INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Styrene degradation in microbial systems is of considerable interest for two reasons. As an important monomer in the manufacturing of polymers, styrene frequently enters our natural environment. Styrene itself is toxic to living systems in fairly low amounts mainly due to membrane-related effects (5). In addition, the immediate degradation product of styrene in human liver, styrene oxide, is a known carcinogen (15). Furthermore, styrene possesses malodorous properties when present in amounts less than 1 ppm (22). All of this makes efficient removal of styrene from the environment highly desirable, and microorganisms have been found to play a key role in this process (16).

Secondly, reactions typically involved in styrene degradation possess significant potential in synthetic organic chemistry. Rhodococcus rhodochrous NCIMB 13259 was shown to degrade styrene via a dioxygenase attack on the aromatic ring, leading to 3-vinylcatechol after rearomatization (54). 3-Vinyl-1,2-cis-dihydroxycyclohexa-3,5-diene, the transformation product of the dioxygenase reaction, served as the starting point for an enantioselective organic synthesis of ()-Zeylena, exemplifying a route to cyclohexene oxides with antitumor potential (26). A variety of bacteria from the genus Pseudomonas (22, 30, 38, 54) have been shown to transform styrene through an attack on the vinylic side chain to styrene oxide and subsequently to phenylacetaldehyde. From a biotechnological point of view, this pathway contains at least two potentially useful reactions for the formation of enantiopure styrene oxide, which is known as a valuable building block in the manufacturing of optically active compounds such as pharmaceuticals (17). These reactions could be either the enantiospecific formation of styrene oxide and/or the kinetic resolution of racemic styrene oxide. The enantioselective oxidation of styrene has been observed in a chemical mutant of the styrene degrader Pseudomonas putida S12 (37), for which no genetic data are available, leading to an enantiomeric excess (e.e.) of more than 98%. Resolution could in theory be catalyzed by styrene oxide hydrolases or styrene oxide isomerases. The latter reaction has been found in bacteria, where styrene oxide is converted to phenylacetaldehyde, but the enantioselectivity of the reaction has not been reported (4) or was low (37). The former reaction has been described for bacterial and fungal enzymes, which enantioselectively hydrolyze one enantiomer of racemic styrene oxide and produce an optically active vicinal diol, while the other styrene oxide enantiomer is left behind (42, 48). Until recently, Pseudomonas fluorescens ST has been the only microorganism in which the genes and enzymes that affect styrene degradation have been investigated in more detail (4). In another study, regulatory and structural genes involved in styrene degradation in Pseudomonas sp. strain Y2 have been analyzed and found to be very similar to the P. fluorescens ST genes (51). Up to now, no data on the enantiospecificities of any of the reactions have been published. Obviously, the concomitant presence of enzymes that produce and consume styrene oxide in one host limits the biotechnological potential. Therefore, we used recombinant strains of Escherichia coli producing exclusively either styrene monooxygenase or styrene oxide isomerase. The genes and biotechnological potential of the cognate enzymes involved in styrene degradation in Pseudomonas sp. strain VLB120 were identified, and this knowledge was exploited to design recombinant biocatalysts suited to the production of (S)-styrene oxide at high e.e.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Media, chemicals, strains, and plasmids. We used routinely Luria-Bertani (LB) broth (Difco, Detroit, Mich.) or M9 mineral medium (46) supplemented with MT trace element solution (43). When necessary, cultures were supplemented with kanamycin (final concentration, 50 mg/liter), ampicillin (150 mg/liter), chloramphenicol (30 mg/liter), or 1 mM indole. IPTG (isopropyl--D-1-thiogalactopyranoside) was added to a final concentration of 0.5 mM to liquid cultures except where otherwise indicated. Cells were routinely grown on horizontal shakers at 200 rpm and 30 or 37°C. Strains and plasmids used in this study are listed in Table 1. Restriction and DNA modification enzymes were obtained from Boehringer Mannheim (Rotkreuz, Switzerland), NEB (Schwalbach, Germany), Gibco (Basel, Switzerland), or Promega (Zurich, Switzerland). Chemicals were obtained from Fluka (Buchs, Switzerland) (styrene, >99%; styrene oxide, >97%; phenylacetic acid, >99%) and Sigma (Buchs, Switzerland) (phenylacetaldehyde, >90%).

Isolation and identification of Pseudomonas sp. strain VLB120. Pseudomonas sp. strain VLB120 was isolated in the area of Stuttgart, Germany, from forest ground with styrene as the sole substrate. Details on its isolation and physiology will be published elsewhere. Sequencing of the first 470 nucleotides of the 5' end of the 16S rRNA revealed that Pseudomonas sp. strain VLB120 is a member of the genus Pseudomonas but does not belong to any of the previously described species. It is most closely related to P. alcaligenes (LMG 1224-T), P. pseudoalcaligenes (LMG 1225-T and DSM 50188-T), P. oleovorans (DSM 1045-T), and P. aureofaciens (LMG 1245-T). All of these strains have a 16S rRNA sequence identity of 98.7% to Pseudomonas sp. strain VLB120.

