Wender & Miller on molecular frontiers

16 July 2009 - The future

Paul Wender and Benjamin Miller in a recent contribution to Nature (DOI) discuss the future of organic synthesis. Here are some of the keywords that sum the article up. Bioinspired synthesis: many of the most important therapeutic drugs in the last 50 years were inspired by biomolecules. Natural products that can only be found in trace amounts can be artificially synthesised in large quantities provided that the so-called step-economy (minimization of the number of chemical procedures in a synthesis) requirement is met. Classic example of offer: the synthesis of cyclooctatetraene from an alkaloid (11 steps and low yield) or from acetylene (1 step with high yield). Another example is Wender's own prostratin semisynthesis. Key strategies for step-economy according to Wender & Miller: the multicomponent reaction and the cascade reaction. And when total synthesis fails to offer an advantage over harvesting from nature then at least the synthesis of analogs should open up the road to new drugs. Example on offer: the laboratory synthesis of bryostatin takes 70 steps in comparison to 30 steps for a pimped-down version.

But hey, wait a minute! In a Science review published just a day later Jesse Li and John Vederas also explore a frontier, that of drug discovery and natural products (DOI). They note that in drug companies, natural product research itself actually is in decline! Although the success rate for natural products (0.3%) is actually higher than that of high-throughput screening (HTS) (0.001%), HTS is all the rage nowadays because all-synthetic chemical libraries are easily available for example with diversity oriented synthesis. Performing HTS on natural sources such as plant extracts has the added disadvantages of intellectual property issues (the plant may be owned by the locals), availability issues (a rare plant may have gone extinct by the time R&D is ready for the next step) and an active ingredient may actually turn out to be a non-patentable known compound (a waste of effort). So what to do!. Li & Vederas list some creative ideas in the field: the creation of genomic libraries based on metagenomics - the automation of biomolecule separation prior to HTS - the automation of structure elucidation e.g. tandem HPLC-MS or even NMR. The generation of analogs is also mentioned: compare for example biomolecule lovastatin with analog zocor. And even more fancy strategies to look out for in the future: Single Molecule Real Time Sequencing, synthetic biology and 454 Sequencing.