Appl Biochem Biotechnol (2014) 174:2864–2874 DOI 10.1007/s12010-014-1232-4
Purification and Characterization of Recombinant Cel7A from Maize Seed Nathan C. Hood & Kendall R. Hood & Susan L. Woodard & Shivakumar P. Devaiah & Tina Jeoh & Lisa Wilken & Zivko Nikolov & Erin Egelkrout & John A. Howard & Elizabeth E. Hood
Received: 10 March 2014 / Accepted: 10 September 2014 / Published online: 24 September 2014 # Springer Science+Business Media New York 2014
Abstract The corn grain biofactory was used to produce Cel7A, an exo-cellulase (cellobiohydrolase I) from Hypocrea jecorina. The enzymatic activity on small molecule substrates was equivalent to its fungal counterpart. The corn grain-derived enzyme is glycosylated and 6 kDa smaller than the native fungal protein, likely due to more sugars added in the glycosylation of the fungal enzyme. Our data suggest that corn seed-derived cellobiohydrolase (CBH) I performs as well as or better than its fungal counterpart in releasing sugars from complex substrates such as pre-treated corn stover or wood. This recombinant protein product can enter and expand current reagent enzyme markets as well as create new markets in textile or pulp processing. The purified protein is now available commercially. Keywords Cel7A . Recombinant protein . Cellobiohydrolase I . Protein purification . Cellulase . Biomass conversion N. C. Hood : K. R. Hood : E. E. Hood (*) Infinite Enzymes, LLC, PO Box 2654, State University, AR 72467, USA e-mail: [email protected] S. L. Woodard Kalon Biotherapeutics, 100 Discovery Drive, Suite 200, College Station, TX 77845, USA S. P. Devaiah Department of Biological Sciences, East Tennessee State University, Johnson City, TN 37614, USA T. Jeoh Department of Biological and Agricultural Engineering, University of California, Davis, CA 95616, USA L. Wilken Biological and Agricultural Engineering, Kansas State University, Manhattan, KS 66506-2906, USA Z. Nikolov Biological and Agricultural Engineering, Texas A&M University, College Station, TX 77843, USA E. Egelkrout : J. A. Howard Applied Biotechnology Institute, Cal Poly Tech Park, San Luis Obispo, CA 93407, USA
Appl Biochem Biotechnol (2014) 174:2864–2874
2865
Introduction Nature has developed effective cellulose hydrolytic enzymes, mostly microbial in origin, for recycling carbon from plant biomass in the environment. To date, many cellulase genes have been cloned and sequenced from a wide variety of bacteria, fungi, and plants [1, 2]. Cellulose is degraded through the synergistic action of two general types of enzymes. Enzymes that cleave the cellulose chain internally are referred to as endo-1,4-β-D-glucanases (E.C. 3.2.1.4) and provide new reducing and nonreducing chain ends on which exo-1,4-β-D-glucanases (cellobiohydrolase, CBH; E.C. 3.2.1.91) can operate [2]. Two types of exo-glucanase have been described that cleave cellulose from either the nonreducing end or the reducing end of the cellulose chain. The product of the exo-glucanase reaction is typically cellobiose, but other short-chain cellooligomers can also be produced. A third activity, β-D-glucosidase (E.C. 3.2.1.21), is required to cleave cellobiose and other oligomers to glucose. The β-D-glucosidase activity is required at 100 to 1,000 times lower concentration than the cellulases. We chose to express cellulases in maize as a first step toward developing a low-cost production system for enzymes used in lignocellulosic degradation [3, 4]. Using transgenic maize as a production system for industrial enzymes is a cost-effective, viable alternative to submerged culture fermentation systems and has been used to successfully produce several enzymes [5–7]. Previous work to express cellulases in plants has been successful for the thermophilic endo-1,4, β-D-glucanase (E1) but has failed to produce cellobiohydrolase (CBH) at levels enabling a commercial enterprise [8–10]. In contrast, the maize seed expression platform produced commercial levels of CBH I [3, 4]. High-level accumulation of cellulases, particularly CBH I, in maize seed has the potential to meet the cost target for the enzymes in the biomass-to-biofuel industry [11]. In this study, we report the purification and characterization of a recombinant Cel7A (CBH I) produced in maize seed and its activity on defined and complex substrates.
Materials and Methods Plant Material Maize (Zea mays L.) plants carrying the CBH I gene from Hypocrea jecorina were generated previously [3]. Expression of the gene is driven by the maize globulin-1 promoter; thus, the protein specifically accumulates in the embryo. Primary transgenic plants were bred to elite inbred varieties to generate production material as described [4]. Extraction of CBH I from Maize Whole seed of transgenic corn (1.5 kg) was pre-ground with a Cuisinart coffee grinder (model: DCG-20-N) in small batches (100 g each) for about 4 s. Samples were then fed through a Tribest Wolfgang Classic grain mill (model: KM-001; Anaheim, CA) producing finely ground corn flour. Flour was mixed with 3 L of extraction buffer (50 mM sodium acetate, pH 5.0) in a large plastic container on ice for 1 h using an electric mixer. The corn slurry was filtered through a 24-cm diameter Buchner funnel with a glass fiber filter and 1,200 g Hyflo® Super Cel® (Celite Corporation, Lompoc, CA). Ammonium sulfate (AS, Ideal Chemical and Supply Company, Memphis, TN) was added slowly to the filtrate with mixing to make a 50 % AS solution. The extract was allowed to precipitate for an hour on ice with mixing. The solution was filtered as before using a 15-cm Buchner funnel. Additional extraction
2866
Appl Biochem Biotechnol (2014) 174:2864–2874
