Cell-Free Translation of Biofuel Enzymes

Abstract

In nature, bacteria and fungi are able to utilize recalcitrant plant materials by secreting a diverse set of enzymes. While genomic sequencing efforts offer exhaustive lists of genes annotated as potential polysaccharide-degrading enzymes, biochemical and functional characterizations of the encoded proteins are still needed to realize the full potential of this natural genomic diversity. This chapter outlines an application of wheat germ cell-free translation to the study of biofuel enzymes using genes from Clostridium thermocellum, a model cellulolytic organism. Since wheat germ extract lacks enzymatic activities that can hydrolyze insoluble polysaccharide substrates and is likewise devoid of enzymes that consume the soluble sugar products, the cell-free translation reactions provide a clean background for production and study of the reactions of biofuel enzymes. Examples of assays performed with individual enzymes or with small sets of enzymes obtained directly from cell-free translation are provided.

1 Introduction

The availability of cheap energy has been a major factor supporting economic development worldwide during the past century [1, 2]. Among the many available energy sources, liquid fuels, which are primarily used for transportation, are the most valuable but are also increasingly problematic, with major perturbations in their supply and price occurring in the mid-1980s and again after 2000. Given this recurring state of affairs, efforts to secure new sources of liquid fuels have intensified worldwide, with ongoing research in academic, industrial, and government settings.

The Great Lakes Bioenergy Research Center is a multi-institution collaborative effort funded by the US Department of Energy, with a mission to perform basic research needed to generate and improve technology to convert cellulosic biomass to ethanol and advanced biofuels. Methods described in this chapter have been developed and carried out in the Great Lakes Bioenergy Research Center in order to evaluate the genomic contents of new isolates from highly cellulolytic environmental niches and to better understand the capabilities of model cellulolytic organisms. This approach is also applicable to engineering of multi-protein mixtures for improved deconstruction of biomass.

1.1 Overview of Biomass Deconstruction Needed for Biofuel Production

The complexities of biomass, particularly the diversity of cellulosic and hemicellulosic substructures that make up a large portion of plants, make the economic utilization of cellulosic biomass a difficult challenge. Currently, biomass is chemically pretreated to achieve a partial decomposition or extraction of lignin, which exposes the entrained polysaccharide substructures and also alters its physical state [3]. After chemical pretreatment, the biomass is subjected to enzymatic hydrolysis to produce soluble sugar solutions suitable for fermentation by microbes into biofuels. Because of its complex structure, many different enzymes are necessary for efficient, high yield deconstruction of biomass [4].

1.2 Enzymes Needed to Deconstruct Cellulose

Natural cellulolytic organisms are known to deconstruct biomass materials into simple sugars [5]. To better understand this natural capability, the DOE Joint Genome Institute and other researchers have sequenced over 5,000 natural organisms and metagenomes, including an increasing number of proven cellulolytic microbes and fungi. Bioinformatics approaches have sorted the protein-coding sequences of these genomes into many putative cellulolytic functions. As one leading example, the Carbohydrate-Active Enzymes database (CAZy, http://www.cazy.org [6]) provides an intensively curated list that is separated into glycoside hydrolase (GH), glycoside transferase (GT), carbohydrate esterase (CE), polysaccharide lyase (PL), and carbohydrate-binding module (CBM) families.

Efficient hydrolysis of pure cellulose requires concerted activities provided by several different members of the GH family [7]. Endocellulases or endoglucanases attack crystalline regions in the interior of cellulose molecules, while exoglucanases or cellobiohydrolases degrade cellulose from either the reducing or nonreducing ends of a glucan strand to produce cellobiose. Cellobiosidases then hydrolyze cellobiose to glucose. Both endo- and exocellulases have associated CBM domains, which increase the efficiency of binding the catalytic GH domain to the insoluble polysaccharide. Related cascades of enzymes react with polysaccharides such as xylan, mannan, and pectin, which collectively represent the hemicellulose fractions. Deconstruction of biomass is potentiated when enzymes capable of reaction with the cellulose and hemicellulose fractions are simultaneously present [8].

Among cellulolytic microorganisms, Clostridium thermocellum, an anaerobic, thermophilic, Gram-positive bacterium, has demonstrated one of the fastest growth rates on crystalline cellulose [9]. Since it can also ferment sugars to ethanol, C. thermocellum has become an important model organism for biofuel research. C. thermocellum secretes dozens of cellulases, xylanases, mannanases, and other polysaccharide-degrading enzymes that allow the efficient conversion of polysaccharides into soluble oligosaccharides. Biochemical [10, 11], transcriptomic [12], and proteomic [13] studies have helped to establish the identities of these enzymes, and many have been purified and studied for their catalytic function and 3-dimensional structures. Additionally, C. thermocellum produces a unique extracellular assembly termed the cellulosome, a multienzyme complex in which many different enzymes are recruited to a protein called scaffoldin through tight-binding interactions of dockerin and cohesin domains [14]. The formation of this large complex is believed to provide high efficiency in polysaccharide degradation by increasing local enzyme concentration. However, because the scaffoldin has a repetitive multi-domain structure of −200 kDa, obtaining pure preparations of it through heterologous expression has been a significant challenge. Consequently, protein engineering has been used to develop more experimentally tractable substitutes for the scaffoldin and various assemblies of enzymes from C. thermocellum [15, 16].

1.3 Overview of Experimental Strategy and Method

In this chapter, we describe an approach for producing engineered fusions of GH families with CBM domains. Figures 1 and 2 show an overview of the plasmid constructs and cloning methods that were applied to genes from C. thermocellum to produce forms of the enzymes that efficiently react with insoluble substrates without a need for the scaffoldin. These general principles are applicable to the study of the catalytic and polysaccharide binding domains from other organisms.

Fig. 1.

Maps of two cloning vectors used in this study. The top portion shows the nucleotide sequence of the SP6 promoter, TMV-omega sequence, the start codon (ATG) and sequence encoding a His6 tag, and the Sgfl restriction site. Circular maps of pEU_HSBC and pEU_HSBC_CBM3a are shown below. In both plasmids, ampiciilin resistance is used as the transformation selection marker. During cloning, the sacB-CAT cassette is replaced by the gene of interest, which can permit positive selection by growth on agar medium containing sucrose if desired. In pEU_HSBC_CBM3a, the 40-amino acid linker and CBM3a coding sequence is placed in frame at the 3′ end of sacB-CAT cassette so that the target protein is expressed as a fusion with the additional linker and CBM3a domain

Fig. 2.

