A Gene Circuit Combining the Endogenous I-E Type CRISPR-Cas System and a Light Sensor to Produce Poly-β-Hydroxybutyric Acid Efficiently

‘Metabolic burden,’ which arises when introducing exogenic synthesizing pathways into a host strain, remains a challenging issue in metabolic engineering. Redirecting metabolic flux from cell growth to product synthesis at an appropriate culture timepoint is ideal for resolving this issue. In this report, we introduce optogenetics—which is capable of precise temporal and spatial control—as a genetic switch, accompanied by the endogenous type I-E CRISPRi system in Escherichia coli (E. coli) to generate a metabolic platform that redirects metabolic flux. Poly-β-hydroxybutyric acid (PHB) production was taken as an example to demonstrate the performance of this platform. A two-to-three-fold increase in PHB content was observed under green light when compared with the production of PHB under red light, confirming the regulatory activity of this platform and its potential to redirect metabolic flux to synthesize target products.


Introduction
Metabolic engineering is a promising research field that can meet the demands of target compound production while meeting social and environmental obligations [1]. The primary strategy of metabolic engineering is rebuilding and optimizing metabolic networks and regulatory pathways within cells to enhance the production of native metabolites or to enable cells to produce new products, including fuels, chemicals, foods, feeds and pharmaceuticals [2,3]. However, the productivity of desired products is not always ideal, mainly because the native state of engineered strains is influenced by modifications to intrinsic metabolic networks, which causes a 'metabolic burden' [4]. Metabolic burden arises because of competition between endogenous and exogenic pathways for substrates and ATP [5]. The downregulation of genes associated with bacterial growth and cell maintenance may increase the titer and yield of target products but, at the same time, decrease the density of engineered bacteria, which may reduce potential productivity levels. Thus, overcoming inefficient production requires resolving the conflict between cell growth and product synthesis. This issue can be remedied by dynamically controlling fermentation during the growth phase, i.e., engineered pathways are inhibited to focus cellular metabolism on biomass build-up and the production phase (in which engineered pathways are induced to biosynthesize the product) [6][7][8].
Several gene expression systems can redirect metabolic flux from mainstream to introduced pathways. Generally, exogenous chemical inducers are combined with soluble transcription factors to achieve the artificial control of gene expression [9], e.g., L-arabinose heterologous muconic acid synthesis in E. coli with a 130% increase in production [33]. Furthermore, Polstein et al. designed a light-activated CRISPR-Cas9 effector (LACE) system. This system was accomplished by fusing light-induced heterodimeric proteins CRY2 and CIB1 to the transactivation domain and catalytically inactive dCas9, respectively, to induce the transcription of endogenous genes in the presence of blue light [34]. When using the type II CRISPRi system, in addition to designing a specific guide RNA for the target gene, an additional plasmid vector expressing dCas9 (4 kb) should be constructed. In contrast, the endogenous type I-E CRISPR-Cas system only requires the expression of the CRISPR array and, thus, can be easily used for internal regulation without causing a metabolic burden. Our group applied IPTG-induced type I-E CRISPR-Cas system to the metabolic engineering of E. coli for the first time [35].
By taking advantage of the green/red light-sensitive CcaS-CcaR system and the endogenous type I-E CRISPRi system [35], we present a metabolic platform for the precise temporal and spatial redirection of metabolic flux and use poly-β-hydroxybutyric acid (PHB) production as an example to test the platform. This metabolic platform functioned well through fermentation tests using the dry weight of cells and PHB productivity as indicators, thereby supporting our concept. Thus, we illustrate a new strategy in metabolic engineering by fusing optogenetic tools and the endogenous CRISPRi system for precise switching between cell growth and product synthesis.

