Construction of T7-Like Expression System in Pseudomonas putida KT2440 to Enhance the Heterologous Expression Level

Pseudomonas putida KT2440 has become an attractive chassis for heterologous expression with the development of effective genetic manipulation tools. Improving the level of transcriptional regulation is particularly important for extending the potential of P. putida KT2440 in heterologous expression. Although many strategies have been applied to enhance the heterologous expression level in P. putida KT2440, it was still at a relatively low level. Herein we constructed a T7-like expression system in P. putida KT2440, mimicking the pET expression system in Escherichia coli, which consisted of T7-like RNA polymerase (MmP1) integrated strain and the corresponding expression vector for the heterologous expression enhancement. With the optimization of the insertion site and the copy number of RNA polymerase (RNAP), the relative fluorescence intensity (RFI) of the super-folder green fluorescent protein (sfGFP) was improved by 1.4-fold in MmP1 RNAP integrated strain. The induction point and IPTG concentration were also optimized. This strategy was extended to the gene-reduced strain EM42 and the expression of sfGFP was improved by 2.1-fold. The optimal RNAP integration site was also used for introducing T7 RNAP in P. putida KT2440 and the expression level was enhanced, indicating the generality of the integration site for the T7 expression system. Compared to other inducible expression systems in KT2440, the heterologous expression level of the Mmp1 system and T7 system were more than 2.5 times higher. Furthermore, the 3.6-fold enhanced expression level of a difficult-to-express nicotinate dehydrogenase from Comamonas testosteroni JA1 verified the efficiency of the T7-like expression system in P. putida KT2440. Taken together, we constructed and optimized the T7-like and T7 expression system in P. putida, thus providing a set of applicable chassis and corresponding plasmids to improve recombinant expression level, expecting to be used for difficult-to-express proteins.

Many efforts have been made to enhance the heterologous expression level in P. putida KT2440 by regulating the transcriptional level or translation level. To improve the integration expression, the constitutive promoter libraries have been constructed (Zobel et al., 2015) and the T7 RNA polymerase has been integrated into the genome of P. putida KT2440 (Herrero et al., 1993;Christina et al., 2012). Additionally, to enhance the plasmid expression, suites of broad-host-expression plasmids (Calero et al., 2016;Cook et al., 2018), a library of synthetic promoters (Elmore et al., 2017), and the dual-inducible duet-expression system have been generated (Gauttam et al., 2020). The modification of the ribosome-binding site (Calero et al., 2016) and codon optimization (Dammeyer et al., 2011) were also carried out to enhance the expression at the translational level. However, the expression level of P. putida KT2440 is still low, limiting the application of P. putida as a chassis for heterologous expression in industry.
T7 RNA polymerase expression system is an important and high-efficient expression system in model strains such as Escherichia coli Shilling et al., 2020). However, the T7 RNA polymerase-mediated expression system is not always efficient for some non-model microbial strains (Zhao et al., 2017). For example, the T7 expression system constructed in P. putida KT2440 was low efficiency and was rarely used (Christina et al., 2012). Recently, broad-host T7-like systems including the MmP1 expression system have been developed in Halomonas sp. TD01 and P. entomophila LAC31 (Zhao et al., 2017). Furthermore, the MmP1 expression system constructed in the plasmid has successfully contributed to the enhancement of heterologous expression of nicotinate dehydrogenase (NDHase) in C. testosteroni CNB-2 (Lu et al., 2020). Given the high efficiency and versatility, it is supposed that the T7-like system might be suitable for the heterologous expression in P. putida KT2440.
In this study, we attempted to construct the T7-like expression system and evaluate the heterologous expression level in P. putida KT2440. MmP1 RNAP cassette was successfully integrated into the genome of P. putida KT2440, which significantly improved the expression level of super-folder green fluorescent protein (sfGFP). The insertion site and the copy number of MmP1 RNAP cassette were then optimized to improve the efficiency of the MmP1 expression system. We successfully obtained the T7-like expression system in P. putida KT2440, which consisted of the MmP1 RNAP integrated strain and corresponding plasmid. We also utilized the strategy to obtain a T7 expression system in P. putida KT2440 with T7 RNAP inserted in the optimal site. These expression systems were much more efficient than those previously reported in P. putida KT2440. The Mmp1 expression system was also constructed in the P. putida EM42. The expression level of difficult-to-express NDHase was successfully enhanced by using the MmP1 expression system in both P. putida KT2440 and P. putida EM42, demonstrating the high efficiency of the expression system.

