Metabolic engineering of "last line antibiotic" colistin in Paenibacillus polymyxa

Abstract

Colistin, also known as polymyxin E, is an antibiotic considered as last-line treatment for infections caused by multidrug-resistant gram-negative bacteria. The clinical application of colistin has high standards on its analog ratio and purity, which poses a high challenge to its production and downstream isolation process. Recently, with the advances in synthetic biology techniques, rational metabolic engineering of colistin natural producers is becoming a promising strategy to address this challenge. This study aimed to perform rational metabolic engineering for colistin production in Paenibacillus polymyxa ATCC 842 to construct an ideal host for the efficient production of colistin with a clean background. In the first section, we analyzed the colistin analogs produced by colistin natural producer strain P. polymyxa ATCC 842 and inactivated its non-essential BGCs. The primary colistin analogs produced by P. polymyxa ATCC 842 were identified as colistin A and colistin B, and the ratio of the analogs complies with the British Pharmacopoeia and European Pharmacopoeia standards for pharmaceutical colistin. Subsequently, by respective inactivation of 8 non-essential biosynthetic gene clusters (BGCs) via CRISPR/Cas9-based gene editing in the chassis strain, we found that the mutant P3 with inactivation of BGC 4, an NRPS-PKS BGC, could dramatically improve secondary metabolites production, which increased the colistin titer by 79% to 315.4mg/L. Additionally, through multiple inactivation of 7 non-essential BGCs, an engineered strain P14 with a clean metabolic background was generated. In the second section, we further improved the chassis strain's colistin titer by strengthening the precursor supply. Nine promoters in P. polymyxa ATCC 842 were characterized, and the promoter PH4 was applied to strengthen the expression of the genes involved in the L-2,4-DABA pathway. By co-overexpression of diaminobutyrate-2-oxoglutarate transaminase ectB and aspartate kinase lysC, an engineered strain P19 with a 649.3 mg/L colistin titer, which was 269% higher than the starting strain, was obtained. Finally, in 5 L bioreactor continuous fermentation, the final colistin titer achieved 1711.7 mg/L. This study represents the pioneering case of metabolic engineering for colistin production in P. polymyxa, its natural producer. We successfully developed an engineered strain, P19, with high colistin production and a clean metabolite background. These findings hold promising implications for advancing the metabolic engineering of NRPS products in non-model Bacillota strains.