Combinatorial metabolic engineering of Pseudomonas putida KT2440 for efficient mineralization of 1,2,3-trichloropropane

An industrial waste, 1,2,3-trichloropropane (TCP), is toxic and extremely recalcitrant to biodegradation. To date, no natural TCP degraders able to mineralize TCP aerobically have been isolated. In this work, we engineered a biosafety Pseudomonas putida strain KT2440 for aerobic mineralization of TCP by implantation of a synthetic biodegradation pathway into the chromosome and further improved TCP mineralization using combinatorial engineering strategies. Initially, a synthetic pathway composed of haloalkane dehalogenase, haloalcohol dehalogenase and epoxide hydrolase was functionally assembled for the conversion of TCP into glycerol in P. putida KT2440. Then, the growth lag-phase of using glycerol as a growth precursor was eliminated by deleting the glpR gene, significantly enhancing the flux of carbon through the pathway. Subsequently, we improved the oxygen sequestering capacity of this strain through the heterologous expression of Vitreoscilla hemoglobin, which makes this strain able to mineralize TCP under oxygen-limited conditions. Lastly, we further improved intracellular energy charge (ATP/ADP ratio) and reducing power (NADPH/NADP+ ratio) by deleting flagella-related genes in the genome of P. putida KT2440. The resulting strain (named KTU-TGVF) could efficiently utilize TCP as the sole source of carbon for growth. Degradation studies in a bioreactor highlight the value of this engineered strain for TCP bioremediation.

thetic biology approaches can be applied for implantation of the synthetic biodegradation pathways into an ideal host strain for bioremediation 20 . In this work, a synthetic pathway for aerobic mineralization of TCP comprising DhaA from Rhodococcus rhodochrous NCIMB 13064, HheC from Agrobacterium radiobacter AD1, and EchA from Agrobacterium radiobacter AD1 was assembled in P. putida KT2440 (Fig. 1). All these genes (dhaA90R, hheC and echA) were integrated into the chromosome of P. putida KT2440 using a previously developed genome editing method with upp as a counter-selectable marker 16 . To improve the efficiency of the synthetic biodegradation pathway, the vgb gene was inserted into the P. putida chromosome, and both the glpR gene and the flagellar operon were knocked out. The resulting mutant strain was designated as P. putida KTU-TGVF. The successful construction of strain KTU-TGVF was verified by PCR and DNA sequencing. The correct amplicons were obtained by PCR with chromosomal DNA of strain KTU-TGVF as the template (Fig. S1). The nucleotide sequences of the inserted DNA fragments on chromosome were in accordance with those of four synthetic gene cassettes (Fig. S2).
The poor activity of DhaA toward TCP and cellular toxicity of TCP and intermediates represent primary bottlenecks of the TCP mineralization pathway, limiting the flux of carbon through the pathway 3, 12 . To overcome the bottlenecks of aerobic TCP mineralization, previous studies focused on engineering of DhaA, the first enzyme of the TCP mineralization pathway. Two DhaA mutants, DhaA31 and DhaA90R, were obtained using computer-assisted directed evolution. Mutant DhaA31 8 showed 26-fold higher catalytic efficiency toward TCP than the wild-type enzyme (DhaAwt: k cat = 0.08 s −1 , K m = 2.2 mM, k cat /K m = 36 M −1 s −1 ; DhaA31: k cat = 1.26 s −1 , K m = 1.2 mM, k cat /K m = 1050 M −1 s −1 ), while the (S)-DCP was accumulated in the pathway because of equimolar production of (R)-DCP and (S)-DCP by the non-enantioselective DhaA31 and the poor activity of HheC toward (S)-DCP. Mutant DhaA90R 9 possessed similar catalytic efficiency toward TCP as the wild-type enzyme (DhaA90R: k cat = 0.16 s −1 , K m = 6.5 mM, k cat /K m = 25 M −1 s −1 ), while DhaA90R converted prochiral TCP predominantly into (R)-DCP (ee 90%), which is the preferred substrate for the enantioselective HheC. In a previous study, the synthetic pathway with DhaA90R reconstructed in E. coli BL21 (DE3) failed to provide sufficient carbon flux to support bacterial growth in minimal medium with TCP 12 . In another study, a novel TCP biodegradation pathway including a chromosome-borne dhaA31 gene and an intrinsic nonselective DCP degradation pathway was reconstructed in a natural DCP-degrading bacterium P. putida MC4. Because of cellular toxicity of TCP and intermediates, the engineered P. putida MC4 showed a prolonged growth lag-phase when cultured in minimal medium with TCP 4 . To construct a practically applicable TCP-degrading bacterium, in this study, a robust environmental bacterium P. putida KT2440, which is resistant to toxic organic solvents, was chosen as a host strain for reconstructing the TCP mineralization pathway. To avoid the accumulation of (S)-DCP in the pathway, in this study, the enantioselective DhaA90R was selected for the dehalogenation of TCP to DCP. As discussed later in detail, the insufficient carbon flux through the pathway caused by the less active DhaA90R was further improved by eliminating the growth lag-phase of using GLY as a carbon source, enhancing aerobic metabolism, and increasing intracellular energy charge and reducing power.
