Engineering of Pseudomonas taiwanensis VLB120 for Constitutive Solvent Tolerance and Increased Specific Styrene Epoxidation Activity

Abstract

The application of whole cells as biocatalysts is often limited by the toxicity of organic solvents, which constitute interesting substrates/products or can be used as a second phase for in situ product removal and as tools to control multistep biocatalysis. Solvent-tolerant bacteria, especially Pseudomonas strains, are proposed as promising hosts to overcome such limitations due to their inherent solvent tolerance mechanisms. However, potential industrial applications suffer from tedious, unproductive adaptation processes, phenotypic variability, and instable solvent-tolerant phenotypes. In this study, genes described to be involved in solvent tolerance were identified in Pseudomonas taiwanensis VLB120, and adaptive solvent tolerance was proven by cultivation in the presence of 1% (vol/vol) toluene. Deletion of ttgV, coding for the specific transcriptional repressor of solvent efflux pump TtgGHI gene expression, led to constitutively solvent-tolerant mutants of P. taiwanensis VLB120 and VLB120ΔC. Interestingly, the increased amount of solvent efflux pumps enhanced not only growth in the presence of toluene and styrene but also the biocatalytic performance in terms of stereospecific styrene epoxidation, although proton-driven solvent efflux is expected to compete with the styrene monooxygenase for metabolic energy. Compared to that of the P. taiwanensis VLB120ΔC parent strain, the maximum specific epoxidation activity of P. taiwanensis VLB120ΔCΔttgV doubled to 67 U/g of cells (dry weight). This study shows that solvent tolerance mechanisms, e.g., the solvent efflux pump TtgGHI, not only allow for growth in the presence of organic compounds but can also be used as tools to improve redox biocatalysis involving organic solvents.

INTRODUCTION

For oxyfunctionalizations in chemical synthesis, the use of metabolically active whole cells as biocatalysts provides several advantages over the use of isolated enzymes. This includes the regeneration of required cofactors, increased enzyme stability, the possibility to combine multiple enzyme reactions, and the deactivation of reactive oxygen species (1–4). However, the efficiency of whole cells as biocatalysts is often limited by the toxicity of apolar substrates and products (5–9), which typically involves detrimental effects on cellular membranes (10). Solvents having a log partition coefficient in an equimolar mixture of octanol and water (P_OW) between 1 and 4 are regarded to be extremely toxic, as they readily intercalate into the membrane (11, 12). To reduce such toxic effects, the introduction of a second water-immiscible phase, functioning as a reservoir and sink for substrates and products, respectively, has been shown to be a suitable tool for biotransformations with toxic organic compounds (13–17). Such a carrier solvent not only minimizes toxic and inhibitory effects of the involved compounds but also enables control of product formation, making use of substrate and product partitioning to shift chemical equilibria and to exploit reaction kinetics for multistep biocatalysis (5, 6, 16, 18). For such two-liquid-phase applications, solvent toxicity limits solvent choice (19). Octanol and toluene, which are solvents with low log P_OW values, are, e.g., favorable for the extraction of highly toxic chemicals with intermediate polarity, such as phenols, lactones, or catechols, but are themselves toxic (14, 20, 21). Also with nontoxic second phases, product toxicity may still restrict the maximally achievable product concentration (22). In these cases, the use of solvent-tolerant bacteria has been proposed as a means to overcome toxicity issues (14, 17, 23).

Some solvent-tolerant bacteria, often Pseudomonads, are able to survive sudden exposure to toxic organic solvents (24–27), and some are even able to grow in the presence of a second phase of toxic solvents in complex media (28–32). Such solvent tolerance is based on a combination of several adaptive mechanisms, which have been intensively investigated and reviewed for Pseudomonas putida S12 and DOT-T1E (14, 33, 34). It was shown that energy-dependent solvent efflux mediated by membrane pumps belonging to the resistance nodulation division (RND) family is one of the main mechanisms counteracting solvent toxicity (35, 36) and acts as an inducible long-term response involving transcriptional regulation (37). For P. putida DOT-T1E, TtgV has been shown by Rojas et al. to repress TtgGHI solvent efflux pump gene expression (26). Accordingly, a knockout of the ttgV gene was found to improve survival of P. putida DOT-T1E upon a short-term solvent shock. Typically, tedious adaptation procedures are required to establish a solvent-tolerant phenotype (20). These procedures often suffer from variable growth behaviors and/or productivity lag phases of several hours (21, 38) and usually require a stepwise increase of the solvent concentration (39, 40). Furthermore, the solvent-tolerant phenotype is reversible and is lost after the removal of solvent stress. Poorly controllable adaptation processes, phenotypic variability, and the instability of the solvent-tolerant phenotype are critical factors for industrial implementation.

In this study, the solvent-tolerant styrene degrader Pseudomonas taiwanensis VLB120 (41) (K. A. K. Köhler, L. M. Blank, B. E. Ebert, C. Rückert, J. Kalinowski, A. Schmid, submitted for publication) was investigated regarding its native tolerance toward toxic compounds such as toluene and styrene and, on the genome level, for homologues of known solvent tolerance genes. With the goal to avoid tedious adaptation, regulatory mutants of P. taiwanensis VLB120 and its styrene oxide isomerase (StyC)-lacking P. taiwanensis VLB120ΔC mutant strain, efficiently catalyzing the conversion of styrene into (S)-styrene oxide (42), were constructed by deletion of the ttgV gene. Solvent tolerance and specific styrene epoxidation characteristics of these P. taiwanensis VLB120ΔC mutants were investigated, giving evidence for constitutive solvent tolerance during growth and improved biocatalyst performance regarding not only stability but, most interestingly, also specific styrene epoxidation activity.

