Toluene Monooxygenase-Catalyzed Epoxidation of Alkenes

Abstract

Several toluene monooxygenase-producing organisms were tested for their ability to oxidize linear alkenes and chloroalkenes three to eight carbons long. Each of the wild-type organisms degraded all of the alkenes that were tested. Epoxides were produced during the oxidation of butene, butadiene, and pentene but not hexene or octadiene. A strain of Escherichia coli expressing the cloned toluene-4-monooxygenase (T4MO) of Pseudomonas mendocina KR1 was able to oxidize butene, butadiene, pentene, and hexene but not octadiene, producing epoxides from all of the substrates that were oxidized. A T4MO-deficient variant of P. mendocina KR1 oxidized alkenes that were five to eight carbons long, but no epoxides were detected, suggesting the presence of multiple alkene-degrading enzymes in this organism. The alkene oxidation rates varied widely (ranging from 0.01 to 0.33 μmol of substrate/min/mg of cell protein) and were specific for each organism-substrate pair. The enantiomeric purity of the epoxide products also varied widely, ranging from 54 to >90% of a single epoxide enantiomer. In the absence of more preferred substrates, such as toluene or alkenes, the epoxides underwent further toluene monooxygenase-catalyzed transformations, forming products that were not identified.

The reactivity of epoxides makes them useful and important intermediates for industrial chemical syntheses, including the production of pharmaceuticals, agrochemicals, and polymers. Although there are many applications for racemic epoxides, which are usually produced by chemical means, the demand for the production of enantiomerically pure feedstocks, including epoxides, has increased in recent years. Between 1996 and 1997, the sale of enantiomerically pure pharmaceuticals increased 21%, and similar increases are expected in the agrochemical and polymer markets (19). This increased interest in enantiomeric purity arises from the observation that the (R) and (S) enantiomers of an individual compound often have appreciably different biological and physical properties. For example, the (S) enantiomer of carvone gives caraway seeds their distinctive odor, whereas the (R) enantiomer is perceived as spearmint (10). Similarly, (S)-thalidomide can cause severe birth defects, while (R)-thalidomide is a safe and effective sedative (10). These dramatic differences have led the U.S. Food and Drug Administration to require that each enantiomer of a racemic drug be tested individually prior to approval of a drug (10), thereby greatly increasing the cost of bringing a new drug to market. By using enantiomerically pure intermediates in drug synthesis, the number of isomers of a new drug that require testing prior to approval can be reduced.

One possible method for producing enantiomerically pure epoxides, that of using enzymes to enantioselectively insert oxygen across the carbon double bonds of alkenes, has been explored, with various degrees of success. The styrene monooxygenase of Pseudomonas sp. strain VLB120 converts styrene to (S)-styrene oxide with an enantiomer excess of >99% (16). Another enzyme, the alkane hydroxylase of Pseudomonas oleovorans, converts 1,7-octadiene to optically active (R)-7,8-epoxy-1-octene with an enantiomeric purity of 92%; the latter is also oxidized by the enzyme, forming (R,R)-1,2-7,8-diepoxyoctane with an enantiomeric purity of 83% (12). These observations led to industrial applications for the synthesis of the drugs Metoprolol and Atenolol (1).

A class of oxygenases with a non-heme diiron cluster at the catalytic center is also known to form epoxides from alkenes. One such enzyme, the soluble methane monooxygenase (MMO) of Methylosinus trichosporium OB3b, mediated the epoxidation of propene, 1-butene, 2-butene, and 1,3-butadiene, but with very low enantiomeric specificity (less than 64% of a single isomer) (15). The epoxides formed from propene, butene, and butadiene were not oxidized further by MMO under the experimental conditions described, but in other experiments, ethylene epoxide and cis-dichloroethylene (cis-DCE) epoxide were shown to be substrates for MMO (20). MMO was unable to oxidize the larger alkenes 1-pentene, cyclohexene, and 3-methyl-butene (15). In contrast, the alkene monooxygenase of Xanthobacter sp. strain Py2 can form epoxides from alkenes and chlorinated alkenes, but with a higher degree of enantiomeric selectivity. For example, the oxidation of 3-chloropropene yielded 80% (S)-3-chloro-1,2-epoxypropene, whereas the oxidation of 1-butene yielded 94% of the (R) isomer (6, 7). The formation of epoxides by other bacterial systems has been reviewed in detail elsewhere (1, 2).

