Construction and Characterization of a Refactored Actinorhodin Biosynthetic Cluster

Next-generation sequencing technologies have enabled single genomes as well as complex environmental samples (metagenomes) to be sequenced at an ever-increasing pace. Advanced bioinformatics analysis of the resulting genomics and transcriptomics data is revealing an unprecedented catalogue of proteins, and pathways that encode a myriad of novel catalytic activities and a vast diversity of novel compounds of potential value for environmental and biomedical applications.

Characterizing the products of predicted biosynthetic pathways remains challenging as (i) Clusters originate from diverse organisms many of which are difficult to culture and manipulate, (ii) Biosynthetic clusters are often under transcriptional or translational repression by unknown regulatory mechanisms, (iii) Biosynthetic clusters are often large and complex in terms of sequence composition, (iv) Molecule production may require the presence of unknown substrates or co-factors.

In order to systematically characterize the products of biosynthetic clusters we are building a scalable pipeline for pathway refactoring which includes automated design, efficient DNA construction and assembly, sequence verification, strain generation and compound detection.

As a proof-of-principle for our pipeline we designed, built and tested a fully refactored actinorhodin pathway rACT, which was synthesized from scratch using synthetic promoters and terminators, in-silico calculated RBSs, and optimized coding sequences. In total rACT comprised 22 Kb of synthetic sequence, which was organized as 5 contiguous operons that did not mimic the natural cluster architecture in terms of orientation, order or operon composition.  The rACT DNA design was carefully screened for sequence features that might compromise synthesis efficiency and any such features present in the original design were removed by codon juggling. The resulting ‘polished’ DNA design was partitioned to optimize the assembly junctions, and synthesized using standard methods. Construct assembly was performed using a two-step process involving Gibson followed by yeast Gap-repair cloning. The integrity and sequence composition of rACT was verified using Pacific Biosciences sequencing technology.

The rACT cluster was integrated into a Streptomyces coelicolor strain lacking the native actinorhodin cluster  (delACT strain).  The resulting rACT strain was cultured and compared to the delACT strain as well as to wildtype S. Coelicolor. Samples were collected at days 2, 4 and 6 of culture in triplicates and subjected to transcriptomics, proteomics and metabolomics analyses.  Actinorhodin as well as all of its metabolic intermediates were detected by LC-MS in the rACT and wildtype strains but not in the delACT strain. Interestingly, actinorhodin production was detected earlier in the rACT strain than in the wildtype strain showing a clear de-coupling of growth and molecule production.

Our design strategy, build pipeline and full OMICS analysis will be presented in detail as we seek to further refine and scale our genomes to molecules pipeline.