Construction of Strains and Plasmids
The strains, plasmids and primers used in this work are listed in Supplementary Tables S1-S3. Plasmids without the crRNA sequence were constructed in E. coli DH5α cells with the sequence of EE-E15 (lab stock, E. coli Top10∆cas3). The strain was attained by knocking out cas3 and substituting the native promoter of the cascade operon with J23119. CcaS consists of an N-terminal transmembrane helix; a GAF domain, which acts as the sensor domain; a linker region (L1); two PAS domains; a second linker region (L2); and a C-terminal histidine kinase (HK) domain. By removing L1 and the two PAS domains and fusing the GAF and HK domains with the truncated linker region, a truncated version of the photoreceptor can be constructed [36]. The plasmid pSR43.6 was used as a template, and the primers #3 F, #3 R, #4 F, #4 R, #10 F and #10 R were used for PCR amplification. The plasmids pSR43.6#3, pSR43.6#4 and pSR43.6#10 were obtained by Gibson assembly. The above plasmids were transformed into DH5α cells by chemical transformation to obtain strains #3, #4 and #10. The plasmid Paracr15A was used as the plasmid construction template. The primers 58.6 primer F and 58.6 primer R were used for PCR amplification, and the high-copy plasmid Paracr58.6 was obtained by Gibson assembly. The vector plasmids pHZ3.1 and pSR43.6 were PCR-amplified with the crRNA gene fragments, respectively, and the plasmids pHZ-crRNA and 43.6-crRNA were obtained by Gibson assembly. The pHZ-crRNA was used as the plasmid construction template, and the primers phbCAB F2 and phbCAB R were used for PCR amplification. The plasmid vector and the phbCAB were assembled by Gibson to obtain the plasmid pHZ-PHBcon. The plasmids Paracr58.6, pHZ-crRNA, 43.6-crRNA and pHZ-PHBcon were transformed into E. coli EE-E15 by chemical transformation to obtain the strains Ara58.6, SphZ, S43.6 and SlcrPHB.

Characterization of the CcaS-CcaR System
Bacteria were inoculated in 2 mL of LB culture medium with chloramphenicol and streptomycin in 24-well plates. The plates were placed in a shaker and cultured at 37 • C and 500 rpm for 12 h. The cells were then placed on ice to stop bacteria growth, and 100 µL of the culture was transferred to wells of a 96-well plate to measure OD 600 . The cultures were diluted to an OD 600 = 0.1 with M9 culture medium (with chloramphenicol and streptomycin) to a final volume of 1 mL. Three parallel repeats were prepared for each sample. The new plates were wrapped with aluminum foil to prevent natural light reaching the cells, and the lighting device was placed in the aluminum foil to ensure that the cells can be illuminated by green/red light. The plate was cultured at 37 • C and 500 rpm. The containers were placed into a microplate reader (Synergy HT, BioTek, Winooski, VT, USA) every hour to measure the cell density and fluorescence. A negative control was used to measure and subtract intrinsic cellular fluorescence.

Characterization of crRNA
Single colonies were inoculated into 50 mL of LB medium with appropriate antibiotics (spectinomycin and chloramphenicol) and cultured at 37 • C and 220 rpm in darkness overnight for 13 h. The samples were diluted to an OD 600 of 0.06-0.07 using 50 mL of M9 medium (containing 1% glycerol and appropriate antibiotics), and these cultures were placed in Erlenmeyer flasks. Three parallel repeats were prepared for each sample. The cultures in the Erlenmeyer flasks were cultured at 37 • C and 220 rpm under the red-light condition and then changed to green light at 0, 5 and 9 h. Then, 200 µL of the samples were placed into 96-well cell culture clusters using the microplate reader to measure the absorbance every 2 h.

Fermentation of PHB
Single E. coli colonies were each inoculated into 6 mL of LB culture medium containing chloramphenicol and streptomycin, and these seeds were cultured for 12 h. Subsequently, 1 mL of the first seed was inoculated into 50 mL of LB culture medium containing chloramphenicol and streptomycin and cultured at 37 • C, 220 rpm for 8 h. Approximately 1 mL of the 50 mL second seed culture was inoculated into 100 mL of M9 culture medium (OD 600 = 0.05) containing chloramphenicol and streptomycin for fermentation. Glucose was the only carbon resource. Fermentation cultures were grown at 37 • C, 180 rpm for 64 h. The OD and glucose consumption were measured every 8 h. OD was detected with an Ultraviolet-visible spectrophotometer and 15 µL of fermentation culture supernatant was used to analyze glucose consumption in a biosensor analyzer. The PHB content was determined by gas chromatography (GC).

