Genome Mining for Natural Products

I've discovered a lot of interesting science in ChemBioChem, a European chemical biology journal. We have more access to more of these journals than most of the Nature journals; it must be because there is some relatively cheap Wiley journal package that comes with Angewandte Chemie, since we basically get all of the European chemistry journals that are published by that group.

But in any case, one pet area that has been fascinating me lately is the convergence of genetics, microbiology, analytical chemistry, and synthetic chemistry to the discovery of new natural products, especially in examining modular biosynthetic polyketide genes. This review that came out a few days ago, Strategies for Discovering New Natural Products through Genome Mining, went through basically what is being done in the field.

The authors explain that natural products discovery has not been favored by pharmaceutical industries for a while as a way to discover new drugs. Most of the 20th century (after penicillin was discovered) is described as the golden age of antibiotics and natural product discovery. It has fallen out of favor because it is time intensive and often leads to discovery of natural products that are already known.

But now, in a post genomics era with so much bioinformatic data out there, we actually have new strategies to approach discovering natural products, especially in examining biosynthetic gene clusters. We can search for PKS (polyketide synthase) domains, we can knockdown genes, and (most exciting to me) you can reconstruct products in vitro to determine if that gene cluster is actually making what you think it is (and then confirm what you have with NMR and MS). Of course, with the reconstruction strategy there are certain problems, like finding the right promoter and so forth, but chemoenzymatic total synthesis to reconstruct biosynthetic pathways has been done in vitro before.

Polyketides are neat from a pharmacological standpoint, because the secondary metabolites tend to have a lot of medicinal properties. They are also neat in that a lot of the genes are modular, so there have been some papers where they re-arrange them essentially doing synthetic chemistry in a cell instead of in an RBF. And sometimes they have neat little genes beyond just PKS genes--I recently wrote a paper for school reviewing research on a gene that was responsible for doing a powerful oxidation to form a chiral THF ring. It seems like a way to both discover new natural products with pharmaceutical value and to generally get a better mechanistic grasp of biosynthesis.

Incorperation of Bioenzymatic Synthesis in Total Synthesis of Natural Products

So I think I've found an area of biological chemistry that really really interests me: using enzymes in total synthesis. It seems to be a relatively new tool to pull out of the bag of synthetic tricks, but I have found papers that use a combo of bioenzymatic snytheses and traditional synthetic organic reactions (in this case, a Stille coupling), as well as a total enzymatic synthesis of this particular set of bacterial products that seem to slow down tumor formation. I'm going to write a paper for biochem about this (focusing on the biochem side instead of the synthetic side...i.e. the relevance of products and why this particular enzymatic transformation is synthetically useful as opposed to the synthetic methodology aspect...but both catch my interest). But it's really handy because enzymes do all sorts of crazy shit that we don't do well...like stereochemical control and massively concerted reactions...and additionally are pretty green (i.e. don't require large amounts of heavy metals that are a bitch to dispose of). They can only be used for very specialized transformations, so there's a lot of work that still needs to be done in this field. Also, they can be a bit of a bitch to purify, but then again, organic reactions in general can be a bitch to purify.

This seems like an area that satisfies the synthetic chemist and the biochemist in me.