CRISPRi-mediated Tunable Control Of Gene Expression Level With Engineered Single-guide RNA in Escherichia Coli

ABSTRACT

Precise control of gene expression is essential for flux redistribution in metabolic pathways. Although the CRISPR interference (CRISPRi) system can effectively repress gene expression at the transcriptional level, it has still been difficult to precisely control the level without loss of specificity or an increase in cell toxicity. In this study, we developed a tunable CRISPRi system that performs transcriptional regulation at various levels. We constructed a single-guide RNA (sgRNA) library targeting repeat, tetraloop, and anti-repeat regions to modulate the binding affinity against dCas9. Each screened sgRNA could regulate the gene expression at a certain level between fully repressing and nonrepressing states (>45-fold). These sgRNAs also enabled modular regulation with various target DNA sequences. We applied this system to redistribute the metabolic flux to produce violacein derivatives in a predictable ratio and optimize lycopene production. This system would help accelerate the flux optimization processes in metabolic engineering and synthetic biology.

cistanche tubulosa-improve immune system

Click here to view  Cistanche Enhance Immunity products

【Ask for more】 Email:cindy.xue@wecistanche.com /  Whats App:  0086 18599088692 /  Wechat:  18599088692

INTRODUCTION

The type II Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)/CRISPR-associated (Cas) system is an RNA-guided DNA cleavage system composed of CRISPR RNA (crRNA), trans-activating CRISPR RNA (tracrRNA) and a Cas9 protein (1,2). Cas9 from Streptococcus pyogenes has been widely used in genome engineering due to its well-characterized properties (3,4). The complex of the spCas9 protein, tracrRNA, and crRNA binds and cleaves the target DNA protospacer complementary to the 20 nucleotides (nt) spacer sequence of the crRNA in the presence of 5’-NGG-3’ protospacer adjacent motif (PAM) sequence (5,6). To simplify the system, chimeric single guide RNA (sgRNA) where crRNA and tracrRNA are joined by an artificial linker called tetraloop can be used (5). It allows for replacing two separate RNAs with a single RNA molecule, making the system more manageable. One of the CRISPR-based systems, CRISPR interference (CRISPRi), utilizes a catalytically inactive Cas9 (dCas9) by replacing certain amino acids from each endonuclease domain (7). This modification makes a ribonucleoprotein (RNP) complex that can direct and bind to target DNA and interferes with RNA polymerase (RNAP) to inhibit transcription initiation or elongation (7). In contrast to a knockout by CRISPR/Cas9 system, CRISPR allows for the reversible knockdown of genes by expressing dCas9 or sgRNA under an inducible promoter, allowing for more flexibility in gene regulation (7). CRISPRi system has effectively silenced target genes hundreds-fold (7,8). Target-specific gene repression enables high-throughput phenotyping using a genome-wide sgRNA pool (9) or redirects carbon flux in engineered bacteria toward the target chemical (10–12). In addition, multiplex CRISPRi has been demonstrated in various hosts using a variety of methods to stably express multiple sgRNAs, including long crRNA arrays, toehold switch integration, and nonrepetitive extralong sgRNA arrays (13–16). These techniques of inhibiting multiple genes simultaneously have been utilized to perform more complex tasks, such as reducing toxic intermediates or blocking competing pathways in bacterial fermentation (13,16–19). However, complete repression is not always desirable for metabolic engineering. Gene expression levels should be precisely controlled to achieve optimal enzyme activity between the growth and production pathways, balancing in intracellular precursor concentrations and avoiding the accumulation of toxic intermediates (20–22). Various techniques have been developed to engineer the repression efficiency of CRISPRi. Several factors were manipulated, such as the amount of sgRNA and dCas9 protein or the choice of sgRNA. Controlling dCas9 expression or both dCas9 and sgRNA expression using an inducible promoter regulated target gene expression by more than 30-fold or 300-fold, respectively (23,24). However, it has been observed that modulating dCas9 concentration can lead to inconsistent repression efficiencies on different protospacers (23). The limited number of available inducible promoters also makes design challenges despite many efforts (25). Another strategy is the insertion of a ligand-specific RNA aptamer into the sgRNA, where a small cognate molecule can expose the spacer, thus enabling dose-dependent control of CRISPRi (26). The ligand-specific sgRNA is helpful if a suitable ligand and aptamer pair is available. The distance from the transcription start site and the orientation of dCas9 binding also affect the efficiency of CRISPRi (7). Adjusting the distance between the sgRNA binding site and the transcription start site can optimize the pathway in metabolic engineering (17,27,28). Hybridization efficiency between the spacer and protospacer sequences also changes RNP complex binding efficiency (29–31). Therefore, mismatches designed outside the seed region to alter the free energy were introduced to diversify CRISPRi efficiency (31,32). However, designing the sgRNA is required for each new target sequence, and achieving a significant difference in free energy through only a mismatch can be challenging. Moreover, there is a possibility of increased off-target binding (33). Nevertheless, it is still limited to precisely repressing multiple gene expressions in a modular manner regardless of the target sequence. In this study, we developed a tunable CRISPRi system that regulates gene expression with high precision and predictability. By iterative fluorescence-based cell sorting of the sgRNA library targeting tetraloop and its flanking region, we achieved transcriptional repression efficiencies up to 45-fold even when the target DNA is changed. This system allows for the predictable redistribution of metabolic flux without manipulating the host genome or synthetic pathways, providing a promising tool for optimizing metabolic pathways.

MATERIALS AND METHODS

Bacterial strains, plasmids, and reagents

Escherichia coli strains and plasmids used in this study are listed in Supplementary Table S1. Primers used in this study are listed in Supplementary Table S2. Cloning procedures are described in the supplementary methods. Target genes and their target DNA sequences for the CRISPR system are shown in Supplementary Table S3. Mach1-T1R was used for general cloning. The library was constructed using NEB 5-alpha Electrocompetent E. coli (NEB). Routine cultures for constructing plasmids and strains were performed at 37◦C in Luria-Bertani (LB) broth or on an LB agar plate containing appropriate antibiotics (34 g/ml chloramphenicol, 50 g/ml spectinomycin). PCR was performed using Q5® High-Fidelity DNA Polymerase (NEB). Plasmids and PCR products were purified using GeneAll® Exprep™ Plasmid SV Kit and Gel SV Kit (GeneAll Biotechnology). Repression efficiency of sgRNA mutants Overnight culture media was refreshed to OD600 of 0.05 and incubated until it reached a value between 0.6 and 0.8. The culture media was then diluted to OD600 of 0.05 with 20 ng/ml of aTc and incubated for 4 h. The fluorescence of a single colony was measured at 575 nm/616 nm using a Hidex Sense microplate reader (Hidex), and the OD600 value was determined using Jenway 7300 spectrophotometer (Jenway). The relative fluorescence was calculated as the fluorescence units per OD600 value, with the value for PC set as 100%.