buffer (400–500 mL) with 50 % ammonium sulfate was used to wash the filter. The filtrate was collected and the precipitate discarded. Additional ammonium sulfate was added to bring the solution to 90 % AS saturation. The mixture was allowed to precipitate overnight at 4 °C. The extract was filtered as above, and the precipitate was rinsed with a solution of cold 90 % ammonium sulfate in extraction buffer. The filtrate was discarded, and the precipitate was then dissolved in 50 mM sodium acetate buffer, pH 4.0, filtered through a Buchner funnel to remove hi-flow, and desalted using tangential flow filtration (TFF, Millipore, Billerica, MA). Membrane diafiltration was accomplished using a 10-kDa MWCO Biomax membrane (polyethersulfone) and two volumes of 50 mM sodium acetate buffer, pH 4.0. The diafiltered extract was then loaded onto a column (1.5 cm×30 cm) containing 50-mL Gigacap S650 resin (TOSOH Bioscience, King of Prussia, PA). The flow-through containing CBH I was collected and buffer exchanged with 20 mM sodium phosphate buffer, pH 7.0, using TFF. The concentrated extract was loaded onto another column (1.5 cm×30 cm) containing 50 mL Q Sepharose Fast Flow resin (GE Healthcare Bio-Sciences, Pittsburgh, PA). The column was washed with 20 mM Na-phosphate buffer, pH 7.0, containing 0.05 M NaCl until little to no protein was detected using absorption at 280 nm using a Synergy HT plate reader with a Take 3 plate (BioTek, Winooski, VT), and then, CBH I was eluted with 20 mM sodium phosphate buffer, pH 7.0, containing 0.3 M NaCl. The purified enzyme was buffer exchanged into the pH 5 extraction buffer. Purified CBH I was formulated as 3.2-M ammonium sulfate solution for stable storage and shipping (Sigma Chemical Co., E6412). Determination of CBH I Specific Activity: The specific activity of CBH I was determined by its ability to break down 4methylumbelliferyl β-D-cellobioside (MUC, Sigma Chemical Co., St. Louis, MO). Ten nanograms of purified CBH I and 60 micrograms of MUC were brought up to 125 μL with extraction buffer. The mixture was incubated in a 50 °C water bath for 2 h. Twenty-five microliters of solution was removed from each well and added into a black microtiter reading plate with 175 μL 0.1 M sodium carbonate to halt the reaction. The plate was read on a BioTek Synergy HT (Winooski, VT) 96-well plate reader with an excitation wavelength of 360 nm and an emission wavelength of 460 nm. A standard curve of 0.1–6 nmol/well 4-methylumbelliferone (MU) was used to calculate the number of micromoles of MU turned over per minute. A standard curve of cellulase (Sigma C-2730) was used when calculating milligrams per milliliter CBH I in the activity assay. Characterization of CBH I Two and four micrograms of purified CBH I were loaded onto a 4–16 % Precise sodium dodecyl sulfate (SDS)-PAGE gel (Thermo Scientific Pierce, Rockford, IL). The samples were run at 100 V and maximum milliampere in Tris-HEPES-SDS buffer (according to manufacturer’s instructions) for 1 h to allow separation. Protein size was determined by comparing to a pre-stained protein ladder (PageRuler Prestained Protein Ladder, Thermo Scientific) and HPLC-purified recombinant CBH I enzyme from maize. Stains for glycosylation were performed with a Pierce Glycoprotein Staining Kit (#24562) as recommended by the manufacturer. Western blots were performed as described previously [3].
Appl Biochem Biotechnol (2014) 174:2864–2874
2867
Determination of Optimal pH for Enzyme Activity Solutions of 200 mM sodium acetate and 200 mM acetic acid were prepared and filtersterilized. Specific amounts of each were combined with purified water to produce 50 mM sodium acetate with a pH ranging from 3.6 to 5.6 [12]. Solutions with pH 3.0, 6.0, and 7.0 were prepared using a phosphate citrate buffer. CBH I specific activity assays were run with the different buffers using purified enzyme. Specific activities were calculated from each assay. One unit of CBH I will turn over 1 μmol of MUC per minute. Extinction Coefficient Prior to lyophilization, purified CBH I TSP was measured with a Bradford assay and A280 using a Take3 plate with a Biotek Synergy HT plate reader. Concentration of CBH I was also measured using the CBH I activity assay. Ten-milliliter solution of purified CBH I in 50 mM sodium acetate buffer, pH 5, was aliquoted into 15-mL conical-bottom screw cap tubes. To measure the amount of ions within the sample, three tubes were filled with 10 mL each of 50 mM sodium acetate buffer, pH 5. Samples were then allowed to dry on a lyophilizer (Freezone 6 Labconco) for at least 24 h. After lyophilization, the tubes were weighed, and the original tube weight was subtracted to calculate sample weight within the tube. The weight of salt ions was also subtracted to calculate the milligram of protein within the sample. Measured and calculated protein weights were compared, and coefficients were calculated by dividing the calculated weight by the measured weight. The coefficient was multiplied by the A280 reading or assay result to obtain corrected values. Cellulose Deconstruction Enzyme amounts were determined by A280 measurement for purified CBH I or by MUC assay and expressed as Mega relative fluorescence units (MRFU). Purified CBH I (approximately 2 mg or 41 MRFU) was mixed with semi-purified E1 (280 MRFU), CBH II (1 MRFU), and βglucosidase (3 μL Sigma C6105/Novozyme 188) in a reaction volume of 1.6 mL in 50 mM sodium acetate buffer, pH 5. Sodium azide was added to inhibit microbial growth. Wood (20 mg) or 250-μL corn stover slurry (approximately 50 mg stover) was added as substrate. The reaction was incubated up to 7 days at 45 °C, with periodic removal of 20-μL samples. Samples at each time point were frozen until analysis at the end of the experiment. The level of glucose release was assayed essentially as described by (http://www.worthington-biochem. com/gop/assay.html, http://www.faizyme.com/assgluo.html). Samples (30 μL) were mixed with 160-μL reaction mixture consisting of 0.1 M potassium phosphate buffer, pH 6.0, 0. 016 % dianisidine, and horseradish peroxidase (Sigma P8125, 0.05 mg/mL). Glucose oxidase (Sigma G6125) was added at 0.05 mg/mL, and the reaction was read over 5 min using a Molecular Dynamics SpectraMax Plus 384.