MATERIALS AND METHODS

Chemicals, bacterial strains, and culture conditions.

All chemicals used in this study were purchased from Sigma-Aldrich (Steinheim, Germany) or Carl-Roth (Karlsruhe, Germany) at the highest purity available. Strains and plasmids used in this study are listed in Table 1. Escherichia coli DH5α and E. coli DH5α λpir were used for cloning purposes. Cells were cultivated either in Luria-Bertani (LB) broth (43) or M9 medium (44) supplemented with 1 ml liter⁻¹ US* trace element solution (45), 2 ml liter⁻¹ 1 M MgSO₄ · 7H₂O, and 0.5% (wt/vol) glucose as the sole carbon source. Antibiotics (50 μg ml⁻¹ kanamycin [Km⁵⁰], 25 μg ml⁻¹ gentamicin [Gm²⁵], or 100 μg ml⁻¹ streptomycin [Sm¹⁰⁰]) were added when appropriate. Single colonies were picked from LB agar plates and transferred into 5 ml LB medium. After 8 h, 500 μl was transferred to 50 ml M9 medium supplemented as described above. Overnight cultures were then used to inoculate experimental cultures to an optical density at 450 nm (OD₄₅₀) of 0.2. Cultivations were performed in screw-cap, baffled Erlenmeyer shaking flasks with a gas/liquid ratio (vol/vol) of 9:1 in a Multitron standard shaker (Infors HT, Bottmingen, Switzerland) at 200 rpm (25-mm amplitude) and 30°C.

TABLE 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Characteristicsa	Reference/source
Strains
E. coli DH5α	supE44 ΔlacU169 (ϕ80 lacZΔM15) hsdR17 recA1 endA1 gyrA96 thi-1 relA1	69
E. coli DH5α λpir strain	λpir lysogen of DH5α	52
P. taiwanensis VLB120	Wild-type Pseudomonas; styrene prototroph	41; Köhler et al., submitted
P. taiwanensis VLB120ΔC	styC-deficient mutant of Pseudomonas taiwanensis VLB120 (styC disrupted by Smr gene); is no longer able to grow on styrene	42
P. taiwanensis VLB120ΔttgV	Pseudomonas taiwanensis VLB120, ttgV-knockout mutant	This work
P. taiwanensis VLB120ΔCΔttgV	Pseudomonas taiwanensis VLB120, ttgV- and styC-knockout mutant	This work
Plasmids
pJQ200SK	Gmr, suicide vector, P15A ori sacB RP4 pBluescriptSK MCS	49
pJQhdp::Km	Kmr Gmr, pJQ200SK with hdp disrupted by kanamycin resistance gene, flanked by loxP recombination sites	N. Wierckx, unpublished data
pJQttgV::Km	Kmr Gmr, pJQ200SK with ttgV disrupted by kanamycin resistance gene, flanked by loxP recombination sites	This work
pJTNcre	Gmr, encodes Cre recombinase	48
pEMG	Kmr oriR6K, lacZα with two flanking I-SceI sites	52
pEMG-ttgV	Kmr, pEMG containing a 1.0-kb EcoRI-XbaI fragment (0.5 kb up- and downstream of ttgV)	This work
pSW-2	Gmr oriRK2 xylS Pm→I-SceI (transcriptional fusion of I-SceI to Pm)	52

^aGm^r, gentamicin resistance; Km^r, kanamycin resistance; Sm^r, streptomycin resistance.

Growth in the presence of organic solvents.

To follow growth in the presence of organic solvents, cultures were grown from an OD₄₅₀ of 0.2 into the early exponential growth phase (∼OD₄₅₀ of 0.5) before the organic phase was added. Depending on the experimental setup, an evaporation container filled with 1 ml styrene or toluene was used to saturate the headspace and minimize solvent loss by evaporation. To avoid interference of organic solvent droplets with OD measurements, samples were diluted 10-fold, well mixed, and stored for 60 s at room temperature for phase separation (39).

Whole-cell activity assays with resting cells.

Resting cell assays were carried out as described previously (41, 42). After the cells had grown to an OD₄₅₀ of 0.5, cells were induced via addition of 1 mM pure styrene to the aqueous phase. Additionally, styrene was continuously provided via the gas phase using an evaporation container filled with 1 ml pure styrene. For induction studies, the amount of styrene added into the aqueous phase was varied, and experiments were done with and without an evaporation container. The induction time was 3 h, as activities did not further increase with longer induction times (42). Cells were harvested by centrifugation at 4°C and 4,500 × g for 10 min (Kendro Heraeus Multifuge, Langenselbold, Germany) and resuspended to a concentration of 0.5 g of cells (dry weight) liter⁻¹ in 0.1 M potassium phosphate buffer (pH 7.4, 10 g liter⁻¹ glucose). Aliquots of 1 ml were transferred into Pyrex tubes and equilibrated for 5 min at 30°C and 300 rpm in a water bath. Styrene epoxidation was started with the addition of 1.5 mM styrene from a 100 mM stock in pure EtOH. After 5, 10, 20, 30, 60, 90, and 120 min, epoxidation reactions were stopped by the addition of 1 ml ice-cold diethyl ether (also serving as the extractant) containing 0.2 mM n-decane as the internal standard and vortexed for 30 s. Subsequently, phases were separated via short centrifugation, and the organic phase was dried over anhydrous sodium sulfate before being transferred into GC vials for analysis. Differing from the previously described whole-cell resting assays (42), the concentration of styrene oxide remaining from the induction process was determined at time point 0 and for activity determination subtracted from the concentrations determined after 5, 10, 20, 30, 60, 90, and 120 min.