In this report, we discuss the epoxidation of alkenes by a subgroup of the non-heme diiron-containing enzymes, the toluene monooxygenases, which hydroxylate the aromatic ring of toluene, forming ortho-, meta-, and para-cresols. The rate and the enantiomeric selectivity of alkene oxidations varied with each enzyme-substrate pair. In some cases, the enantiomeric selectivity of the epoxidation reactions was >90%. We also found evidence that the epoxides formed as a result of alkene oxidation underwent additional toluene monooxygenase-mediated transformations.

MATERIALS AND METHODS

Chemicals.

The chemicals 1-butene (99%), 2-butene (99%, cis-trans mixture), 1,3-butadiene (99%), epoxybutane (99%), 1,3-butadiene monoepoxide (BME) (95%), butadiene diepoxide (97%), epichlorohydrin (99%), 1-pentene (99%), 2-pentene (99%, cis-trans mixture), hexene (97%), octadiene (95%), 2-chloropropene (99%), 2,3-dichloropropene (95%), 1,2-butanediol (95%), toluene (95%), triethylamine, 4-(p-nitrobenzyl)pyridine (PNBP), ethylene glycol, and isopropyl-β-thiogalactopyranoside (IPTG) were obtained from Aldrich Chemicals (Milwaukee, Wis.).

Growth and preparation of cells.

Pseudomonas mendocina KR1 (22), P. mendocina ENVpmx1 (a toluene-4-monooxygenase [T4MO]-deficient mutant of KR1) (K. McClay and R. J. Steffan, Abstr. 97th Gen. Meet. Am. Soc. Microbiol., abstr. K36, p. 348, 1997), Ralstonia pickettii PKO1 (3), Burkholderia cepacia G4 (18), Burkholderia sp. strain ENVBF1 (13), and Pseudomonas sp. strain ENVPC5 (13) were cultured overnight at 30°C in shake flasks containing basal salts medium (8) supplemented with 0.4% sodium glutamate. Toluene was included in the vapor phase of the cultures when the induction of toluene oxygenases was desired. Prior to substrate degradation assays, the cultures were harvested by centrifugation and resuspended in basal salts medium to an optical density at 550 nm (OD₅₅₀) of 2, unless otherwise indicated. A standard curve of optical density versus protein concentration for each strain was used to calculate the amount of protein per milliliter of the resuspended cultures. Escherichia coli DH10B containing plasmid pRS202 (17) was prepared in a similar manner, except that the strain was grown at 37°C in Luria-Bertani medium and was resuspended to an OD₅₅₀ in Luria-Bertani medium of 4 with 0.3 to 1 mM IPTG to induce the expression of T4MO.

Substrate degradation assays.

To determine the substrate range and the rates of alkene oxidation for the various toluene monooxygenase-producing organisms, duplicate 5-ml aliquots of the resuspended cultures were dispensed into 25-ml serum vials and crimp sealed with Teflon-faced septa. Gaseous substrates were added as pure compounds using a gas-tight syringe, and other substrates were added from 20% stock solutions in dimethylformamide (Fig. 1). The amounts of individual substrates added were as follows, unless otherwise indicated: 1-butene, 2-butene, and 1,3-butadiene, 2.2 μM; epoxybutane, BME, and epichlorohydrin, 8 μM; 1- and 2-pentene, 9 μM; hexene, 4 μM; octadiene, 4.5 μM; 2-chloropropene, 11 μM; and 2,3-dichloropropene, 12 μM. The serum vials were then placed horizontally on a rotary shaker operating at 100 rpm. During incubation, the temperature was maintained at between 20 and 22°C for the pseudomonads and at 37°C for E. coli. The serum vials were periodically removed from the shaker, and a 10- to 25-μl portion of the headspace gas was withdrawn through the septa and injected into a gas chromatograph (GC) equipped with a 30-m Vocol column (Supelco Inc., Bellefonte, Pa.) maintained at 160°C and a flame ionization detector. This procedure allowed us to monitor the concentrations of both the alkene substrates and the epoxide products. The same protocol was used to detect the formation of 1,2-butanediol and butadiene diepoxide, except that 1 μl of culture medium was injected into the column instead of headspace gas.