Heterologous expression of the synthetic TCP mineralization pathway in P. putida KT2440. To achieve the optimal expression levels, the optimized gene expression regulatory elements 21 , including a strong constitutive promoter, two Shine-Dalgarno sequences, and a terminator, were designed for controlling the transcription and translation of exogenous genes (Fig. S3). In order to verify whether these exogenous genes (dhaA90R, hheC, echA and vgb) are transcribed to mRNA in P. putida KTU-TGVF, RT-PCR assays were performed. As expected, these target products were obtained by PCR with cDNA or genomic DNA as template. In contrast, no PCR products were obtained with mRNA or ddH 2 O as template (Fig. S4). These results indicated that transcription of these exogenous genes had occurred in P. putida KTU-TGVF.
Function validation for the synthetic TCP mineralization pathway. The functionality of DhaA, HheC and EchA in KTU-TGVF cells was validated by GC-MS analyses of resting-cell transformation products. Resting cells of KTU-T1 containing dhaA90R produced a new chromatographic peak with a retention time (RT) of 11.7 min when incubated with TCP, and this peak was identified as DCP by comparing its MS pattern with that of DCP standard ( Fig. 2A). Transformation of DCP by resting cells of KTU-T2 containing hheC produced a new chromatographic peak with a RT of 5.1 min, which had the same RT and MS pattern as ECH standard (Fig. 2B). After the transformation of ECH by resting cells of KTU-T3 containing echA, a new chromatographic peak appeared at a RT of 12.3 min, the RT and MS pattern of this peak were the same as those of CPD standard (Fig. 2C). Therefore, KTU-T1, KTU-T2 and KTU-T3 cells expressed functional DhaA, HheC and EchA, respectively.
TCP degradation experiment was performed with KTU-TGVF in M9 minimal medium supplemented with 0.5 mM TCP. As shown in Fig. 3A, 0.5 mM TCP was almost completely converted to GLY and the vast majority of GLY were further metabolized within 32 h, as revealed by quantification of DCP, ECH, CPD and GLY throughout the experiment. Accompanying with TCP degradation, cell growth was observed (Fig. 3B), which indicated that this strain could utilize TCP as the sole source of carbon for growth. In contrast, the concentration of TCP did not change and no growth was observed when inoculated with KTU. These results indicated that TCP could be converted to GLY by the heterologous pathway assembled in P. putida KT2440 (Fig. 1). Furthermore, GLY could be utilized as the carbon source for cell growth via the intrinsic pathway of KT2440. Taken together, we concluded that the pathway for aerobic mineralization of TCP was functionally assembled in P. putida KT2440.
Elimination of the growth lag-phase of using GLY as a carbon source. GlpR functions as a negative regulator for controlling the utilization of GLY as a growth precursor in P. putida KT2440 and the regulatory mechanisms have been elucidated 22 . Deletion of glpR may eliminate the growth lag-phase of using GLY as a carbon source. In this work, the strain KTU-TGVF (ΔglpR) could grow aerobically without the lag-phase in minimal medium with TCP as the sole carbon source. In contrast, no growth was observed with the strain KTU-T123 (glpR + ) when using TCP as the sole carbon source. These results suggest that the efficient utilization of GLY may support bacterial growth, alleviate cellular toxicity of TCP and intermediate metabolites and improve the flux of carbon through the TCP mineralization pathway.
