Rhodococcus jostii RHA1, a polychlorinated biphenyl-degrading soil bacterium whose genome has been sequenced, shows lignin degrading activity in two recently developed spectrophotometric assays. Bioinformatic analysis reveals two unannotated peroxidase genes present in the genome of R. jostii RHA1 with sequence similarity to open reading frames in other lignin-degrading microbes. They are members of the Dyp peroxidase family and were annotated as DypA and DypB, on the basis of bioinformatic analysis. Assay of gene deletion mutants using a colorimetric lignin degradation assay reveals that a ΔdypB mutant shows greatly reduced lignin degradation activity, consistent with a role in lignin breakdown. Recombinant DypB protein shows activity in the colorimetric assay and shows Michaelis-Menten kinetic behavior using Kraft lignin as a substrate. DypB is activated by Mn(2+) by 5-23-fold using a range of assay substrates, and breakdown of wheat straw lignocellulose by recombinant DypB is observed over 24-48 h in the presence of 1 mM MnCl(2). Incubation of recombinant DypB with a β-aryl ether lignin model compound shows time-dependent turnover, giving vanillin as a product, indicating that C(α)-C(β) bond cleavage has taken place. This reaction is inhibited by addition of diaphorase, consistent with a radical mechanism for C-C bond cleavage. Stopped-flow kinetic analysis of the DypB-catalyzed reaction shows reaction between the intermediate compound I (397 nm) and either Mn(II) (k(obs) = 2.35 s(-1)) or the β-aryl ether (k(obs) = 3.10 s(-1)), in the latter case also showing a transient at 417 nm, consistent with a compound II intermediate. These results indicate that DypB has a significant role in lignin degradation in R. jostii RHA1, is able to oxidize both polymeric lignin and a lignin model compound, and appears to have both Mn(II) and lignin oxidation sites. This is the first detailed characterization of a recombinant bacterial lignin peroxidase.

The soil bacterium Rhodococcus jostii RHA1 contains two dye-decolorizing peroxidases (DyPs) named according to the subfamily they represent: DypA, predicted to be periplasmic, and DypB, implicated in lignin degradation. Steady-state kinetic studies of these enzymes revealed that they have much lower peroxidase activities than C- and D-type DyPs. Nevertheless, DypA showed 6-fold greater apparent specificity for the anthraquinone dye Reactive Blue 4 (k(cat)/K(m) = 12800 ± 600 M(-1) s(-1)) than either ABTS or pyrogallol, consistent with previously characterized DyPs. By contrast, DypB showed the greatest apparent specificity for ABTS (k(cat)/K(m) = 2000 ± 100 M(-1) s(-1)) and also oxidized Mn(II) (k(cat)/K(m) = 25.1 ± 0.1 M(-1) s(-1)). Further differences were detected using electron paramagnetic resonance (EPR) spectroscopy: while both DyPs contained high-spin (S = (5)/(2)) Fe(III) in the resting state, DypA had a rhombic high-spin signal (g(y) = 6.32, g(x) = 5.45, and g(z) = 1.97) while DypB had a predominantly axial signal (g(y) = 6.09, g(x) = 5.45, and g(z) = 1.99). Moreover, DypA reacted with H(2)O(2) to generate an intermediate with features of compound II (Fe(IV)═O). By contrast, DypB reacted with H(2)O(2) with a second-order rate constant of (1.79 ± 0.06) × 10(5) M(-1) s(-1) to generate a relatively stable green-colored intermediate (t(1/2) ∼ 9 min). While the electron absorption spectrum of this intermediate was similar to that of compound I of plant-type peroxidases, its EPR spectrum was more consistent with a poorly coupled protein-based radical than with an [Fe(IV)═O Por(•)](+) species. The X-ray crystal structure of DypB, determined to 1.4 Å resolution, revealed a hexacoordinated heme iron with histidine and a solvent species occupying axial positions. A solvent channel potentially provides access to the distal face of the heme for H(2)O(2). A shallow pocket exposes heme propionates to the solvent and contains a cluster of acidic residues that potentially bind Mn(II). Insight into the structure and function of DypB facilitates its engineering for the improved degradation of lignocellulose.

Lovastatin, a cyclic nonaketide from Aspergillus terreus, is a hypercholesterolemic agent and a precursor to simvastatin, a semi-synthetic cholesterol-lowering drug. The biosynthesis of the lovastatin backbone (dihydromonacolin L) and the final 2-methylbutyryl decoration have been fully characterized. However, it remains unclear how two central reactions are catalyzed, namely, introduction of the 4a,5-double bond and hydroxylation at C-8. A cytochrome P450 gene, lovA, clustered with polyketide synthase lovB, has been a prime candidate for these reactions, but inability to obtain LovA recombinant enzyme has impeded detailed biochemical analyses. The synthetic codon optimization and/or N-terminal peptide replacement of lovA allowed the lovA expression in yeast (Saccharomyces cerevisiae). Both in vivo feeding and in vitro enzyme assays showed that LovA catalyzed the conversion of dihydromonacolin L acid to monacolin L acid and monacolin J acid, two proposed pathway intermediates in the biosynthesis of lovastatin. LovA was demonstrated to catalyze the regio- and stereo-specific hydroxylation of monacolin L acid to yield monacolin J acid. These results demonstrate that LovA is the single enzyme that performs both of the two elusive oxidative reactions in the lovastatin biosynthesis.

Rhodococcus is a genus of mycolic-acid-containing actinomycetes that utilize a remarkable variety of organic compounds as growth substrates. This degradation helps maintain the global carbon cycle and has increasing applications ranging from the biodegradation of pollutants to the biocatalytic production of drugs and hormones. We have been using Rhodococcus jostii RHA1 as a model organism to understand the catabolic versatility of Rhodococcus and related bacteria. Our approach is exemplified by the discovery of a cluster of genes specifying the catabolism of cholesterol. This degradation proceeds via β-oxidative degradation of the side chain and O2-dependent cleavage of steroid ring A in a process similar to bacterial degradation of aromatic compounds. The pathway is widespread in Actinobacteria and is critical to the pathogenesis of Mycobacterium tuberculosis, arguably the world's most successful pathogen. The close similarity of some of these enzymes with biphenyl- and polychlorinated-biphenyl-degrading enzymes that we have characterized is facilitating inhibitor design. Our studies in RHA1 have also provided important insights into a number of novel metalloenzymes and their biosynthesis, such as acetonitrile hydratase (ANHase), a cobalt-containing enzyme with no significant sequence identity with characterized nitrile hydratases. Molecular genetic and biochemical studies have identified AnhE as a dimeric metallochaperone that delivers cobalt to ANHase, enabling its maturation in vivo. Other metalloenzymes we are characterizing include N-acetylmuramic acid hydroxylase, which catalyzes an unusual hydroxylation of the rhodococcal and mycobacterial peptidoglycan, and 2 RHA1 dye-decolorizing peroxidases. Using molecular genetic and biochemical approaches, we have demonstrated that one of these enzymes is involved in the degradation of lignin. Overall, our studies are providing fundamental insights into a range of catabolic processes that have a wide variety of applications.

No abstract