Design of primer pairs and a schematic of the cloning method, (a) The Cthe_0032 (ManA) gene is used as an example. The PFAM image of the domain structure [39] shows the predicted amino acid residue positions for the signal peptide (residues 1–32), the GH26 domain (residues 180–506, green), and the two dockerin domains (red). The forward primer (b) and reverse primer design (c) for PCR needed for two-step Flexi Vector transfer are shown with nucleotide sequences and corresponding amino acid sequences

Identification of candidate genes includes bioinformatics evaluation and, if available, correlation with experimental results such as transcriptomics or proteomics. Since Flexi Vector cloning uses SgfI and PmeI restriction sites [17], candidate genes containing these restriction sites will require mutagenesis (or gene synthesis) to eliminate them. Fortunately, most genomes contain a low frequency of these restriction sites [17]. Furthermore, since most biofuel enzymes are secreted from the natural host, they contain hydrophobic signal sequences at the N-terminus that should be removed in the design for cell-free translation. This simple excision can dramatically improve the solubility of translated proteins. Bioinformatics tools such as SignalP (http://www.cbs.dtu.dk/services/SignalP/) are useful to identify possible positions for truncation of the sequence encoding a signal peptide from the native gene. In this work, we also removed nucleotide sequences encoding the C-terminal dockerin domains in order to facilitate our studies of catalytic potential outside of the cellulosomal format.

Endo, Sawasaki, and colleagues at Ehime University, Matsuyama, Japan, created an optimized wheat germ cell-free translation plasmid, pEU [18]. For the protocols presented here, the parental plasmid was modified by the UW Center for Eukaryotic Structural Genomics to create pEU-HSBC (Fig. 1). This Flexi Vector cloning-compatible plasmid contains an SP6 promoter at the 5′ end of the critical transcription/translation region, followed by the tobacco mosaic virus translational enhancer (TMV-omega) sequence [19], and then nucleotide sequences encoding a start codon and Gly residue, a His6 purification tag, and an AIA sequence corresponding to the Flexi Vector SgfI restriction site [17]. The empty vector contains a toxic cassette consisting of the SacB gene [20] and chloramphenicol acetyl transferase genes. At the 3′ end, pEU-HSBC provides a stop codon by use of the Flexi Vector PmeI site, which thus yields a candidate protein with the N-terminal tag MGHHHHHHAIA-provided by the vector and the remaining coding sequence produced by the PCR primer design and targeted gene amplification.

Another vector created by the Great Lakes Bioenergy Research Center, pEU-HSBC-CBM3a, creates a C-terminal fusion to a 40 amino acid linker and CBM3a domain from the C. thermocellum scaffoldin protein CipA (Cthe_3077). Upon translation, candidate genes cloned into pEU-HSBC-CBM3a yield an unnatural fusion protein with an N-terminal tag MGHHHHHHAIA–, the candidate domains, and then a fusion to CBM3a, which is known to bind to crystalline cellulose. Genes cloned according to the Flexi Vector method are easily transferred between different Flexi Vector plasmids [17], and the presence of a 3′ pF1K homology sequence increases the transfer efficiency by inhibiting self-ligation of the two plasmid vector backbones during serial transfer of genes.

Primer design has an important role in this effort. By consideration of the domain structure of a gene of interest, domains of interest can be excised from the complete gene to assemble better performing constructs, to test specific aspects of the catalytic capabilities of multi-domain enzymes, or to obtain engineered assemblies. Figure 2 schematizes this primer design process as applied to Cthe_0032, a GH26 protein. In this work, the catalytic domain of the protein was cloned, expressed, and assayed. By use of the 2-stage PCR method, it is possible to build additional functionality at the 5′ and 3′ regions of the cloned gene. For example, inclusion of the coding sequence for a protease recognition site immediately after the SgfI site but before the coding sequence of the gene of interest is useful for removal of N-terminal tags from expressed proteins for structure determination studies [21]. Alternatively, omission of a stop codon from the coding sequence of the gene of interest can be used to create C-terminal fusion proteins such as some of the examples provided here (Table 3, CBM3a fusions).

Table 3.

Summary of assay results obtained for individual C. thermocellum enzymes

Gene locus	Gene name	Domains produced in translation reaction	His6 addeda	CBM3a addedb	Uniprot ID	FP	PASC	Xylan	Mannan	MUG	MUC	MUX	MUM	Measured function (s)
Cthe_0032		GH26			A3DBE4	−	−	−	+	−	−	−	−	Mannanase
Cthe_0040	celI	GH9 CBM3			Q02934	+	+	−	−	−	−	−	−	Endocellulase
Cthe_0212	bglA	GH1			P26208	−	−	−	−	+	−	−	−	β-Glucosidase
Cthe_0269	celA	GH8		Yes	A3DC29	+	+	−	−	−	−	−	−	Endocellulase
Cthe_0270	chiA	GH18			A3DC30	−	−	−	−	−	−	−	−	None detected
Cthe_0405	celL	GH5		Yes	A3DCG4	+	+	−	−	−	+	−	−	Exocellulase
Cthe_0412	celK	CBM_4_9 CelDN GH9		Yes	A3DCH1	−	+	+	−	+	+	−	−	Exocellulase
Cthe_0536	celB	GH5	Yes		P04956	+	+	−	−	−	+	−	−	Exocellulase
Cthe_0543	celF	GH9 CBM3	Yes		P26224	+	+	−	_	−	−	−	−	Endocellulase
Cthe_0578	celR	GH9 CBM3		Yes	A3DCY5	+	−	−	−	+	−	−	−	Endocellulase
Cthe_0625	celQ	GH9 CBM3	Yes		A3DD31	+	+	−	−	−	−	−	−	Endocellulase
Cthe_0797	celE	GH5 Lipase_GDSL	Yes		A3DDK3	+	+	+	+		+	−	−	Exocellulase, mannanase, xylanase
Cthe_0912	xynY	CBM_4_9 GH10 CBM_4_9			A3DDW7	−	−	+	−	−	−	−	−	Xylanase
Cthe_1256	bglB	GH3	Yes		P14002	−	−	−	−	+	+	+	−	β-Glucosidase, β-xylosidase
Cthe_1838	xynC	CBM_4_9 GH10		Yes	A3DGI0	−	−	+	−	−	−	−	−	Xylanase
Cthe_2872	celG	GH5			Q05332	+	+	−	−	−	+	−	−	Exocellulase

