Leveraging synthetic biology for producing bioactive polyketides and non-ribosomal peptides in bacterial heterologous hosts

Abstract

Bacteria have historically been a rich source of natural products (e.g. polyketides and non-ribosomal peptides) that possess medically-relevant activities. Despite extensive discovery programs in both industry and academia, a plethora of biosynthetic pathways remain uncharacterized and the corresponding molecular products untested for potential bioactivities. This knowledge gap comes in part from the fact that many putative natural product producers have not been cultured in conventional laboratory settings in which the corresponding products are produced at detectable levels. Next-generation sequencing technologies are further increasing the knowledge gap by obtaining metagenomic sequence information from complex communities where production of the desired compound cannot be isolated in the laboratory. For these reasons, many groups are turning to synthetic biology to produce putative natural products in heterologous hosts. This strategy depends on the ability to heterologously express putative biosynthetic gene clusters and produce relevant quantities of the corresponding products. Actinobacteria remain the most abundant source of natural products and the most promising heterologous hosts for natural product discovery and production. However, researchers are discovering more natural products from other groups of bacteria, such as myxobacteria and cyanobacteria. Therefore, phylogenetically similar heterologous hosts have become promising candidates for synthesizing these novel molecules. The downside of working with these microbes is the lack of well-characterized genetic tools for optimizing expression of gene clusters and product titers. This review examines heterologous expression of natural product gene clusters in terms of the motivations for this research, the traits desired in an ideal host, tools available to the field, and a survey of recent progress.

Fig. 1. Strategies for discovering novel natural products using heterologous expression of BGCs.

Fig. 2. Synthetic biology tools and physiological traits required for production of polyketides and non-ribosomal peptides from a heterologous host.

Fig. 3. Trends and statistics impacting heterologous expression of BGCs in bacteria. a) Number of publications reporting heterologous expression of polyketide and non-ribosomal peptide BGCs per year. b) Number of sequenced BGCs encoding PKSs and/or NRPSs in the AntiSmash database. Data labels represent value on bar chart. All groups represent a bacterial class, except for the phylum cyanobacteria. Values were determined by building a query on the AntiSmash database searching for clusters of the types “nrps” and/or “pks” for each taxonomic group of interest. The AntiSmash database was accessed on January 6, 2019. c) Number of publications reporting heterologous expression of BGCs according to heterologous host and phylogenetic distance between the native host and heterologous host. Data labels represent total number of publications. d) Number of publications reporting heterologous expression of BGCs according to heterologous host and publication purpose. Data in panels C and D are limited to articles published from 2013 to 2019. Publications were identified via advanced searches in PubMed and Web of Science and supplemented by a manual literature review. The PubMed query was “(((heterologous[Title/Abstract] AND (expression[Title/Abstract] OR production)[Title/Abstract])) AND (polyketide OR nonribosomal peptide OR non-ribosomal peptide)) NOT (fungi[Title/Abstract] OR fungal[Title/Abstract] OR plant[Title/Abstract])”. The Web of Science query was “ALL = (heterologous AND (expression OR production)) AND ALL = (polyketide OR nonribosomal peptide OR non-ribosomal peptide) NOT TS = (fungal OR fungi OR plant)”. Irrelevant publications were manually removed from the search results.

Fig. 4. Structures of select bioactive polyketides and non-ribosomal peptides produced by heterologous hosts.