Growth and maintenance of Pseudomonas sp. strain VLB120. Cells were routinely transferred once per month between M9 and MT mineral medium plates. The plates were stored in an atmosphere saturated with styrene at room temperature. All Pseudomonas sp. strain VLB120 cultures were started from such plates. Cells were inoculated into tubes with M9 mineral medium, and styrene was added to a final concentration of 1.7 mM with only the volume of the liquid phase taken into account. From these tubes, larger cultures were started with the same concentration of styrene. Where necessary, the addition of styrene was repeated after 12 h to increase biomass concentration. To test the growth of Pseudomonas sp. strain VLB120 on different substrates, cells were streaked onto M9 mineral medium plates and these were incubated at room temperature in an atmosphere saturated with the respective substrate. Phenylacetic acid was added to the plate to 0.1% (wt/vol).

Construction of a genomic library of Pseudomonas sp. strain VLB120 in E. coli DH10B. Genomic DNA of Pseudomonas sp. strain VLB120 was prepared from styrene-grown cells by using the Qiagen kit for isolation of genomic DNA (Qiagen, Basel, Switzerland). This preparation was digested with one of the following restriction enzymes: SalI, EcoRI, HindIII, BamHI, XhoI, and KpnI. One microgram of digested DNA was ligated to 50 ng of accordingly digested vector pZero2.1 (Invitrogen, Leek, The Netherlands) (the SalI digest was ligated into the XhoI site of the vector) in 0.5× KGB buffer (50 mM potassium glutamate, 12.5 mM Tris-acetate [pH 7.5], 5 mM magnesium acetate, 25 µg of bovine serum albumin per ml, 0.25 mM -mercaptoethanol) (46) supplemented with 1 mM ATP at 6°C for 24 h. The ligation mixtures were used to transform E. coli DH10B cells made competent by rubidium chloride (46). Cells were plated on LB medium supplemented with kanamycin and indole and incubated at 30°C.

DNA sequencing. The DNA sequence of a 5.7-kb XhoI genomic fragment of Pseudomonas sp. strain VLB120 was determined from deletion clones of pSPW1 or from subclones in pUC18Not by using fluorescently labeled 40 forward and reverse primers (MWG Biotech, Ebersberg, Germany) and a Sequenase kit with 7-deaza-dGTP (Amersham, Zurich, Switzerland), according to a cycle-sequencing protocol. The reaction mixes were separated and analyzed on a LICOR sequencer with BaseImagIR version 2.3 software (MWG Biotech). Subsequent data analysis was performed with the DNAstar software package (DNAstar, Madison, Wisconsin), the BlastP algorithm (1) of the GenBank database at the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/), and the FOLD program of the Genetics Computer Group (GCG) software package for mRNA analysis (GCG, Madison, Wisconsin).

DNA manipulation and constructs. Plasmids for deletion analysis of sty genes were obtained from pSPW1 by digestion with the enzymes indicated in Fig. 1C and by religation. For efficient labeling of translation products, the T7 terminator of pUJ9 was introduced as a 0.4-kb HindIII fragment into the HindIII site of pUC18Not/T7* to yield pPT7T. This plasmid possesses two promoters in tandem (lacZp and the T7 promoter) pointing towards the polylinker of pUC18Not. Either the whole XhoI fragment or isolated parts of it, as indicated in Fig. 1D, were inserted into the polylinker of pPT7T.

[³⁵S]methionine labeling of translation products. Plasmids carrying genes under control of the T7 promoter (Fig. 1D) were introduced into E. coli BL21(DE3) together with plasmid pVLT33, which provided a lacI^q gene to prevent early expression of the T7 RNA polymerase gene of E. coli BL21(DE3). Cells from an overnight culture were transferred to fresh LB medium supplemented with antibiotics and grown at 30°C to an optical density at 600 nm of 0.8. A volume of 2 ml of the culture was centrifuged, and cells were washed in 1.5 ml of M9 medium and resuspended in 1 ml of the same medium supplemented with 0.2% glucose and 0.02% methionine assay medium (Difco). The cells were incubated for 60 min at 30°C, IPTG was added to 0.6 mM, and the cultures continued to grow for 60 min at 30°C. Subsequently, rifampin was added to 600 mg/liter and the cultures were transferred to 42°C for 45 min and then to 30°C for 30 min. Afterwards, cells were incubated for 5 min with 10 µCi of [³⁵S]methionine (Hartmann Analytic, Braunschweig, Germany), collected by centrifugation, washed, and resuspended in sample buffer (150 mM Tris [pH 6.8], 3.5% sodium dodecyl sulfate [SDS], 25% glycerol, 0.06% bromphenol blue, 0.12% -mercaptoethanol). After being heated to 95°C, proteins were separated on an SDS-12% polyacrylamide gel which was subsequently fixed in 25% isopropanol-10% acetic acid-65% water, stained for 10 min with Coomassie brilliant blue (Bio-Rad, Glattbrugg, Switzerland), and destained (46), dried, and analyzed by exposing X-ray film to it.