Molecular biology methods.

The targeted disruption of ttgV (AGZ37957.1) located on the 312-kb megaplasmid pSTY of P. taiwanensis VLB120 (46) was performed as described earlier for gene knockouts in P. putida S12 (47, 48). The gene replacement vector pJQttgV::Km was constructed from pJQ200SK (49) with primers 1 to 4 listed in Table 2, and P. taiwanensis VLB120 was transformed by electroporation (50). After screening for Km resistance and Gm sensitivity by replica plating on LB plates with the respective antibiotics, the disruption of ttgV by the Km^r cassette was confirmed by colony PCR (51). The Km^r cassette was removed by introducing a Cre recombinase located on pJTNcre, which catalyzes site-specific recombination at the loxP sites flanking the Km^r gene. Strains were cured from pJTNcre by repetitive cultivation on unselective LB medium in baffled Erlenmeyer shaking flasks. The construction of P. taiwanensis VLB120ΔttgV was confirmed by sequencing of the disrupted gene at LGC genomics (Berlin, Germany) using primers 5 and 6 (Table 2).

TABLE 2.

Oligonucleotide primers used in this study

Primer	Sequence (5′→3′)a	Characteristics
1	ATAAGAATGCGGCCGCATGAACCAATCAGACGAAGT	Start of ttgV, forward primer, NotI
2	GCTCTAGACAGCTCTCGTTCTGACACGA	Positions 371–390 in ttgV, reverse primer, XbaI
3	GCTCTAGACGTGTGGTGTTCCCGATTGG	Positions 391–410 in ttgV, forward primer, XbaI
4	CGGGATCCCTAGGGCGCTTTCTTTGACG	End of ttgV, reverse primer, BamHI
5	AGTGGCGGCCATAACAGATGC	Positions 27–47 upstream of ttgV, forward primer
6	ATGATGCGCTGTCGTGTCTCC	Positions 35–55 downstream of ttgV, reverse primer
7	CGGAATTCTAGGTACGCGGGTCAATTTG	Positions 506–526 upstream of ttgV, forward primer, EcoRI
8	TTTAACTAGGGCGCTTTCTTGTCTGATTGGTTCATATCTC	Overlap extension reverse primer, 20-bp overlap to beginning of ttgV, 20-bp overlap to end of ttgV
9	GAGATATGAACCAATCAGACAAGAAAGCGCCCTAGTTAAA	Overlap extension forward primer, 20-bp overlap to end of ttgV, 20-bp overlap to beginning of ttgV
10	GCTCTAGATGGCTGCGTAGGCCCTCATAA	Positions 506–527 downstream of ttgV, reverse primer, XbaI
11	GGTCGGTTGTTAGGGTCAGAG	Positions 370–390 upstream of ttgV, forward primer
12	GATCAGACCCGATTCGTCCAC	Positions 339–359 downstream of ttgV, reverse primer
13	CAGGTTTCCCGACTGGAAAGCG	Binding in lacZ, forward primer
14	GCAGTCCGACACCATCAAAA	qPCR forward primer, positions 693–712 in ttgV
15	GATGGCGCGCTCTATATTCTG	qPCR reverse primer, positions 733–753 in ttgV
16	TCCTCTAACGGCGCTGATG	qPCR forward primer, positions 39–57 in ttgG
17	GGCGACGTGGTCTTTTTCA	qPCR reverse primer, positions 75–93 in ttgG
18	CATCGAAGACGCGATTGCT	qPCR forward primer, positions 2304–2322 in gyrB
19	CGCGTCGCCCATCAAG	qPCR reverse primer, positions 2343–2358 in gyrB
20	ATACGGCGCGGCATACA	qPCR forward primer, positions 3885–3901 in rpoB
21	CGTTCACGTCGTCCGACTT	qPCR reverse primer, positions 3925–3943 in rpoB
22	GGGTGCTATCGCCCACTTC	qPCR forward primer, positions 399–417 in rpoD
23	CGGTCATATTCGCCGAGAAT	qPCR reverse primer, positions 436–455 in rpoD

^aRestriction sites are underlined.

Due to the reduced screening effort, a different, recently developed method was chosen to delete ttgV in P. taiwanensis VLB120ΔC (52). A loss of the megaplasmid pSTY due to I-SceI cleavage could be avoided by the addition of Sm¹⁰⁰ due to a resistance cassette disrupting the styC gene (AGZ38050.1) located on the megaplasmid pSTY. The plasmid pEMG-ttgV was constructed by fusing 500 bp of both flanking regions of ttgV, amplified with primers 7 to 10 (Table 2), with overlap extension PCR and ligating the resulting fragment into pEMG using EcoRI and XbaI restriction sites. Screening for successful ligation events was done in E. coli DH5α λpir cells, which were plated on LB agar plates containing Km⁵⁰, 1.0 mM IPTG (isopropyl-β-d-thiogalactopyranoside), and 40 μg liter⁻¹ X-Gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside) for blue-white screening. P. taiwanensis VLB120ΔC cells transformed with pEMG-ttgV were plated on LB agar plates containing Km⁵⁰, and integration of pEMG-ttgV was confirmed by colony PCR. If positive, cells were transformed with pSW-2 to select for the second recombination via I-SceI cleavage and plated on LB agar plates containing Gm²⁵, followed by replica plating on LB agar plates containing Gm²⁵ and Km⁵⁰. Compared to the original method, induction with 15 mM 3-methyl-benzoate was not necessary to obtain colonies able to grow in the presence of Gm²⁵ and not in the presence of Km⁵⁰, in which correct ttgV deletion was confirmed using primers 11 to 13 (Table 2). One positive clone was grown in LB medium for curing of pSW-2, which was verified by replica plating on LB agar plates with and without Gm²⁵.

