Rapid pathway prototyping and engineering using in vitro and in vivo synthetic genome SCRaMbLE-in methods

Abstract

Exogenous pathway optimization and chassis engineering are two crucial methods for heterologous pathway expression. The two methods are normally carried out step-wise and in a trial-and-error manner. Here we report a recombinase-based combinatorial method (termed "SCRaMbLE-in") to tackle both challenges simultaneously. SCRaMbLE-in includes an in vitro recombinase toolkit to rapidly prototype and diversify gene expression at the pathway level and an in vivo genome reshuffling system to integrate assembled pathways into the synthetic yeast genome while combinatorially causing massive genome rearrangements in the host chassis. A set of loxP mutant pairs was identified to maximize the efficiency of the in vitro diversification. Exemplar pathways of β-carotene and violacein were successfully assembled, diversified, and integrated using this SCRaMbLE-in method. High-throughput sequencing was performed on selected engineered strains to reveal the resulting genotype-to-phenotype relationships. The SCRaMbLE-in method proves to be a rapid, efficient, and universal method to fast track the cycle of engineering biology.

Fig. 1

Schematic overview of the SCRaMbLE-in toolkit for metabolic engineering. Once designed, the synthetic pathway of interest will be rapidly prototyped in vitro. The pathway is pre-assembled in such a way that all genes are assembled with respective regulatory elements, except that one key gene is left “promoter-less”, having a recombination site upstream. Three recombination systems, Cre-loxP, VCre-Vlox, and Dre-rox, were developed to integrate a range of promoters upstream to the key gene in the pathway. The assembled library is transformed into yeast and potentially productive pathways are selected and integrated in vivo in the next step. The selected productive pathways are flanked with a pair of loxPsym sites and can subsequently be integrated into the synthetic chromosomes through SCRaMbLE, and simultaneously the synthetic yeast chassis undergoes whole-genome rearrangement to result in strain optimization. The successful integrated strains will be profiled with mass spectrometry analysis and next-generation sequencing

Fig. 2

Integration rate improvement by LE/RE strategy. a LE/RE strategy was used to improve the integration efficiency. The functional domains of a recombination site are divided as a spacer region and two palindromic regions. The recombination between LE and RE mutation site generates a LE:RE double-mutant site and a wild-type loxP site. For in vitro application, a genetic Element of Interest (EoI) flanked by two LE sites is first excised out, and then integrates into an RE site, finally being flanked by a LE/RE site and a loxP site. b Excision rate quantification of palindromic mutant site pairs (LE–LE or RE–RE) and double-mutant site and loxP site (LE:RE-loxP). LE includes lox71, loxJT15, loxJT510, loxJTZ17LE; RE includes lox66, loxJT15RE, loxJT510 RE, loxJTZ17. Error bar represents the standard deviation, n = 3. c Excision rate comparison between same pair recombination group and LE:RE–WT pair recombination group. Center value represents the mean of the excision rates. t test was performed for difference evaluation between the two groups. d Integration rate quantification of selected LE- and RE-mutant sites. The integration rate between loxJT15 and loxJTZ17 is three-fold of that between loxP site. Error bar represents the standard deviation and P value was generated by t test, n = 3. LE, left element with mutation in left palindromic region, RE, right element with mutation in right palindromic region

Fig. 3

In vitro pathway prototying of β-carotene and violacein pathways. a Illustration of gene function in β-carotene synthesis pathway. CrtI was chosen as the target gene for integration regulation in β-carotene pathway. CrtI encodes the desaturase that converts the colorless phytoene to the yellow neurosporene first and then to the red lycopene through four desaturation reactions. Since it is a key enzyme in catalyzing the colorless intermediates to colored molecules, the extent of yellow, red, or orange can indicate the transformation efficiency and visibly display diversification of the pathway with various expression of CrtI. b Promoter integration confirmation and LC-MS quantification of carotenoids in β-carotene pathway. VCre-Vlox system was used for CrtI regulation. Seven promoter-integrated strains were quantified. Error bar represents the standard deviation, n = 3. LC-MS, liquid chromatography and mass spectrometry. c Illustration of gene function in violacein synthesis pathway. VioA was chosen as the target gene for integration regulation in violacein pathway. VioA encodes the flavoenzyme L-tryptophan oxidase that catalyses the incorporation of two molecules of substrate L-tryptophan into indole-3-pyruvic acid imine. It catalyzes the initial step in the violacein synthesis and is important for transforming enough tryptophan substrates for following steps towards full synthesis of the pathway. d Promoter integration confirmation and HPLC quantification of violacein. Both VCre-Vlox system and Cre-loxJT15-loxJTZ17 system were used for VioA regulation. Six promoter-integrated strains were quantified. Error bar represents the standard deviation, n = 3. HPLC, high performance liquid chromatography