Measurement of PHB Production
After fermentation, the cells were harvested by centrifugation at 7000 rpm for 15 min. The bacteria sample was lyophilized for 6 h, and the cell sample was weighed. The bacteria were ground carefully, and 15 mg of the bacteria sample was accurately weighed into a well-sealed small bottle. The weight of the bottle with the bacteria sample was weighed. Concentrated sulfuric acid (150 µL), 850 µL of methanol and 1 mL of trichloromethane were added in that order under a fume hood. The bottles were sealed and placed into a 100 • C oil bath for 1 h. The bottles were then removed, and 1 mL of ultrapure water was added. The bottles were sealed, and the samples were thoroughly mixed by vortex oscillation. The samples were left standing for at least an hour until the liquid separated. The top water phase and lower organic phase were stored. PHB dissolved into the organic solvent. The organic phase (~600 µL) was carefully transferred to a 1 mL vial for a GC analysis to determine the content of PHB.
The chromatographic column of the gas chromatograph is an Rtx-5 capillary column, and the detector is a hydrogen flame ionization detector (FID). The column temperature was kept at 80 • C for 1 min, then the column temperature was raised to 120 • C at the rate of 10 • C/min, and then to 160 • C at the rate of 45 • C/min. The column temperature was kept for 5 min, with a total time of 10.89 min. The FID temperature was 300 • C, the hydrogen flow rate was 40 mL/min, and the air flow rate was 400 mL/min. The injection port temperature was 250 • C, and helium was used as the carrier for a 1:15 split injection.

Optimization of the Optogenetic CcaS/CcaR Sensor
CcaS-CcaR, a green/red photo-reversible two-component signal transduction system (TCS) first discovered in cyanobacteria [37], is used widely in prokaryotes and eukaryotes. The autophosphorylation of CcaS and downstream phosphorylation of CcaR are upregulated under 672 nm green light irradiation, and then CcaR binds to the cpcG2 promoter region, which initiates the expression of downstream genes. The above process is reversed under red light irradiation.
In this study, we expressed the CcaS/CcaR system in E. coli TOP10 to explore and improve its regulatory performance. We constructed three variant strains, #3/#4/#10, by removing different lengths of PAS domain and linker DNA [38,39] (Figure 1a), which were co-characterized with the wild-type system as a control. Upon illumination, all sensors exhibited different expression patterns under red and green irradiation ( Figure S1). Among them, variant #10 had the highest dynamic ranges of green over red light, whereas the difference between the wild-type and #3 was insignificant. The dynamic ranges of green light over red light for the different systems were 2.8 (wild type); 2.7 (#3); and 4.9 (#10). In contrast, #4 exhibited a reverse response when compared with the other two systems (i.e., #3 and #10), with higher expression under red light instead of green light and an approximate 1.5-fold change ( Figure 1b). Thus, we selected the best-performing system, #10, for regulation in the following studies.

