Chemicals and reagents
Veratryl alcohol (3,4-dimethoxybenzyl alcohol, A13396) and ABTS (2,2′-azino-bis[3-ethylbenzothiazoline-6-sulfonic acid] diammonium salt, J65535) were purchased from Alfa Aesar (Haverhill, MA, USA). β-O-4 dimer (1-[3,4,-dimethoxyphenyl]-2-[2-methoxyphenoxy]propane-1,3-diol, AK-40175) was purchased from Ark Pharm (Arlington Heights, IL, USA). DCIP (dichloroindophenol, D1878) was purchased from Sigma-Aldrich (St. Louis, MO, USA).
Glucose oxidase from Aspergillus niger (G2133) and lignin peroxidase from Phanerochaete chrysosporium (42603) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Peroxidase from horseradish (31941) was obtained from SERVA (Heidelberg, Germany).
Protein expression in S. cerevisiae
S. cerevisiae strain JHY69325 (gift of J. Horecka and A. Chu, Stanford Genomic Technology Center) was used as the background strain for all yeast protein expression. Ligninase genes were codon-optimized (Gen9) for expression in S. cerevisiae and synthesized de novo using DNA sequences coding for the mature enzymes sourced from UniProt and/or MycoCosm (Joint Genome Institute) databases. For single-copy expression vectors, a pRS415-based cassette (gift of C. Harvey, Stanford Genomic Technology Center) was used; for multi-copy expression vectors, the 2-micron cassette pCHINT2AL25 was used (gift of C. Harvey). Inducible expression was driven by the ADH2 promoter of S. cerevisiae, and ER targeting and secretion of ligninases were achieved by N-terminal fusion to an evolved variant of the α-mating-factor prepropeptide of S. cerevisiae48. Yeast transformation was carried out using the Frozen-EZ Yeast Transformation II Kit (Zymo Research). Transformant selection was performed using synthetic defined media plates deficient in leucine. Single colonies were picked into 0.5 ml SD-leu media in a 96-well culture plate and incubated overnight with orbital shaking (425 rpm, 30 °C). After centrifugation (600 × g, 10 min), the supernatant was removed, and the cell pellets resuspended in supplemented YPEG media (2% ethanol, 3% glycerol, 70 mM potassium phosphate pH 6.0; for heme peroxidase production, 0.01 mM hemin + 1 mM CaCl2; for cellobiose dehydrogenase production, 0.01 mM hemin; for laccase production, 2 mM CuSO4; no additional supplements for pyranose oxidase or aryl alcohol oxidase production) and incubated for 48 h with orbital shaking (425 rpm, 20 °C). After centrifugation (600 × g, 10 min), the culture supernatant was used for subsequent activity assays at 10% v/v.
Protein expression in N. benthamiana
Agrobacterium-mediated transient expression was performed as described before51. Genes encoding mature ligninases were cloned from S. cerevisiae vectors above into the pEAQ-HT expression cassette49. The signal peptide of the dirigent protein of Sinopodophyllum hexandrum (UniProt A0A059XIK7, residues 1–27) was used to direct protein export to the apoplast. Expression of ligninases was driven by a 35S promoter from cauliflower mosaic virus (CaMV) and a 5′UTR from cowpea mosaic virus (CPMV). Agrobacterium strains harboring the expression cassette and a p19-silencing plasmid were grown on dual-selecting kanamycin-gentamycin LB plates (30 °C, 2 days), incubated in induction buffer (150 μM acetosyringone, 10 mM MgCl2, 10 mM sodium succinate, pH 5.6; 4–5 h) before being infiltrated into the three youngest leaves of 5- to 7-week-old N. benthamiana plants at an OD600 of 0.3 in induction buffer. Four days post-infiltration, apoplastic contents were extracted as previously described with modifications28. Leaves were harvested and submerged in ice-cold extraction buffer (0.1 M sodium acetate, 0.3 M NaCl, pH 4.5); it was observed that 2-(N-morpholino)ethanesulfonic acid (MES) buffer has an inhibitory effect on peroxidase activity (Supplementary Fig. 16). Leaves were subjected to vacuum cycles (>650 mmHg, 3 × 5 min) to infiltrate the leaves with buffer. Leaves were individually placed on a piece of Parafilm and rolled around a pipette tip. This assembly was inserted into a 5-ml plunger-less syringe contained in a 15-ml conical centrifuge tube and centrifuged (1600 × g, 10 min, 4 °C) to produce the apoplastic extract, which was further clarified via centrifugation to remove any plant debris (21,000 × g, 10 min, 4 °C). In experiments featuring diafiltrated apoplast extracts, extracts were diafiltrated at least 600-fold with 20 mM sodium tartrate, pH 4.5, 10% v/v glycerol through Amicon Ultra-4 10-kDa MWCO centrifugal filters units (EMD Millipore). For sample storage, 1 volume of 40% glycerol was added to 3 volumes of extract and stored at −80 °C.