Effects Of Cistanche tubulosa-treat constipation

sgRNA library screening

The same cultural method as the one described above was used. Culture media was diluted to [cell] <10−7 cells/ml in PBS buffer (pH 7.4). The fluorescence distribution was obtained from 100,000 cells from each sample using S3e Cell Sorter (Bio-Rad), and 5000 cells for each fluorescence range were sorted in each iteration. Cells were overnight cultured in fresh media for the next iteration. Samples were obtained after the final iteration and spread to the LB agar plate. Each colony was inoculated into 100 l of LB media and overnight cultured in 96-well microplate shaking at 900 rpm by Hidex Sense microplate reader (Hidex). Overnight culture media was refreshed by diluting 3 l culture media in 97 l fresh media and incubated for 2 h. Culture media was re-diluted by mixing 6 l culture media with 94 l fresh media, and aTc was added. The fluorescence of every single colony was obtained at 575 nm/616 nm by Hidex Sense microplate reader (Hidex) after 6 h of incubation. The OD600 value from the microplate reader was converted to the OD600 value from the spectrophotometer.

RNA extraction

Total RNA was extracted using RNeasy Mini Kit (Qiagen) only when miRNeasy Mini Kit (Qiagen) for sgRNA quantification, following treatment with RNAprotect Bacteria Reagent (Qiagen) and storage at –80◦C. On-column DNA digestion using RNase-Free DNase Set (Qiagen) was carried out during the extraction process. DNA contamination was verified through PCR without reverse transcription. The concentration and purity of each sample were determined using a NanoDrop One (Thermo Scientific), and the quality was evaluated using a DNA 5K/RNA/CZE 24 LabChip (PerkinElmer) on a LabChip GX Touch 24 (PerkinElmer).

cistanche tubulosa-improve immune system

RT-qPCR analysis

Primers for RT-qPCR were either sourced from lab stock or designed using the Primer3Plus tool (34) and listed in Supplementary Table S4. The CysG gene was used as a reference gene (35) for all RT-qPCR experiments. A standard curve was generated by 10-fold diluting a template from 107 copies/ l to 10−1 copies/ l. The limit of quantification was determined by fitting the copy number and positive amplification rate to a sigmoid function and calculating the 95% positive copy number value. The linear dynamic range was defined as the lowest copy number at which the back-calculated Cq variation is less than 35%. Cq values above 30 or those not amplified in non-template control were validated. Standard curve data is available in Supplementary Table S5 and Supplementary Figure S1. A total of 2.5 ng of total RNA was used in a 10 l RT-qPCR reaction with Luna® Universal One-Step RT-qPCR Kit (NEB). RT-PCR was performed in triplicate on StepOnePlus™ RealTime PCR System (Applied Biosystems) with the following steps: (i) reverse transcription at 55◦C for 10 min, followed by denaturation at 95◦C for 1 min; (ii) 40 thermal cycles at 95◦C for 10 s and 60◦C for 1 min; (iii) melt curve analysis at 95◦C for 15 s and 60◦C for 1 min, followed by a ramp from 60◦C to 95◦C at a rate of rate 0.3◦C/min. The relative quantification of RNA was determined using the delta-deltaCt method (36) and analyzed with StepOnePlus™ Software (Applied Biosystems).

RNA-sequencing (RNA-seq) and data analysis

1.5 in biological duplicate. Total RNA was extracted by the abovementioned method, and ribosomal RNA (rRNA) was depleted using Ribo-Zero plus rRNA Depletion Kit (Illumina). cDNA libraries were constructed using KAPA Stranded RNA-Seq Kits (Kapa Biosystems), with an additional size-selection step using AMPure XP (Beckman Coulter). Adapters and primers used in this study are listed in Supplementary Table S2. The concentration was measured using a Qubit 3.0 Fluorometer (Thermo Scientific), and the size distribution of the cDNA was measured using X-Mark LabChip (PerkinElmer) on the LabChip GX Touch 24 instrument. The average cDNA length ranged from 390 to 450 bps. RNA-Seq was performed as >19M reads of 41 bp paired-end using NextSeq 500/550 High Output Kit v2.5 (75 Cycles) (Illumina) on NextSeq 550 (Illumina). Data analysis was performed using the reference genome of E. coli BL21(DE3) complete genome (Genbank CP001509.3) with FastQC (v0.11.9) (http:// www.bioinformatics.babraham.ac.uk/projects/fastqc/) (37), Bowtie (v1.2.3) (38) with trim3 option value of 3, samtools (v1.10) (39), and cufflinks (v2.2.1) (40,41). All samples between replicates had R2 > 0.92 based on FPKM (Fragments Per Kilobase of transcripts per Million mapped reads). Differentially expressed genes were obtained from the threshold of 1.5 (log2-based) and 0.01 (q-value).

Proviolacein and prodeoxyviolacein production and sampling

For violacein and prodeoxyviolacein production, the following medium composition was used: 10.7 g/l K2HPO4, 5.2 g/l KH2PO4,1g/l NaCl, 1 g/l NH4Cl, 1 g/l MgSO4, 0.2 g/l yeast extract, and 10 g/l glucose. An overnight culture media was refreshed to an OD600 of 0.1 and cultured until OD600 reached between 0.6 and 0.8. The culture media was then diluted to OD600 of 0.1 and supplemented with 500 ng/ml aTc, 0.05 mM IPTG, and 0.1 g/l L-tryptophan to a final volume of 5 ml. 300 l of culture media was sampled and centrifuged at 14 000 rpm for 5 min to obtain a cell pellet for HPLC analysis, and 300 l of culture media was sampled for RNA extraction.

HPLC analysis

For violacein biosynthetic pathway analysis, the cell pellet was resuspended in 2-volume methanol, sonicated for 5 min, centrifuged, and filtered through a 0.22-m nylon filter. HPLC analysis followed the method (42) with Poroshell 120 EC-C18, 4.6 × 150 mm, 4 m column (Agilent). Two mobile phases, A with 0.1% formic acid in DDW and B with 0.1% formic acid in acetonitrile, were used at a flow rate of 0.6 ml/min with the following gradient profile: 95% A for 1.5 min, ramping A from 95% to 2% for 4 min, maintaining 2% A for 2 min, ramping A from 2% to 95% for 30 s, and maintaining 95% A for 10 min. Samples were detected using a PDA detector, and the HPLC area was determined at 600 nm for violacein and 610 nm for prodeoxyviolacein.

