Genetic Engineering of Chromobacterium Vaccinii DSM 25150 for Improved Production of FR900359

The depsipeptide FR900359 has been first described in literature in 1988 (Fujioka et al, 1988) to be isolated from a methanol extract of the whole plant of Ardisia crenata. FR900359 can be isolated from the leaves of A. crenata, but the very low quantities and the complex matrix prevent access to sufficient amounts of FR900359 to enable drug development efforts and potential commercial manufacturing. Almost two decades later, it has been discovered that FR900359 is in fact produced by a strictly obligate bacterial endosymbiont, Candidatus Burkholderia crenata, of the plant Ardisia crenata (Carlier et al, 2016). This study identified also the DNA sequence of the biosynthetic gene cluster (BGC) of FR900359. In order to identify alternative and scalable methods for production of FR900359, a genome mining effort on bacterial genomes from both public sequence databases and genome sequences generated from internal efforts at Novartis was initiated. Translated amino acid sequences of the FR900359‑BGC from Candidatus B. crenata were used as query sequence. While the query of public sequence databases did not return highly similar sequences, a gene cluster with very high homology in translated amino acid sequence and identical prediction of protein functions was discovered in the genome of Chromobacterium vaccinii DSM 25150, which had been sequenced internally at Novartis. Here we describe the genetic engineering of Chromobacterium vaccinii DSM 25150 resulting in mutants that exhibit improved production of FR900359 and improved characteristics concerning downstream processing and purification.