Comparative Community Proteomics Demonstrates the Unexpected Importance of Actinobacterial Glycoside Hydrolase Family 12 Protein for Crystalline Cellulose Hydrolysis

Sample collection and enrichment of thermophilic consortia.Compost samples were collected from a free municipal green waste composting program in Berkeley, CA (37°52′08.0″N 122°18′46.1″W), referred to here as Berkeley Green Waste (BGW) compost. The green waste consisted of yard trimmings and discarded food waste from an end-stage compost pile. Samples were transported to the lab at room temperature and stored at 4°C until inoculation. The adaptation of thermophilic consortia to purified substrates was described previously (36). Briefly, microcrystalline cellulose (0.5% wt/vol; Sigma, St. Louis, MO) was the sole supplemented carbon and energy source in 50 ml of M9 medium augmented with vitamins and buffered with 10 mM 2-(N-morpholino)ethanesulfonic acid (MES) at a final pH of 6.5 (37). Approximately 0.5 g of the BGW compost material was inoculated into the initial enrichments. Two biological replicates, referred to as A and B, were incubated in parallel at 60°C under aerobic conditions at 200 rpm. The enrichments were serially passed through three sets of liquid cultures (10% [vol/vol] inoculum), referred to as passages.

Measurement of protein concentration and glycoside hydrolase activity.At the end of each serial passage, DNA was isolated from cell pellets with the FastDNA spin kit for soil and the FastPrep instrument (MP Biomedicals, Santa Ana, CA). Protein concentrations were determined by bicinchoninic (BCA) assay (Pierce BCA protein assay kit; Thermo Scientific, Rockford, IL) methods, using a 96-well plate (200-µl reaction volume) with bovine serum albumin as the standard. Cellulase activity assays were conducted as previously described (38). Briefly, endoglucanase and xylanase activities were assessed by the 3,5-dinitrosalicylic acid (DNS) method, using carboxymethyl cellulose and birchwood xylan as the substrates, respectively, with either glucose or xylose as the standard (39). The enzyme reaction volume was 80 µl followed by 80 µl of DNS solution to measure released reducing sugars. One unit of cellulase or xylanase activity was defined as the amount of crude protein releasing a micromole of reducing sugar per minute per milliliter of supernatant volume. Cellobiohydrolase (pNPC), β-d-glucosidase (pNPG), and β-d-xylosidase (pNPX) activities were determined using their respective p-nitrophenyl sugar substrates. The p-nitrophenyl substrate (90 µl) was incubated with 10 µl of diluted enzyme, incubated for 30 min, and quenched with 50 µl of 2% cold sodium bicarbonate. The absorbance of released p-nitrophenol was measured at 410 nm. Activities using p-nitrophenyl substrates were calculated as micromoles of p-nitrophenol released per minute per milligram of crude protein.

Analysis of metagenomic data sets.Sequencing of the metagenomes was carried out by the Joint Genome Institute as previously described (40). The sequencing reads of four metagenomic data sets, including lineage A passage (2A), lineage A passage 3 (3A), lineage B passage 2 (2B), and lineage B passage 3 (3B), were coassembled using SPAdes 3.6 (41). The coassembled scaffolds were then binned using MaxBin 2.0 (40, 42). A customized Perl script was implemented to remove redundant genomic regions in the binned genomes. Briefly, this script will first self-BLAST each of the binned genomes using BLASTN, detecting distinct regions that were overlapped by >1,000 bp with identity of >95% and leave only one copy of the overlapped regions on the longest scaffold. To analyze the genetic content of the binned genome that was most closely related to Thermobispora bispora, genes were predicted from the recovered genome using Prodigal (43) and compared against the T. bispora DSM 43833 genome downloaded from NCBI (accession ID NC_014165.1) using BLASTX with an E value set to 1e−5 and max_target_seqs set to 1. Glycoside hydrolase (GH) genes were predicted by running HMMER3 (44) on the HMM files downloaded from the dbCAN server (45).

Saccharification with crude supernatants on cellulose.Saccharifications were performed in the presence of 2% (wt/vol) Avicel (Sigma, St. Louis, MO). Each mixture was prepared in 50 mM MES (pH 6.0) with 5 mg protein from crude supernatants per g glucan in biomass to a final volume of 5 ml in a 15-ml Falcon tube. Saccharifications were carried out 60°C in a shaker for 72 h. All hydrolysates were collected via centrifugation at 21,000 × g for 5 min and 0.45-µm-pore filtered to remove large biomass particles prior to sugar analysis. After filtration, samples were kept frozen at −20°C and thawed prior to analysis. Glucose concentrations were measured on an Agilent 1200 series high-performance liquid chromatography (HPLC) system equipped with an Aminex HPX-87H column (Bio-Rad, Hercules, CA) and refractive index detector. Samples were run with an isocratic 4 mM sulfuric acid mobile phase. Sugar concentrations were determined using standards containing both glucose and cellobiose.