Lycopene production and sampling

For lycopene production, the following medium composition was used: 13.56 g/l Na2HPO4,6g/l KH2PO4,2g NH4Cl, 1 g NaCl, 2 ml of 1 M MgSO4, 0.1 ml of 1 M CaCl2 and 4 g/l of glucose (22). 27 g/ml chloramphenicol and 50 g/ml spectinomycin were used. An overnight culture media was refreshed to OD600 of 0.1 to a final volume of 5 ml. When the OD600 reached between 0.5 and 0.6, aTc was added to the final concentration of 500 ng/ml, and when the OD600 reached between 0.9 and 1, IPTG was added to the final concentration of 0.02 mM. Lycopene was extracted using acetone, as described in the reference (22). The concentration of lycopene was obtained from the absorbance at 475 nm using a Jenway 7300 spectrophotometer using a standard curve. 300 l of culture media was sampled for RNA extraction.

cistanche benefits for men-strengthen immune system

RESULTS AND DISCUSSION

The effects of the mutations in sgRNA on the repression efficiency of the CRISPR system

The CRISPRi system can repress gene expression when the dCas9–sgRNA–DNA ternary complex strongly binds together. We hypothesized that the repression efficiency of the CRISPRi system could be changed to a certain level by modulating the RNP complex formation. To quantify the repression efficiency of the CRISPR system, we designed a fluorescence reporting system based on the genetic circuit where dCas9 from S. pyogenes is induced by anhydrotetracycline (aTc), and both reporter gene (mCherry) and sgRNA variants are constitutively expressed (Figure 1A). Once aTc is added into the medium, dCas9–sgRNA– DNA ternary complex blocks the transcription of the reporter gene, and the fluorescence level inversely indicates the repression efficiency of the CRISPR system. The sgRNA contains 96 nucleotides (nts) in total. The spacer region (the first 20 nts) binds to the target DNA sequence, and the other 76 nts contribute to the complex formation and stability. Since mismatches in the spacer region may lose specificity and increase off-target effects in the CRISPR-based system (33,43), the maximum length of the sgRNA that can be engineered is 76 nts. In this case, a library size of 476 (approximately 1045) should be obtained, but it is too large to be constructed. Thus, it is required to narrow the size of the library down by determining specific positions on the sgRNA. We first investigated how mutations in sgRNA at each region affect the repression in the CRISPR system. A previous study reported that mutations in the stem–loop 1 (sg mut01 and sg mut02) led to a significant decrease in the nuclease activity of Cas9, while mutations in the stem–loop 2 (sg mut03) or stem–loop 3 (sg mut04) unaffected its activity (44). A reconstructed sgRNA (sg mut05) containing mutations in the stem–loop 1 region and repeat, tetraloop, anti-repeat, and stem–loop 2 regions also decreased nuclease activity (44). However, in the CRISPRi system, all five sgRNA variants had little effect on repression efficiency compared to the wild-type sgRNA (sg WT), with each variant exhibiting a 2–4% fluorescence level compared to the positive control without a spacer region (sg PC) (Figure 1B). These results suggest that the formation of the dCas9–sgRNA complex is less sensitive to sequence modification in the stem–loop 1 region in the CRISPRi system, in contrast to the CRISPR/Cas9 system. We hypothesized that removing regions related to the stability of the complex can be a starting point to control the expression on multiple levels further. To this end, we constructed a truncated sgRNA (sg trcSL1) lacking the linker, stem–loop 2 and stem–loop 3. This truncated sgRNA has a minimal number of components and has been shown to significantly reduce the nuclease activity in the CRISPR/Cas9 system (5,44). When tested in the CRISPRi system, this defective mutant exhibited an 8% fluorescence level compared to sg PC, suggesting that it could still bind to dCas9 and repress the gene of interest at a certain level even with decreased ternary complex stability (Figure 1B). The structural analysis of the RNP complex suggests that the linker, stem–loop 2, and stem–loop 3 regions of the sgRNA are responsible for stabilizing the complex through their interactions with the nuclease domain of dCas9, which leads to a significant decrease in indel efficiency in the CRISPR/Cas9 system (44). In contrast, the repeat: anti repeat region and stem–loop 1 of the sgRNA mainly interact with the recognition domain and PAM-interacting domain of dCas9 and do not significantly affect the indel efficiency in the CRISPR/Cas9 system (44). This result supports the observation that sg trcSL1 exhibited a significant reduction in nuclease activity in the CRISPR/Cas9 system while only slightly affecting the binding affinity in the CRISPRi system.

Figure 1. System construction and the effect on the repression efficiency by sgRNA mutation. (A) Two plasmids system for the quantification of the repression efficiency. One plasmid with p15A origin contains the dCas9 gene under an aTc-inducible promoter and the mCherry gene under a constitutive promoter. The other plasmid with cloDF13 origin contains wild-type or mutant sgRNA under a constitutive promoter. (B) Relative fluorescence of each sgRNA mutant (sg mut01 ∼ sg mut05 and sg trcSL1) and sg WT compared with sg PC. +1 to + 20 are the spacer sequence that binds to the mCherry gene. sg PC: sgRNA without spacer as a positive control, sg WT: wild-type sgRNA. Error bars indicate standard deviation (n = 3). P-values were determined by unpaired t-test. n.s.: not significant, **P < 0.01

Variating the complex formation efficiency for a multiple level expression control

Tetraloop is an artificial linker that connects crRNA and tracrRNA in the CRISPR-Cas9 system (5). The crystal structure of the ternary complex, composed of dCas9, crRNA, and tracrRNA, has shown that the tetraloop and its 5 and 3 flanking regions have little contact with dCas9 (44). At the same time, repeat:anti-repeat and stem–loop 1 regions are critical for the functional complex (44). As it does not directly interact with dCas9, the tetraloop, and its flanking regions have been explored as potential sites for engineering sgRNA with additional components, such as incorporating aptamer sequences to recruit additional proteins to the sgRNA (45,46). It is plausible that mutations in this region could diversify the binding affinity between sgRNA and dCas9 without entirely disrupting the complex. A 12- mer random library (4 nt of the tetraloop region and 4 nt of 5 and 3 flanking regions, respectively) was constructed using sg trcSL1 as a template to further lower the complex formation efficiency between sgRNA and dCas9 (Figure 2A). We applied a two-step sorting strategy to obtain various sgRNA mutants covering repression efficiency between sg WT (fully-repressing state) and sg PC (non-repressing state). We divided the original library into smaller libraries with different fluorescence intensity ranges, and then individually identified the fluorescence of each sgRNA variant from the smaller library. The fluorescence distribution of the original library obtained by flow cytometry was almost identical to that of sg PC (Figure 2B). However, after three consecutive sorting steps, each library (high, mid, and low) had 63%, 24%, and 14% median fluorescence levels compared to sg PC, respectively (Figure 2B). Eighty-two colonies were isolated from these smaller libraries, with fluorescence levels ranging from 8% to 104% (Figure 2C). These results suggest that mutating tetraloop and its flanking region of the sgRNA can modulate the binding affinity between dCas9 and sgRNA for various repression efficiencies. Reverse-transcription quantitative PCR (RT-qPCR) analysis of 10 sgRNA mutants (sg trcSL1c1–sg trcSL1c10, Supplementary Table S6) with various fluorescence signals between sg WT and sg PC among 82 sgRNA colonies showed that fluorescence levels linearly increased with increasing mCherry transcripts, indicating that our CRISPRi system controls gene expression at the transcriptional level (Supplementary Figure S2). Our tunable CRISPRi system utilizes engineered sgRNAs, which can achieve up to 45-fold repression control, allowing for fine-tuning of the repression efficiency without altering the amounts of dCas9 or sgRNA. A notable advantage of our system is that it can fine-tune the repression efficiency through specific sgRNA variants without the potential for unintended consequences, even though additional cloning is required. Moreover, the growth curve suggests our system does not impose a significant cellular burden (Supplementary Figure S3). Furthermore, RNA-Seq on three sg trcSL1 variants with varying levels of repression efficiency showed that the expression of 51 to 83 genes was changed, in contrast to the 262 genes altered in the sg WT (Supplementary Figure S4). These results suggest that the sg trcSL1 variants had even lower off-target effects than the conventional CRISPRi system. It is particularly relevant in metabolic engineering applications, where optimizing metabolic pathways often involves the manipulation of multiple genes. The ability to fine-tune gene expression without causing negative side effects can significantly accelerate the optimization process and improve the overall efficiency and productivity of the system.