Enzyme activity testing
ABTS activity assays were performed using 4 mM ABTS, 100 μM H2O2, 50 mM sodium tartrate, pH 3.5. Assays for Mn-dependent ABTS oxidation were performed using 4 mM ABTS, 100 μM H2O2, 1.0 mM MnSO4, and 50 mM sodium malonate, pH 4.5. ABTS oxidation kinetics were observed at 414 nm (extinction coefficient 36,000 1/M 1/cm) using a Synergy HTX plate reader at 25 °C. Veratryl alcohol activity38 was measured as the production of veratraldehyde at 310 nm (extinction coefficient 9300 1/M 1/cm) using 20 mM veratryl alcohol, 100 μM H2O2, 50 mM sodium tartrate, pH 3.5, at 25 °C. Manganese-dependent activity39 was measured by Mn(III)-malonate complex formation using 1.0 mM MnSO4 and 100 μM H2O2 in 50 mM sodium malonate (270 nm, 11,590 1/M 1/cm) at 25 °C. Cellobiose dehydrogenase activity was measured at 522 nm using 10% w/v cellobiose, 0.3 mM dichloroindophenol, and 50 mM sodium tartrate, pH 5.0, at 25 °C. Pyranose oxidase activity was measured by coupling to ABTS as above with the inclusion of 1 μg commercial horseradish peroxidase (HRP) and 2% w/v D-glucose in 50 mM sodium acetate, pH 6.0. For all assays, 1 unit of activity is defined as 1 μmol of observable product per liter per minute, and activities are calculated as the maximum observed rate during the initial phase of the enzyme assays.
LC-MS kinetic analysis of dimer oxidation
All reactions contained 20 mM β-O-4 dimer and peroxidase-containing diafiltrated extract from N. benthamiana to 0.2 µM total heme content as determined by the pyridine hemochromagen assay52. Glucose oxidase assays contained 0.4% w/v D-glucose and either 1.0 ng/μl glucose oxidase and 50 mM sodium tartrate pH 3.5, or 0.574 ng/μl glucose oxidase and 50 mM sodium malonate pH 4.5 with 1.0 mM MnSO4, where the glucose oxidase concentration was adjusted between the two pH conditions to keep the rate of peroxide generation constant. Aryl alcohol oxidase assays contained 10 mM benzyl alcohol, 40 U/L (HRP-coupled ABTS activity) of diafiltrated extract of PE-aao(FX9) from N. benthamiana, and 50 mM sodium tartrate pH 4.0. Pyranose oxidase assays contained 0.4% w/v D-glucose, 10 U/L (HRP-coupled ABTS activity) of diafiltrated supernatant of TV-pox from S. cerevisiae, and 50 mM sodium tartrate pH 4.0. Reactions were clarified (21,000 × g, 5 min) and initiated by the addition of peroxide-generating enzyme.
Model lignin dimer LC-MS kinetic assays were performed using an Agilent 6545 UHPLC Q-TOF running in positive mode with a 6-min water-acetonitrile gradient (A: water + 0.1% formic acid, B: acetonitrile + 0.1% formic acid: 0 min, 95% A; 0.2 min, 95% A; 3.65 min, 37.5% A; 3.66 min, 5% A; 4.11 min, 5% A; 4.15 min, 95% A; 5.18 min, 95% A; flow rate 0.8 ml/min) on an Agilent RRHD EclipsePlus 95 Å C18 column (2.1 × 50 mm, 1.8 µm, 1200 bar). Reaction product profiles were measured every 24 min by 1 µl direct injection of reaction vials, which were maintained at 22 °C in the autosampler. Extracted ion counts (EIC) were obtained using the ‘Find by Formula’ function in Agilent MassHunter Qualitative Analysis software, using 35 ppm mass tolerance, 35, 500, and 35 ppm symmetric expansion of values for chromatogram extraction for C9H10O3 (veratraldehyde), C18H20O6 (dehydrodimer), and C11H14O4 (Hibbert ketone), respectively. Possible charge carriers and neutral losses were specified as -e−, +H, +Na, +K, +NH4, and −H2O.