FIG. 1.

Structures of compounds used in this study.

Verification of epoxide formation.

The commercially available epoxides butene epoxide, BME, and butadiene diepoxide were used in GC analyses to determine the retention times of the authentic compounds and to quantitate the conversion of the alkenes to the corresponding epoxides. A modification of the method of van Hylckama et al. (20) was used to verify that the observed peaks that coeluted with the authentic compounds were volatile epoxides; the epoxides were conjugated with PNBP to form intensely colored adducts of the epoxides. Serum vials were prepared as described above, except that prior to sealing, a glass test tube (5 by 50 mm) was placed inside each vial with the opening of the test tube extending out of the liquid. The substrate was then added, and the vials were incubated with moderate shaking at a 45° angle, preventing the culture liquid from entering the test tube. The transformation of the alkenes was monitored by GC analyses. After the rate of transformation decreased significantly or 90% of the alkene substrate was depleted from the headspace, 400 μl of 100 mM PNBP dissolved in ethylene glycol was injected through the septum into the open end of the test tube. The vials were then incubated for 5 h before they were opened, and the epoxide-binding PNBP was withdrawn. The total volume of PNBP solution was combined with an equal volume of acetone-triethylamine (50:50) and mixed rapidly. Epoxide-PNBP adducts were detected by spectral analysis at 400 to 700 nm.

Determining enantiomeric ratios.

To determine the ratio of (R) and (S) isomers of the epoxides formed from the oxidation of the alkenes, samples were analyzed for the presence of the epoxides and the alkenes by GC analyses as described above. When the epoxide concentration neared its maximum or when 80 to 90% of the substrate was oxidized, the enantiomeric ratio of epoxides produced was determined by injecting a sample of the headspace gas into a GC equipped with a chiral separation column (RT-BDEXSE; Restek, Inc., Bellefonte, Pa.) and a flame ionization detector. The column was maintained at 50°C.

RESULTS

Alkene oxidation and identification of oxidation products.

All of the wild-type toluene monooxygenase-producing organisms tested were able to oxidize alkenes. Greater than 95% of the added butadiene (2.2 μmol) was oxidized by strains G4 and ENVBF1 during the first 5 h of incubation, whereas strains KR1 and ENVPC5 oxidized only 50% of the butadiene in 20 h (Fig. 2). Greater than 95% of the added 2-butene (2.2 μmol) was oxidized by all the strains tested, except ENVpmx1, within 20 h, with strains KR1 and ENVPC5 having higher initial degradation rates than strains G4 and ENVBF1 (Table 1). With the T4MO-deficient strain ENVpmx1, the concentrations of butadiene and 2-butene decreased by less than 10% during 20-h incubations. The T4MO clone, E. coli pRS202, oxidized all of the butenes tested.

FIG. 2.

Oxidation of 1,3-butadiene by toluene monooxygenase-producing organisms. Each datum point represents the average of the analysis of duplicate sample vials, with the range indicted by error bars. Symbols: closed circle, P. mendocina KR1; open triangle, ENVpmx1; open square, pRS202; closed triangle, ENVPC5; closed square, ENVBF1; closed diamond, B. cepacia G4.

TABLE 1.