Results Purification of CBH I from Transgenic Corn An outline of the CBH I purification process is illustrated in Fig. 1. The steps include ammonium sulfate precipitation of a crude extract, concentration and desalting of a 90 % ammonium sulfate-precipitated protein cake, and column chromatography first with a Gigacap
2868
Appl Biochem Biotechnol (2014) 174:2864–2874
Fig. 1 Flow diagram of purification process of recombinant CBH I from corn grain. A continuation on the flow diagram indicates which fraction was retained for further processing
column followed by a Q Sepharose column. Fold purification and percent recovery were calculated using crude protein containing CBH I activity as the reference point. Starting with 1.5 kg of corn flour, 3 L of buffer was added to make crude extract. The crude extract contained a total of 639 mg CBH I protein as detected by the CBH I activity assay and 24 g total protein (measured by A280). Fold purification increased at each step (Table 1). Final recovery of CBH I was 63 % using the process described. Losses were minimal at every step. The protein is purified approximately 60-fold as determined by estimates of protein concentration determined using
Appl Biochem Biotechnol (2014) 174:2864–2874
2869
Table 1 Mass balance recovery of CBH I during purification from transgenic corn grain Name
Volume (mL)
Protein (mg)
CBH I (mg)
Fold purification
% CBHI recovery
0
100
Crude
3,400
24,130
639
90 % AS filtrate
4,580
14,434
0
Cake extract
1,066
5,283
594
4.2
93
Concentrate Gigacap
165 229
1,968 481
472 506
9.0 40.0
74 79
Q column
106
251
400
60.0
63
absorbance at 280 nm (Table 1). Recovery allows purification of ~400 mg of CBH I activity from 1.5 kg of grain. A visual documentation of recovery at various steps is shown in Fig. 2. Characterization Measurement of Total Soluble Protein The ability to establish specific activity of CBH I depends on the ability to accurately determine the amount of protein. Thus, total soluble protein was measured using two methods that employ very different analysis criteria: microvolume absorbance measurements at 280 nm and the Bradford assay [13]. The microvolume absorbance measurements were faster and more convenient than the Bradford assay for batch-to-batch total protein determination; thus, microvolume absorbance became our standard method. Both assay methods gave consistent results although the Bradford assay indicated 4- to 9-fold lower protein concentrations than the absorbance at 280 nm (Table 2). Because of extreme differences in measured concentration using these two methods, we compared 10 μg estimated protein amounts on SDS-PAGE (data not shown), and the differences in measured protein concentration were visually apparent. The magnitude of the difference between the protein absorption at 280 nm and Bradford
Fig. 2 SDS-PAGE of CBH I fractions obtained during purification of recombinant CBH I from corn grain. Lanes: 1, MW markers; 2, 20 μg crude protein extract from ground grain; 3, 20 μg concentrated crude extract after 90 % AS precipitation; 4, MW markers; 5, 10 μg partially purified CBH I after Gigacap; 6, 10-μg final product
2870
Appl Biochem Biotechnol (2014) 174:2864–2874
Table 2 Comparison of absorbance (A280) and Bradford methods of protein concentration (mg/mL) determination in crude and purified corn extracts Sample
A280 (mg/mL)
Bradford (mg/mL)
Fold difference
Crude extract
7.5
0.81
9.2
Cake extract
3.1
0.36
8.6
TFF concentrate
13.0
2.73
4.8
Q Sepharose load
5.0
1.12
4.5
Pure CBH I
8.7
2.00
4.4
Pure CBH I x (ext. coef.)
7.0
2.00
3.5
measurements decreased as CBH I was purified (Table 2). To resolve these differences for pure protein, purified CBH I protein was lyophilized, the mass measured, and the result compared to estimated concentrations from absorbance at 280 nm (A280), Bradford, and CBH I activity assay. Correction factors for these assays and an extinction coefficient for the microvolume absorbance were calculated by dividing the milligrams of CBH I measured on the balance by the milligrams calculated from each method. Results obtained from various methods are multiplied by the correction factor or multiplied by one tenth of the extinction coefficient to obtain more accurate values. The correction factors are 1.188 for the CBH I activity assay and 3.98 for the Bradford assay. The extinction coefficient is 7.14 L/g cm for A280. SDS-PAGE and Western Blot Analysis The recombinant CBH I from maize has an apparent molecular weight 6 kDa smaller than native CBH I from H. jecorina when analyzed by SDS-PAGE (Fig. 3a). The broad bands of protein from both sources suggest that the native and recombinant proteins are glycosylated. Indeed, SDS-PAGE treated with Periodic Acid-Schiff stain shows glycosylation of the recombinant maize protein (Fig. 3b). Recombinant CBH I purified from maize and native CBH I purified from H. jecorina were compared by Western blot analysis of separated proteins
Fig. 3 a SDS-PAGE comparison of CBH I from recombinant maize (lane 2) and native H. jecorina (lane 3). Molecular weight markers are shown on either side (lanes 1 and 4). b Periodic Acid-Schiff stain of purified recombinant CBH I from corn grain. Lane 1, 2 μg; lane 2, MW ladder. c Western blot of CBH I. Lanes 1 and 4, MW markers; lane 2, CBH I from H. jecorina; lane 3, recombinant CBH I from maize
Appl Biochem Biotechnol (2014) 174:2864–2874
2871
transferred to a PVDF membrane and detected with anti-CBH I antibody. Enzyme from both sources appears as a single, broad band, confirming that we have purified CBH I from maize (Fig. 3c). CBH I Activity Activity assays were used to quantify CBH I at different steps during purification using 4methylumbelliferyl-β-D-cellobioside (MUC) to provide a measure of the total CBH I in each sample. To ensure accurate estimates of recovery, the amount of CBH I in the starting material had to be measured accurately. Equal concentrations of CBH I by activity assay of crude and purified samples were loaded onto an SDS-polyacrylamide gel. The crude extract showed roughly equivalent concentration of CBH I as 10 μg of purified protein (Fig. 4). Activity of CBH I on a protein basis increased approximately 60-fold with a 63 % recovery. Optimal pH Activity Activity of CBH I was determined across a range of pH values from 3 to 7. Purified CBH I from H. jecorina and maize showed optimal activity at pH values from 3.6 to 5 with a significant decrease in activity below pH 3.5 and above pH 5 (Fig. 5). Activity on Cellulose The ability of the purified CBH I to release glucose from two potential substrates, pre-treated wood and corn stover, was tested. To release glucose from complex substrates, a minimum of three enzymes is needed: an endocellulase (E1), an exo-cellulase (CBH I), and β-glucosidase. Purified CBH I in the presence of E1, CBH II (both derived from corn grain), and βglucosidase was incubated for up to 1 week with substrate, and glucose release was measured
Fig. 4 Reducing SDS-PAGE separation of proteins using HEPES buffer. Loading of CBH I crude extract by activity estimated from the MUC assay. Lanes: 1 and 2, 10 μg and 15 μg purified CBH I, respectively; lane 3, 10 μg CBH I in crude extract; lane 4, MW markers
Appl Biochem Biotechnol (2014) 174:2864–2874
Acvity (% of maximum)
2872
100 80 60 A
40
B C
20 0
2
4
pH
6
8
Fig. 5 pH optimum of CBH I activity from three sources: A HPLC-purified recombinant corn protein, B purified fungal enzyme from H. jecorina, and C purified recombinant corn enzyme using the process described here
using a glucose oxidase-based assay. Purified CBH I allowed significant glucose release in the absence of commercial mixes such as Celluclast. The glucose released was estimated to represent roughly 35 (wood) to 55 % (stover) of the total possible theoretical conversion (Fig. 6). As expected, glucose release from stover was more efficient than from wood. No significant glucose was observed in the enzyme fractions in the absence of substrate nor was any glucose detected in the presence of β-glucosidase only.