RNA isolation and quantification.

Cells of P. taiwanensis VLB120ΔC and VLB120ΔCΔttgV were harvested in the early exponential growth phase (OD₄₅₀ of 0.75, LB medium). Total RNA was isolated using a NucleoSpin RNA II kit by following the protocol provided by the manufacturer (Macherey-Nagel, Düren, Germany). The Ambion TURBO DNA-free kit (Life Technologies, Darmstadt, Germany) was used to remove DNA contaminations. RNA concentration and purity (A₂₆₀/A₂₃₀) were determined with a NanoDrop 2000 spectrophotometer (Peqlab, Erlangen, Germany). Three micrograms of total RNA were transcribed into first-strand cDNA with the GoScript reverse transcription system (Promega, Madison, WI, USA) using random primers, assuming 100% transcription efficiency. Quantitative PCR (qPCR) primers were designed using the Primer Express software (Life Technologies, Darmstadt, Germany; Table 2). mRNA levels of ttgG (AGZ37958.1) and ttgV were quantified in triplicates on a StepOne Plus real-time PCR system (Life Technologies, Darmstadt, Germany) with the appropriate software using an ROX FastStart universal SYBR green master mix (Roche Diagnostics, Mannheim, Germany). Raw data were processed using the comparative threshold cycle (C_T) method (53) with the housekeeping genes gyrB (AGZ32814.1), rpoB (AGZ37372.1), and rpoD (AGZ35950.1) (46) serving as endogenous controls. Relative quantification (RQ) was performed for ttgG transcripts, giving the transcript level in P. taiwanensis VLB120ΔCΔttgV relative to the level in P. taiwanensis VLB120ΔC (53).

Analytical procedures.

Cell (dry weight) concentrations were determined by measuring the optical density at 450 nm (Libra S11 spectrophotometer; Biochrom Ltd., Cambridge, United Kingdom) using a correlation factor of 0.186 g of cells (dry weight)/liter/OD₄₅₀ (54). Styrene and styrene oxide concentrations were quantified on a TRACE GC UltraTM gas chromatograph (Thermo Fisher Scientific Inc., Waltham, MA, USA) equipped with a FactorFour VF-5ms column (30-m length, 0.25-mm diameter, 0.25-µm film thickness) (Varian, Inc., Paolo Alto, CA, USA) using n-decane as the internal standard. The temperature profile was identical to the one published by Kuhn et al. (22). Specific epoxidation activities are given in units (U) per gram of cells (dry weight), with 1 U defined as 1 μmol product formed per minute. Glucose and gluconate concentrations were quantified using a LaChrom Elite high-performance liquid chromatography (HPLC) system (Hitachis High Technologies America, Inc., Pleasonton, CA, USA) equipped with a Trentec 308R-Gel.H ion exclusion column (300 by 8 mm; Trentec Analysentechnik, Gerlingen, Germany) as described earlier (22).

Bioinformatic analyses.

The open-source genome annotation system GenDB (55) and the National Center for Biotechnology Information BLAST server (NCBI) (56) were used to screen the genome of P. taiwanensis VLB120 (46) for homologues of genes known to be involved in solvent tolerance.

Nucleotide sequence accession no.

The nucleotide sequences described in this study are available via GenBank/EMBL under accession no. CP003961 (chromosome) and CP003962 (Megaplasmid pSTY).

RESULTS

P. taiwanensis VLB120 harbors solvent tolerance gene homologues.

To determine the genetic basis for the solvent tolerance of P. taiwanensis VLB120, BLAST analyses were performed based on its recently sequenced genome (46). For this purpose, genes of various E. coli and Pseudomonas strains known to be involved in solvent tolerance, especially those coding for solvent efflux pumps, were aligned with the genome of P. taiwanensis VLB120. Eleven open reading frames (orfs) were identified, of which the gene products showed high amino acid identities to the gene products described to be involved in solvent tolerance (Table 3). orfs with gene products with very high amino acid identities were found for the well-described solvent efflux pumps TtgABC and TtgGHI of P. putida DOT-T1E and SrpABC of P. putida S12, all shown to be involved in solvent extrusion (35, 36, 57). The presence of orfs with gene products with high amino acid identities to the products of respective regulatory genes of P. putida DOT-T1E (ttgR, ttgV, and ttgW), as well as of P. putida S12 (srpR and srpS), additionally suggests a similar regulation of gene expression (26, 58, 59).

TABLE 3.

Genes related to solvent tolerance in P. taiwanensis VLB120

orf	Location on genome	% aa identity to gene producta	Function of protein	Annotation	GenBank accession no.
1	1124417–1125571	95 to TtgA, 96 to ArpA	Periplasmic fusion protein of TtgABC efflux pump	ttgA	AGZ33797.1
2	1121261–1124413	96 to TtgB, 96 to ArpB	Inner membrane transporter of TtgABC efflux pump	ttgB	AGZ33796.1
3	1119810–1121264	96 to TtgC, 96 to ArpC	Outer membrane protein of TtgABC efflux pump	ttgC	AGZ33795.1
4	1125828–1126460	94 to TtgR	Repressor of the ttgABC operon	ttgR	AGZ33798.1
5	40326–41420	92 to TtgG, 92 to SrpA	Periplasmic fusion protein of TtgGHI efflux pump	ttgG	AGZ37958.1
6	41434–44583	98 to TtgH, 98 to SrpB	Inner membrane transporter of TtgGHI efflux pump	ttgH	AGZ37959.1
7	44573–45982	95 to TtgI, 96 to SrpC	Outer membrane protein of TtgGHI efflux pump	ttgI	AGZ37960.1
8	39265–40044	91 to TtgV, 93 to SrpS	Repressor of the ttgGHI operon	ttgV	AGZ37957.1
9	38609–39259	86 to TtgW, 80 to SrpR	Potential antirepressor of TtgV	ttgW	AGZ37956.1
10	3248148–3250457	90 to Cti	Cis/trans-isomerase	cti	AGZ35745.1
11	2494283–2494861	81 to TrgI	Toluene-repressed gene	trgI	AGZ35061.1

^aFor reasons of clarity, only the two best hits are shown.