Figure 2. Structure of the sgRNA library and screening of the library. (A) sgRNA scaffold used in this study. 12-mer containing tetraloop and its flanking sequence is randomly mutated from sg trcSL1 to obtain a sgRNA library. (B) Fluorescence distribution obtained from flow cytometry before(left) and after(right) sorting. Each distribution contains 100,000 cells. (C) Fluorescence of single colonies obtained from (B). Error bars indicate standard deviation (n = 2).

Effects of the quantity of dCas9 and sgRNA on repression efficiency

Since modulating the amount of either dCas9 or sgRNA could change the repression efficiency of the CRISPRi system (23,24), it is essential to examine the impact of dCas9 and sgRNA quantity on repression efficiency. We tested the performance of our system under different conditions by changing the amounts of dCas9 and sgRNA using ten sgRNA mutants (sg trcSL1c1–sg trcSL1c10). Three different concentrations (20, 100, and 500 ng/ml) of the inducer were used for dCas9 expression. Compared to the dCas9 mRNA level when 20 ng/ml of inducer was used, it was further increased by 15-fold (100 ng/ml) and 22-fold (500 ng/ml) (Supplementary Figure S5A). There are concerns about overexpressing dCas9 since a high level of dCas9 could be toxic to cells (47,48) or sometimes not (49). However, the growth rate was reduced by less than 10% when up to 500 ng/ml of aTc was used, according to the growth curves at various inducer concentrations (Supplementary Figure S6). This suggests that the amount of dCas9 did not have a toxic effect on E. coli under the tested conditions. The repression efficiency was marginally changed when the different amount of dCas9 was expressed (Figure 3A). It was consistent with a previous study that the CRISPRi system reached maximum repression efficiency at [aTc] < 5 ng/ml when the same replication origin and promoter were used for the expression of CRISPRi components (24). We set the inducer concentration to 500 ng/ml to sufficiently supply dCas9 for further applications such as multiple gene regulation. The property that our system is less sensitive to the amount of dCas9 has advantages over the strategy of adjusting inducer concentration, as the dose-dependent induction of dCas9 might be varied due to the promoters, plasmid copy numbers, or host strains (50–53). Additionally, using CRISPR-based systems allows for the simultaneous targeting of multiple genes for editing or repression, making it an effective tool for multiplexing gene regulation (54). Our tunable CRISPRi system allows for the precise modulation of repression efficiency through sgRNA variants. With this approach, it would be possible to target multiple genes with varying levels of repression simultaneously. The previous study reported that stronger promoters for sgRNA expression lead to higher repression efficiency (24). To investigate the effect of the amount of sgRNA on repression efficiency in our system, two different promoters having different strengths were used (BBa J23100; J23100 and BBa K2753055; UPJ23119). Promoter UPJ23119 contains a UP element that interacts with E. coli RNA polymerase and expresses 16-fold more transcript than J23100 (55). However, the exact fold change may vary depending on the plasmids, genes, and host strains used. Our results showed that UPJ23119 doubled sgRNA expression in our system and had higher repression efficiency (Figure 3B and Supplementary Figure S5B). These results suggest that adjusting the amount of sgRNA can further expand the dynamic range of multi-level expression and provide precise modulation of repression efficiency.

Figure 3. Characterization and the modularity of the tunable CRISPRi system. (A) Relative fluorescence under different aTc concentrations (20, 100, and 500 ng/ml). (B) Relative fluorescence under different promoters (J23100 and UPJ23119). (C) Relative fluorescence under different protospacers. The x-axis indicates the relative fluorescence of mCherry, and the y-axis indicates the relative fluorescence of lacI33-mCherry (red) and araC33-mCherry (blue). 33 bp of additional sequences from lacI and araC were added to the 5-end of mCherry (56). (D) Free energy of 12-mer and the relative fluorescence. Dashed lines indicate a 95% prediction interval and solid lines indicate a regression line. dG value of sg trcSL1c2 could not be obtained using mFold and thus excluded from the analysis. (A–D) All error bars indicate standard deviation (n = 3).

Modularity of the tunable CRISPRi system

Our approach using specific sgRNA variants to maintain a certain expression level for a target DNA has the potential to be used as a modular system. To demonstrate the feasibility of the tunable CRISPRi system for targeting different protospacers, we constructed two fusion genes (lacI33-mCherry and araC33-mCherry) by adding 33 bps of sequences from lacI and araC to the 5 -end of mCherry (56). The genome of E. coli BL21(DE3) (GenBank: CP001509.3) contains two copies of the protospacer sequence for lacI33- mCherry and one copy for araC33-mCherry. However, these chromosomal sequences do not have PAM sequences and, therefore, cannot be targeted by the dCas9-sgRNA complex (43). When the siRNAs were designed to target each protospacer, the repression efficiency was consistently maintained for all protospacers from lacI33, araC33, and mCherry (Figure 3C). These findings suggest that our tunable CRISPRi system can effectively regulate gene expression at a certain level regardless of the protospacer sequence. One potential explanation for the observed relationship between RNP complex formation and repression efficiency is the change in free energy when the complex is formed. The artificial tetraloop linker and its flanking region contain two A: U and G: C Watson-Crick base pairs and an A: G non-Watson–Crick base pair (44). As the binding of crRNA and tracrRNA are crucial for the complex formation, mutations targeting tetraloop would affect the binding affinity between dCas9 and sgRNA. This affinity can be calculated from the free energy difference between the bounded and unbounded states. However, computing the free energy of bounded states is challenging (57). We supposed that the bounded state’s overall structure and stability do not change since sgRNAs sorted from the library did not entirely disrupt its function. Moreover, the tetraloop and its flanking region have little contact with dCas9 (44). Therefore, the binding affinity between dCas9 and sgRNA would be mainly affected by the stability of the secondary structure of the tetraloop and its flanking sequence. Based on this idea, we computed the free energy of the tetraloop and its flanking sequences using mold (58). The difference in free energy of folding due to the mutations is defined as the difference between the free energy of the mutated and wild-type sequences. It could be directly computed from the free energy of folding of mutants since the wild-type sequence’s folding energy is constant at –4.3 kcal/mol. As the free energy of folding increased, the fluorescence signal increased (Figure 3D). Repression efficiency was highest when the folding energy of the tetraloop-containing region was close to or below 0 kcal/mol and decreased linearly as the folding energy exceeded 0 kcal/mol. This observation suggests that the efficiency of complex formation between dCas9 and sgRNA is determined by the secondary structure of the tetraloop region, which helps explain the modularity of our system. The repression efficiency of an additional 85 sgRNA with different 12-mer sequences from the library was measured to further validate this relation. It also showed the correlation observed from sg trcSL1 variants (Figure 3D and Supplementary Table S7), suggesting that the tetraloop linker plays a crucial role in forming the RNP complex and that mutations in this region can impact the efficiency of the CRISPR system. In addition, this relation shows the possibility of computer-based design of sgRNA variants with desired repression efficiency.

