Spatial organization of heterologous metabolic system in vivo based on TALE

For years, prokaryotic hosts have been widely applied in bio-engineering. However, the confined in vivo enzyme clustering of heterologous metabolic pathways in these organisms often results in low local concentrations of enzymes and substrates, leading to a low productive efficacy. We developed a new method to accelerate a heterologous metabolic system by integrating a transcription activator-like effector (TALE)-based scaffold system into an Escherichia coli chassis. The binding abilities of the TALEs to the artificial DNA scaffold were measured through ChIP-PCR. The effect of the system was determined through a split GFP study and validated through the heterologous production of indole-3-acetic acid (IAA) by incorporating TALE-fused IAA biosynthetic enzymes in E. coli. To the best of our knowledge, we are the first to use the TALE system as a scaffold for the spatial organization of bacterial metabolism. This technique might be used to establish multi-enzymatic reaction programs in a prokaryotic chassis for various applications.

RNA hairpins linked with gRNA. Thus, the application of such a method is still limited. Conrado et al. reported a zinc-finger-based DNA scaffold system to improve catalytic efficiency, indicating a powerful DNA scaffold tool in vivo on the basis of the specific and direct combination between DNA and proteins 17 . However, given its simpler design, higher specificity, and lower toxicity compared with ZFN 18 , the TALEN technique might be more efficient and practical for the clustering of multi-enzymes in vivo on the basis of TALE-fused enzymes and their corresponding DNA scaffolds.
TALE is a family III effector in Xanthomonas that aids the pathogen in infecting various plant species 19 . Different TALEs share a similar domain structure that enables them to bind the genome of the host cell and act as transcriptional effectors. A certain amount (around 1.5-33.5) of tandem transcription activator-like [TAL] repeats, each able to recognize a specific DNA base pair, was determined in the central DNA-binding domain of TALEs 20,21 . Each TAL repeat contains 33-35 highly conserved amino acids, among which the residues at positions 12 and 13 (also known as repeat variable di-residues) confer DNA specificity. This structural characteristic allows the TALE to be utilized in protein-engineering applications. TALENs are created by physically fusing the TALE with the cleavage domain of FokI nucleases. These nucleases are extensively applied in prokaryotic and eukaryotic cells. Other methods for engineering TALEs as transcriptional effectors, chromatin regulators, imagers, and separators have also been reported [22][23][24][25] .
In this study, we generate a new method to increase the production of a heterogeneous metabolic system in prokaryotic cells on the basis of the DNA-binding characteristic of TALEs. This artificial TALE-DNA scaffold system can efficiently gather TALE-fused proteins/enzymes around the DNA scaffolds and enrich local enzyme concentrations. The exact function of this method in accelerating exogenous metabolic reactions was tested by a biosynthesis system of indole-3-acetic acid (IAA). IAA is a plant hormone that regulates various developmental and physiological processes 26 and increases plant protection from external stress 27 via the spatial organization of two enzymes, namely, tryptophan-2-mono-oxygenase (IAAM) and indole-3-acetimide hydrolase (IAAH). To the best of our knowledge, this study provides a novel method to hasten the heterologous metabolic system in prokaryotic cells. The first introduction of TALEs into artificial in vivo scaffold systems can enrich the application of the TALE technique and provide broad insights into the construction of new in vivo multi-enzymatic accelerators with high compatibility and extensibility.