Due to the high amino acid identities found for the products of orfs 1 to 9 of P. taiwanensis VLB120 and the RND efflux pumps and their regulatory systems of P. putida DOT-T1E, respective genes were annotated according to the nomenclature used for P. putida DOT-T1E. Besides the efflux pumps TtgABC and TtgGHI, which typically act as a long-term response to solvent exposure, additional genes were found, which have been proposed to be involved in first-line defense mechanisms against solvent toxicity. A cis/trans isomerase, responsible for the cis- to trans-isomerization of unsaturated fatty acids and the related increase in membrane rigidity (60), the toluene repressed gene trgI, proposed to be involved in altering the outer cell structure (61), and the AGZ33638.1 gene, encoding the multidrug efflux membrane protein PP1272 (62).

These findings on the genomic level confirm the previously described tolerance of P. taiwanensis VLB120 toward styrene (42, 63). The presence of the genes for the TtgGHI efflux pump, involved in toluene extrusion in P. putida DOT-T1E (36), indicates that P. taiwanensis VLB120 may also be tolerant toward solvents with lower log P_OW values than that of styrene (log P_OW of 3.0).

The ttgV deletion mutant of P. taiwanensis VLB120 shows constitutive toluene tolerance, relieving the need for adaptation.

Typically, adaptive solvent-tolerant bacteria, e.g., P. taiwanensis VLB120, which are exposed toward a second phase of a solvent with a low log P_OW, such as toluene, require adaptation to develop a solvent-tolerant phenotype and recover growth. Once adapted, cells start to grow in the presence of the respective solvent without any lag phase (see Fig. S1 in the supplemental material).

With the goal to avoid the necessity for these tedious adaptation procedures and associated lag phases for growth and productivity, the genetic efflux pump regulation was engineered as a strategy to achieve constitutive solvent tolerance. The key role of the TtgGHI and SrpABC efflux pumps of P. putida DOT-T1E (36) and P. putida S12 (35), respectively, in long-term tolerance toward solvents with low log P_OW values qualifies them as promising engineering targets. Rojas and coworkers (26) showed that TtgV functions as a repressor of ttgGHI gene expression in P. putida DOT-T1E. Thus, the ttgV gene was chosen as the target to establish constitutively high TtgGHI levels and was deleted in P. taiwanensis VLB120 and VLB120ΔC, resulting in the P. taiwanensis VLB120ΔttgV and VLB120ΔCΔttgV mutant strains. Reverse transcription (RT)-qPCR confirmed the absence of ttgV mRNA in P. taiwanensis VLB120ΔCΔttgV, while the amount of ttgG mRNA increased more than 200-fold (relative quantification [RQ] of 229, RQ_min of 203, RQ_max of 259) in comparison to that in P. taiwanensis VLB120ΔC. SDS-PAGE analyses of membrane fractions of P. taiwanensis VLB120ΔC and VLB120ΔCΔttgV confirmed increased TtgGHI levels in the latter strain. TtgG (42.5 kDa) and TtgH (114 kDa) became clearly visible in the membrane fraction of only P. taiwanensis VLB120ΔCΔttgV and not in that of P. taiwanensis VLB120ΔC, while TtgI (51 kDa) was not clearly visible, due to interfering host intrinsic proteins (see Fig. S2 in the supplemental material). After cultivation on M9 minimal medium with 5 g liter⁻¹ glucose as the carbon source, the two strains showed virtually the same trends regarding glucose, gluconate, and biomass concentrations (see Fig. S3 in the supplemental material). This indicates that the increased amount of TtgGHI efflux pumps does not have a major impact on the metabolism of P. taiwanensis VLB120ΔttgV. Both the biomass yields on glucose (Y_X/S) and the specific growth rates (µ) of the mutant (Y_X/S = 0.35 g_biomass g_glucose⁻¹, μ = 0.53 h⁻¹) and the wild-type strain (Y_X/S = 0.36 g_biomass g_glucose⁻¹, μ = 0.56 h⁻¹) were nearly identical. Thus, regarding growth physiology, the mutant strain does not feature drawbacks with respect to biotechnological applications.

Growth experiments with LB medium, to which a second phase of 1% (vol/vol) toluene was added in the early exponential growth phase, revealed a big effect of ttgV gene deletion (Fig. 1). While the wild-type strain immediately stopped growing upon toluene addition and remained in an adaptation lag phase for around 24 h, the P. taiwanensis VLB120ΔttgV regulatory mutant continued to grow without showing a lag phase. The deletion of the ttgV gene and the concomitantly increased TtgGHI levels in the membrane not only improved survival upon sudden short time shocks of a toxic organic solvent, as it was shown for P. putida DOT-T1E (26), but also enabled continued growth after the addition of a second phase of toxic solvents, although the growth was not exponential anymore (Fig. 1). Adaptation and growth of wild-type P. taiwanensis VLB120 in eight separate cultures were characterized by a high variability, which may relate to variable initial levels of solvent efflux pump gene expression. In contrast, a low variability was observed among eight separate cultures of P. taiwanensis VLB120ΔttgV, indicating constitutive ttgGHI expression to similar levels in all cultures being less influenced by adaptation phenomena.