^aProteins produced using pEU-His-FV, which adds an N-terminal MGHHHHHHAIA tag

^bProtein produced using pEU-His-FV-CBM3a, which gives a native N-terminus and adds fusion of a 40-amino acid linker followed by the CBM3a domain from Cthe_3077 to the C-terminus

Plasmid DNA purified according to the protocols given is the essential substrate for a transcription reaction, which provides an mRNA that becomes the substrate for cell-free translation. Both the transcription and translation reactions are carried out by robots [22], and the individual translation reactions are then suitable for assay with a wide variety of biofuel relevant substrates, can be combined into higher-order assemblies of enzymes, or can be subjected to additional purification if desired.

1.4 Protein Functional Assays

During our initial work on cell-free translation of biofuel enzymes, we discovered that wheat germ extract had no endogenous capability for hydrolysis of the insoluble substrates cellulose, xylan, or mannan and likewise no competing enzymatic reactions that would consume soluble sugars or contain other products such as NADH that would be produced from coupled enzyme assays. Moreover, a low-level soluble glucosidase activity present in the wheat germ extract was easily inactivated at temperatures higher than 45°C. Thus many proteins translated in wheat germ extract can be assayed for their various cellulolytic activities without a need to obtain purified preparations of target polypeptides.

Table 1 summarizes the substrates used in this work. Assays with small molecule fluorogenic analogs, purified polysaccharides, and biomass materials have all been carried out with enzymes expressed in cell-free translation. In many cases, assays of single enzymes are inadequate for some discovery purposes, particularly when synergistic enzyme activities are required to disassemble the complicated ultrastructure of biomass. With robotic cell-free translation, new samples of enzymes can be rapidly obtained for combinatorial tests of synergistic activity or as replacements for individual enzymes in mixtures with established reactivity. We briefly introduce examples of this capability in the chapter.

Table 1.

Summary of the substrates used in enzyme assays

Substrate (abbreviation)	Assay type	Activity detected	Source
4-Methylumbelliferyl-β-D-glucopyranoside (MUG)	Fluorescence	β-Glucosidase	Sigma-Aldrich (M3633)
4-Methylumbelliferyl-β-D-cellobioside (MUC)	Fluorescence	Cellobiohydrolase	Sigma-Aldrich (M6018)
4-Methyumbelliferyl-β-D-mannopyranoside (MUM)	Fluorescence	β-Mannosidase	Sigma-Aldrich (M0905)
4-Methylumbelliferyl-β-D-xylopyranoside (MUX)	Fluorescence	β-Xylosidase	Sigma-Aldrich (M7008)
Filter paper (FP)	DNS	Cellulase	GE Healthcare (Whatman 10001-185)
Phosphoric acid-swollen cellulose (PASC)	DNS	Cellulase	Prepared as described in Subheading 3.6; also see reference [38]
Birch xylan (xylan)	DNS	Xylanase	Sigma-Aldrich (X0502)
1, 4-β-D-Mannan (mannan)	DNS	Mannanase	Megazyme (90302b)
AFEX-treated switchgrass (AFEX-SG)	DNS	Biomass deconstruction	Great Lakes Bioenergy Research Center
Ionic liquid-treated switchgrass (IL-SG)	DNS	Biomass deconstruction	Joint Bioenergy Institute

The methylumbelliferyl glycosides are small molecule fluorogenic analogs used in enzyme screening to deduce the potential reactivity of biofuel enzymes with oligomeric substrates. Hydrolysis of the glycosidic bond releases the fluorophore 4-methylumbelliferone [23, 24]. The high sensitivity of the fluorescence detection, the short reaction time, and the small amounts of enzyme needed contribute to a broad glycosidase diagnostic utility of the small molecule analogs.

Insoluble substrates are most realistic for biofuel enzymology but also provide increasing experimental challenges [25]. While individual enzymes can often be reliably assayed using purified polysaccharides, studies of biomass most often require the combination of chemical pretreatment and mixtures of synergistic enzymes to provide sufficient reactivity. The determination of enzyme function with insoluble polysaccharides can be accomplished by a variety of methods [26]. While samples from cell-free translation are acceptable for most of these, we focus here on use of the dinitrosalicylic acid (DNS) assay [27] because of its broad applicability to a variety of products, relatively simple methodology, sufficient sensitivity, and low cost. When an enzyme hydrolyzes a polysaccharide, the reducing-end sugar product can react with the DNS reagent, producing a colored adduct with an absorbance maximum of ~540 nm. The color formation permits the use of optical spectroscopy for detection, and the method is well suited for high throughput because the sample handling process can be reduced to relatively few liquid handling steps. In addition, we have adapted a simple protein pull-down assay to assess the function of CBM domains [28] and show examples of this qualitative evaluation using pure cellulose as a binding substrate.

2 Materials

2.1 Equipment

1. The Protemist DT II benchtop cell-free translation robot (CellFree Sciences Co., Ltd., Yokohama, Japan) is used for preparative protein production. This robot uses a bilayer cell-free protein synthesis method [29],

2. The Protemist DT II software (CellFree Sciences Co., Ltd., Yokohama, Japan) is used to program and execute the robotic translation experiment.

3. A spectrometer is used to measure DNA concentrations in plasmid preparations.

4. The Criterion stain-free gel imaging system (Bio-Rad Laboratories, Inc., USA) is used for qualitative and quantitative measurement of protein production (see Note ¹). Criterion Precast 4–20 % gels (#345-0426, Bio-Rad Laboratories, Inc., USA) are used for protein electrophoresis.

5. Image Lab software version 3.0 (Bio-Rad Laboratories, Inc., USA) is used to visualize and quantitate proteins separated by the Criterion imaging system.