Detection of metabolites and determination of enzyme activities in whole-cell assays. For -galactosidase assays, E. coli CC118 was transformed with the plasmids under study. Whole-cell assays for investigating biotransformations were carried out with E. coli JM101. Recombinant strains containing the plasmids under study were incubated in 50 or 100 ml of medium in the presence of the respective antibiotics. At an optical density at 450 nm of 0.3 to 0.4, cells were induced by the addition of IPTG and the incubation was continued for 4 h, during which the cells entered stationary phase. Cells were then either subjected to a -galactosidase assay, as described elsewhere (34), or used for determination of specific activities. In the latter case, cells were harvested and resuspended to a certain cell dry weight (CDW) in 100 mM potassium phosphate buffer, pH 7.4, containing 1% (wt/vol) glucose. Aliquots of 1 ml were distributed in Pyrex tubes and incubated horizontally on a rotary shaker at 200 rpm. After 3 min, substrate was added to a final concentration of 1.5 mM from a 20-fold stock solution in ethanol. The reaction continued for 5 min in the shaker and was then stopped by incubating the sample on ice and by immediately adding 1 ml of ice-cold ether containing 0.1 mM 1-dodecanol as an internal standard. After addition of saturating amounts of sodium chloride, the water phase was extracted by vigorous shaking for 5 min at 30°C and the phases were separated by centrifugation. The organic phase was dried over anhydrous sodium sulfate and analyzed by gas chromatography.

1

Determination of e.e. and enantiomeric ratio. The e.e. was determined from the concentrations of the styrene oxide stereoisomers according to the following equation: e.e. = |(S  R)|/(S + R), with S and R representing the concentrations of the two stereoisomers. To calculate the selectivity of the isomerase reaction, the extent of phenylacetaldehyde formation from racemic styrene oxide was determined and corrected for formation of phenylacetaldehyde due to the injector temperature. Where necessary, the assay time was shortened to ensure that at least 30% of the starting material was left. Under the assumption of treating an irreversible reaction, the enantiomeric ratio, E, for the kinetic resolution of racemic styrene oxide was derived from the e.e. of the remaining substrate and the extent of conversion, as described elsewhere (6).

N-terminal sequencing of StyA. A hexahistidine affinity tag (H₆) was added to the C terminus of StyA to allow preparation of a protein fraction enriched in a modified StyA-H₆. E. coli JM101(p3His) was grown and induced as described for whole-cell assays, harvested by centrifugation, and treated according to the protocol for nondenaturing isolation of hexahistidine-tagged proteins on nickel-nitrilotriacetic acid spin columns (Qiagen). In short, proteins were loaded onto a Ni-affinity column in the presence of 20 mM imidazole to prevent unspecific binding and eluted with 250 mM imidazole. This protocol yielded a fraction which contained approximately 90% StyA-H₆. Separation of proteins in the resulting enriched fraction by SDS-polyacrylamide gel electrophoresis (PAGE) and transfer of the proteins to polyvinylidene difluoride membranes was performed as described previously (31). The N-terminal amino acid sequence of StyA-H₆ was determined on an Applied Biosystems automatic protein sequencer by automated Edman degradation.

Nucleotide sequence accession number. DNA sequences were deposited into publicly accessible databases (the 5.7-kb XhoI fragment was assigned GenBank accession no. AF031161; the 16S rRNA sequence was assigned EMBL accession no. AF224883).

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Growth of Pseudomonas sp. strain VLB120 on different substrates. Pseudomonas sp. strain VLB120 was able to use styrene, styrene oxide, phenylacetaldehyde, and phenylacetic acid as growth substrates when plated on M9 mineral medium. The strain did not grow on toluene, benzene, ethylbenzene, 1-phenylethanol, 2-phenylethanol, or 1,2-phenylethanediol. This growth substrate profile strongly suggests that epoxidation of the vinylic side chain was the first step in styrene degradation.