FIG 1.

Growth of P. taiwanensis VLB120 and VLB120ΔttgV in LB medium containing 1% (vol/vol) toluene. The arrow indicates the addition of toluene. Error bars give the standard deviations from 8 independent cultures.

Styrene and styrene oxide tolerance of ttgV-knockout mutants.

A tolerance of the cells to the substrate and the product is crucial for an efficient whole-cell biotransformation. The biocatalytic applicability of the ttgV-knockout mutants was thus evaluated for the model reaction of toxic styrene to toxic (S)-styrene oxide. Similarly to the addition of toluene, the addition of 1 and 10% (vol/vol) styrene to LB cultures of P. taiwanensis VLB120ΔC immediately resulted in a cease of growth (Fig. 2A and B). In contrast, the ΔttgV-knockout strain, P. taiwanensis VLB120ΔCΔttgV, continued to grow at a decreased rate. When a second phase of styrene (1% [vol/vol]) was present in M9 minimal medium with 5 g liter⁻¹ glucose as the carbon source, P. taiwanensis VLB120ΔCΔttgV supported growth only if yeast extract (data not shown) or tryptone (Fig. 2D) was present. For all cultures, a diauxic behavior was observed, indicating a limited ability of the stressed cell metabolism to switch between the different nutrients provided by the complex medium. The addition of 10 mM styrene oxide leads to complete growth inhibition of Escherichia coli (8). In contrast, both P. taiwanensis VLB120ΔC and VLB120ΔCΔttgV were able to grow under these conditions, albeit at a reduced rate, which was slightly higher for P. taiwanensis VLB120ΔC. In conclusion, the ttgV deletion enabled a significantly increased tolerance toward styrene but not toward styrene oxide, for which a rather high tolerance seems to be inherent to P. taiwanensis VLB120. However, to take advantage of the increased amount of solvent efflux pumps in the cells, the presence of complex medium components, presumably amino acids or protectants, e.g., antioxidants or chelating agents (64), seems to be required.

FIG 2.

Growth of P. taiwanensis VLB120ΔC and VLB120ΔCΔttgV in LB medium in the presence of 1% (vol/vol) (87 mM) styrene (A), 10% (vol/vol) (872 mM) styrene (B), and 10 mM styrene oxide (C) and in M9 medium containing 0.5% glucose, 1% tryptone, and 10% (vol/vol) styrene (D). For the control, the strains were grown in the respective media without the addition of solvents (growth curves are shown in Fig. S4 in the supplemental material). The arrows indicate the time point of styrene and styrene oxide addition. Error bars give the standard deviations from two independent OD measurements. Experiments were repeated at least twice.

The ttgV deletion results in increased specific styrene epoxidation activities.

The high biocatalytic potential of P. taiwanensis VLB120ΔC for the specific epoxidation of styrene to (S)-styrene oxide has been shown both on a small scale with resting cells (42) and in two-liquid-phase biotransformations with growing cells (65). The fact that styrene serves on the one hand as the inducer for styAB expression and on the other hand as the substrate for the epoxidation reaction complicates the differentiation between induction- and biotransformation-related effects. During induction, the cells are exposed not only to toxic styrene but also to accumulating toxic styrene oxide, combined with an increasing amount of NADH channeled into the epoxidation reaction. The increased solvent tolerance of the regulatory mutant P. taiwanensis VLB120ΔCΔttgV might influence the induction process either in a positive way, by extruding toxic styrene and styrene oxide out of the membrane, or in a negative way, by reducing the intracellular availability of the inducer and substrate styrene and channeling a lot of energy into the solvent efflux pumps. To investigate such effects, resting cell assays were performed after induction of cultures with different amounts of styrene. Based on the published induction method (42), the addition of 0.5 to 2 mM styrene to M9 cultures was combined with continuous styrene supply via the gas phase using an evaporation container (see Fig. S5 in the supplemental material). With and without gas-phase exchanges, P. taiwanensis VLB120ΔCΔttgV showed at least 2-fold-higher specific epoxidation rates than P. taiwanensis VLB120ΔC, with maxima at 72 U/g of cells (dry weight) compared to 38 U/g of cells (dry weight). Upon saturation of the gas and the aqueous phases, both growth and specific epoxidation activity were observed only with P. taiwanensis VLB120ΔCΔttgV.

As an alternative induction procedure depending less on the gas-liquid mass transfer, different amounts of styrene were added, being equivalent to concentrations of 0.5 to 5 mM (assuming complete dissolution), into sealed M9 shaking flask cultures, which were incubated for 3 h without using an evaporation container (Fig. 3). With an experimentally determined distribution ratio of styrene between the aqueous and the gaseous phases of 42:58 (aqueous/gaseous) in the applied system, actual styrene concentrations in the aqueous phase ranged from 0.21 to 2.1 mM. Growth of P. taiwanensis VLB120ΔC was affected only minimally by the addition of 0.5 to 1.5 mM styrene, compared to growth in the absence of styrene (data not shown). The addition of ≥2 mM styrene resulted in an immediate stop of growth and a reduction of the biomass concentration, probably due to cell lysis. In contrast, P. taiwanensis VLB120ΔCΔttgV grew more slowly than P. taiwanensis VLB120ΔC at low styrene concentrations, but this growth slowed down only slightly at styrene concentrations of ≥2 mM. The specific epoxidation activity of P. taiwanensis VLB120ΔC increased by increasing the initial amount of styrene from 0.5 to 1 mM, indicating a limitation due to low and decreasing inducer levels. However, a further increase of the inducer concentration reduced the specific epoxidation activity, pointing toward the beginning of toxification by styrene, even though biomass formation was not significantly affected. Cells exposed to ≥2 mM styrene showed no activity at all, indicating, together with the decrease in optical density, toxicity-induced cell death. The specific epoxidation activity of P. taiwanensis VLB120ΔCΔttgV increased from 17.6 U/g of cells (dry weight) (0.5 mM styrene) to 58 U/g of cells (dry weight) (2 mM styrene), again suggesting an inducer limitation at low styrene concentrations. A further increase in the styrene concentration did not influence the specific activity for (S)-styrene oxide formation significantly. A considerable activity decrease was observed only in the presence of 5 mM styrene, when the theoretical styrene concentration in the aqueous phase (2.1 mM) was close to the styrene solubility in water (2.3 mM, at 20°C), at the beginning of the induction phase. Thus, the low activity might be a result of toxicity-induced stress affecting cell viability. As observed for P. taiwanensis VLB120ΔC, biocatalytic activity already dropped down at a specific styrene concentration, at which metabolic capacity still allowed biomass formation.