Fig. 4

Induction and screening of pathway SCRaMbLE-in yeast variants. a A URA3-based counter selection strategy to facilitate the SCRaMbLE-in process. The SCRaMbLE-in device is based on a yeast centromeric plasmid with URA3 marker (pRS416). A BsaI site-flanked RFP cassette and a LEU2 expression cassette were placed between two loxPsym sites. LEU2 is used as positive selection marker for integration and URA3 is used as counter selection marker for non-integrated strains. After SCRaMbLE-in induction, successful integrated colonies were selected on SC-Leu + 5-FOA plates. b Violacein quantification in the violacein pathway SCRaMbLE-in variants. LWy137 is control strain with a single-copy violacein pathway inserted at the HO locus. LWy152, LWy238, and LWy239 are violacein SCRaMbLE-in strains. Error bar represents the standard deviation, n = 3. c β-carotene quantification in β-carotene pathway SCRaMbLE-in variants. LWy212 is control strain with a single-copy β-carotene pathway inserted at the HO locus. LWy215, LWy252, and LWy253 are β-carotene SCRaMbLE-in strains. Error bar represents the standard deviation, n = 3. d Color comparison of continuous SCRaMbLEd variants with violacein integrated pathway. Both darker colony color and lighter colony color were observed for the three SCRaMbLE-in strains with continuous SCRaMbLE. LWy152, LWy238, and LWy239 are SCRaMbLE-in strains; LWy256 and LWy257 are continuous SCRaMbLEd strains from LWy152; LWy258 and LWy259 are continuous SCRaMbLEd strains from LWy238; LWy260 and LWy261 are continuous SCRaMbLEd strains from LWy239. e Violacein quantification of continuous SCRaMbLEd variants with violacein integrated pathway. 152+: increased production compared with source strain LWy152; 152−: decreased production compared with source strain LWy152. Error bar represents the standard deviation, n = 3

Fig. 5

Next-generation sequencing result of violacein SCRaMbLE-in strains. a Genome rearrangements in SCRaMbLE-in strains. Each SCRaMbLEd synII chromosome (770 kb) is represented as a sequence of arrows. The color of each arrow represents the segment number separated by loxPsym sites in the parental synII chromosome and the direction indicates the orientation. The unfilled arrow indicates exogenous violacein synthesis pathway. The red-outlined arrow represents the segment where the centromere is located. Based on the SCRaMbLE events, hierarchical clustering method was performed to cluster the strains to illustrate the component differences on the position and length. Remarkably a chromosome fusion event was observed in LWy256 and the total length of SCRaMbLEd synII is 1544.17 kb. b Large fragment duplications in the SCRaMbLE-in strains. Segment copy number in each strain is indicated as deletion and no copy (gray), one copy with no change (white), duplication as two copies (orange), and multiple duplications with more than three copies (red). The asterisk represents a CEN2 centromere resides in segment 8

Fig. 6

Orthogonal inducible SCRaMbLE systems. a Design of SSR expression device and functional test device. SSR or SSR-LBD fusion protein were assembled into a pRS415-based centromeric vector with LEU2 as selection marker. URA3 marker was used as recombination reporter and flanked by two recombination sites. The reporter device was based on pRS413 centromeric vector with HIS3 as selection marker. b Orthogonal function test of Cre and Dre. Yeast cells with reporter device rox-URA3-rox did not survive with the expression of Dre and cells with reporter device loxP-URA3-loxP could not survive with the expression of Cre on SC-His-Leu-Ura plate. c Orthogonal induction test of CrePBD and DreEBD. Only cells with reporter device loxP-URA3-loxP and CrePBD under the induction of RU486, cells with reporter device rox-URA3-rox and Dre-EBD under the induction of β-estradiol did not survive on SC-His-Leu-Ura plate. PBD progesterone-binding domain, EBD estrogen-binding domain, E+ β-estradiol induction, R+ RU486 induction, SSR site-specific recombinase, LBD ligand-binding domain