center is not involved in any direct interactions with the protein; the metal center works to orient the organic ligands in a conformation that favors binding.