FIG 3.

Induction of P. taiwanensis VLB120ΔC and VLB120ΔCΔttgV with different amounts of styrene. (A) X-fold change in OD₄₅₀ of P. taiwanensis VLB120ΔC and VLB120ΔCΔttgV after induction for 3 h during shake flask cultivation in M9 medium. (B) Maximum specific styrene epoxidation activities obtained in resting cell assays as a function of the amount of styrene used for induction. Cells were incubated in sealed shaking flasks for 3 h with the respective amount of styrene. Styrene (1.5 mM) was supplied as the substrate in resting cell assays. Specific activities were calculated based on styrene oxide formation. Error bars give the standard deviations from two independent OD measurements and the standard deviations from two individual biotransformations.

In conclusion, the regulatory mutant P. taiwanensis VLB120ΔCΔttgV showed significantly higher styrene epoxidation activities than P. taiwanensis VLB120ΔC independently of the induction strategy.

The induction method giving the best results with both strains, the addition of 1 mM styrene combined with styrene supply via the gas phase, was used to analyze the course of specific styrene epoxidation by resting cells of P. taiwanensis VLB120ΔC and VLB120ΔCΔttgV (Fig. 4). The latter strain converted 1.5 mM styrene to 1.38 mM styrene oxide within 60 min, showing an initial specific activity of 67 U/g of cells (dry weight). The strong decrease in the specific activity after 30 min was due to substrate depletion. P. taiwanensis VLB120ΔC showed an initial specific epoxidation activity of ∼30 U/g of cells (dry weight), which only decreased slightly over time. After 120 min, a styrene oxide concentration of 1.35 mM was reached. These results confirm the advantages of the constitutively solvent-tolerant ΔttgV mutant over the nonadapted solvent-tolerant parent strain for the biocatalytic conversion of a toxic organic compound, such as styrene.

FIG 4.

Course of specific styrene epoxidation with resting cells (0.5 g of cells [dry weight]/liter) of P. taiwanensis VLB120ΔC (A) and VLB120ΔCΔttgV (B) over 120 min. Cells were induced during shake flask cultivation by the addition of 1 mM styrene and continuous styrene supply via the gas phase. Styrene (1.5 mM) was supplied as the substrate. Specific activities are calculated based on styrene oxide formation over time. Error bars give the standard deviations from two individual biotransformations.

DISCUSSION

Several Pseudomonas strains show tolerance toward toxic solvents with low log P_OW values, between 2 and 4. Thereby, proton-driven efflux of organic solvents by solvent efflux pumps belonging to the RND family represents by far the most effective mechanism counteracting solvent toxicity (33, 34). Here, a homologue of the major efflux pumps TtgGHI of P. putida DOT-T1E (36) and SrpABC of P. putida S12 (35) has been identified in P. taiwanensis VLB120. The results presented on TtgV in P. taiwanensis VLB120 confirmed the repressing function observed for TtgV in P. putida DOT-T1E and its homologue SrpS in P. putida S12 (26, 27). For P. putida S12, the ISS12 insertion element was proposed to be involved in RND efflux pump regulation (24, 27). Interestingly, an ISS12 homologue as well as a potential binding site within the ttgV sequence (AGZ37957.1) were found in P. taiwanensis VLB120 (data not shown), a further indication for the involvement of mobile elements in efflux pump regulation, as it was proposed for P. putida S12. Besides the ttgGHI genes, further genes reported to be involved in solvent tolerance have been identified in P. taiwanensis VLB120, confirming the previously observed solvent tolerance of P. taiwanensis VLB120 (42, 63) on the genetic level. Growth experiments with P. taiwanensis VLB120 showed that this strain is able to grow after adaptation to the presence of a second phase of toluene (1% [vol/vol]), verifying its solvent tolerance. Similarly, P. putida strains IH-2000, S12, DOT-T1E, and T-57 and Pseudomonas sp. strain BCNU 171 have been shown to be capable to thrive in the presence of a second phase of toluene (28, 29, 32, 66, 67).

