Discovery of an Inducible Toluene Monooxygenase That Cooxidizes 1,4-Dioxane and 1,1-Dichloroethylene in Propanotrophic Azoarcus sp. Strain DD4

Toluene MOs have been well recognized given their robust abilities to degrade a variety of environmental pollutants. Built upon previous research efforts, this study ascertained the untapped capability of a toluene MO in DD4 for effective cooxidation of dioxane and 1,1-DCE, two of the most prevailing yet challenging groundwater contaminants. This report also aligns the induction of a toluene MO with nontoxic and commercially accessible chemicals (e.g., propane and 1-propanol), extending its implications in the field of environmental microbiology and beyond.

ABSTRACT

Cometabolic degradation plays a prominent role in bioremediation of commingled groundwater contamination (e.g., chlorinated solvents and the solvent stabilizer 1,4-dioxane [dioxane]). In this study, we untangled the diversity and catalytic functions of multicomponent monooxygenases in Azoarcus sp. strain DD4, a Gram-negative propanotroph that is effective in degrading dioxane and 1,1-dichloroethylene (1,1-DCE). Using a combination of knockout mutagenesis and heterologous expression, a toluene monooxygenase (MO) encoded by the tmoABCDEF gene cluster was unequivocally proved to be the key enzyme responsible for the cometabolism of both dioxane and 1,1-DCE. Interestingly, in addition to utilizing toluene as a primary substrate, this toluene MO can also oxidize propane into 1-propanol. Expression of this toluene MO in DD4 appears inducible by both substrates (toluene and propane) and their primary hydroxylation products (m-cresol, p-cresol, and 1-propanol). These findings coherently explain why DD4 can grow on propane and express toluene MO for active cooxidation of dioxane and 1,1-DCE. Furthermore, upregulation of tmo transcription by 1-propanol underlines the implication potential of using 1-propanol as an alternative auxiliary substrate for DD4 bioaugmentation. The discovery of this toluene MO in DD4 and its degradation and induction versatility can lead to broad applications, spanning from environmental remediation and water treatment to biocatalysis in green chemistry.

IMPORTANCE Toluene MOs have been well recognized given their robust abilities to degrade a variety of environmental pollutants. Built upon previous research efforts, this study ascertained the untapped capability of a toluene MO in DD4 for effective cooxidation of dioxane and 1,1-DCE, two of the most prevailing yet challenging groundwater contaminants. This report also aligns the induction of a toluene MO with nontoxic and commercially accessible chemicals (e.g., propane and 1-propanol), extending its implications in the field of environmental microbiology and beyond.

INTRODUCTION

As a possible human carcinogen, the common chlorinated solvent stabilizer 1,4-dioxane (dioxane) has emerged with increasing concern nationwide and globally (1, 2). Its extreme hydrophilicity and stable cyclic structure preclude effective treatment by most physical and chemical approaches, such as adsorption, air sparging, and chemical oxidation (3). Recent microbiological efforts have promoted bioremediation as a promising alternative to mitigate dioxane contamination. Most of the characterized dioxane-degrading bacteria are Actinomycetes (e.g., Pseudonocardia dioxanivorans CB1190 [4] and Mycobacterium dioxanotrophicus PH-06 [5]), with relatively low growth rates, low affinity to dioxane, and clumping behaviors, which limit their field applications (6–8). These unfavorable characteristics are exacerbated by the coexistence of 1,1-dichloroethylene (1,1-DCE), a major attenuation product of chlorinated solvents (e.g., 1,1,1-trichloroethane via an abiotic process [9] and trichlorethylene via anaerobic biotransformation [10]), which is frequently detected at dioxane-impacted sites and displays potent inhibition to dioxane biodegradation (11–14). Thus, 1,1-DCE has long been recognized as a stumbling block hindering effective biological removal of dioxane and other coexisting pollutants.

Azoarcus sp. strain DD4 is a Gram-negative bacterium that can cometabolize both dioxane and 1,1-DCE when it is fed with propane as the primary substrate (15). DD4 demonstrates many superior physiological properties well suited for field applications (e.g., fast and planktonic growth) (15). Dioxane and 1,1-DCE biotransformation pathways were predicted based on the degradation metabolites detected using modern mass spectrometry techniques (15). Detection of 2-hydroxyethoxyacetic acid (HEAA) suggests dioxane is oxidized via 2-hydroxylation, a catalytic process that inserts a hydroxyl group into the α-carbon adjacent to the oxygen, enabling the subsequent ring cleavage. As for 1,1-DCE, epoxidation is postulated as the initial step to activate the decomposition (15). Biotransformation of dioxane, 1,1-DCE, and propane are all terminated once DD4 is exposed to acetylene, a suicide inhibitor to bacterial monooxygenases (MOs) (16). Thus, converging lines of evidence demonstrate the pivotal role of MO(s) in initiating the breakdown of dioxane and 1,1-DCE in DD4 (15).