VHb enhances TCP mineralization under oxygen-limited conditions. CO-difference spectrum assays were used to detect VHb activity. The cell extracts from the strain KTU-TGVF showed a characteristic absorption peak at 420 nm when fed with CO, while the peak was not observed with the cell extracts from the strain KTU fed with CO (Fig. S6). These results indicated that the strain KTU-TGVF expressed functional VHb. VHb delivers the O 2 to the respiratory chain, enhancing ATP generation and NADH consumption at low O 2 concentrations. This process could improve the efficiency of bacterial aerobic respiration, increase the carbon flux through the tricarboxylic acid cycle and improve the aerobic growth of bacteria. Heterologous expression of VHb in a variety of hosts has been shown to improve cell growth, protein synthesis, metabolite productivity, and bioremediation under oxygen-limited conditions 23,24 . In a previous study, integrating vgb into the chromosomes of P. aeruginosa and Burkholderia sp. strain DNT could improve growth and degradation of 2,4-dinitrotoluene or benzoic acid under hypoxic conditions 25 . In another study, P. putida KT2440 was engineered for the anoxic biodegradation of 1,3-dichloroprop-1-ene by introducing various genes retrieved from facultative anaerobe and aerotolerant bacteria 26 .
Sufficient oxygen is the key factor for complete oxidative mineralization of TCP. The success of aerobic mineralization of TCP in oxygen-restricted environments will largely depend on the oxygen sequestering capability of TCP-degrading bacteria. In this work, significant growth was observed for the strain KTU-TGVF expressing VHb under oxygen-limited conditions, while no growth was observed for the strain KTU-TGF without VHb expression under oxygen-limited conditions (Fig. 4). These results suggest that VHb may improve the ability of P. putida KT2440 to compete for limited oxygen in hypoxic environments, thus making this bacterium more competitive in actual environments such as a packed-bed bioreactor. This study not only underscores the value of P. putida KT2440 as a versatile biocatalyst for biotransformation under oxygen-limited conditions but also highlights the value of metabolic engineering for expanding the catalytic repertoire of P. putida KT2440.

Increased intracellular ATP and NADPH levels improve aerobic mineralization of TCP.
Both the production and the motion of flagella are energy-consuming processes for the cell. In E. coli, flagellar production consumes about 2% of the biosynthetic energy expenditure of the cell, while flagellar motion demands about 0.1% of the total energy cost 27 . In this work, a non-flagellated strain KTU-TGVF was generated from the original strain KTU-TGV by deleting the flagellar operon in the genome. Accordingly, the flagellar operon knockout mutant strain showed the complete absence of the flagellum structure when observed using TEM after negative staining. In contrast, the flagella were observed in the wild-type strain KT2440 (Fig. S7). These results indicated that the flagellar synthesis was completely blocked in the mutant strain.
It has been reported that the non-flagellated strain KT2440 showed increased intracellular ATP and NADPH levels 28 . In our study, the ATP/ADP ratio in the strain KTU-TGVF was 1.3-fold higher than that in the strain KTU-TGV (Fig. 5A). Since the ATP levels in bacterial cells are tightly regulated, a difference of 30% could significantly influence cellular functions. These results indicated that eliminating the consumption required for flagella synthesis and rotation greatly changed the energy status of the cells.