Optimization of the Optogenetic CcaS/CcaR Sensor
CcaS-CcaR, a green/red photo-reversible two-component signal transduction system (TCS) first discovered in cyanobacteria [37], is used widely in prokaryotes and eukaryotes. The autophosphorylation of CcaS and downstream phosphorylation of CcaR are upregulated under 672 nm green light irradiation, and then CcaR binds to the cpcG2 promoter region, which initiates the expression of downstream genes. The above process is reversed under red light irradiation.
In this study, we expressed the CcaS/CcaR system in E. coli TOP10 to explore and improve its regulatory performance. We constructed three variant strains, #3/#4/#10, by removing different lengths of PAS domain and linker DNA [38,39] (Figure 1a), which were co-characterized with the wild-type system as a control. Upon illumination, all sensors exhibited different expression patterns under red and green irradiation ( Figure S1). Among them, variant #10 had the highest dynamic ranges of green over red light, whereas the difference between the wild-type and #3 was insignificant. The dynamic ranges of green light over red light for the different systems were 2.8 (wild type); 2.7 (#3); and 4.9 (#10). In contrast, #4 exhibited a reverse response when compared with the other two systems (i.e., #3 and #10), with higher expression under red light instead of green light and an approximate 1.5-fold change ( Figure 1b). Thus, we selected the best-performing system, #10, for regulation in the following studies. The fluorescence intensity of cells grown under green light and red light. The fluorescence intensity was acquired 9 h after inoculating the cultures into 24-well plates with a starting OD600 of ~0.1. The cultures were exposed to different light to characterize the different expression levels. The basic OD600 and fluorescence were eliminated. Bars indicate relative fluorescence intensity when compared with the normalized fluorescence level of the wild-type system. Data represent the mean ± SD from three independent experiments (n = 3).
In optimizing the optical sensor, we developed a 24-well plate read-out combined with an LED patch lamp and optical control hardware to achieve greater penetration of light and an earlier K value of the culture (Figure 2a). The ballast is connected to a 220 V power supply, and 24 LED patch lamps match the 24-well plate (Figure 2b). The light illumination from each patch lamp is controlled independently. The illumination intensity of red light was 5.49 w/m 2 , and that of green light was 3.26 w/m 2 (Table S4). The fluorescence intensity of cells grown under green light and red light. The fluorescence intensity was acquired 9 h after inoculating the cultures into 24-well plates with a starting OD 600 of~0.1. The cultures were exposed to different light to characterize the different expression levels. The basic OD 600 and fluorescence were eliminated. Bars indicate relative fluorescence intensity when compared with the normalized fluorescence level of the wild-type system. Data represent the mean ± SD from three independent experiments (n = 3).
In optimizing the optical sensor, we developed a 24-well plate read-out combined with an LED patch lamp and optical control hardware to achieve greater penetration of light and an earlier K value of the culture (Figure 2a). The ballast is connected to a 220 V power supply, and 24 LED patch lamps match the 24-well plate (Figure 2b). The light illumination from each patch lamp is controlled independently. The illumination intensity of red light was 5.49 w/m 2 , and that of green light was 3.26 w/m 2 (Table S4). Biosensors 2022, 12, x FOR PEER REVIEW 6 of 13

Characterization of the Type I-E CRISPRi System
In E. coli, the widely used type II system requires an additional plasmid to express dcas9. In contrast, the endogenous type I CRISPR system is relatively easy to regulate internally without causing metabolic burden [35]. CrRNA in the type I-E system sequences is recognized by the ribonucleoprotein complex, Cascade, during target DNA binding [40,41]. The ribonucleoprotein complex Cascade is composed of a 61-nt crRNA and five different Cas proteins in an uneven stoichiometry, which are encoded by eight genes separately (cas3, cse1, cse2, cas7, cas5, cas6e, cas1, cas2) [42]. In this system, by knocking out cas3 and activating the expression of the CRISPR-associated complex for antiviral defense (Cascade), the DNA-cutting function of the CRISPR system is inactivated, whereas the DNA-binding function is maintained, which can inhibit the expression of target genes ( Figure S2). After exploring the expression of the single light-controlled gene expression system CcaS/R in E. coli, we combined the single light-controlled gene expression system with the endogenous I-E CRISPRi system in E. coli to construct a platform for the dynamic regulation of metabolic flux.
To remove the DNA degradation function and maintain the DNA binding function of the modified type I-E CRISPR system, a special E. coli TOP10 with the Cas3 protein knocked out, and the promoter of Cascade substituted by the promoter J23119 was constructed and the strain named EE-E15 [35] (Figure S2). In order to verify the function of this system, the GFP-expressing plasmid and crRNA expression vector were co-transformed into E. coli TOP10Δcas3, and a series of strains SGFP-Y (Y represents 0, T1 and NT1, which are the spacer names; 0 indicates control, targeting no sites) were constructed to test the function of crRNA targeting different GFP sites (Figure 3a). In the presence of L-arabinose, SGFP-T1 and SGFP-NT1 showed strong inhibition. The results showed that