Selection of clones carrying styrene degradation genes from an E. coli gene library. It is well documented that some mono- and dioxygenases can convert indole in E. coli to indoxyl with subsequent spontaneous formation of indigo (13, 33, 39). Xylene oxygenase from the TOL plasmid is a monooxygenase that catalyzes the oxidation of toluene to benzylalcohol (20). In addition, it oxidizes styrene to styrene oxide (59) and initiates indigo formation in E. coli probably by converting indole to indoxyl (33). In analogy to this, we reasoned that E. coli clones carrying styrene oxidation genes from Pseudomonas sp. strain VLB120 might be able to produce indigo and turn blue when grown on LB medium plates containing indole. Of the six different genomic libraries of Pseudomonas sp. strain VLB120 that we prepared in E. coli DH10B, three yielded blue colonies on LB medium-indole plates after 2 days of incubation (XhoI, BamHI, and KpnI gene libraries). Of these, the library based on XhoI fragments yielded one blue colony per approximately 500 transformants. Plasmid DNA was isolated, and 85% of the 48 colonies contained a 5.7-kb insert in pZero2.1, in the majority of cases together with one of a variety of second inserts. Of the plasmids isolated from these transformants, four conveyed the ability to convert styrene to phenylacetaldehyde to E. coli DHB10 grown in liquid culture with or without induction by IPTG. They all harbored only the 5.7-kb XhoI fragment. One such plasmid was named pSPW1 and served as the basis for further studies. Transformants from the BamHI and the KpnI libraries did not yield clones that showed the desired transformation ability, nor did the restriction patterns of plasmids from such clones match the 5.7-kb XhoI fragment.

Sequence of the 5.7-kb XhoI fragment. The sequence of both DNA strands was determined. Computer analysis of the sequence revealed six ORFs (ORFs 1 to 6) in the same direction as the lacZp promoter (Fig. 1B). ORF 1 did not possess a SD sequence similar to the known consensus sequence (47). For ORF 6, we reasoned that the translational start with the best SD sequence was most likely the actual translation start (AAGGAGN₇ATG). The properties of the ORFs and their predicted gene products are listed in Table 2. We compared the deduced amino acid sequences of the first two ORFs to entries in the GenBank database by using the BlastP algorithm (1), and homologs with more than 20% sequence identity are listed in Table 3. Homologs to proteins encoded by ORFs 3 to 6 have been discussed elsewhere (see Discussion) (4).

View larger version (29K): [in this window] [in a new window]   FIG. 1.   (A) Proposed reaction sequence mediated by enzymes encoded on the 5.7-kb XhoI fragment of Pseudomonas sp. strain VLB120. Compounds, from left to right: styrene, styrene oxide (the asterisk indicates the chiral carbon atom), phenylacetaldehyde, and phenylacetic acid. (B) Restriction site map and genetic structure of the XhoI fragment in pSPW1. Stippled boxes indicate vector pZero2.1-derived sequences. The lacZp promoter is provided by the vector. The putative sty promoter is denoted styAp. The restriction sites external to the two XhoI sites, and therefore external to the insert, were derived from the pZero2.1 polylinker. Arrows indicate annealing sites for PCR primers P1 to P4 (see Materials and Methods). (C) Analysis of the XhoI fragment with different deletion derivatives of pSPW1. The sites used for deletions are indicated (and also presented in panel B). The lacZp promoter remained unaffected by the deletions. Designations of the plasmids resulting from the deletions are indicated on the right. E. coli recombinant strains carrying such plasmids were analyzed for formation of styrene oxide (So) from styrene or of phenylacetaldehyde and 2-phenylethanol (Pl) from styrene oxide or production of indigo (In) after overnight growth on LB medium. Plus signs indicate quantitative conversion of the substrate or blue color for indigo; minus signs indicate no conversion or no blue color in the culture medium. The plus sign in parentheses indicates an at least 12-fold-lower styrene oxide specific activity than that indicated by a plus sign without parentheses (conversion was not complete). (D) pPT7T-derived constructs for T7 RNA polymerase-based labeling of translation products. The fragments were obtained by digestion with the indicated restriction enzymes and introduced into the pPT7T polylinker with either the same sites in the polylinker or sites indicated additionally. The position of the T7 promoter (T7p) and the T7 terminator (stem-loop structure) are given only for one construct but are present in all plasmids. The names of the resulting plasmids are given on the right. Restriction sites relevant for constructions: A, ApaI; Bg, BglII; C, ClaI; E, EcoRI; F, FspI; H, HincII; Nc, NcoI; Nr, NruI; P, PstI; Sl, SalI; Sc, SacI; Sm, SmaI; Sn, SnaBI; St, StuI; Xb, XbaI; Xh, XhoI.

