UC Berkeley

Lactam is an important class of commodity chemicals used in the manufacture of nylons, with millions of tons of production every year.

2-Pyrrolidone is a valuable bulk chemical with myriad applications as a solvent, polymer precursor and active pharmaceutical intermediate. A novel 2-pyrrolidone synthase, ORF27, from Streptomyces aizunensis was identified to catalyze the ring closing dehydration of γ-aminobutyrate. ORF27’s tendency to aggregate was resolved by expression at low temperature and fusion to the maltose binding protein (MBP). Recombinant Escherichia coli was metabolically engineered for the production of 2-pyrrolidone from glutamate by expressing both the genes encoding GadB, a glutamate decarboxylase, and ORF27. Incorporation of a GadB mutant lacking H465 and T466, GadB_ΔHT, improved the efficiency of one-pot 2-pyrrolidone biosynthesis in vivo. When the recombinant E. coli strain expressing the E. coli GadB_ΔHT mutant and the ORF27-MBP fusion was cultured in ZYM-5052 medium containing 9 g/L of L-glutamate, 7.7 g/L of L-glutamate was converted to 1.1 g/L of 2-pyrrolidone within 31 h, achieving 25% molar yield from the consumed substrate.

Caprolactam and valerolactam are important commodity chemicals used in the manufacture of nylons, with millions of tons produced every year. Biological production of these high valued chemicals has not been possible due to a lack of enzymes that will cyclize the ω-amino fatty acid precursors to the corresponding lactams under ambient conditions. In this study, we demonstrated proof of these bioconversions by in vitro enzyme assays. We found that ORF27, an enzyme involved in the biosynthesis of ECO-02301 in Streptomyces aizunensis, has a broad substrate spectrum and can not only cyclize γ-aminobutyric acid into butyrolactam, but also 5-aminovaleric acid (5-AVA) into valerolactam and 6-aminohexanoic acid (6-AHA) into caprolactam. The ORF27 lactam formation reaction was characterized by product analysis, and ORF27’s activity on the three ω-amino fatty acids were compared. Recombinant E. coli expressing ORF27 produced valerolactam and caprolactam when 5-AVA and 5-AHA, respectively, were added to the culture medium. Upon co-expressing ORF27 with a metabolic pathway that produced 5-aminovaleric acid from lysine, we were able to demonstrate production of -valerolactam from lysine or directly from glucose.

Biological production of caprolactam, valerolactam and butyrolactam were enabled by the recent discovery of lactam synthases that cyclize their ω-amino fatty acid precursors to the corresponding lactams. To facilitate strain optimization with regard to product yields, productivities, and titers, it is desirable to develop a high throughput screening system forspecific small-molecule detection and response. We took the chemical informatics concepts used in small molecule drug discovery, and adapted them into a metabolic engineering strategy for targeted scouting of protein sensor candidates, named “Analog Generation towards Catabolizable Chemicals”. We discovered a lactam biosensor based on the ChnR/Pb transcription factor-promoter pair. The microbial biosensor was engineered to be a single plasmid system, demonstrating dose dependent response for caprolactam, valerolactam and butyrolactam with great dynamic range (1.8-3.5) and wide linear range of 1-2 orders of magnitude. The biosensor also showed specificity against intermediates of lactam biosynthetic pathways and therefore could potentially be applied for high throughput metabolic engineering for industrially important lactam biosynthesis.