Biochemical Characterization of a Family 15 Carbohydrate Esterase from a Bacterial Marine Arctic Metagenome

Abstract

Background: The glucuronoyl esterase enzymes of wood-degrading fungi (Carbohydrate Esterase family 15; CE15) form part of the hemicellulolytic and cellulolytic enzyme systems that break down plant biomass, and have possible applications in biotechnology. Homologous enzymes are predicted in the genomes of several bacteria, however these have been much less studied than their fungal counterparts. Here we describe the recombinant production and biochemical characterization of a bacterial CE15 enzyme denoted MZ0003, which was identified by in silico screening of a prokaryotic metagenome library derived from marine Arctic sediment. MZ0003 has high similarity to several uncharacterized gene products of polysaccharide-degrading bacterial species, and phylogenetic analysis indicates a deep evolutionary split between these CE15s and fungal homologs.

Results: MZ0003 appears to differ from previously-studied CE15s in some aspects. Some glucuronoyl esterase activity could be measured by qualitative thin-layer chromatography which confirms its assignment as a CE15, however MZ0003 can also hydrolyze a range of other esters, including p-nitrophenyl acetate, which is not acted upon by some fungal homologs. The structure of MZ0003 also appears to differ as it is predicted to have several large loop regions that are absent in previously studied CE15s, and a combination of homology-based modelling and site-directed mutagenesis indicate its catalytic residues deviate from the conserved Ser-His-Glu triad of many fungal CE15s. Taken together, these results indicate that potentially unexplored diversity exists among bacterial CE15s, and this may be accessed by investigation of the microbial metagenome. The combination of low activity on typical glucuronoyl esterase substrates, and the lack of glucuronic acid esters in the marine environment suggest that the physiological substrate of MZ0003 and its homologs is likely to be different from that of related fungal enzymes.

Fig 1. Phylogenetic tree and structural organisation of CE15s.

A) Maximum Likelihood tree of CE15s from fungi (red text) and bacteria (blue text). A dashed line is used to highlight the deep-branching split between the two major clades, with ‘Clade F’ referring to previously-characterised fungal CE15s and their homologs, and ‘Clade B’ to MZ0003 and related bacterial sequences. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. Bootstrap values above 50% are shown. Evolutionary analyses were conducted in MEGA6. Full species name and gene identifiers of sequences are given in S1B Table) Schematic of the alignment of fungal and bacterial CE15 structures. Secondary structural elements from the M. thermophilia StGE2 sequence are shown with β-strands indicated by arrows, α-helices by oblongs and turns denoted by ‘TT’. Nomenclature for the StGE2 structure is given above, and the prediction for the MZ0003 structure below. Expanded areas 1–3 show the position of loops predicted in the MZ0003 model (discussed below) which are found in related bacterial homologs of ‘Clade B’. Bacterial sequences are indicated by the blue box, and putative loops are shaded yellow. Regions surrounding active site residues are also expanded, with fully conserved residues shaded red, and functionally conserved residues shown in red text. Confirmed active site residues of StGE2 are indicated by red arrows above the sequence, while catalytic residues identified by mutagenesis in MZ0003 are indicated in green on the lower sequence. Residues which were mutated without effect in MZ0003 are indicated with grey arrows. Details of the full sequence alignment are given in S1 Fig.

Fig 2. Specific activity of MZ0003 measured by TLC.

A) Substrates used to detect glucuronoyl esterase activity by TLC. B) De-esterification of methyl 4-O-methyl-D-glucopyranuronate. MZ0003 concentrations used in the assay are: (A) 0 μg, (B) 5 μg, (C) 10 μg, (D) 15 μg, (E) 25 μg, (F) 50 μg in a final volume of 100 μl. C) De-esterification of allyl D-glucuronate (i), benzyl D-glucuronate (ii) and D-glucuronic acid methyl ester (iii) by 50 μg MZ0003 (+), or incubated in the absence of enzyme (-). Activity was measured after 90 minutes at 25°C for all substrates.

Fig 3

Temperature dependence of activity and stability of MZ0003 A) Temperature optimum: activity represents the initial rate of the reaction at each temperature in the presence of 1 M NaCl (upper plot) and no added salt (lower plot), normalised to the highest rate, which was recorded at 35°C in 1 M NaCl. B) Temperature stability: residual activity represents the initial rate of the reaction measured at 35°C after incubation for various lengths of time at the temperatures shown. Values are normalised to the rate recorded without pre-incubation. All data points for assays represent the mean of three independent experiments, and error is given as the standard deviation. C) Thermal unfolding measured by DSC: representative thermogram with buffer subtracted and a sigmoidal baseline fitted.

Fig 4. The effect of salt and pH on MZ0003.

A) Thermal stability (blue dashed line) and activity (red solid line) with varying NaCl concentrations. Stability was determined by DSF in 50 mM HEPES pH 8.0, activity was tested in 0.1 M Tris-HCl pH 8.0 and defined as the percentage relative to the maximum enzyme activity B) Effect of pH on MZ0003 tested at 30°C: MES was used for pH 5.0–6.0, Na-phosphate buffer from pH 6.0–7.5, Tris-HCl buffer for pH 7.5–9.5 and CAPS buffer for pH 9.5–10.5. All data points for assays represent the mean of three separate experiments.

Fig 5. Effect of various metal ions, common inhibitors and denaturants on MZ0003 activity.

A) Activity tested with addition of KCl, LiCl, AgCl₂, NaF, ZnCl₂, CaCl₂, MgCl₂, CoCl2, SnCl₂, CuCl₂, MnCl₂ and CdC_l2 at concentrations of 1 and 10 mM. B) Activity tested with: EDTA, Urea, 2-Mercaptoethanol (2-MeOH), SDS, PMSF and DTT at concentrations of 1 and 10 mM. MZ0003 was incubated with each chemical for 1 hour in 0.1 M TRIS HCl pH 8.0 on ice and then the activity was tested with p-NP acetate at 35°C. All data are the average of three independent experiments, error is the standard deviation.

Fig 6. Structural model of MZ0003 and mutagenesis.

A) Overall structure of MZ0003: α-helix of regions that aligned with the M. thermophilia StGE2 sequence are shown in cyan, β-strands in violet. Large regions, which could not be aligned based on sequence, are shown in yellow. B) Predicted active site of MZ0003 studied by site directed mutagenesis: residues in equivalent positions to the catalytic triad of StGE2 are shown as red sticks. Nearby polar residues in the predicted loop are shown as yellow sticks. C) Esterase activity of mutants on methyl 4-O-methyl-D-glucopyranosyluronate at 25°C measured after 90 minutes by TLC. Activity seen as the appearance of the lower product spot (indicated by P) and the decrease in intensity of the upper substrate spot (S).

Fig 7. Organisation of genes surrounding the CE15 homolog which have functions that may be linked to carbohydrate degradation.

Homologs of MZ0003 are coloured red, genes annotated as carbohydrate degrading enzymes are coloured blue, transporters are coloured yellow and regulatory proteins green. Hypothetical proteins and those with unrelated functions are shown in grey. The figure was produced manually based on graphics from the Artemis genome viewer.