Soluble di-iron monooxygenases (SDIMOs) represent a nonheme bacterial enzyme family (17, 18) known for their substrate versatility and robust capability of degrading a plethora of anthropogenic pollutants. SDIMOs can be divided into six groups based on their sequence similarity and substrate preference (19). In dioxane metabolizers (e.g., CB1190 and PH-06) that can grow with dioxane as the sole carbon and energy source, group-5 tetrahydrofuran (THF) MOs (20) and group-6 propane MOs (21) have been shown to initiate the metabolism of dioxane via 2-hydroxylation. Some other bacteria, like DD4, can degrade dioxane via cometabolism, a fortuitous oxidation process necessitating supplementation with auxiliary substrates (e.g., propane and toluene). A few propane- and toluene-inducible SDIMOs have been reported to be responsible for dioxane cometabolism. Some propanotrophs (e.g., Mycobacterium sp. strain ENV421 [22], Mycobacterium vaccae JOB5 [23], and Rhodococcus ruber ENV425 [24]) expressing group-5 and/or group-6 propane MOs have demonstrated their ability to cooxidize dioxane when fed with propane. Additionally, toluene-2-MO in Burkholderia cepacia G4, toluene-p-MO in Ralstonia pickettii PKO1, and toluene-4-MO in Pseudomonas mendocina KR1 exhibited dioxane degradation abilities in both toluene-induced wild-type (wt) cultures and heterologous clones (6). Biotransformation of 1,1-DCE has also been observed in SDIMO-expressing bacteria (10, 25), such as the methanotrophic strain Methylosinus trichosporium OB3b (26) and the toluene/o-xylene degrader Pseudomonas stutzeri OX1 (27). Hence, it is plausible that one or more SDIMOs are responsible for the cometabolic degradation of dioxane and 1,1-DCE in the propanotrophic isolate DD4.

The complete genome of DD4 has been recently sequenced using the long-read high-accuracy PacBio SMRT sequencing platform (28). In this present study, with the combined assistance of genome annotation, gene knockout, and heterologous expression, the diversity and catalytic functions of SDIMOs in DD4 are examined with a focus on the cometabolic degradation of dioxane and 1,1-DCE. Their roles in the oxidation of propane and other auxiliary substrates are elucidated through the characterization of degradation profiles in knockout mutants, as well as quantitative analysis of key metabolite production. Importantly, regulation of SDIMO expression in DD4 is assessed in association with these primary substrates and their metabolites to align propane assimilation with cometabolic degradation of dioxane and 1,1-DCE (15). This research advances our fundamental knowledge about the catalysis and induction versatility of SDIMOs, promoting the innovation of effective remedial operations that can be tailored to mitigate the cocontamination of dioxane and 1,1-DCE.

RESULTS AND DISCUSSION

DD4 harbors a diversity of SDIMO gene clusters.

Five distinct SDIMO gene clusters (Fig. 1; see also Table S2 in the supplemental material) were identified scattered across the single chromosome of Azoarcus sp. strain DD4 (28). According to the sequence identity and arrangement order of their gene components, these SDIMO gene clusters (Fig. 2) were categorized and given designations, including prmABCD, encoding a group-5 propane MO, bmoXYBZDC, encoding a group-3 butane MO, tmoABCDEF, encoding a group-2 toluene MO, and dmp1KLMNOP and dmp2KLMNOP, encoding two group-1 phenol hydroxylases. On the basis of the 16S rRNA gene sequence analysis, DD4 is most phylogenetically related to Azoarcus sp. strain BH72 and Azoarcus olearius DQS-4 (15). Similarly, the prm and two dmp gene clusters are highly identical to those in BH72 (29) and DQS-4 (30) (mostly >92%) (Table S1), suggesting all three betaproteobacteria can exploit propane and phenolic compounds as their substrates.

FIG 1.

Organization of five SDIMO gene clusters and one copper-based particulate MO gene cluster in Azoarcus sp. strain DD4. ORFs are depicted as gray arrows. Gene names for the SDIMO gene clusters and their flanking genes are described below. SDIMO components are labeled with abbreviated symbols, including alpha (α), beta (β), and gamma (γ) hydroxylase, coupling protein (C), ferredoxin component (F), reductase (R), and protein with unknown function (X). For knockout mutagenesis, deletion fragments are framed in the rectangles with dotted lines.

FIG 2.

Neighbor-joining tree showing the phylogenetic relationship of five SDIMOs (highlighted in red) in Azoarcus sp. strain DD4 with representatives from all six subgroups of SDIMOs. This phylogenetic tree is constructed based on the alignment of amino acid sequences of their alpha subunits. Key SDIMO components are abbreviated as alpha (α), beta (β), and gamma (γ) hydroxylase, coupling protein (C), reductase (R), ferredoxin component (F), and protein with unknown function (X).

Compared to BH72 and DQS-4, DD4 harbors two unique SDIMO genes, bmo and tmo. Closest homologues of bmo- and tmo-encoding proteins have been found in Thauera butanivorans (31–33) and Thauera sp. strain 27 (34, 35), respectively. bmoXYBZDC, encoding butane MO, has been well characterized in Thauera butanivorans regarding its catalytic preference and regiospecificity, which predominantly oxidizes terminal carbons of C-4 and C-5 alkanes (31, 36). In DD4, all six key bmo gene components have been identified, although bmoZ, bmoD, and bmoC were separated by two genes of unknown functions (hp1 and hp2 in Fig. 1 and Table S2). Homologues of these two hypothetical proteins are found in the gammaproteobacteria (e.g., Solimonas and Methylohalobius) rather than betaproteobacteria such as Azoarcus or Thauera. The existence of internal spacer genes in a bmo or SDIMO gene cluster has been sparsely reported. However, the encoded butane MO in DD4 retains its catalytic activity toward butane with the presence of these two inserted genes, as revealed by the following mutagenesis assays.