Radioactive labeling of proteins encoded on the 5.7-kb XhoI fragment. DNA fragments covering either the complete XhoI fragment or the single predicted ORFs were placed under the control of a T7 RNA polymerase-driven promoter (Fig. 1D). [³⁵S]methionine labeling of translation products showed that at least five proteins were encoded on the XhoI fragment (Fig. 2). The vector-only control indicated that transcription termination by the T7 terminator was not complete (Fig. 2, lane 1), since three bands were detected. As the bla gene encoding -lactamase in pPT7T is transcribed from the same strand of DNA as the insert under study, it is likely that these bands correspond to the complete bla gene transcript at 31.6 kDa, the mature -lactamase immediately under it, and a putative degradation product at approximately 27 kDa. An identical pattern of bands derived from the -lactamase primary transcript has been found elsewhere (12). While a plasmid carrying ORF 1 did not yield an identifiable product other than that encoded by the vector-only control (not shown), plasmids carrying ORFs 2 to 6 did. Furthermore, for each ORF a translation product that was in good agreement with protein sizes predicted from the DNA sequence and with proteins synthesized from the original XhoI fragment were identified (Fig. 2, lane 2). Based on this, we considered ORFs 2 to 6 to represent the genes styRABCD with their cognate protein products. ORF 1 was tentatively designated stySc (see below). Interestingly, expression of the sty genes from the XhoI fragment led to prominent bands for StyR, StyB and/or StyC, and StyD, but only to a faint band for StyA. Consistent with this, the StyA band was very weak in a recombinant carrying pT7ST-A (Fig. 2, lane 4). However, from previous SDS-PAGE analysis it was likely that StyA could be expressed to significantly higher levels in E. coli JM101(pSPW1) (data not shown). To investigate whether sequences upstream of styA were involved in determining the weak expression of styA from pT7ST-A, we modified the DNA sequences upstream of the translation start of styA. First, the original translation start of StyA was verified by N-terminal sequencing. It read MKKRIGIVGAGTAGLHLGLF and was identical to the sequence predicted from ORF 3. Subsequently, an EcoRI site together with a modified SD sequence was inserted in front of the ATG codon of styA, removing at the same time the original upstream sequences, resulting in pT7ST-Am carrying styA* (Fig. 3A). The sequence exchange led to a very prominent StyA band with the same apparent size as StyA synthesized from pT7ST-A in labeling experiments (Fig. 2, lanes 4 and 5).

View larger version (59K): [in this window] [in a new window]   FIG. 2.   [35S]methionine labeling of T7 RNA polymerase-based translation products. E. coli BL21(DE3) carrying pPT7T without the insert (lane 1), with the complete 5.7-kb XhoI fragment (lane 2), or with fragments containing only one complete ORF (lanes 3 to 8) was induced with IPTG; proteins synthesized under control of the T7 promoter were labeled with [35S]methionine. Protein extracts were separated on an SDS-12% polyacrylamide gel. Lanes: 1, pPT7T; 2, pT7ST-ScD; 3, pT7ST-R; 4, pT7ST-A; 5, pT7ST-Am; 6, pT7ST-B; 7, pT7ST-C; 8, pT7ST-D.

View larger version (25K): [in this window] [in a new window]   FIG. 3.   Effect of shortening or removing the 163-bp sequence upstream of styA in the presence or absence of the regulatory genes styScR. (A) Structure of the investigated plasmids. Drawing is not to scale. The stippled and hatched boxes represent the wild-type SD sequence and altered SD sequences of styA, respectively. The latter was inserted to provide translational signals. (B) The left graph shows styrene oxide formation in E. coli JM101 cells carrying one of the plasmids, 1 to 3, shown in panel A, measured 4 h after IPTG induction. The right graph shows styrene oxide formation in E. coli JM101 cells carrying pSTFull and pSTHalf in the presence of low-copy-number plasmid pCKST-ScR expressing styScR from lacZp, measured 4 h after IPTG induction.

Deletion analysis of the 5.7-kb XhoI fragment. To correlate the information derived from the DNA sequence for styScRABC with functions of the proteins, we performed a deletion analysis of the 5.7-kb XhoI fragment in E. coli JM101 (Fig. 1C). From the nucleotide sequence, we assumed that styABCD were structural genes for conversion of styrene to phenylacetic acid, with styAB encoding the components of a styrene monooxygenase. In partial agreement with this, we found the minimum DNA element needed for indigo formation in E. coli to consist only of styA; cells carrying pSPW2 or pSPW6 produced indigo, while those carrying pSPW3 did not. Given our hypothesis that the reactions that convert indole and styrene are performed by the same enzyme, this would attribute the oxidation activity to StyA. However, to detect styrene oxidation activity in whole-cell assays, elements upstream of styA had to be present as in E. coli JM101(pSPW6). Since indigo formation and accumulation over time is a much more sensitive indication of styrene monooxygenase activity than styrene oxide formation in whole-cell assays, this indicates that elements upstream of styA can increase specific styrene monooxygenase activity but are not necessarily a structural part of the enzyme. The same conclusion resulted from comparison of the transformation abilities of E. coli JM101(pSPW1), which was able to transform styrene to phenylacetaldehyde, and E. coli JM101(pSPW2), which was not. However, inspection of the results obtained with E. coli JM101 carrying pSPW5 and pSPW6 indicated that StyB was necessary for maximal activity of StyA in the presence of the elements upstream of styA, as it increased the rate of styrene oxide formation at least 12-fold.