6. A shaking device (suggested Microtube Rotisserie (LABQUAKE^®, Barnstead Thermolyne, NH, USA)) is used for enzyme assays.

7. A microplate reader is used for making fluorescence and optical density measurements of samples in either 96-well or 384-well microti ter plate formats (suggested instrument Infinite^® M1000 Pro, TECAN Group Ltd., Switzerland).

2.2 Cloning Reagents

1. 18 MΩ water is used to prepare all buffers and other reagents.

2. The pEU plasmid is the parental plasmid ([18], CellFree Sciences Co., Ltd., Yokohama, Japan). The wheat germ cell-free translation plasmid pEU-HSBC, shown in Fig. 1, was created from this parent.

3. C. thermocellum ATCC 27405 genomic DNA was provided by Dr. Paul Weimer (US Department of Agriculture Dairy Forage Research Center, Madison, WI).

4. Herculase^® II Fusion Enzyme with dNTP Combo (Agilent Technologies, Santa Clara, CA, USA) is used for PCR reactions.

5. 10× Flexi Enzyme Blend (Promega, Madison, WI, USA) containing optimized amounts of SgfI and PmeI endonucleases is used for restriction enzyme digestions.

6. T4 DNA Ligase (New England Biolabs Ltd., UK) is used for ligation reactions.

7. Chemically competent cells (suggested strain Escherichia coli 10G, Lucigen, Middleton, WI, USA) are used for transformation.

8. BigDye^® v2.0 (Applied Biosystems, CA, USA) is used for DNA sequencing reactions.

2.3 Transcription and Translation Reagents

1. A Qiagen Miniprep kit (Qiagen, Germany) is used for plasmid DNA preparations.

2. Proteinase K (Sigma-Aldrich, St. Louis, MO, USA) is used to eliminate residual RNase from plasmid DNA preparations. The 10× Proteinase K buffer consists of 100 mM Tris-HCl, pH 8.0, containing 50 mM EDTA and 1 % (w/v) SDS.

3. Transcription buffer (TB) consists of 400 mM HEPES-KOH, pH 7.8, 100 mM magnesium acetate, 10 mM spermidine hydrochloride, and 50 mM DTT. This buffer is used in cell-free translation reactions carried out in the Protemist DT II benchtop robot.

4. Nucleotide solution (NTPs) consists of 25 mM each of ATP, GTP, CTP, and UTP (Sigma-Aldrich, St. Louis, MO, USA) dissolved in Milli-Q water. This solution is used in the transcription stage of the cell-free translation reaction.

5. SP6 RNA polymerase and RNase inhibitor (Promega, Madison, WI, USA) are used in transcription reactions.

6. Translation buffer (40× solutions 1–4; CellFree Sciences Co., Ltd., Yokohama, Japan) is used with the Protemist DT II robot.

7. The amino acid mixture consists of 2 mM of each of the 20 L-amino acids.

8. Wheat germ cell-free extract, either WEPRO2240 or WEPRO2240H (CellFree Sciences Co., Ltd., Yokohama, Japan), is used in the Protemist DT II robot (see Note ²).

9. A 50 mg/mL solution of creatine kinase (Roche Applied Sciences, Indianapolis, IN, USA) is prepared in Milli-Q water and used in translation reactions.

2.4 Protein Purification Reagents

1. Ni-Sepharose high-performance chromatography resin (suggested resin is GE Flealthcare, Piscataway, NJ, USA) is used for purification of His-tagged proteins from the wheat germ cell-free translation reactions.

2. Purification loading buffer (100 mM MOPS, pH 7.4, containing 300 mM NaCi, 2 mM CaCl₂, and 25 mM imidazole) is used to load sample to the resin.

3. Purification washing buffer (100 mM MOPS, pH 7.4, containing 300 mM NaCl, 2 mM CaCl₂, and 50 mM imidazole) is used to reduce nonspecific binding of proteins to the resin.

4. Purification elution buffer (100 mM MOPS, pH 7.4, containing 300 mM NaCl, 2 mM CaCl₂, and 250 mM imidazole) is used to elute protein from the resin.

5. Laemmli sample buffer (#161-0737, Bio-Rad Laboratories, Inc., USA) is the 2× protein sample buffer.

6. Protein molecular weight size markers (suggested markers are Precision Plus Protein Unstained Standards, Bio-Rad Laboratories, Inc., USA) are used for internal standards in electrophoresis experiments.

7. VIVASPIN 500, 10,000 MWCO PES (Sartorius Stedim Biotech, Goettingen, Germany) is used for concentrating protein samples after purification (see Note ³).

8. U-bottom 96-well plates are from Greiner Bio-One (Monroe, NC).

2.5 Enzyme Assay Substrates

1. The fluorogenic substrates 4-methylumbelliferyl-β-D-glucopyranoside (MUG), 4-methylumbelliferyl-β-D-cellobioside (MUC), 4-methylumbelliferyl-β-D-mannopyranoside (MUM), and 4-methylumbelIiferyl-β-D-xylopyranoside (MUX) (Sigma-Aldrich, St. Louis, MO, USA) are used in small molecule analog assays. These substrates are prepared fresh for each experiment by dissolving an appropriate mass in a 50 mL solution of 0.1 M sodium phosphate buffer, pH 8.0, in order to achieve 0.2 mM substrate concentration.

2. Whatman filter paper 1001-185 (GE Healthcare, UK) is used as a representative substrate for cellulase assays. A standard paper hole punch tool is used to create ~0.7 cm diameter disks, corresponding to a weight of 3.0 mg.

3. Phosphoric acid-swollen cellulose (PASC) is used as a representative substrate for cellulase assays.

4. Birch xylan (Sigma-Aldrich, St. Louis, MO, USA) is used as a substrate for xylanase assays.

5. Mannan (Sigma-Aldrich, St. Louis, MO, USA) is used as a substrate for mannanase assay and mannan pull-down assay.

6. Ammonia fiber expansion (AFEX) switchgrass ground to 1 mm particle size and washed ionic liquid-treated switchgrass are used as biomass substrates (see Notes ⁴ and ⁵).