Specific activity of toluene monooxygenases against alkenes and chlorinated alkenes

Substrate degraded	μmol of substrate degraded/min/mg of total cell protein by the following organisma:
	G4	KR1	ENVPC5	ENVBF1
1,3-Butadiene	0.19	0.07	0.09	0.13
2-Butene	0.14	0.26	0.17	0.12
1-Pentene	0.08	0.21	0.15	0.05
2-Pentene	0.16	0.33	0.18	0.14
2-Chloropropene	ND	0.07	0.18	0.16
2,3-Chloropropene	0.23	0.01	0.12	0.08

^aMeasured following 30 min of incubation, except for 2,3-chloropropene, for which the rate was determined after 60 min of incubation. ND, not determined. 

1-Pentene and 2-pentene were oxidized by all of the wild-type strains tested. In the case of the T4MO knockout mutant ENVpmx1, an extended lag period was followed by the rapid transformation of 2-pentene that exceeded the rates of 2-pentene oxidation observed in the wild-type organisms (Fig. 3). Similarly, hexene and octadiene were also transformed more rapidly by strain ENVpmx1 than by the wild-type organisms. E. coli(pRS202) expressing T4MO oxidized 1-pentene, 2-pentene, and hexene but did not oxidize octadiene.

FIG. 3.

Oxidation of 2-pentene by toluene monooxygenase-producing organisms. Each datum point represents the average of the analysis of duplicate sample vials, with the range indicted by error bars. Symbols: closed circle, P. mendocina KR1; open triangle, ENVpmx1; open square, pRS202; closed triangle, ENVPC5; closed square, ENVBF1; closed diamond, B. cepacia G4.

The chlorinated alkene 2-chloropropene was oxidized by strains KR1, ENVPC5, and ENVBF1. Even though all three of these organisms appear to produce the same class of toluene monooxygenase (as indicated by the appearance of para-cresol in the culture medium when grown on toluene), there were differences in their ability to oxidize this compound. Strains KR1 and ENVPC5 oxidized 2.1 and 4 μmol of substrate in the first 3.5 h of incubation, respectively, and oxidized only an additional 0.54 and 1.0 μmol, respectively, during the following 17 h. Strain ENVBF1 oxidized 2.3 μmol of 2-chloropropene in the first 3 h of incubation and oxidized 6.3 μmol of substrate in the following 17 h. It is not known why ENVBF1 continued to oxidize 2-chloropropene after oxidation by the other two T4MO-producing organisms had ceased.

The ability of the T4MO-producing organisms to oxidize alkenes was adversely affected by the presence of an additional chlorine atom on the substrate molecules. The amounts of 2,3-chloropropene oxidized by strains KR1, ENVPC5, and ENVBF1 were 95.2, 66.0, and 75.6% smaller than the amounts of 2-chloropropene oxidized by the same strains, respectively. Strain G4 was not tested on 2-chloropropene, but it was able to oxidize twice as much 2,3-chloropropene (6.7 μmol) as any of the T4MO-expressing organisms in this assay.

The amounts of the 3- and 4-carbon alkenes oxidized by the cultures were related to the specific activity of the organisms toward the given alkene. For example, 1,3-butadiene was oxidized by strain G4 at an initial rate of 0.19 μmol/min/mg of cell protein (Table 1) and was oxidized to a concentration below the limits of detection within the first 6 h of incubation (Fig. 2). Strain KR1 had an initial oxidation rate of 0.07 μmol/min/mg of cell protein and ultimately oxidized only 1.1 μmol (50%) of butadiene in 20 h. With 2-butene as a substrate, KR1 had an initial oxidation rate of 0.26 μmol/min/mg of cell protein and oxidized all of the 2-butene (2.2 μmol) in 3.5 h. Strain G4 oxidized 2-butene at an initial oxidation rate of only 0.14 μmol/min/mg of cell protein and required 20 h to oxidize >95% of the added compound.