Fig. 6 Activity of purified CBH I from corn grain in combination with crude extracts of E1 endocellulase and βglucosidase on complex substrates. Two milligrams of purified CBH I was added to 20-mg wood or 50-mg corn stover in a 3-mL reaction. Data are shown as milligram of glucose released per milligram of substrate
Appl Biochem Biotechnol (2014) 174:2864–2874
2873
Discussion Plant lignocellulosic biomass is a complex matrix of polymers comprising the polysaccharides cellulose and hemicellulose, and a polyphenolic complex, lignin, as the major structural components [14]. Cellulose, the most abundant biopolymer on earth, is a simple, linear polymer of glucose. However, its semi-crystalline structure is notoriously resistant to hydrolysis by both enzymatic and chemical means. Yet, high yields of glucose from cellulose are critical to any economically viable biomass utilization strategy. Here, we describe the purification and partial characterization of an exo-cellulase, CBH I, produced recombinantly in corn grain for applications in research and in industries such as pulp processing and biomass conversion. To confirm that the enzyme is active and behaves as expected, it was purified and compared to the native protein from H. jecorina. Characteristics of the protein including pH optimum, Western blot reactivity, and enzyme activity were similar. However, the size of the corn-produced protein was 6 kDa smaller than the native fungal protein. This is likely due to the tendency of fungal strains to highly glycosylate secreted proteins [15, 16]. The purification of CBH I from corn grain was relatively straightforward with few steps, and recovery of the purified enzyme was 63 % of the starting material. The Gigacap 650S cation exchange chromatography (CBH I in the flow-through) removed a significant number of native corn proteins and significantly improved CBH I recovery and purification on Q Sepharose. The tools developed for standardizing the validation of purified recombinant CBH I will be applied to additional formulations of the enzyme from corn seed for large-scale, industrial applications to establish similar parameters for concentrated extract and whole defatted germ [11]. These latter formulations are the target formulations of the enzyme for markets such as pulp and textile processing and production of sugars from biomass feedstocks for biofuels and biobased products.
Conclusions Cel7A was expressed in corn embryos and was successfully extracted and purified. The cornderived protein shows similar characteristics as the fungal-derived protein. The protein is now available commercially. Acknowledgments This study was supported by a Phase 1 Small Business Innovation Research Grant from the National Institute for Food and Agriculture (ARKW-2011-00108).
References 1. Schulein, M. (2000). Biochimica et Biophysica Acta, 1543, 239–252. 2. Tomme, P., Warren, R. A., & Gilkes, N. R. (1995). Advances in Microbial Physiology, 37, 1–81. 3. Hood, E. E., Love, R., Lane, J., Bray, J., Clough, R., Pappu, K., Drees, C., Hood, K. R., Yoon, S., Ahmad, A., & Howard, J. A. (2007). Plant Biotechnology Journal, 5, 709–719. 4. Hood, E. E., Devaiah, S. P., Fake, G., Egelkrout, E., Teoh, K., Requesens, D. V., Hayden, C., Hood, K. R., Pappu, K. M., Carroll, J., & Howard, J. A. (2012). Plant Biotechnology Journal, 10, 20–30. 5. Hood, E. E., Bailey, M. R., Beifuss, K., Magallanes-Lundback, M., Horn, M. E., Callaway, E., Drees, C., Delaney, D. E., Clough, R., & Howard, J. A. (2003). Plant Biotechnology Journal, 1, 129–140.
2874
Appl Biochem Biotechnol (2014) 174:2864–2874
6. Hood, E. E., & Woodard, S. (2002). Industrial proteins produced from plants. In E. E. Hood & J. A. Howard (Eds.), Plants as factories for protein production (pp. 119–135). The Netherlands: Kluwer Academic Publishers Dordrecht. 7. Woodard, S. L., Mayor, J. M., Bailey, M. R., Barker, D. K., Love, R. T., Lane, J. R., Delaney, D. E., McComas-Wagner, J. M., Mallubhotla, H. D., Hood, E. E., Dangott, L. J., Tichy, S. E., & Howard, J. A. (2003). Biotechnology and Applied Biochemistry, 38, 123–130. 8. Ziegler, M., Thomas, S. R., & Danna, K. (2000). Molecular Breeding, 6, 37–46. 9. Ziegelhoffer, T., Will, J., & Austin-Phillips, S. (1999). Molecular Breeding, 5, 309–318. 10. Dai, Z., Hooker, B., Anderson, D. B., & Thomas, S. R. (2000). Transgenic Research, 9, 43–54. 11. Howard, J., Nikolov, Z., & Hood, E. (2011). Enzyme production systems for biomass conversion. In E. Hood, P. Nelson, & R. Powell (Eds.), Plant biomass conversion (pp. 227–253). Ames: Wiley-Blackwell. 12. Walpole, G. S. (1914). CCXXXIII. Journal of the Chemical Society, Transactions, 105, 2501–2521. 13. Bradford, M. M. (1976). Analytical Biochemistry, 72, 248–254. 14. Buchanan, B. B., Gruissem, W. and Jones, R. L. (2002) Biochemistry & Molecular Biology of Plants. American Society of Plant Biologists, Rockville, MD 15. Jeoh, T., Michener, W., Himmel, M. E., Decker, S. R., & Adney, W. S. (2008). Biotechnology for Biofuels, 1, 10–22. 16. Adney, W., S., Chou, Y.-C., Decker Stephen, R., Ding, S.-Y., Baker John, O., Kunkel, G., Vinzant Todd, B. and Himmel Michael, E. (2003) Heterologous expression of Trichoderma reesei 1,4-beta-D-glucan cellobiohydrolase (Cel 7A). In Applications of enzymes to lignocellulosics (pp. 403–437). Washington: American Chemical Society
Purification and Characterization of Recombinant Cel7A from Maize Seed Nathan C. Hood & Kendall R. Hood & Susan L. Woodard & Shivakumar P. Devaiah & Tina Jeoh & Lisa Wilken & Zivko Nikolov & Erin Egelkrout & John A. Howard & Elizabeth E. Hood
Received: 10 March 2014 / Accepted: 10 September 2014 / Published online: 24 September 2014 # Springer Science+Business Media New York 2014