Comparative proteomic analysis of thermophilic bacterial enrichment supernatants.Each supernatant was thawed and buffer exchanged through a 3,000 molecular weight cutoff (MWCO) spin column (Millipore, Temecula, CA) into 100 mM NH₄HCO₃ (pH 8). Each sample was then transferred into a fresh, labeled centrifuge tube, and a bicinchoninic acid assay (BCA assay) (Thermo Scientific, Rockford, IL) was performed to determine the protein concentration. Powdered urea was added to the supernatant at a concentration of 8 M along with 10 mM dithiothreitol (DTT). The samples were sonicated and incubated at 60°C for 30 min with constant shaking at 800 rpm. Samples were then diluted 8-fold for preparation for digestion with 100 mM NH₄HCO₃–1 mM CaCl₂, and sequencing-grade modified porcine trypsin (Promega, Madison, WI) was added to all protein samples at a 1:50 (wt/wt) trypsin/protein ratio for 3 h at 37°C. Digested samples were desalted using a 4-probe positive-pressure Gilson GX-274 ASPEC system (Gilson, Inc., Middleton, WI) with Discovery C₁₈ 100-mg/ml solid-phase extraction tubes (Supelco, St. Louis, MO) as previously described (46). The samples were concentrated down to ~30 µl using a Speed Vac, and a final BCA assay was performed to determine the peptide concentration.

The samples were measured and vialed to contain 50 µg each, and the volumes were brought up to 15 µl using 0.5 M tetraethylammonium bromide (TEAB; Sigma, St. Louis, MO) in a low-protein-binding 1.5-ml centrifuge tube. The pH of each sample was measured and brought to over pH 8 using 1 M TEAB. Each vial of 8-plex iTRAQ reagent (AB Sciex, Framingham, MA) was brought to room temperature. The reagents were pulse spun to ensure the contents were collected at the bottom, and 60 µl of isopropanol was added to each reagent vial. The reagents were thoroughly vortexed, spun down, and added to the appropriate sample. Only 4 out of the 8-plex iTRAQ channels were utilized for the current comparison, which included reporter ions 115, 117, 119, and 121, corresponding to supernatant sample passages 2B, 3B, 3A, and 2A, respectively. Samples were vortexed and spun down to incubate at room temperature for 2 h, at which time 100 µl of Nanopure water was added to hydrolyze the sample and incubated for an additional 30 min. The samples were partially dried down in a Speed Vac to remove the organic solvent and then pooled to obtain 1 sample containing all iTRAQ-labeled preparations, followed by C₁₈ extraction (as described above) and BCA assay to determine the final peptide mass for HPLC fractionation.

The sample was diluted to a volume of 900 µl with 10 mM ammonium formate buffer (pH 10.0), and resolved on an XBridge C₁₈ column (250 by 4.6 mm, 5 µM, with 4.6- by 20-mm guard column) (Waters, Milford, MA). Separations were performed at 0.5 ml/min using an Agilent 1100 series HPLC system (Agilent Technologies, Santa Clara, CA) as previously described (47). Each 24th fraction was combined for a total of 24 samples (each with n = 4 fractions pooled), each with 50% acetonitrile rinsing. The fractions were then completely dried down, and 25 µl of 25 mM ammonium bicarbonate was added to each fraction for storage at −20°C until liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis.

All iTRAQ-labeled fractions were analyzed by LC-MS/MS with the LC component utilizing an automated 65-cm by 75-mm-inside-diameter (i.d.) reversed-phase capillary column, HPLC system, and PAL autosampler as previously described (47). The MS component consisted of a Thermo Scientific LTQ-Orbitrap Velos mass spectrometer (Thermo Scientific, San Jose, CA) operated with settings previously described (47)

LC-MS/MS raw data were converted into dta files using Bioworks Cluster 3.2 (Thermo, Fisher Scientific, Cambridge, MA), and the MSGF+ algorithm (48) was used to search MS/MS spectra against genes predicted from all binned metagenomic data sets described above (108,923 entries). The search parameters used were previously described (47). A decoy database searching methodology was used to control the false discovery rate at the unique peptide level to ~0.1% (48). For quantification, peptide reporter ion intensities were captured across all channels and compared by calculating the summed protein intensity values across all fractions. Peptide/protein redundancy was maintained throughout.