The tmo gene cluster in DD4 consists of all six components essential for group-2 toluene MOs (Fig. 1 and Table S2). Immediately downstream of the tmoABCDEF gene cluster, three genes were identified on the same strand of the sequence, including tmoX, tmoS, and tmoT. tmoX encodes a putative membrane channel protein promoting the transport of toluene from outside the cell (37). TmoST likely represents a dual regulation system (38). The presence of toluene or other inducers may trigger the phosphorylation of TmoS. The phosphorylated TmoS then reacts with TmoT and transfers the phosphate to TmoT. When phosphorylated, TmoT binds to the promoter of the tmo gene cluster and facilitates its transcription. These three genes have been identified across the genomes of archetypical toluene degraders, such as toluene dioxygenase (TOD)-expressing Pseudomonas putida F1 and toluene-4-MO-expressing P. mendocina KR1 (38). To our knowledge, a consecutive tmoABCDEFXST gene cluster encoding comprehensive functions covering toluene biocatalysis, transport, and signal transduction has not been previously reported. The integrity of this gene cluster may enable efficient dissemination of toluene catabolism abilities via horizontal gene transfer in toluene-impacted environments.

Deletion of the tmo gene inactivates biotransformation of both dioxane and 1,1-DCE in DD4.

To discern the contribution of SDIMOs to dioxane and 1,1-DCE degradation in DD4, knockout mutants were created using homologous recombination with the sucrose counterselectable suicide vector, pK18mobsacB. We primarily deleted prm, bmo, and tmo genes, given their potential roles in cometabolism of dioxane and chlorinated aliphatic hydrocarbons (CAHs). Successful deletion of target SDIMO genes was verified by PCR and gel electrophoresis in comparison with amplicons using genomic DNA from wild-type DD4 (wt DD4) as the template (Fig. 1 and Fig. S1). Ethanol was used as a carbon source for knockout mutants, since it can not only support the growth of DD4 regardless of the presence or absence of the various SDIMO genes (Fig. S2) but also stimulate active dioxane/1,1-DCE degradation in wt DD4 (Table S3).

Notably, deletion of the tmo gene was conducive to the complete disability of dioxane transformation in the Δtmo DD4 mutant (Fig. 3). When grown with ethanol, wt DD4 was capable of proliferation (Fig. S2) and concurrently degraded the initial 11.0 mg/liter of dioxane to below our method detection limit (MDL) (i.e., 0.1 mg/liter) until day 4 (Fig. 3). Although experiencing a lag phase that lasted for 2 days (Fig. S2), Δprm and Δbmo DD4 mutants were able to resume their growth and transform approximately 79.4% and 77.4% of the initially dosed dioxane, respectively. However, no significant removal of dioxane was observed in the Δtmo DD4 mutant, although continuous cell growth was evidenced by the increase in turbidity of the culture media over 4 days of incubation (Fig. S2). This indicated that an intact tmo gene cluster is essential for the dioxane degradation activity in DD4. This finding was outside our expectations, in part because (i) propane was initially identified as the primary substrate to sustain the dioxane cometabolism of DD4 (15) and (ii) no previous report has revealed group-2 toluene MOs are inducible by propane. To validate our knockout assays, the dioxane degradation capability of wt DD4 was tested when it was fed with toluene as the sole carbon and energy source. As shown in Fig. 4, DD4 can grow on toluene and concurrently degrade dioxane. With two consecutive toluene amendments (∼58 mg/liter per time), DD4 steadily degraded 16.4 mg/liter dioxane within 9 days of incubation. This demonstrated that toluene can sustain the growth of DD4 and stimulate the cometabolic oxidation of dioxane, supporting the contribution of tmo in dioxane cometabolism.

FIG 3.

Dioxane degradation by wild-type and SDIMO-deleted growing DD4 cells that were fed with ethanol (200 mg/liter) at time zero. Negative controls (NC) were prepared without cells. Data are averages from triplicates, and error bars indicate their standard deviations.

FIG 4.

Dioxane degradation by DD4 using toluene as the sole carbon and energy source. To avoid inhibition from the high concentration of toluene, toluene was initially amended at the aqueous phase concentration of 57.7 mg/liter. The same amount of toluene was reamended when toluene was fully consumed on day 4. Error bars indicate standard deviations among triplicates.

Further assays using resting cells pregrown with ethanol indicated this tmo gene is also tied to 1,1-DCE biotransformation in DD4 (Fig. 5). Compared with the rapid 1,1-DCE depletion by wt DD4, no 1,1-DCE removal by the Δtmo DD4 mutant was observed. In contrast, deletion of the prm gene did not significantly affect the biotransformation of 1,1-DCE. The Δbmo knockout mutant exhibited 27.6% less 1,1-DCE removal than wt DD4 over the incubation course of 20 h. This decrease in 1,1-DCE removal is probably attributed to the encoded butane MO taking part in a certain metabolic process(es). Deletion of this bmo gene may negatively affect overall degradation performance of resting cells by diminishing cellular energy efficiency or impeding detoxification response to stressors (e.g., DCE epoxide formed from 1,1-DCE oxidation [15]). This hypothesis is also supported by the observation of the delayed growth incurred to this Δbmo mutant when fed with ethanol (Fig. S2). Furthermore, the Δtmo DD4 mutant carries the complete bmo gene cluster and presumably expresses active butane MO. However, it has completely lost the ability to oxidize 1,1-DCE. Thus, the integrity of the bmo gene cluster appears independent from the catalytic capability of degrading 1,1-DCE in the Δtmo DD4 mutant. Collectively, all these lines of evidence corroborate that the toluene MO encoded by tmo is responsible for cooxidation of dioxane and 1,1-DCE in DD4, precluding the involvement of other SDIMOs. Slowed dioxane or 1,1-DCE removal in prm and bmo knockout mutants was probably due to inadvertent molecular detriments to cellular metabolism.