Abstract The corn grain biofactory was used to produce Cel7A, an exo-cellulase (cellobiohydrolase I) from Hypocrea jecorina. The enzymatic activity on small molecule substrates was equivalent to its fungal counterpart. The corn grain-derived enzyme is glycosylated and 6 kDa smaller than the native fungal protein, likely due to more sugars added in the glycosylation of the fungal enzyme. Our data suggest that corn seed-derived cellobiohydrolase (CBH) I performs as well as or better than its fungal counterpart in releasing sugars from complex substrates such as pre-treated corn stover or wood. This recombinant protein product can enter and expand current reagent enzyme markets as well as create new markets in textile or pulp processing. The purified protein is now available commercially. Keywords Cel7A . Recombinant protein . Cellobiohydrolase I . Protein purification . Cellulase . Biomass conversion N. C. Hood : K. R. Hood : E. E. Hood (*) Infinite Enzymes, LLC, PO Box 2654, State University, AR 72467, USA e-mail: [email protected] S. L. Woodard Kalon Biotherapeutics, 100 Discovery Drive, Suite 200, College Station, TX 77845, USA S. P. Devaiah Department of Biological Sciences, East Tennessee State University, Johnson City, TN 37614, USA T. Jeoh Department of Biological and Agricultural Engineering, University of California, Davis, CA 95616, USA L. Wilken Biological and Agricultural Engineering, Kansas State University, Manhattan, KS 66506-2906, USA Z. Nikolov Biological and Agricultural Engineering, Texas A&M University, College Station, TX 77843, USA E. Egelkrout : J. A. Howard Applied Biotechnology Institute, Cal Poly Tech Park, San Luis Obispo, CA 93407, USA
Appl Biochem Biotechnol (2014) 174:2864–2874
2865
Introduction Nature has developed effective cellulose hydrolytic enzymes, mostly microbial in origin, for recycling carbon from plant biomass in the environment. To date, many cellulase genes have been cloned and sequenced from a wide variety of bacteria, fungi, and plants [1, 2]. Cellulose is degraded through the synergistic action of two general types of enzymes. Enzymes that cleave the cellulose chain internally are referred to as endo-1,4-β-D-glucanases (E.C. 3.2.1.4) and provide new reducing and nonreducing chain ends on which exo-1,4-β-D-glucanases (cellobiohydrolase, CBH; E.C. 3.2.1.91) can operate [2]. Two types of exo-glucanase have been described that cleave cellulose from either the nonreducing end or the reducing end of the cellulose chain. The product of the exo-glucanase reaction is typically cellobiose, but other short-chain cellooligomers can also be produced. A third activity, β-D-glucosidase (E.C. 3.2.1.21), is required to cleave cellobiose and other oligomers to glucose. The β-D-glucosidase activity is required at 100 to 1,000 times lower concentration than the cellulases. We chose to express cellulases in maize as a first step toward developing a low-cost production system for enzymes used in lignocellulosic degradation [3, 4]. Using transgenic maize as a production system for industrial enzymes is a cost-effective, viable alternative to submerged culture fermentation systems and has been used to successfully produce several enzymes [5–7]. Previous work to express cellulases in plants has been successful for the thermophilic endo-1,4, β-D-glucanase (E1) but has failed to produce cellobiohydrolase (CBH) at levels enabling a commercial enterprise [8–10]. In contrast, the maize seed expression platform produced commercial levels of CBH I [3, 4]. High-level accumulation of cellulases, particularly CBH I, in maize seed has the potential to meet the cost target for the enzymes in the biomass-to-biofuel industry [11]. In this study, we report the purification and characterization of a recombinant Cel7A (CBH I) produced in maize seed and its activity on defined and complex substrates.
Materials and Methods Plant Material Maize (Zea mays L.) plants carrying the CBH I gene from Hypocrea jecorina were generated previously [3]. Expression of the gene is driven by the maize globulin-1 promoter; thus, the protein specifically accumulates in the embryo. Primary transgenic plants were bred to elite inbred varieties to generate production material as described [4]. Extraction of CBH I from Maize Whole seed of transgenic corn (1.5 kg) was pre-ground with a Cuisinart coffee grinder (model: DCG-20-N) in small batches (100 g each) for about 4 s. Samples were then fed through a Tribest Wolfgang Classic grain mill (model: KM-001; Anaheim, CA) producing finely ground corn flour. Flour was mixed with 3 L of extraction buffer (50 mM sodium acetate, pH 5.0) in a large plastic container on ice for 1 h using an electric mixer. The corn slurry was filtered through a 24-cm diameter Buchner funnel with a glass fiber filter and 1,200 g Hyflo® Super Cel® (Celite Corporation, Lompoc, CA). Ammonium sulfate (AS, Ideal Chemical and Supply Company, Memphis, TN) was added slowly to the filtrate with mixing to make a 50 % AS solution. The extract was allowed to precipitate for an hour on ice with mixing. The solution was filtered as before using a 15-cm Buchner funnel. Additional extraction
2866
Appl Biochem Biotechnol (2014) 174:2864–2874
