Genomic and metagenomic approaches to natural product chemistry

For many years, natural products have been a primary source of new molecules for the treatment of disease, and microorganisms have been a prolific source of these molecules. Recent studies have indicated, however, that many biosynthetic pathways are present in organisms for which no natural product can be associated, and only a small fraction of the microbial life present in the environment can be grown in culture. This indicates that if methods could be developed for the isolation of these pathways and production of their target molecules in heterologous hosts, great numbers of potentially valuable compounds might be discovered. In these investigations, large insert libraries of two microorganisms were constructed, one a bacterial artificial chromosome (BAC) library, the other a fosmid library, and two large insert fosmid libraries were constructed with DNA isolated from marine environmental samples. A mathematical formula was derived to estimate probabilities of cloning intact biosynthetic pathways with large insert genomic libraries and tested with a computer simulation. This indicated that even large pathways could be cloned intact in large insert libraries, provided there was an adequate size difference between the target pathway and the library inserts, and there was a concomitant increase in the size of the library with the targeting of these larger pathways. In addition, an investigation into a mixed marine culture sample lead to the identification of an unusual relationship between two bacteria for which extended co-culture leads to the production of pyocyanin. However, no useful biosynthetic pathways were located within the genomic libraries. It is concluded that significant improvements would be required to make this approach feasible for larger scale investigations. It is further concluded, on the basis of recent developments in the field, including a reduction in the cost of sequencing, improvements in techniques of whole-genome shotgun sequencing, and the development of recombination based cloning, that the employment of mass sequencing efforts and sequence-driven, recombinationbased cloning, might prove to be a more fruitful and efficient alternative to large-insert library construction for the isolation and expression of these pathways. A possible paradigm for the cloning of pathways on the basis of this technology is proposed.