LC-MS analysis of dimer cleavage extent by peroxidase isozymes
Reactions contained 20 mM β-O-4 dimer and 0.4% w/v D-glucose. Reactions assaying direct substrate oxidation contained 50 mM sodium tartrate, pH 3.5, and 1.0 ng/μl glucose oxidase; reactions assaying Mn-mediated substrate oxidation contained 50 mM sodium malonate, pH 4.5, 1 mM MnSO4, and 0.574 ng/μl glucose oxidase (adjusted to keep reaction rate similar). The amounts of diafiltrated extracts of PO-vp1 and PO-vp3 used in the reactions were normalized to the Mn(II) activity of PC-mnp1 at a reaction concentration of 0.2 µM (total heme content; ~6 U/L). The amounts of diafiltrated extracts of PE-vpl2 and CS-lip1 used in the reaction were normalized to PC-mnp1 by total heme content. Diafiltrated extract of GFP-expressing N. benthamiana was used as a negative control at 1% v/v (total heme content ~0.07 µM) with the addition of 33.3 ng/μl commercial horseradish peroxidase in order to prevent peroxide accumulation. Reactions were clarified (21,000 × g, 5 min) prior to initiation by addition of glucose oxidase or hydrogen peroxide. After 9-h incubation at room temperature, samples were moved to the LC-MS autosampler maintained at 10 °C and analyzed by 1 µl direct injection of the reaction contents on an Agilent 6545 Q-TOF running in positive mode with a 6-min water-acetonitrile gradient (as above) and an Agilent RRHD EclipsePlus 95 Å C18 column (2.1 × 50 mm, 1.8 µm, 1200 bar). EIC values were obtained as above.
Coupled condition optimization
Reactions contained 20 mM β-O-4 dimer, 50 mM sodium tartrate, pH 4.0, and 330 U/L ABTS activity of FPLC-purified53 PE-vpl2 heterologously produced in N. benthamiana. Coupled reactions additionally contained 0.4% w/v D-glucose. Absorbance corresponding to the formation of dehydrodimer and veratraldehyde was measured at 310 nm using a Synergy HTX plate reader and converted to an estimate of total aldehyde produced using the molar extinction coefficient for veratraldehyde (9300 1/M 1/cm). Reactions were initiated by the addition of peroxide or glucose oxidase.
After completion, 1 μl of the reaction was injected on a 6545 Agilent UHPLC Q-TOF running in positive mode with an 8-min water-acetonitrile gradient (0 min, 95% A; 0.2 min, 95% A; 5.65 min, 37.5% A; 5.66 min, 5% A; 6.11 min, 5% A; 6.15 min, 95% A; 7.18 min, 95% A; A: water + 0.1% formic acid, B: acetonitrile + 0.1% formic acid; flow rate 0.8 ml/min) on an Agilent RRHD EclipsePlus 95 Å C18 column (2.1 × 50 mm, 1.8 µm, 1200 bar). EIC values were obtained as above.
FPLC purification of PE-vpl2
Apoplast extracts from 54 leaves of N. benthamiana heterologously producing PE-vpl2 were prepared as above and dialyzed into 20 mM sodium succinate, pH 5.5, using 20 K MWCO 15-ml Slide Cassette G2 diafiltration cassettes. Protein was purified using a 1 ml HiTrap Q HP column (GE Healthcare Systems) running a gradient from 0 to 0.25 M sodium chloride at 1 ml min−1 (Supplementary Fig. 10c). Fractions with veratryl alcohol were pooled and concentrated for use in experiments.
Methylated DHP lignin preparation
Unmethylated DHP lignin was prepared as described previously29. 100 mg sinapyl alcohol and 17 mg coniferyl alcohol were dissolved in 2 ml acetone and diluted into 20 ml 10 mM sodium phosphate (pH 6.5) containing 60 purpurogallin units of peroxidase from horseradish (Serva) (solution 1). Separately, 55 μl 30% hydrogen peroxide was diluted into 20 ml 10 mM sodium phosphate pH 6.5 (solution 2). Solutions 1 and 2 were slowly added using a syringe pump to a round-bottom flask containing 1.5 mg vanillyl alcohol in 10 ml 10 mM sodium phosphate pH 6.5. Addition of the solutions was performed at 2 ml/h at room temperature while stirring under argon, shielded from light using aluminum foil. The reaction was allowed to proceed for a total of 24 h. The reaction products were centrifuged (3200 rpm, 30 min, 4 °C) and washed twice with 50 ml water before lyophilization.
