PRECURSOR-DIRECTED BIOSYNTHESIS OF POLYKETIDE AND NONRIBOSOMAL PEPTIDE NATURAL PRODUCTS - Modern Tools for the Synthesis of Complex Bioactive Molecules - page 487

nonribosomal peptides. Introduction of unnatural priming and extender molecules
offers the advantage of getting complex unnatural molecules with little or no synthetic
effort. Scale-up of the target molecule is straightforward. However, the approach
offers limited opportunities for variation in the final structure. Moreover, the number
of enzymes encountered by these unnatural precursors is large, and the probability
that processing is hindered by enzyme discrimination is significant. Introduction of
the unnatural precursor later along the biosynthetic pathway allows the loading of
more promiscuous enzymes and offers more opportunities for structural divergence
from the wild-type compound. However, this comes at the cost of requiring synthet-
icallymore complexmolecules, and product scale-up is challenging. Judicious choice
of how complex a precursor molecule to use for a desired application is central to
exploiting the full power of this technique for the production of analogues of clinically
relevant natural products. So central is this choice to the technique that the final
section of this chapter containing examples of the successful application of precursor-
directed biosynthesis is split into three parts, exploiting precursors of low, inter-
mediate, and high complexity.
14.3.3. Host Selection and Biosynthetic System Engineering
In an ideal system, a precursor molecule could simply be fed to the fast growing and
easy to culture native host. The precursor would efficiently cross the cell membrane
and a few days later, grams of the desired natural product analogue would be isolated
per liter of culture. This is, of course, rarely the case, as there are many factors to
consider when selecting a host organism for precursor-directed biosynthesis and how
to engineer that system for reasonable titers and clean product profiles.
Clinically relevant natural products have been isolated from a wide variety of
organisms, both prokaryotes and eukaryotes. When considering the use of a precur-
sor-directed approach, one of the first decisions that must be made is that of where to
establish the system. This question has three possible answers, each with significant
advantages: use the native host, heterologously express the system in a nonnative host,
or use purified enzymes in an in vitro system.
The first of these choices, the use of the native host, is the simplest from the
standpoint of initial setup. No knowledge of the genes responsible for the production
of the biosynthetic system is necessary, and feeding an unnatural precursor to the
natural product producing microorganism results in the desired analogue being
isolated in moderate yield. No genetic manipulation of the host is required. If the
biosynthetic gene cluster has been cloned and characterized, it is, of course, possible
to engineer the biosynthetic pathway through mutation of the chromosomal DNA of
the host organism. In practice, however, molecular biology tools are often extremely
limited in the context of most natural product producers. Therefore, if extensive
genetic engineering is to be undertaken, it is desirable to express the biosynthetic
gene cluster in a genetically tractable heterologous host. Examples of heterologous
hosts that have been used for precursor-directed biosynthesis include Streptomyces
coelicolor and Escherichia coli .
Heterologous expression of a biosynthetic gene cluster in a model host requires
considerably more initial effort than simply using the native host. The complete gene