Knocking out the ttgV gene in P. taiwanensis VLB120 enabled not only the survival but also incessant growth upon exposure to high concentrations of toxic toluene and styrene and thus the avoidance of an adaptation phase, which typically is required to achieve solvent tolerance. Such unproductive adaptation phases have previously been identified to be prohibitive for the industrial implementation of biotransformations involving toxic solvents and solvent-tolerant P. putida strains (20). Furthermore, the constitutive solvent tolerance of the P. taiwanensis VLB120ΔttgV and VLB120ΔCΔttgV regulatory mutants now allows storage and reuse of a strain showing a solvent-tolerant phenotype without the risk of variability related to solvent adaptation, as it was observed for wild-type P. taiwanensis VLB120 growing in the presence of a second phase of toluene (see Fig. S1 in the supplemental material). Regarding in situ product removal, which still is a challenge in whole-cell biocatalysis (14), Pseudomonas strains with a constitutive tolerance toward a broad range of solvents with low log P_OW values bring advantages in terms of solvent choice. Not only nontoxic carrier solvents, such as bis(2-ethylhexyl)phthalate (BEHP) in the case of styrene epoxidation (15, 42), but also toxic solvents with log P_OW values of <4 may now be applied as extractive phases. Examples include toluene as an appropriate solvent for the extraction of 4-chlorophenol (substrate) and 4-chlorocatechol (product) (14) or styrene itself in the case of stereospecific styrene epoxidation.

Furthermore, both P. taiwanensis VLB120ΔC and VLB120ΔCΔttgV tolerated high aqueous styrene oxide concentrations and still grew in the presence of 10 mM styrene oxide, whereas E. coli JM101, which has been shown to be a potent host for styrene epoxidation, already lost the ability to grow at a styrene oxide concentration of 6 mM (8).

P. taiwanensis VLB120ΔCΔttgV showed very promising characteristics for stereospecific styrene epoxidation. The constitutive expression of the ttgGHI genes, as reflected by a 200-fold increase of the ttgGHI mRNA level and the higher TtgGHI levels in the membrane (see Fig. S2 in the supplemental material), can be expected to affect the energy metabolism of P. taiwanensis VLB120ΔCΔttgV. To establish the proton motive force necessary to drive the active export of organic solvents (33), respiration competes for NADH with styrene epoxidation and other NADH-dependent reactions (39). Thus, NADH-dependent bioconversions may be limited by NADH availability. Although Pseudomonas strains are able to boost their energy metabolism and meet high energy demands, Blank et al. observed that the specific styrene epoxidation rate of recombinant P. putida DOT-T1E (pTEZ2440) was reduced by 60% in the presence of octanol (39). Additionally, inducer and substrate limitations may occur as a result of active styrene extrusion. In contrast to such effects reducing biocatalyst efficiency, P. taiwanensis VLB120ΔCΔttgV showed considerably higher specific epoxidation activities than P. taiwanensis VLB120ΔC in all the different induction and biotransformation setups tested. Only the slightly reduced biomass formation under nontoxic and subtoxic induction conditions pointed toward an increased energy demand due to the solvent efflux pumps present in the constitutively solvent-tolerant strain (Fig. 3; see also Fig. S5 in the supplemental material), which was not observed when growing in the absence of styrene (see Fig. S3). In the case of toxic induction conditions, the VLB120ΔCΔttgV strain outperformed the VLB120ΔC strain with respect to biomass formation and specific styrene epoxidation activity. For both strains, low styrene concentrations, not toxic to P. taiwanensis VLB120ΔC, resulted in low specific activities due to inducer depletion, before styAB expression was fully induced. With both strains, maximal specific epoxidation activities were reached with the setup described by Park et al. (42), but the published activity of 60 to 70 U/g of cells (dry weight) was not reached with P. taiwanensis VLB120ΔC. In the previous study, no reference sample was taken before the biotransformation was started by styrene addition. In the present study, these reference samples were found to contain styrene oxide originating from the induction process, which was subtracted from the styrene oxide amount in samples taken during the biotransformation. Omitting this subtraction leads to an overestimation of the activity up to 50%, which explains the higher activity reported by Park et al. (42). The VLB120ΔCΔttgV strain showed considerably higher specific epoxidation activities than P. taiwanensis VLB120ΔC in all experimental setups tested. With both strains, the specific epoxidation activity decreased at styrene concentrations slightly below those affecting biomass formation (Fig. 3). Inhibition of the styrene monooxygenase or of the metabolism and thus of NADH regeneration by styrene or styrene oxide are reasonable explanations for this activity decrease. Otto et al. (68) observed product inhibition of isolated StyAB at styrene oxide concentrations above 0.5 mM. On the other hand, it was shown that styrene oxide concentrations below the toxicity limit already affected the metabolic and biocatalytic performance of recombinant E. coli JM101 (pSPZ10) (9). However, as only half the amount of styrene oxide was formed in the induction setups leading to low specific activities compared to that in the induction setups resulting in maximal specific epoxidation activities (see Fig. S6 in the supplemental material), toxic effects of or inhibition by styrene oxide can be excluded as the cause for the reduced activities. Thus, styrene itself seems to cause such inhibition or toxic effects. This was the case for both strains, with the difference that the activity decrease with P. taiwanensis VLB120ΔCΔttgV occurred at much higher styrene concentrations than that with P. taiwanensis VLB120ΔC. Together with the observation that the knockout of ttgV improved the growth behavior in the presence of styrene and not in the presence of styrene oxide, this indicates that the enhanced metabolic and biocatalytic performance of P. taiwanensis VLB120ΔCΔttgV is caused mainly by improved extrusion of styrene rather than of styrene oxide.

This study for the first time reports a strain, P. taiwanensis VLB120ΔCΔttgV, showing constitutive solvent tolerance in terms of growth upon addition of a toxic solvent phase without prior adaptation. This strain also shows higher specific styrene epoxidation activities, making the ΔttgV strain highly interesting for energy-dependent whole-cell biocatalysis involving toxic solvents. In general, the deregulation of solvent efflux pump gene expression is a key target in engineering of controllable whole-cell biocatalysts for applications involving toxic solvents.