7. Sigmacell 20 (Sigma-Aldrich, St. Louis, MO, USA) is used as a substrate for the protein pull-down assay.

2.6 Enzyme Assay Reagents

1. A 50 mM sodium phosphate buffer, pH 8.0, is used for all reactions.

2. The colorimetric reagent 3,5-dinitrosalicylic acid (Sigma-Aldrich, St. Louis, MO, USA) is used in reducing sugar determinations. To prepare the stock reagent solution, 10.6 g of 3,5-dinitrosalicylic acid (Sigma Aldrich, St. Louis, MO) and 19.8 g of NaOH are dissolved in 1 L of Milli-Q water. Then 306 g of Rochelle salts (Sigma-Aldrich, St. Louis, MO, USA), 7.6 mL of phenol, and 8.3 g of sodium metabisulfite (Sigma-Aldrich, St. Louis, MO, USA) are added. The final volume of the reagent solution is adjusted to 1.5 L by addition of Milli-Q water. The reagent stock is stored at 4°C in foil-covered bottles until use.

3. For reducing sugar assays with reactions containing cellulose, β-glucosidase (Lucigen, Middleton, WI) is added to convert all soluble oligosaccharides to glucose. Although not essential, this addition allows the quantitation of total monomer sugar equivalents produced.

2.7 Polyacrylamide Gel Silver Staining Reagents

1. A silver staining method was used to detect protein-polysaccharide interactions in polyacrylamide gels. Reagents used are sodium thiosulfate (Sigma-Aldrich, St. Louis, MO, USA), silver nitrate, AgNO₃ (Sigma-Aldrich, St. Louis, MO, USA), and formaldehyde (Fisher Scientific, Hampton, NH, USA).

3 Methods

3.1 Design of First-Step PCR Primers for Candidate Genes

Figure 2 summarizes how PCR primers are designed by using the C. thermocellum Cthe_0032 gene, Uniprot A3DBE4, as an example (see Note ⁶).

1. The boundaries of the translocation signal peptide (residues 1–32) are used to define the position for truncation at the 5′ end of the gene. This truncation removes the signal peptide, which is not necessary for cell-free translation. According to this criterion, the gene-specific portion of the forward primer is 5′-atgtattcccttcctgtggac-3′ (Fig. 2b).

2. The first-step forward PCR primer is then completed by appending a nucleotide sequence encoding a portion of the tobacco etch virus protease recognition sequence to the 5′ end of the sequence identified in step 1. With this addition, the complete sequence of the first-step forward PCR primer is 5′-AACCTGTACTTCCAGTCCatgtattcccttcctgtggac-3′, indicated as “1st PCR, partial TEV cleavage site + gene-specific sequence” in Fig. 2b.

3. The end of the GH26 domain (residue 506) is used to define the position for truncation at the 3′ end of the gene. This truncation removes the dockerin domain. According to this criterion, the first-step reverse complement primer for the gene-specific sequence is 5′-ttcatccaaggtgattacata-3′.

4. For insertion of a stop codon at the end of the gene-specific sequence, the first-step reverse PCR primer is then completed by appending a nucleotide sequence encoding a stop codon and the PmeI restriction sequence to the 5′ end of the sequence identified in step 3. With this addition, the complete sequence of the first-step reverse PCR primer is 5′-GCTCGAATTCGTTTAAACTAttcatccaaggtgattacata-3′, indicated as “1st PCR, gene-specific sequence + PmeI site” in Fig. 2c. This primer is also shown in Table 2, Cthe_0032R.

5. For creation of a C-terminal fusion protein, the first-step reverse PCR primer is then completed by appending a nucleotide sequence encoding a PmeI restriction sequence to the 5′ end of the sequence identified in step 1. With this addition, the complete sequence of the first-step reverse PCR primer is 5′-GCTCGAATTCGTTTAAACCttcatccaaggtgattacata-3′. This primer is also shown in Table 2, Cthe_0032R_CBM3a.

Table 2.

Oligonucleotides used for cloning

Primer name	Sequencea	Direction
Cthe_0032F	5′-AACCTGTACTTCCAGTCCatgtattcccttcctgtggac-3′	Forward
Cthe_0032R	5′-GCTCGAATTCGTTTAAACTAttcatccaaggtgattacata-3′	Reverse
Cthe_0032R_CBM3a	5′-GCTCGAATTCGTTTAAACCttcatccaaggtgattacata-3′	Reverse
Cthe_0212F	5′-AACCTGTACTTCCAGTCCatgtcaaagataactttccc-3′	Forward
Cthe_0212R	5′-GCTCGAATTCGTTTAAACTAttaaaaaccgttgtttttgatta-3′	Reverse
Cthe_0269F	5′-AACCTGTACTTCCAGTCCatggtgccttttaacacaaa-3′	Forward
Cthe_0269R_CBM3a	5′-GCTCGAATTCGTTTAAACctcctgttatgtacaacaaagtg-3′	Reverse
Cthe_0270F	5′-AACCT GTACTTCCAGTCCatgaaaaaaataccgttactta-3′	Forward
Cthe_0270R	5′-GCTCGAATTCGTTTAAACTAtcaatcatcaacaggtatattgt-3′	Reverse
Cthe_0405F	5′-AACCT GTACTTCCAGTCCatggatccgaacaatgacgactg-3′	Forward
Cthe_0405R_CBM3a	5′-GCTCGAATTCGTTTAAACtccatttgaaccaagaggtatc-3′	Reverse
Cthe_0412F	5′-AACCTGTACTTCCAGTCCatgggacatcatcatcatcatca-3′	Forward
Cthe_0412R_CBM3a	5′-GCT CGAATTCGTTTAAACagctgtaacccatgcaaacg-3′	Reverse
Cthe_0536F	5′-AACCTGTACTTCCAGTCCatgggacatcatcatcatcatca-3′	Forward
Cthe_0536R	5′-GCTCGAATTCGTTTAAACTAtttatacggcaactcactta-3′	Reverse
Cthe_0543F	5′-AACCTGTACTTCCAGTCCatgggacatcatcatcatcatca-3′	Forward
Cthe_0543R	5′-GCTCGAATTCGTTTAAACTAttactgttcagccgggaattttt-3′	Reverse
Cthe_0578F	5′-AACCTGTACTTCCAGTCCatggactataactatggagaag-3′	Forward
Cthe_0578R_CBM3a	5′-GCTCGAATTCGTTTAAACggtaccattgggttctacacc-3′	Reverse
Cthe_0625F	5′-AACCTGTACTTCCAGTCCatgggacatcatcatcatcatca-3′	Forward
Cthe_0625R	5′-GCTCGAATTCGTTTAAACTActattctaccggaaatttatcta-3′	Reverse
Cthe_0797F	5′-AACCTGTACTTCCAGTCCatgggacatcatcatcatcatca-3′	Forward
Cthe_0797R	5′-GCTCGAATTCGTTTAAACTAtcttcaacgccggcacctctc-3′	Reverse
Cthe_0912F	5′-AACCTGTACTTCCAGTCCatgggacatcatcatcatcatca-3′	Forward
Cthe_0912R	5′-GCTCGAATTCGTTTAAACTActccttcgattacagttcc-3′	Reverse
Cthe_1256F	5′-AACCTGTACTTCCAGTCCatggcggtagatatcaagaaaat-3′	Forward
Cthe_1256R	5′-GCTCGAATT CGTTTAAACTAttattccacgttgtttattttgt-3′	Reverse
Cthe_1838F	5′-AACCTGTACTTCCAGTCCatgggacatcatcatcatcatca-3′	Forward
Cthe_1838R_CBM3a	5′-GCTCGAATT CGTTTAAACgttaactatagcataaaatgc-3′	Reverse
Cthe_2872F	5′-AACCTGTACTTCCAGTCCatgggacatcatcatcatcatca-3′	Forward
Cthe_2872R	5′-GCTCGAATTCGTTTAAACTAtccgtagtactcgcccagggaaa-3′	Reverse
2nd universal primerF	5′-GGTTGCGATCGCCGAAAACCTGTACTTCCAG-3′	Forward
2nd universal primerR	5′-GTGTGAGCTCGAATTCGTTTAAACC-3′	Reverse