The extent of pentene and halogenated propene oxidation achieved by the individual organisms was not well correlated with the initial rates of oxidation of the compounds. Strain KR1 had the highest initial 2-pentene oxidation rate (0.33 μmol/min/mg of cell protein). This rate was more than twice the initial rate of 2-pentene oxidation by G4 and BF1 (Table 1), yet BF1 and G4 oxidized more 2-pentene than KR1 during a 20-h incubation (Fig. 3). The T4MO-deficient mutant of KR1, strain ENVpmx1, oxidized 5- to 8-carbon alkenes efficiently, apparently utilizing an alternate enzyme system. Strain ENVPC5 had a higher initial oxidation rate than ENVBF1 for 2- and 2,3-chloropropene (Table 1) but ultimately oxidized less of these substrates over a 20-h incubation (data not shown).

Epoxide formation and degradation.

GC analysis showed that during 1,3-butadiene, 2-butene, 1-pentene, and 2-pentene oxidation by the wild-type organisms and E. coli(pRS202), a transient secondary peak appeared on chromatograms. A similar peak was observed during the oxidation of hexene by pRS202 but not when the wild-type organisms oxidized this substrate. These secondary peaks increased in proportion to the amount of alkene oxidized and then decreased following the depletion of the alkene. The peak formed during the oxidation of 1-butene and butadiene coeluted with the commercially available butene epoxide and BME. To verify that this peak and the corresponding peaks formed during the oxidation of the other alkenes were epoxides, they were conjugated with PNBP as described in Materials and Methods. Spectral analysis showed that the PNBP conjugates of the products of 1-butene and butadiene oxidation had absorbance maxima identical to those of the conjugates of the purchased epoxides and agreed closely with the data obtained for other epoxides (4, 20). The conjugates of the pentenes and 2-butene had similar absorbance spectra. From these data it was concluded that the secondary peaks were epoxides.

Even though the T4MO-deficient mutant ENVpmx1 efficiently oxidized pentene, hexene, and octadiene, no epoxides were detected during the oxidation of any of these compounds by this strain. Similarly, no epoxides were detected during the oxidation of hexene, octadiene, 2-chloropropene, or 2,3-chloropropene by the wild-type organisms.

The stoichiometry of epoxide formation was evaluated by incubating toluene-induced G4 with 4.4 μmol of 1,3-butadiene and monitoring both butadiene and BME concentrations. Initially, there was a nearly stoichiometric conversion of butadiene to BME (>95%), followed by a decrease in BME concentration after the parental compound was depleted (Fig. 4). In contrast, when G4 was incubated with 2.2 μmol of 2-butene, only 43% of the 2-butene oxidized could be detected as the epoxide product. Efforts to detect 2-butene-1-ol in liquid media were not successful.

FIG. 4.

Stoichiometric conversion of 1,3-butadiene to monoepoxide by T2MO-expressing B. cepacia G4. Each datum point represents the average of the analysis of duplicate sample vials, with the range indicated by error bars. Symbols: closed circle, 1,3-butadiene; open circle, BME.

To determine if the disappearance of the epoxides was caused by chemical or enzymatic reactions, BME was incubated with both toluene-induced and uninduced cultures of strain G4. Induced cells incubated with pure BME oxidized 8 μmol of the substrate in the first 50 min of incubation, whereas uninduced cells of G4 oxidized less than 1 μmol in the same time period. When toluene-induced cells of G4 were incubated in the presence of both toluene and BME, they degraded 0.5 μmol of BME in 50 min (Fig. 5). A similar inhibition of BME transformation was observed when BME was coincubated with butadiene. The oxidation of both toluene and butadiene was unaffected by the presence of BME, but butadiene oxidation was inhibited by toluene (data not shown). Similar results were obtained with ENVBF1. The uninduced cells degraded less than 1.3 μmol of BME in 24 h, whereas the induced cells oxidized 13.4 μmol. Epichlorohydrin and butene epoxide also served as substrates for toluene monooxygenases (Table 2).

FIG. 5.

