Characterization of a C—C Bond Hydrolase from Sphingomonas wittichii RW1 with Novel Specificities towards Polychlorinated Biphenyl Metabolites

Chlorinated aromatic compounds, such as polychlorinated biphenyls (PCBs) and dibenzofurans, are among the most widespread, toxic, and/or persistent environmental pollutants. The Bph and Dxn/Dbf pathways responsible for the aerobic bacterial catabolism of biphenyl and dibenzofuran, respectively, have been extensively studied, in part due to their potential for remediating environments contaminated with the polychlorinated compounds (2, 10). The pathways share three homologous enzymes (Fig. ​(Fig.1):1): a multicomponent dioxygenase that catalyzes the initial step of ring hydroxylation, an extradiol dioxygenase that catalyzes oxygenolytic cleavage of the catecholic intermediate at the meta position to yield 2-hydroxy-6-oxo-6-phenylhexa-2,4-dienoate (HOPDA) or ortho-substituted HOPDA, and a meta-cleavage product (MCP) hydrolase that catalyzes an unusual C—C bond cleavage of the HOPDA. FIG. 1. Microbial catabolism of dibenzofuran and biphenyl by the Dbf and Bph pathways, respectively. Carbon atoms 1, 6, and 8 of 8-OH HOPDA are numbered. The Dbf pathway enzymes are as follows: DxnA1 and DxnA2, large and small subunits of dioxin dioxygenase, ... A general limitation of bacterial catabolic pathways for the degradation of complex industrial mixtures is that one or more pathway enzymes lack the requisite broad substrate specificity to degrade all man-made (xenobiotic) congeners of the naturally occurring compounds, and some enzymes are susceptible to inhibition by the chlorinated metabolites. Studies have identified MCP hydrolases as a bottleneck in the Bph pathway, as summarized below, as well as in the toluene catabolic pathway (11). Substrate specificity studies of MCP hydrolases from potent PCB degraders, BphDLB400 from Burkholderia xenovorans LB400 and BphDP6 from Rhodococcus globerulus P6, revealed that these enzymes are inefficient in transforming HOPDA chlorinated at the 3 or 4 position (30, 31). In addition, these poorly transformed chlorinated metabolites are potential competitive inhibitors of the enzymes, thereby reducing the efficiency of transformation of all congeners, including unsubstituted HOPDA. Consistent with these studies, HOPDAs accumulate during the bacterial degradation of some PCBs (12). Interest in the inhibition of MCP hydrolases is further heightened with the recent discovery of HsaDH37Rv of Mycobacterium tuberculosis H37Rv (35). Formerly annotated as BphD, HsaDH37Rv appears to be involved in steroid metabolism during the survival of M. tuberculosis in the macrophage, hydrolyzing 4,5-9,10-diseco-3-hydroxy-5,9,17-trioxoandrosta-1(10),2-diene-4-oic acid (4,9-DSHA). MCP hydrolases are members of the α/β-hydrolase superfamily (28). The hydrolyzed C—C bond lies between a dienoate (C-1 to C-5) (Fig. ​(Fig.1)1) and a carbonyl (C-6). Hydrolysis is mediated by conserved Ser-His-Asp residues reminiscent of the catalytic triad of proteases. The proposed catalytic mechanism involves His-mediated ketonization of the dienoate (18, 23), followed by Ser-mediated hydrolysis at C-6. It is unclear whether the catalytic Ser attacks C-6 directly to form an acyl-enzyme intermediate or whether it activates a water molecule which then attacks C-6 (9, 18, 23, 24). Crystal structures have been reported for five MCP hydrolases: BphDLB400 (18), BphDRHA1 from Rhodococcus sp. strain RHA1 (26), CumDIP01 from Pseudomonas fluorescens IP01, involved in cumene degradation (13), CarCJ3 from Janthinobacterium sp. strain J3, involved in carbazole degradation (14), and MhpC (7). All comprise a “core domain,” consisting of an eight-stranded β-sheet flanked by α-helices, and a “lid domain,” which occurs as an insertion into the core domain. The active site, located between the core and lid domains, includes the conserved Ser, His, and Asp residues and is divided into two subsites: P (polar) and NP (nonpolar), located on each side of the catalytic serine. The P subsite binds the dienoate moiety, and the NP subsite accommodates the C-6 substituent, a phenyl ring in the case of HOPDA. Sphingomonas wittichii RW1, one of the best-characterized dibenzofuran-degrading bacteria (37), is able to grow on either dibenzofuran or dibenzo-p-dioxin as a sole organic substrate (36). In addition, the strain partially transforms 2-Cl, 3-Cl, 4-Cl, and 2,3-diCl dibenzofurans to the corresponding chlorinated hydroxysalicylates (36). It is not entirely clear which enzymes are responsible for these transformations, as the bacterium possesses multiple homologues of key catabolic enzymes. For example, Armengaud and coworkers isolated four genes that encode α subunits of ring-hydroxylating dioxygenases and two genes that encode β subunits (2). The bacterium also contains at least two extradiol dioxygenases and three MCP hydrolases (2, 4). Two of the MCP hydrolases, DxnB and DxnB2 (previously H1 and H2), were purified and shown to hydrolyze HOPDA and 8-hydroxy HOPDA (4). In addition, the gene encoding a third potential MCP hydrolase was cloned (2). DxnB2 has been reported to be more related to MCP hydrolases involved in the catabolism of monocyclic aromatic compounds than to enzymes such as BphD, which are involved in the degradation of bicyclic aromatics (15). This result, derived from amino acid sequence comparisons, is somewhat surprising as the substrates of DxnB and BphD differ by a single hydroxyl group on the phenyl ring. Better characterization of the DxnB enzymes may clarify the phylogenetic relationships of MCP hydrolases and provide insight into why S. wittichii RW1 contains multiple isozymes. Herein, we report the cloning, sequencing, and heterologous expression of dxnB2. Heterologously produced DxnB2 was purified, and steady-state kinetic studies revealed that the enzyme possesses useful activities towards chlorinated HOPDAs. We generated a structure-based sequence alignment of MCP hydrolases and used this to perform phylogenetic analyses. The results identified a previously unrecognized class of MCP hydrolases and revealed possible structural determinants of substrate specificity in the various classes.