Characterization of the Type I-E CRISPRi System
In E. coli, the widely used type II system requires an additional plasmid to express dcas9. In contrast, the endogenous type I CRISPR system is relatively easy to regulate internally without causing metabolic burden [35]. CrRNA in the type I-E system sequences is recognized by the ribonucleoprotein complex, Cascade, during target DNA binding [40,41]. The ribonucleoprotein complex Cascade is composed of a 61-nt crRNA and five different Cas proteins in an uneven stoichiometry, which are encoded by eight genes separately (cas3, cse1, cse2, cas7, cas5, cas6e, cas1, cas2) [42]. In this system, by knocking out cas3 and activating the expression of the CRISPR-associated complex for antiviral defense (Cascade), the DNA-cutting function of the CRISPR system is inactivated, whereas the DNA-binding function is maintained, which can inhibit the expression of target genes ( Figure S2). After exploring the expression of the single light-controlled gene expression system CcaS/R in E. coli, we combined the single light-controlled gene expression system with the endogenous I-E CRISPRi system in E. coli to construct a platform for the dynamic regulation of metabolic flux.
To remove the DNA degradation function and maintain the DNA binding function of the modified type I-E CRISPR system, a special E. coli TOP10 with the Cas3 protein knocked out, and the promoter of Cascade substituted by the promoter J23119 was constructed and the strain named EE-E15 [35] (Figure S2). In order to verify the function of this system, the GFP-expressing plasmid and crRNA expression vector were co-transformed into E. coli TOP10∆cas3, and a series of strains SGFP-Y (Y represents 0, T1 and NT1, which are the spacer names; 0 indicates control, targeting no sites) were constructed to test the function of crRNA targeting different GFP sites (Figure 3a). In the presence of L-arabinose, SGFP-T1 and SGFP-NT1 showed strong inhibition. The results showed that repression among different sites varied widely, and the spacer positioned on the promotor region presented a high repression level (Figure 3b). repression among different sites varied widely, and the spacer positioned on the promotor region presented a high repression level (Figure 3b).

Construction of the Type-I-E CRISPR System for Regulating gltA Expression
We focused on citrate synthase as a model candidate for metabolic flux regulation. Citrate synthase catalyzes acetyl-CoA and oxaloacetate to form citric acid, which is the rate-limiting step in the TCA cycle and crucial for bacterial growth [43,44]. Thus, regulating the expression of gltA, which codes for citrate synthase, may affect central metabolism. In contrast to previous work using the imported exogenous CRISPRi system (CRISPR/dCas9) [30,45], an endogenous CRISPRi system, type I-E CRISPRi system in E. coli was applied to convert the growth phase to the production phase without causing extra metabolic burden and interference. Therefore, we used the type-I-E CRISPR system for regulating gltA expression to control metabolic flux (Figure 4a).
In characterizing the type I-E CRISPR system, we constructed strain EE-E15 with Cas3 knocked out. As demonstrated in other studies, targeting different sites of gene can lead to a different regulation effect for gltA, and because the spacer targeting the latter of the two promoters of the gltA gene showed the greatest effect [35,46]. Therefore, we selected this promoter as the target site and constructed a medium-copy plasmid (Paracr15A) and a high-copy plasmid (Paracr58.6) containing crRNA targeting gltA to inhibit the expression of gltA ( Figure S3). Initially, we characterized this regulator with chemical substances. Inducing with 0.2% L-arabinose resulted in an apparent suppression of strain growth when compared with the strain grown without L-arabinose (Figure 4b). This observation indicated that the type I-E CRISPRi system affects cell growth by producing crRNA targeting gltA. Strains with aracr58.6 showed a slower growth rate, which may be caused by the high leakage of the high-copy plasmid.
We then combined the CcaS #10/CcaR optogenetic sensor with crRNA to investigate the regulation effect of targeting gltA through switching light. We constructed the medium-copy plasmid (43.6-crRNA) and high-copy plasmid (pHZ-crRNA) with a specific