Inhibition of butadiene and BME transformation by toluene in B. cepacia G4. Each datum point represents the average of the analysis of duplicate sample vials, with the range indicated with error bars. Symbols: closed circle, butadiene; open circle, butadiene in the presence of toluene; closed square, BME; open square, monoepoxide in the presence of toluene. The initial concentration of toluene in samples was 20 μmol. When the quantity of toluene was reduced to 10 to 11 μmol, an additional 10 μmol of toluene was added through the septa.

TABLE 2.

Toluene monooxygenase-catalyzed degradation of epoxides during a 150-min incubation

Strain	Oxygenase isoform	Mean (range) μmol of the following substrate degradeda:
		ECH	BE	BME
G4	T2MO	2.5 (0.2)	0.1 (0.2)	5.2 (0.2)
ENVBF1	T4MO	1.5 (0.4)	0.9 (0.3)	2.4 (0.5)
KR1	T4MO	0.93 (0.6)	1.1 (0.4)	0.4 (0.6)
PKO1	T3MO	ND	0.2 (0.0)	3.5 (0.1)

^aEight micromoles of each substrate was added to 5-ml cell suspensions (OD₅₅₀, 0.5). Duplicate assays were performed. ECH, epichlorohydrin; BE, butene epoxide; ND, not determined. 

Enantiomeric ratios of biologically produced epoxides.

The epoxides formed from the oxidation of 1-pentene and 1,3-butadiene were analyzed by chiral chromatography, and the results of these analyses are presented in Table 3. The enantiomeric selectivity observed during the formation of BME from 1,3-butadiene by the various toluene monooxygenases differed, but all of the oxygenases tested favored the production of the (S)-enantiomer. The highest selectivity occurred with strain G4 (toluene-2-monooxygenase [T2MO]) [91.9% (S)], and the lowest selectivity occurred with strain ENVPC5 (T4MO) [67.3% (S)].

TABLE 3.

Enantiomeric ratio of monoepoxides produced from terminal alkenes

Strain	Oxygenase isoform	Enantiomeric ratio of epoxides produceda
		1,3-Butadiene			1-Pentene
		(R)	(S)	No. (range) of assays	(R)	(S)	No. (range) of assays
G4	T2MO	8	92	6 (3)	100	0	2 (0)
ENVBF1	T4MO	22	78	2 (3)	66	34	2 (0)
ENVPC5	T4MO	33	67	2 (0)	72	28	1 (NA)
KR1	T4MO	33	67	7 (1)	54	46	5 (0)
PKO1	T3MO	20	80	1 (NA)	65	35	2 (2)

^aThe enantiomeric ratios of commercially available BME were 24% (R) and 76% (S) (±2%). NA, not applicable. 

The oxidation of 1-pentene showed greater variation in the observed selectivity, but in each case, a larger percentage of the epoxide formed was of the (R)-enantiomer. The product distribution ranged from 100% (R)-enantiomer formed by strain G4 to a low of 54.2% (R)-enantiomer formed by strain KR1 (Table 3).

DISCUSSION

Several non-heme diiron-containing monooxygenases form epoxides as the primary product of halogenated ethene oxidation (4, 21). In this study, we report the oxidation of 4-, 5-, and 6-carbon alkenes to their corresponding epoxides by several toluene monooxygenases. The enzymes tested oxidize toluene at the ortho, meta, or para position; each has considerable amino acid sequence homology to the hydroxylases of MMO, the alkene monooxygenase of Xanthobacter sp. strain PY2, and the alkene monooxygenase of Rhodococcus rhodochrous B276 (23), all of which are known to oxidize alkenes to their corresponding epoxides. We also screened the toluene dioxygenases of Pseudomonas putida F1 (24) and Pseudomonas sp. strain GZ-9 and found that although they oxidized alkenes, no epoxides were formed (data not shown). These findings are consistent with those of Lange and Wackett (11), who showed that the toluene dioxygenase of P. putida F1 oxidizes a wide range of alkenes and haloalkenes, forming alcohols, diols, and ketones.