Functional analysis of StyB and StyC. To further investigate the role of StyB, we carried out a series of experiments using plasmids carrying styA*, thus eliminating all of the upstream sequences interfering with styAB expression (see above). Genes styA* and styB were provided on the same or different plasmids, and expression from both genes was driven from the lac promoter. Table 4 shows clearly that in E. coli JM101, the presence of styA* alone was sufficient to oxidize styrene (experiment 3), while expression of styB alone did not lead to detectable styrene oxide formation (experiment 2). Nevertheless, to achieve high levels of activity, the presence of styB was also required (experiment 4 [Table 4]). These results confirmed that gene products other than those of styAB do not play a structural role in styrene monooxygenase function. As shown in experiments 5 and 6 (Table 4), providing styAB on a high-copy-number plasmid increased styrene oxide formation rates relative to those achieved with a low-copy-number plasmid. However, the correct molar ratio of the two proteins appeared to play a role in the efficiency of styrene oxide formation (experiments 4 and 5 [Table 4]). Although these data qualitatively confirm results obtained for equivalent experiments with genes of P. fluorescens ST (4), it is important to note that in contrast to results of the earlier study, the data presented in this report were derived from genetic elements containing only the genes of interest, i.e., styA* and styB, without major portions of upstream DNA.

Enantioselectivity of styrene oxide formation. Having clarified the functions of the various structural genes, we determined the enantiospecificity as a key parameter of the cognate enzymes to evaluate their usefulness for biocatalysis. Using E. coli JM101(pSPW5), which carries styScRAB under lacZp promoter control, we have investigated the e.e. of styrene oxide formed in whole-cell biotransformations and found it to be greater than 99% (Fig. 4C) by chiral gas chromatography. (S)-Styrene oxide produced by E. coli JM101(pBG63) expressing the genes for TOL plasmid-derived xylene oxygenase served as a reference (Fig. 4B) (58). The enantiomer formed by styrene monooxygenase eluted at the same time as the major enantiomer formed by xylene oxygenase, indicating that the product consisted of at least 99.5% (S)-styrene oxide.

View larger version (19K): [in this window] [in a new window]   FIG. 4.   Production of (S)-styrene oxide by StyAB. (A) Separation of 1.5 mM racemic styrene oxide on a chiral cyclodextrin column. The peak at 18.7 min is phenylacetaldehyde, which can be formed due to the injector temperature. (B) (S)-Styrene oxide with an e.e. of 93% formed by E. coli JM101(pBG63), which produces xylene oxygenase and served as a reference. (C) (S)-Styrene oxide is formed with an e.e. of >99% by E. coli JM101(pSPW5), which produces styrene monooxygenase. Columns were intentionally overloaded to obtain a signal for the least abundant enantiomer in panels B and C.

Enantiomeric ratio of styrene oxide consumption. E. coli JM101(pT7ST-C) recombinants synthesizing styrene oxide isomerase from a high-copy-number plasmid showed a very high reaction rate (experiment 8 [Table 4]). However, the enzyme demonstrated only a slight preference for the transformation of the S enantiomer of racemic styrene oxide. From the enantiomeric excess of remaining styrene oxide and the extent of conversion when terminating whole-cell assays before the complete consumption of styrene oxide, we determined the enantiomeric ratio, E, of styrene oxide conversion to be 1.4.

Expression of the styrene monooxygenase structural genes. To clarify why we could not obtain measurable styrene oxide formation by E. coli JM101(pSPW2), we constructed two high-copy-number plasmids carrying styAB under lacZp promoter control using either the SacI site 163 bp upstream (pSTFull) or the NcoI site 77 bp upstream (pSTHalf) of the translational start of styA (Fig. 3A) and compared the resulting styrene monooxygenase expression to that from pT7ST-ABm. The latter construct contained the same modified ribosome binding site as pT7ST-Am (see above). As shown in the left panel of Fig. 3B, we were unable to detect formation of styrene oxide with either of the first two constructs, but specific activities were very high with the construct in which all of the upstream DNA sequence had been eliminated. Furthermore, SDS-PAGE analysis of whole-cell extracts from IPTG-induced E. coli JM101 carrying the respective plasmids showed a prominent band at 47 kDa corresponding to the predicted size of StyA with either the original plasmid pSPW1 or pT7ST-ABm (data not shown). This excludes the possibility that the new SD sequence is the sole cause of the differences of translation efficiency between pSTFull or pSTHalf and pT7ST-ABm. No band at 47 kDa was visible with either pSTFull or pSTHalf (results not shown). Apparently, the presence of the 163 bp of wild-type DNA upstream of styA leads to a lack of detectable styrene oxide formation in recombinant strains.