buffer (400–500 mL) with 50 % ammonium sulfate was used to wash the filter. The filtrate was collected and the precipitate discarded. Additional ammonium sulfate was added to bring the solution to 90 % AS saturation. The mixture was allowed to precipitate overnight at 4 °C. The extract was filtered as above, and the precipitate was rinsed with a solution of cold 90 % ammonium sulfate in extraction buffer. The filtrate was discarded, and the precipitate was then dissolved in 50 mM sodium acetate buffer, pH 4.0, filtered through a Buchner funnel to remove hi-flow, and desalted using tangential flow filtration (TFF, Millipore, Billerica, MA). Membrane diafiltration was accomplished using a 10-kDa MWCO Biomax membrane (polyethersulfone) and two volumes of 50 mM sodium acetate buffer, pH 4.0. The diafiltered extract was then loaded onto a column (1.5 cm×30 cm) containing 50-mL Gigacap S650 resin (TOSOH Bioscience, King of Prussia, PA). The flow-through containing CBH I was collected and buffer exchanged with 20 mM sodium phosphate buffer, pH 7.0, using TFF. The concentrated extract was loaded onto another column (1.5 cm×30 cm) containing 50 mL Q Sepharose Fast Flow resin (GE Healthcare Bio-Sciences, Pittsburgh, PA). The column was washed with 20 mM Na-phosphate buffer, pH 7.0, containing 0.05 M NaCl until little to no protein was detected using absorption at 280 nm using a Synergy HT plate reader with a Take 3 plate (BioTek, Winooski, VT), and then, CBH I was eluted with 20 mM sodium phosphate buffer, pH 7.0, containing 0.3 M NaCl. The purified enzyme was buffer exchanged into the pH 5 extraction buffer. Purified CBH I was formulated as 3.2-M ammonium sulfate solution for stable storage and shipping (Sigma Chemical Co., E6412). Determination of CBH I Specific Activity: The specific activity of CBH I was determined by its ability to break down 4methylumbelliferyl β-D-cellobioside (MUC, Sigma Chemical Co., St. Louis, MO). Ten nanograms of purified CBH I and 60 micrograms of MUC were brought up to 125 μL with extraction buffer. The mixture was incubated in a 50 °C water bath for 2 h. Twenty-five microliters of solution was removed from each well and added into a black microtiter reading plate with 175 μL 0.1 M sodium carbonate to halt the reaction. The plate was read on a BioTek Synergy HT (Winooski, VT) 96-well plate reader with an excitation wavelength of 360 nm and an emission wavelength of 460 nm. A standard curve of 0.1–6 nmol/well 4-methylumbelliferone (MU) was used to calculate the number of micromoles of MU turned over per minute. A standard curve of cellulase (Sigma C-2730) was used when calculating milligrams per milliliter CBH I in the activity assay. Characterization of CBH I Two and four micrograms of purified CBH I were loaded onto a 4–16 % Precise sodium dodecyl sulfate (SDS)-PAGE gel (Thermo Scientific Pierce, Rockford, IL). The samples were run at 100 V and maximum milliampere in Tris-HEPES-SDS buffer (according to manufacturer’s instructions) for 1 h to allow separation. Protein size was determined by comparing to a pre-stained protein ladder (PageRuler Prestained Protein Ladder, Thermo Scientific) and HPLC-purified recombinant CBH I enzyme from maize. Stains for glycosylation were performed with a Pierce Glycoprotein Staining Kit (#24562) as recommended by the manufacturer. Western blots were performed as described previously [3].
Appl Biochem Biotechnol (2014) 174:2864–2874
2867
Determination of Optimal pH for Enzyme Activity Solutions of 200 mM sodium acetate and 200 mM acetic acid were prepared and filtersterilized. Specific amounts of each were combined with purified water to produce 50 mM sodium acetate with a pH ranging from 3.6 to 5.6 [12]. Solutions with pH 3.0, 6.0, and 7.0 were prepared using a phosphate citrate buffer. CBH I specific activity assays were run with the different buffers using purified enzyme. Specific activities were calculated from each assay. One unit of CBH I will turn over 1 μmol of MUC per minute. Extinction Coefficient Prior to lyophilization, purified CBH I TSP was measured with a Bradford assay and A280 using a Take3 plate with a Biotek Synergy HT plate reader. Concentration of CBH I was also measured using the CBH I activity assay. Ten-milliliter solution of purified CBH I in 50 mM sodium acetate buffer, pH 5, was aliquoted into 15-mL conical-bottom screw cap tubes. To measure the amount of ions within the sample, three tubes were filled with 10 mL each of 50 mM sodium acetate buffer, pH 5. Samples were then allowed to dry on a lyophilizer (Freezone 6 Labconco) for at least 24 h. After lyophilization, the tubes were weighed, and the original tube weight was subtracted to calculate sample weight within the tube. The weight of salt ions was also subtracted to calculate the milligram of protein within the sample. Measured and calculated protein weights were compared, and coefficients were calculated by dividing the calculated weight by the measured weight. The coefficient was multiplied by the A280 reading or assay result to obtain corrected values. Cellulose Deconstruction Enzyme amounts were determined by A280 measurement for purified CBH I or by MUC assay and expressed as Mega relative fluorescence units (MRFU). Purified CBH I (approximately 2 mg or 41 MRFU) was mixed with semi-purified E1 (280 MRFU), CBH II (1 MRFU), and βglucosidase (3 μL Sigma C6105/Novozyme 188) in a reaction volume of 1.6 mL in 50 mM sodium acetate buffer, pH 5. Sodium azide was added to inhibit microbial growth. Wood (20 mg) or 250-μL corn stover slurry (approximately 50 mg stover) was added as substrate. The reaction was incubated up to 7 days at 45 °C, with periodic removal of 20-μL samples. Samples at each time point were frozen until analysis at the end of the experiment. The level of glucose release was assayed essentially as described by (http://www.worthington-biochem. com/gop/assay.html, http://www.faizyme.com/assgluo.html). Samples (30 μL) were mixed with 160-μL reaction mixture consisting of 0.1 M potassium phosphate buffer, pH 6.0, 0. 016 % dianisidine, and horseradish peroxidase (Sigma P8125, 0.05 mg/mL). Glucose oxidase (Sigma G6125) was added at 0.05 mg/mL, and the reaction was read over 5 min using a Molecular Dynamics SpectraMax Plus 384.