FIG 5.

Degradation of 1,1-DCE by resting cells of wild-type and SDIMO gene-deleted DD4 that were precultured with ethanol. Negative controls (NC) were prepared without cells. Data are averages from triplicates, and error bars indicate their standard deviations.

Toluene MO in DD4 can initiate the oxidation of toluene, dioxane, 1,1-DCE, and propane.

To further characterize its catalytic function in DD4, this toluene MO was heterologously expressed in E. coli BL21(DE3). After induction with isopropyl-β-d-thiogalactopyranoside (IPTG), BL21(DE3) transformed with pET-tmo, the high-expression vector that carries the whole tmoABCDEF gene cluster, can produce a catalytically active toluene MO complex. This toluene MO-expressing transformant can catalyze the oxidation of its primary substrate, toluene (Fig. 6A). This was also supported by the formation of blue colonies cultured in LB media owing to the ability of toluene MOs to convert indole to indigo (39, 40) (Fig. S3). Furthermore, transformation of dioxane, 1,1-DCE, and propane was positively detected in the toluene MO-expressing transformant (Fig. 6). In contrast, the control BL21(DE3) transformant carrying the empty plasmid pET-28a(+) could neither turn blue (as an indication of no indigo generation) (Fig. S3) nor degrade any of the tested substrate compounds (Fig. 6). This heterologous expression assay unequivocally demonstrates the catalysis of toluene, dioxane, 1,1-DCE, and even propane by this group-2 toluene MO in DD4.

FIG 6.

Degradation of toluene (A), dioxane (B), 1,1-DCE (C), and propane (D) by E. coli BL21(DE3) heterologously expressing DD4 toluene MO. Control experiments were prepared with BL21(DE3) carrying the empty vector pET-28a(+). All transformants were induced with 0.5 mM IPTG prior to substrate exposure. For toluene and 1,1-DCE, aqueous-phase concentrations are calculated at equilibrium based on their Henry’s Law constants. Concentrations of dioxane and propane are considered in the aqueous and gas phases, respectively. Error bars indicate standard deviations among triplicates.

The expressed toluene MO was most efficient in transforming toluene at a rate of 14.7 μg toluene/mg protein/h. The initial biotransformation rate for dioxane in the first 2 h was estimated as 1.3 μg dioxane/mg protein/h (Fig. 6B). We noticed that the dioxane oxidation rate of this transformant was at least 1 order of magnitude lower than that exhibited by resting wt DD4 cells (23 μg dioxane/mg protein/h) (15), possibly due to the production of insoluble protein aggregates that lack of biocatalyst activity, as observed in Fig. S4.

For 1,1-DCE, the instant biotransformation rate was estimated as 12.0 μg 1,1-DCE/mg protein/h in the toluene MO-expressing transformant (Fig. 6C). 1,1-DCE cometabolism has been reported in bacteria expressing SDIMOs. Canada et al. found that the toluene-2-MO, classified as a group-1 SDIMO, and its shuffle mutants from B. cepacia G4 toluene MO can degrade 1,1-DCE (41). In subgroup 3, methane MO can oxidize 1,1-DCE into 1,1-DCE epoxide in M. trichosporium OB3b (26). Further, belonging to the same SDIMO subgroup 2 as the toluene MO of DD4, the toluene/o-xylene MO of P. stutzeri OX1 can cooxidize 1,1-DCE and many other recalcitrant CAHs (e.g., PCE, TCE, and chloroform) while releasing free chloride as an indication of mineralization (27, 42). However, no significant dioxane removal was observed in an E. coli transformant expressing this specific toluene/o-xylene MO (6).

It is surprising that the toluene MO in DD4 can also degrade propane with a transformation rate of 2.1 μg propane/mg protein/h (Fig. 6D). No prior studies have reported propane or other short-chain alkanes as a substrate for group-2 toluene MOs. The ability to oxidize propane by the toluene MO in DD4 partially explains why propane is effective in sustaining the growth of DD4 and the transformation of dioxane and 1,1-DCE at the same time.

Toluene MO in DD4 catalyzes two successive toluene oxidation steps.