Applying a tunable CRISPRi system for flux redistribution in the biosynthetic pathway of violacein in E. coli

Redistributing the metabolic flux is an important process in metabolic engineering to maximize the yield of target products (59,60). Because our system could precisely control gene expression by targeting any protospacer, we chose the violacein biosynthetic pathway as a model system for pathway redistribution. Violacein is an indole derivative compound that exhibits a deep purple color, originally produced by Chromobacterium violaceum (Figure 4A) (61). This compound is synthesized from L-tryptophan by five enzymes encoded by vioA, vioB, vioE, vioD, and vioC (62). Along with violacein (V), other indole derivatives such as proviolacein (PV), deoxyviolacein (DV), and prodeoxyviolacein (PDV) with different colors can also be produced (Figure 4A) (62). The production of violacein and deoxy violacein is an enzymatic process mediated by vioC, while the production of PV and PDV is a non-enzymatic process. We expressed the violacein biosynthetic pathway under the IPTG inducible promoter (63), and different trcSL1 variants were designed to target vioD to redirect the flux. It could effectively modulate the expression of vioD at the transcriptional level (Figure 4B). vioC was repressed by sg WT and showed high variation (Supplementary Figure S7), but the absolute copy number was kept at low levels (>100-fold) compared to other genes when calculated from the standard curves. In contrast, the expression of other unregulated enzymes in the violacein pathway (vioA, vioB, and vioE) remained unchanged. These results suggest that our system can specifically target and modulate the expression of individual genes without affecting other genes in the pathway. Direct quantification of PV and PDV was not possible due to the lack of commercially available standards for HPLC analysis. Thus we measured the titers in arbitrary units using the HPLC peak area. The titer of the product reached a maximum at 8 h after induction and then gradually decreased (Supplementary Figure S8), possibly due to the conversion of the product to other side products, such as chromoviridans, by an unknown mechanism (42,64). The observed minor peaks on the HPLC spectrum indicated the presence of additional minor side products. Previous research has demonstrated that the logarithm of the titer of each product in arbitrary units could be predicted by linear combinations of the promoter strength of each gene involved in the violacein biosynthetic pathway (42). This log-linear regression model was also valid in our system, and increased vioD transcripts were associated with increased PV production (Figure 4B). It is known that engineering the promoter or 5 -UTR of each gene involved in the violacein biosynthetic pathway enabled flux redistribution and pathway optimization (65,66). Our system could work as an alternative method illustrating the transcription control without manipulating promoters involved in the pathway using synthetic sgRNAs, which are short and can be assembled with one-step methods. Screening through the tunable CRISPRi system makes it possible to redirect the pathway with a predictable flux before directly manipulating host strains or gene cassettes, potentially accelerating the process.

Applying the tunable CRISPRi system for balancing carbon flux in lycopene-producing E. coli

The efficient production of valuable compounds in metabolic engineering requires balancing the carbon flux. The methylerythritol 4-phosphate (MEP) pathway is a metabolic pathway that uses simple precursors such as pyruvate and glyceraldehyde 3-phosphate (GAP) to produce isoprenoid building blocks, dimethylallyl pyrophosphate and isopentenyl pyrophosphate (67,68). These molecules synthesize geranyl pyrophosphate, which is converted to farnesyl pyrophosphate (FPP) (67,68). Lycopene, a highly valued carotenoid with potent antioxidant properties and diverse applications in various industries (69), is synthesized from FPP via a series of enzymatic reactions from crtE, crtB, and crtI (Figure 4C) (22,70). The GAP precursor is essential not only for the MEP pathway but also for the glycolysis pathway (22). Decreased expression of the gapA gene, which encodes the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in the glycolysis pathway, has been shown to enhance lycopene production by increasing the GAP pool (22). Therefore, we targeted gapA with five variants, including three sg trcSL1 variants (sg trcSL1c2, sg trcSL1c5, and sg trcSL1c6) that showed various repression efficiency, sg PC, and sg WT. However, gapA is an essential gene that cannot be completely knocked down (71,72). Attempts to repress gapA with conventional CRISPRi using sg WT were unsuccessful even with the low leakiness of dcas9 under tet promoter (73). On the other hand, multiple-level repression of gapA gene expression using the tunable CRISPRi system was achieved without growth defect and resulted in a 2.7-fold increase with sg trcSL1c6 in lycopene production compared with a strain fully expressing gapA gene after 15 h of cultivation (Figure 4D and Supplementary Figure S9). It was reported that only 10% of the natural expression of the gapA gene leads to an increased pool of GAP without a significant change in downstream metabolites in the glycolysis pathway, such as 2-phosphoglycerate and 3-phosphoglycerate (22). Collectively, these results demonstrate that tunable CRISPRi can precisely repress gene expression at multiple levels without chromosome engineering.