Except for the energy consumption, a considerable amount of reducing power, mainly in the form of NADPH, is indispensable for basic anabolic processes, such as the biosynthesis of building blocks for biomass 29 . In this work, the strain KTU-TGVF had a 1.2-fold higher NADPH/NADP + ratio than did the strain KTU-TGV (Fig. 5B). Furthermore, the growth kinetics of the strain KTU-TGVF and KTU-TGV were compared. The strain KTU-TGVF showed a higher growth rate in M9 minimal medium containing 0.5 mM TCP than did the strain KTU-TGV. The KTU-TGVF and KTU-TGV cultures reached a maximum OD 600 of 0.104 and 0.088 at 32 h, respectively (Fig. S8). Obviously, the strain KTU-TGVF could utilize TCP more efficiently as the sole carbon source for growth, which may be a reflection of increased resistance to cellular toxicity of TCP and intermediate  metabolites. Increased resistance to toxic substrates may be due in part to enhanced anabolic capability resulting from the increase of ATP and NADPH levels within the cell. Taken together, the efficient mineralization of TCP by the strain KTU-TGVF may be attributed to the efficient utilization of GLY as a growth precursor and the increased intracellular ATP and NADPH levels.

Treatment of TCP-contaminated water in a lab-scale bioreactor. The flagella-deficient strain
KTU-TGVF showed a distinct change in biofilm formation. The quantities of biofilm formed by the strain KTU-TGVF were 4.3-and 9.5-fold higher at 24 and 96 h, respectively, than those formed by the strain KTU-TGV, as judged by the CV staining assay (Fig. S9), when these strains were grown in M9 minimal medium containing 0.4% (w/v) glucose. The observation is in agreement with a previous report 28 , in which the non-flagellated P. putida KT2440 cells form more biofilm than the wild-type KT2440 cells. Recently, the biofilm formation capability of P. putida KT2440 was improved for boosting biodegradation of haloalkanes by the modulation of the intracellular c-di-GMP level 30 . Biofilm offers protection against a hostile environment and helps bacteria persist within the environment. Therefore, the strain KTU-TGVF has potential for use in a packed-bed bioreactor treating TCP-contaminated water because its high biofilm formation ability is favorable for the immobilization of cells on the support material in wastewater treatment.
To explore the feasibility of using strain KTU-TGVF for TCP bioremediation, we designed a packed-bed bioreactor and tested the removal of TCP from wastewater streams in a consecutive process (Fig. S10). KTU-TGVF cells were grown in M9 minimal medium containing 0.5 mM TCP plus 0.4% glucose to OD 600 = 0.4 and were subsequently inoculated into the reactor for immobilization on ceramic rings. Previous studies have shown that ceramic rings as the support material are suitable for the immobilization of microbial cells 4,31 . After inoculation, the reactor was operated under fed-batch conditions for 4 days to promote attachment of the cells to the support material. Within a period of 4 to 18 days, TCP (0.05 to 0.2 mM) was supplied continuously at a rate of 0.1 ml/min. During the operating period, the maintenance of a high hydraulic retention time (HRT, 133 h) may be favorable for cell growth and biofilm formation. HRT is considered as an important operating parameter directly influencing the bioreactor performance 4, 31 . After 18 days, the influent concentration of TCP was maintained at 0.2 mM, and the reactor was continuously operated for another 30 days. The reactor performance was evaluated by continuous monitoring of the effluent concentrations of TCP, DCP and chloride (Fig. 6). Under these operating conditions, the removal efficiencies of TCP were 95 to 97% and quantitative release of chloride (TCP/chloride, molar ratio 1:3) was observed. We detected minute quantities of DCP in the reactor effluent. When inoculated with the strain KTU, the TCP effluent concentration remained unchanged relative to the influent concentration.
Cells from the reactor effluent were spread on LB agar plates and colonies formed were checked for their identity. All 20 colonies tested were identified as P. putida KT2440 by the 16 S rRNA gene sequencing. Furthermore, both the four gene insertions (dhaA90R + , hheC + , echA + , vgb + ) and the two gene deletions (∆glpR, ∆flagellar operon) were detected by PCR from all 20 colonies tested, and the detection results of one colony are shown in Fig. S11. These results demonstrated that the inoculated KTU-TGVF cells thrived and accounted for the observed TCP degradation during a 30-day operating period in the bioreactor. In the future, the efficient TCP-mineralizing strain KTU-TGVF coupled with a field-scale reactor has enormous potential to be applied for the treatment of industrial wastewater containing TCP.

