Metagenomic scaffolds enable combinatorial lignin transformation

Abstract

Engineering the microbial transformation of lignocellulosic biomass is essential to developing modern biorefining processes that alleviate reliance on petroleum-derived energy and chemicals. Many current bioprocess streams depend on the genetic tractability of Escherichia coli with a primary emphasis on engineering cellulose/hemicellulose catabolism, small molecule production, and resistance to product inhibition. Conversely, bioprocess streams for lignin transformation remain embryonic, with relatively few environmental strains or enzymes implicated. Here we develop a biosensor responsive to monoaromatic lignin transformation products compatible with functional screening in E. coli. We use this biosensor to retrieve metagenomic scaffolds sourced from coal bed bacterial communities conferring an array of lignin transformation phenotypes that synergize in combination. Transposon mutagenesis and comparative sequence analysis of active clones identified genes encoding six functional classes mediating lignin transformation phenotypes that appear to be rearrayed in nature via horizontal gene transfer. Lignin transformation activity was then demonstrated for one of the predicted gene products encoding a multicopper oxidase to validate the screen. These results illuminate cellular and community-wide networks acting on aromatic polymers and expand the toolkit for engineering recombinant lignin transformation based on ecological design principles.

Keywords: environmental genomics; synthetic biology.

Fig. 1.

PemrR-GFP biosensor discovery and characterization. (A) Screening E. coli intragenic regions with monaromatic lignin transformation products including vanillin, vanillic acid, p-coumaric acid, vanillyl alcohol, and veratryl alcohol. (B) Relative reporter signal after incubation with 0.5 mM of select benzene derivatives for 2 h. Tree represents hierarchical clustering of the compound similarity using the single linkage algorithm. (C) Reporter sensitivity after 2 h. (D) Monitoring of in vitro lignin oxidation by DypB N246A in the presence of glucose oxidase, manganese, and hydrogen peroxide. Controls did not contain manganese. Error bars represent 95% confidence intervals (n = 3).

Fig. 2.

Profiling monoaromatic compounds by GC-MS. (A) Intensity of lignin-related monoaromatic compounds in culture supernatant as a ratio of the maximum. All values are normalized to a control strain harboring an empty fosmid and an internal standard. Clones were incubated with both SF-HKL and HP-L in minimal media. (B) Chromatogram showing production of vanillin on incubation of the clone 183_01_D18 with HP-L. The empty fosmid control and an authentic standard of vanillin are shown for the comparison. (C) Difference chromatogram showing CopA-catalyzed transformation of HP-L in the presence and absence of CuSO₄.

Fig. 3.

Genetic context maps for select active fosmids. Functional classes related to lignin degradation, Carbohydrate-active enzymes (CAZy) auxiliary enzymes, mobile genetic element associations, transposon mutagenesis insertions, and tRNAs are annotated. Transposon mutagenesis insertions that caused a statistically significant decrease in GFP signal were identified using a z-score ratio. The G+C ratio for every 200 nucleotides and gene abundance determined by mapping more than 500 million illumina reads sourced from the coal bed milieu is displayed as well. Connections represent protein homologs with minimum 50% identity and an E-value of 10E-20.