Results Purification of CBH I from Transgenic Corn An outline of the CBH I purification process is illustrated in Fig. 1. The steps include ammonium sulfate precipitation of a crude extract, concentration and desalting of a 90 % ammonium sulfate-precipitated protein cake, and column chromatography first with a Gigacap
2868
Appl Biochem Biotechnol (2014) 174:2864–2874
Fig. 1 Flow diagram of purification process of recombinant CBH I from corn grain. A continuation on the flow diagram indicates which fraction was retained for further processing
column followed by a Q Sepharose column. Fold purification and percent recovery were calculated using crude protein containing CBH I activity as the reference point. Starting with 1.5 kg of corn flour, 3 L of buffer was added to make crude extract. The crude extract contained a total of 639 mg CBH I protein as detected by the CBH I activity assay and 24 g total protein (measured by A280). Fold purification increased at each step (Table 1). Final recovery of CBH I was 63 % using the process described. Losses were minimal at every step. The protein is purified approximately 60-fold as determined by estimates of protein concentration determined using
Appl Biochem Biotechnol (2014) 174:2864–2874
2869
Table 1 Mass balance recovery of CBH I during purification from transgenic corn grain Name
Volume (mL)
Protein (mg)
CBH I (mg)
Fold purification
% CBHI recovery
0
100
Crude
3,400
24,130
639
90 % AS filtrate
4,580
14,434
0
Cake extract
1,066
5,283
594
4.2
93
Concentrate Gigacap
165 229
1,968 481
472 506
9.0 40.0
74 79
Q column
106
251
400
60.0
63
absorbance at 280 nm (Table 1). Recovery allows purification of ~400 mg of CBH I activity from 1.5 kg of grain. A visual documentation of recovery at various steps is shown in Fig. 2. Characterization Measurement of Total Soluble Protein The ability to establish specific activity of CBH I depends on the ability to accurately determine the amount of protein. Thus, total soluble protein was measured using two methods that employ very different analysis criteria: microvolume absorbance measurements at 280 nm and the Bradford assay [13]. The microvolume absorbance measurements were faster and more convenient than the Bradford assay for batch-to-batch total protein determination; thus, microvolume absorbance became our standard method. Both assay methods gave consistent results although the Bradford assay indicated 4- to 9-fold lower protein concentrations than the absorbance at 280 nm (Table 2). Because of extreme differences in measured concentration using these two methods, we compared 10 μg estimated protein amounts on SDS-PAGE (data not shown), and the differences in measured protein concentration were visually apparent. The magnitude of the difference between the protein absorption at 280 nm and Bradford
Fig. 2 SDS-PAGE of CBH I fractions obtained during purification of recombinant CBH I from corn grain. Lanes: 1, MW markers; 2, 20 μg crude protein extract from ground grain; 3, 20 μg concentrated crude extract after 90 % AS precipitation; 4, MW markers; 5, 10 μg partially purified CBH I after Gigacap; 6, 10-μg final product
2870
Appl Biochem Biotechnol (2014) 174:2864–2874
Table 2 Comparison of absorbance (A280) and Bradford methods of protein concentration (mg/mL) determination in crude and purified corn extracts Sample
A280 (mg/mL)
Bradford (mg/mL)
Fold difference
Crude extract
7.5
0.81
9.2
Cake extract
3.1
0.36
8.6
TFF concentrate
13.0
2.73
4.8
Q Sepharose load
5.0
1.12
4.5
Pure CBH I
8.7
2.00
4.4
Pure CBH I x (ext. coef.)
7.0
2.00
3.5
measurements decreased as CBH I was purified (Table 2). To resolve these differences for pure protein, purified CBH I protein was lyophilized, the mass measured, and the result compared to estimated concentrations from absorbance at 280 nm (A280), Bradford, and CBH I activity assay. Correction factors for these assays and an extinction coefficient for the microvolume absorbance were calculated by dividing the milligrams of CBH I measured on the balance by the milligrams calculated from each method. Results obtained from various methods are multiplied by the correction factor or multiplied by one tenth of the extinction coefficient to obtain more accurate values. The correction factors are 1.188 for the CBH I activity assay and 3.98 for the Bradford assay. The extinction coefficient is 7.14 L/g cm for A280. SDS-PAGE and Western Blot Analysis The recombinant CBH I from maize has an apparent molecular weight 6 kDa smaller than native CBH I from H. jecorina when analyzed by SDS-PAGE (Fig. 3a). The broad bands of protein from both sources suggest that the native and recombinant proteins are glycosylated. Indeed, SDS-PAGE treated with Periodic Acid-Schiff stain shows glycosylation of the recombinant maize protein (Fig. 3b). Recombinant CBH I purified from maize and native CBH I purified from H. jecorina were compared by Western blot analysis of separated proteins
Fig. 3 a SDS-PAGE comparison of CBH I from recombinant maize (lane 2) and native H. jecorina (lane 3). Molecular weight markers are shown on either side (lanes 1 and 4). b Periodic Acid-Schiff stain of purified recombinant CBH I from corn grain. Lane 1, 2 μg; lane 2, MW ladder. c Western blot of CBH I. Lanes 1 and 4, MW markers; lane 2, CBH I from H. jecorina; lane 3, recombinant CBH I from maize
Appl Biochem Biotechnol (2014) 174:2864–2874
2871
transferred to a PVDF membrane and detected with anti-CBH I antibody. Enzyme from both sources appears as a single, broad band, confirming that we have purified CBH I from maize (Fig. 3c). CBH I Activity Activity assays were used to quantify CBH I at different steps during purification using 4methylumbelliferyl-β-D-cellobioside (MUC) to provide a measure of the total CBH I in each sample. To ensure accurate estimates of recovery, the amount of CBH I in the starting material had to be measured accurately. Equal concentrations of CBH I by activity assay of crude and purified samples were loaded onto an SDS-polyacrylamide gel. The crude extract showed roughly equivalent concentration of CBH I as 10 μg of purified protein (Fig. 4). Activity of CBH I on a protein basis increased approximately 60-fold with a 63 % recovery. Optimal pH Activity Activity of CBH I was determined across a range of pH values from 3 to 7. Purified CBH I from H. jecorina and maize showed optimal activity at pH values from 3.6 to 5 with a significant decrease in activity below pH 3.5 and above pH 5 (Fig. 5). Activity on Cellulose The ability of the purified CBH I to release glucose from two potential substrates, pre-treated wood and corn stover, was tested. To release glucose from complex substrates, a minimum of three enzymes is needed: an endocellulase (E1), an exo-cellulase (CBH I), and β-glucosidase. Purified CBH I in the presence of E1, CBH II (both derived from corn grain), and βglucosidase was incubated for up to 1 week with substrate, and glucose release was measured
Fig. 4 Reducing SDS-PAGE separation of proteins using HEPES buffer. Loading of CBH I crude extract by activity estimated from the MUC assay. Lanes: 1 and 2, 10 μg and 15 μg purified CBH I, respectively; lane 3, 10 μg CBH I in crude extract; lane 4, MW markers
Appl Biochem Biotechnol (2014) 174:2864–2874
Acvity (% of maximum)
2872
100 80 60 A
40
B C
20 0
2
4
pH
6
8
Fig. 5 pH optimum of CBH I activity from three sources: A HPLC-purified recombinant corn protein, B purified fungal enzyme from H. jecorina, and C purified recombinant corn enzyme using the process described here
using a glucose oxidase-based assay. Purified CBH I allowed significant glucose release in the absence of commercial mixes such as Celluclast. The glucose released was estimated to represent roughly 35 (wood) to 55 % (stover) of the total possible theoretical conversion (Fig. 6). As expected, glucose release from stover was more efficient than from wood. No significant glucose was observed in the enzyme fractions in the absence of substrate nor was any glucose detected in the presence of β-glucosidase only.