Results
Evaluation of the binding ability of TALE-GFP1/2 to the plasmid DNA scaffold. ChIP-PCR analysis was conducted to evaluate whether the TALE-GFP1/2 fusion protein can effectively target the binding motifs (BMs) on the plasmid DNA scaffold. To perform the assay, the synthetic biological parts of the promoter Plac, the ribosome binding site (RBS), TALE1/2/3, split GFP1/2, the terminator, and the corresponding scaffold1/2 were assembled in order by using the BioBrick ™ standard assembly method and then inserted into the plasmid backbone of pSB1C3. The plasmids ( Supplementary Fig. S1a online) pT1-GFP1-S1, pT2-GFP2-S2, pT3-GFP2-S1, pT1-GFP1 (no-scaffold control), pT2-GFP2 (no-scaffold control), and pT3-GFP2 (no-scaffold control) were verified by sequencing and then transformed into the Escherichia coli chassis. The primers used for ChIP-PCR were forward P1 and reverse P2 for GFP1 amplification to detect TALE1-GFP1 binding and forward P3 and reverse P4 for GFP2 amplification to detect TALE2/3-GFP2 binding (Fig. 1a). To optimize the culture temperature during isopropylthio-β -galactoside (IPTG) induction, the TALE1-GFP1-Scaffold1 system was used as a representative for the ChIP-PCR assay. All of the culture temperatures (20 °C, 25 °C, and 30 °C) were suitable for TALE-GFP expression and binding to corresponding scaffolds, among which the 25 °C group showed the strongest binding signal. These results may be attributed to the assumption that a cultivation temperature of 25 °C is relatively appropriate for exogenous protein folding and scaffold stabilization in E. coli. Thus, we selected 25 °C as the culture temperature for subsequent experiments (Fig. 1b). As shown in Fig. 1c, 471 and 251 bp DNA fragments were amplified from the precipitates of the TALE1-GFP1-Scaffold1 and TALE2-GFP2-Scaffold2/TALE3-GFP2-Scaffold1 groups, respectively, by using anti-GFP antibody. By contrast, the negative control immunoprecipitations that did not use antibodies (beads only) or adopted the normal rabbit IgG showed no amplification signal. Meanwhile, the no-scaffold control groups did not display any amplification signal. The amplified fragments were confirmed by sequencing. These results indicate that TALE-GFP1/2 can specifically bind to the corresponding plasmid DNA BMs in vivo.
Split GFP assay for determining the effectiveness of the TALE-DNA scaffold system. The effect of the TALE-DNA scaffold system on the in vivo clustering of functional proteins fused with orthogonal interaction domains was examined using the split GFP assay. To perform this assay, the synthetic biological parts of Plac, RBS, TALE1, GFP1, TALE2/3, GFP2, the terminator, and the corresponding scaffold1/2/3 were assembled in order using the BioBrick ™ standard assembly method and then inserted into the plasmid backbone of pSB1C3. The plasmids pT1-G1/T3-G2-S1, pT1-G1/T2-G2-S2, pT1-G1/T2-G2-S3 (designed as shown in Fig. 2a), pT1-G1/T3-G2 (without scaffold1 control), and pT1-G1/T2-G2 (without scaffold2/3 control) ( Supplementary  Fig. S1b online) were constructed and then verified by sequencing. E. coli BL21 (DE3) transformed with such plasmids were cultured and induced by IPTG overnight. The groups without IPTG supplementation were designated as uninduced controls. Subsequently, bacterial samples were harvested, and corresponding fluorescence intensities (abbreviated as FI, Ex: 488 nm; Em: 538 nm) were determined by the Fluoroskan Ascent FL (Thermo Scientific). As shown in Fig. 2b, the FI/OD 600 value in the TALE1-GFP1/TALE3-GFP2-Scaffold1 group was significantly higher than that in the no-scaffold1 control group upon IPTG induction (p = 0.0052). In parallel, the FI/OD 600 values in the TALE1-GFP1/TALE2-GFP2-Scaffold2 and TALE1-GFP1/TALE2-GFP2-Scaffold3 groups were significantly higher than that in the no-scaffold2/3 control group upon IPTG induction (p = 0.0100 and p = 0.0027, respectively). Both semi-quantitative reverse-transcription (semi-qRT-PCR) and real-time quantitative RT-PCR (qRT-PCR) analyses showed that there are no statistical significances in expression of GFP1 (for scaffold1 group p = 0.2928, for scaffold2 group p = 0.6690, for scaffold3 group p = 0.5502) and GFP2 (for scaf-fold1 group p = 0.4745, for scaffold2 group p = 0.7877, for scaffold3 group p = 0.6177) between scaffold groups and no-scaffold control groups. This result indicated that the increase in green FI in the scaffold system was not caused by the expression variation of GFP1 or GFP2 (Fig. 2c, Fig. S2a,S2b). These findings suggest that the TALE-DNA scaffold system is an efficient device for the clustering and ordering of different proteins fused with TALE proteins.