Figure 4. Application of tunable CRISPRi system for violacein biosynthetic pathway. (A) Violacein biosynthetic pathway and the plasmid construction. Five genes (vioA, vioB, vioC, vioD, and vioE) responsible for producing violacein were cloned into a dCas9-expressing plasmid. Four major products (violacein, proviolacein, deoxyviolacein, and prodeoxyviolacein) are produced. sg WT for vioC and five sg trcSL1 variants (sg trcSL1c2, sg trcSL1c5, sg trcSL1c6, sg trcSL1c8 and sg trcSL1c9) for vioD, showing different repression efficiencies, were cloned into the other plasmid. Trp: L-tryptophan, V: violacein, PV: proviolacein, PDV: prodeoxyviolacein, DV: deoxyviolacein, PDVA: protodeoxyviolaceinic acid, PVA: protoviolaceinic acid. (B) Relation between vioD transcripts, mCherry fluorescence, and titer of PDV and PV. The x-axis indicates the relative vioD RNA obtained from RT-qPCR, the y-axis on the left indicates the relative mCherry fluorescence from Figure 3, and the y-axis on the right indicates the ratio of the HPLC area. Samples with the highest vioD RNA value were set to 100% for normalization. (C) Lycopene biosynthetic pathway and the plasmid construction. Three genes (crtE, crtB, crtI) responsible for producing lycopene were cloned into a dCas9-expressing plasmid. gapA was targeted with sg trcSL1 variants (sg trcSL1c2, sg trcSL1c5, sg trcSL1c6) and sg PC was cloned into the other plasmid. Glc: Glucose, GAP: glyceraldehyde 3-phosphate, 1,3-BPG: 1,3-bisphosphoglycerate, DMAPP: dimethylallyl pyrophosphate, IPP: isopentenyl pyrophosphate, FPP: farnesyl pyrophosphate. (D) Relation between gapA transcripts, mCherry fluorescence, and lycopene production. The x-axis indicates the relative gap RNA obtained from RT-qPCR, the y-axis on the left indicates the relative mCherry fluorescence from Figure 3, and the y-axis on the right indicates the titer of lycopene. (B, D) All error bars indicate standard deviation (n = 3).

CONCLUSION

We designed a tunable CRISPRi system that could precisely control the gene expression at the desired level. By mutating the tetraloop and its flanking region, synthetic sgRNAs targeting various protospacer sequences could regulate the gene at a certain level. We successfully applied this system to redirect the metabolic flux of interest in the violacein biosynthetic pathway to produce indole derivatives in a predictable ratio. Additionally, we have shown that the use of tunable CRISPRi led to a 2.7-fold increase in lycopene production, highlighting the versatility of our system for improving the production of other valuable compounds. This system would accelerate the flux optimization processes before permanently manipulating host genetic information, such as replacing the promoter or the 5-UTR sequences.

cistanche tubulosa-improve immune system

REFERENCES

1. Sapranauskas,R., Gasiunas,G., Fremaux,C., Barrangou,R., Horvath,P. and Siksnys,V. (2011) The Streptococcus thermophilus CRISPR/Cas system provides immunity in Escherichia coli. Nucleic Acids Res., 39, 9275–9282. 

2. Gasiunas,G., Barrangou,R., Horvath,P. and Siksnys,V. (2012) Cas9-crRNA ribonucleoprotein complex mediates specific DNA cleavage for adaptive immunity in bacteria. Proc. Natl. Acad. Sci. U.S.A., 109, E2579–E2586. 

3. Cong, L., Ran,F.A., Cox, D., Lin,S., Barretto, R., Habib,N., Hsu, P.D., Wu,X., Jiang, W., Marraffini, L.A. et al. (2013) Multiplex genome engineering using CRISPR/Cas systems. Science, 339, 819–823. 

4. Ran, F.A., Hsu,P.D., Wright,J., Agarwala,V., Scott,D.A. and Zhang,F. (2013) Genome engineering using the CRISPR-Cas9 system. Nat. Protoc., 8, 2281–2308. 

5. Jinek,M., Chylinski,K., Fonfara,I., Hauer,M., Doudna,J.A. and Charpentier,E. (2012) A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity. Science, 337, 816–821. 

6. Anders,C., Niewoehner,O., Duerst,A. and Jinek,M. (2014) Structural basis of PAM-dependent target DNA recognition by the Cas9 endonuclease. Nature, 513, 569–573. 

7. Qi,L.S., Larson, M.H., Gilbert,L.A., Doudna,J.A., Weissman,J.S., Arkin,A.P. and Lim,W.A. (2013) Repurposing CRISPR as an RNA-guided platform for sequence-specific control of gene expression. Cell, 152, 1173–1183. 

8. Larson,M.H., Gilbert,L.A., Wang,X., Lim,W.A., Weissman,J.S. and Qi,L.S. (2013) CRISPR interference (CRISPRi) for sequence-specific control of gene expression. Nat. Protoc., 8, 2180–2196. 

9. McGlincy,N.J., Meacham,Z.A., Reynaud,K.K., Muller,R., Baum,R. and Ingolia,N.T. (2021) A genome-scale CRISPR interference guide library enables comprehensive phenotypic profiling in yeast. BMC Genomics, 22, 205. 

10. Yoon,J. and Woo,H.M. (2018) CRISPR interference-mediated metabolic engineering of Corynebacterium glutamicum for homo-butyrate production. Biotechnol. Bioeng., 115, 2067–2074. 

11. Cleto,S., Jensen,J.V., Wendisch,V.F. and Lu,T.K. (2016) Corynebacterium glutamicum metabolic engineering with CRISPR interference (CRISPRi). ACS Synth. Biol., 5, 375–385. 

12. Zhang,G.Q., Ren,X.N., Liang,X.H., Wang,Y.Q., Feng,D.X., Zhang,Y.J., Xian,M. and Zou,H.B. (2021) Improving the microbial production of amino acids: from conventional approaches to recent trends. Biotechnol. Bioproc. E, 26, 708–727. 

13. Kim,S.K., Seong,W., Han,G.H., Lee,D.H. and Lee,S.G. (2017) CRISPR interference-guided multiplex repression of endogenous competing pathway genes for redirecting metabolic flux in Escherichia coli. Microb. Cell Fact., 16, 188. 

14. Ellis,N.A., Kim,B., Tung,J. and Machner,M.P. (2021) A multiplex CRISPR interference tool for virulence gene interrogation in Legionella pneumophila. Commun. Biol., 4, 157. 

15. Siu,K.H. and Chen,W. (2019) Riboregulated toehold-gated gRNA for programmable CRISPR-Cas9 function. Nat. Chem. Biol., 15, 217–220. 

16. Reis,A.C., Halper,S.M., Vezeau,G.E., Cetnar,D.P., Hossain,A., Clauer,P.R. and Salis,H.M. (2019) Simultaneous repression of multiple bacterial genes using nonrepetitive extra-long sgRNA arrays. Nat. Biotechnol., 37, 1294–1301. 

17. Tian,T., Kang,J.W., Kang,A. and Lee,T.S. (2019) Redirecting metabolic flux via combinatorial multiplex CRISPRi-mediated repression for isopentenol production in Escherichia coli. ACS Synth. Biol., 8, 391–402. 

18. Gauttam,R., Seibold,G.M., Mueller,P., Weil,T., Weiss,T., Handrick,R. and Eikmanns,B.J. (2019) A simple dual-inducible CRISPR interference system for multiple gene targeting in Corynebacterium glutamicum. Plasmid, 103, 25–35. 

19. Yao,L., Cengic,I., Anfelt,J. and Hudson,E.P. (2016) Multiple gene repression in cyanobacteria using CRISPRi. Acs Synth. Biol., 5, 207–212. 