MATERIALS AND METHODS
Strains, Culture Conditions, and Chemicals E. coli strain DH5α was used for molecular cloning and plasmid propagation throughout this study. P. putida KT2440 and EM42 were used for the characterization of T7 and MmP1 RNAPpromoter pairs. Luria-Bertani (LB) medium was used for culturing E. coli and P. putida. Antibiotics were supplemented at the following concentrations when needed: for E. coli, 50 μg/ml kanamycin, 15 μg/ml tetracycline; for P. putida, 100 μg/ml kanamycin, 25 μg/ml tetracycline, 100 μg/ml spectinomycin, 50 μg/ml gentamicin. The inducer for P trc promoter was isopropyl β-D-1-thiogalactopyranoside (IPTG) and 3methylbenzoate was for Xyls/P m , arabinose was for AraC/ P araB , and rhamnose was for RhaS/P rha . E. coli was grown at 37°C, and P. putida KT2440 and EM42 were incubated at 30°C.
The DNA polymerase PrimeSTAR ® Max was purchased from Takara Bio Inc (Dalian, China). The Green Taq Mix and the One Step Cloning Kit were purchased from Vazyme Biotechnology (Nanjing, China). The Fast Digest DpnI and the PageRuler ™ Prestained Protein Ladder 10-180 kDa were purchased from Thermo Scientific (United States).

Construction of Plasmids
All the strains and plasmids used in this study are listed in Supplementary Table 2. Plasmid pSEVA-gRNAF was used as the backbone for plasmid construction. Plasmid pSEVA64-vdh was derived from pSEVA-gRNAF by replacing the original N20 sequence with "GTGGAACGCCCCGGTGATAC" via inverse PCR using primers vdh-gF/vdh-gR. The 800 bp length upstream and downstream homologous arms of vdh acquired from the Pseudomonas Genome Database were amplified using primers vdh-1F/vdh-ZR, vdh-ZF/vdh-2R, and connected via overlap-extension (SOE) PCR. The N20 sequence was designed by CasOT (Xiao et al., 2014). A similar method was applied to construct the plasmids pSEVA64-5003 and pSEVA64-5007. T7-like RNAP cassette and promoter were cloned from the plasmid pMmP1-p321, which were donated by the Chen team (Zhao et al., 2017). The RNAP module of T7 and MmP1 was inserted between the upstream homologous arm and the downstream homologous arm to generate the gene inserting vector pSEVA64-T7-vdh, pSEVA64-MmP1-vdh, pSEVA64-MmP1-5003, pSEVA64-T7-5003, and pSEVA64-Mmp1-5007. The verification plasmids containing RNAP modules and their promoter modules (the reporter protein is the sfGFP) were assembled on the plasmid backbone pSEVA64 (Martínez-Garćía et al., 2015) using One Step Cloning Kit to create the plasmids pSEVA64-DE-pRO, pSEVA64-MmP1-pRO. Removed the RNAP module of pSEVA64-DE-pRO, pSEVA64-MmP1-pRO to create the plasmids pSEVA64-DE-sfGFP and pSEVA64-MmP1-sfGFP. Similarly, the other five inducible expression systems (LacI q -P trc , LacI q -P trc , Xyls/P m , AraC/P araB , RhaS/P rha ) were assembled on the upstream of sfGFP with the plasmid backbone pSEVA64 to create the plasmid pSEVA644-sfGFP, pSEVA644R-sfGFP, pSEVA648-sfGFP, pSEVA6410-sfGFP, and pSEVA64-Rha-sfGFP.

Construction of Platform Strains
The platform strains including KTVT and KTVM were constructed by integrating the RNAP module of T7 and MmP1 into the vdh site of the genome of P. putida KT2440, respectively. The integration was carried out by CRISPR/Cas9 system (Sun et al., 2018). The integrated strain KTCM, KTFM, EMCM, EMFM were constructed by integrating the RNAP module of MmP1 into the PP_5003 and the PP_5007 site of P. putida KT2440 and EM42, respectively. The other four P. putida KT2440 mutants, KTCVM, KTVFM, KTCFM, KTCFVM were constructed by the different combinations of the integration sites. The integrated strain KTCT was generated by introducing the T7 RNAP cassette into the PP_5003 site of P. putida KT2440.

Coarse Characterization
Plasmids pSEVA64-MmP1-pRO and pSEVA64-MmP1-sfGFP were transformed into P. putida KT2440, KTVM, respectively, using the electroporation method ( Figure 1). Plasmids pSEVA64-DE-pRO and pSEVA64-DE-sfGFP were transformed into P. putida KT2440, KTVT, respectively. The candidate strains were screened in selection plates and identified by colony PCR. These strains were incubated in LB medium at 30°C, and 1 mM IPTG was added for the induction expression when the optical density reached OD 600 0.4 (about 3 h). After 24 h of incubation, cells were harvested by centrifugation at 4°C and 12,000 × g for 10 min. The cell pellets were washed twice with phosphate buffer (pH 7.4) and then diluted two times. 200 μ of the diluted liquid was added into the 96-well microtiter plates and measured the absolute fluorescence intensity (AFI) and cell density (OD 600 ) by a microplate reader using the excitation wavelength at 488 nm and fluorescence emission at 510 nm. Finally, the relative fluorescence intensity (RFI) was calculated by the AFI/OD 600 .