Role of the TALE-DNA scaffold system in IAA production. To investigate the function of the TALE-DNA scaffold system in heterologous metabolic pathways, we adopted a prototype of the IAA synthetic pathway by fusing the IAAM and IAAH enzymes to two TALEs (TALE1 and TALE2, Fig. 3a,b). After performing a construction process similar to that described above, the plasmids pT1-IAAM/T2-IAAH-S2 and pT1-IAAM/ T2-IAAH-S3 ( Supplementary Fig. S1c online) were verified by sequencing and then employed to transform E. coli BL21 (DE3), which were cultured and induced with IPTG overnight. The pT1-IAAM/T2-IAAH transformation group was applied as the no-scaffold control. Sandwich enzyme-linked immunosorbent analysis (ELISA) revealed that the IAA productions in the T1-IAAM/T2-IAAH-S2 (9.457 μ mol/L) and T1-IAAM/T2-IAAH-S3 (6.715 μ mol/L) groups were approximately 9.6-fold (p = 0.0086) and 6.8-fold (p = 0.0028) higher than that in the no-scaffold T1-IAAM/T2-IAAH control group (0.986 μ mol/L), respectively (Fig. 3c). Notably, the Scaffold3 group with larger adjacent BM intervals than the Scaffold2 group showed a lower increasing yield efficiency (p = 0.0305), indicating a distance-dependent pattern in the TALE-DNA scaffold system. Both semi-qRT-PCR and qRT-PCR analyses showed that there are no statistical significances in expression of IAAM (for scaffold2 group p = 0.4196, for scaffold3 group p = 0.6543) and IAAH (for scaffold2 group p = 0.1079, for scaffold3 group p = 0.7730) between scaffold groups and no-scaffold control group. This result indicated that the increased IAA production in the scaffold system was not due to the expression variation of IAAM and IAAH ( Fig. 3d and Fig. S2c). On the basis of these results, the TALE-DNA scaffold system was demonstrated to increase IAA production effectively through an IAAM-IAAH metabolic pathway. Thus, the TALE-DNA scaffold system can efficiently accelerate the rates of heterologous metabolic pathways in prokaryotic chassis.

Discussion
Given their sequence-specific DNA-binding abilities, TALEs have been extensively used for genome editing, transcriptional regulation, epigenetic modification, visualization of genomic regions, and locus-specific ChIP 28 . All of these applications are focused on genomic research. However, whether the locus-specific DNA-binding behavior of TALEs could be utilized in biochemical and bioengineering studies, such as the hastening of heterologous metabolic reactions in vivo, remains undetermined. For this purpose, we first constructed expression plasmids containing both the sequences encoding the rationally designed TALE proteins fused with split GFPs or specific enzymes and the corresponding DNA BMs. After transforming into E. coli, both the split GFP and IAA production assays demonstrated that the TALE-DNA scaffold system can be efficiently used to cluster and order TALE-fused proteins, as well as accelerate the rates of heterologous metabolic pathways in prokaryotic chassis.
Previous artificial scaffolds for pathway organization have rapidly become an important approach to provide control over pathway fluxes by altering local enzyme stoichiometry. Consequently, the TALE-based DNA scaffold system developed in this study could be a novel approach to achieve this purpose. The TALE-based DNA scaffold system may possess more advantages in several aspects, such as compatibility, stabilization, expandability, and predictability. TALE is a natural component of the bacterial infection system. Therefore, the TALE-based DNA scaffold system may exhibit substantial compatibility with prokaryotic cells. This feature renders this system suitable for in vivo integration into prokaryotic biofactories. The TALE-based DNA scaffold system also exhibits greater extensibility as compared with previously developed in vivo protein-based and RNA-based scaffolds. In vivo protein scaffolds have encountered the problems of limited metazoan signaling interaction proteins and complexity in folding and function. Although the in vivo RNA scaffold system can offer greater predictability and may avoid much of the undesirable crosstalk that is prevalent in protein-based scaffolds, this system still showed degradable properties and fewer well-characterized RNA binding domains of high binding affinity. By contrast, each TALE can specifically bind to a DNA sequence of only 10-20 bp in length. Therefore, tandem DNA BMs can recruit a series of TALE-fused enzymes, which may theoretically allow unlimited numbers of enzymes to be spatially organized. Meanwhile, TALE-based genome manipulation has been confirmed to be available for a diverse array of eukaryotic species, such as mouse, zebrafish, Caenorhabditis elegans, and Drosophila 29-32 . This advance suggests the potential of the TALE-DNA scaffold system to be used in broader chassis. Thus, aside from other scaffold systems used to accelerate heterologous metabolic pathways, the TALE-DNA scaffold system theoretically possesses a high extensibility both in long multi-enzymatic chains and in numerous host species. However, this supposition remains to be confirmed in future studies.