Conclusions
In this work, an efficient TCP-mineralizing strain was created using combinatorial engineering strategies. In brief, we first integrated a synthetic pathway for aerobic TCP mineralization into the chromosome of P. putida KT2440 by functional assembly of various TCP-degrading enzymes from different bacteria. We then enhanced the utilization of TCP as a carbon source by eliminating the growth lag-phase of using GLY as a growth precursor. We additionally engineered the strain with the ability to mineralize TCP under oxygen-limited conditions through the heterologous expression of VHb. Lastly, we further improved the utilization of TCP by increasing intracellular ATP and NADPH levels resulting from deletion of the flagellar operon. We envision that these combinatorial engineering approaches could provide insights into devising engineering strategies to improve the degradation of anthropogenic chemicals by P. putida.
E. coli strains were grown at 37 °C in LB medium supplemented with 50 μg/ml kanamycin (Kan) if necessary. P. putida strains were grown at 30 °C in M9 minimal medium 16 containing 0.5 mM TCP as well as in LB medium 32 supplemented with 50 μg/ml Kan or 20 μg/ml 5-fluorouracil (5-FU) when required. DNA manipulation and strain constructions. Nine P. putida mutant strains with multiple gene insertions and/or deletions were constructed as described previously using a chromosomal scarless modification strategy 19 and a suicide plasmid pK18mobsacB 33 . The detailed procedures for the construction of these nine P. putida mutant strains are described in Supplementary methods. The detailed information on the inserted exogenous genes (dhaA90R, hheC, echA and vgb) and deleted target genes (glpR and flagellar operon) is shown in Table 1. These four gene cassettes (dhaA90R, hheC, echA and vgb) were chemically synthesized by Genscript Co. Ltd., Nanjing, China and their nucleotide sequences are shown in Fig. S2. The strains, plasmids, and primers used in this study are listed in Table 2.
Transformation of plasmid into P. putida was carried out using the electroporation method 34 . Screening of P. putida mutant strains was performed as described previously 19 . In brief, the single-crossover recombinants were screened by incubating at 30 °C for 24 h on LB agar plates supplemented with 50 μg/ml Kan. The selected recombinants were then incubated at 30 °C for 24 h in LB medium. To further screen the double-crossover recombinants, the culture broths that had been diluted to 10 −5 were spread on LB agar plates supplemented with 20 μg/ml 5-FU. The selected recombinants showing 5-FU r and Kan s were further checked by PCR using primers listed in Table 2. All the constructed mutant strains were validated by DNA sequencing.

RT-PCR assays. Total RNA was extracted from KTU-TGVF cells using a RNApure Bacteria Kit (CWBIO,
China) and treated with DNase I at 37 °C for 30 min. RNA quality was checked by an Agilent 2100 Bioanalyzer. cDNA was synthesized with 0.5 μg total RNA as template using a PrimeScript RT Master Mix Kit (TaKaRa, Japan). PCR was carried out with cDNA as the template using PrimeSTAR HS DNA polymerase (TaKaRa) and primers listed in Table 2 on an ABI 2720 thermal cycler (Applied Biosystems). Genomic DNA, mRNA or ddH 2 O was used as the template in the control reactions. PCR products were detected by agarose gel electrophoresis.   Table 2. Strains, plasmids, and primers used in this study.

Resting-cell transformation assays.
organic layer was recovered carefully. The extraction process was repeated three times. The organic layers were pooled and then dried over anhydrous Na 2 SO 4 . Then, the concentrations of TCP, DCP, ECH and CPD were measured by GC-MS. GLY was quantified using a free GLY assay kit (Sigma, USA). Samples of 1 μl (diluted if necessary) were injected directly for GC-MS analysis, which was obtained on an Agilent Technologies 7890A-5975C (Agilent Technologies, Palo Alto, CA, USA). The GC was equipped with a HP-5 capillary column (30 m × 0.25 mm × 0.25 µm) and operated in split ratio of 20:1. Helium (>99.999%) was used as carrier gas with a constant flow rate of 0.6 ml/min. The GC conditions were as follows: injector temperature of 250 °C, oven temperature maintained at 35 °C for 6 min and then increased to 250 °C at a rate of 10 °C/min, and interface temperature of 230 °C. The mass spectrum conditions were as follows: ionization energy of 70 eV, scan range of 30-300 amu, and ion chamber temperature of 250 °C with tungsten filament used for ionization of molecules. The concentration of metabolites was determined using a standard curve of peak area versus concentration. For identification of metabolites, the mass spectra of metabolites were compared with those of their authentic standards or the same compound available in the National Institute of Standard Technology library, USA.