All of the toluene monooxygenase-producing strains that we examined catalyzed the epoxidation of short-chain alkenes. When expressed in E. coli, the T4MO of KR1 formed epoxides from 1-butene, 2-butene, 1,3-butadiene, 1-pentene, 2-pentene, and 1-hexene, but octadiene was not oxidized. The wild-type organisms R. pickettii PKO1, B. cepacia G4, P. mendocina KR1, Pseudomonas sp. strain ENVBF1, and Pseudomonas sp. strain ENVPC5 oxidized all of the nonhalogenated alkenes tested. However, the oxidation of hexene and octadiene by the wild-type organisms did not lead to the formation of epoxides.

Although 2-chloropropene and 2,3-chloropropene were oxidized by the strains tested, no epoxides were detected by GC analysis. Nonetheless, we suspect that epoxides were formed. The epoxides of 2-chloropropene and 2,3-chloropropene would have a chlorine atom bonded to one of the epoxide ring carbons, and such an arrangement would result in an unstable epoxide that would undergo rapid chemical hydrolysis, thereby preventing accumulation and detection. Similar results have been described for the oxidation of various halogenated alkenes, such as trichloroethylene (TCE) and vinyl chloride (4, 20).

The rates of alkene oxidation observed during this study varied between the substrates and the strains tested, yet there was no apparent correlation between the regiospecificity of toluene oxidation and the rate of alkene oxidation by these enzymes. However, because whole-cell assays were used in this study, the oxidation rates observed could have been affected by factors other than the kinetic characteristics of the toluene monooxygenases. For example, differences in membrane compositions or transport mechanisms or the presence of alternative alkene oxidation pathways could affect the observed transformation rates. This situation was demonstrated by comparing the rates of alkene degradation by P. mendocina KR1, the T4MO-deficient mutant ENVpmx1, and E. coli(pRS202) expressing T4MO. Essentially no 1,3-butadiene or 2-butene was oxidized by ENVpmx1 in the first 5 h of incubation, whereas KR1 and E. coli(pRS202) oxidized 30 to 77% of the butadiene and nearly 100% of the 2-butene. Because T4MO was shown to facilitate short-chain alkene degradation in KR1, these results were not surprising. However, ENVpmx1 oxidized 2-pentene, hexene, and octadiene more rapidly than either E. coli(pRS202) or KR1. Although we previously have shown that strain KR1 can oxidize alkanes (13), presumably by using a separate alkane oxidation pathway, it was surprising that the T4MO-deficient mutant would oxidize a greater quantity of the alkenes than the wild-type strain, which has both functional pathways available to oxidize alkenes. However, ENVpmx1 does contain the lux operon, which KR1 lacks. The light-producing apparatus of the lux operon catalyzes the oxidation of aldehydes (5), and it is possible that an enzyme present in KR1 converts pentene, hexene, and octadiene to aldehydes, which are then rapidly utilized by the lux genes, thereby participating in the metabolism of the alkenes. We observed that the alkenes used in this study induced the expression of the T4MO::lux transcriptional fusion present in ENVpmx1 (data not shown), ensuring that the products of the lux genes would be present throughout the assay.

Another possible explanation for the difference in observed alkene oxidation rates between these two strains is that the epoxides that result from alkene oxidations damage T4MO-producing cells. It has been reported that TCE and vinyl chloride epoxides are harmful to cells, presumably because they react with cellular components (4, 14). Even though unhalogenated epoxides are more stable than halogenated ethenes, with a half-life on the order of days rather than seconds (20), it is still possible that they have a similar toxic effect. If ENVpmx1 does not produce epoxides during alkene oxidation, the overall health of the cells may be better than that of toluene monooxygenase-expressing cultures, allowing faster degradation of the alkenes.