20. Lim,H.G., Noh,M.H., Jeong,J.H., Park,S. and Jung,G.Y. (2016) Optimum rebalancing of the 3-hydroxypropionic acid production pathway from glycerol in Escherichia coli. ACS Synth. Biol., 5, 1247–1255. 

21. Yang,Y., Lin,Y., Li,L., Linhardt,R.J. and Yan,Y. (2015) Regulating malonyl-CoA metabolism via synthetic antisense RNAs for enhanced biosynthesis of natural products. Metab. Eng., 29, 217–226. 

22. Jung,J., Lim,J.H., Kim,S.Y., Im,D.K., Seok,J.Y., Lee,S.V., Oh,M.K. and Jung,G.Y. (2016) Precise precursor rebalancing for isoprenoids production by fine control of gapA expression in Escherichia coli. Metab. Eng., 38, 401–408. 

23. Li,X.T., Jun,Y., Erickstad,M.J., Brown,S.D., Parks,A., Court,D.L. and Jun,S. (2016) tCRISPRi: tunable and reversible, one-step control of gene expression. Sci. Rep., 6, 39076. 

24. Fontana,J., Dong,C., Ham,J.Y., Zalatan,J.G. and Carothers,J.M. (2018) Regulated expression of sgRNAs tunes CRISPRi in E. coli. Biotechnol. J., 13, e1800069. 

25. Kim,N.M., Sinnott,R.W. and Sandoval,N.R. (2020) Transcription factor-based biosensors and inducible systems in non-model bacteria: current progress and future directions. Curr. Opin. Biotechnol., 64, 39–46. 

26. Kundert,K., Lucas,J.E., Watters,K.E., Fellmann,C., Ng,A.H., Heineike,B.M., Fitzsimmons,C.M., Oakes,B.L., Qu,J., Prasad,N. et al. (2019) Controlling CRISPR-Cas9 with ligand-activated and ligand-deactivated sgRNAs. Nat. Commun., 10, 2127. 

27. Ferreira,R., Skrekas,C., Hedin,A., Sanchez,B.J., Siewers,V., Nielsen,J. and David,F. (2019) Model-assisted fine-tuning of central carbon metabolism in yeast through dCas9-based regulation. ACS Synth. Biol., 8, 2457–2463. 

28. Su,T., Guo,Q., Zheng,Y., Liang,Q., Wang,Q. and Qi,Q. (2019) Fine-tuning of hemB using CRISPRi for increasing 5-aminolevulinic acid production in Escherichia coli. Front. Microbiol., 10, 1731. 

29. Zheng,T., Hou,Y., Zhang,P., Zhang,Z., Xu,Y., Zhang,L., Niu,L., Yang,Y., Liang,D., Yi,F. et al. (2017) Profiling single-guide RNA specificity reveals a mismatch sensitive core sequence. Sci. Rep., 7, 40638. 

30. Jiang,W., Bikard,D., Cox,D., Zhang,F. and Marraffini,L.A. (2013) RNA-guided editing of bacterial genomes using CRISPR-Cas systems. Nat. Biotechnol., 

31, 233–239. 31. van Gestel,J., Hawkins,J.S., Todor,H. and Gross,C.A. (2021) Computational pipeline for designing guide RNAs for mismatch-CRISPRi. STAR Protoc, 2, 100521. 

32. Hawkins,J.S., Silvis,M.R., Koo,B.M., Peters,J.M., Osadnik,H., Jost,M., Hearne,C.C., Weissman,J.S., Todor,H. and Gross,C.A. (2020) Mismatch-CRISPRi reveals the co-varying expression-fitness relationships of essential genes in Escherichia coli and Bacillus subtilis. Cell Syst., 11, 523–535. 

33. Fortin,J.-P., Tan,J., Gascoigne,K.E., Haverty,P.M., Forrest,W.F., Costa,M.R. and Martin,S.E. (2019) Multiple-gene targeting and mismatch tolerance can confound analysis of genome-wide pooled CRISPR screens. Genome Biol., 20, 21. 

34. Rozen,S. and Skaletsky,H. (2000) Primer3 on the WWW for general users and for biologist programmers. Methods Mol. Biol., 132, 365–386. 

35. Zhou,K., Zhou,L., Lim,Q., Zou,R., Stephanopoulos,G. and Too,H.P. (2011) Novel reference genes for quantifying transcriptional responses of Escherichia coli to protein overexpression by quantitative PCR. BMC Mol. Biol., 12, 18. 

36. Livak,K.J. and Schmittgen,T.D. (2001) Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods, 25, 402–408. 

37. Andrews,S. (2010) FastQC: A Quality Control Tool for High Throughput Sequence Data. v0.11.9 ed. 

38. Langmead,B., Trapnell,C., Pop,M. and Salzberg,S.L. (2009) Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol., 10, R25. 

39. Danecek,P., Bonfield,J.K., Liddle,J., Marshall,J., Ohan,V., Pollard,M.O., Whitwham,A., Keane,T., McCarthy,S.A., Davies,R.M. et al. (2021) Twelve years of SAMtools and BCFtools. Gigascience, 10, giab008. 

40. Trapnell,C., Williams,B.A., Pertea,G., Mortazavi,A., Kwan,G., van Baren,M.J., Salzberg,S.L., Wold,B.J. and Pachter,L. (2010) Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nat. Biotechnol., 28, 511–515. 

41. Trapnell,C., Hendrickson,D.G., Sauvageau,M., Goff,L., Rinn,J.L. and Pachter,L. (2013) Differential analysis of gene regulation at transcript resolution with RNA-seq. Nat. Biotechnol., 31, 46–53. 

42. Lee,M.E., Aswani,A., Han,A.S., Tomlin,C.J. and Dueber,J.E. (2013) Expression-level optimization of a multi-enzyme pathway in the absence of a high-throughput assay. Nucleic Acids Res., 41, 10668–10678. 

43. Hsu,P.D., Scott,D.A., Weinstein,J.A., Ran,F.A., Konermann,S., Agarwala,V., Li,Y., Fine,E.J., Wu,X., Shalem,O. et al. (2013) DNA targeting specificity of RNA-guided Cas9 nucleases. Nat. Biotechnol., 31, 827–832. 

44. Nishimasu,H., Ran,F., Hsu,P., Konermann,S., Shehata,S., Dohmae,N., Ishitani,R., Zhang,F. and Nureki,O. (2014) Crystal structure of Cas9 in complex with Guide RNA and target DNA. Cell, 156, 935–949. 

45. Fu,Y., Rocha,P.P., Luo,V.M., Raviram,R., Deng,Y., Mazzoni,E.O. and Skok,J.A. (2016) CRISPR-dCas9 and sgRNA scaffolds enable dual-colour live imaging of satellite sequences and repeat-enriched individual loci. Nat. Commun., 7, 11707. 