^aLowercase letters indicate the gene-specific sequence

3.2 PCR Cloning and Plasmid Preparation

1. A two-step PCR method is used to amplify genes for this cloning method (see Note ⁷). The first-step PCR is performed using the forward and reverse primer pairs indicated in Table 2, with a reverse primer designed to create either a catalytic domain alone or a fusion protein. Each primer possesses ~20 nucleotides of gene-specific sequence and ~20 nucleotide overhang, respectively. The C. thermocellum genomic DNA is used as the amplification template. The PCR reaction contains 0.5 μL of 1 μg/μL C. thermocellum genomic DNA, 1.25 μL of 10 μM primer pair, 0.5 μL of 100 mM dNTP, 10 μL of 5× PCR buffer, and 1 μL of Herculase DNA polymerase in a total volume of 50 μL. The first-step PCR reaction cycle consists of 95°C for 5 min followed by 95°C for 30 s, a ramp from 50 to 60°C for 30 s, and 72°C for Xmin for 35 cycles, where X is increased by 1 min for each 1,000 base pairs in the template, and then the reaction cycle is ended by 72°C for 5 min.

2. The second-step PCR is performed using the first PCR product as a DNA template and the forward and reverse universal primers (Table 2). The forward universal primer is indicated in Fig. 2b as “2nd PCR with SgfI + TEV cleavage site,” while the reverse universal primer is indicated as “2nd PCR PmeI site.”

3. The PCR reaction contains 5 μL of 1st PCR reaction product, 2.5 μL of 10 μM primer pair, 2 μL of 100 mM dNTP, 20 μL of 5× PCR buffer, and 2 μL of Herculase DNA polymerase in a total volume of 100 μL. The second-step PCR reaction cycle consists of 95°C for 5 min followed by 95°C for 30 s, ramp from 50 to 60°C for 30 s, and 72°C for Xmin for 35 cycles, where X is increased by 1 min for each 1,000 base pairs in the template, and then the reaction cycle is ended by 72°C for 5 min.

4. Plasmid vector DNA (pEU-HSBC-CBM3a, see Note ⁸) and PCR product are digested using the Flexi Enzyme Blend. A typical reaction contains 3.5 μL of the DNA product, 0.5 μL of the Flexi Enzyme Blend, and 1 μL of buffer in a total volume of 5 μL. The reaction is carried out at 37°C for 60 min and followed by 65°C for 20 min.

5. Ligation of the digested PCR product and the plasmid vector DNA (pEU-HSBC-CBM3a) is carried out using T4 DNA Ligase with 1 mM dATP in l× ligation buffer for 12 h at 16°C.

6. The ligated sample is transformed into chemically competent E. coli 10G cells by a standard heat shock transformation protocol. Transformants are selected on Luria Bertani agar plates containing 100 μg/mL of ampicillin and 5% (w/v) of sucrose.

7. Single colonies are screened by colony PCR methods to identify transformants that contain an insert of the approximately correct size. Transformants with the correct insert size are then submitted for DNA sequencing, in this case at the University of Wisconsin Biotechnology Center. The nucleotide sequences are analyzed by software comparison of predicted and desired sequences (suggested DNASTAR software, v.10, DNASTAR, Madison, WI, USA).

8. A sequence-verified plasmid carrying the gene of interest is prepared for transcription by using a Qiagen Miniprep kit. The purified plasmid DNA is treated with proteinase K in 1 × Proteinase K buffer at 37°C for 1 h to eliminate residual RNase. After the digestion, the plasmid preparation is treated by phenol/chloroform extraction. The plasmid preparation is recovered from the aqueous phase of the phenol/chloroform extraction by ethanol precipitation with a 2.5× volume addition of absolute ethanol supplemented with a 1/10 volume of 3 M ammonium acetate. The precipitation reaction is incubated at −80°C for 1 h. The precipitated DNA is recovered by centrifugation at 16,100×g for 15 min. The amount of DNA is measured by spectrometry using A₂₆₀ measurement. Based on this measurement, the concentration of plasmid DNA is adjusted to 1 μg/μL.

3.3 Protein Cell-Free Translation

1. A 5 μL aliquot of a 1 μg/μL preparation of purified plasmid DNA is used for each transcription reaction carried out in the Protemist DT II.