Yeager et al. (21) recently reported that the T2MO of G4 produces epoxides during the oxidation of ethene and propene, with 96% of the propene being converted to propene oxide. We obtained similar results, with 104% (±10%) of the butadiene being converted to BME. When 2-butene was oxidized by G4, however, the epoxide did not accumulate stoichiometrically. Although all of the added 2-butene was depleted, only 43% was recovered as the corresponding epoxide, and the remaining 57% was unaccounted for (data not shown). Similarly, when MMO oxidized 2-butene, 54% of the oxidized substrate was converted to the epoxide, while the remaining 46% was converted to 2-butene-1-ol (15). Because toluene monooxygenases also can hydroxylate methyl groups (17), it is possible that a similar reaction occurred in our 2-butene oxidation assays, thereby limiting the amount of epoxide generated.

As stated previously, the accumulation of epoxides was transient. The disappearance of the epoxides could be the result of either biological or chemical reactions. Previous experiments with MMO found that no diepoxide was formed from butadiene, and the potential for MMO-catalyzed hydrolysis of BME was not suggested (15). It was suggested, however, that BME may be too large to fit into the active site of MMO. Conversely, van Hylckama et al. (20) found that both ethene epoxide and cis-DCE epoxide served as substrates for MMO, although the products of the reaction were not identified. Here, when toluene-induced and uninduced cultures of G4 or ENVBF1 were incubated with BME, the uninduced cells did not oxidize BME, whereas the induced cells did. Furthermore, the presence of toluene and butadiene inhibited the degradation of BME. These results show that although BME may undergo slow chemical hydrolysis in the presence of water at a neutral pH, toluene monooxygenase-expressing cells catalyze a more rapid oxidation of BME. Thus, it is likely that toluene monooxygenase acts as the catalyst. Epichlorohydrin and butane epoxide were also degraded by toluene monooxygenase-expressing organisms. We were unable to determine if the transformation of the epoxides led to the hydrolysis of the epoxide ring, forming the corresponding diol, or if the oxygenases hydroxylated the compounds at one of the acyclic carbons.

Because of the growing interest in enantiomerically pure feedstocks for both industrial and pharmaceutical chemical syntheses, we examined the enantiomeric selectivity of the epoxidation reactions catalyzed by the various toluene monooxygenases with 1,3-butadiene and 1-pentene as substrates. In some cases, the toluene monooxygenases catalyzed epoxidation reactions with a high degree of enantiomeric selectivity (Table 3). For example, when the T2MO of G4 oxidized 1-pentene, only one enantiomer of pentene epoxide could be detected. Thus, strain G4 may be a candidate catalyst for producing pentene epoxide for use in pharmaceutical syntheses.

In this study, we examined variants of toluene monooxygenase that oxidize the aromatic ring of toluene at all three possible positions with three representatives, KR1, ENVPC5, and ENVBF1, of the T4MO variety. (In a previous report [13], we suggested that the toluene monooxygenase of ENVBF1 was a T2MO, based on oxygen consumption studies following growth on toluene, but we have since discovered that the cloned ENVBF1 toluene oxygenase genes produce p-cresol during toluene oxidation.) When 1,3-butadiene served as the substrate, the three toluene monooxygenase isoforms that oxidize chloroform (13) had the lowest degree of enantiomeric selectivity. The enantiomeric selectivities were similar to that of MMO [36% (R) and 64% (S)], which also oxidizes chloroform (9). Therefore, it appears that the enzyme plasticity that allows these related enzymes to oxidize a wide range of substrates, including chloroform, may limit their applicability for producing highly pure epoxide enantiomers.

With the exception of the oxidations catalyzed by the T2MO of G4, the enantiomeric selectivities of the oxidations catalyzed by the wild-type toluene monooxygenases are probably too low for them to be useful for the commercial production of enantiomerically pure epoxides. However, it might be possible to improve the enantiomeric selectivity of these reactions through site-directed mutagenesis of toluene monooxygenases. More research, including an evaluation of product yields, stability, and toxicity, is needed to further understand the potential of the toluene monooxygenases as biocatalysts for the production of enantiomerically pure epoxides.