Regulatory functions encoded by stySc and styR. Amino acid sequence similarities for styScR-encoded assumed or observed proteins suggested that the gene products of these two ORFs represent a two-component regulatory system. With the results obtained so far in mind, it appeared likely that the two proteins together form a positive control system for styrene degradation that also alleviates the block in styrene monooxygenase synthesis, as observed in E. coli JM101(pSPW2). To investigate this further, we repeated the experiments with pSTFull and pSTHalf in the presence of styScR under lacZp promoter control on the low-copy-number plasmid pCKO1 (Fig. 3B, right panel). When styScR were expressed, the production of styrene oxide by StyAB from pSTFull was clearly measurable, while it was not detectable when pSTHalf was present. Part of this effect was also observed with a translational fusion of styA including the upstream region from the SacI site to a lacZ reporter gene. The -galactosidase activity derived from the reporter gene was increased 21-fold in E. coli CC118(pCKST-ScR, pSPW20) containing styScR compared to E. coli CC118(pCKO1, pSPW20) without the regulatory genes (Fig. 5B, columns 1 and 2). We investigated the functions of styScR in more detail by providing stySc, styR, and the styA-lacZ fusion on different plasmids (Fig. 5B, columns 3 to 6). Basal activity of the reporter protein was found when only stySc was expressed (450 Miller units). However, expression of styR increased -galactosidase activity 38-fold and expression of stySc and styR at the same time increased the activity 115-fold, relative to stySc expression only.

View larger version (23K): [in this window] [in a new window]   FIG. 5.   Involvement of stySc and styR in transcriptional regulation of sty structural genes. (A) Plasmids used for analysis. The basic replicon determining the plasmid copy number is indicated in brackets. The hatched elements represent the styAp promoter and the part of styA that has been fused to the 'lacZ gene of pUJ9. Drawing is not to scale. (B) Transcriptional activity of IPTG-induced E. coli CC118 carrying different plasmid combinations. Left panel, experiment with styScR present on one plasmid; right panel: experiments with stySc and styR present on two different plasmids. Minus signs indicate that the vector without the insert was present as a control; plus signs indicate the presence of the plasmid shown on the left.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Biotechnological potential of styrene oxide monooxygenase. Enantioselective biotransformations of styrene to styrene oxide or kinetic resolution of racemic styrene oxide and its derivatives have received a lot of attention in recent years (36, 37, 41, 42, 55, 58). P. putida S12 and Mycobacterium sp. strain E3 were shown to produce the two possible stereoisomers of styrene oxide with enantiomeric excesses greater than 98%, albeit with different reaction rates: 200 U/g of CDW for a mutant P. putida S12 and 5 U/g of CDW for Mycobacterium sp. strain E3 (37). Interestingly, E. coli recombinants carrying styAB of Pseudomonas sp. strain VLB120 produce the same enantiomer as the P. putida S12 mutant and E. coli recombinants carrying xylMA from the TOL plasmid upper pathway, the S enantiomer (21, 58). The enantiomeric excess of the reaction catalyzed by StyAB is more than 99% (Fig. 4) and is thus significantly better than that of the reaction catalyzed by xylene oxygenase (58). Furthermore, reaction rates could be increased 8-fold compared to E. coli recombinants synthesizing xylene oxygenase (Fig. 3; Table 4) (58). These results make recombinants synthesizing StyAB promising biocatalysts for future applications.

Biotechnological potential of styrene oxide isomerase. A second route to optically pure styrene oxide utilizes kinetic resolution of racemic mixtures of styrene oxide. This route has been investigated especially for fungal epoxide hydrolases, where it was proven to produce excellent optical yields (42). Bearing in mind that the first step of styrene transformation in VLB120 is enantioselective and that microbial systems for the enantioselective conversion of epoxides had been described previously (56, 57), we investigated whether the second step, conversion of the epoxide, was also enantioselective. Our results suggest that this is not the case for the Pseudomonas sp. strain VLB120-derived enzyme StyC. Both styrene oxide enantiomers are converted, with only slightly different reaction rates. The poor enantiomeric ratio, E, of 1.4 makes styrene oxide isomerase from Pseudomonas sp. strain VLB120 a poor enzyme for kinetic resolution of styrene oxide enantiomers. A similar observation was made for P. putida S12 grown on phenylacetic acid, which converted (S)-styrene oxide slightly faster than (R)-styrene oxide (37).