Construction of the Type-I-E CRISPR System for Regulating gltA Expression
We focused on citrate synthase as a model candidate for metabolic flux regulation. Citrate synthase catalyzes acetyl-CoA and oxaloacetate to form citric acid, which is the rate-limiting step in the TCA cycle and crucial for bacterial growth [43,44]. Thus, regulating the expression of gltA, which codes for citrate synthase, may affect central metabolism. In contrast to previous work using the imported exogenous CRISPRi system (CRISPR/dCas9) [30,45], an endogenous CRISPRi system, type I-E CRISPRi system in E. coli was applied to convert the growth phase to the production phase without causing extra metabolic burden and interference. Therefore, we used the type-I-E CRISPR system for regulating gltA expression to control metabolic flux (Figure 4a).
In characterizing the type I-E CRISPR system, we constructed strain EE-E15 with Cas3 knocked out. As demonstrated in other studies, targeting different sites of gene can lead to a different regulation effect for gltA, and because the spacer targeting the latter of the two promoters of the gltA gene showed the greatest effect [35,46]. Therefore, we selected this promoter as the target site and constructed a medium-copy plasmid (Paracr15A) and a highcopy plasmid (Paracr58.6) containing crRNA targeting gltA to inhibit the expression of gltA ( Figure S3). Initially, we characterized this regulator with chemical substances. Inducing with 0.2% L-arabinose resulted in an apparent suppression of strain growth when compared with the strain grown without L-arabinose (Figure 4b). This observation indicated that the type I-E CRISPRi system affects cell growth by producing crRNA targeting gltA. Strains with aracr58.6 showed a slower growth rate, which may be caused by the high leakage of the high-copy plasmid.
continuously (Figure 4c). Bacteria growth had obvious palliation after switching to the green light at 5 h (K/2 value). In contrast, there was a negligible growth difference between the strains induced at 9 h and the control sample illuminated with red light. Thus, we hypothesize that during the stationary phase, there was effectively no TCA cycle inhibition to repress the cell growth, which may be caused by the metabolic flow of the cells [8]. The metabolic pathway to the synthetic product. Two acetyl-CoA will condense to produce acetoacetyl-CoA, which is catalyzed by acetoacetyl-CoA reductase to generate (R)-3-hydroxybutyryl-CoA. Then, it is polymerized by PHA polymerase to form PHB. (b) Comparison of the regulation effect between different plasmid copy numbers. Ara58.6 is the strain with the aracr58.6 plasmid, and Ara15A is the strain with the aracr15A plasmid. (c) The effect of different regulation times by switching red light to green light. Green and red represent continuous exposure to green and red light, respectively, whereas X h (X = 5.9) represents switching the light at X hours after starting the culture. Data represent the mean ± SD from three independent experiments (n = 3). The metabolic pathway to the synthetic product. Two acetyl-CoA will condense to produce acetoacetyl-CoA, which is catalyzed by acetoacetyl-CoA reductase to generate (R)-3-hydroxybutyryl-CoA. Then, it is polymerized by PHA polymerase to form PHB. (b) Comparison of the regulation effect between different plasmid copy numbers. Ara58.6 is the strain with the aracr58.6 plasmid, and Ara15A is the strain with the aracr15A plasmid. (c) The effect of different regulation times by switching red light to green light. Green and red represent continuous exposure to green and red light, respectively, whereas X h (X = 5.9) represents switching the light at X hours after starting the culture. Data represent the mean ± SD from three independent experiments (n = 3).
We then combined the CcaS #10/CcaR optogenetic sensor with crRNA to investigate the regulation effect of targeting gltA through switching light. We constructed the mediumcopy plasmid (43.6-crRNA) and high-copy plasmid (pHZ-crRNA) with a specific crRNA spacer and transformed both plasmids into Top10∆cas3 cells to gain the strains S43.6 and SphZ, respectively. Upon green or red illumination, the two groups showed variance in growth ( Figure S4). The growth of SphZ decreased under green light, and the growth inhibition of S43.6 was not clearly observed, which may be because of the low expression of crRNA and the leakage of CcaS/CcaR under red light. Thus, we selected the high-copy plasmid (pHZ-crRNA) in the subsequent studies.
We then investigated the regulation effects of light-controlled endogenous CRISPRi targeting gltA at different induction times by switching red light to green light at 0, 5 and 9 h. The strains exposed to red light grew better than those exposed to green illumination continuously (Figure 4c). Bacteria growth had obvious palliation after switching to the green light at 5 h (K/2 value). In contrast, there was a negligible growth difference between the strains induced at 9 h and the control sample illuminated with red light. Thus, we hypothesize that during the stationary phase, there was effectively no TCA cycle inhibition to repress the cell growth, which may be caused by the metabolic flow of the cells [8].

