A robust gene-stacking method utilizing yeast assembly for plant synthetic biology

Abstract

The advent and growth of synthetic biology has demonstrated its potential as a promising avenue of research to address many societal needs. However, plant synthetic biology efforts have been hampered by a dearth of DNA part libraries, versatile transformation vectors and efficient assembly strategies. Here, we describe a versatile system (named jStack) utilizing yeast homologous recombination to efficiently assemble DNA into plant transformation vectors. We demonstrate how this method can facilitate pathway engineering of molecules of pharmaceutical interest, production of potential biofuels and shuffling of disease-resistance traits between crop species. Our approach provides a powerful alternative to conventional strategies for stacking genes and traits to address many impending environmental and agricultural challenges.

Figure 1. jStack hierarchical assembly.

(a) A library of Level 0 parts composed of linkers (linker, L), promoters (Prom., P), coding sequences (CDS; C) and terminators (Term., T) are available. Level 1 assembly consists of digesting desired Level 0 parts along with their intermediate vector backbone (containing a complementary antibiotic selection; chloramphenicol, Cm) with a Type IIS restriction enzyme followed by ligation facilitated by compatible sticky ends. Type IIS enzymes generating compatible sticky ends are denoted by the tiny black arrows above Level 0 plasmids. A GFP dropout-cassette is cut out from the intermediate vector backbone allowing for green/white selection, where all the white colonies have correctly assembled Level 1 constructs (illustrated in top right grey box). Compatibility of liberated ends by the Type IIS restriction enzymes (illustrated with coloured box at the end of each DNA part) allows assembling of all DNA parts in the correct order. (b) Various Level 1 constructs with compatible linker and terminator sequences that have homology to one another are digested with one of three flanking rare Type II restriction enzyme cutsites (NotI, SbfI or AscI). These are all transformed into yeast with a PmlI linearized pYB vector, which loses the Ura3 dropout-cassette. Homology between linker and terminator sequences allows DNA assembly via in vivo yeast recombination in the correct order, circularizing the pYB vector. The loss of the Ura3 cassette and the presence of a Leu2 marker on the backbone confer the resistance to 5-FOA and autotrophy for leucine, respectively; thus, any colonies that grow on minimal SD-Leu supplemented with 5-FOA, have recombined the various DNA fragments together into the pYB vector.

Figure 2. Assembling multi-gene pathways into pYB vectors with jStack.

(a) Stacking various reporter genes together into one transfer DNA: β-glucuronidase (GUS), GFP, DsRed. Promoters driving each gene are labelled and denoted by black arrows in front of coding sequences. Stable soybean root transformant lines were generated via Agrobacterium rhizogenes. Transformed roots exhibited GUS activity and GFP and DsRed fluorescence. Scale bar represents 20 μm. (b) Optimizing engineered metabolic pathways of a potential biofuel with gene stacking. Stacking various upstream enzymes in conjunction with new organellar targeting increased Bisabolene yields in planta—values are labelled in the graph. By stacking upstream enzymes such as chloroplast-localized farnesyl pyrophosphate synthase (CpFPPS) and chloroplast-localized 1-Deoxy-D-xylulose 5-phosphate synthase (CpDXS) with either cytosolic or plastid-targeted bisabolene synthase (BisSyn and CpBisSyn, respectively), increased yields of bisabolene were observed. Error bars indicate standard deviation, n=3. Mevalonate (MVA), isopentenyl pyrophosphate (IPP), farnesyl pyrophosphate (FPP), pyruvate (Pyr), glyceraldehyde 3-phosphate (G3P). (c) Engineering heterologous bacterial metabolites into plants. The cytotoxic molecule violacein is found in various soil bacteria and requires five enzymes (VioA, VioB, VioC, VioD and VioE) for its biosynthesis from tryptophan. All the five genes were stacked via yeast assembly and infiltrated into Nicotiana benthamiana leaves. Violacein production was observed via LC-MS methods.

Figure 3. High-throughput assembly of R gene clusters and synthetic gene clusters.

Chromosomal positions of all nine R gene clusters (R1–R9) across the Arabidopsis genome that display the canonical head-to-head inverted tandem arrangement. Dotted lines represent the PCR amplicons that contain overlapping homologous regions to facilitate yeast homologous recombination into pYB vectors. The number of fragments chosen to assemble a whole gene cluster was dictated by the size of the different gene clusters. Below, assembly of a synthetic gene cluster fusing two R gene clusters together using jStack assembly. Dotted lines over the cartoon schematic of the synthetic gene cluster represent the overlapping PCR amplicons used in jStack assembly of the construct.