A translational synthetic biology platform for rapid access to gram-scale quantities of novel drug-like molecules

Abstract

Plants are an excellent source of drug leads. However availability is limited by access to source species, low abundance and recalcitrance to chemical synthesis. Although plant genomics is yielding a wealth of genes for natural product biosynthesis, the translation of this genetic information into small molecules for evaluation as drug leads represents a major bottleneck. For example, the yeast platform for artemisinic acid production is estimated to have taken >150 person years to develop. Here we demonstrate the power of plant transient transfection technology for rapid, scalable biosynthesis and isolation of triterpenes, one of the largest and most structurally diverse families of plant natural products. Using pathway engineering and improved agro-infiltration methodology we are able to generate gram-scale quantities of purified triterpene in just a few weeks. In contrast to heterologous expression in microbes, this system does not depend on re-engineering of the host. We next exploit agro-infection for quick and easy combinatorial biosynthesis without the need for generation of multi-gene constructs, so affording an easy entrée to suites of molecules, some new-to-nature, that are recalcitrant to chemical synthesis. We use this platform to purify a suite of bespoke triterpene analogs and demonstrate differences in anti-proliferative and anti-inflammatory activity in bioassays, providing proof of concept of this system for accessing and evaluating medicinally important bioactives. Together with new genome mining algorithms for plant pathway discovery and advances in plant synthetic biology, this advance provides new routes to synthesize and access previously inaccessible natural products and analogs and has the potential to reinvigorate drug discovery pipelines.

Keywords: Combinatorial biosynthesis; Drug discovery; Synthetic biology; Terpenes; Transient plant expression technology; Triterpenoids.

Fig. 1

Co-expression with tHMGR gives enhanced triterpene levels a, Biosynthesis of triterpenes occurs via the mevalonate pathway. b, β-Amyrin content of tobacco leaves co-expressing SAD1 β-amyrin synthase with GFP, tHMGR, FPS, SQS or SQE (mean, three biological replicates ± s.e; control, GFP only); *, P < 0.05; n.s., not significant. c, Total ion chromatograms (TICs) for extracts from leaves expressing SAD1 with either GFP or tHMGR. d, Oxygenation of the β-amyrin scaffold to 12,13-epoxy, 16-hydroxy-β-amyrin (EpHβA) by CYP51H10. e, EpHβA content of leaves expressing SAD1 and CYP51H10 with GFP or tHMGR (mean, three biological replicates ± s.e.); ***, P < 0.0001. f, Total ion chromatograms (TICs) for the data shown in e. IS, internal standard (coprostan-3-ol).

Fig. 2

Generation of gram quantities of triterpene using vacuum infiltration a, Vacuum infiltration of N. benthamiana plants. Plants are retained by a bespoke holder, inverted into a bath containing 10 L of A. tumefaciens suspension, and a vacuum applied. Upon release of the vacuum the infiltration process is complete. b, GFP expression in leaves from a vacuum-infiltrated plant 5 days after infiltration (leaves arranged from top left to bottom right in descending order of their height on the plant). The youngest leaves (top left) were formed post-infiltration. c, β-Amyrin purified from vacuum-infiltrated plants following transient expression.

Fig. 3

Combinatorial biosynthesis of oxygenated triterpenes in N. benthamiana. The major products of combinatorial biosynthesis are shown in the outer circle. The structures of compounds 11–17 were determined by NMR following initial GC-MS analysis (Figs S8-13; Tables S1-8). Structures 7–10 were inferred by GC-MS, based on comparison with previously reported compounds (Table S1; Figs S4-7). GC-MS also revealed a further 30 new unidentified products across the different enzyme combinations (Table S1). Similar results were obtained when β-amyrin synthase was co-expressed with the different CYP combinations in yeast (Fig. S14).

Fig. 4

Evaluation of biological activity. a, Compounds used. b, Anti-proliferative assays. IC₅₀ values (µM) for the human cancer cell line HL60 are shown (means, three biological replicates ± s.d.). c, Anti-inflammatory activity. TNFα release was measured by ELISA following lipopolysaccharide stimulation of THP-1 cells. Triterpenes were used at 100 μM, with the exception of bardoxolone methyl (100 nM). Values are relative to control DMSO-treated cells (means, three biological replicates ± s.d.). *, P < 0.05; n.s., not significant.