The catalytic function of this new toluene MO in DD4 was further characterized using the constructed expression clone. After the toluene MO-expressing transformant was exposed to toluene for 20 h, three metabolites (p-cresol, m-cresol, and 4-methylcatechol [4-MC]) were detected by high-performance liquid chromatography (HPLC) (Fig. S5A). No 3-methylcatechol (3-MC) or other metabolites were noticeably detected. Negative controls with the BL21(DE3) transformant carrying the empty vector exhibited no significant abiotic loss of toluene over the course of the experiment (Fig. S5B). Detection of both p-cresol and m-cresol revealed the ability of toluene MO in DD4 to attack both the para and meta carbons of toluene via hydroxylation. 4-MC was only detected at 20 h, indicating the toluene MO can insert a second hydroxyl group into p-cresol and/or m-cresol. To further verify this secondary hydroxylation by the toluene MO, toluene MO-expressing transformant cells were exposed to p-cresol and m-cresol, respectively. Interestingly, 4-MC was observed as the only metabolite of either p-cresol or m-cresol via the oxidation of toluene MO (Fig. S5C and E). Although there was no significant abiotic loss of p- and m-cresol in the BL21(DE3) transformant carrying the empty vector (Fig. S5D and S5F), disappearance of 4-MC was observed in the transformant clones with or without the tmo gene cluster (Fig. S6). This decay of 4-MC was probably due to the biodegradation by the BL21(DE3) host cells (43) and/or abiotic transformation, explaining the major molar discrepancy between the production of 4-MC and the reduction of p- or m-cresol, as observed in Fig. S5C and E.

Accompanying the disappearance of initial 0.86 mM toluene, the production of p-cresol (0.72 mM), m-cresol (0.06 mM), and 4-methylcatechol (0.09 mM) was observed at 20 h (Fig. S5A). As shown in Fig. S5C and E, degradation rates of p-cresol and m-cresol to 4-MC were indistinguishable (P > 0.05). Assuming conversion rates of m-cresol and p-cresol to 4-MC were identical, stoichiometric generation of p- and m-cresol at a ratio of 12:1 was estimated from the oxidation of toluene. Therefore, toluene MO in DD4 is a multihydroxylating enzyme that oxidizes toluene to p-cresol (mainly) and m-cresol followed by the insertion of a second hydroxyl group to form 4-MC. The ring structure of 4-MC can be further broken down by catechol 2,3-dioxygenase (C23O) via metacleavage, entering the TCA cycle as depicted in Fig. S7. The DD4 genome contains three copies of C23O genes along with other genes that participate in downstream metabolism (Table S4).

Phylogenetic analysis of α subunits revealed the toluene MO in DD4 is more closely related to the archetypic toluene-p-MO in PKO1 (71%) and toluene/o-xylene MO in OX1 (72%) than toluene-4-MO in KR1 (69%) and toluene-2-MO in G4 (26%) (Fig. 2 and Table S5). This was also echoed by the sequence identity analysis for other MO subunits (Table S5). In good agreement with their evolutionary relationship, catalytic versatility exhibited by the DD4 toluene MO is somewhat more similar to that of toluene-p-MO in PKO1 and toluene/o-xylene MO in OX1. The recombinant toluene-p-MO from PKO1 was shown to oxidize toluene at both para and meta positions. The formation ratio of p- and m-cresol was determined as around 9:1 stoichiometry with the assistance of gas chromatography and nuclear magnetic resonance analyses (44). However, this toluene-p-MO was not able to further oxidize p- or m-cresol in PKO1 (45). In OX1, the toluene/o-xylene MO can not only attack toluene at all three positions (para and ortho primarily) but also proceed with the secondary hydroxylation to generate 4-MC (from p- and m-cresols) and 3-MC (from o-cresol) (46, 47). In contrast, toluene-4-MO in KR1 (48) and toluene-2-MO in G4 (49) can only attack toluene at the para and ortho position, respectively, exhibiting relatively high regiospecificity.

With the recent discovery of more group 2 toluene MOs, their regiospecificity and substrate range are of great variance. For instance, benzene MO in Pseudomonas aeruginosa JI104 and toluene MO in Burkholderia cepacia AA1 (Fig. 2) are closely related to the toluene MO in DD4 with high sequence identities (79% and 76%, respectively). The benzene MO in JI104 exhibited relaxed regiospecificity toward toluene, producing three metabolites, o-, m-, and p-cresol, at a ratio of 1:2:1.7 (50). However, the toluene MO in AA1 is highly specific, which may only catalyze at the meta position (51). Thus, categorization of the catalytic functions of group-2 SDIMOs on the basis of their refined phylogenies would be ambiguous (Fig. 2). Further molecular characterization and computational efforts are under way to unravel the catalytic diversity of group-2 SDIMOs in association with their enzyme configurations and key residues that govern substrate interactions.

Expression of toluene MO is inducible by both substrates and their hydroxylated products.

DD4 resting cells harvested from toluene, m-cresol, p-cresol, propane, or 1-propanol can degrade both dioxane and 1,1-DCE (Table S3), whereas no significant dioxane or 1,1-DCE degradation was observed in DD4 cells cultured with 4-MC, 3-MC, propionaldehyde, propionic acid, succinate, or pyruvate. To gain further insight into tmo regulation in DD4, differential expression of tmoA, encoding the large hydroxylase subunit of the toluene MO, was examined using reverse transcription-quantitative PCR (RT-qPCR). Notably, tmoA expression was upregulated not only by growth substrates (i.e., toluene and propane) but also by their primary hydroxylated metabolites (i.e., p-cresol, m-cresol, and 1-propanol) (Fig. 7). When DD4 was fed with toluene, the highest level of tmoA expression ([4.81 ± 0.90]-fold) was observed compared with that of the control cells cultivated on pyruvate. Significant upregulation of tmoA was also detected in DD4 cells pregrown on propane ([2.93 ± 0.47]-fold), p-cresol ([2.79 ± 1.03]-fold), m-cresol ([2.60 ± 0.19]-fold), or 1-propanol ([2.16 ± 0.30]-fold). When grown with downstream metabolites of 1-propanol or cresol oxidation (e.g., propionaldehyde, propionic acid, or 4-MC), no positive expression of tmoA was observed compared to that of pyruvate-fed cells. The induction of toluene MO expression in DD4 revealed by our RT-qPCR results is in good agreement with the observation of active cometabolism of dioxane and 1,1-DCE after DD4 was fed with toluene, propane, and their primary hydroxylated metabolites but not the secondary metabolite, 4-MC, or other lower pathway products (e.g., propionaldehyde) (Table S3).