2. The transcription mixture (45.25 μL total volume per reaction) consists of 29 μL of water, 10 μL of 5× of transcription buffer (CellFree Sciences Co., Ltd., Japan), 5 μL of 25 mM NTPs, 0.625 μL of SP6 RNA polymerase, and 0.625 μL of RNase inhibitor enzymes (80,000 units/mL).

3. The translation mixture (59.2 μL total volume per reaction) consists of 56 μL of wheat germ extract (CellFree Sciences Co., Ltd., Japan), 3 μL of 2 mM amino acid mix, and 0.2 μL of a 20 μg/μL solution of creatine kinase (Roche, USA).

4. The l× translation buffer (1.1 mL per reaction) containing ATP, GTP, and 20 amino acids is prepared from a SUB-AMIX (CFS-SUB, CellFree Sciences Co., Ltd., Japan).

5. The plasmid DNA, transcription mixture, translation mixture, and translation buffer are placed onto the appropriate positions of the deck of the Protemist™ DT II robot as described in the instrumental manual. The Protemist™ DT II carries out fully automated transcription and bilayer mode translation reactions in 28 h duty cycle [29].

3.4 Verification of Translated Protein in Wheat Germ Extract

1. Translated proteins are analyzed by using denaturing gel electrophoresis (suggested Bio-Rad Criterion system, see Notes ¹, ¹², ¹³). A 5 pL sample of the translation reaction and an equal volume of 2× protein sample buffer are mixed and heated at 95°C for 10 min. Electrophoresis is performed in constant voltage mode at 220 V for 50 min in l× Criterion buffer system.

3.5 Purification of His-Tagged Proteins

1. Ni-Sepharose resin is equilibrated with water to make an ~50% (v/v) slurry.

2. For each translation reaction from Protemist DT II robot, place 50 μL of the Ni-Sepharose slurry into each well of the filter plate, retained in the U-bottom 96-well plate. Add 100 μL of the Ni binding/washing buffer to each well. This combination of a filter plate and the U-bottom plates is the purification plate.

3. Add 100 μL of the supernatant from the soluble translation product plate to each well of the purification plate.

4. Mix the solution in a plate shaker for 10 min at room temperature, taking caution to not spill or cross-contaminate the wells.

5. Centrifuge the plate (suggested JS 5.9 rotor and Avanti J-30I centrifuge for 1 min at 3,640 rpm, 2,500×g, and ambient temperature).

6. Add 150 μL of Ni binding/washing buffer to wash out non-specifically bound proteins. Gently mix the solution using a plate shaker for 10 s at room temperature. Centrifuge the plate. Repeat this step three times. The filtrate should not contain the translated protein, but can be retained for analysis of binding efficiency and product yield if needed.

7. Place the filter plate onto a new U-bottom 96-well plate. Add 50 μL of the Ni elution buffer into each well of the purification plate. Use a plate shaker to gently mix the solution for 1 min at room temperature to elute the bound proteins.

8. Centrifuge the plate and save the filtrate for analysis by denaturing gel electrophoresis (Fig. 3) and enzymatic assays.

Fig. 3.

SDS-PAGE analysis (a) of the wheat germ cell-free translation of C. thermocellum proteins and (b) purification of proteins containing an N-terminal His-tag. Lane 1, Cthe_0032; lane 2, Cthe_0212; lane 3, Cthe_0269; lane 4, Cthe_0270; lane 5, Cthe_0536; lane 6, Cthe_0543; lane 7, Cthe_0625; lane 8, Cthe_0797; lane 9, Cthe_0912; lane 10, Cthe_1256; lane 11, Cthe_1838; lane 12, Cthe_2872. The domains included in the translated proteins and results of catalytic assays on these samples are provided in Table 3

3.6 Preparation of Phosphoric Acid-Swollen Cellulose

1. Sigmacell 50 microcrystalline cellulose (20 g, Sigma, St. Louis, MO) is swollen in 800 g of cold (0°C) 80 % phosphoric acid, with rapid stirring with a plastic rod for ~1 h in an ice bath.

2. The cellulose is diluted with 2 L of cold water, thoroughly mixed, and allowed to settle, and the overlying liquid is removed by siphoning. This washing and settling procedure is repeated several times to reduce the acid content.

3. The cellulose slurry is neutralized with solid NaHC0₃, rinsed, decanted as above, and then secured inside bags formed from nylon-reinforced paper towelling. These bags are filled with ~1 L of distilled water, and the excess liquid is squeezed off; this process is repeated 20 times. The bags are then sealed, suspended in buckets containing 5 L of cold deionized water, and dialyzed for 10 days, with frequent changes of water. Prior to each change of water, the bags are tightly hand squeezed to facilitate removal of the equilibrated solutions.

4. After completion of dialysis (when the phosphate content reaches < 1 μg/L) the cellulose is lyophilized (see Note ⁹).

3.7 Fluorogenic Substrate Hydrolysis Assay

• 5A 10 μL aliquot of cell-free translation reaction containing a translated enzyme is mixed with the fluorogenic substrate and assay buffer to give a total reaction volume of 100 μL. These reactions are conveniently prepared in a 96-well plate and include technical replicates.
• 6The fluorogenic reaction is carried out at 37°C for 15 min in plate reader. The fluorescence measurement is performed using excitation at 360 nm and emission detection at 460 nm. A negative control is made using a blank wheat germ translation reaction.
• 7Results from assays with a panel of fluorogenic substrates are summarized in Table 3. Enzymes that give a positive fluorescence increase after subtraction of the background fluorescence in the wheat germ extract control with the fluorogenic substrate are assigned to have positive reactivity with the indicated substrate (see Note ¹⁰).

3.8 Reducing Sugar Assays

1. A 50 μL aliquot of a cell-free translation reaction containing a translated enzyme is mixed with 50 μL of a 20 mg/mL suspension of pure polysaccharide substrate in 50 mM phosphate buffer, pH 8.0 (see Note ¹⁰). Pure polysaccharides suitable for use with this method include PASC, xylan, and mannan. These samples are conveniently prepared in a 96-well microtiter plate and sealed with TempPlate sealing film (USA Scientific, Ocala, FL).