Demonstrating Light-Controlled Redirection of Metabolic Flux
We used our system to regulate the PHB synthetic pathway as a proof of concept for the practical application of the light-mediated metabolic flux redirection system. The PHB production pathway competes with the TCA cycle for the common substrate acetyl-CoA. Thus, green light was used to activate the I-E CRISPRi system. This system targeted the gltA gene, which is the key gene of the TCA cycle. We inhibited the expression of gltA at different growth stages to balance the metabolic flux between growth and product production to improve the target product yield. We constructed the strain SlcrPHB, which expressed the phbCAB cluster constitutively and transcript crRNA targeting gltA under the control of the CcaS/CcaR system ( Figure 5). crRNA was expressed at different growth stages (lag, exponential, stationary phases) during PHB fermentation by switching red to green light irradiation. The groups were name experimental group A (green light irradiation at 0 h inhibition of gltA); experimental group B (red light converted to green light irradiation at 5 h inhibition of gltA); experimental group C (red light converted to green light irradiation at 13.5 h inhibition of gltA); and control group D (red light irradiation, no inhibition of gltA). To provide different corresponding lighting for PHB fermentation, the inner wall of each light box was surrounded by a circle of LEDs, and the maximum changes in radiance were related to the number of LEDs.

Demonstrating Light-Controlled Redirection of Metabolic Flux
We used our system to regulate the PHB synthetic pathway as a proof of concept for the practical application of the light-mediated metabolic flux redirection system. The PHB production pathway competes with the TCA cycle for the common substrate acetyl-CoA. Thus, green light was used to activate the I-E CRISPRi system. This system targeted the gltA gene, which is the key gene of the TCA cycle. We inhibited the expression of gltA at different growth stages to balance the metabolic flux between growth and product production to improve the target product yield. We constructed the strain SlcrPHB, which expressed the phbCAB cluster constitutively and transcript crRNA targeting gltA under the control of the CcaS/CcaR system ( Figure 5). crRNA was expressed at different growth stages (lag, exponential, stationary phases) during PHB fermentation by switching red to green light irradiation. The groups were name experimental group A (green light irradiation at 0 h inhibition of gltA); experimental group B (red light converted to green light irradiation at 5 h inhibition of gltA); experimental group C (red light converted to green light irradiation at 13.5 h inhibition of gltA); and control group D (red light irradiation, no inhibition of gltA). To provide different corresponding lighting for PHB fermentation, the inner wall of each light box was surrounded by a circle of LEDs, and the maximum changes in radiance were related to the number of LEDs. Figure 5. The light-controlled CRISPR system regulates the metabolic pathway of PHB production. The crRNA was expressed on a plasmid and Cascade was activated for expression with the constitutive promoter J23119. This Cascade mediates the maturation of crRNA and forms a complex with crRNA to target gene sites, thus, disturbing transcription.
During fermentation, there was no significant difference in the OD600 values between SlcrPHB growing under different light conditions (Figure 6a). This observation indicated that the suppression of gltA did not affect the growth of E. coli, which was probably because PHB had a positive effect on cell growth and offset the detrimental impact of gltA knockdown [47].
Moreover, compared with that of control group D under red light irradiation, the PHB contents for experimental group A and experimental group B increased by 2.6-and 2.7-fold, respectively (Figure 6b). This result suggests that inhibiting gltA was essential for improving the content of PHB. In contrast, the PHB production of experimental group B was 1.12 and 1.35 times that of experimental group A and experimental group C, respectively, possessing the highest productivity and content (Figure 6c). This result is probably because suppressing gltA expression during the lag phase causes detrimental effects on intrinsic metabolism. In contrast, the inhibition of gltA during the stationary phase exerts During fermentation, there was no significant difference in the OD 600 values between SlcrPHB growing under different light conditions (Figure 6a). This observation indicated that the suppression of gltA did not affect the growth of E. coli, which was probably because PHB had a positive effect on cell growth and offset the detrimental impact of gltA knockdown [47]. thase has not been used frequently in metabolic engineering because of the adverse effects on cell growth [8]. Previous research showed nearly no E. coli growth after knocking out gltA, which can be attributed to the inhibition of the TCA cycle [48]. However, endogenous and synthetic pathways compete for the important metabolic precursor, acetyl-CoA. Therefore, the light-controlled conditional inhibition of gltA serves as a promising strategy to redirect metabolic flux from the TCA cycle to desired pathways.