Optimization of Induction Condition of Different Inducible Expression Systems in P. putida KT2440
For the Mmp1 expression system, the RNAP-integrated strains harboring the corresponding verification plasmids were grown in The verified plasmids were into wild-type strains and RNAP integrated strains, respectively; (c) Shake flask experiments were carried out by inoculating with 2% of the seed cultures at 30°C; (d) Cells were harvested and diluted two times. 200 μl of the diluted liquid was added into the 96-well microtiter plates and measured the absolute fluorescence intensity (AFI) and cell density (OD 600 ) by microplate reader using the excitation wavelength at 488 nm and fluorescence emission at 510 nm. Finally, the relative fluorescence intensity (RFI) was calculated by the AFI/OD 600 .; (C)Construction of different copy numbers of MmP1 RNAP integrated in the genome of P. putida.

The NDHase Activity Assay
The engineered strains (KTCM and EMCM), as well as the wild-type strains (KT2440 and EM42) harboring pSEVA64-JA-sfGFP plasmid, were cultured in 5 ml LB overnight at 30°C. 1 ml of the seed cultures were used to inoculate in 250 ml shake flasks. Then 0.5 mM IPTG was added at 4 h after inoculation for the induction expression at 18°C for 20 h. After the shake flask fermentation, 4 ml of the samples were harvested by centrifugation at 4°C and 4,000 rpm for 10 min. The cell pellets were resuspended with 1 ml of 0.1 M 3-cyanpyridine. To measure NDHase activity, 0.05 mM PMS was added into 1 ml resuspension with magnetic stirring for 2 h at 30°C. Cells were heated at 99°C for 5 min to terminate the reaction. Highperformance liquid chromatography (HPLC) with a Pntulips QS-C18 column (4.6 mm × 250 mm, 5 μm) was used to detect the production of 6-hydroxy-3-cyanopyridine (6-HCP). Detection was performed using a wavelength of 260 nm with 1 ml/min of the liquid phase containing acetonitrile and water (10:90 v/v). One unit (U) of NDHase activity was defined as the amount of enzyme required to produce 1 μmol of 6-HCP in 1 min under the test conditions.

Construction of MmP1 Expression System in P. putida KT2440
To study the effect of T7-like expression system containing T7like RNAP cassette and T7-like promoter module in P. putida KT2440, the Mmp1 RNAP cassette was inserted either in the genome or plasmid. The Mmp1 RNAP cassette was first integrated into the vdh site (locus tag is PP_3357) in the genome by CRISPR/Cas9, resulting in the integrated strain KTVM ( Figure 1A and Supplementary Figure 1). The corresponding plasmid containing P MmP1 promoter and sfgfp was transformed into the integrated strain KTVM to examine the effect of constructed system. The RNAP cassette coupled with sfgfp was also inserted in the plasmid pSEVA64, which was then transformed into P. putida KT2440.
The absolute fluorescence intensity (AFI) and cell density (OD 600 ) were determined after shake-flask fermentation ( Figure 1B). When the MmP1 RNAP cassette was expressed in the plasmid, the RFI value was slightly higher than that inserted into the genome (Figure 2). However, the large-size expression plasmid might decrease the transformation frequency (Walker and Klaenhammer, 2002), affect the quantity of the plasmid DNA isolated from the host strain (Yang and Yang, 2012), and was difficult to construct than smaller size plasmids. Although RNAP cassette inserted in the plasmid provided relatively higher sfGFP expression, considering the RNAP integrated into the genome was genetically stable and convenient to manipulate as the development of the genome editing technology. Hence, the MmP1 RNAP cassette integrated into the genome of P. putida KT2440 was chosen for further enhancing the heterologous expression.
Optimization of the Mmp1 Expression System in P. putida KT2440

Optimization of the Insertion Site of MmP1 RNAP Cassette
The insertion sites were optimized to further improve the MmP1 expression system in P. putida KT2440  Table 3). Three nonessential genes including phaC1, phaF, vdh were selected as the insertion sites, which have been used to integrate other heterologous genes (Gong et al., 2016). The modified strains were named KTCM, KTFM, and KTVM with RNAP cassette inserted in phaC1, phaF, and vdh, respectively. Interestingly, the expression of sfGFP was improved 2-fold in KTCM relative to that in KTVM. Compared to the plasmid expression style, the expression of the sfGFP in KTCM was increased by 1.4-fold (Figure 2), whereas that in KTFM was decreased by 3.3-fold.