FIG 7.

Degradation pathways of toluene and propane initiated by the toluene MO of DD4. Expression fold changes of the tmoA gene are indicated below each compound that DD4 was fed with as the sole carbon and energy source. Data were normalized to a treatment in which DD4 was fed with pyruvate. Green boxes highlight significant upregulation of the tmoA gene compared to the pyruvate control. The 16S rRNA gene of DD4 was used as the housekeeping gene for error control.

Note that the RT-qPCR analysis used in our study is inadequate to rule out two possibilities. First, induction by toluene and propane may be partially attributed to their primary metabolites (e.g., cresols and 1-propanol). When DD4 cells are grown with toluene or propane, their primary metabolites are formed. Although these intermediate metabolites are transient and subject to rapid degradation, they may induce the expression of the tmo gene cluster to some extent. Second, along with this toluene MO, DD4 expresses other MOs (e.g., group-1 phenol hydroxylases, group-5 propane MO, group-3 butane MO, and particulate MO) that may also oxidize toluene or propane (Fig. S8 and detailed discussion in the supplemental material), generating cresols or 1-propanol, which in turn can enhance the transcription of tmo. In these two scenarios, the production of inducible metabolites by the toluene MO and other MOs in DD4 complicates the analysis of the extract regulation triggered by toluene or propane. Thus, further research utilizing quantitative induction assays such as a Δtmo mutant fused with a green fluorescent protein (GFP) reporter system or in vitro assays with purified regulatory proteins and promoter DNA is needed to differentiate between these alternatives. However, our RT-qPCR results qualitatively prove that toluene and propane are both inducers of the toluene MO, since their exposures resulted in the upregulation of tmoA transcription to significantly higher levels (P < 0.05) than those of their corresponding metabolites (e.g., p- and m-cresols and 1-propanol, respectively) (Fig. 7). Further, when RNA was extracted for RT-qPCR analysis, cells were harvested before half of the dosed substrate was degraded, ensuring induction resulted predominantly from the much greater exposure of the substrate (toluene or propane) than its metabolites. Our results also preclude the direct induction from dioxane and 1,1-DCE, because (i) DD4 will not grow with these two compounds without the supplement of carbon substrates and (ii) DD4 grown with pyruvate or succinate (noninducers) will not degrade dioxane or 1,1-DCE (Table S3).

Nonetheless, this study provides multiple lines of evidence to establish a short-chain gaseous alkane (i.e., propane) as both the substrate and inducer of a group-2 toluene MO, although it is not unusual for SDIMOs to be flexibly regulated by a wide range of substrates and their hydroxylated products. Toluene-2-MO from G4 can be induced by toluene, phenol, and TCE (52). In PKO1, toluene-p-MO expression was upregulated by toluene, benzene, and ethylbenzene (53), as well as m-cresol (54). Similarly, toluene-4-MO from KR1 can be positively regulated by toluene, CAHs (e.g., TCE, perchloroethylene, cis-1,2-dichloroethylene, and chlorothene) (55), and p-cresol (38). Elevated toluene degradation activities were also observed when KR1 was fed with medium-chain alkanes, such as hexane, pentane, and octane (55). For the toluene/o-xylene MO in OX1, toluene was a less effective inducer than one of the hydroxylated products, o-cresol (56). However, none of these previous studies have revealed the induction of group-2 toluene MOs by propane, which is applicable for bioremediation (e.g., biostimulation and bioaugmentation) in the field.

Importantly, upregulation of tmo by 1-propanol should be also noted, considering the implication potential of using 1-propanol as an alternative auxiliary substrate for DD4 bioaugmentation. Compared to propane, 1-propanol is a water-miscible liquid, which can be advantageous for streamlining the engineering efforts for field injection. As 1-propanol is not a substrate for this toluene MO, growth with 1-propanol is unlikely to pose competitive inhibition to fortuitous degradation of dioxane or 1,1-DCE, which is a common practice concern when primary substrates (e.g., propane) are used to stimulate cometabolism (57). In addition, 1-propanol is well suited for scalable applications, since it is nontoxic, commercially available, and relatively inexpensive. Collectively, this study presents adequate evidence establishing a link between a newly identified group-2 toluene MO with propane (a short-chain alkane gas) and 1-propanol (a primary alcohol) degradation and/or induction. This extends our horizon in understanding the substrate range and induction of SDIMOs for potential exploitation of their catalytic functions in natural and engineered systems.