Fig. 6 Activity of purified CBH I from corn grain in combination with crude extracts of E1 endocellulase and βglucosidase on complex substrates. Two milligrams of purified CBH I was added to 20-mg wood or 50-mg corn stover in a 3-mL reaction. Data are shown as milligram of glucose released per milligram of substrate
Appl Biochem Biotechnol (2014) 174:2864–2874
2873
Discussion Plant lignocellulosic biomass is a complex matrix of polymers comprising the polysaccharides cellulose and hemicellulose, and a polyphenolic complex, lignin, as the major structural components [14]. Cellulose, the most abundant biopolymer on earth, is a simple, linear polymer of glucose. However, its semi-crystalline structure is notoriously resistant to hydrolysis by both enzymatic and chemical means. Yet, high yields of glucose from cellulose are critical to any economically viable biomass utilization strategy. Here, we describe the purification and partial characterization of an exo-cellulase, CBH I, produced recombinantly in corn grain for applications in research and in industries such as pulp processing and biomass conversion. To confirm that the enzyme is active and behaves as expected, it was purified and compared to the native protein from H. jecorina. Characteristics of the protein including pH optimum, Western blot reactivity, and enzyme activity were similar. However, the size of the corn-produced protein was 6 kDa smaller than the native fungal protein. This is likely due to the tendency of fungal strains to highly glycosylate secreted proteins [15, 16]. The purification of CBH I from corn grain was relatively straightforward with few steps, and recovery of the purified enzyme was 63 % of the starting material. The Gigacap 650S cation exchange chromatography (CBH I in the flow-through) removed a significant number of native corn proteins and significantly improved CBH I recovery and purification on Q Sepharose. The tools developed for standardizing the validation of purified recombinant CBH I will be applied to additional formulations of the enzyme from corn seed for large-scale, industrial applications to establish similar parameters for concentrated extract and whole defatted germ [11]. These latter formulations are the target formulations of the enzyme for markets such as pulp and textile processing and production of sugars from biomass feedstocks for biofuels and biobased products.
Conclusions Cel7A was expressed in corn embryos and was successfully extracted and purified. The cornderived protein shows similar characteristics as the fungal-derived protein. The protein is now available commercially. Acknowledgments This study was supported by a Phase 1 Small Business Innovation Research Grant from the National Institute for Food and Agriculture (ARKW-2011-00108).
References 1. Schulein, M. (2000). Biochimica et Biophysica Acta, 1543, 239–252. 2. Tomme, P., Warren, R. A., & Gilkes, N. R. (1995). Advances in Microbial Physiology, 37, 1–81. 3. Hood, E. E., Love, R., Lane, J., Bray, J., Clough, R., Pappu, K., Drees, C., Hood, K. R., Yoon, S., Ahmad, A., & Howard, J. A. (2007). Plant Biotechnology Journal, 5, 709–719. 4. Hood, E. E., Devaiah, S. P., Fake, G., Egelkrout, E., Teoh, K., Requesens, D. V., Hayden, C., Hood, K. R., Pappu, K. M., Carroll, J., & Howard, J. A. (2012). Plant Biotechnology Journal, 10, 20–30. 5. Hood, E. E., Bailey, M. R., Beifuss, K., Magallanes-Lundback, M., Horn, M. E., Callaway, E., Drees, C., Delaney, D. E., Clough, R., & Howard, J. A. (2003). Plant Biotechnology Journal, 1, 129–140.
2874
Appl Biochem Biotechnol (2014) 174:2864–2874
6. Hood, E. E., & Woodard, S. (2002). Industrial proteins produced from plants. In E. E. Hood & J. A. Howard (Eds.), Plants as factories for protein production (pp. 119–135). The Netherlands: Kluwer Academic Publishers Dordrecht. 7. Woodard, S. L., Mayor, J. M., Bailey, M. R., Barker, D. K., Love, R. T., Lane, J. R., Delaney, D. E., McComas-Wagner, J. M., Mallubhotla, H. D., Hood, E. E., Dangott, L. J., Tichy, S. E., & Howard, J. A. (2003). Biotechnology and Applied Biochemistry, 38, 123–130. 8. Ziegler, M., Thomas, S. R., & Danna, K. (2000). Molecular Breeding, 6, 37–46. 9. Ziegelhoffer, T., Will, J., & Austin-Phillips, S. (1999). Molecular Breeding, 5, 309–318. 10. Dai, Z., Hooker, B., Anderson, D. B., & Thomas, S. R. (2000). Transgenic Research, 9, 43–54. 11. Howard, J., Nikolov, Z., & Hood, E. (2011). Enzyme production systems for biomass conversion. In E. Hood, P. Nelson, & R. Powell (Eds.), Plant biomass conversion (pp. 227–253). Ames: Wiley-Blackwell. 12. Walpole, G. S. (1914). CCXXXIII. Journal of the Chemical Society, Transactions, 105, 2501–2521. 13. Bradford, M. M. (1976). Analytical Biochemistry, 72, 248–254. 14. Buchanan, B. B., Gruissem, W. and Jones, R. L. (2002) Biochemistry & Molecular Biology of Plants. American Society of Plant Biologists, Rockville, MD 15. Jeoh, T., Michener, W., Himmel, M. E., Decker, S. R., & Adney, W. S. (2008). Biotechnology for Biofuels, 1, 10–22. 16. Adney, W., S., Chou, Y.-C., Decker Stephen, R., Ding, S.-Y., Baker John, O., Kunkel, G., Vinzant Todd, B. and Himmel Michael, E. (2003) Heterologous expression of Trichoderma reesei 1,4-beta-D-glucan cellobiohydrolase (Cel 7A). In Applications of enzymes to lignocellulosics (pp. 403–437). Washington: American Chemical Society