Growth conditions and protein expression.Genes encoding the T. bispora hydrolytic cellulase sequences from enrichment cultures were synthesized by GenScript (Piscataway, NJ). Sequences contain a His tag terminus and were codon optimized for expression in E. coli, and the signal peptides were removed. Genes were provided in the entry vector pEOpt3 and were cloned using Golden Gate assembly (49) into E. coli DH10B. All reagents were purchased from New England Biolabs (Ipswich, MA). Briefly, the desired gene was digested out of the entry vector, ligated into a new destination vector, and transformed into an E. coli expression strain. Digestion was completed by incubating the gene plus entry vector (0.1 µg) with destination vector (50 ng pFil-B), 1× BSA, 1 mM ATP, 1× CutSmart buffer, and BsaI HP restriction enzyme with up to 10 µl of water at 37°C for 1 h. Next, T4 ligase (1 µl) was added to the digested product, and the mixture was incubated with 25 cycles of 37°C for 3 min and 16°C for 4 min, 50°C for 10 min, and 80°C for 10 min and then cooled to 4°C. The product was then transformed into chemically competent E. coli DH10B cells for storage and again into chemically competent E. coli NEB Express for heterologous expression of proteins. Starter cultures (50 ml) of E. coli NEB Express-harboring plasmids were grown overnight in LB medium containing 25 µg/ml kanamycin at 37°C and shaken at 200 rpm in rotary shakers. Expression was performed in Terrific Broth with 2% glycerol, 25 µg/ml kanamycin, and 2 mM MgSO₄. Starter cultures were used to inoculate 1 liter of expression medium in a 2-liter baffled Erlenmeyer flask and incubated at 18°C while shaking (200 rpm) and induced with 500 µM isopropyl-β-d-1-thiogalactopyranoside (IPTG). Following induction, cultures were again incubated at 18°C. At 22 h, cultures were divided and centrifuged at 15,500 × g for 30 min. Cell pellets were resuspended in 25 ml 50 mM HEPES plus 150 mM NaCl plus 20 mM imidazole (pH 7.4) and homogenized with an EmulsiFlex-C3 instrument (Avestein, Inc., Ontario, Canada). Lysates were collected via centrifugation at 75,000 × g for 30 min and 0.45-µm-pore filtered to remove large particles prior to purification. Polyhistidine-tagged proteins were purified on Ni-nitrilotriacetic acid (NTA) resin (Thermo Scientific, Rockford, IL) and stored at 4°C until ready for use. The proteins were >90% pure as visualized by SDS-PAGE.

Saccharification with purified proteins on biomass.Saccharifications using purified proteins were conducted on 2% (wt/vol) Avicel. Each mixture was prepared in 50 mM MES (pH 6.0) with 40 mg total protein per g glucan in biomass to a final volume of 100 µl in a 96-well microtiter plate. Saccharifications were carried out 60°C in a shaker for 72 h. Samples were harvested and analyzed by HPLC as described above.

NIMS assay.Time course reactions were set up by incubating 25 µl enzyme solution of the GH12, GH6_exo, and GH48 proteins (60 mg/g glucan) in a volume of 500 µl 100 mM NaOAc (pH 5.5) and 5 mg Avicel or phosphoric acid-swollen cellulose (PASC) (50). One Whatman no. 1 filter paper disc (GE Healthcare, Pittsburgh, PA), which was produced with a hole puncher and averaged ~3.5 mg, was added to each reaction mixture. The reaction mixtures were incubated at 60°C for 8 h, and 2-µl aliquots were removed at the indicated time points and analyzed by NIMS as previously described with ¹³C-labeled glucose as a standard (24). The processivity measurement was performed with Whatman no. 1 filter paper as previously described (51).

Accession number(s).The four metagenomes can be accessed at the JGI IMG website (http://img.jgi.doe.gov/) with IMG genome ID 3300001232 (passage 2A), 3300005157 (passage 3A), 3300000906 (passage 2B), and 3300005137 (passage 3B). The gene sequences and plasmid constructs for the T. bispora glycoside hydrolases GH12 (JPUB_007416), GH6_exo (JPUB_007418), GH48 (JPUB_007420), and GH6_endo (JPUB_007422) are available from the public version of the JBEI Registry (https://public-registry.jbei.org) and are physically available from the authors and/or Addgene (http://www.addgene.org) upon request. The mass spectrometry proteomics data have been deposited in the ProteomeXchange Consortium (52) (http://www.proteomexchange.org/) via the PRIDE partner repository with the data set identifier PXD004204.