Conclusions
In this report, we demonstrated that a light-controlled gene expression system can be used for the metabolic flux regulation in E. coli. After characterizing the effects of different Moreover, compared with that of control group D under red light irradiation, the PHB contents for experimental group A and experimental group B increased by 2.6-and 2.7-fold, respectively (Figure 6b). This result suggests that inhibiting gltA was essential for improving the content of PHB. In contrast, the PHB production of experimental group B was 1.12 and 1.35 times that of experimental group A and experimental group C, respectively, possessing the highest productivity and content (Figure 6c). This result is probably because suppressing gltA expression during the lag phase causes detrimental effects on intrinsic metabolism. In contrast, the inhibition of gltA during the stationary phase exerts weaker effects on cell growth, whereas a lower level of the carbon source is available for PHB synthesis. This observation indicated that suppressing gltA expression during the exponential phase may be a feasible option to optimize PHB production, which is consistent with the work by Soma et al. [8], where later induction resulted in a smaller increase in the concentration of acetate and product accumulation. As fermentation scales, expressing crRNA during the exponential phase may, to some extent, increase profit.
We demonstrated that optogenetic tools are suitable for the metabolic flux regulation of a prokaryotic system. The effect of multi-light channels applied in metabolic engineering to inhibit the expression of acetate-producing genes has been demonstrated previously [30], indicating that introducing optogenetic control in fermentations should increase the yield of target products. We altered light at various time points to accomplish the conditional knockdown of a critical gene to switch the bacterial phase, which distinguishes our work from previous research efforts. Using this approach, bacteria accumulated biomass during the red light-induced growth phase and then, under green light illumination at a particular time point, transformed to the production phase. We can maximize the production of valuable chemicals by regulating the timing of the switch. Moreover, we tuned the expression of citrate synthase to increase the efficiency and universality because we controlled bacterial growth. The knockout or knockdown of citrate synthase has not been used frequently in metabolic engineering because of the adverse effects on cell growth [8]. Previous research showed nearly no E. coli growth after knocking out gltA, which can be attributed to the inhibition of the TCA cycle [48]. However, endogenous and synthetic pathways compete for the important metabolic precursor, acetyl-CoA. Therefore, the light-controlled conditional inhibition of gltA serves as a promising strategy to redirect metabolic flux from the TCA cycle to desired pathways.

Conclusions
In this report, we demonstrated that a light-controlled gene expression system can be used for the metabolic flux regulation in E. coli. After characterizing the effects of different CcaS variants in E. coli, a platform that dynamically regulates metabolic flux was constructed by combining the CcaS #10 variant with the endogenous I-E CRISPRi system, and PHB production was taken as an example to illustrate the performance of this metabolic platform. PHB productivity was found to differ by 2-3-fold from green light to red light irradiation, indicating the successful use of the platform. Moreover, we showed that the time point for induction gave different target product yields. Therefore, inhibiting gltA at three different time points, the lag phase, logarithmic phase and stationary phase, can significantly improve the effectiveness and universality of light-controlled CRISPR switches. We demonstrated the possibility of combining light-controlled elements with the endogenous I-E CRISPRi system for dynamic regulation studies, providing new research ideas and tools for the precise allocation of metabolic flux.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/bios12080642/s1, Table S1: Strains used in this study; Table S2: Plasmids used in this study; Table S3: Primers used in this study; Table S4: Parameters of LED strips; Figure S1: The relative fluorescence of different type of CcaS-CcaR; Figure S2: The modification of endogenous CRISPRi system; Figure S3. sgRNA sequence targeting gltA; Figure S4: The regulation effect of different copy-number in light-controlled system.