Toluene MOs are known to be valuable for bioremediation given their versatile abilities to degrade common pollutants, such as aromatic compounds, CAHs, and polycyclic aromatic hydrocarbons (PAHs) (55, 58–60). However, the application of toluene MO-expressing bacteria for cometabolic bioremediation has been greatly hindered by the necessity of using toluene as the substrate, incurring risks of discharging this hazardous chemical into the environment. Although the genetic and biochemical induction mechanisms need to be further refined, this study uncovered a spectrum of nonhazardous gaseous and liquid substrates (e.g., propane and 1-propanol) that can induce the expression of the toluene MO in DD4. Concurrently, the promiscuity of the toluene MO allows the cometabolic degradation of dioxane, 1,1-DCE, toluene, and some other persistent pollutants that may cooccur in the field. Overall, our findings provide a foundation for the development of intrinsic bioremediation strategies with diverse auxiliary substrate options to enhance the cleanup effectiveness and field engineering practicality.

MATERIALS AND METHODS

Bacterial strains, plasmids, and growth conditions.

The Azoarcus and E. coli strains and the plasmids used in this study are listed in Table S1 in the supplemental material. E. coli was grown in LB at 37°C by following standard protocols (61). Azoarcus sp. strain DD4 was routinely grown aerobically at 30°C in VM-ethanol medium (62). For substrate degradation studies with either growing or resting cells or reverse transcription-quantitative PCR (RT-qPCR) analysis, nitrate mineral salts (NMS) medium supplemented with the indicated substrates was used as the growth medium.

Knockout mutagenesis of prm, bmo, and tmo genes in DD4.

Knockout mutants for prm, bmo, and tmo genes were constructed using in-frame deletion of respective gene clusters. Primers used to produce the homologous recombination arms are listed in Table S1. Upstream and downstream regions of the target genes were amplified by primers containing restriction sites and cloned in sequence into pBBRMCS-2 to generate the plasmid containing the deletion modification. The recombination insert was further subcloned into pK18mobsacB to generate the suicide plasmid that carries the recombinant sequence, which was then transferred into wt DD4 by triparental conjugation with the assistance of the helper strain E. coli(pRK2013). After incubation for 2 days at 30°C, cells were scraped from KON plates (63), suspended in KON liquid medium, and eventually transferred to VM-ethanol plates containing 25 μg/ml kanamycin. Single colonies were obtained by restreaking on VM-ethanol plates, and then successful exconjugants were confirmed using colony PCR. Individual exconjugants were further selected for double recombination on 10% sucrose in VM-ethanol agar without antibiotics. After being grown at 30°C for 48 h, strains with positive deletions of prm, bmo, and tmo gene clusters were examined by gel electrophoresis. Details regarding the plasmid construction, triparental conjugation, and exconjugant selection are provided in the supplemental material.

Catalytic assays using growing cells.

To identify the catalytic functions of three SDIMOs encoded by prm, bmo, and tmo, concentrations of corresponding substrates were monitored in the growing wt or deletion mutant cells. Unless otherwise stated, all assays were operated in 160-ml serum vials in triplicate. The initial 0.1 mg protein biomass of wt or mutant DD4 was inoculated into 20 ml NMS medium supplemented with individual substrates, including propane, butane, toluene, and ethanol. The serum vials were sealed with butyl rubber stoppers and crimped with aluminum caps before being aerobically cultured at 30°C. For propane and butane degradation assays, an initial concentration of 20 mg/liter propane (instrument grade, purity of >99.5%; Airgas) or butane (chemically pure, purity of >99.0%; Airgas) in the gas phase was injected. To avoid the inhibitory effects derived from high exposure of toluene, the initial aqueous phase concentration of toluene used was not allowed to exceed 80 mg/liter. To cultivate DD4 mutants, ethanol was used as the primary substrate at an initial concentration of 200 mg/liter. Abiotic controls were prepared without cell suspensions. At the selected incubation intervals, 700 μl of aqueous or 100 μl of headspace sample was removed and analyzed for the disappearance of the amended substrate by gas chromatography-flame ionization detection (GC-FID). For the MO inhibition observation, allylthiourea (ATU) was dosed to a concentration of 25 μM in the NMS medium. Acetylene was injected to achieve 10% (vol/vol) of the headspace volume. Concentrations of acetylene were monitored by GC-FID to confirm no gas leakage during incubation.

To assess cometabolism of dioxane, wt DD4 or knockout mutants were grown in NMS initially supplemented with 200 mg/liter ethanol (as the substrate) and 10 mg/liter dioxane. Dioxane concentrations were monitored over time by GC-FID. Analytical approaches regarding biomass and chemical measurements are detailed in the supplemental material.

Catalytic assays using resting cells.

To test the transformation capabilities for 1,1-DCE, wt or mutant DD4 clones were pregrown with ethanol (200 mg/liter) in NMS media. The transformation assay was conducted in 35-ml serum vials containing 4.5 ml of 1× phosphate-buffered saline (PBS) and 0.5 ml of cell suspension. The initial biomass was estimated as 1.5 mg total protein per vial. In the beginning, 1,1-DCE was dosed at 3.0 mg/liter as the equilibrium aqueous-phase concentration. Abiotic controls were prepared without cell suspensions. At selected incubation times, 100-μl-headspace samples were removed and analyzed for the disappearance of 1,1-DCE by GC-FID.

Heterologously expression of toluene MO.