2. The reaction is carried out at 60°C for 20 h. A 30 μL aliquot of the reaction and a 60 μL aliquot of DNS reagent are combined in a 96-well PCR plate. After mixing, the sample is heated at 95°C for 10 min.

3. The absorbance at 540 nm is used to determine the amount of reducing sugar present (Fig. 4a). A water blank containing DNS reagent is used for the background subtraction. The amount of reducing sugar produced is determined by comparison to standard curves generated using the corresponding pure monomer sugar, i.e., glucose, xylose, or mannose.

4. Results of the DNS assay with pure polysaccharides are shown in Fig. 4a and summarized in Table 3.

Fig. 4.

Representative endpoint assays of reducing sugar formation from C. thermocellum enzymes prepared by robotic cell-free translation and assayed without further purification. The reactions were carried out in a 384-well microtiter plate. The substrates used are indicated, and detection was by use of the DNS reagent, which reacts with reducing-end sugars, yielding a change in color from pale yellow to red (increase in absorbance at 540 nm). (a) Reactions of a set of individual cellulase and xylanase enzymes with pure polysaccharides, where each row contains the results for the panel of enzymes indicated with a different substrate, (b) The same set of single enzymes was tested for reaction with AFEX-SG and IL-SG. Table 3 gives a summary of the assay results, (c) Synergistic improvement in biomass deconstruction obtained from binary combinations of a cellulase with a xylanase. The influence of β-glucosidase in alleviation of product inhibition is also demonstrated

3.9 Biomass Assays

1. A 50 μL aliquot of cell-free translation reaction containing a translated enzyme is mixed with 50 μL of a 20 mg/mL suspension of pretreated biomass in 50 mM phosphate buffer, pH 8.0 (see Note ¹⁰). Biomass substrates used with this method include AFEX-SG and IL-SG.

2. The reaction is carried out at 60°C for 20 h. A 30 μL aliquot of the reaction and a 60 μL aliquot of DNS reagent are combined in a 96-well PCR plate. After mixing, the sample is heated at 95°C for 10 min.

3. Results of the DNS assay with pretreated biomass are shown in Fig. 4b and summarized in Table 3.

4. For combinatorial assays, 25 μL of two separate cell-free translation reactions are mixed with 50 μL of a 20 mg/mL suspension of pretreated biomass in 50 mM phosphate buffer, pH 8.0. The reaction is carried out as described in step 2 above (see Note ¹¹).

3.10 Protein-Polysaccharide Pull-Down Assay

1. Enzymes purified as described in Fig. 3 are used to test for binding to insoluble substrates Sigmacell 20 and PASC (Fig. 5). A 5 μL aliquot of purified enzyme is mixed with 1 mg of each substrate in 5 μL of total reaction in 50 mM phosphate buffer, pH 7.0, in a 96-well PCR plate. Reactions of protein without substrate are also prepared as a control for protein precipitation caused by aggregation or denaturation. Samples are incubated for 1 h at room temperature and then spun at 5,000 for 15 min at 4°C. The supernatant and pellet are carefully separated, and 10 μL of 2× sample buffer is added to the supernatant, while the pellet is suspended in 5 μL of 2× sample buffer and an equal volume of water. The samples are heated for 5 min at 95°C.

2. Denatured samples are separated using polyacrylamide gel electrophoresis and analyzed as described above.

3. The gels are placed into fixing solution (50% methanol, 12% acetic acid, 0.5 mL/L 37% formaldehyde) for 1 h.

4. The gels are washed in 50% ethanol three times for 5 min and pretreated with a 0.2 g/L solution of Na₂S₂O₃·5H₂O for 1 min.

5. The gels are rinsed with Milli-Q water 3× for 20 s and stained in a solution containing 2 g/L AgNO₃ and 0.75 mL/L of 37 % formaldehyde for 30 min.

6. The gels are rinsed again with Milli-Q water 3× for 20 s; developed in a solution containing 60 g/L Na₂CO₃, 0.5 mL/L of 37 % formaldehyde, and 4 mg/L of Na₂S₂O₃·5H₂O for 5–10 min; and rinsed with Milli-Q water 2× for 5 s.

7. The development is stopped with 50 % methanol (v/v) and 12% (w/v) acetic acid for 10 min. The gel is washed in 50 % (v/v) methanol for 20 min and imaged.

Fig. 5.

Pull-down assays of enzymes using Sigmacell 20 as the binding substrate. Five C. thermocellum proteins that have reactivity with insoluble cellulose were produced in cell-free translation reactions, purified as described, and tested for the ability to bind to cellulose. After introduction of the cellulose to the translation mixture and incubation, the pull-down sample was separated into pellet (P) and supernatant (S) fractions by centrifugation. Proteins with cellulose binding capacity were enriched in the pellet fraction after centrifugation (marked by stars). Soluble and pellet fractions were treated with denaturing buffer, separated by gel electrophoresis, and visualized by silver staining. For each protein, the control lane provides information on translated protein that pelleted in the absence of cellulose. Cthe_0536 lacks a CBM domain and did not bind to cellulose, while Cthe_0543 contained a natural CBM3c domain and was bound. Cthe_0625 also contained a natural CBM3c domain, but did not bind to cellulose. This second CBM3c domain has a different primary sequence that the similarly named domain in Cthe_0543. Surprisingly, Cthe_0797 formed a soluble complex with an endogenous wheat germ protein (marked by triangle); this interaction was broken when Cthe_0797 bound to cellulose. The multi-domain protein Cthe_1256 also exhibited binding to cellulose. In this figure, Cthe_0269 and Cthe_0412 were obtained by recombinant E. coli protein expression system. The Cthe_0269 is engineered such that additional CBM3a domain derived from Cthe_3077 scaffoldin is present at C-terminus. A native form of Cthe_0412 possesses a CBM4_9 domain at N-terminus. Pull-down assays are carried out for testing their binding capacity to two different forms of cellulose, Sigmacell 20 and PASC. Results indicate that both enzymes bind to two forms of cellulose. Apparently, Cthe_0269 binds more tightly to substrates than Cthe_0412, indicating that CBM3a binds to cellulose material more effectively than CBM4_9 domain