Aromatic Dimer Dehydrogenases from Novosphingobium aromaticivorans Reduce Monoaromatic Diketones

Abstract

Lignin is a potential source of valuable chemicals, but its chemical depolymerization results in a heterogeneous mixture of aromatics and other products. Microbes could valorize depolymerized lignin by converting multiple substrates into one or a small number of products. In this study, we describe the ability of Novosphingobium aromaticivorans to metabolize 1-(4-hydroxy-3-methoxyphenyl)propane-1,2-dione (G-diketone), an aromatic Hibbert diketone that is produced during formic acid-catalyzed lignin depolymerization. By assaying genome-wide transcript levels from N. aromaticivorans during growth on G-diketone and other chemically-related aromatics, we hypothesized that the Lig dehydrogenases, previously characterized as oxidizing β-O-4 linkages in aromatic dimers, were involved in G-diketone metabolism by N. aromaticivorans. Using purified N. aromaticivorans Lig dehydrogenases, we found that LigL, LigN, and LigD each reduced the Cα ketone of G-diketone in vitro but with different substrate specificities and rates. Furthermore, LigL, but not LigN or LigD, also reduced the Cα ketone of 2-hydroxy-1-(4-hydroxy-3-methoxyphenyl)propan-1-one (GP-1) in vitro, a derivative of G-diketone with the Cβ ketone reduced, when GP-1 was provided as a substrate. The newly identified activity of these Lig dehydrogenases expands the potential range of substrates utilized by N. aromaticivorans beyond what has been previously recognized. This is beneficial both for metabolizing a wide range of natural and non-native depolymerized lignin substrates and for engineering microbes and enzymes that are active with a broader range of aromatic compounds. IMPORTANCE Lignin is a major plant polymer composed of aromatic units that have value as chemicals. However, the structure and composition of lignin have made it difficult to use this polymer as a renewable source of industrial chemicals. Bacteria like Novosphingobium aromaticivorans have the potential to make chemicals from lignin not only because of their natural ability to metabolize a variety of aromatics but also because there are established protocols to engineer N. aromaticivorans strains to funnel lignin-derived aromatics into valuable products. In this work, we report a newly discovered activity of previously characterized dehydrogenase enzymes with a chemically modified by-product of lignin depolymerization. We propose that the activity of N. aromaticivorans enzymes with both native lignin aromatics and those produced by chemical depolymerization will expand opportunities for producing industrial chemicals from the heterogenous components of this abundant plant polymer.

Keywords: Lignin; Novosphingobium; aromatic dehydrogenases; aromatic metabolism; ketone reduction; sphingomonads.

FIG 1

Growth and metabolism of N. aromaticivorans DSM12444, 12444ΔSacB strain during growth on G-diketone and glucose. Triplicate cultures were grown in SMB with 0.5 g/liter glucose and 0.418 g/liter G-diketone. In each panel, error bars represent the standard deviation. (A) Increases in N. aromaticivorans cell density as monitored by Klett colorimeter units. (B) Extracellular concentrations of G-diketone, GP-1, and threo-GD identified and quantified via HPLC-MS and HPLC-UV (Fig. S1, see text).

FIG 2

Changes in transcript abundance for indicated genes when cells are grown in the presence of G-type aromatics. Each plot displays the log₂ fold change in reads per kilobase million (RPKM) compared with glucose-grown N. aromaticivorans cells showing genes identified as encoding enzymes involved in aromatic metabolism. Black stars above a transcript with a significant change in levels (*q < 0.05, **q < 0.01, ***q < 0.001) compared with cells grown in the presence of glucose alone. Bars in each panel are colored to denote steps in aromatic metabolism that gene products are known to function (dimer degradation, aromatic ring processing, side chain processing/demethylation).

FIG 3

Time-dependent loss of G-diketone and GP-1 in vitro when incubated with recombinant LigL, LigN, and LigD, with and without NADH. 0.5 mmol/liter of G-diketone (A) or GP-1 (B) was incubated with each in enzyme with and without 2 mmol/liter NADH. Addition of NADH initiated the reaction. Samples were incubated in the dark at 30°C. Concentrations of G-diketone and GP-1 was measured using HPLC-MS.

FIG 4

GC-MS analysis of derivatized aromatic substrates and in vitro reaction products of individual LigLND dehydrogenases with G-diketone and GP-1. GC-MS analysis of derivatized aromatic substrates and enzyme reaction products after indicated Lig dehydrogenases were incubated for 24 h at 30°C with the G-diketone and NADH (A, C) or GP-1 and NADH (B, D).

FIG 5

Kinetic parameters of LigL, LigN and LigD dehydrogenases with indicated aromatic substrates. Shown are the measured K_cat and apparent K_m using recombinant LigL, LigN, and LigD enzymes with the indicated aromatic substrates and either NADH (G-diketone) or NAD⁺ (GGE, GD) as a cofactor.

FIG 6

Model for G-diketone metabolism by N. aromaticivorans. We hypothesize that the indicated Lig dehydrogenases initiate degradation of G-diketone, reducing the Cα ketone to GP-2. GP-2 and GP-1, as Hibberts ketones, can spontaneously interconvert; the question marks indicate that we cannot rule out the existence of enzymes that produce GP-1. The figure also indicates that the LigL dehydrogenase reduced GP-2 to 1-(4-hydroxy-3-methoxyphenyl)propane-1,2-diol (GD). In this model, one or more unknown enzymes, indicated by the question mark, are used to produce vanillin from GP-1 or GD.