VHb detection. VHb activity was detected by carbon monoxide (CO)-difference spectra. The complex consisting of CO and reduced hemoglobin has a characteristic peak at 420 nm 35 . KTU-TGVF or KTU cells were suspended in 50 mM potassium phosphate buffer (pH 7.2) and broken using an ultrasonic disruptor on ice (200 W for 15 min with cycles of sonication of 10 s each and 15 s pause). The crude extracts were centrifuged at 10,000 rpm and 4 °C for 15 min to remove cell debris. The soluble cellular fraction was exposed to CO for 2 min. Then, hemoglobin levels were determined by CO-difference spectra using a SHIMADZU UV-1800 spectrophotometer (SHIMADZU, Japan). Determination of the ATP/ADP and NADPH/NADP + molar ratios. Cells were grown in M9 minimal medium containing 0.2% (w/v) glucose at 30 °C for 12 or 20 h. The intracellular levels of ADP and ATP were measured using an ADP/ATP ratio bioluminescence assay kit (BioVision, USA). The intracellular concentrations of NADP + and NADPH were measured using an EnzyChrom ™ NADP/NADPH assay kit (BioAssay Systems, USA).
Biofilm formation assays. Biofilm formation was quantified using the crystal violet (CV) staining assay 30 .
Cells were grown overnight at 30 °C and 170 rpm in M9 minimal medium supplemented with 0.4% (w/v) glucose. The OD 600 of the cultures was adjusted to 0.05 before inoculating 200 μl into 96-well microtiter plates. The plates were incubated at 30 °C without shaking for 24 or 96 h, after which the culture broth was removed from the plates and the OD 600 was measured to estimate the planktonic cells in the culture. The plates were washed three times with H 2 O to remove all non-adhered cells, and stained with 0.1% (w/v) CV for 30 min at room temperature. After discarding the dye, the plates were washed with H 2 O, dried and the remaining stain was dissolved in 33% (v/v) acetic acid. The plates were agitated gently for 1 h. Biofilm was quantified by measuring the absorbance at 590 nm (A 590 ) using a BIOLOG microplate reader. The biofilm index was calculated as the ratio of biofilm formation to planktonic cell density (A 590 /OD 600 ).

Degradation of TCP by KTU-TGVF in a lab-scale bioreactor.
A lab-scale reactor was designed for treatment of TCP-contaminated water (Fig. S10), which was mainly composed of a cylindrical vessel of polymethyl methacrylate with a volumetric capacity of 2 liters and a working volume of 0.8 liter. The reactor was filled with 500 g ceramic Raschig rings (outer diameter, 8 mm; internal diameter, 4 mm; length, 8 mm; Pingxiang Tower Packing Co.) as packing material. TCP-contaminated water was introduced into the reactor using a peristaltic pump from a glass vessel (0.5 liter) containing TCP in water. The airflow was regulated with a gas flow controller. The pH was controlled by adding 0.1 M NaOH when necessary.
For startup, the reactor was inoculated with a 800 ml suspension of cells grown on 0.5 mM TCP plus 0.4% glucose (OD 600 = 0.4), which was left for 4 days without a supply of TCP-contaminated water. During this period, air was introduced into the reactor at a rate of 0.1 ml/min. After 4 days of inoculation, a wastewater stream containing TCP was introduced continuously, and the influent concentration of TCP was increased gradually from 0.05 to 0.2 mM. From day 18 onward, about 0.2 mM TCP was introduced into the reactor at a rate of 0.1 ml/min, and the flow rate was maintained during the operating period. The airflow was maintained at 1 ml/min to provide sufficient oxygen for aerobic mineralization of TCP. The effluent concentrations of TCP, DCP, and chloride ions were measured by GC-MS and a chloride assay kit (Sigma, USA) during a 30-day operating period.