According to previous reports, the importance of the precise order and spacing of enzymes in the context of pathway enhancement remains debatable 33 . Accordingly, our results indicated that the influence of interval length between adjacent TALE-binding motifs to the TALE system was fused-enzyme/protein type-dependent. For instance, the IAA production system showed a distance-dependent pattern, whereas the split-GFP system did not. These differences may be attributed to the fact that the IAA production system essentially depended on an enzyme-catalytic mechanism, whereas the split-GFP system was essentially influenced by a protein interaction mechanism. As indicated by the scaffold-mediated metabolite micro-domain hypothesis 34 , different scaffold architectures with different parameters, i.e., enzyme distance, may significantly affect the probability of the intermediate's downstream reaction and the subsequent metabolic flux dynamics in the micro-domain. This hypothesis explains the distance-dependent pattern of IAA production. However, as long as the distance between the two portions of GFP is within a tolerable range for their interaction in the split GFP assay, the effect of subtle distance variation between these parts might be insignificant. Meanwhile, the enzyme-catalytic process might amplify the distance effect, whereas the split GFP assay could not exert such an effect. Nevertheless, the space effect on different metabolic systems requires further investigations based on TALE-DNA scaffolds in future studies to completely elucidate such mechanisms.
To the best of our knowledge, this study is the first to use TALEs in an in vivo artificial heterologous metabolic accelerating system in a prokaryotic chassis. Our study demonstrated that locus-specific DNA-binding tools could be further expanded into broader areas, such as biochemical engineering, in addition to traditional genome studies. We anticipate that this novel heterologous metabolic system accelerator would be widely applied both in prokaryotic and eukaryotic biofactories for the bioengineering of various metabolic products, such as biomass, biofuels, and biomaterials, as well as for sewage disposal and waste treatment. (c) IAA production was determined by sandwich ELISA after overnight culture of E. coli with IPTG induction. After TMB staining and reaction termination, the color was measured spectrophotometrically at 450 nm wavelength. IAA concentrations in the samples were then determined by comparing their OD 450 to standard curves. The relative IAA concentration of the TALE1-IAAM/TALE2-IAAH no-scaffold control group was set arbitrarily at 1.0, and the levels of the other groups were adjusted correspondingly. Data represent the results obtained from at least three independent experiments. *p < 0.05, **p < 0.01. (d) Semi-qRT-PCR analysis of IAAM and IAAH expression in different TALE-IAA-scaffold groups. The cDNA sequence of 16S rRNA was amplified as the standard.
Scientific RepoRts | 6:26065 | DOI: 10.1038/srep26065 Methods Strains and plasmids. E. coli DH5α (Takara) was used for routine subcloning, and E. coli BL21 (DE3) was used for protein expression in accordance with the manufacturer's instructions. The plasmids were constructed on the basis of the BioBrick standard assembly using pSB1C3 as the backbone containing tandem restriction sites for EcoRI, PstI, SpeI, and XbaI digestion. The plasmids of pSB1C3-Plac (BBa_R0010), pSB1C3-IAA biosynthetic genes (BBa_K515100), and pSB1C3-Terminator (BBa_B0015) were supplied by the registry of standard biological parts from the iGEM Foundation. All plasmid manipulations were conducted using standard cloning techniques.
Design of the DNA scaffold. Two 14 bp TALE BMs (BM1: GGAGGCACCGGTGG and BM2: GATAAACACCTTTC) were designed on the basis of the sequence of an unrelated gene to avoid homology with the E. coli genome. Considering the typical in vivo B-type DNA structure with helical turns of 10 bp in length, we designed the interval sequences between two 14 bp BMs to be 6 bp or 16 bp long to ensure that the adjacent fusing enzymes are in the same spatial direction. No interval sequence (only a 5′ -T for TALE binding) was added to the "head-to-head" pattern to allow the fused enzymes to be as close as possible. The devices exhibiting the "BM1-interval-BM2-interval" pattern were synthesized with repetitions of more than 10 times (11 times) and then integrated into the plasmid ( Supplementary Fig. S3 online).
Construction of TALE-coding plasmid. TALE1, TALE2, and TALE3 were designed on the basis of the DNA scaffold BMs of BM1, BM2, and reverse BM2, respectively. The TALEs were constructed with the Golden Gate TALEN and TAL Effector Kit 2.0 by Addgene ( Supplementary Fig. S4 online). The fragments expressing TALE modules were subcloned into a pSB1C3 plasmid flanked by the restriction sites EcoRI and XbaI at the 5′ end and SpeI and PstI at the 3′ end. The linker sequence ACTAGA for the fusion of enzymes with the TALEs was introduced by the isoschizomers of SpeI and XbaI. The primers employed to identify the sequences containing the TALE modules are shown in Supplementary Table S1 online. Plasmid DNA was extracted following the Miniprep protocol (Qiagen) and then sequenced on the MegaBACE 1000 system (GE Healthcare) by using a DYEnamic ET dye terminator cycle sequencing kit (Pharmacia). Full-length cDNA fragments were assembled by a catabolite gene activator protein (CAP 3.0).