Characterization and evolution of anthranilate 1,2-dioxygenase from Acinetobacter sp. strain ADP1

Abstract

The two-component anthranilate 1,2-dioxygenase of the bacterium Acinetobacter sp. strain ADP1 was expressed in Escherichia coli and purified to homogeneity. This enzyme converts anthranilate (2-aminobenzoate) to catechol with insertion of both atoms of O(2) and consumption of one NADH. The terminal oxygenase component formed an alpha(3)beta(3) hexamer of 54- and 19-kDa subunits. Biochemical analyses demonstrated one Rieske-type [2Fe-2S] center and one mononuclear nonheme iron center in each large oxygenase subunit. The reductase component, which transfers electrons from NADH to the oxygenase component, was found to contain approximately one flavin adenine dinucleotide and one ferredoxin-type [2Fe-2S] center per 39-kDa monomer. Activities of the combined components were measured as rates and quantities of NADH oxidation, substrate disappearance, product appearance, and O(2) consumption. Anthranilate conversion to catechol was stoichiometrically coupled to NADH oxidation and O(2) consumption. The substrate analog benzoate was converted to a nonaromatic benzoate 1,2-diol with similarly tight coupling. This latter activity is identical to that of the related benzoate 1, 2-dioxygenase. A variant anthranilate 1,2-dioxygenase, previously found to convey temperature sensitivity in vivo because of a methionine-to-lysine change in the large oxygenase subunit, was purified and characterized. The purified M43K variant, however, did not hydroxylate anthranilate or benzoate at either the permissive (23 degrees C) or nonpermissive (39 degrees C) growth temperatures. The wild-type anthranilate 1,2-dioxygenase did not efficiently hydroxylate methylated or halogenated benzoates, despite its sequence similarity to broad-substrate specific dioxygenases that do. Phylogenetic trees of the alpha and beta subunits of these terminal dioxygenases that act on natural and xenobiotic substrates indicated that the subunits of each terminal oxygenase evolved from a common ancestral two-subunit component.

FIG. 1

Degradation of anthranilate and benzoate in Acinetobacter sp. strain ADP1 via the β-ketoadipate pathway (19). Relevant compounds and roles of the component AntDO (AntAB and AntC) proteins are indicated. The corresponding BenDO (Ben) components and reactions are enclosed within dashed boxes.

FIG. 2

SDS-PAGE of AntAB, AntC, and the M43K AntAB variant. Lane 1, total protein of noninduced TOP10F′(pBAC209); lane 2, total protein of TOP10F′(pBAC209) with AntAB induced; lane 3, purified AntAB; lane 4, total protein of noninduced DH5αF′(pBAC208); lane 5, total protein of DH5αF′(pBAC208) with AntC induced; lane 6, purified AntC; lane 7, total protein of noninduced TOP10F′(pDMK3). Lanes 8 and 9 are the cell extract and pellet, respectively, following sonication and centrifugation of TOP10F′(pDMK3) with M43K AntAB induced. Numbers at left correspond to molecular masses (in kilodaltons) corresponding to the adjacent markers (lane MW).

FIG. 3

UV-visible absorption spectra of oxidized and enzymatically reduced AntAB (12 μM α₃β₃) and oxidized M43K AntAB (15 μM α₃β₃) (inset) (A) and oxidized AntC reconstituted with FAD or as isolated in normal room light (FAD depleted) (B).

FIG. 4

EPR spectra of reduced AntAB and AntC. Purified samples of protein (250 μM) were reduced anaerobically with excess sodium dithionite (0.5 mM). Spectra were recorded under the following conditions: temperature, 10 K; microwave frequency, 9.59 GHz; modulation amplitude, 6.366 G; microwave power, 4 mW.

FIG. 5

Phylogenetic trees of the α subunits (a), β subunits (b), and reductases (c) of known and putative class IB DOs. Also included were two α subunit components from different ARHDO classes (Table 1): NDO (NdoB, class III, M23914 [27]) and phthalate DO (Pht3, class IA, D13229 [34]). An outgroup (Pht3) was used only for panel a. Abbreviations: Abs, 2-aminobenzenesulfonate DO of Alcaligenes sp. strain AF109074) (29); Ant, AntDO of Acinetobacter sp. strain ADP1 (6); Ant(Pa), putative sequence of P. aeruginosa PAO1; Ben, BenDO of Acinetobacter sp. strain ADP1 (32); Ben(Pa), putative sequence of P. aeruginosa PAO1; Ben(Pp), sequence of P. putida PRS2000 (8); Cbd, 2-halobenzoate 1,2-DO of B. cepacia (X79076 [16]); Tft, 2,4,5-trichlorophenoxyacetic acid oxygenase of B. cepacia (U11420 [9]); Xyl, toluate 1,2-DO of P. putida (PIR A41659 [17]). The numbers at branch points indicate the confidence (in percent) as determined by bootstrap analysis with 100 replicates. The scale bar indicates the relative phylogenetic distances measured as number of amino acid substitutions per site.

FIG. 6

Sequence of ADP1 AntA and a consensus sequence derived from an alignment of the α subunits from nine class IB ARHDOs and the class III NDO (Table 1 and Fig. 5). The consensus sequence indicates residues that are identical in the nine class IB sequences. Bold residues are also identical in NdoB of NDO. Residues known to furnish ligands to the Rieske and mononuclear iron sites in NdoB are boxed. The italicized asparagine at position 214 of the consensus sequence is a possible ligand to the mononuclear iron site (24) and is conserved in nine of the sequences but not in AbsA.