46. Khosravi,S., Schindele,P., Gladilin,E., Dunemann,F., Rutten,T., Puchta,H. and Houben,A. (2020) Application of Aptamers Improves CRISPR-Based Live Imaging of Plant Telomeres. Front. Plant Sci., 11, 1254. 

47. Nielsen,A.A. and Voigt,C.A. (2014) Multi-input CRISPR/Cas genetic circuits that interface host regulatory networks. Mol. Syst. Biol., 10, 763. 

48. Lee,Y.J., Hoynes-O’Connor,A., Leong,M.C. and Moon,T.S. (2016) Programmable control of bacterial gene expression with the 44, 2462–2473. 

49. Vigouroux,A., Oldewurtel,E., Cui,L., Bikard,D. and van Teeffelen,S. (2018) Tuning dCas9’s ability to block transcription enables robust, noiseless knockdown of bacterial genes. Mol. Syst. Biol., 14, e7899. 

50. Zhang,Y., Shang,X., Lai,S., Zhang,G., Liang,Y. and Wen,T. (2012) Development and application of an arabinose-inducible expression system by facilitating inducer uptake in Corynebacterium glutamicum. Appl. Environ. Microbiol., 78, 5831–5838. 

51. Thompson,M.G., Sedaghatian,N., Barajas,J.F., Wehrs,M., Bailey,C.B., Kaplan,N., Hillson,N.J., Mukhopadhyay,A. and Keasling,J.D. (2018) Isolation and characterization of novel mutations in the pSC101 origin that increase copy number. Sci Rep-UK, 8, 1590. 52. Fricke,P.M., Link,T., Gatgens,J., Sonntag,C., Otto,M., Bott,M. and Polen,T. (2020) A tunable L-arabinose-inducible expression plasmid for the acetic acid bacterium Gluconobacter oxydans. Appl. Microbiol. Biotechnol., 104, 9267–9282. 

53. Nguyen,T.T., Nguyen,N.H., Kim,Y., Kim,J.R. and Park,S. (2021) In vivo characterization of the inducible promoter system of 3-hydroxypropionic dehydrogenase in Pseudomonas denitrificans. Biotechnol Bioproc E, 26, 612–620. 

54. McCarty,N.S., Graham,A.E., Studena,L. and Ledesma-Amaro,R. (2020) Multiplexed CRISPR technologies for gene editing and transcriptional regulation. Nat. Commun., 11, 1281. 

55. Yan,Q. and Fong,S.S. (2017) Study of in vitro transcriptional binding effects and noise using constitutive promoters combined with UP element sequences in Escherichia coli. J. Biol. Eng., 11, 33. 

56. Seo,S.W., Yang,J.S., Kim,I., Yang,J., Min,B.E., Kim,S. and Jung,G.Y. (2013) Predictive design of mRNA translation initiation region to control prokaryotic translation efficiency. Metab. Eng., 15, 67–74. 57. Kappel,K., Jarmoskaite,I., Vaidyanathan,P.P., Greenleaf,W.J., Herschlag,D. and Das,R. (2019) Blind tests of RNA-protein binding affinity prediction. Proc. Natl. Acad. Sci. U.S.A., 116, 8336–8341. 

58. Zuker,M. (2003) Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res., 31, 3406–3415. 

59. Xiong,B., Zhu,Y., Tian,D., Jiang,S., Fan,X., Ma,Q., Wu,H. and Xie,X. (2021) Flux redistribution of central carbon metabolism for efficient production of l-tryptophan in Escherichia coli. Biotechnol. Bioeng., 118, 1393–1404. 

60. Yao,R., Li,J., Feng,L., Zhang,X. and Hu,H. (2019) (13)C metabolic flux analysis-guided metabolic engineering of Escherichia coli for improved acetol production from glycerol. Biotechnol. Biofuels, 12, 29. 

61. Duran,N. and Menck,C.F. (2001) Chromobacterium violaceum: a review of pharmacological and industiral perspectives. Crit. Rev. Microbiol., 27, 201–222. 

62. Balibar,C.J. and Walsh,C.T. (2006) In vitro biosynthesis of violacein from L-tryptophan by the enzymes VioA-E from Chromobacterium violaceum. Biochemistry, 45, 15444–15457. 

63. Gwon,D.A., Seok,J.Y., Jung,G.Y. and Lee,J.W. (2021) Biosensor-assisted adaptive laboratory evolution for violacein production. Int. J. Mol. Sci., 22, 6594. 

64. Sanchez,C., Brana,A.F., Mendez,C. and Salas,J.A. (2006) Reevaluation of the violacein biosynthetic pathway and its relationship to indolocarbazole biosynthesis. Chem. Bio. Chem., 7, 1231–1240. 

65. Jones,J.A., Vernacchio,V.R., Lachance,D.M., Lebovich,M., Fu,L., Shirke,A.N., Schultz,V.L., Cress,B., Linhardt,R.J. and Koffas,M.A. (2015) ePathOptimize: a combinatorial approach for transcriptional balancing of metabolic pathways. Sci. Rep., 5, 11301. 

66. Zhang,Y., Chen,H., Zhang,Y., Yin,H., Zhou,C. and Wang,Y. (2021) Direct RBS Engineering of the biosynthetic gene cluster for efficient productivity of violaceins in E. coli. Microb. Cell Fact., 20, 38. 

67. Hunter,W.N. (2007) The non-mevalonate pathway of isoprenoid precursor biosynthesis. J. Biol. Chem., 282, 21573–21577. 

68. Rohmer,M. (1999) The discovery of a mevalonate-independent pathway for isoprenoid biosynthesis in bacteria, algae and higher plants. Nat. Prod. Rep., 16, 565–574. 

69. Muller,L., Caris-Veyrat,C., Lowe,G. and Bohm,V. (2016) Lycopene and its antioxidant role in the prevention of cardiovascular diseases––a critical review. Crit. Rev. Food Sci. Nutr., 56, 1868–1879. 

70. Miura,Y., Kondo,K., Saito,T., Shimada,H., Fraser,P.D. and Misawa,N. (1998) Production of the carotenoids lycopene, beta-carotene, and astaxanthin in the food yeast Candida utilis. Appl. Environ. Microbiol., 64, 1226–1229. 

71. Goodall,E.C.A., Robinson,A., Johnston,I.G., Jabbari,S., Turner,K.A., Cunningham,A.F., Lund,P.A., Cole,J.A. and Henderson,I.R. (2018) The essential genome of Escherichia coli K-12. Mbio, 9, e02096-17. 

72. Baba,T., Ara,T., Hasegawa,M., Takai,Y., Okumura,Y., Baba,M., Datsenko,K.A., Tomita,M., Wanner,B.L. and Mori,H. (2006) Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol. Syst. Biol., 2, 2006.0008. 

73. Bertram,R. and Hillen,W. (2008) The application of Tet repressor in prokaryotic gene regulation and expression. Microb. Biotechnol., 1, 2–16.