Methylation was performed using trimethylsilyldiazomethane (TMSD) as previously described54. Twenty milligrams of DHP lignin was dissolved in 360 μl N,N-dimethylformamide in a 1.6 ml SafeLock Eppendorf microcentrifuge tube. Forty microliters of methanol, 24.26 μl N,N-diisopropylethylamine, and 64 μl TMSD (2.2 M in n-hexanes) were sequentially added. The reaction was performed at room temperature with rotation for 18 h. The reaction contents were transferred across 6 microcentrifuge tubes and each diluted 17.5-fold with water before centrifugation (21,000 × g, 10 min). The pellets were washed four times with 1 ml water before lyophilization.
DHP lignin depolymerization
Depolymerization reactions contained 0.66 μM FPLC-purified PE-vpl2, 200 μg/ml DHP lignin, 10 mM veratryl alcohol, and 0.25% v/v Tween-20 in 10 mM sodium acetate, pH 4.5. Enzyme was replaced with an equal volume of buffer (15 mM sodium succinate and 10% v/v glycerol, pH 5.6) for reactions without enzyme. One hundred micromolar hydrogen peroxide was added every 1.5 h for 6 additions in total. Reactions were performed with rotation in Protein Lo-Bind SafeLock microcentrifuge tubes (Eppendorf) at room temperature. Conditions were the same for reactions involving unmethylated and methylated DHP lignin.
Gel permeation chromatography
Depolymerization reactions were lyophilized and reconstituted in 125 μl N,N-dimethylformamide containing 0.1 M lithium bromide. Samples were clarified (21,000 × g, 10 min) and 60 μl injected on two PSS-PolarSil columns in series with N,N-dimethylformamide containing 0.1 M lithium bromide as the running solvent and a flow rate of 1 ml/min. Lignin elution was monitored using a diode-array detector at 310 nm.
Methoxybenzene activity testing
Reactions contained 0.4 mM methoxybenzene substrate, 0.66 μΜ FPLC-purified PE-vpl2 or commercial peroxidase from horseradish (Serva), and 0.2 mM hydrogen peroxide in 50 mM sodium tartrate, pH 3.5. Enzyme concentrations were normalized using spectrophotometric measurements of heme absorbance at 410 nm and extinction coefficients of 133 and 105 mM−1 cm−1 for PE-vpl2 and horseradish peroxidase, respectively. Reactions were carried out at room temperature for 1 h before injection of 5 μl on an Agilent 6520 Q-TOF LC-MS running in positive mode with an 25-min water-acetonitrile gradient (0 min, 97% A; 1 min, 97% A; 20 min, 3% A; 22 min, 3% A; 22.5 min, 97% A; 25 min, 97% A; A: water + 0.1% formic acid, B: acetonitrile + 0.1% formic acid; flow rate 0.4 ml/min) on an Agilent RRHD EclipsePlus 95 Å C18 column (2.1 × 50 mm, 1.8 µm, 1200 bar). For all substrates except 1,4-dimethoxybenzene, substrate conversion was measured as difference in MS peak area corresponding to the methoxybenzene substrate between enzyme and no-enzyme conditions. For 1,4-dimethoxybenzene, substrate conversion was measured using UV peak area at 270 nm. In experiments involving 1,2,3,5-tetramethoxybenzene, substrate conversion extent was found to be dependent on the concentration of horseradish peroxidase used. Given this sensitivity to enzyme concentration, our data should not be used to estimate the absolute redox potential of either enzyme assayed here. Redox potentials of methoxybenzene substrates used here were previously determined in non-aqueous conditions as polarographic half-wave potentials using a rotating platinum electrode50.
Statistics and reproducibility
All in vitro enzyme reactions were separately assembled and performed in triplicate. For protein production in N. benthamiana, three leaves on three separately grown plants were chosen at random as biological replicates. Plants were selected from the same batch planted on the same date. Spectroscopic activity measurements were performed as technical triplicates for each individual leaf in N. benthamiana and activity levels were calculated as the average of the three individual leaves. GFP-expressing leaf controls were included in each round of protein production in N. benthamiana to control for variability between plant batches. For protein production in S. cerevisiae, three individual colonies were selected at random from each vector transformation as biological replicates and were used as such for spectroscopic activity measurements. All data points represent the arithmetic mean of three independent replicates and error bars represent one standard deviation. A two-tailed t-test with a p-value of 0.05 was used to determine statistical significance relative to GFP controls in activity level measurements after exclusion of outliers.
Reporting summary
Further information on research design is available in the Nature Research Reporting Summary linked to this article.