A 4,997-bp fragment of the tmoABCDEF gene cluster was amplified using TMO/pETF and TMO/pETR as primers and the genomic DNA of DD4 as the template. PCR mixtures (50 μl) consisted of 1× PCR buffer, 100 nM deoxynucleotide triphosphates, 250 nM either primer, 1 U of Pfu polymerase (Thermo, Carlsbad, CA), and 10 ng of the DD4 genomic DNA. Thermocycling conditions were 98°C for 5 min, 30 cycles of 98°C for 20 s and 62°C for 6 min, and then 72°C for 10 min at the end. Restriction sites NdeI and NheI were introduced at the 5′ ends of the forward and reverse primers, respectively (Table S1). The PCR amplicon and vector pET-28a(+) were both digested with NheI and NdeI (New England Biolabs, Ipswich, MA). After gel purification, the linearized plasmid and PCR insert were ligated at a 1:3 (plasmid/insert) ratio at 16°C overnight with T4 DNA ligase (New England Biolabs, Ipswich, MA). The ligation mixture (1 μl) then was used to transform electrocompetent E. coli DH5α cells to generate pET-tmo. Recombinant plasmid pET-tmo then was harvested using the Zyppy plasmid miniprep kit (Zymo, Irvine, CA) and transformed into E. coli BL21(DE3) via electroporation with the MicroPulser electroporator (Bio-Rad, Hercules, CA). For enzyme expression, BL21(DE3) transformants with the empty vector pET-28a(+) or recombinant plasmid pET-tmo were grown in LB medium containing 25 μg/ml kanamycin at 30°C. When the optical density at 600 nm (OD₆₀₀) increased to 0.6, IPTG (0.5 mM) was added to induce protein expression for an additional 4 h at 30°C before cell harvesting. The production of toluene MO components in cell lysates was examined using SDS-PAGE analysis as detailed in the supplemental material.

To examine their degradation activities, IPTG-induced transformants BL21(DE3) (pET-tmo) and BL21(DE3) [pET-28a(+)] were washed after growth in LB medium and suspended in 5 ml 1× PBS in 35-ml serum vials sealed with rubber caps. The initial biomass of transformants was 1.5 mg total protein per vial. The initial amount of toluene, dioxane, 1,1-DCE, and propane was 1.43, 0.49, 0.84, and 2.28 μmol, respectively, in each bottle equivalent to 26.4, 2.73, and 8.6 mg/liter toluene, 1,1-DCE, and dioxane in the aqueous phase and 3.5 mg/liter propane in the gas phase based on the equilibrium following their Henry’s Law constants (64). Abiotic controls were prepared without cell suspensions. Treatments were all incubated at 30°C while being shaken at 175 rpm and performed in triplicate. The disappearance of the four added compounds was monitored by GC-FID. Instant degradation rates were calculated by averaging substrate disappearance in triplicate within the first 2 h, which was further normalized by the initial biomass as measured by total protein using the Bradford assay (12, 14).

RNA extraction and RT-qPCR analysis.

To quantify the expression levels of the tmo gene cluster under different substrate conditions, RT-qPCR was conducted using tmoA as the target gene and the 16S rRNA gene of DD4 as the housekeeping gene to normalize experimental variance. Primers specific for tmoA and 16S rRNA genes in DD4 were designed and are listed in Table S1. A broad array of substrates and metabolites was first screened to assess if they can support the growth of DD4 and also stimulate dioxane and 1,1-DCE biodegradation activities. Growth of DD4 was monitored by the OD at 600 nm when cultivated in NMS amended with individual substrates (50 mg/liter) at 30°C for 24 h. After growth with different substrates, cells were washed and concentrated to an OD₆₀₀ of 1.5 in 5 ml 1× PBS and tested for dioxane or 1,1-DCE removal after exposure for 1 h. Based upon these screening results, we selected toluene, m-cresol, p-cresol, 3-methylcatechol (3-MC), 4-methylcatechol (4-MC), propane, 1-propanol, propionaldehyde, and propionic acid for further RT-qPCR analysis considering their relevance to toluene and propane oxidation and distinct effects in inducing dioxane and 1,1-DCE biodegradation. DD4 cells were cultured using these selected compounds at an initial dosage of 2.0 mM as the sole carbon source at 30°C while being shaken at 150 rpm. As a control, pyruvate treatment was conducted as the control in parallel. When grown to the exponential phase, DD4 cells were harvested for total RNA extraction before substrate concentrations dropped below half the initial dosage.

The total RNA was extracted using the PureLink RNA minikit (Thermo, Carlsbad, CA) according to the manufacturer's protocol, in combination with an on-column PureLink DNase treatment (Thermo, Carlsbad, CA) to eliminate interference from DNA. cDNA was synthesized using the high-capacity cDNA reverse transcription kit (Thermo, Carlsbad, CA) and then purified using the DNA Clean & Concentrator-5 kit (Zymo, Irvine, CA). RT-qPCR mixtures contained 1 μl of diluted cDNA (5 ng/μl), 10 μl of 2× Power SYBR green PCR master mix (Thermo, Carlsbad, CA), 0.3 μM forward and reverse primers, and DNA-free water to a total volume of 20 μl. RT-qPCR was performed with a QuantStudio 3 real-time PCR system (Thermo, Carlsbad, CA) with the following temperature setup: 95°C for 10 min and then 40 cycles of 95°C for 15 s and 60°C for 1 min. Differential gene expression was quantified using the 2^−ΔΔCq method (8), and the expression fold change was calculated with the following formula: ΔΔC_q, target gene = (C_q, target gene − C_q, housekeeping gene) treatment − (C_q, target gene − C_q, housekeeping gene) control.