Genetics of styrene degradation in Pseudomonas sp. strain VLB120. Combination of the sequence data and biotransformation studies suggests that Pseudomonas sp. strain VLB120 transforms styrene to styrene oxide through the action of a two-component styrene monooxygenase, StyAB, in which StyA is the epoxidizing component and StyB might function as an enhancer of monooxygenase activity, either as a provider of reducing equivalents whose action might be replaced to a small degree by unspecific reductases of the E. coli host or in an up to now poorly understood mode similar to the action of the HpaC protein of E. coli W (44). As a second step, styrene oxide is isomerized to phenylacetaldehyde by the styrene oxide isomerase StyC, and phenylacetaldehyde is very likely to be converted by the phenylacetaldehyde dehydrogenase StyD to phenylacetic acid. Recently, a similar system from P. fluorescens ST has been analyzed at the sequence and functional levels (4). In the regions of the structural genes, the systems have a high degree of sequence identity (between 91.8 and 96.6%). The equivalent enzymes appear to catalyze identical reactions. To this end, it is noteworthy that although the specific activity of phenylacetaldehyde formation was high in E. coli JM101(pT7ST-C) (experiment 8 [Table 4]), we did not observe a specific protein band by SDS-PAGE (not shown) indicating high specific activity of the enzyme. We assume that formation of small amounts of 2-phenylethanol, which was observed in deletion analyses, is not part of the activity of StyC. We did not detect this reaction product when E. coli JM101(pT7ST-C) was added to whole-cell assays under conditions of excess styrene oxide. Possibly, 2-phenylethanol formation is catalyzed by enzymes of the host. The rate of 2-phenylethanol formation remained thus below the detection level when small amounts (5.6 mg/liter) of cell material were provided but not when cells were added to 2.5 g/liter, as was done in the deletion analyses. In agreement with this, E. coli JM109 was found to convert phenylacetaldehyde to 2-phenylethanol (4). Furthermore, P. putida S12 also appears to convert styrene oxide to phenylacetaldehyde with no formation of 2-phenylethanol (37).

Regulation of styrene degradation. The protein homologs of StySc and StyR (Table 3) suggest that these proteins constitute a two-component regulatory system. StySc possesses sequence motifs that are typical for signal-transmitting domains of sensors, especially the H box, containing the catalytic histidine typical for its function as a histidine kinase, as well as the N, D, F, and G boxes (49). Given the small size of the protein, which leaves no space for a signal-receiving domain, stySc is likely to be the truncated 3' part of a larger gene, styS.

Functional analysis of the regulatory genes. No StySc protein was detected in T7 expression experiments. Given the predicted size of StySc (27 kDa), it is possible that the band corresponding to the degradation product of -lactamase comigrates with the respective band in the T7 expression experiment. Alternatively, the signal emitted from StySc could be too low, relative to the intensity of the -lactamase-related bands, to be detected. On the other hand, it appears reasonable to assume that stySc encodes a functional polypeptide that is produced in E. coli. It has been shown that liberated domains of transcriptional regulators in two-component regulatory systems may retain their activity in a constitutive fashion (2, 35), and expression of stySc augmented transcriptional activity of the putative styA promoter when styR was expressed at the same time (Fig. 5).

Analysis of the region between styR and styA. Genes styR and styA are separated by 180 bp, which apparently prevents the use of heterologous promoters for the synthesis of considerable levels of styrene oxygenase (Fig. 3). From the transcriptional activity assays involving the HincII fragment containing this region (Fig. 5), we postulate that StyR binds to this region and initiates transcription of the downstream genes, thus constituting the styAp promoter.

View larger version (24K): [in this window] [in a new window]   FIG. 6.   Analysis of the region between styR and styA. (A) Schematic representation of the region as it is present in pSTFull, which contains the complete 163 bp between the SacI site and the ATG codon of styA, including the vector-derived lacZp. Vector- and insert-derived sequences are indicated. Sequence length numbers refer to the two segments for which mRNA secondary structures were predicted. A putative LacZ fusion peptide is terminated by an insert-derived in-frame stop codon. Known (+1, for lacZa-mRNA) or assumed (lightly shaded arrow, for styA-mRNA) mRNA synthesis starts are indicated. The inverted repeats of the tod box are indicated by repetitive carets, while the larger repeats present in the sty system are indicated by open arrows. (B) Predicted mRNA secondary structures when either the region between the SacI site upstream of the 163-bp sequence and the HincII site downstream of the ATG of styA or the region between the NcoI and the HincII site is transcribed. Free energies for the entire structures are given. Relevant sequences are boxed. The tod box sequence is printed in boldface type, and the AUG codon of styA mRNA is, additionally, italicized. (C) Amino acid sequence comparison of the putative protein regions involved directly in DNA binding for StyR and TodT, with a postulated consensus sequence for the equivalent region of class 3 response regulators derived from the alignment of 19 proteins (see text). Boxed letters indicate identical residues between StyR and TodT; italicized letters indicate identical residues in all three sequences.