Optimization of the Copy Number of the MmP1 RNAP Cassette
We also investigated the effect of the copy number of MmP1 RNAP cassette integrated into P. putida KT2440 on expression. Two copies of RNAP integrated strains were constructed including KTCFM with insertion sites of phaC1 and phaF, KTCVM with insertion sites of phaC1 and vdh, KTFVM with insertion sites of phaF and vdh, and three copies of RNAP integration strain, KTCFVM was constructed with the insertion sites of phaC1, phaF, and vdh ( Figure 1C). Among them, the two-copy strain KTCVM displayed the highest expression level of sfGFP, 1.3-fold higher than the RNAP plasmid expression system (Figure 3). However, it was still slightly lower than the KTCM (Figure 2). Thus, the strain KTCM was selected for further investigation, and the MmP1 expression system in P. putida KT2440 consisting of the host strain KTCM and the plasmid pSEVA64-MmP1-sfGFP carrying P MmP1 promoter corresponding to MmP1 RNAP.

Construction of the MmP1 Expression System in P. putida EM42
To verify the versatility of the MmP1 expression system, we extended the MmP1 expression system to the gene-reduced strain EM42. The optimal RNAP integration site (phaC1) was utilized in EM42 resulting in the integrated strain EMCM. As a result, the RFI in EMCM was about 2.1-fold higher than that of the MmP1 RNAP plasmid expression in EM42 (Figure 4).   To test whether the strategy of increasing heterologous expression level applied to T7 RNAP, the corresponding plasmid containing sfGFP was transformed into the integrated strains KTVT and KTCT. It turns out that the optimal insertion site was also effective for T7 RNAP. As shown in Figure 5, the RFI in KTCT was 1.4-fold higher than that in KTVT.

Comparing the Optimized MmP1
Expression System and the T7 Expression System With Other Expression Systems in P. putida KT2440 The optimized MmP1 and T7 expression systems were compared with other inducible expression systems in P. putida KT2440 to understand its effect. The induction conditions were first optimized. The optimization of the induction condition of KT2440 was performed under different induction points (0-8 h) by using the MmP1 expression system. As shown in Supplementary Figure 4A, when IPTG was added at the induction point 4 h after inoculation, the final RFI value was highest, which was determined to be the optimal induced point (Supplementary Figure 4A). Different inducer concentrations were tested for the effect on expression level and the optimal inducer concentration for six inducible expression systems in P. putida was 0.5 mM IPTG for MmP1 expression system, 1 mM IPTG for LacI q /ptac expression system, 2 mM IPTG for LacI q /ptrc expression system, 80 mM arabinose for AraC/P araB expression system, 5 mM rhamnose for RhaS/P rha expression system, 0.25 mM 3-methylbenzoate for Xyls/ P m expression system (Supplementary Figure 4). The optimal condition of the T7 expression system was considered the same as that of the MmP1 expression system. The optimized induction conditions were then utilized to compare the expression level of different inducible systems in P. putida KT2440. As shown in Figure 6, a high expression level was achieved by the MmP1 expression system showing 2.5-fold, 2.6-fold, 6.6-fold, 9.3-fold, and 10.1fold improvement compared to RhaS/P rha, LacI q -P trc AraC/P araB Xyls/P m , and LacI q -P tac, respectively. The expression level of the optimized T7 expression system was slightly higher than MmP1 by 1.2-fold. The expression level of RhaS/P rha was followed by the LacI q -P trc expression system and the AraC/P araB expression system was lower than the LacI q -P trc expression system, almost half of the RhaS/P rha expression system. LacI q -P tac and Xyls/P m showed a weak expression level. All expression systems showed more than 10-fold improved sfGFP expression in induced conditions compared to uninduced conditions, except for Xyls/P m , specifically, the ratio of induced expression to uninduced expression was 73, 82, 43, 35, 123, and 15 for T7, Mmp1, RhaS/P rha , LacI q -P trc , AraC/P araB , and LacI q -P tac expression system, respectively.

Verification of the Effectiveness of the MmP1 Expression System by the Expression of NDHase
The difficult-to-express NDHase from C. testosteroni JA1 was expressed using the MmP1 expression system both in KTCM and EMCM compared with KT2440 (Supplementary Figure 5). The activity of the NDHase using 3-cyanopyridine as substrate in the KTCM was 82.1 U/L, 3.6-fold higher than that in KT2440. The activity of the NDHase in the EMCM was 50.4 U/L, which is 2.2-fold higher than that in EM42 (Figure 7), exhibiting the excellent performance of the MmP1 expression system.

DISCUSSION
In previous research, the T7-like system MmP1 dramatically increased the GFP expression level in P. entomophila LAC31 (Zhao et al., 2017). Here, we would like to explore whether the T7-like system integrated  Frontiers in Chemistry | www.frontiersin.org July 2021 | Volume 9 | Article 664967